[ { "path": ".gitignore", "content": "*.aux\n*.bbl\n*.bcf\n*.blg\n*.log\n*.out\n*.run.xml\n*.toc\n*.synctex.gz\ndata/\n**.idea/\n**__pycache__/\n*.swp\n*.~lock.*ods#\nend2end_neural_el_env/\ndeep-ed/data/\nlocal/\n" }, { "path": "Examples _ End-to-End Neural Entity Linking.ipynb", "content": "{\n \"cells\": [\n {\n \"cell_type\": \"markdown\",\n \"metadata\": {},\n \"source\": [\n \"## Try on your own input\\n\",\n \"The following notebook shows how to practice with the code repository with you example text.\\n\",\n \"**Pre-requisites**\\n\",\n \"- Make sure you have installed and downloaded everything as the [README](https://github.com/dalab/end2end_neural_el#trying-the-system-on-random-user-input-text) mentions on the github page.\\n\",\n \"- Once you follow the instructions you will have the server running at https://localhost:5555.\\n\",\n \"- To quickly. run the following cell.- \"\n ]\n },\n {\n \"cell_type\": \"code\",\n \"execution_count\": 15,\n \"metadata\": {},\n \"outputs\": [\n {\n \"data\": {\n \"text/plain\": [\n \"\"\n ]\n },\n \"execution_count\": 15,\n \"metadata\": {},\n \"output_type\": \"execute_result\"\n }\n ],\n \"source\": [\n \"import requests, json\\n\",\n \"requests.post(\\\"http://localhost:5555\\\") ## if Response is [200] then it means the server is running\"\n ]\n },\n {\n \"cell_type\": \"code\",\n \"execution_count\": 16,\n \"metadata\": {},\n \"outputs\": [\n {\n \"name\": \"stdout\",\n \"output_type\": \"stream\",\n \"text\": [\n \"[[17, 7, \\\"Germany\\\"], [49, 6, \\\"Angela_Merkel\\\"], [0, 5, \\\"Barack_Obama\\\"]]\"\n ]\n }\n ],\n \"source\": [\n \"## Check using curl. (it is not nessasary to run the code)\\n\",\n \"!curl -X POST --header 'Content-Type: application/json' --header 'Accept: application/json' -d \\\"{ \\\\\\\"text\\\\\\\": \\\\\\\"Obama will visit Germany and have a meeting with Merkel tomorrow.\\\\\\\", \\\\\\\"spans\\\\\\\": [] }\\\" 'http://localhost:5555'\"\n ]\n },\n {\n \"cell_type\": \"markdown\",\n \"metadata\": {},\n \"source\": [\n \"## Utility Functions\"\n ]\n },\n {\n \"cell_type\": \"code\",\n \"execution_count\": 17,\n \"metadata\": {},\n \"outputs\": [],\n \"source\": [\n \"import pprint\\n\",\n \"import pandas as pd\\n\",\n \"from IPython.display import Markdown\\n\",\n \"\\n\",\n \"\\n\",\n \"def query(text):\\n\",\n \" ## Takes the input string and passes it to the service and gets the reponse back.\\n\",\n \" myjson = { \\\"text\\\": text, \\\"spans\\\": [] }\\n\",\n \" r = requests.post(\\\"http://localhost:5555\\\", json=myjson)\\n\",\n \" return json.loads(r.content)\\n\",\n \"\\n\",\n \"def printmd(string):\\n\",\n \" ## displays the annotated/tagged input text in jupyter's Markdown format\\n\",\n \" display(Markdown(string))\\n\",\n \" \\n\",\n \" \\n\",\n \"def format_index_output(text):\\n\",\n \" ## main function which sends the input text to the service, gets the response back and formats the output\\n\",\n \" ## in a presentable form to evaluate.\\n\",\n \" \\n\",\n \" ents = query(text)\\n\",\n \" ents.sort(key=lambda tup: tup[0], reverse=True) \\n\",\n \" for i, ent in enumerate(ents):\\n\",\n \" text = text[:ent[0]] + '['+text[ent[0]:ent[0]+ent[1]]+'](https://en.wikipedia.org/wiki/'+ ent[2] +')' + text[ent[0]+ent[1]:]\\n\",\n \" \\n\",\n \" # Replace $ sign : Quick fix since $ sign is a keyword in jupyter markdowns\\n\",\n \" text = text.replace(\\\"$\\\",\\\"\\\\\\\\$\\\")\\n\",\n \" \\n\",\n \" printmd(text)\"\n ]\n },\n {\n \"cell_type\": \"code\",\n \"execution_count\": 24,\n \"metadata\": {},\n \"outputs\": [],\n \"source\": [\n \"text = [\\\"Eminem is the best rapper of all time. MMLP was his best album, Eminem and Dre have produced this album\\\",\\n\",\n \" \\\"Pakistan has one of the best teams in cricket. The pakistani squad for ICC Cricket World Cup has Shahid Afridi and Shoaib Akhtar\\\",\\n\",\n \" \\\"KIEV: Separatist rebels have not fulfilled conditions like handing back border posts or laying down their weapons, Ukraine’s president said on Monday in a phone call with the leaders of Russia, Germany and France as he pondered whether to extend a ceasefire.The call between President Petro Poroshenko, Russia’s Vladimir Putin, Germany’s Chancellor Angela Merkel and France’s Francois Hollande took place as an expiration deadline neared for Ukraine’s shaky, unilateral ceasefire\\\"\\n\",\n \" ]\"\n ]\n },\n {\n \"cell_type\": \"code\",\n \"execution_count\": 27,\n \"metadata\": {\n \"scrolled\": true\n },\n \"outputs\": [\n {\n \"data\": {\n \"text/markdown\": [\n \"[Eminem](https://en.wikipedia.org/wiki/Eminem) is the best rapper of all time. [MMLP](https://en.wikipedia.org/wiki/The_Marshall_Mathers_LP) was his best album, [Eminem](https://en.wikipedia.org/wiki/Eminem) and [Dre](https://en.wikipedia.org/wiki/Dr._Dre) have produced this album\"\n ],\n \"text/plain\": [\n \"\"\n ]\n },\n \"metadata\": {},\n \"output_type\": \"display_data\"\n },\n {\n \"data\": {\n \"text/markdown\": [\n \"Pakistan has one of the best teams in cricket. The pakistani squad for [ICC Cricket World Cup](https://en.wikipedia.org/wiki/Cricket_World_Cup) has [Shahid Afridi](https://en.wikipedia.org/wiki/Shahid_Afridi) and Shoaib Akhtar\"\n ],\n \"text/plain\": [\n \"\"\n ]\n },\n \"metadata\": {},\n \"output_type\": \"display_data\"\n },\n {\n \"data\": {\n \"text/markdown\": [\n \"[KIEV](https://en.wikipedia.org/wiki/Kiev): Separatist rebels have not fulfilled conditions like handing back border posts or laying down their weapons, [Ukraine](https://en.wikipedia.org/wiki/Ukraine)’s president said on Monday in a phone call with the leaders of [Russia](https://en.wikipedia.org/wiki/Russia), [Germany](https://en.wikipedia.org/wiki/Germany) and [France](https://en.wikipedia.org/wiki/France) as he pondered whether to extend a ceasefire.The call between President [Petro Poroshenko](https://en.wikipedia.org/wiki/Petro_Poroshenko), [Russia](https://en.wikipedia.org/wiki/Russia)’s [Vladimir Putin](https://en.wikipedia.org/wiki/Vladimir_Putin), [Germany](https://en.wikipedia.org/wiki/Germany)’s Chancellor [Angela Merkel](https://en.wikipedia.org/wiki/Angela_Merkel) and [France](https://en.wikipedia.org/wiki/France)’s [Francois Hollande](https://en.wikipedia.org/wiki/François_Hollande) took place as an expiration deadline neared for [Ukraine](https://en.wikipedia.org/wiki/Ukraine)’s shaky, unilateral ceasefire\"\n ],\n \"text/plain\": [\n \"\"\n ]\n },\n \"metadata\": {},\n \"output_type\": \"display_data\"\n }\n ],\n \"source\": [\n \"for t in text:\\n\",\n \" format_index_output(t)\"\n ]\n },\n {\n \"cell_type\": \"code\",\n \"execution_count\": null,\n \"metadata\": {},\n \"outputs\": [],\n \"source\": []\n }\n ],\n \"metadata\": {\n \"kernelspec\": {\n \"display_name\": \"Python [conda env:end2end_nel]\",\n \"language\": \"python\",\n \"name\": \"conda-env-end2end_nel-py\"\n },\n \"language_info\": {\n \"codemirror_mode\": {\n \"name\": \"ipython\",\n \"version\": 3\n },\n \"file_extension\": \".py\",\n \"mimetype\": \"text/x-python\",\n \"name\": \"python\",\n \"nbconvert_exporter\": \"python\",\n \"pygments_lexer\": \"ipython3\",\n \"version\": \"3.6.8\"\n }\n },\n \"nbformat\": 4,\n \"nbformat_minor\": 2\n}\n" }, { "path": "LICENSE", "content": "\n Apache License\n Version 2.0, January 2004\n http://www.apache.org/licenses/\n\n TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION\n\n 1. Definitions.\n\n \"License\" shall mean the terms and conditions for use, reproduction,\n and distribution as defined by Sections 1 through 9 of this document.\n\n \"Licensor\" shall mean the copyright owner or entity authorized by\n the copyright owner that is granting the License.\n\n \"Legal Entity\" shall mean the union of the acting entity and all\n other entities that control, are controlled by, or are under common\n control with that entity. For the purposes of this definition,\n \"control\" means (i) the power, direct or indirect, to cause the\n direction or management of such entity, whether by contract or\n otherwise, or (ii) ownership of fifty percent (50%) or more of the\n outstanding shares, or (iii) beneficial ownership of such entity.\n\n \"You\" (or \"Your\") shall mean an individual or Legal Entity\n exercising permissions granted by this License.\n\n \"Source\" form shall mean the preferred form for making modifications,\n including but not limited to software source code, documentation\n source, and configuration files.\n\n \"Object\" form shall mean any form resulting from mechanical\n transformation or translation of a Source form, including but\n not limited to compiled object code, generated documentation,\n and conversions to other media types.\n\n \"Work\" shall mean the work of authorship, whether in Source or\n Object form, made available under the License, as indicated by a\n copyright notice that is included in or attached to the work\n (an example is provided in the Appendix below).\n\n \"Derivative Works\" shall mean any work, whether in Source or Object\n form, that is based on (or derived from) the Work and for which the\n editorial revisions, annotations, elaborations, or other modifications\n represent, as a whole, an original work of authorship. For the purposes\n of this License, Derivative Works shall not include works that remain\n separable from, or merely link (or bind by name) to the interfaces of,\n the Work and Derivative Works thereof.\n\n \"Contribution\" shall mean any work of authorship, including\n the original version of the Work and any modifications or additions\n to that Work or Derivative Works thereof, that is intentionally\n submitted to Licensor for inclusion in the Work by the copyright owner\n or by an individual or Legal Entity authorized to submit on behalf of\n the copyright owner. For the purposes of this definition, \"submitted\"\n means any form of electronic, verbal, or written communication sent\n to the Licensor or its representatives, including but not limited to\n communication on electronic mailing lists, source code control systems,\n and issue tracking systems that are managed by, or on behalf of, the\n Licensor for the purpose of discussing and improving the Work, but\n excluding communication that is conspicuously marked or otherwise\n designated in writing by the copyright owner as \"Not a Contribution.\"\n\n \"Contributor\" shall mean Licensor and any individual or Legal Entity\n on behalf of whom a Contribution has been received by Licensor and\n subsequently incorporated within the Work.\n\n 2. Grant of Copyright License. Subject to the terms and conditions of\n this License, each Contributor hereby grants to You a perpetual,\n worldwide, non-exclusive, no-charge, royalty-free, irrevocable\n copyright license to reproduce, prepare Derivative Works of,\n publicly display, publicly perform, sublicense, and distribute the\n Work and such Derivative Works in Source or Object form.\n\n 3. Grant of Patent License. Subject to the terms and conditions of\n this License, each Contributor hereby grants to You a perpetual,\n worldwide, non-exclusive, no-charge, royalty-free, irrevocable\n (except as stated in this section) patent license to make, have made,\n use, offer to sell, sell, import, and otherwise transfer the Work,\n where such license applies only to those patent claims licensable\n by such Contributor that are necessarily infringed by their\n Contribution(s) alone or by combination of their Contribution(s)\n with the Work to which such Contribution(s) was submitted. If You\n institute patent litigation against any entity (including a\n cross-claim or counterclaim in a lawsuit) alleging that the Work\n or a Contribution incorporated within the Work constitutes direct\n or contributory patent infringement, then any patent licenses\n granted to You under this License for that Work shall terminate\n as of the date such litigation is filed.\n\n 4. Redistribution. You may reproduce and distribute copies of the\n Work or Derivative Works thereof in any medium, with or without\n modifications, and in Source or Object form, provided that You\n meet the following conditions:\n\n (a) You must give any other recipients of the Work or\n Derivative Works a copy of this License; and\n\n (b) You must cause any modified files to carry prominent notices\n stating that You changed the files; and\n\n (c) You must retain, in the Source form of any Derivative Works\n that You distribute, all copyright, patent, trademark, and\n attribution notices from the Source form of the Work,\n excluding those notices that do not pertain to any part of\n the Derivative Works; and\n\n (d) If the Work includes a \"NOTICE\" text file as part of its\n distribution, then any Derivative Works that You distribute must\n include a readable copy of the attribution notices contained\n within such NOTICE file, excluding those notices that do not\n pertain to any part of the Derivative Works, in at least one\n of the following places: within a NOTICE text file distributed\n as part of the Derivative Works; within the Source form or\n documentation, if provided along with the Derivative Works; or,\n within a display generated by the Derivative Works, if and\n wherever such third-party notices normally appear. The contents\n of the NOTICE file are for informational purposes only and\n do not modify the License. You may add Your own attribution\n notices within Derivative Works that You distribute, alongside\n or as an addendum to the NOTICE text from the Work, provided\n that such additional attribution notices cannot be construed\n as modifying the License.\n\n You may add Your own copyright statement to Your modifications and\n may provide additional or different license terms and conditions\n for use, reproduction, or distribution of Your modifications, or\n for any such Derivative Works as a whole, provided Your use,\n reproduction, and distribution of the Work otherwise complies with\n the conditions stated in this License.\n\n 5. Submission of Contributions. Unless You explicitly state otherwise,\n any Contribution intentionally submitted for inclusion in the Work\n by You to the Licensor shall be under the terms and conditions of\n this License, without any additional terms or conditions.\n Notwithstanding the above, nothing herein shall supersede or modify\n the terms of any separate license agreement you may have executed\n with Licensor regarding such Contributions.\n\n 6. Trademarks. This License does not grant permission to use the trade\n names, trademarks, service marks, or product names of the Licensor,\n except as required for reasonable and customary use in describing the\n origin of the Work and reproducing the content of the NOTICE file.\n\n 7. Disclaimer of Warranty. Unless required by applicable law or\n agreed to in writing, Licensor provides the Work (and each\n Contributor provides its Contributions) on an \"AS IS\" BASIS,\n WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or\n implied, including, without limitation, any warranties or conditions\n of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A\n PARTICULAR PURPOSE. You are solely responsible for determining the\n appropriateness of using or redistributing the Work and assume any\n risks associated with Your exercise of permissions under this License.\n\n 8. Limitation of Liability. In no event and under no legal theory,\n whether in tort (including negligence), contract, or otherwise,\n unless required by applicable law (such as deliberate and grossly\n negligent acts) or agreed to in writing, shall any Contributor be\n liable to You for damages, including any direct, indirect, special,\n incidental, or consequential damages of any character arising as a\n result of this License or out of the use or inability to use the\n Work (including but not limited to damages for loss of goodwill,\n work stoppage, computer failure or malfunction, or any and all\n other commercial damages or losses), even if such Contributor\n has been advised of the possibility of such damages.\n\n 9. Accepting Warranty or Additional Liability. While redistributing\n the Work or Derivative Works thereof, You may choose to offer,\n and charge a fee for, acceptance of support, warranty, indemnity,\n or other liability obligations and/or rights consistent with this\n License. However, in accepting such obligations, You may act only\n on Your own behalf and on Your sole responsibility, not on behalf\n of any other Contributor, and only if You agree to indemnify,\n defend, and hold each Contributor harmless for any liability\n incurred by, or claims asserted against, such Contributor by reason\n of your accepting any such warranty or additional liability.\n\n END OF TERMS AND CONDITIONS\n\n APPENDIX: How to apply the Apache License to your work.\n\n To apply the Apache License to your work, attach the following\n boilerplate notice, with the fields enclosed by brackets \"[]\"\n replaced with your own identifying information. (Don't include\n the brackets!) The text should be enclosed in the appropriate\n comment syntax for the file format. We also recommend that a\n file or class name and description of purpose be included on the\n same \"printed page\" as the copyright notice for easier\n identification within third-party archives.\n\n Copyright [yyyy] [name of copyright owner]\n\n Licensed under the Apache License, Version 2.0 (the \"License\");\n you may not use this file except in compliance with the License.\n You may obtain a copy of the License at\n\n http://www.apache.org/licenses/LICENSE-2.0\n\n Unless required by applicable law or agreed to in writing, software\n distributed under the License is distributed on an \"AS IS\" BASIS,\n WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n See the License for the specific language governing permissions and\n limitations under the License.\n" }, { "path": "code/__init__.py", "content": "" }, { "path": "code/evaluation/metrics.py", "content": "import numpy as np\nfrom collections import defaultdict\nfrom operator import itemgetter\nimport tensorflow as tf\n\n\nclass Evaluator(object):\n def __init__(self, threshold, name):\n self.threshold = threshold\n self.name = name\n self.TP = defaultdict(int) # docid -> counter\n self.FP = defaultdict(int) # docid -> counter\n self.FN = defaultdict(int) # docid -> counter\n self.docs = set() # set with all the docid encountered\n self.gm_num = 0\n\n def gm_add(self, gm_in_batch):\n self.gm_num += gm_in_batch\n\n def check_tp(self, score, docid):\n if score >= self.threshold:\n self.docs.add(docid)\n self.TP[docid] += 1\n return True\n return False\n\n def check_fp(self, score, docid):\n if score >= self.threshold:\n self.docs.add(docid)\n self.FP[docid] += 1\n return True\n return False\n\n def check_fn(self, score, docid):\n if score < self.threshold:\n self.docs.add(docid)\n self.FN[docid] += 1\n return True\n return False\n\n def _score_computation(self, el_mode):\n micro_tp, micro_fp, micro_fn = 0, 0, 0\n macro_pr, macro_re = 0, 0\n\n for docid in self.docs:\n tp, fp, fn = self.TP[docid], self.FP[docid], self.FN[docid]\n micro_tp += tp\n micro_fp += fp\n micro_fn += fn\n\n doc_precision = tp / (tp + fp + 1e-6)\n macro_pr += doc_precision\n\n doc_recall = tp / (tp + fn + 1e-6)\n macro_re += doc_recall\n\n if el_mode is False:\n assert(self.gm_num == micro_tp + micro_fn)\n\n micro_pr = 100 * micro_tp / (micro_tp + micro_fp + 1e-6)\n micro_re = 100 * micro_tp / (micro_tp + micro_fn + 1e-6)\n micro_f1 = 2*micro_pr*micro_re / (micro_pr + micro_re + 1e-6)\n\n macro_pr = 100 * macro_pr / len(self.docs)\n macro_re = 100 * macro_re / len(self.docs)\n macro_f1 = 2*macro_pr*macro_re / (macro_pr + macro_re + 1e-6)\n\n return micro_pr, micro_re, micro_f1, macro_pr, macro_re, macro_f1\n\n def print_log_results(self, tf_writer, eval_cnt, el_mode):\n micro_pr, micro_re, micro_f1, macro_pr, macro_re, macro_f1 = self._score_computation(el_mode)\n\n print(\"micro\", \"P: %.1f\" % micro_pr, \"\\tR: %.1f\" % micro_re, \"\\tF1: %.1f\" % micro_f1)\n print(\"macro\", \"P: %.1f\" % macro_pr, \"\\tR: %.1f\" % macro_re, \"\\tF1: %.1f\" % macro_f1)\n\n if tf_writer is None:\n return micro_f1, macro_f1\n\n\n name = self.name+\" macro\"\n writer_name = \"el_\" if el_mode else \"ed_\"\n summary = tf.Summary(value=[tf.Summary.Value(tag=name, simple_value=macro_f1)])\n tf_writer[writer_name+\"f1\"].add_summary(summary, eval_cnt)\n summary = tf.Summary(value=[tf.Summary.Value(tag=name, simple_value=macro_pr)])\n tf_writer[writer_name+\"pr\"].add_summary(summary, eval_cnt)\n summary = tf.Summary(value=[tf.Summary.Value(tag=name, simple_value=macro_re)])\n tf_writer[writer_name+\"re\"].add_summary(summary, eval_cnt)\n\n name = self.name+\" micro\"\n writer_name = \"el_\" if el_mode else \"ed_\"\n summary = tf.Summary(value=[tf.Summary.Value(tag=name, simple_value=micro_f1)])\n tf_writer[writer_name+\"f1\"].add_summary(summary, eval_cnt)\n summary = tf.Summary(value=[tf.Summary.Value(tag=name, simple_value=micro_pr)])\n tf_writer[writer_name+\"pr\"].add_summary(summary, eval_cnt)\n summary = tf.Summary(value=[tf.Summary.Value(tag=name, simple_value=micro_re)])\n tf_writer[writer_name+\"re\"].add_summary(summary, eval_cnt)\n\n return micro_f1, macro_f1\n\n def print_log_results_old(self, tf_writer, eval_cnt, el_mode):\n micro_tp, micro_fp, micro_fn = 0, 0, 0\n macro_pr, macro_re = 0, 0\n\n try:\n valid_macro_prec_cnt = 0\n valid_macro_recall_cnt = 0\n for docid in self.docs:\n tp, fp, fn = self.TP[docid], self.FP[docid], self.FN[docid]\n micro_tp += tp\n micro_fp += fp\n micro_fn += fn\n\n if tp + fp > 0:\n doc_precision = tp / (tp + fp)\n macro_pr += doc_precision\n valid_macro_prec_cnt += 1\n if tp + fn > 0:\n doc_recall = tp / (tp + fn)\n macro_re += doc_recall\n valid_macro_recall_cnt += 1\n\n if el_mode is False:\n assert(self.gm_num == micro_tp + micro_fn)\n\n micro_pr = 100 * micro_tp / (micro_tp + micro_fp) if (micro_tp + micro_fp) > 0 else 0\n micro_re = 100 * micro_tp / (micro_tp + micro_fn) if (micro_tp + micro_fn) > 0 else 0\n micro_f1 = 2*micro_pr*micro_re / (micro_pr + micro_re) if (micro_pr + micro_re) > 0 else 0\n\n macro_pr = 100 * macro_pr / valid_macro_prec_cnt if valid_macro_prec_cnt > 0 else 0\n macro_re = 100 * macro_re / valid_macro_recall_cnt if valid_macro_recall_cnt > 0 else 0\n macro_f1 = 2*macro_pr*macro_re / (macro_pr + macro_re) if (macro_pr + macro_re) > 0 else 0\n except ZeroDivisionError:\n print(\"Exception! ZeroDivisionError in print results!\\nmicro_tp, micro_fp, micro_fn = \", micro_tp,\n micro_fp, micro_fn)\n\n print(\"micro\", \"P: %.1f\" % micro_pr, \"\\tR: %.1f\" % micro_re, \"\\tF1: %.1f\" % micro_f1)\n print(\"macro\", \"P: %.1f\" % macro_pr, \"\\tR: %.1f\" % macro_re, \"\\tF1: %.1f\" % macro_f1)\n\n if tf_writer is None:\n print(\"len(self.docs)={}\\tvalid_macro_prec_cnt={}\\tvalid_macro_recall_cnt={}\".format(\n len(self.docs), valid_macro_prec_cnt, valid_macro_recall_cnt))\n return micro_f1, macro_f1\n\n\n name = self.name+\" macro\"\n writer_name = \"el_\" if el_mode else \"ed_\"\n summary = tf.Summary(value=[tf.Summary.Value(tag=name, simple_value=macro_f1)])\n tf_writer[writer_name+\"f1\"].add_summary(summary, eval_cnt)\n summary = tf.Summary(value=[tf.Summary.Value(tag=name, simple_value=macro_pr)])\n tf_writer[writer_name+\"pr\"].add_summary(summary, eval_cnt)\n summary = tf.Summary(value=[tf.Summary.Value(tag=name, simple_value=macro_re)])\n tf_writer[writer_name+\"re\"].add_summary(summary, eval_cnt)\n\n name = self.name+\" micro\"\n writer_name = \"el_\" if el_mode else \"ed_\"\n summary = tf.Summary(value=[tf.Summary.Value(tag=name, simple_value=micro_f1)])\n tf_writer[writer_name+\"f1\"].add_summary(summary, eval_cnt)\n summary = tf.Summary(value=[tf.Summary.Value(tag=name, simple_value=micro_pr)])\n tf_writer[writer_name+\"pr\"].add_summary(summary, eval_cnt)\n summary = tf.Summary(value=[tf.Summary.Value(tag=name, simple_value=micro_re)])\n tf_writer[writer_name+\"re\"].add_summary(summary, eval_cnt)\n\n return micro_f1, macro_f1\n\nclass StrongMatcher(object):\n \"\"\"is initialized with the gm_gt_list i.e. a list of tuples\n (begin_idx, end_idx, gt) and from the list of tuples it builds a set of tuples\n that will help us answer if our prediction matches with a tuple from the\n ground truth\"\"\"\n def __init__(self, b_e_gt_iterator):\n self.data = set() # of tuples (begin_idx, end_idx, gt)\n for t in b_e_gt_iterator:\n self.data.add(t)\n\n def check(self, t):\n \"\"\"returns True if tuple matches with ground truth else False\"\"\"\n return True if t in self.data else False\n\n\nclass WeakMatcher(object):\n \"\"\"is initialized with the gm_gt_list i.e. a list of tuples\n (begin_idx, end_idx, gt) and from the list of tuples it builds a data structure\n that will help us answer if our prediction matches with a tuple from the\n ground truth.\n structure used: a dict with key the gt and value a list of tuples\n (begin_idx, end_idx). So i compare the predicted triplet (b,e,ent_id)\n with all the ground truth triplets and check\n if they overlap (weak matching) and return True or False.\n e.g. 4 -> [(5,7), (13,14)] \"\"\"\n def __init__(self, b_e_gt_iterator):\n self.data = defaultdict(list)\n for b, e, gt in b_e_gt_iterator:\n self.data[gt].append((b, e))\n\n def check(self, t):\n # here the t comes from filtereds_spans[1:] so begin_idx, end_idxm best_cand_id but name it gt in the code\n s, e, gt = t\n if gt in self.data:\n for s2, e2 in self.data[gt]:\n if s<=s2 and e<=e2 and s2=s2 and e>=e2 and s=e2:\n return True\n elif s>=s2 and e<=e2:\n return True\n return False\n\n\nclass FNStrongMatcher(object):\n \"\"\"when initialized it takes our algorithms predictions\n (score, begin_idx, end_idx, ent_id) list and builds a dictionary.\n later we use it to check what score we have given to the ground truth i.e.\n gold mention plus the correct entity.\n structure used: a dict with key (begin_idx, end_idx, ent_id) --> given_score\n by my algorithm\"\"\"\n def __init__(self, filtered_spans):\n self.data = dict()\n for score, b, e, ent_id in filtered_spans:\n self.data[(b, e, ent_id)] = score\n\n def check(self, t):\n \"\"\"t are tuples (begin_idx, end_idx, gt) from gm_gt_list. I check\n if the ground truth is in my predictions and return the given score.\"\"\"\n return self.data[t] if t in self.data else -10000\n\n\nclass FNWeakMatcher(object):\n \"\"\"when initialized it takes our algorithms predictions\n (score, begin_idx, end_idx, ent_id) list and builds a data structure.\n later we use it to check what score we have given to the ground truth i.e.\n gold mention plus the correct entity.\n structure used: # a dict with key the gt and value a list of tuples\n (begin_idx, end_idx, given_score). So i compare the ground truth triplet (s,e,gt)\n with all the spans that my algorithm has linked to the same entity (gt) and check\n if they overlap (data matching) and return the highest score.\n e.g. 4 -> [(5,7, 0.2), (13,14, 0.3)] \"\"\"\n def __init__(self, filtered_spans):\n self.data = defaultdict(list)\n for score, b, e, ent_id in filtered_spans:\n self.data[ent_id].append((b, e, score))\n\n def check(self, t):\n \"\"\"t are tuples (begin_idx, end_idx, gt) from gm_gt_list. I check\n if the ground truth has overlap with some of my predictions and return\n the highest given score.\"\"\"\n s, e, gt = t\n best_score = -10000\n if gt in self.data:\n for s2, e2, score in self.data[gt]:\n if s<=s2 and e<=e2 and s2=s2 and e>=e2 and s=e2:\n best_score = max(best_score, score)\n elif s>=s2 and e<=e2:\n best_score = max(best_score, score)\n return best_score\n\n\ndef _filtered_spans_and_gm_gt_list(b, final_scores, cand_entities_len, cand_entities,\n begin_span, end_span, spans_len,\n begin_gm, end_gm, ground_truth,\n ground_truth_len, words_len):\n spans = []\n for i in range(spans_len[b]): # candidate span\n begin_idx = begin_span[b][i]\n end_idx = end_span[b][i]\n\n best_cand_id = -1\n best_cand_score = -10000\n for j in range(cand_entities_len[b][i]): # how many candidate entities we have for this span\n score = final_scores[b][i][j]\n if score > best_cand_score:\n best_cand_score = score\n best_cand_id = cand_entities[b][i][j]\n\n spans.append((best_cand_score, begin_idx, end_idx, best_cand_id))\n\n # now filter this list of spans based on score. from the overlapping ones keep the one\n # with the highest score.\n spans = sorted(spans, reverse=True) # highest score lowest score\n filtered_spans = []\n claimed = np.full(words_len[b], False, dtype=bool) # initially all words are free to select\n for span in spans:\n best_cand_score, begin_idx, end_idx, best_cand_id = span\n if not np.any(claimed[begin_idx:end_idx]) and best_cand_id > 0:\n # nothing is claimed so take it TODO this > 0 condition is it correct???\n claimed[begin_idx:end_idx] = True\n filtered_spans.append(span)\n\n gm_gt_list = [(begin_gm[b][i], end_gm[b][i], ground_truth[b][i]) for i in range(ground_truth_len[b])]\n\n return filtered_spans, gm_gt_list\n\n\ndef threshold_calculation(final_scores, cand_entities_len, cand_entities,\n begin_span, end_span, spans_len, begin_gm, end_gm, ground_truth,\n ground_truth_len, words_len, chunk_id, el_mode):\n tp_fp_batch_scores = []\n fn_batch_scores = []\n if el_mode is False:\n begin_gm = begin_span\n end_gm = end_span\n for b in range(final_scores.shape[0]): # batch\n filtered_spans, gm_gt_list = _filtered_spans_and_gm_gt_list(b, final_scores, cand_entities_len, cand_entities,\n begin_span, end_span, spans_len, begin_gm, end_gm, ground_truth, ground_truth_len, words_len)\n matcher = WeakMatcher(gm_gt_list) if el_mode else StrongMatcher(gm_gt_list)\n for t in filtered_spans:\n if matcher.check(t[1:]):\n tp_fp_batch_scores.append((t[0], 1)) # (score, TP)\n else:\n tp_fp_batch_scores.append((t[0], 0)) # (score, FP)\n\n # now check for the fn\n matcher = FNWeakMatcher(filtered_spans) if el_mode else FNStrongMatcher(filtered_spans)\n for t in gm_gt_list:\n score = matcher.check(t)\n fn_batch_scores.append(score)\n\n return tp_fp_batch_scores, fn_batch_scores\n\n\ndef metrics_calculation(evaluator, final_scores, cand_entities_len, cand_entities,\n begin_span, end_span, spans_len,\n begin_gm, end_gm, ground_truth,\n ground_truth_len, words_len, chunk_id, el_mode):\n if el_mode is False:\n begin_gm = begin_span\n end_gm = end_span\n # for each candidate span find which is the cand entity with the highest score\n for b in range(final_scores.shape[0]): # batch\n filtered_spans, gm_gt_list = _filtered_spans_and_gm_gt_list(b, final_scores, cand_entities_len, cand_entities,\n begin_span, end_span, spans_len, begin_gm, end_gm, ground_truth, ground_truth_len, words_len)\n matcher = WeakMatcher(gm_gt_list) if el_mode else StrongMatcher(gm_gt_list)\n docid = chunk_id[b].split(b\"&*\", 1)[0] # b'947testa_CRICKET&*0&*0' to b'947testa_CRICKET'\n # TODO remove this and the assertion\n evaluator.gm_add(len(gm_gt_list))\n for t in filtered_spans:\n if matcher.check(t[1:]):\n evaluator.check_tp(t[0], docid)\n else:\n evaluator.check_fp(t[0], docid)\n\n # now check for the fn\n matcher = FNWeakMatcher(filtered_spans) if el_mode else FNStrongMatcher(filtered_spans)\n for t in gm_gt_list:\n score = matcher.check(t)\n evaluator.check_fn(score, docid)\n\n\ndef metrics_calculation_and_prediction_printing(evaluator, final_scores,\n cand_entities_len, cand_entities,\n begin_span, end_span, spans_len,\n begin_gm, end_gm, ground_truth,\n ground_truth_len, words_len, chunk_id,\n words, chars, chars_len,\n scores_l, global_pairwise_scores, scores_names_l,\n el_mode, printPredictions=None):\n if el_mode is False:\n begin_gm = begin_span\n end_gm = end_span\n # for each candidate span find which is the cand entity with the highest score\n for b in range(final_scores.shape[0]): # batch\n spans = []\n for i in range(spans_len[b]): # candidate span\n begin_idx = begin_span[b][i]\n end_idx = end_span[b][i]\n\n best_cand_id = -1\n best_cand_score = -10000\n best_cand_similarity_score = -10000\n best_cand_position = -1\n scores_text = \"invalid\"\n for j in range(cand_entities_len[b][i]): # how many candidate entities we have for this span\n score = final_scores[b][i][j]\n if score > best_cand_score:\n best_cand_score = score\n best_cand_id = cand_entities[b][i][j]\n scores_text = ' '.join([scores_name + \"=\" + str(score[b][i][j]) for scores_name, score in zip(scores_names_l, scores_l)])\n # best_cand_similarity_score = similarity_scores[b][i][j]\n best_cand_position = j\n\n span_num = i\n spans.append((best_cand_score, begin_idx, end_idx, best_cand_id,\n scores_text, best_cand_position, span_num))\n\n # now filter this list of spans based on score. from the overlapping ones keep the one\n # with the highest score.\n spans = sorted(spans, reverse=True) # highest score lowest score\n filtered_spans = []\n claimed = np.full(words_len[b], False, dtype=bool) # initially all words are free to select\n for span in spans:\n best_cand_score, begin_idx, end_idx, best_cand_id = span[:4]\n if not np.any(claimed[begin_idx:end_idx]) and best_cand_id > 0:\n # nothing is claimed so take it TODO this > 0 condition is it correct???\n claimed[begin_idx:end_idx] = True\n filtered_spans.append(span)\n\n # now traverse all the filtered spans and compare them with the gold mentions\n # for each tuple of filtered_spans check for tp or fp\n gm_gt_list = [(begin_gm[b][i], end_gm[b][i], ground_truth[b][i]) for i in range(ground_truth_len[b])]\n matcher = WeakMatcher(gm_gt_list) if el_mode else StrongMatcher(gm_gt_list)\n\n docid = chunk_id[b].split(b\"&*\", 1)[0]\n evaluator.gm_add(len(gm_gt_list))\n\n tp_pred = []\n fp_pred = []\n fn_pred = []\n gt_minus_fn_pred = [] # gt_minus_fn_pred + fn_pred create the gm_gt_list\n for t in filtered_spans:\n if matcher.check(t[1:4]):\n if evaluator.check_tp(t[0], docid):\n tp_pred.append(t)\n else:\n if evaluator.check_fp(t[0], docid):\n fp_pred.append(t)\n\n # now check for the fn\n temp = [t[:4] for t in filtered_spans]\n matcher = FNWeakMatcher(temp) if el_mode else FNStrongMatcher(temp)\n for gm_num, t in enumerate(gm_gt_list):\n score = matcher.check(t)\n if evaluator.check_fn(score, docid):\n fn_pred.append((gm_num, *t))\n else:\n gt_minus_fn_pred.append((gm_num, *t))\n\n if printPredictions is not None:\n gmask = global_pairwise_scores[0][b] if global_pairwise_scores else None\n entity_embeddings = global_pairwise_scores[1][b] if global_pairwise_scores else None\n printPredictions.process_sample(str(chunk_id[b]),\n tp_pred, fp_pred, fn_pred, gt_minus_fn_pred,\n words[b], words_len[b],\n chars[b], chars_len[b],\n cand_entities[b], cand_entities_len[b],\n final_scores[b], filtered_spans,\n [score[b] for score in scores_l], scores_names_l,\n gmask, entity_embeddings)\n\n\n" }, { "path": "code/evaluation/metrics_old.py", "content": "import numpy as np\nfrom collections import defaultdict\nfrom operator import itemgetter\nimport tensorflow as tf\n\n\nclass Evaluator_aux(object):\n def __init__(self, threshold, name):\n self.threshold = threshold\n self.name = name\n self.TP = defaultdict(int) # docid -> counter\n self.FP = defaultdict(int) # docid -> counter\n self.FN = defaultdict(int) # docid -> counter\n self.docs = set() # set with all the docid encountered\n\n def check_tp(self, score, docid):\n if score >= self.threshold:\n self.docs.add(docid)\n self.TP[docid] += 1\n\n def check_fp(self, score, docid):\n if score >= self.threshold:\n self.docs.add(docid)\n self.FP[docid] += 1\n\n def check_fn(self, score, docid):\n if score < self.threshold:\n self.docs.add(docid)\n self.FN[docid] += 1\n\n def print_results(self):\n micro_tp, micro_fp, micro_fn = 0, 0, 0\n macro_pr, macro_re = 0, 0\n try:\n valid_macro_prec_cnt = 0\n valid_macro_recall_cnt = 0\n for docid in self.docs:\n tp, fp, fn = self.TP[docid], self.FP[docid], self.FN[docid]\n micro_tp += tp\n micro_fp += fp\n micro_fn += fn\n\n if tp + fp > 0:\n doc_precision = tp / (tp + fp)\n macro_pr += doc_precision\n valid_macro_prec_cnt += 1\n if tp + fn > 0:\n doc_recall = tp / (tp + fn)\n macro_re += doc_recall\n valid_macro_recall_cnt += 1\n\n\n micro_pr = micro_tp / (micro_tp + micro_fp) if (micro_tp + micro_fp) > 0 else 0\n micro_re = micro_tp / (micro_tp + micro_fn) if (micro_tp + micro_fn) > 0 else 0\n micro_f1 = 2*micro_pr*micro_re / (micro_pr + micro_re) if (micro_pr + micro_re) > 0 else 0\n\n macro_pr = macro_pr / valid_macro_prec_cnt if valid_macro_prec_cnt > 0 else 0\n macro_re = macro_re / valid_macro_recall_cnt if valid_macro_recall_cnt > 0 else 0\n macro_f1 = 2*macro_pr*macro_re / (macro_pr + macro_re) if (macro_pr + macro_re) > 0 else 0\n except ZeroDivisionError:\n print(\"Exception! ZeroDivisionError in print results!\\nmicro_tp, micro_fp, micro_fn = \", micro_tp,\n micro_fp, micro_fn)\n\n print(self.name, \"thr\", self.threshold)\n print(\"micro\", \"P:\", micro_pr, \"\\tR:\", micro_re, \"\\tF1:\", micro_f1)\n print(\"macro\", \"P:\", macro_pr, \"\\tR:\", macro_re, \"\\tF1:\", macro_f1)\n\n return micro_f1, macro_f1, self.threshold\n\nclass Evaluator(object):\n def __init__(self, weak_thr=None, strong_thr=None, name=\"\"):\n self.weak_evaluators = []\n self.strong_evaluators = []\n self.name = name\n for thr in weak_thr:\n self.weak_evaluators.append(Evaluator_aux(thr, \"data\"))\n for thr in strong_thr:\n self.strong_evaluators.append(Evaluator_aux(thr, \"strong\"))\n\n def weak_check_tp(self, score, docid):\n #list(map(lambda x: x.check_tp(score, docid), self.weak_evaluators))\n for x in self.weak_evaluators:\n x.check_tp(score, docid)\n\n def weak_check_fp(self, score, docid):\n #map(lambda x: x.check_fp(score, docid), self.weak_evaluators)\n for x in self.weak_evaluators:\n x.check_fp(score, docid)\n\n def weak_check_fn(self, score, docid):\n #map(lambda x: x.check_fn(score, docid), self.weak_evaluators)\n for x in self.weak_evaluators:\n x.check_fn(score, docid)\n\n def strong_check_tp(self, score, docid):\n #map(lambda x: x.check_tp(score, docid), self.strong_evaluators)\n for x in self.strong_evaluators:\n x.check_tp(score, docid)\n\n def strong_check_fp(self, score, docid):\n #map(lambda x: x.check_fp(score, docid), self.strong_evaluators)\n for x in self.strong_evaluators:\n x.check_fp(score, docid)\n\n def strong_check_fn(self, score, docid):\n #map(lambda x: x.check_fn(score, docid), self.strong_evaluators)\n for x in self.strong_evaluators:\n x.check_fn(score, docid)\n\n def print_log_results(self, writer, eval_cnt):\n weak_scores = [x.print_results() for x in self.weak_evaluators]\n strong_scores = [x.print_results() for x in self.strong_evaluators]\n if writer is not None:\n for micro_f1, macro_f1, threshold in weak_scores:\n name = self.name+\" data \" + str(threshold)\n summary = tf.Summary(value=[tf.Summary.Value(tag=name+\" micro_f1\",\n simple_value=micro_f1)])\n writer.add_summary(summary, eval_cnt)\n summary = tf.Summary(value=[tf.Summary.Value(tag=name+\" macro_f1\",\n simple_value=macro_f1)])\n writer.add_summary(summary, eval_cnt)\n\n for micro_f1, macro_f1, threshold in strong_scores:\n name = self.name+\" strong \" + str(threshold)\n summary = tf.Summary(value=[tf.Summary.Value(tag=name+\" micro_f1\",\n simple_value=micro_f1)])\n writer.add_summary(summary, eval_cnt)\n summary = tf.Summary(value=[tf.Summary.Value(tag=name+\" macro_f1\",\n simple_value=macro_f1)])\n writer.add_summary(summary, eval_cnt)\n\n result_list = weak_scores if self.weak_evaluators != [] else strong_scores\n return max(result_list, key=itemgetter(0))[0]\n\n\nclass WeakStrongMatching(object):\n def __init__(self, b_e_gt_iterator):\n self.exact = set() # of tuples (begin_idx, end_idx, gt)\n self.weak = defaultdict(list) # a map with key the gt and value a list of tuples\n # e.g. 4 -> [(5,7), (13,14)]\n # so in order to check for data match a search my gt if in the\n # data dictionary and if yes then check one by one overlap with\n # the span\n\n for b, e, gt in b_e_gt_iterator:\n self.exact.add((b, e, gt))\n self.weak[gt].append((b, e))\n\n def strong_check(self, t):\n return True if t in self.exact else False\n\n def weak_check(self, t):\n s, e, gt = t\n if gt in self.weak:\n for s2, e2 in self.weak[gt]:\n if s<=s2 and e<=e2 and s2=s2 and e>=e2 and s=e2:\n return True\n elif s>=s2 and e<=e2:\n return True\n\n return False\n\n\nclass FNWeakStrongMatching(object):\n def __init__(self, filtered_spans):\n self.weak = defaultdict(list) # a map with key the gt and value a list of tuples\n # e.g. 4 -> [(5,7, 0.2), (13,14, 0.3)]\n # so in order to check for data match a search my gt if in the\n # data dictionary and if yes then check one by one overlap with\n # the span\n\n for score, b, e, ent_id in filtered_spans:\n self.weak[ent_id].append((b, e, score))\n\n\n def strong_check(self, t):\n s, e, gt = t\n best_score = -10000\n if gt in self.weak:\n for s2, e2, score in self.weak[gt]:\n if s==s2 and e==e2:\n best_score = max(best_score, score)\n return best_score\n\n\n def weak_check(self, t):\n s, e, gt = t\n best_score = -10000\n if gt in self.weak:\n for s2, e2, score in self.weak[gt]:\n if s<=s2 and e<=e2 and s2=s2 and e>=e2 and s=e2:\n best_score = max(best_score, score)\n elif s>=s2 and e<=e2:\n best_score = max(best_score, score)\n return best_score\n\n\ndef validation_scores_calculation(evaluator, final_scores, cand_entities_len, cand_entities,\n begin_span, end_span, spans_len,\n begin_gm, end_gm, ground_truth,\n ground_truth_len, words_len, chunk_id, test_mode):\n if test_mode is False:\n begin_gm = begin_span\n end_gm = end_span\n # for each candidate span find which is the cand entity with the highest score\n for b in range(final_scores.shape[0]): # batch\n spans = []\n for i in range(spans_len[b]): # candidate span\n begin_idx = begin_span[b][i]\n end_idx = end_span[b][i]\n\n best_cand_id = -1\n best_cand_score = -10000\n for j in range(cand_entities_len[b][i]): # how many candidate entities we have for this span\n score = final_scores[b][i][j]\n if score > best_cand_score:\n best_cand_score = score\n best_cand_id = cand_entities[b][i][j]\n\n spans.append((best_cand_score, begin_idx, end_idx, best_cand_id))\n\n # now filter this list of spans based on score. from the overlapping ones keep the one\n # with the highest score.\n spans = sorted(spans, reverse=True) # highest score lowest score\n filtered_spans = []\n claimed = np.full(words_len[b], False, dtype=bool) # initially all words are free to select\n for span in spans:\n best_cand_score, begin_idx, end_idx, best_cand_id = span\n if not np.any(claimed[begin_idx:end_idx]) and best_cand_id > 0:\n # nothing is claimed so take it TODO this > 0 condition is it correct???\n claimed[begin_idx:end_idx] = True\n filtered_spans.append(span)\n\n\n # now traverse all the filtered spans and compare them with the gold mentions\n # for each tuple of filtered_spans check for tp or fp\n gm_gt_list = [(begin_gm[b][i], end_gm[b][i], ground_truth[b][i]) for i in range(ground_truth_len[b])]\n matcher = WeakStrongMatching(gm_gt_list)\n\n # b'947testa_CRICKET&*0&*0' to b'947testa_CRICKET'\n docid = chunk_id[b].split(b\"&*\", 1)[0]\n for t in filtered_spans:\n if matcher.strong_check(t[1:]):\n evaluator.strong_check_tp(t[0], docid)\n else:\n evaluator.strong_check_fp(t[0], docid)\n\n if matcher.weak_check(t[1:]):\n evaluator.weak_check_tp(t[0], docid)\n else:\n evaluator.weak_check_fp(t[0], docid)\n\n # now check for the fn\n matcher = FNWeakStrongMatching(filtered_spans)\n for t in gm_gt_list:\n score = matcher.strong_check(t)\n evaluator.strong_check_fn(score, docid)\n\n score = matcher.weak_check(t)\n evaluator.weak_check_fn(score, docid)\n\n\ndef evaluation_scores_calculation(evaluator, final_scores, cand_entities_len, cand_entities,\n begin_span, end_span, spans_len,\n begin_gm, end_gm, ground_truth,\n ground_truth_len, words_len, chunk_id, similarity_scores,\n words, chars, chars_len, cand_entities_scores,\n test_mode, printPredictions=None):\n if test_mode is False:\n begin_gm = begin_span\n end_gm = end_span\n # for each candidate span find which is the cand entity with the highest score\n for b in range(final_scores.shape[0]): # batch\n spans = []\n for i in range(spans_len[b]): # candidate span\n begin_idx = begin_span[b][i]\n end_idx = end_span[b][i]\n\n best_cand_id = -1\n best_cand_score = -10000\n best_cand_similarity_score = -10000\n best_cand_position = -1\n for j in range(cand_entities_len[b][i]): # how many candidate entities we have for this span\n score = final_scores[b][i][j]\n if score > best_cand_score:\n best_cand_score = score\n best_cand_id = cand_entities[b][i][j]\n best_cand_similarity_score = similarity_scores[b][i][j]\n best_cand_position = j\n\n spans.append((best_cand_score, begin_idx, end_idx, best_cand_id,\n best_cand_similarity_score, best_cand_position))\n\n # now filter this list of spans based on score. from the overlapping ones keep the one\n # with the highest score.\n spans = sorted(spans, reverse=True) # highest score lowest score\n filtered_spans = []\n claimed = np.full(words_len[b], False, dtype=bool) # initially all words are free to select\n for span in spans:\n best_cand_score, begin_idx, end_idx, best_cand_id, _, _ = span\n if not np.any(claimed[begin_idx:end_idx]) and best_cand_id > 0:\n # nothing is claimed so take it TODO this > 0 condition is it correct???\n claimed[begin_idx:end_idx] = True\n filtered_spans.append(span)\n\n # now traverse all the filtered spans and compare them with the gold mentions\n # for each tuple of filtered_spans check for tp or fp\n gm_gt_list = [(begin_gm[b][i], end_gm[b][i], ground_truth[b][i]) for i in range(ground_truth_len[b])]\n matcher = WeakStrongMatching(gm_gt_list)\n\n # b'947testa_CRICKET&*0&*0' to b'947testa_CRICKET'\n docid = chunk_id[b].split(b\"&*\", 1)[0]\n\n tp_pred = []\n fp_pred = []\n fn_pred = []\n thr = printPredictions.thr if printPredictions is not None else 0.2\n for t in filtered_spans:\n if matcher.strong_check(t[1:-2]):\n evaluator.strong_check_tp(t[0], docid)\n else:\n evaluator.strong_check_fp(t[0], docid)\n\n if matcher.weak_check(t[1:-2]):\n evaluator.weak_check_tp(t[0], docid)\n if t[0] >= thr:\n tp_pred.append(t)\n else:\n evaluator.weak_check_fp(t[0], docid)\n if t[0] >= thr:\n fp_pred.append(t)\n\n # now check for the fn\n matcher = FNWeakStrongMatching(\n [t[:-2] for t in filtered_spans])\n for t in gm_gt_list:\n score = matcher.strong_check(t)\n evaluator.strong_check_fn(score, docid)\n\n score = matcher.weak_check(t)\n evaluator.weak_check_fn(score, docid)\n if score < thr:\n fn_pred.append(t)\n\n if printPredictions is not None:\n printPredictions.process_sample(chunk_id[b], gm_gt_list,\n tp_pred, fp_pred, fn_pred,\n words[b], words_len[b],\n chars[b], chars_len[b],\n cand_entities[b], cand_entities_scores[b])\n" }, { "path": "code/evaluation/print_predictions (copy).py", "content": "\nfrom termcolor import colored\nimport pickle\nfrom preprocessing.util import load_wikiid2nnid, reverse_dict, load_wiki_name_id_map\nfrom collections import defaultdict\nimport operator\n\nclass PrintPredictions(object):\n def __init__(self, output_folder, predictions_folder, entity_extension=None):\n self.thr = None\n self.output_folder = output_folder\n self.predictions_folder = predictions_folder\n with open(output_folder+\"word_char_maps.pickle\", 'rb') as handle:\n _, self.id2word, _, self.id2char, _, _ = pickle.load(handle)\n\n self.nnid2wikiid = reverse_dict(load_wikiid2nnid(entity_extension), unique_values=True)\n _, self.wiki_id_name_map = load_wiki_name_id_map()\n self.extra_info = \"\"\n\n def map_entity(self, nnid):\n wikiid = self.nnid2wikiid[nnid]\n wikiname = self.wiki_id_name_map[wikiid].replace(' ', '_') if wikiid != \"\" else \"\"\n return \"{} {}\".format(wikiid, wikiname)\n\n def process_file(self, el_mode, name, opt_thr):\n self.thr = opt_thr\n self.el_mode = el_mode\n filepath = self.predictions_folder + (\"el/\" if el_mode else \"ed/\") + name\n self.fout = open(filepath, \"w\")\n\n def file_ended(self):\n self.fout.close()\n\n def process_sample(self, chunkid,\n tp_pred, fp_pred, fn_pred, gt_minus_fn_pred,\n words, words_len, chars, chars_len,\n cand_entities, log_cand_entities_scores, cand_entities_len,\n final_scores, similarity_scores):\n \"\"\"words: [None] 1d the words of a sample, words_len: scalar,\n chars: [None, None] 2d words, chars of each word, chars_len: [None] for each word\n the length in terms of characters.\n cand_entities: [None, None] gold_mentions, candidates for each gm,\n cand_entitites_len: [None] how many cand ent each gm has.\"\"\"\n reconstructed_words = []\n for i in range(words_len):\n word = words[i]\n if word != 0:\n reconstructed_words.append(self.id2word[word])\n else: # \n word_chars = []\n for j in range(chars_len[i]):\n word_chars.append(self.id2char[chars[i][j]])\n reconstructed_words.append(''.join(word_chars))\n\n text_tags = defaultdict(list)\n gt_legend = []\n\n for mylist, mycolor in zip([gt_minus_fn_pred, fn_pred], [\"green\", \"red\"]):\n for i, (gm_num, b, e, gt) in enumerate(mylist, 1):\n text_tags[b].append((1, colored(\"[{}\".format(i), mycolor)))\n text_tags[e].append((0, colored(\"]\", mycolor)))\n\n gt_text = \"\"\n if self.el_mode is False: # find the position and the score of the ground truth\n gt_text = \"gt not in candidate entities (recall miss)\"\n for j in range(cand_entities_len[gm_num]):\n if cand_entities[gm_num][j] == gt:\n gt_text = \"gt_p_e_m_pos={}, gt_logpem_score={}\".format(j,\n log_cand_entities_scores[gm_num][j])\n break\n text = colored(\"{}: {} {}\".format(i, self.map_entity(gt), gt_text), mycolor)\n gt_legend.append(text)\n\n tp_legend = []\n tp_pred = sorted(tp_pred, key=operator.itemgetter(1))\n for i, (score, b, e, nnid, sim_score, p_e_m_pos, span_num) in enumerate(tp_pred, 1):\n text_tags[b].append((1, colored(\"[{}\".format(i), \"blue\")))\n text_tags[e].append((0, colored(\"]\", \"blue\")))\n\n text = colored(\"{}: {}, score={}, sim_score={}, logpem={}, pem_pos={}\".format(i,\n self.map_entity(nnid), score, sim_score,\n log_cand_entities_scores[span_num][p_e_m_pos], p_e_m_pos), \"blue\")\n tp_legend.append(text)\n\n fp_legend = []\n fp_pred = sorted(fp_pred, key=operator.itemgetter(1))\n if len(fp_pred) > 0:\n fpWeakMatcherLogging = FPWeakMatcherLogging(self, fn_pred+gt_minus_fn_pred,\n cand_entities, cand_entities_len, log_cand_entities_scores,\n final_scores, similarity_scores)\n for i, (score, b, e, nnid, sim_score, p_e_m_pos, span_num) in enumerate(fp_pred, 1):\n text_tags[b].append((1, colored(\"[{}\".format(i), \"magenta\")))\n text_tags[e].append((0, colored(\"]\", \"magenta\")))\n\n fp_gt_text = fpWeakMatcherLogging.check(b, e, span_num)\n text = colored(\"{}: {}, score={}, sim_score={}, logpem={}, pem_pos={} {} \".format(i,\n self.map_entity(nnid), score, sim_score,\n log_cand_entities_scores[span_num][p_e_m_pos], p_e_m_pos,\n fp_gt_text), \"magenta\")\n fp_legend.append(text)\n\n final_acc = [\"new sample \" + chunkid+\"\\n\"]\n for i in range(words_len+1):\n final_acc.extend([text for _, text in sorted(text_tags[i])])\n if i < words_len:\n final_acc.append(reconstructed_words[i])\n self.fout.write(\" \".join(final_acc)+\"\\n\")\n self.fout.write(\"\\n\".join(gt_legend + tp_legend + fp_legend))\n self.fout.write(\"\\n\")\n\n\nclass FPWeakMatcherLogging(object):\n \"\"\"is initialized with the gm_gt_list i.e. a list of tuples\n (begin_idx, end_idx, gt) and from the list of tuples it builds a data structure. We already\n know that our tuple doesn't match a ground truth. Now we want to find out what exactly happens.\n cases: 1)) doesn't overlap with any gm 2)) overlap with one or more gm. In this case for each gm\n that it overlaps with find a) which is the gt of this gm, b) final_score, sim_score, p_e_m position\n of the gt in my fp tuple.\n structure used: just a list of (begin_idx, end_idx, gt) tuples.\n This one is used only during evaluation.py from the\n metrics_calculation_and_prediction_printing in order to produce logging text\n for the fp\"\"\"\n def __init__(self, printPredictions, b_e_gt_iterator, cand_entities, cand_entities_len,\n log_cand_entities_scores, final_scores, similarity_scores):\n self.printPredictions = printPredictions\n self.data = b_e_gt_iterator\n self.cand_entities = cand_entities\n self.cand_entities_len = cand_entities_len\n self.log_cand_entities_scores = log_cand_entities_scores\n self.final_scores = final_scores\n self.similarity_scores = similarity_scores\n\n def check(self, s, e, span_num):\n # all the above information that i have for my best_cand_id, now i have to find them\n # for the gt of the gm that overlap with my fp tuple.\n # compare my tuple s, e with all the gm\n\n acc = []\n for (gm_num, s2, e2, gt) in self.data:\n overlap = False # overlap with this specific gm of the for loop\n if s<=s2 and e<=e2 and s2=s2 and e>=e2 and s=e2:\n overlap = True\n elif s>=s2 and e<=e2:\n overlap = True\n\n if not overlap:\n continue\n\n # add to the text accumulator the info for this gt\n # find gt_score, gt_similarity_score, gt_cand_position\n # check all the candidate entities of this span and find where is the gt\n # of course we may not find it at all (recall miss)\n gt_cand_position = -1\n for j in range(self.cand_entities_len[span_num]):\n if self.cand_entities[span_num][j] == gt:\n gt_cand_position = j\n break\n\n if gt_cand_position >= 0:\n acc.append(\"| {}, score={}, sim_score={}, logpem={}, pem_pos={}\".format(\n self.printPredictions.map_entity(gt),\n self.final_scores[span_num][gt_cand_position],\n self.similarity_scores[span_num][gt_cand_position],\n self.log_cand_entities_scores[span_num][gt_cand_position],\n gt_cand_position))\n else:\n acc.append(\"| {}, recall miss\".format(self.printPredictions.map_entity(gt)))\n\n if acc == []:\n acc.append(\"| no overlap with gm\")\n\n return ' '.join(acc)\n" }, { "path": "code/evaluation/print_predictions.py", "content": "\nfrom termcolor import colored\nimport pickle\nfrom preprocessing.util import load_wikiid2nnid, reverse_dict, load_wiki_name_id_map\nfrom collections import defaultdict\nimport operator\nimport numpy as np\n\n\nclass GMBucketingResults(object):\n def __init__(self, gm_bucketing_pempos):\n gm_bucketing_pempos.append(200) # [0,1,2,7,200]\n self.gm_buckets = gm_bucketing_pempos\n self.gm_cnt = defaultdict(int) # how many gold mentions fall in that frquency. one counter for each bucket\n self.fn_cnt = defaultdict(int) # for many false negative in that bucket\n self.fn_nowinnermatch_cnt = defaultdict(int) # from the fn we exclude the ones that our winner was identical\n # to gt even if we decided not to annotate in the end\n self.gm_to_gt_unique_mapping = 0 # for this gold mention we have only one candidate entity which is the gt.\n\n def reinitialize(self):\n self.gm_cnt = defaultdict(int)\n self.fn_cnt = defaultdict(int)\n self.fn_nowinnermatch_cnt = defaultdict(int)\n self.gm_to_gt_unique_mapping = 0\n\n def process_fn(self, pos, match_with_winner, num_of_cand_entities):\n if pos == 0 and num_of_cand_entities == 1:\n self.gm_to_gt_unique_mapping += 1\n for t in self.gm_buckets:\n if pos <= t:\n self.gm_cnt[t] += 1\n self.fn_cnt[t] += 1\n if not match_with_winner:\n self.fn_nowinnermatch_cnt[t] += 1\n break\n\n def process_tp(self, pos, num_of_cand_entities):\n if pos == 0 and num_of_cand_entities == 1:\n self.gm_to_gt_unique_mapping += 1\n for t in self.gm_buckets:\n if pos <= t:\n self.gm_cnt[t] += 1\n break\n\n def print(self):\n print(\"gm_to_gt_unique_mapping =\", self.gm_to_gt_unique_mapping)\n for t in self.gm_buckets:\n print(str(t), \"]\", \"gm_cnt=\", str(self.gm_cnt[t]),\n \"solved=%.1f\" % (100*(self.gm_cnt[t] - self.fn_cnt[t])/self.gm_cnt[t]),\n \"winner_match=%.1f\" % (100*(self.gm_cnt[t] - self.fn_nowinnermatch_cnt[t])/self.gm_cnt[t]))\n\n\n\nclass PrintPredictions(object):\n def __init__(self, output_folder, predictions_folder, entity_extension=None, gm_bucketing_pempos=None,\n print_global_voters=False, print_global_pairwise_scores=False):\n self.thr = None\n self.output_folder = output_folder\n self.predictions_folder = predictions_folder\n with open(output_folder+\"word_char_maps.pickle\", 'rb') as handle:\n _, self.id2word, _, self.id2char, _, _ = pickle.load(handle)\n\n self.nnid2wikiid = reverse_dict(load_wikiid2nnid(entity_extension), unique_values=True)\n _, self.wiki_id_name_map = load_wiki_name_id_map()\n self.extra_info = \"\"\n self.gm_bucketing = GMBucketingResults(gm_bucketing_pempos) if gm_bucketing_pempos else None\n self.print_global_pairwise_scores = print_global_pairwise_scores\n self.print_global_voters = print_global_voters\n\n def map_entity(self, nnid, onlyname=False):\n wikiid = self.nnid2wikiid[nnid]\n wikiname = self.wiki_id_name_map[wikiid].replace(' ', '_') if wikiid != \"\" else \"\"\n return wikiname if onlyname else \"{} {}\".format(wikiid, wikiname)\n\n def process_file(self, el_mode, name, opt_thr):\n self.thr = opt_thr\n self.el_mode = el_mode\n filepath = self.predictions_folder + (\"el/\" if el_mode else \"ed/\") + name\n self.fout = open(filepath, \"w\")\n if self.gm_bucketing:\n self.gm_bucketing.reinitialize()\n\n def file_ended(self):\n self.fout.close()\n if self.gm_bucketing:\n self.gm_bucketing.print()\n\n def scores_text(self, scores_l, scores_names_l, i, j):\n return ' '.join([scores_name + \"=\" + str(score[i][j]) for scores_name, score in zip(scores_names_l, scores_l)])\n\n def process_sample(self, chunkid,\n tp_pred, fp_pred, fn_pred, gt_minus_fn_pred,\n words, words_len, chars, chars_len,\n cand_entities, cand_entities_len,\n final_scores, filtered_spans, scores_l, scores_names_l, gmask, entity_embeddings):\n \"\"\"words: [None] 1d the words of a sample, words_len: scalar,\n chars: [None, None] 2d words, chars of each word, chars_len: [None] for each word\n the length in terms of characters.\n cand_entities: [None, None] gold_mentions, candidates for each gm,\n cand_entitites_len: [None] how many cand ent each gm has.\n filtered_spans = [span1, span2,...] sorted in terms of score. each span is a tuple\n (score, begin_idx, end_idx, best_nnid, simil_score, best_position 1-30, span_num)\n tp_pred and fp_pred is also a list of spans like above and it is also sorted for score.\n fn_pred is a [(gm_num, begin_gm, end_gm, gt)]\"\"\"\n reconstructed_words = []\n for i in range(words_len):\n word = words[i]\n if word != 0:\n reconstructed_words.append(self.id2word[word])\n else: # \n word_chars = []\n for j in range(chars_len[i]):\n word_chars.append(self.id2char[chars[i][j]])\n reconstructed_words.append(''.join(word_chars))\n\n span_num_b_e_gt = sorted(fn_pred+gt_minus_fn_pred)\n\n text_tags = defaultdict(list)\n gt_legend = []\n if len(fn_pred) > 0:\n fnWeakMatcherLogging = FNWeakMatcherLogging(self, filtered_spans, cand_entities,\n cand_entities_len, final_scores, scores_l, scores_names_l,\n reconstructed_words, self.gm_bucketing, gmask, entity_embeddings,\n span_num_b_e_gt)\n for mylist, mycolor in zip([gt_minus_fn_pred, fn_pred], [\"green\", \"red\"]):\n for i, (gm_num, b, e, gt) in enumerate(mylist, 1):\n text_tags[b].append((1, colored(\"[{}\".format(i), mycolor)))\n text_tags[e].append((0, colored(\"]\", mycolor)))\n\n gt_text = \"\"\n if self.el_mode is False: # find the position and the score of the ground truth\n gt_text = \"gt not in candidate entities (recall miss)\"\n for j in range(cand_entities_len[gm_num]):\n if cand_entities[gm_num][j] == gt:\n gt_text = \"gt_p_e_m_pos={}, {}\".format(j,\n self.scores_text(scores_l, scores_names_l, gm_num, j))\n break\n text = \"{}: {} {}\".format(i, self.map_entity(gt), gt_text)\n if mycolor == \"red\":\n text += fnWeakMatcherLogging.check(gm_num, b, e, gt)\n text = colored(text, mycolor)\n gt_legend.append(text)\n\n tp_legend = []\n tp_pred = sorted(tp_pred, key=operator.itemgetter(1))\n for i, (score, b, e, nnid, scores_text, p_e_m_pos, span_num) in enumerate(tp_pred, 1):\n text_tags[b].append((1, colored(\"[{}\".format(i), \"blue\")))\n text_tags[e].append((0, colored(\"]\", \"blue\")))\n\n text = colored(\"{}: {}, score={}, {}, pem_pos={}\".format(i,\n self.map_entity(nnid), score, scores_text, p_e_m_pos), \"blue\")\n tp_legend.append(text)\n if self.gm_bucketing:\n self.gm_bucketing.process_tp(p_e_m_pos, cand_entities_len[span_num])\n\n fp_legend = []\n fp_pairwise_scores_legend = []\n fp_pred = sorted(fp_pred, key=operator.itemgetter(1))\n if len(fp_pred) > 0:\n fpWeakMatcherLogging = FPWeakMatcherLogging(self, span_num_b_e_gt, #fn_pred+gt_minus_fn_pred,\n cand_entities, cand_entities_len,\n final_scores, scores_l, scores_names_l, reconstructed_words, self.gm_bucketing,\n gmask, entity_embeddings)\n for i, (score, b, e, nnid, scores_text, p_e_m_pos, span_num) in enumerate(fp_pred, 1):\n text_tags[b].append((1, colored(\"[{}\".format(i), \"magenta\")))\n text_tags[e].append((0, colored(\"]\", \"magenta\")))\n\n fp_gt_text, pairwise_score_text = fpWeakMatcherLogging.check(b, e, span_num, p_e_m_pos)\n text = \"{}: {}, score={}, {}, pem_pos={} {} \".format(i,\n self.map_entity(nnid), score, scores_text, p_e_m_pos,\n fp_gt_text)\n fp_legend.append(colored(text, \"magenta\"))\n fp_pairwise_scores_legend.append(\"\\n\"+text)\n fp_pairwise_scores_legend.append(pairwise_score_text)\n\n final_acc = [\"new sample \" + chunkid+\"\\n\"+self.extra_info+\"\\n\"]\n for i in range(words_len+1):\n final_acc.extend([text for _, text in sorted(text_tags[i])])\n if i < words_len:\n final_acc.append(reconstructed_words[i])\n self.fout.write(\" \".join(final_acc)+\"\\n\")\n if self.print_global_voters:\n self.fout.write(\"global score voters and weights:\\n\")\n gmask_print_string = self.print_gmask(gmask, span_num_b_e_gt, reconstructed_words, cand_entities)\n self.fout.write(gmask_print_string+\"\\n\")\n self.fout.write(\"\\n\".join(gt_legend + tp_legend + fp_legend))\n if self.print_global_pairwise_scores:\n self.fout.write(colored(\"\\n\".join(fp_pairwise_scores_legend), \"grey\"))\n self.fout.write(\"\\n\")\n\n def print_gmask(self, gmask, span_num_b_e_gt, reconstructed_words, cand_entities):\n i = 0\n document_gmask_acc = []\n for span_num, b, e, gt in span_num_b_e_gt:\n assert(i == span_num)\n text_acc = [\"mention {} {}: \".format(span_num, ' '.join(reconstructed_words[b:e]))]\n for cand_ent_pos in range(gmask.shape[1]):\n mask_value = gmask[span_num][cand_ent_pos]\n assert(mask_value >= 0)\n if mask_value > 0:\n text_acc.append(\"{} {:.2f} | \".format(self.map_entity(cand_entities[span_num][cand_ent_pos]),\n mask_value))\n i += 1\n document_gmask_acc.append(' '.join(text_acc))\n return '\\n'.join(document_gmask_acc)\n\n\nclass FPWeakMatcherLogging(object):\n \"\"\"is initialized with the gm_gt_list i.e. a list of tuples\n (begin_idx, end_idx, gt) and from the list of tuples it builds a data structure. We already\n know that our tuple doesn't match a ground truth. Now we want to find out what exactly happens.\n cases: 1)) doesn't overlap with any gm 2)) overlap with one or more gm. In this case for each gm\n that it overlaps with find a) which is the gt of this gm, b) final_score, sim_score, p_e_m position\n of the gt in my fp tuple.\n structure used: just a list of (begin_idx, end_idx, gt) tuples.\n This one is used only during evaluation.py from the\n metrics_calculation_and_prediction_printing in order to produce logging text\n for the fp\"\"\"\n def __init__(self, printPredictions, span_num_b_e_gt, cand_entities, cand_entities_len,\n final_scores, scores_l, scores_names_l, reconstructed_words, gm_bucketing=None,\n gmask=None, entity_embeddings=None):\n self.printPredictions = printPredictions\n self.data = span_num_b_e_gt\n self.cand_entities = cand_entities\n self.cand_entities_len = cand_entities_len\n self.final_scores = final_scores\n self.scores_l = scores_l\n self.scores_names_l = scores_names_l\n self.reconstructed_words = reconstructed_words\n self.gm_bucketing = gm_bucketing\n self.gmask = gmask\n self.entity_embeddings = entity_embeddings\n\n def check(self, s, e, span_num, winner_pos=None):\n # all the above information that i have for my best_cand_id, now i have to find them\n # for the gt of the gm that overlap with my fp tuple.\n # compare my tuple s, e with all the gm\n\n acc = []\n pairwise_scores_text = \"\"\n for (gm_num, s2, e2, gt) in self.data:\n overlap = False # overlap with this specific gm of the for loop\n if s<=s2 and e<=e2 and s2=s2 and e>=e2 and s=e2:\n overlap = True\n elif s>=s2 and e<=e2:\n overlap = True\n\n if not overlap:\n continue\n\n # add to the text accumulator the info for this gt\n # find gt_score, gt_similarity_score, gt_cand_position\n # check all the candidate entities of this fp span and find where is the gt\n # of course we may not find it at all (recall miss)\n gt_cand_position = -1\n for j in range(self.cand_entities_len[span_num]):\n if self.cand_entities[span_num][j] == gt:\n gt_cand_position = j\n break\n\n if gt_cand_position >= 0:\n acc.append(\"| {}, score={}, {}, pem_pos={}\".format(\n self.printPredictions.map_entity(gt),\n self.final_scores[span_num][gt_cand_position],\n self.printPredictions.scores_text(self.scores_l, self.scores_names_l, span_num, gt_cand_position),\n gt_cand_position))\n else:\n acc.append(\"| {}, recall miss\".format(self.printPredictions.map_entity(gt)))\n if self.printPredictions.print_global_pairwise_scores:\n pairwise_scores_text = print_global_pairwise_voting(self.gmask, self.data, self.reconstructed_words,\n self.cand_entities, self.printPredictions,\n self.entity_embeddings, span_num, winner_pos,\n gt_cand_position)\n\n if acc == []:\n acc.append(\"| no overlap with gm\")\n\n return ' '.join(acc), pairwise_scores_text\n\n\nclass FNWeakMatcherLogging(object):\n \"\"\" This is used to produce text for the FN.\n From the filtered spans i.e. the spans that we keep that do not overlap with each other\n filtered_spans: [(best_cand_score, begin_idx, end_idx, best_cand_id,\n scores_text, best_cand_position, span_num),(),...]\"\"\"\n def __init__(self, printPredictions, filtered_spans, cand_entities, cand_entities_len,\n final_scores, scores_l, scores_names_l, reconstructed_words, gm_bucketing=None,\n gmask=None, entity_embeddings=None, span_num_b_e_gt=None):\n self.printPredictions = printPredictions\n self.data = filtered_spans\n self.cand_entities = cand_entities\n self.cand_entities_len = cand_entities_len\n self.scores_l = scores_l\n self.scores_names_l = scores_names_l\n self.final_scores = final_scores\n self.reconstructed_words = reconstructed_words\n self.gm_bucketing = gm_bucketing\n self.gmask = gmask\n self.entity_embeddings = entity_embeddings\n self.span_num_b_e_gt = span_num_b_e_gt\n\n def check(self, gm_num, s, e, gt):\n # now I will compare each possible span of filtered_spans and for each of them if they overlap\n # with the fn mention I will print what was their winner entity (even if it was below the\n # threshold) and what was the assigned score to the gt (if it was a candidate)\n acc = []\n for (best_cand_score, s2, e2, best_cand_id, scores_text, best_cand_position, span_num) in self.data:\n overlap = False # overlap with this specific filtered span\n if s<=s2 and e<=e2 and s2=s2 and e>=e2 and s=e2:\n overlap = True\n elif s>=s2 and e<=e2:\n overlap = True\n\n if not overlap:\n continue\n\n # add to the text accumulator the info for this filtered span\n # print winner of this span info plus gt info: find gt_score, gt_cand_position\n # check all the candidate entities of this span and find where is the gt\n # of course we may not find it at all (recall miss)\n gt_cand_position = -1\n for j in range(self.cand_entities_len[span_num]):\n if self.cand_entities[span_num][j] == gt:\n gt_cand_position = j\n break\n\n assert(abs(best_cand_score - self.final_scores[span_num][best_cand_position]) < 0.001)\n acc.append(\"[span: {} winner: {}, score={}, {}, pem_pos={}\".format(\n ' '.join(self.reconstructed_words[s2:e2]),\n self.printPredictions.map_entity(best_cand_id),\n best_cand_score,\n self.printPredictions.scores_text(self.scores_l, self.scores_names_l, span_num, best_cand_position),\n best_cand_position))\n if gt_cand_position >= 0:\n acc.append(\" | gt: {}, score={}, {}, pem_pos={} ]\".format(\n self.printPredictions.map_entity(gt),\n self.final_scores[span_num][gt_cand_position],\n self.printPredictions.scores_text(self.scores_l, self.scores_names_l, span_num, gt_cand_position),\n gt_cand_position))\n if self.gm_bucketing:\n self.gm_bucketing.process_fn(gt_cand_position, best_cand_id == gt, self.cand_entities_len[span_num])\n else:\n acc.append(\" | {}, recall miss\".format(self.printPredictions.map_entity(gt)))\n if False and self.printPredictions.print_global_pairwise_scores:\n acc.append(print_global_pairwise_voting(self.gmask, self.span_num_b_e_gt, self.reconstructed_words,\n self.cand_entities, self.printPredictions,\n self.entity_embeddings, span_num, best_cand_position,\n gt_cand_position))\n\n if acc == []:\n acc.append(\" | no overlap with any filtered span\")\n\n return ' '.join(acc)\n\n\n\n# TODO for ed it works well. for EL we have more spans than just the gold mentions that vote.\n# pass as parameters the begin_spans, end_spans from metrics.py\ndef print_global_pairwise_voting(gmask, span_num_b_e_gt, reconstructed_words, cand_entities, printPredictions,\n entity_embeddings, span_num, winner_pos, gt_pos):\n i = 0\n return_acc = [\"'winner & gt' score given by each global voter\"]\n winner_score_sum = 0\n gt_score_sum = 0\n voters_cnt = 0\n for other_span, b, e, _ in span_num_b_e_gt:\n assert(i == other_span)\n if other_span == span_num:\n i += 1\n continue #only the other spans vote\n mention_acc = [\"mention {} {}: \".format(other_span, ' '.join(reconstructed_words[b:e]))]\n for cand_ent_pos in range(gmask.shape[1]):\n mask_value = gmask[other_span][cand_ent_pos]\n assert(mask_value >= 0)\n if mask_value > 0:\n winner_score = np.dot(entity_embeddings[other_span][cand_ent_pos],\n entity_embeddings[span_num][winner_pos]) * mask_value\n gt_score = np.dot(entity_embeddings[other_span][cand_ent_pos],\n entity_embeddings[span_num][gt_pos]) * mask_value\n winner_score_sum += winner_score\n gt_score_sum += gt_score\n voters_cnt += 1\n mention_acc.append(\"{} {:.2f} & {:.2f} |\".format(\n printPredictions.map_entity(cand_entities[other_span][cand_ent_pos], onlyname=True),\n winner_score, gt_score))\n i += 1\n return_acc.append(' '.join(mention_acc))\n return_acc.append(\"global winner_score_avg = {:.2f} gt_score_avg = {:.2f}\".format(\n winner_score_sum/voters_cnt, gt_score_sum/voters_cnt))\n return '\\n'.join(return_acc)\n" }, { "path": "code/evaluation/summarize_all_experiments.py", "content": "import argparse\nimport os\n\n\ndef process_experiment(ed_acc, el_acc, training_name):\n if not os.path.exists(os.path.join(training_name, \"log.txt\")):\n print(\"File doesn't exists: \", os.path.join(training_name, \"log.txt\"))\n return\n if file_is_used(os.path.join(training_name, \"log.txt\")):\n print(\"File is being used by another process. Skip it.\", os.path.join(training_name, \"log.txt\"))\n return\n with open(os.path.join(training_name, \"log.txt\"), \"r\") as fin:\n print(\"file: \", training_name+\"/log.txt\")\n best = dict()\n best[\"ed_dev_f1\"] = 0\n best[\"el_dev_f1\"] = 0\n best[\"ed_test_f1\"] = 0\n best[\"el_test_f1\"] = 0\n\n mode = \"\"\n for line in fin:\n line = line.rstrip()\n if line.startswith(\"args.eval_cnt\"):\n eval_cnt = line[line.rfind(' ')+1:]\n elif line.startswith(\"Evaluating ED datasets\"):\n mode = \"ed\"\n elif line.startswith(\"Evaluating EL datasets\"):\n mode = \"el\"\n elif line.startswith(args.dev_set): #(\"aida_dev.txt\"):\n try:\n micro_line = next(fin)\n macro_line = next(fin)\n line = macro_line if args.macro_or_micro == \"macro\" else micro_line\n dev_f1 = float(line.split()[-1])\n dev_pr = float(line.split()[2])\n dev_re = float(line.split()[4])\n if dev_f1 > best[mode+\"_dev_f1\"]:\n best[mode+\"_dev_f1\"] = dev_f1\n best[mode+\"_dev_pr\"] = dev_pr\n best[mode+\"_dev_re\"] = dev_re\n best[mode+\"_eval_cnt\"] = eval_cnt\n\n # now read forward the test results\n #assert(next(fin).startswith(args.test_set)) #(\"aida_test.txt\"))\n next_line = next(fin)\n while not next_line.startswith(args.test_set):\n next_line = next(fin)\n assert(next_line.startswith(args.test_set)) #(\"aida_test.txt\"))\n micro_line = next(fin)\n macro_line = next(fin)\n line = macro_line if args.macro_or_micro == \"macro\" else micro_line\n best[mode+\"_test_f1\"] = float(line.split()[-1])\n best[mode+\"_test_pr\"] = float(line.split()[2])\n best[mode+\"_test_re\"] = float(line.split()[4])\n\n except StopIteration:\n break\n\n path = training_name[training_name.find(base_folder)+len(base_folder):]\n # print the scores for this log file\n #fixed_no_wikidump_entvecsl2/checkpoints/model-7 model-30.meta\n if \"ed_eval_cnt\" in best:\n checkpoint_text = \"checkpoint_yes\" if os.path.exists(training_name + \"/checkpoints/ed/model-{}.meta\".format(best[\"ed_eval_cnt\"])) else \"checkpoint_no\"\n ed_acc.append((best[\"ed_dev_f1\"], best[\"ed_test_f1\"], path,\n best[\"ed_test_pr\"], best[\"ed_test_re\"], best[\"ed_eval_cnt\"], checkpoint_text, training_name))\n if \"el_eval_cnt\" in best:\n checkpoint_text = \"checkpoint_yes\" if os.path.exists(training_name + \"/checkpoints/el/model-{}.meta\".format(best[\"el_eval_cnt\"])) else \"checkpoint_no\"\n el_acc.append((best[\"el_dev_f1\"], best[\"el_test_f1\"], path,\n best[\"el_test_pr\"], best[\"el_test_re\"], best[\"el_eval_cnt\"], checkpoint_text, training_name))\n\n\ndef process_folder(ed_acc, el_acc, training_name):\n \"\"\"training_name may be a folder with one experiment or a group folder containing many experiment. In\n the second case do recursion on all the subfolders.\"\"\"\n training_name_suffix = os.path.basename(os.path.normpath(training_name))\n if training_name_suffix.startswith(\"group_\") or training_name_suffix.startswith(\"reduced\") or\\\n training_name_suffix.startswith(\"ensemble_\"):\n # then it is a group folder so do recursion on all your subfolders\n d = training_name\n subfolders = [os.path.join(d, o) for o in os.listdir(d)\n if os.path.isdir(os.path.join(d, o))]\n for subfolder in subfolders:\n process_folder(ed_acc, el_acc, subfolder)\n else:\n process_experiment(ed_acc, el_acc, training_name)\n\n\ndef file_is_used(filepath):\n from subprocess import check_output, Popen, PIPE, DEVNULL, STDOUT\n try:\n lsout = Popen(['lsof', filepath], stdout=PIPE, shell=False, stderr=DEVNULL)\n check_output([\"grep\", filepath], stdin=lsout.stdout, shell=False)\n return True\n except:\n #check_output will throw an exception here if it won't find any process using that file\n return False\n\n\ndef main():\n ed_acc = []\n el_acc = []\n if args.group_folder_path:\n process_folder(ed_acc, el_acc, args.group_folder_path)\n else:\n d = base_folder\n # if the base folder is itself an experiment i.e. contains a training_folder and an all_spans_training_folder\n print(os.listdir(d))\n if len([o for o in os.listdir(d) if o in [\"all_spans_training_folder\", \"training_folder\"]]) > 0:\n experiment_names = [d]\n else:\n experiment_names = [os.path.join(d, o) for o in os.listdir(d)\n if os.path.isdir(os.path.join(d, o))]\n print(\"experiment_names =\", experiment_names)\n for experiment_name in experiment_names:\n training_names = []\n for temp in [\"training_folder\", \"all_spans_training_folder\"]:\n d = os.path.join(experiment_name, temp)\n if not os.path.exists(d):\n continue\n training_names.extend([os.path.join(d, o) for o in os.listdir(d)\n if os.path.isdir(os.path.join(d, o))])\n #print(training_names)\n for training_name in training_names:\n process_folder(ed_acc, el_acc, training_name)\n\n ed_acc = sorted(ed_acc, reverse=True)\n el_acc = sorted(el_acc, reverse=True)\n print(\"Dev_score, Test_score, path, test_precision, test_recall, eval_cnt, checkp_existence\")\n print(\"ED Best Scores:\")\n for t in ed_acc:\n print('\\t'.join(map(str, t[:-1])))\n print(\"\\n\\n\\nEL Best Scores:\")\n for t in el_acc:\n print('\\t'.join(map(str, t[:-1])))\n\n\ndef _parse_args():\n parser = argparse.ArgumentParser()\n parser.add_argument(\"--base_folder\", default=\"../../data/tfrecords/\")\n parser.add_argument(\"--macro_or_micro\", default=\"macro\")\n parser.add_argument(\"--dev_set\", default=\"aida_dev.txt\")\n parser.add_argument(\"--test_set\", default=\"aida_test.txt\")\n parser.add_argument(\"--group_folder_path\", default=None)\n return parser.parse_args()\n\n\nif __name__ == \"__main__\":\n args = _parse_args()\n base_folder = os.path.abspath(args.base_folder)\n print(\"base_folder =\", base_folder)\n print(\"group_folder =\", args.group_folder_path)\n main()\n" }, { "path": "code/gerbil/build_entity_universe.py", "content": "\nimport pickle\nfrom nltk.tokenize import word_tokenize\nimport preprocessing.prepro_util as prepro_util\nfrom preprocessing.util import load_wikiid2nnid, reverse_dict, load_wiki_name_id_map, FetchCandidateEntities\nimport os\nimport model.config as config\n\n\nclass BuildEntityUniverse(object):\n def __init__(self):\n self.entities_universe = set()\n self.fetchCandidateEntities = FetchCandidateEntities(Struct())\n prepro_util.args = Struct()\n\n def process(self, text, given_spans):\n # if we wanted to find entities for ed only then restrict it to given_spans instead of all spans\n chunk_words = word_tokenize(text)\n myspans = prepro_util.SamplesGenerator.all_spans(chunk_words)\n for left, right in myspans:\n cand_ent, _ = self.fetchCandidateEntities.process(chunk_words[left:right])\n # cand_ent is a list of strings (i.e. wikiids are still strings) not nums\n if cand_ent:\n self.entities_universe.update(cand_ent)\n\n def flush_entity_universe(self):\n print(\"len(self.entities_universe) =\", len(self.entities_universe))\n entities_folder = config.base_folder+\"data/entities/extension_entities/\"\n _, wiki_id_name_map = load_wiki_name_id_map()\n if not os.path.exists(entities_folder):\n os.makedirs(entities_folder)\n\n def dump_entities(entity_set, name):\n with open(entities_folder + name+\".pickle\", 'wb') as handle:\n pickle.dump(entity_set, handle)\n with open(entities_folder + name+\".txt\", \"w\") as fout:\n for ent_id in entity_set:\n fout.write(ent_id + \"\\t\" + wiki_id_name_map[ent_id].replace(' ', '_') + \"\\n\")\n\n dump_entities(self.entities_universe, \"entities_universe\")\n # now calculate the expansion i.e. from this universe omit the ones that we have already trained\n extension_entity_set = set()\n wikiid2nnid = load_wikiid2nnid()\n for wikiid in self.entities_universe:\n if wikiid not in wikiid2nnid:\n extension_entity_set.add(wikiid)\n\n print(\"len(extension_entity_set) =\", len(extension_entity_set))\n dump_entities(extension_entity_set, \"extension_entities\")\n\n\nclass Struct(object):\n def __init__(self):\n self.p_e_m_choice = \"yago\"\n self.cand_ent_num = 30\n self.lowercase_p_e_m = False\n self.lowercase_spans = False\n self.max_mention_width = 10\n self.spans_separators = [\".\"]\n\n\nif __name__ == \"__main__\":\n pass\n\n" }, { "path": "code/gerbil/gerbil_recall_calculation.py", "content": "\nimport argparse\nimport os\nimport preprocessing.util as util\nimport rdflib\n\n\nclass ProcessDataset(object):\n def __init__(self):\n self.entityNameIdMap = util.EntityNameIdMap()\n self.entityNameIdMap.init_gerbil_compatible_ent_id()\n self.unknown_ent_name = dict()\n self.no_english_uri = dict()\n self.all_gm_cnt = dict()\n self.englishuri_gm_cnt = dict()\n self.valid_gms = dict()\n\n def process(self, filepath, filename):\n #the name of the dataset. just extract the last part of path\n unknown_ent_name = 0\n print(\"Dataset\", filename, filepath)\n g = rdflib.Graph()\n dataset = g.parse(filepath)\n print(\"graph has %s statements.\" % len(g))\n print(g.serialize(format='n3'))\n\n\n\n\n\n\n\n\n\n\n\n\ndef main():\n if not os.path.exists(\"/home/master_thesis_share/data/gerbil/nif_original_datasets/\"):\n os.makedirs(\"/home/master_thesis_share/data/gerbil/nif_original_datasets/\")\n paths_names = [(\"/home/gerbil/gerbil_data/datasets/N3/Reuters-128.ttl\", \"n3_reuters_128\")]\n #paths_names = [(\"/home/gerbil/gerbil_data/datasets/oke-challenge2016/evaluation-data/task1/evaluation-dataset-task1.ttl\", \"oke2016eval\")]\n processDataset = ProcessDataset()\n for filepath, filename in paths_names:\n processDataset.process(filepath, filename)\n\n\nif __name__ == \"__main__\":\n main()\n\n" }, { "path": "code/gerbil/nn_processing.py", "content": "\nfrom model.model_ablations import Model\nfrom time import sleep\nimport tensorflow as tf\nimport pickle\nfrom nltk.tokenize import word_tokenize\nimport preprocessing.prepro_util as prepro_util\nfrom evaluation.metrics import _filtered_spans_and_gm_gt_list\nimport numpy as np\nfrom preprocessing.util import load_wikiid2nnid, reverse_dict, load_wiki_name_id_map, FetchCandidateEntities, \\\n load_persons, FetchFilteredCoreferencedCandEntities\nimport string\n\nclass StreamingSamples(object):\n def __init__(self):\n # those are not used here\n #self.chunk_id, self.ground_truth, self.ground_truth_len, self.begin_gm, self.end_gm\n #self.cand_entities_labels,\n \"\"\"only those are used:\n self.words, self.words_len, self.chars, self.chars_len, \\\n self.begin_span, self.end_span, self.spans_len, \\\n self.cand_entities, self.cand_entities_scores, \\\n self.cand_entities_len\"\"\"\n self.sample = None\n self.empty = True\n\n def new_sample(self, sample):\n self.sample = sample\n self.empty = False\n\n def gen(self):\n while True:\n if not self.empty:\n self.empty = True\n yield self.sample\n else:\n print(\"sleep\")\n sleep(0.5)\n\n\"\"\"\n DT_INT64, DT_INT64, \tDT_INT64, DT_INT64, \tDT_INT64, \tDT_INT64, \t DT_INT64, \tDT_INT64, \t DT_FLOAT, \t\t DT_INT64, \n words, words_len, chars, chars_len, \tbegin_span, \tend_span, spans_len, cand_entities, cand_entities_scores, cand_entities_len, \n\n[?,?], [?], [?,?,?], [?,?], [?,?], [?,?], [?], [?,?,?], [?,?,?], [?,?], \nwords, words_len, chars, chars_len, begin_span, end_span, spans_len, cand_entities, cand_entities_scores, cand_entities_len, \n\"\"\"\n\n\"\"\"(tf.TensorShape([None, None]), tf.TensorShape([None]), tf.TensorShape([None, None, None]), tf.TensorShape([None, None]),\ntf.TensorShape([None, None]), tf.TensorShape([None, None]), tf.TensorShape([None]),\ntf.TensorShape([None, None, None]), tf.TensorShape([None, None, None]), tf.TensorShape([None, None])))\n \n\"\"\"\n\nclass NNProcessing(object):\n def __init__(self, train_args, args):\n self.args = args\n # input pipeline\n self.streaming_samples = StreamingSamples()\n ds = tf.data.Dataset.from_generator(\n self.streaming_samples.gen, (tf.int64, tf.int64, tf.int64, tf.int64, #words, words_len, chars, chars_len\n tf.int64, tf.int64, tf.int64, # begin_span, end_span, span_len\n tf.int64, tf.float32, tf.int64), #cand_entities, cand_entities_scores, cand_entities_len\n (tf.TensorShape([None]), tf.TensorShape([]), tf.TensorShape([None, None]), tf.TensorShape([None]),\n tf.TensorShape([None]), tf.TensorShape([None]), tf.TensorShape([]),\n tf.TensorShape([None, None]), tf.TensorShape([None, None]), tf.TensorShape([None])))\n next_element = ds.make_one_shot_iterator().get_next()\n # batch size = 1 i expand the dims now to match the training that has batch dimension\n next_element = [tf.expand_dims(t, 0) for t in next_element]\n next_element = [None, *next_element[:-1], None, next_element[-1],\n None, None, None, None]\n\n # restore model\n print(\"loading Model:\", train_args.output_folder)\n model = Model(train_args, next_element)\n model.build()\n checkpoint_path = model.restore_session(\"el\" if args.el_mode else \"ed\")\n self.model = model\n if args.hardcoded_thr:\n self.thr = args.hardcoded_thr\n print(\"threshold used:\", self.thr)\n else:\n # optimal threshold recovery from log files.\n # based on the checkpoint selected look at the log file for threshold (otherwise recompute it)\n self.thr = retrieve_optimal_threshold_from_logfile(train_args.output_folder, checkpoint_path, args.el_mode)\n print(\"optimal threshold selected = \", self.thr)\n\n if args.running_mode == \"el_mode\":\n args.el_mode = True\n elif args.running_mode == \"ed_mode\":\n args.el_mode = False\n\n # convert text to tensors for the NN\n with open(args.experiment_folder+\"word_char_maps.pickle\", 'rb') as handle:\n self.word2id, _, self.char2id, _, _, _ = pickle.load(handle)\n\n self.wikiid2nnid = load_wikiid2nnid(extension_name=args.entity_extension)\n self.nnid2wikiid = reverse_dict(self.wikiid2nnid, unique_values=True)\n _, self.wiki_id_name_map = load_wiki_name_id_map()\n\n with open(args.experiment_folder+\"prepro_args.pickle\", 'rb') as handle:\n self.prepro_args = pickle.load(handle)\n if args.lowercase_spans_pem:\n self.prepro_args.lowercase_p_e_m = True\n self.prepro_args.lowercase_spans = True\n print(\"prepro_args:\", self.prepro_args)\n self.prepro_args.persons_coreference = args.persons_coreference\n self.prepro_args.persons_coreference_merge = args.persons_coreference_merge\n self.fetchFilteredCoreferencedCandEntities = FetchFilteredCoreferencedCandEntities(self.prepro_args)\n prepro_util.args = self.prepro_args\n\n self.special_tokenized_words = {\"``\", '\"', \"''\"}\n self.special_words_assertion_errors = 0\n self.gm_idx_errors = 0\n if self.args.el_with_stanfordner_and_our_ed:\n from nltk.tag import StanfordNERTagger\n self.st = StanfordNERTagger(\n '../data/stanford_core_nlp/stanford-ner-2018-02-27/classifiers/english.all.3class.distsim.crf.ser.gz',\n '../data/stanford_core_nlp/stanford-ner-2018-02-27/stanford-ner.jar', encoding='utf-8')\n self.from_myspans_to_given_spans_map_errors = 0\n\n def process(self, text, given_spans):\n self.given_spans = sorted(given_spans) if not self.args.el_mode else given_spans\n self.fetchFilteredCoreferencedCandEntities.init_coref(self.args.el_mode)\n original_words, chunk_words, startidx2wordnum, endidx2wordnum, words2charidx = \\\n self.map_words_to_char_positions(text)\n\n if self.args.el_with_stanfordner_and_our_ed: # el test, use stanford ner to extract spans and decide with our ed system\n self.given_spans, myspans = self.stanford_ner_spans(chunk_words, words2charidx)\n # from given given_spans (start, length) in characters convert them to given_spans in word num\n elif not self.args.el_mode: # simple ed mode. use the spans provided\n myspans = []\n for span in self.given_spans:\n try:\n start, length = span\n end = start+length\n if start not in startidx2wordnum:\n start = self.nearest_idx(start, startidx2wordnum.keys())\n if end not in endidx2wordnum:\n end = self.nearest_idx(end, endidx2wordnum.keys())\n if (start, end-start) != span:\n print(\"given span:\", text[span[0]:span[0]+span[1]], \" new span:\",\n text[start:end])\n myspans.append((startidx2wordnum[start], endidx2wordnum[end]+1))\n except KeyError:\n print(\"Exception KeyError!!!!\")\n print(\"original_words =\", original_words)\n print(\"chunk_words =\", chunk_words)\n print(\"start={}, length={}, left={}, span={}, right={}\".format(start, length,\n text[start-30:start], text[start:start+length], text[start+length:start+length+30]))\n print(\"text =\", text)\n print(\"start= {}\".format(\"in\" if start in startidx2wordnum else \"out\"))\n print(\"end= {}\".format(\"in\" if start + length in endidx2wordnum else \"out\"))\n else: # simple el mode so consider all possible given_spans\n myspans = prepro_util.SamplesGenerator.all_spans(chunk_words)\n # at this point whether we do ed or el by stanfordner_plus_our_ed we must have myspans [word_num_begin, word_num_end)\n # and self.given_spans which are the same spans but with characters [begin_char, length)\n\n begin_spans, end_spans, cand_entities, cand_entities_scores = [], [], [], []\n for left, right in myspans:\n cand_ent, scores = self.fetchFilteredCoreferencedCandEntities.process(left, right, chunk_words)\n cand_ent_filtered, scores_filtered = [], []\n if cand_ent is not None and scores is not None:\n for e, s in zip(cand_ent, scores):\n if e in self.wikiid2nnid:\n cand_ent_filtered.append(self.wikiid2nnid[e])\n scores_filtered.append(s)\n\n if cand_ent_filtered:\n begin_spans.append(left)\n end_spans.append(right)\n cand_entities.append(cand_ent_filtered)\n cand_entities_scores.append(scores_filtered)\n\n if begin_spans == []:\n return [] # this document has no annotation\n\n words = []\n chars = []\n for word in chunk_words:\n words.append(self.word2id[word] if word in self.word2id\n else self.word2id[\"\"])\n chars.append([self.char2id[c] if c in self.char2id else self.char2id[\"\"]\n for c in word])\n chars_len = [len(word) for word in chars]\n new_sample = (words, len(words), list_of_lists_to_2darray(chars), chars_len,\n begin_spans, end_spans, len(begin_spans),\n list_of_lists_to_2darray(cand_entities),\n list_of_lists_to_2darray(cand_entities_scores),\n [len(t) for t in cand_entities])\n self.streaming_samples.new_sample(new_sample)\n\n result_l = self.model.sess.run([self.model.final_scores, self.model.cand_entities_len,\n self.model.cand_entities, self.model.begin_span, self.model.end_span,\n self.model.spans_len], feed_dict={self.model.dropout: 1})\n filtered_spans, _ = _filtered_spans_and_gm_gt_list(0, *result_l, None, None, None, [0], [len(words)])\n\n # based on final_scores and thr return annotations. also translate my given_spans to char given_spans\n print(\"self.special_words_assertion_errors =\", self.special_words_assertion_errors)\n print(\"gm_idx_errors =\", self.gm_idx_errors)\n\n response = []\n for span in filtered_spans:\n score, begin_idx, end_idx, nnid = span\n if score >= self.thr:\n self._add_response_span(response, span, words2charidx)\n print(\"persons_mentions_seen =\", self.fetchFilteredCoreferencedCandEntities.persons_mentions_seen)\n return response\n\n def map_words_to_char_positions(self, text):\n original_words = word_tokenize(text)\n words2charidx = []\n idx = 0\n startidx2wordnum, endidx2wordnum = None, None\n if not self.args.el_mode:\n startidx2wordnum = dict()\n endidx2wordnum = dict()\n given_span_pos = -1\n span_start, span_len = -100, 0\n span_end = -100\n\n chunk_words = [] # correct the special words\n word_num = 0\n\n def insert_word(start, end):\n chunk_words.append(text[start:end])\n words2charidx.append((start, end)) # [..)\n if not self.args.el_mode:\n startidx2wordnum[start] = word_num\n endidx2wordnum[end] = word_num\n\n for word in original_words:\n original_word = word\n if word in self.special_tokenized_words:\n smallest_idx = len(text)\n for special_word in self.special_tokenized_words:\n start = text.find(special_word, idx)\n if start != -1 and start < smallest_idx:\n word = special_word\n smallest_idx = start\n if word != '\"':\n pass\n #print(\"special word replacement: \", original_words[max(0, word_num-2):word_num+2], \"new word:\", word)\n start = text.find(word, idx)\n if start == -1 or start > idx + 10:\n print(\"Assertion Error! in words2charidx. word={}, original_word={}\".format(word, original_word),\n \"near_text={}\\nsnippet={}\".format(text[idx:idx+20], text[idx-50:idx+50]))\n self.special_words_assertion_errors += 1\n for special_word in self.special_tokenized_words:\n start = text.find(special_word, idx)\n print(\"idx=\", idx, \"special_word =\", special_word, \"start=\", start)\n else:\n end = start + len(word)\n idx = end\n if self.args.el_mode:\n insert_word(start, end)\n word_num += 1\n else:\n while span_end <= start and given_span_pos < len(self.given_spans)-1:\n given_span_pos += 1\n span_start, span_len = self.given_spans[given_span_pos]\n span_end = span_start + span_len\n\n inserted_flag = False\n for boundary in [span_start, span_end]:\n if start < boundary < end:\n print(\"given spans fix. original text: \", text)\n print(\"original word: \", text[start:end], word)\n print(\"new split: \", text[start:boundary], \" and \", text[boundary:end])\n insert_word(start, boundary)\n word_num += 1\n insert_word(boundary, end)\n word_num += 1\n print(words2charidx)\n print(startidx2wordnum)\n print(endidx2wordnum)\n inserted_flag = True\n if not inserted_flag:\n insert_word(start, end)\n word_num += 1\n return original_words, chunk_words, startidx2wordnum, endidx2wordnum, words2charidx\n\n def nearest_idx(self, key, values):\n self.gm_idx_errors += 1\n # find the value in values that is nearest to key\n nearest_value = None\n min_distance = 1e+6\n for value in values:\n if abs(key - value) < min_distance:\n nearest_value = value\n min_distance = abs(key-value)\n return nearest_value\n\n def _add_response_span(self, response, span, words2charidx):\n score, begin_idx, end_idx, nnid = span\n start = words2charidx[begin_idx][0] # the word begin_idx starts at this character\n end = words2charidx[end_idx-1][1] # the word begin_idx starts at this character\n wikiid = self.nnid2wikiid[nnid]\n wikiname = self.wiki_id_name_map[wikiid].replace(' ', '_')\n if not self.args.el_mode: # try to match it with a given span\n start, end = self.nearest_given_span(start, end)\n response.append((start, end-start, wikiname))\n\n def nearest_given_span(self, begin_idx, end_idx): # [begin_idx, end_idx) end_idx points to the next character after mention\n min_distance = 1e+6\n nearest_idxes = (-1, -1)\n for (start, length) in self.given_spans:\n distance = abs(begin_idx - start) + abs(end_idx - (start + length))\n if distance < min_distance:\n nearest_idxes = (start, start + length)\n min_distance = distance\n if min_distance > 0:\n self.from_myspans_to_given_spans_map_errors += 1\n return nearest_idxes\n\n def stanford_ner_spans(self, words_l, words2charidx):\n \"\"\"returns a list of tuples (start_idx, length)\"\"\"\n tags = self.st.tag(words_l)\n begin_spans, end_spans, prev_tag = [], [], 'O'\n for i, (_, tag) in enumerate(tags):\n if tag == 'O' and prev_tag != 'O':\n end_spans.append(i)\n elif tag == 'O' and prev_tag == 'O':\n pass\n elif tag != 'O' and prev_tag == 'O':\n begin_spans.append(i)\n elif tag != 'O' and prev_tag == tag:\n pass\n elif tag != 'O' and prev_tag != tag: # and prev_tag != 'O'\n end_spans.append(i)\n begin_spans.append(i)\n prev_tag = tag\n\n char_spans = [] # (begin_char, length)\n word_spans = [] # [begin_word, end_word)\n for bw, ew in zip(begin_spans, end_spans):\n word_spans.append((bw, ew))\n bc = words2charidx[bw][0]\n ec = words2charidx[ew-1][1]\n char_spans.append((bc, ec - bc))\n return char_spans, word_spans\n\n\ndef list_of_lists_to_2darray(a):\n # with padding zeros\n b = np.zeros([len(a), len(max(a, key=lambda x: len(x)))])\n for i, j in enumerate(a):\n b[i][0:len(j)] = j\n return b\n\n\ndef retrieve_optimal_threshold_from_logfile(model_folder, checkpoint_path, el_mode):\n eval_cnt = checkpoint_path[checkpoint_path.rfind(\"-\")+1:] # fixed_no_wikidump_entvecsl2/checkpoints/model-7\n print(\"eval_cnt from checkpoint_path =\", eval_cnt)\n with open(model_folder+\"log.txt\", \"r\") as fin:\n line = next(fin).strip()\n while line != \"args.eval_cnt = \" + eval_cnt:\n line = next(fin).strip()\n line = next(fin).strip()\n while line != \"Evaluating {} datasets\".format(\"EL\" if el_mode else \"ED\"):\n line = next(fin).strip()\n line = next(fin).strip() # Best validation threshold = -0.112 with F1=91.8\n line = line.split()\n assert line[3] == \"=\" and line[5] == \"with\", line\n return float(line[4])\n\n\nif __name__ == \"__main__\":\n pass\n" }, { "path": "code/gerbil/nn_processing_correct.py", "content": "\nfrom model.model_ablations import Model\nfrom time import sleep\nimport tensorflow as tf\nimport pickle\nfrom nltk.tokenize import word_tokenize\nfrom preprocessing.prepro_util import SamplesGenerator\nfrom evaluation.metrics import _filtered_spans_and_gm_gt_list\nimport numpy as np\nfrom preprocessing.util import load_wikiid2nnid, reverse_dict, load_wiki_name_id_map, FetchCandidateEntities, load_persons\n\nclass StreamingSamples(object):\n def __init__(self):\n # those are not used here\n #self.chunk_id, self.ground_truth, self.ground_truth_len, self.begin_gm, self.end_gm\n #self.cand_entities_labels,\n \"\"\"only those are used:\n self.words, self.words_len, self.chars, self.chars_len, \\\n self.begin_span, self.end_span, self.spans_len, \\\n self.cand_entities, self.cand_entities_scores, \\\n self.cand_entities_len\"\"\"\n self.sample = None\n self.empty = True\n\n def new_sample(self, sample):\n self.sample = sample\n self.empty = False\n\n def gen(self):\n while True:\n if not self.empty:\n self.empty = True\n yield self.sample\n else:\n print(\"sleep\")\n sleep(0.5)\n\n\"\"\"\n DT_INT64, DT_INT64, \tDT_INT64, DT_INT64, \tDT_INT64, \tDT_INT64, \t DT_INT64, \tDT_INT64, \t DT_FLOAT, \t\t DT_INT64, \n words, words_len, chars, chars_len, \tbegin_span, \tend_span, spans_len, cand_entities, cand_entities_scores, cand_entities_len, \n\n[?,?], [?], [?,?,?], [?,?], [?,?], [?,?], [?], [?,?,?], [?,?,?], [?,?], \nwords, words_len, chars, chars_len, begin_span, end_span, spans_len, cand_entities, cand_entities_scores, cand_entities_len, \n\"\"\"\n\n\"\"\"(tf.TensorShape([None, None]), tf.TensorShape([None]), tf.TensorShape([None, None, None]), tf.TensorShape([None, None]),\ntf.TensorShape([None, None]), tf.TensorShape([None, None]), tf.TensorShape([None]),\ntf.TensorShape([None, None, None]), tf.TensorShape([None, None, None]), tf.TensorShape([None, None])))\n \n\"\"\"\n\nclass NNProcessing(object):\n def __init__(self, train_args, args):\n self.args = args\n # input pipeline\n self.streaming_samples = StreamingSamples()\n ds = tf.data.Dataset.from_generator(\n self.streaming_samples.gen, (tf.int64, tf.int64, tf.int64, tf.int64, #words, words_len, chars, chars_len\n tf.int64, tf.int64, tf.int64, # begin_span, end_span, span_len\n tf.int64, tf.float32, tf.int64), #cand_entities, cand_entities_scores, cand_entities_len\n (tf.TensorShape([None]), tf.TensorShape([]), tf.TensorShape([None, None]), tf.TensorShape([None]),\n tf.TensorShape([None]), tf.TensorShape([None]), tf.TensorShape([]),\n tf.TensorShape([None, None]), tf.TensorShape([None, None]), tf.TensorShape([None])))\n next_element = ds.make_one_shot_iterator().get_next()\n # batch size = 1 i expand the dims now to match the training that has batch dimension\n next_element = [tf.expand_dims(t, 0) for t in next_element]\n next_element = [None, *next_element[:-1], None, next_element[-1],\n None, None, None, None]\n\n # restore model\n print(\"loading Model:\", train_args.output_folder)\n model = Model(train_args, next_element)\n model.build()\n checkpoint_path = model.restore_session(\"el\" if args.el_mode else \"ed\")\n self.model = model\n if args.hardcoded_thr:\n self.thr = args.hardcoded_thr\n print(\"threshold used:\", self.thr)\n else:\n # optimal threshold recovery from log files.\n # based on the checkpoint selected look at the log file for threshold (otherwise recompute it)\n self.thr = retrieve_optimal_threshold_from_logfile(train_args.output_folder, checkpoint_path, args.el_mode)\n print(\"optimal threshold selected = \", self.thr)\n\n if args.running_mode == \"el_mode\":\n args.el_mode = True\n elif args.running_mode == \"ed_mode\":\n args.el_mode = False\n # else it remains as it is\n if args.manual_thr:\n self.thr = args.manual_thr\n print(\"threshold manually overriden = \", self.thr)\n\n # convert text to tensors for the NN\n with open(args.experiment_folder+\"word_char_maps.pickle\", 'rb') as handle:\n self.word2id, _, self.char2id, _, _, _ = pickle.load(handle)\n\n self.wikiid2nnid = load_wikiid2nnid(extension_name=args.entity_extension)\n self.nnid2wikiid = reverse_dict(self.wikiid2nnid, unique_values=True)\n _, self.wiki_id_name_map = load_wiki_name_id_map()\n\n with open(args.experiment_folder+\"prepro_args.pickle\", 'rb') as handle:\n self.prepro_args = pickle.load(handle)\n if args.lowercase_spans_pem:\n self.prepro_args.lowercase_p_e_m = True\n self.prepro_args.lowercase_spans = True\n print(\"prepro_args:\", self.prepro_args)\n self.fetchCandidateEntities = FetchCandidateEntities(self.prepro_args)\n\n self.special_tokenized_words = {\"``\", '\"', \"''\"}\n self.special_words_assertion_errors = 0\n self.gm_idx_errors = 0\n if self.args.el_with_stanfordner_and_our_ed:\n from nltk.tag import StanfordNERTagger\n self.st = StanfordNERTagger(\n '/home/programs/stanford-ner-2018-02-27/classifiers/english.all.3class.distsim.crf.ser.gz',\n '/home/programs/stanford-ner-2018-02-27/stanford-ner.jar', encoding='utf-8')\n if self.args.persons_coreference:\n self.persons_wikiids = load_persons()\n self.persons_mentions_seen = set()\n\n def process(self, text, given_spans):\n self.persons_mentions_seen = set()\n original_words = word_tokenize(text)\n #print(\"tokenized chunk_words =\", chunk_words)\n words2charidx = []\n idx = 0\n self.given_spans = given_spans\n #chunk_words = parsing_errors(chunk_words)\n # do not use this, not useful for ed either #el has slightly better results without using this\n #aida_test 0.8126 without, 0.8114 with.\n if not self.args.el_mode:\n startidx2wordnum = dict()\n endidx2wordnum = dict()\n chunk_words = [] # correct the special words\n for word_num, word in enumerate(original_words):\n original_word = word\n if word in self.special_tokenized_words:\n smallest_idx = len(text)\n for special_word in self.special_tokenized_words:\n start = text.find(special_word, idx)\n if start != -1 and start < smallest_idx:\n word = special_word\n smallest_idx = start\n if word != '\"':\n pass\n #print(\"special word replacement: \", original_words[max(0, word_num-2):word_num+2], \"new word:\", word)\n start = text.find(word, idx)\n if start == -1 or start > idx + 10:\n print(\"Assertion Error! in words2charidx. word={}, original_word={}\".format(word, original_word),\n \"near_text={}\\nsnippet={}\".format(text[idx:idx+20], text[idx-50:idx+50]))\n self.special_words_assertion_errors += 1\n for special_word in self.special_tokenized_words:\n start = text.find(special_word, idx)\n print(\"idx=\", idx, \"special_word =\", special_word, \"start=\", start)\n else:\n chunk_words.append(word)\n end = start + len(word)\n idx = end\n assert(len(words2charidx) == word_num)\n words2charidx.append((start, end)) # [..)\n if not self.args.el_mode:\n startidx2wordnum[start] = word_num\n endidx2wordnum[end] = word_num\n\n if self.args.el_with_stanfordner_and_our_ed: # el test, use stanford ner to extract spans and decide with our ed system\n self.given_spans, myspans = self.stanford_ner_spans(chunk_words, words2charidx)\n # from given given_spans (start, length) in characters convert them to given_spans in word num\n elif not self.args.el_mode: # simple ed mode. use the spans provided\n self.given_spans = sorted(self.given_spans)\n myspans = []\n for span in self.given_spans:\n try:\n start, length = span\n end = start+length\n if start not in startidx2wordnum:\n start = self.nearest_idx(start, startidx2wordnum.keys())\n if end not in endidx2wordnum:\n end = self.nearest_idx(end, endidx2wordnum.keys())\n if (start, end-start) != span:\n print(\"given span:\", text[span[0]:span[0]+span[1]], \" new span:\",\n text[start:end])\n myspans.append((startidx2wordnum[start], endidx2wordnum[end]+1))\n except KeyError:\n print(\"Exception KeyError!!!!\")\n print(\"original_words =\", original_words)\n print(\"chunk_words =\", chunk_words)\n print(\"start={}, length={}, left={}, span={}, right={}\".format(start, length,\n text[start-30:start], text[start:start+length], text[start+length:start+length+30]))\n print(\"text =\", text)\n print(\"start= {}\".format(\"in\" if start in startidx2wordnum else \"out\"))\n print(\"end= {}\".format(\"in\" if start + length in endidx2wordnum else \"out\"))\n else: # simple el mode\n # consider all possible given_spans\n myspans = SamplesGenerator.all_spans(chunk_words)\n # at this point whether we do ed or el by stanfordner_plus_oured we must have myspans [word_num_b, word_num_end)\n # and self.given_spans which are those spans but with characters [begin_char, length)\n\n begin_spans, end_spans, cand_entities, cand_entities_scores = [], [], [], []\n for left, right in myspans:\n span_text = ' '.join(chunk_words[left:right])\n cand_ent, scores = self.fetchCandidateEntities.process(span_text)\n if self.args.persons_coreference:\n coreference_supermention = self.find_corefence_person(span_text)\n if coreference_supermention:\n print(\"original text:\", chunk_words[max(0, left-4):max(len(chunk_words), right+4)])\n if not self.args.persons_coreference_merge:\n cand_ent, scores = self.fetchCandidateEntities.process(coreference_supermention)\n else: # merge with cand_ent and scores\n cand_ent2, scores2 = self.fetchCandidateEntities.process(coreference_supermention)\n temp1 = list(zip(scores, cand_ent))\n temp2 = list(zip(scores2, cand_ent2))\n temp3 = sorted(temp1 + temp2, reverse=True)\n scores, cand_ent = map(list, zip(*temp3[:self.prepro_args.cand_ent_num]))\n\n # ['Obama_e', 'ent2', 'ent3'] , [0.9, 0.2, 0.8]\n # filter out entities that are not in our universe (and its corresponding scores)\n # then encode it from wikiid2nnid\n # similar to prepro_util._encode_cand_entities_and_labels\n cand_ent_filtered, scores_filtered = [], []\n if cand_ent is not None and scores is not None:\n if self.args.persons_coreference and not coreference_supermention and \\\n cand_ent[0] in self.persons_wikiids and len(span_text) >= 3:\n self.persons_mentions_seen.add(span_text)\n for e, s in zip(cand_ent, scores):\n if e in self.wikiid2nnid:\n cand_ent_filtered.append(self.wikiid2nnid[e])\n scores_filtered.append(s)\n\n if cand_ent_filtered:\n begin_spans.append(left)\n end_spans.append(right)\n cand_entities.append(cand_ent_filtered)\n cand_entities_scores.append(scores_filtered)\n\n if begin_spans == []:\n return [] # this document has no annotation\n\n words = []\n chars = []\n for word in chunk_words:\n words.append(self.word2id[word] if word in self.word2id\n else self.word2id[\"\"])\n chars.append([self.char2id[c] if c in self.char2id else self.char2id[\"\"]\n for c in word])\n chars_len = [len(word) for word in chars]\n new_sample = (words, len(words), list_of_lists_to_2darray(chars), chars_len,\n begin_spans, end_spans, len(begin_spans),\n list_of_lists_to_2darray(cand_entities),\n list_of_lists_to_2darray(cand_entities_scores),\n [len(t) for t in cand_entities])\n self.streaming_samples.new_sample(new_sample)\n\n result_l = self.model.sess.run([self.model.final_scores, self.model.cand_entities_len,\n self.model.cand_entities, self.model.begin_span, self.model.end_span,\n self.model.spans_len], feed_dict={self.model.dropout: 1})\n filtered_spans, _ = _filtered_spans_and_gm_gt_list(0, *result_l, None, None, None, [0], [len(words)])\n\n # based on final_scores and thr return annotations. also translate my given_spans to char given_spans\n\n print(\"self.special_words_assertion_errors =\", self.special_words_assertion_errors)\n print(\"gm_idx_errors =\", self.gm_idx_errors)\n\n if self.args.each_entity_only_once or self.args.each_mention_only_once or \\\n self.args.omit_first_sentence:\n return self.custom_response(filtered_spans, text, words2charidx, chunk_words)\n response = []\n for span in filtered_spans:\n score, begin_idx, end_idx, nnid = span\n if score >= self.thr:\n self._add_response_span(response, span, words2charidx)\n print(\"self.persons_mentions_seen =\", self.persons_mentions_seen)\n return response\n\n def find_corefence_person(self, span_text):\n \"\"\"if span_text is substring of another person mention found before. it should be\n substring of words. so check next character and previous character to be non alphanumeric\"\"\"\n if len(span_text) < 3:\n return None\n for mention in self.persons_mentions_seen:\n idx = mention.find(span_text)\n if idx != -1:\n #print(\"find_coreference_person substring initial match\")\n if len(mention) == len(span_text):\n continue # they are identical so no point in substituting them\n if idx > 0 and mention[idx-1].isalpha():\n continue\n if idx < len(mention) - 1 and mention[idx+1].isalpha():\n continue\n print(\"persons coreference, before:\", span_text, \"after:\", mention)\n return mention\n return None\n\n\n def nearest_idx(self, key, values):\n self.gm_idx_errors += 1\n # find the value in values that is nearest to key\n nearest_value = None\n min_distance = 1e+6\n for value in values:\n if abs(key - value) < min_distance:\n nearest_value = value\n min_distance = abs(key-value)\n return nearest_value\n\n def _add_response_span(self, response, span, words2charidx):\n score, begin_idx, end_idx, nnid = span\n start = words2charidx[begin_idx][0] # the word begin_idx starts at this character\n end = words2charidx[end_idx-1][1] # the word begin_idx starts at this character\n wikiid = self.nnid2wikiid[nnid]\n wikiname = self.wiki_id_name_map[wikiid].replace(' ', '_')\n if not self.args.el_mode: # try to match it with a given span\n start, end = self.nearest_given_span(start, end)\n response.append((start, end-start, wikiname))\n\n def nearest_given_span(self, begin_idx, end_idx): # [begin_idx, end_idx) end_idx points to the next character after mention\n min_distance = 1e+6\n nearest_idxes = (-1, -1)\n for (start, length) in self.given_spans:\n distance = abs(begin_idx - start) + abs(end_idx - (start + length))\n if distance < min_distance:\n nearest_idxes = (start, start + length)\n min_distance = distance\n return nearest_idxes\n\n def custom_response(self, filtered_spans, text, words2charidx, chunk_words):\n from operator import itemgetter\n filtered_spans = sorted(filtered_spans, key=itemgetter(1))\n response = []\n # omit title\n if self.args.omit_first_sentence:\n start = text.find(self.args.first_sentence_separator)\n if start > 100: # this dataset doesn't have a title so do not omit anything\n start = 0\n print(\"omit first sentence:\", text[:start])\n for i, span in enumerate(filtered_spans):\n if words2charidx[span[1]][0] > start:\n break\n print(\"omitted annotations:\", [chunk_words[span[1]:span[2]] for span in filtered_spans[:i] if span[0] > self.thr])\n filtered_spans = filtered_spans[i:]\n\n # each entity only once\n if self.args.each_entity_only_once:\n used_entities = set()\n for span in filtered_spans:\n score, begin_idx, end_idx, nnid = span\n if score >= self.thr and nnid not in used_entities:\n self._add_response_span(response, span, words2charidx)\n used_entities.add(nnid)\n elif self.args.each_mention_only_once:\n used_mentions = set()\n for span in filtered_spans:\n score, begin_idx, end_idx, nnid = span\n mention = text[words2charidx[begin_idx][0]:words2charidx[end_idx-1][1]]\n if score >= self.thr and mention not in used_mentions:\n self._add_response_span(response, span, words2charidx)\n used_mentions.add(mention)\n\n return response\n\n def stanford_ner_spans(self, words_l, words2charidx):\n \"\"\"returns a list of tuples (start_idx, length)\"\"\"\n tags = self.st.tag(words_l)\n begin_spans, end_spans, prev_tag = [], [], 'O'\n for i, (_, tag) in enumerate(tags):\n if tag == 'O' and prev_tag != 'O':\n end_spans.append(i)\n elif tag == 'O' and prev_tag == 'O':\n pass\n elif tag != 'O' and prev_tag == 'O':\n begin_spans.append(i)\n elif tag != 'O' and prev_tag == tag:\n pass\n elif tag != 'O' and prev_tag != tag: # and prev_tag != 'O'\n end_spans.append(i)\n begin_spans.append(i)\n prev_tag = tag\n\n char_spans = [] # (begin_char, length)\n word_spans = [] # [begin_word, end_word)\n for bw, ew in zip(begin_spans, end_spans):\n word_spans.append((bw, ew))\n bc = words2charidx[bw][0]\n ec = words2charidx[ew-1][1]\n char_spans.append((bc, ec - bc))\n return char_spans, word_spans\n\n\n\"\"\" not used\ndef parsing_errors(chunk_words):\n # check each chunk_word if it is alpha. if not then try to split it\n # 'U.S' '.' merge them to U.S.\n # alpha notalaphcharacter alpha do not split them\n temp = []\n i = 0\n while i < len(chunk_words):\n word = chunk_words[i]\n if word == 'U.S.':\n temp.append(word)\n elif word == 'U.S' and chunk_words[i+1] == '.':\n temp.append('U.S.')\n i += 1\n elif word.isalpha():\n temp.append(word)\n else:\n temp_idx = 0\n for c in word:\n if not c.isalpha():\n break\n temp_idx += 1\n if 0 < temp_idx < len(word) and not word[temp_idx+1:].isalpha():\n temp.append(word[:temp_idx])\n temp.append(word[temp_idx:])\n else:\n temp.append(word)\n i += 1\n return temp\n\"\"\"\n\ndef list_of_lists_to_2darray(a):\n # with padding zeros\n b = np.zeros([len(a), len(max(a, key=lambda x: len(x)))])\n for i, j in enumerate(a):\n b[i][0:len(j)] = j\n return b\n\n\ndef retrieve_optimal_threshold_from_logfile(model_folder, checkpoint_path, el_mode):\n eval_cnt = checkpoint_path[checkpoint_path.rfind(\"-\")+1:] # fixed_no_wikidump_entvecsl2/checkpoints/model-7\n print(\"eval_cnt from checkpoint_path =\", eval_cnt)\n with open(model_folder+\"log.txt\", \"r\") as fin:\n line = next(fin).strip()\n while line != \"args.eval_cnt = \" + eval_cnt:\n line = next(fin).strip()\n line = next(fin).strip()\n while line != \"Evaluating {} datasets\".format(\"EL\" if el_mode else \"ED\"):\n line = next(fin).strip()\n line = next(fin).strip() # Best validation threshold = -0.112 with F1=91.8\n line = line.split()\n assert line[3] == \"=\" and line[5] == \"with\", line\n return float(line[4])\n\n\n\nif __name__ == \"__main__\":\n debug = True" }, { "path": "code/gerbil/server.py", "content": "from http.server import BaseHTTPRequestHandler, HTTPServer\nimport json\nimport argparse\nimport model.config as config\nfrom gerbil.nn_processing import NNProcessing\nfrom model.util import load_train_args\nfrom gerbil.build_entity_universe import BuildEntityUniverse\n\n\nclass GetHandler(BaseHTTPRequestHandler):\n\n def do_POST(self):\n content_length = int(self.headers['Content-Length'])\n post_data = self.rfile.read(content_length)\n self.send_response(200)\n self.end_headers()\n\n if args.build_entity_universe:\n buildEntityUniverse.process(*read_json(post_data))\n response = []\n else:\n response = nnprocessing.process(*read_json(post_data))\n\n print(\"response in server.py code:\\n\", response)\n self.wfile.write(bytes(json.dumps(response), \"utf-8\"))\n return\n\n\ndef read_json(post_data):\n data = json.loads(post_data.decode(\"utf-8\"))\n #print(\"received data:\", data)\n text = data[\"text\"]\n spans = [(int(j[\"start\"]), int(j[\"length\"])) for j in data[\"spans\"]]\n return text, spans\n\n\ndef _parse_args():\n parser = argparse.ArgumentParser()\n parser.add_argument(\"--experiment_name\", default=\"per_document_no_wikidump\",\n help=\"under folder data/tfrecords/\")\n parser.add_argument(\"--training_name\", default=\"doc_fixed_nowiki_evecsl2dropout\")\n parser.add_argument(\"--all_spans_training\", type=bool, default=False)\n parser.add_argument(\"--el_mode\", dest='el_mode', action='store_true')\n parser.add_argument(\"--ed_mode\", dest='el_mode', action='store_false')\n parser.set_defaults(el_mode=True)\n\n parser.add_argument(\"--running_mode\", default=None, help=\"el_mode or ed_mode, so\"\n \"we can restore an ed_mode model and run it for el\")\n\n parser.add_argument(\"--lowercase_spans_pem\", type=bool, default=False)\n\n parser.add_argument(\"--entity_extension\", default=None, help=\"extension_entities or extension_entities_all etc\")\n\n # those are for building the entity set\n parser.add_argument(\"--build_entity_universe\", type=bool, default=False)\n parser.add_argument(\"--hardcoded_thr\", type=float, default=None, help=\"0, 0.2\")\n parser.add_argument(\"--el_with_stanfordner_and_our_ed\", type=bool, default=False)\n\n parser.add_argument(\"--persons_coreference\", type=bool, default=False)\n parser.add_argument(\"--persons_coreference_merge\", type=bool, default=False)\n\n args = parser.parse_args()\n if args.persons_coreference_merge:\n args.persons_coreference = True\n print(args)\n if args.build_entity_universe:\n return args, None\n\n temp = \"all_spans_\" if args.all_spans_training else \"\"\n args.experiment_folder = config.base_folder+\"data/tfrecords/\" + args.experiment_name+\"/\"\n\n args.output_folder = config.base_folder+\"data/tfrecords/\" + \\\n args.experiment_name+\"/{}training_folder/\".format(temp) + \\\n args.training_name+\"/\"\n\n train_args = load_train_args(args.output_folder, \"gerbil\")\n train_args.entity_extension = args.entity_extension\n\n print(train_args)\n return args, train_args\n\n\ndef terminate():\n tee.close()\n if args.build_entity_universe:\n buildEntityUniverse.flush_entity_universe()\n else:\n print(\"from_myspans_to_given_spans_map_errors:\", nnprocessing.from_myspans_to_given_spans_map_errors)\n\n\nif __name__ == \"__main__\":\n args, train_args = _parse_args()\n if args.build_entity_universe:\n buildEntityUniverse = BuildEntityUniverse()\n else:\n nnprocessing = NNProcessing(train_args, args)\n server = HTTPServer(('localhost', 5555), GetHandler)\n print('Starting server at http://localhost:5555')\n from model.util import Tee\n tee = Tee('server.txt', 'w')\n try:\n server.serve_forever()\n except KeyboardInterrupt:\n terminate()\n exit(0)\n" }, { "path": "code/model/base_model.py", "content": "import os\nimport tensorflow as tf\n\n\nclass BaseModel(object):\n\n def __init__(self, args):\n self.args = args\n self.sess = None\n self.ed_saver = None\n self.el_saver = None\n\n def reinitialize_weights(self, scope_name):\n \"\"\"Reinitializes the weights of a given layer\"\"\"\n variables = tf.contrib.framework.get_variables(scope_name)\n init = tf.variables_initializer(variables)\n self.sess.run(init)\n\n def add_train_op(self, lr_method, lr, loss, clip=-1):\n \"\"\"Defines self.train_op that performs an update on a batch\n Args:\n lr_method: (string) sgd method, for example \"adam\"\n lr: (tf.placeholder) tf.float32, learning rate\n loss: (tensor) tf.float32 loss to minimize\n clip: (python float) clipping of gradient. If < 0, no clipping\n \"\"\"\n _lr_m = lr_method.lower() # lower to make sure\n\n with tf.variable_scope(\"train_step\"):\n if _lr_m == 'adam': # sgd method\n optimizer = tf.train.AdamOptimizer(lr)\n elif _lr_m == 'adagrad':\n optimizer = tf.train.AdagradOptimizer(lr)\n elif _lr_m == 'sgd':\n optimizer = tf.train.GradientDescentOptimizer(lr)\n elif _lr_m == 'rmsprop':\n optimizer = tf.train.RMSPropOptimizer(lr)\n else:\n raise NotImplementedError(\"Unknown method {}\".format(_lr_m))\n\n if clip > 0: # gradient clipping if clip is positive\n grads, vs = zip(*optimizer.compute_gradients(loss))\n grads, gnorm = tf.clip_by_global_norm(grads, clip)\n self.train_op = optimizer.apply_gradients(zip(grads, vs))\n else:\n self.train_op = optimizer.minimize(loss)\n\n def initialize_session(self):\n \"\"\"Defines self.sess and initialize the variables\"\"\"\n print(\"Initializing tf session\")\n #config = tf.ConfigProto()\n #config.gpu_options.allow_growth = True\n #self.sess = tf.Session(config=config)\n self.sess = tf.Session()\n #from tensorflow.python import debug as tf_debug\n #self.sess = tf_debug.LocalCLIDebugWrapperSession(tf.Session())\n self.sess.run(tf.global_variables_initializer())\n self.ed_saver = tf.train.Saver(var_list=self.checkpoint_variables(), max_to_keep=self.args.checkpoints_num)\n self.el_saver = tf.train.Saver(var_list=self.checkpoint_variables(), max_to_keep=self.args.checkpoints_num)\n\n def restore_session(self, option=\"latest\"):\n \"\"\"option: 'latest', 'ed', 'el' so it chooses the latest checkpoint for ed for el or from\n both of them if it is 'latest' (this is used in continue_training'\"\"\"\n \"\"\"restores from the latest checkpoint of this folder\"\"\"\n assert(option in [\"latest\", \"ed\", \"el\"])\n if hasattr(self.args, 'checkpoint_model_num') and self.args.checkpoint_model_num is not None:\n assert(option != \"latest\") # it is either ed or el\n checkpoint_path = self.args.checkpoints_folder + option + \"/model-{}\".format(self.args.checkpoint_model_num)\n else:\n if option == \"ed\":\n checkpoint_path = self.my_latest_checkpoint(self.args.checkpoints_folder+\"ed/\")\n elif option == \"el\":\n checkpoint_path = self.my_latest_checkpoint(self.args.checkpoints_folder+\"el/\")\n elif option == \"latest\":\n print(\"Reloading the latest trained model...(either ed or el)\")\n ed = self.my_latest_checkpoint(self.args.checkpoints_folder+\"ed/\")\n el = self.my_latest_checkpoint(self.args.checkpoints_folder+\"el/\")\n ed_eval_cnt = int(ed[ed.rfind('-') + 1:])\n el_eval_cnt = int(el[el.rfind('-') + 1:])\n if ed_eval_cnt >= el_eval_cnt:\n checkpoint_path = self.my_latest_checkpoint(self.args.checkpoints_folder+\"ed/\")\n option = \"ed\"\n else:\n checkpoint_path = self.my_latest_checkpoint(self.args.checkpoints_folder+\"el/\")\n option = \"el\"\n print(\"Using checkpoint: {}\".format(checkpoint_path))\n self.sess = tf.Session()\n self.ed_saver = tf.train.Saver(var_list=self.checkpoint_variables(), max_to_keep=self.args.checkpoints_num)\n self.el_saver = tf.train.Saver(var_list=self.checkpoint_variables(), max_to_keep=self.args.checkpoints_num)\n\n saver = self.ed_saver if option == \"ed\" else self.el_saver\n saver.restore(self.sess, checkpoint_path)\n self.init_embeddings()\n print(\"Finished loading checkpoint.\")\n return checkpoint_path\n\n def my_latest_checkpoint(self, folder_path): # model-9.meta\n files = [name for name in os.listdir(folder_path) if name.startswith(\"model\") and name.endswith(\"meta\")]\n max_epoch = max([int(name[len(\"model-\"):-len(\".meta\")]) for name in files])\n return folder_path + \"model-\" + str(max_epoch)\n\n def save_session(self, eval_cnt, save_ed_flag, save_el_flag):\n \"\"\"Saves session = weights\"\"\"\n for save_flag, category in zip([save_ed_flag, save_el_flag], [\"ed\", \"el\"]):\n if save_flag is False:\n continue\n checkpoints_folder = self.args.checkpoints_folder + category + \"/\"\n if not os.path.exists(checkpoints_folder):\n os.makedirs(checkpoints_folder)\n print(\"saving session checkpoint for {}...\".format(category))\n checkpoint_prefix = os.path.join(checkpoints_folder, \"model\")\n saver = self.ed_saver if category == \"ed\" else self.el_saver\n save_path = saver.save(self.sess, checkpoint_prefix, global_step=eval_cnt)\n print(\"Checkpoint saved in file: %s\" % save_path)\n\n def close_session(self):\n \"\"\"Closes the session\"\"\"\n self.sess.close()\n\n def _restore_list(self):\n return [n for n in tf.global_variables()\n if n.name != 'entities/_entity_embeddings:0']\n\n def checkpoint_variables(self):\n \"\"\"omit word embeddings and entity embeddings from being stored in checkpoint when they are fixed\n in order to save disk space. word emb are always fixed, entity emb are fixed when\n args.train_ent_vecs == False\"\"\"\n omit_variables = ['words/_word_embeddings:0']\n if not self.args.train_ent_vecs:\n omit_variables.append('entities/_entity_embeddings:0')\n variables = [n for n in tf.global_variables() if n.name not in omit_variables]\n print(\"checkpoint variables to restore:\", variables)\n return variables\n\n def find_variable_handler_by_name(self, var_name):\n for n in tf.global_variables():\n if n.name == var_name:\n return n" }, { "path": "code/model/config.py", "content": "base_folder = \"../\"\nspans_separators = [\".\"] #maybe also try ['.', ',', ';']\n\nunk_ent_id = \"0\"\n\n" }, { "path": "code/model/ed_model_original.py", "content": "import numpy as np\nimport pickle\nimport tensorflow as tf\nimport model.config as config\nfrom .base_model import BaseModel\nimport model.util as util\n\n\nclass EDModel(BaseModel):\n def __init__(self, args, next_element):\n super().__init__(args)\n self.chunk_id, self.words, self.words_len, self.chars, self.chars_len,\\\n self.begin_span, self.end_span, self.spans_len,\\\n self.cand_entities, self.cand_entities_scores, self.cand_entities_labels,\\\n self.cand_entities_len, self.ground_truth, self.ground_truth_len,\\\n self.begin_gm, self.end_gm = next_element\n\n self.begin_span = tf.cast(self.begin_span, tf.int32)\n self.end_span = tf.cast(self.end_span, tf.int32)\n self.words_len = tf.cast(self.words_len, tf.int32) # TODO new command caution if it is breaking old models\n with open(config.base_folder +\"data/tfrecords/\" + self.args.experiment_name+\n \"/word_char_maps.pickle\", 'rb') as handle:\n _, id2word, _, id2char, _, _ = pickle.load(handle)\n self.nwords = len(id2word)\n self.nchars = len(id2char)\n if self.args.cand_ent_num_restriction:\n self.cand_entities = tf.slice(self.cand_entities, [0, 0, 0], [-1, -1, self.args.cand_ent_num_restriction])\n self.cand_entities_scores = tf.slice(self.cand_entities_scores, [0, 0, 0], [-1, -1, self.args.cand_ent_num_restriction])\n self.cand_entities_labels = tf.slice(self.cand_entities_labels, [0, 0, 0], [-1, -1, self.args.cand_ent_num_restriction])\n self.cand_entities_len = tf.minimum(self.cand_entities_len, self.args.cand_ent_num_restriction)\n self.ffnn_l2normalization_op_list = []\n\n if not tf.__version__.startswith(\"1.4\"):\n temp_shape = tf.shape(self.cand_entities_len)\n temp = tf.sequence_mask(tf.reshape(self.cand_entities_len, [-1]), tf.shape(self.cand_entities_scores)[2], dtype=tf.float32)\n self.loss_mask = tf.reshape(temp, [temp_shape[0], temp_shape[1], tf.shape(temp)[-1]])\n else:\n self.loss_mask = tf.sequence_mask(self.cand_entities_len, tf.shape(self.cand_entities_scores)[2], dtype=tf.float32) # 30 candidates\n\n\n def add_placeholders(self):\n \"\"\"Define placeholders = entries to computational graph\"\"\"\n \"\"\"\n # shape = (batch size, max length of sentence in batch)\n self.words = tf.placeholder(tf.int64, shape=[None, None],\n name=\"words\")\n\n # shape = (batch size)\n self.words_len = tf.placeholder(tf.int64, shape=[None],\n name=\"words_len\")\n\n # shape = (batch size, max length of sentence, max length of word)\n self.chars = tf.placeholder(tf.int64, shape=[None, None, None],\n name=\"chars\")\n\n # shape = (batch_size, max_length of sentence)\n self.chars_len = tf.placeholder(tf.int64, shape=[None, None],\n name=\"chars_len\")\n\n # shape = (batch_size, max number of candidate spans in one of the batch sentences)\n self.begin_span = tf.placeholder(tf.int64, shape=[None, None],\n name=\"begin_span\")\n self.end_span = tf.placeholder(tf.int64, shape=[None, None],\n name=\"end_span\")\n\n # shape = (batch size)\n self.spans_len = tf.placeholder(tf.int64, shape=[None],\n name=\"spans_len\")\n\n # shape = (batch size, max number of candidate spans, max number of cand entitites)\n self.cand_entities = tf.placeholder(tf.int64, shape=[None, None, None],\n name=\"cand_entities\")\n self.cand_entities_scores = tf.placeholder(tf.float32, shape=[None, None, None],\n name=\"cand_entities_scores\")\n self.cand_entities_labels = tf.placeholder(tf.int64, shape=[None, None, None],\n name=\"cand_entities_labels\")\n # shape = (batch_size, max number of candidate spans)\n self.cand_entities_len = tf.placeholder(tf.int64, shape=[None, None],\n name=\"cand_entities_len\")\n\n # shape = (batch_size, max number of candidate spans)\n self.ground_truth = tf.placeholder(tf.int64, shape=[None, None],\n name=\"ground_truth\")\n\n # shape = (batch_size)\n self.ground_truth_len = tf.placeholder(tf.int64, shape=[None],\n name=\"ground_truth_len\")\n\n # the next two placeholders are not useful. TODO remove them\n # shape = (batch_size, max number of gold mentions)\n \n self.begin_gm = tf.placeholder(tf.int64, shape=[None, None],\n name=\"begin_gm\")\n self.end_gm = tf.placeholder(tf.int64, shape=[None, None],\n name=\"end_gm\")\n \"\"\"\n\n # hyper parameters\n self.dropout = tf.placeholder(dtype=tf.float32, shape=[], name=\"dropout\")\n self.lr = tf.placeholder(dtype=tf.float32, shape=[], name=\"lr\")\n\n\n def init_embeddings(self):\n print(\"\\n!!!! init embeddings !!!!\\n\")\n # read the numpy file\n embeddings_nparray = np.load(config.base_folder +\"data/tfrecords/\" + self.args.experiment_name+\n \"/embeddings_array.npy\")\n self.sess.run(self.word_embedding_init, feed_dict={self.word_embeddings_placeholder: embeddings_nparray})\n\n entity_embeddings_nparray = util.load_ent_vecs(self.args)\n self.sess.run(self.entity_embedding_init, feed_dict={self.entity_embeddings_placeholder: entity_embeddings_nparray})\n\n def add_embeddings_op(self):\n\n \"\"\"Defines self.word_embeddings\"\"\"\n #with tf.device('/cpu:0'), tf.name_scope(\"embedding\"):\n with tf.variable_scope(\"words\"):\n _word_embeddings = tf.Variable(\n tf.constant(0.0, shape=[self.nwords, 300]),\n name=\"_word_embeddings\",\n dtype=tf.float32,\n trainable=False)\n\n self.word_embeddings_placeholder = tf.placeholder(tf.float32, [self.nwords, 300])\n self.word_embedding_init = _word_embeddings.assign(self.word_embeddings_placeholder)\n\n word_embeddings = tf.nn.embedding_lookup(_word_embeddings,\n self.words, name=\"word_embeddings\")\n self.pure_word_embeddings = word_embeddings\n #print(\"word_embeddings (after lookup) \", word_embeddings)\n\n with tf.variable_scope(\"chars\"):\n if self.args.use_chars:\n # get char embeddings matrix\n _char_embeddings = tf.get_variable(\n name=\"_char_embeddings\",\n dtype=tf.float32,\n shape=[self.nchars, self.args.dim_char], trainable=True)\n char_embeddings = tf.nn.embedding_lookup(_char_embeddings,\n self.chars, name=\"char_embeddings\")\n\n # put the time dimension on axis=1\n s = tf.shape(char_embeddings)\n char_embeddings = tf.reshape(char_embeddings,\n shape=[s[0] * s[1], s[-2], self.args.dim_char])\n char_lengths = tf.reshape(self.chars_len, shape=[s[0] * s[1]])\n\n # bi lstm on chars\n cell_fw = tf.contrib.rnn.LSTMCell(self.args.hidden_size_char,\n state_is_tuple=True)\n cell_bw = tf.contrib.rnn.LSTMCell(self.args.hidden_size_char,\n state_is_tuple=True)\n _output = tf.nn.bidirectional_dynamic_rnn(\n cell_fw, cell_bw, char_embeddings,\n sequence_length=char_lengths, dtype=tf.float32)\n\n # read and concat output\n _, ((_, output_fw), (_, output_bw)) = _output\n output = tf.concat([output_fw, output_bw], axis=-1)\n\n # shape = (batch size, max sentence length, char hidden size)\n output = tf.reshape(output,\n shape=[s[0], s[1], 2 * self.args.hidden_size_char])\n #print(\"output after char lstm \", output)\n word_embeddings = tf.concat([word_embeddings, output], axis=-1)\n #print(\"word_embeddings with char after concatenation \", word_embeddings)\n # (batch, words, 300+2*100)\n\n self.word_embeddings = tf.nn.dropout(word_embeddings, self.dropout)\n\n with tf.variable_scope(\"entities\"):\n from preprocessing.util import load_wikiid2nnid\n self.nentities = len(load_wikiid2nnid(extension_name=self.args.entity_extension))\n _entity_embeddings = tf.Variable(\n tf.constant(0.0, shape=[self.nentities, 300]),\n name=\"_entity_embeddings\",\n dtype=tf.float32,\n trainable=self.args.train_ent_vecs)\n\n self.entity_embeddings_placeholder = tf.placeholder(tf.float32, [self.nentities, 300])\n self.entity_embedding_init = _entity_embeddings.assign(self.entity_embeddings_placeholder)\n\n self.entity_embeddings = tf.nn.embedding_lookup(_entity_embeddings, self.cand_entities,\n name=\"entity_embeddings\")\n self.pure_entity_embeddings = self.entity_embeddings\n if self.args.ent_vecs_regularization.startswith(\"l2\"): # 'l2' or 'l2dropout'\n self.entity_embeddings = tf.nn.l2_normalize(self.entity_embeddings, dim=3)\n if self.args.ent_vecs_regularization == \"dropout\" or \\\n self.args.ent_vecs_regularization == \"l2dropout\":\n self.entity_embeddings = tf.nn.dropout(self.entity_embeddings, self.dropout)\n\n #print(\"entity_embeddings = \", self.entity_embeddings)\n\n\n def add_context_emb_op(self):\n with tf.variable_scope(\"context-bi-lstm\"):\n cell_fw = tf.contrib.rnn.LSTMCell(self.args.hidden_size_lstm)\n cell_bw = tf.contrib.rnn.LSTMCell(self.args.hidden_size_lstm)\n (output_fw, output_bw), _ = tf.nn.bidirectional_dynamic_rnn(\n cell_fw, cell_bw, self.word_embeddings,\n sequence_length=self.words_len, dtype=tf.float32)\n output = tf.concat([output_fw, output_bw], axis=-1)\n self.context_emb = tf.nn.dropout(output, self.dropout)\n #print(\"context_emb = \", self.context_emb) # [batch, num_mentions, 600]\n\n\n\n def add_span_emb_op(self):\n mention_emb_list = []\n # span embedding based on boundaries (start, end) and head mechanism. but do that on top of contextual bilistm\n # output or on top of original word+char embeddings. this flag determines that. Of course by default head\n # is on top of word+char emb instead of bilstm output.\n boundaries_input_vecs = self.word_embeddings if self.args.span_boundaries_from_wordemb else self.context_emb\n\n if self.args.span_emb.find(\"boundaries\") != -1:\n mention_start_emb = tf.gather_nd(boundaries_input_vecs, tf.stack(\n [tf.tile(tf.expand_dims(tf.range(tf.shape(self.begin_span)[0]), 1), [1, tf.shape(self.begin_span)[1]]),\n self.begin_span], 2))\n #mention_start_emb = tf.gather(text_outputs, mention_starts) # [num_mentions, emb]\n mention_emb_list.append(mention_start_emb)\n\n mention_end_emb = tf.gather_nd(boundaries_input_vecs, tf.stack(\n [tf.tile(tf.expand_dims(tf.range(tf.shape(self.begin_span)[0]), 1), [1, tf.shape(self.begin_span)[1]]),\n tf.nn.relu(self.end_span-1)], 2)) # -1 because the end of span in exclusive [start, end)\n # relu so the 0 don't become -1\n #mention_end_emb = tf.gather(text_outputs, mention_ends) # [num_mentions, emb]\n mention_emb_list.append(mention_end_emb)\n #print(\"mention_start_emb = \", mention_start_emb)\n #print(\"mention_end_emb = \", mention_end_emb)\n\n mention_width = self.end_span - self.begin_span # [batch, num_mentions]\n # TODO remove the comment code below\n \"\"\"\n if self.args.use_features:\n mention_width_index = mention_width - 1 # [num_mentions]\n mention_width_emb = tf.gather(tf.get_variable(\"mention_width_embeddings\", [self.args[\"max_mention_width\"],\n self.args[\"feature_size\"]]),\n mention_width_index) # [batch, num_mentions, emb]\n mention_width_emb = tf.nn.dropout(mention_width_emb, self.dropout)\n #print(\"mention_width_emb = \", mention_width_emb)\n mention_emb_list.append(mention_width_emb)\n \"\"\"\n\n if self.args.span_emb.find(\"head\") != -1: # here the attention is computed\n self.max_mention_width = tf.minimum(self.args.max_mention_width,\n tf.reduce_max(self.end_span - self.begin_span))\n mention_indices = tf.range(self.max_mention_width) + \\\n tf.expand_dims(self.begin_span, 2) # [batch, num_mentions, max_mention_width]\n mention_indices = tf.minimum(tf.shape(self.word_embeddings)[1] - 1,\n mention_indices) # [batch, num_mentions, max_mention_width]\n #print(\"mention_indices = \", mention_indices)\n batch_index = tf.tile(tf.expand_dims(tf.expand_dims(tf.range(tf.shape(mention_indices)[0]), 1), 2),\n [1, tf.shape(mention_indices)[1], tf.shape(mention_indices)[2]])\n mention_indices = tf.stack([batch_index, mention_indices], 3)\n # [batch, num_mentions, max_mention_width, [row,col] ] 4d tensor\n\n # this means that head will be either the same as boundaries or boundaries from bilstm and head from wordemb\n head_input_vecs = boundaries_input_vecs if self.args.model_heads_from_bilstm else self.word_embeddings\n mention_text_emb = tf.gather_nd(head_input_vecs, mention_indices)\n # [batch, num_mentions, max_mention_width, 500 ] 4d tensor\n #print(\"mention_text_emb = \", mention_text_emb)\n\n with tf.variable_scope(\"head_scores\"):\n # from [batch, max_sent_len, 600] to [batch, max_sent_len, 1]\n self.head_scores = util.projection(boundaries_input_vecs, 1, model=self)\n # [batch, num_mentions, max_mention_width, 1]\n mention_head_scores = tf.gather_nd(self.head_scores, mention_indices)\n #print(\"mention_head_scores = \", mention_head_scores)\n if not tf.__version__.startswith(\"1.4\"):\n temp_shape = tf.shape(mention_width)\n temp = tf.sequence_mask(tf.reshape(mention_width, [-1]), self.max_mention_width, dtype=tf.float32)\n temp_mask = tf.reshape(temp, [temp_shape[0], temp_shape[1], tf.shape(temp)[-1]])\n else:\n temp_mask = tf.sequence_mask(mention_width, self.max_mention_width, dtype=tf.float32)\n mention_mask = tf.expand_dims(temp_mask, 3) # [batch, num_mentions, max_mention_width, 1]\n mention_mask = tf.minimum(1.0, tf.maximum(self.args.zero, mention_mask)) # 1e-3\n mention_attention = tf.nn.softmax(mention_head_scores + tf.log(mention_mask),\n dim=2) # [batch, num_mentions, max_mention_width, 1]\n mention_head_emb = tf.reduce_sum(mention_attention * mention_text_emb, 2) # [batch, num_mentions, emb]\n #print(\"mention_head_emb = \", mention_head_emb)\n mention_emb_list.append(mention_head_emb)\n\n self.span_emb = tf.concat(mention_emb_list, 2) # [batch, num_mentions, emb i.e. 1700]\n #print(\"span_emb = \", self.span_emb)\n\n def add_lstm_score_op(self):\n #print(\"cand_entities = \", self.cand_entities)\n with tf.variable_scope(\"span_emb_ffnn\"):\n # [batch, num_mentions, 300]\n if self.args.span_emb_ffnn[0] == 0:\n span_emb_projected = util.projection(self.span_emb, 300, model=self)\n else:\n hidden_layers, hidden_size = self.args.span_emb_ffnn[0], self.args.span_emb_ffnn[1]\n span_emb_projected = util.ffnn(self.span_emb, hidden_layers, hidden_size, 300,\n self.dropout if self.args.ffnn_dropout else None, model=self)\n #print(\"span_emb_projected = \", span_emb_projected)\n scores = tf.matmul(tf.expand_dims(span_emb_projected, 2), self.entity_embeddings, transpose_b=True)\n #print(\"scores = \", scores)\n self.similarity_scores = tf.squeeze(scores, axis=2) # [batch, num_mentions, 1, 30]\n #print(\"scores = \", self.similarity_scores) # [batch, num_mentions, 30]\n\n def add_local_attention_op(self):\n # shape=(b, num_of_spans, 30, 300)\n attention_entity_emb = self.pure_entity_embeddings if self.args.attention_ent_vecs_no_regularization else self.entity_embeddings\n with tf.variable_scope(\"attention\"):\n K = self.args.attention_K\n left_mask = self._sequence_mask_v13(self.begin_span, K)\n #left_mask = tf.sequence_mask(self.begin_span, K, dtype=tf.float32)\n right_mask = self._sequence_mask_v13(tf.expand_dims(self.words_len, 1) - self.end_span, K)\n #right_mask = tf.sequence_mask(tf.expand_dims(self.words_len, 1) - self.end_span, # number of words on the right\n # K, dtype=tf.float32) # but maximum i get K not more\n ctxt_mask = tf.concat([left_mask, right_mask], 2) # [batch, num_of_spans, 2*K]\n ctxt_mask = tf.log(tf.minimum(1.0, tf.maximum(self.args.zero, ctxt_mask)))\n # T, T, T, F, F | T, T, F, F, F\n # -1, -2, -3, -4, -5 +0, +1, +2, +3, +4\n\n leftctxt_indices = tf.maximum(0, tf.range(-1, -K - 1, -1) + \\\n tf.expand_dims(self.begin_span, 2)) # [batch, num_mentions, K]\n rightctxt_indices = tf.minimum(tf.shape(self.pure_word_embeddings)[1] - 1, tf.range(K) + \\\n tf.expand_dims(self.end_span, 2)) # [batch, num_mentions, K]\n ctxt_indices = tf.concat([leftctxt_indices, rightctxt_indices], 2) # [batch, num_mentions, 2*K]\n\n batch_index = tf.tile(tf.expand_dims(tf.expand_dims(tf.range(tf.shape(ctxt_indices)[0]), 1), 2),\n [1, tf.shape(ctxt_indices)[1], tf.shape(ctxt_indices)[2]])\n ctxt_indices = tf.stack([batch_index, ctxt_indices], 3)\n # [batch, num_of_spans, 2*K, 2] the last dimension is row,col for gather_nd\n # [batch, num_of_spans, 2*K, [row,col]]\n\n att_x_w = self.pure_word_embeddings # [batch, max_sent_len, 300]\n if self.args.attention_on_lstm and self.args.nn_components.find(\"lstm\") != -1:\n # [batch, max_sent_len, 600] hidden_size_of_lstm*2 so project it to 300\n # TODO maybe omit projection if already in 300 dimention? but projection allows transormation...\n att_x_w = util.projection(self.context_emb, 300, model=self) # if tf.shape(self.context_emb)[-1] != 300 else self.context_emb\n\n ctxt_word_emb = tf.gather_nd(att_x_w, ctxt_indices)\n # [batch, num_of_spans, 2K, emb_size] emb_size = 300 only pure word emb used\n # and not after we add char emb and dropout\n\n\n x_c_voters = attention_entity_emb\n # restrict the number of entities that participate in the forming of the x_c context vector\n if self.args.attention_retricted_num_of_entities:\n x_c_voters = tf.slice(attention_entity_emb, [0, 0, 0, 0],\n [-1, -1, self.args.attention_retricted_num_of_entities, -1])\n if self.args.attention_use_AB:\n att_A = tf.get_variable(\"att_A\", [300])\n x_c_voters = att_A * x_c_voters\n # [b, num_of_spans, 2*K, 300] mul [b, num_of_spans, 30, 300] instead of 30 it can be the reduced number of entities\n scores = tf.matmul(ctxt_word_emb, x_c_voters, transpose_b=True) # [b, spans, 2K, 30]\n scores = tf.reduce_max(scores, reduction_indices=[-1]) # max score of each word for each span acquired from any cand entity\n\n scores = scores + ctxt_mask # some words are not valid out of window\n # so we assign to them very low score\n\n top_values, _ = tf.nn.top_k(scores, self.args.attention_R)\n # [batch, num_of_spans, R]\n #R_value = tf.reduce_min(top_values, axis=-1)\n R_value = top_values[:, :, -1] # [batch, num_of_spans]\n # same as above command but probably faster\n R_value = tf.maximum(self.args.zero, R_value) # so to avoid keeping words that\n # have max score with any of the entities <=0 (also score = 0 can have words with\n # padding candidate entities)\n\n threshold = tf.tile(tf.expand_dims(R_value, 2), [1, 1, 2 * K])\n # [batch, num_of_spans, 2K]\n scores = scores - tf.to_float(((scores - threshold) < 0)) * 50 # 50 where score=thr\n scores = tf.nn.softmax(scores, dim=2) # [batch, num_of_spans, 2K]\n scores = tf.expand_dims(scores, 3) # [batch, num_of_spans, 2K, 1]\n # [batch, num_of_spans, 2K, 1] * [batch, num_of_spans, 2K, emb_size]\n # = [batch, num_of_spans, 2K, emb_size]\n x_c = tf.reduce_sum(scores * ctxt_word_emb, 2) # = [batch, num_of_spans, emb_size]\n if self.args.attention_use_AB:\n att_B = tf.get_variable(\"att_B\", [300])\n x_c = att_B * x_c\n x_c = tf.expand_dims(x_c, 3) # [batch, num_of_spans, emb_size, 1]\n # [batch, num_of_spans, 30, emb_size=300] mul with [batch, num_of_spans, emb_size, 1]\n x_e__x_c = tf.matmul(attention_entity_emb, x_c) # [batch, num_of_spans, 30, 1]\n x_e__x_c = tf.squeeze(x_e__x_c, axis=3) # [batch, num_of_spans, 30]\n self.attention_scores = x_e__x_c\n\n\n def custom_pem(self, log=True, buckets_boundaries=None, bucketing_name=\"pem_embeddings\"):\n if buckets_boundaries:\n return self._pem_bucketing_embeddings(buckets_boundaries, bucketing_name)\n elif self.args.pem_without_log:\n return self.cand_entities_scores\n else:\n return tf.log(tf.minimum(1.0, tf.maximum(self.args.zero, self.cand_entities_scores)))\n\n\n def _pem_bucketing_embeddings(self, buckets_boundaries, bucketing_name):\n from tensorflow.python.ops import math_ops\n bucketized_pem = math_ops._bucketize(self.cand_entities_scores, boundaries=buckets_boundaries)\n with tf.variable_scope(bucketing_name):\n _pem_embeddings = tf.get_variable(\n name=\"pem_embeddings_var\",\n dtype=tf.float32,\n shape=[len(buckets_boundaries) + 1, 1], trainable=True)\n pem_embeddings = tf.nn.embedding_lookup(_pem_embeddings, bucketized_pem, name=\"pem_embeddings_lookup\")\n # pem_embeddings = Tensor(\"pem_embeddings/pem_embeddings_lookup:0\", shape=(?, ?, ?, 1), dtype=float32)\n return tf.squeeze(pem_embeddings, axis=3)\n\n def add_cand_ent_scores_op(self):\n \"\"\"\n # TODO remove this code and this option (without p_e_m) just to experiment once...\n if hasattr(self.args, 'no_p_e_m_usage') and self.args.no_p_e_m_usage:\n self.final_scores = self.similarity_scores\n return\n \"\"\"\n #now add the cand_entity_scores maybe also some extra features and through a simple ffnn\n stack_values = []\n if self.args.nn_components.find(\"lstm\") != -1:\n stack_values.append(self.similarity_scores)\n if self.args.nn_components.find(\"pem\") != -1:\n # TODO rename to pem_scores\n self.log_cand_entities_scores = self.custom_pem(\n self.args.pem_without_log, self.args.pem_buckets_boundaries)\n stack_values.append(self.log_cand_entities_scores)\n if self.args.nn_components.find(\"attention\") != -1:\n stack_values.append(self.attention_scores)\n \"\"\"\n if len(stack_values) == 1:\n # since only one scalar omit the final ffnn\n self.final_scores = stack_values[0]\n return\n \"\"\"\n scalar_predictors = tf.stack(stack_values, 3)\n #print(\"scalar_predictors = \", scalar_predictors) #[batch, num_mentions, 30, 2]\n\n with tf.variable_scope(\"similarity_and_prior_ffnn\"):\n # [batch, num_mentions, 30, 1] squeeze to [batch, num_mentions, 30]\n if self.args.final_score_ffnn[0] == 0:\n self.final_scores = util.projection(scalar_predictors, 1, model=self)\n else:\n hidden_layers, hidden_size = self.args.final_score_ffnn[0], self.args.final_score_ffnn[1]\n self.final_scores = util.ffnn(scalar_predictors, hidden_layers, hidden_size, 1,\n self.dropout if self.args.ffnn_dropout else None, model=self)\n #self.final_scores = tf.squeeze(util.ffnn(scalar_predictors, 1, 100, 1, self.dropout), axis=3)\n self.final_scores = tf.squeeze(self.final_scores, axis=3)\n #print(\"final_scores = \", self.final_scores)\n\n def add_global_voting_op(self):\n with tf.variable_scope(\"global_voting\"):\n #self.final_scores_before_global = self.final_scores\n # TODO important change\n #self.final_scores_before_global = tf.where(tf.less(self.loss_mask, 1e-3), -3.0, self.final_scores)\n self.final_scores_before_global = - (1 - self.loss_mask) * 50 + self.final_scores\n if self.args.global_topkfromallspans:\n batch_num = tf.shape(self.final_scores)[0]\n spans_num = tf.shape(self.final_scores)[1] # num of spans\n cand_ent_num = tf.shape(self.final_scores)[2] # 30\n new_size = spans_num * cand_ent_num\n temp = tf.diag(tf.ones([spans_num]))\n temp = tf.tile(tf.expand_dims(temp, axis=2), [1, 1, cand_ent_num])\n temp = tf.reshape(temp, [spans_num, new_size])\n mask = tf.reshape(tf.tile(tf.expand_dims(temp, axis=1), [1, cand_ent_num, 1]), [new_size, new_size])\n mask = 1 - mask\n\n all_entities = tf.reshape(self.pure_entity_embeddings, [batch_num, new_size, 300])\n all_scores = tf.matmul(all_entities, all_entities, transpose_b=True) # [batch, new_size, new_size]\n filtered_scores = all_scores * mask\n\n top_values, _ = tf.nn.top_k(filtered_scores, self.args.global_topkfromallspans)\n # [batch, new_size, K]\n if self.args.global_topkfromallspans_onlypositive:\n top_values = tf.maximum(top_values, self.args.zero)\n # so to avoid keeping cand ent that have score < of this value even if they are the\n self.global_voting_scores = tf.reduce_mean(top_values, axis=2) # [batch, new_size]\n self.global_voting_scores = tf.reshape(self.global_voting_scores, [batch_num, spans_num, cand_ent_num])\n else:\n if self.args.global_gmask_unambigious:\n gmask = self._sequence_mask_v13(tf.equal(self.cand_entities_len, 1), tf.shape(self.final_scores)[2])\n elif not self.args.global_topk:\n gmask = tf.to_float(((self.final_scores_before_global - self.args.global_thr) >= 0)) # [b,s,30]\n else:\n top_values, _ = tf.nn.top_k(self.final_scores_before_global, self.args.global_topk)\n # [batch, num_of_spans, K]\n K_value = top_values[:, :, -1] # [batch, num_of_spans]\n #if hasattr(self.args, 'global_topkthr'):\n if self.args.global_topkthr:\n K_value = tf.maximum(self.args.global_topkthr, K_value)\n # so to avoid keeping cand ent that have score < of this value even if they are the\n # top for this span. 30\n threshold = tf.tile(tf.expand_dims(K_value, 2), [1, 1, tf.shape(self.final_scores)[-1]])\n # [batch, num_of_spans, 30]\n gmask = tf.to_float(((self.final_scores_before_global - threshold) >= 0))\n gmask = gmask * self.loss_mask\n if self.args.global_mask_scale_each_mention_voters_to_one:\n temp = tf.reduce_sum(gmask, axis=2, keep_dims=True) # [batch, num_of_spans, 1]\n temp = tf.where(tf.less(temp, 1e-4), temp, 1. / (temp + 1e-4))\n gmask = gmask * temp\n elif self.args.global_gmask_based_on_localscore:\n \"\"\"\n temp_masked_local_scores = gmask * self.final_scores_before_global # [batch, num_of_spans, 30]\n # per last dimension (30) add the smallest score if negative otherwise keep it like that to zero\n # so we only have positive mask scores\n offset = tf.reduce_min(temp_masked_local_scores, axis=2) # [batch, num_of_spans]\n offset = tf.expand_dims(tf.minimum(0.0, offset), axis=2) # [batch, num_of_spans, 1]\n gmask = (temp_masked_local_scores - offset) * gmask # [batch, num_of_spans, 30]\n gmask = tf.maximum(0.0, gmask)\n \"\"\"\n gmask = gmask * tf.nn.softmax(self.final_scores_before_global)\n self.gmask = gmask\n\n masked_entity_emb = self.pure_entity_embeddings * tf.expand_dims(gmask, axis=3) # [b,s,30,300] * [b,s,30,1]\n if self.args.new_all_voters_emb:\n batch_size = tf.shape(masked_entity_emb)[0]\n all_voters_emb = tf.reduce_sum(tf.reshape(masked_entity_emb, [batch_size, -1, 300]), axis=1,\n keep_dims=True) # [b, 1, 300]\n else:\n all_voters_emb = tf.reduce_sum(tf.reshape(masked_entity_emb, [-1, 300]), axis=0) # [300]\n span_voters_emb = tf.reduce_sum(masked_entity_emb, axis=2) # [batch, num_of_spans, 300]\n valid_voters_emb = all_voters_emb - span_voters_emb\n #[300] - [batch, spans, 300] = [batch, spans, 300] (broadcasting)\n if self.args.global_norm_or_mean == \"norm\":\n valid_voters_emb = tf.nn.l2_normalize(valid_voters_emb, dim=2)\n else:\n all_voters_num = tf.reduce_sum(gmask) # scalar\n span_voters_num = tf.reduce_sum(gmask, axis=2) # [batch, spans]\n valid_voters_emb = valid_voters_emb / tf.expand_dims(all_voters_num - span_voters_num, axis=2)\n\n self.global_voting_scores = tf.squeeze(tf.matmul(self.pure_entity_embeddings, tf.expand_dims(valid_voters_emb, axis=3)), axis=3)\n # [b,s,30,300] matmul [b,s,300,1] --> [b,s,30,1]-->[b,s,30]\n\n # TODO here i check the self.args.stage2_nn_components\n stack_values = []\n if self.args.stage2_nn_components.find(\"pem\") != -1:\n # TODO rename to pem_scores\n self.gpem_scores = self.custom_pem(\n self.args.gpem_without_log, self.args.gpem_buckets_boundaries)\n stack_values.append(self.gpem_scores)\n if self.args.stage2_nn_components.find(\"local\") != -1:\n stack_values.append(self.final_scores_before_global)\n stack_values.append(self.global_voting_scores)\n scalar_predictors = tf.stack(stack_values, 3)\n #print(\"scalar_predictors = \", scalar_predictors) #[b, s, 30, 2]\n with tf.variable_scope(\"psi_and_global_ffnn\"):\n # [batch, num_mentions, 30, 1] squeeze to [batch, num_mentions, 30]\n if self.args.global_score_ffnn[0] == 0:\n self.final_scores = util.projection(scalar_predictors, 1, model=self)\n else:\n hidden_layers, hidden_size = self.args.global_score_ffnn[0], self.args.global_score_ffnn[1]\n self.final_scores = util.ffnn(scalar_predictors, hidden_layers, hidden_size, 1,\n self.dropout if self.args.ffnn_dropout else None, model=self)\n self.final_scores = tf.squeeze(self.final_scores, axis=3)\n #print(\"final_scores = \", self.final_scores)\n\n\n def add_loss_op(self):\n cand_entities_labels = tf.cast(self.cand_entities_labels, tf.float32)\n loss1 = cand_entities_labels * tf.nn.relu(self.args.gamma_thr - self.final_scores)\n loss2 = (1 - cand_entities_labels) * tf.nn.relu(self.final_scores)\n self.loss = loss1 + loss2\n if self.args.nn_components.find(\"global\") != -1 and not self.args.global_one_loss:\n loss3 = cand_entities_labels * tf.nn.relu(self.args.gamma_thr - self.final_scores_before_global)\n loss4 = (1 - cand_entities_labels) * tf.nn.relu(self.final_scores_before_global)\n self.loss = loss1 + loss2 + loss3 + loss4\n #print(\"loss_mask = \", loss_mask)\n self.loss = self.loss_mask * self.loss\n self.loss = tf.reduce_sum(self.loss)\n # for tensorboard\n #tf.summary.scalar(\"loss\", self.loss)\n\n\n def build(self):\n self.add_placeholders()\n self.add_embeddings_op()\n if self.args.nn_components.find(\"lstm\") != -1:\n self.add_context_emb_op()\n self.add_span_emb_op()\n self.add_lstm_score_op()\n if self.args.nn_components.find(\"attention\") != -1:\n self.add_local_attention_op()\n self.add_cand_ent_scores_op()\n if self.args.nn_components.find(\"global\") != -1:\n self.add_global_voting_op()\n if self.args.running_mode.startswith(\"train\"):\n self.add_loss_op()\n # Generic functions that add training op\n self.add_train_op(self.args.lr_method, self.lr, self.loss, self.args.clip)\n self.merged_summary_op = tf.summary.merge_all()\n\n if self.args.running_mode == \"train_continue\":\n self.restore_session(\"latest\")\n elif self.args.running_mode == \"train\":\n self.initialize_session() # now self.sess is defined and vars are init\n self.init_embeddings()\n\n # if we run the evaluate.py script then we should call explicitly the model.restore(\"ed\")\n # or model.restore(\"el\"). here it doesn't initialize or restore values for the evaluate.py\n # case.\n\n def _sequence_mask_v13(self, mytensor, max_width):\n \"\"\"mytensor is a 2d tensor\"\"\"\n if not tf.__version__.startswith(\"1.4\"):\n temp_shape = tf.shape(mytensor)\n temp = tf.sequence_mask(tf.reshape(mytensor, [-1]), max_width, dtype=tf.float32)\n temp_mask = tf.reshape(temp, [temp_shape[0], temp_shape[1], max_width]) # tf.shape(temp)[-1]])\n else:\n temp_mask = tf.sequence_mask(mytensor, max_width, dtype=tf.float32)\n return temp_mask\n\n\n" }, { "path": "code/model/ensemble_eval.py", "content": "import argparse\nimport pickle\nimport model.config as config\nimport os\nimport tensorflow as tf\nfrom model.model_ablations import Model\nimport model.train as train\nfrom evaluation.metrics import Evaluator, metrics_calculation_and_prediction_printing, threshold_calculation\nimport model.reader as reader\nfrom model.util import load_train_args\n\ndef validation_loss_calculation(filename, opt_thr, el_mode):\n if args.predictions_folder is not None:\n printPredictions.process_file(el_mode, filename, opt_thr)\n\n ensemble_fixed = []\n ensemble_acc = [] # final_scores and similarity_scores. all the rest are fixed\n for model_num, model_folder in enumerate(args.output_folder): # for all ensemble models\n model, handles = create_input_pipeline(el_mode, model_folder,\n [filename])\n retrieve_l = [model.final_scores, model.similarity_scores,\n model.cand_entities_len, model.cand_entities,\n model.begin_span, model.end_span, model.spans_len,\n model.begin_gm, model.end_gm,\n model.ground_truth, model.ground_truth_len,\n model.words_len, model.chunk_id,\n # model.similarity_scores,\n model.words, model.chars, model.chars_len,\n model.log_cand_entities_scores]\n elem_idx = 0\n while True:\n try:\n result_l = model.sess.run(\n retrieve_l, feed_dict={model.input_handle_ph: handles[0], model.dropout: 1})\n if model_num == 0:\n ensemble_fixed.append(result_l[2:])\n ensemble_acc.append(result_l[:2])\n else:\n ensemble_acc[elem_idx][0] += result_l[0]\n ensemble_acc[elem_idx][1] += result_l[1]\n\n elem_idx += 1\n except tf.errors.OutOfRangeError:\n break\n model.close_session()\n\n evaluator = Evaluator(opt_thr, name=filename)\n number_of_models = len(args.output_folder)\n for (final_scores, similarity_scores), fixed in zip(ensemble_acc, ensemble_fixed):\n final_scores /= number_of_models\n similarity_scores /= number_of_models\n\n metrics_calculation_and_prediction_printing(evaluator,\n final_scores, similarity_scores, *fixed, el_mode,\n printPredictions=printPredictions)\n\n if printPredictions:\n printPredictions.file_ended()\n print(filename)\n micro_f1, macro_f1 = evaluator.print_log_results(None, -1, el_mode)\n return macro_f1\n\n\ndef optimal_thr_calc(el_mode):\n filenames = args.el_datasets if el_mode else args.ed_datasets\n val_datasets = args.el_val_datasets if el_mode else args.ed_val_datasets\n\n ensemble_fixed = []\n ensemble_acc = [] # final_scores and similarity_scores. all the rest are fixed\n for model_num, model_folder in enumerate(args.output_folder): # for all ensemble models\n model, handles = create_input_pipeline(el_mode, model_folder,\n [filenames[i] for i in val_datasets])\n\n retrieve_l = (model.final_scores, model.cand_entities_len, model.cand_entities,\n model.begin_span, model.end_span, model.spans_len,\n model.begin_gm, model.end_gm,\n model.ground_truth, model.ground_truth_len,\n model.words_len, model.chunk_id)\n elem_idx = 0\n for dataset_handle in handles: # 1, 4 for each validation dataset\n while True:\n try:\n result_l = model.sess.run(\n retrieve_l, feed_dict={model.input_handle_ph: dataset_handle, model.dropout: 1})\n if model_num == 0:\n ensemble_fixed.append(result_l[1:])\n ensemble_acc.append(result_l[0])\n else:\n ensemble_acc[elem_idx] += result_l[0]\n\n elem_idx += 1\n except tf.errors.OutOfRangeError:\n break\n model.close_session()\n\n number_of_models = len(args.output_folder)\n tp_fp_scores_labels = []\n fn_scores = []\n for final_scores, fixed in zip(ensemble_acc, ensemble_fixed):\n final_scores /= number_of_models\n\n tp_fp_batch, fn_batch = threshold_calculation(final_scores, *fixed, el_mode)\n tp_fp_scores_labels.extend(tp_fp_batch)\n fn_scores.extend(fn_batch)\n\n return train.optimal_thr_calc_aux(tp_fp_scores_labels, fn_scores)\n\n\ndef create_input_pipeline(el_mode, model_folder, filenames):\n tf.reset_default_graph()\n folder = config.base_folder+\"data/tfrecords/\" + args.experiment_name + (\"/test/\" if el_mode else \"/train/\")\n datasets = []\n for file in filenames:\n datasets.append(reader.test_input_pipeline([folder+file], args))\n\n input_handle_ph = tf.placeholder(tf.string, shape=[], name=\"input_handle_ph\")\n iterator = tf.contrib.data.Iterator.from_string_handle(\n input_handle_ph, datasets[0].output_types, datasets[0].output_shapes)\n next_element = iterator.get_next()\n\n train_args = load_train_args(args.output_folder, \"ensemble_eval\")\n\n print(\"loading Model:\", model_folder)\n #train_args.evaluation_script = True\n train_args.entity_extension = args.entity_extension\n model = Model(train_args, next_element)\n model.build()\n #print(\"model train_args:\", model.args)\n #print(\"model checkpoint_folder:\", model.args.checkpoints_folder)\n model.input_handle_ph = input_handle_ph\n model.restore_session(\"el\" if el_mode else \"ed\")\n\n #iterators, handles = from_datasets_to_iterators_and_handles(model.sess, datasets)\n iterators = []\n handles = []\n for dataset in datasets:\n #iterator = dataset.make_initializable_iterator() # one shot iterators fits better here\n iterator = dataset.make_one_shot_iterator()\n iterators.append(iterator)\n handles.append(model.sess.run(iterator.string_handle()))\n return model, handles\n\n\ndef evaluate():\n for el_mode in [False, True]:\n filenames = args.el_datasets if el_mode else args.ed_datasets\n if filenames:\n print(\"Evaluating {} datasets\".format(\"EL\" if el_mode else \"ED\"))\n opt_thr, _ = optimal_thr_calc(el_mode)\n # TODO check the following lines\n results = []\n #for test_handle, test_name, test_it in zip(datasets, names):\n for filename in filenames:\n f1_score = validation_loss_calculation(filename, opt_thr, el_mode=el_mode)\n results.append(f1_score)\n\ndef _parse_args():\n parser = argparse.ArgumentParser()\n parser.add_argument(\"--experiment_name\", default=\"alldatasets_perparagr\", #\"standard\",\n help=\"under folder data/tfrecords/\")\n parser.add_argument(\"--training_name\", help=\"under folder data/tfrecords/\")\n parser.add_argument(\"--predictions_folder\", default=None, help=\"full or relative path. where to print\"\n \"the result files.\")\n\n parser.add_argument(\"--ed_datasets\", default=\"aida_train.txt_z_aida_dev.txt_z_aida_test.txt_z_\"\n \"ace2004.txt_z_aquaint.txt_z_clueweb.txt_z_msnbc.txt_z_wikipedia.txt\")\n parser.add_argument(\"--el_datasets\", default=\"aida_train.txt_z_aida_dev.txt_z_aida_test.txt_z_\"\n \"ace2004.txt_z_aquaint.txt_z_clueweb.txt_z_msnbc.txt_z_wikipedia.txt\")\n\n parser.add_argument(\"--ed_val_datasets\", default=\"1\")\n parser.add_argument(\"--el_val_datasets\", default=\"1\")\n\n parser.add_argument(\"--all_spans_training\", help=\"y_y_n_y_n\")\n parser.add_argument(\"--entity_extension\", default=None, help=\"extension_entities or extension_entities_all etc\")\n args = parser.parse_args()\n\n args.ed_datasets = args.ed_datasets.split('_z_') if args.ed_datasets != \"\" else None\n args.el_datasets = args.el_datasets.split('_z_') if args.el_datasets != \"\" else None\n args.ed_val_datasets = [int(x) for x in args.ed_val_datasets.split('_')]\n args.el_val_datasets = [int(x) for x in args.el_val_datasets.split('_')]\n\n args.training_name = args.training_name.split('_z_')\n args.all_spans_training = [\"all_spans_\" if x == 'y' else \"\" for x in args.all_spans_training.split('_')]\n assert(len(args.training_name) == len(args.all_spans_training))\n\n if args.predictions_folder is not None and not os.path.exists(args.predictions_folder):\n os.makedirs(args.predictions_folder)\n if args.predictions_folder is not None and not os.path.exists(args.predictions_folder+\"ed/\"):\n os.makedirs(args.predictions_folder+\"ed/\")\n if args.predictions_folder is not None and not os.path.exists(args.predictions_folder+\"el/\"):\n os.makedirs(args.predictions_folder+\"el/\")\n\n args.output_folder = []\n for training_name, prefix in zip(args.training_name, args.all_spans_training):\n args.output_folder.append(config.base_folder+\"data/tfrecords/\" + \\\n args.experiment_name+\"/{}training_folder/\".format(prefix) + \\\n training_name+\"/\")\n\n args.batch_size = 1\n print(args)\n return args\n\n\nif __name__ == \"__main__\":\n args = _parse_args()\n printPredictions = None\n if args.predictions_folder is not None:\n from evaluation.print_predictions import PrintPredictions\n printPredictions = PrintPredictions(config.base_folder+\"data/tfrecords/\"+\n args.experiment_name+\"/\", args.predictions_folder)\n evaluate()\n\n" }, { "path": "code/model/evaluate.py", "content": "import argparse\nimport pickle\nimport model.config as config\nimport os\nimport tensorflow as tf\nfrom model.model_ablations import Model\nfrom evaluation.metrics import Evaluator, metrics_calculation_and_prediction_printing\nimport model.train as train\nfrom model.util import load_train_args\n\n\ndef validation_loss_calculation(model, iterator, dataset_handle, opt_thr, el_mode, name=\"\"):\n if args.print_predictions:\n printPredictions.process_file(el_mode, name, opt_thr)\n model.sess.run(iterator.initializer)\n evaluator = Evaluator(opt_thr, name=name)\n\n while True:\n try:\n retrieve_l = [model.final_scores,\n model.cand_entities_len, model.cand_entities,\n model.begin_span, model.end_span, model.spans_len,\n model.begin_gm, model.end_gm,\n model.ground_truth, model.ground_truth_len,\n model.words_len, model.chunk_id,\n model.words, model.chars, model.chars_len]\n scores_retrieve_l, scores_names_l = [], []\n if model.args.nn_components.find(\"lstm\") != -1:\n scores_retrieve_l.append(model.similarity_scores)\n scores_names_l.append(\"lstm\")\n if model.args.nn_components.find(\"pem\") != -1:\n scores_retrieve_l.append(model.log_cand_entities_scores)\n scores_names_l.append(\"logpem\")\n if model.args.nn_components.find(\"attention\") != -1:\n scores_retrieve_l.append(model.attention_scores)\n scores_names_l.append(\"attention\")\n if model.args.nn_components.find(\"global\") != -1:\n scores_retrieve_l.append(model.final_scores_before_global)\n scores_names_l.append(\"before_global\")\n scores_retrieve_l.append(model.global_voting_scores)\n scores_names_l.append(\"global_voting\")\n global_pairwise_scores = []\n if args.print_global_voters:\n global_pairwise_scores.append(model.gmask)\n global_pairwise_scores.append(model.pure_entity_embeddings)\n\n retrieve_l.append(scores_retrieve_l)\n retrieve_l.append(global_pairwise_scores)\n result_l = model.sess.run(\n retrieve_l, feed_dict={model.input_handle_ph: dataset_handle, model.dropout: 1})\n metrics_calculation_and_prediction_printing(evaluator, *result_l, scores_names_l, el_mode,\n printPredictions=printPredictions)\n\n except tf.errors.OutOfRangeError:\n if args.print_predictions:\n printPredictions.file_ended()\n print(name)\n micro_f1, macro_f1 = evaluator.print_log_results(None, -1, el_mode)\n break\n return macro_f1\n\n# identical with the train.compute_ed_el_scores\ndef compute_ed_el_scores(model, handles, names, iterators, el_mode):\n if args.hardcoded_thr:\n opt_thr = args.hardcoded_thr\n print(\"hardcoded threshold used:\", opt_thr)\n else:\n # first compute the optimal threshold based on validation datasets.\n opt_thr, _ = train.optimal_thr_calc(model, handles, iterators, el_mode)\n # give the opt_thr and the projection variables to the PrintPredictions for insight\n if printPredictions:\n printPredictions.extra_info = print_thr_and_ffnn_values(model, opt_thr)\n\n results = []\n for test_handle, test_name, test_it in zip(handles, names, iterators):\n f1_score = validation_loss_calculation(model, test_it, test_handle, opt_thr,\n el_mode=el_mode, name=test_name)\n results.append(f1_score)\n return results\n\n\ndef evaluate():\n\n ed_datasets, ed_names = train.create_el_ed_pipelines(gmonly_flag=True, filenames=args.ed_datasets, args=args)\n el_datasets, el_names = train.create_el_ed_pipelines(gmonly_flag=False, filenames=args.el_datasets, args=args)\n\n input_handle_ph = tf.placeholder(tf.string, shape=[], name=\"input_handle_ph\")\n sample_dataset = ed_datasets[0] if ed_datasets != [] else el_datasets[0]\n iterator = tf.data.Iterator.from_string_handle(\n input_handle_ph, sample_dataset.output_types, sample_dataset.output_shapes)\n next_element = iterator.get_next()\n\n model = Model(train_args, next_element)\n model.build()\n model.input_handle_ph = input_handle_ph # just for convenience so i can access it from everywhere\n print(tf.global_variables())\n if args.p_e_m_algorithm:\n model.final_scores = model.cand_entities_scores\n\n\n def ed_el_dataset_handles(sess, datasets):\n test_iterators = []\n test_handles = []\n for dataset in datasets:\n test_iterator = dataset.make_initializable_iterator()\n test_iterators.append(test_iterator)\n test_handles.append(sess.run(test_iterator.string_handle()))\n return test_iterators, test_handles\n\n\n for el_mode, datasets, names in zip([False, True], [ed_datasets, el_datasets], [ed_names, el_names]):\n if names == []:\n continue\n model.restore_session(\"el\" if el_mode else \"ed\")\n #print_variables_values(model)\n\n with model.sess as sess:\n print(\"Evaluating {} datasets\".format(\"EL\" if el_mode else \"ED\"))\n iterators, handles = ed_el_dataset_handles(sess, datasets)\n compute_ed_el_scores(model, handles, names, iterators, el_mode=el_mode)\n\n\n# TODO delete this function\ndef print_variables_values(model):\n var_names = ['similarity_and_prior_ffnn/output_weights:0',\n 'similarity_and_prior_ffnn/output_bias:0']\n for var_name in var_names:\n var_handle = [v for v in tf.global_variables() if v.name == var_name][0]\n print(var_name)\n print(model.sess.run(var_handle))\n\n\ndef print_thr_and_ffnn_values(model, opt_thr):\n result = \"opt_thr={}, nn_components={}\\n\".format(opt_thr, model.args.nn_components)\n var_names, var_print_names = [], []\n if model.args.final_score_ffnn[0] == 0:\n var_names.extend(['similarity_and_prior_ffnn/output_weights:0',\n 'similarity_and_prior_ffnn/output_bias:0'])\n var_print_names.extend(['lstm_pem_attention_weights', 'lstm_pem_attention_bias'])\n if model.args.nn_components.find(\"global\") != -1 and model.args.global_score_ffnn[0] == 0:\n # print only if simple projection, otherwise it could be too many variables\n var_names.extend(['global_voting/psi_and_global_ffnn/output_weights:0',\n 'global_voting/psi_and_global_ffnn/output_bias:0'])\n var_print_names.extend(['psi_globalscore_weights', 'psi_globalscore_bias'])\n for var_name, print_name in zip(var_names, var_print_names):\n var_handle = [v for v in tf.global_variables() if v.name == var_name][0]\n result += print_name + \"=\" + str(model.sess.run(var_handle))\n return result\n\n\ndef _parse_args():\n parser = argparse.ArgumentParser()\n parser.add_argument(\"--experiment_name\", default=\"alldatasets_perparagr\", #\"standard\",\n help=\"under folder data/tfrecords/\")\n parser.add_argument(\"--training_name\", help=\"under folder data/tfrecords/\")\n parser.add_argument(\"--checkpoint_model_num\", default=None,\n help=\"e.g. give '7' or '4' if you want checkpoints/model-7 or model-4 \"\n \"to be restored\")\n\n parser.add_argument(\"--print_predictions\", dest='print_predictions', action='store_true',\n help=\"prints for each dataset the predictions to a file and compares with ground\"\n \"truth and simple baselines.\")\n parser.add_argument(\"--no_print_predictions\", dest='print_predictions', action='store_false')\n parser.set_defaults(print_predictions=True)\n parser.add_argument(\"--print_global_voters\", type=bool, default=False)\n parser.add_argument(\"--print_global_pairwise_scores\", type=bool, default=False)\n\n parser.add_argument(\"--ed_datasets\", default=\"aida_train.txt_z_aida_dev.txt_z_aida_test.txt_z_\"\n \"ace2004.txt_z_aquaint.txt_z_clueweb.txt_z_msnbc.txt_z_wikipedia.txt\")\n parser.add_argument(\"--el_datasets\", default=\"aida_train.txt_z_aida_dev.txt_z_aida_test.txt_z_\"\n \"ace2004.txt_z_aquaint.txt_z_clueweb.txt_z_msnbc.txt_z_wikipedia.txt\")\n\n parser.add_argument(\"--ed_val_datasets\", default=\"1\")\n parser.add_argument(\"--el_val_datasets\", default=\"1\")\n\n parser.add_argument(\"--p_e_m_algorithm\", type=bool, default=False,\n help=\"Baseline. Doesn't use the NN but only the p_e_m dictionary for\"\n \"its predictions.\")\n parser.add_argument(\"--all_spans_training\", type=bool, default=False)\n parser.add_argument(\"--entity_extension\", default=None, help=\"extension_entities or extension_entities_all etc\")\n\n parser.add_argument(\"--debug\", type=bool, default=False)\n parser.add_argument(\"--gm_bucketing_pempos\", default=None, help=\"0_1_2_7 will create bins 0, 1, 2, [3,7], [8,inf)\")\n parser.add_argument(\"--hardcoded_thr\", type=float, default=None)\n args = parser.parse_args()\n\n temp = \"all_spans_\" if args.all_spans_training else \"\"\n args.output_folder = config.base_folder+\"data/tfrecords/\" + \\\n args.experiment_name+\"/{}training_folder/\".format(temp)+ \\\n args.training_name+\"/\"\n args.checkpoints_folder = args.output_folder + \"checkpoints/\"\n args.predictions_folder = args.output_folder + \"predictions/\"\n\n if args.p_e_m_algorithm:\n args.predictions_folder = args.output_folder + \"p_e_m_predictions/\"\n\n if args.print_predictions and not os.path.exists(args.predictions_folder):\n os.makedirs(args.predictions_folder)\n if args.print_predictions and not os.path.exists(args.predictions_folder+\"ed/\"):\n os.makedirs(args.predictions_folder+\"ed/\")\n if args.print_predictions and not os.path.exists(args.predictions_folder+\"el/\"):\n os.makedirs(args.predictions_folder+\"el/\")\n\n train_args = load_train_args(args.output_folder, \"evaluate\")\n\n args.ed_datasets = args.ed_datasets.split('_z_') if args.ed_datasets != \"\" else None\n args.el_datasets = args.el_datasets.split('_z_') if args.el_datasets != \"\" else None\n args.ed_val_datasets = [int(x) for x in args.ed_val_datasets.split('_')]\n args.el_val_datasets = [int(x) for x in args.el_val_datasets.split('_')]\n args.gm_bucketing_pempos = [int(x) for x in args.gm_bucketing_pempos.split('_')] if args.gm_bucketing_pempos else []\n\n print(args)\n return args, train_args\n\n\nif __name__ == \"__main__\":\n args, train_args = _parse_args()\n print(\"train_args:\\n\", train_args)\n # TODO do this argument transfering in the load_train_args instead of train.py, evaluate.py, ensemble_evaluate.py\n train_args.checkpoint_model_num = args.checkpoint_model_num\n train_args.entity_extension = args.entity_extension\n train.args = args\n args.batch_size = train_args.batch_size\n printPredictions = None\n if args.print_predictions:\n from evaluation.print_predictions import PrintPredictions\n printPredictions = PrintPredictions(config.base_folder+\"data/tfrecords/\"+\n args.experiment_name+\"/\", args.predictions_folder, args.entity_extension,\n args.gm_bucketing_pempos, args.print_global_voters,\n args.print_global_pairwise_scores)\n evaluate()\n\n" }, { "path": "code/model/model.py", "content": "\n# b9d87f7 on Mar 21 Nikolaos Kolitsas ffnn dropout and some minor modif in evaluate to accept entity extension\n# ed_model_21_march\nimport numpy as np\nimport pickle\nimport tensorflow as tf\nimport model.config as config\nfrom .base_model import BaseModel\nimport model.util as util\n\n\nclass Model(BaseModel):\n\n def __init__(self, args, next_element):\n super().__init__(args)\n self.chunk_id, self.words, self.words_len, self.chars, self.chars_len,\\\n self.begin_span, self.end_span, self.spans_len,\\\n self.cand_entities, self.cand_entities_scores, self.cand_entities_labels,\\\n self.cand_entities_len, self.ground_truth, self.ground_truth_len,\\\n self.begin_gm, self.end_gm = next_element\n\n self.begin_span = tf.cast(self.begin_span, tf.int32)\n self.end_span = tf.cast(self.end_span, tf.int32)\n self.words_len = tf.cast(self.words_len, tf.int32)\n \"\"\"\n self.words: tf.int64, shape=[None, None] # shape = (batch size, max length of sentence in batch)\n self.words_len: tf.int32, shape=[None], # shape = (batch size)\n self.chars: tf.int64, shape=[None, None, None], # shape = (batch size, max length of sentence, max length of word)\n self.chars_len: tf.int64, shape=[None, None], # shape = (batch_size, max_length of sentence)\n self.begin_span: tf.int32, shape=[None, None], # shape = (batch_size, max number of candidate spans in one of the batch sentences)\n self.end_span: tf.int32, shape=[None, None],\n self.spans_len: tf.int64, shape=[None], # shape = (batch size)\n self.cand_entities: tf.int64, shape=[None, None, None], # shape = (batch size, max number of candidate spans, max number of cand entitites)\n self.cand_entities_scores: tf.float32, shape=[None, None, None],\n self.cand_entities_labels: tf.int64, shape=[None, None, None],\n # shape = (batch_size, max number of candidate spans)\n self.cand_entities_len: tf.int64, shape=[None, None],\n self.ground_truth: tf.int64, shape=[None, None], # shape = (batch_size, max number of candidate spans)\n self.ground_truth_len: tf.int64, shape=[None], # shape = (batch_size)\n self.begin_gm: tf.int64, shape=[None, None], # shape = (batch_size, max number of gold mentions)\n self.end_gm = tf.placeholder(tf.int64, shape=[None, None],\n \"\"\"\n with open(config.base_folder +\"data/tfrecords/\" + self.args.experiment_name +\n \"/word_char_maps.pickle\", 'rb') as handle:\n _, id2word, _, id2char, _, _ = pickle.load(handle)\n self.nwords = len(id2word)\n self.nchars = len(id2char)\n\n self.loss_mask = self._sequence_mask_v13(self.cand_entities_len, tf.shape(self.cand_entities_scores)[2])\n\n def add_placeholders(self):\n \"\"\"Define placeholders = entries to computational graph\"\"\"\n self.dropout = tf.placeholder(dtype=tf.float32, shape=[], name=\"dropout\")\n self.lr = tf.placeholder(dtype=tf.float32, shape=[], name=\"lr\")\n\n def init_embeddings(self):\n print(\"\\n!!!! init embeddings !!!!\\n\")\n # read the numpy file\n embeddings_nparray = np.load(config.base_folder +\"data/tfrecords/\" + self.args.experiment_name +\n \"/embeddings_array.npy\")\n self.sess.run(self.word_embedding_init, feed_dict={self.word_embeddings_placeholder: embeddings_nparray})\n\n entity_embeddings_nparray = util.load_ent_vecs(self.args)\n self.sess.run(self.entity_embedding_init, feed_dict={self.entity_embeddings_placeholder: entity_embeddings_nparray})\n\n def add_embeddings_op(self):\n \"\"\"Defines self.word_embeddings\"\"\"\n with tf.variable_scope(\"words\"):\n _word_embeddings = tf.Variable(\n tf.constant(0.0, shape=[self.nwords, 300]),\n name=\"_word_embeddings\",\n dtype=tf.float32,\n trainable=False)\n\n self.word_embeddings_placeholder = tf.placeholder(tf.float32, [self.nwords, 300])\n self.word_embedding_init = _word_embeddings.assign(self.word_embeddings_placeholder)\n\n word_embeddings = tf.nn.embedding_lookup(_word_embeddings,\n self.words, name=\"word_embeddings\")\n self.pure_word_embeddings = word_embeddings\n #print(\"word_embeddings (after lookup) \", word_embeddings)\n\n with tf.variable_scope(\"chars\"):\n if self.args.use_chars:\n # get char embeddings matrix\n _char_embeddings = tf.get_variable(\n name=\"_char_embeddings\",\n dtype=tf.float32,\n shape=[self.nchars, self.args.dim_char], trainable=True)\n char_embeddings = tf.nn.embedding_lookup(_char_embeddings,\n self.chars, name=\"char_embeddings\")\n\n # char_embeddings: tf.float32, shape=[None, None, None, dim_char],\n # shape = (batch size, max length of sentence, max length of word, dim_char)\n # put the time dimension on axis=1\n s = tf.shape(char_embeddings)\n char_embeddings = tf.reshape(char_embeddings, shape=[s[0] * s[1], s[-2], self.args.dim_char])\n # (batch*sent_length, characters of word, dim_char)\n\n char_lengths = tf.reshape(self.chars_len, shape=[s[0] * s[1]])\n # shape = (batch_size*max_length of sentence)\n\n # bi lstm on chars\n cell_fw = tf.contrib.rnn.LSTMCell(self.args.hidden_size_char, state_is_tuple=True)\n cell_bw = tf.contrib.rnn.LSTMCell(self.args.hidden_size_char, state_is_tuple=True)\n _output = tf.nn.bidirectional_dynamic_rnn(\n cell_fw, cell_bw, char_embeddings,\n sequence_length=char_lengths, dtype=tf.float32)\n\n # read and concat output\n _, ((_, output_fw), (_, output_bw)) = _output\n output = tf.concat([output_fw, output_bw], axis=-1)\n\n # shape = (batch size, max sentence length, char hidden size)\n output = tf.reshape(output, shape=[s[0], s[1], 2 * self.args.hidden_size_char])\n #print(\"output after char lstm \", output)\n word_embeddings = tf.concat([word_embeddings, output], axis=-1) # concatenate word and char embeddings\n #print(\"word_embeddings with char after concatenation \", word_embeddings)\n # (batch, words, 300+2*100)\n self.word_embeddings = tf.nn.dropout(word_embeddings, self.dropout)\n\n with tf.variable_scope(\"entities\"):\n from preprocessing.util import load_wikiid2nnid\n self.nentities = len(load_wikiid2nnid(extension_name=self.args.entity_extension))\n _entity_embeddings = tf.Variable(\n tf.constant(0.0, shape=[self.nentities, 300]),\n name=\"_entity_embeddings\",\n dtype=tf.float32,\n trainable=self.args.train_ent_vecs)\n\n self.entity_embeddings_placeholder = tf.placeholder(tf.float32, [self.nentities, 300])\n self.entity_embedding_init = _entity_embeddings.assign(self.entity_embeddings_placeholder)\n\n self.entity_embeddings = tf.nn.embedding_lookup(_entity_embeddings, self.cand_entities,\n name=\"entity_embeddings\")\n self.pure_entity_embeddings = self.entity_embeddings\n if self.args.ent_vecs_regularization.startswith(\"l2\"): # 'l2' or 'l2dropout'\n self.entity_embeddings = tf.nn.l2_normalize(self.entity_embeddings, dim=3)\n # not necessary since i do normalization in the entity embed creation as well, just for safety\n if self.args.ent_vecs_regularization == \"dropout\" or \\\n self.args.ent_vecs_regularization == \"l2dropout\":\n self.entity_embeddings = tf.nn.dropout(self.entity_embeddings, self.dropout)\n #print(\"entity_embeddings = \", self.entity_embeddings)\n\n def add_context_emb_op(self):\n \"\"\"this method creates the bidirectional LSTM layer (takes input the v_k vectors and outputs the\n context-aware word embeddings x_k)\"\"\"\n with tf.variable_scope(\"context-bi-lstm\"):\n cell_fw = tf.contrib.rnn.LSTMCell(self.args.hidden_size_lstm)\n cell_bw = tf.contrib.rnn.LSTMCell(self.args.hidden_size_lstm)\n (output_fw, output_bw), _ = tf.nn.bidirectional_dynamic_rnn(\n cell_fw, cell_bw, self.word_embeddings,\n sequence_length=self.words_len, dtype=tf.float32)\n output = tf.concat([output_fw, output_bw], axis=-1)\n self.context_emb = tf.nn.dropout(output, self.dropout)\n #print(\"context_emb = \", self.context_emb) # [batch, words, 300]\n\n def add_span_emb_op(self):\n mention_emb_list = []\n # span embedding based on boundaries (start, end) and head mechanism. but do that on top of contextual bilistm\n # output or on top of original word+char embeddings. this flag determines that. The parer reports results when\n # using the contextual lstm emb as it achieves better score. Used for ablation studies.\n boundaries_input_vecs = self.word_embeddings if self.args.span_boundaries_from_wordemb else self.context_emb\n\n # the span embedding is modeled by g^m = [x_q; x_r; \\hat(x)^m] (formula (2) of paper)\n # \"boundaries\" mean use x_q and x_r. \"head\" means use also the head mechanism \\hat(x)^m (formula (3))\n if self.args.span_emb.find(\"boundaries\") != -1:\n # shape (batch, num_of_cand_spans, emb)\n mention_start_emb = tf.gather_nd(boundaries_input_vecs, tf.stack(\n [tf.tile(tf.expand_dims(tf.range(tf.shape(self.begin_span)[0]), 1), [1, tf.shape(self.begin_span)[1]]),\n self.begin_span], 2)) # extracts the x_q embedding for each candidate span\n # the tile command creates a 2d tensor with the batch information. first lines contains only zeros, second\n # line ones etc... because the begin_span tensor has the information which word inside this sentence is the\n # beginning of the candidate span.\n mention_emb_list.append(mention_start_emb)\n\n mention_end_emb = tf.gather_nd(boundaries_input_vecs, tf.stack(\n [tf.tile(tf.expand_dims(tf.range(tf.shape(self.begin_span)[0]), 1), [1, tf.shape(self.begin_span)[1]]),\n tf.nn.relu(self.end_span-1)], 2)) # -1 because the end of span in exclusive [start, end)\n # relu so that the 0 doesn't become -1 of course no valid candidate span end index is zero since [0,0) is empty\n mention_emb_list.append(mention_end_emb)\n #print(\"mention_start_emb = \", mention_start_emb)\n #print(\"mention_end_emb = \", mention_end_emb)\n\n mention_width = self.end_span - self.begin_span # [batch, num_mentions] the width of each candidate span\n\n if self.args.span_emb.find(\"head\") != -1: # here the attention is computed\n # here the \\hat(x)^m is computed (formula (2) and (3))\n self.max_mention_width = tf.minimum(self.args.max_mention_width,\n tf.reduce_max(self.end_span - self.begin_span))\n mention_indices = tf.range(self.max_mention_width) + \\\n tf.expand_dims(self.begin_span, 2) # [batch, num_mentions, max_mention_width]\n mention_indices = tf.minimum(tf.shape(self.word_embeddings)[1] - 1,\n mention_indices) # [batch, num_mentions, max_mention_width]\n #print(\"mention_indices = \", mention_indices)\n batch_index = tf.tile(tf.expand_dims(tf.expand_dims(tf.range(tf.shape(mention_indices)[0]), 1), 2),\n [1, tf.shape(mention_indices)[1], tf.shape(mention_indices)[2]])\n mention_indices = tf.stack([batch_index, mention_indices], 3)\n # [batch, num_mentions, max_mention_width, [row,col] ] 4d tensor\n\n # for the boundaries we had the option to take them either from x_k (output of bilstm) or from v_k\n # the head is derived either from the same option as boundaries or from the v_k.\n head_input_vecs = boundaries_input_vecs if self.args.model_heads_from_bilstm else self.word_embeddings\n mention_text_emb = tf.gather_nd(head_input_vecs, mention_indices)\n # [batch, num_mentions, max_mention_width, 500 ] 4d tensor\n #print(\"mention_text_emb = \", mention_text_emb)\n\n with tf.variable_scope(\"head_scores\"):\n # from [batch, max_sent_len, 300] to [batch, max_sent_len, 1]\n self.head_scores = util.projection(boundaries_input_vecs, 1)\n # [batch, num_mentions, max_mention_width, 1]\n mention_head_scores = tf.gather_nd(self.head_scores, mention_indices)\n # print(\"mention_head_scores = \", mention_head_scores)\n\n # depending on tensorflow version we do the same with different operations (since each candidate span is not\n # of the same length we mask out the invalid indices created above (mention_indices)).\n temp_mask = self._sequence_mask_v13(mention_width, self.max_mention_width)\n # still code for masking invalid indices for the head computation\n mention_mask = tf.expand_dims(temp_mask, 3) # [batch, num_mentions, max_mention_width, 1]\n mention_mask = tf.minimum(1.0, tf.maximum(self.args.zero, mention_mask)) # 1e-3\n # formula (3) computation\n mention_attention = tf.nn.softmax(mention_head_scores + tf.log(mention_mask),\n dim=2) # [batch, num_mentions, max_mention_width, 1]\n mention_head_emb = tf.reduce_sum(mention_attention * mention_text_emb, 2) # [batch, num_mentions, emb]\n #print(\"mention_head_emb = \", mention_head_emb)\n mention_emb_list.append(mention_head_emb)\n\n self.span_emb = tf.concat(mention_emb_list, 2) # [batch, num_mentions, emb i.e. 1700] formula (2) concatenation\n #print(\"span_emb = \", self.span_emb)\n\n def add_lstm_score_op(self):\n with tf.variable_scope(\"span_emb_ffnn\"):\n # [batch, num_mentions, 300]\n # the span embedding can have different size depending on the chosen hyperparameters. We project it to 300\n # dims to match the entity embeddings (formula 4)\n if self.args.span_emb_ffnn[0] == 0:\n span_emb_projected = util.projection(self.span_emb, 300)\n else:\n hidden_layers, hidden_size = self.args.span_emb_ffnn[0], self.args.span_emb_ffnn[1]\n span_emb_projected = util.ffnn(self.span_emb, hidden_layers, hidden_size, 300,\n self.dropout if self.args.ffnn_dropout else None)\n #print(\"span_emb_projected = \", span_emb_projected)\n # formula (6) computation. this is the lstm score\n scores = tf.matmul(tf.expand_dims(span_emb_projected, 2), self.entity_embeddings, transpose_b=True)\n #print(\"scores = \", scores)\n self.similarity_scores = tf.squeeze(scores, axis=2) # [batch, num_mentions, 1, 30]\n #print(\"scores = \", self.similarity_scores) # [batch, num_mentions, 30]\n\n def add_local_attention_op(self):\n attention_entity_emb = self.pure_entity_embeddings if self.args.attention_ent_vecs_no_regularization else self.entity_embeddings\n with tf.variable_scope(\"attention\"):\n K = self.args.attention_K\n left_mask = self._sequence_mask_v13(self.begin_span, K) # number of words on the left (left window)\n right_mask = self._sequence_mask_v13(tf.expand_dims(self.words_len, 1) - self.end_span, K)\n # number of words on the right. of course i don't get more than K even if more words exist.\n ctxt_mask = tf.concat([left_mask, right_mask], 2) # [batch, num_of_spans, 2*K]\n ctxt_mask = tf.log(tf.minimum(1.0, tf.maximum(self.args.zero, ctxt_mask)))\n # T, T, T, F, F | T, T, F, F, F\n # -1, -2, -3, -4, -5 +0, +1, +2, +3, +4\n\n leftctxt_indices = tf.maximum(0, tf.range(-1, -K - 1, -1) +\n tf.expand_dims(self.begin_span, 2)) # [batch, num_mentions, K]\n rightctxt_indices = tf.minimum(tf.shape(self.pure_word_embeddings)[1] - 1, tf.range(K) +\n tf.expand_dims(self.end_span, 2)) # [batch, num_mentions, K]\n ctxt_indices = tf.concat([leftctxt_indices, rightctxt_indices], 2) # [batch, num_mentions, 2*K]\n\n batch_index = tf.tile(tf.expand_dims(tf.expand_dims(tf.range(tf.shape(ctxt_indices)[0]), 1), 2),\n [1, tf.shape(ctxt_indices)[1], tf.shape(ctxt_indices)[2]])\n ctxt_indices = tf.stack([batch_index, ctxt_indices], 3)\n # [batch, num_of_spans, 2*K, 2] the last dimension is row,col for gather_nd\n # [batch, num_of_spans, 2*K, [row,col]]\n\n att_x_w = self.pure_word_embeddings # [batch, max_sent_len, 300]\n if self.args.attention_on_lstm and self.args.nn_components.find(\"lstm\") != -1:\n # ablation: here the attention is computed on the output of the lstm layer x_k instead of using the\n # pure word2vec vectors. (word2vec used in paper).\n att_x_w = util.projection(self.context_emb, 300) # if tf.shape(self.context_emb)[-1] != 300 else self.context_emb\n\n ctxt_word_emb = tf.gather_nd(att_x_w, ctxt_indices)\n # [batch, num_of_spans, 2K, emb_size] emb_size = 300 only pure word emb used (word2vec)\n # and not after we add char emb and dropout\n\n # in this implementation we don't use the diagonal A and B arrays that are mentioned in\n # Ganea and Hoffmann 2017 (only used in the ablations)\n temp = attention_entity_emb\n if self.args.attention_use_AB:\n att_A = tf.get_variable(\"att_A\", [300])\n temp = att_A * attention_entity_emb\n scores = tf.matmul(ctxt_word_emb, temp, transpose_b=True)\n scores = tf.reduce_max(scores, reduction_indices=[-1]) # max score of each word for each span acquired from any cand entity\n scores = scores + ctxt_mask # some words are not valid out of window so we assign to them very low score\n top_values, _ = tf.nn.top_k(scores, self.args.attention_R)\n # [batch, num_of_spans, R]\n R_value = top_values[:, :, -1] # [batch, num_of_spans]\n R_value = tf.maximum(self.args.zero, R_value) # so to avoid keeping words that\n # have max score with any of the entities <=0 (also score = 0 can have words with\n # padding candidate entities)\n\n threshold = tf.tile(tf.expand_dims(R_value, 2), [1, 1, 2 * K])\n # [batch, num_of_spans, 2K]\n scores = scores - tf.to_float(((scores - threshold) < 0)) * 50 # 50 where score=thr\n scores = tf.nn.softmax(scores, dim=2) # [batch, num_of_spans, 2K]\n scores = tf.expand_dims(scores, 3) # [batch, num_of_spans, 2K, 1]\n # [batch, num_of_spans, 2K, 1] * [batch, num_of_spans, 2K, emb_size]\n # = [batch, num_of_spans, 2K, emb_size]\n x_c = tf.reduce_sum(scores * ctxt_word_emb, 2) # = [batch, num_of_spans, emb_size]\n if self.args.attention_use_AB:\n att_B = tf.get_variable(\"att_B\", [300])\n x_c = att_B * x_c\n x_c = tf.expand_dims(x_c, 3) # [batch, num_of_spans, emb_size, 1]\n # [batch, num_of_spans, 30, emb_size=300] mul with [batch, num_of_spans, emb_size, 1]\n x_e__x_c = tf.matmul(attention_entity_emb, x_c) # [batch, num_of_spans, 30, 1]\n x_e__x_c = tf.squeeze(x_e__x_c, axis=3) # [batch, num_of_spans, 30]\n self.attention_scores = x_e__x_c\n\n def add_cand_ent_scores_op(self):\n self.log_cand_entities_scores = tf.log(tf.minimum(1.0, tf.maximum(self.args.zero, self.cand_entities_scores)))\n stack_values = []\n if self.args.nn_components.find(\"lstm\") != -1:\n stack_values.append(self.similarity_scores)\n if self.args.nn_components.find(\"pem\") != -1:\n stack_values.append(self.log_cand_entities_scores)\n if self.args.nn_components.find(\"attention\") != -1:\n stack_values.append(self.attention_scores)\n\n scalar_predictors = tf.stack(stack_values, 3)\n #print(\"scalar_predictors = \", scalar_predictors) # [batch, num_mentions, 30, 3]\n\n with tf.variable_scope(\"similarity_and_prior_ffnn\"):\n if self.args.final_score_ffnn[0] == 0:\n self.final_scores = util.projection(scalar_predictors, 1) # [batch, num_mentions, 30, 1]\n else:\n hidden_layers, hidden_size = self.args.final_score_ffnn[0], self.args.final_score_ffnn[1]\n self.final_scores = util.ffnn(scalar_predictors, hidden_layers, hidden_size, 1,\n self.dropout if self.args.ffnn_dropout else None)\n self.final_scores = tf.squeeze(self.final_scores, axis=3) # squeeze to [batch, num_mentions, 30]\n #print(\"final_scores = \", self.final_scores)\n\n def add_global_voting_op(self):\n with tf.variable_scope(\"global_voting\"):\n self.final_scores_before_global = - (1 - self.loss_mask) * 50 + self.final_scores\n gmask = tf.to_float(((self.final_scores_before_global - self.args.global_thr) >= 0)) # [b,s,30]\n\n masked_entity_emb = self.pure_entity_embeddings * tf.expand_dims(gmask, axis=3) # [b,s,30,300] * [b,s,30,1]\n batch_size = tf.shape(masked_entity_emb)[0]\n all_voters_emb = tf.reduce_sum(tf.reshape(masked_entity_emb, [batch_size, -1, 300]), axis=1,\n keep_dims=True) # [b, 1, 300]\n span_voters_emb = tf.reduce_sum(masked_entity_emb, axis=2) # [batch, num_of_spans, 300]\n valid_voters_emb = all_voters_emb - span_voters_emb\n # [b, 1, 300] - [batch, spans, 300] = [batch, spans, 300] (broadcasting)\n # [300] - [batch, spans, 300] = [batch, spans, 300] (broadcasting)\n valid_voters_emb = tf.nn.l2_normalize(valid_voters_emb, dim=2)\n\n self.global_voting_scores = tf.squeeze(tf.matmul(self.pure_entity_embeddings, tf.expand_dims(valid_voters_emb, axis=3)), axis=3)\n # [b,s,30,300] matmul [b,s,300,1] --> [b,s,30,1]-->[b,s,30]\n\n scalar_predictors = tf.stack([self.final_scores_before_global, self.global_voting_scores], 3)\n #print(\"scalar_predictors = \", scalar_predictors) #[b, s, 30, 2]\n with tf.variable_scope(\"psi_and_global_ffnn\"):\n if self.args.global_score_ffnn[0] == 0:\n self.final_scores = util.projection(scalar_predictors, 1)\n else:\n hidden_layers, hidden_size = self.args.global_score_ffnn[0], self.args.global_score_ffnn[1]\n self.final_scores = util.ffnn(scalar_predictors, hidden_layers, hidden_size, 1,\n self.dropout if self.args.ffnn_dropout else None)\n # [batch, num_mentions, 30, 1] squeeze to [batch, num_mentions, 30]\n self.final_scores = tf.squeeze(self.final_scores, axis=3)\n #print(\"final_scores = \", self.final_scores)\n\n def add_loss_op(self):\n cand_entities_labels = tf.cast(self.cand_entities_labels, tf.float32)\n loss1 = cand_entities_labels * tf.nn.relu(self.args.gamma_thr - self.final_scores)\n loss2 = (1 - cand_entities_labels) * tf.nn.relu(self.final_scores)\n self.loss = loss1 + loss2\n if self.args.nn_components.find(\"global\") != -1 and not self.args.global_one_loss:\n loss3 = cand_entities_labels * tf.nn.relu(self.args.gamma_thr - self.final_scores_before_global)\n loss4 = (1 - cand_entities_labels) * tf.nn.relu(self.final_scores_before_global)\n self.loss = loss1 + loss2 + loss3 + loss4\n #print(\"loss_mask = \", loss_mask)\n self.loss = self.loss_mask * self.loss\n self.loss = tf.reduce_sum(self.loss)\n # for tensorboard\n #tf.summary.scalar(\"loss\", self.loss)\n\n def build(self):\n self.add_placeholders()\n self.add_embeddings_op()\n if self.args.nn_components.find(\"lstm\") != -1:\n self.add_context_emb_op()\n self.add_span_emb_op()\n self.add_lstm_score_op()\n if self.args.nn_components.find(\"attention\") != -1:\n self.add_local_attention_op()\n self.add_cand_ent_scores_op()\n if self.args.nn_components.find(\"global\") != -1:\n self.add_global_voting_op()\n if self.args.running_mode.startswith(\"train\"):\n self.add_loss_op()\n # Generic functions that add training op\n self.add_train_op(self.args.lr_method, self.lr, self.loss, self.args.clip)\n self.merged_summary_op = tf.summary.merge_all()\n\n if self.args.running_mode == \"train_continue\":\n self.restore_session(\"latest\")\n elif self.args.running_mode == \"train\":\n self.initialize_session() # now self.sess is defined and vars are init\n self.init_embeddings()\n\n # if we run the evaluate.py script then we should call explicitly the model.restore(\"ed\")\n # or model.restore(\"el\"). here it doesn't initialize or restore values for the evaluate.py\n # case.\n\n def _sequence_mask_v13(self, mytensor, max_width):\n \"\"\"mytensor is a 2d tensor\"\"\"\n if not tf.__version__.startswith(\"1.4\"):\n temp_shape = tf.shape(mytensor)\n temp = tf.sequence_mask(tf.reshape(mytensor, [-1]), max_width, dtype=tf.float32)\n temp_mask = tf.reshape(temp, [temp_shape[0], temp_shape[1], tf.shape(temp)[-1]])\n else:\n temp_mask = tf.sequence_mask(mytensor, max_width, dtype=tf.float32)\n return temp_mask\n\n\n" }, { "path": "code/model/model_ablations.py", "content": "import numpy as np\nimport pickle\nimport tensorflow as tf\nimport model.config as config\n\nfrom .base_model import BaseModel\nimport model.util as util\n\nclass Model(BaseModel):\n\n def __init__(self, args, next_element):\n super().__init__(args)\n self.chunk_id, self.words, self.words_len, self.chars, self.chars_len,\\\n self.begin_span, self.end_span, self.spans_len,\\\n self.cand_entities, self.cand_entities_scores, self.cand_entities_labels,\\\n self.cand_entities_len, self.ground_truth, self.ground_truth_len,\\\n self.begin_gm, self.end_gm = next_element\n\n self.begin_span = tf.cast(self.begin_span, tf.int32)\n self.end_span = tf.cast(self.end_span, tf.int32)\n self.words_len = tf.cast(self.words_len, tf.int32) # TODO new command caution if it is breaking old models\n \"\"\"\n self.words: tf.int64, shape=[None, None] # shape = (batch size, max length of sentence in batch)\n self.words_len: tf.int32, shape=[None], # shape = (batch size)\n self.chars: tf.int64, shape=[None, None, None], # shape = (batch size, max length of sentence, max length of word)\n self.chars_len: tf.int64, shape=[None, None], # shape = (batch_size, max_length of sentence)\n self.begin_span: tf.int32, shape=[None, None], # shape = (batch_size, max number of candidate spans in one of the batch sentences)\n self.end_span: tf.int32, shape=[None, None],\n self.spans_len: tf.int64, shape=[None], # shape = (batch size)\n self.cand_entities: tf.int64, shape=[None, None, None], # shape = (batch size, max number of candidate spans, max number of cand entitites)\n self.cand_entities_scores: tf.float32, shape=[None, None, None],\n self.cand_entities_labels: tf.int64, shape=[None, None, None],\n # shape = (batch_size, max number of candidate spans)\n self.cand_entities_len: tf.int64, shape=[None, None],\n self.ground_truth: tf.int64, shape=[None, None], # shape = (batch_size, max number of candidate spans)\n self.ground_truth_len: tf.int64, shape=[None], # shape = (batch_size)\n self.begin_gm: tf.int64, shape=[None, None], # shape = (batch_size, max number of gold mentions)\n self.end_gm = tf.placeholder(tf.int64, shape=[None, None],\n \"\"\"\n with open(config.base_folder +\"data/tfrecords/\" + self.args.experiment_name+\n \"/word_char_maps.pickle\", 'rb') as handle:\n _, id2word, _, id2char, _, _ = pickle.load(handle)\n self.nwords = len(id2word)\n self.nchars = len(id2char)\n\n if self.args.cand_ent_num_restriction:\n self.cand_entities = tf.slice(self.cand_entities, [0, 0, 0], [-1, -1, self.args.cand_ent_num_restriction])\n self.cand_entities_scores = tf.slice(self.cand_entities_scores, [0, 0, 0], [-1, -1, self.args.cand_ent_num_restriction])\n self.cand_entities_labels = tf.slice(self.cand_entities_labels, [0, 0, 0], [-1, -1, self.args.cand_ent_num_restriction])\n self.cand_entities_len = tf.minimum(self.cand_entities_len, self.args.cand_ent_num_restriction)\n self.ffnn_l2normalization_op_list = []\n self.loss_mask = self._sequence_mask_v13(self.cand_entities_len, tf.shape(self.cand_entities_scores)[2])\n\n def add_placeholders(self):\n self.dropout = tf.placeholder(dtype=tf.float32, shape=[], name=\"dropout\")\n self.lr = tf.placeholder(dtype=tf.float32, shape=[], name=\"lr\")\n\n def init_embeddings(self):\n print(\"\\n!!!! init embeddings !!!!\\n\")\n # read the numpy file\n embeddings_nparray = np.load(config.base_folder +\"data/tfrecords/\" + self.args.experiment_name+\n \"/embeddings_array.npy\")\n self.sess.run(self.word_embedding_init, feed_dict={self.word_embeddings_placeholder: embeddings_nparray})\n\n entity_embeddings_nparray = util.load_ent_vecs(self.args)\n self.sess.run(self.entity_embedding_init, feed_dict={self.entity_embeddings_placeholder: entity_embeddings_nparray})\n\n def add_embeddings_op(self):\n\n \"\"\"Defines self.word_embeddings\"\"\"\n #with tf.device('/cpu:0'), tf.name_scope(\"embedding\"):\n with tf.variable_scope(\"words\"):\n _word_embeddings = tf.Variable(\n tf.constant(0.0, shape=[self.nwords, 300]),\n name=\"_word_embeddings\",\n dtype=tf.float32,\n trainable=False)\n\n self.word_embeddings_placeholder = tf.placeholder(tf.float32, [self.nwords, 300])\n self.word_embedding_init = _word_embeddings.assign(self.word_embeddings_placeholder)\n\n word_embeddings = tf.nn.embedding_lookup(_word_embeddings,\n self.words, name=\"word_embeddings\")\n self.pure_word_embeddings = word_embeddings\n #print(\"word_embeddings (after lookup) \", word_embeddings)\n\n with tf.variable_scope(\"chars\"):\n if self.args.use_chars:\n # get char embeddings matrix\n _char_embeddings = tf.get_variable(\n name=\"_char_embeddings\",\n dtype=tf.float32,\n shape=[self.nchars, self.args.dim_char], trainable=True)\n char_embeddings = tf.nn.embedding_lookup(_char_embeddings,\n self.chars, name=\"char_embeddings\")\n\n # char_embeddings: tf.float32, shape=[None, None, None, dim_char],\n # shape = (batch size, max length of sentence, max length of word, dim_char)\n # put the time dimension on axis=1\n s = tf.shape(char_embeddings)\n char_embeddings = tf.reshape(char_embeddings,\n shape=[s[0] * s[1], s[-2], self.args.dim_char])\n char_lengths = tf.reshape(self.chars_len, shape=[s[0] * s[1]])\n\n # bi lstm on chars\n cell_fw = tf.contrib.rnn.LSTMCell(self.args.hidden_size_char, state_is_tuple=True)\n cell_bw = tf.contrib.rnn.LSTMCell(self.args.hidden_size_char, state_is_tuple=True)\n _output = tf.nn.bidirectional_dynamic_rnn(\n cell_fw, cell_bw, char_embeddings,\n sequence_length=char_lengths, dtype=tf.float32)\n\n # read and concat output\n _, ((_, output_fw), (_, output_bw)) = _output\n output = tf.concat([output_fw, output_bw], axis=-1)\n\n # shape = (batch size, max sentence length, char hidden size)\n output = tf.reshape(output, shape=[s[0], s[1], 2 * self.args.hidden_size_char])\n #print(\"output after char lstm \", output)\n word_embeddings = tf.concat([word_embeddings, output], axis=-1)\n #print(\"word_embeddings with char after concatenation \", word_embeddings)\n # (batch, words, 300+2*100)\n self.word_embeddings = tf.nn.dropout(word_embeddings, self.dropout)\n\n with tf.variable_scope(\"entities\"):\n from preprocessing.util import load_wikiid2nnid\n self.nentities = len(load_wikiid2nnid(extension_name=self.args.entity_extension))\n _entity_embeddings = tf.Variable(\n tf.constant(0.0, shape=[self.nentities, 300]),\n name=\"_entity_embeddings\",\n dtype=tf.float32,\n trainable=self.args.train_ent_vecs)\n\n self.entity_embeddings_placeholder = tf.placeholder(tf.float32, [self.nentities, 300])\n self.entity_embedding_init = _entity_embeddings.assign(self.entity_embeddings_placeholder)\n\n self.entity_embeddings = tf.nn.embedding_lookup(_entity_embeddings, self.cand_entities,\n name=\"entity_embeddings\")\n self.pure_entity_embeddings = self.entity_embeddings\n if self.args.ent_vecs_regularization.startswith(\"l2\"): # 'l2' or 'l2dropout'\n self.entity_embeddings = tf.nn.l2_normalize(self.entity_embeddings, dim=3)\n if self.args.ent_vecs_regularization == \"dropout\" or \\\n self.args.ent_vecs_regularization == \"l2dropout\":\n self.entity_embeddings = tf.nn.dropout(self.entity_embeddings, self.dropout)\n #print(\"entity_embeddings = \", self.entity_embeddings)\n\n def add_context_emb_op(self):\n \"\"\"this method creates the bidirectional LSTM layer (takes input the v_k vectors and outputs the\n context-aware word embeddings x_k)\"\"\"\n with tf.variable_scope(\"context-bi-lstm\"):\n cell_fw = tf.contrib.rnn.LSTMCell(self.args.hidden_size_lstm)\n cell_bw = tf.contrib.rnn.LSTMCell(self.args.hidden_size_lstm)\n (output_fw, output_bw), _ = tf.nn.bidirectional_dynamic_rnn(\n cell_fw, cell_bw, self.word_embeddings,\n sequence_length=self.words_len, dtype=tf.float32)\n output = tf.concat([output_fw, output_bw], axis=-1)\n self.context_emb = tf.nn.dropout(output, self.dropout)\n #print(\"context_emb = \", self.context_emb) # [batch, words, 300]\n\n def add_span_emb_op(self):\n mention_emb_list = []\n # span embedding based on boundaries (start, end) and head mechanism. but do that on top of contextual bilistm\n # output or on top of original word+char embeddings. this flag determines that. The parer reports results when\n # using the contextual lstm emb as it achieves better score. Used for ablation studies.\n boundaries_input_vecs = self.word_embeddings if self.args.span_boundaries_from_wordemb else self.context_emb\n\n # the span embedding is modeled by g^m = [x_q; x_r; \\hat(x)^m] (formula (2) of paper)\n # \"boundaries\" mean use x_q and x_r. \"head\" means use also the head mechanism \\hat(x)^m (formula (3))\n if self.args.span_emb.find(\"boundaries\") != -1:\n # shape (batch, num_of_cand_spans, emb)\n mention_start_emb = tf.gather_nd(boundaries_input_vecs, tf.stack(\n [tf.tile(tf.expand_dims(tf.range(tf.shape(self.begin_span)[0]), 1), [1, tf.shape(self.begin_span)[1]]),\n self.begin_span], 2)) # extracts the x_q embedding for each candidate span\n # the tile command creates a 2d tensor with the batch information. first lines contains only zeros, second\n # line ones etc... because the begin_span tensor has the information which word inside this sentence is the\n # beginning of the candidate span.\n mention_emb_list.append(mention_start_emb)\n\n mention_end_emb = tf.gather_nd(boundaries_input_vecs, tf.stack(\n [tf.tile(tf.expand_dims(tf.range(tf.shape(self.begin_span)[0]), 1), [1, tf.shape(self.begin_span)[1]]),\n tf.nn.relu(self.end_span-1)], 2)) # -1 because the end of span in exclusive [start, end)\n # relu so that the 0 doesn't become -1 of course no valid candidate span end index is zero since [0,0) is empty\n mention_emb_list.append(mention_end_emb)\n #print(\"mention_start_emb = \", mention_start_emb)\n #print(\"mention_end_emb = \", mention_end_emb)\n\n mention_width = self.end_span - self.begin_span # [batch, num_mentions] the width of each candidate span\n\n if self.args.span_emb.find(\"head\") != -1: # here the attention is computed\n # here the \\hat(x)^m is computed (formula (2) and (3))\n self.max_mention_width = tf.minimum(self.args.max_mention_width,\n tf.reduce_max(self.end_span - self.begin_span))\n mention_indices = tf.range(self.max_mention_width) + \\\n tf.expand_dims(self.begin_span, 2) # [batch, num_mentions, max_mention_width]\n mention_indices = tf.minimum(tf.shape(self.word_embeddings)[1] - 1,\n mention_indices) # [batch, num_mentions, max_mention_width]\n #print(\"mention_indices = \", mention_indices)\n batch_index = tf.tile(tf.expand_dims(tf.expand_dims(tf.range(tf.shape(mention_indices)[0]), 1), 2),\n [1, tf.shape(mention_indices)[1], tf.shape(mention_indices)[2]])\n mention_indices = tf.stack([batch_index, mention_indices], 3)\n # [batch, num_mentions, max_mention_width, [row,col] ] 4d tensor\n\n # for the boundaries we had the option to take them either from x_k (output of bilstm) or from v_k\n # the head is derived either from the same option as boundaries or from the v_k.\n head_input_vecs = boundaries_input_vecs if self.args.model_heads_from_bilstm else self.word_embeddings\n mention_text_emb = tf.gather_nd(head_input_vecs, mention_indices)\n # [batch, num_mentions, max_mention_width, 400 ] 4d tensor\n #print(\"mention_text_emb = \", mention_text_emb)\n\n with tf.variable_scope(\"head_scores\"):\n # from [batch, max_sent_len, 300] to [batch, max_sent_len, 1]\n self.head_scores = util.projection(boundaries_input_vecs, 1, model=self)\n # [batch, num_mentions, max_mention_width, 1]\n mention_head_scores = tf.gather_nd(self.head_scores, mention_indices)\n #print(\"mention_head_scores = \", mention_head_scores) #\n\n # depending on tensorflow version we do the same with different operations (since each candidate span is not\n # of the same length we mask out the invalid indices created above (mention_indices)).\n temp_mask = self._sequence_mask_v13(mention_width, self.max_mention_width)\n # still code for masking invalid indices for the head computation\n mention_mask = tf.expand_dims(temp_mask, 3) # [batch, num_mentions, max_mention_width, 1]\n mention_mask = tf.minimum(1.0, tf.maximum(self.args.zero, mention_mask)) # 1e-3\n # formula (3) computation\n mention_attention = tf.nn.softmax(mention_head_scores + tf.log(mention_mask),\n dim=2) # [batch, num_mentions, max_mention_width, 1]\n mention_head_emb = tf.reduce_sum(mention_attention * mention_text_emb, 2) # [batch, num_mentions, emb]\n #print(\"mention_head_emb = \", mention_head_emb)\n mention_emb_list.append(mention_head_emb)\n\n self.span_emb = tf.concat(mention_emb_list, 2) # [batch, num_mentions, emb i.e. 1700] formula (2) concatenation\n #print(\"span_emb = \", self.span_emb)\n\n def add_lstm_score_op(self):\n #print(\"cand_entities = \", self.cand_entities)\n with tf.variable_scope(\"span_emb_ffnn\"):\n # [batch, num_mentions, 300]\n # the span embedding can have different size depending on the chosen hyperparameters. We project it to 300\n # dims to match the entity embeddings (formula 4)\n if self.args.span_emb_ffnn[0] == 0:\n span_emb_projected = util.projection(self.span_emb, 300, model=self)\n else:\n hidden_layers, hidden_size = self.args.span_emb_ffnn[0], self.args.span_emb_ffnn[1]\n span_emb_projected = util.ffnn(self.span_emb, hidden_layers, hidden_size, 300,\n self.dropout if self.args.ffnn_dropout else None, model=self)\n #print(\"span_emb_projected = \", span_emb_projected)\n # formula (6) computation. this is the lstm score\n scores = tf.matmul(tf.expand_dims(span_emb_projected, 2), self.entity_embeddings, transpose_b=True)\n #print(\"scores = \", scores)\n self.similarity_scores = tf.squeeze(scores, axis=2) # [batch, num_mentions, 1, 30]\n #print(\"scores = \", self.similarity_scores) # [batch, num_mentions, 30]\n\n def add_local_attention_op(self):\n # shape=(b, num_of_spans, 30, 300)\n attention_entity_emb = self.pure_entity_embeddings if self.args.attention_ent_vecs_no_regularization else self.entity_embeddings\n with tf.variable_scope(\"attention\"):\n K = self.args.attention_K\n left_mask = self._sequence_mask_v13(self.begin_span, K)\n right_mask = self._sequence_mask_v13(tf.expand_dims(self.words_len, 1) - self.end_span, K)\n # number of words on the right. of course i don't get more than K even if more words exist.\n ctxt_mask = tf.concat([left_mask, right_mask], 2) # [batch, num_of_spans, 2*K]\n ctxt_mask = tf.log(tf.minimum(1.0, tf.maximum(self.args.zero, ctxt_mask)))\n # T, T, T, F, F | T, T, F, F, F\n # -1, -2, -3, -4, -5 +0, +1, +2, +3, +4\n\n leftctxt_indices = tf.maximum(0, tf.range(-1, -K - 1, -1) + \\\n tf.expand_dims(self.begin_span, 2)) # [batch, num_mentions, K]\n rightctxt_indices = tf.minimum(tf.shape(self.pure_word_embeddings)[1] - 1, tf.range(K) + \\\n tf.expand_dims(self.end_span, 2)) # [batch, num_mentions, K]\n ctxt_indices = tf.concat([leftctxt_indices, rightctxt_indices], 2) # [batch, num_mentions, 2*K]\n\n batch_index = tf.tile(tf.expand_dims(tf.expand_dims(tf.range(tf.shape(ctxt_indices)[0]), 1), 2),\n [1, tf.shape(ctxt_indices)[1], tf.shape(ctxt_indices)[2]])\n ctxt_indices = tf.stack([batch_index, ctxt_indices], 3)\n # [batch, num_of_spans, 2*K, 2] the last dimension is row,col for gather_nd\n # [batch, num_of_spans, 2*K, [row,col]]\n\n att_x_w = self.pure_word_embeddings # [batch, max_sent_len, 300]\n if self.args.attention_on_lstm and self.args.nn_components.find(\"lstm\") != -1:\n # ablation: here the attention is computed on the output of the lstm layer x_k instead of using the\n # pure word2vec vectors. (word2vec used in paper).\n # [batch, max_sent_len, hidden_size_of_lstm*2] project it to 300\n att_x_w = util.projection(self.context_emb, 300, model=self) # if tf.shape(self.context_emb)[-1] != 300 else self.context_emb\n\n ctxt_word_emb = tf.gather_nd(att_x_w, ctxt_indices)\n # [batch, num_of_spans, 2K, emb_size] emb_size = 300 only pure word emb used\n # and not after we add char emb and dropout\n\n x_c_voters = attention_entity_emb\n # restrict the number of entities that participate in the forming of the x_c context vector\n if self.args.attention_retricted_num_of_entities:\n x_c_voters = tf.slice(attention_entity_emb, [0, 0, 0, 0],\n [-1, -1, self.args.attention_retricted_num_of_entities, -1])\n if self.args.attention_use_AB:\n att_A = tf.get_variable(\"att_A\", [300])\n x_c_voters = att_A * x_c_voters\n # [b, num_of_spans, 2*K, 300] mul [b, num_of_spans, 30, 300] instead of 30 it can be the reduced number of entities\n scores = tf.matmul(ctxt_word_emb, x_c_voters, transpose_b=True) # [b, spans, 2K, 30]\n scores = tf.reduce_max(scores, reduction_indices=[-1]) # max score of each word for each span acquired from any cand entity\n scores = scores + ctxt_mask # some words are not valid out of window so we assign to them very low score\n top_values, _ = tf.nn.top_k(scores, self.args.attention_R)\n # [batch, num_of_spans, R]\n R_value = top_values[:, :, -1] # [batch, num_of_spans]\n R_value = tf.maximum(self.args.zero, R_value) # so to avoid keeping words that\n # have max score with any of the entities <=0 (also score = 0 can have words with\n # padding candidate entities)\n\n threshold = tf.tile(tf.expand_dims(R_value, 2), [1, 1, 2 * K])\n # [batch, num_of_spans, 2K]\n scores = scores - tf.to_float(((scores - threshold) < 0)) * 50 # 50 where score=thr\n scores = tf.nn.softmax(scores, dim=2) # [batch, num_of_spans, 2K]\n scores = tf.expand_dims(scores, 3) # [batch, num_of_spans, 2K, 1]\n # [batch, num_of_spans, 2K, 1] * [batch, num_of_spans, 2K, emb_size]\n # = [batch, num_of_spans, 2K, emb_size]\n x_c = tf.reduce_sum(scores * ctxt_word_emb, 2) # = [batch, num_of_spans, emb_size]\n if self.args.attention_use_AB:\n att_B = tf.get_variable(\"att_B\", [300])\n x_c = att_B * x_c\n x_c = tf.expand_dims(x_c, 3) # [batch, num_of_spans, emb_size, 1]\n # [batch, num_of_spans, 30, emb_size=300] mul with [batch, num_of_spans, emb_size, 1]\n x_e__x_c = tf.matmul(attention_entity_emb, x_c) # [batch, num_of_spans, 30, 1]\n x_e__x_c = tf.squeeze(x_e__x_c, axis=3) # [batch, num_of_spans, 30]\n self.attention_scores = x_e__x_c\n\n def custom_pem(self, log=True, buckets_boundaries=None, bucketing_name=\"pem_embeddings\"):\n if buckets_boundaries:\n return self._pem_bucketing_embeddings(buckets_boundaries, bucketing_name)\n elif self.args.pem_without_log:\n return self.cand_entities_scores\n else:\n return tf.log(tf.minimum(1.0, tf.maximum(self.args.zero, self.cand_entities_scores)))\n\n def _pem_bucketing_embeddings(self, buckets_boundaries, bucketing_name):\n from tensorflow.python.ops import math_ops\n bucketized_pem = math_ops._bucketize(self.cand_entities_scores, boundaries=buckets_boundaries)\n with tf.variable_scope(bucketing_name):\n _pem_embeddings = tf.get_variable(\n name=\"pem_embeddings_var\",\n dtype=tf.float32,\n shape=[len(buckets_boundaries) + 1, 1], trainable=True)\n pem_embeddings = tf.nn.embedding_lookup(_pem_embeddings, bucketized_pem, name=\"pem_embeddings_lookup\")\n # pem_embeddings = Tensor(\"pem_embeddings/pem_embeddings_lookup:0\", shape=(?, ?, ?, 1), dtype=float32)\n return tf.squeeze(pem_embeddings, axis=3)\n\n def add_cand_ent_scores_op(self):\n # now add the cand_entity_scores maybe also some extra features and through a simple ffnn\n stack_values = []\n if self.args.nn_components.find(\"lstm\") != -1:\n stack_values.append(self.similarity_scores)\n if self.args.nn_components.find(\"pem\") != -1:\n # TODO rename to pem_scores\n self.log_cand_entities_scores = self.custom_pem(\n self.args.pem_without_log, self.args.pem_buckets_boundaries)\n stack_values.append(self.log_cand_entities_scores)\n if self.args.nn_components.find(\"attention\") != -1:\n stack_values.append(self.attention_scores)\n if len(stack_values) == 1:\n # since only one scalar omit the final ffnn\n self.final_scores = stack_values[0]\n return\n scalar_predictors = tf.stack(stack_values, 3)\n #print(\"scalar_predictors = \", scalar_predictors) #[batch, num_mentions, 30, 2]\n\n with tf.variable_scope(\"similarity_and_prior_ffnn\"):\n if self.args.final_score_ffnn[0] == 0:\n self.final_scores = util.projection(scalar_predictors, 1, model=self)\n else:\n hidden_layers, hidden_size = self.args.final_score_ffnn[0], self.args.final_score_ffnn[1]\n self.final_scores = util.ffnn(scalar_predictors, hidden_layers, hidden_size, 1,\n self.dropout if self.args.ffnn_dropout else None, model=self)\n self.final_scores = tf.squeeze(self.final_scores, axis=3)\n #print(\"final_scores = \", self.final_scores)\n\n def add_global_voting_op(self):\n with tf.variable_scope(\"global_voting\"):\n self.final_scores_before_global = - (1 - self.loss_mask) * 50 + self.final_scores\n if self.args.global_topkfromallspans:\n batch_num = tf.shape(self.final_scores)[0]\n spans_num = tf.shape(self.final_scores)[1] # num of spans\n cand_ent_num = tf.shape(self.final_scores)[2] # 30\n new_size = spans_num * cand_ent_num\n temp = tf.diag(tf.ones([spans_num]))\n temp = tf.tile(tf.expand_dims(temp, axis=2), [1, 1, cand_ent_num])\n temp = tf.reshape(temp, [spans_num, new_size])\n mask = tf.reshape(tf.tile(tf.expand_dims(temp, axis=1), [1, cand_ent_num, 1]), [new_size, new_size])\n mask = 1 - mask\n\n all_entities = tf.reshape(self.pure_entity_embeddings, [batch_num, new_size, 300])\n all_scores = tf.matmul(all_entities, all_entities, transpose_b=True) # [batch, new_size, new_size]\n filtered_scores = all_scores * mask\n\n top_values, _ = tf.nn.top_k(filtered_scores, self.args.global_topkfromallspans)\n # [batch, new_size, K]\n if self.args.global_topkfromallspans_onlypositive:\n top_values = tf.maximum(top_values, self.args.zero)\n # so to avoid keeping cand ent that have score < of this value even if they are the\n self.global_voting_scores = tf.reduce_mean(top_values, axis=2) # [batch, new_size]\n self.global_voting_scores = tf.reshape(self.global_voting_scores, [batch_num, spans_num, cand_ent_num])\n else:\n if self.args.global_gmask_unambigious:\n gmask = self._sequence_mask_v13(tf.equal(self.cand_entities_len, 1), tf.shape(self.final_scores)[2])\n elif not self.args.global_topk:\n gmask = tf.to_float(((self.final_scores_before_global - self.args.global_thr) >= 0)) # [b,s,30]\n else:\n top_values, _ = tf.nn.top_k(self.final_scores_before_global, self.args.global_topk)\n # [batch, num_of_spans, K]\n K_value = top_values[:, :, -1] # [batch, num_of_spans]\n #if hasattr(self.args, 'global_topkthr'):\n if self.args.global_topkthr:\n K_value = tf.maximum(self.args.global_topkthr, K_value)\n # so to avoid keeping cand ent that have score < of this value even if they are the\n # top for this span. 30\n threshold = tf.tile(tf.expand_dims(K_value, 2), [1, 1, tf.shape(self.final_scores)[-1]])\n # [batch, num_of_spans, 30]\n gmask = tf.to_float(((self.final_scores_before_global - threshold) >= 0))\n gmask = gmask * self.loss_mask\n if self.args.global_mask_scale_each_mention_voters_to_one:\n temp = tf.reduce_sum(gmask, axis=2, keep_dims=True) # [batch, num_of_spans, 1]\n temp = tf.where(tf.less(temp, 1e-4), temp, 1. / (temp + 1e-4))\n gmask = gmask * temp\n elif self.args.global_gmask_based_on_localscore:\n gmask = gmask * tf.nn.softmax(self.final_scores_before_global)\n self.gmask = gmask\n\n masked_entity_emb = self.pure_entity_embeddings * tf.expand_dims(gmask, axis=3) # [b,s,30,300] * [b,s,30,1]\n batch_size = tf.shape(masked_entity_emb)[0]\n all_voters_emb = tf.reduce_sum(tf.reshape(masked_entity_emb, [batch_size, -1, 300]), axis=1,\n keep_dims=True) # [b, 1, 300]\n span_voters_emb = tf.reduce_sum(masked_entity_emb, axis=2) # [batch, num_of_spans, 300]\n valid_voters_emb = all_voters_emb - span_voters_emb\n # [b, 1, 300] - [batch, spans, 300] = [batch, spans, 300] (broadcasting)\n # [300] - [batch, spans, 300] = [batch, spans, 300] (broadcasting)\n if self.args.global_norm_or_mean == \"norm\":\n valid_voters_emb = tf.nn.l2_normalize(valid_voters_emb, dim=2)\n else:\n all_voters_num = tf.reduce_sum(gmask) # scalar\n span_voters_num = tf.reduce_sum(gmask, axis=2) # [batch, spans]\n valid_voters_emb = valid_voters_emb / tf.expand_dims(all_voters_num - span_voters_num, axis=2)\n\n self.global_voting_scores = tf.squeeze(tf.matmul(self.pure_entity_embeddings, tf.expand_dims(valid_voters_emb, axis=3)), axis=3)\n # [b,s,30,300] matmul [b,s,300,1] --> [b,s,30,1]-->[b,s,30]\n\n stack_values = []\n if self.args.stage2_nn_components.find(\"pem\") != -1:\n # TODO rename to pem_scores\n self.gpem_scores = self.custom_pem(\n self.args.gpem_without_log, self.args.gpem_buckets_boundaries)\n stack_values.append(self.gpem_scores)\n if self.args.stage2_nn_components.find(\"local\") != -1:\n stack_values.append(self.final_scores_before_global)\n stack_values.append(self.global_voting_scores)\n scalar_predictors = tf.stack(stack_values, 3)\n #print(\"scalar_predictors = \", scalar_predictors) #[b, s, 30, 2]\n with tf.variable_scope(\"psi_and_global_ffnn\"):\n if self.args.global_score_ffnn[0] == 0:\n self.final_scores = util.projection(scalar_predictors, 1, model=self)\n else:\n hidden_layers, hidden_size = self.args.global_score_ffnn[0], self.args.global_score_ffnn[1]\n self.final_scores = util.ffnn(scalar_predictors, hidden_layers, hidden_size, 1,\n self.dropout if self.args.ffnn_dropout else None, model=self)\n # [batch, num_mentions, 30, 1] squeeze to [batch, num_mentions, 30]\n self.final_scores = tf.squeeze(self.final_scores, axis=3)\n #print(\"final_scores = \", self.final_scores)\n\n def add_loss_op(self):\n cand_entities_labels = tf.cast(self.cand_entities_labels, tf.float32)\n loss1 = cand_entities_labels * tf.nn.relu(self.args.gamma_thr - self.final_scores)\n loss2 = (1 - cand_entities_labels) * tf.nn.relu(self.final_scores)\n self.loss = loss1 + loss2\n if self.args.nn_components.find(\"global\") != -1 and not self.args.global_one_loss:\n loss3 = cand_entities_labels * tf.nn.relu(self.args.gamma_thr - self.final_scores_before_global)\n loss4 = (1 - cand_entities_labels) * tf.nn.relu(self.final_scores_before_global)\n self.loss = loss1 + loss2 + loss3 + loss4\n #print(\"loss_mask = \", loss_mask)\n self.loss = self.loss_mask * self.loss\n self.loss = tf.reduce_sum(self.loss)\n # for tensorboard\n #tf.summary.scalar(\"loss\", self.loss)\n\n def build(self):\n self.add_placeholders()\n self.add_embeddings_op()\n if self.args.nn_components.find(\"lstm\") != -1:\n self.add_context_emb_op()\n self.add_span_emb_op()\n self.add_lstm_score_op()\n if self.args.nn_components.find(\"attention\") != -1:\n self.add_local_attention_op()\n self.add_cand_ent_scores_op()\n if self.args.nn_components.find(\"global\") != -1:\n self.add_global_voting_op()\n if self.args.running_mode.startswith(\"train\"):\n self.add_loss_op()\n # Generic functions that add training op\n self.add_train_op(self.args.lr_method, self.lr, self.loss, self.args.clip)\n self.merged_summary_op = tf.summary.merge_all()\n\n if self.args.running_mode == \"train_continue\":\n self.restore_session(\"latest\")\n elif self.args.running_mode == \"train\":\n self.initialize_session() # now self.sess is defined and vars are init\n self.init_embeddings()\n\n # if we run the evaluate.py script then we should call explicitly the model.restore(\"ed\")\n # or model.restore(\"el\"). here it doesn't initialize or restore values for the evaluate.py\n # case.\n\n def _sequence_mask_v13(self, mytensor, max_width):\n \"\"\"mytensor is a 2d tensor\"\"\"\n if not tf.__version__.startswith(\"1.4\"):\n temp_shape = tf.shape(mytensor)\n temp = tf.sequence_mask(tf.reshape(mytensor, [-1]), max_width, dtype=tf.float32)\n temp_mask = tf.reshape(temp, [temp_shape[0], temp_shape[1], max_width]) # tf.shape(temp)[-1]])\n else:\n temp_mask = tf.sequence_mask(mytensor, max_width, dtype=tf.float32)\n return temp_mask\n\n\n" }, { "path": "code/model/reader.py", "content": "import tensorflow as tf\nimport argparse\nimport model.config as config\nimport pickle\n\ndef parse_sequence_example(serialized):\n sequence_features={\n \"words\": tf.FixedLenSequenceFeature([], dtype=tf.int64), # in order to have a vector. if i put [1] it will probably\n # be a matrix with just one column\n \"chars\": tf.VarLenFeature(tf.int64),\n \"chars_len\": tf.FixedLenSequenceFeature([], dtype=tf.int64),\n \"begin_span\": tf.FixedLenSequenceFeature([], dtype=tf.int64),\n \"end_span\": tf.FixedLenSequenceFeature([], dtype=tf.int64),\n \"cand_entities\": tf.VarLenFeature(tf.int64),\n \"cand_entities_scores\": tf.VarLenFeature(tf.float32),\n \"cand_entities_labels\": tf.VarLenFeature(tf.int64),\n \"cand_entities_len\": tf.FixedLenSequenceFeature([], dtype=tf.int64),\n \"ground_truth\": tf.FixedLenSequenceFeature([], dtype=tf.int64)\n }\n if True:\n sequence_features[\"begin_gm\"] = tf.FixedLenSequenceFeature([], dtype=tf.int64)\n sequence_features[\"end_gm\"] = tf.FixedLenSequenceFeature([], dtype=tf.int64)\n\n context, sequence = tf.parse_single_sequence_example(\n serialized,\n context_features={\n \"chunk_id\": tf.FixedLenFeature([], dtype=tf.string),\n \"words_len\": tf.FixedLenFeature([], dtype=tf.int64),\n \"spans_len\": tf.FixedLenFeature([], dtype=tf.int64),\n \"ground_truth_len\": tf.FixedLenFeature([], dtype=tf.int64)\n },\n sequence_features=sequence_features)\n\n return context[\"chunk_id\"], sequence[\"words\"], context[\"words_len\"],\\\n tf.sparse_tensor_to_dense(sequence[\"chars\"]), sequence[\"chars_len\"],\\\n sequence[\"begin_span\"], sequence[\"end_span\"], context[\"spans_len\"],\\\n tf.sparse_tensor_to_dense(sequence[\"cand_entities\"]),\\\n tf.sparse_tensor_to_dense(sequence[\"cand_entities_scores\"]),\\\n tf.sparse_tensor_to_dense(sequence[\"cand_entities_labels\"]),\\\n sequence[\"cand_entities_len\"],\\\n sequence[\"ground_truth\"], context[\"ground_truth_len\"],\\\n sequence[\"begin_gm\"], sequence[\"end_gm\"]\n #return context, sequence\n\n\ndef count_records_of_one_epoch(trainfiles):\n filename_queue = tf.train.string_input_producer(trainfiles, num_epochs=1)\n reader = tf.TFRecordReader()\n key, serialized_example = reader.read(filename_queue)\n with tf.Session() as sess:\n sess.run(\n tf.variables_initializer(\n tf.global_variables() + tf.local_variables()\n )\n )\n\n # Start queue runners\n coord = tf.train.Coordinator()\n threads = tf.train.start_queue_runners(sess=sess, coord=coord)\n\n counter = 0\n try:\n while not coord.should_stop():\n fetch_vals = sess.run((key))\n #print(fetch_vals)\n counter += 1\n except tf.errors.OutOfRangeError:\n pass\n except KeyboardInterrupt:\n print(\"Training stopped by Ctrl+C.\")\n finally:\n coord.request_stop()\n coord.join(threads)\n print(\"number of tfrecords in trainfiles = \", counter)\n return counter\n\n\ndef train_input_pipeline(filenames, args):\n dataset = tf.data.TFRecordDataset(filenames)\n #dataset = tf.contrib.data.TFRecordDataset(filenames)\n dataset = dataset.map(parse_sequence_example)\n #dataset = dataset.map(parse_sequence_example, num_parallel_calls=3)\n dataset = dataset.repeat()\n dataset = dataset.shuffle(buffer_size=args.shuffle_capacity)\n dataset = dataset.padded_batch(args.batch_size, dataset.output_shapes)\n return dataset\n\n\ndef test_input_pipeline(filenames, args):\n dataset = tf.data.TFRecordDataset(filenames)\n dataset = dataset.map(parse_sequence_example)\n dataset = dataset.padded_batch(args.batch_size, dataset.output_shapes)\n return dataset\n\nif __name__ == \"__main__\":\n count_records_of_one_epoch([\"/home/master_thesis_share/data/tfrecords/\"\n \"wikiRLTD_perparagr_wthr_6_cthr_201/\"\n \"train/wikidumpRLTD.txt\"])\n\n\n" }, { "path": "code/model/train.py", "content": "import argparse\nimport model.reader as reader\nimport model.config as config\nimport os\nimport tensorflow as tf\nfrom evaluation.metrics import Evaluator, metrics_calculation, threshold_calculation\nimport time\nimport pickle\nimport numpy as np\nfrom model.util import load_train_args\n\ndef create_training_pipelines(args):\n folder = \"../data/tfrecords/\" + args.experiment_name + \\\n (\"/allspans/\" if args.all_spans_training else \"/gmonly/\")\n training_dataset = reader.train_input_pipeline([folder + file for file in args.train_datasets], args)\n return training_dataset\n\n\ndef create_el_ed_pipelines(gmonly_flag, filenames, args):\n if filenames is None:\n return [], []\n\n folder = config.base_folder+\"data/tfrecords/\" + args.experiment_name + (\"/gmonly/\" if gmonly_flag else \"/allspans/\")\n test_datasets = []\n for file in filenames:\n test_datasets.append(reader.test_input_pipeline([folder+file], args))\n\n return test_datasets, filenames\n\n\ndef tensorboard_writers(graph):\n tf_writers = dict()\n tf_writers[\"train\"] = tf.summary.FileWriter(args.summaries_folder + 'train/', graph)\n tf_writers[\"ed_pr\"] = tf.summary.FileWriter(args.summaries_folder + 'ed_pr/')\n tf_writers[\"ed_re\"] = tf.summary.FileWriter(args.summaries_folder + 'ed_re/')\n tf_writers[\"ed_f1\"] = tf.summary.FileWriter(args.summaries_folder + 'ed_f1/')\n #tf_writers[\"ed_pr@1\"] = tf.summary.FileWriter(args.summaries_folder + 'ed_pr@1/')\n\n tf_writers[\"el_pr\"] = tf.summary.FileWriter(args.summaries_folder + 'el_pr/')\n tf_writers[\"el_re\"] = tf.summary.FileWriter(args.summaries_folder + 'el_re/')\n tf_writers[\"el_f1\"] = tf.summary.FileWriter(args.summaries_folder + 'el_f1/')\n return tf_writers\n\n\ndef validation_loss_calculation(model, iterator, dataset_handle, opt_thr, el_mode, name=\"\"):\n # name is the name of the dataset e.g. aida_test.txt, aquaint.txt\n # Run one pass over the validation dataset.\n model.sess.run(iterator.initializer)\n evaluator = Evaluator(opt_thr, name=name)\n while True:\n try:\n retrieve_l = [model.final_scores, model.cand_entities_len, model.cand_entities,\n model.begin_span, model.end_span, model.spans_len,\n model.begin_gm, model.end_gm,\n model.ground_truth, model.ground_truth_len,\n model.words_len, model.chunk_id]\n result_l = model.sess.run(\n retrieve_l, feed_dict={model.input_handle_ph: dataset_handle, model.dropout: 1})\n metrics_calculation(evaluator, *result_l, el_mode)\n\n except tf.errors.OutOfRangeError:\n print(name)\n micro_f1, macro_f1 = evaluator.print_log_results(model.tf_writers, args.eval_cnt, el_mode)\n break\n return micro_f1, macro_f1\n\n\n\ndef optimal_thr_calc(model, handles, iterators, el_mode):\n val_datasets = args.el_val_datasets if el_mode else args.ed_val_datasets\n tp_fp_scores_labels = []\n fn_scores = []\n for val_dataset in val_datasets: # 1, 4\n dataset_handle = handles[val_dataset]\n iterator = iterators[val_dataset]\n model.sess.run(iterator.initializer)\n while True:\n try:\n retrieve_l = [model.final_scores, model.cand_entities_len, model.cand_entities,\n model.begin_span, model.end_span, model.spans_len,\n model.begin_gm, model.end_gm,\n model.ground_truth, model.ground_truth_len,\n model.words_len, model.chunk_id]\n result_l = model.sess.run(\n retrieve_l, feed_dict={model.input_handle_ph: dataset_handle, model.dropout: 1})\n tp_fp_batch, fn_batch = threshold_calculation(*result_l, el_mode)\n tp_fp_scores_labels.extend(tp_fp_batch)\n fn_scores.extend(fn_batch)\n except tf.errors.OutOfRangeError:\n break\n return optimal_thr_calc_aux(tp_fp_scores_labels, fn_scores)\n\n\ndef optimal_thr_calc_aux(tp_fp_scores_labels, fn_scores):\n # based on tp_fp_scores and fn_scores calculate optimal threshold\n tp_fp_scores_labels = sorted(tp_fp_scores_labels) # low --> high\n fn_scores = sorted(fn_scores)\n tp, fp = 0, 0\n fn_idx = len(fn_scores) # from [0, fn_idx-1] is fn. [fn_idx, len(fn_scores)) isn't.\n # initially i start with a very high threshold which means I reject everything, hence tp, fp =0\n # and all the gold mentions are fn. so fn = len(fn_scores)\n best_thr = tp_fp_scores_labels[-1][0]+1 # the highest (rightmost) possible threshold + 1 (so everything is rejected)\n best_f1 = -1\n # whatever is on the right or at the position we point is included in the tp, fp\n # whatever is on the left remains to be processed-examined\n tp_fp_idx = len(tp_fp_scores_labels) # similar to fn_idx\n while tp_fp_idx > 0: # if we point to 0 then nothing on the left to examine (smaller thresholds)\n # from right to left loop\n tp_fp_idx -= 1\n new_thr, label = tp_fp_scores_labels[tp_fp_idx]\n tp += label\n fp += (1 - label)\n while tp_fp_idx > 0 and tp_fp_scores_labels[tp_fp_idx-1][0] == new_thr:\n tp_fp_idx -= 1\n new_thr, label = tp_fp_scores_labels[tp_fp_idx]\n tp += label\n fp += (1 - label)\n\n while fn_idx > 0 and fn_scores[fn_idx-1] >= new_thr: # move left one position\n fn_idx -= 1\n assert( 0 <= tp <= len(tp_fp_scores_labels) and\n 0 <= fp <= len(tp_fp_scores_labels) and\n 0 <= fn_idx <= len(fn_scores))\n precision = 100 * tp / (tp + fp + 1e-6)\n recall = 100 * tp / (tp + fn_idx + 1e-6)\n f1 = 2 * precision * recall / (precision + recall + 1e-6)\n assert(0 <= precision <= 100 and 0 <= recall <= 100 and 0 <= f1 <= 100)\n if f1 > best_f1:\n best_f1 = f1\n best_thr = new_thr\n\n print('Best validation threshold = %.3f with F1=%.1f ' % (best_thr, best_f1))\n return best_thr, best_f1\n\n\ndef compute_ed_el_scores(model, handles, names, iterators, el_mode):\n # first compute the optimal threshold based on validation datasets.\n if args.hardcoded_thr:\n opt_thr = args.hardcoded_thr\n else:\n opt_thr, val_f1 = optimal_thr_calc(model, handles, iterators, el_mode)\n\n micro_results = []\n macro_results = []\n for test_handle, test_name, test_it in zip(handles, names, iterators):\n micro_f1, macro_f1 = validation_loss_calculation(model, test_it, test_handle, opt_thr,\n el_mode=el_mode, name=test_name)\n micro_results.append(micro_f1)\n macro_results.append(macro_f1)\n\n val_datasets = args.el_val_datasets if el_mode else args.ed_val_datasets\n if not args.hardcoded_thr and len(val_datasets) == 1 and abs(micro_results[val_datasets[0]] - val_f1) > 0.1:\n print(\"ASSERTION ERROR: optimal threshold f1 calculalation differs from normal\"\n \"f1 calculation!!!!\", val_f1, \" and \", micro_results[val_datasets[0]])\n return micro_results\n\n\ndef train():\n training_dataset = create_training_pipelines(args)\n\n ed_datasets, ed_names = create_el_ed_pipelines(gmonly_flag=True, filenames=args.ed_datasets, args=args)\n el_datasets, el_names = create_el_ed_pipelines(gmonly_flag=False, filenames=args.el_datasets, args=args)\n\n input_handle_ph = tf.placeholder(tf.string, shape=[], name=\"input_handle_ph\")\n iterator = tf.contrib.data.Iterator.from_string_handle(\n input_handle_ph, training_dataset.output_types, training_dataset.output_shapes)\n next_element = iterator.get_next()\n #print(next_element)\n\n if args.ablations:\n from model.model_ablations import Model\n else:\n from model.model import Model\n model = Model(args, next_element)\n model.build()\n model.input_handle_ph = input_handle_ph # just for convenience so i can access it from everywhere\n #print(tf.global_variables())\n\n tf_writers = tensorboard_writers(model.sess.graph)\n model.tf_writers = tf_writers # for accessing convenience\n\n # The `Iterator.string_handle()` method returns a tensor that can be evaluated\n # and used to feed the `handle` placeholder.\n with model.sess as sess:\n\n def ed_el_dataset_handles(datasets):\n test_iterators = []\n test_handles = []\n for dataset in datasets:\n test_iterator = dataset.make_initializable_iterator()\n test_iterators.append(test_iterator)\n test_handles.append(sess.run(test_iterator.string_handle()))\n return test_iterators, test_handles\n\n training_iterator = training_dataset.make_one_shot_iterator()\n training_handle = sess.run(training_iterator.string_handle())\n\n ed_iterators, ed_handles = ed_el_dataset_handles(ed_datasets)\n el_iterators, el_handles = ed_el_dataset_handles(el_datasets)\n\n\n # Loop forever, alternating between training and validation.\n best_ed_score = 0\n best_el_score = 0\n termination_ed_score = 0\n termination_el_score = 0\n nepoch_no_imprv = 0 # for early stopping\n train_step = 0\n while True:\n total_train_loss = 0\n # for _ in range(args.steps_before_evaluation): # for training based on training steps\n wall_start = time.time()\n while ( (time.time() - wall_start) / 60 ) <= args.evaluation_minutes:\n train_step += 1\n if args.ffnn_l2maxnorm:\n sess.run(model.ffnn_l2normalization_op_list)\n _, loss = sess.run([model.train_op, model.loss], feed_dict={input_handle_ph: training_handle, model.dropout: args.dropout,\n model.lr: model.args.lr})\n total_train_loss += loss\n\n args.eval_cnt += 1\n summary = tf.Summary(value=[tf.Summary.Value(tag=\"total_train_loss\", simple_value=total_train_loss)])\n tf_writers[\"train\"].add_summary(summary, args.eval_cnt)\n\n print(\"args.eval_cnt = \", args.eval_cnt)\n summary = sess.run(model.merged_summary_op)\n tf_writers[\"train\"].add_summary(summary, args.eval_cnt)\n\n wall_start = time.time()\n comparison_ed_score = comparison_el_score = -0.1\n if ed_names:\n print(\"Evaluating ED datasets\")\n ed_scores = compute_ed_el_scores(model, ed_handles, ed_names, ed_iterators, el_mode=False)\n comparison_ed_score = np.mean(np.array(ed_scores)[args.ed_val_datasets])\n if el_names:\n print(\"Evaluating EL datasets\")\n el_scores = compute_ed_el_scores(model, el_handles, el_names, el_iterators, el_mode=True)\n comparison_el_score = np.mean(np.array(el_scores)[args.el_val_datasets])\n print(\"Evaluation duration in minutes: \", (time.time() - wall_start) / 60)\n\n #comparison_ed_score = (ed_scores[1] + ed_scores[4]) / 2 # aida_dev + acquaint\n #comparison_score = ed_scores[1] # aida_dev\n if model.args.lr_decay > 0:\n model.args.lr *= model.args.lr_decay # decay learning rate\n text = \"\"\n best_ed_flag = False\n best_el_flag = False\n # otherwise not significant improvement 75.2 to 75.3 micro_f1 of aida_dev\n if comparison_ed_score >= best_ed_score + 0.1: # args.improvement_threshold:\n text = \"- new best ED score!\" + \" prev_best= \" + str(best_ed_score) +\\\n \" new_best= \" + str(comparison_ed_score)\n best_ed_flag = True\n best_ed_score = comparison_ed_score\n if comparison_el_score >= best_el_score + 0.1: #args.improvement_threshold:\n text += \"- new best EL score!\" + \" prev_best= \" + str(best_el_score) +\\\n \" new_best= \" + str(comparison_el_score)\n best_el_flag = True\n best_el_score = comparison_el_score\n if best_ed_flag or best_el_flag: # keep checkpoint\n print(text)\n if args.nocheckpoints is False:\n model.save_session(args.eval_cnt, best_ed_flag, best_el_flag)\n # check for termination now.\n if comparison_ed_score >= termination_ed_score + args.improvement_threshold\\\n or comparison_el_score >= termination_el_score + args.improvement_threshold:\n print(\"significant improvement. reset termination counter\")\n termination_ed_score = comparison_ed_score\n termination_el_score = comparison_el_score\n nepoch_no_imprv = 0\n else:\n nepoch_no_imprv += 1\n if nepoch_no_imprv >= args.nepoch_no_imprv:\n print(\"- early stopping {} epochs without \"\n \"improvement\".format(nepoch_no_imprv))\n terminate()\n break\n\n\ndef _parse_args():\n parser = argparse.ArgumentParser()\n parser.add_argument(\"--experiment_name\", default=\"alldatasets_perparagr\", #\"standard\",\n help=\"under folder data/tfrecords/\")\n parser.add_argument(\"--training_name\", default=None,\n help=\"under folder data/tfrecords/\")\n parser.add_argument(\"--shuffle_capacity\", type=int, default=500)\n parser.add_argument(\"--debug\", type=bool, default=False)\n\n parser.add_argument(\"--nepoch_no_imprv\", type=int, default=5)\n parser.add_argument(\"--improvement_threshold\", type=float, default=0.3, help=\"if improvement less than this then\"\n \"it is considered not significant and we have early stopping.\")\n parser.add_argument(\"--clip\", type=int, default=-1, help=\"if negative then no clipping\")\n parser.add_argument(\"--lr_decay\", type=float, default=-1.0, help=\"if negative then no decay\")\n parser.add_argument(\"--lr\", type=float, default=0.001)\n parser.add_argument(\"--lr_method\", default=\"adam\")\n parser.add_argument(\"--batch_size\", type=int, default=10)\n parser.add_argument(\"--dropout\", type=float, default=0.5)\n parser.add_argument(\"--train_ent_vecs\", dest='train_ent_vecs', action='store_true')\n parser.add_argument(\"--no_train_ent_vecs\", dest='train_ent_vecs', action='store_false')\n parser.set_defaults(train_ent_vecs=False)\n\n parser.add_argument(\"--steps_before_evaluation\", type=int, default=10000)\n parser.add_argument(\"--evaluation_minutes\", type=int, default=15, help=\"every this number of minutes pause\"\n \" training and run an evaluation epoch\")\n parser.add_argument(\"--dim_char\", type=int, default=100)\n parser.add_argument(\"--hidden_size_char\", type=int, default=100, help=\"lstm on chars\")\n parser.add_argument(\"--hidden_size_lstm\", type=int, default=300, help=\"lstm on word embeddings\")\n\n parser.add_argument(\"--use_chars\", dest=\"use_chars\", action='store_true', help=\"use character embeddings or not\")\n parser.add_argument(\"--no_use_chars\", dest=\"use_chars\", action='store_false')\n parser.set_defaults(use_chars=True)\n parser.add_argument(\"--model_heads_from_bilstm\", type=bool, default=False,\n help=\"use the bilstm vectors for the head instead of the word embeddings\")\n parser.add_argument(\"--span_boundaries_from_wordemb\", type=bool, default=False, help=\"instead of using the \"\n \"output of contextual bilstm for start and end of span we use word+char emb\")\n parser.add_argument(\"--span_emb\", default=\"boundaries_head\", help=\"boundaries for start and end, and head\")\n\n\n parser.add_argument(\"--max_mention_width\", type=int, default=10)\n parser.add_argument(\"--use_features\", type=bool, default=False, help=\"like mention width\")\n parser.add_argument(\"--feature_size\", type=int, default=20) # each width is represented by a vector of that size\n\n\n parser.add_argument(\"--ent_vecs_regularization\", default=\"l2dropout\", help=\"'no', \"\n \"'dropout', 'l2', 'l2dropout'\")\n\n parser.add_argument(\"--span_emb_ffnn\", default=\"0_0\", help=\"int_int the first int\"\n \"indicates the number of hidden layers and the second the hidden size\"\n \"so 2_100 means 2 hidden layers of width 100 and then projection to output size\"\n \". 0_0 means just projecting without hidden layers\")\n parser.add_argument(\"--final_score_ffnn\", default=\"1_100\", help=\"int_int look span_emb_ffnn\")\n\n\n parser.add_argument(\"--gamma_thr\", type=float, default=0.2)\n\n parser.add_argument(\"--nocheckpoints\", type=bool, default=False)\n parser.add_argument(\"--checkpoints_num\", type=int, default=1, help=\"maximum number of checkpoints to keep\")\n\n parser.add_argument(\"--ed_datasets\", default=\"\")\n parser.add_argument(\"--ed_val_datasets\", default=\"1\", help=\"based on these datasets pick the optimal\"\n \"gamma thr and also consider early stopping\")\n #--ed_val_datasets=1_4 # aida_dev, aquaint\n parser.add_argument(\"--el_datasets\", default=\"\")\n parser.add_argument(\"--el_val_datasets\", default=\"1\") #--el_val_datasets=1_4 # aida_dev, aquaint\n\n parser.add_argument(\"--train_datasets\", default=\"aida_train.txt\")\n #--train_datasets=aida_train.txt_z_wikidumpRLTD.txt\n\n parser.add_argument(\"--continue_training\", type=bool, default=False,\n help=\"if true then just restore the previous command line\"\n \"arguments and continue the training in exactly the\"\n \"same way. so only the experiment_name and \"\n \"training_name are used from here. Retrieve values from\"\n \"latest checkpoint.\")\n parser.add_argument(\"--onleohnard\", type=bool, default=False)\n\n parser.add_argument(\"--comment\", default=\"\", help=\"put any comment here that describes your experiment\"\n \", for logging purposes only.\")\n\n parser.add_argument(\"--all_spans_training\", type=bool, default=False)\n parser.add_argument(\"--fast_evaluation\", type=bool, default=False, help=\"if all_spans training then evaluate only\"\n \"on el tests, corresponding if gm training evaluate only on ed tests.\")\n\n\n parser.add_argument(\"--entity_extension\", default=None, help=\"extension_entities or extension_entities_all etc\")\n\n parser.add_argument(\"--nn_components\", default=\"pem_lstm\", help=\"each option is one scalar, then these are fed to\"\n \"the final ffnn and we have the final score. choose any combination you want: e.g\"\n \"pem_lstm_attention_global, pem_attention, lstm_attention, pem_lstm_global, etc\")\n parser.add_argument(\"--attention_K\", type=int, default=100, help=\"K from left and K from right, in total 2K\")\n parser.add_argument(\"--attention_R\", type=int, default=30, help=\"hard attention\")\n parser.add_argument(\"--attention_use_AB\", type=bool, default=False)\n parser.add_argument(\"--attention_on_lstm\", type=bool, default=False, help=\"instead of using attention on\"\n \"original pretrained word embedding. use it on vectors or lstm, \"\n \"needs also projection now the context vector x_c to 300 dimensions\")\n parser.add_argument(\"--attention_ent_vecs_no_regularization\", type=bool, default=False)\n parser.add_argument(\"--attention_retricted_num_of_entities\", type=int, default=None,\n help=\"instead of using 30 entities for creating the context vector we use only\"\n \"the top x number of entities for reducing noise.\")\n parser.add_argument(\"--global_thr\", type=float, default=0.1) # 0.0, 0.05, -0.05, 0.2\n parser.add_argument(\"--global_mask_scale_each_mention_voters_to_one\", type=bool, default=False)\n parser.add_argument(\"--global_topk\", type=int, default=None)\n parser.add_argument(\"--global_gmask_based_on_localscore\", type=bool, default=False) # new\n parser.add_argument(\"--global_topkthr\", type=float, default=None) # 0.0, 0.05, -0.05, 0.2\n parser.add_argument(\"--global_score_ffnn\", default=\"1_100\", help=\"int_int look span_emb_ffnn\")\n parser.add_argument(\"--global_one_loss\", type=bool, default=False)\n parser.add_argument(\"--global_norm_or_mean\", default=\"norm\")\n parser.add_argument(\"--global_topkfromallspans\", type=int, default=None)\n parser.add_argument(\"--global_topkfromallspans_onlypositive\", type=bool, default=False)\n parser.add_argument(\"--global_gmask_unambigious\", type=bool, default=False)\n\n parser.add_argument(\"--hardcoded_thr\", type=float, default=None, help=\"if this is specified then we don't calculate\"\n \"optimal threshold based on the dev dataset but use this one.\")\n parser.add_argument(\"--ffnn_dropout\", dest=\"ffnn_dropout\", action='store_true')\n parser.add_argument(\"--no_ffnn_dropout\", dest=\"ffnn_dropout\", action='store_false')\n parser.set_defaults(ffnn_dropout=True)\n parser.add_argument(\"--ffnn_l2maxnorm\", type=float, default=None, help=\"if positive\"\n \" then bound the Frobenius norm <= value for the weight tensor of the \"\n \"hidden layers and the output layer of the FFNNs\")\n parser.add_argument(\"--ffnn_l2maxnorm_onlyhiddenlayers\", type=bool, default=False)\n\n parser.add_argument(\"--cand_ent_num_restriction\", type=int, default=None, help=\"for reducing memory usage and\"\n \"avoiding OOM errors in big NN I can reduce the number of candidate ent for each span\")\n # --ed_datasets= --el_datasets=\"aida_train.txt_z_aida_dev.txt\" which means i can leave something empty\n # and i can also put \"\" in the cla\n\n parser.add_argument(\"--no_p_e_m_usage\", type=bool, default=False, help=\"use similarity score instead of \"\n \"final score for prediction\")\n parser.add_argument(\"--pem_without_log\", type=bool, default=False)\n parser.add_argument(\"--pem_buckets_boundaries\", default=None,\n help=\"example: 0.03_0.1_0.2_0.3_0.4_0.5_0.6_0.7_0.8_0.9_0.99\")\n # the following two command line arguments\n parser.add_argument(\"--gpem_without_log\", type=bool, default=False)\n parser.add_argument(\"--gpem_buckets_boundaries\", default=None,\n help=\"example: 0.03_0.1_0.2_0.3_0.4_0.5_0.6_0.7_0.8_0.9_0.99\")\n parser.add_argument(\"--stage2_nn_components\", default=\"local_global\", help=\"each option is one scalar, then these are fed to\"\n \"the final ffnn and we have the final score. choose any combination you want: e.g\"\n \"pem_local_global, pem_global, local_global, global, etc\")\n parser.add_argument(\"--ablations\", type=bool, default=False)\n args = parser.parse_args()\n\n if args.training_name is None:\n from datetime import datetime\n args.training_name = \"{:%d_%m_%Y____%H_%M}\".format(datetime.now())\n\n temp = \"all_spans_\" if args.all_spans_training else \"\"\n args.output_folder = config.base_folder+\"data/tfrecords/\" + \\\n args.experiment_name+\"/{}training_folder/\".format(temp)+\\\n args.training_name+\"/\"\n\n if args.continue_training:\n print(\"continue training...\")\n train_args = load_train_args(args.output_folder, \"train_continue\")\n return train_args\n args.running_mode = \"train\" # \"evaluate\" \"ensemble_eval\" \"gerbil\"\n\n if os.path.exists(args.output_folder) and not args.continue_training:\n print(\"!!!!!!!!!!!!!!\\n\"\n \"experiment: \", args.output_folder, \"already exists and args.continue_training=False.\"\n \"folder will be deleted in 20 seconds. Press CTRL+C to prevent it.\")\n time.sleep(20)\n import shutil\n shutil.rmtree(args.output_folder)\n\n if not os.path.exists(args.output_folder):\n os.makedirs(args.output_folder)\n\n args.checkpoints_folder = args.output_folder + \"checkpoints/\"\n if args.onleohnard:\n args.checkpoints_folder = \"/cluster/home/nkolitsa/checkpoints_folder/\"+\\\n args.experiment_name + \"/\" + args.training_name + \"/\"\n\n args.summaries_folder = args.output_folder + \"summaries/\"\n if not os.path.exists(args.summaries_folder):\n os.makedirs(args.summaries_folder)\n\n args.ed_datasets = args.ed_datasets.split('_z_') if args.ed_datasets != \"\" else None\n args.el_datasets = args.el_datasets.split('_z_') if args.el_datasets != \"\" else None\n args.train_datasets = args.train_datasets.split('_z_') if args.train_datasets != \"\" else None\n\n args.ed_val_datasets = [int(x) for x in args.ed_val_datasets.split('_')]\n args.el_val_datasets = [int(x) for x in args.el_val_datasets.split('_')]\n\n args.span_emb_ffnn = [int(x) for x in args.span_emb_ffnn.split('_')]\n args.final_score_ffnn = [int(x) for x in args.final_score_ffnn.split('_')]\n args.global_score_ffnn = [int(x) for x in args.global_score_ffnn.split('_')]\n\n args.eval_cnt = 0\n args.zero = 1e-6\n\n if args.pem_buckets_boundaries:\n args.pem_buckets_boundaries = [float(x) for x in args.pem_buckets_boundaries.split('_')]\n if args.gpem_buckets_boundaries:\n args.gpem_buckets_boundaries = [float(x) for x in args.gpem_buckets_boundaries.split('_')]\n\n if args.fast_evaluation:\n if args.all_spans_training: # destined for el so omit the evaluation on ed\n args.ed_datasets = None\n else:\n args.el_datasets = None\n return args\n\n\ndef log_args(filepath):\n with open(filepath, \"w\") as fout:\n attrs = vars(args)\n # {'kids': 0, 'name': 'Dog', 'color': 'Spotted', 'age': 10, 'legs': 2, 'smell': 'Alot'}\n fout.write('\\n'.join(\"%s: %s\" % item for item in attrs.items()))\n\n with open(args.output_folder+\"train_args.pickle\", 'wb') as handle:\n pickle.dump(args, handle)\n\n\ndef terminate():\n tee.close()\n with open(args.output_folder+\"train_args.pickle\", 'wb') as handle:\n pickle.dump(args, handle)\n\n\nif __name__ == \"__main__\":\n args = _parse_args()\n print(args)\n log_args(args.output_folder+\"train_args.txt\")\n from model.util import Tee\n tee = Tee(args.output_folder+'log.txt', 'a')\n try:\n train()\n except KeyboardInterrupt:\n terminate()\n\n\n" }, { "path": "code/model/util.py", "content": "import tensorflow as tf\nimport numpy as np\nimport model.config as config\n\ndef projection(inputs, output_size, initializer=None, model=None):\n return ffnn(inputs, 0, -1, output_size, dropout=None, output_weights_initializer=initializer, model=model)\n\ndef shape(x, dim):\n return x.get_shape()[dim].value or tf.shape(x)[dim]\n\ndef ffnn(inputs, num_hidden_layers, hidden_size, output_size, dropout, output_weights_initializer=None, model=None):\n l2maxnorm = model.args.ffnn_l2maxnorm if model else None\n if l2maxnorm:\n dropout = None # either dropout or l2normalization not both\n\n if len(inputs.get_shape()) > 2:\n # from [batch, max_sentence_length, emb i.e.500] to [batch*max_sentence_length, emb]\n current_inputs = tf.reshape(inputs, [-1, shape(inputs, -1)])\n else:\n current_inputs = inputs\n\n for i in range(num_hidden_layers):\n hidden_weights = tf.get_variable(\"hidden_weights_{}\".format(i), [shape(current_inputs, 1), hidden_size])\n hidden_bias = tf.get_variable(\"hidden_bias_{}\".format(i), [hidden_size])\n variable_summaries(hidden_weights)\n variable_summaries(hidden_bias)\n if l2maxnorm:\n temp = hidden_weights / tf.maximum(tf.norm(hidden_weights)/l2maxnorm, 1)\n model.ffnn_l2normalization_op_list.append(hidden_weights.assign(temp))\n temp = hidden_bias / tf.maximum(tf.norm(hidden_bias)/l2maxnorm, 1)\n model.ffnn_l2normalization_op_list.append(hidden_bias.assign(temp))\n\n current_outputs = tf.nn.relu(tf.matmul(current_inputs, hidden_weights) + hidden_bias)\n\n if dropout is not None:\n current_outputs = tf.nn.dropout(current_outputs, dropout)\n current_inputs = current_outputs\n\n # 500 1\n output_weights = tf.get_variable(\"output_weights\", [shape(current_inputs, 1), output_size], initializer=output_weights_initializer)\n output_bias = tf.get_variable(\"output_bias\", [output_size])\n variable_summaries(output_weights)\n variable_summaries(output_bias)\n if l2maxnorm and not model.args.ffnn_l2maxnorm_onlyhiddenlayers:\n temp = output_weights / tf.maximum(tf.norm(output_weights)/l2maxnorm, 1)\n model.ffnn_l2normalization_op_list.append(output_weights.assign(temp))\n temp = output_bias / tf.maximum(tf.norm(output_bias)/l2maxnorm, 1)\n model.ffnn_l2normalization_op_list.append(output_bias.assign(temp))\n\n outputs = tf.matmul(current_inputs, output_weights) + output_bias\n #print(\"model/util variable name = \", output_weights.name, output_bias.name)\n\n if len(inputs.get_shape()) == 3:\n outputs = tf.reshape(outputs, [shape(inputs, 0), shape(inputs, 1), output_size])\n elif len(inputs.get_shape()) == 4:\n outputs = tf.reshape(outputs, [shape(inputs, 0), shape(inputs, 1), shape(inputs, 2), output_size])\n elif len(inputs.get_shape()) > 4:\n raise ValueError(\"FFNN with rank {} not supported\".format(len(inputs.get_shape())))\n return outputs\n\n\n\ndef variable_summaries(var):\n \"\"\"Attach a lot of summaries to a Tensor (for TensorBoard visualization).\"\"\"\n name = \"_\" + var.name.split(\"/\")[-1].split(\":\")[0]\n #print(\"name of summary\", name)\n with tf.name_scope('summaries'+name):\n mean = tf.reduce_mean(var)\n tf.summary.scalar('mean', mean)\n with tf.name_scope('stddev'):\n stddev = tf.sqrt(tf.reduce_mean(tf.square(var - mean)))\n tf.summary.scalar('stddev', stddev)\n tf.summary.scalar('max', tf.reduce_max(var))\n tf.summary.scalar('min', tf.reduce_min(var))\n\n\nimport sys\n# https://stackoverflow.com/questions/616645/how-do-i-duplicate-sys-stdout-to-a-log-file-in-python\n# http://web.archive.org/web/20141016185743/https://mail.python.org/pipermail/python-list/2007-May/460639.html\nclass Tee(object):\n\n def __init__(self, name, mode):\n self.file = open(name, mode)\n self.stdout = sys.stdout\n sys.stdout = self\n\n def close(self):\n if self.stdout is not None:\n sys.stdout = self.stdout\n self.stdout = None\n if self.file is not None:\n self.file.close()\n self.file = None\n\n def write(self, data):\n self.file.write(data)\n self.stdout.write(data)\n\n def flush(self):\n self.file.flush()\n self.stdout.flush()\n\n def __del__(self):\n self.close()\n\n\ndef _correct_train_args_leohnard_dalab(train_args, model_folder):\n \"\"\"if the experiment is executed on leohnard then the path starts with\n /cluster/ whereas on dalabgpu startswith /local/ also folders may have\n been moved to a different position so I should correct the paths.\"\"\"\n if train_args.output_folder != model_folder:\n train_args.output_folder = model_folder\n train_args.checkpoints_folder = model_folder + \"checkpoints/\"\n train_args.summaries_folder = model_folder + \"summaries/\"\n train_args.inconsistent_model_folder = True\n\nimport pickle\n\ndef load_train_args(output_folder, running_mode):\n \"\"\"running_mode: train, train_continue, evaluate, ensemble_eval, gerbil\"\"\"\n with open(output_folder+\"train_args.pickle\", 'rb') as handle:\n train_args = pickle.load(handle)\n _correct_train_args_leohnard_dalab(train_args, output_folder)\n # update checkpoint train_args to be compatible with new arguments added in code\n train_args.running_mode = running_mode\n if not hasattr(train_args, 'nn_components') and hasattr(train_args, 'pem_lstm_attention'):\n train_args.nn_components = train_args.pem_lstm_attention\n if not hasattr(train_args, 'nn_components'):\n train_args.nn_components = \"pem_lstm\"\n if not hasattr(train_args, 'model_heads_from_bilstm'):\n train_args.model_heads_from_bilstm = False\n if not hasattr(train_args, 'zero'):\n train_args.zero = 1e-3 # for compatibility with old experiments\n if not hasattr(train_args, 'attention_use_AB'):\n train_args.attention_use_AB = False\n if not hasattr(train_args, 'attention_on_lstm'):\n train_args.attention_on_lstm = False\n if not hasattr(train_args, 'span_emb'):\n train_args.span_emb = \"boundaries\"\n if train_args.model_heads:\n train_args.span_emb += \"_head\"\n if not hasattr(train_args, 'span_boundaries_from_wordemb'):\n train_args.span_boundaries_from_wordemb = False\n if not hasattr(train_args, 'attention_ent_vecs_no_regularization'):\n train_args.attention_ent_vecs_no_regularization = False\n if not hasattr(train_args, 'global_topkthr'):\n train_args.global_topkthr = None\n if not hasattr(train_args, 'global_topkfromallspans'):\n train_args.global_topkfromallspans = None\n if not hasattr(train_args, 'hardcoded_thr'):\n train_args.hardcoded_thr = None\n if not hasattr(train_args, 'global_one_loss'):\n train_args.global_one_loss = False\n if not hasattr(train_args, 'global_norm_or_mean'):\n train_args.global_norm_or_mean = \"norm\"\n if not hasattr(train_args, 'ffnn_dropout'):\n train_args.ffnn_dropout = True\n if not hasattr(train_args, 'cand_ent_num_restriction'):\n train_args.cand_ent_num_restriction = None\n if not hasattr(train_args, 'ffnn_l2maxnorm'):\n train_args.ffnn_l2maxnorm = None\n if not hasattr(train_args, 'ffnn_l2maxnorm_onlyhiddenlayers'):\n train_args.ffnn_l2maxnorm_onlyhiddenlayers = False\n if not hasattr(train_args, 'pem_without_log'):\n train_args.pem_without_log = False\n if not hasattr(train_args, 'global_mask_scale_each_mention_voters_to_one'):\n train_args.global_mask_scale_each_mention_voters_to_one = False\n if not hasattr(train_args, 'pem_buckets_boundaries'):\n train_args.pem_buckets_boundaries = None\n if not hasattr(train_args, 'global_gmask_based_on_localscore'):\n train_args.global_gmask_based_on_localscore = False\n if not hasattr(train_args, 'attention_retricted_num_of_entities'):\n train_args.attention_retricted_num_of_entities = None\n\n if not hasattr(train_args, 'stage2_nn_components'):\n train_args.stage2_nn_components = \"local_global\"\n if not hasattr(train_args, 'global_gmask_unambigious'):\n train_args.global_gmask_unambigious = False\n return train_args\n\n\ndef load_ent_vecs(args):\n entity_embeddings_nparray = np.load(config.base_folder + \"data/entities/ent_vecs/ent_vecs.npy\")\n entity_embeddings_nparray[0] = 0\n if hasattr(args, 'entity_extension') and args.entity_extension is not None:\n entity_extension = np.load(config.base_folder +\"data/entities/\"+args.entity_extension+\n \"/ent_vecs/ent_vecs.npy\")\n entity_embeddings_nparray = np.vstack((entity_embeddings_nparray, entity_extension))\n return entity_embeddings_nparray\n\n\n\n\nif __name__ == \"__main__\":\n path = \"/home/master_thesis_share/data/tfrecords/per_document_no_wikidump/all_spans_training_folder/group_global/c50h50_lstm150_nohead_attR10K100_fffnn0_0_glthrm005_glffnn0_0v1/train_args.pickle\"\n with open(path, 'rb') as handle:\n train_args = pickle.load(handle)\n train_args.global_topk = None\n with open(path, 'wb') as handle:\n pickle.dump(train_args, handle)\n\n\n\n\n\n" }, { "path": "code/preprocessing/__init__.py", "content": "" }, { "path": "code/preprocessing/aida_insight.py", "content": "import model.config as config\n\n\ndef process_file(filename):\n entities = set()\n mentions = set()\n with open(config.base_folder+\"data/new_datasets/\"+filename) as fin:\n inmention = False\n mention_acc = []\n for line in fin:\n line = line.rstrip() # omit the '\\n' character\n if line.startswith('MMSTART_'):\n ent_id = line[8:] # assert that ent_id in wiki_name_id_map\n entities.add(ent_id)\n inmention = True\n mention_acc = []\n elif line == 'MMEND':\n inmention = False\n mentions.add(' '.join(mention_acc))\n elif inmention:\n mention_acc.append(line)\n return entities, mentions\n\n\ndef main():\n train_entities, train_mentions = process_file(\"aida_train.txt\")\n dev_entities, dev_mentions = process_file(\"aida_dev.txt\")\n test_entities, test_mentions = process_file(\"aida_test.txt\")\n\n print(\"len(train_entities) =\", len(train_entities))\n print(\"len(dev_entities) =\", len(dev_entities))\n print(\"len(test_entities) =\", len(test_entities))\n\n print(\"train_entities.intersection(dev_entities) =\", len(train_entities.intersection(dev_entities)))\n print(\"train_entities.intersection(test_entities) =\", len(train_entities.intersection(test_entities)))\n print(\"dev_entities.intersection(test_entities) =\", len(dev_entities.intersection(test_entities)))\n\n\n print(\"len(train_mentions) =\", len(train_mentions))\n print(\"len(dev_mentions) =\", len(dev_mentions))\n print(\"len(test_mentions) =\", len(test_mentions))\n\n print(\"train_mentions.intersection(dev_mentions) =\", len(train_mentions.intersection(dev_mentions)))\n print(\"train_mentions.intersection(test_mentions) =\", len(train_mentions.intersection(test_mentions)))\n print(\"dev_mentions.intersection(test_mentions) =\", len(dev_mentions.intersection(test_mentions)))\n\nif __name__ == \"__main__\":\n main()\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n" }, { "path": "code/preprocessing/bridge_code_lua/ent_vecs_from_txt_to_npy.py", "content": "import numpy as np\nfrom preprocessing.util import load_wikiid2nnid\nimport model.config as config\n\n\ndef keep_only_new_entities(ent_vecs, folder):\n main_wikiid2nnid = load_wikiid2nnid()\n additional_wikiids = []\n rows_to_extract = []\n with open(folder + \"wikiid2nnid/wikiid2nnid.txt\", \"r\") as fin, \\\n open(folder + \"wikiid2nnid/additional_wikiids.txt\", \"w\") as fout:\n for line in fin:\n ent_id, nnid = line.split('\\t')\n nnid = int(nnid) - 1 # torch starts from 1 instead of zero\n if ent_id not in main_wikiid2nnid:\n additional_wikiids.append(ent_id)\n rows_to_extract.append(nnid)\n fout.write(ent_id+\"\\n\")\n print(\"additional entities =\", len(additional_wikiids))\n return ent_vecs[rows_to_extract]\n\n\ndef main():\n folder = config.base_folder + (\"data/entities/extension_entities/\" if args.entity_extension else \"data/entities/\")\n print(\"folder =\", folder)\n ent_vecs = np.loadtxt(folder + \"ent_vecs/ent_vecs.txt\")\n #ent_vecs.shape # (484048, 300)\n if args.entity_extension:\n ent_vecs = keep_only_new_entities(ent_vecs, folder)\n np.save(folder + \"ent_vecs/ent_vecs.npy\", ent_vecs)\n\n\ndef _parse_args():\n import argparse\n parser = argparse.ArgumentParser()\n parser.add_argument(\"--entity_extension\", default=False)\n args = parser.parse_args()\n return args\n\n\nif __name__ == \"__main__\":\n args = _parse_args()\n main()\n" }, { "path": "code/preprocessing/bridge_code_lua/ent_vecs_to_txt.lua", "content": "require 'torch'\nrequire 'nn'\nrequire 'os'\nel_path = os.getenv(\"EL_PATH\")\ncmd = torch.CmdLine()\ncmd:option('-ent_vecs_filepath', '/home/end2end_neural_el/data/entities/ent_vecs/ent_vecs__ep_147.t7', 'the file paht for the entity vectors with best score.')\nopt = cmd:parse(arg or {})\nfolder_name, file_name = string.match(opt.ent_vecs_filepath, \"(.-)([^/]*)$\")\nprint(\"folder_name: \" .. folder_name)\nprint(\"file_name: \" .. file_name)\n--print(opt.ent_vecs_filepath)\n\nent_vecs = torch.load(opt.ent_vecs_filepath)\nprint('lines = ' .. ent_vecs:size(1))\nprint('columns = ' .. ent_vecs:size(2))\nent_vecs = nn.Normalize(2):forward(ent_vecs:double()) -- Needs to be normalized to have norm 1.\n-- print them to normal txt file\n--out = assert(io.open(el_path .. \"/data/entities/ent_vecs/ent_vecs.txt\", \"w\")) -- open a file for serialization\nout = assert(io.open(folder_name .. \"ent_vecs.txt\", \"w\")) -- open a file for serialization\nsplitter = \" \"\nfor i=1,ent_vecs:size(1) do\n for j=1,ent_vecs:size(2) do\n out:write(ent_vecs[i][j])\n if j == ent_vecs:size(2) then\n out:write(\"\\n\")\n else\n out:write(splitter)\n end\n end\nend\nout:close()\n\n\n" }, { "path": "code/preprocessing/extra.py", "content": "import pickle\nfrom collections import defaultdict\nimport numpy as np\nimport time\nimport sys\nimport os\n\nimport preprocessing.config as config\nfrom preprocessing.util import *\n\n\ndef vocabulary_count_wiki():\n word_freq = defaultdict(int)\n char_freq = defaultdict(int) # how many times each character is encountered\n\n doc_cnt = 0\n #with open(base_folder + \"data/mydata/tokenized_toy_wiki_dump.txt\") as fin:\n with open(config.base_folder+\"data/basic_data/tokenizedWiki.txt\") as fin:\n for line in fin:\n if line.startswith(''):\n doc_cnt += 1\n if doc_cnt % 5000 == 0:\n print(\"document counter: \", doc_cnt)\n continue\n if line.startswith('') or line.startswith('*NL*'):\n continue\n\n line = line[:-1] # omit the '\\n' character\n word_freq[line.lower() if config.lowercase_emb else line] += 1\n for c in line:\n char_freq[c] += 1\n\n # print some statistics\n print(\"some frequency statistics. The bins are [...) \")\n for d, name, edges in zip([word_freq, char_freq], [\"word\", \"character\"],\n [[1, 2, 3, 6, 11, 21, 31, 51, 76, 101, 201, np.inf],\n [1, 6, 11, 21, 51, 101, 201, 501, 1001, 2001, np.inf]]):\n hist_values, _ = np.histogram(list(d.values()), edges)\n cum_sum = np.cumsum(hist_values[::-1])\n print(name, \" frequency histogram, edges: \", edges)\n print(\"absolute values: \", hist_values)\n print(\"absolute cumulative (right to left): \", cum_sum[::-1])\n print(\"probabilites cumulative (right to left):\", (cum_sum / np.sum(hist_values))[::-1])\n\n with open(config.base_folder+\"data/vocabulary/frequencies{}.pickle\".format(\"_lowercase\"\n if config.lowercase_emb else \"\"), 'wb') as handle:\n pickle.dump((word_freq, char_freq), handle)\n\n\n\n\ndef entity_count_wiki_aux():\n with open(config.base_folder+\"data/entities/entity_frequencies.pickle\", 'rb') as handle:\n ent_freq = pickle.load(handle)\n import operator\n sorted_ = sorted(ent_freq.items(), key=operator.itemgetter(1), reverse=True)\n _, wiki_id_name_map = load_wiki_name_id_map()\n wiki_id_name_map[\"0\"] = \"unk_entity\"\n ent_id_errors = 0\n with open(config.base_folder+\"data/entities/entity_frequencies.txt\", \"w\") as fout:\n for ent_id, freq in sorted_:\n if ent_id not in wiki_id_name_map:\n print(\"ent_id: \", ent_id, \" not in wiki_name_id_map\")\n ent_id_errors += 1\n else:\n fout.write(ent_id+\"\\t\"+wiki_id_name_map[ent_id]+\"\\t\"+str(freq)+\"\\n\")\n print(\"ent_id_errors: \", ent_id_errors)\n\n\n# checks for entities\n# equivalent to vocabulary_count but for entities. executed only once\n# counts how many hyperlinks point to each entity and prints statistics\ndef entity_count_wiki():\n ent_freq = defaultdict(int)\n\n hyperlink2EntityId = EntityNameIdMap()\n hyperlink2EntityId.init_hyperlink2id()\n doc_cnt = 0\n #with open(base_folder + \"data/mydata/tokenized_toy_wiki_dump.txt\") as fin:\n with open(config.base_folder+\"data/basic_data/tokenizedWiki.txt\") as fin:\n for line in fin:\n if line.startswith(''):\n doc_cnt += 1\n if doc_cnt % 5000 == 0:\n print(\"document counter: \", doc_cnt)\n continue\n if line.startswith(' 100:\n # break\n print(\"end of p_e_m reading. wall time:\", (time.time() - wall_start)/60, \" minutes\")\n print(\"p_e_m_errors: \", p_e_m_errors)\n\n with open(config.base_folder+\"data/serializations/p_e_m_disamb_redirect_wikinameid_maps.pickle\", 'wb') as handle:\n pickle.dump((p_e_m, config.cand_ent_num, disambiguations_ids, disambiguations_titles, redirections, \\\n wiki_name_id_map, wiki_id_name_map, wiki_name_id_map_lower), handle)\n return p_e_m, disambiguations_ids, redirections, wiki_name_id_map\n" }, { "path": "code/preprocessing/old/prepro_datasets1_once.py", "content": "import pickle\nimport argparse\nimport time\n#import datetime\nfrom collections import defaultdict\n\n#import sys\n#import os\n#sys.path.append(os.path.abspath('../aux'))\n\n\nword_freq = defaultdict(int)\nchar_freq = defaultdict(int) # how many times each character is encountered\np_e_m_position = defaultdict(int) # for the gold mentions where the ground truth appears in\n # the p_e_m dictionary\nlist_of_p_e_m_position = []\n\ndef vocabulary_count_wiki():\n\n doc_cnt = 0\n #with open(base_folder + \"data/mydata/tokenized_toy_wiki_dump.txt\") as fin:\n with open(config.base_folder+\"data/basic_data/tokenizedWiki.txt\") as fin:\n for line in fin:\n if line.startswith(''):\n doc_cnt += 1\n if doc_cnt % 5000 == 0:\n print(\"document counter: \", doc_cnt)\n continue\n if line.startswith('') or line.startswith('*NL*'):\n continue\n\n line = line[:-1] # omit the '\\n' character\n word_freq[line.lower() if config.lowercase_emb else line] += 1\n for c in line:\n char_freq[c] += 1\n\n # print some statistics\n print(\"some frequency statistics. The bins are [...) \")\n for d, name, edges in zip([word_freq, char_freq], [\"word\", \"character\"],\n [[1, 2, 3, 6, 11, 21, 31, 51, 76, 101, 201, np.inf],\n [1, 6, 11, 21, 51, 101, 201, 501, 1001, 2001, np.inf]]):\n hist_values, _ = np.histogram(list(d.values()), edges)\n cum_sum = np.cumsum(hist_values[::-1])\n print(name, \" frequency histogram, edges: \", edges)\n print(\"absolute values: \", hist_values)\n print(\"absolute cumulative (right to left): \", cum_sum[::-1])\n print(\"probabilites cumulative (right to left):\", (cum_sum / np.sum(hist_values))[::-1])\n\n with open(config.base_folder+\"data/vocabulary/frequencies{}.pickle\".format(\"_lowercase\"\n if config.lowercase_emb else \"\"), 'wb') as handle:\n pickle.dump((word_freq, char_freq), handle)\n\n\ndef process_aida_aux(filepath):\n with open(filepath) as fin:\n line_cnt = 0\n for line in fin:\n line = line.rstrip()\n if line.startswith(\"-DOCSTART-\"):\n continue\n #if line == \"\":\n # print(\"*NL*\")\n\n line = line.split('\\t')\n\n word_freq[line.lower() if args.lowercase_emb else line] += 1\n for c in line:\n char_freq[c] += 1\n\n line_cnt += 1\n if line_cnt == 30:\n break\n\n\n\ndef process_aida():\n process_aida_aux(datasets+\"AIDA/aida_train.txt\")\n process_aida_aux(datasets+\"AIDA/testa_testb_aggregate_original\")\n\n\ndef process_other_datasets(folder_path):\n pass\n\n\n\n\n\ndef _parse_args():\n parser = argparse.ArgumentParser()\n parser.add_argument(\"--datasets\", default=\"/home/master_thesis_share/data/basic_data/test_datasets/\")\n parser.add_argument(\"--lowercase_emb\", type=bool, default=False,\n help=\"if true then lowercase word for counting and embedding\")\n return parser.parse_args()\n\nif __name__ == \"__main__\":\n args = _parse_args()\n datasets = args.datasets\n\n process_aida()\n '''\n process_other_datasets(datasets+\"/wned-datasets/ace2004/\")\n process_other_datasets(datasets+\"/wned-datasets/aquaint/\")\n process_other_datasets(datasets+\"/wned-datasets/clueweb/\")\n process_other_datasets(datasets+\"/wned-datasets/msnbc/\")\n process_other_datasets(datasets+\"/wned-datasets/wikipedia/\")\n '''\n\n\n\n\n\n\n\n" }, { "path": "code/preprocessing/old/preprocess1.py", "content": "import pickle\nimport argparse\nfrom collections import defaultdict\nfrom nltk.tokenize import sent_tokenize\nfrom nltk.tokenize import word_tokenize\nimport numpy as np\nimport time\n\n# TODO put all these in a config file\ndata_folder = \"../data/mydata/\" # \"../data/basic_data/\"\ndev_ratio = 0.1\ntest_ratio = 0.05\nper_sentence = False\nper_paragraph = False # or per article, or per sentence\nper_article = True\n\nword_freq_thr = 20 # word frequency threshold (if freq less than that then replace it with the\n # special word . if it has higher than that frequency but not in Word2Vec again\n # replace it with \nchar_freq_thr = 100 # character frequency threshold (if encountered less times than that then\n # replace it with the char \n\nlowercase_wordemb = False # before we look for its emb we lowercase\n\n\n\n\ndef main():\n word_freq = defaultdict(int)\n char_freq = defaultdict(int) # how many times each character is encountered\n entity_freq = defaultdict(int) # how many hyperlinks point to this entity\n ent_name_id_map = dict() # just to verify that it is the same with the wiki_name_id_map.txt\n\n\n import re\n docid_title_pattern = re.compile(r'curid=(\\d+)\" title=\"(.*)\">')\n hyperlink_pattern = re.compile(r'(.*?)(.*?)')\n\n\n doc_cnt = 0\n# with open(data_folder + \"toy_wiki_dump.txt\") as fin: # \"textWithAnchorsFromAllWikipedia2014Feb.txt\") as f:\n with open(\"../data/basic_data/textWithAnchorsFromAllWikipedia2014Feb.txt\") as fin:\n for line in fin:\n if line.startswith(''):\n doc_cnt += 1\n if doc_cnt % 500 == 0:\n print(\"document counter: \", doc_cnt)\n continue\n result = re.findall(hyperlink_pattern, line)\n clean_text = []\n for res in result:\n clean_text.append(res[0])\n clean_text.append(res[2])\n clean_text.append(line.rsplit('', 1)[-1]) # text after last hyperlink\n clean_text = ''.join(clean_text)\n for w in word_tokenize(clean_text):\n word_freq[w] += 1\n for c in clean_text:\n char_freq[c] += 1\n\n # print some statistics\n print(\"some frequency statistics. The bins are [...) \")\n for d, name, edges in zip([word_freq, char_freq], [\"word\", \"character\"],\n [[1, 2, 3, 6, 11, 21, 31, 51, 76, 101, 201, np.inf],\n [1, 6, 11, 21, 51, 101, 201, 501, 1001, 2001, np.inf]]):\n hist_values, _ = np.histogram(list(d.values()), edges)\n cum_sum = np.cumsum(hist_values[::-1])\n print(name, \" frequency histogram, edges: \", edges)\n print(\"absolute values: \", hist_values)\n print(\"absolute cumulative (right to left): \", cum_sum[::-1])\n print(\"probabilites cumulative (right to left):\", (cum_sum / np.sum(hist_values))[::-1])\n\n\n\n\ndef re_faster(line):\n result = []\n pass\n\ndef profiler():\n\n\n import re\n docid_title_pattern = re.compile(r'curid=(\\d+)\" title=\"(.*)\">')\n hyperlink_pattern = re.compile(r'(.*?)(.*?)')\n\n wall_start = time.time()\n experiments_times = []\n doc_cnt = 0\n for tokenization_flag, counting_flag, name in [(False, False, \"clean text\"), (True, False, \"clean text & tokenization\"), (True, True, \"clean text & tokenization & counting\")]: \n word_freq = defaultdict(int)\n char_freq = defaultdict(int) # how many times each character is encountered\n entity_freq = defaultdict(int) # how many hyperlinks point to this entity\n ent_name_id_map = dict() # just to verify that it is the same with the wiki_name_id_map.txt\n with open(\"../data/basic_data/textWithAnchorsFromAllWikipedia2014Feb.txt\") as fin:\n for line in fin:\n if line.startswith(''):\n doc_cnt += 1\n if doc_cnt % 500 == 0:\n print(\"document counter: \", doc_cnt)\n if doc_cnt == 10000:\n doc_cnt = 0\n break\n continue\n result = re.findall(hyperlink_pattern, line)\n clean_text = []\n for res in result:\n clean_text.append(res[0])\n clean_text.append(res[2])\n clean_text.append(line.rsplit('', 1)[-1]) # text after last hyperlink\n clean_text = ''.join(clean_text)\n \n if tokenization_flag:\n all_tokens = word_tokenize(clean_text)\n if counting_flag: \n for w in all_tokens:\n word_freq[w] += 1\n for c in clean_text:\n char_freq[c] += 1\n\t\n cur_exp_time = (time.time() - wall_start)/60\n print(\"wall time for \", name, \" : \", cur_exp_time, \" minutes\") \n experiments_times.append(cur_exp_time)\n wall_start = time.time()\n\n print(\"clean text time: \", experiments_times[0])\n print(\"tokenization time: \", experiments_times[1]-experiments_times[0])\n print(\"counting time: \", experiments_times[2]-experiments_times[1])\n\n'''\nclean text time: 2.8340925057729085\ntokenization time: 2.915117104848226\ncounting time: 0.2761443416277567\n'''\n\n\n'''\n\n # first pass through the corpus. now we just count the frequency of words, chars and entities\n #text_chunk = [] # it can be a sentence, paragraph or the whole article\n #text_chunk = ''.join([`num` for num in xrange(loop_count)])\n chunks = [] # this is where we accumulate all the data samples. but since the\n # corpus is too big we flush it at the end of each document that we read\n for line in fin:\n if line.startswith(''):\n # TODO flush the chunks to file train dev test\n # read consecutive lines until we form the desired chunking article, par, sent\n if per_sentence:\n cur_chunks = sent_tokenize(line)\n\n\n\n # in the second pass we do that (after encoding)\n #if line.startswith(''):\n # TODO flush the chunks to file train dev test\n\n'''\n\n\n\nif __name__ == \"__main__\":\n #main()\n profiler()\n" }, { "path": "code/preprocessing/old/preprocessv2.py", "content": "import pickle\nimport argparse\nfrom collections import defaultdict\nimport numpy as np\nimport time\nimport sys\nimport os\n\n# TODO put all these in a args file\nimport config\nimport util\n\n\n\n\ndef build_word_char_maps():\n with open(config.base_folder+\"data/vocabulary/frequencies.pickle\", 'rb') as handle:\n word_freq, char_freq = pickle.load(handle)\n word2id = dict()\n id2word = dict()\n char2id = dict()\n id2char = dict()\n\n import gensim\n model = gensim.models.KeyedVectors.load_word2vec_format(\n \"/home/master_thesis_share/data/basic_data/wordEmbeddings/Word2Vec/GoogleNews-vectors-negative300.bin\", binary=True)\n embedding_dim = len(model['queen'])\n\n wcnt = 0 # unknown word\n word2id[\"\"] = wcnt\n id2word[wcnt] = \"\"\n wcnt += 1\n ccnt = 0 # unknown character\n char2id[\"\"] = ccnt\n id2char[ccnt] = \"\"\n ccnt += 1\n\n for word in word_freq: # for every word in the corpus\n if word in model: # has a pretrained embeddings\n if config.word_freq_thr and word_freq[word] < config.word_freq_thr:\n pass # rare word, so don't put it in our vocabulary\n else:\n word2id[word] = wcnt\n id2word[wcnt] = word\n wcnt += 1\n\n for c in char_freq:\n if config.char_freq_thr and char_freq[c] < config.char_freq_thr:\n pass # rare character, so don't put it in our vocabulary\n else:\n char2id[c] = ccnt\n id2char[ccnt] = c\n ccnt += 1\n # the space ' ' and the EOL '\\n' are not inside since they are discarded by the tokenizer\n # maybe they are useful later?\n assert(len(word2id) == wcnt)\n assert(len(char2id) == ccnt)\n print(\"words in vocabulary: \", wcnt)\n print(\"characters in vocabulary: \", ccnt)\n with open(config.base_folder+\"data/vocabulary/word_char_maps.pickle\", 'wb') as handle:\n pickle.dump((word2id, id2word, char2id, id2char, config.word_freq_thr, config.char_freq_thr), handle)\n\n return word2id, id2word, char2id, id2char, config.word_freq_thr, config.char_freq_thr\n\n\n\n\ndef samples_load(filename):\n with open(filename, \"rb\") as f:\n while True:\n try:\n yield pickle.load(f)\n except EOFError:\n break\n\n# https://stackoverflow.com/questions/20716812/saving-and-loading-multiple-objects-in-pickle-file\n#samples = load(myfilename)\n\nclass Encoding(object):\n def __init__(self):\n self.entity_freq = defaultdict(int) # how many hyperlinks point to this entity\n self.ent_name_id_map = dict() # just to verify that it is the same with the wiki_name_id_map.txt\n self.ent_id_name_map = dict()\n with open(config.base_folder+\"data/vocabulary/word_char_maps.pickle\", 'rb') as handle:\n self.word2id, _, self.char2id, _, _, _ = pickle.load(handle)\n #_word2id, id2word, _char2id, id2char, word_freq_thr, char_freq_thr = pickle.load(handle)\n\n with open(config.base_folder+\"data/serializations/p_e_m_disamb_redirect_wikinameid_maps.pickle\", 'rb') as handle:\n self.p_e_m, _, self.disambiguations_ids, _, self.redirections, self.wiki_name_id_map, _,\\\n self.wiki_name_id_map_lower = pickle.load(handle)\n # p_e_m, args.cand_ent_num, disambiguations_ids, disambiguations_titles, redirections,\\\n # wiki_name_id_map, wiki_id_name_map\n\n self.unk_ent_id = 0\n #assert(0 not in self.wiki_id_name_map)\n\n self.chunk_ending = {''}\n if config.per_paragraph:\n self.chunk_ending.add('*NL*')\n if config.per_sentence:\n self.chunk_ending.add('*NL*')\n self.chunk_ending.add('.')\n\n\n def new_chunk(self):\n self.chunk_words = []\n self.chunk_chars = [] # a list of lists\n self.begin_idx = [] # the start positions of mentions\n self.end_idx = [] # the end positions of mentions\n self.candidate_entities = [] # a list of lists\n\n def process_word(self, word):\n pass\n\n def serialize(self):\n if self.begin_idx != []: # this chunk has something to train on\n self.serialize_tfrecords()\n #pickle.dump(((docid, par_cnt, sent_cnt), self.chunk_words, self.chunk_chars,\n # self.begin_idx, self.end_idx, self.candidate_entities), handle)\n self.new_chunk()\n\n def serialize_tfrecords(self):\n pass\n\n\n pass\n\n def encode_wikidump(self):\n doc_cnt = 0\n #with open(base_folder+\"data/basic_data/tokenizedWiki.txt\") as fin,\\\n with open(config.base_folder + \"data/mydata/tokenized_toy_wiki_dump.txt\") as fin,\\\n open(config.base_folder+\"data/basic_data/samples.pickle\", 'wb') as handle:\n self.new_chunk()\n\n # paragraph and sentence counter are not actually useful. only for debugging purposes.\n par_cnt = 0 # paragraph counter (useful if we work per paragraph)\n sent_cnt = 0 # sentence counter (useful if we work per sentence)\n self.in_mention = False\n for line in fin:\n line = line[:-1] # omit the '\\n' character\n if line.startswith('':\n doc_cnt += 1\n if doc_cnt % 5000 == 0:\n print(\"document counter: \", doc_cnt)\n continue\n\n if line.startswith('') or line.startswith('*NL*'):\n continue\n\n\n #if line.startswith('') or line.startswith('*NL*'):\n\n\n\n\n\n\n'''\n\n # first pass through the corpus. now we just count the frequency of words, chars and entities\n #text_chunk = [] # it can be a sentence, paragraph or the whole article\n #text_chunk = ''.join([`num` for num in xrange(loop_count)])\n chunks = [] # this is where we accumulate all the data samples. but since the\n # corpus is too big we flush it at the end of each document that we read\n for line in fin:\n if line.startswith(''):\n # TODO flush the chunks to file train dev test\n # read consecutive lines until we form the desired chunking article, par, sent\n if per_sentence:\n cur_chunks = sent_tokenize(line)\n\n\n\n # in the second pass we do that (after encoding)\n #if line.startswith(''):\n # TODO flush the chunks to file train dev test\n\n'''\n\n\n\nif __name__ == \"__main__\":\n # util.tokenize_p_e_m()\n # util.vocabulary_count()\n util.p_e_m_disamb_redirect_wikinameid_maps()\n # the above commands can be executed only once. the results are serialized so no need to\n # execute them every time.\n\n #build_word_char_maps()\n\n #temp = Encoding()\n #temp.encode_wikidump()\n #main()\n" }, { "path": "code/preprocessing/p_e_m.py", "content": "import pickle\nfrom collections import defaultdict\nfrom nltk.tokenize import word_tokenize\nimport time\nimport sys\nimport os\nimport operator\n\nimport preprocessing.config as config\nimport preprocessing.util as util\n\n\n# TODO delete this one, no longer usefull since there is another one that also handles the conflicts\ndef tokenize_p_e_m():\n '''\n tokenizes the mention of the p(e|m) dictionary so that it is consistent with our\n tokenized corpus (,.;' all these symbols separate from the previous word with whitespace)\n it only modifies the mention and nothing else.\n '''\n #for dict_file in [\"prob_wikipedia_p_e_m.txt\", \"prob_yago_crosswikis_wikipedia_p_e_m.txt\",\n # \"prob_crosswikis_wikipedia_p_e_m.txt\"]:\n for dict_file in [\"prob_yago_crosswikis_wikipedia_p_e_m.txt\"]:\n with open(config.base_folder+\"data/p_e_m/\"+dict_file) as fin, \\\n open(config.base_folder+\"data/p_e_m/\" + \"tokenized/\"+dict_file, \"w\") as fout:\n diff_cnt = 0\n for line in fin:\n mention, rest = line.split('\\t', 1)\n if len(mention.split()) > 1:\n tokenized_mention = ' '.join(word_tokenize(mention))\n else:\n tokenized_mention = mention\n if mention != tokenized_mention:\n diff_cnt += 1\n #print(mention, \" --------> \", tokenized_mention)\n fout.write(tokenized_mention + \"\\t\" + rest)\n print(dict_file, \"diff_cnt = \", diff_cnt)\n\n\ndef print_p_e_m_dictionary_to_file(p_e_m, full_filepath):\n _, wiki_id_name_map = util.load_wiki_name_id_map()\n with open(full_filepath, \"w\") as fout:\n for mention, entities in p_e_m.items():\n out_acc = []\n # entities is a defaultdict(int)\n # so items returns ent2: 10, ent54:20, ent3:2\n sorted_ = sorted(entities.items(), key=operator.itemgetter(1), reverse=True)\n # a list of tuples [(ent54,20), (ent2,10), (ent3,2)]\n total_freq = 0\n for ent_id, prob in sorted_:\n if len(out_acc) > 100: # at most 100 candidate entities\n break\n total_freq += prob\n out_acc.append(','.join([ent_id, str(prob),\n wiki_id_name_map[ent_id].replace(' ', '_')]))\n fout.write(mention + \"\\t\" + str(total_freq) + \"\\t\" + \"\\t\".join(out_acc) + \"\\n\")\n\n\ndef tokenize_p_e_m_and_merge_conflicts(filename, yago=False):\n \"\"\"takes as input a p_e_m with absolute frequency, tokenizes the mention, handles conflicts\n (same mention after tokenization) with merging. execute that on wiki, crosswiki, yago\n absolute frequency files -> output again absolute frequency.\"\"\"\n p_e_m = defaultdict(lambda: defaultdict(int))\n with open(config.base_folder+\"data/p_e_m/\"+filename) as fin:\n diff_cnt = 0\n conflicts_cnt = 0\n for line in fin:\n line = line.rstrip()\n l = line.split(\"\\t\")\n mention = l[0]\n tokenized_mention = ' '.join(word_tokenize(mention))\n\n if mention != tokenized_mention:\n diff_cnt += 1\n if tokenized_mention in p_e_m:\n conflicts_cnt += 1\n #print(mention, \" --------> \", tokenized_mention)\n for e in l[2:]:\n if yago:\n ent_id, _ = e.split(',', 1)\n ent_id = ent_id.strip() # not necessary\n freq = 1\n else:\n ent_id, freq, _ = e.split(',', 2)\n ent_id = ent_id.strip() # not necessary\n freq = int(freq)\n p_e_m[tokenized_mention][ent_id] += freq\n\n print(\"conflicts from tokenization counter: \", conflicts_cnt)\n print_p_e_m_dictionary_to_file(p_e_m, config.base_folder+\"data/p_e_m/tokenized/\"+filename)\n\n\ndef from_freq_to_prob(filename):\n with open(config.base_folder+\"data/p_e_m/tokenized/\"+filename) as fin, \\\n open(config.base_folder+\"data/p_e_m/tokenized/prob_\"+filename, \"w\") as fout:\n p_e_m_errors = 0\n for line in fin:\n line = line.rstrip()\n try:\n l = line.split(\"\\t\")\n total_freq = int(l[1])\n out_acc = [l[0], str(1)] # mention and total_prob\n for e in l[2:]:\n ent_id, freq, title = e.split(',', 2)\n out_acc.append(','.join([ent_id, str(int(freq)/total_freq), title]))\n fout.write('\\t'.join(out_acc) + \"\\n\")\n except Exception as esd:\n exc_type, exc_obj, exc_tb = sys.exc_info()\n fname = os.path.split(exc_tb.tb_frame.f_code.co_filename)[1]\n print(exc_type, fname, exc_tb.tb_lineno)\n p_e_m_errors += 1\n print(\"error in line: \", repr(line))\n\n\ndef merge_two_prob_dictionaries(filename1, filename2, newfilename):\n \"\"\"merge two p_e_m dictionaries that are already in probabilities to a new one again\n with probabilities.\"\"\"\n p_e_m = defaultdict(lambda: defaultdict(float))\n for filename in [filename1, filename2]:\n with open(config.base_folder+\"data/p_e_m/tokenized/\"+filename) as fin:\n for line in fin:\n line = line.rstrip()\n l = line.split(\"\\t\")\n mention = l[0]\n for e in l[2:]:\n ent_id, prob, _ = e.split(',', 2)\n ent_id = ent_id.strip() # not necessary\n prob = float(prob)\n #p_e_m[mention][ent_id] = min(1, p_e_m[mention][ent_id] + prob)\n p_e_m[mention][ent_id] = p_e_m[mention][ent_id] + prob # without min\n # even without restricting it still the range of values is [0,2]\n\n print_p_e_m_dictionary_to_file(p_e_m, config.base_folder+\"data/p_e_m/tokenized/\"\n + newfilename)\n\n\nif __name__ == \"__main__\":\n tokenize_p_e_m()\n\n #tokenize_p_e_m_and_merge_conflicts(\"wikipedia_p_e_m.txt\")\n #tokenize_p_e_m_and_merge_conflicts(\"crosswikis_wikipedia_p_e_m.txt\")\n #tokenize_p_e_m_and_merge_conflicts(\"yago_p_e_m.txt\", yago=True)\n\n #from_freq_to_prob(\"wikipedia_p_e_m.txt\")\n #from_freq_to_prob(\"crosswikis_wikipedia_p_e_m.txt\")\n #from_freq_to_prob(\"yago_p_e_m.txt\")\n\n \"\"\"\n merge_two_prob_dictionaries(\"prob_crosswikis_wikipedia_p_e_m.txt\",\n \"prob_yago_p_e_m.txt\",\n \"prob_yago_crosswikis_wikipedia_p_e_m.txt\")\n \"\"\"\n\n # vocabulary_count_wiki()\n # entity_count_wiki()\n # load_p_e_m()\n # from_freq_to_prob()\n" }, { "path": "code/preprocessing/prepro_aida.py", "content": "import argparse\nimport os\nimport preprocessing.util as util\n\ndef process_aida(in_filepath, out_filepath):\n\n # _, wiki_id_name_map = util.load_wiki_name_id_map(lowercase=False)\n #_, wiki_id_name_map = util.entity_name_id_map_from_dump()\n entityNameIdMap = util.EntityNameIdMap()\n entityNameIdMap.init_compatible_ent_id()\n unknown_gt_ids = 0 # counter of ground truth entity ids that are not in the wiki_name_id.txt\n ent_id_changes = 0\n with open(in_filepath) as fin, open(out_filepath, \"w\") as fout:\n in_mention = False # am i inside a mention span or not\n first_document = True\n for line in fin:\n l = line.split('\\t')\n if in_mention and not (len(l) == 7 and l[1]=='I'):\n # if I am in mention but the current line does not continue the previous mention\n # then print MMEND and be in state in_mention=FALSE\n fout.write(\"MMEND\\n\")\n in_mention = False\n\n if line.startswith(\"-DOCSTART-\"):\n if not first_document:\n fout.write(\"DOCEND\\n\")\n # line = \"-DOCSTART- (967testa ATHLETICS)\\n\"\n doc_title = line[len(\"-DOCSTART- (\"): -2]\n fout.write(\"DOCSTART_\"+doc_title.replace(' ', '_')+\"\\n\")\n first_document = False\n elif line == \"\\n\":\n fout.write(\"*NL*\\n\")\n elif len(l) == 7 and l[1] == 'B': # this is a new mention\n wiki_title = l[4]\n wiki_title = wiki_title[len(\"http://en.wikipedia.org/wiki/\"):].replace('_', ' ')\n new_ent_id = entityNameIdMap.compatible_ent_id(wiki_title, l[5])\n if new_ent_id is not None:\n if new_ent_id != l[5]:\n ent_id_changes += 1\n #print(line, \"old ent_id: \" + l[5], \" new_ent_id: \", new_ent_id)\n fout.write(\"MMSTART_\"+new_ent_id+\"\\n\") # TODO check here if entity id is inside my wikidump\n # if not then omit this mention\n fout.write(l[0]+\"\\n\") # write the word\n in_mention = True\n else:\n unknown_gt_ids += 1\n fout.write(l[0]+\"\\n\") # write the word\n print(line)\n else:\n # words that continue a mention len(l) == 7: and l[1]=='I'\n # or normal word outside of mention, or in mention without disambiguation (len(l) == 4)\n fout.write(l[0].rstrip()+\"\\n\")\n fout.write(\"DOCEND\\n\") # for the last document\n print(\"process_aida unknown_gt_ids: \", unknown_gt_ids)\n print(\"process_aida ent_id_changes: \", ent_id_changes)\n\n\ndef split_dev_test(in_filepath):\n with open(in_filepath) as fin, open(args.output_folder+\"temp_aida_dev\", \"w\") as fdev,\\\n open(args.output_folder+\"temp_aida_test\", \"w\") as ftest:\n fout = fdev\n for line in fin:\n if line.startswith(\"-DOCSTART-\") and line.find(\"testb\") != -1:\n fout = ftest\n fout.write(line)\n\n\ndef create_necessary_folders():\n if not os.path.exists(args.output_folder):\n os.makedirs(args.output_folder)\n\ndef _parse_args():\n parser = argparse.ArgumentParser()\n parser.add_argument(\"--aida_folder\", default=\"../data/basic_data/test_datasets/AIDA/\")\n parser.add_argument(\"--output_folder\", default=\"../data/new_datasets/\")\n return parser.parse_args()\n\nif __name__ == \"__main__\":\n args = _parse_args()\n create_necessary_folders()\n process_aida(args.aida_folder+\"aida_train.txt\", args.output_folder+\"aida_train.txt\")\n\n split_dev_test(args.aida_folder+\"testa_testb_aggregate_original\")\n process_aida(args.output_folder+\"temp_aida_dev\", args.output_folder+\"aida_dev.txt\")\n process_aida(args.output_folder+\"temp_aida_test\", args.output_folder+\"aida_test.txt\")\n\n os.remove(args.output_folder + \"temp_aida_dev\")\n os.remove(args.output_folder + \"temp_aida_test\")" }, { "path": "code/preprocessing/prepro_aida_tokenize.py", "content": "import argparse\nimport os\nimport preprocessing.util as util\nfrom subprocess import call\n\ndef process_aida(in_filepath, out_filepath):\n\n # _, wiki_id_name_map = util.load_wiki_name_id_map(lowercase=False)\n #_, wiki_id_name_map = util.entity_name_id_map_from_dump()\n entityNameIdMap = util.EntityNameIdMap()\n entityNameIdMap.init_compatible_ent_id()\n unknown_gt_ids = 0 # counter of ground truth entity ids that are not in the wiki_name_id.txt\n ent_id_changes = 0\n text_acc = []\n with open(in_filepath) as fin, open(args.output_folder+\"tokenize_\"+out_filepath, \"w\") as fout:\n in_mention = False # am i inside a mention span or not\n first_document = True\n for line in fin:\n l = line.split('\\t')\n if in_mention and not (len(l) == 7 and l[1]=='I'):\n # if I am in mention but the current line does not continue the previous mention\n # then print MMEND and be in state in_mention=FALSE\n #fout.write(\"MMEND\\n\")\n text_acc.append(\"MMEND\")\n in_mention = False\n\n if line.startswith(\"-DOCSTART-\"):\n if not first_document:\n #fout.write(\"DOCEND\\n\")\n text_acc.append(\"DOCEND\")\n # line = \"-DOCSTART- (967testa ATHLETICS)\\n\"\n doc_title = line[len(\"-DOCSTART- (\"): -2]\n #fout.write(\"DOCSTART_\"+doc_title.replace(' ', '_')+\"\\n\")\n text_acc.append(\"DOCSTART_\"+doc_title.replace(' ', '_'))\n first_document = False\n elif line == \"\\n\":\n #fout.write(\"*NL*\\n\")\n text_acc.append(\"\\n\")\n elif len(l) == 7 and l[1] == 'B': # this is a new mention\n wiki_title = l[4]\n wiki_title = wiki_title[len(\"http://en.wikipedia.org/wiki/\"):].replace('_', ' ')\n new_ent_id = entityNameIdMap.compatible_ent_id(wiki_title, l[5])\n if new_ent_id is not None:\n if new_ent_id != l[5]:\n ent_id_changes += 1\n #print(line, \"old ent_id: \" + l[5], \" new_ent_id: \", new_ent_id)\n #fout.write(\"MMSTART_\"+new_ent_id+\"\\n\") # TODO check here if entity id is inside my wikidump\n text_acc.append(\"MMSTART_\"+new_ent_id)\n # if not then omit this mention\n #fout.write(l[0]+\"\\n\") # write the word\n text_acc.append(l[0]) # write the word\n in_mention = True\n else:\n unknown_gt_ids += 1\n #fout.write(l[0]+\"\\n\") # write the word\n text_acc.append(l[0]) # write the word\n print(line)\n else:\n # words that continue a mention len(l) == 7: and l[1]=='I'\n # or normal word outside of mention, or in mention without disambiguation (len(l) == 4)\n #fout.write(l[0].rstrip()+\"\\n\")\n text_acc.append(l[0].rstrip())\n #fout.write(\"DOCEND\\n\") # for the last document\n text_acc.append(\"DOCEND\") # for the last document\n fout.write(' '.join(text_acc))\n print(\"process_aida unknown_gt_ids: \", unknown_gt_ids)\n print(\"process_aida ent_id_changes: \", ent_id_changes)\n print(\"now tokenize with stanford tokenizer\")\n tokenize_command = 'cd {}; java -cp \"*\" ' \\\n 'edu.stanford.nlp.process.PTBTokenizer -options \"tokenizeNLs=True\" < {} > {}'.format(\n args.stanford_tokenizer_folder, args.output_folder+\"tokenize_\"+out_filepath, args.output_folder+out_filepath)\n print(tokenize_command)\n call(tokenize_command, shell=True)\n\n\ndef split_dev_test(in_filepath):\n with open(in_filepath) as fin, open(args.output_folder+\"temp_aida_dev\", \"w\") as fdev,\\\n open(args.output_folder+\"temp_aida_test\", \"w\") as ftest:\n fout = fdev\n for line in fin:\n if line.startswith(\"-DOCSTART-\") and line.find(\"testb\") != -1:\n fout = ftest\n fout.write(line)\n\n\ndef create_necessary_folders():\n if not os.path.exists(args.output_folder):\n os.makedirs(args.output_folder)\n\ndef _parse_args():\n parser = argparse.ArgumentParser()\n parser.add_argument(\"--aida_folder\", default=\"/home/master_thesis_share/data/basic_data/test_datasets/AIDA/\")\n parser.add_argument(\"--output_folder\", default=\"/home/master_thesis_share/data/new_datasets/\")\n parser.add_argument(\"--stanford_tokenizer_folder\",\n default=\"/home/programs/stanford_core_nlp/stanford-corenlp-full-2017-06-09\")\n return parser.parse_args()\n\nif __name__ == \"__main__\":\n args = _parse_args()\n create_necessary_folders()\n process_aida(args.aida_folder+\"aida_train.txt\", \"aida_train.txt\")\n\n split_dev_test(args.aida_folder+\"testa_testb_aggregate_original\")\n process_aida(args.output_folder+\"temp_aida_dev\", \"aida_dev.txt\")\n process_aida(args.output_folder+\"temp_aida_test\", \"aida_test.txt\")\n\n os.remove(args.output_folder + \"temp_aida_dev\")\n os.remove(args.output_folder + \"temp_aida_test\")\n os.remove(args.output_folder + \"tokenize_aida_train.txt\")\n os.remove(args.output_folder + \"tokenize_aida_dev.txt\")\n os.remove(args.output_folder + \"tokenize_aida_test.txt\")\n" }, { "path": "code/preprocessing/prepro_gerbil_datasets.py", "content": "import argparse\nimport os\nimport preprocessing.util as util\nfrom nltk.tokenize import word_tokenize\n\n\nclass ProcessDataset(object):\n def __init__(self):\n self.entityNameIdMap = util.EntityNameIdMap()\n self.entityNameIdMap.init_gerbil_compatible_ent_id()\n self.unknown_ent_name = dict()\n self.no_english_uri = dict()\n self.all_gm_cnt = dict()\n self.englishuri_gm_cnt = dict()\n self.valid_gms = dict()\n\n def process(self, filepath):\n #the name of the dataset. just extract the last part of path\n dataset = os.path.basename(os.path.normpath(filepath))\n unknown_ent_name = 0\n no_english_uri = 0\n all_gm_cnt = 0\n englishuri_gm_cnt = 0 # has an english uri\n valid_gms = 0 # the ones that we use in the end\n with open(filepath, \"r\") as fin, open(args.output_folder+dataset+\".txt\", \"w\") as fout:\n for doc_name_line in fin:\n doc_name_line = \"DOCSTART_\" + \"\".join(filter(str.isalnum, doc_name_line.rstrip()[len(\"DOCSTART_\"):]))\n assert(doc_name_line.startswith(\"DOCSTART_\"))\n # accumulates the text of one document until we process all mentions then outputs to fout\n text = next(fin)\n text = text.rstrip()[len(\"text: \"):] # clear text\n if text == \"\":\n # dataset error. try to read next line. assert that next line is not annotations or docstart\n text = next(fin)\n gm_num = int(next(fin).rsplit(\" \", 1)[1])\n all_gm_cnt += gm_num\n annotations = []\n for _ in range(gm_num):\n # http://en.wikipedia.org/wiki/Fundraising\n #m_start, length, uri_list = next(fin).split(\", \", 2)\n #m_start = int(m_start[1:])\n m_start, length, uri_list = next(fin).split(\"_z_\", 2)\n m_start = int(m_start)\n m_end = m_start + int(length)\n #for uri in uri_list.rstrip()[1:-2].split(\", \"):\n for uri in uri_list.rstrip()[1:-1].split(\", \"):\n if uri.startswith(\"http://en.wikipedia.org\"):\n break\n if not uri.startswith(\"http://en.wikipedia.org\"):\n # skip this annotation since it doesn't have a link from english wikipedia\n no_english_uri += 1\n print(uri_list.rstrip())\n else:\n annotations.append((m_start, m_end, uri))\n annotations = sorted(annotations) # sort gm on start point\n englishuri_gm_cnt += len(annotations)\n\n text_acc = [doc_name_line.rstrip()]\n m_start_pr = 0 # previous mention start position\n m_end_pr = 0 # previous mention end position\n for (m_start, m_end, uri) in annotations:\n text_acc.append(text[m_end_pr:m_start])\n ent_id = self.entityNameIdMap.gerbil_compatible_ent_id(uri)\n if ent_id is not None:\n text_acc.append(\"MMSTART_\"+ent_id)\n text_acc.append(text[m_start:m_end])\n text_acc.append(\"MMEND\")\n valid_gms += 1\n else: # if not then omit this mention\n unknown_ent_name += 1\n text_acc.append(text[m_start:m_end])\n m_start_pr = m_start\n m_end_pr = m_end\n # add text after the last mention\n text_acc.append(text[m_end_pr:])\n text_acc.append(\"DOCEND\")\n doc_text = ' '.join(text_acc)\n doc_text = word_tokenize(doc_text)\n #doc_text = fix_tokenization(doc_text)\n doc_text = [word if word not in [\"``\", \"''\"] else '\"' for word in doc_text]\n fout.write('\\n'.join(doc_text) + \"\\n\")\n self.unknown_ent_name[dataset] = unknown_ent_name\n self.no_english_uri[dataset] = no_english_uri\n self.all_gm_cnt[dataset] = all_gm_cnt\n self.englishuri_gm_cnt[dataset] = englishuri_gm_cnt\n self.valid_gms[dataset] = valid_gms\n\n def process_readable(self, filepath):\n dataset = os.path.basename(os.path.normpath(filepath))\n with open(filepath, \"r\") as fin, open(args.output_folder+dataset+\".txt\", \"w\") as fout:\n for doc_name_line in fin:\n doc_name_line = \"DOCSTART_\" + \"\".join(filter(str.isalnum, doc_name_line.rstrip()[len(\"DOCSTART_\"):]))\n # accumulates the text of one document until we process all mentions then outputs to fout\n text = next(fin)\n text = text.rstrip()[len(\"text: \"):] # clear text\n if text == \"\":\n # dataset error. try to read next line. assert that next line is not annotations or docstart\n text = next(fin)\n gm_num = int(next(fin).rsplit(\" \", 1)[1])\n annotations = []\n for _ in range(gm_num):\n # http://en.wikipedia.org/wiki/Fundraising\n m_start, length, uri_list = next(fin).split(\"_z_\", 2)\n m_start = int(m_start)\n m_end = m_start + int(length)\n english_url_l = []\n for uri in uri_list.rstrip()[1:-1].split(\", \"):\n if uri.startswith(\"http://en.wikipedia.org\"):\n temp = \"http://en.wikipedia.org/wiki/\"\n assert(uri.startswith(temp))\n english_url_l.append(uri[len(temp):])\n if not english_url_l:\n # this annotation doesn't have a link from english wikipedia\n english_url_l.append(\"no wiki url\")\n annotations.append((m_start, m_end, english_url_l))\n annotations = sorted(annotations) # sort gm on start point\n\n text_acc = [doc_name_line.rstrip()]\n m_start_pr = 0 # previous mention start position\n m_end_pr = 0 # previous mention end position\n for cnt, (m_start, m_end, uri_l) in enumerate(annotations, 1):\n text_acc.append(text[m_end_pr:m_start])\n text_acc.append(\"[\"+str(cnt))\n text_acc.append(text[m_start:m_end])\n text_acc.append(\"]\")\n m_start_pr = m_start\n m_end_pr = m_end\n # add text after the last mention\n text_acc.append(text[m_end_pr:])\n doc_text = ' '.join(text_acc)\n fout.write(doc_text + \"\\n\")\n for cnt, (_, _, uri_l) in enumerate(annotations, 1):\n fout.write(str(cnt) + \": \" + str(uri_l) + \"\\n\")\n\ndef fix_tokenizatVion(doc_text):\n result = []\n for word in doc_text:\n if word in [\"``\", \"''\"]:\n result.append('\"')\n else:\n result.append(word)\n return result\n\n\ndef create_necessary_folders():\n if not os.path.exists(args.output_folder):\n os.makedirs(args.output_folder)\n\n\ndef _parse_args():\n parser = argparse.ArgumentParser()\n parser.add_argument(\"--other_datasets_folder\",\n default=\"/home/master_thesis_share/data/gerbil/gerbil_datasets/raw/\")\n parser.add_argument(\"--output_folder\", default=\"/home/master_thesis_share/data/gerbil/gerbil_datasets/readable_datasets/\")\n parser.add_argument(\"--human_readable_output\", type=bool, default=True)\n return parser.parse_args()\n\n\ndef get_immediate_files(a_dir):\n return [name for name in os.listdir(a_dir)\n if os.path.isfile(os.path.join(a_dir, name))]\n\n\ndef main():\n processDataset = ProcessDataset()\n for dataset in get_immediate_files(args.other_datasets_folder):\n if not args.human_readable_output:\n processDataset.process(os.path.join(args.other_datasets_folder, dataset))\n else:\n processDataset.process_readable(os.path.join(args.other_datasets_folder, dataset))\n print(\"Dataset\", dataset, \"done.\")\n print(\"processDataset.all_gm_cnt =\", processDataset.all_gm_cnt)\n print(\"processDataset.englishuri_gm_cnt =\", processDataset.englishuri_gm_cnt)\n print(\"processDataset.valid_gms =\", processDataset.valid_gms) # valid gold mentions used\n print(\"processDataset.no_english_uri =\", processDataset.no_english_uri)\n print(\"processDataset.unknown_ent_name =\", processDataset.unknown_ent_name)\n\n\nif __name__ == \"__main__\":\n args = _parse_args()\n create_necessary_folders()\n main()\n\n" }, { "path": "code/preprocessing/prepro_other_datasets.py", "content": "import argparse\nimport os\nimport preprocessing.util as util\nimport xml.etree.ElementTree as ET\nfrom subprocess import call\n\n\nclass ProcessDataset(object):\n def __init__(self):\n self.entityNameIdMap = util.EntityNameIdMap()\n self.entityNameIdMap.init_compatible_ent_id()\n\n def process(self, dataset_folder):\n #the name of the dataset. just extract the last part of path\n dataset = os.path.basename(os.path.normpath(dataset_folder))\n print(\"processing dataset:\", dataset)\n xml_filepath = os.path.join(dataset_folder, \"{}.xml\".format(dataset))\n rawtext_folder = os.path.join(dataset_folder, \"RawText\")\n if not (os.path.exists(rawtext_folder) and os.path.exists(xml_filepath)):\n print(\"Dataset \", dataset, \"is not in proper format.\")\n return\n tree = ET.parse(xml_filepath)\n root = tree.getroot()\n\n unknown_ent_name = 0\n with open(args.output_folder+\"temp_\"+dataset, \"w\") as fout:\n for document in root.findall('document'):\n docName = document.get('docName')\n with open(os.path.join(rawtext_folder, docName)) as fin:\n text = fin.read() # the whole raw text\n # accumulates the text of one document until we process all mentions then outputs to fout\n text_acc = []\n text_acc.append(\"DOCSTART_\"+docName.replace(' ', '_').replace('.', '_'))\n m_start_pr = 0 # previous mention start position\n m_end_pr = 0 # previous mention end position\n for annotation in document.findall('annotation'):\n m_start = int(annotation.find('offset').text) # inclusive\n m_end = m_start + int(annotation.find('length').text) # exclusive\n\n text_acc.append(text[m_end_pr:m_start])\n ent_id = self.entityNameIdMap.compatible_ent_id(\n name=annotation.find('wikiName').text\n )\n if ent_id is not None:\n text_acc.append(\"MMSTART_\"+ent_id)\n text_acc.append(text[m_start:m_end])\n text_acc.append(\"MMEND\")\n else: # if not then omit this mention\n unknown_ent_name += 1\n text_acc.append(text[m_start:m_end])\n print(\"unknown entity name: \", annotation.find('wikiName').text)\n m_start_pr = m_start\n m_end_pr = m_end\n # add text after the last mention\n text_acc.append(text[m_end_pr:])\n text_acc.append(\"DOCEND\\n\")\n fout.write(' '.join(text_acc))\n\n print(\"unknown_ent_name =\", unknown_ent_name)\n tokenize_command = 'cd {}; java -cp \"*\" '\\\n 'edu.stanford.nlp.process.PTBTokenizer -options \"tokenizeNLs=True\" < {} > {}'.format(\n args.stanford_tokenizer_folder, os.path.abspath(args.output_folder+\"temp_\"+dataset),\n os.path.abspath(args.output_folder+dataset+\".txt\"))\n print(tokenize_command)\n call(tokenize_command, shell=True)\n os.remove(args.output_folder+\"temp_\"+dataset)\n\n\n\ndef create_necessary_folders():\n if not os.path.exists(args.output_folder):\n os.makedirs(args.output_folder)\n\ndef _parse_args():\n parser = argparse.ArgumentParser()\n parser.add_argument(\"--other_datasets_folder\",\n default=\"../data/basic_data/test_datasets/wned-datasets/\")\n parser.add_argument(\"--output_folder\", default=\"../data/new_datasets/\")\n parser.add_argument(\"--stanford_tokenizer_folder\",\n default=\"../data/stanford_core_nlp/stanford-corenlp-full-2017-06-09\")\n return parser.parse_args()\n\ndef get_immediate_subdirectories(a_dir):\n return [name for name in os.listdir(a_dir)\n if os.path.isdir(os.path.join(a_dir, name))]\n\ndef main():\n processDataset = ProcessDataset()\n for dataset in get_immediate_subdirectories(args.other_datasets_folder):\n processDataset.process(os.path.join(args.other_datasets_folder, dataset))\n print(\"Dataset \", dataset, \"done.\")\n\nif __name__ == \"__main__\":\n args = _parse_args()\n create_necessary_folders()\n main()\n\n\n" }, { "path": "code/preprocessing/prepro_util.py", "content": "import pickle\nfrom collections import defaultdict, namedtuple\nimport numpy as np\nimport argparse\n\nimport os\nimport model.config as config\nimport preprocessing.util as util\nfrom termcolor import colored\nimport tensorflow as tf\n\n\nclass VocabularyCounter(object):\n \"\"\"counts the frequency of each word and each character in the corpus. With each\n file that it processes it increases the counters. So one frequency vocab for all the files\n that it processes.\"\"\"\n def __init__(self, lowercase_emb=False):\n import gensim\n self.model = gensim.models.KeyedVectors.load_word2vec_format(\n config.base_folder+\"data/basic_data/wordEmbeddings/Word2Vec/GoogleNews-vectors-negative300.bin\", binary=True)\n \"\"\"lowercase_emb=False if True then we lowercase the word for counting of\n frequencies and hence for finding the pretrained embedding.\"\"\"\n self.word_freq = defaultdict(int)\n self.char_freq = defaultdict(int) # how many times each character is encountered\n self.lowercase_emb = lowercase_emb\n self.not_in_word2vec_cnt = 0\n self.all_words_cnt = 0\n\n def add(self, filepath):\n \"\"\"the file must be in the new dataset format.\"\"\"\n with open(filepath) as fin:\n for line in fin:\n if line.startswith(\"DOCSTART_\") or line.startswith(\"DOCEND\") or\\\n line.startswith(\"MMSTART_\") or line.startswith(\"MMEND\") or \\\n line.startswith(\"*NL*\"):\n continue\n line = line.rstrip() # omit the '\\n' character\n word = line.lower() if self.lowercase_emb else line\n self.all_words_cnt += 1\n if word not in self.model:\n self.not_in_word2vec_cnt += 1\n else:\n self.word_freq[word] += 1\n for c in line:\n self.char_freq[c] += 1\n\n def print_statistics(self, word_edges=None,\n char_edges=None):\n \"\"\"Print some statistics about word and char frequency.\"\"\"\n if word_edges is None:\n word_edges = [1, 2, 3, 6, 11, 21, 31, 51, 76, 101, 201, np.inf]\n if char_edges is None:\n char_edges = [1, 6, 11, 21, 51, 101, 201, 501, 1001, 2001, np.inf]\n print(\"not_in_word2vec_cnt = \", self.not_in_word2vec_cnt)\n print(\"all_words_cnt = \", self.all_words_cnt)\n print(\"some frequency statistics. The bins are [...) \")\n for d, name, edges in zip([self.word_freq, self.char_freq], [\"word\", \"character\"], [word_edges, char_edges]):\n hist_values, _ = np.histogram(list(d.values()), edges)\n cum_sum = np.cumsum(hist_values[::-1])\n print(name, \" frequency histogram, edges: \", edges)\n print(\"absolute values: \", hist_values)\n print(\"absolute cumulative (right to left): \", cum_sum[::-1])\n print(\"probabilites cumulative (right to left):\", (cum_sum / np.sum(hist_values))[::-1])\n\n def serialize(self, folder=None, name=\"vocab_freq.pickle\"):\n if folder is None:\n folder = config.base_folder+\"data/vocabulary/\"\n if not os.path.exists(folder):\n os.makedirs(folder)\n with open(folder+name, 'wb') as handle:\n pickle.dump((self.word_freq, self.char_freq), handle)\n\n def count_datasets_vocabulary(self):\n new_dataset_folder = config.base_folder+\"data/new_datasets/\"\n \"\"\"\n datasets = ['aida_train.txt', 'aida_dev.txt', 'aida_test.txt', 'ace2004.txt',\n 'aquaint.txt', 'clueweb.txt', 'msnbc.txt', 'wikipedia.txt']\n \"\"\"\n for dataset in util.get_immediate_files(new_dataset_folder):\n dataset = os.path.basename(os.path.normpath(dataset))\n print(\"Processing dataset: \", dataset)\n self.add(new_dataset_folder+dataset)\n self.print_statistics()\n self.serialize(folder=config.base_folder+\"data/vocabulary/\",\n name=\"vocab_freq.pickle\")\n\n\ndef build_word_char_maps():\n output_folder = config.base_folder+\"data/tfrecords/\"+args.experiment_name+\"/\"\n if not os.path.exists(output_folder):\n os.makedirs(output_folder)\n with open(config.base_folder+\"data/vocabulary/vocab_freq.pickle\", 'rb') as handle:\n word_freq, char_freq = pickle.load(handle)\n word2id = dict()\n id2word = dict()\n char2id = dict()\n id2char = dict()\n\n wcnt = 0 # unknown word\n word2id[\"\"] = wcnt\n id2word[wcnt] = \"\"\n wcnt += 1\n ccnt = 0 # unknown character\n char2id[\"\"] = ccnt\n id2char[ccnt] = \"\"\n ccnt += 1\n\n # for every word in the corpus (we have already filtered out the words that are not in word2vec)\n for word in word_freq:\n if word_freq[word] >= args.word_freq_thr:\n word2id[word] = wcnt\n id2word[wcnt] = word\n wcnt += 1\n\n for c in char_freq:\n if char_freq[c] >= args.char_freq_thr:\n char2id[c] = ccnt\n id2char[ccnt] = c\n ccnt += 1\n assert(len(word2id) == wcnt)\n assert(len(char2id) == ccnt)\n print(\"words in vocabulary: \", wcnt)\n print(\"characters in vocabulary: \", ccnt)\n with open(output_folder+\"word_char_maps.pickle\", 'wb') as handle:\n pickle.dump((word2id, id2word, char2id, id2char, args.word_freq_thr,\n args.char_freq_thr), handle)\n\n import gensim\n model = gensim.models.KeyedVectors.load_word2vec_format(\n config.base_folder+\"data/basic_data/wordEmbeddings/Word2Vec/GoogleNews-vectors-negative300.bin\", binary=True)\n embedding_dim = len(model['queen'])\n embeddings_array = np.empty((wcnt, embedding_dim)) # id2emb\n embeddings_array[0] = np.zeros(embedding_dim)\n for i in range(1, wcnt):\n embeddings_array[i] = model[id2word[i]]\n np.save(output_folder+'embeddings_array.npy', embeddings_array)\n return word2id, char2id\n\n\ndef build_word_char_maps_restore():\n output_folder = config.base_folder+\"data/tfrecords/\"+args.experiment_name+\"/\"\n with open(output_folder+\"word_char_maps.pickle\", 'rb') as handle:\n word2id, _, char2id, _, _, _ = pickle.load(handle)\n return word2id, char2id\n\n\nclass Chunker(object):\n def __init__(self):\n self.separator = args.chunking\n self.chunk_ending = {'DOCEND'}\n if self.separator == \"per_paragraph\":\n self.chunk_ending.add('*NL*')\n if self.separator == \"per_sentence\":\n self.chunk_ending.add('.')\n self.chunk_ending.add('*NL*')\n self.parsing_errors = 0\n\n def new_chunk(self):\n self.chunk_words = []\n self.begin_gm = [] # the starting positions of gold mentions\n self.end_gm = [] # the end positions of gold mentions\n self.ground_truth = [] # list with the correct entity ids\n\n def compute_result(self, docid):\n chunk_id = docid\n if self.separator == \"per_paragraph\":\n chunk_id = chunk_id + \"&*\" + str(self.par_cnt)\n if self.separator == \"per_sentence\":\n chunk_id = chunk_id + \"&*\" + str(self.par_cnt) + \"&*\" + str(self.sent_cnt)\n result = (chunk_id, self.chunk_words, self.begin_gm, self.end_gm, self.ground_truth)\n\n # correctness checks. not necessary\n no_errors_flag = True\n if len(self.begin_gm) != len(self.end_gm) or \\\n len(self.begin_gm) != len(self.ground_truth):\n no_errors_flag = False\n for b, e in zip(self.begin_gm, self.end_gm):\n if e <= b or b >= len(self.chunk_words) or e > len(self.chunk_words):\n no_errors_flag = False\n\n self.new_chunk()\n if no_errors_flag == False:\n self.parsing_errors += 1\n print(\"chunker parse error: \", result)\n return None\n else:\n return result\n\n def process(self, filepath):\n with open(filepath) as fin:\n self.new_chunk()\n docid = \"\"\n # paragraph and sentence counter are not actually useful. only for debugging purposes.\n self.par_cnt = 0 # paragraph counter (useful if we work per paragraph)\n self.sent_cnt = 0 # sentence counter (useful if we work per sentence)\n for line in fin:\n line = line.rstrip() # omit the '\\n' character\n if line in self.chunk_ending:\n if len(self.chunk_words) > 0: # if we have continues *NL* *NL* do not return empty chunks\n temp = self.compute_result(docid)\n if temp is not None:\n yield temp\n # do not add the chunk separator, no use\n if line == '*NL*':\n self.par_cnt += 1\n self.sent_cnt = 0\n if line == '.':\n self.sent_cnt += 1\n elif line == '*NL*':\n self.par_cnt += 1\n self.sent_cnt = 0\n # do not add this in our words list\n elif line == '.':\n self.sent_cnt += 1\n self.chunk_words.append(line)\n elif line.startswith('MMSTART_'):\n ent_id = line[8:] # assert that ent_id in wiki_name_id_map\n self.ground_truth.append(ent_id)\n self.begin_gm.append(len(self.chunk_words))\n elif line == 'MMEND':\n self.end_gm.append(len(self.chunk_words))\n elif line.startswith('DOCSTART_'):\n docid = line[9:]\n self.par_cnt = 0\n self.sent_cnt = 0\n else:\n self.chunk_words.append(line)\n\n print(filepath, \" chunker parsing errors: \", self.parsing_errors)\n self.parsing_errors = 0\n\n\n\nGmonlySample = namedtuple(\"GmonlySample\",\n [\"chunk_id\", \"chunk_words\", 'begin_gm', \"end_gm\",\n \"ground_truth\", \"cand_entities\", \"cand_entities_scores\"])\nAllspansSample = namedtuple(\"AllspansSample\",\n [\"chunk_id\", \"chunk_words\", \"begin_spans\", \"end_spans\",\n \"ground_truth\", \"cand_entities\", \"cand_entities_scores\",\n \"begin_gm\", \"end_gm\"])\n\n\nclass SamplesGenerator(object):\n def __init__(self, mode=\"allspans\"):\n self.mode = mode\n self._generator = Chunker()\n self.fetchFilteredCoreferencedCandEntities = util.FetchFilteredCoreferencedCandEntities(args)\n self.all_gm_misses = 0\n self.all_gt_misses = 0\n self.all_gm = 0 # all the gm encountered in all the datasets\n\n def set_gmonly_mode(self):\n self.mode = \"gmonly\"\n\n def set_allspans_mode(self):\n self.mode = \"allspans\"\n\n def is_gmonly_mode(self):\n return True if self.mode == \"gmonly\" else False\n\n def is_allspans_mode(self):\n return True if self.mode == \"allspans\" else False\n\n def process(self, filepath):\n if self.is_allspans_mode():\n return self._process_allspans(filepath)\n else:\n return self._process_gmonly(filepath)\n\n def _process_allspans(self, filepath):\n gt_misses = 0\n gm_misses = 0\n gm_this_file = 0 # how many gold mentions are in this document - dataset. so we can find percentage for misses\n\n max_mention_width_violations = 0\n for chunk in self._generator.process(filepath):\n self.fetchFilteredCoreferencedCandEntities.init_coref(el_mode=True)\n begin_spans = []\n end_spans = []\n cand_entities = [] # list of lists candidate entities\n cand_entities_scores = []\n chunk_id, chunk_words, begin_gm, end_gm, ground_truth = chunk\n gm_this_file += len(begin_gm)\n for left, right in self.all_spans(chunk_words):\n cand_ent, scores = self.fetchFilteredCoreferencedCandEntities.process(left, right, chunk_words)\n if cand_ent is not None:\n begin_spans.append(left)\n end_spans.append(right)\n cand_entities.append(cand_ent)\n cand_entities_scores.append(scores)\n\n if args.calculate_stats:\n # check if gold mentions are inside the candidate spans and if yes check if ground truth is in cand ent.\n gm_spans = list(zip(begin_gm, end_gm)) # [(3, 5), (10, 11), (15, 18)]\n all_spans = list(zip(begin_spans, end_spans))\n for i, gm_span in enumerate(gm_spans):\n if gm_span not in all_spans:\n gm_misses += 1\n #print(\"gm not in spans\\t\\t\\t\", colored(' '.join(chunk_words[gm_span[0]:gm_span[1]]), 'red'))\n elif ground_truth[i] not in cand_entities[all_spans.index(gm_span)]:\n gt_misses += 1\n #print(\"gt not in cand ent\", colored(' '.join(chunk_words[gm_span[0]:gm_span[1]]), 'green'))\n #print(\"gt: \", ground_truth[i], \"cand_ent: \", cand_entities[all_spans.index(gm_span)])\n\n for b, e in zip(begin_gm, end_gm):\n if e - b > args.max_mention_width:\n max_mention_width_violations += 1\n\n if begin_spans: # there are candidate spans in the processed text\n yield AllspansSample(chunk_id, chunk_words, begin_spans, end_spans,\n ground_truth, cand_entities, cand_entities_scores,\n begin_gm, end_gm)\n if args.calculate_stats:\n print(\"max_mention_width_violations :\", max_mention_width_violations)\n print(\"gt_misses\", gt_misses)\n print(\"gm_misses\", gm_misses)\n print(\"gm_this_file: \", gm_this_file)\n print(\"recall % : \", (1 - (gm_misses+gt_misses)/gm_this_file)*100, \" %\")\n self.all_gt_misses += gt_misses\n self.all_gm_misses += gm_misses\n self.all_gm += gm_this_file\n\n @staticmethod\n def all_spans(chunk_words):\n # this function produces all possible text spans that do not include spans separators (fullstops).\n # divide the list of words to lists of lists based on spans_separator.\n # e.g. if chunk_words is for the whole document divide it to sentences (a list of\n # sentences) since no span extend above a fullstop.\n separation_indexes = []\n spans_separator = set(config.spans_separators)\n for idx, word in enumerate(chunk_words):\n if word in spans_separator:\n separation_indexes.append(idx)\n\n separation_indexes.append(len(chunk_words))\n\n def all_spans_aux(begin_idx, end_idx):\n for left_idx in range(begin_idx, end_idx):\n for length in range(1, args.max_mention_width + 1):\n if left_idx + length > end_idx:\n break\n yield left_idx, left_idx + length\n\n begin_idx = 0\n for end_idx in separation_indexes:\n for left, right in all_spans_aux(begin_idx, end_idx):\n # print(left, right, chunk_words[left:right])\n # print(left, right, ' '.join(chunk_words[left:right])\n yield left, right\n begin_idx = end_idx + 1\n\n def _process_gmonly(self, filepath):\n gt_misses = 0\n gm_misses = 0\n gm_this_file = 0\n max_mention_width_violations = 0\n for chunk in self._generator.process(filepath):\n self.fetchFilteredCoreferencedCandEntities.init_coref(el_mode=False)\n cand_entities = [] # list of lists candidate entities\n cand_entities_scores = []\n chunk_id, chunk_words, begin_gm, end_gm, ground_truth = chunk\n gm_this_file += len(begin_gm)\n for left, right, gt in zip(begin_gm, end_gm, ground_truth):\n cand_ent, scores = self.fetchFilteredCoreferencedCandEntities.process(left, right, chunk_words)\n if cand_ent is None:\n gm_misses += 1\n cand_ent, scores = [], []\n #print(\"gm not in p_e_m\\t\\t\\t\", colored(' '.join(chunk_words[left:right]), 'red'))\n elif args.calculate_stats and gt not in cand_ent:\n gt_misses += 1\n #print(\"gt not in cand ent\", colored(' '.join(chunk_words[left:right]), 'green'))\n #print(\"gt: \", gt, \"cand_ent: \", cand_ent)\n\n if right - left > args.max_mention_width:\n max_mention_width_violations += 1\n\n #print(' '.join(chunk_words[left:right])\n #print(cand_ent, scores)\n cand_entities.append(cand_ent)\n cand_entities_scores.append(scores)\n\n if begin_gm: #not emtpy\n yield GmonlySample(chunk_id, chunk_words, begin_gm, end_gm, ground_truth,\n cand_entities, cand_entities_scores)\n\n if args.calculate_stats:\n print(\"max_mention_width_violations :\", max_mention_width_violations)\n print(\"gt_misses\", gt_misses)\n print(\"gm_misses\", gm_misses)\n print(\"gm_this_file\", gm_this_file)\n print(\"recall % : \", (1 - (gm_misses+gt_misses)/gm_this_file)*100, \" %\")\n self.all_gt_misses += gt_misses\n self.all_gm_misses += gm_misses\n self.all_gm += gm_this_file\n\n\nSampleEncoded = namedtuple(\"SampleEncoded\",\n [\"chunk_id\",\n \"words\", 'words_len', # list, scalar\n 'chars', 'chars_len', # list of lists, list\n 'begin_spans', \"end_spans\", 'spans_len', # the first 2 are lists, last is scalar\n \"cand_entities\", \"cand_entities_scores\", 'cand_entities_labels', # lists of lists\n 'cand_entities_len', # list\n \"ground_truth\", \"ground_truth_len\",\n 'begin_gm', 'end_gm']) # list\n\n\nclass EncoderGenerator(object):\n \"\"\"receives samples Train or Test samples and encodes everything to numbers ready to\n be transformed to tfrecords. Also filters out candidate entities that are not in the\n entity universe.\"\"\"\n def __init__(self):\n self._generator = SamplesGenerator()\n self._word2id, self._char2id = build_word_char_maps()\n #self._word2id, self._char2id = build_word_char_maps_restore() # alternative\n self._wikiid2nnid = util.load_wikiid2nnid(args.entity_extension)\n\n def set_gmonly_mode(self):\n self._generator.set_gmonly_mode()\n\n def set_allspans_mode(self):\n self._generator.set_allspans_mode()\n\n def is_gmonly_mode(self):\n return self._generator.is_gmonly_mode()\n\n def is_allspans_mode(self):\n return self._generator.is_allspans_mode()\n\n def process(self, filepath):\n ground_truth_errors_cnt = 0\n cand_entities_not_in_universe_cnt = 0\n samples_with_errors = 0\n for sample in self._generator.process(filepath):\n words = []\n chars = []\n for word in sample.chunk_words:\n words.append(self._word2id[word] if word in self._word2id\n else self._word2id[\"\"])\n chars.append([self._char2id[c] if c in self._char2id else self._char2id[\"\"]\n for c in word])\n chars_len = [len(word) for word in chars]\n\n ground_truth_enc = [self._wikiid2nnid[gt] if gt in self._wikiid2nnid else self._wikiid2nnid[\"\"]\n for gt in sample.ground_truth]\n ground_truth_errors_cnt += ground_truth_enc.count(self._wikiid2nnid[\"\"]) # it is always zero\n\n #print(colored(\"New sample\", 'red'))\n #print(sample)\n if len(sample.begin_gm) != len(sample.end_gm) or \\\n len(sample.begin_gm) != len(ground_truth_enc):\n samples_with_errors += 1\n continue\n if isinstance(sample, GmonlySample):\n cand_entities, cand_entities_scores, cand_entities_labels, not_in_universe_cnt = \\\n self._encode_cand_entities_and_labels(\n sample.cand_entities, sample.cand_entities_scores, sample.ground_truth)\n\n yield SampleEncoded(chunk_id=sample.chunk_id,\n words=words, words_len=len(words),\n chars=chars, chars_len=chars_len,\n begin_spans=sample.begin_gm, end_spans=sample.end_gm, spans_len=len(sample.begin_gm),\n cand_entities=cand_entities, cand_entities_scores=cand_entities_scores,\n cand_entities_labels=cand_entities_labels,\n cand_entities_len=[len(t) for t in cand_entities],\n ground_truth=ground_truth_enc, ground_truth_len=len(sample.ground_truth),\n begin_gm=[], end_gm=[])\n\n elif isinstance(sample, AllspansSample):\n if len(sample.begin_spans) != len(sample.end_spans):\n samples_with_errors += 1\n continue\n # for each span i have the gt or the value -1 if this span is not a gm\n # and then i work in the same way as above\n span_ground_truth = []\n gm_spans = list(zip(sample.begin_gm, sample.end_gm)) # [(3, 5), (10, 11), (15, 18)]\n for left, right in zip(sample.begin_spans, sample.end_spans):\n if (left, right) in gm_spans:\n span_ground_truth.append(sample.ground_truth[gm_spans.index((left, right))])\n else:\n span_ground_truth.append(-1) # this span is not a gm\n cand_entities, cand_entities_scores, cand_entities_labels, not_in_universe_cnt = \\\n self._encode_cand_entities_and_labels(\n sample.cand_entities, sample.cand_entities_scores, span_ground_truth)\n\n yield SampleEncoded(chunk_id=sample.chunk_id,\n words=words, words_len=len(words),\n chars=chars, chars_len=chars_len,\n begin_spans=sample.begin_spans, end_spans=sample.end_spans, spans_len=len(sample.begin_spans),\n cand_entities=cand_entities, cand_entities_scores=cand_entities_scores,\n cand_entities_labels=cand_entities_labels,\n cand_entities_len=[len(t) for t in cand_entities],\n ground_truth=ground_truth_enc, ground_truth_len=len(sample.ground_truth),\n begin_gm=sample.begin_gm, end_gm=sample.end_gm)\n\n cand_entities_not_in_universe_cnt += not_in_universe_cnt\n print(\"ground_truth_errors_cnt =\", ground_truth_errors_cnt)\n print(\"cand_entities_not_in_universe_cnt =\", cand_entities_not_in_universe_cnt)\n print(\"encoder samples_with_errors =\", samples_with_errors)\n\n\n\n def _encode_cand_entities_and_labels(self, cand_entities_p, cand_entities_scores_p,\n ground_truth_p):\n \"\"\"receives cand_entities (list of lists), and ground_truth (list) and does the following:\n 1) removes cand ent that are not in our universe\n 2) creates a label 0, 1 if this candidate is correct or not (i.e. if the span is indeed a\n gold mention (row of candidate entities array) and this specific candidate entity (column\n of candidate entities array) is correct. Returns the filtered cand_entities\n and the corresponding label (they have the same shape)\"\"\"\n cand_entities = []\n cand_entities_scores = []\n cand_entities_labels = []\n not_in_universe_cnt = 0\n for cand_ent_l, cand_scores_l, gt in zip(cand_entities_p, cand_entities_scores_p, ground_truth_p):\n ent_l = []\n score_l = []\n label_l = []\n for cand_ent, score in zip(cand_ent_l, cand_scores_l):\n if cand_ent in self._wikiid2nnid: # else continue, this entity not in our universe\n ent_l.append(self._wikiid2nnid[cand_ent])\n score_l.append(score)\n label_l.append(1 if cand_ent == gt else 0)\n else:\n not_in_universe_cnt += 1\n cand_entities.append(ent_l)\n cand_entities_scores.append(score_l)\n cand_entities_labels.append(label_l)\n return cand_entities, cand_entities_scores, cand_entities_labels, not_in_universe_cnt\n\n\nclass TFRecordsGenerator(object):\n def __init__(self):\n self._generator = EncoderGenerator()\n\n def set_gmonly_mode(self):\n self._generator.set_gmonly_mode()\n\n def set_allspans_mode(self):\n self._generator.set_allspans_mode()\n\n def is_gmonly_mode(self):\n return self._generator.is_gmonly_mode()\n\n def is_allspans_mode(self):\n return self._generator.is_allspans_mode()\n\n @staticmethod\n def _to_sequence_example(sample):\n def _bytes_feature(value):\n return tf.train.Feature(bytes_list=tf.train.BytesList(value=[value]))\n # Those two create a simple feature. The first a simple feature with one integer, whereas the second a simple\n # list of integers as one feature.\n def _int64_feature(value):\n \"\"\"value is a simple integer.\"\"\"\n return tf.train.Feature(int64_list=tf.train.Int64List(value=[value]))\n\n def _int64list_feature(value):\n \"\"\"value is a list of integers.\"\"\"\n return tf.train.Feature(int64_list=tf.train.Int64List(value=value))\n\n def _int64_feature_list(values):\n \"\"\" values is a list of integers like the words (words = [2,4,6,8,10])\n a feature list where each feature has only one number (a list with fixed\n number of elements, specifically only one)\"\"\"\n return tf.train.FeatureList(feature=[_int64_feature(v) for v in values])\n\n def _int64list_feature_list(values):\n \"\"\" like the chars = [[1,2,3], [4,5], [6], [7,8], [9,10,11,12]] a feature list where each feature can have variable\n number of ements\"\"\"\n return tf.train.FeatureList(feature=[_int64list_feature(v) for v in values])\n\n def _floatlist_feature_list(values):\n \"\"\" like the chars = [[0.1,0.2,0.3], [0.4,0.5]] a feature list where each feature can have variable\n number of ements\"\"\"\n def _floatlist_feature(value):\n \"\"\"value is a list of integers.\"\"\"\n return tf.train.Feature(float_list=tf.train.FloatList(value=value))\n return tf.train.FeatureList(feature=[_floatlist_feature(v) for v in values])\n\n context = tf.train.Features(feature={\n \"chunk_id\": _bytes_feature(sample.chunk_id.encode('utf-8')),\n \"words_len\": _int64_feature(sample.words_len),\n \"spans_len\": _int64_feature(sample.spans_len),\n \"ground_truth_len\": _int64_feature(sample.ground_truth_len)\n })\n feature_list = {\n \"words\": _int64_feature_list(sample.words),\n \"chars\": _int64list_feature_list(sample.chars),\n \"chars_len\": _int64_feature_list(sample.chars_len),\n \"begin_span\": _int64_feature_list(sample.begin_spans),\n \"end_span\": _int64_feature_list(sample.end_spans),\n \"cand_entities\": _int64list_feature_list(sample.cand_entities),\n \"cand_entities_scores\": _floatlist_feature_list(sample.cand_entities_scores),\n \"cand_entities_labels\": _int64list_feature_list(sample.cand_entities_labels),\n \"cand_entities_len\": _int64_feature_list(sample.cand_entities_len),\n \"ground_truth\": _int64_feature_list(sample.ground_truth)\n }\n if isinstance(sample, SampleEncoded):\n feature_list[\"begin_gm\"] = _int64_feature_list(sample.begin_gm)\n feature_list[\"end_gm\"] = _int64_feature_list(sample.end_gm)\n feature_lists = tf.train.FeatureLists(feature_list=feature_list)\n\n sequence_example = tf.train.SequenceExample(context=context, feature_lists=feature_lists)\n return sequence_example\n\n\n def process(self, filepath):\n print(\"processing file: \", filepath)\n #the name of the dataset. just extract the last part of path\n filename = os.path.basename(os.path.normpath(filepath))[:-4] # omit the '.txt'\n output_folder = config.base_folder+\"data/tfrecords/\"+args.experiment_name+\"/\"\n output_folder += \"gmonly/\" if self.is_gmonly_mode() else \"allspans/\"\n if not os.path.exists(output_folder):\n os.makedirs(output_folder)\n writer = tf.python_io.TFRecordWriter(output_folder+filename)\n records_cnt = 0\n for sample in self._generator.process(filepath):\n #print(sample)\n sequence_example = self._to_sequence_example(sample)\n # write it to file\n if sequence_example is not None:\n writer.write(sequence_example.SerializeToString())\n records_cnt += 1\n writer.close()\n print(\"records_cnt = \", records_cnt)\n\n\ndef create_tfrecords():\n new_dataset_folder = config.base_folder+\"data/new_datasets/\"\n datasets = [os.path.basename(os.path.normpath(d)) for d in util.get_immediate_files(new_dataset_folder)]\n print(\"datasets: \", datasets)\n\n tfrecords_generator = TFRecordsGenerator()\n tfrecords_generator.set_gmonly_mode()\n for file in datasets:\n tfrecords_generator.process(filepath=new_dataset_folder+file)\n tfrecords_generator.set_allspans_mode()\n for file in datasets:\n tfrecords_generator.process(filepath=new_dataset_folder+file)\n\n\nclass PrintSamples(object):\n def __init__(self, only_misses=True):\n _, self.wiki_id_name_map = util.load_wiki_name_id_map()\n self.only_misses = only_misses\n\n def print_candidates(self, ent_ids_list):\n \"\"\"takes as input a list of ent_id and returns a string. This string has each ent_id\n together with the corresponding name (in the name withspaces are replaced by underscore)\n and candidates are separated with a single space. e.g. ent_id,Barack_Obama ent_id2,US_President\"\"\"\n acc = []\n for ent_id in ent_ids_list:\n acc.append(ent_id + \",\" + self.wiki_id_name_map[ent_id].replace(' ', '_'))\n return ' '.join(acc)\n\n def print_sample(self, sample):\n chunk_words, begin_gm, end_gm, ground_truth, cand_entities = \\\n sample.chunk_words, sample.begin_gm, sample.end_gm, sample.ground_truth, sample.cand_entities\n if isinstance(sample, GmonlySample):\n misses_idx = []\n for i, (gt, cand_ent) in enumerate(zip(ground_truth, cand_entities)):\n if gt not in cand_ent:\n misses_idx.append(i) # miss detected\n\n if self.only_misses and misses_idx:\n print(colored(\"New sample\", 'red'))\n print(' '.join(chunk_words))\n for i in misses_idx:\n message = ' '.join(chunk_words[begin_gm[i]:end_gm[i]]) + \"\\tgt=\" + \\\n self.print_candidates([ground_truth[i]]) + \\\n \"\\tCandidates: \" + self.print_candidates(cand_entities[i])\n print(colored(message, 'yellow'))\n if self.only_misses == False:\n print(colored(\"New sample\", 'red'))\n print(' '.join(chunk_words))\n for i in range(len(begin_gm)):\n message = ' '.join(chunk_words[begin_gm[i]:end_gm[i]]) + \"\\tgt=\" + \\\n self.print_candidates([ground_truth[i]]) + \\\n \"\\tCandidates: \" + self.print_candidates(cand_entities[i])\n print(colored(message, 'yellow' if i in misses_idx else 'white'))\n elif isinstance(sample, AllspansSample):\n begin_spans, end_spans = sample.begin_spans, sample.end_spans\n gm_spans = list(zip(begin_gm, end_gm)) # [(3, 5), (10, 11), (15, 18)]\n all_spans = list(zip(begin_spans, end_spans))\n print(colored(\"New sample\", 'red'))\n print(' '.join(chunk_words))\n for i, gm_span in enumerate(gm_spans):\n if gm_span not in all_spans:\n message = ' '.join(chunk_words[begin_gm[i]:end_gm[i]]) + \"\\tgt=\" + \\\n self.print_candidates([ground_truth[i]]) + \"\\tgm_miss\"\n print(colored(message, 'magenta'))\n elif ground_truth[i] not in cand_entities[all_spans.index(gm_span)]:\n message = ' '.join(chunk_words[begin_gm[i]:end_gm[i]]) + \"\\tgt=\" + \\\n self.print_candidates([ground_truth[i]]) + \"\\tgt_miss Candidates: \" + \\\n self.print_candidates(cand_entities[all_spans.index(gm_span)])\n print(colored(message, 'yellow'))\n\n if self.only_misses == False:\n # then also print all the spans and their candidate entities\n for left, right, cand_ent in zip(begin_spans, end_spans, cand_entities):\n # if span is a mention and includes gt then green color, otherwise white\n if (left, right) in gm_spans and ground_truth[gm_spans.index((left, right))] in cand_ent:\n message = ' '.join(chunk_words[left:right]) + \"\\tgt=\" + \\\n self.print_candidates([ground_truth[gm_spans.index((left, right))]]) + \\\n \"\\tgm_gt_hit Candidates: \" + \\\n self.print_candidates(cand_ent)\n print(colored(message, 'green'))\n else:\n message = ' '.join(chunk_words[left:right]) + \\\n \"\\t not a mention Candidates: \" + \\\n self.print_candidates(cand_ent)\n print(colored(message, 'white'))\n\n\ndef create_entity_universe(gmonly_files=None, allspans_files=None, printSamples=None):\n new_dataset_folder = config.base_folder+\"data/new_datasets/\"\n if gmonly_files is None:\n gmonly_files = []\n if allspans_files is None:\n allspans_files = ['aida_train.txt', 'aida_dev.txt', 'aida_test.txt', 'ace2004.txt',\n 'aquaint.txt', 'clueweb.txt', 'msnbc.txt', 'wikipedia.txt']\n print(\"gmonly_files: \", gmonly_files)\n print(\"allspans_files: \", allspans_files)\n\n def create_entity_universe_aux(generator, datasets):\n entities_universe = set()\n for dataset in datasets:\n print(\"Processing dataset: \", dataset)\n for sample in generator.process(filepath=new_dataset_folder+dataset):\n entities_universe.update(*sample.cand_entities)\n entities_universe.update(sample.ground_truth)\n if printSamples:\n printSamples.print_sample(sample)\n\n print(\"Overall statistics: \")\n print(\"all_gm_misses: \", generator.all_gm_misses)\n print(\"all_gt_misses: \", generator.all_gt_misses)\n print(\"all_gm: \", generator.all_gm)\n print(\"recall % : \", (1 - (generator.all_gm_misses+generator.all_gt_misses)/generator.all_gm)*100, \" %\")\n print(\"len(entities_universe):\\t\\t\\t\", colored(len(entities_universe), 'red'))\n return entities_universe\n\n gmonly_entities, allspans_entities = set(), set()\n samplesGenerator = SamplesGenerator()\n if gmonly_files:\n print(\"gmonly files statistics: \")\n samplesGenerator.set_gmonly_mode()\n gmonly_entities = create_entity_universe_aux(samplesGenerator, gmonly_files)\n if allspans_files:\n print(\"Test files statistics: \")\n samplesGenerator.set_allspans_mode()\n allspans_entities = create_entity_universe_aux(samplesGenerator, allspans_files)\n\n all_entities = gmonly_entities | allspans_entities\n print(\"len(all_entities) = \", len(all_entities))\n\n # print the entities of our universe to a file together with the name\n with open(config.base_folder+\"data/entities/entities_universe.txt\", \"w\") as fout:\n _, wiki_id_name_map = util.load_wiki_name_id_map()\n for ent_id in all_entities:\n fout.write(ent_id + \"\\t\" + wiki_id_name_map[ent_id].replace(' ', '_') + \"\\n\")\n\n return all_entities\n\n\ndef create_necessary_folders():\n if not os.path.exists(config.base_folder+\"data/tfrecords/\"):\n os.makedirs(config.base_folder+\"data/tfrecords/\")\n\n\ndef _parse_args():\n parser = argparse.ArgumentParser()\n parser.add_argument(\"--chunking\", default=\"per_document\",\n help=\"per_sentence or per_paragraph or per_article\"\n \"per_document: each document is processed as one example\"\n \"per_paragraph: each paragraph is processed as a separate example\")\n\n parser.add_argument(\"--p_e_m_choice\", default=\"yago\",\n help=\"'wiki' p(e|m) constructed only from wikipedia dump (prob_wikipedia_p_e_m.txt file),\\\n 'crosswiki' constructed from wikipedia dump + crosswiki (prob_crosswikis_wikipedia_p_e_m.txt),\\\n 'yago' (prob_yago_crosswikis_wikipedia_p_e_m.txt)\")\n parser.add_argument(\"--cand_ent_num\", type=int, default=30,\n help=\"how many candidate entities to keep for each mention\")\n parser.add_argument(\"--lowercase_p_e_m\", type=bool, default=False)\n parser.add_argument(\"--lowercase_spans\", type=bool, default=False)\n parser.add_argument(\"--calculate_stats\", type=bool, default=True)\n\n parser.add_argument(\"--experiment_name\", default=\"corefmerge\",\n help=\"under folder data/tfrecords/\")\n parser.add_argument(\"--include_wikidumpRLTD\", type=bool, default=False)\n parser.add_argument(\"--word_freq_thr\", type=int, default=1,\n help=\"words that have freq less than this are not included in our\"\n \"vocabulary.\")\n parser.add_argument(\"--char_freq_thr\", type=int, default=1)\n\n parser.add_argument(\"--max_mention_width\", type=int, default=10, help=\"in allspans mode consider all spans with\"\n \"length <= to this value as candidate entities to be linked\")\n parser.add_argument(\"--entity_extension\", default=None, help=\"extension_entities or extension_entities_all etc\")\n parser.add_argument(\"--persons_coreference\", type=bool, default=True)\n parser.add_argument(\"--persons_coreference_merge\", type=bool, default=True)\n parser.add_argument(\"--create_entity_universe\", type=bool, default=False)\n return parser.parse_args()\n\n\ndef log_args(folderpath):\n if not os.path.exists(folderpath):\n os.makedirs(folderpath)\n with open(folderpath+\"prepro_args.txt\", \"w\") as fout:\n attrs = vars(args)\n fout.write('\\n'.join(\"%s: %s\" % item for item in attrs.items()))\n with open(folderpath+\"prepro_args.pickle\", 'wb') as handle:\n pickle.dump(args, handle)\n\n\nif __name__ == \"__main__\":\n args = _parse_args()\n print(args)\n create_necessary_folders()\n log_args(config.base_folder+\"data/tfrecords/\"+args.experiment_name+\"/\")\n\n vocabularyCounter = VocabularyCounter()\n vocabularyCounter.count_datasets_vocabulary()\n if args.create_entity_universe:\n create_entity_universe(gmonly_files=[], allspans_files=['aida_train.txt', 'aida_dev.txt', 'aida_test.txt' # ])\n , 'ace2004.txt', 'aquaint.txt', 'msnbc.txt'])\n else:\n create_tfrecords()\n\n\n\n\n\n\n\n\n\n\n\n\n\n" }, { "path": "code/preprocessing/prepro_wikidump.py", "content": "import argparse\nimport os\nimport sys\nimport preprocessing.util as util\nimport preprocessing.config as config\nimport traceback\n\n\ndef wikidump_to_new_format():\n doc_cnt = 0\n hyperlink2EntityId = util.EntityNameIdMap()\n hyperlink2EntityId.init_hyperlink2id()\n if args.debug:\n infilepath = config.base_folder + \"data/mydata/tokenized_toy_wiki_dump2.txt\"\n outfilepath = args.out_folder+\"toy_wikidump.txt\"\n else:\n infilepath = config.base_folder+\"data/basic_data/tokenizedWiki.txt\"\n outfilepath = args.out_folder+\"wikidump.txt\"\n with open(infilepath) as fin,\\\n open(outfilepath, \"w\") as fout:\n in_mention = False\n for line in fin:\n line = line.rstrip() # omit the '\\n' character\n if line.startswith('':\n fout.write(\"DOCEND\\n\")\n doc_cnt += 1\n if doc_cnt % 5000 == 0:\n print(\"document counter: \", doc_cnt)\n elif line == '':\n if in_mention:\n fout.write(\"MMEND\\n\")\n in_mention = False\n else:\n fout.write(line+\"\\n\")\n\n\ndef subset_wikidump_only_relevant_mentions():\n # consider only the RLTD entities (484048). take them from the files\n entities_universe = set()\n with open(\"/home/other_projects/deep_ed/data/generated/nick/\"\n \"wikiid2nnid.txt\") as fin:\n for line in fin:\n ent_id = line.split('\\t')[0]\n entities_universe.add(ent_id)\n\n # filter wikidump\n doc_cnt = 0\n mention_errors = 0\n if args.debug:\n infilepath = args.out_folder+\"toy_wikidump.txt\"\n outfilepath = args.out_folder+\"toy_wikidumpRLTD.txt\"\n else:\n infilepath = args.out_folder+\"wikidump.txt\"\n outfilepath = args.out_folder+\"wikidumpRLTD.txt\"\n with open(infilepath) as fin, open(outfilepath, \"w\") as fout:\n in_mention_acc = []\n for line in fin:\n if line.startswith('DOCSTART_'):\n document_acc = [line]\n paragraph_acc = []\n paragraph_relevant = False\n in_mention_acc = []\n elif line == '*NL*\\n': # the or not necessary.\n # there is always a *NL* before DOCEND\n # end of paragraph so check if relevant\n if in_mention_acc:\n in_mention_acc.append(line)\n else:\n paragraph_acc.append(line) # normal word\n if in_mention_acc: # we have a parsing error resulting to enter a mention but\n #print(\"in_mention_acc\", in_mention_acc)\n mention_errors += 1 # never detecting the end of it so still in_mention\n paragraph_acc.extend(in_mention_acc[1:])\n #print(\"paragraph_acc\", paragraph_acc)\n in_mention_acc = []\n if paragraph_relevant:\n try:\n assert(len(paragraph_acc) >= 4) # MMSTART, word, MMEND *NL*orDOCEND\n assert(len(document_acc) >= 1)\n document_acc.extend(paragraph_acc)\n except AssertionError:\n _, _, tb = sys.exc_info()\n traceback.print_tb(tb) # Fixed format\n tb_info = traceback.extract_tb(tb)\n filename, line, func, text = tb_info[-1]\n print('An error occurred on line {} in statement {}'.format(line, text))\n print(\"in_mention_acc\", in_mention_acc)\n print(\"paragraph_acc\", paragraph_acc)\n\n\n paragraph_acc = []\n paragraph_relevant = False\n elif line == \"DOCEND\\n\":\n assert(in_mention_acc == []) # because there is always an *NL* before DOCEND\n if len(document_acc) > 1:\n document_acc.append(line)\n fout.write(''.join(document_acc))\n document_acc = [] # those 3 commands are not necessary\n paragraph_acc = []\n paragraph_relevant = False\n\n doc_cnt += 1\n if doc_cnt % 5000 == 0:\n print(\"document counter: \", doc_cnt)\n elif line.startswith('MMSTART_'):\n if in_mention_acc: # not a parsing error resulting to enter a mention but\n #print(\"in_mention_acc\", in_mention_acc)\n mention_errors += 1 # never detecting the end of it so still in_mention\n paragraph_acc.extend(in_mention_acc[1:])\n #print(\"paragraph_acc\", paragraph_acc)\n in_mention_acc = []\n ent_id = line.rstrip()[8:] # assert that ent_id in wiki_name_id_map\n if ent_id in entities_universe:\n paragraph_relevant = True\n in_mention_acc.append(line)\n elif line == 'MMEND\\n':\n if in_mention_acc:\n in_mention_acc.append(line)\n paragraph_acc.extend(in_mention_acc)\n in_mention_acc = []\n # else this mention is not in our universe so we don't accumulate it.\n else:\n if in_mention_acc:\n in_mention_acc.append(line)\n else:\n paragraph_acc.append(line) # normal word\n\n print(\"mention_errors =\", mention_errors)\n\n\n\n\ndef _parse_args():\n parser = argparse.ArgumentParser()\n parser.add_argument(\"--entities_universe_file\",\n default=\"/home/master_thesis_share/data/entities/entities_universe.txt\")\n parser.add_argument(\"--out_folder\", default=\"/home/master_thesis_share/data/new_datasets/wikidump/\")\n parser.add_argument(\"--debug\", type=bool, default=False)\n return parser.parse_args()\n\nif __name__ == \"__main__\":\n args = _parse_args()\n\n #if args.debug:\n # wikidump_to_new_format()\n\n subset_wikidump_only_relevant_mentions()\n" }, { "path": "code/preprocessing/util.py", "content": "import pickle\nfrom collections import defaultdict\nimport numpy as np\nimport time\nimport sys\n#print(\"preprocessing/util.py print sys.path\")\n#print(sys.path)\nimport os\nimport string\n\n\n# import preprocessing.config as config\nimport model.config as config\n\n# methods below are executed every time because they are fast or because their result\n# depend on the args\n\ndef load_entities_universe():\n entities_universe = set()\n # TODO this path is hardcoded. these mapping files should be transfered in ./data folder\n with open(\"/home/other_projects/deep_ed/data/generated/nick/\"\n \"wikiid2nnid.txt\") as fin:\n for line in fin:\n ent_id = line.split('\\t')[0]\n entities_universe.add(ent_id)\n return entities_universe\n\n\ndef load_wikiid2nnid(extension_name=None):\n \"\"\"returns a map from wiki id to neural network id (for the entity embeddings)\"\"\"\n wikiid2nnid = dict() # wikiid is string, nnid is integer\n with open(config.base_folder+\"data/entities/wikiid2nnid/wikiid2nnid.txt\") as fin:\n for line in fin:\n ent_id, nnid = line.split('\\t')\n wikiid2nnid[ent_id] = int(nnid) - 1 # torch starts from 1 instead of zero\n assert(wikiid2nnid[\"1\"] == 0)\n assert(-1 not in wikiid2nnid)\n wikiid2nnid[\"\"] = 0\n del wikiid2nnid[\"1\"]\n #print(len(wikiid2nnid))\n\n if extension_name:\n load_entity_extension(wikiid2nnid, extension_name)\n return wikiid2nnid\n\n\ndef load_entity_extension(wikiid2nnid, extension_name):\n filepath = config.base_folder + \"data/entities/\" + extension_name + \"/wikiid2nnid/additional_wikiids.txt\"\n max_nnid = max(wikiid2nnid.values())\n assert(len(wikiid2nnid) - 1 == max_nnid)\n with open(filepath) as fin:\n line_cnt = 1\n for line in fin:\n ent_id = line.strip()\n if ent_id in wikiid2nnid: # if extension entities has overlap with the normal entities set\n wikiid2nnid[ent_id + \"dupl\"] = max_nnid + line_cnt # this vector is duplicate and is never going to be used\n else:\n wikiid2nnid[ent_id] = max_nnid + line_cnt\n line_cnt += 1\n print(\"original entities: \", max_nnid + 1, \" extension entities: \", len(wikiid2nnid) - (max_nnid+1))\n\n\ndef reverse_dict(d, unique_values=False):\n new_d = dict()\n for k, v in d.items():\n if unique_values:\n assert(v not in new_d)\n new_d[v] = k\n return new_d\n\n#def p_e_m_disamb_redirect_wikinameid_maps():\ndef load_redirections(lowercase=None):\n if lowercase is None:\n lowercase = config.lowercase_maps\n wall_start = time.time()\n redirections = dict()\n with open(config.base_folder + \"data/basic_data/wiki_redirects.txt\") as fin:\n redirections_errors = 0\n for line in fin:\n line = line.rstrip()\n try:\n old_title, new_title = line.split(\"\\t\")\n if lowercase:\n old_title, new_title = old_title.lower(), new_title.lower()\n redirections[old_title] = new_title\n except ValueError:\n redirections_errors += 1\n\n print(\"load redirections. wall time:\", (time.time() - wall_start)/60, \" minutes\")\n print(\"redirections_errors: \", redirections_errors)\n return redirections\n\ndef load_disambiguations():\n wall_start = time.time()\n disambiguations_ids = set()\n #disambiguations_titles = set()\n disambiguations_errors = 0\n with open(config.base_folder + \"data/basic_data/wiki_disambiguation_pages.txt\") as fin:\n for line in fin:\n line = line.rstrip()\n try:\n article_id, title = line.split(\"\\t\")\n disambiguations_ids.add(article_id)\n #disambiguations_titles.add(title)\n except ValueError:\n disambiguations_errors += 1\n print(\"load disambiguations. wall time:\", (time.time() - wall_start)/60, \" minutes\")\n print(\"disambiguations_errors: \", disambiguations_errors)\n return disambiguations_ids\n\ndef load_persons():\n wiki_name_id_map, _ = load_wiki_name_id_map()\n persons_wikiids = set()\n not_found_cnt = 0\n with open(config.base_folder + \"data/basic_data/persons.txt\") as fin:\n for line in fin:\n line = line.strip()\n if line in wiki_name_id_map:\n persons_wikiids.add(wiki_name_id_map[line])\n else:\n not_found_cnt += 1\n #print(\"not found:\", repr(line))\n print(\"persons not_found_cnt:\", not_found_cnt)\n return persons_wikiids\n\ndef load_wiki_name_id_map(lowercase=False, filepath=None):\n wall_start = time.time()\n wiki_name_id_map = dict()\n wiki_id_name_map = dict()\n wiki_name_id_map_errors = 0\n duplicate_names = 0 # different lines in the doc with the same title\n duplicate_ids = 0 # with the same id\n if filepath is None:\n filepath = config.base_folder + \"data/basic_data/wiki_name_id_map.txt\"\n disambiguations_ids = load_disambiguations()\n with open(filepath) as fin:\n for line in fin:\n line = line.rstrip()\n try:\n wiki_title, wiki_id = line.split(\"\\t\")\n if wiki_id in disambiguations_ids:\n continue\n if lowercase:\n wiki_title = wiki_title.lower()\n\n if wiki_title in wiki_name_id_map:\n duplicate_names += 1\n if wiki_id in wiki_id_name_map:\n duplicate_ids += 1\n\n wiki_name_id_map[wiki_title] = wiki_id\n wiki_id_name_map[wiki_id] = wiki_title\n except ValueError:\n wiki_name_id_map_errors += 1\n print(\"load wiki_name_id_map. wall time:\", (time.time() - wall_start)/60, \" minutes\")\n print(\"wiki_name_id_map_errors: \", wiki_name_id_map_errors)\n print(\"duplicate names: \", duplicate_names)\n print(\"duplicate ids: \", duplicate_ids)\n return wiki_name_id_map, wiki_id_name_map\n\n\n\n\n\n\n\n\n\n\nclass FetchCandidateEntities(object):\n \"\"\"takes as input a string or a list of words and checks if it is inside p_e_m\n if yes it returns the candidate entities otherwise it returns None.\n it also checks if string.lower() inside p_e_m and if string.lower() inside p_e_m_low\"\"\"\n def __init__(self, args):\n self.lowercase_spans = args.lowercase_spans\n self.lowercase_p_e_m = args.lowercase_p_e_m\n self.p_e_m, self.p_e_m_low, self.mention_total_freq = custom_p_e_m(\n cand_ent_num=args.cand_ent_num,\n lowercase_p_e_m=args.lowercase_p_e_m)\n\n\n def process(self, span):\n \"\"\"span can be either a string or a list of words\"\"\"\n if isinstance(span, list):\n span = ' '.join(span)\n title = span.title()\n # 'obama 44th president of united states'.title() # 'Obama 44Th President Of United States'\n title_freq = self.mention_total_freq[title] if title in self.mention_total_freq else 0\n span_freq = self.mention_total_freq[span] if span in self.mention_total_freq else 0\n\n if title_freq == 0 and span_freq == 0:\n if self.lowercase_spans and span.lower() in self.p_e_m:\n return map(list, zip(*self.p_e_m[span.lower()]))\n elif self.lowercase_p_e_m and span.lower() in self.p_e_m_low:\n return map(list, zip(*self.p_e_m_low[span.lower()]))\n else:\n return None, None\n\n else:\n if span_freq > title_freq:\n return map(list, zip(*self.p_e_m[span]))\n else:\n return map(list, zip(*self.p_e_m[title]))\n\n # from [('ent1', 0.4), ('ent2', 0.3), ('ent3', 0.3)] to\n # ('ent1', 'ent2', 'ent3') and (0.4, 0.3, 0.3)\n # after map we have lists i.e. ['ent1', 'ent2', 'ent3'] , [0.4, 0.3, 0.3]\n\nclass FetchFilteredCoreferencedCandEntities(object):\n def __init__(self, args):\n self.args = args\n self.fetchCandidateEntities = FetchCandidateEntities(args)\n self.el_mode = True\n if args.persons_coreference:\n self.persons_wikiids = load_persons()\n self.persons_mentions_seen = list()\n\n def init_coref(self, el_mode):\n self.persons_mentions_seen = list()\n self.el_mode = el_mode\n\n def process(self, left, right, chunk_words):\n left_right_words = [chunk_words[left-1] if left - 1 >= 0 else None,\n chunk_words[right] if right <= len(chunk_words)-1 else None] if self.el_mode else None\n span_text = ' '.join(chunk_words[left:right])\n cand_ent, scores = self.fetchCandidateEntities.process(span_text)\n if self.args.persons_coreference:\n coreference_supermention = self.find_corefence_person(span_text, left_right_words)\n if coreference_supermention:\n #print(\"original text:\", chunk_words[max(0, left-4):min(len(chunk_words), right+4)])\n if not self.args.persons_coreference_merge:\n cand_ent, scores = self.fetchCandidateEntities.process(coreference_supermention)\n else: # merge with cand_ent and scores\n cand_ent2, scores2 = self.fetchCandidateEntities.process(coreference_supermention)\n temp1 = list(zip(scores, cand_ent)) if scores and cand_ent else []\n temp2 = list(zip(scores2, cand_ent2)) if scores2 and cand_ent2 else []\n temp3 = sorted(temp1 + temp2, reverse=True)\n scores, cand_ent = map(list, zip(*temp3[:self.args.cand_ent_num]))\n\n if cand_ent is not None and scores is not None:\n if self.args.persons_coreference and not coreference_supermention and \\\n cand_ent[0] in self.persons_wikiids and len(span_text) >= 3:\n if not self.el_mode or span_text == span_text.title() or span_text == string.capwords(span_text):\n self.persons_mentions_seen.append(span_text)\n return cand_ent, scores\n\n def find_corefence_person(self, span_text, left_right_words):\n \"\"\"if span_text is substring of another person's mention found before. it should be\n substring of words. so check next and previous characters to be non alphanumeric\"\"\"\n if len(span_text) < 3:\n return None\n if left_right_words: # this check is only for allspans mode not for gmonly.\n if left_right_words[0] and left_right_words[0][0].isupper() or \\\n left_right_words[1] and left_right_words[1][0].isupper():\n # if the left or the right word has uppercased its first letter then do not search for coreference\n # since most likely it is a subspan of a mention.\n # This condition gives no improvement to Gerbil results even a very slight decrease (0.02%)\n return None\n for mention in reversed(self.persons_mentions_seen):\n idx = mention.find(span_text)\n if idx != -1:\n if len(mention) == len(span_text):\n continue # they are identical so no point in substituting them\n if idx > 0 and mention[idx-1].isalpha():\n continue\n if idx + len(span_text) < len(mention) and mention[idx+len(span_text)].isalpha():\n continue\n #print(\"persons coreference, before:\", span_text, \"after:\", mention)\n return mention\n return None\n\n\nclass EntityNameIdMap(object):\n def __init__(self):\n pass\n\n def init_compatible_ent_id(self):\n self.wiki_name_id_map, self.wiki_id_name_map = load_wiki_name_id_map(lowercase=False)\n\n def init_gerbil_compatible_ent_id(self):\n self.wiki_name_id_map, self.wiki_id_name_map = load_wiki_name_id_map(lowercase=False)\n self.redirections = load_redirections(lowercase=False)\n\n def init_hyperlink2id(self):\n self.wiki_name_id_map, self.wiki_id_name_map = load_wiki_name_id_map(lowercase=False)\n self.wiki_name_id_map_l, _ = load_wiki_name_id_map(lowercase=True)\n self.redirections = load_redirections(lowercase=False)\n self.disambiguations = load_disambiguations()\n self.hyperlinks_to_dismabiguation_pages = 0\n\n def hyperlink2id(self, line):\n \"\"\" gets as input the raw line:\n \\n\n '\\n'\n \"\"\"\n #line = '\\n'\n line = line.rstrip()\n hyperlink_text = line[9:-2]\n #print(repr(hyperlink_text))\n hyperlink_text = hyperlink_text.replace('\\xa0', ' ').strip()\n\n for title in [hyperlink_text, hyperlink_text.title()]:\n #look for redirection\n if title in self.redirections:\n title = self.redirections[title]\n if title in self.wiki_name_id_map:\n return self.wiki_name_id_map[title]\n\n if hyperlink_text.lower() in self.wiki_name_id_map_l:\n return self.wiki_name_id_map_l[hyperlink_text.lower()]\n else:\n return config.unk_ent_id\n\n def is_valid_entity_id(self, ent_id):\n return ent_id in self.wiki_id_name_map\n\n def compatible_ent_id(self, name=None, ent_id=None):\n \"\"\"takes as input the name and the entity id found in the dataset. If the entity id\n is also in our wiki_name_id_map then this means that this concept-entity also exist in\n out world and with the same id. If the id is not found in our world then we search for\n the name if it is inside the wiki_name_id_map. if yes then we have the same concept\n in our world but with different id so we return and use our own id from now on.\n if neither the id nor the name is in wiki_name_id_map then we return None i.e.\n unknown concept so skip it from the dataset.\"\"\"\n if ent_id is not None and ent_id in self.wiki_id_name_map:\n return ent_id\n elif name is not None and name in self.wiki_name_id_map:\n return self.wiki_name_id_map[name]\n else:\n return None\n\n def gerbil_compatible_ent_id(self, uri):\n from urllib.parse import unquote\n title = unquote(uri)\n title = title[len(\"http://en.wikipedia.org/wiki/\"):].replace('_', ' ')\n if title in self.redirections:\n title = self.redirections[title]\n if title in self.wiki_name_id_map:\n return self.wiki_name_id_map[title]\n else:\n print(\"unknown entity. title_searched:\", repr(title), \" original uri:\", repr(uri))\n return None\n\n\n\n\n\ndef custom_p_e_m(cand_ent_num=15, allowed_entities_set=None,\n lowercase_p_e_m=False):\n \"\"\"Args:\n cand_ent_num: how many candidate entities to keep for each mention\n allowed_entities_set: restrict the candidate entities to only this set. for example\n the most frequent 1M entities. First this restiction applies and then the cand_ent_num.\"\"\"\n wall_start = time.time()\n p_e_m = dict() # for each mention we have a list of tuples (ent_id, score)\n mention_total_freq = dict() # for each mention of the p_e_m we store the total freq\n # this will help us decide which cand entities to take\n p_e_m_errors = 0\n entityNameIdMap = EntityNameIdMap()\n entityNameIdMap.init_compatible_ent_id()\n incompatible_ent_ids = 0\n with open(config.base_folder + 'data/basic_data/prob_yago_crosswikis_wikipedia_p_e_m.txt') as fin:\n duplicate_mentions_cnt = 0\n clear_conflict_winner = 0 # both higher absolute frequency and longer cand list\n not_clear_conflict_winner = 0 # higher absolute freq but shorter cand list\n for line in fin:\n line = line.rstrip()\n try:\n temp = line.split(\"\\t\")\n mention, entities = temp[0], temp[2:]\n absolute_freq = int(temp[1])\n res = []\n for e in entities:\n if len(res) >= cand_ent_num:\n break\n ent_id, score, _ = map(str.strip, e.split(',', 2))\n #print(ent_id, score)\n if not entityNameIdMap.is_valid_entity_id(ent_id):\n incompatible_ent_ids += 1\n elif allowed_entities_set is not None and \\\n ent_id not in allowed_entities_set:\n pass\n else:\n res.append((ent_id, float(score)))\n if res:\n if mention in p_e_m:\n duplicate_mentions_cnt += 1\n #print(\"duplicate mention: \", mention)\n if absolute_freq > mention_total_freq[mention]:\n if len(res) > len(p_e_m[mention]):\n clear_conflict_winner += 1\n else:\n not_clear_conflict_winner += 1\n p_e_m[mention] = res\n mention_total_freq[mention] = absolute_freq\n else:\n p_e_m[mention] = res # for each mention we have a list of tuples (ent_id, score)\n mention_total_freq[mention] = absolute_freq\n\n except Exception as esd:\n exc_type, exc_obj, exc_tb = sys.exc_info()\n fname = os.path.split(exc_tb.tb_frame.f_code.co_filename)[1]\n print(exc_type, fname, exc_tb.tb_lineno)\n p_e_m_errors += 1\n print(\"error in line: \", repr(line))\n\n print(\"duplicate_mentions_cnt: \", duplicate_mentions_cnt)\n print(\"end of p_e_m reading. wall time:\", (time.time() - wall_start)/60, \" minutes\")\n print(\"p_e_m_errors: \", p_e_m_errors)\n print(\"incompatible_ent_ids: \", incompatible_ent_ids)\n\n if not lowercase_p_e_m: # do not build lowercase dictionary\n return p_e_m, None, mention_total_freq\n\n wall_start = time.time()\n # two different p(e|m) mentions can be the same after lower() so we merge the two candidate\n # entities lists. But the two lists can have the same candidate entity with different score\n # we keep the highest score. For example if \"Obama\" mention gives 0.9 to entity Obama and\n # OBAMA gives 0.7 then we keep the 0.9 . Also we keep as before only the cand_ent_num entities\n # with the highest score\n p_e_m_lowercased = defaultdict(lambda: defaultdict(int))\n\n for mention, res in p_e_m.items():\n l_mention = mention.lower()\n # if l_mention != mention and l_mention not in p_e_m:\n # the same so do nothing already exist in dictionary\n # e.g. p(e|m) has Obama and obama. So when i convert Obama to lowercase\n # I find that obama already exist so i will prefer this.\n if l_mention not in p_e_m:\n for r in res:\n ent_id, score = r\n p_e_m_lowercased[l_mention][ent_id] = max(score, p_e_m_lowercased[l_mention][ent_id])\n\n print(\"end of p_e_m lowercase. wall time:\", (time.time() - wall_start)/60, \" minutes\")\n\n import operator\n p_e_m_lowercased_trim = dict()\n for mention, ent_score_map in p_e_m_lowercased.items():\n sorted_ = sorted(ent_score_map.items(), key=operator.itemgetter(1), reverse=True)\n p_e_m_lowercased_trim[mention] = sorted_[:cand_ent_num]\n\n return p_e_m, p_e_m_lowercased_trim, mention_total_freq\n\n\ndef get_immediate_files(a_dir):\n return [name for name in os.listdir(a_dir)\n if os.path.isfile(os.path.join(a_dir, name))]\n\n\nif __name__ == \"__main__\":\n #load_wikiid2nnid()\n load_persons()\n pass\n #pass\n # vocabulary_count_wiki()\n\n #entity_count_wiki()\n #entity_count_wiki_aux()\n\n # load_p_e_m()\n\n #entity_name_id_map_from_dump()\n #compare_name_id_maps()\n #test_wiki_name_id_map_txt_conflicts_when_lowering():\n\n" }, { "path": "code/script", "content": "time python -m preprocessing.prepro_util --p_e_m_choice=yago --cand_ent_num=5 \\\n --allowed_entities_set_top=1000000 --lowercase_spans=True --lowercase_p_e_m=True >> screen_output/screen_out_4\ntime python -m preprocessing.prepro_util --p_e_m_choice=yago --cand_ent_num=10 \\\n --allowed_entities_set_top=1000000 --lowercase_spans=True --lowercase_p_e_m=True >> screen_output/screen_out_4\ntime python -m preprocessing.prepro_util --p_e_m_choice=yago --cand_ent_num=15 \\\n --allowed_entities_set_top=1000000 --lowercase_spans=True --lowercase_p_e_m=True >> screen_output/screen_out_4\ntime python -m preprocessing.prepro_util --p_e_m_choice=yago --cand_ent_num=30 \\\n --allowed_entities_set_top=1000000 --lowercase_spans=True --lowercase_p_e_m=True >> screen_output/screen_out_4\ntime python -m preprocessing.prepro_util --p_e_m_choice=yago --cand_ent_num=40 \\\n --allowed_entities_set_top=1000000 --lowercase_spans=True --lowercase_p_e_m=True >> screen_output/screen_out_4\n\n\ntime python -m preprocessing.prepro_util --p_e_m_choice=crosswiki --cand_ent_num=5 \\\n --allowed_entities_set_top=1000000 --lowercase_spans=True --lowercase_p_e_m=True >> screen_output/screen_out_4\ntime python -m preprocessing.prepro_util --p_e_m_choice=crosswiki --cand_ent_num=10 \\\n --allowed_entities_set_top=1000000 --lowercase_spans=True --lowercase_p_e_m=True >> screen_output/screen_out_4\ntime python -m preprocessing.prepro_util --p_e_m_choice=crosswiki --cand_ent_num=15 \\\n --allowed_entities_set_top=1000000 --lowercase_spans=True --lowercase_p_e_m=True >> screen_output/screen_out_4\ntime python -m preprocessing.prepro_util --p_e_m_choice=crosswiki --cand_ent_num=30 \\\n --allowed_entities_set_top=1000000 --lowercase_spans=True --lowercase_p_e_m=True >> screen_output/screen_out_4\n\n\ntime python -m preprocessing.prepro_util --p_e_m_choice=wiki --cand_ent_num=5 \\\n --allowed_entities_set_top=1000000 --lowercase_spans=True --lowercase_p_e_m=True >> screen_output/screen_out_4\ntime python -m preprocessing.prepro_util --p_e_m_choice=wiki --cand_ent_num=10 \\\n --allowed_entities_set_top=1000000 --lowercase_spans=True --lowercase_p_e_m=True >> screen_output/screen_out_4\ntime python -m preprocessing.prepro_util --p_e_m_choice=wiki --cand_ent_num=15 \\\n --allowed_entities_set_top=1000000 --lowercase_spans=True --lowercase_p_e_m=True >> screen_output/screen_out_4\ntime python -m preprocessing.prepro_util --p_e_m_choice=wiki --cand_ent_num=30 \\\n --allowed_entities_set_top=1000000 --lowercase_spans=True --lowercase_p_e_m=True >> screen_output/screen_out_4\n\n\n" }, { "path": "create_entity_vectors.md", "content": "# Creating Entity Vectors\nAs it is already mentioned, we have created entity vectors for 502661 entities from many\ndifferent popular datasets. \n\n## Description\nThe paper that describes how the entity vectors are created and the corresponding code can be found \n[here](https://github.com/dalab/deep-ed). That code is written in Lua and Torch, so in the current project \nwe provide some wrapper code that helps in creating the entity vectors through Python and for Tensorflow \nand also assists in detecting which entity vectors we need to create for the corpus at hand. But you can\nalways directly use the original lua code and modify it for your needs.\n\nThis code is like glue code that connects this new project with an old one and also provides code\n for processing the different dataset's formats. So it is not something I would recommend to spend time\n on. Specifically, \nthe following functionality is provided:\n- The first step in creating the entity vectors is to decide which entities we want in our universe. \n The original Lua code considers only the candidate entities (from the p(e|m) dictionary) for the \n gold mention of the evaluation datasets (AIDA, ACE2004, AQUAINT, MSNBC, CLUEWEB). In this project, since\n we work on end to end entity linking we consider all possible text spans (and not only the gold mentions) \n and for each span we take the first 30 candidate entities (first in terms of p(e|m) score). However, since\n the available datasets are given in different formats we have written code for each one of them:\n - first prepro_aida.py and prepro_other_datasets.py is executed in order to transform all of them in a\n common and simplified format. Then we run the code``preprocessing.prepro_util --create_entity_universe=True`` \n which processes all these documents (considers all possible spans and for each span all candidate entities and returns back \n a set of wiki ids i.e. the wiki ids for which we must create entity vectors). Those wiki ids are\n stored in the file data/entities/entities_universe.txt\n - ``gerbil/server.py --build_entity_uninverse=True`` for the format of the all the gerbil datasets. \n The Gerbil platform reads the different datasets and sends the tested documents in a common format. \n This code accepts the documents sent by gerbil and does the same processing as described above i.e.\n considers all possible spans and for each span all candidate entities and returns back a set of wiki ids. Those wiki ids\n are stored in the file data/entities/extension_entities/extension_entities.txt\n\n So depending on the format of the corpus that we are interested in we ran the corresponding code. \n The gerbil format is actually simple text so you can use this format for arbitrary text, look section\n [\"Trying the system on random user input text\"](readme.md#Trying-the-system-on-random-user-input-text)\n \n It is advisable to merge the files entities_universe.txt and extension_entities.txt so that we execute the \n commands of the next section only once and all the entity vectors are trained together.\n \n **Alternative:** directly edit the file data/entities/entities_universe.txt and \n put there all the wiki ids for which we want entity vectors. \n \n## Commands to be executed for recreating all the entities used for the paper experiments\nCreating a set with the wikipedia ids of all the candidate entities from all spans for the\ndatasets: AIDA-TRAINING, AIDA_TEST-A, AIDA-TEST-B, ACE2004, AQUAINT, MSNBC (file entities_universe.txt is created)\n```\nEL_PATH=/home/end2end_neural_el\ncd end2end_neural_el/\npython -m preprocessing.prepro_util --create_entity_universe=True\n```\n\nAdditional entities for gerbil evaluation:\n\nNow we create another set of wikipedia ids necessary for the evaluation on all the other datasets of the Gerbil platform\n(DBpediaSpotlight, Derczynski, ERD2014, GERDAQ-Dev, GERDAQ-Test, GERDAQ-TrainingA, GERDAQ-TrainingB, KORE50, \nMicroposts2016-Dev, Microposts2016-Test, Microposts2016-Train, N3-RSS-500, N3-Reuters-128,\nOKE 2015 Task1, OKE 2016 Task1). \n\n\nDownload Gerbil from https://github.com/dice-group/gerbil Find the datasets that you are interested in\nand place them in the appropriate locations so that gerbil can use them (follow gerbil instructions).\nOn one terminal execute:\n```\ncd gerbil/ \n./start.sh\n```\nOn another terminal execute:\n```\ncd end2end_neural_el/gerbil-SpotWrapNifWS4Test/\nmvn clean -Dmaven.tomcat.port=1235 tomcat:run\n```\n\nOn a third terminal execute:\n```\ncd end2end_neural_el/code \npython -m gerbil.server --build_entity_uninverse=True\n```\nOpen the url http://localhost:1234/gerbil\n- Configure experiment\n- In URI field write: http://localhost:1235/gerbil-spotWrapNifWS4Test/myalgorithm\n- Name: whatever you wish\n- Press add annotator button\n- Select datasets that you wish to create entity vectors\n- Run experiment\n- Terminate with ctrl+C the third terminal (the one that runs the command: python -m gerbil.server)\n when all the datasets have been processed (you can see the results on the gerbil web-page)\n Here instead of replying we just process the text the gerbil sends to extract the\n candidate entities and we always return an empty response. So the scores will be zero.\n(file extension_entities.txt has been created). \n\nWe can now merge the files entities_universe.txt and extension_entities.txt to simplify the process and train\nthem all together. Alternatively, we can train the two sets of entity vectors separately and use the extension entities\non the fly only when needed for the evaluation on the specific datasets.\n\nNow create the entity vectors:\n-\tFollow instructions from https://github.com/dalab/deep-ed\n- the code of this project is under end2end_neural_el/deep-ed/\n\tThere are only some minor changes at the deep-ed-master/entities/relatedness/relatedness.lua\n\tto read our file of entities (the entities for whom we want to create the entity vectors). \n\tReplace lines 277-278 with 280-281 for the extension entities.\n\t \n\n\nAfter you follow the instructions from https://github.com/dalab/deep-ed do the following in \norder to get the entity vectors from the torch-lua framework to python-tensorflow\n- pick the epoch that has the best scores and copy this file to our project\n```\ncp $EL_PATH/deep-ed/data/generated/ent_vecs/ent_vecs__ep_147.t7 $EL_PATH/data/entities/ent_vecs/ \n```\n- converts the file to simple txt format\n```\nend2end_neural_el/code$ th preprocessing/bridge_code_lua/ent_vecs_to_txt.lua -ent_vecs_filepath /home/end2end_neural_el/data/entities/ent_vecs/ent_vecs__ep_147.t7\n```\n- and then to numpy array ready to be used from tensorflow\n```\nend2end_neural_el/code$ python -m preprocessing.bridge_code_lua.ent_vecs_from_txt_to_npy\n```\n\nNow the same for the extension entities (optional steps):\n- Replace lines 277-278 with 280-281 in relatedness.lua file.\n- Delete the t7 files, otherwise the changed relatedness.lua script will not be executed \n(in order to work on the extension entities this time) \n```\nrm end2end_neual_el/deep-ed/data/generated/*t7\n```\n- Repeat the steps of https://github.com/dalab/deep-ed starting from step 13 (no need to repeat the previous ones)\n- After the training is finished execute the following commands (similar to before)\n```\nmkdir -p $EL_PATH/data/entities/extension_entities/ent_vecs/\ncp $EL_PATH/deep-ed/data/generated/ent_vecs/ent_vecs__ep_54.t7 $EL_PATH/data/entities/extension_entities/ent_vecs/\nend2end_neural_el/code$ th preprocessing/bridge_code_lua/ent_vecs_to_txt.lua -ent_vecs_filepath /home/end2end_neural_el/data/entities/extension_entities/ent_vecs/ent_vecs__ep_54.t7 \nend2end_neural_el/code$ python -m preprocessing.bridge_code_lua.ent_vecs_from_txt_to_npy --entity_extension=True\n```\n\n## Files Description\n\n***In data/entities/:\n\n- entities_universe.txt: This file contains the wiki ids of the entities that we want to build entity vectors.\n In our case we were creating this file with the command ``python -m preprocessing.prepro_util --create_entity_universe=True``\n i.e. top 30 candidate entities for all spans of the datasets AIDA-TRAINING, AIDA_TEST-A, AIDA-TEST-B, ACE2004, AQUAINT, MSNBC.\n This file in our case contains 470105 wiki ids 1. The deep-ed (lua) project then reads this file and \n create entity vectors for all of them plus some extra entities used for the evaluation of the training procedure\n (Ceccarelli et al., 2013) (deep-ed-master/entities/relatedness/relatedness.lua) and hence sum to 484048 entities. When we\n run the Gerbil evaluation we add some \"extension entities\" necessary for the rest of the datasets, so we add 18613 for a\n total of 502661 entities. \n \n 1 These entities were created a year ago, but if I remember correctly, the lines 719 and 721 were: \n gmonly_files = ['aida_train.txt']\n allspans_files = ['aida_dev.txt', 'aida_test.txt', 'ace2004.txt', 'aquaint.txt', 'clueweb.txt', 'msnbc.txt', 'wikipedia.txt']\n\n- wikiid2nnid/*.txt (484,048 lines each): maps between wikiids and nnids (rows in the embedding array). \n\n\n***in data/basic_data/:\n- wiki_name_id_map.txt (4’399,390 lines): a list of wiki_name – wiki_id pairs for *all* the \nwikipages in the dump, but not including the disambiguation and redirect pages.\nThe disambiguation and redirect pages are not included since the result of our algorithm should be a concrete and \nspecific entity not a disambiguation page.\n- prob_yago_crosswikis_wikipedia_p_e_m.txt (21’587,744 lines): \n This pem dictionary is created by the whole wikipedia dump + crosswiki dataset + yago dataset.\nFor more details look at [this](https://arxiv.org/pdf/1704.04920.pdf) paper section 6 \"Candidate Selection\". \n\"It is computed by averaging probabilities from two indexes build from mention entity hyperlink count \nstatistics from Wikipedia and a large Web corpus (Spitkovsky and Chang, 2012). Moreover, we \nadd the YAGO dictionary of (Hoffart et al., 2011), where each candidate receives a uniform prior.\"\nAlso in terms of code it is implemented in the following files deep-ed/deep-ed-master/data_gen/gen_p_e_m/.\nFor instructions of executing this code please look at [this](https://github.com/dalab/deep-ed) link\n\n\n\n\n" }, { "path": "deep-ed/deep-ed-master/LICENSE", "content": "\n Apache License\n Version 2.0, January 2004\n http://www.apache.org/licenses/\n\n TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION\n\n 1. Definitions.\n\n \"License\" shall mean the terms and conditions for use, reproduction,\n and distribution as defined by Sections 1 through 9 of this document.\n\n \"Licensor\" shall mean the copyright owner or entity authorized by\n the copyright owner that is granting the License.\n\n \"Legal Entity\" shall mean the union of the acting entity and all\n other entities that control, are controlled by, or are under common\n control with that entity. For the purposes of this definition,\n \"control\" means (i) the power, direct or indirect, to cause the\n direction or management of such entity, whether by contract or\n otherwise, or (ii) ownership of fifty percent (50%) or more of the\n outstanding shares, or (iii) beneficial ownership of such entity.\n\n \"You\" (or \"Your\") shall mean an individual or Legal Entity\n exercising permissions granted by this License.\n\n \"Source\" form shall mean the preferred form for making modifications,\n including but not limited to software source code, documentation\n source, and configuration files.\n\n \"Object\" form shall mean any form resulting from mechanical\n transformation or translation of a Source form, including but\n not limited to compiled object code, generated documentation,\n and conversions to other media types.\n\n \"Work\" shall mean the work of authorship, whether in Source or\n Object form, made available under the License, as indicated by a\n copyright notice that is included in or attached to the work\n (an example is provided in the Appendix below).\n\n \"Derivative Works\" shall mean any work, whether in Source or Object\n form, that is based on (or derived from) the Work and for which the\n editorial revisions, annotations, elaborations, or other modifications\n represent, as a whole, an original work of authorship. For the purposes\n of this License, Derivative Works shall not include works that remain\n separable from, or merely link (or bind by name) to the interfaces of,\n the Work and Derivative Works thereof.\n\n \"Contribution\" shall mean any work of authorship, including\n the original version of the Work and any modifications or additions\n to that Work or Derivative Works thereof, that is intentionally\n submitted to Licensor for inclusion in the Work by the copyright owner\n or by an individual or Legal Entity authorized to submit on behalf of\n the copyright owner. For the purposes of this definition, \"submitted\"\n means any form of electronic, verbal, or written communication sent\n to the Licensor or its representatives, including but not limited to\n communication on electronic mailing lists, source code control systems,\n and issue tracking systems that are managed by, or on behalf of, the\n Licensor for the purpose of discussing and improving the Work, but\n excluding communication that is conspicuously marked or otherwise\n designated in writing by the copyright owner as \"Not a Contribution.\"\n\n \"Contributor\" shall mean Licensor and any individual or Legal Entity\n on behalf of whom a Contribution has been received by Licensor and\n subsequently incorporated within the Work.\n\n 2. Grant of Copyright License. Subject to the terms and conditions of\n this License, each Contributor hereby grants to You a perpetual,\n worldwide, non-exclusive, no-charge, royalty-free, irrevocable\n copyright license to reproduce, prepare Derivative Works of,\n publicly display, publicly perform, sublicense, and distribute the\n Work and such Derivative Works in Source or Object form.\n\n 3. Grant of Patent License. Subject to the terms and conditions of\n this License, each Contributor hereby grants to You a perpetual,\n worldwide, non-exclusive, no-charge, royalty-free, irrevocable\n (except as stated in this section) patent license to make, have made,\n use, offer to sell, sell, import, and otherwise transfer the Work,\n where such license applies only to those patent claims licensable\n by such Contributor that are necessarily infringed by their\n Contribution(s) alone or by combination of their Contribution(s)\n with the Work to which such Contribution(s) was submitted. If You\n institute patent litigation against any entity (including a\n cross-claim or counterclaim in a lawsuit) alleging that the Work\n or a Contribution incorporated within the Work constitutes direct\n or contributory patent infringement, then any patent licenses\n granted to You under this License for that Work shall terminate\n as of the date such litigation is filed.\n\n 4. Redistribution. You may reproduce and distribute copies of the\n Work or Derivative Works thereof in any medium, with or without\n modifications, and in Source or Object form, provided that You\n meet the following conditions:\n\n (a) You must give any other recipients of the Work or\n Derivative Works a copy of this License; and\n\n (b) You must cause any modified files to carry prominent notices\n stating that You changed the files; and\n\n (c) You must retain, in the Source form of any Derivative Works\n that You distribute, all copyright, patent, trademark, and\n attribution notices from the Source form of the Work,\n excluding those notices that do not pertain to any part of\n the Derivative Works; and\n\n (d) If the Work includes a \"NOTICE\" text file as part of its\n distribution, then any Derivative Works that You distribute must\n include a readable copy of the attribution notices contained\n within such NOTICE file, excluding those notices that do not\n pertain to any part of the Derivative Works, in at least one\n of the following places: within a NOTICE text file distributed\n as part of the Derivative Works; within the Source form or\n documentation, if provided along with the Derivative Works; or,\n within a display generated by the Derivative Works, if and\n wherever such third-party notices normally appear. The contents\n of the NOTICE file are for informational purposes only and\n do not modify the License. You may add Your own attribution\n notices within Derivative Works that You distribute, alongside\n or as an addendum to the NOTICE text from the Work, provided\n that such additional attribution notices cannot be construed\n as modifying the License.\n\n You may add Your own copyright statement to Your modifications and\n may provide additional or different license terms and conditions\n for use, reproduction, or distribution of Your modifications, or\n for any such Derivative Works as a whole, provided Your use,\n reproduction, and distribution of the Work otherwise complies with\n the conditions stated in this License.\n\n 5. Submission of Contributions. Unless You explicitly state otherwise,\n any Contribution intentionally submitted for inclusion in the Work\n by You to the Licensor shall be under the terms and conditions of\n this License, without any additional terms or conditions.\n Notwithstanding the above, nothing herein shall supersede or modify\n the terms of any separate license agreement you may have executed\n with Licensor regarding such Contributions.\n\n 6. Trademarks. This License does not grant permission to use the trade\n names, trademarks, service marks, or product names of the Licensor,\n except as required for reasonable and customary use in describing the\n origin of the Work and reproducing the content of the NOTICE file.\n\n 7. Disclaimer of Warranty. Unless required by applicable law or\n agreed to in writing, Licensor provides the Work (and each\n Contributor provides its Contributions) on an \"AS IS\" BASIS,\n WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or\n implied, including, without limitation, any warranties or conditions\n of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A\n PARTICULAR PURPOSE. You are solely responsible for determining the\n appropriateness of using or redistributing the Work and assume any\n risks associated with Your exercise of permissions under this License.\n\n 8. Limitation of Liability. In no event and under no legal theory,\n whether in tort (including negligence), contract, or otherwise,\n unless required by applicable law (such as deliberate and grossly\n negligent acts) or agreed to in writing, shall any Contributor be\n liable to You for damages, including any direct, indirect, special,\n incidental, or consequential damages of any character arising as a\n result of this License or out of the use or inability to use the\n Work (including but not limited to damages for loss of goodwill,\n work stoppage, computer failure or malfunction, or any and all\n other commercial damages or losses), even if such Contributor\n has been advised of the possibility of such damages.\n\n 9. Accepting Warranty or Additional Liability. While redistributing\n the Work or Derivative Works thereof, You may choose to offer,\n and charge a fee for, acceptance of support, warranty, indemnity,\n or other liability obligations and/or rights consistent with this\n License. However, in accepting such obligations, You may act only\n on Your own behalf and on Your sole responsibility, not on behalf\n of any other Contributor, and only if You agree to indemnify,\n defend, and hold each Contributor harmless for any liability\n incurred by, or claims asserted against, such Contributor by reason\n of your accepting any such warranty or additional liability.\n\n END OF TERMS AND CONDITIONS\n\n APPENDIX: How to apply the Apache License to your work.\n\n To apply the Apache License to your work, attach the following\n boilerplate notice, with the fields enclosed by brackets \"[]\"\n replaced with your own identifying information. (Don't include\n the brackets!) The text should be enclosed in the appropriate\n comment syntax for the file format. We also recommend that a\n file or class name and description of purpose be included on the\n same \"printed page\" as the copyright notice for easier\n identification within third-party archives.\n\n Copyright [yyyy] [name of copyright owner]\n\n Licensed under the Apache License, Version 2.0 (the \"License\");\n you may not use this file except in compliance with the License.\n You may obtain a copy of the License at\n\n http://www.apache.org/licenses/LICENSE-2.0\n\n Unless required by applicable law or agreed to in writing, software\n distributed under the License is distributed on an \"AS IS\" BASIS,\n WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n See the License for the specific language governing permissions and\n limitations under the License.\n" }, { "path": "deep-ed/deep-ed-master/README.md", "content": "# Source code for \"Deep Joint Entity Disambiguation with Local Neural Attention\"\n\n[O-E. Ganea and T. Hofmann, full paper @ EMNLP 2017](https://arxiv.org/abs/1704.04920)\n\nSlides and poster can be accessed [here](http://people.inf.ethz.ch/ganeao/).\n\n## Pre-trained entity embeddings\n\nEntity embeddings trained with our method using Word2Vec 300 dimensional pre-trained word vectors (GoogleNews-vectors-negative300.bin). They have norm 1 and are restricted only to entities appearing in the training, validation and test sets described in our paper. Available [here](https://polybox.ethz.ch/index.php/s/sH2JSB2c1OSj7yv).\n\n## Full set of annotations made by one of our global models\n\nSee file our_system_annotations.txt . Best to visualize together with its color scheme in a bash terminal. Contains the full set of annotations for the following datasets:\n\n```\n$ cat our_system_annotations.txt | grep 'Micro ' \n==> AQUAINT AQUAINT ; EPOCH = 307: Micro recall = 88.03% ; Micro F1 = 89.51%\t\n==> MSNBC MSNBC ; EPOCH = 307: Micro recall = 93.29% ; Micro F1 = 93.65%\t\n==> ACE04 ACE04 ; EPOCH = 307: Micro recall = 84.05% ; Micro F1 = 86.92%\t\n==> aida-B aida-B ; EPOCH = 307: Micro recall = 92.08% ; Micro F1 = 92.08%\t\n==> aida-A aida-A ; EPOCH = 307: Micro recall = 91.00% ; Micro F1 = 91.01%\t\n```\n\nGlobal model was trained on AIDA-train with pre-trained entity embeddings trained on Wikipedia. See details of how to run our code below.\n\nDetailed statistics per dataset as in table 6 of our paper can be accessed:\n\n```\n$ cat our_system_annotations.txt | grep -A20 'Micro ' \n```\n\n\n## How to run the system and reproduce our results\n\n1) Install [Torch](http://torch.ch/)\n\n\n2) Install torch libraries: cudnn, cutorch, [tds](https://github.com/torch/tds), gnuplot, xlua\n\n```luarocks install lib_name```\n\nCheck that each of these libraries can be imported in a torch terminal.\n\n\n3) Create a $DATA_PATH directoy (will be assumed to end in '/' in the next steps). Create a directory $DATA_PATH/generated/ that will contain all files generated in the next steps.\n\n\n4) Download data files needed for training and testing from [this link](https://drive.google.com/uc?id=0Bx8d3azIm_ZcbHMtVmRVc1o5TWM&export=download).\n Download basic_data.zip, unzip it and place the basic_data directory in $DATA_PATH/. All generated files will be build based on files in this basic_data/ directory.\n\n\n5) Download pre-trained Word2Vec vectors GoogleNews-vectors-negative300.bin.gz from https://code.google.com/archive/p/word2vec/.\nUnzip it and place the bin file in the folder $DATA_PATH/basic_data/wordEmbeddings/Word2Vec.\n\n\nNow we start creating additional data files needed in our pipeline:\n\n6) Create wikipedia_p_e_m.txt: \n\n```th data_gen/gen_p_e_m/gen_p_e_m_from_wiki.lua -root_data_dir $DATA_PATH```\n\n \n7) Merge wikipedia_p_e_m.txt and crosswikis_p_e_m.txt : \n\n```th data_gen/gen_p_e_m/merge_crosswikis_wiki.lua -root_data_dir $DATA_PATH```\n\n\n8) Create yago_p_e_m.txt: \n\n```th data_gen/gen_p_e_m/gen_p_e_m_from_yago.lua -root_data_dir $DATA_PATH ```\n\n\n9) Create a file ent_wiki_freq.txt with entity frequencies: \n\n```th entities/ent_name2id_freq/e_freq_gen.lua -root_data_dir $DATA_PATH```\n\n\n10) Generate all entity disambiguation datasets in a CSV format needed in our training stage: \n\n```\nmkdir $DATA_PATH/generated/test_train_data/\nth data_gen/gen_test_train_data/gen_all.lua -root_data_dir $DATA_PATH\n```\n \nVerify the statistics of these files as explained in the header comments of the files gen_ace_msnbc_aquaint_csv.lua and gen_aida_test.lua .\n\n\n11) Create training data for learning entity embeddings:\n\n i) From Wiki canonical pages: \n\n```th data_gen/gen_wiki_data/gen_ent_wiki_w_repr.lua -root_data_dir $DATA_PATH```\n\n ii) From context windows surrounding Wiki hyperlinks: \n\n```th data_gen/gen_wiki_data/gen_wiki_hyp_train_data.lua -root_data_dir $DATA_PATH```\n\n\n12) Compute the unigram frequency of each word in the Wikipedia corpus: \n\n```th words/w_freq/w_freq_gen.lua -root_data_dir $DATA_PATH```\n\n\n13) Compute the restricted training data for learning entity embeddings by using only candidate entities from the relatedness datasets and all ED sets:\n i) From Wiki canonical pages: \n\n```th entities/relatedness/filter_wiki_canonical_words_RLTD.lua -root_data_dir $DATA_PATH```\n\n ii) From context windows surrounding Wiki hyperlinks: \n\n```th entities/relatedness/filter_wiki_hyperlink_contexts_RLTD.lua -root_data_dir $DATA_PATH```\n\nAll files in the $DATA_PATH/generated/ folder containing the substring \"_RLTD\" are restricted to this set of entities (should contain 276030 entities).\n\nYour $DATA_PATH/generated/ folder should now contain the files : \n\n```\n$DATA_PATH/generated $ ls -lah ./\ntotal 147G\n9.5M all_candidate_ents_ed_rltd_datasets_RLTD.t7\n775M crosswikis_wikipedia_p_e_m.txt\n5.0M empty_page_ents.txt\n520M ent_name_id_map.t7\n95M ent_wiki_freq.txt\n7.3M relatedness_test.t7\n8.9M relatedness_validate.t7\n220 test_train_data\n1.5G wiki_canonical_words_RLTD.txt\n8.4G wiki_canonical_words.txt\n88G wiki_hyperlink_contexts.csv\n48G wiki_hyperlink_contexts_RLTD.csv\n329M wikipedia_p_e_m.txt\n11M word_wiki_freq.txt\n749M yago_p_e_m.txt\n\n$DATA_PATH//generated/test_train_data $ ls -lah ./\ntotal 124M\n14M aida_testA.csv\n13M aida_testB.csv\n50M aida_train.csv\n723K wned-ace2004.csv\n1.6M wned-aquaint.csv\n31M wned-clueweb.csv\n1.6M wned-msnbc.csv\n15M wned-wikipedia.csv\n```\n\n14) Now we train entity embeddings for the restricted set of entities (written in all_candidate_ents_ed_rltd_datasets_RLTD.t7). This is the step described in Section 3 of our paper.\n\nTo check the full list of parameters run: \n\n```th entities/learn_e2v/learn_a.lua -help```\n\nOptimal parameters (see entities/learn_e2v/learn_a.lua):\n\n```\noptimization = 'ADAGRAD'\nlr = 0.3\nbatch_size = 500\nword_vecs = 'w2v'\nnum_words_per_ent = 20\nnum_neg_words = 5\nunig_power = 0.6\nentities = 'RLTD'\nloss = 'maxm'\ndata = 'wiki-canonical-hyperlinks'\nnum_passes_wiki_words = 200\nhyp_ctxt_len = 10\n```\n\nTo run the embedding training on one GPU: \n\n```\nmkdir $DATA_PATH/generated/ent_vecs\nCUDA_VISIBLE_DEVICES=0 th entities/learn_e2v/learn_a.lua -root_data_dir $DATA_PATH |& tee log_train_entity_vecs\n```\n\nWarning: This code is not sufficiently optimized to run at maximum speed and GPU usage. Sorry for the inconvenience. It only uses the main thread to load data and perform word embedding lookup. It can be made to run much faster.\n\nDuring training, you will see (in the log file log_train_entity_vecs) the validation score on the entity relatedness dataset of the current set of entity embeddings. After around 24 hours (this code is not optimized!), you will see no improvement and can thus stop the training script. Pick the set of saved entity vectors from the folder generated/ent_vecs/ corresponding to the best validation score on the entity relatedness dataset which is the sum of all validation metrics (the TOTAL VALIDATION column). In our paper, we reported in Table 1 the results on the test set corresponding to this best validation score. \n\nYou should get some numbers similar to the following (may vary a little bit due to random initialization):\n\n```\nEntity Relatedness quality measure:\t\nmeasure =\tNDCG1\tNDCG5\tNDCG10\tMAP\tTOTAL VALIDATION\t\nour (vald) =\t0.681\t0.639\t0.671\t0.619\t2.610\t\nour (test) =\t0.650\t0.609\t0.641\t0.579\t\nYamada'16 =\t0.59\t0.56\t0.59\t0.52\t\nWikiMW =\t0.54\t0.52\t0.55\t0.48\n==> saving model to $DATA_PATH/generated/ent_vecs/ent_vecs__ep_69.t7\t\n```\n\nWe will call the name of this file with entity embeddings as $ENTITY_VECS. In our case, it is 'ent_vecs__ep_69.t7'\n\nThis code uses a simple initialization of entity embeddings based on the average of entities' title words (excluding stop words). This helps speed-up training and avoiding getting stuck in local minima. We found that using a random initialization might result in a slight quality decrease for ED only (up to 1%), requiring also a longer training time until reaching the same quality on entity relatedness (~60 hours).` \n\n15) Run the training for the global/local ED neural network. Arguments file: ed/args.lua . To list all arguments: \n\n```th ed/ed.lua -help```\n\nCommand to run the training: \n\n```\nmkdir $DATA_PATH/generated/ed_models/\nmkdir $DATA_PATH/generated/ed_models/training_plots/\nCUDA_VISIBLE_DEVICES=0 th ed/ed.lua -root_data_dir $DATA_PATH -ent_vecs_filename $ENTITY_VECS -model 'global' |& tee log_train_ed\n```\n\nLet it train for at least 48 hours (or 400 epochs as defined in our code), until the validation accuracy does not improve any more or starts dropping. As we wrote in the paper, we stop learning if\nthe validation F1 does not increase after 500 full epochs of the AIDA train dataset. Validation F1 can be following using the command:\n\n```cat log_train_ed | grep -A20 'Micro F1' | grep -A20 'aida-A'```\n\nThe best ED models will be saved in the folder generated/ed_models. This will only happen after the model gets > 90% F1 score on validation set (see test.lua).\n\nStatistics, weights and scors will be written in the log_train_ed file. Plots of micro F1 scores on all the validation and test set will be written in the folder $DATA_PATH/generated/ed_models/training_plots/ .\n\nResults variability: Results reported in our ED paper in Tables 3 and 4 are averaged over different runs of the ED neural architecture learning, but using the same set of entity embeddings. However, we found that the variance of ED results based on different trainings of entity embeddings might be a little higher, up to 0.5%.\n\n16) After training is terminated, one can re-load and test the best ED model using the command:\n\n```\nCUDA_VISIBLE_DEVICES=0 th ed/test/test_one_loaded_model.lua -root_data_dir $DATA_PATH -model global -ent_vecs_filename $ENTITY_VECS -test_one_model_file $ED_MODEL_FILENAME\n```\n\nwhere $ED_MODEL_FILENAME is a file in $DATA_PATH/generated/ed_models/ .\n\n\n17) Enjoy!\n" }, { "path": "deep-ed/deep-ed-master/data_gen/gen_p_e_m/gen_p_e_m_from_wiki.lua", "content": "-- Generate p(e|m) index from Wikipedia\n-- Run: th data_gen/gen_p_e_m/gen_p_e_m_from_wiki.lua -root_data_dir $DATA_PATH\n\ncmd = torch.CmdLine()\ncmd:option('-root_data_dir', '', 'Root path of the data, $DATA_PATH.')\ncmd:text()\nopt = cmd:parse(arg or {})\nassert(opt.root_data_dir ~= '', 'Specify a valid root_data_dir path argument.')\n\n\nrequire 'torch'\ndofile 'utils/utils.lua'\ndofile 'data_gen/parse_wiki_dump/parse_wiki_dump_tools.lua'\ntds = tds or require 'tds' \n\nprint('\\nComputing Wikipedia p_e_m')\n\nit, _ = io.open(opt.root_data_dir .. 'basic_data/textWithAnchorsFromAllWikipedia2014Feb.txt')\nline = it:read()\n\nwiki_e_m_counts = tds.Hash()\n\n-- Find anchors, e.g. anarchism\nlocal num_lines = 0\nlocal parsing_errors = 0\nlocal list_ent_errors = 0\nlocal diez_ent_errors = 0\nlocal disambiguation_ent_errors = 0\nlocal num_valid_hyperlinks = 0\n\nwhile (line) do\n num_lines = num_lines + 1\n if num_lines % 5000000 == 0 then\n print('Processed ' .. num_lines .. ' lines. Parsing errs = ' .. \n parsing_errors .. ' List ent errs = ' .. \n list_ent_errors .. ' diez errs = ' .. diez_ent_errors .. \n ' disambig errs = ' .. disambiguation_ent_errors .. \n ' . Num valid hyperlinks = ' .. num_valid_hyperlinks)\n end\n\n if not line:find(' b.freq end)\n\n local str = ''\n local total_freq = 0\n for _,el in pairs(tbl) do\n str = str .. el.ent_wikiid .. ',' .. el.freq \n str = str .. ',' .. get_ent_name_from_wikiid(el.ent_wikiid):gsub(' ', '_') .. '\\t'\n total_freq = total_freq + el.freq\n end\n ouf:write(mention .. '\\t' .. total_freq .. '\\t' .. str .. '\\n')\n\n -- NICK\n str = ''\n for _,el in pairs(tbl) do\n str = str .. el.ent_wikiid .. ',' .. (el.freq/total_freq) \n str = str .. ',' .. get_ent_name_from_wikiid(el.ent_wikiid):gsub(' ', '_') .. '\\t'\n end\n ouf2:write(mention .. '\\t' .. 1 .. '\\t' .. str .. '\\n') \n\nend\nouf:flush()\nio.close(ouf)\n\nouf2:flush()\nio.close(ouf2)\n\nprint(' Done sorting and writing.')\n\n\n\n" }, { "path": "deep-ed/deep-ed-master/data_gen/gen_p_e_m/gen_p_e_m_from_yago.lua", "content": "-- Generate p(e|m) index from Wikipedia\n-- Run: th data_gen/gen_p_e_m/gen_p_e_m_from_yago.lua -root_data_dir $DATA_PATH\n\ncmd = torch.CmdLine()\ncmd:option('-root_data_dir', '', 'Root path of the data, $DATA_PATH.')\ncmd:text()\nopt = cmd:parse(arg or {})\nassert(opt.root_data_dir ~= '', 'Specify a valid root_data_dir path argument.')\n\n\nrequire 'torch'\ndofile 'utils/utils.lua'\ndofile 'data_gen/gen_p_e_m/unicode_map.lua'\nif not get_redirected_ent_title then\n dofile 'data_gen/indexes/wiki_redirects_index.lua'\nend\nif not get_ent_name_from_wikiid then\n dofile 'entities/ent_name2id_freq/ent_name_id.lua'\nend\n\ntds = tds or require 'tds' \n\nprint('\\nComputing YAGO p_e_m')\n\nlocal it, _ = io.open(opt.root_data_dir .. 'basic_data/p_e_m_data/aida_means.tsv')\nlocal line = it:read()\n\nlocal num_lines = 0\nlocal wiki_e_m_counts = tds.Hash()\n\nwhile (line) do\n num_lines = num_lines + 1\n if num_lines % 5000000 == 0 then\n print('Processed ' .. num_lines .. ' lines.')\n end\n local parts = split(line, '\\t')\n assert(table_len(parts) == 2)\n assert(parts[1]:sub(1,1) == '\"')\n assert(parts[1]:sub(parts[1]:len(),parts[1]:len()) == '\"')\n \n local mention = parts[1]:sub(2, parts[1]:len() - 1)\n mention = trim1(mention) -- NICK\n local ent_name = parts[2]\n ent_name = string.gsub(ent_name, '&', '&')\n ent_name = string.gsub(ent_name, '"', '\"')\n while ent_name:find('\\\\u') do\n local x = ent_name:find('\\\\u')\n local code = ent_name:sub(x, x + 5)\n assert(unicode2ascii[code], code)\n replace = unicode2ascii[code]\n if(replace == \"%\") then\n\t replace = \"%%\"\n end\n ent_name = string.gsub(ent_name, code, replace)\n end\n \n ent_name = preprocess_ent_name(ent_name)\n local ent_wikiid = get_ent_wikiid_from_name(ent_name, true)\n if ent_wikiid ~= unk_ent_wikiid and mention ~= \"\" then -- NICK\n if not wiki_e_m_counts[mention] then\n wiki_e_m_counts[mention] = tds.Hash()\n end\n wiki_e_m_counts[mention][ent_wikiid] = 1\n end\n \n line = it:read()\nend\n\n\nprint('Now sorting and writing ..')\nout_file = opt.root_data_dir .. 'generated/yago_p_e_m.txt'\nouf = assert(io.open(out_file, \"w\"))\n\nfor mention, list in pairs(wiki_e_m_counts) do\n local str = ''\n local total_freq = 0\n for ent_wikiid, _ in pairs(list) do\n str = str .. ent_wikiid .. ',' .. get_ent_name_from_wikiid(ent_wikiid):gsub(' ', '_') .. '\\t'\n total_freq = total_freq + 1\n end\n ouf:write(mention .. '\\t' .. total_freq .. '\\t' .. str .. '\\n')\nend\nouf:flush()\nio.close(ouf)\n\nprint(' Done sorting and writing.')\n" }, { "path": "deep-ed/deep-ed-master/data_gen/gen_p_e_m/merge_crosswikis_wiki.lua", "content": "-- Merge Wikipedia and Crosswikis p(e|m) indexes\n-- Run: th data_gen/gen_p_e_m/merge_crosswikis_wiki.lua -root_data_dir $DATA_PATH\n\ncmd = torch.CmdLine()\ncmd:option('-root_data_dir', '', 'Root path of the data, $DATA_PATH.')\ncmd:text()\nopt = cmd:parse(arg or {})\nassert(opt.root_data_dir ~= '', 'Specify a valid root_data_dir path argument.')\n\n\nrequire 'torch'\ndofile 'utils/utils.lua'\ndofile 'entities/ent_name2id_freq/ent_name_id.lua'\n\nprint('\\nMerging Wikipedia and Crosswikis p_e_m')\n\ntds = tds or require 'tds' \nmerged_e_m_counts = tds.Hash()\n\nprint('Process Wikipedia')\nit, _ = io.open(opt.root_data_dir .. 'generated/wikipedia_p_e_m.txt')\nline = it:read()\n\nwhile (line) do\n local parts = split(line, \"\\t\")\n local mention = parts[1] \n\n if (not mention:find('Wikipedia')) and (not mention:find('wikipedia')) then\n if not merged_e_m_counts[mention] then\n merged_e_m_counts[mention] = tds.Hash()\n end \n\n local total_freq = tonumber(parts[2])\n assert(total_freq, line)\n local num_ents = table_len(parts)\n for i = 3, num_ents do\n local ent_str = split(parts[i], \",\")\n local ent_wikiid = tonumber(ent_str[1])\n assert(ent_wikiid)\n local freq = tonumber(ent_str[2])\n assert(freq)\n\n if not merged_e_m_counts[mention][ent_wikiid] then\n merged_e_m_counts[mention][ent_wikiid] = 0\n end\n merged_e_m_counts[mention][ent_wikiid] = merged_e_m_counts[mention][ent_wikiid] + freq\n end\n end\n line = it:read()\nend\n\n\nprint('Process Crosswikis')\n\nit, _ = io.open(opt.root_data_dir .. 'basic_data/p_e_m_data/crosswikis_p_e_m.txt')\nline = it:read()\n\nwhile (line) do\n local parts = split(line, \"\\t\")\n local mention = trim1(parts[1]) -- NICK not sure if it is necessary\n\n if mention ~= \"\" and (not mention:find('Wikipedia')) and (not mention:find('wikipedia')) then\n if not merged_e_m_counts[mention] then\n merged_e_m_counts[mention] = tds.Hash()\n end \n\n local total_freq = tonumber(parts[2])\n assert(total_freq)\n local num_ents = table_len(parts)\n for i = 3, num_ents do\n local ent_str = split(parts[i], \",\")\n local ent_wikiid = tonumber(ent_str[1])\n assert(ent_wikiid)\n local freq = tonumber(ent_str[2])\n assert(freq)\n\n if not merged_e_m_counts[mention][ent_wikiid] then\n merged_e_m_counts[mention][ent_wikiid] = 0\n end\n merged_e_m_counts[mention][ent_wikiid] = merged_e_m_counts[mention][ent_wikiid] + freq\n end\n end\n line = it:read()\nend\n\n\nprint('Now sorting and writing ..')\nout_file = opt.root_data_dir .. 'generated/crosswikis_wikipedia_p_e_m.txt'\nouf = assert(io.open(out_file, \"w\"))\n-- NICK\nout_file2 = opt.root_data_dir .. 'generated/prob_crosswikis_wikipedia_p_e_m.txt'\nouf2 = assert(io.open(out_file2, \"w\"))\n\nfor mention, list in pairs(merged_e_m_counts) do\n if mention:len() >= 1 then\n local tbl = {}\n for ent_wikiid, freq in pairs(list) do\n table.insert(tbl, {ent_wikiid = ent_wikiid, freq = freq})\n end\n table.sort(tbl, function(a,b) return a.freq > b.freq end)\n\n local str = ''\n local total_freq = 0\n local num_ents = 0\n for _,el in pairs(tbl) do\n if is_valid_ent(el.ent_wikiid) then\n str = str .. el.ent_wikiid .. ',' .. el.freq \n str = str .. ',' .. get_ent_name_from_wikiid(el.ent_wikiid):gsub(' ', '_') .. '\\t'\n num_ents = num_ents + 1\n total_freq = total_freq + el.freq\n\n if num_ents >= 100 then -- At most 100 candidates\n break\n end\n end\n end\n ouf:write(mention .. '\\t' .. total_freq .. '\\t' .. str .. '\\n')\n\n -- NICK\n str = ''\n num_ents = 0\n for _,el in pairs(tbl) do\n if is_valid_ent(el.ent_wikiid) then\n str = str .. el.ent_wikiid .. ',' .. (el.freq/total_freq) \n str = str .. ',' .. get_ent_name_from_wikiid(el.ent_wikiid):gsub(' ', '_') .. '\\t'\n num_ents = num_ents + 1\n\n if num_ents >= 100 then -- At most 100 candidates\n break\n end\n end\n end\n ouf2:write(mention .. '\\t' .. 1 .. '\\t' .. str .. '\\n') \n -- NICK end\n\n end\nend\nouf:flush()\nio.close(ouf)\n\nouf2:flush()\nio.close(ouf2)\n\nprint(' Done sorting and writing.')\n\n" }, { "path": "deep-ed/deep-ed-master/data_gen/gen_p_e_m/unicode_map.lua", "content": "local unicode = {'\\\\u00bb', '\\\\u007d', '\\\\u00a1', '\\\\u0259', '\\\\u0641', '\\\\u0398', '\\\\u00fd', '\\\\u0940', '\\\\u00f9', '\\\\u02bc', '\\\\u00f6', '\\\\u00f8', '\\\\u0107', '\\\\u0648', '\\\\u0105', '\\\\u002c', '\\\\u6768', '\\\\u0160', '\\\\u015b', '\\\\u00c0', '\\\\u266f', '\\\\u0430', '\\\\u0141', '\\\\u1ea3', '\\\\u00df', '\\\\u2212', '\\\\u8650', '\\\\u012d', '\\\\u1e47', '\\\\u00c5', '\\\\u00ab', '\\\\u0226', '\\\\u0930', '\\\\u04a4', '\\\\u030d', '\\\\u0631', '\\\\u207f', '\\\\u00bf', '\\\\u2010', '\\\\u1e6e', '\\\\u00cd', '\\\\u00c6', '\\\\u30e8', '\\\\u1e63', '\\\\u6a5f', '\\\\u03c3', '\\\\u00d5', '\\\\u0644', '\\\\u2020', '\\\\u0104', '\\\\u010a', '\\\\u013c', '\\\\u0123', '\\\\u0159', '\\\\u1e6f', '\\\\u003a', '\\\\u06af', '\\\\u00fc', '\\\\u4e09', '\\\\u0028', '\\\\u03b2', '\\\\u2103', '\\\\u0191', '\\\\u03bc', '\\\\u00d1', '\\\\u207a', '\\\\u79d2', '\\\\u6536', '\\\\u1ed3', '\\\\u0329', '\\\\u0196', '\\\\u00fb', '\\\\u0435', '\\\\u01b0', '\\\\u007e', '\\\\u1e62', '\\\\u0181', '\\\\u1ea7', '\\\\u2011', '\\\\u03c9', '\\\\u201d', '\\\\u0165', '\\\\u0422', '\\\\u1e33', '\\\\u0144', '\\\\u00fa', '\\\\u1ed5', '\\\\u0632', '\\\\u0643', '\\\\u1ea1', '\\\\u011e', '\\\\u062a', '\\\\u00ee', '\\\\u00c2', '\\\\u016d', '\\\\u003d', '\\\\u2202', '\\\\u2605', '\\\\u0112', '\\\\u73cd', '\\\\u03a1', '\\\\u0182', '\\\\u00d2', '\\\\u0153', '\\\\u016f', '\\\\u00de', '\\\\u00a3', '\\\\u1e45', '\\\\u1ef1', '\\\\u4e45', '\\\\u06cc', '\\\\u1ea2', '\\\\u0152', '\\\\uff09', '\\\\u0219', '\\\\u0457', '\\\\u0283', '\\\\u1e5a', '\\\\u064e', '\\\\u0164', '\\\\u0116', '\\\\u018b', '\\\\u1ec3', '\\\\u00b4', '\\\\u002b', '\\\\u7248', '\\\\u0937', '\\\\u203a', '\\\\u4eba', '\\\\u002f', '\\\\u0136', '\\\\u01bc', '\\\\u017b', '\\\\u00a9', '\\\\u03a9', '\\\\u00f2', '\\\\u2026', '\\\\u00c7', '\\\\u0969', '\\\\u0198', '\\\\u011f', '\\\\u00e0', '\\\\u0126', '\\\\u018a', '\\\\u1edf', '\\\\u005e', '\\\\u03b4', '\\\\u0137', '\\\\u01f5', '\\\\u1e34', '\\\\u007b', '\\\\u00f3', '\\\\u01c0', '\\\\u00f1', '\\\\u1ef9', '\\\\u03c8', '\\\\ub8e8', '\\\\u0119', '\\\\u014f', '\\\\uff5e', '\\\\u016c', '\\\\u0358', '\\\\u529f', '\\\\u2606', '\\\\u00e4', '\\\\u012b', '\\\\u00c9', '\\\\u0173', '\\\\u092f', '\\\\u5957', '\\\\u1e49', '\\\\u1ec1', '\\\\u019f', '\\\\u02bb', '\\\\u0399', '\\\\u01c1', '\\\\u03d5', '\\\\u017a', '\\\\u1e0d', '\\\\u0148', '\\\\u01a4', '\\\\u00a2', '\\\\u011c', '\\\\u1e92', '\\\\u01b1', '\\\\u0443', '\\\\u00f4', '\\\\u2122', '\\\\u82e5', '\\\\u0967', '\\\\u1eaf', '\\\\u013e', '\\\\u1e46', '\\\\u03b1', '\\\\u884c', '\\\\u0328', '\\\\u0021', '\\\\u00aa', '\\\\u014d', '\\\\u002e', '\\\\u00cb', '\\\\u062f', '\\\\u0102', '\\\\u0155', '\\\\u00cf', '\\\\u0446', '\\\\u1e80', '\\\\u003b', '\\\\u6c38', '\\\\u0103', '\\\\u1e6c', '\\\\u203c', '\\\\u00dc', '\\\\u00b3', '\\\\u0145', '\\\\u0122', '\\\\u8fdb', '\\\\u015e', '\\\\u017c', '\\\\u043f', '\\\\u0442', '\\\\u0629', '\\\\u6176', '\\\\u1edd', '\\\\u2018', '\\\\u00ea', '\\\\u0060', '\\\\u0147', '\\\\u00e7', '\\\\u00dd', '\\\\u00d6', '\\\\u043a', '\\\\u00ec', '\\\\ufb01', '\\\\u0124', '\\\\u1e5f', '\\\\u0431', '\\\\u1e94', '\\\\u1ea8', '\\\\u01b3', '\\\\u018f', '\\\\u0627', '\\\\u1ef3', '\\\\u091f', '\\\\u03a6', '\\\\u674e', '\\\\u016b', '\\\\u039b', '\\\\u2032', '\\\\u002a', '\\\\u2033', '\\\\u00b7', '\\\\u00ce', '\\\\u2075', '\\\\u043c', '\\\\u2116', '\\\\u1e6d', '\\\\u00be', '\\\\u0171', '\\\\u0433', '\\\\u0635', '\\\\u0640', '\\\\u01a1', '\\\\u00a7', '\\\\u019d', '\\\\u301c', '\\\\u00f5', '\\\\u0187', '\\\\u1ef6', '\\\\u1e0c', '\\\\u1ec5', '\\\\u01b8', '\\\\u00f0', '\\\\u1e93', '\\\\u00eb', '\\\\u03bd', '\\\\u0108', '\\\\u010d', '\\\\u5229', '\\\\u2080', '\\\\u00c8', '\\\\u9ece', '\\\\u0917', '\\\\u85cf', '\\\\u1edb', '\\\\u1e25', '\\\\u045b', '\\\\u1ec9', '\\\\u1ecd', '\\\\u0633', '\\\\u2022', '\\\\u01e8', '\\\\u0197', '\\\\u2019', '\\\\u0421', '\\\\u1ecb', '\\\\u9910', '\\\\uc2a4', '\\\\u1e31', '\\\\u00c1', '\\\\u1ea9', '\\\\u1eeb', '\\\\u01f4', '\\\\u01c2', '\\\\u0146', '\\\\u0162', '\\\\ufb02', '\\\\u01ac', '\\\\u0025', '\\\\u015c', '\\\\u01ce', '\\\\u01c3', '\\\\u0179', '\\\\u01e5', '\\\\u58eb', '\\\\u745e', '\\\\uff08', '\\\\u016a', '\\\\u03a5', '\\\\u039c', '\\\\u00bd', '\\\\u0169', '\\\\u55f7', '\\\\u89d2', '\\\\u00e6', '\\\\u9752', '\\\\u005b', '\\\\u003f', '\\\\u041f', '\\\\u06a9', '\\\\u0027', '\\\\u1ebf', '\\\\u0646', '\\\\u0130', '\\\\u1eb1', '\\\\u0395', '\\\\u03b3', '\\\\u01d4', '\\\\u00ed', '\\\\u041a', '\\\\u0149', '\\\\u0143', '\\\\u010f', '\\\\u1e24', '\\\\u0121', '\\\\u1ecf', '\\\\u06c1', '\\\\u00d3', '\\\\u0029', '\\\\u02bf', '\\\\u010e', '\\\\u1e0e', '\\\\u01d2', '\\\\u00ff', '\\\\u00fe', '\\\\u03a0', '\\\\u1ebc', '\\\\u2153', '\\\\u00e1', '\\\\u00ca', '\\\\u012c', '\\\\u017d', '\\\\u0110', '\\\\u266d', '\\\\u0639', '\\\\u014e', '\\\\u1e43', '\\\\u0026', '\\\\u20ac', '\\\\u0024', '\\\\u011d', '\\\\u003e', '\\\\u0163', '\\\\u0939', '\\\\u221a', '\\\\u00e3', '\\\\u65f6', '\\\\u0118', '\\\\u0101', '\\\\u0628', '\\\\u221e', '\\\\u1ed1', '\\\\u0393', '\\\\u00c4', '\\\\u0161', '\\\\u00e9', '\\\\u0220', '\\\\u0115', '\\\\u002d', '\\\\u03c0', '\\\\u0177', '\\\\u00ba', '\\\\u0158', '\\\\u01a7', '\\\\u0117', '\\\\u043b', '\\\\u00d7', '\\\\u00e2', '\\\\u0175', '\\\\u0420', '\\\\u0391', '\\\\u00b2', '\\\\u014c', '\\\\u2013', '\\\\u00b9', '\\\\u1ef7', '\\\\u064a', '\\\\u0301', '\\\\u95a2', '\\\\u0113', '\\\\u013b', '\\\\u094d', '\\\\u03b5', '\\\\u1eef', '\\\\u2c6b', '\\\\u00da', '\\\\u00d8', '\\\\u0432', '\\\\u0109', '\\\\u00d9', '\\\\u00d4', '\\\\u011b', '\\\\u0303', '\\\\u0392', '\\\\u1eed', '\\\\u0444', '\\\\u026a', '\\\\u0218', '\\\\u00ef', '\\\\u1ed9', '\\\\u00b0', '\\\\u010c', '\\\\uac00', '\\\\u02be', '\\\\u2012', '\\\\u5baa', '\\\\u00e8', '\\\\u1ebd', '\\\\u30fb', '\\\\u0127', '\\\\u010b', '\\\\u0131', '\\\\u1ebb', '\\\\u0150', '\\\\u0327', '\\\\u0100', '\\\\u1ee7', '\\\\u1ed7', '\\\\u0129', '\\\\u00c3', '\\\\u003c', '\\\\u2260', '\\\\u0106', '\\\\u6625', '\\\\u0184', '\\\\u1eb5', '\\\\u4fdd', '\\\\u00b1', '\\\\u021b', '\\\\u014b', '\\\\uff0d', '\\\\u1e2a', '\\\\u00e5', '\\\\u017e', '\\\\u011a', '\\\\u1eab', '\\\\u200e', '\\\\u1e35', '\\\\u1e5b', '\\\\u2192', '\\\\u0040', '\\\\u1eb7', '\\\\u01b2', '\\\\u5b58', '\\\\u201c', '\\\\u015f', '\\\\u01e6', '\\\\u0111', '\\\\u738b', '\\\\u03a7', '\\\\u1ead', '\\\\u1ec7', '\\\\u0324', '\\\\u2665', '\\\\ub9c8', '\\\\u6bba', '\\\\u0151', '\\\\u2661', '\\\\u03ba', '\\\\ua784', '\\\\u2014', '\\\\u1ee9', '\\\\u0120', '\\\\u012a', '\\\\u7433', '\\\\u0134', '\\\\u039a', '\\\\u1ee3', '\\\\u1ea5', '\\\\u1ee5', '\\\\u0142', '\\\\u043e', '\\\\u01eb', '\\\\u0440', '\\\\u03a3', '\\\\u093e', '\\\\u00d0', '\\\\u092e', '\\\\u00b5', '\\\\u013d', '\\\\u1ecc', '\\\\u0394', '\\\\u00bc', '\\\\u01d0', '\\\\u015a', '\\\\u02b9', '\\\\u0645', '\\\\u043d', '\\\\u00cc'}\n\nlocal ascii = {'»', '}', '¡', 'ə', 'ف', 'Θ', 'ý', 'ी', 'ù', 'ʼ', 'ö', 'ø', 'ć', 'و', 'ą', ',', '杨', 'Š', 'ś', 'À', '♯', 'а', 'Ł', 'ả', 'ß', '−', '虐', 'ĭ', 'ṇ', 'Å', '«', 'Ȧ', 'र', 'Ҥ', 'ʼ', 'ر', 'ⁿ', '¿', '‐', 'Ṯ', 'Í', 'Æ', 'ヨ', 'ṣ', '機', 'σ', 'Õ', 'ل', '†', 'Ą', 'Ċ', 'ļ', 'ģ', 'ř', 'ṯ', ':', 'گ', 'ü', '三', '(', 'β', '℃', 'Ƒ', 'μ', 'Ñ', '⁺', '秒', '收', 'ồ', '̩', 'Ɩ', 'û', 'е', 'ư', '~', 'Ṣ', 'Ɓ', 'ầ', '‑', 'ω', '”', 'ť', 'Т', 'ḳ', 'ń', 'ú', 'ổ', 'ز', 'ك', 'ạ', 'Ğ', 'ت', 'î', 'Â', 'ŭ', '=', '∂', '★', 'Ē', '珍', 'Ρ', 'Ƃ', 'Ò', 'œ', 'ů', 'Þ', '£', 'ṅ', 'ự', '久', 'ی', 'Ả', 'Œ', '）', 'ș', 'ї', 'ʃ', 'Ṛ', 'َ', 'Ť', 'Ė', 'Ƌ', 'ể', '´', '+', '版', 'ष', '›', '人', '/', 'Ķ', 'Ƽ', 'Ż', '©', 'Ω', 'ò', '…', 'Ç', '३', 'Ƙ', 'ğ', 'à', 'Ħ', 'Ɗ', 'ở', '^', 'δ', 'ķ', 'ǵ', 'Ḵ', '{', 'ó', 'ǀ', 'ñ', 'ỹ', 'ψ', '루', 'ę', 'ŏ', '～', 'Ŭ', '͘', '功', '☆', 'ä', 'ī', 'É', 'ų', 'य', '套', 'ṉ', 'ề', 'Ɵ', 'ʻ', 'Ι', 'ǁ', 'ϕ', 'ź', 'ḍ', 'ň', 'Ƥ', '¢', 'Ĝ', 'Ẓ', 'Ʊ', 'у', 'ô', '™', '若', '१', 'ắ', 'ľ', 'Ṇ', 'α', '行', '̨', '!', 'ª', 'ō', '.', 'Ë', 'د', 'Ă', 'ŕ', 'Ï', 'ц', 'Ẁ', ';', '永', 'ă', 'Ṭ', '‼', 'Ü', '³', 'Ņ', 'Ģ', '进', 'Ş', 'ż', 'п', 'т', 'ة', '慶', 'ờ', '‘', 'ê', '`', 'Ň', 'ç', 'Ý', 'Ö', 'к', 'ì', 'fi', 'Ĥ', 'ṟ', 'б', 'Ẕ', 'Ẩ', 'Ƴ', 'Ə', 'ا', 'ỳ', 'ट', 'Φ', '李', 'ū', 'Λ', '′', '*', '″', '·', 'Î', '⁵', 'м', '№', 'ṭ', '¾', 'ű', 'г', 'ص', 'ـ', 'ơ', '§', 'Ɲ', '〜', 'õ', 'Ƈ', 'Ỷ', 'Ḍ', 'ễ', 'Ƹ', 'ð', 'ẓ', 'ë', 'ν', 'Ĉ', 'č', '利', '₀', 'È', '黎', 'ग', '藏', 'ớ', 'ḥ', 'ћ', 'ỉ', 'ọ', 'س', '•', 'Ǩ', 'Ɨ', '’', 'С', 'ị', '餐', '스', 'ḱ', 'Á', 'ẩ', 'ừ', 'Ǵ', 'ǂ', 'ņ', 'Ţ', 'fl', 'Ƭ', '%', 'Ŝ', 'ǎ', 'ǃ', 'Ź', 'ǥ', '士', '瑞', '（', 'Ū', 'Υ', 'Μ', '½', 'ũ', '嗷', '角', 'æ', '青', '[', '?', 'П', 'ک', '\\'', 'ế', 'ن', 'İ', 'ằ', 'Ε', 'γ', 'ǔ', 'í', 'К', 'ʼn', 'Ń', 'ď', 'Ḥ', 'ġ', 'ỏ', 'ہ', 'Ó', ')', 'ʿ', 'Ď', 'Ḏ', 'ǒ', 'ÿ', 'þ', 'Π', 'Ẽ', '⅓', 'á', 'Ê', 'Ĭ', 'Ž', 'Đ', '♭', 'ع', 'Ŏ', 'ṃ', '&', '€', '$', 'ĝ', '>', 'ţ', 'ह', '√', 'ã', '时', 'Ę', 'ā', 'ب', '∞', 'ố', 'Γ', 'Ä', 'š', 'é', 'Ƞ', 'ĕ', '-', 'π', 'ŷ', 'º', 'Ř', 'Ƨ', 'ė', 'л', '×', 'â', 'ŵ', 'Р', 'Α', '²', 'Ō', '–', '¹', 'ỷ', 'ي', '́', '関', 'ē', 'Ļ', '्', 'ε', 'ữ', 'Ⱬ', 'Ú', 'Ø', 'в', 'ĉ', 'Ù', 'Ô', 'ě', '̃', 'Β', 'ử', 'ф', 'ɪ', 'Ș', 'ï', 'ộ', '°', 'Č', '가', 'ʾ', '‒', '宪', 'è', 'ẽ', '・', 'ħ', 'ċ', 'ı', 'ẻ', 'Ő', '̧', 'Ā', 'ủ', 'ỗ', 'ĩ', 'Ã', '<', '≠', 'Ć', '春', 'Ƅ', 'ẵ', '保', '±', 'ț', 'ŋ', '－', 'Ḫ', 'å', 'ž', 'Ě', 'ẫ', '‎', 'ḵ', 'ṛ', '→', '@', 'ặ', 'Ʋ', '存', '“', 'ş', 'Ǧ', 'đ', '王', 'Χ', 'ậ', 'ệ', '̤', '♥', '마', '殺', 'ő', '♡', 'κ', 'Ꞅ', '—', 'ứ', 'Ġ', 'Ī', '琳', 'Ĵ', 'Κ', 'ợ', 'ấ', 'ụ', 'ł', 'о', 'ǫ', 'р', 'Σ', 'ा', 'Ð', 'म', 'µ', 'Ľ', 'Ọ', 'Δ', '¼', 'ǐ', 'Ś', 'ʹ', 'م', 'н', 'Ì'}\n\ndofile 'utils/utils.lua'\nassert(table_len(unicode) == table_len(ascii))\n\nunicode2ascii = {}\nfor i,letter in pairs(ascii) do\n unicode2ascii[unicode[i]] = ascii[i]\n unicode2ascii['\\\\u0022'] = '\"'\n unicode2ascii['\\\\u0023'] = '#'\n unicode2ascii['\\\\u005c'] = '\\\\'\n unicode2ascii['\\\\u00a0'] = ''\nend\n\n" }, { "path": "deep-ed/deep-ed-master/data_gen/gen_test_train_data/gen_ace_msnbc_aquaint_csv.lua", "content": "-- Generate test data from the ACE/MSNBC/AQUAINT datasets by keeping the context and\n-- entity candidates for each annotated mention\n\n-- Format: \n-- doc_name \\t doc_name \\t mention \\t left_ctxt \\t right_ctxt \\t CANDIDATES \\t [ent_wikiid,p_e_m,ent_name]+ \\t GT: \\t pos,ent_wikiid,p_e_m,ent_name\n\n-- Stats:\n--cat wned-ace2004.csv | wc -l\n--257\n--cat wned-ace2004.csv | grep -P 'GT:\\t-1' | wc -l\n--20\n--cat wned-ace2004.csv | grep -P 'GT:\\t1,' | wc -l\n--217\n\n--cat wned-aquaint.csv | wc -l\n--727\n--cat wned-aquaint.csv | grep -P 'GT:\\t-1' | wc -l\n--33\n--cat wned-aquaint.csv | grep -P 'GT:\\t1,' | wc -l\n--604\n\n--cat wned-msnbc.csv | wc -l\n--656\n--cat wned-msnbc.csv | grep -P 'GT:\\t-1' | wc -l\n--22\n--cat wned-msnbc.csv | grep -P 'GT:\\t1,' | wc -l\n--496\n\nif not ent_p_e_m_index then\n require 'torch'\n dofile 'data_gen/indexes/wiki_redirects_index.lua'\n dofile 'data_gen/indexes/yago_crosswikis_wiki.lua'\n dofile 'utils/utils.lua'\nend\n\ntds = tds or require 'tds'\n\nlocal function gen_test_ace(dataset) \n\n print('\\nGenerating test data from ' .. dataset .. ' set ')\n\n local path = opt.root_data_dir .. 'basic_data/test_datasets/wned-datasets/' .. dataset .. '/'\n\n out_file = opt.root_data_dir .. 'generated/test_train_data/wned-' .. dataset .. '.csv'\n ouf = assert(io.open(out_file, \"w\"))\n\n annotations, _ = io.open(path .. dataset .. '.xml')\n \n local num_nonexistent_ent_id = 0\n local num_correct_ents = 0\n\n local cur_doc_text = ''\n local cur_doc_name = ''\n\n local line = annotations:read()\n while (line) do\n if (not line:find('document docName=\\\"')) then\n if line:find('') then\n line = annotations:read()\n local x,y = line:find('')\n local z,t = line:find('')\n local cur_mention = line:sub(y + 1, z - 1)\n cur_mention = string.gsub(cur_mention, '&', '&')\n \n line = annotations:read()\n x,y = line:find('')\n z,t = line:find('')\n cur_ent_title = ''\n if not line:find('') then\n cur_ent_title = line:sub(y + 1, z - 1)\n end\n \n line = annotations:read()\n x,y = line:find('')\n z,t = line:find('')\n local offset = 1 + tonumber(line:sub(y + 1, z - 1))\n \n line = annotations:read()\n x,y = line:find('')\n z,t = line:find('')\n local length = tonumber(line:sub(y + 1, z - 1))\n length = cur_mention:len()\n\n line = annotations:read()\n if line:find('') then\n line = annotations:read()\n end\n \n assert(line:find(''))\n \n offset = math.max(1, offset - 10)\n while (cur_doc_text:sub(offset, offset + length - 1) ~= cur_mention) do\n-- print(cur_mention .. ' ---> ' .. cur_doc_text:sub(offset, offset + length - 1))\n offset = offset + 1\n end\n \n cur_mention = preprocess_mention(cur_mention)\n \n if cur_ent_title ~= 'NIL' and cur_ent_title ~= '' and cur_ent_title:len() > 0 then\n local cur_ent_wikiid = get_ent_wikiid_from_name(cur_ent_title)\n if cur_ent_wikiid == unk_ent_wikiid then\n num_nonexistent_ent_id = num_nonexistent_ent_id + 1\n print(green(cur_ent_title))\n else\n num_correct_ents = num_correct_ents + 1\n end\n \n assert(cur_mention:len() > 0)\n local str = cur_doc_name .. '\\t' .. cur_doc_name .. '\\t' .. cur_mention .. '\\t'\n \n local left_words = split_in_words(cur_doc_text:sub(1, offset - 1))\n local num_left_words = table_len(left_words)\n local left_ctxt = {}\n for i = math.max(1, num_left_words - 100 + 1), num_left_words do\n table.insert(left_ctxt, left_words[i])\n end\n if table_len(left_ctxt) == 0 then\n table.insert(left_ctxt, 'EMPTYCTXT')\n end\n str = str .. table.concat(left_ctxt, ' ') .. '\\t'\n\n local right_words = split_in_words(cur_doc_text:sub(offset + length))\n local num_right_words = table_len(right_words)\n local right_ctxt = {}\n for i = 1, math.min(num_right_words, 100) do\n table.insert(right_ctxt, right_words[i])\n end\n if table_len(right_ctxt) == 0 then\n table.insert(right_ctxt, 'EMPTYCTXT')\n end\n str = str .. table.concat(right_ctxt, ' ') .. '\\tCANDIDATES\\t'\n\n\n -- Entity candidates from p(e|m) dictionary\n if ent_p_e_m_index[cur_mention] and #(ent_p_e_m_index[cur_mention]) > 0 then\n\n local sorted_cand = {}\n for ent_wikiid,p in pairs(ent_p_e_m_index[cur_mention]) do\n table.insert(sorted_cand, {ent_wikiid = ent_wikiid, p = p})\n end\n table.sort(sorted_cand, function(a,b) return a.p > b.p end)\n\n local candidates = {}\n local gt_pos = -1\n for pos,e in pairs(sorted_cand) do\n if pos <= 100 then\n table.insert(candidates, e.ent_wikiid .. ',' .. string.format(\"%.3f\", e.p) .. ',' .. get_ent_name_from_wikiid(e.ent_wikiid))\n if e.ent_wikiid == cur_ent_wikiid then\n gt_pos = pos\n end\n else\n break\n end\n end\n str = str .. table.concat(candidates, '\\t') .. '\\tGT:\\t'\n\n if gt_pos > 0 then\n ouf:write(str .. gt_pos .. ',' .. candidates[gt_pos] .. '\\n')\n else\n if cur_ent_wikiid ~= unk_ent_wikiid then\n ouf:write(str .. '-1,' .. cur_ent_wikiid .. ',' .. cur_ent_title .. '\\n')\n else\n ouf:write(str .. '-1\\n')\n end\n end\n else\n if cur_ent_wikiid ~= unk_ent_wikiid then\n ouf:write(str .. 'EMPTYCAND\\tGT:\\t-1,' .. cur_ent_wikiid .. ',' .. cur_ent_title .. '\\n')\n else\n ouf:write(str .. 'EMPTYCAND\\tGT:\\t-1\\n')\n end\n end\n\n end\n end\n else\n local x,y = line:find('document docName=\\\"')\n local z,t = line:find('\\\">')\n cur_doc_name = line:sub(y + 1, z - 1)\n cur_doc_name = string.gsub(cur_doc_name, '&', '&')\n \n local it,_ = io.open(path .. 'RawText/' .. cur_doc_name)\n cur_doc_text = ''\n local cur_line = it:read()\n while cur_line do\n cur_doc_text = cur_doc_text .. cur_line .. ' '\n cur_line = it:read()\n end\n cur_doc_text = string.gsub(cur_doc_text, '&', '&')\n end\n line = annotations:read()\n end\n\n ouf:flush()\n io.close(ouf)\n\n print('Done ' .. dataset .. '.')\n print('num_nonexistent_ent_id = ' .. num_nonexistent_ent_id .. '; num_correct_ents = ' .. num_correct_ents)\nend\n\n\ngen_test_ace('wikipedia')\ngen_test_ace('clueweb')\ngen_test_ace('ace2004')\ngen_test_ace('msnbc') \ngen_test_ace('aquaint')\n" }, { "path": "deep-ed/deep-ed-master/data_gen/gen_test_train_data/gen_aida_test.lua", "content": "-- Generate test data from the AIDA dataset by keeping the context and\n-- entity candidates for each annotated mention\n\n-- Format: \n-- doc_name \\t doc_name \\t mention \\t left_ctxt \\t right_ctxt \\t CANDIDATES \\t [ent_wikiid,p_e_m,ent_name]+ \\t GT: \\t pos,ent_wikiid,p_e_m,ent_name\n\n-- Stats:\n--cat aida_testA.csv | wc -l\n--4791\n--cat aida_testA.csv | grep -P 'GT:\\t-1' | wc -l\n--43\n--cat aida_testA.csv | grep -P 'GT:\\t1,' | wc -l\n--3401\n\n--cat aida_testB.csv | wc -l\n--4485\n--cat aida_testB.csv | grep -P 'GT:\\t-1' | wc -l\n--19\n--cat aida_testB.csv | grep -P 'GT:\\t1,' | wc -l\n--3084\n\nif not ent_p_e_m_index then\n require 'torch'\n dofile 'data_gen/indexes/wiki_redirects_index.lua'\n dofile 'data_gen/indexes/yago_crosswikis_wiki.lua'\n dofile 'utils/utils.lua'\nend\n\ntds = tds or require 'tds'\n\nprint('\\nGenerating test data from AIDA set ')\n\nit, _ = io.open(opt.root_data_dir .. 'basic_data/test_datasets/AIDA/testa_testb_aggregate_original')\n\nout_file_A = opt.root_data_dir .. 'generated/test_train_data/aida_testA.csv'\nout_file_B = opt.root_data_dir .. 'generated/test_train_data/aida_testB.csv'\n\nouf_A = assert(io.open(out_file_A, \"w\"))\nouf_B = assert(io.open(out_file_B, \"w\"))\n\nlocal ouf = ouf_A\n\nlocal num_nme = 0\nlocal num_nonexistent_ent_title = 0\nlocal num_nonexistent_ent_id = 0\nlocal num_nonexistent_both = 0\nlocal num_correct_ents = 0\nlocal num_total_ents = 0\n\nlocal cur_words_num = 0\nlocal cur_words = {}\nlocal cur_mentions = {}\nlocal cur_mentions_num = 0\n\nlocal cur_doc_name = ''\n\nlocal function write_results()\n -- Write results:\n if cur_doc_name ~= '' then\n local header = cur_doc_name .. '\\t' .. cur_doc_name .. '\\t'\n for _, hyp in pairs(cur_mentions) do\n assert(hyp.mention:len() > 0, line) \n local mention = hyp.mention\n local str = header .. hyp.mention .. '\\t'\n\n local left_ctxt = {}\n for i = math.max(0, hyp.start_off - 100), hyp.start_off - 1 do\n table.insert(left_ctxt, cur_words[i])\n end\n if table_len(left_ctxt) == 0 then\n table.insert(left_ctxt, 'EMPTYCTXT')\n end\n str = str .. table.concat(left_ctxt, ' ') .. '\\t'\n\n local right_ctxt = {}\n for i = hyp.end_off + 1, math.min(cur_words_num, hyp.end_off + 100) do\n table.insert(right_ctxt, cur_words[i])\n end\n if table_len(right_ctxt) == 0 then\n table.insert(right_ctxt, 'EMPTYCTXT')\n end\n str = str .. table.concat(right_ctxt, ' ') .. '\\tCANDIDATES\\t'\n\n -- Entity candidates from p(e|m) dictionary\n if ent_p_e_m_index[mention] and #(ent_p_e_m_index[mention]) > 0 then\n\n local sorted_cand = {}\n for ent_wikiid,p in pairs(ent_p_e_m_index[mention]) do\n table.insert(sorted_cand, {ent_wikiid = ent_wikiid, p = p})\n end\n table.sort(sorted_cand, function(a,b) return a.p > b.p end)\n\n local candidates = {}\n local gt_pos = -1\n for pos,e in pairs(sorted_cand) do\n if pos <= 100 then\n table.insert(candidates, e.ent_wikiid .. ',' .. string.format(\"%.3f\", e.p) .. ',' .. get_ent_name_from_wikiid(e.ent_wikiid))\n if e.ent_wikiid == hyp.ent_wikiid then\n gt_pos = pos\n end\n else\n break\n end\n end\n str = str .. table.concat(candidates, '\\t') .. '\\tGT:\\t'\n\n if gt_pos > 0 then\n ouf:write(str .. gt_pos .. ',' .. candidates[gt_pos] .. '\\n')\n else\n if hyp.ent_wikiid ~= unk_ent_wikiid then\n ouf:write(str .. '-1,' .. hyp.ent_wikiid .. ',' .. get_ent_name_from_wikiid(hyp.ent_wikiid) .. '\\n')\n else\n ouf:write(str .. '-1\\n')\n end\n end\n else\n if hyp.ent_wikiid ~= unk_ent_wikiid then\n ouf:write(str .. 'EMPTYCAND\\tGT:\\t-1,' .. hyp.ent_wikiid .. ',' .. get_ent_name_from_wikiid(hyp.ent_wikiid) .. '\\n')\n else\n ouf:write(str .. 'EMPTYCAND\\tGT:\\t-1\\n')\n end\n end\n end\n end\nend\n\n\n\nlocal line = it:read()\nwhile (line) do\n if (not line:find('-DOCSTART-')) then\n local parts = split(line, '\\t')\n local num_parts = table_len(parts)\n assert(num_parts == 0 or num_parts == 1 or num_parts == 4 or num_parts == 7 or num_parts == 6, line)\n if num_parts > 0 then\n if num_parts == 4 and parts[2] == 'B' then\n num_nme = num_nme + 1\n end\n \n if (num_parts == 7 or num_parts == 6) and parts[2] == 'B' then\n \n -- Find current mention. A few hacks here.\n local cur_mention = preprocess_mention(parts[3])\n \n local x,y = parts[5]:find('/wiki/')\n local cur_ent_title = parts[5]:sub(y + 1)\n local cur_ent_wikiid = tonumber(parts[6])\n local index_ent_title = get_ent_name_from_wikiid(cur_ent_wikiid)\n local index_ent_wikiid = get_ent_wikiid_from_name(cur_ent_title)\n \n local final_ent_wikiid = index_ent_wikiid\n if final_ent_wikiid == unk_ent_wikiid then\n final_ent_wikiid = cur_ent_wikiid\n end\n \n if (index_ent_title == cur_ent_title and cur_ent_wikiid == index_ent_wikiid) then\n num_correct_ents = num_correct_ents + 1\n elseif (index_ent_title ~= cur_ent_title and cur_ent_wikiid ~= index_ent_wikiid) then\n num_nonexistent_both = num_nonexistent_both + 1\n elseif index_ent_title ~= cur_ent_title then\n assert(cur_ent_wikiid == index_ent_wikiid)\n num_nonexistent_ent_title = num_nonexistent_ent_title + 1\n else\n assert(index_ent_title == cur_ent_title)\n assert(cur_ent_wikiid ~= index_ent_wikiid)\n num_nonexistent_ent_id = num_nonexistent_ent_id + 1\n end\n \n num_total_ents = num_total_ents + 1 -- Keep even incorrect links\n\n cur_mentions_num = cur_mentions_num + 1\n cur_mentions[cur_mentions_num] = {}\n cur_mentions[cur_mentions_num].mention = cur_mention\n cur_mentions[cur_mentions_num].ent_wikiid = final_ent_wikiid\n cur_mentions[cur_mentions_num].start_off = cur_words_num + 1\n cur_mentions[cur_mentions_num].end_off = cur_words_num + table_len(split(parts[3], ' '))\n end\n \n local words_on_this_line = split_in_words(parts[1])\n for _,w in pairs(words_on_this_line) do\n table.insert(cur_words, modify_uppercase_phrase(w))\n cur_words_num = cur_words_num + 1\n end\n end\n \n else\n assert(line:find('-DOCSTART-'))\n write_results()\n \n if cur_doc_name:find('testa') and line:find('testb') then\n ouf = ouf_B\n print('Done validation testA : ')\n print('num_nme = ' .. num_nme .. '; num_nonexistent_ent_title = ' .. num_nonexistent_ent_title)\n print('num_nonexistent_ent_id = ' .. num_nonexistent_ent_id .. '; num_nonexistent_both = ' .. num_nonexistent_both)\n print('num_correct_ents = ' .. num_correct_ents .. '; num_total_ents = ' .. num_total_ents)\n end\n \n local words = split_in_words(line)\n for _,w in pairs(words) do\n if w:find('testa') or w:find('testb') then\n cur_doc_name = w\n break\n end\n end \n cur_words = {}\n cur_words_num = 0\n cur_mentions = {}\n cur_mentions_num = 0\n end\n \n line = it:read()\nend\n\nwrite_results()\n\nouf_A:flush()\nio.close(ouf_A)\nouf_B:flush()\nio.close(ouf_B)\n\n\nprint(' Done AIDA.')\nprint('num_nme = ' .. num_nme .. '; num_nonexistent_ent_title = ' .. num_nonexistent_ent_title)\nprint('num_nonexistent_ent_id = ' .. num_nonexistent_ent_id .. '; num_nonexistent_both = ' .. num_nonexistent_both)\nprint('num_correct_ents = ' .. num_correct_ents .. '; num_total_ents = ' .. num_total_ents)\n" }, { "path": "deep-ed/deep-ed-master/data_gen/gen_test_train_data/gen_aida_train.lua", "content": "-- Generate train data from the AIDA dataset by keeping the context and\n-- entity candidates for each annotated mention\n\n-- Format: \n-- doc_name \\t doc_name \\t mention \\t left_ctxt \\t right_ctxt \\t CANDIDATES \\t [ent_wikiid,p_e_m,ent_name]+ \\t GT: \\t pos,ent_wikiid,p_e_m,ent_name\n\nif not ent_p_e_m_index then\n require 'torch'\n dofile 'data_gen/indexes/wiki_redirects_index.lua'\n dofile 'data_gen/indexes/yago_crosswikis_wiki.lua'\n dofile 'utils/utils.lua'\n tds = tds or require 'tds'\nend\n\nprint('\\nGenerating train data from AIDA set ')\n\nit, _ = io.open(opt.root_data_dir .. 'basic_data/test_datasets/AIDA/aida_train.txt')\n\nout_file = opt.root_data_dir .. 'generated/test_train_data/aida_train.csv'\nouf = assert(io.open(out_file, \"w\"))\n\nlocal num_nme = 0\nlocal num_nonexistent_ent_title = 0\nlocal num_nonexistent_ent_id = 0\nlocal num_nonexistent_both = 0\nlocal num_correct_ents = 0\nlocal num_total_ents = 0\n\nlocal cur_words_num = 0\nlocal cur_words = {}\nlocal cur_mentions = {}\nlocal cur_mentions_num = 0\n\nlocal cur_doc_name = ''\n\nlocal function write_results()\n -- Write results:\n if cur_doc_name ~= '' then\n local header = cur_doc_name .. '\\t' .. cur_doc_name .. '\\t'\n for _, hyp in pairs(cur_mentions) do\n assert(hyp.mention:len() > 0, line)\n local str = header .. hyp.mention .. '\\t'\n\n local left_ctxt = {}\n for i = math.max(0, hyp.start_off - 100), hyp.start_off - 1 do\n table.insert(left_ctxt, cur_words[i])\n end\n if table_len(left_ctxt) == 0 then\n table.insert(left_ctxt, 'EMPTYCTXT')\n end\n str = str .. table.concat(left_ctxt, ' ') .. '\\t'\n\n local right_ctxt = {}\n for i = hyp.end_off + 1, math.min(cur_words_num, hyp.end_off + 100) do\n table.insert(right_ctxt, cur_words[i])\n end\n if table_len(right_ctxt) == 0 then\n table.insert(right_ctxt, 'EMPTYCTXT')\n end\n str = str .. table.concat(right_ctxt, ' ') .. '\\tCANDIDATES\\t'\n\n -- Entity candidates from p(e|m) dictionary\n if ent_p_e_m_index[hyp.mention] and #(ent_p_e_m_index[hyp.mention]) > 0 then\n\n local sorted_cand = {}\n for ent_wikiid,p in pairs(ent_p_e_m_index[hyp.mention]) do\n table.insert(sorted_cand, {ent_wikiid = ent_wikiid, p = p})\n end\n table.sort(sorted_cand, function(a,b) return a.p > b.p end)\n\n local candidates = {}\n local gt_pos = -1\n for pos,e in pairs(sorted_cand) do\n if pos <= 100 then\n table.insert(candidates, e.ent_wikiid .. ',' .. string.format(\"%.3f\", e.p) .. ',' .. get_ent_name_from_wikiid(e.ent_wikiid))\n if e.ent_wikiid == hyp.ent_wikiid then\n gt_pos = pos\n end\n else\n break\n end\n end\n str = str .. table.concat(candidates, '\\t') .. '\\tGT:\\t'\n\n if gt_pos > 0 then\n ouf:write(str .. gt_pos .. ',' .. candidates[gt_pos] .. '\\n')\n end\n end\n end\n end\nend\n\n\n\nlocal line = it:read()\nwhile (line) do\n if (not line:find('-DOCSTART-')) then\n local parts = split(line, '\\t')\n local num_parts = table_len(parts)\n assert(num_parts == 0 or num_parts == 1 or num_parts == 4 or num_parts == 7 or num_parts == 6, line)\n if num_parts > 0 then\n if num_parts == 4 and parts[2] == 'B' then\n num_nme = num_nme + 1\n end\n \n if (num_parts == 7 or num_parts == 6) and parts[2] == 'B' then\n \n -- Find current mention. A few hacks here.\n \n local cur_mention = preprocess_mention(parts[3])\n \n local x,y = parts[5]:find('/wiki/')\n local cur_ent_title = parts[5]:sub(y + 1)\n local cur_ent_wikiid = tonumber(parts[6])\n local index_ent_title = get_ent_name_from_wikiid(cur_ent_wikiid)\n local index_ent_wikiid = get_ent_wikiid_from_name(cur_ent_title)\n \n local final_ent_wikiid = index_ent_wikiid\n if final_ent_wikiid == unk_ent_wikiid then\n final_ent_wikiid = cur_ent_wikiid\n end\n \n if (index_ent_title == cur_ent_title and cur_ent_wikiid == index_ent_wikiid) then\n num_correct_ents = num_correct_ents + 1\n elseif (index_ent_title ~= cur_ent_title and cur_ent_wikiid ~= index_ent_wikiid) then\n num_nonexistent_both = num_nonexistent_both + 1\n elseif index_ent_title ~= cur_ent_title then\n assert(cur_ent_wikiid == index_ent_wikiid)\n num_nonexistent_ent_title = num_nonexistent_ent_title + 1\n else\n assert(index_ent_title == cur_ent_title)\n assert(cur_ent_wikiid ~= index_ent_wikiid)\n num_nonexistent_ent_id = num_nonexistent_ent_id + 1\n end\n \n num_total_ents = num_total_ents + 1 -- Keep even incorrect links\n\n cur_mentions_num = cur_mentions_num + 1\n cur_mentions[cur_mentions_num] = {}\n cur_mentions[cur_mentions_num].mention = cur_mention\n cur_mentions[cur_mentions_num].ent_wikiid = final_ent_wikiid\n cur_mentions[cur_mentions_num].start_off = cur_words_num + 1\n cur_mentions[cur_mentions_num].end_off = cur_words_num + table_len(split(parts[3], ' '))\n end\n \n local words_on_this_line = split_in_words(parts[1])\n for _,w in pairs(words_on_this_line) do\n table.insert(cur_words, modify_uppercase_phrase(w))\n cur_words_num = cur_words_num + 1\n end\n end\n \n else\n assert(line:find('-DOCSTART-'))\n write_results()\n \n cur_doc_name = line:sub(13)\n \n cur_words = {}\n cur_words_num = 0\n cur_mentions = {}\n cur_mentions_num = 0\n end\n \n line = it:read()\nend\n\nwrite_results()\n\nouf:flush()\nio.close(ouf)\n\n\nprint(' Done AIDA.')\nprint('num_nme = ' .. num_nme .. '; num_nonexistent_ent_title = ' .. num_nonexistent_ent_title)\nprint('num_nonexistent_ent_id = ' .. num_nonexistent_ent_id .. '; num_nonexistent_both = ' .. num_nonexistent_both)\nprint('num_correct_ents = ' .. num_correct_ents .. '; num_total_ents = ' .. num_total_ents)\n" }, { "path": "deep-ed/deep-ed-master/data_gen/gen_test_train_data/gen_all.lua", "content": "-- Generates all training and test data for entity disambiguation.\n\nif not ent_p_e_m_index then\n require 'torch'\n dofile 'data_gen/indexes/wiki_redirects_index.lua'\n dofile 'data_gen/indexes/yago_crosswikis_wiki.lua'\n dofile 'utils/utils.lua'\n tds = tds or require 'tds'\nend\n\ndofile 'data_gen/gen_test_train_data/gen_aida_test.lua'\ndofile 'data_gen/gen_test_train_data/gen_aida_train.lua'\ndofile 'data_gen/gen_test_train_data/gen_ace_msnbc_aquaint_csv.lua'" }, { "path": "deep-ed/deep-ed-master/data_gen/gen_wiki_data/gen_ent_wiki_w_repr.lua", "content": "if not opt then\n cmd = torch.CmdLine()\n cmd:option('-root_data_dir', '', 'Root path of the data, $DATA_PATH.')\n cmd:text()\n opt = cmd:parse(arg or {})\n assert(opt.root_data_dir ~= '', 'Specify a valid root_data_dir path argument.') \nend\n\n\nrequire 'torch'\ndofile 'utils/utils.lua'\ndofile 'data_gen/parse_wiki_dump/parse_wiki_dump_tools.lua'\ndofile 'entities/ent_name2id_freq/e_freq_index.lua'\ntds = tds or require 'tds' \n\nprint('\\nExtracting text only from Wiki dump. Output is wiki_canonical_words.txt containing on each line an Wiki entity with the list of all words in its canonical Wiki page.')\n\nit, _ = io.open(opt.root_data_dir .. 'basic_data/textWithAnchorsFromAllWikipedia2014Feb.txt')\n\nout_file = opt.root_data_dir .. 'generated/wiki_canonical_words.txt'\nouf = assert(io.open(out_file, \"w\"))\n\nline = it:read()\n\n-- Find anchors, e.g. anarchism\nlocal num_lines = 0\nlocal num_valid_ents = 0\nlocal num_error_ents = 0 -- Probably list or disambiguation pages.\n\nlocal empty_valid_ents = get_map_all_valid_ents()\n\nlocal cur_words = ''\nlocal cur_ent_wikiid = -1\n\nwhile (line) do\n num_lines = num_lines + 1\n if num_lines % 5000000 == 0 then\n print('Processed ' .. num_lines .. ' lines. Num valid ents = ' .. num_valid_ents .. '. Num errs = ' .. num_error_ents)\n end\n\n if (not line:find('')) then\n _, text, _, _ , _ , _ = extract_text_and_hyp(line, false)\n local words = split_in_words(text)\n cur_words = cur_words .. table.concat(words, ' ') .. ' '\n \n elseif line:find(' 0 and cur_words ~= '') then\n if cur_ent_wikiid ~= unk_ent_wikiid and is_valid_ent(cur_ent_wikiid) then\n ouf:write(cur_ent_wikiid .. '\\t' .. get_ent_name_from_wikiid(cur_ent_wikiid) .. '\\t' .. cur_words .. '\\n')\n empty_valid_ents[cur_ent_wikiid] = nil\n num_valid_ents = num_valid_ents + 1\n else\n num_error_ents = num_error_ents + 1\n end\n end\n \n cur_ent_wikiid = extract_page_entity_title(line)\n cur_words = ''\n end\n \n line = it:read()\nend\nouf:flush()\nio.close(ouf)\n\n-- Num valid ents = 4126137. Num errs = 332944\nprint(' Done extracting text only from Wiki dump. Num valid ents = ' .. num_valid_ents .. '. Num errs = ' .. num_error_ents)\n\n\nprint('Create file with all entities with empty Wikipedia pages.')\nlocal empty_ents = {}\nfor ent_wikiid, _ in pairs(empty_valid_ents) do\n table.insert(empty_ents, {ent_wikiid = ent_wikiid, f = get_ent_freq(ent_wikiid)})\nend\ntable.sort(empty_ents, function(a,b) return a.f > b.f end)\n\nlocal ouf2 = assert(io.open(opt.root_data_dir .. 'generated/empty_page_ents.txt', \"w\"))\nfor _,x in pairs(empty_ents) do\n ouf2:write(x.ent_wikiid .. '\\t' .. get_ent_name_from_wikiid(x.ent_wikiid) .. '\\t' .. x.f .. '\\n')\nend\nouf2:flush()\nio.close(ouf2)\nprint(' Done')" }, { "path": "deep-ed/deep-ed-master/data_gen/gen_wiki_data/gen_wiki_hyp_train_data.lua", "content": "-- Generate training data from Wikipedia hyperlinks by keeping the context and\n-- entity candidates for each hyperlink\n\n-- Format: \n-- ent_wikiid \\t ent_name \\t mention \\t left_ctxt \\t right_ctxt \\t CANDIDATES \\t [ent_wikiid,p_e_m,ent_name]+ \\t GT: \\t pos,ent_wikiid,p_e_m,ent_name\n\nif not opt then\n cmd = torch.CmdLine()\n cmd:option('-root_data_dir', '', 'Root path of the data, $DATA_PATH.')\n cmd:text()\n opt = cmd:parse(arg or {})\n assert(opt.root_data_dir ~= '', 'Specify a valid root_data_dir path argument.') \nend\n\nrequire 'torch'\ndofile 'data_gen/parse_wiki_dump/parse_wiki_dump_tools.lua'\ndofile 'data_gen/indexes/yago_crosswikis_wiki.lua'\ntds = tds or require 'tds' \n\nprint('\\nGenerating training data from Wiki dump')\n\nit, _ = io.open(opt.root_data_dir .. 'basic_data/textWithAnchorsFromAllWikipedia2014Feb.txt')\n\nout_file = opt.root_data_dir .. 'generated/wiki_hyperlink_contexts.csv'\nouf = assert(io.open(out_file, \"w\"))\n\n-- Find anchors, e.g. anarchism\nlocal num_lines = 0\nlocal num_valid_hyp = 0\n\nlocal cur_words_num = 0\nlocal cur_words = {}\nlocal cur_mentions = {}\nlocal cur_mentions_num = 0\nlocal cur_ent_wikiid = -1\n\nlocal line = it:read()\nwhile (line) do\n num_lines = num_lines + 1\n if num_lines % 1000000 == 0 then\n print('Processed ' .. num_lines .. ' lines. Num valid hyp = ' .. num_valid_hyp)\n end\n\n -- If it's a line from the Wiki page, add its text words and its hyperlinks\n if (not line:find('')) then\n list_hyp, text, _, _ , _ , _ = extract_text_and_hyp(line, true)\n \n local words_on_this_line = split_in_words(text)\n local num_added_hyp = 0\n local line_mentions = {}\n for _,w in pairs(words_on_this_line) do\n local wstart = string_starts(w, 'MMSTART')\n local wend = string_starts(w, 'MMEND')\n if (not wstart) and (not wend) then\n table.insert(cur_words, w)\n cur_words_num = cur_words_num + 1\n elseif wstart then\n local mention_idx = tonumber(w:sub(1 + ('MMSTART'):len()))\n assert(mention_idx, w)\n line_mentions[mention_idx] = {start_off = (cur_words_num + 1), end_off = -1}\n elseif wend then\n num_added_hyp = num_added_hyp + 1\n local mention_idx = tonumber(w:sub(1 + ('MMEND'):len()))\n assert(mention_idx, w)\n assert(line_mentions[mention_idx])\n line_mentions[mention_idx].end_off = cur_words_num\n end\n end\n \n assert(table_len(list_hyp) == num_added_hyp, line .. ' :: ' .. text .. ' :: ' .. num_added_hyp .. ' ' .. table_len(list_hyp))\n for _,hyp in pairs(list_hyp) do\n assert(line_mentions[hyp.cnt])\n cur_mentions_num = cur_mentions_num + 1\n cur_mentions[cur_mentions_num] = {}\n cur_mentions[cur_mentions_num].mention = hyp.mention\n cur_mentions[cur_mentions_num].ent_wikiid = hyp.ent_wikiid\n cur_mentions[cur_mentions_num].start_off = line_mentions[hyp.cnt].start_off\n cur_mentions[cur_mentions_num].end_off = line_mentions[hyp.cnt].end_off\n end\n \n elseif line:find(' 0 then\n assert(hyp.mention:len() > 0, line)\n local str = header .. hyp.mention .. '\\t'\n\n local left_ctxt = {}\n for i = math.max(0, hyp.start_off - 100), hyp.start_off - 1 do\n table.insert(left_ctxt, cur_words[i])\n end\n if table_len(left_ctxt) == 0 then\n table.insert(left_ctxt, 'EMPTYCTXT')\n end\n str = str .. table.concat(left_ctxt, ' ') .. '\\t'\n\n local right_ctxt = {}\n for i = hyp.end_off + 1, math.min(cur_words_num, hyp.end_off + 100) do\n table.insert(right_ctxt, cur_words[i])\n end\n if table_len(right_ctxt) == 0 then\n table.insert(right_ctxt, 'EMPTYCTXT')\n end\n str = str .. table.concat(right_ctxt, ' ') .. '\\tCANDIDATES\\t'\n \n -- Entity candidates from p(e|m) dictionary\n local unsorted_cand = {}\n for ent_wikiid,p in pairs(ent_p_e_m_index[hyp.mention]) do\n table.insert(unsorted_cand, {ent_wikiid = ent_wikiid, p = p})\n end\n table.sort(unsorted_cand, function(a,b) return a.p > b.p end)\n \n local candidates = {}\n local gt_pos = -1\n for pos,e in pairs(unsorted_cand) do\n if pos <= 32 then\n table.insert(candidates, e.ent_wikiid .. ',' .. string.format(\"%.3f\", e.p) .. ',' .. get_ent_name_from_wikiid(e.ent_wikiid))\n if e.ent_wikiid == hyp.ent_wikiid then\n gt_pos = pos\n end\n else\n break\n end\n end\n str = str .. table.concat(candidates, '\\t') .. '\\tGT:\\t'\n \n if gt_pos > 0 then\n num_valid_hyp = num_valid_hyp + 1\n ouf:write(str .. gt_pos .. ',' .. candidates[gt_pos] .. '\\n')\n end\n end\n end\n end\n \n cur_ent_wikiid = extract_page_entity_title(line)\n cur_words = {}\n cur_words_num = 0\n cur_mentions = {}\n cur_mentions_num = 0\n end\n \n line = it:read()\nend\nouf:flush()\nio.close(ouf)\n\nprint(' Done generating training data from Wiki dump. Num valid hyp = ' .. num_valid_hyp)\n" }, { "path": "deep-ed/deep-ed-master/data_gen/indexes/wiki_disambiguation_pages_index.lua", "content": "-- Loads the link disambiguation index from Wikipedia\n\nif not opt then\n cmd = torch.CmdLine()\n cmd:option('-root_data_dir', '', 'Root path of the data, $DATA_PATH.')\n cmd:text()\n opt = cmd:parse(arg or {})\n assert(opt.root_data_dir ~= '', 'Specify a valid root_data_dir path argument.') \nend\n\n\nrequire 'torch'\ndofile 'utils/utils.lua'\ntds = tds or require 'tds' \n\nprint('==> Loading disambiguation index')\nit, _ = io.open(opt.root_data_dir .. 'basic_data/wiki_disambiguation_pages.txt')\nline = it:read()\n\nwiki_disambiguation_index = tds.Hash()\nwhile (line) do\n parts = split(line, \"\\t\")\n assert(tonumber(parts[1]))\n wiki_disambiguation_index[tonumber(parts[1])] = 1\n line = it:read()\nend\n\nassert(wiki_disambiguation_index[579])\nassert(wiki_disambiguation_index[41535072])\n\nprint(' Done loading disambiguation index')\n" }, { "path": "deep-ed/deep-ed-master/data_gen/indexes/wiki_redirects_index.lua", "content": "-- Loads the link redirect index from Wikipedia\n\nif not opt then\n cmd = torch.CmdLine()\n cmd:option('-root_data_dir', '', 'Root path of the data, $DATA_PATH.')\n cmd:text()\n opt = cmd:parse(arg or {})\n assert(opt.root_data_dir ~= '', 'Specify a valid root_data_dir path argument.') \nend\n\nrequire 'torch'\ndofile 'utils/utils.lua'\ntds = tds or require 'tds' \n\nprint('==> Loading redirects index')\nit, _ = io.open(opt.root_data_dir .. 'basic_data/wiki_redirects.txt')\nline = it:read()\n\nlocal wiki_redirects_index = tds.Hash()\nwhile (line) do\n parts = split(line, \"\\t\")\n wiki_redirects_index[parts[1]] = parts[2]\n line = it:read()\nend\n\nassert(wiki_redirects_index['Coercive'] == 'Coercion')\nassert(wiki_redirects_index['Hosford, FL'] == 'Hosford, Florida')\n\nprint(' Done loading redirects index')\n\n\nfunction get_redirected_ent_title(ent_name)\n if wiki_redirects_index[ent_name] then\n return wiki_redirects_index[ent_name]\n else\n return ent_name\n end\nend\n" }, { "path": "deep-ed/deep-ed-master/data_gen/indexes/yago_crosswikis_wiki.lua", "content": "-- Loads the merged p(e|m) index.\nif not opt then\n cmd = torch.CmdLine()\n cmd:option('-root_data_dir', '', 'Root path of the data, $DATA_PATH.')\n cmd:text()\n opt = cmd:parse(arg or {})\n assert(opt.root_data_dir ~= '', 'Specify a valid root_data_dir path argument.') \nend\n\n\nrequire 'torch'\ntds = tds or require 'tds' \n\ndofile 'utils/utils.lua'\ndofile 'entities/ent_name2id_freq/ent_name_id.lua'\n\nent_p_e_m_index = tds.Hash()\n\nmention_lower_to_one_upper = tds.Hash()\n\nmention_total_freq = tds.Hash()\n\nlocal crosswikis_textfilename = opt.root_data_dir .. 'generated/crosswikis_wikipedia_p_e_m.txt'\nprint('==> Loading crosswikis_wikipedia from file ' .. crosswikis_textfilename)\nlocal it, _ = io.open(crosswikis_textfilename)\nlocal line = it:read()\n\nlocal num_lines = 0\nwhile (line) do\n num_lines = num_lines + 1\n if num_lines % 2000000 == 0 then\n print('Processed ' .. num_lines .. ' lines. ')\n end\n\n local parts = split(line , '\\t')\n local mention = parts[1]\n\n local total = tonumber(parts[2])\n assert(total)\n if total >= 1 then\n ent_p_e_m_index[mention] = tds.Hash()\n mention_lower_to_one_upper[mention:lower()] = mention\n mention_total_freq[mention] = total\n local num_parts = table_len(parts)\n for i = 3, num_parts do\n local ent_str = split(parts[i], ',')\n local ent_wikiid = tonumber(ent_str[1])\n local freq = tonumber(ent_str[2])\n assert(ent_wikiid)\n assert(freq)\n ent_p_e_m_index[mention][ent_wikiid] = freq / (total + 0.0) -- not sorted\n end\n end\n line = it:read()\nend\n\n\nlocal yago_textfilename = opt.root_data_dir .. 'generated/yago_p_e_m.txt'\nprint('==> Loading yago index from file ' .. yago_textfilename)\nit, _ = io.open(yago_textfilename)\nline = it:read()\n\nnum_lines = 0\nwhile (line) do\n num_lines = num_lines + 1\n if num_lines % 2000000 == 0 then\n print('Processed ' .. num_lines .. ' lines. ')\n end\n\n local parts = split(line , '\\t')\n local mention = parts[1]\n\n local total = tonumber(parts[2])\n assert(total)\n if total >= 1 then\n mention_lower_to_one_upper[mention:lower()] = mention\n if not mention_total_freq[mention] then\n mention_total_freq[mention] = total\n else\n mention_total_freq[mention] = total + mention_total_freq[mention]\n end\n \n local yago_ment_ent_idx = tds.Hash()\n local num_parts = table_len(parts)\n for i = 3, num_parts do\n local ent_str = split(parts[i], ',')\n local ent_wikiid = tonumber(ent_str[1])\n local freq = 1\n assert(ent_wikiid)\n yago_ment_ent_idx[ent_wikiid] = freq / (total + 0.0) -- not sorted\n end\n \n if not ent_p_e_m_index[mention] then\n ent_p_e_m_index[mention] = yago_ment_ent_idx\n else\n for ent_wikiid,prob in pairs(yago_ment_ent_idx) do\n if not ent_p_e_m_index[mention][ent_wikiid] then\n ent_p_e_m_index[mention][ent_wikiid] = 0.0\n end\n ent_p_e_m_index[mention][ent_wikiid] = math.min(1.0, ent_p_e_m_index[mention][ent_wikiid] + prob)\n end\n \n end\n \n end\n line = it:read()\nend\n\n\n-- nikos code. serialize the dictionary to a file\nif true then -- make it false so it is not executed every time\n print('Now sorting and writing ..')\n out_file = opt.root_data_dir .. 'generated/prob_yago_crosswikis_wikipedia_p_e_m.txt'\n ouf = assert(io.open(out_file, \"w\"))\n \n for mention, list in pairs(ent_p_e_m_index) do\n if mention:len() >= 1 then\n local tbl = {}\n -- freq in the following lines is actually prob (probability and not absolute count)\n for ent_wikiid, freq in pairs(list) do\n table.insert(tbl, {ent_wikiid = ent_wikiid, freq = freq})\n end\n table.sort(tbl, function(a,b) return a.freq > b.freq end)\n \n local str = ''\n local total_freq = 0\n local num_ents = 0\n for _,el in pairs(tbl) do\n -- is this entity is in the entity name - > entity id map. entities/ent_name2id_freq/ent_name_id.lua if e_id_name.ent_wikiid2name[ent_wikiid] then\n if is_valid_ent(el.ent_wikiid) then \n str = str .. el.ent_wikiid .. ',' .. el.freq \n str = str .. ',' .. get_ent_name_from_wikiid(el.ent_wikiid):gsub(' ', '_') .. '\\t'\n num_ents = num_ents + 1\n total_freq = total_freq + el.freq\n \n if num_ents >= 100 then -- At most 100 candidates\n break\n end\n end\n end\n ouf:write(mention .. '\\t' .. mention_total_freq[mention] .. '\\t' .. str .. '\\n')\n end -- instead of writing the total_freq (the probability which is [0,2] and no\n -- usefull information, i write the absolute total freq.\n end\n ouf:flush()\n io.close(ouf)\n \n print(' Done sorting and writing.')\nend\n-- end of my code\n\n\n\n\n\nassert(ent_p_e_m_index['Dejan Koturovic'] and ent_p_e_m_index['Jose Luis Caminero'])\n\n-- Function used to preprocess a given mention such that it has higher \n-- chance to have at least one valid entry in the p(e|m) index.\nfunction preprocess_mention(m)\n assert(ent_p_e_m_index and mention_total_freq)\n local cur_m = modify_uppercase_phrase(m)\n if (not ent_p_e_m_index[cur_m]) then\n cur_m = m\n end\n if (mention_total_freq[m] and (mention_total_freq[m] > mention_total_freq[cur_m])) then\n cur_m = m -- Cases like 'U.S.' are handed badly by modify_uppercase_phrase\n end\n -- If we cannot find the exact mention in our index, we try our luck to find it in a case insensitive index.\n if (not ent_p_e_m_index[cur_m]) and mention_lower_to_one_upper[cur_m:lower()] then\n cur_m = mention_lower_to_one_upper[cur_m:lower()]\n end\n return cur_m\nend\n\n\nprint(' Done loading index')\n\n\n\n\n\n\n\n\n\n\n\n\n" }, { "path": "deep-ed/deep-ed-master/data_gen/parse_wiki_dump/parse_wiki_dump_tools.lua", "content": "-- Utility functions to extract the text and hyperlinks from each page in the Wikipedia corpus.\n\nif not table_len then\n dofile 'utils/utils.lua'\nend\nif not get_redirected_ent_title then\n dofile 'data_gen/indexes/wiki_redirects_index.lua'\nend\nif not get_ent_name_from_wikiid then\n dofile 'entities/ent_name2id_freq/ent_name_id.lua'\nend\n\n\nfunction extract_text_and_hyp(line, mark_mentions)\n local list_hyp = {} -- (mention, entity) pairs\n local text = ''\n local list_ent_errors = 0\n local parsing_errors = 0\n local disambiguation_ent_errors = 0\n local diez_ent_errors = 0\n \n local end_end_hyp = 0\n local begin_end_hyp = 0\n local begin_start_hyp, end_start_hyp = line:find('', end_start_hyp + 1)\n if next_quotes then\n local ent_name = line:sub(end_start_hyp + 1, next_quotes - 1)\n begin_end_hyp, end_end_hyp = line:find('', end_quotes + 1)\n if begin_end_hyp then\n local mention = line:sub(end_quotes + 1, begin_end_hyp - 1)\n local mention_marker = false\n \n local good_mention = true\n good_mention = good_mention and (not mention:find('Wikipedia'))\n good_mention = good_mention and (not mention:find('wikipedia'))\n good_mention = good_mention and (mention:len() >= 1)\n \n if good_mention then\n local i = ent_name:find('wikt:')\n if i == 1 then\n ent_name = ent_name:sub(6)\n end\n ent_name = preprocess_ent_name(ent_name)\n\n i = ent_name:find('List of ')\n if (not i) or (i ~= 1) then\n if ent_name:find('#') then\n diez_ent_errors = diez_ent_errors + 1\n else\n local ent_wikiid = get_ent_wikiid_from_name(ent_name, true)\n if ent_wikiid == unk_ent_wikiid then\n disambiguation_ent_errors = disambiguation_ent_errors + 1\n else\n -- A valid (entity,mention) pair\n num_mentions = num_mentions + 1\n table.insert(list_hyp, {mention = mention, ent_wikiid = ent_wikiid, cnt = num_mentions})\n if mark_mentions then\n mention_marker = true\n end\n end\n end\n else\n list_ent_errors = list_ent_errors + 1\n end\n end\n \n if (not mention_marker) then\n text = text .. ' ' .. mention .. ' '\n else\n text = text .. ' MMSTART' .. num_mentions .. ' ' .. mention .. ' MMEND' .. num_mentions .. ' '\n end\n else \n parsing_errors = parsing_errors + 1\n begin_start_hyp = nil\n end\n else \n parsing_errors = parsing_errors + 1\n begin_start_hyp = nil\n end\n \n if begin_start_hyp then\n begin_start_hyp, end_start_hyp = line:find(' ' .. line:sub(startoff + 1, startquotes - 1))\n local starttitlestartoff, starttitleendoff = line:find(' title=\"')\n local endtitleoff, _ = line:find('\">')\n local ent_name = line:sub(starttitleendoff + 1, endtitleoff - 1)\n if (ent_wikiid ~= get_ent_wikiid_from_name(ent_name, true)) then\n -- Most probably this is a disambiguation or list page\n local new_ent_wikiid = get_ent_wikiid_from_name(ent_name, true)\n-- print(red('Error in Wiki dump: ' .. line .. ' ' .. ent_wikiid .. ' ' .. new_ent_wikiid))\n return new_ent_wikiid\n end\n return ent_wikiid\nend\n\n\nlocal test_line_4 = ''\n\nprint(extract_page_entity_title(test_line_4))\n" }, { "path": "deep-ed/deep-ed-master/ed/args.lua", "content": "-- We add params abbreviations to the abbv map if they are important hyperparameters\n-- used to differentiate between different methods.\nabbv = {}\n\ncmd = torch.CmdLine()\ncmd:text()\ncmd:text('Deep Joint Entity Disambiguation w/ Local Neural Attention')\ncmd:text('Command line options:')\n\n---------------- runtime options:\ncmd:option('-root_data_dir', '', 'Root path of the data, $DATA_PATH.')\n\ncmd:option('-unit_tests', false, 'Run unit tests or not')\n\n---------------- CUDA:\ncmd:option('-type', 'cudacudnn', 'Type: cuda | float | cudacudnn')\n\n---------------- train data:\ncmd:option('-store_train_data', 'RAM', 'Where to read the training data from: RAM (tensors) | DISK (text, parsed all the time)')\n\n---------------- loss:\ncmd:option('-loss', 'max-margin', 'Loss: nll | max-margin')\nabbv['-loss'] = ''\n\n---------------- optimization:\ncmd:option('-opt', 'ADAM', 'Optimization method: SGD | ADADELTA | ADAGRAD | ADAM')\nabbv['-opt'] = ''\n\ncmd:option('-lr', 1e-4, 'Learning rate. Will be divided by 10 after validation F1 >= 90%.')\nabbv['-lr'] = 'lr'\n\ncmd:option('-batch_size', 1, 'Batch size in terms of number of documents.')\nabbv['-batch_size'] = 'bs'\n\n---------------- word vectors\ncmd:option('-word_vecs', 'w2v', 'Word vectors type: glove | w2v (Word2Vec)')\nabbv['-word_vecs'] = ''\n\n---------------- entity vectors\ncmd:option('-entities', 'RLTD', 'Which entity vectors to use, either just those that appear as candidates in all datasets, or all. All is impractical when storing on GPU. RLTD | ALL')\n\ncmd:option('-ent_vecs_filename', 'ent_vecs__ep_228.t7', 'File name containing entity vectors generated with entities/learn_e2v/learn_a.lua.')\n \n---------------- context\ncmd:option('-ctxt_window', 100, 'Number of context words at the left plus right of each mention')\nabbv['-ctxt_window'] = 'ctxtW'\n\ncmd:option('-R', 25, 'Hard attention threshold: top R context words are kept, the rest are discarded.')\nabbv['-R'] = 'R'\n\n---------------- model\ncmd:option('-model', 'global', 'ED model: local | global')\n\ncmd:option('-nn_pem_interm_size', 100, 'Number of hidden units in the f function described in Section 4 - Local score combination.')\nabbv['-nn_pem_interm_size'] = 'nnPEMintermS'\n\n-------------- model regularization:\ncmd:option('-mat_reg_norm', 1, 'Maximum norm of columns of matrices of the f network.')\nabbv['-mat_reg_norm'] = 'matRegNorm'\n\n---------------- global model parameters\ncmd:option('-lbp_iter', 10, 'Number iterations of LBP hard-coded in a NN. Referred as T in the paper.')\nabbv['-lbp_iter'] = 'lbpIt'\n\ncmd:option('-lbp_damp', 0.5, 'Damping factor for LBP')\nabbv['-lbp_damp'] = 'lbpDamp'\n\n----------------- reranking of candidates\ncmd:option('-num_cand_before_rerank', 30, '')\nabbv['-num_cand_before_rerank'] = 'numERerank'\n\ncmd:option('-keep_p_e_m', 4, '')\nabbv['-keep_p_e_m'] = 'keepPEM'\n\ncmd:option('-keep_e_ctxt', 3, '')\nabbv['-keep_e_ctxt'] = 'keepEC'\n\n----------------- coreference:\ncmd:option('-coref', true, 'Coreference heuristic to match persons names.')\nabbv['-coref'] = 'coref'\n\n------------------ test one model with saved pretrained parameters\ncmd:option('-test_one_model_file', '', 'Saved pretrained model filename from folder $DATA_PATH/generated/ed_models/.')\n\n------------------ banner:\ncmd:option('-banner_header', '', ' Banner header to be used for plotting')\n\ncmd:text()\nopt = cmd:parse(arg or {})\n\n-- Whether to save the current ED model or not during training.\n-- It will become true after the model gets > 90% F1 score on validation set (see test.lua).\nopt.save = false \n\n-- Creates a nice banner from the command line arguments\nfunction get_banner(arg, abbv)\n local num_args = table_len(arg)\n local banner = opt.banner_header \n\n if opt.model == 'global' then\n banner = banner .. 'GLB'\n else\n banner = banner .. 'LCL'\n end\n \n for i = 1,num_args do\n if abbv[arg[i]] then\n banner = banner .. '|' .. abbv[arg[i]] .. '=' .. tostring(opt[arg[i]:sub(2)])\n end\n end\n return banner\nend\n\n\nfunction serialize_params(arg)\n local num_args = table_len(arg)\n local str = opt.banner_header \n if opt.banner_header:len() > 0 then\n str = str .. '|'\n end\n \n str = str .. 'model=' .. opt.model\n\n for i = 1,num_args do\n if abbv[arg[i]] then\n str = str .. '|' .. arg[i]:sub(2) .. '=' .. tostring(opt[arg[i]:sub(2)])\n end\n end\n return str\nend\n\nbanner = get_banner(arg, abbv)\n\nparams_serialized = serialize_params(arg)\nprint('PARAMS SERIALIZED: ' .. params_serialized)\nprint('BANNER : ' .. banner .. '\\n')\n\nassert(params_serialized:len() < 255, 'Parameters string length should be < 255.')\n\n\nfunction extract_args_from_model_title(title)\n local x,y = title:find('model=')\n local parts = split(title:sub(x), '|')\n for _,part in pairs(parts) do\n if string.find(part, '=') then\n local components = split(part, '=')\n opt[components[1]] = components[2]\n if tonumber(components[2]) then\n opt[components[1]] = tonumber(components[2])\n end\n if components[2] == 'true' then\n opt[components[1]] = true\n end\n if components[2] == 'false' then\n opt[components[1]] = false\n end\n end\n end\nend\n" }, { "path": "deep-ed/deep-ed-master/ed/ed.lua", "content": "require 'optim'\nrequire 'torch'\nrequire 'gnuplot'\nrequire 'nn'\nrequire 'xlua'\n\ntds = tds or require 'tds'\ndofile 'utils/utils.lua'\n\nprint('\\n' .. green('==========> TRAINING of ED NEURAL MODELS <==========') .. '\\n')\n \ndofile 'ed/args.lua'\n\nprint('===> RUN TYPE (CPU/GPU): ' .. opt.type)\n \ntorch.setdefaulttensortype('torch.FloatTensor')\nif string.find(opt.type, 'cuda') then\n print('==> switching to CUDA (GPU)')\n require 'cunn'\n require 'cutorch'\n require 'cudnn'\n cudnn.benchmark = true \n cudnn.fastest = true\nelse\n print('==> running on CPU')\nend\n\ndofile 'utils/logger.lua'\ndofile 'entities/relatedness/relatedness.lua'\ndofile 'entities/ent_name2id_freq/ent_name_id.lua'\ndofile 'entities/ent_name2id_freq/e_freq_index.lua'\ndofile 'words/load_w_freq_and_vecs.lua' -- w ids\ndofile 'words/w2v/w2v.lua'\ndofile 'entities/pretrained_e2v/e2v.lua'\ndofile 'ed/minibatch/build_minibatch.lua'\ndofile 'ed/minibatch/data_loader.lua'\ndofile 'ed/models/model.lua'\ndofile 'ed/loss.lua'\ndofile 'ed/train.lua'\ndofile 'ed/test/test.lua'\n\ngeom_unit_tests() -- Show some entity examples\ncompute_relatedness_metrics(entity_similarity) -- UNCOMMENT\n\ntrain_and_test()\n" }, { "path": "deep-ed/deep-ed-master/ed/loss.lua", "content": "if opt.loss == 'nll' then\n criterion = nn.CrossEntropyCriterion()\nelse\n -- max-margin with margin parameter = 0.01\n criterion = nn.MultiMarginCriterion(1, torch.ones(max_num_cand), 0.01)\nend\n\nif string.find(opt.type, 'cuda') then\n criterion = criterion:cuda()\nend" }, { "path": "deep-ed/deep-ed-master/ed/minibatch/build_minibatch.lua", "content": "-- Builds and fills a minibatch. In our case, minibatches are variable sized because they\n-- contain all mentions in opt.batch_size documents.\n\nassert(unk_ent_wikiid)\n\n-- For train & test.\nfunction empty_minibatch_with_ids(num_mentions)\n local inputs = {}\n -- ctxt_w_ids\n inputs[1] = {}\n inputs[1][1] = torch.ones(num_mentions, opt.ctxt_window):int():mul(unk_w_id)\n -- TO BE FILLED : inputs[1][2] = CTXT WORD VECTORS : num_mentions x opt.ctxt_window x ent_vecs_size\n \n -- e_wikiids\n inputs[2] = {}\n inputs[2][1] = torch.ones(num_mentions, opt.num_cand_before_rerank):int():mul(unk_ent_thid)\n -- TO BE FILLED : inputs[2][2] = ENT VECTORS: num_mentions x opt.num_cand_before_rerank x ent_vecs_size\n \n -- p(e|m)\n inputs[3] = torch.zeros(num_mentions, opt.num_cand_before_rerank) \n \n return inputs\nend\n\n\n-- Parse context words:\nlocal function parse_context(parts) \n local ctxt_word_ids = torch.ones(opt.ctxt_window):int():mul(unk_w_id)\n \n local lc = parts[4]\n local lc_words = split(lc, ' ')\n local lc_words_size = table_len(lc_words)\n local j = opt.ctxt_window / 2\n local i = lc_words_size\n while (j >= 1 and i >= 1) do\n while (i >= 2 and get_id_from_word(lc_words[i]) == unk_w_id) do\n i = i - 1\n end\n ctxt_word_ids[j] = get_id_from_word(lc_words[i])\n j = j - 1\n i = i - 1\n end\n \n local rc = parts[5]\n local rc_words = split(rc, ' ')\n local rc_words_size = table_len(rc_words)\n j = (opt.ctxt_window / 2) + 1\n i = 1\n while (j <= opt.ctxt_window and i <= rc_words_size) do\n while (i < rc_words_size and get_id_from_word(rc_words[i]) == unk_w_id) do\n i = i + 1\n end\n ctxt_word_ids[j] = get_id_from_word(rc_words[i])\n j = j + 1\n i = i + 1\n end\n \n return ctxt_word_ids\nend\n\n\n---------------------- Entity candidates: ----------------\nlocal function parse_num_cand_and_grd_trth(parts)\n assert(parts[6] == 'CANDIDATES') \n if parts[7] == 'EMPTYCAND' then\n return 0\n else\n local num_cand = 1\n while parts[7 + num_cand] ~= 'GT:' do\n num_cand = num_cand + 1\n end\n return num_cand\n end\nend\n\n--- Collect the grd truth label:\n-- @return grd_trth_idx, grd_trth_ent_wikiid, grd_trth_prob\nlocal function get_grd_trth(parts, num_cand, for_training)\n assert(parts[7 + math.max(1, num_cand)] == 'GT:')\n local grd_trth_str = parts[8 + math.max(1, num_cand)]\n local grd_trth_parts = split(grd_trth_str, ',')\n local grd_trth_idx = tonumber(grd_trth_parts[1])\n if grd_trth_idx ~= -1 then\n assert(grd_trth_idx >= 1 and grd_trth_idx <= num_cand)\n assert(grd_trth_str == grd_trth_idx .. ',' .. parts[6 + grd_trth_idx])\n else\n assert(not for_training)\n end \n local grd_trth_prob = 0\n if grd_trth_idx > 0 then\n grd_trth_prob = math.min(1.0, math.max(1e-3, tonumber(grd_trth_parts[3])))\n end\n local grd_trth_ent_wikiid = unk_ent_wikiid\n if table_len(grd_trth_parts) >= 2 then\n grd_trth_ent_wikiid = tonumber(grd_trth_parts[2])\n end\n assert(grd_trth_ent_wikiid and grd_trth_prob)\n return grd_trth_idx, grd_trth_ent_wikiid, grd_trth_prob\nend\n\n\n-- @return grd_trth_idx, grd_trth_ent_wikiid, ent_cand_wikiids, log_p_e_m, log_p_e\nfunction parse_candidate_entities(parts, for_training, orig_max_num_cand)\n -- Num of entity candidates\n local num_cand = parse_num_cand_and_grd_trth(parts)\n \n -- Ground truth index in the set of entity candidates, wikiid, p(e|m)\n local grd_trth_idx, grd_trth_ent_wikiid, grd_trth_prob = get_grd_trth(parts, num_cand, for_training)\n \n -- When including coreferent mentions this might happen\n if grd_trth_idx > orig_max_num_cand then \n grd_trth_idx = -1\n end\n \n -- P(e|m) prior:\n local log_p_e_m = torch.ones(orig_max_num_cand):mul(-1e8)\n \n -- Vector of entity candidates ids\n local ent_cand_wikiids = torch.ones(orig_max_num_cand):int():mul(unk_ent_wikiid)\n \n -- Parse all candidates\n for cand_index = 1,math.min(num_cand, orig_max_num_cand) do\n local cand_parts = split(parts[6 + cand_index], ',')\n local cand_ent_wikiid = tonumber(cand_parts[1])\n assert(cand_ent_wikiid)\n ent_cand_wikiids[cand_index] = cand_ent_wikiid\n \n assert(for_training or get_thid(cand_ent_wikiid) ~= unk_ent_thid) -- RLTD entities have valid id \n \n local cand_p_e_m = math.min(1.0, math.max(1e-3, tonumber(cand_parts[2])))\n log_p_e_m[cand_index] = torch.log(cand_p_e_m)\n local cand_p_e = get_ent_freq(cand_ent_wikiid)\n end\n \n -- Reinsert grd truth for training only\n if grd_trth_idx == -1 and for_training then\n assert(num_cand >= orig_max_num_cand)\n grd_trth_idx = orig_max_num_cand\n ent_cand_wikiids[grd_trth_idx] = grd_trth_ent_wikiid\n log_p_e_m[grd_trth_idx] = torch.log(grd_trth_prob)\n end\n \n -- Sanity checks:\n assert(log_p_e_m[1] == torch.max(log_p_e_m), line)\n assert(log_p_e_m[2] == torch.max(log_p_e_m:narrow(1, 2, orig_max_num_cand - 1)), line)\n if (grd_trth_idx ~= -1) then\n assert(grd_trth_ent_wikiid ~= unk_ent_wikiid)\n assert(ent_cand_wikiids[grd_trth_idx] == grd_trth_ent_wikiid)\n assert(log_p_e_m[grd_trth_idx] == torch.log(grd_trth_prob))\n else \n assert(not for_training)\n end\n \n return grd_trth_idx, grd_trth_ent_wikiid, ent_cand_wikiids, log_p_e_m\nend\n\nlocal function get_cand_ent_thids(minibatch_tensor)\n return minibatch_tensor[2][1]\nend\n\n\n\n-- Fills in the minibatch and returns the idx of the grd truth entity:\nfunction process_one_line(line, minibatch_tensor, mb_index, for_training)\n local parts = split(line, '\\t') \n \n -- Ctxt word ids:\n local ctxt_word_ids = parse_context(parts)\n minibatch_tensor[1][1][mb_index] = ctxt_word_ids\n \n -- Entity candidates:\n local grd_trth_idx, grd_trth_ent_wikiid, ent_cand_wikiids, log_p_e_m = \n parse_candidate_entities(parts, for_training, opt.num_cand_before_rerank)\n minibatch_tensor[2][1][mb_index] = get_ent_thids(ent_cand_wikiids)\n assert(grd_trth_idx == -1 or (get_wikiid_from_thid(minibatch_tensor[2][1][mb_index][grd_trth_idx]) == grd_trth_ent_wikiid))\n \n -- log p(e|m):\n minibatch_tensor[3][mb_index] = log_p_e_m\n \n return grd_trth_idx\nend\n\n\n-- Reranking of entity candidates:\n---- Keeps only 4 candidates, top 2 candidates from p(e|m) and top 2 from \nmax_num_cand = opt.keep_p_e_m + opt.keep_e_ctxt \n\nlocal function rerank(minibatch_tensor, targets, for_training)\n local num_mentions = get_cand_ent_thids(minibatch_tensor):size(1)\n local new_targets = torch.ones(num_mentions):mul(-1)\n \n -- Average of word vectors in a window of (at most) size 50 around the mention.\n local ctxt_vecs = minibatch_tensor[1][2]\n if opt.ctxt_window > 50 then \n ctxt_vecs = ctxt_vecs:narrow(2, math.floor(opt.ctxt_window / 2) - 25 + 1, 50)\n end\n ctxt_vecs = ctxt_vecs:sum(2):view(num_mentions, ent_vecs_size)\n \n local new_log_p_e_m = correct_type(torch.ones(num_mentions, max_num_cand):mul(-1e8))\n local new_ent_cand_wikiids = torch.ones(num_mentions, max_num_cand):int():mul(unk_ent_wikiid) \n local new_ent_cand_vecs = correct_type(torch.zeros(num_mentions, max_num_cand, ent_vecs_size))\n \n for k = 1,num_mentions do\n local ent_vecs = minibatch_tensor[2][2][k]\n local scores = ent_vecs * ctxt_vecs[k]\n assert(scores:size(1) == opt.num_cand_before_rerank)\n local _,ctxt_indices = torch.sort(scores, true)\n \n local added_indices = {} \n for j = 1,opt.keep_e_ctxt do\n added_indices[ctxt_indices[j]] = 1\n end\n local j = 1\n while table_len(added_indices) < max_num_cand do\n added_indices[j] = 1\n j = j + 1\n end\n \n local new_grd_trth_idx = -1\n local i = 1\n for idx,_ in pairs(added_indices) do\n new_ent_cand_wikiids[k][i] = minibatch_tensor[2][1][k][idx]\n new_ent_cand_vecs[k][i] = minibatch_tensor[2][2][k][idx]\n new_log_p_e_m[k][i] = minibatch_tensor[3][k][idx]\n \n if idx == targets[k] then\n assert(minibatch_tensor[2][1][k][idx] ~= unk_ent_wikiid)\n new_grd_trth_idx = i\n end\n i = i + 1\n end\n\n -- Grd trth:\n if targets[k] > 0 and (new_grd_trth_idx == -1) then\n assert(targets[k] > opt.keep_p_e_m)\n-- if not for_training then\n-- print('Grd trth idx became -1') ------------> 79 times on AIDA test B for max_num_cand = 2 + 2\n-- end\n if for_training then -- Reinsert the grd truth entity for training only\n new_ent_cand_wikiids[k][1] = minibatch_tensor[2][1][k][targets[k]]\n new_ent_cand_vecs[k][1] = minibatch_tensor[2][2][k][targets[k]]\n new_log_p_e_m[k][1] = minibatch_tensor[3][k][targets[k]]\n new_grd_trth_idx = 1\n end\n end\n \n new_targets[k] = new_grd_trth_idx\n end\n \n minibatch_tensor[2][1] = new_ent_cand_wikiids\n minibatch_tensor[2][2] = new_ent_cand_vecs\n minibatch_tensor[3] = new_log_p_e_m\n \n return minibatch_tensor, new_targets\nend\n\n\n-- Convert mini batch to correct type (e.g. move data to GPU):\n-- ATT: Since lookup_table:forward() cannot be called twice without changing the output,\n-- we have to keep the order here: \nfunction minibatch_to_correct_type(minibatch_tensor, targets, for_training)\n -- ctxt_w_vecs : num_mentions x opt.ctxt_window x ent_vecs_size\n minibatch_tensor[1][2] = word_lookup_table:forward(minibatch_tensor[1][1])\n \n -- ent_vecs : num_mentions x max_num_cand x ent_vecs_size\n minibatch_tensor[2][2] = ent_lookup_table:forward(minibatch_tensor[2][1])\n \n -- log p(e|m) : num_mentions x max_num_cand\n minibatch_tensor[3] = correct_type(minibatch_tensor[3])\n \n return rerank(minibatch_tensor, targets, for_training)\nend\n\n\n--- Used in data_loader.lua\nfunction minibatch_table2tds(inputs_table)\n local inputs = tds.Hash()\n inputs[1] = tds.Hash()\n inputs[1][1] = inputs_table[1][1]\n inputs[2] = tds.Hash()\n inputs[2][1] = inputs_table[2][1]\n inputs[3] = inputs_table[3]\n return inputs\nend\n\nfunction minibatch_tds2table(inputs_tds)\n local inputs = {}\n inputs[1] = {}\n inputs[1][1] = inputs_tds[1][1]\n inputs[2] = {}\n inputs[2][1] = inputs_tds[2][1]\n inputs[3] = inputs_tds[3]\n return inputs\nend\n\n\n-- Utility functions used in test file for debug/vizualization\nfunction get_cand_ent_wikiids(minibatch_tensor)\n local num_mentions = get_cand_ent_thids(minibatch_tensor):size(1)\n assert(get_cand_ent_thids(minibatch_tensor):size(2) == max_num_cand)\n local t = torch.zeros(num_mentions, max_num_cand):int()\n for i = 1, num_mentions do\n for j = 1, max_num_cand do\n t[i][j] = get_wikiid_from_thid(get_cand_ent_thids(minibatch_tensor)[i][j])\n end\n end\n return t\nend\n\nfunction get_log_p_e_m(minibatch_tensor)\n return minibatch_tensor[3]\nend\n\nfunction get_ctxt_word_ids(minibatch_tensor)\n return minibatch_tensor[1][1]\nend\n" }, { "path": "deep-ed/deep-ed-master/ed/minibatch/data_loader.lua", "content": "-- Data loader for training of ED models.\n\ntrain_file = opt.root_data_dir .. 'generated/test_train_data/aida_train.csv'\nit_train, _ = io.open(train_file)\n\nprint('==> Loading training data with option ' .. opt.store_train_data)\nlocal function one_doc_to_minibatch(doc_lines)\n -- Create empty mini batch:\n local num_mentions = #doc_lines\n assert(num_mentions > 0)\n\n local inputs = empty_minibatch_with_ids(num_mentions)\n local targets = torch.zeros(num_mentions)\n\n -- Fill in each example:\n for i = 1, num_mentions do\n local target = process_one_line(doc_lines[i], inputs, i, true)\n targets[i] = target\n assert(target >= 1 and target == targets[i])\n end\n\n return inputs, targets \nend\n\nif opt.store_train_data == 'RAM' then\n all_docs_inputs = tds.Hash()\n all_docs_targets = tds.Hash()\n doc2id = tds.Hash()\n id2doc = tds.Hash()\n\n local cur_doc_lines = tds.Hash()\n local prev_doc_id = nil\n\n local line = it_train:read()\n while line do\n local parts = split(line, '\\t')\n local doc_name = parts[1]\n if not doc2id[doc_name] then\n if prev_doc_id then\n local inputs, targets = one_doc_to_minibatch(cur_doc_lines)\n all_docs_inputs[prev_doc_id] = minibatch_table2tds(inputs)\n all_docs_targets[prev_doc_id] = targets\n end\n local cur_docid = 1 + #doc2id\n id2doc[cur_docid] = doc_name\n doc2id[doc_name] = cur_docid\n cur_doc_lines = tds.Hash()\n prev_doc_id = cur_docid\n end\n cur_doc_lines[1 + #cur_doc_lines] = line\n line = it_train:read()\n end\n if prev_doc_id then\n local inputs, targets = one_doc_to_minibatch(cur_doc_lines)\n all_docs_inputs[prev_doc_id] = minibatch_table2tds(inputs)\n all_docs_targets[prev_doc_id] = targets\n end \n assert(#doc2id == #all_docs_inputs, #doc2id .. ' ' .. #all_docs_inputs)\n\nelse\n all_doc_lines = tds.Hash()\n doc2id = tds.Hash()\n id2doc = tds.Hash()\n\n local line = it_train:read()\n while line do\n local parts = split(line, '\\t')\n local doc_name = parts[1]\n if not doc2id[doc_name] then\n local cur_docid = 1 + #doc2id\n id2doc[cur_docid] = doc_name\n doc2id[doc_name] = cur_docid\n all_doc_lines[cur_docid] = tds.Hash()\n end\n all_doc_lines[doc2id[doc_name]][1 + #all_doc_lines[doc2id[doc_name]]] = line\n line = it_train:read()\n end\n assert(#doc2id == #all_doc_lines)\nend\n\n\nget_minibatch = function()\n -- Create empty mini batch:\n local inputs = nil\n local targets = nil\n\n if opt.store_train_data == 'RAM' then\n local random_docid = math.random(#id2doc)\n inputs = minibatch_tds2table(all_docs_inputs[random_docid])\n targets = all_docs_targets[random_docid]\n else\n local doc_lines = all_doc_lines[math.random(#id2doc)]\n inputs, targets = one_doc_to_minibatch(doc_lines)\n end\n\n -- Move data to GPU:\n inputs, targets = minibatch_to_correct_type(inputs, targets, true)\n targets = correct_type(targets)\n\n return inputs, targets\nend\n\nprint(' Done loading training data.')\n" }, { "path": "deep-ed/deep-ed-master/ed/models/SetConstantDiag.lua", "content": "-- Torch layer that receives as input a squared matrix and sets its diagonal to a constant value.\n\nlocal SetConstantDiag, parent = torch.class('nn.SetConstantDiag', 'nn.Module')\n\nfunction SetConstantDiag:__init(constant_scalar, ip)\n parent.__init(self)\n assert(type(constant_scalar) == 'number', 'input is not scalar!')\n self.constant_scalar = constant_scalar\n \n -- default for inplace is false\n self.inplace = ip or false\n if (ip and type(ip) ~= 'boolean') then\n error('in-place flag must be boolean')\n end\n if not opt then\n opt = {}\n opt.type = 'double'\n dofile 'utils/utils.lua'\n end\nend\n\nfunction SetConstantDiag:updateOutput(input)\n assert(input:dim() == 3)\n assert(input:size(2) == input:size(3))\n local n = input:size(3)\n local prod_mat = torch.ones(n,n) - torch.eye(n) \n prod_mat = correct_type(prod_mat)\n local sum_mat = torch.eye(n):mul(self.constant_scalar)\n sum_mat = correct_type(sum_mat)\n if self.inplace then\n input:cmul(torch.repeatTensor(prod_mat, input:size(1), 1, 1))\n input:add(torch.repeatTensor(sum_mat, input:size(1), 1, 1))\n self.output:set(input)\n else\n self.output:resizeAs(input)\n self.output:copy(input)\n self.output:cmul(torch.repeatTensor(prod_mat, input:size(1), 1, 1))\n self.output:add(torch.repeatTensor(sum_mat, input:size(1), 1, 1))\n end\n return self.output\nend \n\nfunction SetConstantDiag:updateGradInput(input, gradOutput)\n local n = input:size(3)\n local prod_mat = torch.ones(n,n) - torch.eye(n) \n prod_mat = correct_type(prod_mat)\n if self.inplace then\n self.gradInput:set(gradOutput)\n else\n self.gradInput:resizeAs(gradOutput)\n self.gradInput:copy(gradOutput)\n end\n self.gradInput:cmul(torch.repeatTensor(prod_mat, input:size(1), 1, 1))\n return self.gradInput\nend" }, { "path": "deep-ed/deep-ed-master/ed/models/linear_layers.lua", "content": "-- Define all parametrized layers: linear layers (diagonal matrices) A,B,C + network f\n\nfunction new_linear_layer(out_dim)\n cmul = nn.CMul(out_dim)\n -- init weights with ones to speed up convergence\n cmul.weight = torch.ones(out_dim)\n return cmul\nend\n\n---- Create shared weights\nA_linear = new_linear_layer(ent_vecs_size)\n\n-- Local ctxt bilinear weights\nB_linear = new_linear_layer(ent_vecs_size)\n\n-- Used only in the global model\nC_linear = new_linear_layer(ent_vecs_size)\n\nf_network = nn.Sequential()\n :add(nn.Linear(2,opt.nn_pem_interm_size))\n :add(nn.ReLU())\n :add(nn.Linear(opt.nn_pem_interm_size,1))\n\n\nfunction regularize_f_network()\n if opt.mat_reg_norm < 10 then\n for i = 1,f_network:size() do\n if f_network:get(i).weight and (f_network:get(i).weight:norm() > opt.mat_reg_norm) then\n f_network:get(i).weight:mul(opt.mat_reg_norm / f_network:get(i).weight:norm())\n end\n if f_network:get(i).bias and (f_network:get(i).bias:norm() > opt.mat_reg_norm) then\n f_network:get(i).bias:mul(opt.mat_reg_norm / f_network:get(i).bias:norm())\n end\n end\n end\nend\n\nfunction pack_saveable_weights()\n local linears = nn.Sequential():add(A_linear):add(B_linear):add(C_linear):add(f_network)\n return linears:float()\nend\n\nfunction unpack_saveable_weights(saved_linears)\n A_linear = saved_linears:get(1)\n B_linear = saved_linears:get(2)\n C_linear = saved_linears:get(3)\n f_network = saved_linears:get(4)\nend\n\n\nfunction print_net_weights()\n print('\\nNetwork norms of parameter weights :')\n print('A (attention mat) = ' .. A_linear.weight:norm())\n print('B (ctxt embedding) = ' .. B_linear.weight:norm())\n print('C (pairwise mat) = ' .. C_linear.weight:norm())\n\n if opt.mat_reg_norm < 10 then\n print('f_network norm = ' .. f_network:get(1).weight:norm() .. ' ' ..\n f_network:get(1).bias:norm() .. ' ' .. f_network:get(3).weight:norm() .. ' ' ..\n f_network:get(3).bias:norm()) \n else\n p,gp = f_network:getParameters()\n print('f_network norm = ' .. p:norm())\n end\nend\n" }, { "path": "deep-ed/deep-ed-master/ed/models/model.lua", "content": "dofile 'ed/models/SetConstantDiag.lua'\ndofile 'ed/models/linear_layers.lua'\ndofile 'ed/models/model_local.lua'\ndofile 'ed/models/model_global.lua'\n\nfunction get_model(num_mentions)\n local model_ctxt, additional_local_submodels = local_model(num_mentions, A_linear, B_linear)\n local model = model_ctxt\n \n if opt.model == 'global' then\n model = global_model(num_mentions, model_ctxt, C_linear, f_network)\n else\n assert(opt.model == 'local')\n end\n \n return model, additional_local_submodels \nend\n" }, { "path": "deep-ed/deep-ed-master/ed/models/model_global.lua", "content": "-- Definition of the neural network used for global (joint) ED. Section 5 of our paper. \n-- It unrolls a fixed number of LBP iterations allowing training of the CRF potentials using backprop.\n-- To run a simple unit test that checks the forward and backward passes, just run :\n-- th ed/models/model_global.lua\n\nif not opt then -- unit tests\n dofile 'ed/models/model_local.lua'\n dofile 'ed/models/SetConstantDiag.lua'\n dofile 'ed/models/linear_layers.lua' \n \n opt.lbp_iter = 10\n opt.lbp_damp = 0.5\n opt.model = 'global'\nend\n\n\nfunction global_model(num_mentions, model_ctxt, param_C_linear, param_f_network)\n \n assert(num_mentions)\n assert(model_ctxt)\n assert(param_C_linear)\n assert(param_f_network)\n \n local unary_plus_pairwise = nn.Sequential()\n :add(nn.ParallelTable()\n :add(nn.Sequential() -- Pairwise scores s_{ij}(y_i, y_j)\n :add(nn.View(num_mentions * max_num_cand, ent_vecs_size)) -- e_vecs\n :add(nn.ConcatTable()\n :add(nn.Identity())\n :add(param_C_linear)\n )\n :add(nn.MM(false, true)) -- s_{ij}(y_i, y_j) is s[i][y_i][j][y_j] = \n :add(nn.View(num_mentions, max_num_cand, num_mentions, max_num_cand))\n :add(nn.MulConstant(2.0 / num_mentions, true))\n )\n :add(nn.Sequential() -- Unary scores s_j(y_j)\n :add(nn.Replicate(num_mentions * max_num_cand, 1))\n :add(nn.Reshape(num_mentions, max_num_cand, num_mentions, max_num_cand))\n )\n )\n :add(nn.CAddTable()) -- q[i][y_i][j][y_j] = s_j(y_j) + s_{ij}(y_i, y_j): num_mentions x max_num_cand x num_mentions x max_num_cand\n \n \n -- Input is q[i][y_i][j] : num_mentions x max_num_cand x num_mentions\n local messages_one_round = nn.ConcatTable()\n :add(nn.SelectTable(1)) -- 1. unary_plus_pairwise : num_mentions, max_num_cand, num_mentions, max_num_cand\n :add(nn.Sequential()\n :add(nn.ConcatTable()\n :add(nn.Sequential()\n :add(nn.SelectTable(2)) -- old messages\n :add(nn.Exp())\n :add(nn.MulConstant(1.0 - opt.lbp_damp, false))\n )\n :add(nn.Sequential()\n :add(nn.ParallelTable()\n :add(nn.Identity()) -- unary plus pairwise\n :add(nn.Sequential() -- old messages: num_mentions, max_num_cand, num_mentions\n :add(nn.Sum(3)) -- g[i][y_i] := \\sum_{k != i} q[i][y_i][k]\n :add(nn.Replicate(num_mentions * max_num_cand, 1))\n :add(nn.Reshape(num_mentions, max_num_cand, num_mentions, max_num_cand))\n )\n )\n :add(nn.CAddTable()) -- s_{j}(y_j) + s_{ij}(y_i, y_j) + g[j][y_j] : num_mentions, max_num_cand, num_mentions, max_num_cand\n :add(nn.Max(4)) -- unnorm_q[i][y_i][j] : num_mentions x max_num_cand x num_mentions\n :add(nn.Transpose({2,3})) \n :add(nn.View(num_mentions * num_mentions, max_num_cand))\n :add(nn.LogSoftMax()) -- normalization: \\sum_{y_i} exp(q[i][y_i][j]) = 1\n :add(nn.View(num_mentions, num_mentions, max_num_cand))\n :add(nn.Transpose({2,3})) \n :add(nn.Transpose({1,2}))\n :add(nn.SetConstantDiag(0, true)) -- we make q[i][y_i][i] = 0, \\forall i and y_i\n :add(nn.Transpose({1,2})) \n :add(nn.Exp())\n :add(nn.MulConstant(opt.lbp_damp, false))\n )\n )\n :add(nn.CAddTable()) -- 2. messages for next round: num_mentions, max_num_cand, num_mentions\n :add(nn.Log())\n )\n\n\n local messages_all_rounds = nn.Sequential()\n messages_all_rounds:add(nn.Identity())\n for i = 1, opt.lbp_iter do\n messages_all_rounds:add(messages_one_round:clone('weight','bias','gradWeight','gradBias'))\n end\n\n local model_gl = nn.Sequential()\n :add(nn.ConcatTable()\n :add(nn.Sequential()\n :add(nn.SelectTable(2))\n :add(nn.SelectTable(2)) -- e_vecs : num_mentions, max_num_cand, ent_vecs_size\n )\n :add(model_ctxt) -- unary scores : num_mentions, max_num_cand\n )\n :add(nn.ConcatTable()\n :add(nn.Sequential()\n :add(unary_plus_pairwise)\n :add(nn.ConcatTable()\n :add(nn.Identity())\n :add(nn.Sequential()\n :add(nn.Max(4)) \n :add(nn.MulConstant(0, false)) -- first_round_zero_messages\n )\n )\n :add(messages_all_rounds)\n :add(nn.SelectTable(2))\n :add(nn.Sum(3)) -- \\sum_{j} msgs[i][y_i]: num_mentions x max_num_cand\n )\n :add(nn.SelectTable(2)) -- unary scores : num_mentions x max_num_cand\n )\n :add(nn.CAddTable()) \n :add(nn.LogSoftMax()) -- belief[i][y_i] (lbp marginals in log scale)\n\n \n -- Combine lbp marginals with log p(e|m) using the simple f neural network\n local pem_layer = nn.SelectTable(3)\n model_gl = nn.Sequential()\n :add(nn.ConcatTable()\n :add(nn.Sequential()\n :add(model_gl)\n :add(nn.View(num_mentions * max_num_cand, 1))\n )\n :add(nn.Sequential()\n :add(pem_layer)\n :add(nn.View(num_mentions * max_num_cand, 1))\n ) \n )\n :add(nn.JoinTable(2))\n :add(param_f_network)\n :add(nn.View(num_mentions, max_num_cand))\n\n ------- Cuda conversions:\n if string.find(opt.type, 'cuda') then\n model_gl = model_gl:cuda()\n end\n\n return model_gl\nend\n\n\n--- Unit tests\nif unit_tests_now then\n print('\\n Global network model unit tests:')\n local num_mentions = 13\n \n local inputs = {}\n -- ctxt_w_vecs\n inputs[1] = {}\n inputs[1][1] = torch.ones(num_mentions, opt.ctxt_window):int():mul(unk_w_id)\n inputs[1][2] = torch.randn(num_mentions, opt.ctxt_window, ent_vecs_size)\n -- e_vecs\n inputs[2] = {}\n inputs[2][1] = torch.ones(num_mentions, max_num_cand):int():mul(unk_ent_thid)\n inputs[2][2] = torch.randn(num_mentions, max_num_cand, ent_vecs_size)\n -- p(e|m)\n inputs[3] = torch.log(torch.rand(num_mentions, max_num_cand))\n\n local model_ctxt, _ = local_model(num_mentions, A_linear, B_linear, opt)\n\n local model_gl = global_model(num_mentions, model_ctxt, C_linear, f_network)\n local outputs = model_gl:forward(inputs)\n print(outputs)\n print('MIN: ' .. torch.min(outputs) .. ' MAX: ' .. torch.max(outputs))\n assert(outputs:size(1) == num_mentions and outputs:size(2) == max_num_cand)\n print('Global FWD success!')\n\n model_gl:backward(inputs, torch.randn(num_mentions, max_num_cand))\n print('Global BKWD success!')\n \n parameters,gradParameters = model_gl:getParameters()\n print(parameters:size())\n print(gradParameters:size())\nend" }, { "path": "deep-ed/deep-ed-master/ed/models/model_local.lua", "content": "-- Definition of the local neural network with attention used for local (independent per each mention) ED. \n-- Section 4 of our paper.\n-- To run a simple unit test that checks the forward and backward passes, just run :\n-- th ed/models/model_local.lua\n\nif not opt then -- unit tests\n unit_tests_now = true\n dofile 'utils/utils.lua'\n require 'nn'\n opt = {type = 'double', ctxt_window = 100, R = 25, model = 'local', nn_pem_interm_size = 100}\n \n word_vecs_size = 300\n ent_vecs_size = 300\n max_num_cand = 6\n unk_ent_wikiid = 1\n unk_ent_thid = 1\n unk_w_id = 1\n dofile 'ed/models/linear_layers.lua' \n word_lookup_table = nn.LookupTable(5, ent_vecs_size)\n ent_lookup_table = nn.LookupTable(5, ent_vecs_size)\nelse\n word_lookup_table = nn.LookupTable(w2vutils.M:size(1), ent_vecs_size)\n word_lookup_table.weight = w2vutils.M\n\n ent_lookup_table = nn.LookupTable(e2vutils.lookup:size(1), ent_vecs_size)\n ent_lookup_table.weight = e2vutils.lookup\n\n if string.find(opt.type, 'cuda') then\n word_lookup_table = word_lookup_table:cuda()\n ent_lookup_table = ent_lookup_table:cuda()\n end\nend\n\nassert(word_vecs_size == 300 and ent_vecs_size == 300)\n\n\n----------------- Define the model\nfunction local_model(num_mentions, param_A_linear, param_B_linear)\n \n assert(num_mentions)\n assert(param_A_linear)\n assert(param_B_linear)\n \n model = nn.Sequential()\n \n ctxt_embed_and_ent_lookup = nn.ConcatTable()\n :add(nn.Sequential()\n :add(nn.SelectTable(1))\n :add(nn.SelectTable(2)) -- 1 : Context words W : num_mentions x opt.ctxt_window x ent_vecs_size\n )\n :add(nn.Sequential()\n :add(nn.SelectTable(2))\n :add(nn.SelectTable(2)) -- 2 : Candidate entity vectors E : num_mentions x max_num_cand x ent_vecs_size\n )\n :add(nn.SelectTable(3)) -- 3 : log p(e|m) : num_mentions x max_num_cand\n \n model:add(ctxt_embed_and_ent_lookup)\n\n \n local mem_weights_p_2 = nn.ConcatTable()\n :add(nn.Identity()) -- 1 : {W, E, logp(e|m)}\n :add(nn.Sequential()\n :add(nn.ConcatTable()\n :add(nn.Sequential()\n :add(nn.SelectTable(2))\n :add(nn.View(num_mentions * max_num_cand, ent_vecs_size)) -- E : num_mentions x max_num_cand x ent_vecs_size\n :add(param_A_linear)\n :add(nn.View(num_mentions, max_num_cand, ent_vecs_size)) -- (A*) E \n )\n :add(nn.SelectTable(1)) --- W : num_mentions x opt.ctxt_window x ent_vecs_size\n )\n :add(nn.MM(false, true)) -- 2 : E^t * A * W : num_mentions x max_num_cand x opt.ctxt_window\n )\n model:add(mem_weights_p_2)\n\n\n local mem_weights_p_3 = nn.ConcatTable()\n :add(nn.SelectTable(1)) -- 1 : {W, E, logp(e|m)}\n :add(nn.Sequential()\n :add(nn.SelectTable(2))\n :add(nn.Max(2)) --- 2 : max(word-entity scores) : num_mentions x opt.ctxt_window\n ) \n\n model:add(mem_weights_p_3)\n\n\n local mem_weights_p_4 = nn.ConcatTable()\n :add(nn.SelectTable(1)) -- 1 : {W, E, logp(e|m)}\n :add(nn.Sequential()\n :add(nn.SelectTable(2))\n -- keep only top K scored words\n :add(nn.ConcatTable()\n :add(nn.Identity()) -- all w-e scores\n :add(nn.Sequential() \n :add(nn.View(num_mentions, opt.ctxt_window, 1))\n :add(nn.TemporalDynamicKMaxPooling(opt.R))\n :add(nn.Min(2)) -- k-th largest w-e score\n :add(nn.View(num_mentions))\n :add(nn.Replicate(opt.ctxt_window, 2))\n )\n )\n :add(nn.ConcatTable() -- top k w-e scores (the rest are set to -infty)\n :add(nn.SelectTable(2)) -- k-th largest w-e score that we substract and then add again back after nn.Threshold\n :add(nn.Sequential()\n :add(nn.CSubTable())\n :add(nn.Threshold(0, -50, true))\n )\n )\n :add(nn.CAddTable())\n :add(nn.SoftMax()) -- 2 : sigma (attention weights normalized): num_mentions x opt.ctxt_window\n :add(nn.View(num_mentions, opt.ctxt_window, 1))\n ) \n\n model:add(mem_weights_p_4)\n\n\n local ctxt_full_embeddings = nn.ConcatTable()\n :add(nn.SelectTable(1)) -- 1 : {W, E, logp(e|m)}\n :add(nn.Sequential()\n :add(nn.ConcatTable()\n :add(nn.Sequential()\n :add(nn.SelectTable(1))\n :add(nn.SelectTable(1)) -- W\n )\n :add(nn.SelectTable(2)) -- sigma\n )\n :add(nn.MM(true, false)) \n :add(nn.View(num_mentions, ent_vecs_size)) -- 2 : ctxt embedding = (W * B)^\\top * sigma : num_mentions x ent_vecs_size\n )\n\n model:add(ctxt_full_embeddings) \n\n\n local entity_context_sim_scores = nn.ConcatTable()\n :add(nn.SelectTable(1)) -- 1 : {W, E, logp(e|m)}\n :add(nn.Sequential()\n :add(nn.ConcatTable()\n :add(nn.Sequential()\n :add(nn.SelectTable(1))\n :add(nn.SelectTable(2)) -- E\n )\n :add(nn.Sequential()\n :add(nn.SelectTable(2))\n :add(param_B_linear)\n :add(nn.View(num_mentions, ent_vecs_size, 1)) -- context vectors\n )\n )\n :add(nn.MM()) --> 2. context * E^T\n :add(nn.View(num_mentions, max_num_cand))\n )\n \n model:add(entity_context_sim_scores)\n\n if opt.model == 'local' then\n model = nn.Sequential()\n :add(nn.ConcatTable()\n :add(nn.Sequential()\n :add(model)\n :add(nn.SelectTable(2)) -- context - entity similarity scores\n :add(nn.View(num_mentions * max_num_cand, 1))\n )\n :add(nn.Sequential()\n :add(nn.SelectTable(3)) -- log p(e|m) scores\n :add(nn.View(num_mentions * max_num_cand, 1))\n ) \n )\n :add(nn.JoinTable(2))\n :add(f_network)\n :add(nn.View(num_mentions, max_num_cand))\n else\n model:add(nn.SelectTable(2)) -- context - entity similarity scores\n end\n \n \n ------------- Visualizing weights: \n\n -- sigma (attention weights normalized): num_mentions x opt.ctxt_window\n local model_debug_softmax_word_weights = nn.Sequential()\n :add(ctxt_embed_and_ent_lookup)\n :add(mem_weights_p_2)\n :add(mem_weights_p_3)\n :add(mem_weights_p_4)\n :add(nn.SelectTable(2))\n :add(nn.View(num_mentions, opt.ctxt_window))\n \n ------- Cuda conversions:\n if string.find(opt.type, 'cuda') then\n model = model:cuda()\n model_debug_softmax_word_weights = model_debug_softmax_word_weights:cuda()\n end\n\n local additional_local_submodels = {\n model_final_local = model,\n model_debug_softmax_word_weights = model_debug_softmax_word_weights, \n }\n \n return model, additional_local_submodels \nend\n\n\n--- Unit tests\nif unit_tests_now then\n print('Network model unit tests:')\n local num_mentions = 13\n \n local inputs = {}\n \n -- ctxt_w_vecs\n inputs[1] = {}\n inputs[1][1] = torch.ones(num_mentions, opt.ctxt_window):int():mul(unk_w_id)\n inputs[1][2] = torch.randn(num_mentions, opt.ctxt_window, ent_vecs_size)\n -- e_vecs\n inputs[2] = {}\n inputs[2][1] = torch.ones(num_mentions, max_num_cand):int():mul(unk_ent_thid)\n inputs[2][2] = torch.randn(num_mentions, max_num_cand, ent_vecs_size)\n -- p(e|m)\n inputs[3] = torch.zeros(num_mentions, max_num_cand) \n \n local model, additional_local_submodels = local_model(num_mentions, A_linear, B_linear, opt)\n \n print(additional_local_submodels.model_debug_softmax_word_weights:forward(inputs):size())\n \n local outputs = model:forward(inputs)\n assert(outputs:size(1) == num_mentions and outputs:size(2) == max_num_cand)\n print('FWD success!')\n\n model:backward(inputs, torch.randn(num_mentions, max_num_cand))\n print('BKWD success!')\n \n parameters,gradParameters = model:getParameters()\n print(parameters:size())\n print(gradParameters:size()) \nend\n" }, { "path": "deep-ed/deep-ed-master/ed/test/check_coref.lua", "content": "-- Runs our trivial coreference resolution method and outputs the new set of \n-- entity candidates. Used for debugging the coreference resolution method.\n\nif not opt then\n cmd = torch.CmdLine()\n cmd:option('-root_data_dir', '', 'Root path of the data, $DATA_PATH.')\n cmd:text()\n opt = cmd:parse(arg or {})\n assert(opt.root_data_dir ~= '', 'Specify a valid root_data_dir path argument.') \nend\n\n\nrequire 'torch'\ndofile 'utils/utils.lua'\n\ntds = tds or require 'tds'\n\ndofile 'entities/ent_name2id_freq/ent_name_id.lua'\ndofile 'ed/test/coref.lua'\n\nfile = opt.root_data_dir .. 'generated/test_train_data/aida_testB.csv'\n\nopt = {}\nopt.coref = true\n\n it, _ = io.open(file)\n local all_doc_lines = tds.Hash()\n local line = it:read()\n while line do\n local parts = split(line, '\\t')\n local doc_name = parts[1]\n if not all_doc_lines[doc_name] then\n all_doc_lines[doc_name] = tds.Hash()\n end\n all_doc_lines[doc_name][1 + #all_doc_lines[doc_name]] = line\n line = it:read()\n end\n -- Gather coreferent mentions to increase accuracy.\n build_coreference_dataset(all_doc_lines, 'aida-B')\n" }, { "path": "deep-ed/deep-ed-master/ed/test/coref_persons.lua", "content": "-- Given a dataset, try to retrieve better entity candidates\n-- for ambiguous mentions of persons. For example, suppose a document \n-- contains a mention of a person called 'Peter Such' that can be easily solved with \n-- the current system. Now suppose that, in the same document, there \n-- exists a mention 'Such' referring to the same person. For this \n-- second highly ambiguous mention, retrieving the correct entity in \n-- top K candidates would be very hard. We adopt here a simple heuristical strategy of \n-- searching in the same document all potentially coreferent mentions that strictly contain \n-- the given mention as a substring. If such mentions exist and they refer to \n-- persons (contain at least one person candidate entity), then the ambiguous \n-- shorter mention gets as candidates the candidates of the longer mention.\n\ntds = tds or require 'tds'\nassert(get_ent_wikiid_from_name)\nprint('==> Loading index of Wiki entities that represent persons.')\n\nlocal persons_ent_wikiids = tds.Hash()\nfor line in io.lines(opt.root_data_dir .. 'basic_data/p_e_m_data/persons.txt') do\n local ent_wikiid = get_ent_wikiid_from_name(line, true)\n if ent_wikiid ~= unk_ent_wikiid then\n persons_ent_wikiids[ent_wikiid] = 1\n end\nend\n\nfunction is_person(ent_wikiid)\n return persons_ent_wikiids[ent_wikiid]\nend\n\nprint(' Done loading persons index. Size = ' .. #persons_ent_wikiids)\n\n\nlocal function mention_refers_to_person(m, mention_ent_cand)\n local top_p = 0\n local top_ent = -1\n for e_wikiid, p_e_m in pairs(mention_ent_cand[m]) do\n if p_e_m > top_p then\n top_ent = e_wikiid\n top_p = p_e_m\n end\n end\n return is_person(top_ent)\nend\n\n\nfunction build_coreference_dataset(dataset_lines, banner)\n if (not opt.coref) then\n return dataset_lines\n else\n \n -- Create new entity candidates\n local coref_dataset_lines = tds.Hash()\n for doc_id, lines_map in pairs(dataset_lines) do \n \n coref_dataset_lines[doc_id] = tds.Hash()\n \n -- Collect entity candidates for each mention.\n local mention_ent_cand = {}\n for _,sample_line in pairs(lines_map) do\n local parts = split(sample_line, \"\\t\")\n assert(doc_id == parts[1])\n local mention = parts[3]:lower()\n if not mention_ent_cand[mention] then\n mention_ent_cand[mention] = {}\n assert(parts[6] == 'CANDIDATES') \n if parts[7] ~= 'EMPTYCAND' then\n local num_cand = 1\n while parts[6 + num_cand] ~= 'GT:' do\n local cand_parts = split(parts[6 + num_cand], ',')\n local cand_ent_wikiid = tonumber(cand_parts[1])\n local cand_p_e_m = tonumber(cand_parts[2])\n assert(cand_p_e_m >= 0, cand_p_e_m) \n assert(cand_ent_wikiid)\n mention_ent_cand[mention][cand_ent_wikiid] = cand_p_e_m\n num_cand = num_cand + 1\n end\n end\n end\n end\n \n -- Find coreferent mentions\n for _,sample_line in pairs(lines_map) do\n local parts = split(sample_line, \"\\t\")\n assert(doc_id == parts[1])\n local mention = parts[3]:lower()\n \n assert(mention_ent_cand[mention])\n assert(parts[table_len(parts) - 1] == 'GT:')\n \n -- Grd trth infos\n local grd_trth_parts = split(parts[table_len(parts)], ',')\n local grd_trth_idx = tonumber(grd_trth_parts[1])\n assert(grd_trth_idx == -1 or table_len(grd_trth_parts) >= 4, sample_line)\n local grd_trth_entwikiid = -1\n if table_len(grd_trth_parts) >= 3 then\n grd_trth_entwikiid = tonumber(grd_trth_parts[2])\n end\n assert(grd_trth_entwikiid)\n \n -- Merge lists of entity candidates\n local added_list = {}\n local num_added_mentions = 0\n for m,_ in pairs(mention_ent_cand) do\n local stupid_lua_pattern = string.gsub(mention, '%.', '%%%.')\n stupid_lua_pattern = string.gsub(stupid_lua_pattern, '%-', '%%%-')\n if m ~= mention and (string.find(m, ' ' .. stupid_lua_pattern) or string.find(m, stupid_lua_pattern .. ' ')) and mention_refers_to_person(m, mention_ent_cand) then\n \n if banner == 'aida-B' then\n print(blue('coref mention = ' .. m .. ' replaces original mention = ' .. mention) ..\n ' ; DOC = ' .. doc_id)\n end\n \n num_added_mentions = num_added_mentions + 1\n for e_wikiid, p_e_m in pairs(mention_ent_cand[m]) do\n if not added_list[e_wikiid] then\n added_list[e_wikiid] = 0.0\n end\n added_list[e_wikiid] = added_list[e_wikiid] + p_e_m\n end\n end\n end\n \n -- Average:\n for e_wikiid, _ in pairs(added_list) do\n added_list[e_wikiid] = added_list[e_wikiid] / num_added_mentions\n end\n \n -- Merge the two lists\n local merged_list = mention_ent_cand[mention]\n if num_added_mentions > 0 then\n merged_list = added_list\n end \n \n local sorted_list = {}\n for ent_wikiid,p in pairs(merged_list) do\n table.insert(sorted_list, {ent_wikiid = ent_wikiid, p = p})\n end\n table.sort(sorted_list, function(a,b) return a.p > b.p end)\n \n -- Write the new line\n local str = parts[1] .. '\\t' .. parts[2] .. '\\t' .. parts[3] .. '\\t' .. parts[4] .. '\\t' \n .. parts[5] .. '\\t' .. parts[6] .. '\\t' \n \n if table_len(sorted_list) == 0 then\n str = str .. 'EMPTYCAND\\tGT:\\t-1'\n if grd_trth_entwikiid ~= unk_ent_wikiid then\n str = str .. ',' .. grd_trth_entwikiid .. ',' ..\n get_ent_name_from_wikiid(grd_trth_entwikiid)\n end\n else\n local candidates = {}\n local gt_pos = -1\n for pos,e in pairs(sorted_list) do\n if pos <= 100 then\n table.insert(candidates, e.ent_wikiid .. ',' ..\n string.format(\"%.3f\", e.p) .. ',' .. get_ent_name_from_wikiid(e.ent_wikiid))\n if e.ent_wikiid == grd_trth_entwikiid then\n gt_pos = pos\n end\n else\n break\n end\n end\n str = str .. table.concat(candidates, '\\t') .. '\\tGT:\\t'\n\n if gt_pos > 0 then\n str = str .. gt_pos .. ',' .. candidates[gt_pos]\n else\n str = str .. '-1'\n if grd_trth_entwikiid ~= unk_ent_wikiid then\n str = str .. ',' .. grd_trth_entwikiid .. ',' ..\n get_ent_name_from_wikiid(grd_trth_entwikiid)\n end\n end\n end\n \n coref_dataset_lines[doc_id][1 + #coref_dataset_lines[doc_id]] = str\n end \n end\n \n assert(#dataset_lines == #coref_dataset_lines)\n for doc_id, lines_map in pairs(dataset_lines) do\n assert(table_len(dataset_lines[doc_id]) == table_len(coref_dataset_lines[doc_id]))\n end\n \n return coref_dataset_lines\n end\nend" }, { "path": "deep-ed/deep-ed-master/ed/test/ent_freq_stats_test.lua", "content": "-- Statistics of annotated entities based on their frequency in Wikipedia corpus \n-- Table 6 (left) from our paper\nlocal function ent_freq_to_key(f)\n if f == 0 then\n return '0'\n elseif f == 1 then\n return '1'\n elseif f <= 5 then\n return '2-5'\n elseif f <= 10 then\n return '6-10'\n elseif f <= 20 then\n return '11-20'\n elseif f <= 50 then\n return '21-50'\n else\n return '50+'\n end\nend\n\n\nfunction new_ent_freq_map()\n local m = {}\n m['0'] = 0.0\n m['1'] = 0.0\n m['2-5'] = 0.0\n m['6-10'] = 0.0 \n m['11-20'] = 0.0 \n m['21-50'] = 0.0 \n m['50+'] = 0.0 \n return m\nend\n\nfunction add_freq_to_ent_freq_map(m, f)\n m[ent_freq_to_key(f)] = m[ent_freq_to_key(f)] + 1\nend\n\nfunction print_ent_freq_maps_stats(smallm, bigm)\n print(' ===> entity frequency stats :')\n for k,_ in pairs(smallm) do\n local perc = 0\n if bigm[k] > 0 then \n perc = 100.0 * smallm[k] / bigm[k]\n end\n assert(perc <= 100)\n print('freq = ' .. k .. ' : num = ' .. bigm[k] .. \n ' ; correctly classified = ' .. smallm[k] .. \n ' ; perc = ' .. string.format(\"%.2f\", perc))\n end\n print('')\nend\n\n \n " }, { "path": "deep-ed/deep-ed-master/ed/test/ent_p_e_m_stats_test.lua", "content": "-- Statistics of annotated entities based on their p(e|m) prio\n-- Table 6 (right) from our paper\n\nlocal function ent_prior_to_key(f)\n if f <= 0.001 then\n return '<=0.001'\n elseif f <= 0.003 then\n return '0.001-0.003'\n elseif f <= 0.01 then\n return '0.003-0.01'\n elseif f <= 0.03 then\n return '0.01-0.03'\n elseif f <= 0.1 then\n return '0.03-0.1'\n elseif f <= 0.3 then\n return '0.1-0.3'\n else\n return '0.3+'\n end\nend\n\n\nfunction new_ent_prior_map()\n local m = {}\n m['<=0.001'] = 0.0\n m['0.001-0.003'] = 0.0\n m['0.003-0.01'] = 0.0\n m['0.01-0.03'] = 0.0 \n m['0.03-0.1'] = 0.0 \n m['0.1-0.3'] = 0.0 \n m['0.3+'] = 0.0 \n return m\nend\n\nfunction add_prior_to_ent_prior_map(m, f)\n m[ent_prior_to_key(f)] = m[ent_prior_to_key(f)] + 1\nend\n\nfunction print_ent_prior_maps_stats(smallm, bigm)\n print(' ===> entity p(e|m) stats :') \n for k,_ in pairs(smallm) do\n local perc = 0\n if bigm[k] > 0 then \n perc = 100.0 * smallm[k] / bigm[k]\n end\n assert(perc <= 100)\n print('p(e|m) = ' .. k .. ' : num = ' .. bigm[k] .. \n ' ; correctly classified = ' .. smallm[k] .. \n ' ; perc = ' .. string.format(\"%.2f\", perc))\n end\nend\n\n \n " }, { "path": "deep-ed/deep-ed-master/ed/test/test.lua", "content": "dofile 'ed/test/coref_persons.lua'\ndofile 'ed/test/ent_freq_stats_test.lua'\ndofile 'ed/test/ent_p_e_m_stats_test.lua'\n\ntestAccLogger = Logger(opt.root_data_dir .. 'generated/ed_models/training_plots/f1-' .. banner)\n\n---------------------------- Define all datasets -----------------------------\n\ndatasets = {}\n\ndatasets['aida-A'] = opt.root_data_dir .. 'generated/test_train_data/aida_testA.csv' -- Validation set\ndatasets['aida-B'] = opt.root_data_dir .. 'generated/test_train_data/aida_testB.csv'\ndatasets['MSNBC'] = opt.root_data_dir .. 'generated/test_train_data/wned-msnbc.csv'\ndatasets['AQUAINT'] = opt.root_data_dir .. 'generated/test_train_data/wned-aquaint.csv'\ndatasets['ACE04'] = opt.root_data_dir .. 'generated/test_train_data/wned-ace2004.csv'\n\n------- Uncomment the following lines if you want to test on more datasets during training (will be slower).\n--datasets['train-aida'] = opt.root_data_dir .. 'generated/test_train_data/aida_train.csv'\n--datasets['CLUEWEB'] = opt.root_data_dir .. 'generated/test_train_data/wned-clueweb.csv'\n--datasets['WNED-WIKI'] = opt.root_data_dir .. 'generated/test_train_data/wned-wikipedia.csv'\n\n\nlocal classes = {}\nfor i = 1,max_num_cand do\n table.insert(classes, i)\nend\n\n--------------------------- Functions ------------------------------------------\nlocal function get_dataset_lines(banner)\n it, _ = io.open(datasets[banner])\n local all_doc_lines = tds.Hash()\n local line = it:read()\n while line do\n local parts = split(line, '\\t')\n local doc_name = parts[1]\n if not all_doc_lines[doc_name] then\n all_doc_lines[doc_name] = tds.Hash()\n end\n all_doc_lines[doc_name][1 + #all_doc_lines[doc_name]] = line\n line = it:read()\n end\n -- Gather coreferent mentions to increase accuracy.\n return build_coreference_dataset(all_doc_lines, banner)\nend\n\n\nlocal function get_dataset_num_non_empty_candidates(dataset_lines)\n local num_predicted = 0\n for doc_id, lines_map in pairs(dataset_lines) do \n for _,sample_line in pairs(lines_map) do\n local parts = split(sample_line, '\\t')\n if parts[7] ~= 'EMPTYCAND' then\n num_predicted = num_predicted + 1\n end\n end\n end\n return num_predicted\nend\n\nlocal function test_one(banner, f1_scores, epoch)\n collectgarbage(); collectgarbage();\n -- Load dataset lines\n local dataset_lines = get_dataset_lines(banner)\n \n local dataset_num_mentions = 0\n for doc_id, lines_map in pairs(dataset_lines) do \n dataset_num_mentions = dataset_num_mentions + #lines_map\n end \n print('\\n===> ' .. banner .. '; num mentions = ' .. dataset_num_mentions)\n\n local time = sys.clock()\n\n confusion = optim.ConfusionMatrix(classes) \n confusion:zero()\n \n xlua.progress(0, dataset_num_mentions)\n \n local num_true_positives = 0.0\n local grd_ent_freq_map = new_ent_freq_map()\n local correct_classified_ent_freq_map = new_ent_freq_map()\n local grd_ent_prior_map = new_ent_prior_map()\n local correct_classified_ent_prior_map = new_ent_prior_map()\n \n local num_mentions_without_gold_ent_in_candidates = 0\n local both_pem_ours = 0 -- num mentions solved both by argmax p(e|m) and our global model\n local only_pem_not_ours = 0 -- num mentions solved by argmax p(e|m), but not by our global model\n local only_ours_not_pem = 0 -- num mentions solved by our global model, but not by argmax p(e|m)\n local not_ours_not_pem = 0 -- num mentions not solved neither by our model nor by argmax p(e|m)\n \n local processed_docs = 0\n local processed_mentions = 0\n \n for doc_id, doc_lines in pairs(dataset_lines) do\n processed_docs = processed_docs + 1\n local num_mentions = #doc_lines\n processed_mentions = processed_mentions + num_mentions\n local inputs = empty_minibatch_with_ids(num_mentions)\n local targets = torch.zeros(num_mentions)\n local mentions = {}\n for k = 1, num_mentions do\n local sample_line = doc_lines[k]\n local parts = split(sample_line, '\\t')\n mentions[k] = parts[3]\n local target = process_one_line(sample_line, inputs, k, false)\n targets[k] = target \n end\n inputs, targets = minibatch_to_correct_type(inputs, targets, false)\n \n -- NN forward pass:\n model, additional_local_submodels = get_model(num_mentions)\n model:evaluate()\n local preds = model:forward(inputs):float() \n \n \n --------- Subnetworks used to print debug weights and scores :\n -- num_mentions x num_ctxt_vecs\n debug_softmax_word_weights = additional_local_submodels.model_debug_softmax_word_weights:forward(inputs):float() \n -- num_mentions, max_num_cand:\n final_local_scores = additional_local_submodels.model_final_local:forward(inputs):float() \n\n -- Process results:\n local all_ent_wikiids = get_cand_ent_wikiids(inputs)\n \n for k = 1, num_mentions do\n local pred = preds[k] -- vector of size : max_num_cand\n assert (torch.norm(pred) ~= math.nan)\n\n -- Ignore unk entities (padding entities):\n local ent_wikiids = all_ent_wikiids[k] -- vector of size : max_num_cand\n for i = 1,max_num_cand do\n if ent_wikiids[i] == unk_ent_wikiid then\n pred[i] = -1e8 --> -infinity\n end\n end\n \n -- PRINT DEBUG SCORES: Show network weights and scores for entities with a valid gold label.\n if (targets[k] > 0) then\n local log_p_e_m = get_log_p_e_m(inputs)\n\n if (k == 1) then\n print('\\n')\n print(blue('============================================'))\n print(blue('============ DOC : ' .. doc_id .. ' ================'))\n print(blue('============================================'))\n end\n\n -- Winning entity\n local _, argmax_idx = torch.max(pred, 1)\n local win_idx = argmax_idx[1] -- the actual number\n local ent_win = ent_wikiids[win_idx]\n local ent_win_name = get_ent_name_from_wikiid(ent_win)\n local ent_win_log_p_e_m = log_p_e_m[k][win_idx]\n local ent_win_local = final_local_scores[k][win_idx]\n \n -- Just some sanity check\n local best_pred, best_pred_idxs = topk(pred, 1)\n if (pred[best_pred_idxs[1]] ~= best_pred[1]) then\n print(pred)\n end\n \n assert(pred[best_pred_idxs[1]] == best_pred[1])\n assert(pred[best_pred_idxs[1]] == pred[win_idx])\n \n -- Grd trth entity\n local grd_idx = targets[k]\n local ent_grd = ent_wikiids[grd_idx]\n local ent_grd_name = get_ent_name_from_wikiid(ent_grd)\n local ent_grd_log_p_e_m = log_p_e_m[k][grd_idx]\n local ent_grd_local = final_local_scores[k][grd_idx]\n \n if win_idx ~= grd_idx then\n assert(ent_win ~= ent_grd)\n print('\\n====> ' .. red('INCORRECT ANNOTATION') .. \n ' : mention = ' .. skyblue(mentions[k]) ..\n ' ==> ENTITIES (OURS/GOLD): ' .. red(ent_win_name) .. ' <---> ' .. green(ent_grd_name))\n else\n assert(ent_win == ent_grd)\n local mention_str = 'mention = ' .. mentions[k]\n if math.exp(ent_grd_log_p_e_m) >= 0.99 then\n mention_str = yellow(mention_str)\n end \n print('\\n====> ' .. green('CORRECT ANNOTATION') .. \n ' : ' .. mention_str ..\n ' ==> ENTITY: ' .. green(ent_grd_name))\n end\n\n print(\n 'SCORES: global= ' .. nice_print_red_green(pred[win_idx], pred[grd_idx]) ..\n '; local()= ' .. nice_print_red_green(ent_win_local, ent_grd_local) ..\n '; log p(e|m)= ' .. nice_print_red_green(ent_win_log_p_e_m, ent_grd_log_p_e_m)\n )\n \n -- Print top attended ctxt words and their attention weights:\n local str_words = '\\nTop context words (sorted by attention weight, only non-zero weights - top R words): \\n'\n local ctxt_word_ids = get_ctxt_word_ids(inputs) -- num_mentions x opt.ctxt_window\n local best_scores, best_word_idxs = topk(debug_softmax_word_weights[k], opt.ctxt_window)\n local seen_unk_w_id = false\n for wk = 1,opt.ctxt_window do \n local w_idx_ctxt = best_word_idxs[wk]\n assert(w_idx_ctxt >= 1 and w_idx_ctxt <= opt.ctxt_window)\n local w_id = ctxt_word_ids[k][w_idx_ctxt]\n if w_id ~= unk_w_id or (not seen_unk_w_id) then\n if w_id == unk_w_id then\n seen_unk_w_id = true\n end\n local w = get_word_from_id(w_id)\n local score = debug_softmax_word_weights[k][w_idx_ctxt]\n assert(score == best_scores[wk])\n \n if score > 0.001 then \n str_words = str_words .. w .. '[' .. string.format(\"%.3f\", score) .. ']; '\n end\n end\n end\n print(str_words) \n end ----------------- Done printing scores and weights\n\n \n -- Count how many of the winning entities do not have a valid ent vector\n local _, argmax_idx = torch.max(pred, 1)\n if targets[k] > 0 and get_thid(ent_wikiids[argmax_idx[1]]) == unk_ent_thid then \n print(pred)\n print(ent_wikiids)\n print('\\n\\n' .. red('!!!!Entity w/o vec: ' .. ent_wikiids[argmax_idx[1]] .. ' line = ' .. doc_lines[k]))\n os.exit()\n end\n\n -- Accumulate statistics about the annotations of our system.\n if (targets[k] > 0) then\n \n local log_p_e_m = get_log_p_e_m(inputs)\n \n local nn_is_good = true\n local pem_is_good = true\n \n -- Grd trth entity\n local grd_idx = targets[k]\n \n for j = 1,max_num_cand do\n if j ~= grd_idx and pred[grd_idx] < pred[j] then\n assert(ent_wikiids[j] ~= unk_ent_wikiid)\n nn_is_good = false\n end\n if j ~= grd_idx and log_p_e_m[k][grd_idx] < log_p_e_m[k][j] then\n assert(ent_wikiids[j] ~= unk_ent_wikiid)\n pem_is_good = false\n end \n end\n \n if nn_is_good and pem_is_good then\n both_pem_ours = both_pem_ours + 1\n elseif nn_is_good then\n only_ours_not_pem = only_ours_not_pem + 1\n elseif pem_is_good then\n only_pem_not_ours = only_pem_not_ours + 1\n else\n not_ours_not_pem = not_ours_not_pem + 1\n end\n\n assert(ent_wikiids[targets[k]] ~= unk_ent_wikiid, ' Something is terribly wrong here ')\n confusion:add(pred, targets[k])\n\n -- Update number of true positives\n local _, argmax_idx = torch.max(pred, 1)\n local winning_entiid = ent_wikiids[argmax_idx[1]]\n local grd_entiid = ent_wikiids[targets[k]]\n \n local grd_ent_freq = get_ent_freq(grd_entiid)\n add_freq_to_ent_freq_map(grd_ent_freq_map, grd_ent_freq)\n \n local grd_ent_prior = math.exp(log_p_e_m[k][grd_idx])\n add_prior_to_ent_prior_map(grd_ent_prior_map, grd_ent_prior)\n if winning_entiid == grd_entiid then\n add_freq_to_ent_freq_map(correct_classified_ent_freq_map, grd_ent_freq)\n add_prior_to_ent_prior_map(correct_classified_ent_prior_map, grd_ent_prior)\n num_true_positives = num_true_positives + 1\n end\n\n else -- grd trth is not between the given set of candidates, so we cannot be right\n num_mentions_without_gold_ent_in_candidates = num_mentions_without_gold_ent_in_candidates + 1\n \n confusion:add(torch.zeros(max_num_cand), max_num_cand)\n end\n end\n -- disp progress\n xlua.progress(processed_mentions, dataset_num_mentions) \n end -- done with this mini batch\n\n -- Now plotting results\n time = sys.clock() - time\n time = time / dataset_num_mentions\n print(\"==> time to test 1 sample = \" .. (time*1000) .. 'ms')\n\n confusion:__tostring__()\n\n -- We refrain from solving mentions without at least one candidate\n local precision = 100.0 * num_true_positives / get_dataset_num_non_empty_candidates(dataset_lines)\n local recall = 100.0 * num_true_positives / dataset_num_mentions\n assert(math.abs(recall - confusion.totalValid * 100.0) < 0.01, 'Difference in recalls.')\n local f1 = 2.0 * precision * recall / (precision + recall)\n f1_scores[banner] = f1\n\n local f1_str = red(string.format(\"%.2f\", f1) .. '%')\n if banner == 'aida-A' and f1 >= 90.20 then\n f1_str = green(string.format(\"%.2f\", f1) .. '%')\n end\n \n print('==> '.. red(banner) .. ' ' .. banner .. ' ; EPOCH = ' .. epoch .. \n ': Micro recall = ' .. string.format(\"%.2f\", confusion.totalValid * 100.0) .. '%' .. \n ' ; Micro F1 = ' .. f1_str)\n \n -- Lower learning rate if we got close to minimum\n if banner == 'aida-A' and f1 >= 90 then\n opt.lr = 1e-5\n end\n \n -- We slow down training a little bit if we passed the 90% F1 score threshold.\n -- And we start saving (good quality) models from now on.\n if banner == 'aida-A' then\n if f1 >= 90.0 then\n opt.save = true\n else\n opt.save = false\n end\n end\n \n print_ent_freq_maps_stats(correct_classified_ent_freq_map, grd_ent_freq_map)\n print_ent_prior_maps_stats(correct_classified_ent_prior_map, grd_ent_prior_map)\n \n print(' num_mentions_w/o_gold_ent_in_candidates = ' ..\n num_mentions_without_gold_ent_in_candidates .. \n ' total num mentions in dataset = ' .. dataset_num_mentions)\n \n print(' percentage_mentions_w/o_gold_ent_in_candidates = ' ..\n string.format(\"%.2f\", 100.0 * num_mentions_without_gold_ent_in_candidates / dataset_num_mentions) .. '%; ' ..\n ' percentage_mentions_solved : ' ..\n 'both_pem_ours = ' .. string.format(\"%.2f\", 100.0 * both_pem_ours / dataset_num_mentions) .. '%; ' ..\n ' only_pem_not_ours = ' .. string.format(\"%.2f\", 100.0 * only_pem_not_ours / dataset_num_mentions) .. '%; ' ..\n ' only_ours_not_pem = ' .. string.format(\"%.2f\", 100.0 * only_ours_not_pem / dataset_num_mentions) .. '%; ' ..\n ' not_ours_not_pem = ' .. string.format(\"%.2f\", 100.0 * not_ours_not_pem / dataset_num_mentions) .. '%')\nend\n\n\n-------- main test function:\nfunction test(epoch)\n if not epoch then\n epoch = 0\n end\n \n print('\\n\\n=====> TESTING <=== ') \n f1_scores = {}\n for banner,dataset_file in pairs(datasets) do\n test_one(banner, f1_scores, epoch)\n end\n\n -- Plot accuracies\n if train_and_test then\n num_batch_train_between_plots = num_batches_per_epoch\n \n testAccLogger:add(f1_scores)\n styles = {}\n for banner,_ in pairs(f1_scores) do\n styles[banner] = '-'\n end\n testAccLogger:style(styles)\n testAccLogger:plot('F1', 'x ' .. num_batch_train_between_plots .. ' mini-batches')\n end\nend" }, { "path": "deep-ed/deep-ed-master/ed/test/test_one_loaded_model.lua", "content": "-- Test one single ED model trained using ed/ed.lua\n\n-- Run: CUDA_VISIBLE_DEVICES=0 th ed/test/test_one_loaded_model.lua -root_data_dir $DATA_PATH -model global -ent_vecs_filename $ENTITY_VECS -test_one_model_file $ED_MODEL_FILENAME\nrequire 'optim'\nrequire 'torch'\nrequire 'gnuplot'\nrequire 'nn'\nrequire 'xlua'\ntds = tds or require 'tds'\n\ndofile 'utils/utils.lua'\nprint('\\n' .. green('==========> Test a pre-trained ED neural model <==========') .. '\\n')\n\ndofile 'ed/args.lua'\n\nprint('===> RUN TYPE: ' .. opt.type)\ntorch.setdefaulttensortype('torch.FloatTensor')\nif string.find(opt.type, 'cuda') then\n print('==> switching to CUDA (GPU)')\n require 'cunn'\n require 'cutorch'\n require 'cudnn'\n cudnn.benchmark = true \n cudnn.fastest = true\nelse\n print('==> running on CPU')\nend\n\nextract_args_from_model_title(opt.test_one_model_file)\n\ndofile 'utils/logger.lua'\ndofile 'entities/relatedness/relatedness.lua'\ndofile 'entities/ent_name2id_freq/ent_name_id.lua'\ndofile 'entities/ent_name2id_freq/e_freq_index.lua'\ndofile 'words/load_w_freq_and_vecs.lua'\ndofile 'words/w2v/w2v.lua'\ndofile 'entities/pretrained_e2v/e2v.lua'\ndofile 'ed/minibatch/build_minibatch.lua'\ndofile 'ed/models/model.lua'\ndofile 'ed/test/test.lua'\n\nlocal saved_linears = torch.load(opt.root_data_dir .. 'generated/ed_models/' .. opt.test_one_model_file)\nunpack_saveable_weights(saved_linears)\n\ntest()\n" }, { "path": "deep-ed/deep-ed-master/ed/train.lua", "content": "-- Training of ED models.\n\nif opt.opt == 'SGD' then\n optimState = {\n learningRate = opt.lr,\n momentum = 0.9,\n learningRateDecay = 5e-7\n }\n optimMethod = optim.sgd\n\nelseif opt.opt == 'ADAM' then -- See: http://cs231n.github.io/neural-networks-3/#update \n optimState = {\n learningRate = opt.lr,\n }\n optimMethod = optim.adam\n\nelseif opt.opt == 'ADADELTA' then -- See: http://cs231n.github.io/neural-networks-3/#update \n -- Run with default parameters, no need for learning rate and other stuff\n optimState = {}\n optimConfig = {} \n optimMethod = optim.adadelta\n \nelseif opt.opt == 'ADAGRAD' then -- See: http://cs231n.github.io/neural-networks-3/#update \n optimMethod = optim.adagrad\n optimState = {\n learningRate = opt.lr \n } \n \nelse\n error('unknown optimization method')\nend\n----------------------------------------------------------------------\n\n-- Each batch is one document, so we test/validate/save the current model after each set of \n-- 5000 documents. Since aida-train contains 946 documents, this is equivalent with 5 full epochs.\nnum_batches_per_epoch = 5000 \n\nfunction train_and_test()\n \n print('\\nDone testing for ' .. banner)\n print('Params serialized = ' .. params_serialized) \n \n -- epoch tracker\n epoch = 1\n \n local processed_so_far = 0\n \n local f_bs = 0\n local gradParameters_bs = nil\n\n while true do\n local time = sys.clock()\n print('\\n')\n print('One epoch = ' .. (num_batches_per_epoch / 1000) .. ' full passes over AIDA-TRAIN in our case.')\n print(green('==> TRAINING EPOCH #' .. epoch .. ' <=='))\n \n print_net_weights()\n \n local processed_mentions = 0\n for batch_index = 1,num_batches_per_epoch do\n -- Read one mini-batch from one data_thread:\n local inputs, targets = get_minibatch()\n \n local num_mentions = targets:size(1)\n processed_mentions = processed_mentions + num_mentions\n \n local model, _ = get_model(num_mentions)\n model:training()\n\n -- Retrieve parameters and gradients:\n -- extracts and flattens all model's parameters into a 1-dim vector\n parameters,gradParameters = model:getParameters()\n gradParameters:zero()\n\n -- Just in case:\n collectgarbage()\n collectgarbage()\n\n -- Reset gradients\n gradParameters:zero()\n\n -- Evaluate function for complete mini batch\n\n local outputs = model:forward(inputs)\n assert(outputs:size(1) == num_mentions and outputs:size(2) == max_num_cand)\n local f = criterion:forward(outputs, targets)\n\n -- Estimate df/dW\n local df_do = criterion:backward(outputs, targets)\n\n model:backward(inputs, df_do)\n\n if opt.batch_size == 1 or batch_index % opt.batch_size == 1 then\n gradParameters_bs = gradParameters:clone():zero()\n f_bs = 0\n end\n \n gradParameters_bs:add(gradParameters)\n f_bs = f_bs + f\n \n if opt.batch_size == 1 or batch_index % opt.batch_size == 0 then\n \n gradParameters_bs:div(opt.batch_size)\n f_bs = f_bs / opt.batch_size\n \n -- Create closure to evaluate f(X) and df/dX\n local feval = function(x)\n return f_bs, gradParameters_bs\n end\n\n -- Optimize on current mini-batch\n optimState.learningRate = opt.lr\n optimMethod(feval, parameters, optimState)\n \n -- Regularize the f_network with projected SGD.\n regularize_f_network()\n end\n \n -- Display progress\n processed_so_far = processed_so_far + num_mentions\n if processed_so_far > 100000000 then\n processed_so_far = processed_so_far - 100000000\n end\n xlua.progress(processed_so_far, 100000000) \n end \n\n -- Measure time taken\n time = sys.clock() - time\n time = time / processed_mentions\n \n print(\"\\n==> time to learn 1 sample = \" .. (time*1000) .. 'ms')\n \n -- Test:\n test(epoch)\n print('\\nDone testing for ' .. banner)\n print('Params serialized = ' .. params_serialized)\n \n -- Save the current model:\n if opt.save then\n local filename = opt.root_data_dir .. 'generated/ed_models/' .. params_serialized .. '|ep=' .. epoch\n print('==> saving model to '..filename)\n torch.save(filename, pack_saveable_weights())\n end\n \n -- Next epoch\n epoch = epoch + 1\n end\nend\n" }, { "path": "deep-ed/deep-ed-master/ent_vecs_scores.txt", "content": "\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.499\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.495\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.534\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.475\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.003\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.467\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.476\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.510\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.442\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_3.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.562\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.547\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.583\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.522\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.214\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.532\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.524\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.554\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.486\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_6.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.583\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.569\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.601\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.541\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.294\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.554\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.541\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.570\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.502\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_9.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.612\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.580\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.615\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.555\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.363\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.575\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.556\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.584\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.517\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_12.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.624\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.587\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.620\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.562\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.393\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.588\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.562\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.589\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.522\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_15.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.638\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.596\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.628\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.572\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.434\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.595\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.571\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.598\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.530\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_18.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.633\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.598\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.630\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.574\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.434\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.598\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.572\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.600\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.533\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_21.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.629\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.601\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.632\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.575\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.437\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.603\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.576\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.600\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.535\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_24.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.641\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.604\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.635\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.579\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.458\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.611\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.575\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.602\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.538\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_27.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.636\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.606\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.638\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.580\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.461\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.619\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.578\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.605\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.542\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_30.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.649\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.607\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.639\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.583\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.478\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.616\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.580\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.606\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.542\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_33.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.638\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.607\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.639\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.582\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.467\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.613\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.580\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.606\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.542\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_36.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.645\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.610\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.641\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.584\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.479\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.610\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.580\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.606\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.542\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_39.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.652\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.610\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.641\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.585\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.487\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.617\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.581\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.607\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.544\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_42.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.643\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.611\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.642\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.585\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.482\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.621\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.583\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.610\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.546\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_45.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.646\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.611\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.642\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.586\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.485\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.624\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.586\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.612\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.548\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_48.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.648\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.611\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.642\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.586\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.487\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.623\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.585\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.612\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.548\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_51.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.649\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.613\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.644\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.588\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.493\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.619\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.588\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.614\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.548\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_54.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.647\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.612\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.644\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.587\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.490\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.622\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.588\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.613\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.548\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_57.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.646\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.613\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.644\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.588\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.491\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.626\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.587\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.613\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.548\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_60.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.642\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.614\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.645\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.587\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.488\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.622\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.585\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.613\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.547\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_63.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.672\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.631\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.663\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.609\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.575\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.641\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.607\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.636\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.570\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_66.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.677\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.632\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.665\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.610\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.585\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.644\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.609\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.637\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.573\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_69.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.675\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.634\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.667\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.613\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.589\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.634\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.610\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.640\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.576\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_72.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.675\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.635\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.666\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.612\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.588\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.637\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.608\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.640\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.576\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_75.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.672\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.638\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.668\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.613\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.591\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.638\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.609\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.640\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.577\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_78.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.669\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.636\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.669\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.615\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.590\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.637\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.610\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.642\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.580\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_81.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.672\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.637\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.671\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.616\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.596\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.633\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.609\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.644\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.577\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_84.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.663\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.634\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.667\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.611\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.575\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.618\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.606\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.638\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.569\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_87.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.667\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.639\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.672\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.615\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.594\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.625\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.612\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.644\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.577\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_90.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.666\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.639\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.674\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.617\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.596\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.627\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.610\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.643\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.576\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_93.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.675\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.641\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.674\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.619\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.609\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.631\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.610\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.642\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.577\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_96.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.671\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.638\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.672\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.617\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.598\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.636\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.606\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.637\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.575\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_99.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.673\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.639\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.673\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.617\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.602\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.641\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.609\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.640\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.578\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_102.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.670\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.639\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.670\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.614\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.594\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.640\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.606\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.635\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.573\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_105.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.689\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.643\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.672\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.620\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.625\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.652\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.611\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.639\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.577\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_108.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.675\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.641\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.671\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.617\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.604\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.652\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.610\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.638\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.577\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_111.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.680\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.640\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.671\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.616\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.607\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.649\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.610\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.638\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.577\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_114.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.675\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.642\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.673\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.617\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.607\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.647\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.610\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.639\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.578\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_117.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.681\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.642\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.674\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.620\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.618\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.641\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.610\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.640\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.577\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_120.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.683\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.645\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.676\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.621\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.624\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.646\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.611\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.641\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.579\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_123.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.669\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.641\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.674\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.616\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.599\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.631\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.605\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.636\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.571\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_126.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.674\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.642\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.676\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.620\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.613\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.634\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.612\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.642\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.578\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_129.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.675\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.642\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.676\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.620\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.612\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.634\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.608\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.640\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.576\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_132.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.682\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.644\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.677\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.623\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.626\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.638\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.609\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.641\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.576\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_135.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.678\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.640\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.674\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.619\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.612\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.635\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.607\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.639\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.574\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_138.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.680\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.643\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.677\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.621\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.621\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.642\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.613\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.643\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.579\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_141.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.675\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.642\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.675\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.619\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.612\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.639\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.609\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.638\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.575\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_144.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.683\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.645\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.676\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.621\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.626\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.640\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.610\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.639\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.575\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_147.t7\u001b[0m\t\r\n--\n\u001b[0m\u001b[33mmeasure =\u001b[39m\u001b[0m\t\u001b[0mNDCG1\u001b[0m\t\u001b[0mNDCG5\u001b[0m\t\u001b[0mNDCG10\u001b[0m\t\u001b[0mMAP\u001b[0m\t\u001b[0mTOTAL VALIDATION\u001b[0m\t\r\n\u001b[0m\u001b[33mour (vald) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.675\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.644\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.674\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m0.619\u001b[39m\u001b[0m\t\u001b[0m\u001b[34m2.613\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mour (test) =\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.643\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.610\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.639\u001b[39m\u001b[0m\t\u001b[0m\u001b[31m0.576\u001b[39m\u001b[0m\t\r\n\u001b[0m\u001b[33mYamada'16 =\u001b[39m\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.56\u001b[0m\t\u001b[0;36m0.59\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\r\n\u001b[0m\u001b[33mWikiMW =\u001b[39m\u001b[0m\t\u001b[0;36m0.54\u001b[0m\t\u001b[0;36m0.52\u001b[0m\t\u001b[0;36m0.55\u001b[0m\t\u001b[0;36m0.48\u001b[0m\t\r\n\u001b[0m==> saving model to /local/home/nkolitsa/end2end_neural_el/deep-ed/data/generated/ent_vecs/ent_vecs__ep_150.t7\u001b[0m\t\r\n" }, { "path": "deep-ed/deep-ed-master/entities/ent_name2id_freq/e_freq_gen.lua", "content": "-- Creates a file that contains entity frequencies.\n\nif not opt then\n cmd = torch.CmdLine()\n cmd:option('-root_data_dir', '', 'Root path of the data, $DATA_PATH.')\n cmd:text()\n opt = cmd:parse(arg or {})\n assert(opt.root_data_dir ~= '', 'Specify a valid root_data_dir path argument.') \nend\n\n\nrequire 'torch'\ntds = tds or require 'tds' \n\ndofile 'utils/utils.lua'\ndofile 'entities/ent_name2id_freq/ent_name_id.lua'\n\nentity_freqs = tds.Hash()\n\nlocal num_lines = 0\nit, _ = io.open(opt.root_data_dir .. 'generated/crosswikis_wikipedia_p_e_m.txt')\nline = it:read()\n\nwhile (line) do\n num_lines = num_lines + 1\n if num_lines % 2000000 == 0 then\n print('Processed ' .. num_lines .. ' lines. ')\n end\n \n local parts = split(line , '\\t')\n local num_parts = table_len(parts)\n for i = 3, num_parts do\n local ent_str = split(parts[i], ',')\n local ent_wikiid = tonumber(ent_str[1])\n local freq = tonumber(ent_str[2])\n assert(ent_wikiid)\n assert(freq)\n \n if (not entity_freqs[ent_wikiid]) then\n entity_freqs[ent_wikiid] = 0\n end\n entity_freqs[ent_wikiid] = entity_freqs[ent_wikiid] + freq\n end\n line = it:read()\nend\n\n\n-- Writing word frequencies\nprint('Sorting and writing')\nsorted_ent_freq = {}\nfor ent_wikiid,freq in pairs(entity_freqs) do\n if freq >= 10 then\n table.insert(sorted_ent_freq, {ent_wikiid = ent_wikiid, freq = freq})\n end\nend\n\ntable.sort(sorted_ent_freq, function(a,b) return a.freq > b.freq end)\n\nout_file = opt.root_data_dir .. 'generated/ent_wiki_freq.txt'\nouf = assert(io.open(out_file, \"w\"))\ntotal_freq = 0\nfor _,x in pairs(sorted_ent_freq) do\n ouf:write(x.ent_wikiid .. '\\t' .. get_ent_name_from_wikiid(x.ent_wikiid) .. '\\t' .. x.freq .. '\\n')\n total_freq = total_freq + x.freq\nend\nouf:flush()\nio.close(ouf)\n\nprint('Total freq = ' .. total_freq .. '\\n')\n" }, { "path": "deep-ed/deep-ed-master/entities/ent_name2id_freq/e_freq_index.lua", "content": "-- Loads an index containing entity -> frequency pairs. \n-- TODO: rewrite this file in a simpler way (is complicated because of some past experiments).\ntds = tds or require 'tds' \n\nprint('==> Loading entity freq map') \n\nlocal ent_freq_file = opt.root_data_dir .. 'generated/ent_wiki_freq.txt'\n\nmin_freq = 1\ne_freq = tds.Hash()\ne_freq.ent_f_start = tds.Hash()\ne_freq.ent_f_end = tds.Hash()\ne_freq.total_freq = 0\ne_freq.sorted = tds.Hash()\n\ncur_start = 1\ncnt = 0\nfor line in io.lines(ent_freq_file) do\n local parts = split(line, '\\t')\n local ent_wikiid = tonumber(parts[1])\n local ent_f = tonumber(parts[3])\n assert(ent_wikiid)\n assert(ent_f)\n \n if (not rewtr) or rewtr.reltd_ents_wikiid_to_rltdid[ent_wikiid] then\n e_freq.ent_f_start[ent_wikiid] = cur_start\n e_freq.ent_f_end[ent_wikiid] = cur_start + ent_f - 1\n e_freq.sorted[cnt] = ent_wikiid\n cur_start = cur_start + ent_f\n cnt = cnt + 1\n end\nend\n\ne_freq.total_freq = cur_start - 1\n\nprint(' Done loading entity freq index. Size = ' .. cnt)\n\nfunction get_ent_freq(ent_wikiid) \n if e_freq.ent_f_start[ent_wikiid] then\n return e_freq.ent_f_end[ent_wikiid] - e_freq.ent_f_start[ent_wikiid] + 1\n end\n return 0\nend\n\n" }, { "path": "deep-ed/deep-ed-master/entities/ent_name2id_freq/ent_name_id.lua", "content": "------------------ Load entity name-id mappings ------------------\n-- Each entity has:\n-- a) a Wikipedia URL referred as 'name' here\n-- b) a Wikipedia ID referred as 'ent_wikiid' or 'wikiid' here\n-- c) an ID that will be used in the entity embeddings lookup table. Referred as 'ent_thid' or 'thid' here.\n\ntds = tds or require 'tds' -- saves lots of memory for ent_name_id.lua. Mem overflow with normal {}\nlocal rltd_only = false\nif opt and opt.entities and opt.entities ~= 'ALL' then\n assert(rewtr.reltd_ents_wikiid_to_rltdid, 'Import relatedness.lua before ent_name_id.lua')\n rltd_only = true\nend\n\n-- Unk entity wikid:\nunk_ent_wikiid = 1\n\nlocal entity_wiki_txtfilename = opt.root_data_dir .. 'basic_data/wiki_name_id_map.txt'\nlocal entity_wiki_t7filename = opt.root_data_dir .. 'generated/ent_name_id_map.t7'\nif rltd_only then\n entity_wiki_t7filename = opt.root_data_dir .. 'generated/ent_name_id_map_RLTD.t7'\nend\n\nprint('==> Loading entity wikiid - name map') \n\nlocal e_id_name = nil\n\nif paths.filep(entity_wiki_t7filename) then\n print(' ---> from t7 file: ' .. entity_wiki_t7filename)\n e_id_name = torch.load(entity_wiki_t7filename)\n\nelse\n print(' ---> t7 file NOT found. Loading from disk (slower). Out f = ' .. entity_wiki_t7filename)\n dofile 'data_gen/indexes/wiki_disambiguation_pages_index.lua'\n print(' Still loading entity wikiid - name map ...') \n \n e_id_name = tds.Hash()\n \n -- map for entity name to entity wiki id\n e_id_name.ent_wikiid2name = tds.Hash()\n e_id_name.ent_name2wikiid = tds.Hash()\n\n -- map for entity wiki id to tensor id. Size = 4.4M\n if not rltd_only then\n e_id_name.ent_wikiid2thid = tds.Hash()\n e_id_name.ent_thid2wikiid = tds.Hash()\n end\n \n local cnt = 0\n local cnt_freq = 0\n for line in io.lines(entity_wiki_txtfilename) do\n local parts = split(line, '\\t')\n local ent_name = parts[1]\n local ent_wikiid = tonumber(parts[2]) \n \n if (not wiki_disambiguation_index[ent_wikiid]) then\n if (not rltd_only) or rewtr.reltd_ents_wikiid_to_rltdid[ent_wikiid] then\n e_id_name.ent_wikiid2name[ent_wikiid] = ent_name\n e_id_name.ent_name2wikiid[ent_name] = ent_wikiid\n end\n if not rltd_only then\n cnt = cnt + 1\n e_id_name.ent_wikiid2thid[ent_wikiid] = cnt\n e_id_name.ent_thid2wikiid[cnt] = ent_wikiid \n end\n end\n end\n\n if not rltd_only then\n cnt = cnt + 1\n e_id_name.ent_wikiid2thid[unk_ent_wikiid] = cnt\n e_id_name.ent_thid2wikiid[cnt] = unk_ent_wikiid\n end\n e_id_name.ent_wikiid2name[unk_ent_wikiid] = 'UNK_ENT'\n e_id_name.ent_name2wikiid['UNK_ENT'] = unk_ent_wikiid\n\n torch.save(entity_wiki_t7filename, e_id_name)\nend\n\nif not rltd_only then\n unk_ent_thid = e_id_name.ent_wikiid2thid[unk_ent_wikiid]\nelse\n unk_ent_thid = rewtr.reltd_ents_wikiid_to_rltdid[unk_ent_wikiid]\nend\n\n------------------------ Functions for wikiids and names-----------------\nfunction get_map_all_valid_ents()\n local m = tds.Hash()\n for ent_wikiid, _ in pairs(e_id_name.ent_wikiid2name) do\n m[ent_wikiid] = 1\n end\n return m \nend\n\nis_valid_ent = function(ent_wikiid)\n if e_id_name.ent_wikiid2name[ent_wikiid] then\n return true\n end\n return false\nend\n\n\nget_ent_name_from_wikiid = function(ent_wikiid)\n local ent_name = e_id_name.ent_wikiid2name[ent_wikiid]\n if (not ent_wikiid) or (not ent_name) then\n return 'NIL'\n end\n return ent_name\nend\n\npreprocess_ent_name = function(ent_name)\n ent_name = trim1(ent_name)\n ent_name = string.gsub(ent_name, '&', '&')\n ent_name = string.gsub(ent_name, '"', '\"')\n ent_name = ent_name:gsub('_', ' ')\n ent_name = first_letter_to_uppercase(ent_name)\n if get_redirected_ent_title then\n ent_name = get_redirected_ent_title(ent_name)\n end\n return ent_name\nend\n\nget_ent_wikiid_from_name = function(ent_name, not_verbose)\n local verbose = (not not_verbose)\n ent_name = preprocess_ent_name(ent_name)\n local ent_wikiid = e_id_name.ent_name2wikiid[ent_name]\n if (not ent_wikiid) or (not ent_name) then\n if verbose then\n print(red('Entity ' .. ent_name .. ' not found. Redirects file needs to be loaded for better performance.'))\n end\n return unk_ent_wikiid\n end\n return ent_wikiid\nend\n\n------------------------ Functions for thids and wikiids -----------------\n-- ent wiki id -> thid\nget_thid = function (ent_wikiid)\n if rltd_only then\n ent_thid = rewtr.reltd_ents_wikiid_to_rltdid[ent_wikiid]\n else\n ent_thid = e_id_name.ent_wikiid2thid[ent_wikiid]\n end\n if (not ent_wikiid) or (not ent_thid) then\n return unk_ent_thid\n end\n return ent_thid\nend\n\ncontains_thid = function (ent_wikiid)\n if rltd_only then\n ent_thid = rewtr.reltd_ents_wikiid_to_rltdid[ent_wikiid]\n else\n ent_thid = e_id_name.ent_wikiid2thid[ent_wikiid]\n end\n if ent_wikiid == nil or ent_thid == nil then\n return false\n end\n return true\nend\n\nget_total_num_ents = function()\n if rltd_only then\n assert(table_len(rewtr.reltd_ents_wikiid_to_rltdid) == rewtr.num_rltd_ents)\n return table_len(rewtr.reltd_ents_wikiid_to_rltdid)\n else\n return #e_id_name.ent_thid2wikiid\n end\nend\n \nget_wikiid_from_thid = function(ent_thid)\n if rltd_only then\n ent_wikiid = rewtr.reltd_ents_rltdid_to_wikiid[ent_thid]\n else\n ent_wikiid = e_id_name.ent_thid2wikiid[ent_thid]\n end \n if ent_wikiid == nil or ent_thid == nil then\n return unk_ent_wikiid\n end\n return ent_wikiid\nend\n\n-- tensor of ent wiki ids --> tensor of thids\nget_ent_thids = function (ent_wikiids_tensor)\n local ent_thid_tensor = ent_wikiids_tensor:clone()\n if ent_wikiids_tensor:dim() == 2 then\n for i = 1,ent_thid_tensor:size(1) do\n for j = 1,ent_thid_tensor:size(2) do\n ent_thid_tensor[i][j] = get_thid(ent_wikiids_tensor[i][j])\n end\n end\n elseif ent_wikiids_tensor:dim() == 1 then\n for i = 1,ent_thid_tensor:size(1) do\n ent_thid_tensor[i] = get_thid(ent_wikiids_tensor[i])\n end\n else\n print('Tensor with > 2 dimentions not supported')\n os.exit()\n end\n return ent_thid_tensor\nend\n\nprint(' Done loading entity name - wikiid. Size thid index = ' .. get_total_num_ents())\n" }, { "path": "deep-ed/deep-ed-master/entities/learn_e2v/4EX_wiki_words.lua", "content": "-- Small dataset. Used for debugging / unit testing\n\nif not is_stop_word_or_number then\n dofile 'words/stop_words.lua'\nend\n\nent_lines_4EX = {} \n\nent_lines_4EX['Japan national football team'] = '993546\tJapan national football team\tJapan national football team Japan is one of the most successful football teams in Asia having qualified for the last five consecutive FIFA World Cups with second round advancements in 2002 2010 and having won the AFC Asian Cup a record four times in 1992 2000 2004 2011 To this they add a 2001 FIFA Confederations Cup second place The Japanese team is commonly known by the fans and media as or as abbreviated expressions Although the team does not have an official nickname as such it is often known by the name of the manager For example under Takeshi Okada the team was known as Recently the team has been known or nicknamed as the Samurai Blue while news media still refer it to by manager s last name as or in short History Japan s first major achievement in international football came in the 1968 Summer Olympics in Mexico City where the team won the bronze medal Although this result earned the sport increased recognition in Japan the absence of a professional domestic league hindered its growth and Japan would not qualify for the FIFA World Cup until 30 years later In 1991 the owners of the semi-professional Japan Soccer League agreed to disband the league and re-form as the professional J League partly to raise the sport s profile and to strengthen the national team program With the launch of the new league in 1993 interest in football and the national team grew However in its first attempt to qualify with professional players Japan narrowly missed a ticket to the 1994 FIFA World Cup after failing to beat Iraq in the final match of the qualification round remembered by fans as the Agony of Doha The nation s first ever FIFA World Cup appearance was in 1998 where they lost all three matches Japan s first two fixtures went 1–0 in favor of Argentina and Croatia despite playing well in both games Their campaign ended with an unexpected 2–1 defeat to rank outsiders Jamaica Four years later Japan co-hosted the 2002 FIFA World Cup with South Korea Despite being held to a 2–2 draw by Belgium in their opening game the Japanese team advanced to the second round with a 1–0 win over Russia and a 2–0 victory against Tunisia However they subsequently exited the tournament during the Round of 16 after losing 1–0 to eventual third-place finishers Turkey On June 8 2005 Japan qualified for the 2006 FIFA World Cup in Germany its third consecutive World Cup by beating North Korea 2–0 on neutral ground However Japan failed to advance to the Round of 16 after finishing the group without a win losing to Australia 1–3 drawing Croatia 0–0 and losing to Brazil 1–4 Japan has had considerably more success in the Asian Cup taking home the winner s trophy in four of the last six finals in 1992 2000 2004 and 2011 Their principal continental rivals are South Korea followed by Saudi Arabia and most recently Australia Japan is the only team from outside the Americas to participate in the Copa América having been invited in 1999 and 2011 During the 2010 FIFA World Cup qualification in the fourth round of the Asian Qualifiers Japan became the first team other than the host South Africa to qualify after defeating Uzbekistan 1–0 away Japan was put in Group E along with the Netherlands Denmark and Cameroon Japan won its opening game of the 2010 FIFA World Cup defeating Cameroon 1–0 but subsequently lost to the Netherlands 0–1 before defeating Denmark 3–1 to advance to the next round against Paraguay In the first knockout round Japan were eliminated from the competition following penalties after a 0–0 draw against Paraguay After the World Cup head coach Takeshi Okada resigned He was replaced by former Juventus and AC Milan coach Alberto Zaccheroni In his first few matches Japan recorded victories over Guatemala 2–1 and Paraguay 1–0 as well as one of their best ever results – a 1–0 victory over Argentina At the start of 2011 Japan participated in the 2011 AFC Asian Cup in Qatar On 29 January they beat Australia 1–0 in the final after extra time their fourth Asian Cup triumph and allowing them to qualify for FIFA Confederations Cup Japan then started their road to World Cup 2014 Brazil with numerous qualifiers Throughout they suffered only two losses to Uzbekistan and Jordan and drawing against Australia Afterwards on October 12 Japan picked up a historic 1–0 victory over France a team they had never before defeated After a 1–1 draw with Australia they qualified for the 2014 FIFA World Cup becoming the first nation outside of Brazil who is hosting the tournament to qualify Japan started their 2013 FIFA Confederations Cup with a 3–0 loss to Brazil They were then eliminated from the competition after losing to Italy 3–4 but received praise for their style of play in the match They lost their final game 1-2 against Mexico and finished 4th place in Group A in the 2013 FIFA Confederations Cup One month later in the EAFF East Asian Cup they started out with a 3-3 draw to China They then beat Australia 3-2 and beat South Korea 2-1 in the 3rd and final match in the 2013 EAFF East Asian Cup to claim their first title in history Team image Fan Chanting Japanese national team supporters are known for chanting Nippon Ole Nippon is the Japanese term for Japan at home matches Kits and colours Japan s current kit is provided by Adidas the team s official apparel sponsor The home kit consists of a Navy blue jersey with a red line down the center with all support for Japan faintly written on it navy blue shorts with bright blue patches on the side and navy blue socks with a red line down the center The away kit consists of a white jersey white shorts and white socks all with In 2011 Japan switched temporarily the color of the numbers from white to gold Prior to Adidas the team s official apparel sponsor were the Japanese brand Asics and Puma The national team kit design has gone through several alterations in the past In the early 80s the kit was white with blue trim When Japan was coached by Kenzo Yokoyama 1988–1991 the kits were red and white matching the colors of Japan s national flag The kits worn for the 1992 AFC Asian Cup consisted of white stripes stilized to form a wing with red diamonds During Japan s first World Cup appearance in 1998 the national team kits were blue jerseys with red and white flame designs on the sleeves and designed by JFA with the sponsor alternating each year between Asics Puma and Adidas Japan uses blue and white rather than red and white due to a superstition In its first major international competition the 1936 Summer Olympics Japan used a blue kit in the match against Sweden and Japan won the match by a score of 3–2 Also the Japanese Football Association logo has some yellow it represents the fair play honesty in Japanese tradition all surrounding by blue on the jersey that means youth in Japanese tradition that also explains the colours of the uniform which could be translated as the fair play purpose supported on the power of youth Sponsorship Japan has one of the highest sponsorship incomes for a national squad In 2006 their sponsorship income amounted to over 16 5 million pounds Primary sponsors include Adidas Kirin Saison Card International FamilyMart JAL Mitsui Sumitomo Insurance Sony Asahi Shinbun Konami Mizuho Financial and Audi Mascot The mascots are Karappe and Karara two Yatagarasu wearing the Japan national football team uniform The mascots were designed by Japanese manga artist Susumu Matsushita Each year when a new kit is launched the mascots change uniforms Players Current squad Squad selected for the friendlies matches vs Netherlands on 16 November and Belgium on 19 November 2013 Recent call-ups The following players have also been called up to the Japan squad within the last 12 months Competitions Olympic Games Since 1992 the Olympic team has been drawn from a squad with a maximum of three players over 23 years of age and the achievements of this team are not generally regarded as part of the national team s records nor are the statistics credited to the players international records Copa América Japan is the only team from outside the Americas to participate in the Copa América having been invited in both 1999 and 2011 However Japan declined their invitation on May 16 2011 after events related with the Tōhoku earthquake and difficulty to release some Japanese players from European teams to play as a replacement On May 17 2011 CONMEBOL invited Costa Rica to replace Japan in the competition the Costa Rican Football Federation accepted their invitation later that day'\n\nent_lines_4EX['Leicestershire'] = '61153\tLeicestershire\tLeicestershire Leicestershire or abbreviation Leics is a landlocked county in the English Midlands It takes its name from the City of Leicester traditionally its administrative centre although the City of Leicester unitary authority is today administered separately from the rest of Leicestershire The county borders Nottinghamshire to the north Lincolnshire to the north-east Rutland to the east Northamptonshire to the south-east Warwickshire to the south-west Staffordshire to the west and Derbyshire to the north-west The border with Warwickshire is Watling Street the A5 The county has a population of just under 1 million with over half the population living in Leicester s built-up area History Leicestershire was recorded in the Domesday Book in four wapentakes Guthlaxton Framland Goscote and Gartree These later became hundreds with the division of Goscote into West Goscote and East Goscote and the addition of Sparkenhoe hundred In 1087 the first recorded use of the name was as Laegrecastrescir Leicestershire s external boundaries have changed little since the Domesday Survey The Measham - Donisthorpe exclave of Derbyshire has been exchanged for the Netherseal area and the urban expansion of Market Harborough has caused Little Bowden previously in Northamptonshire to be annexed In 1974 the Local Government Act 1972 abolished the county borough status of Leicester city and the county status of neighbouring Rutland converting both to administrative districts of Leicestershire These actions were reversed on 1 April 1997 when Rutland and the City of Leicester became unitary authorities Rutland became a distinct Ceremonial County once again although it continues to be policed by Leicestershire Constabulary The symbol of the county council Leicestershire County Cricket Club and Leicester City FC is the fox Leicestershire is considered to be the birthplace of fox hunting as it is known today Hugo Meynell who lived in Quorn is known as the father of fox hunting Melton Mowbray and Market Harborough have associations with fox hunting as has neighbouring Rutland Geography The River Soar rises to the east of Hinckley in the far south of the county and flows northward through Leicester before emptying into the River Trent at the point where Derbyshire Leicestershire and Nottinghamshire meet A large part of the north-west of the county around Coalville forms part of the new National Forest area extending into Derbyshire and Staffordshire The highest point of the county is Bardon Hill at 278 metres 912 ft which is also a Marilyn Demographics The population of Leicestershire excluding the Leicester unitary authority is 609 578 people 2001 census The county covers an area of 2 084 km2 804 sq mi Its largest population centre is the city of Leicester followed by the town of Loughborough Other large towns include Ashby-de-la-Zouch Coalville Hinckley Market Harborough Melton Mowbray Oadby Wigston and Lutterworth Some of the larger of Leicestershire s villages are Birstall population 11 400 in 2004 Broughton Astley Castle Donington Kibworth Beauchamp along with Kibworth Harcourt Great Glen Ibstock Countesthorpe and Kegworth One of the most rapidly expanding villages is Anstey which has recently seen a large number of development schemes The United Kingdom Census 2001 showed a total resident population for Leicester of 279 921 a 0 5 decrease from the 1991 census Approximately 62 000 were aged under 16 199 000 were aged 16–74 and 19 000 aged 75 and over 76 9 of Leicester s population claim they have been born in the UK according to the 2001 UK Census Mid-year estimates for 2006 indicate that the population of the City of Leicester stood at 289 700 making Leicester the most populous city in East Midlands The population density is and for every 100 females there were 92 9 males Of those aged 16–74 in Leicester 38 5 had no academic qualifications significantly higher than 28 9 in all of England 23 0 of Leicester s residents were born outside of the United Kingdom more than double than the English average of 9 2 Economy Engineering Engineering has long been an important part of the economy of Leicestershire John Taylor Bellfounders continues a history of bellfounding in Loughborough since the 14th century In 1881 John Taylors cast the largest bell in Britain Great Paul for St Paul s Cathedral in London Norman Underwood have been making sand cast sheet lead roofing and stained glass since 1825 working on many of England s major cathedrals and historic buildings including Salisbury Cathedral Windsor Castle Westminster Abbey Hampton Court Palace and Chatsworth House Snibston Discovery Park is built on one of three coal mines that operated in Coalville from the 1820s until 1986 Abbey Pumping Station houses four enormous steam powered beam engines built in Leicester in the 1890s in the Vulcan factory owned by Josiah Gimson whose son Ernest Gimson was an influential furniture designer and architect of the English arts and crafts movement Engineering companies today include sports car maker Noble Automotive Ltd in Barwell Triumph Motorcycles in Hinckley Jones Shipman machine tools Metalfacture Ltd sheet metal work Richards Engineering foundry equipment Transmon Engineering materials handling equipment Trelleborg Industrial AVS in Beaumont Leys industrial suspension components Parker Plant quarrying equipment Aggregate Industries UK construction materials Infotec in Ashby-de-la-Zouch electronic information display boards Alstec in Whetstone Leicestershire airport baggage handling systems and Brush Traction railway locomotives in Loughborough Local commitment to nurturing the upcoming cadre of British engineers includes apprenticeship schemes with local companies and academic-industrial connections with the engineering departments at Leicester University De Montfort University and Loughborough University The Systems Engineering Innovation Centre and Centre for Excellence for low carbon and fuel cell technologies are both based at Loughborough University Private sector research and development organisations include PERA - the technology based consultancy in Melton Mowbray and MIRA - the automotive research and development centre based on the outskirts of Hinckley Automotive and aerospace engineers use the test facilities at Mallory Park and Bruntingthorpe Aerodrome and proving ground On 18 October 2007 the last airworthy Avro Vulcan was flown from Bruntingthorpe Aerodrome after 10 years of restoration there by aerospace engineers of the Vulcan Operating Company Trade associations There are several trade associations with their head offices based in Leicestershire including the Ergonomics Society the European Construction Institute the Institute of Diagnostic Engineers the Pre-cast Flooring Federation the Concrete Pipe Association the Timber Packaging Pallet Confederation and the National Association of Wood Shaving Sawdust Merchants Contractors Farming Leicestershire has a long history of livestock farming which continues today Robert Bakewell farmer 1725–1795 of Dishley near Loughborough was a revolutionary in the field of selective breeding Bakewell s Leicester Longwool sheep was much prized by farmers across the British Empire and is today a heritage breed admired all over the world Commercial and rare breeds associated with the descendants of Bakewell s sheep include the English Leicester Border Leicester Bluefaced Leicester Scotch mule and Welsh halfbred In 2006 in Leicestershire and Rutland there were 6 450 people working as farmers managers and farm labourers on 2 719 farms with of farmed land The animal population was 122 284 cattle 57 059 pigs and 314 214 sheep Source DEFRA The Leicestershire County Show is held on the first Bank Holiday in May each year and includes animal showings trade exhibitions and show jumping Melton Mowbray Market is an important regional livestock market Field Sports remain an important part of the rural economy of Leicestershire with stables kennels and gunsmiths based in the county Thatched roofs are built and maintained by members of Rutland Leicestershire Master Thatchers Association Food and drink Stilton Red Leicester cheese and the pork pie are three of Leicestershire s most famous contributions to English cuisine Leicestershire food producers include Claybrooke mill one of the very few commercially working watermills left in Britain producing a range of over 40 flours meat from rare and minority breeds from Brockleby s Christmas turkey and goose from Seldom Seen Farm Two dairies produce Red Leicester cheese in the county Long Clawson and the Leicestershire Handmade Cheese Company All natural non-alcoholic fruit cordials and presse drinks are made by Belvoir Fruit Farms and sold in supermarkets across Britain Swithland Spring Water is sourced from the Charnwood hills Breweries in Leicestershire and Rutland are listed on the Leicester CAMRA website The county s largest beer brewer is Everards and there are several microbreweries such as Belvoir Brewery in Old Dalby Parish Brewery in Burrough on the Hill Wicked Hathern Brewery in Loughborough the Gas Dog Brewery at Somerby near Melton Ellis Wood brewery in Hinckley and the Pig Pub Brewery in Claybrooke Magna near Lutterworth Vineyards in Leicestershire include Chevelswarde Vineyard Lutterworth Welland Valley Vineyard Market Harborough and Eglantine Loughborough Melton Mowbray Sloe Gin is a liqueur with a distinctive flavour Various markets are held across the county Leicester Market is the largest outdoor covered marketplace in Europe and among the products on sale are fruit and vegetables sold by enthusiastic market stallholders who shout out their prices and fresh fish and meat in the Indoor Market The annual East Midlands Food Drink Festival held in Melton Mowbray had over 200 exhibitors and 20 000 visitors attending in 2007 making it the largest British regional food festival Food processing in the city and county includes popular British fish and chip shop pie Pukka Pies who are based in Syston Walkers Midshire Foods part of the Samworth Brothers group makes sausages and pies in its Beaumont Leys factories Samworth Brothers has operations in Leicestershire and Cornwall Ginsters making a range of products from sandwiches to desserts for UK retailers under their brands as well the company s own portfolio of brands including Dickinson Morris producers of pork pies and Melton Hunt Cake Walkers crisps are made in Beaumont Leys using Lincolnshire potatoes United Biscuits have their distribution centre in Ashby-de-la-Zouch as well as a snacks factory producing brands such as Hula Hoops Skips snack Nik Naks and Space Raiders and they also have a biscuit factory in Wigston The Masterfoods UK factory at Melton Mowbray produces petfood for brands such as Cesar Kitekat PAL Pedigree Sheba Whiskas Aquarian and Trill Hand made chocolates are produced by Chocolate Perfection in Ashby-de-la-Zouch Some 15 major Indian food manufacturers are based in Leicester including Sara Foods Mayur Foods Cofresh Snack Foods Ltd Farsan Apni Roti and Spice n Tice The Mithai Indian sweet market is catered for by award winning Indian restaurants – for instance the vegetable samosas approved by the Vegetarian Society sold at The Sharmilee on Belgrave Road The growing market for Indian food has afforded new opportunities to long standing local companies for example the Long Clawson dairy a co-operative manufacturer of Stilton cheese now also makes Paneer cheese used in the Indian dish Mattar Paneer Leicestershire food exported abroad includes cheese from the Long Clawson dairy which is sold in supermarkets in Canada and the United States via a network of distributors coordinated by Taunton based company Somerdale Belvoir Fruit Farms cordials and pressé drinks are sold on the United States east coast in Wegmans Food Markets World Market Harris Teeter Dean DeLuca and in specialized British food stores such as Myers of Keswick New York City and the British Pantry near Washington D C The annual Leicestershire Rutland Restaurant Awards has several categories including Leicestershire Rutland Restaurant of the Year Best Asian Restaurant Best Service Best Newcomer Best Fine Dining Restaurant Best Value for Money Best Drinks Wine List Best Local Produce Menu Best Gastro Pub Best Neighbourhood Restaurant Best Business Lunch and Leicestershire Rutland Young Chef of the Year See also Leicester food drink Clothing Leicester and Leicestershire has had a traditional industry of knitwear hosiery and footwear and the sheep on the county s coat of arms is recognition of this The rich history of the East Midlands knitting knitwear industry is chronicled on the Knitting Together website The local manufacturing industry which began with hand knitting in the Middle Ages and was fully industrialized by the end of the 19th century survived until the end of the 20th century through retailers buying UK sourced products and government measures such as the protection of the Multi Fibre Arrangement which ended in 2004 Cheaper global competition coupled with the 1999 slump in the UK fashion retail sector led to the end of much of the cheaper clothing manufacturing industry Today Leicestershire companies focus on high quality clothing and specialty textiles One such company is Pantherella who make socks at their Hallaton Street factory off Saffron Lane which are sold in high-end department stores around the world including in the UK Harrods Selfridges and John Lewis and in the US in Nordstrom Bergdorf Goodman and Neiman Marcus Other local companies manufacture knitwear such as Commando Knitwear of Wigston and others specialize in technical textiles for industrial or medical purposes Clothing and fabric for the British Asian community is made here - for example the shop Saree Mandir sells silk saree s and salwar suits for women whose design patterns closely follow contemporary Indian trends The Knitting Industries Federation continues to be based in Leicestershire On the creative side the design centre for Next clothing is in Enderby and the design centre for George Clothing Asda Walmart is in Lutterworth De Montfort University has in the form of its Fashion and Contour Design course a leading design department for female underwear It also has the only UK University courses in Footwear Design providing future designers for local shoemakers Shoefayre Stead and Simpson and Shoe Zone who all have their headquarters in the county Gola also originates from the county Healthcare University Hospitals Leicester NHS Trust employs around 11 000 at its three hospitals in the city and county the Glenfield the General and the Royal Infirmary Leicestershire Partnership NHS Trust employs around 5 600 staff providing mental health learning disability and community health services in the city and county These services are commissioned by the three newly established Clinical Commissioning Groups led by local GP s The British Psychological Society and the Institute of Occupational Safety Health IOSH based in Wigston have their head offices in Leicestershire Biomedical industries Pharmaceutical biotechnology and medical instrument manufacturing companies include 3M Bridgehead International in Melton Fisher Scientific in Loughborough and Ashfield Healthcare in Ashby de-la-Zouch Freight and distribution Transportation links are good East Midlands Airport is one mile 1 6 km south of Castle Donington next to the M1 in north-west Leicestershire and is the second largest freight airport in the United Kingdom after London Heathrow DHL Aviation have a large purpose built facility at EMA and courier companies UPS and TNT also use the airport as a base Lufthansa Cargo is also a regular user of East Midlands and the airport is a primary hub for Royal Mail The M1 is Leicestershire s other important transport hub The start of the M6 and part of the A14 briefly intersect with the southern tip of Leicestershire Many large retail companies have huge warehouses at the Magna Park complex near Lutterworth The Widdowson Group make use of J21a of the M1 to provide warehousing transportation freight forwarding garage services and LGV HGV training Pall-Ex of Ellistown provide automated palletised freight distribution services from their location off Junction 22 of the M1 The Midland Main Line provides important connections to Yorkshire and London and the Birmingham–Stansted Line is essentially Leicestershire s east–west connection from Hinckley to Melton Other Ibstock based developer Wilson Bowden was bought in 2007 by Barratt Developments plc in a GBP2 2 billion deal Charles Street Buildings Leicester and Jelson Homes are two other successful Leicester based property companies Syston based Dunelm Mill is a growing home furnishings retailer The company started in 1979 as a family business selling curtains from a Leicester market stall whose first store opened in Churchgate Leicester in 1984 In 2006 Dunelm opened its 80th store and the company floated on the stock market placing the company s founders the Adderley family among Britain s most successful entrepreneurs Hamilton based LPC Group manufactures more than 600 million toilet rolls and kitchen towel rolls per year in its Leicestershire factories Toy car company Corgi have their European operation at the Meridian Business Park although the toys are now manufactured in China and the company is owned by Margate based Hornby Hairdresser Barrie Hedley operates three Barrie Stephen salons in the city and county and has been a finalist in the British hairdressing awards 2004 2005 and 2006 In 2007 Hedley won the Entrepreneur of the year at the Leicestershire Business Awards Lumbers of Market Street Leicester was a finalist in the Independent Retailer category of the UK Jewellery Awards 2007 Ulverscroft Large Print Books of Anstey Leicestershire are a leading publisher of books for the visually impaired Leicestershire is twinned with Kilkenny Ireland Leicester s Cultural Quarter is an ambitious plan to drive the regeneration of a large run-down area of the City It has delivered A new venue for the performing arts Curve Creative workspaces for artists designers LCB Depot and a Digital Media Centre A huge number of Creative and Media businesses have thrived in the region such as the digital agency Cite http www cite co uk In addition the area now has much-improved streets pavements and open spaces with integrated artworks As part of a 2002 marketing campaign the plant conservation charity Plantlife chose the Foxglove as the county flower Financial and business services Financial and business service companies with operations in Leicestershire include Alliance Leicester Royal Bank of Scotland State Bank of India HSBC and PricewaterhouseCoopers Pension provision company Mattioli Woods employs 170 people at its Grove Park Enderby HQ and has a reputation for employing graduates directly from Leicestershire Universities Companies that have their head office in the area include Next clothing and British Gas Business The Institute of Credit Management the European Association of Trade Mark Owners and the Point of Purchase Advertising International POPAI are based in Leicestershire Key stakeholders promoting economic development formed Leicester Leicestershire Economic Partnership in 2011 Leicestershire Chamber of Commerce is another good source for business advice Business awards The Leicestershire Business Awards has categories including Investing in Leicestershire Contribution to the Community and Entrepreneur of the Year Recent Leicestershire winners of the Queen s Award for Enterprise are listed on the Lord Lieutenant s website Statistics This is a chart of trend of regional gross value added of the non-metropolitan county of Leicestershire and Rutland it does not include the City of Leicester at current basic prices published pp  240–253 by Office for National Statistics with figures in millions of British Pounds Sterling Local government County Hall situated in Glenfield about 3 miles 5 km north-west of Leicester city centre is the seat of Leicestershire County Council and the headquarters of the county authority The City of Leicester is administered from offices in Leicester itself and the City Council meets at Leicester Town Hall Below the County Council seven district councils for a second tier of local government Education Publicly funded secondary schools in Leicestershire are comprehensive The schools are segregated by age in some areas to ages 10–14 middle schools and 14–16 upper schools or 14–18 upper schools which also provide sixth form education The schools compared with other LEAs have large numbers on the roll with school enrollment often 2000 and more For Melton and Blaby districts although there is division by middle and upper schools there is only one upper school in either district giving no choice of school However it should be noted that many students of Lutterworth College in Harborough District actually hail from Blaby district Charnwood has the largest school population – four times the size of the Melton district In 2007 the best-performing state school at GCSE was Beauchamp College in Oadby No comprehensives in Leicestershire LEA were rated as poor performers unlike in some neighbouring counties In 2007 7 800 pupils took GCSE exams For A-levels the best comprehensive school in the county was the De Lisle Catholic Science College in Loughborough The best schools overall at A-level were the two private single-sex schools in Loughborough Loughborough Grammar School and Loughborough High School GCSE results by district council of pupils gaining 5 grades A-C in 2007 including English and Maths 46 8 was the England average compared to Leicestershire s 48 9 Private schools Private schools in Leicestershire include Leicester Grammar School mixed Leicester High School for Girls girls Loughborough Grammar School boys Loughborough High School girls Fairfield Preparatory School primary school – mixed Welbeck College military 6th form college – mixed Ratcliffe College Roman Catholic – mixed Grace Dieu Manor School Roman Catholic – mixed Stoneygate school primary school – mixed and Stoneygate College mixed Our Lady s Convent School OLCS Roman Catholic - girls Further education Leicester College offers among others courses in catering cookery hospitality and leisure plumbing electrician carpentry and joinery building trades and gas motor vehicle maintenance computing business design and media and print Stephenson College Coalville offers among others courses in construction building trades and gas motor vehicle maintenance and repair beauty computing business sport and coaching care and complementary therapy Farming sector training Brooksby Melton College provides apprenticeships and further education training courses in animal care countryside equine fisheries and land based service engineering at their Brooksby campus Music education Soar Valley Music Centre offers further education courses in music performance and production Higher education Leicestershire has three universities the University of Leicester Loughborough University and De Montfort University Educational associations Several educational associations have their head offices in Leicestershire including the Mathematical Association the Association of School and College Leaders the Association for College Management the Girls Schools Association the National Adult School Association the National Institute of Adult Continuing Education and the Headmasters Headmistresses Conference Sporting associations A number of UK sporting bodies have their head offices in Leicestershire including the Institute of Sports Recreation Management the Institute of Swimming Teachers Coaches the English Volleyball Association the Great Britain Wheelchair Basketball Association the British Hang Gliding and Paragliding Association the British Judo Association the British Parachute Association the British Triathlon Federation the Amateur Swimming Association the British Gliding Association the British Motorcycle Federation the English Indoor Bowls Association the Youth Sports Trust and the British Isles Bowls Council Music The full range of music is performed in the county from early medieval European and Asian classical music folk jazz blues rock and pop The major Download Festival a hard rock and metal festival is hosted at Donington Park Symphony orchestras The Philharmonia Orchestra Leicester Symphony Orchestra and the internationally famous Leicestershire Schools Symphony Orchestra are three of the larger orchestras in the county Amateur orchestras Leicestershire Sinfonia the Loughborough Orchestra the Charnwood orchestra the Coalville Light Orchestra and the Soar Valley Music Centre Orchestra Choirs and choral societies Leicester based choirs include the Leicester Bach Choir Broom Leys Choral Society Whitwick Cantamici the Cecilian Singers Charnwood Choral Society Coalville and District Male Voice Choir Coro Nostro Chamber Choir Humberstone Choral Society Kainé Gospel Choir Kingfisher Chorale Leicester Church Music Consort Leicester City Male Voice Choir Leicester Philharmonic Choir Leicestershire Chorale Loughborough Male Voice Choir Meridian Singers Newtown Linford mixed voice choir Red Leicester choir the Scarlet choir Shepshed Singers Synergy Community Choir Wigston and district male voice choir Unity Community Choir and the Peepul Choir Early music The Longsdale Consort perform music of the renaissance and baroque periods Leicester Recorder Society Music shops Stores selling sheet music and musical instruments in Leicestershire include Sona Rupa Indian Sheehans Music Instruments Intasound Music Ltd ABC Music Market Harborough and MH Music'\n\nent_lines_4EX['Leicestershire County Cricket Club'] = '1622318\tLeicestershire County Cricket Club\tLeicestershire County Cricket Club Leicestershire County Cricket Club is one of the 18 major county clubs which make up the English and Welsh national cricket structure representing the historic county of Leicestershire It has also been representative of the county of Rutland Its limited overs team is called the Leicestershire Foxes Their kit colours are red with black trim in the Clydesdale Bank 40 and black with red trim in the t20 The shirt sponsors are Oval Insurance Broking with Highcross Leicester shopping centre on the top reverse side of the shirt The club is based at Grace Road Leicester and have also played home games at Aylestone Road in Leicester at Hinckley Loughborough Melton Mowbray Ashby-de-la-Zouch and in Coalville inside the traditional county boundaries and at Uppingham and Oakham over the border in Rutland Leicestershire are in the second divisions of the County Championship and in Group C of the Pro40 one day league They recently finished bottom of the County Championship for the first time since the introduction of two divisions Their best showing in recent years has been in the Twenty20 Cup with the Foxes winning the trophy three times in eight years Honours Second XI honours 1 Bain Hogg Trophy – 2nd 11 one day competition – 1996 History Earliest cricket Cricket may not have reached the county until well into the 18th century A notice in the Leicester Journal dated 17 August 1776 is the earliest known mention of cricket in Leicestershire But it was only a few years after that before a Leicestershire and Rutland Cricket Club was taking part in important matches mainly against Nottingham Cricket Club and Marylebone Cricket Club MCC This club was prominent from 1781 until the beginning of the 19th century 19th century Little more is heard of Leicestershire cricket until the formation of the present club on 25 March 1879 Essex CCC versus Leicestershire CCC at Leyton on 14 15 16 May 1894 was the initial first-class match played by either club In 1895 the County Championship was restructured into a 14-team competition with the introduction of Essex Leicestershire and Warwickshire CCC Early and mid 20th century Leicestershire s first 70 years were largely spent in lower table mediocrity with few notable exceptions In 1953 the motivation of secretary-captain Charles Palmer lifted the side fleetingly to third place but most of the rest of the 1950s was spent propping up the table or thereabouts Start of improvement The late 1950s and the 1960s Change came in the late 1950s with the recruitment of the charismatic Willie Watson at the end of a distinguished career with England and Yorkshire Watson s run gathering sparked the home-grown Maurice Hallam into becoming one of England s best opening batsmen In bowling Leicestershire had an erratically successful group of seamers in Terry Spencer Brian Boshier John Cotton and Jack van Geloven plus the spin of John Savage Another change was in the captaincy Tony Lock the former England and Surrey spinner who had galvanised Western Australia The 1970s and the first golden era Ray Illingworth again from Yorkshire instilled self-belief to the extent that the county took its first ever trophy in 1972 the Benson Hedges Cup with Chris Balderstone man of the match This was start of the first golden era as the first of five trophies in five years and included Leicestershire s first ever County Championship title in 1975 A couple of runners up spots were also thrown in The game when Leicestershire won their first ever County Championship on 15 September 1975 marked something of a personal triumph for Chris Balderstone Batting on 51 not out against Derbyshire at Chesterfield after close of play he changed into his football kit to play for Doncaster Rovers in an evening match 30 miles away a 1–1 draw with Brentford Thus he is the only player to have played League Football and first class cricket on the same day He then returned to Chesterfield to complete a century the following morning and take three wickets to wrap up the title To add to that season s success for Leicestershire was a second Benson Hedges victory The 1980s A runners up spot in the 1982 County Championship brought some respectability but the decade s only first class silverware was in the 1985 Benson Hedges Cup with Balderstone still on board making him the most successful trophy winner in the club s history with six Success in the late 1990s Leicestershire won the county championship in 1996 and again in 1998 This was an amazing achievement considering the resources of the club compared to other county teams This Leicestershire side led by Jack Birkenshaw and James Whitaker used team spirit and togetherness to get the best out of a group of players who were either discarded from other counties or brought through the Leicestershire ranks This team did not have many stars but Aftab Habib Darren Maddy Vince Wells Jimmy Ormond Alan Mullally and Chris Lewis all had chances for England West Indian all-rounder Phil Simmons was also named as one of Wisden s Cricketers of the year in 1997 while playing for the club International players England Australia Bangladesh India New Zealand Pakistan South Africa West Indies Records Most first-class runs for Leicestershire br Qualification – 17000 runs Most first-class wickets for Leicestershire br Qualification – 600 wickets Most first team winners medal for Leicestershire Batting Best partnership for each wicket county championship Sub Academy The Leicestershire Sub Academy is designed for young cricketers who have potential to play at the highest level It is also called the EPP Emerging Player Programme Many players who are involved in this set up move on to the LCCC academy where they will play matches against academies from other counties'\n\nent_lines_4EX['Barack Obama'] = '534366\tBarack Obama\tBarack Obama Barack Hussein Obama II born August 4 1961 is the 44th and current President of the United States the first African American to hold the office Born in Honolulu Hawaii Obama is a graduate of Columbia University and Harvard Law School where he was president of the Harvard Law Review He was a community organizer in Chicago before earning his law degree He worked as a civil rights attorney and taught constitutional law at the University of Chicago Law School from 1992 to 2004 He served three terms representing the 13th District in the Illinois Senate from 1997 to 2004 running unsuccessfully for the United States House of Representatives in 2000 In 2004 Obama received national attention during his campaign to represent Illinois in the United States Senate with his victory in the March Democratic Party primary his keynote address at the Democratic National Convention in July and his election to the Senate in November He began his presidential campaign in 2007 and in 2008 after a close primary campaign against Hillary Rodham Clinton he won sufficient delegates in the Democratic Party primaries to receive the presidential nomination He then defeated Republican nominee John McCain in the general election and was inaugurated as president on January 20 2009 Nine months after his election Obama was named the 2009 Nobel Peace Prize laureate During his first two years in office Obama signed into law economic stimulus legislation in response to the Great Recession in the form of the American Recovery and Reinvestment Act of 2009 and the Tax Relief Unemployment Insurance Reauthorization and Job Creation Act of 2010 Other major domestic initiatives in his first term include the Patient Protection and Affordable Care Act often referred to as Obamacare the Dodd–Frank Wall Street Reform and Consumer Protection Act and the Don t Ask Don t Tell Repeal Act of 2010 In foreign policy Obama ended U S military involvement in the Iraq War increased U S troop levels in Afghanistan signed the New START arms control treaty with Russia ordered U S military involvement in Libya and ordered the military operation that resulted in the death of Osama bin Laden He later became the first sitting U S president to publicly support same-sex marriage In November 2010 the Republicans regained control of the House of Representatives as the Democratic Party lost a total of 63 seats and after a lengthy debate over federal spending and whether or not to raise the nation s debt limit Obama signed the Budget Control Act of 2011 and the American Taxpayer Relief Act of 2012 Obama was re-elected president in November 2012 defeating Republican nominee Mitt Romney and was sworn in for a second term on January 20 2013 During his second term Obama has promoted domestic policies related to gun control in response to the Sandy Hook Elementary School shooting has called for full equality for LGBT Americans and his administration filed briefs which urged the Supreme Court to strike down the Defense of Marriage Act of 1996 and California s Proposition 8 as unconstitutional In foreign policy Obama has continued the process of ending U S combat operations in Afghanistan Early life and career Obama was born on August 4 1961 at Kapiolani Maternity Gynecological Hospital now Kapiolani Medical Center for Women and Children in Honolulu Hawaii and is the first President to have been born in Hawaii His mother Stanley Ann Dunham was born in Wichita Kansas and was of mostly English ancestry His father Barack Obama Sr was a Luo from Nyang’oma Kogelo Kenya Obama s parents met in 1960 in a Russian class at the University of Hawaiʻi at Mānoa where his father was a foreign student on scholarship The couple married in Wailuku on Maui on February 2 1961 and separated when Obama s mother moved with their newborn son to Seattle Washington in late August 1961 to attend the University of Washington for one year In the meantime Obama Sr completed his undergraduate economics degree in Hawaii in June 1962 then left to attend graduate school at Harvard University on a scholarship Obama s parents divorced in March 1964 Obama Sr returned to Kenya in 1964 where he remarried he visited Barack in Hawaii only once in 1971 He died in an automobile accident in 1982 when his son was 21 years old In 1963 Dunham met Lolo Soetoro an Indonesian East–West Center graduate student in geography at the University of Hawaii and the couple were married on Molokai on March 15 1965 After two one-year extensions of his J-1 visa Lolo returned to Indonesia in 1966 followed sixteen months later by his wife and stepson in 1967 with the family initially living in a Menteng Dalam neighborhood in the Tebet subdistrict of south Jakarta then from 1970 in a wealthier neighborhood in the Menteng subdistrict of central Jakarta From ages six to ten Obama attended local Indonesian-language schools St Francis of Assisi Catholic School for two years and Besuki Public School for one and a half years supplemented by English-language Calvert School homeschooling by his mother In 1971 Obama returned to Honolulu to live with his maternal grandparents Madelyn and Stanley Dunham and with the aid of a scholarship attended Punahou School a private college preparatory school from fifth grade until his graduation from high school in 1979 Obama lived with his mother and sister in Hawaii for three years from 1972 to 1975 while his mother was a graduate student in anthropology at the University of Hawaii Obama chose to stay in Hawaii with his grandparents for high school at Punahou when his mother and sister returned to Indonesia in 1975 to begin anthropology field work His mother spent most of the next two decades in Indonesia divorcing Lolo in 1980 and earning a PhD in 1992 before dying in 1995 in Hawaii following treatment for ovarian cancer and uterine cancer Of his early childhood Obama recalled That my father looked nothing like the people around me—that he was black as pitch my mother white as milk—barely registered in my mind He described his struggles as a young adult to reconcile social perceptions of his multiracial heritage Reflecting later on his years in Honolulu Obama wrote The opportunity that Hawaii offered—to experience a variety of cultures in a climate of mutual respect—became an integral part of my world view and a basis for the values that I hold most dear Obama has also written and talked about using alcohol marijuana and cocaine during his teenage years to push questions of who I was out of my mind Obama was also a member of the choom gang a self-named group of friends that spent time together and occasionally smoked marijuana Following high school Obama moved to Los Angeles in 1979 to attend Occidental College In February 1981 he made his first public speech calling for Occidental to participate in the disinvestment from South Africa in response to its policy of apartheid In mid-1981 Obama traveled to Indonesia to visit his mother and half-sister Maya and visited the families of college friends in Pakistan and India for three weeks Later in 1981 he transferred as a junior to Columbia College Columbia University in New York City where he majored in political science with a specialty in international relations and graduated with a Bachelor of Arts in 1983 He worked for a year at the Business International Corporation then at the New York Public Interest Research Group Community organizer and Harvard Law School Two years after graduating Obama was hired in Chicago as director of the Developing Communities Project DCP a church-based community organization originally comprising eight Catholic parishes in Roseland West Pullman and Riverdale on Chicago s South Side He worked there as a community organizer from June 1985 to May 1988 He helped set up a job training program a college preparatory tutoring program and a tenants rights organization in Altgeld Gardens Obama also worked as a consultant and instructor for the Gamaliel Foundation a community organizing institute ref In mid-1988 he traveled for the first time in Europe for three weeks and then for five weeks in Kenya where he met many of his paternal relatives for the first time He returned to Kenya in 1992 with his fiancée Michelle and his half-sister Auma He returned to Kenya in August 2006 for a visit to his father s birthplace a village near Kisumu in rural western Kenya In late 1988 Obama entered Harvard Law School He was selected as an editor of the Harvard Law Review at the end of his first year and president of the journal in his second year During his summers he returned to Chicago where he worked as an associate at the law firms of Sidley Austin in 1989 and Hopkins Sutter in 1990 After graduating with a J D magna cum laude from Harvard in 1991 he returned to Chicago Obama s election as the first black president of the Harvard Law Review gained national media attention and led to a publishing contract and advance for a book about race relations which evolved into a personal memoir The manuscript was published in mid-1995 as Dreams from My Father University of Chicago Law School and civil rights attorney In 1991 Obama accepted a two-year position as Visiting Law and Government Fellow at the University of Chicago Law School to work on his first book He then taught at the University of Chicago Law School for twelve years—as a Lecturer from 1992 to 1996 and as a Senior Lecturer from 1996 to 2004—teaching constitutional law From April to October 1992 Obama directed Illinois s Project Vote a voter registration campaign with ten staffers and seven hundred volunteer registrars it achieved its goal of registering 150 000 of 400 000 unregistered African Americans in the state leading Crain s Chicago Business to name Obama to its 1993 list of 40 under Forty powers to be In 1993 he joined Davis Miner Barnhill Galland a 13-attorney law firm specializing in civil rights litigation and neighborhood economic development where he was an associate for three years from 1993 to 1996 then of counsel from 1996 to 2004 His law license became inactive in 2007 From 1994 to 2002 Obama served on the boards of directors of the Woods Fund of Chicago which in 1985 had been the first foundation to fund the Developing Communities Project and of the Joyce Foundation He served on the board of directors of the Chicago Annenberg Challenge from 1995 to 2002 as founding president and chairman of the board of directors from 1995 to 1999 Legislative career 1997–2008 State Senator 1997–2004 Obama was elected to the Illinois Senate in 1996 succeeding State Senator Alice Palmer as Senator from Illinois s 13th District which at that time spanned Chicago South Side neighborhoods from Hyde Park  – Kenwood south to South Shore and west to Chicago Lawn Once elected Obama gained bipartisan support for legislation that reformed ethics and health care laws He sponsored a law that increased tax credits for low-income workers negotiated welfare reform and promoted increased subsidies for childcare In 2001 as co-chairman of the bipartisan Joint Committee on Administrative Rules Obama supported Republican Governor Ryan s payday loan regulations and predatory mortgage lending regulations aimed at averting home foreclosures Obama was reelected to the Illinois Senate in 1998 defeating Republican Yesse Yehudah in the general election and was reelected again in 2002 In 2000 he lost a Democratic primary race for Illinois s 1st congressional district in the United States House of Representatives to four-term incumbent Bobby Rush by a margin of two to one In January 2003 Obama became chairman of the Illinois Senate s Health and Human Services Committee when Democrats after a decade in the minority regained a majority He sponsored and led unanimous bipartisan passage of legislation to monitor racial profiling by requiring police to record the race of drivers they detained and legislation making Illinois the first state to mandate videotaping of homicide interrogations During his 2004 general election campaign for U S Senate police representatives credited Obama for his active engagement with police organizations in enacting death penalty reforms Obama resigned from the Illinois Senate in November 2004 following his election to the U S Senate U S Senate campaign In May 2002 Obama commissioned a poll to assess his prospects in a 2004 U S Senate race he created a campaign committee began raising funds and lined up political media consultant David Axelrod by August 2002 Obama formally announced his candidacy in January 2003 Obama was an early opponent of the George W Bush administration s 2003 invasion of Iraq On October 2 2002 the day President Bush and Congress agreed on the joint resolution authorizing the Iraq War Obama addressed the first high-profile Chicago anti-Iraq War rally and spoke out against the war He addressed another anti-war rally in March 2003 and told the crowd that it s not too late to stop the war Decisions by Republican incumbent Peter Fitzgerald and his Democratic predecessor Carol Moseley Braun to not participate in the election resulted in wide-open Democratic and Republican primary contests involving fifteen candidates In the March 2004 primary election Obama won in an unexpected landslide—which overnight made him a rising star within the national Democratic Party started speculation about a presidential future and led to the reissue of his memoir Dreams from My Father In July 2004 Obama delivered the keynote address at the 2004 Democratic National Convention seen by 9 1 million viewers His speech was well received and elevated his status within the Democratic Party Obama s expected opponent in the general election Republican primary winner Jack Ryan withdrew from the race in June 2004 Six weeks later Alan Keyes accepted the Republican nomination to replace Ryan In the November 2004 general election Obama won with 70 percent of the vote U S Senator 2005–2008 Obama was sworn in as a senator on January 3 2005 becoming the only Senate member of the Congressional Black Caucus CQ Weekly characterized him as a loyal Democrat based on analysis of all Senate votes in 2005–2007 Obama announced on November 13 2008 that he would resign his Senate seat on November 16 2008 before the start of the lame-duck session to focus on his transition period for the presidency Legislation Obama cosponsored the Secure America and Orderly Immigration Act He introduced two initiatives that bore his name Lugar–Obama which expanded the Nunn–Lugar Cooperative Threat Reduction concept to conventional weapons and the Federal Funding Accountability and Transparency Act of 2006 which authorized the establishment of USAspending gov a web search engine on federal spending On June 3 2008 Senator Obama—along with Senators Tom Carper Tom Coburn and John McCain —introduced follow-up legislation Strengthening Transparency and Accountability in Federal Spending Act of 2008 Obama sponsored legislation that would have required nuclear plant owners to notify state and local authorities of radioactive leaks but the bill failed to pass in the full Senate after being heavily modified in committee Regarding tort reform Obama voted for the Class Action Fairness Act of 2005 and the FISA Amendments Act of 2008 which grants immunity from civil liability to telecommunications companies complicit with NSA warrantless wiretapping operations In December 2006 President Bush signed into law the Democratic Republic of the Congo Relief Security and Democracy Promotion Act marking the first federal legislation to be enacted with Obama as its primary sponsor In January 2007 Obama and Senator Feingold introduced a corporate jet provision to the Honest Leadership and Open Government Act which was signed into law in September 2007 Obama also introduced Deceptive Practices and Voter Intimidation Prevention Act a bill to criminalize deceptive practices in federal elections and the Iraq War De-Escalation Act of 2007 neither of which was signed into law Later in 2007 Obama sponsored an amendment to the Defense Authorization Act to add safeguards for personality-disorder military discharges This amendment passed the full Senate in the spring of 2008 He sponsored the Iran Sanctions Enabling Act supporting divestment of state pension funds from Iran s oil and gas industry which has not passed committee and co-sponsored legislation to reduce risks of nuclear terrorism Obama also sponsored a Senate amendment to the State Children s Health Insurance Program providing one year of job protection for family members caring for soldiers with combat-related injuries Committees Obama held assignments on the Senate Committees for Foreign Relations Environment and Public Works and Veterans Affairs through December 2006 In January 2007 he left the Environment and Public Works committee and took additional assignments with Health Education Labor and Pensions and Homeland Security and Governmental Affairs He also became Chairman of the Senate s subcommittee on European Affairs As a member of the Senate Foreign Relations Committee Obama made official trips to Eastern Europe the Middle East Central Asia and Africa He met with Mahmoud Abbas before Abbas became President of the Palestinian National Authority and gave a speech at the University of Nairobi in which he condemned corruption within the Kenyan government Presidential campaigns 2008 presidential campaign On February 10 2007 Obama announced his candidacy for President of the United States in front of the Old State Capitol building in Springfield Illinois The choice of the announcement site was viewed as symbolic because it was also where Abraham Lincoln delivered his historic House Divided speech in 1858 Obama emphasized issues of rapidly ending the Iraq War increasing energy independence and reforming the health care system in a campaign that projected themes of hope and change A large number of candidates entered the Democratic Party presidential primaries The field narrowed to a duel between Obama and Senator Hillary Rodham Clinton after early contests with the race remaining close throughout the primary process but with Obama gaining a steady lead in pledged delegates due to better long-range planning superior fundraising dominant organizing in caucus states and better exploitation of delegate allocation rules On June 7 2008 Clinton ended her campaign and endorsed Obama On August 23 Obama announced his selection of Delaware Senator Joe Biden as his vice presidential running mate Obama selected Biden from a field speculated to include former Indiana Governor and Senator Evan Bayh and Virginia Governor Tim Kaine At the Democratic National Convention in Denver Colorado Hillary Clinton called for her supporters to endorse Obama and she and Bill Clinton gave convention speeches in his support Obama delivered his acceptance speech not at the center where the Democratic National Convention was held but at Invesco Field at Mile High to a crowd of over 75 000 the speech was viewed by over 38 million people worldwide During both the primary process and the general election Obama s campaign set numerous fundraising records particularly in the quantity of small donations On June 19 2008 Obama became the first major-party presidential candidate to turn down public financing in the general election since the system was created in 1976 John McCain was nominated as the Republican candidate and the two engaged in three presidential debates in September and October 2008 On November 4 Obama won the presidency with 365 electoral votes to 173 received by McCain Obama won 52 9 of the popular vote to McCain s 45 7 He became the first African American to be elected president Obama delivered his victory speech before hundreds of thousands of supporters in Chicago s Grant Park 2012 presidential campaign On April 4 2011 Obama announced his re-election campaign for 2012 in a video titled It Begins with Us that he posted on his website and filed election papers with the Federal Election Commission As the incumbent president he ran virtually unopposed in the Democratic Party presidential primaries and on April 3 2012 Obama had secured the 2778 convention delegates needed to win the Democratic nomination At the Democratic National Convention in Charlotte North Carolina former President Bill Clinton formally nominated Obama and Joe Biden as the Democratic Party candidates for president and vice president in the general election in which their main opponents were Republicans Mitt Romney the former governor of Massachusetts and Representative Paul Ryan of Wisconsin On November 6 2012 Obama won 332 electoral votes exceeding the 270 required for him to be re-elected as president With more than 51 of the popular vote Obama became the first Democratic president since Franklin D Roosevelt to twice win the majority of the popular vote President Obama addressed supporters and volunteers at Chicago s McCormick Place after his reelection and said Tonight you voted for action not politics as usual You elected us to focus on your jobs not ours And in the coming weeks and months I am looking forward to reaching out and working with leaders of both parties Presidency First days The inauguration of Barack Obama as the 44th President took place on January 20 2009 In his first few days in office Obama issued executive orders and presidential memoranda directing the U S military to develop plans to withdraw troops from Iraq He ordered the closing of the Guantanamo Bay detention camp but Congress prevented the closure by refusing to appropriate the required funds and preventing moving any Guantanamo detainee into the U S or to other countries Obama reduced the secrecy given to presidential records He also revoked President George W Bush s restoration of President Ronald Reagan s Mexico City Policy prohibiting federal aid to international family planning organizations that perform or provide counseling about abortion Domestic policy The first bill signed into law by Obama was the Lilly Ledbetter Fair Pay Act of 2009 relaxing the statute of limitations for equal-pay lawsuits Five days later he signed the reauthorization of the State Children s Health Insurance Program SCHIP to cover an additional 4 million uninsured children In March 2009 Obama reversed a Bush-era policy which had limited funding of embryonic stem cell research and pledged to develop strict guidelines on the research Obama appointed two women to serve on the Supreme Court in the first two years of his Presidency Sonia Sotomayor nominated by Obama on May 26 2009 to replace retiring Associate Justice David Souter was confirmed on August 6 2009 becoming the first Hispanic Supreme Court Justice Elena Kagan nominated by Obama on May 10 2010 to replace retiring Associate Justice John Paul Stevens was confirmed on August 5 2010 bringing the number of women sitting simultaneously on the Court to three for the first time in American history On September 30 2009 the Obama administration proposed new regulations on power plants factories and oil refineries in an attempt to limit greenhouse gas emissions and to curb global warming On October 8 2009 Obama signed the Matthew Shepard and James Byrd Jr Hate Crimes Prevention Act a measure that expands the 1969 United States federal hate-crime law to include crimes motivated by a victim s actual or perceived gender sexual orientation gender identity or disability On March 30 2010 Obama signed the Health Care and Education Reconciliation Act a reconciliation bill which ends the process of the federal government giving subsidies to private banks to give out federally insured loans increases the Pell Grant scholarship award and makes changes to the Patient Protection and Affordable Care Act In a major space policy speech in April 2010 Obama announced a planned change in direction at NASA the U S space agency He ended plans for a return of human spaceflight to the moon and development of the Ares I rocket Ares V rocket and Constellation program in favor of funding Earth science projects a new rocket type and research and development for an eventual manned mission to Mars and ongoing missions to the International Space Station On December 22 2010 Obama signed the Don t Ask Don t Tell Repeal Act of 2010 fulfilling a key promise made in the 2008 presidential campaign to end the Don t ask don t tell policy of 1993 that had prevented gay and lesbian people from serving openly in the United States Armed Forces President Obama s 2011 State of the Union Address focused on themes of education and innovation stressing the importance of innovation economics to make the United States more competitive globally He spoke of a five-year freeze in domestic spending eliminating tax breaks for oil companies and reversing tax cuts for the wealthiest Americans banning congressional earmarks and reducing healthcare costs He promised that the United States would have one million electric vehicles on the road by 2015 and would be 80 reliant on clean electricity As a candidate for the Illinois state senate Obama had said in 1996 that he favored legalizing same-sex marriage but by the time of his run for the U S senate in 2004 he said that while he supported civil unions and domestic partnerships for same-sex partners for strategic reasons he opposed same-sex marriages On May 9 2012 shortly after the official launch of his campaign for re-election as president Obama said his views had evolved and he publicly affirmed his personal support for the legalization of same-sex marriage becoming the first sitting U S president to do so During his second inaugural address on January 21 2013 Obama called for full equality for gay Americans Our journey is not complete until our gay brothers and sisters are treated like anyone else under the law — for if we are truly created equal then surely the love we commit to one another must be equal as well This was a historic moment being the first time that a president mentioned gay rights or the word gay in an inaugural address In 2013 the Obama administration filed briefs which urged the Supreme Court to rule in favor of same-sex couples in the cases of Hollingsworth v Perry regarding same-sex marriage and United States v Windsor regarding the Defense of Marriage Act Economic policy On February 17 2009 Obama signed the American Recovery and Reinvestment Act of 2009 a 787 billion economic stimulus package aimed at helping the economy recover from the deepening worldwide recession The act includes increased federal spending for health care infrastructure education various tax breaks and incentives and direct assistance to individuals which is being distributed over the course of several years In March Obama s Treasury Secretary Timothy Geithner took further steps to manage the financial crisis including introducing the Public-Private Investment Program for Legacy Assets which contains provisions for buying up to two trillion dollars in depreciated real estate assets Obama intervened in the troubled automotive industry in March 2009 renewing loans for General Motors and Chrysler to continue operations while reorganizing Over the following months the White House set terms for both firms bankruptcies including the sale of Chrysler to Italian automaker Fiat and a reorganization of GM giving the U S government a temporary 60 equity stake in the company with the Canadian government taking a 12 stake In June 2009 dissatisfied with the pace of economic stimulus Obama called on his cabinet to accelerate the investment He signed into law the Car Allowance Rebate System known colloquially as Cash for Clunkers that temporarily boosted the economy Although spending and loan guarantees from the Federal Reserve and the Treasury Department authorized by the Bush and Obama administrations totaled about 11 5 trillion only 3 trillion had been spent by the end of November 2009 However Obama and the Congressional Budget Office predicted that the 2010 budget deficit will be 1 5 trillion or 10 6 of the nation s gross domestic product GDP compared to the 2009 deficit of 1 4 trillion or 9 9 of GDP For 2011 the administration predicted the deficit will slightly shrink to 1 34 trillion while the 10-year deficit will increase to 8 53 trillion or 90 of GDP The most recent increase in the U S debt ceiling to 16 4 trillion was signed into law on January 26 2012 On August 2 2011 after a lengthy congressional debate over whether to raise the nation s debt limit Obama signed the bipartisan Budget Control Act of 2011 The legislation enforces limits on discretionary spending until 2021 establishes a procedure to increase the debt limit creates a Congressional Joint Select Committee on Deficit Reduction to propose further deficit reduction with a stated goal of achieving at least 1 5 trillion in budgetary savings over 10 years and establishes automatic procedures for reducing spending by as much as 1 2 trillion if legislation originating with the new joint select committee does not achieve such savings By passing the legislation Congress was able to prevent a U S government default on its obligations As it did throughout 2008 the unemployment rate rose in 2009 reaching a peak in October at 10 0 and averaging 10 0 in the fourth quarter Following a decrease to 9 7 in the first quarter of 2010 the unemployment rate fell to 9 6 in the second quarter where it remained for the rest of the year Between February and December 2010 employment rose by 0 8 which was less than the average of 1 9 experienced during comparable periods in the past four employment recoveries By November 2012 the unemployment rate fell to 7 7 decreasing to 7 3 in the second quarter of 2013 GDP growth returned in the third quarter of 2009 expanding at a rate of 1 6 followed by a 5 0 increase in the fourth quarter Growth continued in 2010 posting an increase of 3 7 in the first quarter with lesser gains throughout the rest of the year In July 2010 the Federal Reserve expressed that although economic activity continued to increase its pace had slowed and chairman Ben Bernanke stated that the economic outlook was unusually uncertain Overall the economy expanded at a rate of 2 9 in 2010 The Congressional Budget Office and a broad range of economists credit Obama s stimulus plan for economic growth The CBO released a report stating that the stimulus bill increased employment by 1–2 1 million while conceding that It is impossible to determine how many of the reported jobs would have existed in the absence of the stimulus package Although an April 2010 survey of members of the National Association for Business Economics showed an increase in job creation over a similar January survey for the first time in two years 73 of 68 respondents believed that the stimulus bill has had no impact on employment Within a month of the 2010 midterm elections Obama announced a compromise deal with the Congressional Republican leadership that included a temporary two-year extension of the 2001 and 2003 income tax rates a one-year payroll tax reduction continuation of unemployment benefits and a new rate and exemption amount for estate taxes The compromise overcame opposition from some in both parties and the resulting 858 billion Tax Relief Unemployment Insurance Reauthorization and Job Creation Act of 2010 passed with bipartisan majorities in both houses of Congress before Obama signed it on December 17 2010 In December 2013 Obama declared that growing income inequality is a defining challenge of our time and called on Congress to bolster the safety net and raise wages This came on the heels of the nationwide strikes of fast-food workers and Pope Francis criticism of inequality and trickle-down economics Environmental policy In July 2013 Obama expressed reservations and stated he would reject the Keystone XL pipeline if it increased carbon pollution or greenhouse emmisions Obama s advisers called for a halt to petroleum exploration in the Arctic in January 2013 Gulf of Mexico oil spill On April 20 2010 an explosion destroyed an offshore drilling rig at the Macondo Prospect in the Gulf of Mexico causing a major sustained oil leak The well s operator BP initiated a containment and cleanup plan and began drilling two relief wells intended to stop the flow Obama visited the Gulf on May 2 among visits by members of his cabinet and again on May 28 and June 4 On May 22 he announced a federal investigation and formed a bipartisan commission to recommend new safety standards after a review by Secretary of the Interior Ken Salazar and concurrent Congressional hearings On May 27 he announced a six-month moratorium on new deepwater drilling permits and leases pending regulatory review As multiple efforts by BP failed some in the media and public expressed confusion and criticism over various aspects of the incident and stated a desire for more involvement by Obama and the federal government Health care reform Obama called for Congress to pass legislation reforming health care in the United States a key campaign promise and a top legislative goal He proposed an expansion of health insurance coverage to cover the uninsured to cap premium increases and to allow people to retain their coverage when they leave or change jobs His proposal was to spend 900 billion over 10 years and include a government insurance plan also known as the public option to compete with the corporate insurance sector as a main component to lowering costs and improving quality of health care It would also make it illegal for insurers to drop sick people or deny them coverage for pre-existing conditions and require every American to carry health coverage The plan also includes medical spending cuts and taxes on insurance companies that offer expensive plans On July 14 2009 House Democratic leaders introduced a 1 017-page plan for overhauling the U S health care system which Obama wanted Congress to approve by the end of 2009 After much public debate during the Congressional summer recess of 2009 Obama delivered a speech to a joint session of Congress on September 9 where he addressed concerns over the proposals In March 2009 Obama lifted a ban on using federal funds for stem cell research On November 7 2009 a health care bill featuring the public option was passed in the House On December 24 2009 the Senate passed its own bill—without a public option—on a party-line vote of 60–39 On March 21 2010 the Patient Protection and Affordable Care Act passed by the Senate in December was passed in the House by a vote of 219 to 212 Obama signed the bill into law on March 23 2010 The Patient Protection and Affordable Care Act includes health-related provisions to take effect over four years including expanding Medicaid eligibility for people making up to 133 of the federal poverty level FPL starting in 2014 subsidizing insurance premiums for people making up to 400 of the FPL 88 000 for family of four in 2010 so their maximum out-of-pocket payment for annual premiums will be from 2 to 9 5 of income providing incentives for businesses to provide health care benefits prohibiting denial of coverage and denial of claims based on pre-existing conditions establishing health insurance exchanges prohibiting annual coverage caps and support for medical research According to White House and Congressional Budget Office figures the maximum share of income that enrollees would have to pay would vary depending on their income relative to the federal poverty level The costs of these provisions are offset by taxes fees and cost-saving measures such as new Medicare taxes for those in high-income brackets taxes on indoor tanning cuts to the Medicare Advantage program in favor of traditional Medicare and fees on medical devices and pharmaceutical companies there is also a tax penalty for those who do not obtain health insurance unless they are exempt due to low income or other reasons In March 2010 the Congressional Budget Office estimated that the net effect of both laws will be a reduction in the federal deficit by 143 billion over the first decade The law faced several legal challenges primarily based on the argument that an individual mandate requiring Americans to buy health insurance was unconstitutional On June 28 2012 the Supreme Court ruled by a 5–4 vote in National Federation of Independent Business v Sebelius that the Commerce Clause does not allow the government to require people to buy health insurance but the mandate was constitutional under the US Congress s taxing authority Energy policy needed for energy policy Prior content moved to Environmental Policy Gun control On January 16 2013 one month after the Sandy Hook Elementary School shooting President Obama signed 23 executive orders and outlined a series of sweeping proposals regarding gun control He urged Congress to reintroduce an expired ban on military-style assault weapons such as those used in several recent mass shootings impose limits on ammunition magazines to 10 rounds introduce background checks on all gun sales pass a ban on possession and sale of armor-piercing bullets introduce harsher penalties for gun-traffickers especially unlicensed dealers who buy arms for criminals and approving the appointment of the head of the federal Bureau of Alcohol Tobacco Firearms and Explosives for the first time since 2006 2010 midterm elections Obama called the November 2 2010 election where the Democratic Party lost 63 seats in and control of the House of Representatives humbling and a shellacking He said that the results came because not enough Americans had felt the effects of the economic recovery Foreign policy In February and March 2009 Vice President Joe Biden and Secretary of State Hillary Rodham Clinton made separate overseas trips to announce a new era in U S foreign relations with Russia and Europe using the terms break and reset to signal major changes from the policies of the preceding administration Obama attempted to reach out to Arab leaders by granting his first interview to an Arab cable TV network Al Arabiya On March 19 Obama continued his outreach to the Muslim world releasing a New Year s video message to the people and government of Iran In April Obama gave a speech in Ankara Turkey which was well received by many Arab governments On June 4 2009 Obama delivered a speech at Cairo University in Egypt calling for A New Beginning in relations between the Islamic world and the United States and promoting Middle East peace On June 26 2009 in response to the Iranian government s actions towards protesters following Iran s 2009 presidential election Obama said The violence perpetrated against them is outrageous We see it and we condemn it On July 7 while in Moscow he responded to a Vice President Biden comment on a possible Israeli military strike on Iran by saying We have said directly to the Israelis that it is important to try and resolve this in an international setting in a way that does not create major conflict in the Middle East On September 24 2009 Obama became the first sitting U S president to preside over a meeting of the United Nations Security Council In March 2010 Obama took a public stance against plans by the government of Israeli Prime Minister Benjamin Netanyahu to continue building Jewish housing projects in predominantly Arab neighborhoods of East Jerusalem During the same month an agreement was reached with the administration of Russian President Dmitry Medvedev to replace the 1991 Strategic Arms Reduction Treaty with a new pact reducing the number of long-range nuclear weapons in the arsenals of both countries by about one-third Obama and Medvedev signed the New START treaty in April 2010 and the U S Senate ratified it in December 2010 On December 6 2011 he instructed agencies to consider LGBT rights when issuing financial aid to foreign countries Obama criticized Russia s law discriminating against gays joining other western leaders in the boycott of the 2014 Winter Olympics in Russia Notably his Secretary of State John Kerry declared the end of the Monroe Doctrine in November 2013 Iraqi withdrawal On February 27 2009 Obama announced that combat operations in Iraq would end within 18 months His remarks were made to a group of Marines preparing for deployment to Afghanistan Obama said Let me say this as plainly as I can by August 31 2010 our combat mission in Iraq will end The Obama administration scheduled the withdrawal of combat troops to be completed by August 2010 decreasing troop s levels from 142 000 while leaving a transitional force of about 50 000 in Iraq until the end of 2011 On August 19 2010 the last U S combat brigade exited Iraq Remaining troops transitioned from combat operations to counter-terrorism and the training equipping and advising of Iraqi security forces On August 31 2010 Obama announced that the United States combat mission in Iraq was over On October 21 2011 President Obama announced that all U S troops would leave Iraq in time to be home for the holidays War in Afghanistan Early in his presidency Obama moved to bolster U S troop strength in Afghanistan He announced an increase to U S troop levels of 17 000 in February 2009 to stabilize a deteriorating situation in Afghanistan an area he said had not received the strategic attention direction and resources it urgently requires He replaced the military commander in Afghanistan General David D McKiernan with former Special Forces commander Lt Gen Stanley A McChrystal in May 2009 indicating that McChrystal s Special Forces experience would facilitate the use of counterinsurgency tactics in the war On December 1 2009 Obama announced the deployment of an additional 30 000 military personnel to Afghanistan and proposed to begin troop withdrawals 18 months from that date this took place in July 2011 David Petraeus replaced McChrystal in June 2010 after McChrystal s staff criticized White House personnel in a magazine article In February 2013 Obama said the U S military would reduce the troop level in Afghanistan from 68 000 to 34 000 US troops by February 2014 Israel Obama referred to the bond between the United States and Israel as unbreakable During the initial years of the Obama administration the U S increased military cooperation with Israel including increased military aid re-establishment of the U S -Israeli Joint Political Military Group and the Defense Policy Advisory Group and an increase in visits among high-level military officials of both countries The Obama administration asked Congress to allocate money toward funding the Iron Dome program in response to the waves of Palestinian rocket attacks on Israel In 2011 the United States vetoed a Security Council resolution condemning Israeli settlements with the United States being the only nation to do so Obama supports the two-state solution to the Arab–Israeli conflict based on the 1967 borders with land swaps In 2013 one journalist reported that in Obama s view with each new settlement announcement Netanyahu is moving his country down a path toward near-total isolation War in Libya In March 2011 as forces loyal to Muammar Gaddafi advanced on rebels across Libya calls for a no-fly zone came from around the world including Europe the Arab League and a resolution passed unanimously by the U S Senate In response to the unanimous passage of United Nations Security Council Resolution 1973 on March 17 Gaddafi—who had previously vowed to show no mercy to the rebels of Benghazi—announced an immediate cessation of military activities yet reports came in that his forces continued shelling Misrata The next day on Obama s orders the U S military took part in air strikes to destroy the Libyan government s air defense capabilities to protect civilians and enforce a no-fly-zone including the use of Tomahawk missiles B-2 Spirits and fighter jets Six days later on March 25 by unanimous vote of all of its 28 members NATO took over leadership of the effort dubbed Operation Unified Protector Some Representatives questioned whether Obama had the constitutional authority to order military action in addition to questioning its cost structure and aftermath Osama bin Laden Starting with information received in July 2010 intelligence developed by the CIA over the next several months determined what they believed to be the location of Osama bin Laden in a large compound in Abbottabad Pakistan a suburban area 35 miles from Islamabad CIA head Leon Panetta reported this intelligence to President Obama in March 2011 Meeting with his national security advisers over the course of the next six weeks Obama rejected a plan to bomb the compound and authorized a surgical raid to be conducted by United States Navy SEALs The operation took place on May 1 2011 resulting in the death of bin Laden and the seizure of papers computer drives and disks from the compound DNA testing identified bin Laden s body which was buried at sea several hours later Within minutes of the President s announcement from Washington DC late in the evening on May 1 there were spontaneous celebrations around the country as crowds gathered outside the White House and at New York City s Ground Zero and Times Square Reaction to the announcement was positive across party lines including from former presidents Bill Clinton and George W Bush and from many countries around the world Cultural and political image Obama s family history upbringing and Ivy League education differ markedly from those of African-American politicians who launched their careers in the 1960s through participation in the civil rights movement Expressing puzzlement over questions about whether he is black enough Obama told an August 2007 meeting of the National Association of Black Journalists that we re still locked in this notion that if you appeal to white folks then there must be something wrong Obama acknowledged his youthful image in an October 2007 campaign speech saying I wouldn t be here if time and again the torch had not been passed to a new generation Obama is frequently referred to as an exceptional orator During his pre-inauguration transition period and continuing into his presidency Obama has delivered a series of weekly Internet video addresses According to the Gallup Organization Obama began his presidency with a 68 approval rating before gradually declining for the rest of the year and eventually bottoming out at 41 in August 2010 a trend similar to Ronald Reagan s and Bill Clinton s first years in office He experienced a small poll bounce shortly after the death of Osama bin Laden which lasted until around June 2011 when his approval numbers dropped back to where they were prior to the operation His approval ratings rebounded around the same time as his re-election in 2012 with polls showing an average job approval of 52 shortly after his second inauguration Polls show strong support for Obama in other countries and before being elected President he met with prominent foreign figures including former British Prime Minister Tony Blair Italy s Democratic Party leader and Mayor of Rome Walter Veltroni and French President Nicolas Sarkozy In a February 2009 poll conducted in Western Europe and the U S by Harris Interactive for France 24 and the International Herald Tribune Obama was rated as the most respected world leader as well as the most powerful In a similar poll conducted by Harris in May 2009 Obama was rated as the most popular world leader as well as the one figure most people would pin their hopes on for pulling the world out of the economic downturn Obama won Best Spoken Word Album Grammy Awards for abridged audiobook versions of Dreams from My Father in February 2006 and for The Audacity of Hope in February 2008 His concession speech after the New Hampshire primary was set to music by independent artists as the music video Yes We Can which was viewed 10 million times on YouTube in its first month and received a Daytime Emmy Award In December 2008 Time magazine named Obama as its Person of the Year for his historic candidacy and election which it described as the steady march of seemingly impossible accomplishments He was again named Person of the Year in 2012 On October 9 2009 the Norwegian Nobel Committee announced that Obama had won the 2009 Nobel Peace Prize for his extraordinary efforts to strengthen international diplomacy and cooperation between peoples Obama accepted this award in Oslo Norway on December 10 2009 with deep gratitude and great humility The award drew a mixture of praise and criticism from world leaders and media figures Obama is the fourth U S president to be awarded the Nobel Peace Prize and the third to become a Nobel laureate while in office According to November 2013 polls Obama s average approval rating fell to 41 thus setting a new low for the president Family and personal life In a 2006 interview Obama highlighted the diversity of his extended family It s like a little mini-United Nations he said I ve got relatives who look like Bernie Mac and I ve got relatives who look like Margaret Thatcher Obama has a half-sister with whom he was raised Maya Soetoro-Ng the daughter of his mother and her Indonesian second husband and seven half-siblings from his Kenyan father s family – six of them living Obama s mother was survived by her Kansas-born mother Madelyn Dunham until her death on November 2 2008 two days before his election to the Presidency Obama also has roots in Ireland he met with his Irish cousins in Moneygall in May 2011 In Dreams from My Father Obama ties his mother s family history to possible Native American ancestors and distant relatives of Jefferson Davis President of the Confederate States of America during the American Civil War In his youth Obama was known as Barry among his family and friends but asked to be addressed with his given name during his college years Besides his native English Obama speaks some basic Indonesian having learned the language during his four childhood years in Jakarta He plays basketball a sport he participated in as a member of his high school s varsity team he is left-handed Obama is a supporter of the Chicago White Sox and he threw out the first pitch at the 2005 ALCS when he was still a senator In 2009 he threw out the ceremonial first pitch at the all star game while wearing a White Sox jacket He is also primarily a Chicago Bears football fan in the NFL but in his childhood and adolescence was a fan of the Pittsburgh Steelers and rooted for them ahead of their victory in Super Bowl XLIII 12 days after he took office as President In 2011 Obama invited the 1985 Chicago Bears to the White House the team had not visited the White House after their Super Bowl win in 1986 due to the Space Shuttle Challenger disaster In June 1989 Obama met Michelle Robinson when he was employed as a summer associate at the Chicago law firm of Sidley Austin Assigned for three months as Obama s adviser at the firm Robinson joined him at several group social functions but declined his initial requests to date They began dating later that summer became engaged in 1991 and were married on October 3 1992 The couple s first daughter Malia Ann was born on July 4 1998 followed by a second daughter Natasha Sasha on June 10 2001 The Obama daughters attended the private University of Chicago Laboratory Schools When they moved to Washington D C in January 2009 the girls started at the private Sidwell Friends School The Obamas have a Portuguese Water Dog named Bo a gift from Senator Ted Kennedy Applying the proceeds of a book deal the family moved in 2005 from a Hyde Park Chicago condominium to a 1 6 million house in neighboring Kenwood Chicago The purchase of an adjacent lot—and sale of part of it to Obama by the wife of developer campaign donor and friend Tony Rezko —attracted media attention because of Rezko s subsequent indictment and conviction on political corruption charges that were unrelated to Obama In December 2007 Money estimated the Obama family s net worth at 1 3 million See also ref Their 2009 tax return showed a household income of 5 5 million—up from about 4 2 million in 2007 and 1 6 million in 2005—mostly from sales of his books On his 2010 income of 1 7 million he gave 14 to non-profit organizations including 131 000 to Fisher House Foundation a charity assisting wounded veterans families allowing them to reside near where the veteran is receiving medical treatments As per his 2012 financial disclosure Obama may be worth as much as 10 million Obama tried to quit smoking several times sometimes using nicotine replacement therapy and in early 2010 Michelle Obama said that he had successfully quit smoking Religious views Obama is a Christian whose religious views developed in his adult life He wrote in The Audacity of Hope that he was not raised in a religious household He described his mother raised by non-religious parents whom Obama has specified elsewhere as non-practicing Methodists and Baptists as being detached from religion yet in many ways the most spiritually awakened person that I have ever known He described his father as a confirmed atheist by the time his parents met and his stepfather as a man who saw religion as not particularly useful Obama explained how through working with black churches as a community organizer while in his twenties he came to understand the power of the African-American religious tradition to spur social change In an interview with the evangelical periodical Christianity Today Obama stated I am a Christian and I am a devout Christian I believe in the redemptive death and resurrection of Jesus Christ I believe that that faith gives me a path to be cleansed of sin and have eternal life On September 27 2010 Obama released a statement commenting on his religious views saying I m a Christian by choice My family didn t—frankly they weren t folks who went to church every week And my mother was one of the most spiritual people I knew but she didn t raise me in the church So I came to my Christian faith later in life and it was because the precepts of Jesus Christ spoke to me in terms of the kind of life that I would want to lead—being my brothers and sisters keeper treating others as they would treat me Obama met Trinity United Church of Christ pastor Rev Jeremiah Wright in October 1987 and became a member of Trinity in 1992 He resigned from Trinity in May 2008 during his first presidential campaign after controversial statements by Wright were publicized After a prolonged effort to find a church to attend regularly in Washington Obama announced in June 2009 that his primary place of worship would be the Evergreen Chapel at Camp David External links Official Other'\n\nent_lines_4EX['Germany'] = '11867\tGermany\tGermany Germany officially the Federal Republic of Germany is a federal parliamentary republic in western - central Europe The country consists of 16 states and its capital and largest city is Berlin Germany covers an area of and has a largely temperate seasonal climate With 80 6 million inhabitants it is the most populous member state in the European Union Germany is the major economic and political power of the European continent and a historic leader in many cultural theoretical and technical fields Various Germanic tribes occupied what is now northern Germany and southern Scandinavia since classical antiquity A region named Germania was documented by the Romans before AD 100 During the Migration Period that coincided with the decline of the Roman Empire the Germanic tribes expanded southward and established kingdoms throughout much of Europe Beginning in the 10th century German territories formed a central part of the Holy Roman Empire ref During the 16th century northern German regions became the centre of the Protestant Reformation Occupied during the Napoleonic Wars the rise of Pan-Germanism inside the German Confederation resulted in the unification of most of the German states in 1871 into the German Empire which was dominated by Prussia After the German Revolution of 1918–1919 and the subsequent military surrender in World War I the Empire was replaced by the parliamentary Weimar Republic in 1918 with some of its territory partitioned in the Treaty of Versailles Despite its lead in many scientific and cultural fields at this time Germany nonetheless endured significant economic and political instability which intensified during the Great Depression and contributed to the establishment of the Third Reich in 1933 The subsequent rise of fascism led to World War II After 1945 Germany was divided by allied occupation and evolved into two states East Germany and West Germany In 1990 the country was reunified Germany has the world s fourth-largest economy by nominal GDP and the fifth-largest by purchasing power parity As a global leader in several industrial and technological sectors it is the second-largest exporter and third-largest importer of goods It is a developed country with a very high standard of living featuring comprehensive social security that includes the world s oldest universal health care system Known for its rich cultural and political history Germany has been the home of many influential philosophers music composers scientists and inventors Germany was a founding member of the European Community in 1957 which became the EU in 1993 It is part of the Schengen Area and has been a member of the eurozone since 1999 Germany is a great power in regional and global affairs and is a member of the United Nations NATO the G8 the G20 the OECD and the Council of Europe Etymology The English word Germany derives from the Latin Germania which came into use after Julius Caesar adopted it for the peoples east of the Rhine The German term Deutschland originally diutisciu land the German lands is derived from deutsch descended from Old High German diutisc popular i e belonging to the diot or diota people originally used to distinguish the language of the common people from Latin and its Romance descendants This in turn descends from Proto-Germanic þiudiskaz popular see also the Latinised form Theodiscus derived from þeudō descended from Proto-Indo-European tewtéh₂- people History Germanic tribes and Frankish Empire The Germanic tribes are thought to date from the Nordic Bronze Age or the Pre-Roman Iron Age From southern Scandinavia and north Germany they expanded south east and west from the 1st century BC coming into contact with the Celtic tribes of Gaul as well as Iranian Baltic and Slavic tribes in Central and Eastern Europe Under Augustus Rome began to invade Germania an area extending roughly from the Rhine to the Ural Mountains In AD 9 three Roman legions led by Publius Quinctilius Varus were defeated by the Cheruscan leader Arminius By AD 100 when Tacitus wrote Germania Germanic tribes had settled along the Rhine and the Danube the Limes Germanicus occupying most of the area of modern Germany Austria southern Bavaria and the western Rhineland however were Roman provinces In the 3rd century a number of large West Germanic tribes emerged Alemanni Franks Chatti Saxons Frisii Sicambri and Thuringii Around 260 the Germanic peoples broke into Roman-controlled lands After the invasion of the Huns in 375 and with the decline of Rome from 395 Germanic tribes moved further south-west Simultaneously several large tribes formed in what is now Germany and displaced the smaller Germanic tribes Large areas known since the Merovingian period as Austrasia were occupied by the Franks and Northern Germany was ruled by the Saxons and Slavs Holy Roman Empire On 25 December 800 the Frankish king Charlemagne was crowned emperor and founded the Carolingian Empire which was divided in 843 The Holy Roman Empire resulted from the eastern portion of this division Its territory stretched from the Eider River in the north to the Mediterranean coast in the south Under the reign of the Ottonian emperors 919–1024 several major duchies were consolidated and the German king Otto I was crowned Holy Roman Emperor of these regions in 962 In 996 Gregory V became the first German Pope appointed by his cousin Otto III whom he shortly after crowned Holy Roman Emperor The Holy Roman Empire absorbed northern Italy and Burgundy under the reign of the Salian emperors 1024–1125 although the emperors lost power through the Investiture Controversy Under the Hohenstaufen emperors 1138–1254 the German princes increased their influence further south and east into territories inhabited by Slavs preceding German settlement in these areas and further east Ostsiedlung Northern German towns grew prosperous as members of the Hanseatic League Starting with the Great Famine in 1315 then the Black Death of 1348–50 the population of Germany plummeted The edict of the Golden Bull in 1356 provided the basic constitution of the empire and codified the election of the emperor by seven prince-electors who ruled some of the most powerful principalities and archbishoprics Martin Luther publicised The Ninety-Five Theses in 1517 in Wittenberg challenging the beliefs of the Roman Catholic Church and initiating the Protestant Reformation A separate Lutheran church became the official religion in many German states after 1530 Religious conflict led to the Thirty Years War 1618–1648 which devastated German lands The population of the German states was reduced by about 30 The Peace of Westphalia 1648 ended religious warfare among the German states but the empire was de facto divided into numerous independent principalities In the 18th century the Holy Roman Empire consisted of approximately 1 800 such territories From 1740 onwards dualism between the Austrian Habsburg Monarchy and the Kingdom of Prussia dominated German history In 1806 the Imperium was overrun and dissolved as a result of the Napoleonic Wars German Confederation and Empire Following the fall of Napoleon the Congress of Vienna convened in 1814 and founded the German Confederation Deutscher Bund a loose league of 39 sovereign states Disagreement with restoration politics partly led to the rise of liberal movements followed by new measures of repression by Austrian statesman Metternich The Zollverein a tariff union furthered economic unity in the German states National and liberal ideals of the French Revolution gained increasing support among many especially young Germans The Hambach Festival in May 1832 was a main event in support of German unity freedom and democracy In the light of a series of revolutionary movements in Europe which established a republic in France intellectuals and commoners started the Revolutions of 1848 in the German states King Frederick William IV of Prussia was offered the title of Emperor but with a loss of power he rejected the crown and the proposed constitution leading to a temporary setback for the movement Conflict between King William I of Prussia and the increasingly liberal parliament erupted over military reforms in 1862 and the king appointed Otto von Bismarck the new Minister President of Prussia Bismarck successfully waged war on Denmark in 1864 Prussian victory in the Austro-Prussian War of 1866 enabled him to create the North German Confederation Norddeutscher Bund and to exclude Austria formerly the leading German state from the federation s affairs After the French defeat in the Franco-Prussian War the German Empire was proclaimed in 1871 in Versailles uniting all scattered parts of Germany except Austria or Lesser Germany With almost two-thirds of its territory and population Prussia was the dominating constituent of the new state the Hohenzollern King of Prussia ruled as its concurrent Emperor and Berlin became its capital In the period following the unification of Germany Bismarck s foreign policy as Chancellor of Germany under Emperor William I secured Germany s position as a great nation by forging alliances isolating France by diplomatic means and avoiding war As a result of the Berlin Conference in 1884 Germany claimed several colonies including German East Africa German South-West Africa Togo and Cameroon Under Wilhelm II however Germany like other European powers took an imperialistic course leading to friction with neighbouring countries Most alliances in which Germany had previously been involved were not renewed and new alliances excluded the country The assassination of Austria s crown prince on 28 June 1914 triggered World War I Germany as part of the Central Powers suffered defeat against the Allies in one of the bloodiest conflicts of all time An estimated two million German soldiers died in World War I The German Revolution broke out in November 1918 and Emperor Wilhelm II and all German ruling princes abdicated An armistice ended the war on 11 November and Germany was forced to sign the Treaty of Versailles in June 1919 The treaty was perceived in Germany as a humiliating continuation of the war and is often cited as an influence in the rise of Nazism Weimar Republic and Nazi Regime At the beginning of the German Revolution in November 1918 Germany was declared a republic However the struggle for power continued with radical-left Communists seizing power in Bavaria The revolution came to an end on 11 August 1919 when the democratic Weimar Constitution was signed by President Friedrich Ebert An era of increasing national confidence a very liberal cultural life and decade of economic prosperity followed - known as the Golden Twenties Suffering from the Great Depression of 1929 the harsh peace conditions dictated by the Treaty of Versailles and a long succession of unstable governments Germans increasingly lacked identification with the government in the early 1930s This was exacerbated by a widespread right-wing or stab-in-the-back legend which argued that Germany had lost World War I because of those who wanted to overthrow the government The Weimar government was accused of betraying Germany by signing the Versailles Treaty By 1932 the German Communist Party and the Nazi Party controlled the majority of Parliament fuelled by discontent with the Weimar government After a series of unsuccessful cabinets President Paul von Hindenburg appointed Adolf Hitler as Chancellor of Germany on 30 January 1933 On 27 February 1933 the Reichstag building went up in flames and a consequent emergency decree abrogated basic citizens rights An enabling act passed in parliament gave Hitler unrestricted legislative power Only the Social Democratic Party voted against it while Communist MPs had already been imprisoned Using his powers to crush any actual or potential resistance Hitler established a centralised totalitarian state within months Industry was revitalised with a focus on military rearmament In 1935 Germany reacquired control of the Saar and in 1936 military control of the Rhineland both of which had been lost in the Treaty of Versailles In 1938 Austria was annexed and in 1939 Czechoslovakia was brought under German control The invasion of Poland was prepared through the Molotov–Ribbentrop pact and Operation Himmler On 1 September 1939 the German Wehrmacht launched a blitzkrieg on Poland which was swiftly occupied by Germany and by the Soviet Red Army The UK and France declared war on Germany marking the beginning of World War II As the war progressed Germany and its allies quickly gained control of most of continental Europe and North Africa though plans to force the United Kingdom to an armistice or surrender failed On 22 June 1941 Germany broke the Molotov–Ribbentrop pact and invaded the Soviet Union Japan s attack on Pearl Harbor led Germany to declare war on the United States The Battle of Stalingrad forced the German army to retreat on the Eastern front In September 1943 Germany s ally Italy surrendered and German troops were forced to defend an additional front in Italy D-Day opened a Western front as Allied forces advanced towards German territory On 8 May 1945 the German armed forces surrendered after the Red Army occupied Berlin In what later became known as The Holocaust the Third Reich regime had enacted policies directly subjugating many dissidents and minorities Millions of people were murdered by the Nazis during the Holocaust including several million Jews Romani people Slavic people Soviet POWs people with mental and or physical disabilities Jehovah s Witnesses homosexuals and members of the political and religious opposition World War II was responsible for more than 40 million dead in Europe The war casualties for Germany are estimated at 5 3 million German soldiers millions of German civilians and losing the war resulted in large territorial losses the expulsion of about 15 million ethnic Germans from former eastern territories of Germany and other formerly occupied European countries mass rape of German women and the destruction of numerous major cities The Nuremberg trials of Nazi war criminals were held after World War II East and West Germany After the surrender of Germany the remaining German territory and Berlin were partitioned by the Allies into four military occupation zones Together these zones accepted more than 6 5 million of the ethnic Germans expelled from eastern areas The western sectors controlled by France the United Kingdom and the United States were merged on 23 May 1949 to form the Federal Republic of Germany Bundesrepublik Deutschland on 7 October 1949 the Soviet Zone became the German Democratic Republic Deutsche Demokratische Republik or DDR They were informally known as West Germany and East Germany East Germany selected East Berlin as its capital while West Germany chose Bonn as a provisional capital to emphasise its stance that the two-state solution was an artificial and temporary status quo West Germany established as a federal parliamentary republic with a social market economy was allied with the United States the UK and France Konrad Adenauer was elected the first Federal Chancellor Bundeskanzler of Germany in 1949 and remained in office until 1963 Under his and Ludwig Erhard s leadership the country enjoyed prolonged economic growth beginning in the early 1950s that became famous as the economic miracle West Germany joined NATO in 1955 and was a founding member of the European Economic Community in 1957 East Germany was an Eastern Bloc state under political and military control by the USSR via the latter s occupation forces and the Warsaw Pact Though East Germany claimed to be a democracy political power was exercised solely by leading members Politbüro of the communist-controlled Socialist Unity Party of Germany SED supported by the Stasi an immense secret service and a variety of sub-organisations controlling every aspect of society A Soviet-style command economy was set up the GDR later became a Comecon state While East German propaganda was based on the benefits of the GDR s social programmes and the alleged constant threat of a West German invasion many of its citizens looked to the West for freedom and prosperity The Berlin Wall built in 1961 to stop East Germans from escaping to West Germany became a symbol of the Cold War hence its fall in 1989 following democratic reforms in Poland and Hungary became a symbol of the Fall of Communism German Reunification and Die Wende Tensions between East and West Germany were reduced in the early 1970s by Chancellor Willy Brandt s In summer 1989 Hungary decided to dismantle the Iron Curtain and open the borders causing the emigration of thousands of East Germans to West Germany via Hungary This had devastating effects on the GDR where regular mass demonstrations received increasing support The East German authorities unexpectedly eased the border restrictions allowing East German citizens to travel to the West originally intended to help retain East Germany as a state the opening of the border actually led to an acceleration of the Wende reform process This culminated in the Two Plus Four Treaty a year later on 12 September 1990 under which the four occupying powers renounced their rights under the Instrument of Surrender and Germany regained full sovereignty This permitted German reunification on 3 October 1990 with the accession of the five re-established states of the former GDR new states or neue Länder German Reunification and the EU Based on the Berlin Bonn Act adopted on 10 March 1994 Berlin once again became the capital of the reunified Germany while Bonn obtained the unique status of a Bundesstadt federal city retaining some federal ministries The relocation of the government was completed in 1999 Since reunification Germany has taken a more active role in the European Union and NATO Germany sent a peacekeeping force to secure stability in the Balkans and sent a force of German troops to Afghanistan as part of a NATO effort to provide security in that country after the ousting of the Taliban These deployments were controversial since after the war Germany was bound by domestic law only to deploy troops for defence roles In 2005 Angela Merkel became the first female Chancellor of Germany as the leader of a grand coalition Germany hosted the 2007 G8 summit in Heiligendamm Mecklenburg In 2009 a liberal - conservative coalition under Merkel assumed leadership of the country Geography Germany is in Western and Central Europe with Denmark bordering to the north Poland and the Czech Republic to the east Austria and Switzerland to the south France and Luxembourg to the southwest and Belgium and the Netherlands to the northwest It lies mostly between latitudes 47° and 55° N the tip of Sylt is just north of 55° and longitudes 5° and 16° E The territory covers consisting of of land and of water It is the seventh largest country by area in Europe and the 62nd largest in the world Elevation ranges from the mountains of the Alps highest point the Zugspitze at in the south to the shores of the North Sea Nordsee in the northwest and the Baltic Sea Ostsee in the northeast The forested uplands of central Germany and the lowlands of northern Germany lowest point Wilstermarsch at below sea level are traversed by such major rivers as the Rhine Danube and Elbe Glaciers are found in the Alpine region but are experiencing deglaciation Significant natural resources are iron ore coal potash timber lignite uranium copper natural gas salt nickel arable land and water Climate Most of Germany has a temperate seasonal climate in which humid westerly winds predominate The country is situated in between the oceanic Western European and the continental Eastern European climate The climate is moderated by the North Atlantic Drift the northern extension of the Gulf Stream This warmer water affects the areas bordering the North Sea consequently in the northwest and the north the climate is oceanic Germany gets an average of precipitation per year Rainfall occurs year-round with no obligatory dry season Winters are mild and summers tend to be warm temperatures can exceed The east has a more continental climate winters can be very cold and summers very warm and longer dry periods can occur Central and southern Germany are transition regions which vary from moderately oceanic to continental In addition to the maritime and continental climates that predominate over most of the country the Alpine regions in the extreme south and to a lesser degree some areas of the Central German Uplands have a mountain climate characterised by lower temperatures and greater precipitation Biodiversity The territory of Germany can be subdivided into two ecoregions European-Mediterranean montane mixed forests and Northeast-Atlantic shelf marine the majority of Germany is covered by either arable land 34 or forest and woodland 30 1 only 13 4 of the area consists of permanent pastures 11 8 is covered by settlements and streets Plants and animals are those generally common to middle Europe Beeches oaks and other deciduous trees constitute one-third of the forests conifers are increasing as a result of reforestation Spruce and fir trees predominate in the upper mountains while pine and larch are found in sandy soil There are many species of ferns flowers fungi and mosses Wild animals include deer wild boar mouflon fox badger hare and small numbers of beavers The blue cornflower was once a German national symbol The 14 national parks in Germany include the Jasmund National Park the Vorpommern Lagoon Area National Park the Müritz National Park the Wadden Sea National Parks the Harz National Park the Hainich National Park the Saxon Switzerland National Park the Bavarian Forest National Park and the Berchtesgaden National Park In addition there are 14 Biosphere Reserves as well as 98 nature parks More than 400 registered zoos and animal parks operate in Germany which is believed to be the largest number in any country The Berlin Zoo opened in 1844 is the oldest zoo in Germany and presents the most comprehensive collection of species in the world Politics Germany is a federal parliamentary representative democratic republic The German political system operates under a framework laid out in the 1949 constitutional document known as the Grundgesetz Basic Law Amendments generally require a two-thirds majority of both chambers of parliament the fundamental principles of the constitution as expressed in the articles guaranteeing human dignity the separation of powers the federal structure and the rule of law are valid in perpetuity The president is the head of state and invested primarily with representative responsibilities and powers He is elected by the Bundesversammlung federal convention an institution consisting of the members of the Bundestag and an equal number of state delegates The second-highest official in the German order of precedence is the Bundestagspräsident President of the Bundestag who is elected by the Bundestag and responsible for overseeing the daily sessions of the body The third-highest official and the head of government is the Chancellor who is appointed by the Bundespräsident after being elected by the Bundestag The chancellor currently Angela Merkel is the head of government and exercises executive power similar to the role of a Prime Minister in other parliamentary democracies Federal legislative power is vested in the parliament consisting of the Bundestag Federal Diet and Bundesrat Federal Council which together form the legislative body The Bundestag is elected through direct elections by proportional representation mixed-member The members of the Bundesrat represent the governments of the sixteen federated states and are members of the state cabinets Since 1949 the party system has been dominated by the Christian Democratic Union and the Social Democratic Party of Germany So far every chancellor has been a member of one of these parties However the smaller liberal Free Democratic Party which had members in the Bundestag from 1949 to 2013 and the Alliance 90 The Greens which has had seats in parliament since 1983 have also played important roles Minor parties such as The Left Free Voters and the Pirate Party are represented in some state parliaments Law Germany has a civil law system based on Roman law with some references to Germanic law The Bundesverfassungsgericht Federal Constitutional Court is the German Supreme Court responsible for constitutional matters with power of judicial review Germany s supreme court system called Oberste Gerichtshöfe des Bundes is specialised for civil and criminal cases the highest court of appeal is the inquisitorial Federal Court of Justice and for other affairs the courts are the Federal Labour Court the Federal Social Court the Federal Finance Court and the Federal Administrative Court The Völkerstrafgesetzbuch regulates the consequences of crimes against humanity genocide and war crimes and gives German courts universal jurisdiction in some circumstances Criminal and private laws are codified on the national level in the Strafgesetzbuch and the Bürgerliches Gesetzbuch respectively The German penal system is aimed towards rehabilitation of the criminal and the protection of the general public Except for petty crimes which are tried before a single professional judge and serious political crimes all charges are tried before mixed tribunals on which lay judges sit side by side with professional judges Many of the fundamental matters of administrative law remain in the jurisdiction of the states though most states base their own laws in that area on the 1976 Verwaltungsverfahrensgesetz Administrative Proceedings Act covering important points of administrative law The Oberverwaltungsgerichte are the highest level of administrative jurisdiction concerning the state administrations unless the question of law concerns federal law or state law identical to federal law In such cases final appeal to the Federal Administrative Court is possible Constituent states Germany comprises sixteen states which are collectively referred to as Länder ref Each state has its own state constitution and is largely autonomous in regard to its internal organisation Because of differences in size and population the subdivisions of these states vary especially as between city states Stadtstaaten and states with larger territories Flächenländer For regional administrative purposes five states namely Baden-Württemberg Bavaria Hesse North Rhine-Westphalia and Saxony consist of a total of 22 Government Districts Regierungsbezirke Germany is divided into 403 districts Kreise at a municipal level these consist of 301 rural districts and 102 urban districts Foreign relations Germany has a network of 229 diplomatic missions abroad and maintains relations with more than 190 countries it is the largest contributor to the budget of the European Union providing 20 and the third largest contributor to the UN providing 8 Germany is a member of NATO the Organisation of Economic Co-operation and Development OECD the G8 the G20 the World Bank and the International Monetary Fund IMF It has played a leading role in the European Union since its inception and has maintained a strong alliance with France since the end of World War II Germany seeks to advance the creation of a more unified European political defence and security apparatus The development policy of the Federal Republic of Germany is an independent area of German foreign policy It is formulated by the Federal Ministry for Economic Cooperation and Development BMZ and carried out by the implementing organisations The German government sees development policy as a joint responsibility of the international community It is the world s third biggest aid donor after the United States and France During the Cold War Germany s partition by the Iron Curtain made it a symbol of East–West tensions and a political battleground in Europe However Willy Brandt s Ostpolitik was a key factor in the détente of the 1970s In 1999 Chancellor Gerhard Schröder s government defined a new basis for German foreign policy by taking part in the NATO decisions surrounding the Kosovo War and by sending German troops into combat for the first time since World War II The governments of Germany and the United States are close political allies The 1948 Marshall Plan and strong cultural ties have crafted a strong bond between the two countries although Schröder s vocal opposition to the Iraq War suggested the end of Atlanticism and a relative cooling of German-American relations The two countries are also economically interdependent 8 8 of German exports are US-bound and 6 6 of German imports originate from the US Military Germany s military the Bundeswehr is organised into Heer Army Marine Navy Luftwaffe Air Force Bundeswehr Joint Medical Service and Streitkräftebasis Joint Support Service branches The role of the Bundeswehr is described in the Constitution of Germany Art 87a as absolutely defensive only Its only active role before 1990 was the Katastropheneinsatz disaster control Within the Bundeswehr it helped after natural disasters both in Germany and abroad After 1990 the international situation changed from East-West confrontation to one of general uncertainty and instability Today after a ruling of the Federal Constitutional Court in 1994 the term defense has been defined to not only include protection of the borders of Germany but also crisis reaction and conflict prevention or more broadly as guarding the security of Germany anywhere in the world military spending was an estimated 1 3 of the country s GDP which is low in a ranking of all countries in absolute terms German military expenditure is the 9th highest in the world In peacetime the Bundeswehr is commanded by the Minister of Defence If Germany went to war which according to the constitution is allowed only for defensive purposes the Chancellor would become commander-in-chief of the Bundeswehr Until 2011 military service was compulsory for men at age 18 and conscripts served six-month tours of duty conscientious objectors could instead opt for an equal length of Zivildienst civilian service or a six-year commitment to voluntary emergency services like a fire department or the Red Cross On 1 July 2011 conscription was officially suspended and replaced with a voluntary service Since 2001 women may serve in all functions of service without restriction but they have not been subject to conscription There are presently some 17 500 women on active duty and a number of female reservists Economy Germany has a social market economy with a highly skilled labour force a large capital stock a low level of corruption and a high level of innovation It has the largest and most powerful national economy in Europe the fourth largest by nominal GDP in the world Field listing – GDP official exchange rate ref the fifth largest by PPP Field listing – GDP PPP exchange rate ref and was the biggest net contributor to the EU budget in 2011 The service sector contributes approximately 71 of the total GDP industry 28 and agriculture 1 The official average national unemployment rate in June 2013 was 6 6 However the official average national unemployment rate also includes people with a part-time job that are looking for a full-time job The unofficial average national unemployment rate in 2011 was 5 7 Germany is an advocate of closer European economic and political integration Its commercial policies are increasingly determined by agreements among European Union EU members and by EU legislation Germany introduced the common European currency the euro on 1 January 2002 Its monetary policy is set by the European Central Bank which is headquartered in Frankfurt Two decades after German reunification standards of living and per capita incomes remain significantly higher in the states of the former West Germany than in the former East The modernisation and integration of the eastern German economy is a long-term process scheduled to last until the year 2019 with annual transfers from west to east amounting to roughly 80 billion In January 2009 the German government approved a €50 billion economic stimulus plan to protect several sectors from a downturn and a subsequent rise in unemployment rates Of the world s 500 largest stock-market-listed companies measured by revenue in 2010 the Fortune Global 500 37 are headquartered in Germany 30 Germany-based companies are included in the DAX the German stock market index Well-known global brands are Mercedes-Benz BMW SAP Siemens Volkswagen Adidas Audi Allianz Porsche Bayer Bosch and Nivea Germany is recognised for its specialised small and medium enterprises Around 1 000 of these companies are global market leaders in their segment and are labelled hidden champions The list includes the largest German companies by revenue in 2011 Infrastructure With its central position in Europe Germany is a transport hub This is reflected in its dense and modern transport networks The motorway Autobahn network ranks as the third-largest worldwide in length and is known for its lack of a general speed limit Germany has established a polycentric network of high-speed trains The InterCityExpress or ICE network of the Deutsche Bahn serves major German cities as well as destinations in neighbouring countries with speeds up to 300 kph 186 mph The largest German airports are Frankfurt Airport and Munich Airport both hubs of Lufthansa while Air Berlin has hubs at Berlin Tegel and Düsseldorf Other major airports include Berlin Schönefeld Hamburg Cologne Bonn and Leipzig Halle Both airports in Berlin will be consolidated at a site adjacent to Berlin Schönefeld which will become Berlin Brandenburg Airport Germany was the world s sixth-largest consumer of energy and 60 of its primary energy was imported Government policy promotes energy conservation and renewable energy commercialisation Energy efficiency has been improving since the early 1970s the government aims to meet the country s electricity demands using 40 renewable sources by 2020 and 100 by 2050 In 2010 energy sources were oil 33 7 coal including lignite 22 9 natural gas 21 8 nuclear 10 8 hydro-electric and wind power 1 5 and other renewable sources 7 9 In 2000 the government and the nuclear power industry agreed to phase out all nuclear power plants by 2021 Germany is committed to the Kyoto protocol and several other treaties promoting biodiversity low emission standards recycling and the use of renewable energy and supports sustainable development at a global level The German government has initiated wide-ranging emission reduction activities and the country s overall emissions are falling Nevertheless the country s greenhouse gas emissions were the highest in the EU Science and technology Germany s achievements in the sciences have been significant and research and development efforts form an integral part of the economy The Nobel Prize has been awarded to 104 German laureates For most of the 20th century German laureates had more awards than those of any other nation especially in the sciences physics chemistry and physiology or medicine The work of Albert Einstein and Max Planck was crucial to the foundation of modern physics which Werner Heisenberg and Max Born developed further They were preceded by such key physicists as Hermann von Helmholtz Joseph von Fraunhofer and Gabriel Daniel Fahrenheit among others Wilhelm Röntgen discovered X-rays and was the first winner of the Nobel Prize in Physics in 1901 Otto Hahn was a pioneer in the fields of radioactivity and radiochemistry and discovered nuclear fission while Ferdinand Cohn and Robert Koch were founders of microbiology Numerous mathematicians were born in Germany including Carl Friedrich Gauss David Hilbert Bernhard Riemann Gottfried Leibniz Karl Weierstrass Hermann Weyl and Felix Klein Research institutions in Germany include the Max Planck Society the Helmholtz Association and the Fraunhofer Society The Gottfried Wilhelm Leibniz Prize is granted to ten scientists and academics every year With a maximum of €2 5 million per award it is one of highest endowed research prizes in the world Germany has been the home of many famous inventors and engineers such as Johannes Gutenberg credited with the invention of movable type printing in Europe Hans Geiger the creator of the Geiger counter and Konrad Zuse who built the first fully automatic digital computer German inventors engineers and industrialists such as Count Ferdinand von Zeppelin Otto Lilienthal Gottlieb Daimler Rudolf Diesel Hugo Junkers and Karl Benz helped shape modern automotive and air transportation technology Aerospace engineer Wernher von Braun developed the first space rocket and later on was a prominent member of NASA and developed the Saturn V Moon rocket which paved the way for the success of the US Apollo programme Heinrich Rudolf Hertz s work in the domain of electromagnetic radiation was pivotal to the development of modern telecommunication Germany is one of the leading countries in developing and using green technologies Companies specialising in green technology have an estimated turnover of €200 billion Key sectors of Germany s green technology industry are power generation sustainable mobility material efficiency energy efficiency waste management and recycling and sustainable water management Demographics With a population of 80 2 million according to the May 2011 census Germany is the most populous country in the European Union the second most populous country in Europe after Russia and ranks as the 16th most populous country in the world Its population density stands at 225 inhabitants per square kilometre The overall life expectancy in Germany at birth is 80 19 years 77 93 years for males and 82 58 years for females The fertility rate of 1 41 children born per woman 2011 estimates or 8 33 births per 1000 inhabitants is one of the lowest in the world Since the 1970s Germany s death rate has continuously exceeded its birth rate The Federal Statistical Office of Germany has forecast that the population could shrink to between 65 and 70 million by 2060 depending on the level of net migration Details on the methodology detailed tables etc are provided at ref However such forecasts have often been proven wrong in the past and Germany is currently witnessing increased birth rates and migration rates since the beginning of the 2010s It is notably experiencing a strong increase in the number of well-educated migrants In 2012 300 000 more immigrants than emigrants were reported in Germany Germans by nationality make up 92 3 of the population of Germany and about six million foreign citizens 7 7 of the population were registered in Germany Regarding ethnic background 20 of the country s residents or more than 16 million people were of foreign or partially foreign descent including persons descending or partially descending from ethnic German repatriates 96 of whom lived in the former West Germany or Berlin In 2010 2 3 million families with children under 18 years were living in Germany in which at least one parent had foreign roots They represented 29 of the total of 8 1 million families with minor children Compared with 2005 – the year when the microcensus started to collect detailed information on the population with a migrant background – the proportion of migrant families has risen by 2 percentage points Most of the families with a migrant background live in the western part of Germany In 2010 the proportion of migrant families in all families was 32 in the pre-unification territory of the Federal Republic This figure was more than double that in the new Länder including Berlin where it stood at 15 Families with a migrant background more often have three or more minor children in the household than families without a migrant background In 2010 about 15 of the families with a migrant background contained three or more minor children as compared with just 9 of the families without a migrant background The United Nations Population Fund lists Germany as host to the third-highest number of international migrants worldwide about 5 or 10 million of all 191 million migrants As a consequence of restrictions to Germany s formerly rather unrestricted laws on asylum and immigration the number of immigrants seeking asylum or claiming German ethnicity mostly from the former Soviet Union has been declining steadily since 2000 In 2009 20 of the population had immigrant roots the highest since 1945 the largest national group was from Turkey 2 5 million followed by Italy 776 000 and Poland 687 000 About 3 million Aussiedler —ethnic Germans mainly from the former eastern bloc—have resettled in Germany since 1987 Large numbers of people with full or significant German ancestry are found in the United States Brazil Argentina and Canada Most ethnic minorities especially those of non-European origin reside in large urban areas like Berlin Hamburg Frankfurt Rhine-Main Rhine-Ruhr Rhine-Neckar and Munich The percentage of non-Germans and immigrants is rather low in rural areas and small towns especially in the East German states of the former GDR territory Germany is home to the third-highest number of international migrants worldwide Ethnic composition in 2010 Germany has a number of large cities There are 11 officially recognised metropolitan regions in Germany – and since 2006 34 potential cities were identified which can be called a Regiopolis The largest conurbation is the Rhine-Ruhr region 11 7 million including Düsseldorf the capital of North Rhine-Westphalia Cologne Bonn Dortmund Essen Duisburg and Bochum Religion a href Christianity Christianity a is the largest religion in Germany with around 51 5 million adherents 62 8 in 2008 Relative to the whole population 30 0 of Germans are a href Catholic Church Catholics a 29 9 are a href Protestantism Protestants a belonging to the a href Evangelical Church in Germany Evangelical Church in Germany a EKD and the remaining Christians belong to smaller denominations each with less than 0 5 of the German population Protestantism is concentrated in the north and east and a href Roman Catholicism in Germany Roman Catholicism a is concentrated in the south and west 1 6 of the country s overall population declare themselves Orthodox Christians The second largest religion is Islam with an estimated 3 8 to 4 3 million adherents 4 6 to 5 2 followed by Buddhism with 250 000 and Judaism with around 200 000 adherents 0 3 Hinduism has some 90 000 adherents 0 1 All other religious communities in Germany have fewer than 50 000 adherents Of the roughly 4 million Muslims most are Sunnis and Alevites from Turkey but there are a small number of Shi ites and other denominations German Muslims a large portion of whom are of Turkish origin lack full official state recognition of their religious community Germany has Europe s third largest Jewish population after France and the United Kingdom Approximately 50 of the Buddhists in Germany are Asian immigrants Germans with no stated religious adherence make up 34 1 of the population and are concentrated in the former East Germany and major metropolitan areas German reunification in 1990 greatly increased the country s non-religious population a legacy of the state atheism of the previously Soviet -controlled East Christian church membership has decreased in recent decades particularly among Protestants Languages German is the official and predominant spoken language in Germany It is one of 23 official languages in the European Union and one of the three working languages of the European Commission Recognised native minority languages in Germany are Danish Low German Sorbian Romany and Frisian they are officially protected by the European Charter for Regional or Minority Languages The most used immigrant languages are Turkish Kurdish Polish the Balkan languages and Russian 67 of German citizens claim to be able to communicate in at least one foreign language and 27 in at least two languages other than their own ref Standard German is a West Germanic language and is closely related to and classified alongside English Low German Dutch and the Frisian languages To a lesser extent it is also related to the East extinct and North Germanic languages Most German vocabulary is derived from the Germanic branch of the Indo-European language family Significant minorities of words are derived from Latin and Greek with a smaller amount from French and most recently English known as Denglisch German is written using the Latin alphabet German dialects traditional local varieties traced back to the Germanic tribes are distinguished from varieties of standard German by their lexicon phonology and syntax Education Over 99 of Germans age 15 and above are estimated to be able to read and write Responsibility for educational supervision in Germany is primarily organised within the individual federal states Since the 1960s a reform movement attempted to unify secondary education in a Gesamtschule comprehensive school several West German states later simplified their school system to two or three tiers A system of apprenticeship called Duale Ausbildung dual education allows pupils in vocational training to learn in a company as well as in a state-run vocational school This successful model is highly regarded and reproduced all around the world Optional kindergarten education is provided for all children between three and six years old after which school attendance is compulsory for at least nine years Primary education usually lasts for four to six years and public schools are not stratified at this stage In contrast secondary education includes three traditional types of schools focused on different levels of academic ability the Gymnasium enrols the most gifted children and prepares students for university studies the Realschule for intermediate students lasts six years the Hauptschule prepares pupils for vocational education The general entrance requirement for university is Abitur a qualification normally based on continuous assessment during the last few years at school and final examinations however there are a number of exceptions and precise requirements vary depending on the state the university and the subject Germany s universities are recognised internationally in the Academic Ranking of World Universities ARWU for 2008 six of the top 100 universities in the world are in Germany and 18 of the top 200 Most of the German universities are public institutions charging tuition fees of only around €60 per semester and up to €500 in the state of Niedersachsen for each student Thus academic education is open to most citizens and studying is increasingly common in Germany Although the dual education system that combines practical and theoretical educations and does not lead to academic degrees is typical for Germany and recognised as a role model for other countries The oldest universities of Germany are also among the oldest and best regarded in the world with Heidelberg University being the oldest established in 1386 and in continuous operation since then It is followed by Leipzig University 1409 Rostock University 1419 Greifswald University 1456 Freiburg University 1457 LMU Munich 1472 and the University of Tübingen 1477 Most German universities focus more on teaching than on research Research is mostly exhibited in independent institutes that are embedded in academic clusters such as within Max Planck Fraunhofer Leibniz and Helmholtz institutes This German specialisation is rarely reflected in academic rankings which is the reason why German universities seem to be underperforming according to some of the ratings such as ARWU Some German universities are shifting their focus more towards research though especially the institutes of technology Health Germany has the world s oldest universal health care system dating back to Bismarck s social legislation in 1883 Currently the population is covered by a fairly comprehensive health insurance plan provided by statute Certain groups of people lifetime officials self-employed persons employees with high income can opt out of the plan and switch to a private insurance contract Previously these groups could also choose to do without insurance but this option was dropped in 2009 According to the World Health Organization Germany s health care system was 77 government-funded and 23 privately funded In 2005 Germany spent 11 of its GDP on health care Germany ranked 20th in the world in life expectancy with 77 years for men and 82 years for women and it had a very low infant mortality rate 4 per 1 000 live births the principal cause of death was cardiovascular disease at 41 followed by malignant tumours at 26 about 82 000 Germans had been infected with HIV AIDS and 26 000 had died from the disease cumulatively since 1982 This article may incorporate text from this source which is in the public domain ref According to a 2005 survey 27 of German adults are smokers Culture From its roots culture in German states has been shaped by major intellectual and popular currents in Europe both religious and secular Historically Germany has been called Das Land der Dichter und Denker the land of poets and thinkers because of the major role its famous writers and philosophers have played in the development of Western thought and culture The federated states are in charge of the cultural institutions There are 240 subsidised theatres hundreds of symphonic orchestras thousands of museums and over 25 000 libraries spread in Germany These cultural opportunities are enjoyed by many there are over 91 million German museum visits every year annually 20 million go to theatres and operas 3 6 million per year listen to the symphonic orchestras As of 2012 the UNESCO inscribed 37 properties in Germany on the World Heritage List Germany has established a high level of gender equality promotes disability rights and is legally and socially tolerant towards homosexuals Gays and lesbians can legally adopt their partner s biological children and civil unions have been permitted since 2001 Germany has also changed its attitude towards immigrants since the mid-1990s the government and the majority of Germans have begun to acknowledge that controlled immigration should be allowed based on qualification standards Germany has been named the world s second most valued nation among 50 countries in 2010 A global opinion poll for the BBC revealed that Germany is recognised for having the most positive influence in the world in 2011 and for being the most positively viewed nation in the world in 2013 Art Numerous German painters have enjoyed international prestige through their work in diverse artistic styles Albrecht Dürer Hans Holbein the Younger Matthias Grünewald and Lucas Cranach the Elder were important artists of the Renaissance Peter Paul Rubens and Johann Baptist Zimmermann of Baroque Caspar David Friedrich and Carl Spitzweg of Romanticism Max Liebermann of Impressionism and Max Ernst of Surrealism Several German artist groups formed in the 20th century such as the November Group or Die Brücke The Bridge and Der Blaue Reiter The Blue Rider in Expressionism The New Objectivity arose as a counter-style to it during the Weimar Republic After WWII main movements of Neo-expressionism performance art and Conceptual art evolved with notable artists such as Joseph Beuys Gerhard Richter Jörg Immendorff HA Schult Aris Kalaizis Neo Rauch New Leipzig School and Andreas Gursky photography Major art exhibitions and festivals in Germany are the documenta transmediale and Art Cologne Music German classical music comprises works by some of the world s most well-known composers including Ludwig van Beethoven Johann Sebastian Bach Wolfgang Amadeus Mozart Johannes Brahms Richard Wagner and Richard Strauss Germany is the largest music market in Europe and third largest in the world German popular music of the 20th and 21st century includes the movements of Neue Deutsche Welle Nena Alphaville Ostrock City Keimzeit Metal Rock Rammstein Scorpions Punk Die Ärzte Die Toten Hosen Pop rock Beatsteaks Tokio Hotel Indie Tocotronic Blumfeld and Hip Hop Die Fantastischen Vier Deichkind Especially the German Electronic music gained global influence with Kraftwerk being a pioneer group in this genre and the Minimal and Techno scenes in Germany being very popular e g Paul van Dyk Tomcraft Paul Kalkbrenner and Scooter Architecture Architectural contributions from Germany include the Carolingian and Ottonian styles which were precursors of Romanesque Brick Gothic in medieval times and Brick Expressionism in modern times are two distinctive styles that developed in Germany Also in Renaissance and Baroque art regional and typically German elements evolved e g Weser Renaissance and Dresden Baroque Among many renowned Baroque masters were Pöppelmann Balthasar Neumann Knobelsdorff and the Asam brothers Germany is especially renowned for its timber frame old towns with many well-kept examples to be found along the German Timber-Frame Road leading from the very south of Germany to Northern Germany and its coasts When industrialisation spread across Europe Classicism and a distinctive style of historism developed in Germany sometimes referred to as Gründerzeit style due to the economical boom years at the end of the 19th century Resort architecture and Spa architecture are sub-styles that evolved since the 18th century in Germany with the first modern spas and Seaside resorts of Europe Many architects formed this era with Schinkel Semper Stüler von Gärtner Schwechten and Lipsius among them Jugendstil became a dominant architectural style at the turn of the 19th to the 20th century with a strong influence of the Art Nouveau movement The Art Deco movement did not gain much influence in Germany instead the Expressionist architecture spread across the country with e g Höger Mendelsohn Böhm and Schumacher being influential architects Germany was particularly important in the early modern movement - it is the home of the Bauhaus movement founded by Walter Gropius And thus Germany is a cradle of modern architecture Ludwig Mies van der Rohe became one of the world s most renowned architects in the second half of the 20th century He conceived of the glass façade skyscraper Renowned contemporary architects include Hans Kollhoff Helmut Jahn Behnisch Albert Speer Junior Frei Otto Oswald Mathias Ungers Gottfried Böhm Stephan Braunfels and Anna Heringer Literature and philosophy German literature can be traced back to the Middle Ages and the works of writers such as Walther von der Vogelweide and Wolfram von Eschenbach Well-known German authors include Johann Wolfgang von Goethe Friedrich Schiller Gotthold Ephraim Lessing and Theodor Fontane The collections of folk tales published by the Brothers Grimm popularised German folklore on an international level Influential authors of the 20th century include Gerhart Hauptmann Thomas Mann Hermann Hesse Heinrich Böll and Günter Grass German-speaking book publishers produce some 700 million books every year with about 80 000 titles nearly 60 000 of them new Germany comes third in quantity of books published after the English-speaking book market and the People s Republic of China The Frankfurt Book Fair is the most important in the world for international deals and trading with a tradition spanning over 500 years German philosophy is historically significant Gottfried Leibniz s contributions to rationalism the enlightenment philosophy by Immanuel Kant the establishment of classical German idealism by Johann Gottlieb Fichte Georg Wilhelm Friedrich Hegel and Friedrich Wilhelm Joseph Schelling Arthur Schopenhauer s composition of metaphysical pessimism the formulation of communist theory by Karl Marx and Friedrich Engels Friedrich Nietzsche s development of perspectivism Gottlob Frege s contributions to the dawn of analytic philosophy Martin Heidegger s works on Being and the development of the Frankfurt school by Max Horkheimer Theodor Adorno Herbert Marcuse and Jürgen Habermas have been particularly influential In the 21st century Germany has contributed to the development of contemporary analytic philosophy in continental Europe Media German cinema dates back to the earliest years of the medium it has made major technical and artistic contributions to film as with the work of Max Skladanowsky who showed the first film sequences ever to an audience in 1895 Early German cinema was particularly influential with German expressionists such as Robert Wiene and Friedrich Wilhelm Murnau Director Fritz Lang s Metropolis 1927 is referred to as the first modern science-fiction film In 1930 the Austrian-American Josef von Sternberg directed The Blue Angel the first major German sound film During the 1970s and 1980s New German Cinema directors such as Volker Schlöndorff Werner Herzog Wim Wenders and Rainer Werner Fassbinder put West German cinema on the international stage The annual European Film Awards ceremony is held every other year in Berlin home of the European Film Academy EFA the Berlin Film Festival held annually since 1951 is one of the world s foremost film festivals In the 21st century several German movies have had international success such as Nowhere in Africa 2001 Das Experiment 2001 Good Bye Lenin 2003 Gegen die Wand Head-On 2004 Der Untergang Downfall 2004 2006 The Baader Meinhof Complex 2008 The Wave 2008 The White Ribbon 2009 Pandorum 2009 Soul Kitchen 2009 Animals United 2010 Combat Girls 2011 and Cloud Atlas 2012 The Academy Award for Best Foreign Language Film went to the German production Die Blechtrommel The Tin Drum in 1979 to Nowhere in Africa in 2002 and to Das Leben der Anderen The Lives of Others in 2007 Germany s television market is the largest in Europe with some 34 million TV households Around 90 of German households have cable or satellite TV with a variety of free-to-view public and commercial channels The most watched television broadcast of all-time in Germany was the Germany vs Spain semi-fnal game of the 2010 FIFA World Cup Nine out of ten of the top ten most watched television broadcasts of all-time in Germany feature the German national football team Cuisine German cuisine varies from region to region The southern regions of Bavaria and Swabia for instance share a culinary culture with Switzerland and Austria In all regions meat is often eaten in sausage form Organic food has gained a market share of ca 2 and is expected to increase further Although wine is becoming more popular in many parts of Germany the national alcoholic drink is beer German beer consumption per person is declining but at 121 4 litres in 2009 it is still among the highest in the world The Michelin guide has awarded nine restaurants in Germany three stars the highest designation while 15 more received two stars German restaurants have become the world s second-most decorated after France Pork beef and poultry are the main varieties of meat consumed in Germany with pork being the most popular The average person in Germany will consume up to of meat in a year Among poultry chicken is most common although duck goose and turkey are also enjoyed Game meats especially boar rabbit and venison are also widely available all year round Trout is the most common freshwater fish on the German menu pike carp and European perch also are listed frequently Vegetables are often used in stews or vegetable soups but are also served as side dishes Carrots turnips spinach peas beans broccoli and many types of cabbage are very common A wide variety of cakes and tarts are served throughout the country most commonly made with fresh fruit Apples plums strawberries and cherries are used regularly in cakes Cheesecake is also very popular often made with quark Schwarzwälder Kirschtorte Black Forest cake made with cherries is probably the most well-known example of a wide variety of typically German tortes filled with whipped or butter cream Sports Twenty-seven million Germans are members of a sports club and an additional twelve million pursue sports individually Association football is the most popular sport With more than 6 3 million official members the German Football Association Deutscher Fußball-Bund is the largest sports organisation of its kind worldwide The Bundesliga the top league of German football is the most popular sports league in Germany and attracts the second highest average attendance of any professional sports league in the world The German national football team won the FIFA World Cup in 1954 1974 and 1990 and the UEFA European Football Championship in 1972 1980 and 1996 Germany hosted the FIFA World Cup in 1974 and 2006 and the UEFA European Football Championship in 1988 Among the most well-known footballers are Franz Beckenbauer Gerd Müller Jürgen Klinsmann Lothar Matthäus and Oliver Kahn Other popular spectator sports include handball volleyball basketball ice hockey and tennis Germany is one of the leading motor sports countries in the world Constructors like BMW and Mercedes are prominent manufacturers in motor sport Additionally Porsche has won the 24 Hours of Le Mans an annual endurance race held in France 16 times and Audi has won it 11 times The Formula One driver Michael Schumacher has set many motor sport records during his career having won more Formula One World Drivers Championships and more Formula One races than any other driver he is one of the highest paid sportsmen in history Historically German sportsmen have been successful contenders in the Olympic Games ranking third in an all-time Olympic Games medal count combining East and West German medals In the 2008 Summer Olympics Germany finished fifth in the medal count while in the 2006 Winter Olympics they finished first Germany has hosted the Summer Olympic Games twice in Berlin in 1936 and in Munich in 1972 The Winter Olympic Games took place in Germany once in 1936 in the twin towns of Garmisch and Partenkirchen'\n\nent_lines_4EX['Queen (band)'] = '42010\tQueen (band)\tQueen band Queen are a British rock band formed in London in 1970 originally consisting of Freddie Mercury lead vocals piano Brian May guitar vocals John Deacon bass guitar and Roger Taylor drums vocals Queen s earliest works were influenced by progressive rock hard rock and heavy metal but the band gradually ventured into more conventional and radio-friendly works incorporating further diverse styles into their music Before joining Queen Brian May and Roger Taylor had been playing together in a band named Smile with bassist Tim Staffell Freddie Mercury then known as Farrokh Freddie Bulsara was a fan of Smile and encouraged them to experiment with more elaborate stage and recording techniques after Staffell s departure in 1970 Mercury himself joined the band shortly thereafter changed the name of the band to Queen and adopted his familiar stage name John Deacon was recruited prior to recording their eponymous debut album 1973 Queen enjoyed success in the UK with their debut and its follow-up Queen II 1974 but it was the release of Sheer Heart Attack 1974 and A Night at the Opera 1975 that gained the band international success The latter featured Bohemian Rhapsody which stayed at number one in the UK Singles Chart for nine weeks it charted at number one in several other territories and gave the band their first top ten hit on the US Billboard Hot 100 Their 1977 album News of the World contained two of rock s most recognisable anthems We Will Rock You and We Are the Champions By the early 1980s Queen were one of the biggest stadium rock bands in the world and their performance at 1985 s Live Aid is widely regarded as one of the greatest in rock history In 1991 Mercury died of bronchopneumonia a complication of AIDS and Deacon retired in 1997 Since then May and Taylor have infrequently performed together including a collaboration with Paul Rodgers under the name Queen Paul Rodgers which ended in May 2009 The band have released a total of 18 number one albums 18 number one singles and 10 number one DVDs Estimates of their album sales generally range from 150 million to 300 million albums making them one of the world s best-selling music artists They received the Outstanding Contribution to British Music Award from the British Phonographic Industry in 1990 and were inducted into the Rock and Roll Hall of Fame in 2001 History Early days 1968–1974 In 1968 guitarist Brian May a student at London s Imperial College and bassist Tim Staffell decided to form a band May placed an advertisement on the college notice board for a Mitch Mitchell Ginger Baker type drummer Roger Taylor a young dental student auditioned and got the job The group called themselves Smile While attending Ealing Art College Tim Staffell became friends with Farrokh Bulsara a fellow student who had assumed the English name of Freddie Bulsara felt that he and the band had the same tastes and soon became a keen fan of Smile In late 1970 after Staffell left to join the band Humpy Bong the remaining Smile members encouraged by Bulsara changed their name to Queen and continued working together When asked about the name Bulsara explained I thought up the name Queen It s just a name but it s very regal obviously and it sounds splendid It s a strong name very universal and immediate It had a lot of visual potential and was open to all sorts of interpretations I was certainly aware of gay connotations but that was just one facet of it The band had a number of bass players during this period who did not fit with the band s chemistry It was not until February 1971 that they settled on John Deacon and began to rehearse for their first album They recorded four of their own songs Liar Keep Yourself Alive The Night Comes Down and Jesus for a demo tape no record companies were interested It was also around this time Freddie changed his surname to Mercury inspired by the line Mother Mercury look what they ve done to me in the song My Fairy King On 2 July 1971 Queen played their first show in the classic line-up of Mercury May Deacon and Taylor at a Surrey college outside London Having attended art college Mercury also designed Queen s logo called the Queen crest shortly before the release of the band s first album The logo combines the zodiac signs of all four members two lions for Leo Deacon and Taylor a crab for Cancer May and two fairies for Virgo Mercury The lions embrace a stylised letter Q the crab rests atop the letter with flames rising directly above it and the fairies are each sheltering below a lion There is also a crown inside the Q and the whole logo is over-shadowed by an enormous phoenix The whole symbol bears a passing resemblance to the Royal coat of arms of the United Kingdom particularly with the lion supporters The original logo as found on the reverse-side of the first album cover was a simple line drawing but more intricate colour versions were used on later sleeves In 1972 Queen entered discussions with Trident Studios after being spotted at De La Lane Studios by John Antony and after discussions were offered a management deal by Norman Sheffield under Neptune Productions a subsidiary of Trident to manage the band and enable them to use the facilities at Trident to record new material whilst the management search for a record label to sign Queen This suited both parties at the time as Trident were expanding into management and Queen under the deal were able to make use of the hi-tech recording facilities shared by bands at the time such as The Beatles and Elton John to produce new material However Trident found it difficult to find a label for a band bearing a name with such connotation during the early 1970s In July 1973 Queen finally under under a Trident EMI deal released their eponymous debut album an effort influenced by the heavy metal and progressive rock of the day The album was received well by critics Gordon Fletcher of Rolling Stone said their debut album is superb and Chicago s Daily Herald called it an above average debut It drew little mainstream attention and the lead single Keep Yourself Alive a Brian May composition sold poorly Retrospectively Keep Yourself Alive is cited as the highlight of the album and in 2008 Rolling Stone ranked it 31st in the 100 Greatest Guitar Songs Of All Time describing it as an entire album s worth of riffs crammed into a single song The album was certified gold in the UK and the US The group s second LP Queen II was released in 1974 and features rock photographer Mick Rock s iconic image of the band on the cover This image would be used as the basis for the 1975 Bohemian Rhapsody music video production The album reached number five on the British album chart and became the first Queen album to chart in the UK The Freddie Mercury-written lead single Seven Seas of Rhye reached number ten in the UK giving the band their first hit The album is the first real testament to the band s distinctive layered sound and features long complex instrumental passages fantasy-themed lyrics and musical virtuosity Aside from its only single the album also included the song The March of the Black Queen a six-minute epic which lacks a chorus The Daily Vault described the number as menacing Critical reaction was mixed the Winnipeg Free Press while praising the band s debut album described Queen II as a over-produced monstrosity Allmusic has described the album as a favourite among the band s hardcore fans and it is the first of three Queen albums to feature in the book 1001 Albums You Must Hear Before You Die Sheer Heart Attack to A Night at the Opera 1974–1976 After the band s six-night stand at New York s Uris Theatre in May 1974 Brian May collapsed and was diagnosed as having hepatitis While recuperating May was initially absent when the band started work on their third album but he returned midway through the recording process Released in 1974 Sheer Heart Attack reached number two in the United Kingdom sold well throughout Europe and went gold in the United States It gave the band their first real experience of international success and was a hit on both sides of the Atlantic The album experimented with a variety of musical genres including British music hall heavy metal ballads ragtime and Caribbean At this point Queen started to move away from the progressive tendencies of their first two releases into a more radio-friendly song-oriented style Sheer Heart Attack introduced new sound and melody patterns that would be refined on their next album A Night at the Opera The single Killer Queen from Sheer Heart Attack reached number two on the British charts and became their first US hit reaching number 12 on the Billboard Hot 100 It combines camp vaudeville and British music hall with May s guitar virtuosity The album s second single Now I m Here a more traditional hard rock composition was a number eleven hit in Britain while the high speed rocker Stone Cold Crazy featuring May s uptempo riffs is a precursor to speed metal In recent years the album has received acclaim from music publications In 2006 Classic Rock ranked it number 28 in The 100 Greatest British Rock Albums Ever and in 2007 Mojo ranked it No 88 in The 100 Records That Changed the World It is also the second of three Queen albums to feature in the book 1001 Albums You Must Hear Before You Die In 1975 the band left for a world tour with each member in Zandra Rhodes -created costumes and accompanied with banks of lights and effects They toured the US as headliners and played in Canada for the first time In September later that year after an acromonious split with Trident the band negotiated themselves out of their Trident Studios contract and searched for new management One of the options they considered was an offer from Led Zeppelin s manager Peter Grant Grant wanted them to sign with Led Zeppelin s own production company Swan Song Records The band found the contract unacceptable and instead contacted Elton John s manager John Reid who accepted the position In late 1975 Queen recorded and released A Night at the Opera taking its name from the popular Marx Brothers movie At the time it was the most expensive album ever produced Like its predecessor the album features diverse musical styles and experimentation with stereo sound In The Prophet s Song an eight-minute epic the middle section is a canon with simple phrases layered to create a full-choral sound The Mercury penned ballad Love of My Life featured a harp and overdubbed vocal harmonies The album was very successful in Britain and went triple platinum in the United States The British public voted it the 13th greatest album of all time in a 2004 Channel 4 poll It has also ranked highly in international polls in a worldwide Guinness poll it was voted the 19th greatest of all time while an ABC poll saw the Australian public vote it the 28th greatest of all time A Night at the Opera has frequently appeared in greatest albums lists reflecting the opinions of critics Among other accolades it was ranked number 16 in Q Magazine s The 50 Best British Albums Ever in 2004 and number 11 in Rolling Stone s The 100 Greatest Albums of All Time as featured in their Mexican edition in 2004 It was also placed at No  230 on Rolling Stone magazine s list of The 500 Greatest Albums of All Time in 2003 A Night at the Opera is the third and final Queen album to be featured in the book 1001 Albums You Must Hear Before You Die The album also featured the hit single Bohemian Rhapsody which was number one in the UK for nine weeks and is the third-best-selling single of all time in the UK surpassed only by Band Aid s Do They Know It s Christmas and Elton John s Candle in the Wind 1997 —making it the best selling commercial single in the UK It also reached number nine in the United States a 1992 re-release reached number two on Billboard for five weeks It is the only single ever to sell a million copies on two separate occasions and became the Christmas number one twice in the UK the only single ever to do so Bohemian Rhapsody has been voted numerous times the greatest song of all time The band decided to make a video to help go with the single and hired Trilion a subsidiary of the former management company Trident Studios using new technology to create the video the result is generally considered to have been the first true music video ever produced The album s first track Death on Two Legs is said to be written by Mercury about Norman Sheffield and the former management at Trident who helped make the video so popular Although other bands including The Beatles had made short promotional films or videos of songs prior to this generally those were specifically made to be aired on specific television shows The second single from the album You re My Best Friend the second song composed by John Deacon and his first single peaked at number sixteen in the United States and went on to become a worldwide Top Ten hit The band s A Night at the Opera Tour began in November 1975 and covered Europe the United States Japan and Australia A Day at the Races to Live Killers 1976–1979 By 1976 Queen were back in the studio recording A Day at the Races which is often regarded as a sequel album to A Night at the Opera It again borrowed the name of a Marx Brothers movie and its cover was similar to that of A Night at the Opera a variation on the same Queen Crest The most recognisable of the Marx Brothers Groucho Marx invited Queen to visit him in his Los Angeles home in March 1977 there the band thanked him in person and performed 39 a cappella Musically A Day at the Races was by both fans and critics standards a strong effort reaching number one in the UK and Japan and number five in the US The major hit on the album was Somebody to Love a gospel-inspired song in which Mercury May and Taylor multi-tracked their voices to create a 100-voice gospel choir The song went to number two in the United Kingdom and number thirteen on the US singles chart The album also featured one of the band s heaviest songs Brian May s Tie Your Mother Down which became a staple of their live shows During 1976 Queen played one of their most famous gigs a free concert in Hyde Park London It set an attendance record with 150 000 people confirmed in the audience On 1 December 1976 Queen were the intended guests on London s early evening Today programme but they pulled out at the last-minute which saw their late replacement on the show EMI labelmate the Sex Pistols give their seminal interview During the A Day at the Races Tour in 1977 Queen performed sold-out shows at Madison Square Garden New York in February and Earls Court London in June The band s sixth studio album News of the World was released in 1977 which has gone four times platinum in the United States and twice in the UK The album contained many songs tailor-made for live performance including two of rock s most recognisable anthems We Will Rock You and the rock ballad We Are the Champions both of which became enduring international sports anthems and the latter reached number four in the United States Queen commenced the News of the World Tour in October 1977 and Robert Hilburn of the Los Angeles Times called this concert tour the band s most spectacularly staged and finely honed show In 1978 Queen toured the US and Canada and spent much of 1979 touring in Europe and Japan They released their first live album Live Killers in 1979 it went platinum twice in the United States Queen also released the very successful single Crazy Little Thing Called Love a rockabilly inspired song done in the style of Elvis Presley The song made the top 10 in many countries topped the Australian ARIA Charts for seven consecutive weeks and was the band s first number one single in the United States where it topped the Billboard Hot 100 for four weeks Having written the song on guitar and played rhythm on the record Mercury played rhythm guitar while performing the song live which was the first time he ever played guitar in concert In December 1979 Queen played the opening night at the Concert for the People of Kampuchea in London having accepted a request by the event s organiser Paul McCartney The Game to The Works 1980–1984 Queen began their 1980s career with The Game It featured the singles Crazy Little Thing Called Love and Another One Bites the Dust both of which reached number one in the United States After attending a Queen concert in Los Angeles Michael Jackson suggested to Mercury backstage that Another One Bites the Dust be released as a single and in October 1980 it spent three weeks at number one The album topped the Billboard 200 for five weeks and sold over four million copies in the US It was also the first appearance of a synthesiser on a Queen album Heretofore their albums featured a distinctive No Synthesisers sleeve note The note is widely assumed to reflect an anti-synth pro- hard -rock stance by the band but was later revealed by producer Roy Thomas Baker to be an attempt to clarify that those albums multi-layered solos were created with guitars not synths as record company executives kept assuming at the time In September 1980 Queen performed three sold-out shows at Madison Square Garden In 1980 Queen also released the soundtrack they had recorded for Flash Gordon At the 1981 American Music Awards on 30 January Another One Bites the Dust won the award for Favorite Pop Rock Single and Queen were nominated for Favorite Pop Rock Band Duo or Group In February 1981 Queen travelled to South America as part of The Game Tour and became the first major rock band to play in Latin American stadiums The tour included five shows in Argentina one of which drew the largest single concert crowd in Argentine history with an audience of 300 000 in Buenos Aires and two concerts at the Morumbi Stadium in São Paulo Brazil where they played to an audience of more than 131 000 people in the first night then the largest paying audience for a single band anywhere in the world and more than 120 000 people the following night In October of the same year Queen performed for more than 150 000 fans on 9 October at Monterrey Estadio Universitario and 17 and 18 at Puebla Estadio Zaragoza Mexico On 24 and 25 November Queen played two sell out nights at the Montreal Forum Quebec Canada One of Mercury s most notable performances of The Game s final track Save Me took place in Montreal and the concert is recorded in the live album Queen Rock Montreal Queen worked with David Bowie on the single Under Pressure The first-time collaboration with another artist was spontaneous as Bowie happened to drop by the studio while Queen were recording Upon its release the song was extremely successful reaching number one in the UK and featuring at number 31 on VH1 s 100 Greatest Songs of the 80s Later in 1981 Queen released their first compilation album titled Greatest Hits which showcased the group s highlights from 1974–1981 It is the best-selling album in UK Chart history and has spent 450 weeks in the UK Album Chart The album is certified eight times platinum in the United States and has sold over 25 million copies worldwide Taylor became the first member of the band to release his own solo album in 1981 titled Fun in Space In 1982 the band released the album Hot Space a departure from their trademark seventies sound this time being a mixture of rock pop rock dance funk and R B Most of the album was recorded in Munich during the most turbulent period in the band s history and Taylor and May lamented the new sound with both being very critical of the influence Mercury s personal manager Paul Prenter had on the singer May was also scathing of Prenter who was Mercury s manager from the early 1980s to 1984 for being dismissive of the importance of radio stations such as the US networks and their vital connection between the artist and the community and for denying them access to Mercury The band stopped touring North America after their Hot Space Tour as their success there had waned although they would perform on American television for the only time during the eighth season premiere of Saturday Night Live Queen left Elektra Records their label in the United States Canada Japan Australia and New Zealand and signed onto EMI Capitol Records After working steadily for over ten years Queen decided that they would not perform any live shows in 1983 During this time they recorded a new album at the Record Plant Studios Los Angeles and Musicland Studios Munich and several members of the band explored side projects and solo work May released a mini-album titled Star Fleet Project on which he collaborated with Eddie Van Halen In February 1984 Queen released their eleventh studio album The Works which included the successful singles Radio Ga Ga Hammer to Fall and I Want to Break Free Despite these hit singles the album failed to do well in the United States while in the UK it went triple platinum and remained in the album chart for two years That year Queen began The Works Tour the first tour to feature keyboardist Spike Edney as an extra live musician The tour featured nine sold-out dates in October in Bophuthatswana South Africa at the arena in Sun City Upon returning to England they were the subject of outrage having played there during the height of apartheid and in violation of worldwide divestment efforts and a United Nations cultural boycott The band responded to the critics by stating that they were playing music for fans in that country and they also stressed that the concerts were played before integrated audiences Queen donated to a school for the deaf and blind as a philanthropic gesture but were fined by the Musicians Union and placed on the United Nations blacklisted artists Live Aid and later years 1985–1990 In January 1985 the band headlined two nights of the first Rock in Rio festival at Rio de Janeiro Brazil and played in front of over 300 000 people each night The Boston Globe described it as a mesmerising performance A selection of highlights of both nights was released on VHS with the title Queen Live in Rio and was later broadcast on MTV in the US In April and May 1985 Queen completed the Works Tour with sold-out shows in Australia and Japan At Live Aid held at Wembley on 13 July 1985 in front of the biggest-ever TV audience of 1 9 billion Queen performed some of their greatest hits during which the sold-out stadium audience of 72 000 people clapped sang and swayed in unison The show s organisers Bob Geldof and Midge Ure other musicians such as Elton John Cliff Richard and Dave Grohl and various music journalists commented that Queen stole the show An industry poll in 2005 named it the greatest rock performance of all time When interviewed for Mojo magazine the band said the most amazing sight at Live Aid was to see the audience clapping to Radio Ga Ga Brian May stated I d never seen anything like that in my life and it wasn t calculated either We understood our audience and played to them but that was one of those weird accidents because of the music video I remember thinking oh great they ve picked it up and then I thought this is not a Queen audience This is a general audience who ve bought tickets before they even knew we were on the bill And they all did it How did they know Nobody told them to do it The band now revitalised by the response to Live Aid — a shot in the arm Roger Taylor called it Also a limited-edition boxed set containing all Queen albums to date was released under the title of The Complete Works The package included previously unreleased material most notably Queen s non-album single of Christmas 1984 titled Thank God It s Christmas In early 1986 Queen recorded the album A Kind of Magic containing several reworkings of songs written for the fantasy action film Highlander The album was very successful producing a string of hits including the title track A Kind of Magic Also charting from the album were Friends Will Be Friends Who Wants to Live Forever featuring an orchestra conducted by Michael Kamen and the de facto theme from Highlander Princes of the Universe In summer of 1986 Queen went on their final tour with Freddie Mercury A sold-out tour in support of A Kind of Magic once again they hired Spike Edney leading to him being dubbed the unofficial fifth member The Magic Tour s highlight was at Wembley Stadium in London and resulted in the live double album Queen at Wembley released on CD and as a live concert DVD which has gone five times platinum in the US and four times platinum in the UK Queen could not book Wembley for a third night but they did play at Knebworth Park The show sold out within two hours and over 120 000 fans packed the park for what was Queen s final live performance with Mercury Queen began the tour at the Råsunda Stadium in Stockholm Sweden and during the tour the band performed a concert at Slane Castle Ireland in front of an audience of 95 000 which broke the venue s attendance record The band also played behind the Iron Curtain when they performed to a crowd of 80 000 at the Népstadion in Budapest in what was one of the biggest rock concerts ever held in Eastern Europe More than one million people saw Queen on the tour—400 000 in the United Kingdom alone a record at the time After working on various solo projects during 1988 including Mercury s collaboration with Montserrat Caballé Barcelona the band released The Miracle in 1989 The album continued the direction of A Kind of Magic using a pop-rock sound mixed with a few heavy numbers It spawned the European hits I Want It All Breakthru The Invisible Man Scandal and The Miracle The Miracle also began a change in direction of Queen s songwriting philosophy Since the band s beginning nearly all songs had been written by and credited to a single member with other members adding minimally With The Miracle the band s songwriting became more collaborative and they vowed to credit the final product only to Queen as a group Mercury illness death and tribute 1988–1992 After fans noticed Mercury s increasingly gaunt appearance in 1988 rumours began to spread that Mercury was suffering from AIDS Mercury flatly denied this insisting he was merely exhausted and too busy to provide interviews The band decided to continue making albums starting with The Miracle in 1989 and continuing with Innuendo in 1991 Despite his deteriorating health the lead singer continued to contribute For the last two albums made while Mercury was still alive the band credited all songs to Queen rather than specific members of the group freeing them of internal conflict and differences In 1990 Mercury made his final public appearance when he joined the rest of Queen to collect the BRIT Award for Outstanding Contribution to British Music Innuendo was released in early 1991 with an eponymous number 1 UK hit and other charting singles including The Show Must Go On Mercury was increasingly ill and could barely walk when the band recorded The Show Must Go On in 1990 Because of this May had concerns about whether he was physically capable of singing it Recalling Mercury s successful performance May states he went in and killed it completely lacerated that vocal The band s second greatest hits compilation Greatest Hits II followed in October 1991 which is the eighth best-selling album of all time in the UK and has sold 16 million copies worldwide On 23 November 1991 in a prepared statement made on his deathbed Mercury confirmed that he had AIDS Within 24 hours of the statement he died of bronchial pneumonia which was brought on as a complication of AIDS His funeral service on 27 November in Kensal Green West London was private and held in accordance with the Zoroastrian religious faith of his family Bohemian Rhapsody was re-released as a single shortly after Mercury s death with These Are the Days of Our Lives as the double A-side The music video for These Are the Days of Our Lives contain Mercury s final scenes in front of the camera The single went to number one in the UK remaining there for five weeks – the only recording to top the Christmas chart twice and the only one to be number one in four different years 1975 1976 1991 and 1992 Initial proceeds from the single – approximately £1 000 000 – were donated to the Terrence Higgins Trust Queen s popularity was stimulated in the United States when Bohemian Rhapsody was featured in the 1992 comedy film Wayne s World Its inclusion helped the song reach number two on the Billboard Hot 100 for five weeks in 1992 it remained in the Hot 100 for over 40 weeks and won the band an MTV Award at the 1992 MTV Video Music Awards The compilation album Classic Queen also reached number four on the Billboard 200 and is certified three times platinum in the US Wayne s World footage was used to make a new music video for Bohemian Rhapsody with which the band and management were delighted On 20 April 1992 The Freddie Mercury Tribute Concert was held at London s Wembley Stadium to a 72 000-strong crowd Performers including Def Leppard Robert Plant Guns N Roses Elton John David Bowie George Michael Annie Lennox Seal Extreme and Metallica performed various Queen songs along with the three remaining Queen members The concert is listed in the Guinness Book of Records as The largest rock star benefit concert as it was televised to over 1 2 billion viewers worldwide and raised over £20 000 000 for AIDS charities Made in Heaven to 46664 Concert 1995–2003 Queen s last album featuring Mercury titled Made in Heaven was finally released in 1995 four years after his death Featuring tracks such as Too Much Love Will Kill You and Heaven for Everyone it was constructed from Mercury s final recordings in 1991 material left over from their previous studio albums and re-worked material from May Taylor and Mercury s solo albums Both stages of recording before and after Mercury s death were completed at the band s studio in Montreux Switzerland The album reached No  1 on the UK charts immediately following its release and has sold 20 million copies worldwide On 25 November 1996 a statue of Mercury was unveiled in Montreux overlooking Lake Geneva almost five years to the day since his death In 1997 Queen returned to the studio to record No-One but You Only the Good Die Young It was released as a bonus track on the Queen Rocks compilation album later that year The song was later released as a single reaching number 13 in the UK chart In January 1997 Queen performed The Show Must Go On live with Elton John and the Béjart Ballet in Paris on a night in which Freddie Mercury was remembered and it marked the last performance and public appearance of John Deacon who chose to retire The Paris concert was only the second time Queen had played live since Mercury s death and prompted Elton John to urge them to perform again Brian May and Roger Taylor performed together at several award ceremonies and charity concerts sharing vocals with various guest singers During this time they were billed as Queen followed by the guest singer s name In 1998 the duo appeared at Luciano Pavarotti s benefit concert with May performing Too Much Love Will Kill You with Pavarotti later playing Radio Ga Ga We Will Rock You and We Are the Champions with Zucchero They again attended and performed at Pavarotti s benefit concert in Modena Italy in May 2003 Several of the guest singers recorded new versions of Queen s hits under the Queen name such as Robbie Williams providing vocals for We Are the Champions for the soundtrack of A Knight s Tale 2001 In 1999 a Greatest Hits III album was released This featured among others Queen Wyclef Jean on a rap version of Another One Bites the Dust A live version of Somebody to Love by George Michael and a live version of The Show Must Go On with Elton John were also featured in the album By this point Queen s vast amount of record sales made them the second best selling artist in the UK of all time behind The Beatles In 2002 Queen were awarded the 2 207th star on the Hollywood Walk of Fame which is located at 6358 Hollywood Blvd On 29 November 2003 May and Taylor performed at the 46664 Concert hosted by Nelson Mandela at Green Point Stadium Cape Town in order to raise awareness of the spread of HIV AIDS in South Africa May and Taylor spent time at Mandela s home discussing how Africa s problems might be approached and two years later the band was made ambassadors for the 46664 cause Queen Paul Rodgers 2004–2009 At the end of 2004 May and Taylor announced that they would reunite and return to touring in 2005 with Paul Rodgers founder and former lead singer of Free and Bad Company Brian May s website also stated that Rodgers would be featured with Queen as Queen Paul Rodgers not replacing Mercury The retired John Deacon would not be participating In November 2004 Queen were among the inaugural inductees into the UK Music Hall of Fame and the award ceremony was the first event at which Rodgers joined May and Taylor as vocalist Between 2005 and 2006 Queen Paul Rodgers embarked on a world tour which was the first time Queen toured since their last tour with Freddie Mercury in 1986 The band s drummer Roger Taylor commented We never thought we would tour again Paul came along by chance and we seemed to have a chemistry Paul is just such a great singer He s not trying to be Freddie The first leg was in Europe the second in Japan and the third in the US in 2006 Queen received the inaugural VH1 Rock Honors at the Mandalay Bay Events Center in Las Vegas Nevada on 25 May 2006 The Foo Fighters paid homage to the band in performing Tie Your Mother Down to open the ceremony before being joined on stage by May Taylor and Paul Rodgers who played a selection of Queen hits On 15 August 2006 Brian May confirmed through his website and fan club that Queen Paul Rodgers would begin producing their first studio album beginning in October to be recorded at a secret location Queen Paul Rodgers performed at the Nelson Mandela 90th Birthday Tribute held in Hyde Park London on 27 June 2008 to commemorate Mandela s ninetieth birthday and again promote awareness of the HIV AIDS pandemic The first Queen Paul Rodgers album titled The Cosmos Rocks was released in Europe on 12 September 2008 and in the United States on 28 October 2008 Following the release of the album the band again went on a tour through Europe opening on Kharkiv s Freedom Square in front of 350 000 Ukrainian fans The 12 September concert in Ukraine was later released on DVD The tour then moved to Russia and the band performed two sold-out shows at the Moscow Arena Having completed the first leg of its extensive European tour which saw the band play 15 sold-out dates across nine countries the UK leg of the tour sold out within 90 minutes of going on sale and included three London dates the first of which was The O2 Arena on 13 October The last leg of the tour took place in South America and included a sold-out concert at the Estadio José Amalfitani Buenos Aires Queen and Paul Rodgers officially split up without animosity on 12 May 2009 Rodgers stated My arrangement with was similar to my arrangement with Jimmy Page in The Firm in that it was never meant to be a permanent arrangement Rodgers did not rule out the possibility of working with Queen again Departure from EMI 40th Anniversary 2009–present On 20 May 2009 May and Taylor performed We Are the Champions live on the season finale of American Idol with winner Kris Allen and runner-up Adam Lambert providing a vocal duet In mid-2009 after the split of Queen Paul Rodgers the Queen online website announced a new greatest hits compilation named Absolute Greatest The album was released on 16 November and peaked at number 3 in the official UK Chart The album contains 20 of Queen s biggest hits spanning their entire career and was released in four different formats single disc double disc with commentary double disc with feature book and a vinyl record Prior to its release a competition was run by Queen online to guess the track listing as a promotion for the album On 30 October 2009 May wrote a fanclub letter on his website stating that Queen had no intentions to tour in 2010 but that there was a possibility of a performance He was quoted as saying The greatest debate though is always about when we will next play together as Queen At the moment in spite of the many rumours that are out there we do not have plans to tour in 2010 The good news though is that Roger and I have a much closer mutual understanding these days—privately and professionally and all ideas are carefully considered Music is never far away from us As I write there is an important one-off performance on offer in the USA and it remains to be decided whether we will take up this particular challenge Every day doors seem to open and every day we interact perhaps more than ever before with the world outside It is a time of exciting transition in Rock music and in The Business It s good that the pulse still beats On 15 November 2009 May and Taylor performed Bohemian Rhapsody live on the British TV show The X Factor alongside the finalists On 7 May 2010 May and Taylor announced that they were quitting their record label EMI after almost 40 years On 20 August 2010 Queen s manager Jim Beach put out a Newsletter stating that the band had signed a new contract with Universal Music During an interview for Hardtalk on the BBC on 22 September May confirmed that the band s new deal was with Island Records a subsidiary of Universal For the first time since the late 1980s Queen s catalogue will have the same distributor worldwide as their US home Hollywood Records is currently distributed by Universal for a time in the late 1980s Queen was on EMI-owned Capitol Records in the US On 14 March 2011 which marked the band s 40th anniversary Queen s first five albums were re-released in the UK and some other territories as remastered deluxe editions the US versions were released on 17 May The second five albums of Queen s back catalogue were released worldwide on 27 June with the exception of the US and Canada 27 September The final five were released in the UK on 5 September In May 2011 Jane s Addiction vocalist Perry Farrell noted that Queen are currently scouting their once former and current live bassist Chris Chaney to join the band Farrell stated I have to keep Chris away from Queen who want him and they re not gonna get him unless we re not doing anything Then they can have him In the same month Paul Rodgers stated he may tour with Queen again in the near future At the 2011 Broadcast Music Incorporated BMI Awards held in London on 4 October Queen received the BMI Icon Award in recognition for their airplay success in the US At the 2011 MTV Europe Music Awards on 6 November Queen received the Global Icon Award which Katy Perry presented to Brian May Queen closed the awards ceremony with Adam Lambert on vocals performing The Show Must Go On We Will Rock You and We Are the Champions The collaboration garnered a positive response from both fans and critics resulting in speculation about future projects together In October 2011 it was announced that Queen will be recording a new album featuring lost demos of Mercury on vocals Brian May confirmed that he and Taylor are working their way through the band s old material to compile a selection of unreleased songs for the forthcoming album May also revealed that a series of duets that Mercury recorded with Michael Jackson are to be released in 2012 On 25 and 26 April May and Taylor appeared on the eleventh series of American Idol at the Nokia Theatre Los Angeles performing a Queen medley with the six finalists on the first show and the following day performed Somebody to Love with the Queen Extravaganza band Queen were scheduled to headline Sonisphere at Knebworth on 7 July 2012 with Adam Lambert before the festival was cancelled Queen s final concert with Freddie Mercury was in Knebworth in 1986 Brian May commented It s a worthy challenge for us and I m sure Adam would meet with Freddie s approval Queen expressed disappointment at the cancellation and released a statement to the effect that they were looking to find another venue It was later announced that Queen Adam Lambert would play two shows at the Hammersmith Apollo London on 11 and 12 July 2012 Both shows sold out within 24 hours of tickets going on open sale A third London date was scheduled for 14 July On 30 June Queen Lambert performed in Kiev Ukraine at a joint concert with Elton John for the Elena Pinchuk ANTIAIDS Foundation Queen also performed with Lambert on 3 July 2012 at Moscow s Olympic Stadium and on 7 July 2012 at the Municipal Stadium in Wroclaw Poland On 12 August 2012 Queen performed at the closing ceremony of the 2012 Summer Olympics in London The performance at London s Olympic Stadium opened with a special remastered video clip of Freddie Mercury on stage performing his call and response routine during their 1986 concert at Wembley Stadium Following this May performed part of the Brighton Rock solo before being joined by Taylor and solo artist Jessie J for a performance of We Will Rock You On 20 September 2013 Queen Adam Lambert performed at the iHeartRadio Music Festival at the MGM Grand Hotel Casino in Las Vegas Artistry Musical style The band drew artistic influence from many other British rock acts at the time such as The Beatles Led Zeppelin Pink Floyd The Who Black Sabbath Slade Deep Purple and David Bowie Queen composed music that drew inspiration from many different genres of music often with a tongue-in-cheek attitude The genres they have been associated with include progressive rock glam rock hard rock heavy metal pop rock psychedelic rock blues rock and dance disco Queen also wrote songs that were inspired by genres that are not typically associated with rock such as ragtime opera gospel vaudeville and folk In 1963 the teenage Brian May and his father custom-built his signature guitar Red Special which was purposely designed to feedback Sonic experimentation figured heavily in Queen s songs A distinctive characteristic of Queen s music are the vocal harmonies which are usually composed of the voices of May Mercury and Taylor best heard on the studio albums A Night at the Opera and A Day at the Races Some of the ground work for the development of this sound can be attributed to their former producer Roy Thomas Baker and their engineer Mike Stone Besides vocal harmonies Queen were also known for multi-tracking voices to imitate the sound of a large choir through overdubs For instance according to Brian May there are over 180 vocal overdubs in Bohemian Rhapsody Many Queen songs were also written with audience participation in mind such as We Will Rock You and We Are the Champions Influence Queen have been recognised as having made significant contributions to such genres as hard rock and heavy metal amongst others Hence the band have been cited as an influence by many other musicians Moreover like their music the bands and artists that have claimed to be influenced by Queen are diverse and span different generations countries and genres Some of the musicians that have cited the band as an influence include Anthrax Nirvana Def Leppard Dream Theater Extreme Trivium Foo Fighters Franz Ferdinand George Michael Green Day Guns N Roses Iron Maiden Journey Kansas Katy Perry Keane Lady Gaga Manic Street Preachers Meat Loaf Metallica Mika Muse Mötley Crüe My Chemical Romance Panic at the Disco Queensrÿche Van Halen Radiohead Robbie Williams Trent Reznor Steve Vai Sum 41 Styx Midge Ure The Flaming Lips The Killers and The Smashing Pumpkins Queen have been cited as a major influence on the neo-classical metal genre by Swedish guitarist Yngwie Malmsteen Metallica recorded a cover version of Stone Cold Crazy which first appeared on the album in 1990 and won their first Grammy Award for Best Metal Performance in 1991 In the early 1970s Queen helped spur the heavy metal genre s evolution by discarding much of its blues influence the New Wave of British Heavy Metal in addition they fused the music genre with a punk rock sensibility and an increasing emphasis on speed Legacy In 2002 Queen s Bohemian Rhapsody was voted the UK s favourite hit of all time in a poll conducted by the Guinness World Records British Hit Singles Book and in 2004 the song was inducted into the Grammy Hall of Fame Acclaimed for their stadium rock in 2005 an industry poll ranked Queen s performance at Live Aid in 1985 as the best live act in history In 2007 they were also voted the greatest British band in history by BBC Radio 2 listeners As of 2005 according to the Guinness Book of World Records Queen albums have spent a total of 1 322 weeks twenty-six years on the UK Album Charts more time than any other musical act Also in 2005 with the release of their live album with Paul Rodgers Queen moved into third place on the list of acts with the most aggregate time spent on the British record charts In 2006 the Greatest Hits album was the all-time best-selling album in UK Chart history with sales of 5 407 587 copies over 604 295 more copies than its nearest competitor The Beatles Sgt Pepper s Lonely Hearts Club Band Their Greatest Hits II album is the eighth best seller with sales of 3 746 404 copies The band have released a total of eighteen number one albums eighteen number one singles and ten number one DVDs worldwide making them one of the world s best-selling music artists Queen have sold over 150 million albums with some estimates in excess of 300 million albums worldwide including 34 5 million in the United States alone as of 2004 Inducted into the Rock and Roll Hall of Fame in 2001 the band is also the only group in which every member has composed more than one chart-topping single and all four members of Queen were inducted into the Songwriters Hall of Fame in 2003 In 2009 We Will Rock You and We Are the Champions were inducted into the Grammy Hall of Fame and the latter was voted the world s favourite song in a global music poll Queen are one of the most bootlegged bands ever according to Nick Weymouth who manages the band s official website A 2001 survey discovered the existence of 12 225 websites dedicated to Queen bootlegs the highest number for any band Bootleg recordings have contributed to the band s popularity in certain countries where Western music is censored such as Iran In a project called Queen The Top 100 Bootlegs many of these have been made officially available to download for a nominal fee from Queen s website with profits going to the Mercury Phoenix Trust Rolling Stone ranked Queen at number 52 on its list of the 100 Greatest Artists of All Time while ranking Mercury the 18th greatest singer and May the twenty-sixth greatest guitarist Queen were named 13th on VH1 s 100 Greatest Artists of Hard Rock list and in 2010 were ranked 17th on VH1 s 100 Greatest Artists of All Time list Visitors to Play com voted them as the best music artist of the last 60 years in their list of Top 100 Music Artists in 2013 Gigwise readers chose Queen as the best band of past 60 years In other media Musical theatre In May 2002 a musical or rock theatrical based on the songs of Queen titled We Will Rock You opened at the Dominion Theatre on London s West End The musical was written by British comedian and author Ben Elton in collaboration with Brian May and Roger Taylor and produced by Robert De Niro It has since been staged in many cities around the world Following the Las Vegas premiere on 8 September 2004 Queen were inducted into the Hollywood RockWalk in Sunset Boulevard Los Angeles The original London production was scheduled to close on Saturday 7 October 2006 at the Dominion Theatre but due to public demand the show has now been extended indefinitely We Will Rock You has become the longest running musical ever to run at this prime London theatre overtaking the previous record holder the Grease musical Brian May has confirmed that they are considering writing a sequel to the musical The musical toured around the UK in 2009 playing at Manchester Palace Theatre Sunderland Empire Birmingham Hippodrome Bristol Hippodrome and Edinburgh Playhouse The launch of the musical coincided with Queen Elizabeth II s Golden Jubilee As part of the Jubilee celebrations Brian May performed a guitar solo of God Save the Queen as featured on Queen s A Night at the Opera from the roof of Buckingham Palace The recording of this performance was used as video for the same song on the 30th Anniversary DVD edition of A Night at the Opera Sean Bovim created Queen at the Ballet a tribute to Freddie Mercury which uses Queen s music as a soundtrack for the show s dancers who interpret the stories behind tracks such as Bohemian Rhapsody Radio Ga Ga and Killer Queen Queen s music also appears in the Off-Broadway production Power Balladz most notably the song We Are the Champions with the show s two performers believing the song was the apex of artistic achievement in its day Digital realm In conjunction with Electronic Arts Queen released the computer game in 1998 The music itself—tracks from Queen s vast catalogue in many cases remixed into new instrumental versions—was by and large well received but the game experience was hampered by poor gameplay Adding to the problem was an extremely long development time resulting in graphic elements that already seemed outdated by the time of release Under the supervision of May and Taylor numerous restoration projects have been under way involving Queen s lengthy audio and video catalogue DVD releases of their 1986 Wembley concert titled Live At Wembley Stadium 1982 Milton Keynes concert Queen on Fire – Live at the Bowl and two Greatest Video Hits Volumes 1 and 2 spanning the 1970s and 1980s have seen the band s music remixed into 5 1 and DTS surround sound So far only two of the band s albums A Night at the Opera and The Game have been fully remixed into high-resolution multichannel surround on DVD-Audio A Night at the Opera was re-released with some revised 5 1 mixes and accompanying videos in 2005 for the 30th anniversary of the album s original release CD DVD-Video set In 2007 a Blu-ray edition of Queen s previously released concerts Queen Rock Montreal Live Aid was released marking their first project in 1080p HD Queen have been featured multiple times in the Guitar Hero franchise a cover of Killer Queen in the original Guitar Hero We Are The Champions Fat Bottomed Girls and the Paul Rodgers collaboration C-lebrity in a track pack for Guitar Hero World Tour Under Pressure with David Bowie in Guitar Hero 5 I Want It All in Stone Cold Crazy in and Bohemian Rhapsody in On 13 October 2009 Brian May revealed there was talk going on behind the scenes about a dedicated Queen Rock Band game Queen have also been featured multiple times in the Rock Band franchise a track pack of 10 songs which are compatible with Rock Band Rock Band 2 and Rock Band 3 three of those are also compatible with Lego Rock Band Their hit Bohemian Rhapsody was featured in Rock Band 3 with full harmony and keys support The band also appeared in the video game Lego Rock Band as playable Lego avatars In March 2009 Sony Computer Entertainment released a Queen branded version of the company s karaoke franchise SingStar The game which is available on PlayStation 2 and PlayStation 3 is titled SingStar Queen and has 25 songs on the PS3 and 20 on the PS2 We Will Rock You and other songs by Queen also appear in DJ Hero One Vision was featured on the successful video game Grand Theft Auto IV on the fictional radio station Liberty Rock Radio 97 8 while Radio Ga Ga features on Grand Theft Auto V character trailer for Michael and the game s soundtrack Film and television Queen contributed music directly to the movies Flash Gordon 1980 with Flash as the theme song and Highlander the original 1986 film with A Kind of Magic One Year of Love Who Wants to Live Forever Hammer to Fall and the theme Princes of the Universe which was also used as the theme of the 1992–1998 In the United States Bohemian Rhapsody was re-released as a single in 1992 after appearing in the comedy film Wayne s World The single subsequently reached number two on the Billboard Hot 100 with The Show Must Go On as the first track on the single and helped rekindle the band s popularity in North America Several films have featured their songs performed by other artists A version of Somebody to Love was done by Anne Hathaway in the 2004 film Ella Enchanted In 2006 Brittany Murphy also recorded a cover of the same song for the 2006 movie Happy Feet In 2001 a version of The Show Must Go On was performed by Jim Broadbent and Nicole Kidman in the movie musical Moulin Rouge The closing credits of A Knight s Tale released in 2001 has a version of We Are the Champions performed by Robbie Williams and Queen the introduction to the same movie features We Will Rock You played by the medieval audience We Are the Champions also features in The Mighty Ducks trilogy Chicken Little 2005 and the 2008 film What Happens in Vegas In 1992 the film Gladiator featured snippets of We Will Rock You performed by Warrant whereas their full version was released as a single In 2004 Don t Stop Me Now was featured in the bar fight scene in the cult movie Shaun of the Dead and You re My Best Friend played during the end credits as well as during the 2006 Jennifer Aniston romantic comedy The Break-Up In May 2004 the Japanese live-action version of Sailor Moon called Pretty Guardian Sailor Moon during Act 29 Minako s Rival Mio Kuroki is a Transfer Student used I Was Born To Love You in a volleyball game scene featuring the show s hero Usagi Tsukino Sailor Moon I Was Born to Love You was used as the theme song of the Japanese television drama Pride on Fuji Television in 2004 starring Takuya Kimura and Yūko Takeuchi The show s soundtrack also contained other songs by Queen including We Will Rock You We Are the Champions and Bohemian Rhapsody Since featuring in Grosse Pointe Blank in 1997 Under Pressure has appeared in a number of comedies such as 40 Days and 40 Nights in 2002 and Ben Stiller s 2007 film The Heartbreak Kid Don t Stop Me Now has featured in the BBC television show Top Gear and in 2005 the song was voted as The Greatest Driving Song Ever by the shows viewers The song was also featured in Fox s cartoon series American Dad during the first season episode Roger n Me on Fox on 23 April 2006 This was the first of three appearances of Queen songs the second being Another One Bites the Dust in 42-Year-Old Virgin and Bicycle Race in Jack s Back Keeping in the tradition since Season Five of naming each season s episodes after songs from a famous 1970s era rock band Led Zeppelin for the fifth season The Who for the sixth and The Rolling Stones for the seventh the eighth and final season of That 70s Show consisted of episodes named after Queen songs Bohemian Rhapsody served as the season premiere Fox television show The Simpsons has made storylines which have featured Queen songs such as We Will Rock You We Are the Champions both sung by Homer and You re My Best Friend We Are the Champions has also featured in South Park episode Stanley s Cup season 5 of The Sopranos Tony s ringtone and the first season of Malcolm in the Middle The British teen drama Skins had two Queen songs on its show Escape from the Swamp features on series 2 episode 7 and Don t Stop Me Now on series 3 episode 2 On 11 April 2006 Brian May and Roger Taylor appeared on the American singing contest television show American Idol Each contestant was required to sing a Queen song during that week of the competition Songs which appeared on the show included Bohemian Rhapsody Fat Bottomed Girls The Show Must Go On Who Wants to Live Forever and Innuendo Brian May later criticised the show for editing specific scenes one of which made the group s time with contestant Ace Young look negative despite it being the opposite Taylor and May again appeared on the American Idol Season 8 finale in May 2009 performing We Are the Champions with finalists Adam Lambert and Kris Allen On 15 November 2009 Brian May and Roger Taylor appeared on the singing contest television show X Factor in the UK In the autumn of 2009 the Fox television show Glee featured the fictional high school s show choir singing Somebody to Love as their second act performance in the episode The Rhodes Not Taken The performance was included on the show s soundtrack CD and is available as a single via digital download In June 2010 the choir performed Another One Bites the Dust in the episode Funk In May 2012 the choir performed We Are the Champions in the episode Nationals and the song features in In September 2010 Brian May announced in a BBC interview that Sacha Baron Cohen is to play Mercury in a film of the same name TIME commented with approval on his singing ability and visual similarity to Mercury The motion picture is being written by Peter Morgan who had been nominated for Oscars for his screenplays The Queen and Frost Nixon The film which is being co-produced by Robert De Niro s TriBeCa Productions will focus on Queen s formative years and the period leading up to the celebrated performance at the 1985 Live Aid concert'\n\nent_lines_4EX['Apple'] = '18978754\tApple\tApple The apple is the pomaceous fruit of the apple tree species Malus domestica in the rose family Rosaceae It is one of the most widely cultivated tree fruits and the most widely known of the many members of genus Malus that are used by humans Apples grow on small deciduous trees The tree originated in Central Asia where its wild ancestor Malus sieversii is still found today Apples have been grown for thousands of years in Asia and Europe and were brought to North America by European colonists Apples have been present in the mythology and religions of many cultures including Norse Greek and Christian traditions In 2010 the fruit s genome was decoded leading to new understandings of disease control and selective breeding in apple production There are more than 7 500 known cultivars of apples resulting in a range of desired characteristics Different cultivars are bred for various tastes and uses including cooking fresh eating and cider production Domestic apples are generally propagated by grafting although wild apples grow readily from seed Trees are prone to a number of fungal bacterial and pest problems which can be controlled by a number of organic and non-organic means About 69 million tons of apples were grown worldwide in 2010 and China produced almost half of this total The United States is the second-leading producer with more than 6 of world production Turkey is third followed by Italy India and Poland Apples are often eaten raw but can also be found in many prepared foods especially desserts and drinks Many beneficial health effects are thought to result from eating apples however two forms of allergies are seen to various proteins found in the fruit Botanical information The apple forms a tree that is small and deciduous generally standing tall in cultivation and up to in the wild When cultivated the size shape and branch density is determined by rootstock selection and trimming method The leaves are alternately arranged dark green-colored simple ovals with serrated margins and slightly downy undersides Blossoms are produced in spring simultaneously with the budding of the leaves and are produced on spurs and some long shoots The flowers are white with a pink tinge that gradually fades five petaled with an inflorescence consisting of a cyme with 4–6 flowers The central flower of the inflorescence is called the king bloom it opens first and can develop a larger fruit The fruit matures in autumn and varieties exist with a wide range of sizes Commercial growers aim to produce an apple that is in diameter due to market preference Some consumers especially those in Japan prefer a larger apple while apples below are generally used for making juice and have little fresh market value The skin of ripe apples is generally red yellow green or pink although many bi- or tri-colored varieties may be found The skin may also be wholly or partly russeted i e rough and brown The skin is covered in a protective layer of epicuticular wax The flesh is generally pale yellowish-white though pink or yellow flesh is also known Wild ancestors The original wild ancestor of Malus domestica was Malus sieversii found growing wild in the mountains of Central Asia in southern Kazakhstan Kyrgyzstan Tajikistan and Xinjiang China Cultivation of the species most likely beginning on the forested flanks of the Tian Shan mountains progressed over a long period of time and permitted secondary introgression of genes from other species into the open-pollinated seeds Significant exchange with Malus sylvestris the crabapple resulted in current populations of apples to be more related to crabapples than to the more morphologically similar progenitor Malus sieversii In strains without recent admixture the contribution of the latter predominates Genome In 2010 an Italian-led consortium announced they had decoded the complete genome of the apple in collaboration with horticultural genomicists at Washington State University using the Golden delicious variety It had about 57 000 genes the highest number of any plant genome studied to date and more genes than the human genome about 30 000 This new understanding of the apple genome will help scientists in identifying genes and gene variants that contribute to resistance to disease and drought and other desirable characteristics Understanding the genes behind these characteristics will allow scientists to perform more knowledgeable selective breeding Decoding the genome also provided proof that Malus sieversii was the wild ancestor of the domestic apple—an issue that had been long-debated in the scientific community History The center of diversity of the genus Malus is in eastern Turkey The apple tree was perhaps the earliest tree to be cultivated and its fruits have been improved through selection over thousands of years Alexander the Great is credited with finding dwarfed apples in Kazakhstan in Asia in 328 BCE those he brought back to Macedonia might have been the progenitors of dwarfing root stocks Winter apples picked in late autumn and stored just above freezing have been an important food in Asia and Europe for millennia Apples were brought to North America by colonists in the 17th century and the first apple orchard on the North American continent was planted in Boston by Reverend William Blaxton in 1625 The only apples native to North America are crab apples which were once called common apples Apple varieties brought as seed from Europe were spread along Native American trade routes as well as being cultivated on Colonial farms An 1845 United States apples nursery catalogue sold 350 of the best varieties showing the proliferation of new North American varieties by the early 19th century In the 20th century irrigation projects in Eastern Washington began and allowed the development of the multibillion dollar fruit industry of which the apple is the leading product Until the 20th century farmers stored apples in frostproof cellars during the winter for their own use or for sale Improved transportation of fresh apples by train and road replaced the necessity for storage In the 21st century long-term storage again came into popularity as controlled atmosphere facilities were used to keep apples fresh year-round Controlled atmosphere facilities use high humidity and low oxygen and carbon dioxide levels to maintain fruit freshness Cultural aspects Germanic paganism In Norse mythology the goddess Iðunn is portrayed in the Prose Edda written in the 13th century by Snorri Sturluson as providing apples to the gods that give them eternal youthfulness English scholar H R Ellis Davidson links apples to religious practices in Germanic paganism from which Norse paganism developed She points out that buckets of apples were found in the Oseberg ship burial site in Norway and that fruit and nuts Iðunn having been described as being transformed into a nut in Skáldskaparmál have been found in the early graves of the Germanic peoples in England and elsewhere on the continent of Europe which may have had a symbolic meaning and that nuts are still a recognized symbol of fertility in southwest England Davidson notes a connection between apples and the Vanir a tribe of gods associated with fertility in Norse mythology citing an instance of eleven golden apples being given to woo the beautiful Gerðr by Skírnir who was acting as messenger for the major Vanir god Freyr in stanzas 19 and 20 of Skírnismál Davidson also notes a further connection between fertility and apples in Norse mythology in chapter 2 of the Völsunga saga when the major goddess Frigg sends King Rerir an apple after he prays to Odin for a child Frigg s messenger in the guise of a crow drops the apple in his lap as he sits atop a mound Rerir s wife s consumption of the apple results in a six-year pregnancy and the Caesarean section birth of their son—the hero Völsung Further Davidson points out the strange phrase Apples of Hel used in an 11th-century poem by the skald Thorbiorn Brúnarson She states this may imply that the apple was thought of by Brúnarson as the food of the dead Further Davidson notes that the potentially Germanic goddess Nehalennia is sometimes depicted with apples and that parallels exist in early Irish stories Davidson asserts that while cultivation of the apple in Northern Europe extends back to at least the time of the Roman Empire and came to Europe from the Near East the native varieties of apple trees growing in Northern Europe are small and bitter Davidson concludes that in the figure of Iðunn we must have a dim reflection of an old symbol that of the guardian goddess of the life-giving fruit of the other world Greek mythology Apples appear in many religious traditions often as a mystical or forbidden fruit One of the problems identifying apples in religion mythology and folktales is that the word apple was used as a generic term for all foreign fruit other than berries including nuts as late as the 17th century For instance in Greek mythology the Greek hero Heracles as a part of his Twelve Labours was required to travel to the Garden of the Hesperides and pick the golden apples off the Tree of Life growing at its center The Greek goddess of discord Eris became disgruntled after she was excluded from the wedding of Peleus and Thetis In retaliation she tossed a golden apple inscribed Καλλίστη Kalliste sometimes transliterated Kallisti For the most beautiful one into the wedding party Three goddesses claimed the apple Hera Athena and Aphrodite Paris of Troy was appointed to select the recipient After being bribed by both Hera and Athena Aphrodite tempted him with the most beautiful woman in the world Helen of Sparta He awarded the apple to Aphrodite thus indirectly causing the Trojan War The apple was thus considered in ancient Greece to be sacred to Aphrodite and to throw an apple at someone was to symbolically declare one s love and similarly to catch it was to symbolically show one s acceptance of that love An epigram claiming authorship by Plato states Atalanta also of Greek mythology raced all her suitors in an attempt to avoid marriage She outran all but Hippomenes also known as Melanion a name possibly derived from melon the Greek word for both apple and fruit in general who defeated her by cunning not speed Hippomenes knew that he could not win in a fair race so he used three golden apples gifts of Aphrodite the goddess of love to distract Atalanta It took all three apples and all of his speed but Hippomenes was finally successful winning the race and Atalanta s hand The forbidden fruit in the Garden of Eden Though the forbidden fruit of Eden in the Book of Genesis is not identified popular Christian tradition has held that it was an apple that Eve coaxed Adam to share with her The origin of the popular identification with a fruit unknown in the Middle East in biblical times is found in confusion between the Latin words mālum an apple and mălum an evil each of which is normally written malum The tree of the forbidden fruit is called the tree of the knowledge of good and evil in and the Latin for good and evil is bonum et malum Renaissance painters may also have been influenced by the story of the golden apples in the Garden of Hesperides As a result in the story of Adam and Eve the apple became a symbol for knowledge immortality temptation the fall of man into sin and sin itself The larynx in the human throat has been called Adam s apple because of a notion that it was caused by the forbidden fruit remaining in the throat of Adam The apple as symbol of sexual seduction has been used to imply human sexuality possibly in an ironic vein Cultivars There are more than 7 500 known cultivars of apples Cultivars vary in their yield and the ultimate size of the tree even when grown on the same rootstock Different cultivars are available for temperate and subtropical climates The UK s National Fruit Collection which is the responsibility of the Department of Environment Food and Rural Affairs has a collection of over 2 000 accessions in Kent The University of Reading which is responsible for developing the UK national collection database provides access to search the national collection The University of Reading s work is part of the European Cooperative Programme for Plant Genetic Resources of which there are 38 countries participating in the Malus Pyrus work group The UK s national fruit collection database contains a wealth of information on the characteristics and origin of many apples including alternative names for what is essentially the same genetic apple variety Most of these cultivars are bred for eating fresh dessert apples though some are cultivated specifically for cooking cooking apples or producing cider Cider apples are typically too tart and astringent to eat fresh but they give the beverage a rich flavor that dessert apples cannot Commercially popular apple cultivars are soft but crisp Other desired qualities in modern commercial apple breeding are a colorful skin absence of russeting ease of shipping lengthy storage ability high yields disease resistance common apple shape and developed flavor Modern apples are generally sweeter than older cultivars as popular tastes in apples have varied over time Most North Americans and Europeans favor sweet subacid apples but tart apples have a strong minority following Extremely sweet apples with barely any acid flavor are popular in Asia and especially India Old cultivars are often oddly shaped russeted and have a variety of textures and colors Some find them to have a better flavor than modern cultivars but may have other problems which make them commercially unviable from low yield disease susceptibility or poor tolerance for storage or transport A few old cultivars are still produced on a large scale but many have been preserved by home gardeners and farmers that sell directly to local markets Many unusual and locally important cultivars with their own unique taste and appearance exist apple conservation campaigns have sprung up around the world to preserve such local cultivars from extinction In the United Kingdom old cultivars such as Cox s Orange Pippin and Egremont Russet are still commercially important even though by modern standards they are low yielding and susceptible to disease Cultivation Breeding In the wild apples grow readily from seeds However like most perennial fruits apples are ordinarily propagated asexually by grafting This is because seedling apples are an example of extreme heterozygotes in that rather than inheriting DNA from their parents to create a new apple with those characteristics they are instead significantly different from their parents Triploid varieties have an additional reproductive barrier in that 3 sets of chromosomes cannot be divided evenly during meiosis yielding unequal segregation of the chromosomes aneuploids Even in the case when a triploid plant can produce a seed apples are an example it occurs infrequently and seedlings rarely survive Because apples do not breed true when planted as seeds grafting is generally used to produce new apple trees The rootstock used for the bottom of the graft can be selected to produce trees of a large variety of sizes as well as changing the winter hardiness insect and disease resistance and soil preference of the resulting tree Dwarf rootstocks can be used to produce very small trees less than high at maturity which bear fruit earlier in their life cycle than full size trees Dwarf rootstocks for apple trees can be traced as far back as 300 BC to the area of Persia and Asia Minor Alexander the Great sent samples of dwarf apple trees to Aristotle s Lyceum Dwarf rootstocks became common by the 15th century and later went through several cycles of popularity and decline throughout the world The majority of the rootstocks used today to control size in apples were developed in England in the early 1900s The East Malling Research Station conducted extensive research into rootstocks and today their rootstocks are given an M prefix to designate their origin Rootstocks marked with an MM prefix are Malling-series varieties later crossed with trees of the Northern Spy variety in Merton England Most new apple cultivars originate as seedlings which either arise by chance or are bred by deliberately crossing cultivars with promising characteristics The words seedling pippin and kernel in the name of an apple cultivar suggest that it originated as a seedling Apples can also form bud sports mutations on a single branch Some bud sports turn out to be improved strains of the parent cultivar Some differ sufficiently from the parent tree to be considered new cultivars Since the 1930s the Excelsior Experiment Station at the University of Minnesota has introduced a steady progression of important apples that are widely grown both commercially and by local orchardists throughout Minnesota and Wisconsin Its most important contributions have included Haralson which is the most widely cultivated apple in Minnesota Wealthy Honeygold and Honeycrisp Apples have been acclimatized in Ecuador at very high altitudes where they provide crops twice per year because of constant temperate conditions year-round Pollination Apples are self-incompatible they must cross-pollinate to develop fruit During the flowering each season apple growers often utilize pollinators to carry pollen Honey bees are most commonly used Orchard mason bees are also used as supplemental pollinators in commercial orchards Bumblebee queens are sometimes present in orchards but not usually in enough quantity to be significant pollinators There are four to seven pollination groups in apples depending on climate One cultivar can be pollinated by a compatible cultivar from the same group or close A with A or A with B but not A with C or D Varieties are sometimes classified by the day of peak bloom in the average 30-day blossom period with pollenizers selected from varieties within a 6-day overlap period Maturation and harvest Cultivars vary in their yield and the ultimate size of the tree even when grown on the same rootstock Some cultivars if left unpruned will grow very large which allows them to bear much more fruit but makes harvesting very difficult Depending on the tree density number of trees planted per unit surface area mature trees typically bear of apples each year though productivity can be close to zero in poor years Apples are harvested using three-point ladders that are designed to fit amongst the branches Trees grafted on dwarfing rootstocks will bear about of fruit per year Crops ripen at different times of the year according to the variety of apple Varieties that yield their crop in the summer include Gala Golden Supreme McIntosh Transparent Primate Sweet Bough and Duchess fall producers include Fuji Jonagold Golden Delicious Red Delicious Chenango Gravenstein Wealthy McIntosh Snow and Blenheim winter producers include Winesap Granny Smith King Wagener Swayzie Greening and Tolman Sweet Storage Commercially apples can be stored for some months in controlled-atmosphere chambers to delay ethylene -induced ripening Apples are commonly stored in chambers with higher concentrations of carbon dioxide and high air filtration This prevents ethylene concentrations from rising to higher amounts and preventing ripening from occurring too quickly Ripening continues when the fruit is removed from storage For home storage most varieties of apple can be held for approximately two weeks when kept at the coolest part of the refrigerator i e below 5 °C Some types including the Granny Smith and Fuji can be stored up to a year without significant degradation Pests and diseases Apple trees are susceptible to a number of fungal and bacterial diseases and insect pests Many commercial orchards pursue an aggressive program of chemical sprays to maintain high fruit quality tree health and high yields A trend in orchard management is the use of organic methods These ban the use of some pesticides though some older pesticides are allowed Organic methods include for instance introducing its natural predator to reduce the population of a particular pest A wide range of pests and diseases can affect the plant three of the more common diseases pests are mildew aphids and apple scab Among the most serious disease problems are fireblight a bacterial disease and Gymnosporangium rust and black spot two fungal diseases Codling moths and apple maggots are two other pests which affect apple trees Young apple trees are also prone to mammal pests like mice and deer which feed on the soft bark of the trees especially in winter Production About 69 million tonnes of apples were grown worldwide in 2010 with China producing almost half of this total The United States is the second-leading producer with more than 6 of world production The largest exporters of apples in 2009 were China the U S Turkey Poland Italy Iran and India while the biggest importers in the same year were Russia Germany the UK and the Netherlands In the United States more than 60 of all the apples sold commercially are grown in Washington Imported apples from New Zealand and other more temperate areas are competing with U S production and increasing each year Most of Australia s apple production is for domestic consumption Imports from New Zealand have been disallowed under quarantine regulations for fireblight since 1921 Other countries with a significant production are Russia Ukraine Argentina Germany and Japan Human consumption Apples are often eaten raw The whole fruit including the skin is suitable for human consumption except for the seeds which may affect some consumers The core is often not eaten and is discarded Varieties bred for raw consumption are termed dessert or table apples Apples can be canned or juiced They are milled or pressed to produce apple juice which may be drunk unfiltered called apple cider in North America or filtered The juice can be fermented to make cider called hard cider in North America ciderkin and vinegar Through distillation various alcoholic beverages can be produced such as applejack Calvados and apfelwein Apple seed oil and pectin may also be produced Popular uses Apples are an important ingredient in many desserts such as apple pie apple crumble apple crisp and apple cake They are often eaten baked or stewed and they can also be dried and eaten or reconstituted soaked in water alcohol or some other liquid for later use Puréed apples are generally known as apple sauce Apples are also made into apple butter and apple jelly They are also used cooked in meat dishes Sliced apples turn brown with exposure to air due to the conversion of natural phenolic substances into melanin upon exposure to oxygen Different cultivars vary in their propensity to brown after slicing Sliced fruit can be treated with acidulated water to prevent this effect Organic production Organic apples are commonly produced in the United States Organic production is difficult in Europe though a few orchards have done so with commercial success using disease-resistant cultivars A light coating of kaolin which forms a physical barrier to some pests also helps prevent apple sun scalding Allergy One form of apple allergy often found in northern Europe is called birch-apple syndrome and is found in people who are also allergic to birch pollen Allergic reactions are triggered by a protein in apples that is similar to birch pollen and people affected by this protein can also develop allergies to other fruits nuts and vegetables Reactions which entail oral allergy syndrome OAS generally involve itching and inflammation of the mouth and throat but in rare cases can also include life-threatening anaphylaxis This reaction only occurs when raw fruit is consumed—the allergen is neutralized in the cooking process The variety of apple maturity and storage conditions can change the amount of allergen present in individual fruits Long storage times can increase the amount of proteins that cause birch-apple syndrome In other areas such as the Mediterranean some individuals have adverse reactions to apples because of their similarity to peaches This form of apple allergy also includes OAS but often has more severe symptoms such as vomiting abdominal pain and urticaria and can be life-threatening Individuals with this form of allergy can also develop reactions to other fruits and nuts Cooking does not break down the protein causing this particular reaction so affected individuals can eat neither raw nor cooked apples Freshly harvested over-ripe fruits tend to have the highest levels of the protein that causes this reaction Breeding efforts have yet to produce a hypoallergenic fruit suitable for either of the two forms of apple allergy Toxicity of seeds The seeds of apples contain small amounts of amygdalin a sugar and cyanide compound known as a cyanogenic glycoside Ingesting small amounts of apple seeds will cause no ill effects but in extremely large doses can cause adverse reactions There is only one known case of fatal cyanide poisoning from apple seeds in this case the individual chewed and swallowed one cup of seeds It may take several hours before the poison takes effect as cyanogenic glycosides must be hydrolyzed before the cyanide ion is released Nutrition The proverb An apple a day keeps the doctor away addressing the health effects of the fruit dates from 19th century Wales Preliminary research suggests that apples may reduce the risk of colon cancer prostate cancer and lung cancer Apple peels contain ursolic acid which in rat studies increases skeletal muscle and brown fat and decreases white fat obesity glucose intolerance and fatty liver disease According to the United States Department of Agriculture a typical apple serving weighs 242 grams and contains 126 calories with significant dietary fiber and vitamin C content Apple peels are a source of various phytochemicals with unknown nutritional value and possible antioxidant activity in vitro The predominant phenolic phytochemicals in apples are quercetin epicatechin and procyanidin B2 Apple juice concentrate has been found in mice to increase the production of the neurotransmitter acetylcholine Other studies have shown an alleviation of oxidative damage and cognitive decline in mice after the administration of apple juice Fruit flies fed an apple extract lived 10 longer than other flies fed a normal diet'\n\nent_lines_4EX['Apple Inc.'] = '856\tApple Inc.\tApple Inc Apple Inc is an American multinational corporation headquartered in Cupertino California that designs develops and sells consumer electronics computer software and personal computers Its best-known hardware products are the Mac line of computers the iPod media player the iPhone smartphone and the iPad tablet computer Its consumer software includes the OS X and iOS operating systems the iTunes media browser the Safari web browser and the iLife and iWork creativity and productivity suites Apple was founded by Steve Jobs Steve Wozniak and Ronald Wayne on April 1 1976 to develop and sell personal computers It was incorporated as Apple Computer Inc on January 3 1977 and was renamed as Apple Inc on January 9 2007 to reflect its shifted focus towards consumer electronics Apple is the world s second-largest information technology company by revenue after Samsung Electronics and the world s third-largest mobile phone maker after Samsung and Nokia Fortune magazine named Apple the most admired company in the United States in 2008 and in the world from 2008 to 2012 On September 30 2013 Apple surpassed Coca-Cola to become the world s most valuable brand in the Omnicom Group s Best Global Brands report However the company has received criticism for its contractors labor practices and for Apple s own environmental and business practices As of May 2013 Apple maintains 408 retail stores in fourteen countries as well as the online Apple Store and iTunes Store the latter of which is the world s largest music retailer Apple is the largest publicly traded corporation in the world by market capitalization with an estimated value of US 415 billion as of March 2013 As of September 29 2012 the company had 72 800 permanent full-time employees and 3 300 temporary full-time employees worldwide Its worldwide annual revenue in 2012 totalled 156 billion In May 2013 Apple entered the top ten of the Fortune 500 list of companies for the first time rising 11 places above its 2012 ranking to take the sixth position History 1976–80 Founding and incorporation Apple was established on April 1 1976 by Steve Jobs Steve Wozniak and Ronald Wayne to sell the Apple I personal computer kit a computer single handedly designed by Wozniak The kits were hand-built by Wozniak and first shown to the public at the Homebrew Computer Club The Apple I was sold as a motherboard with CPU RAM and basic textual-video chips which is less than what is today considered a complete personal computer The Apple I went on sale in July 1976 and was market-priced at 666 66 in dollars adjusted for inflation Apple was incorporated January 3 1977 without Wayne who sold his share of the company back to Jobs and Wozniak for US 800 Multi-millionaire Mike Markkula provided essential business expertise and funding of 250 000 during the incorporation of Apple During the first five years of operations revenues doubled every four months an average growth rate of 700 The Apple II also invented by Wozniak was introduced on April 16 1977 at the first West Coast Computer Faire It differed from its major rivals the TRS-80 and Commodore PET due to its character cell-based color graphics and an open architecture While early models used ordinary cassette tapes as storage devices they were superseded by the introduction of a 5 1 4 inch floppy disk drive and interface the Disk II The Apple II was chosen to be the desktop platform for the first killer app of the business world VisiCalc a spreadsheet program VisiCalc created a business market for the Apple II and gave home users compatibility with the office an additional reason to buy an Apple II Apple was a distant third place to Commodore and Tandy until VisiCalc came along By the end of the 1970s Apple had a staff of computer designers and a production line The company introduced the Apple III in May 1980 in an attempt to compete with IBM and Microsoft in the business and corporate computing market Jobs and several Apple employees including Jef Raskin visited Xerox PARC in December 1979 to see the Xerox Alto Xerox granted Apple engineers three days of access to the PARC facilities in return for the option to buy 100 000 shares 800 000 split-adjusted shares of Apple at the pre-IPO price of 10 a share Jobs was immediately convinced that all future computers would use a graphical user interface GUI and development of a GUI began for the Apple Lisa On December 12 1980 Apple went public at 22 per share generating more capital than any IPO since Ford Motor Company in 1956 and instantly creating more millionaires about 300 than any company in history 1981–89 Success with Macintosh Apple began working on the Apple Lisa in 1978 In 1982 Jobs was pushed from the Lisa team due to infighting Jobs took over Jef Raskin s low-cost-computer project the Macintosh A race broke out between the Lisa team and the Macintosh team over which product would ship first Lisa won the race in 1983 and became the first personal computer sold to the public with a GUI but was a commercial failure due to its high price tag and limited software titles In 1984 Apple next launched the Macintosh Its debut was announced by the now famous 1 5 million television commercial 1984 It was directed by Ridley Scott and was aired during the third quarter of Super Bowl XVIII on January 22 1984 It is now hailed as a watershed event for Apple s success and a masterpiece The Macintosh initially sold well but follow-up sales were not strong due to its high price and limited range of software titles The Macintosh was the first personal computer to be sold without a programming language at all The machine s fortunes changed with the introduction of the LaserWriter the first PostScript laser printer to be sold at a reasonable price and PageMaker an early desktop publishing package It has been suggested that the combination of these three products was responsible for the creation of the desktop publishing market The Mac was particularly powerful in the desktop publishing market due to its advanced graphics capabilities which had necessarily been built in to create the intuitive Macintosh GUI In 1985 a power struggle developed between Jobs and CEO John Sculley who had been hired two years earlier The Apple board of directors instructed Sculley to contain Jobs and limit his ability to launch expensive forays into untested products Rather than submit to Sculley s direction Jobs attempted to oust him from his leadership role at Apple Sculley found out that Jobs had been attempting to organize a coup and called a board meeting at which Apple s board of directors sided with Sculley and removed Jobs from his managerial duties Jobs resigned from Apple and founded NeXT Inc the same year The Macintosh Portable was introduced in 1989 and was designed to be just as powerful as a desktop Macintosh but weighed a bulky with a 12-hour battery life After the Macintosh Portable Apple introduced the PowerBook in 1991 The same year Apple introduced System 7 a major upgrade to the operating system which added color to the interface and introduced new networking capabilities It remained the architectural basis for Mac OS until 2001 The success of the PowerBook and other products brought increasing revenue For some time Apple was doing incredibly well introducing fresh new products and generating increasing profits in the process The magazine MacAddict named the period between 1989 and 1991 as the first golden age of the Macintosh Following the success of the Macintosh LC Apple introduced the Centris line a low-end Quadra and the ill-fated Performa line that was sold with an overwhelming number of configurations and software bundles to avoid competing with the various consumer outlets such as Sears Price Club and Wal-Mart the primary dealers for these models Consumers ended up confused and did not understand the difference between models 1990–99 Decline and restructuring During this time Apple experimented with a number of other failed consumer targeted products including digital cameras portable CD audio players speakers video consoles and TV appliances Enormous resources were also invested in the problem-plagued Newton division based on John Sculley s unrealistic market forecasts Ultimately none of these products helped as Apple s market share and stock prices continued to slide Apple saw the Apple II series as too expensive to produce while taking away sales from the low-end Macintosh In 1990 Apple released the Macintosh LC with a single expansion slot for the Apple IIe Card to migrate Apple II users to the Macintosh platform Apple stopped selling the Apple IIe in 1993 Microsoft continued to gain market share with Windows focusing on delivering software to cheap commodity personal computers while Apple was delivering a richly engineered but expensive experience Apple relied on high profit margins and never developed a clear response Instead they sued Microsoft for using a graphical user interface similar to the Apple Lisa in Apple Computer Inc v Microsoft Corporation The lawsuit dragged on for years before it was finally dismissed At the same time a series of major product flops and missed deadlines sullied Apple s reputation and Sculley was replaced as CEO by Michael Spindler By the early 1990s Apple was developing alternative platforms to the Macintosh such as the A UX Apple had also begun to experiment in providing a Mac-only online portal which they called eWorld developed in collaboration with America Online and designed as a Mac-friendly alternative to other online services such as CompuServe The Macintosh platform was itself becoming outdated because it was not built for multitasking and several important software routines were programmed directly into the hardware In addition Apple was facing competition from OS 2 and UNIX vendors such as Sun Microsystems The Macintosh would need to be replaced by a new platform or reworked to run on more powerful hardware In 1994 Apple allied with IBM and Motorola in the AIM alliance The goal was to create a new computing platform the PowerPC Reference Platform which would use IBM and Motorola hardware coupled with Apple s software The AIM alliance hoped that PReP s performance and Apple s software would leave the PC far behind thus countering Microsoft The same year Apple introduced the Power Macintosh the first of many Apple computers to use Motorola s PowerPC processor In 1996 Michael Spindler was replaced by Gil Amelio as CEO Gil Amelio made many changes at Apple including extensive layoffs After numerous failed attempts to improve Mac OS first with the Taligent project then later with Copland and Gershwin Amelio chose to purchase NeXT and its NeXTSTEP operating system bringing Steve Jobs back to Apple as an advisor On July 9 1997 Gil Amelio was ousted by the board of directors after overseeing a three-year record-low stock price and crippling financial losses Jobs became the interim CEO and began restructuring the company s product line At the 1997 Macworld Expo Steve Jobs announced that Apple would join Microsoft to release new versions of Microsoft Office for the Macintosh and that Microsoft made a 150 million investment in non-voting Apple stock On November 10 1997 Apple introduced the Apple Online Store tied to a new build-to-order manufacturing strategy On August 15 1998 Apple introduced a new all-in-one computer reminiscent of the Macintosh 128K the iMac The iMac design team was led by Jonathan Ive who would later design the iPod and the iPhone The iMac featured modern technology and a unique design and sold almost 800 000 units in its first five months Through this period Apple purchased several companies to create a portfolio of professional and consumer-oriented digital production software In 1998 Apple announced the purchase of Macromedia s Final Cut software signaling its expansion into the digital video editing market The following year Apple released two video editing products iMovie for consumers and for professionals Final Cut Pro which has gone on to be a significant video-editing program with 800 000 registered users in early 2007 In 2002 Apple purchased Nothing Real for their advanced digital compositing application Shake as well as Emagic for their music productivity application Logic which led to the development of their consumer-level GarageBand application iPhoto s release the same year completed the iLife suite 2000–06 Return to profitability Mac OS X based on NeXT s OPENSTEP and BSD Unix was released on March 24 2001 after several years of development Aimed at consumers and professionals alike Mac OS X aimed to combine the stability reliability and security of Unix with the ease of use afforded by an overhauled user interface To aid users in migrating from Mac OS 9 the new operating system allowed the use of OS 9 applications through Mac OS X s Classic environment On May 19 2001 Apple opened the first official Apple Retail Stores in Virginia and California On July 9 they bought Spruce Technologies a DVD authoring company On October 23 of the same year Apple announced the iPod portable digital audio player and started selling it on November 10 The product was phenomenally successful — over 100 million units were sold within six years In 2003 Apple s iTunes Store was introduced offering online music downloads for 0 99 a song and integration with the iPod The service quickly became the market leader in online music services with over 5 billion downloads by June 19 2008 Since 2001 Apple s design team has progressively abandoned the use of translucent colored plastics first used in the iMac G3 This began with the titanium PowerBook and was followed by the white polycarbonate iBook and the flat-panel iMac At the Worldwide Developers Conference keynote address on June 6 2005 Steve Jobs announced that Apple would begin producing Intel -based Mac computers in 2006 On January 10 2006 the new MacBook Pro and iMac became the first Apple computers to use Intel s Core Duo CPU By August 7 2006 Apple had transitioned the entire Mac product line to Intel chips over one year sooner than announced The Power Mac iBook and PowerBook brands were retired during the transition the Mac Pro MacBook and MacBook Pro became their respective successors On April 29 2009 The Wall Street Journal reported that Apple was building its own team of engineers to design microchips Apple introduced Boot Camp to help users install Windows XP or Windows Vista on their Intel Macs alongside Mac OS X Apple s success during this period was evident in its stock price Between early 2003 and 2006 the price of Apple s stock increased more than tenfold from around 6 per share split-adjusted to over 80 In January 2006 Apple s market cap surpassed that of Dell Nine years prior Dell s CEO Michael Dell said that if he ran Apple he would shut it down and give the money back to the shareholders Although Apple s market share in computers had grown it remained far behind competitors using Microsoft Windows with only about 8 of desktops and laptops in the US 2007–10 Success with mobile devices Apple achieved widespread success with its iPhone iPod Touch and iPad products which introduced innovations in mobile phones portable music players and personal computers respectively In addition the implementation of a store for the purchase of software applications represented a new business model Touch screens had been invented and seen in mobile devices before but Apple was the first to achieve mass market adoption of such a user interface that included particular pre-programmed touch gestures Delivering his keynote speech at the Macworld Expo on January 9 2007 Jobs announced that Apple Computer Inc would from that point on be known as Apple Inc because computers were no longer the main focus of the company which had shifted its emphasis to mobile electronic devices The event also saw the announcement of the iPhone and the Apple TV In an article posted on Apple s website on February 6 2007 Steve Jobs wrote that Apple would be willing to sell music on the iTunes Store without digital rights management DRM which would allow tracks to be played on third-party players if record labels would agree to drop the technology On April 2 2007 Apple and EMI jointly announced the removal of DRM technology from EMI s catalog in the iTunes Store effective in May Other record labels followed later that year In July of the following year Apple launched the App Store to sell third-party applications for the iPhone and iPod Touch Within a month the store sold 60 million applications and brought in 1 million daily on average with Jobs speculating that the App Store could become a billion-dollar business for Apple Three months later it was announced that Apple had become the third-largest mobile handset supplier in the world due to the popularity of the iPhone On December 16 2008 Apple announced that after over 20 years of attending Macworld 2009 would be the last year Apple would be attending the Macworld Expo and that Phil Schiller would deliver the 2009 keynote in lieu of the expected Jobs Almost exactly one month later on January 14 2009 an internal Apple memo from Jobs announced that he would be taking a six-month leave of absence until the end of June 2009 to allow him to better focus on his health and to allow the company to better focus on its products without having the rampant media speculating about his health Despite Jobs absence Apple recorded its best non-holiday quarter Q1 FY 2009 during the recession with a revenue of 8 16 billion and a profit of 1 21 billion After years of speculation and multiple rumored leaks Apple announced a large screen tablet-like media device known as the iPad on January 27 2010 The iPad runs the same touch based operating system that the iPhone uses and many of the same iPhone apps are compatible with the iPad This gave the iPad a large app catalog on launch even with very little development time before the release Later that year on April 3 2010 the iPad was launched in the US and sold more than 300 000 units on that day reaching 500 000 by the end of the first week In May of the same year Apple s market cap exceeded that of competitor Microsoft for the first time since 1989 Apple released the iPhone 4 which introduced video calling multitasking and a new uninsulated stainless steel design which acts as the phone s antenna Because of this antenna implementation some iPhone 4 users reported a reduction in signal strength when the phone is held in specific ways After a large amount of media coverage including mainstream news organizations Apple held a press conference where they offered buyers a free rubber bumper case which had been proven to eliminate the signal reduction issue Later that year Apple again refreshed its iPod line of MP3 players which introduced a multi-touch iPod Nano iPod Touch with FaceTime and iPod Shuffle with buttons which brought back the buttons of earlier generations In October 2010 Apple shares hit an all-time high eclipsing 300 Additionally on October 20 Apple updated their MacBook Air laptop iLife suite of applications and unveiled Mac OS X Lion the last version with the name Mac OS X On January 6 2011 the company opened their Mac App Store a digital software distribution platform similar to the existing iOS App Store Apple was featured in the documentary Something Ventured which premiered in 2011 2011–12 Steve Jobs death On January 17 2011 Jobs announced in an internal Apple memo that he would take another medical leave of absence for an indefinite period to allow him to focus on his health Chief operating officer Tim Cook assumed Jobs day-to-day operations at Apple although Jobs would still remain involved in major strategic decisions for the company Apple became the most valuable consumer-facing brand in the world In June 2011 Steve Jobs surprisingly took the stage and unveiled iCloud an online storage and syncing service for music photos files and software which replaced MobileMe Apple s previous attempt at content syncing This would be the last product launch Jobs would attend before his death It has been argued that Apple has achieved such efficiency in its supply chain that the company operates as a monopsony one buyer many sellers in that it can dictate terms to its suppliers In July 2011 due to the American debt-ceiling crisis Apple s financial reserves were briefly larger than those of the U S Government On August 24 2011 Jobs resigned his position as CEO of Apple He was replaced by Tim Cook and Jobs became Apple s chairman Prior to this Apple did not have a chairman and instead had two co-lead directors Andrea Jung and Arthur D Levinson who continued with those titles until Levinson became Chairman of the Board in November On October 4 2011 Apple announced the iPhone 4S which included an improved camera with 1080p video recording a dual core A5 chip capable of 7 times faster graphics than the A4 an intelligent software assistant named Siri and cloud-sourced data with iCloud The iPhone 4S was officially released on October 14 2011 On October 5 2011 Apple announced that Jobs had died marking the end of an era for Apple Inc On October 29 2011 Apple purchased C3 Technologies a mapping company for 240 million becoming the third mapping company Apple has purchased On January 10 2012 Apple paid 500 million to acquire Anobit an Israeli hardware company that developed and supplies a proprietary memory signal processing technology that improves the performance of flash-memory used in iPhones and iPads On January 19 2012 Apple s Phil Schiller introduced iBooks Textbooks for iOS and iBook Author for Mac OS X in New York City This was the first major announcement by Apple since the passing of Steve Jobs who stated in his biography that he wanted to reinvent the textbook and education The third-generation iPad was announced on March 7 2012 It includes a Retina display a new CPU a five megapixel camera and 1080p video recording On July 24 2012 during a conference call with investors Tim Cook said that he loved India but that Apple was going to expect larger opportunities outside of India citing the reason as the 30 sourcing requirement from India On August 20 2012 Apple s rising stock rose the company s value to a world-record 624 billion This beat the non-inflation-adjusted record for market capitalization set by Microsoft in 1999 On August 24 2012 a US jury ruled that Samsung should pay Apple 1 05 billion £665m in damages in an intellectual property lawsuit Samsung said they will appeal the court ruling Samsung subsequently prevailed on its motion to vacate this damages award which the Court reduced by 450 million The Court further granted Samsung s request for a new trial On September 12 2012 Apple unveiled the iPhone 5 featuring an enlarged screen more powerful processors and running iOS 6 The latter includes a new mapping application replacing Google Maps that has attracted some criticism It was made available on September 21 2012 and became Apple s biggest iPhone launch with over 2 million pre-orders pushing back the delivery date to late October On October 23 2012 Apple unveiled the iPad Mini which features a 7 9-inch screen in contrast to the iPad s 9 7-inch screen Apple also released a third-generation 13-inch MacBook Pro with a Retina display the fourth-generation iPad featuring a faster processor and a Lightning dock connector and new iMac and Mac Mini computers After the launch of Apple s iPad Mini and fourth generation iPad on November 3 2012 Apple announced that they had sold 3 million iPads in three days of the launch but it did not mention the sales figures of specific iPad models On November 10 2012 Apple confirmed a global settlement that would dismiss all lawsuits between Apple and HTC up to that date in favor of a ten-year license agreement for current and future patents between the two companies It is predicted that Apple will make 280 million a year from this deal with HTC In December 2012 in a TV interview for NBC s Rock Center and also aired on the Today morning show Apple CEO Tim Cook said that in 2013 the company will produce one of its existing lines of Mac computers in the United States In January 2013 Cook stated that he expected China to overtake the US as Apple s biggest market 2013–present Acquisitions and expansions In March 2013 Apple announced a patent for an augmented reality AR system that can identify objects in a live video stream and present information corresponding to these objects through a computer-generated information layer overlaid on top of the real-world image At the Worldwide Developer s Conference on June 10 2013 Apple announced the seventh iOS operating system alongside OS X Mavericks the tenth version of Mac OS X and a new Internet radio service called iTunes Radio iOS 7 and OS X Mavericks are both expected to be released during fall 2013 while the iTunes Radio Service which will be integrated with Apple s personal voice-assistant software program Siri is scheduled for release in the second half of 2013 The radio service features more than 200 stations according to company s statement On July 2 2013 Apple announced the recruitment of Paul Deneve Belgian President and CEO of Yves Saint Laurent to Apple s top ranks A spokesperson for the company stated We re thrilled to welcome Paul Deneve to Apple He ll be working on special projects as a vice president reporting directly to Tim Cook Alongside Google vice-president Vint Cerf and AT T CEO Randall Stephenson Cook attended a closed-door summit held by President Obama on August 8 2013 in regard to government surveillance and the Internet in the wake of the Edward Snowden NSA incident A report on August 22 2013 confirmed that Apple acquired Embark Inc a small Silicon Valley-based mapping company Embark builds free transit apps to help smartphone users navigate public transportation in U S cities such as New York San Francisco and Chicago Following the confirmation of the acquisition an Apple spokesperson explained Apple buys smaller technology companies from time to time and we generally do not discuss our purpose or plans In November 2012 Embark claimed that over 500 000 people used its apps An anonymous Apple employee revealed to the Bloomberg media publication that the opening of a Tokyo Japan store is planned for 2014 The construction of the store will be completed in February 2014 but as of August 29 2013 Takashi Takebayashi a Tokyo-based spokesman for Apple has not made any comment to the media A Japanese analyst has stated For Apple the Japanese market is appealing in terms of quantity and price There is room to expand tablet sales and a possibility the Japanese market expands if Apple’s mobile carrier partners increase On October 1 2013 Apple India executives unveiled a plan to expand further into the Indian market following Cook s acknowledgment of the country in July 2013 when sales results showed that iPhone sales in India grew 400 during the second quarter of 2013 In attendance at the confidential meeting were 20 CEOs and senior executives from telecom and electronic retail companies A mid-October 2013 announcement revealed that Burberry executive Angela Ahrendts will commence as a senior vice president at Apple in mid-2014 Ahrendts oversaw Burberry s digital strategy for almost eight years and during her tenure sales increased to about US 3 2 billion 2 billion pounds and shares gained more than threefold In a company wide memo sent on the morning of October 15 2013 Cook explained the decision to hire Ahrendts She shares our values and our focus on innovation She places the same strong emphasis as we do on the customer experience She cares deeply about people and embraces our view that our most important resource and our soul is our people She believes in enriching the lives of others and she is wicked smart On November 24 2013 Apple Inc confirmed the purchase of PrimeSense an Israeli 3D sensing company based in Tel Aviv In December 2013 Apple Inc purchased social analytics firm Topsy one of a small number of firms with real-time access to the messages that appear on Twitter every tweet published since 2006 is within its scope Debra Aho Williamson an analyst with EMarketer Inc explained “A key point is they are one of the few companies that has access to the Twitter fire hose and can do real-time analysis of the trends and discussions happening on Twitter ” While an exact amount is still unknown the deal was apparently worth more than US 200 million according to people with knowledge of the secret deal and Apple spokespeople have refused to disclose the purpose of the acquisition On December 6 2013 Apple Inc launched iBeacon technology across its 254 U S retail stores Using Bluetooth wireless technology iBeacon senses the user s exact location within the Apple store and sends the user messages about products events and other information tailored to the user s location iBeacon works as long as the user has downloaded the Apple Store app and has given Apple permission to track them Products Mac Apple sells a variety of computer accessories for Macs including Thunderbolt Display Magic Mouse Magic Trackpad Wireless Keyboard Battery Charger the AirPort wireless networking products and Time Capsule iPad On January 27 2010 Apple introduced their much-anticipated media tablet the iPad running a modified version of iOS It offers multi-touch interaction with multimedia formats including newspapers magazines ebooks textbooks photos movies videos of TV shows music word processing documents spreadsheets videogames and most existing iPhone apps It also includes a mobile version of Safari for web browsing as well as access to the App Store iTunes Library iBookstore contacts and notepad Content is downloadable via Wi-Fi and optional 3G service or synced through the user s computer AT T was initially the sole US provider of 3G wireless access for the iPad On March 2 2011 Apple introduced the iPad 2 which had a faster processor and a camera on the front and back It also added support for optional 3G service provided by Verizon in addition to the existing offering by AT T However the availability of the iPad 2 has been limited as a result of the devastating earthquake and ensuing tsunami in Japan in March 2011 On March 7 2012 Apple introduced the third-generation iPad marketed as the new iPad It added LTE service from AT T or Verizon the upgraded A5X processor and the Retina display 2048 by 1536 resolution originally implemented on the iPhone 4 and iPhone 4S The dimensions and form factor remained relatively unchanged with the new iPad being a fraction thicker and heavier than the previous version and minor positioning changes On October 23 2012 Apple s fourth-generation iPad came out marketed as the iPad with Retina display It added the upgraded A6X processor and replaced the traditional 30-pin dock connector with the all-digital Lightning connector The iPad Mini was also introduced with a reduced 7 9-inch display and featuring much of the same internal specifications as the iPad 2 On October 22 2013 Apple introduced the iPad Air It added the new 64 bit Apple-A7 processor The iPad mini with Retina Display was also introduced featuring the Apple-A7 processor as well Since its launch iPad users have downloaded three billion apps while the total number of App Store downloads is over 25 billion iPod On October 23 2001 Apple introduced the iPod digital music player Several updated models have since been introduced and the iPod brand is now the market leader in portable music players by a significant margin with more than 350 million units shipped Apple has partnered with Nike to offer the Nike iPod Sports Kit enabling runners to synchronize and monitor their runs with iTunes and the Nike website Apple currently sells four variants of the iPod iPhone At the Macworld Conference Expo in January 2007 Steve Jobs introduced the long-anticipated iPhone a convergence of an Internet-enabled smartphone and iPod The first-generation iPhone was released on June 29 2007 for 499 4 GB and 599 8 GB with an AT T contract On February 5 2008 it was updated to have 16 GB of memory in addition to the 8 GB and 4 GB models It combined a 2 5G quad band GSM and EDGE cellular phone with features found in handheld devices running scaled-down versions of Apple s Mac OS X dubbed iPhone OS later renamed iOS with various Mac OS X applications such as Safari and Mail It also includes web-based and Dashboard apps such as Google Maps and Weather The iPhone features a touchscreen display Bluetooth and Wi-Fi both b and g At Worldwide Developers Conference WWDC on June 9 2008 Apple announced the iPhone 3G It was released on July 11 2008 with a reduced price of 199 for the 8 GB version and 299 for the 16 GB version This version added support for 3G networking and assisted-GPS navigation The flat silver back and large antenna square of the original model were eliminated in favor of a curved glossy black or white back Software capabilities were improved with the release of the App Store providing applications for download that were compatible with the iPhone On April 24 2009 the App Store surpassed one billion downloads At WWDC on June 8 2009 Apple announced the iPhone 3GS It provided an incremental update to the device including faster internal components support for faster 3G speeds video recording capbility and voice control At WWDC on June 7 2010 Apple announced the redesigned iPhone 4 It features a 960x640 display the Apple A4 processor also used in the iPad a gyroscope for enhanced gaming 5MP camera with LED flash front-facing VGA camera and FaceTime video calling Shortly after its release reception issues were discovered by consumers due to the stainless steel band around the edge of the device which also serves as the phone s cellular signal and Wi-Fi antenna The issue was corrected by a Bumper Case distributed by Apple for free to all owners for a few months In June 2011 Apple overtook Nokia to become the world s biggest smartphone maker by volume On October 4 2011 Apple unveiled the iPhone 4S which was first released on October 14 2011 It features the Apple A5 processor and is the first model offered by Sprint joining AT T and Verizon Wireless as the United States carriers offering iPhone models On October 19 2011 Apple announced an agreement with C Spire Wireless to sell the iPhone 4S with that carrier in the near future marking the first time the iPhone was officially supported on a regional carrier s network Another notable feature of the iPhone 4S was Siri voice assistant technology which Apple had acquired in 2010 as well as other features including an updated 8MP camera with new optics Apple sold 4 million iPhone 4S phones in the first three days of availability On September 12 2012 Apple introduced the iPhone 5 It added a 4-inch display 4G LTE connectivity and the upgraded Apple A6 chip among several other improvements Two million iPhones were sold in the first twenty-four hours of pre-ordering and over five million handsets were sold in the first three days of its launch A patent filed in July 2013 revealed the development of a new iPhone battery system that uses location data in combination with data on the user s habits to moderate the handsets power settings accordingly Apple is working towards a power management system that will provide features such as the ability of the iPhone to estimate the length of time a user will be away from a power source to modify energy usage and a detection function that adjusts the charging rate to best suit the type of power source that is being used In March 2013 one of the largest cellular phone companies in the United States T-Mobile announced that it would begin selling the iPhone 5 on April 12 The announcement of the iPhone came with the announcement that the company would begin implementing 4G cellular service for its users Upon the launch of the iPhone 5S and iPhone 5C Apple sold over nine million devices in the first three days of its launch which sets a new record for first-weekend smartphone sales This was the first time that Apple has simultaneously launched two models and the inclusion of China in the list of markets contributed to the record sales result On October 15 2013 U S Cellular the United States fifth largest cell phone provider announced that it would in fact begin to carry the iPhone It is the last of the five major carriers including AT T Verizon Sprint and T-Mobile to acquire the phone The phone went on sale on November 8 at U S Cellular stores around the country The finalization of a deal between Apple and China Mobile the world s largest mobile network was announced in late December 2013 The multi-year agreement provides iPhone access to over 760 million China Mobile subscribers Apple TV At the 2007 Macworld conference Jobs demonstrated the Apple TV previously known as the iTV a set-top video device intended to bridge the sale of content from iTunes with high-definition televisions The device links up to a user s TV and syncs either via Wi-Fi or a wired network with one computer s iTunes library and streams from an additional four The Apple TV originally incorporated a 40 GB hard drive for storage includes outputs for HDMI and component video and plays video at a maximum resolution of 720p On May 31 2007 a 160 GB drive was released alongside the existing 40 GB model and on January 15 2008 a software update was released which allowed media to be purchased directly from the Apple TV In September 2009 Apple discontinued the original 40 GB Apple TV and now continues to produce and sell the 160 GB Apple TV On September 1 2010 alongside the release of the new line of iPod devices for the year Apple released a completely redesigned Apple TV The new device is 1 4 the size runs quieter and replaces the need for a hard drive with media streaming from any iTunes library on the network along with 8 GB of flash memory to cache media downloaded Apple with the Apple TV has added another device to its portfolio that runs on its A4 processor along with the iPad and the iPhone The memory included in the device is the half of the iPhone 4 at 256 MB the same as the iPad iPhone 3GS third and fourth-generation iPod Touch It has HDMI out as the only video out source Features include access to the iTunes Store to rent movies and TV shows purchasing has been discontinued streaming from internet video sources including YouTube and Netflix and media streaming from an iTunes library Apple also reduced the price of the device to 99 A third generation of the device was introduced at an Apple event on March 7 2012 with new features such as higher resolution 1080p and a new user interface Software Apple develops its own operating system to run on Macs OS X the latest version being OS X Mavericks version 10 9 Apple also independently develops computer software titles for its OS X operating system Much of the software Apple develops is bundled with its computers An example of this is the consumer-oriented iLife software package that bundles iMovie iPhoto and GarageBand For presentation page layout and word processing iWork is available which includes Keynote Pages and Numbers iTunes QuickTime media player and Software Update are available as free downloads for both OS X and Windows Apple also offers a range of professional software titles Their range of server software includes the operating system OS X Server Apple Remote Desktop a remote systems management application and Xsan a Storage Area Network file system For the professional creative market there is Aperture for professional RAW -format photo processing Final Cut Pro a video production suite Logic Pro a comprehensive music toolkit and Motion an advanced effects composition program Apple also offers online services with iCloud which provides cloud storage and syncing for a wide range of data including email contacts calendars photos and documents It also offers iOS device backup and is able to integrate directly with third-party apps for even greater functionality iCloud is the fourth generation of online services provided by Apple and was preceded by MobileMe Mac and iTools all which met varying degrees of success Corporate identity Logo According to Steve Jobs Apple was so named because Jobs was coming back from an apple farm and he was on a fruitarian diet He thought the name was fun spirited and not intimidating Apple s first logo designed by Ron Wayne depicts Sir Isaac Newton sitting under an apple tree It was almost immediately replaced by Rob Janoff s rainbow Apple the now-familiar rainbow-colored silhouette of an apple with a bite taken out of it Janoff presented Jobs with several different monochromatic themes for the bitten logo and Jobs immediately took a liking to it While Jobs liked the logo he insisted it be in color to humanize the company The logo was designed with a bite so that it would not be confused with a cherry The colored stripes were conceived to make the logo more accessible and to represent the fact the Apple II could generate graphics in color This logo is often erroneously referred to as a tribute to Alan Turing with the bite mark a reference to his method of suicide Both Janoff and Apple deny any homage to Turing in the design of the logo On August 27 1999 the following year after the iMac G3 was introduced Apple officially dropped the rainbow scheme and began to use monochromatic themes nearly identical in shape to its previous rainbow incarnation on various products packaging and advertising An Aqua -themed version of the monochrome logo was used from 1999 to 2003 and a Glass-themed version was used from 2007 to 2013 With the release of iOS 7 and OS X Mavericks in fall of 2013 the logo appears flat and white with no glossy effects Steve Jobs and Steve Wozniak were Beatles fans but Apple Inc had trademark issues with Apple Corps Ltd a multimedia company started by the Beatles in 1967 involving their name and logo This resulted in a series of lawsuits and tension between the two companies These issues ended with settling of their most recent lawsuit in 2007 Advertising Apple s first slogan Byte into an Apple was coined in the late 1970s From 1997–2002 the slogan Think Different was used in advertising campaigns and is still closely associated with Apple Apple also has slogans for specific product lines — for example iThink therefore iMac was used in 1998 to promote the iMac and Say hello to iPhone has been used in iPhone advertisements Hello was also used to introduce the original Macintosh Newton iMac hello again and iPod Since the introduction of the Macintosh in 1984 with the 1984 Super Bowl commercial to the more modern Get a Mac adverts Apple has been recognized in the past for its efforts towards effective advertising and marketing for its products though its advertising was criticized for the claims made by some later campaigns particularly the 2005 Power Mac ads and iPhone ads in Britain Apple s product commercials gained fame for launching musicians into stardom as a result of their eye-popping graphics and catchy tunes First the company popularized Canadian singer Feist s 1234 song in its ad campaign Later Apple used the song New Soul by French-Israeli singer-songwriter Yael Naïm to promote the MacBook Air The debut single shot to the top of the charts and sold hundreds of thousands of copies in a span of weeks Brand loyalty Apple s brand loyalty is considered unusual for any product At one time Apple evangelists were actively engaged by the company but this was after the phenomenon was already firmly established Apple evangelist Guy Kawasaki has called the brand fanaticism something that was stumbled upon Apple has however supported the continuing existence of a network of Mac User Groups in most major and many minor centers of population where Mac computers are available Mac users would meet at the European Apple Expo and the San Francisco Macworld Conference Expo trade shows where Apple traditionally introduced new products each year to the industry and public until Apple pulled out of both events While the conferences continue Apple does not have official representation there Mac developers in turn continue to gather at the annual Apple Worldwide Developers Conference Apple Store openings can draw crowds of thousands with some waiting in line as much as a day before the opening or flying in from other countries for the event The New York City Fifth Avenue Cube store had a line as long as half a mile a few Mac fans took the opportunity of the setting to propose marriage The Ginza opening in Tokyo was estimated in the thousands with a line exceeding eight city blocks John Sculley told The Guardian newspaper in 1997 People talk about technology but Apple was a marketing company It was the marketing company of the decade Research in 2002 by NetRatings indicate that the average Apple consumer was usually more affluent and better educated than other PC company consumers The research indicated that this correlation could stem from the fact that on average Apple Inc products are more expensive than other PC products Corporate affairs During the Mac s early history Apple generally refused to adopt prevailing industry standards for hardware instead creating their own This trend was largely reversed in the late 1990s beginning with Apple s adoption of the PCI bus in the 7500 8500 9500 Power Macs Apple has since adopted USB AGP HyperTransport Wi-Fi and other industry standards in its computers and was in some cases a leader in the adoption of standards such as USB FireWire is an Apple-originated standard that has seen widespread industry adoption after it was standardized as IEEE 1394 Ever since the first Apple Store opened Apple has sold third-party accessories For instance at one point Nikon and Canon digital cameras were sold inside the store Adobe one of Apple s oldest software partners also sells its Mac-compatible software as does Microsoft who sells Microsoft Office for the Mac Books from John Wiley Sons who publishes the For Dummies series of instructional books are a notable exception however The publisher s line of books were banned from Apple Stores in 2005 because Steve Jobs disagreed with their decision to publish an unauthorized Jobs biography iCon After the launch of the iBookstore Apple stopped selling physical books both online and at the Apple Retail Stores Headquarters Apple Inc s world corporate headquarters are located in the middle of Silicon Valley at 1–6 Infinite Loop Cupertino California This Apple campus has six buildings that total and was built in 1993 by Sobrato Development Cos In 2006 Apple announced its intention to build a second campus on assembled from various contiguous plots east of N Wolfe Road between Pruneridge Avenue and Vallco Parkway Later acquisitions increased this to 175 acres The new campus also in Cupertino will be about east of the current campus The new campus building will be designed by Norman Foster On October 15 2013 it was announced that the Cupertino City Council has approved the proposed spaceship design campus On June 7 2011 Steve Jobs gave a presentation to Cupertino City Council detailing the architectural design of the new building and its environs The new campus is planned to house up to 13 000 employees in one central four-storied circular building with a café for 3 000 sitting people integrated surrounded by extensive landscape with parking mainly underground and the rest centralized in a parking structure The new campus will be built on the former HP headquarters next to Interstate 280 The morning of the announcement Apple CEO Tim Cook tweeted Our home for innovation and creativity for decades to come Cupertino City Council Gives Unanimous Approval for Apple s New Campus The 2 8 million square foot facility which will include Steve Jobs s original designs for a fitness center and corporate auditorium will be able to house 14 000 employees and will have enough parking to accommodate almost all of them Apple s headquarters for Europe the Middle East and Africa EMEA are located in Cork in the south of Ireland The facility which opened in 1980 was Apple s first location outside of the United States Apple Sales International which deals with all of Apple s international sales outside of the USA is located at Apple s campus in Cork along with Apple Distribution International which similarly deals with Apple s international distribution network On April 20 2012 Apple announced the addition of 500 new jobs to its European headquarters This will bring the total workforce from around 2 800 to 3 300 employees The company will build a new office block on its Hollyhill Campus to accommodate the additional staff Corporate culture Apple was one of several highly successful companies founded in the 1970s that bucked the traditional notions of what a corporate culture should look like in organizational hierarchy flat versus tall casual versus formal attire etc Other highly successful firms with similar cultural aspects from the same period include Southwest Airlines and Microsoft Originally the company stood in opposition to staid competitors like IBM by default thanks to the influence of its founders Steve Jobs often walked around the office barefoot even after Apple was a Fortune 500 company By the time of the 1984 TV ad this trait had become a key way the company attempted to differentiate itself from its competitors According to a 2011 report in Fortune this has resulted in a corporate culture more akin to a startup rather than a multinational corporation As the company has grown and been led by a series of chief executives each with his own idea of what Apple should be some of its original character has arguably been lost but Apple still has a reputation for fostering individuality and excellence that reliably draws talented people into its employ This was especially after Jobs return To recognize the best of its employees Apple created the Apple Fellows program awarding individuals who made extraordinary technical or leadership contributions to personal computing while at the company The Apple Fellowship has so far been awarded to a few individuals including Bill Atkinson Steve Capps Rod Holt Alan Kay Guy Kawasaki Al Alcorn Don Norman Rich Page and Steve Wozniak Apple is also known for strictly enforcing accountability Each project has a directly responsible individual or DRI in Apple jargon As an example when iOS senior vice president Scott Forstall refused to sign Apple s official apology for numerous errors in the redesigned Maps app he was forced to resign Numerous employees of Apple have cited that projects without Jobs involvement often took longer than projects with his involvement At Apple employees are specialists who are not exposed to functions outside their area of expertise Jobs saw this as a means of having best-in-class employees in every role For instance Ron Johnson who was Senior Vice President of Retail Operations until November 1 2011 was responsible for site selection in-store service and store layout yet he had no control of the inventory in his stores which is done company wide by then-COO and now CEO Tim Cook who has a background in supply-chain management This is the opposite of General Electric s corporate culture which has created well-rounded managers Under the leadership of Tim Cook who joined the company in 1998 and ascended to his present position as CEO Apple has developed an extremely efficient and effective supply chain which has been ranked as the world s best for the four years 2007–2010 The company s manufacturing procurement and logistics enables it to execute massive product launches without having to maintain large profit-sapping inventories Apple s profit margins have been 40 percent compared with 10–20 percent for most other hardware companies in 2011 Cook s catchphrase to describe his focus on the company s operational edge is “Nobody wants to buy sour milk” The company previously advertised its products as being made in America up to the late 1990s however as a result of outsourcing initiatives in the 2000s almost all of its manufacturing is now done abroad According to a report by the New York Times Apple insiders believe the vast scale of overseas factories as well as the flexibility diligence and industrial skills of foreign workers have so outpaced their American counterparts that “Made in the U S A ” is no longer a viable option for most Apple products Unlike other major US companies Apple has a relatively simple compensation policy for executives which does not include perks that other CEOs enjoy such as country club fees and private use of company aircraft The company usually grants stock options to executives every other year A media article published in July 2013 provided details about Apple s At-Home Apple Advisors customer support program that serves as the corporation s call center The advisors are employed within the U S and work remotely after undergoing a four-week training program that also serves as a testing period The advisors earn between US 9 and 12 per hour and receive intensive management to ensure a high quality of customer support Employee relations A class-action lawsuit alleging that the company suppressed employee compensation has been filed against Apple in a California federal district court Litigation Apple has been a participant in various legal proceedings and claims since it began operation and like its competitors and peers engages in litigation trying legal cases before the courts in its normal course of business for a variety of reasons In particular Apple is known for and promotes itself as actively and aggressively enforcing its intellectual property interests Some examples include Apple v Samsung Apple v Microsoft Motorola v Apple Apple Corps v Apple Computer Finance In its fiscal year ending in September 2011 Apple Inc reported a total of 108 billion in annual revenues – a significant increase from its 2010 revenues of 65 billion – and nearly 82 billion in cash reserves Apple achieved these results while losing market share in certain product categories On March 19 2012 Apple announced plans for a 2 65-per-share dividend beginning in fourth quarter of 2012 per approval by their board of directors On September 2012 Apple reached a record share price of more than 705 and closed at above 700 With 936 596 000 outstanding shares as of June 30 2012 it had a market capitalization of about 660 billion At the time this was the highest nominal market capitalization ever reached by a publicly traded company surpassing a record set by Microsoft in 1999 Environmental record Climate change and clean energy On April 21 2011 Greenpeace released a report highlighting the fact that data centers consumed up to 2 of all global electricity and this amount was projected to increase Phil Radford of Greenpeace said “we are concerned that this new explosion in electricity use could lock us into old polluting energy sources instead of the clean energy available today ” On April 17 2012 following a Greenpeace protest of Apple Apple Inc released a statement committing to ending its use of coal and shifting to 100 clean energy In 2013 Apple announced it was using 100 renewable energy to power their data centers and overall 75 of its power comes from renewable sources In 2010 Climate Counts a nonprofit organization dedicated to directing consumers toward the greenest companies gave Apple a score of 52 points out of a possible 100 which puts Apple in their top category Striding This was an increase from May 2008 when Climate Counts only gave Apple 11 points out of 100 which placed the company last among electronics companies at which time Climate Counts also labeled Apple with a stuck icon adding that Apple at the time was a choice to avoid for the climate conscious consumer Toxics Greenpeace has campaigned against Apple because of various environmental issues including a global end-of-life take-back plan non-recyclable hardware components and toxins within iPhone hardware Since 2003 Greenpeace has campaigned against Apple s use of particular chemicals in its products more specifically the inclusion of PVC and BFRs in their devices On May 2 2007 Steve Jobs released a report announcing plans to eliminate PVC and BFRs by the end of 2008 Apple has since eliminated PVC and BFRs from its product range becoming the first laptopmaker to do so In the first edition of the Greenpeace Green Electronics Guide released in August 2006 Apple only scored 2 7 10 The Environmental Protection Agency rates Apple highest amongst producers of notebooks and fairly well compared to producers of desktop computers and LCD displays In June 2007 Apple upgraded the MacBook Pro replacing cold cathode fluorescent lamp CCFL backlit LCD displays with mercury -free LED backlit LCD displays and arsenic -free glass and has since done this for all notebooks Apple has also left out BFRs and PVCs in various internal components Apple offers information about emissions materials and electrical usage concerning each product In June 2009 Apple s iPhone 3GS was free of PVC arsenic BFRs and had an efficient power adapter In October 2009 Apple upgraded the iMac and MacBook replacing the cold cathode fluorescent lamp CCFL backlit LCD displays with mercury-free LED backlit LCD displays and arsenic-free glass This means all Apple computers have mercury free LED backlit displays arsenic-free glass and are without PVC cables All Apple computers also have EPEAT Gold status In October 2011 Chinese authorities ordered an Apple supplier to close part of its plant in Suzhou after residents living nearby raised significant environmental concerns In November 2011 Apple featured in Greenpeace s Guide to Greener Electronics which ranks electronics manufacturers on sustainability climate and energy policy and how green their products are The company ranked fourth of fifteen electronics companies moving up five places from the previous year with a score of 4 6 10 down from 4 9 Greenpeace praises Apple s sustainability noting that the company exceeded its 70 global recycling goal in 2010 It continues to score well on the products rating with all Apple products now being free of PVC vinyl plastic and brominated flame retardants However the guide criticizes Apple on the Energy criteria for not seeking external verification of its greenhouse gas emissions data and for not setting out any targets to reduce emissions In January 2012 Apple announced plans and requested that their cable maker Volex begin producing halogen-free USB and power cables In June 2012 Apple Inc withdrew its products from the Electronic Product Environmental Assessment Tool EPEAT certification system but reversed this decision in July Labor practices In 2006 the Mail on Sunday reported on the working conditions that existed at factories in China where the contract manufacturers Foxconn and Inventec produced the iPod The article stated that one complex of factories that assembles the iPod among other items had over 200 000 workers that lived and worked in the factory with employees regularly working more than 60 hours per week The article also reported that workers made around 100 per month and were required to live pay for rent and food from the company which generally amounted to a little over half of workers earnings Apple immediately launched an investigation and worked with their manufacturers to ensure acceptable working conditions In 2007 Apple started yearly audits of all its suppliers regarding worker s rights slowly raising standards and pruning suppliers that did not comply Yearly progress reports have been published since 2008 In 2010 workers in China planned to sue iPhone contractors over poisoning by a cleaner used to clean LCD screens One worker claimed that he and his coworkers had not been informed of possible occupational illnesses After a spate of suicides in a Foxconn facility in China making iPads and iPhones albeit at a lower rate than in China as a whole workers were forced to sign a legally binding document guaranteeing that they would not kill themselves In 2011 Apple admitted that its suppliers child labor practices in China had worsened Workers in factories producing Apple products have also been exposed to n- hexane a neurotoxin that is a cheaper alternative than alcohol for cleaning the products In 2013 China Labor Watch said it found violations of the law and of Apple s pledges about working conditions at facilities operated by Pegatron including discrimination against ethnic minorities and women withholding employees pay excessive work hours poor living conditions health and safety problems and pollution Tax practices Apple created subsidiaries in low-tax places such as the Republic of Ireland the Netherlands Luxembourg and the British Virgin Islands to cut the taxes it pays around the world According to the New York Times in the 1980s Apple was among the first tech companies to designate overseas salespeople in high-tax countries in a manner that allowed the company to sell on behalf of low-tax subsidiaries on other continents sidestepping income taxes In the late 1980s Apple was a pioneer of an accounting technique known as the Double Irish With a Dutch Sandwich which reduces taxes by routing profits through Irish subsidiaries and the Netherlands and then to the Caribbean British Conservative Party Member of Parliament Charlie Elphicke published research on October 30 2012 which showed that some multinational companies including Apple Inc were making billions of pounds of profit in the UK but were paying an effective tax rate to the UK Treasury of only 3 percent well below standard corporation tax He followed this research by calling on the Chancellor of the Exchequer George Osborne to force these multinationals which also included Google and Coca-Cola to state the effective rate of tax they pay on their UK revenues Elphicke also said that government contracts should be withheld from multinationals who do not pay their fair share of UK tax Charitable causes As of 2012 Apple is listed as a partner of the Product RED campaign together with other brands such as Nike Girl American Express and Converse The campaign s mission is to prevent the transmission of HIV from mother to child by 2015 its byline is Fighting For An AIDS Free Generation In November 2012 Apple donated 2 5 million to the American Red Cross to aid relief efforts after Hurricane Sandy'\n\n\nent_names_4EX = {} \nfor name,_ in pairs(ent_lines_4EX) do\n table.insert(ent_names_4EX, name)\nend\n\n\nent_wiki_words_4EX = {}\nfor ent_name,line in pairs(ent_lines_4EX) do\n ent_wiki_words_4EX[ent_name] = {}\n local parts = split(line, '\\t')\n local wiki_words = split(parts[3], ' ')\n\n -- Compute tf's\n for _,w in pairs(wiki_words) do\n if (not is_stop_word_or_number(w)) then\n if (not ent_wiki_words_4EX[ent_name][w]) then\n ent_wiki_words_4EX[ent_name][w] = 0\n end\n ent_wiki_words_4EX[ent_name][w] = ent_wiki_words_4EX[ent_name][w] + 1\n end\n end\nend\n" }, { "path": "deep-ed/deep-ed-master/entities/learn_e2v/batch_dataset_a.lua", "content": "dofile 'utils/utils.lua'\n\nif opt.entities == 'ALL' then\n wiki_words_train_file = opt.root_data_dir .. 'generated/wiki_canonical_words.txt'\n wiki_hyp_train_file = opt.root_data_dir .. 'generated/wiki_hyperlink_contexts.csv'\nelse\n wiki_words_train_file = opt.root_data_dir .. 'generated/wiki_canonical_words_RLTD.txt'\n wiki_hyp_train_file = opt.root_data_dir .. 'generated/wiki_hyperlink_contexts_RLTD.csv'\nend\n\nwiki_words_it, _ = io.open(wiki_words_train_file)\nwiki_hyp_it, _ = io.open(wiki_hyp_train_file)\n\nassert(opt.num_passes_wiki_words)\nlocal train_data_source = 'wiki-canonical'\nlocal num_passes_wiki_words = 1\n\nlocal function read_one_line()\n if train_data_source == 'wiki-canonical' then\n line = wiki_words_it:read()\n else\n assert(train_data_source == 'wiki-canonical-hyperlinks')\n line = wiki_hyp_it:read()\n end\n if (not line) then\n if num_passes_wiki_words == opt.num_passes_wiki_words then\n train_data_source = 'wiki-canonical-hyperlinks'\n print('\\n\\n' .. 'Start training on Wiki Hyperlinks' .. '\\n\\n')\n end\n print('Training file is done. Num passes = ' .. num_passes_wiki_words .. '. Reopening.')\n num_passes_wiki_words = num_passes_wiki_words + 1\n if train_data_source == 'wiki-canonical' then\n wiki_words_it, _ = io.open(wiki_words_train_file)\n line = wiki_words_it:read()\n else\n wiki_hyp_it, _ = io.open(wiki_hyp_train_file)\n line = wiki_hyp_it:read()\n end\n end \n return line\nend\n\nlocal line = nil\n\nlocal function patch_of_lines(num)\n local lines = {}\n local cnt = 0\n assert(num > 0)\n \n while cnt < num do\n line = read_one_line()\n cnt = cnt + 1\n table.insert(lines, line)\n end\n\n assert(table_len(lines) == num)\n return lines\nend\n\n\nfunction get_minibatch()\n -- Create empty mini batch:\n local lines = patch_of_lines(opt.batch_size)\n local inputs = empty_minibatch() \n local targets = correct_type(torch.ones(opt.batch_size, opt.num_words_per_ent))\n\n -- Fill in each example:\n for i = 1,opt.batch_size do\n local sample_line = lines[i] -- load new example line\n local target = process_one_line(sample_line, inputs, i)\n targets[i]:copy(target)\n end\n \n --- Minibatch post processing:\n postprocess_minibatch(inputs, targets)\n targets = targets:view(opt.batch_size * opt.num_words_per_ent) \n\n -- Special target for the NEG and NCE losses\n if opt.loss == 'neg' or opt.loss == 'nce' then\n nce_targets = torch.ones(opt.batch_size * opt.num_words_per_ent, opt.num_neg_words):mul(-1)\n for j = 1,opt.batch_size * opt.num_words_per_ent do\n nce_targets[j][targets[j]] = 1\n end\n targets = nce_targets\n end\n\n return inputs, targets\nend\n" }, { "path": "deep-ed/deep-ed-master/entities/learn_e2v/e2v_a.lua", "content": "-- Entity embeddings utilities\n\nassert(opt.num_words_per_ent)\n\n-- Word lookup:\ngeom_w2v_M = w2vutils.M:float()\n\n-- Stats:\nlocal num_invalid_ent_wikiids = 0\nlocal total_ent_wiki_vec_requests = 0\nlocal last_wrote = 0\nlocal function invalid_ent_wikiids_stats(ent_thid)\n total_ent_wiki_vec_requests = total_ent_wiki_vec_requests + 1\n if ent_thid == unk_ent_thid then\n num_invalid_ent_wikiids = num_invalid_ent_wikiids + 1\n end\n if (num_invalid_ent_wikiids % 15000 == 0 and num_invalid_ent_wikiids ~= last_wrote) then\n last_wrote = num_invalid_ent_wikiids\n local perc = 100.0 * num_invalid_ent_wikiids / total_ent_wiki_vec_requests\n print(red('*** Perc invalid ent wikiids = ' .. perc .. ' . Absolute num = ' .. num_invalid_ent_wikiids))\n end\nend\n\n-- ent id -> vec\nfunction geom_entwikiid2vec(ent_wikiid)\n local ent_thid = get_thid(ent_wikiid)\n assert(ent_thid)\n invalid_ent_wikiids_stats(ent_thid)\n local ent_vec = nn.Normalize(2):forward(lookup_ent_vecs.weight[ent_thid]:float())\n return ent_vec\nend\n\n-- ent name -> vec\nlocal function geom_entname2vec(ent_name)\n assert(ent_name)\n return geom_entwikiid2vec(get_ent_wikiid_from_name(ent_name))\nend\n\nfunction entity_similarity(e1_wikiid, e2_wikiid)\n local e1_vec = geom_entwikiid2vec(e1_wikiid)\n local e2_vec = geom_entwikiid2vec(e2_wikiid)\n return e1_vec * e2_vec\nend\n\n\nlocal function geom_top_k_closest_words(ent_name, ent_vec, k)\n local tf_map = ent_wiki_words_4EX[ent_name]\n local w_not_found = {}\n for w,_ in pairs(tf_map) do\n if tf_map[w] >= 10 then\n w_not_found[w] = tf_map[w]\n end\n end\n \n distances = geom_w2v_M * ent_vec\n\n local best_scores, best_word_ids = topk(distances, k)\n local returnwords = {}\n local returndistances = {}\n for i = 1,k do\n local w = get_word_from_id(best_word_ids[i])\n if is_stop_word_or_number(w) then\n table.insert(returnwords, red(w))\n elseif tf_map[w] then\n if tf_map[w] >= 15 then\n table.insert(returnwords, yellow(w .. '{' .. tf_map[w] .. '}'))\n else\n table.insert(returnwords, skyblue(w .. '{' .. tf_map[w] .. '}'))\n end\n w_not_found[w] = nil\n else\n table.insert(returnwords, w)\n end\n assert(best_scores[i] == distances[best_word_ids[i]], best_scores[i] .. ' ' .. distances[best_word_ids[i]])\n table.insert(returndistances, distances[best_word_ids[i]])\n end\n return returnwords, returndistances, w_not_found\nend\n\n\nlocal function geom_most_similar_words_to_ent(ent_name, k)\n local ent_wikiid = get_ent_wikiid_from_name(ent_name)\n local k = k or 1\n local ent_vec = geom_entname2vec(ent_name)\n assert(math.abs(1 - ent_vec:norm()) < 0.01 or ent_vec:norm() == 0, ':::: ' .. ent_vec:norm())\n \n print('\\nTo entity: ' .. blue(ent_name) .. '; vec norm = ' .. ent_vec:norm() .. ':')\n neighbors, scores, w_not_found = geom_top_k_closest_words(ent_name, ent_vec, k)\n print(green('WORDS MODEL: ') .. list_with_scores_to_str(neighbors, scores))\n \n local str = yellow('WORDS NOT FOUND: ')\n for w,tf in pairs(w_not_found) do\n if tf >= 20 then\n str = str .. yellow(w .. '{' .. tf .. '}; ')\n else\n str = str .. w .. '{' .. tf .. '}; '\n end\n end\n print('\\n' .. str)\n print('============================================================================')\nend\n\n\n-- Unit tests :\nfunction geom_unit_tests()\n print('\\n' .. yellow('Words to Entity Similarity test:')) \n for i=1,table_len(ent_names_4EX) do\n geom_most_similar_words_to_ent(ent_names_4EX[i], 200)\n end\nend\n" }, { "path": "deep-ed/deep-ed-master/entities/learn_e2v/learn_a.lua", "content": "-- Training of entity embeddings. \n\n-- To run:\n-- i) delete all _RLTD files\n-- ii) th entities/relatedness/filter_wiki_canonical_words_RLTD.lua ; th entities/relatedness/filter_wiki_hyperlink_contexts_RLTD.lua\n-- iii) th entities/learn_e2v/learn_a.lua -root_data_dir /path/to/your/ed_data/files/\n\n-- Training of entity vectors\nrequire 'optim'\nrequire 'torch'\nrequire 'gnuplot'\nrequire 'nn'\nrequire 'xlua'\n\ndofile 'utils/utils.lua'\n\ncmd = torch.CmdLine()\ncmd:text()\ncmd:text('Learning entity vectors')\ncmd:text()\ncmd:text('Options:')\n\n---------------- runtime options:\ncmd:option('-type', 'cudacudnn', 'Type: double | float | cuda | cudacudnn')\n\ncmd:option('-root_data_dir', '', 'Root path of the data, $DATA_PATH.')\n\ncmd:option('-optimization', 'ADAGRAD', 'Optimization method: RMSPROP | ADAGRAD | ADAM | SGD')\n\ncmd:option('-lr', 0.3, 'Learning rate')\n\ncmd:option('-batch_size', 500, 'Mini-batch size (1 = pure stochastic)')\n\ncmd:option('-word_vecs', 'w2v', '300d word vectors type: glove | w2v')\n\ncmd:option('-num_words_per_ent', 20, 'Num positive words sampled for the given entity at ' .. \n 'each iteration.')\n\ncmd:option('-num_neg_words', 5, 'Num negative words sampled for each positive word.')\n\ncmd:option('-unig_power', 0.6, 'Negative sampling unigram power (0.75 used in Word2Vec).')\n\ncmd:option('-entities', 'RLTD',\n 'Set of entities for which we train embeddings: 4EX (tiny, for debug) | ' .. \n 'RLTD (restricted set) | ALL (all Wiki entities, too big to fit on a single GPU)')\n\ncmd:option('-init_vecs_title_words', true, 'whether the entity embeddings should be initialized with the average of ' .. \n 'title word embeddings. Helps to speed up convergence speed of entity embeddings learning.')\n\ncmd:option('-loss', 'maxm', 'Loss function: nce (noise contrastive estimation) | ' .. \n 'neg (negative sampling) | is (importance sampling) | maxm (max-margin)')\n\ncmd:option('-data', 'wiki-canonical-hyperlinks', 'Training data: wiki-canonical (only) | ' ..\n 'wiki-canonical-hyperlinks')\n\n-- Only when opt.data = wiki-canonical-hyperlinks\ncmd:option('-num_passes_wiki_words', 200, 'Num passes (per entity) over Wiki canonical pages before ' ..\n 'changing to using Wiki hyperlinks.')\n\ncmd:option('-hyp_ctxt_len', 10, 'Left and right context window length for hyperlinks.')\n\ncmd:option('-banner_header', '', 'Banner header')\n\ncmd:text()\nopt = cmd:parse(arg or {})\n\nbanner = '' .. opt.banner_header .. ';obj-' .. opt.loss .. ';' .. opt.data\nif opt.data ~= 'wiki-canonical' then\n banner = banner .. ';hypCtxtL-' .. opt.hyp_ctxt_len\n banner = banner .. ';numWWpass-' .. opt.num_passes_wiki_words\nend\nbanner = banner .. ';WperE-' .. opt.num_words_per_ent \nbanner = banner .. ';' .. opt.word_vecs .. ';negW-' .. opt.num_neg_words\nbanner = banner .. ';ents-' .. opt.entities .. ';unigP-' .. opt.unig_power \nbanner = banner .. ';bs-' .. opt.batch_size .. ';' .. opt.optimization .. '-lr-' .. opt.lr\n\nprint('\\n' .. blue('BANNER : ' .. banner))\n\nprint('\\n===> RUN TYPE: ' .. opt.type) \n\ntorch.setdefaulttensortype('torch.FloatTensor')\nif string.find(opt.type, 'cuda') then\n print('==> switching to CUDA (GPU)')\n require 'cunn'\n require 'cutorch'\n require 'cudnn'\n cudnn.benchmark = true \n cudnn.fastest = true\nelse\n print('==> running on CPU')\nend\n\ndofile 'utils/logger.lua'\ndofile 'entities/relatedness/relatedness.lua'\ndofile 'entities/ent_name2id_freq/ent_name_id.lua'\ndofile 'words/load_w_freq_and_vecs.lua'\ndofile 'words/w2v/w2v.lua'\ndofile 'entities/learn_e2v/minibatch_a.lua'\ndofile 'entities/learn_e2v/model_a.lua' \ndofile 'entities/learn_e2v/e2v_a.lua'\ndofile 'entities/learn_e2v/batch_dataset_a.lua'\n\nif opt.loss == 'neg' or opt.loss == 'nce' then\n criterion = nn.SoftMarginCriterion()\nelseif opt.loss == 'maxm' then\n criterion = nn.MultiMarginCriterion(1, torch.ones(opt.num_neg_words), 0.1)\nelseif opt.loss == 'is' then\n criterion = nn.CrossEntropyCriterion()\nend\n\nif string.find(opt.type, 'cuda') then\n criterion = criterion:cuda()\nend\n\n\n----------------------------------------------------------------------\nif opt.optimization == 'ADAGRAD' then -- See: http://cs231n.github.io/neural-networks-3/#update \n dofile 'utils/optim/adagrad_mem.lua'\n optimMethod = adagrad_mem\n optimState = {\n learningRate = opt.lr \n }\nelseif opt.optimization == 'RMSPROP' then -- See: cs231n.github.io/neural-networks-3/#update \n dofile 'utils/optim/rmsprop_mem.lua'\n optimMethod = rmsprop_mem\n optimState = {\n learningRate = opt.lr\n }\nelseif opt.optimization == 'SGD' then -- See: http://cs231n.github.io/neural-networks-3/#update \n optimState = {\n learningRate = opt.lr,\n learningRateDecay = 5e-7\n }\n optimMethod = optim.sgd \nelseif opt.optimization == 'ADAM' then -- See: http://cs231n.github.io/neural-networks-3/#update \n optimState = {\n learningRate = opt.lr,\n }\n optimMethod = optim.adam\nelse\n error('unknown optimization method')\nend\n\n\n----------------------------------------------------------------------\nfunction train_ent_vecs()\n print('Training entity vectors w/ params: ' .. banner)\n \n -- Retrieve parameters and gradients:\n -- extracts and flattens all model's parameters into a 1-dim vector\n parameters,gradParameters = model:getParameters()\n gradParameters:zero()\n\n local processed_so_far = 0\n if opt.entities == 'ALL' then\n num_batches_per_epoch = 4000\n elseif opt.entities == 'RLTD' then\n num_batches_per_epoch = 2000\n elseif opt.entities == '4EX' then\n num_batches_per_epoch = 400\n end\n local test_every_num_epochs = 1 \n local save_every_num_epochs = 3\n \n -- epoch tracker\n epoch = 1\n\n -- Initial testing:\n geom_unit_tests() -- Show some examples \n \n print('Training params: ' .. banner)\n\n -- do one epoch\n print('\\n==> doing epoch on training data:')\n print(\"==> online epoch # \" .. epoch .. ' [batch size = ' .. opt.batch_size .. ']')\n\n while true do\n \n local time = sys.clock() \n print(green('\\n===> TRAINING EPOCH #' .. epoch .. '; num batches ' .. num_batches_per_epoch .. ' <==='))\n \n local avg_loss_before_opt_per_epoch = 0.0\n local avg_loss_after_opt_per_epoch = 0.0\n \n for batch_index = 1,num_batches_per_epoch do \n -- Read one mini-batch from one data_thread:\n inputs, targets = get_minibatch() \n \n -- Move data to GPU:\n minibatch_to_correct_type(inputs)\n targets = correct_type(targets)\n \n -- create closure to evaluate f(X) and df/dX\n local feval = function(x)\n -- get new parameters\n if x ~= parameters then\n parameters:copy(x)\n end\n\n -- reset gradients\n gradParameters:zero()\n\n -- evaluate function for complete mini batch \n local outputs = model:forward(inputs)\n assert(outputs:size(1) == opt.batch_size * opt.num_words_per_ent and \n outputs:size(2) == opt.num_neg_words)\n\n local f = criterion:forward(outputs, targets)\n \n -- estimate df/dW\n local df_do = criterion:backward(outputs, targets)\n local gradInput = model:backward(inputs, df_do)\n\n -- return f and df/dX\n return f,gradParameters\n end\n \n -- Debug info:\n local loss_before_opt = criterion:forward(model:forward(inputs), targets)\n avg_loss_before_opt_per_epoch = avg_loss_before_opt_per_epoch + loss_before_opt\n \n -- Optimize on current mini-batch\n optimMethod(feval, parameters, optimState)\n \n local loss_after_opt = criterion:forward(model:forward(inputs), targets)\n avg_loss_after_opt_per_epoch = avg_loss_after_opt_per_epoch + loss_after_opt\n if loss_after_opt > loss_before_opt then\n print(red('!!!!!! LOSS INCREASED: ' .. loss_before_opt .. ' --> ' .. loss_after_opt))\n end\n \n -- Display progress\n train_size = 17000000 ---------- 4 passes over the Wiki entity set\n processed_so_far = processed_so_far + opt.batch_size\n if processed_so_far > train_size then\n processed_so_far = processed_so_far - train_size\n end\n xlua.progress(processed_so_far, train_size)\n end\n \n avg_loss_before_opt_per_epoch = avg_loss_before_opt_per_epoch / num_batches_per_epoch\n avg_loss_after_opt_per_epoch = avg_loss_after_opt_per_epoch / num_batches_per_epoch\n print(yellow('\\nAvg loss before opt = ' .. avg_loss_before_opt_per_epoch .. \n '; Avg loss after opt = ' .. avg_loss_after_opt_per_epoch))\n\n -- time taken\n time = sys.clock() - time\n time = time / (num_batches_per_epoch * opt.batch_size)\n print(\"==> time to learn 1 full entity = \" .. (time*1000) .. 'ms')\n \n geom_unit_tests() -- Show some entity examples\n \n -- Various testing measures:\n if (epoch % test_every_num_epochs == 0) then\n if opt.entities ~= '4EX' then\n compute_relatedness_metrics(entity_similarity)\n end\n end\n\n -- Save model:\n if (epoch % save_every_num_epochs == 0) then\n print('==> saving model to ' .. opt.root_data_dir .. 'generated/ent_vecs/ent_vecs__ep_' .. epoch .. '.t7')\n torch.save(opt.root_data_dir .. 'generated/ent_vecs/ent_vecs__ep_' .. epoch .. '.t7', nn.Normalize(2):forward(lookup_ent_vecs.weight:float()))\n end\n \n print('Training params: ' .. banner)\n \n -- next epoch\n epoch = epoch + 1\n end\nend\n\ntrain_ent_vecs()\n" }, { "path": "deep-ed/deep-ed-master/entities/learn_e2v/minibatch_a.lua", "content": "assert(opt.entities == '4EX' or opt.entities == 'ALL' or opt.entities == 'RLTD', opt.entities)\n\nfunction empty_minibatch()\n local ctxt_word_ids = torch.ones(opt.batch_size, opt.num_words_per_ent, opt.num_neg_words):mul(unk_w_id)\n local ent_component_words = torch.ones(opt.batch_size, opt.num_words_per_ent):int()\n local ent_wikiids = torch.ones(opt.batch_size):int()\n local ent_thids = torch.ones(opt.batch_size):int()\n return {{ctxt_word_ids}, {ent_component_words}, {ent_thids, ent_wikiids}}\nend\n\n-- Get functions:\nfunction get_pos_and_neg_w_ids(minibatch) \n return minibatch[1][1]\nend\nfunction get_pos_and_neg_w_vecs(minibatch) \n return minibatch[1][2]\nend\nfunction get_pos_and_neg_w_unig_at_power(minibatch) \n return minibatch[1][3]\nend\nfunction get_ent_wiki_w_ids(minibatch) \n return minibatch[2][1]\nend\nfunction get_ent_wiki_w_vecs(minibatch) \n return minibatch[2][2]\nend\nfunction get_ent_thids_batch(minibatch) \n return minibatch[3][1]\nend\nfunction get_ent_wikiids(minibatch) \n return minibatch[3][2]\nend\n\n\n-- Fills in the minibatch and returns the grd truth word index per each example.\n-- An example in our case is an entity, a positive word sampled from \\hat{p}(e|m) \n-- and several negative words sampled from \\hat{p}(w)^\\alpha.\nfunction process_one_line(line, minibatch, mb_index)\n if opt.entities == '4EX' then\n line = ent_lines_4EX[ent_names_4EX[math.random(1, table_len(ent_names_4EX))]]\n end\n \n local parts = split(line, '\\t')\n local num_parts = table_len(parts)\n \n if num_parts == 3 then ---------> Words from the Wikipedia canonical page\n assert(table_len(parts) == 3, line)\n ent_wikiid = tonumber(parts[1])\n words_plus_stop_words = split(parts[3], ' ')\n \n else --------> Words from Wikipedia hyperlinks\n assert(num_parts >= 9, line .. ' --> ' .. num_parts)\n assert(parts[6] == 'CANDIDATES', line)\n\n local last_part = parts[num_parts]\n local ent_str = split(last_part, ',')\n ent_wikiid = tonumber(ent_str[2])\n \n words_plus_stop_words = {}\n local left_ctxt_w = split(parts[4], ' ')\n local left_ctxt_w_num = table_len(left_ctxt_w)\n for i = math.max(1, left_ctxt_w_num - opt.hyp_ctxt_len + 1), left_ctxt_w_num do \n table.insert(words_plus_stop_words, left_ctxt_w[i])\n end\n local right_ctxt_w = split(parts[5], ' ')\n local right_ctxt_w_num = table_len(right_ctxt_w)\n for i = 1, math.min(right_ctxt_w_num, opt.hyp_ctxt_len) do \n table.insert(words_plus_stop_words, right_ctxt_w[i])\n end \n end\n \n assert(ent_wikiid)\n local ent_thid = get_thid(ent_wikiid)\n assert(get_wikiid_from_thid(ent_thid) == ent_wikiid)\n get_ent_thids_batch(minibatch)[mb_index] = ent_thid\n assert(get_ent_thids_batch(minibatch)[mb_index] == ent_thid)\n \n get_ent_wikiids(minibatch)[mb_index] = ent_wikiid\n \n\n -- Remove stop words from entity wiki words representations.\n local positive_words_in_this_iter = {}\n local num_positive_words_this_iter = 0\n for _,w in pairs(words_plus_stop_words) do\n if contains_w(w) then\n table.insert(positive_words_in_this_iter, w)\n num_positive_words_this_iter = num_positive_words_this_iter + 1\n end\n end \n \n -- Try getting some words from the entity title if the canonical page is empty.\n if num_positive_words_this_iter == 0 then \n local ent_name = parts[2]\n words_plus_stop_words = split_in_words(ent_name)\n for _,w in pairs(words_plus_stop_words) do\n if contains_w(w) then\n table.insert(positive_words_in_this_iter, w)\n num_positive_words_this_iter = num_positive_words_this_iter + 1\n end\n end \n \n -- Still empty ? Get some random words then.\n if num_positive_words_this_iter == 0 then\n table.insert(positive_words_in_this_iter, get_word_from_id(random_unigram_at_unig_power_w_id()))\n end\n end\n \n local targets = torch.zeros(opt.num_words_per_ent):int()\n \n -- Sample some negative words:\n get_pos_and_neg_w_ids(minibatch)[mb_index]:apply(\n function(x) \n -- Random negative words sampled sampled from \\hat{p}(w)^\\alpha.\n return random_unigram_at_unig_power_w_id() \n end\n )\n\n -- Sample some positive words:\n for i = 1,opt.num_words_per_ent do\n local positive_w = positive_words_in_this_iter[math.random(1, num_positive_words_this_iter)]\n local positive_w_id = get_id_from_word(positive_w)\n \n -- Set the positive word in a random position. Remember that index (used in training).\n local grd_trth = math.random(1, opt.num_neg_words)\n get_ent_wiki_w_ids(minibatch)[mb_index][i] = positive_w_id\n assert(get_ent_wiki_w_ids(minibatch)[mb_index][i] == positive_w_id)\n targets[i] = grd_trth\n get_pos_and_neg_w_ids(minibatch)[mb_index][i][grd_trth] = positive_w_id\n end\n\n return targets\nend\n\n\n-- Fill minibatch with word and entity vectors:\nfunction postprocess_minibatch(minibatch, targets)\n \n minibatch[1][1] = get_pos_and_neg_w_ids(minibatch):view(opt.batch_size * opt.num_words_per_ent * opt.num_neg_words)\n minibatch[2][1] = get_ent_wiki_w_ids(minibatch):view(opt.batch_size * opt.num_words_per_ent)\n \n -- ctxt word vecs\n minibatch[1][2] = w2vutils:lookup_w_vecs(get_pos_and_neg_w_ids(minibatch))\n \n minibatch[1][3] = torch.zeros(opt.batch_size * opt.num_words_per_ent * opt.num_neg_words)\n minibatch[1][3]:map(minibatch[1][1], function(_,w_id) return get_w_unnorm_unigram_at_power(w_id) end)\nend\n\n\n-- Convert mini batch to correct type (e.g. move data to GPU):\nfunction minibatch_to_correct_type(minibatch)\n minibatch[1][1] = correct_type(minibatch[1][1])\n minibatch[2][1] = correct_type(minibatch[2][1])\n minibatch[1][2] = correct_type(minibatch[1][2])\n minibatch[1][3] = correct_type(minibatch[1][3])\n minibatch[3][1] = correct_type(minibatch[3][1])\nend" }, { "path": "deep-ed/deep-ed-master/entities/learn_e2v/model_a.lua", "content": "-- Definition of the neural network used to learn entity embeddings.\n-- To run a simple unit test that checks the forward and backward passes, just run :\n-- th entities/learn_e2v/model_a.lua\n\nif not opt then -- unit tests\n unit_tests = true\n dofile 'utils/utils.lua'\n require 'nn'\n cmd = torch.CmdLine()\n cmd:option('-type', 'double', 'type: double | float | cuda | cudacudnn')\n cmd:option('-batch_size', 7, 'mini-batch size (1 = pure stochastic)')\n cmd:option('-num_words_per_ent', 100, 'num positive words per entity per iteration.')\n cmd:option('-num_neg_words', 25, 'num negative words in the partition function.')\n cmd:option('-loss', 'nce', 'nce | neg | is | maxm')\n cmd:option('-init_vecs_title_words', false, 'whether the entity embeddings should be initialized with the average of title word embeddings. Helps to speed up convergence speed of entity embeddings learning.')\n opt = cmd:parse(arg or {})\n word_vecs_size = 5\n ent_vecs_size = word_vecs_size\n lookup_ent_vecs = nn.LookupTable(100, ent_vecs_size)\nend -- end unit tests\n\n\nif not unit_tests then \n ent_vecs_size = word_vecs_size\n \n -- Init ents vectors\n print('\\n==> Init entity embeddings matrix. Num ents = ' .. get_total_num_ents())\n lookup_ent_vecs = nn.LookupTable(get_total_num_ents(), ent_vecs_size)\n \n -- Zero out unk_ent_thid vector for unknown entities.\n lookup_ent_vecs.weight[unk_ent_thid]:copy(torch.zeros(ent_vecs_size))\n \n -- Init entity vectors with average of title word embeddings. \n -- This would help speed-up training.\n if opt.init_vecs_title_words then\n print('Init entity embeddings with average of title word vectors to speed up learning.')\n for ent_thid = 1,get_total_num_ents() do\n local init_ent_vec = torch.zeros(ent_vecs_size)\n local ent_name = get_ent_name_from_wikiid(get_wikiid_from_thid(ent_thid))\n words_plus_stop_words = split_in_words(ent_name)\n local num_words_title = 0\n for _,w in pairs(words_plus_stop_words) do\n if contains_w(w) then -- Remove stop words.\n init_ent_vec:add(w2vutils.M[get_id_from_word(w)]:float())\n num_words_title = num_words_title + 1\n end\n end\n \n if num_words_title > 0 then\n if num_words_title > 3 then\n assert(init_ent_vec:norm() > 0, ent_name)\n end\n init_ent_vec:div(num_words_title)\n end\n \n if init_ent_vec:norm() > 0 then\n lookup_ent_vecs.weight[ent_thid]:copy(init_ent_vec)\n end\n end\n end\n\n collectgarbage(); collectgarbage();\n print(' Done init.')\nend\n\n---------------- Model Definition --------------------------------\ncosine_words_ents = nn.Sequential()\n :add(nn.ConcatTable()\n :add(nn.Sequential()\n :add(nn.SelectTable(1))\n :add(nn.SelectTable(2)) -- ctxt words vectors\n :add(nn.Normalize(2))\n :add(nn.View(opt.batch_size, opt.num_words_per_ent * opt.num_neg_words, ent_vecs_size)))\n :add(nn.Sequential()\n :add(nn.SelectTable(3))\n :add(nn.SelectTable(1))\n :add(lookup_ent_vecs) -- entity vectors\n :add(nn.Normalize(2))\n :add(nn.View(opt.batch_size, 1, ent_vecs_size))))\n :add(nn.MM(false, true))\n :add(nn.View(opt.batch_size * opt.num_words_per_ent, opt.num_neg_words))\n \nmodel = nn.Sequential()\n :add(cosine_words_ents)\n :add(nn.View(opt.batch_size * opt.num_words_per_ent, opt.num_neg_words))\n \n\nif opt.loss == 'is' then\n model = nn.Sequential()\n :add(nn.ConcatTable()\n :add(model)\n :add(nn.Sequential()\n :add(nn.SelectTable(1))\n :add(nn.SelectTable(3)) -- unigram distributions at power\n :add(nn.Log())\n :add(nn.View(opt.batch_size * opt.num_words_per_ent, opt.num_neg_words))))\n :add(nn.CSubTable()) \n\nelseif opt.loss == 'nce' then\n model = nn.Sequential()\n :add(nn.ConcatTable()\n :add(model)\n :add(nn.Sequential()\n :add(nn.SelectTable(1))\n :add(nn.SelectTable(3)) -- unigram distributions at power\n :add(nn.MulConstant(opt.num_neg_words - 1))\n :add(nn.Log())\n :add(nn.View(opt.batch_size * opt.num_words_per_ent, opt.num_neg_words))))\n :add(nn.CSubTable())\nend\n \n---------------------------------------------------------------------------------------------\n\n------- Cuda conversions:\nif string.find(opt.type, 'cuda') then\n model = model:cuda() --- This has to be called always before cudnn.convert\nend\n\nif string.find(opt.type, 'cudacudnn') then\n cudnn.convert(model, cudnn)\nend\n\n\n--- Unit tests\nif unit_tests then\n print('Network model unit tests:')\n local inputs = {}\n \n inputs[1] = {}\n inputs[1][1] = correct_type(torch.ones(opt.batch_size * opt.num_words_per_ent * opt.num_neg_words)) -- ctxt words\n \n inputs[2] = {}\n inputs[2][1] = correct_type(torch.ones(opt.batch_size * opt.num_words_per_ent)) -- ent wiki words\n \n inputs[3] = {}\n inputs[3][1] = correct_type(torch.ones(opt.batch_size)) -- ent th ids\n inputs[3][2] = torch.ones(opt.batch_size) -- ent wikiids\n \n -- ctxt word vecs\n inputs[1][2] = correct_type(torch.ones(opt.batch_size * opt.num_words_per_ent * opt.num_neg_words, word_vecs_size))\n\n inputs[1][3] = correct_type(torch.randn(opt.batch_size * opt.num_words_per_ent * opt.num_neg_words))\n\n local outputs = model:forward(inputs)\n \n assert(outputs:size(1) == opt.batch_size * opt.num_words_per_ent and\n outputs:size(2) == opt.num_neg_words)\n print('FWD success!')\n\n model:backward(inputs, correct_type(torch.randn(opt.batch_size * opt.num_words_per_ent, opt.num_neg_words)))\n print('BKWD success!')\nend\n" }, { "path": "deep-ed/deep-ed-master/entities/pretrained_e2v/check_ents.lua", "content": "if not opt then\n cmd = torch.CmdLine()\n cmd:option('-root_data_dir', '', 'Root path of the data, $DATA_PATH.')\n cmd:option('-ent_vecs_filename', 'ent_vecs__ep_228.t7', 'File name containing entity vectors generated with entities/learn_e2v/learn_a.lua.')\n cmd:text()\n opt = cmd:parse(arg or {})\n assert(opt.root_data_dir ~= '', 'Specify a valid root_data_dir path argument.') \nend\n\n\nrequire 'optim'\nrequire 'torch'\nrequire 'gnuplot'\nrequire 'nn'\nrequire 'xlua'\n\ndofile 'utils/utils.lua'\ndofile 'ed/args.lua'\n\ntds = require 'tds'\n\n\nprint('===> RUN TYPE: ' .. opt.type)\ntorch.setdefaulttensortype('torch.FloatTensor')\nif string.find(opt.type, 'cuda') then\n print('==> switching to CUDA (GPU)')\n require 'cunn'\n require 'cutorch'\n require 'cudnn'\n cudnn.benchmark = true \n cudnn.fastest = true\nelse\n print('==> running on CPU')\nend\n\ndofile 'utils/logger.lua'\ndofile 'entities/relatedness/relatedness.lua'\ndofile 'entities/ent_name2id_freq/ent_name_id.lua'\ndofile 'entities/ent_name2id_freq/e_freq_index.lua'\ndofile 'words/load_w_freq_and_vecs.lua'\ndofile 'entities/pretrained_e2v/e2v.lua'" }, { "path": "deep-ed/deep-ed-master/entities/pretrained_e2v/e2v.lua", "content": "-- Loads pre-trained entity vectors trained using the file entity/learn_e2v/learn_a.lua\n\nassert(opt.ent_vecs_filename)\nprint('==> Loading pre-trained entity vectors: e2v from file ' .. opt.ent_vecs_filename)\n\nassert(opt.entities == 'RLTD', 'Only RLTD entities are currently supported. ALL entities would blow the GPU memory.')\n\n-- Defining variables:\nent_vecs_size = 300\n\ngeom_w2v_M = w2vutils.M:float()\n\ne2vutils = {}\n\n-- Lookup table: ids -> tensor of vecs\ne2vutils.lookup = torch.load(opt.root_data_dir .. 'generated/ent_vecs/' .. opt.ent_vecs_filename)\ne2vutils.lookup = nn.Normalize(2):forward(e2vutils.lookup) -- Needs to be normalized to have norm 1.\n\nassert(e2vutils.lookup:size(1) == get_total_num_ents() and\n e2vutils.lookup:size(2) == ent_vecs_size, e2vutils.lookup:size(1) .. ' ' .. get_total_num_ents())\nassert(e2vutils.lookup[unk_ent_thid]:norm() == 0, e2vutils.lookup[unk_ent_thid]:norm())\n\n-- ent wikiid -> vec\ne2vutils.entwikiid2vec = function(self, ent_wikiid)\n local thid = get_thid(ent_wikiid)\n return self.lookup[thid]:float()\nend\nassert(torch.norm(e2vutils:entwikiid2vec(unk_ent_wikiid)) == 0)\n\n\ne2vutils.entname2vec = function (self,ent_name)\n assert(ent_name)\n return e2vutils:entwikiid2vec(get_ent_wikiid_from_name(ent_name))\nend\n\n-- Entity similarity based on cosine distance (note that entity vectors are normalized).\nfunction entity_similarity(e1_wikiid, e2_wikiid)\n local e1_vec = e2vutils:entwikiid2vec(e1_wikiid)\n local e2_vec = e2vutils:entwikiid2vec(e2_wikiid)\n return e1_vec * e2_vec\nend\n\n\n-----------------------------------------------------------------------\n---- Some unit tests to understand the quality of these embeddings ----\n-----------------------------------------------------------------------\nlocal function geom_top_k_closest_words(ent_name, ent_vec, k)\n local tf_map = ent_wiki_words_4EX[ent_name]\n local w_not_found = {}\n for w,_ in pairs(tf_map) do\n if tf_map[w] >= 10 then\n w_not_found[w] = tf_map[w]\n end\n end\n \n distances = geom_w2v_M * ent_vec\n\n local best_scores, best_word_ids = topk(distances, k)\n local returnwords = {}\n local returndistances = {}\n for i = 1,k do\n local w = get_word_from_id(best_word_ids[i])\n if get_w_id_freq(best_word_ids[i]) >= 200 then\n local w_freq_str = '[fr=' .. get_w_id_freq(best_word_ids[i]) .. ']'\n if is_stop_word_or_number(w) then\n table.insert(returnwords, red(w .. w_freq_str))\n elseif tf_map[w] then\n if tf_map[w] >= 15 then\n table.insert(returnwords, yellow(w .. w_freq_str .. '{tf=' .. tf_map[w] .. '}'))\n else\n table.insert(returnwords, skyblue(w .. w_freq_str .. '{tf=' .. tf_map[w] .. '}'))\n end\n w_not_found[w] = nil\n else\n table.insert(returnwords, w .. w_freq_str)\n end\n assert(best_scores[i] == distances[best_word_ids[i]], best_scores[i] .. ' ' .. distances[best_word_ids[i]])\n table.insert(returndistances, distances[best_word_ids[i]])\n end\n end\n return returnwords, returndistances, w_not_found\nend\n\n\nlocal function geom_most_similar_words_to_ent(ent_name, k)\n local ent_wikiid = get_ent_wikiid_from_name(ent_name)\n local k = k or 1\n local ent_vec = e2vutils:entname2vec(ent_name)\n assert(math.abs(1 - ent_vec:norm()) < 0.01 or ent_vec:norm() == 0, ':::: ' .. ent_vec:norm())\n \n print('\\nTo entity: ' .. blue(ent_name) .. '; vec norm = ' .. ent_vec:norm() .. ':')\n neighbors, scores, w_not_found = geom_top_k_closest_words(ent_name, ent_vec, k)\n print(green('TOP CLOSEST WORDS: ') .. list_with_scores_to_str(neighbors, scores))\n \n local str = yellow('WORDS NOT FOUND: ')\n for w,tf in pairs(w_not_found) do\n if tf >= 20 then\n str = str .. yellow(w .. '{' .. tf .. '}; ')\n else\n str = str .. w .. '{' .. tf .. '}; '\n end\n end\n print('\\n' .. str)\n print('============================================================================')\nend\n\nfunction geom_unit_tests()\n print('\\n' .. yellow('TOP CLOSEST WORDS to a given entity based on cosine distance:')) \n print('For each word, we show the unigram frequency [fr] and the cosine similarity.')\n print('Infrequent words [fr < 500] have noisy embeddings, thus should be trusted less.')\n print('WORDS NOT FOUND contains frequent words from the Wikipedia canonical page that are not found in the TOP CLOSEST WORDS list.')\n \n for i=1,table_len(ent_names_4EX) do\n geom_most_similar_words_to_ent(ent_names_4EX[i], 300)\n end\nend\n\nprint(' Done reading e2v data. Entity vocab size = ' .. e2vutils.lookup:size(1))\n\n\n" }, { "path": "deep-ed/deep-ed-master/entities/pretrained_e2v/e2v_txt_reader.lua", "content": "print('==> loading e2v')\n\nlocal V = torch.ones(get_total_num_ents(), ent_vecs_size):mul(1e-10) -- not zero because of cosine_distance layer\n\nlocal cnt = 0\nfor line in io.lines(e2v_txtfilename) do\n cnt = cnt + 1\n if cnt % 1000000 == 0 then\n print('=======> processed ' .. cnt .. ' lines')\n end\n\n local parts = split(line, ' ')\n assert(table_len(parts) == ent_vecs_size + 1)\n local ent_wikiid = tonumber(parts[1])\n local vec = torch.zeros(ent_vecs_size)\n for i=1,ent_vecs_size do\n vec[i] = tonumber(parts[i + 1])\n end \n \n if (contains_thid(ent_wikiid)) then \n V[get_thid(ent_wikiid)] = vec\n else\n print('Ent id = ' .. ent_wikiid .. ' does not have a vector. ')\n end\nend\n\nprint(' Done loading entity vectors. Size = ' .. cnt .. '\\n')\n\nprint('Writing t7 File for future usage. Next time Ent2Vec will load faster!')\ntorch.save(e2v_t7filename, V)\nprint(' Done saving.\\n')\n\nreturn V" }, { "path": "deep-ed/deep-ed-master/entities/relatedness/filter_wiki_canonical_words_RLTD.lua", "content": "\nif not opt then\n cmd = torch.CmdLine()\n cmd:option('-root_data_dir', '', 'Root path of the data, $DATA_PATH.')\n cmd:text()\n opt = cmd:parse(arg or {})\n assert(opt.root_data_dir ~= '', 'Specify a valid root_data_dir path argument.') \nend\n\n\ndofile 'utils/utils.lua'\ndofile 'entities/relatedness/relatedness.lua'\n\ninput = opt.root_data_dir .. 'generated/wiki_canonical_words.txt'\n\noutput = opt.root_data_dir .. 'generated/wiki_canonical_words_RLTD.txt'\nouf = assert(io.open(output, \"w\"))\n\nprint('\\nStarting dataset filtering.')\n\nlocal cnt = 0\nfor line in io.lines(input) do\n cnt = cnt + 1\n if cnt % 500000 == 0 then\n print(' =======> processed ' .. cnt .. ' lines')\n end\n\n local parts = split(line, '\\t')\n assert(table_len(parts) == 3)\n\n local ent_wikiid = tonumber(parts[1])\n local ent_name = parts[2]\n assert(ent_wikiid)\n \n if rewtr.reltd_ents_wikiid_to_rltdid[ent_wikiid] then\n ouf:write(line .. '\\n')\n end \nend \n\nouf:flush() \nio.close(ouf)\n" }, { "path": "deep-ed/deep-ed-master/entities/relatedness/filter_wiki_hyperlink_contexts_RLTD.lua", "content": "-- Filter all training data s.t. only candidate entities and ground truth entities for which\n-- we have a valid entity embedding are kept.\n\nif not opt then\n cmd = torch.CmdLine()\n cmd:option('-root_data_dir', '', 'Root path of the data, $DATA_PATH.')\n cmd:text()\n opt = cmd:parse(arg or {})\n assert(opt.root_data_dir ~= '', 'Specify a valid root_data_dir path argument.') \nend\n\n\ndofile 'utils/utils.lua'\ndofile 'entities/relatedness/relatedness.lua'\n\ninput = opt.root_data_dir .. 'generated/wiki_hyperlink_contexts.csv'\n\noutput = opt.root_data_dir .. 'generated/wiki_hyperlink_contexts_RLTD.csv'\nouf = assert(io.open(output, \"w\"))\n\nprint('\\nStarting dataset filtering.')\nlocal cnt = 0\nfor line in io.lines(input) do\n cnt = cnt + 1\n if cnt % 50000 == 0 then\n print(' =======> processed ' .. cnt .. ' lines')\n end\n\n local parts = split(line, '\\t')\n local grd_str = parts[table_len(parts)]\n assert(parts[table_len(parts) - 1] == 'GT:')\n local grd_str_parts = split(grd_str, ',')\n\n local grd_pos = tonumber(grd_str_parts[1])\n assert(grd_pos)\n \n local grd_ent_wikiid = tonumber(grd_str_parts[2])\n assert(grd_ent_wikiid)\n \n if rewtr.reltd_ents_wikiid_to_rltdid[grd_ent_wikiid] then\n assert(parts[6] == 'CANDIDATES')\n \n local output_line = parts[1] .. '\\t' .. parts[2] .. '\\t' .. parts[3] .. '\\t' .. parts[4] .. '\\t' .. parts[5] .. '\\t' .. parts[6] .. '\\t'\n \n local new_grd_pos = -1\n local new_grd_str_without_idx = nil\n \n local i = 1\n local added_ents = 0\n while (parts[6 + i] ~= 'GT:') do\n local str = parts[6 + i]\n local str_parts = split(str, ',')\n local ent_wikiid = tonumber(str_parts[1])\n if rewtr.reltd_ents_wikiid_to_rltdid[ent_wikiid] then\n added_ents = added_ents + 1\n output_line = output_line .. str .. '\\t'\n end\n if (i == grd_pos) then\n assert(ent_wikiid == grd_ent_wikiid, 'Error for: ' .. line)\n new_grd_pos = added_ents\n new_grd_str_without_idx = str\n end\n \n i = i + 1\n end\n \n assert(new_grd_pos > 0)\n output_line = output_line .. 'GT:\\t' .. new_grd_pos .. ',' .. new_grd_str_without_idx\n \n ouf:write(output_line .. '\\n')\n end \nend \n\nouf:flush() \nio.close(ouf)\n" }, { "path": "deep-ed/deep-ed-master/entities/relatedness/relatedness.lua", "content": "-- The code in this file does two things:\n-- a) extracts and puts the entity relatedness dataset in two maps (reltd_validate and\n-- reltd_test). Provides functions to evaluate entity embeddings on this dataset\n-- (Table 1 in our paper).\n-- b) extracts all entities that appear in any of the ED (as mention candidates) or\n-- entity relatedness datasets. These are placed in an object called rewtr that will \n-- be used to restrict the set of entities for which we want to train entity embeddings \n-- (done with the file entities/learn_e2v/learn_a.lua).\n\nif not opt then\n cmd = torch.CmdLine()\n cmd:option('-root_data_dir', '', 'Root path of the data, $DATA_PATH.')\n cmd:text()\n opt = cmd:parse(arg or {})\n assert(opt.root_data_dir ~= '', 'Specify a valid root_data_dir path argument.') \nend\n\ndofile 'utils/utils.lua'\ntds = tds or require 'tds'\n\nif not ent_lines_4EX then\n -- Load a few pre-selected entities. For debug and unit tests.\n dofile 'entities/learn_e2v/4EX_wiki_words.lua'\nend\n\n----------------------- Some function definitions ------------\n\n-- Loads the entity relatedness dataset (validation and test parts) in a map called reltd.\n-- Format: reltd = {query_id q -> (query_entity e1, entity_candidates cand) }\n-- cand = {e2 -> label}, where label is binary, if the candidate entity is related to e1\nfunction load_reltd_set(rel_t7filename, rel_txtfilename, set_type)\n print('==> Loading relatedness ' .. set_type)\n if not paths.filep(rel_t7filename) then\n print(' ---> t7 file NOT found. Loading relatedness ' .. set_type .. ' from txt file instead (slower).')\n local reltd = tds.Hash()\n for line in io.lines(rel_txtfilename) do\n local parts = split(line, ' ')\n local label = tonumber(parts[1])\n assert(label == 0 or label == 1)\n \n local t = split(parts[2], ':')\n local q = tonumber(t[2])\n \n local i = 2\n while parts[i] ~= '#' do\n i = i + 1\n end\n i = i + 1\n\n ents = split(parts[i] , '-')\n e1 = tonumber(ents[1])\n e2 = tonumber(ents[2])\n\n if not reltd[q] then\n reltd[q] = tds.Hash()\n reltd[q].e1 = e1\n reltd[q].cand = tds.Hash()\n end\n reltd[q].cand[e2] = label\n end\n\n print(' Done loading relatedness ' .. set_type .. '. Num queries = ' .. table_len(reltd) .. '\\n')\n print('Writing t7 File for future usage. Next time relatedness dataset will load faster!')\n torch.save(rel_t7filename, reltd)\n print(' Done saving.')\n return reltd\n else\n print(' ---> from t7 file.')\n return torch.load(rel_t7filename)\n end\nend\n\n-- Extracts all entities in the relatedness set, either candidates or :\nlocal function extract_reltd_ents(reltd)\n local reltd_ents_direct = tds.Hash()\n for _,v in pairs(reltd) do\n reltd_ents_direct[v.e1] = 1\n for e2,_ in pairs(v.cand) do \n reltd_ents_direct[e2] = 1\n end\n end\n return reltd_ents_direct\nend\n\n-- computes rltd scores based on a given entity_sim function\nlocal function compute_e2v_rltd_scores(reltd, entity_sim)\n local scores = {}\n for q,_ in pairs(reltd) do\n scores[q] = {}\n for e2,_ in pairs(reltd[q].cand) do\n local aux = {}\n aux.e2 = e2\n aux.score = entity_sim(reltd[q].e1, e2)\n table.insert(scores[q], aux)\n end\n table.sort(scores[q], function(a,b) return a.score > b.score end)\n end\n return scores\nend\n\n-- computes rltd scores based on ground truth labels\nlocal function compute_ideal_rltd_scores(reltd)\n local scores = {}\n for q,_ in pairs(reltd) do\n scores[q] = {}\n for e2,label in pairs(reltd[q].cand) do\n local aux = {}\n aux.e2 = e2\n aux.score = label\n table.insert(scores[q], aux)\n end\n table.sort(scores[q], function(a,b) return a.score > b.score end)\n end\n return scores\nend\n\n-- Mean Average Precision: \n-- https://en.wikipedia.org/wiki/Information_retrieval#Mean_average_precision\nlocal function compute_MAP(scores, reltd)\n local sum_avgp = 0.0\n local num_queries = 0\n for q,_ in pairs(scores) do\n local avgp = 0.0\n local num_rel_ents_so_far = 0\n local num_ents_so_far = 0.0\n for _,c in pairs(scores[q]) do\n local e2 = c.e2\n local label = reltd[q].cand[e2]\n num_ents_so_far = num_ents_so_far + 1.0\n if (label == 1) then\n num_rel_ents_so_far = num_rel_ents_so_far + 1\n local precision = num_rel_ents_so_far / num_ents_so_far\n avgp = avgp + precision\n end\n end\n avgp = avgp / num_rel_ents_so_far\n sum_avgp = sum_avgp + avgp\n num_queries = num_queries + 1\n end\n\n assert(num_queries == table_len(reltd))\n return sum_avgp / num_queries\nend\n\n-- NDCG: https://en.wikipedia.org/wiki/Discounted_cumulative_gain#Normalized_DCG\nlocal function compute_DCG(k, q, scores_q, reltd)\n local dcg = 0.0\n local i = 0\n for _,c in pairs(scores_q) do\n local label = reltd[q].cand[c.e2]\n i = i + 1\n if (label == 1) and i <= k then\n dcg = dcg + (1.0 / math.log(math.max(2,i) + 0.0, 2))\n end\n end\n return dcg\nend\n\nlocal function compute_NDCG(k, all_table)\n local sum_ndcg = 0.0\n local num_queries = 0\n for q,_ in pairs(all_table.scores) do\n local dcg = compute_DCG(k, q, all_table.scores[q], all_table.reltd)\n local idcg = compute_DCG(k, q, all_table.ideals_rltd_scores[q], all_table.reltd)\n assert(dcg <= idcg, dcg .. ' ' .. idcg)\n sum_ndcg = sum_ndcg + (dcg / idcg)\n num_queries = num_queries + 1\n end\n\n assert(num_queries == table_len(all_table.reltd))\n return sum_ndcg / num_queries\nend\n\nlocal function compute_relatedness_metrics_from_maps(entity_sim, validate_set, test_set)\n print(yellow('Entity Relatedness quality measure:'))\n \n collectgarbage(); collectgarbage();\n local ideals_rltd_validate_scores = compute_ideal_rltd_scores(validate_set)\n collectgarbage(); collectgarbage();\n local ideals_rltd_test_scores = compute_ideal_rltd_scores(test_set)\n collectgarbage(); collectgarbage();\n \n assert(math.abs(-1 + compute_MAP(ideals_rltd_validate_scores, validate_set)) < 0.001) \n collectgarbage(); collectgarbage(); \n assert(math.abs(-1 + compute_MAP(ideals_rltd_test_scores, test_set)) < 0.001) \n\n collectgarbage(); collectgarbage();\n local scores_validate = compute_e2v_rltd_scores(validate_set, entity_sim)\n collectgarbage(); collectgarbage();\n local scores_test = compute_e2v_rltd_scores(test_set, entity_sim)\n collectgarbage(); collectgarbage();\n\n local validate_table = {}\n validate_table.scores = scores_validate\n validate_table.ideals_rltd_scores = ideals_rltd_validate_scores\n validate_table.reltd = validate_set\n local test_table = {}\n test_table.scores = scores_test\n test_table.ideals_rltd_scores = ideals_rltd_test_scores\n test_table.reltd = test_set\n\n local map_validate = compute_MAP(scores_validate, validate_set)\n collectgarbage(); collectgarbage();\n local ndcg_1_validate = compute_NDCG(1, validate_table)\n collectgarbage(); collectgarbage();\n local ndcg_5_validate = compute_NDCG(5, validate_table)\n collectgarbage(); collectgarbage();\n local ndcg_10_validate = compute_NDCG(10, validate_table)\n collectgarbage(); collectgarbage();\n local total = map_validate + ndcg_1_validate + ndcg_5_validate + ndcg_10_validate\n\n local map_validate_str = blue_num_str(map_validate)\n local ndcg_1_validate_str = blue_num_str(ndcg_1_validate)\n local ndcg_5_validate_str = blue_num_str(ndcg_5_validate)\n local ndcg_10_validate_str = blue_num_str(ndcg_10_validate) \n local total_str = blue_num_str(total)\n\n local map_test = red(string.format(\"%.3f\", compute_MAP(scores_test, test_set)))\n collectgarbage(); collectgarbage();\n local ndcg_1_test = red(string.format(\"%.3f\", compute_NDCG(1, test_table)))\n collectgarbage(); collectgarbage();\n local ndcg_5_test = red(string.format(\"%.3f\", compute_NDCG(5, test_table)))\n collectgarbage(); collectgarbage();\n local ndcg_10_test = red(string.format(\"%.3f\", compute_NDCG(10, test_table)))\n collectgarbage(); collectgarbage();\n\n print(yellow('measure ='), 'NDCG1' , 'NDCG5', 'NDCG10', 'MAP', 'TOTAL VALIDATION')\n print(yellow('our (vald) ='), ndcg_1_validate_str, ndcg_5_validate_str, ndcg_10_validate_str, map_validate_str, total_str) \n print(yellow('our (test) ='), ndcg_1_test, ndcg_5_test, ndcg_10_test, map_test) \n print(yellow('Yamada\\'16 ='), 0.59, 0.56, 0.59, 0.52) \n print(yellow('WikiMW ='), 0.54, 0.52, 0.55, 0.48) \nend\n\n\n--------------------------------------------------------------\n------------------------ Main code ---------------------------\n--------------------------------------------------------------\nrel_test_txtfilename = opt.root_data_dir .. 'basic_data/relatedness/test.svm'\nrel_test_t7filename = opt.root_data_dir .. 'generated/relatedness_test.t7'\n\nrel_validate_txtfilename = opt.root_data_dir .. 'basic_data/relatedness/validate.svm'\nrel_validate_t7filename = opt.root_data_dir .. 'generated/relatedness_validate.t7'\n\nlocal reltd_validate = load_reltd_set(rel_validate_t7filename, rel_validate_txtfilename, 'validate')\nlocal reltd_test = load_reltd_set(rel_test_t7filename, rel_test_txtfilename, 'test')\n\nlocal reltd_ents_direct_validate = extract_reltd_ents(reltd_validate)\nlocal reltd_ents_direct_test = extract_reltd_ents(reltd_test)\n\n\nlocal rewtr_t7filename = opt.root_data_dir .. 'generated/all_candidate_ents_ed_rltd_datasets_RLTD.t7'\nprint('==> Loading relatedness thid tensor')\nif not paths.filep(rewtr_t7filename) then\n print(' ---> t7 file NOT found. Loading reltd_ents_wikiid_to_rltdid from txt file instead (slower).')\n\n -- Gather the restricted set of entities for which we train entity embeddings:\n local rltd_all_ent_wikiids = tds.Hash()\n \n -- 1) From the relatedness dataset\n for ent_wikiid,_ in pairs(reltd_ents_direct_validate) do\n rltd_all_ent_wikiids[ent_wikiid] = 1\n end\n for ent_wikiid,_ in pairs(reltd_ents_direct_test) do\n rltd_all_ent_wikiids[ent_wikiid] = 1\n end\n \n -- 1.1) From a small dataset (used for debugging / unit testing).\n for _,line in pairs(ent_lines_4EX) do\n local parts = split(line, '\\t')\n assert(table_len(parts) == 3)\n ent_wikiid = tonumber(parts[1])\n assert(ent_wikiid)\n rltd_all_ent_wikiids[ent_wikiid] = 1\n end\n\n -- end2end_neural_el start insert my entity universe\n entities_file_path = '../../data/entities/entities_universe.txt'\n wikiid2nnid_folder_path = '../../data/entities/wikiid2nnid/'\n \n --entities_file_path = '../../data/entities/extension_entities/extension_entities.txt'\n --wikiid2nnid_folder_path = '../../data/entities/extension_entities/wikiid2nnid/'\n\n it, _ = io.open(entities_file_path)\n local line = it:read()\n while line do\n local parts = split(line, '\\t')\n local ent_wikiid = tonumber(parts[1])\n assert(ent_wikiid)\n rltd_all_ent_wikiids[ent_wikiid] = 1\n line = it:read()\n end\n -- end2end_neural_el end\n\n--[=====[ \n -- 2) From all ED datasets:\n local files = {'aida_train.csv', 'aida_testA.csv', 'aida_testB.csv', \n 'wned-aquaint.csv', 'wned-msnbc.csv', 'wned-ace2004.csv',\n 'wned-clueweb.csv', 'wned-wikipedia.csv'}\n for _,f in pairs(files) do\n it, _ = io.open(opt.root_data_dir .. 'generated/test_train_data/' .. f)\n local line = it:read()\n while line do\n local parts = split(line, '\\t')\n assert(parts[6] == 'CANDIDATES')\n assert(parts[table_len(parts) - 1] == 'GT:')\n if parts[7] ~= 'EMPTYCAND' then\n for i = 7, table_len(parts) - 2 do\n local p = split(parts[i], ',')\n local ent_wikiid = tonumber(p[1])\n assert(ent_wikiid)\n rltd_all_ent_wikiids[ent_wikiid] = 1\n end\n\n local p = split(parts[table_len(parts)], ',')\n if table_len(p) >= 2 then\n local ent_wikiid = tonumber(p[2])\n assert(ent_wikiid)\n end\n end\n line = it:read()\n end\n end\n--]=====]\n\n -- Insert unk_ent_wikiid\n local unk_ent_wikiid = 1\n rltd_all_ent_wikiids[unk_ent_wikiid] = 1\n \n -- Sort all wikiids\n local sorted_rltd_all_ent_wikiids = tds.Vec()\n for ent_wikiid,_ in pairs(rltd_all_ent_wikiids) do\n sorted_rltd_all_ent_wikiids:insert(ent_wikiid) -- push back to the vector\n end\n sorted_rltd_all_ent_wikiids:sort(function(a,b) return a < b end) -- sort the vector\n \n local reltd_ents_wikiid_to_rltdid = tds.Hash()\n for rltd_id,wikiid in pairs(sorted_rltd_all_ent_wikiids) do\n reltd_ents_wikiid_to_rltdid[wikiid] = rltd_id\n end\n \n rewtr = tds.Hash()\n rewtr.reltd_ents_wikiid_to_rltdid = reltd_ents_wikiid_to_rltdid\n rewtr.reltd_ents_rltdid_to_wikiid = sorted_rltd_all_ent_wikiids\n rewtr.num_rltd_ents = #sorted_rltd_all_ent_wikiids\n \n -- end2end_neural_el print these to file as well\n print('Now printing to file the wikiid to torch ids mappings...')\n out_file = wikiid2nnid_folder_path .. 'wikiid2nnid.txt'\n --out_file = opt.root_data_dir .. 'generated/wikiid2nnid/wikiid2nnid.txt' -- 'generated/my_wikiid_nnid_map/reltd_ents_wikiid_to_rltdid.txt'\n ouf = assert(io.open(out_file, \"w\"))\n for k, v in pairs(reltd_ents_wikiid_to_rltdid) do\n -- print(k, v)\n ouf:write(k .. '\\t' .. v .. '\\n')\n end\n ouf:flush()\n io.close(ouf) \n out_file = wikiid2nnid_folder_path .. 'nnid2wikiid.txt'\n --out_file = opt.root_data_dir .. 'generated/wikiid2nnid/nnid2wikiid.txt'\n ouf = assert(io.open(out_file, \"w\"))\n for k, v in pairs(sorted_rltd_all_ent_wikiids) do\n -- print(k, v)\n ouf:write(k .. '\\t' .. v .. '\\n')\n end\n ouf:flush()\n io.close(ouf) \n -- end2end_neural_el\n\n\n print('Writing reltd_ents_wikiid_to_rltdid to t7 File for future usage.')\n torch.save(rewtr_t7filename, rewtr)\n print(' Done saving.')\nelse\n print(' ---> from t7 file.')\n rewtr = torch.load(rewtr_t7filename)\nend\n\nprint(' Done loading relatedness sets. Num queries test = ' .. table_len(reltd_test) .. \n '. Num queries valid = ' .. table_len(reltd_validate) ..\n '. Total num ents restricted set = ' .. rewtr.num_rltd_ents)\n\n-- Main function that computes results for the entity relatedness dataset (Table 1 of \n-- our paper) given any entity similarity function as input. \nfunction compute_relatedness_metrics(entity_sim)\n compute_relatedness_metrics_from_maps(entity_sim, reltd_validate, reltd_test)\nend\n" }, { "path": "deep-ed/deep-ed-master/log_train_entity_vecs", "content": "" }, { "path": "deep-ed/deep-ed-master/our_system_annotations.txt", "content": "=====> TESTING <=== \t\n\n===> AQUAINT; num mentions = 727\t\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19981109_0440.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = the Hague ==> ENTITY: \u001b[32mThe Hague\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.070:-0.070[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.048]; ambassador[0.044]; ICJ[0.043]; Court[0.043]; government[0.042]; Europe[0.042]; interfere[0.042]; political[0.042]; Monday[0.042]; filed[0.042]; Tuesday[0.042]; International[0.041]; saying[0.041]; cooperation[0.041]; economic[0.041]; protest[0.041]; protest[0.041]; Free[0.041]; organizations[0.041]; affairs[0.041]; Justice[0.040]; said[0.040]; Foreign[0.040]; Prague[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Prague ==> ENTITY: \u001b[32mPrague\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.053:-0.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPrague[0.048]; Germany[0.043]; Czech[0.043]; Czech[0.043]; Soviet[0.042]; Tuesday[0.042]; Munich[0.042]; set[0.042]; Europe[0.042]; Europe[0.042]; Friday[0.041]; Monday[0.041]; English[0.041]; August[0.041]; ambassador[0.041]; began[0.041]; began[0.041]; protest[0.041]; protest[0.041]; sponsored[0.041]; Radio[0.041]; Radio[0.041]; Iraq[0.040]; Iraq[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cold War\u001b[39m ==> ENTITY: \u001b[32mCold War\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.046]; countries[0.045]; Iraq[0.044]; Iraq[0.044]; economic[0.043]; American[0.043]; Iran[0.043]; Iran[0.043]; Communist[0.043]; communism[0.043]; Europe[0.042]; cooperation[0.042]; period[0.041]; government[0.041]; political[0.041]; Germany[0.040]; following[0.040]; West[0.040]; collapse[0.040]; service[0.040]; program[0.039]; weeks[0.039]; Prague[0.039]; Prague[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Munich, Germany\u001b[39m ==> ENTITY: \u001b[32mMunich\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.107:0.107[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurope[0.044]; Prague[0.044]; Prague[0.044]; Prague[0.044]; American[0.043]; Czech[0.043]; Czech[0.043]; headquarters[0.042]; War[0.042]; communism[0.042]; service[0.041]; government[0.041]; government[0.041]; August[0.041]; vj[0.041]; Friday[0.041]; began[0.041]; began[0.041]; sponsored[0.041]; ambassador[0.041]; program[0.040]; Soviet[0.040]; following[0.040]; earlier[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.143:0.143[0]; log p(e|m)= -1.207:-1.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlanguage[0.046]; language[0.046]; American[0.044]; English[0.044]; Prague[0.043]; Prague[0.043]; Czech[0.043]; Germany[0.043]; Soviet[0.043]; Europe[0.042]; Europe[0.042]; ambassador[0.042]; internal[0.041]; service[0.041]; radio[0.040]; radio[0.040]; countries[0.040]; Radio[0.040]; Radio[0.040]; economic[0.040]; assistance[0.039]; August[0.039]; Munich[0.039]; Netherlands[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = the Netherlands ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.103:0.103[0]; log p(e|m)= -0.176:-0.176[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.049]; Europe[0.048]; Europe[0.048]; day[0.045]; approved[0.044]; saying[0.044]; government[0.043]; government[0.043]; English[0.043]; Munich[0.043]; service[0.043]; ambassador[0.043]; August[0.043]; minute[0.043]; weeks[0.043]; affairs[0.042]; International[0.042]; level[0.042]; Hague[0.042]; internal[0.042]; Czech[0.042]; Czech[0.042]; Tuesday[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = collapse of communism\u001b[39m ==> ENTITY: \u001b[32mRevolutions of 1989\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.047]; government[0.045]; Europe[0.045]; Communist[0.043]; Prague[0.043]; Germany[0.043]; Czech[0.042]; program[0.042]; period[0.042]; American[0.042]; Iron[0.042]; War[0.041]; Iraq[0.041]; Iraq[0.041]; countries[0.041]; following[0.041]; West[0.041]; news[0.040]; August[0.040]; weeks[0.040]; Munich[0.040]; East[0.040]; Iran[0.040]; Iran[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Soviet ==> ENTITY: \u001b[32mSoviet Union\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.098:-0.098[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncommunism[0.044]; Communist[0.043]; War[0.043]; August[0.042]; language[0.042]; Germany[0.042]; program[0.042]; American[0.042]; ambassador[0.042]; Europe[0.042]; Prague[0.042]; Prague[0.042]; Prague[0.042]; following[0.041]; West[0.041]; protest[0.041]; Iraq[0.041]; Iraq[0.041]; cooperation[0.041]; Iran[0.041]; Iran[0.041]; Czech[0.041]; Czech[0.041]; Farsi[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = broadcast ==> ENTITY: \u001b[32mBroadcasting\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.201:-0.201[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbroadcast[0.050]; radio[0.046]; radio[0.046]; news[0.044]; service[0.044]; Radio[0.044]; saying[0.043]; Europe[0.042]; Europe[0.042]; Germany[0.041]; Iraq[0.041]; Iraq[0.041]; transmitting[0.041]; English[0.041]; countries[0.040]; said[0.039]; Tuesday[0.039]; Iran[0.039]; Iran[0.039]; Iran[0.039]; Iran[0.039]; Iran[0.039]; Iran[0.039]; Monday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iran ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.171:0.171[0]; log p(e|m)= -0.138:-0.138[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.052]; Iran[0.052]; Iran[0.052]; Iran[0.052]; saying[0.045]; Kharrazi[0.045]; government[0.043]; reported[0.043]; Europe[0.043]; ambassador[0.043]; quoted[0.042]; said[0.042]; organizations[0.042]; Foreign[0.041]; International[0.041]; assistance[0.041]; Radio[0.041]; protest[0.041]; protest[0.041]; internal[0.041]; Monday[0.040]; global[0.040]; English[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kamal Kharrazi\u001b[39m ==> ENTITY: \u001b[32mKamal Kharazi\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.046]; Iran[0.046]; Iran[0.046]; Iran[0.046]; Iran[0.046]; Iraq[0.045]; Foreign[0.044]; American[0.042]; ambassador[0.042]; affairs[0.042]; International[0.041]; Monday[0.041]; Minister[0.041]; weeks[0.041]; Tuesday[0.040]; reported[0.040]; Europe[0.040]; cooperation[0.039]; Daily[0.039]; English[0.039]; language[0.039]; political[0.039]; government[0.039]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iron Curtain\u001b[39m ==> ENTITY: \u001b[32mIron Curtain\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.047]; Europe[0.044]; Communist[0.043]; countries[0.043]; Germany[0.043]; American[0.042]; communism[0.042]; West[0.042]; cooperation[0.042]; Prague[0.041]; Prague[0.041]; War[0.041]; economic[0.041]; vj[0.041]; East[0.041]; broadcast[0.041]; broadcast[0.041]; government[0.041]; government[0.041]; political[0.041]; period[0.041]; Czech[0.041]; Czech[0.041]; service[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.174:0.174[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.052]; government[0.044]; government[0.044]; cooperation[0.043]; countries[0.043]; American[0.042]; Iran[0.042]; Iran[0.042]; Iran[0.042]; Iran[0.042]; Iran[0.042]; War[0.041]; economic[0.041]; said[0.041]; Foreign[0.040]; assistance[0.040]; West[0.040]; saying[0.040]; protest[0.040]; Kharrazi[0.040]; following[0.040]; Europe[0.040]; political[0.040]; Germany[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = ambassador ==> ENTITY: \u001b[32mAmbassador\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.101:0.101[0]; log p(e|m)= -0.027:-0.027[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.046]; cooperation[0.046]; Europe[0.045]; Europe[0.045]; Foreign[0.044]; broadcast[0.044]; broadcast[0.044]; Iran[0.044]; Iran[0.044]; Iran[0.044]; Iran[0.044]; Iran[0.044]; Iran[0.044]; Iran[0.044]; Iraq[0.044]; Iraq[0.044]; organizations[0.043]; economic[0.043]; Germany[0.042]; Netherlands[0.042]; International[0.042]; saying[0.041]; global[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Communist ==> ENTITY: \u001b[32mCommunism\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.393:-0.393[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.046]; communism[0.045]; Europe[0.045]; countries[0.045]; American[0.044]; government[0.043]; economic[0.043]; political[0.043]; Germany[0.042]; Czech[0.042]; Prague[0.041]; Prague[0.041]; cooperation[0.041]; following[0.040]; Iraq[0.040]; Iraq[0.040]; service[0.040]; August[0.040]; moved[0.040]; uncensored[0.040]; West[0.040]; news[0.040]; Iran[0.039]; Iran[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Radio Free Europe\u001b[39m ==> ENTITY: \u001b[32mRadio Free Europe/Radio Liberty\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbroadcast[0.045]; quoted[0.043]; said[0.043]; radio[0.043]; radio[0.043]; ambassador[0.042]; recalled[0.042]; Prague[0.042]; Prague[0.042]; Czech[0.042]; set[0.042]; American[0.042]; internal[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; International[0.041]; Monday[0.041]; Tuesday[0.040]; reported[0.040]; affairs[0.040]; political[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Farsi ==> ENTITY: \u001b[32mPersian language\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.171:-0.171[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlanguage[0.048]; language[0.048]; English[0.046]; broadcast[0.043]; broadcast[0.043]; Iran[0.042]; Iran[0.042]; Iran[0.042]; Iran[0.042]; American[0.041]; Iraq[0.041]; Iraq[0.041]; radio[0.041]; station[0.041]; station[0.041]; say[0.041]; say[0.041]; countries[0.040]; Radio[0.040]; West[0.040]; cooperation[0.039]; daily[0.039]; government[0.039]; government[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = ICJ\u001b[39m ==> ENTITY: \u001b[32mInternational Court of Justice\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCourt[0.045]; saying[0.044]; Hague[0.043]; Justice[0.042]; Iraq[0.042]; Iraq[0.042]; trial[0.042]; Foreign[0.042]; International[0.042]; political[0.042]; Europe[0.041]; Europe[0.041]; government[0.041]; government[0.041]; Netherlands[0.041]; filed[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; ambassador[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Czech government\u001b[39m ==> ENTITY: \u001b[32mPolitics of the Czech Republic\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.115:0.115[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nICJ[0.044]; government[0.044]; Czech[0.044]; political[0.043]; ambassador[0.043]; cooperation[0.043]; Prague[0.043]; Prague[0.043]; Prague[0.043]; vj[0.043]; Communist[0.042]; countries[0.041]; program[0.041]; organizations[0.041]; Germany[0.041]; Munich[0.041]; communism[0.041]; Cold[0.040]; ti[0.040]; Europe[0.040]; said[0.040]; Curtain[0.040]; protest[0.040]; Soviet[0.040]; \t\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19980917_0818.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = cult ==> ENTITY: \u001b[32mCult\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.526:-0.526[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncult[0.049]; cult[0.049]; Cult[0.046]; Attorneys[0.043]; court[0.042]; prosecutors[0.042]; prosecutors[0.042]; allowed[0.041]; people[0.041]; accused[0.041]; charges[0.041]; charges[0.041]; lawyers[0.041]; charged[0.041]; trial[0.040]; trial[0.040]; trial[0.040]; said[0.040]; denounced[0.040]; witnesses[0.040]; March[0.040]; testimony[0.040]; leaders[0.040]; thousands[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tokyo ==> ENTITY: \u001b[32mTokyo\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.051:0.051[0]; log p(e|m)= -0.093:-0.093[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMarch[0.045]; Thursday[0.045]; Watanabe[0.045]; charges[0.043]; charges[0.043]; Shoko[0.043]; court[0.042]; court[0.042]; subways[0.042]; chief[0.041]; Kyo[0.041]; Cult[0.041]; attack[0.041]; attack[0.041]; thousands[0.041]; Osamu[0.041]; charged[0.040]; trial[0.040]; trial[0.040]; trial[0.040]; leaders[0.040]; Asahara[0.040]; Asahara[0.040]; allowed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = prosecution ==> ENTITY: \u001b[32mProsecutor\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.039:-0.039[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrial[0.048]; proceedings[0.044]; court[0.044]; court[0.044]; prosecutors[0.044]; prosecutors[0.044]; Court[0.043]; lawyer[0.042]; charged[0.042]; accused[0.041]; saying[0.041]; said[0.041]; witness[0.041]; timing[0.041]; defense[0.040]; defense[0.040]; told[0.040]; told[0.040]; similar[0.040]; allowed[0.040]; quoted[0.040]; statement[0.040]; say[0.040]; April[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mgassing\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mChemical warfare\u001b[39m <---> \u001b[32mGas chamber\u001b[39m\t\nSCORES: global= 0.243:0.238[\u001b[31m0.005\u001b[39m]; local()= 0.137:0.131[\u001b[31m0.006\u001b[39m]; log p(e|m)= -2.112:-0.329[\u001b[32m1.783\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngas[0.052]; sarin[0.046]; nerve[0.043]; attack[0.042]; attack[0.042]; attack[0.042]; attack[0.042]; gassing[0.042]; said[0.042]; killed[0.041]; thousands[0.041]; trial[0.041]; trial[0.041]; trial[0.041]; Thursday[0.040]; charges[0.040]; charges[0.040]; people[0.040]; key[0.040]; accused[0.040]; accused[0.040]; told[0.040]; told[0.040]; witnesses[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sarin attack\u001b[39m ==> ENTITY: \u001b[32mSarin gas attack on the Tokyo subway\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsubway[0.046]; Tokyo[0.045]; Tokyo[0.045]; subways[0.045]; attack[0.045]; attack[0.045]; Aum[0.045]; nerve[0.045]; Asahara[0.044]; Asahara[0.044]; Asahara[0.044]; cult[0.043]; cult[0.043]; cult[0.043]; allowed[0.042]; Cult[0.042]; followers[0.042]; cultist[0.042]; charges[0.042]; charges[0.042]; Watanabe[0.042]; Watanabe[0.042]; Watanabe[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mdefense\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMilitary\u001b[39m <---> \u001b[32mDefense (legal)\u001b[39m\t\nSCORES: global= 0.247:0.246[\u001b[31m0.000\u001b[39m]; local()= 0.091:0.182[\u001b[32m0.091\u001b[39m]; log p(e|m)= -1.398:-1.082[\u001b[32m0.317\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrial[0.044]; trial[0.044]; trial[0.044]; lawyers[0.044]; Attorneys[0.043]; prosecution[0.042]; attack[0.042]; attack[0.042]; attack[0.042]; attack[0.042]; lawyer[0.042]; proceedings[0.042]; court[0.041]; court[0.041]; court[0.041]; prosecutors[0.041]; prosecutors[0.041]; prosecutors[0.041]; charges[0.041]; charges[0.041]; allowed[0.041]; negligent[0.040]; key[0.040]; contradicted[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = nerve gas\u001b[39m ==> ENTITY: \u001b[32mNerve agent\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.091:0.091[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsarin[0.053]; attack[0.044]; attack[0.044]; sickened[0.043]; thousands[0.042]; gassing[0.042]; tolerated[0.041]; said[0.041]; Aum[0.041]; chief[0.041]; killed[0.041]; Tokyo[0.041]; charged[0.041]; subways[0.041]; allowed[0.041]; abnormal[0.040]; Asahara[0.040]; Asahara[0.040]; defense[0.040]; people[0.040]; accused[0.040]; cult[0.040]; cult[0.040]; cult[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Aum Shinri Kyo\u001b[39m ==> ENTITY: \u001b[32mAum Shinrikyo\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncult[0.044]; cult[0.044]; cult[0.044]; masterminding[0.043]; attack[0.042]; attack[0.042]; attack[0.042]; Cult[0.042]; Asahara[0.042]; Asahara[0.042]; Asahara[0.042]; sarin[0.042]; charges[0.041]; charges[0.041]; nerve[0.041]; Tokyo[0.041]; Tokyo[0.041]; Thursday[0.041]; March[0.041]; March[0.041]; subway[0.040]; accused[0.040]; accused[0.040]; denounced[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shoko Asahara\u001b[39m ==> ENTITY: \u001b[32mShoko Asahara\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAum[0.043]; Asahara[0.043]; Asahara[0.043]; trial[0.043]; trial[0.043]; trial[0.043]; cult[0.042]; cult[0.042]; cult[0.042]; subway[0.042]; charges[0.042]; charges[0.042]; Tokyo[0.042]; Tokyo[0.042]; sarin[0.042]; subways[0.041]; masterminding[0.041]; attack[0.041]; attack[0.041]; attack[0.041]; nerve[0.041]; Cult[0.040]; charged[0.040]; lawyers[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tokyo District Court\u001b[39m ==> ENTITY: \u001b[32mTokyo District Court\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncourt[0.051]; court[0.051]; March[0.049]; March[0.049]; April[0.049]; negligent[0.047]; Thursday[0.046]; statement[0.046]; proceedings[0.045]; charges[0.045]; charges[0.045]; trial[0.045]; trial[0.045]; trial[0.045]; charged[0.044]; refused[0.043]; cw[0.043]; saying[0.043]; Shoko[0.043]; accused[0.043]; Aum[0.043]; chief[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = guru ==> ENTITY: \u001b[32mGuru\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.078:0.078[0]; log p(e|m)= -0.023:-0.023[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nguru[0.046]; people[0.046]; Thursday[0.044]; leaders[0.043]; lawyers[0.043]; said[0.042]; leader[0.042]; Cult[0.042]; cult[0.042]; cult[0.042]; cult[0.042]; March[0.042]; thousands[0.042]; unfair[0.041]; chief[0.041]; accused[0.041]; charges[0.041]; charges[0.041]; lawyer[0.041]; trial[0.040]; trial[0.040]; denounced[0.040]; testimony[0.040]; charged[0.039]; \t\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19990827_0137.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = pornographic ==> ENTITY: \u001b[32mPornography\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.129:0.129[0]; log p(e|m)= -0.134:-0.134[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsex[0.049]; videos[0.048]; sexual[0.046]; material[0.043]; release[0.043]; circulated[0.043]; inappropriate[0.042]; included[0.042]; says[0.041]; conduct[0.041]; Charlotte[0.041]; Misuse[0.041]; creates[0.040]; mail[0.040]; mail[0.040]; sending[0.040]; tolerate[0.040]; reported[0.040]; company[0.040]; company[0.040]; performance[0.040]; employees[0.040]; employees[0.040]; person[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mhostile environment\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIslamophobia\u001b[39m <---> \u001b[32mHostile environment sexual harassment\u001b[39m\t\nSCORES: global= 0.275:0.265[\u001b[31m0.009\u001b[39m]; local()= 0.077:0.132[\u001b[32m0.055\u001b[39m]; log p(e|m)= 0.000:-1.427[\u001b[31m1.427\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntolerate[0.044]; intimidating[0.044]; interferes[0.043]; conduct[0.043]; inappropriate[0.043]; employee[0.043]; employee[0.043]; fired[0.042]; fired[0.042]; fired[0.042]; University[0.042]; employees[0.042]; employees[0.042]; newspaper[0.041]; reported[0.041]; sexual[0.041]; messages[0.041]; nation[0.041]; people[0.041]; sending[0.040]; creates[0.040]; heavy[0.040]; week[0.040]; week[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fortune 100\u001b[39m ==> ENTITY: \u001b[32mFortune 500\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCompanies[0.048]; companies[0.047]; company[0.046]; company[0.046]; company[0.046]; employees[0.045]; employee[0.043]; employee[0.043]; corporation[0.042]; seeking[0.041]; Charlotte[0.041]; University[0.041]; based[0.040]; number[0.040]; number[0.040]; past[0.040]; systems[0.040]; reported[0.039]; division[0.039]; included[0.039]; increasing[0.039]; bank[0.039]; office[0.038]; office[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = home-equity\u001b[39m ==> ENTITY: \u001b[32mHome equity\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.120:0.120[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUniversity[0.045]; bank[0.044]; bank[0.044]; Companies[0.043]; performance[0.043]; seeking[0.043]; person[0.042]; creates[0.042]; spot[0.042]; employees[0.042]; companies[0.042]; volume[0.041]; office[0.041]; office[0.041]; division[0.041]; number[0.041]; number[0.041]; employee[0.041]; employee[0.041]; based[0.041]; said[0.041]; called[0.040]; included[0.040]; heavy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = e-mail ==> ENTITY: \u001b[32mEmail\u001b[39m\t\nSCORES: global= 0.239:0.239[0]; local()= 0.137:0.137[0]; log p(e|m)= -2.263:-2.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsending[0.047]; mail[0.046]; mail[0.046]; videos[0.043]; employees[0.043]; employees[0.043]; Mail[0.042]; pornographic[0.042]; circulated[0.042]; material[0.042]; environment[0.041]; company[0.041]; company[0.041]; office[0.041]; included[0.041]; release[0.040]; week[0.040]; largest[0.040]; work[0.040]; says[0.040]; inappropriate[0.040]; creates[0.040]; said[0.039]; details[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = server\u001b[39m ==> ENTITY: \u001b[32mServer (computing)\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncompanies[0.046]; employee[0.045]; employee[0.045]; based[0.045]; mails[0.044]; messages[0.044]; employees[0.044]; work[0.044]; work[0.044]; stack[0.044]; company[0.043]; company[0.043]; company[0.043]; number[0.043]; number[0.043]; included[0.043]; Companies[0.043]; office[0.042]; office[0.042]; videos[0.042]; having[0.042]; release[0.042]; volume[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = The Charlotte Observer\u001b[39m ==> ENTITY: \u001b[32mThe Charlotte Observer\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnewspaper[0.051]; said[0.047]; University[0.047]; Friday[0.046]; Friday[0.046]; told[0.046]; week[0.046]; week[0.046]; reported[0.046]; nation[0.045]; CHARLOTTE[0.045]; seven[0.045]; videos[0.045]; AP[0.045]; past[0.045]; joins[0.044]; says[0.044]; Fortune[0.044]; Mail[0.044]; division[0.044]; sixth[0.044]; work[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = sexual harassment\u001b[39m ==> ENTITY: \u001b[32mSexual harassment\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.142:0.142[0]; log p(e|m)= -0.005:-0.005[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nalleging[0.046]; employee[0.045]; employee[0.045]; claims[0.044]; claims[0.044]; week[0.043]; seeking[0.043]; messages[0.042]; action[0.041]; lawsuits[0.041]; told[0.041]; shown[0.041]; policy[0.041]; work[0.041]; newspaper[0.041]; called[0.041]; computer[0.041]; legal[0.040]; environment[0.040]; Union[0.040]; number[0.040]; number[0.040]; joins[0.040]; taking[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = First Union ==> ENTITY: \u001b[32mFirst Union\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.161:0.161[0]; log p(e|m)= -0.038:-0.038[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBank[0.047]; bank[0.046]; bank[0.046]; Corp[0.045]; CHARLOTTE[0.044]; Union[0.044]; Charlotte[0.044]; company[0.042]; University[0.042]; largest[0.042]; nation[0.042]; office[0.041]; employees[0.041]; employees[0.041]; week[0.041]; seven[0.041]; included[0.040]; Friday[0.040]; Friday[0.040]; sixth[0.040]; sending[0.039]; reported[0.038]; mail[0.038]; mail[0.038]; \t\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19981119_0585.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = bear ==> ENTITY: \u001b[32mBear\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.297:-0.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhunting[0.048]; bear[0.048]; deer[0.044]; sheep[0.043]; male[0.043]; fishing[0.042]; black[0.042]; bruin[0.042]; pounds[0.042]; livestock[0.042]; areas[0.041]; nicknamed[0.041]; giant[0.041]; taste[0.041]; western[0.041]; western[0.041]; kilograms[0.040]; devoured[0.040]; said[0.040]; said[0.040]; said[0.040]; cattle[0.039]; cattle[0.039]; inhabited[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dracula ==> ENTITY: \u001b[32mDracula\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.483:-0.483[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDracula[0.048]; Dracula[0.048]; Dracula[0.048]; vampire[0.045]; named[0.042]; inhabited[0.042]; appetite[0.042]; Transylvania[0.041]; black[0.041]; devoured[0.041]; bears[0.041]; bears[0.041]; nicknamed[0.041]; legend[0.041]; count[0.041]; hunting[0.040]; hunting[0.040]; cold[0.040]; hibernate[0.040]; kills[0.040]; giant[0.039]; apparently[0.039]; sheep[0.039]; sheep[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mdictator\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFrancisco Franco\u001b[39m <---> \u001b[32mDictator\u001b[39m\t\nSCORES: global= 0.246:0.242[\u001b[31m0.003\u001b[39m]; local()= 0.095:0.098[\u001b[32m0.002\u001b[39m]; log p(e|m)= -1.604:-0.503[\u001b[32m1.102\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncommunist[0.045]; Ceausescu[0.043]; said[0.043]; said[0.043]; decade[0.042]; State[0.042]; inhabited[0.042]; law[0.042]; Romania[0.042]; Romania[0.042]; Romania[0.042]; past[0.042]; named[0.042]; executed[0.041]; apparently[0.041]; weeks[0.041]; living[0.041]; western[0.041]; western[0.041]; Nicolae[0.041]; toppled[0.041]; Dracula[0.041]; Dracula[0.041]; state[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = weaned ==> ENTITY: \u001b[32mWeaning\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.173:0.173[0]; log p(e|m)= -0.031:-0.031[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncubs[0.045]; cows[0.044]; kilograms[0.043]; hibernate[0.043]; pounds[0.043]; appetite[0.043]; seven[0.042]; seven[0.042]; sheep[0.042]; male[0.042]; weeks[0.042]; weeks[0.042]; State[0.041]; fed[0.041]; meat[0.041]; meat[0.041]; meat[0.041]; grow[0.041]; cold[0.041]; onset[0.040]; bears[0.040]; bears[0.040]; bears[0.040]; protected[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Transylvania ==> ENTITY: \u001b[32mTransylvania\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.166:0.166[0]; log p(e|m)= -0.067:-0.067[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwestern[0.048]; western[0.048]; inhabited[0.046]; Romania[0.046]; Romania[0.046]; Romania[0.046]; regions[0.045]; peasants[0.044]; state[0.044]; forested[0.043]; apparently[0.043]; Ceausescu[0.043]; areas[0.042]; moved[0.042]; Dracula[0.042]; Dracula[0.042]; Dracula[0.042]; Dracula[0.042]; named[0.042]; executed[0.042]; decade[0.042]; communist[0.042]; Peasants[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = livestock\u001b[39m ==> ENTITY: \u001b[32mLivestock\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.166:0.166[0]; log p(e|m)= -0.010:-0.010[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncattle[0.046]; cattle[0.046]; sheep[0.045]; sheep[0.045]; weather[0.045]; cows[0.045]; areas[0.044]; regions[0.043]; forested[0.042]; cold[0.041]; western[0.041]; western[0.041]; bear[0.040]; bear[0.040]; taste[0.040]; appetite[0.040]; pounds[0.040]; hunting[0.040]; state[0.040]; reserves[0.040]; weeks[0.039]; weeks[0.039]; past[0.039]; past[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Apuseni Mountains\u001b[39m ==> ENTITY: \u001b[32mApuseni Mountains\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.097:0.097[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwestern[0.050]; regions[0.047]; forested[0.046]; str[0.044]; Romania[0.043]; Romania[0.043]; State[0.043]; rangers[0.042]; generations[0.041]; living[0.041]; past[0.041]; seven[0.041]; onset[0.041]; Nicolae[0.041]; communist[0.040]; weather[0.040]; said[0.040]; sheep[0.040]; gj[0.040]; meat[0.039]; meat[0.039]; meat[0.039]; dictator[0.039]; cold[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nicolae Ceausescu\u001b[39m ==> ENTITY: \u001b[32mNicolae Ceaușescu\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRomania[0.047]; Romania[0.047]; Romania[0.047]; communist[0.045]; dictator[0.044]; State[0.043]; Nicolae[0.042]; decade[0.042]; toppled[0.042]; executed[0.041]; ago[0.041]; onset[0.040]; head[0.040]; said[0.040]; said[0.040]; said[0.040]; weeks[0.040]; weeks[0.040]; living[0.040]; past[0.040]; weather[0.040]; sheep[0.040]; gj[0.040]; Transylvania[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = vampire ==> ENTITY: \u001b[32mVampire\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.161:0.161[0]; log p(e|m)= -0.137:-0.137[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDracula[0.047]; Dracula[0.047]; Dracula[0.047]; Dracula[0.047]; devoured[0.043]; apparently[0.042]; nicknamed[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; bears[0.041]; bears[0.041]; bear[0.041]; bear[0.041]; legend[0.041]; inhabited[0.040]; creating[0.040]; kills[0.040]; hungry[0.040]; giant[0.040]; black[0.040]; named[0.040]; Transylvania[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = cattle ==> ENTITY: \u001b[32mCattle\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.126:0.126[0]; log p(e|m)= -0.013:-0.013[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlivestock[0.047]; cattle[0.047]; sheep[0.046]; bear[0.043]; bear[0.043]; giant[0.043]; kilograms[0.042]; black[0.042]; seven[0.042]; hunting[0.042]; western[0.041]; western[0.041]; male[0.041]; legend[0.041]; devoured[0.040]; deer[0.040]; appetite[0.040]; named[0.040]; nicknamed[0.040]; weighs[0.040]; pounds[0.039]; hungry[0.039]; lodged[0.039]; creating[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Romania ==> ENTITY: \u001b[32mRomania\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.212:-0.212[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRomania[0.051]; western[0.044]; western[0.044]; weeks[0.043]; Nicolae[0.043]; seven[0.042]; black[0.042]; regions[0.042]; state[0.041]; Transylvania[0.041]; bear[0.041]; bear[0.041]; ready[0.041]; moved[0.041]; said[0.041]; said[0.041]; said[0.041]; peasants[0.040]; official[0.040]; past[0.040]; Peasants[0.040]; Thursday[0.040]; named[0.040]; nicknamed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = communist ==> ENTITY: \u001b[32mCommunism\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.101:0.101[0]; log p(e|m)= -0.083:-0.083[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRomania[0.047]; Romania[0.047]; Romania[0.047]; Ceausescu[0.044]; western[0.043]; western[0.043]; decade[0.043]; state[0.042]; regions[0.041]; moved[0.041]; past[0.041]; Nicolae[0.041]; Nicolae[0.041]; partial[0.041]; dictator[0.041]; seven[0.040]; toppled[0.040]; said[0.040]; said[0.040]; living[0.040]; weeks[0.040]; ago[0.039]; executed[0.039]; State[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = kilograms\u001b[39m ==> ENTITY: \u001b[32mKilogram\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.001:-0.001[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npounds[0.060]; weighs[0.053]; cattle[0.049]; cattle[0.049]; livestock[0.048]; bear[0.047]; bear[0.047]; seven[0.047]; seven[0.047]; weeks[0.046]; weeks[0.046]; quicker[0.046]; giant[0.046]; onset[0.046]; head[0.046]; communist[0.046]; count[0.046]; decade[0.046]; deer[0.046]; complaints[0.046]; weather[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = rangers ==> ENTITY: \u001b[32mPark ranger\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.612:-0.612[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhunting[0.049]; grow[0.047]; hunter[0.047]; inhabited[0.047]; protected[0.047]; areas[0.046]; forested[0.046]; leaving[0.045]; decade[0.045]; weeks[0.045]; quicker[0.045]; living[0.045]; seven[0.045]; Dracula[0.045]; ago[0.045]; bears[0.045]; bears[0.045]; bears[0.045]; reserve[0.045]; generations[0.044]; State[0.044]; weather[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = peasants ==> ENTITY: \u001b[32mPeasant\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.033:-0.033[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPeasants[0.045]; inhabited[0.044]; areas[0.043]; western[0.042]; western[0.042]; weeks[0.042]; cattle[0.042]; cattle[0.042]; bear[0.042]; bear[0.042]; forested[0.041]; black[0.041]; Romania[0.041]; Romania[0.041]; livestock[0.041]; state[0.041]; Thursday[0.041]; said[0.041]; said[0.041]; said[0.041]; creating[0.041]; official[0.041]; Nicolae[0.040]; taste[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = hibernate ==> ENTITY: \u001b[32mHibernation\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.138:0.138[0]; log p(e|m)= -0.351:-0.351[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncold[0.045]; cubs[0.044]; bears[0.043]; bears[0.043]; bears[0.043]; bear[0.043]; hunting[0.042]; hunting[0.042]; weather[0.042]; forested[0.042]; appetite[0.041]; cattle[0.041]; State[0.041]; partial[0.041]; hungry[0.041]; living[0.041]; quicker[0.041]; seven[0.041]; seven[0.041]; livestock[0.041]; weeks[0.040]; weeks[0.040]; grow[0.040]; Mountains[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = sheep\u001b[39m ==> ENTITY: \u001b[32mSheep\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.128:0.128[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlivestock[0.047]; cattle[0.046]; cattle[0.046]; bear[0.043]; bear[0.043]; hunting[0.042]; Romania[0.042]; Romania[0.042]; taste[0.042]; pounds[0.041]; forested[0.041]; deer[0.041]; black[0.041]; regions[0.041]; hibernate[0.041]; western[0.040]; western[0.040]; cold[0.040]; areas[0.040]; said[0.040]; said[0.040]; said[0.040]; weather[0.040]; legend[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = omnivorous\u001b[39m ==> ENTITY: \u001b[32mOmnivore\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.167:0.167[0]; log p(e|m)= -0.002:-0.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhibernate[0.046]; forested[0.046]; bears[0.045]; bears[0.045]; bears[0.045]; protected[0.044]; meat[0.044]; meat[0.044]; meat[0.044]; hungry[0.044]; bear[0.044]; living[0.044]; appetite[0.044]; areas[0.043]; pounds[0.043]; State[0.042]; said[0.042]; said[0.042]; said[0.042]; cold[0.042]; kilograms[0.042]; apparently[0.042]; inhabited[0.042]; \t\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19980903_1073.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.046]; president[0.045]; president[0.045]; Vienna[0.044]; visit[0.044]; Federal[0.043]; Austrian[0.042]; Austrian[0.042]; Klestil[0.042]; Klestil[0.042]; Secretary[0.042]; Chancellor[0.041]; Moscow[0.041]; paid[0.041]; summit[0.040]; government[0.040]; statement[0.040]; Thursday[0.040]; Albright[0.040]; Albright[0.040]; Albright[0.040]; Albright[0.040]; day[0.040]; Thomas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVienna[0.045]; Viktor[0.043]; State[0.043]; art[0.042]; art[0.042]; art[0.042]; art[0.042]; art[0.042]; art[0.042]; Jewish[0.042]; Austria[0.042]; Austria[0.042]; Nazis[0.042]; Thursday[0.041]; officials[0.041]; officials[0.041]; Austrian[0.041]; Austrian[0.041]; Europe[0.041]; works[0.041]; World[0.041]; launched[0.041]; visit[0.041]; Wolfgang[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Adolf Hitler\u001b[39m ==> ENTITY: \u001b[32mAdolf Hitler\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.113:0.113[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJewish[0.049]; Holocaust[0.049]; Germany[0.044]; Europe[0.044]; war[0.042]; Austria[0.042]; Austria[0.042]; Wolfgang[0.042]; victims[0.042]; American[0.042]; War[0.041]; nations[0.041]; nation[0.041]; worldwide[0.041]; museums[0.041]; World[0.040]; World[0.040]; States[0.040]; States[0.040]; current[0.040]; June[0.040]; drive[0.040]; following[0.039]; Association[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.094:0.094[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJewish[0.047]; States[0.046]; States[0.046]; American[0.046]; nations[0.046]; World[0.045]; Austria[0.045]; Austria[0.045]; Europe[0.044]; day[0.044]; said[0.043]; Moscow[0.042]; Thursday[0.042]; United[0.042]; United[0.042]; origin[0.042]; June[0.042]; State[0.042]; Thomas[0.042]; current[0.042]; statement[0.041]; Association[0.041]; paid[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nazis\u001b[39m ==> ENTITY: \u001b[32mNazism\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.010:-0.010[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHolocaust[0.052]; Wolfgang[0.047]; Austria[0.047]; Austria[0.047]; visit[0.045]; victims[0.045]; statement[0.043]; government[0.043]; Austrian[0.043]; Austrian[0.043]; president[0.043]; president[0.043]; drive[0.042]; said[0.042]; President[0.042]; Thursday[0.042]; day[0.042]; States[0.042]; States[0.042]; works[0.042]; paid[0.042]; effort[0.042]; Thomas[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.048]; Austria[0.048]; Europe[0.045]; Hitler[0.045]; War[0.044]; war[0.043]; nation[0.042]; Holocaust[0.041]; Jewish[0.041]; World[0.041]; World[0.041]; States[0.041]; States[0.041]; museums[0.041]; American[0.041]; nations[0.040]; following[0.040]; worldwide[0.040]; June[0.040]; Adolf[0.040]; United[0.040]; United[0.040]; dollars[0.040]; current[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Viktor Klima\u001b[39m ==> ENTITY: \u001b[32mViktor Klima\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.046]; president[0.046]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Wolfgang[0.043]; Chancellor[0.042]; Klestil[0.042]; Klestil[0.042]; State[0.042]; government[0.042]; Secretary[0.042]; meetings[0.041]; June[0.041]; officials[0.041]; officials[0.041]; visit[0.041]; Austrian[0.040]; paid[0.040]; Vienna[0.040]; said[0.040]; Minister[0.040]; Europe[0.040]; Directors[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = stolen art\u001b[39m ==> ENTITY: \u001b[32mArt theft\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.113:0.113[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstolen[0.046]; works[0.044]; Museum[0.044]; art[0.043]; art[0.043]; art[0.043]; Nazis[0.043]; Austrian[0.043]; Austrian[0.043]; Austria[0.042]; Thomas[0.042]; Holocaust[0.042]; State[0.042]; paid[0.042]; origin[0.041]; Chancellor[0.040]; Vienna[0.040]; effort[0.040]; Thursday[0.040]; identify[0.040]; identify[0.040]; owned[0.040]; Klestil[0.039]; Klestil[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Madeleine Albright\u001b[39m ==> ENTITY: \u001b[32mMadeleine Albright\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSecretary[0.047]; President[0.046]; visit[0.043]; summit[0.043]; Foreign[0.043]; Albright[0.042]; Albright[0.042]; Albright[0.042]; States[0.042]; statement[0.042]; State[0.042]; president[0.042]; president[0.042]; told[0.041]; told[0.041]; commission[0.040]; commission[0.040]; Thursday[0.040]; Klima[0.040]; said[0.040]; Chancellor[0.040]; officials[0.040]; officials[0.040]; issue[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wolfgang Schuessel\u001b[39m ==> ENTITY: \u001b[32mWolfgang Schüssel\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.046]; Austria[0.046]; President[0.046]; Chancellor[0.044]; government[0.042]; Austrian[0.042]; meetings[0.042]; Klestil[0.042]; Klestil[0.042]; said[0.042]; Federal[0.042]; State[0.042]; recommended[0.041]; Europe[0.041]; Adolf[0.041]; visit[0.041]; current[0.041]; Minister[0.040]; Klima[0.040]; statement[0.040]; June[0.040]; Vienna[0.040]; officials[0.040]; officials[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Thomas Klestil\u001b[39m ==> ENTITY: \u001b[32mThomas Klestil\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKlestil[0.048]; Austria[0.046]; Austria[0.046]; Jewish[0.046]; visit[0.045]; President[0.045]; Austrian[0.045]; Austrian[0.045]; Federal[0.044]; president[0.043]; president[0.043]; said[0.043]; State[0.043]; Secretary[0.043]; Vienna[0.042]; Wolfgang[0.042]; Moscow[0.042]; Klima[0.042]; statement[0.042]; Chancellor[0.041]; government[0.041]; Holocaust[0.041]; Holocaust[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Secretary of State ==> ENTITY: \u001b[32mSecretary of State\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.161:0.161[0]; log p(e|m)= -1.174:-1.174[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.044]; Foreign[0.043]; president[0.043]; president[0.043]; government[0.043]; Minister[0.042]; Federal[0.042]; Albright[0.042]; Albright[0.042]; Albright[0.042]; Albright[0.042]; paid[0.042]; visit[0.042]; Chancellor[0.042]; States[0.041]; said[0.041]; commission[0.041]; commission[0.041]; Thursday[0.041]; Thomas[0.040]; meetings[0.040]; day[0.040]; officials[0.040]; officials[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = World Jewish Congress\u001b[39m ==> ENTITY: \u001b[32mWorld Jewish Congress\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.129:0.129[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHolocaust[0.047]; President[0.044]; statement[0.043]; told[0.042]; World[0.042]; Germany[0.042]; dollars[0.042]; States[0.042]; States[0.042]; States[0.042]; Vienna[0.041]; Austria[0.041]; Austria[0.041]; Hitler[0.041]; American[0.041]; war[0.041]; commission[0.041]; issue[0.041]; Adolf[0.041]; Europe[0.041]; Foreign[0.041]; Klestil[0.041]; Klestil[0.041]; Wolfgang[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Holocaust\u001b[39m ==> ENTITY: \u001b[32mThe Holocaust\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.103:0.103[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNazis[0.046]; Federal[0.045]; said[0.044]; victims[0.043]; Austria[0.042]; Austria[0.042]; Thursday[0.042]; tells[0.042]; works[0.042]; Museum[0.042]; officials[0.042]; officials[0.042]; day[0.041]; Thomas[0.041]; stolen[0.041]; stolen[0.041]; visit[0.041]; commission[0.040]; commission[0.040]; told[0.040]; told[0.040]; Viktor[0.040]; effort[0.040]; statement[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Association of Art Museum Directors\u001b[39m ==> ENTITY: \u001b[32mAssociation of Art Museum Directors\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmuseums[0.046]; Museum[0.046]; worldwide[0.045]; American[0.045]; States[0.045]; States[0.045]; States[0.045]; Europe[0.044]; art[0.044]; art[0.044]; art[0.044]; art[0.044]; art[0.044]; Congress[0.043]; nation[0.042]; Germany[0.042]; nations[0.042]; Austria[0.042]; Austria[0.042]; collections[0.042]; launched[0.041]; current[0.041]; Adolf[0.041]; \t\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19981020_1367.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = John Glenn ==> ENTITY: \u001b[32mJohn Glenn\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.041:-0.041[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGlenn[0.045]; June[0.043]; April[0.043]; April[0.043]; orbit[0.042]; orbit[0.042]; July[0.042]; July[0.042]; spacecraft[0.042]; spacecraft[0.042]; shuttle[0.042]; moon[0.042]; Challenger[0.041]; space[0.041]; space[0.041]; space[0.041]; flight[0.041]; land[0.041]; Apollo[0.041]; Apollo[0.041]; Apollo[0.041]; Neil[0.041]; seven[0.040]; Alan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mir ==> ENTITY: \u001b[32mMir\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.253:0.253[0]; log p(e|m)= -0.108:-0.108[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nshuttle[0.044]; shuttle[0.044]; Mir[0.043]; Mir[0.043]; flight[0.043]; space[0.042]; space[0.042]; space[0.042]; space[0.042]; space[0.042]; Russian[0.042]; Soviet[0.042]; spacecraft[0.041]; Soyuz[0.041]; mission[0.041]; mission[0.041]; Atlantis[0.041]; Space[0.041]; station[0.041]; orbit[0.041]; orbit[0.041]; orbit[0.041]; Discovery[0.041]; Apollo[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Soyuz spacecraft\u001b[39m ==> ENTITY: \u001b[32mSoyuz (spacecraft)\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.242:0.242[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nspace[0.044]; space[0.044]; space[0.044]; spacecraft[0.044]; launch[0.044]; shuttle[0.044]; shuttle[0.044]; rocket[0.043]; orbit[0.043]; orbit[0.043]; orbit[0.043]; Space[0.042]; flight[0.042]; launched[0.041]; launched[0.041]; mission[0.041]; Atlantis[0.041]; Apollo[0.040]; Apollo[0.040]; Apollo[0.040]; station[0.040]; pad[0.039]; moon[0.039]; seven[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = launch pad ==> ENTITY: \u001b[32mLaunch pad\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.045:-0.045[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nshuttle[0.048]; space[0.044]; space[0.044]; space[0.044]; Space[0.043]; spacecraft[0.043]; spacecraft[0.043]; flight[0.042]; Challenger[0.042]; orbit[0.041]; orbit[0.041]; orbit[0.041]; Apollo[0.041]; Apollo[0.041]; Apollo[0.041]; repair[0.041]; March[0.040]; Soyuz[0.040]; II[0.040]; moon[0.040]; Columbia[0.039]; April[0.039]; April[0.039]; June[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Roger Chaffee\u001b[39m ==> ENTITY: \u001b[32mRoger B. Chaffee\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.229:0.229[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGlenn[0.048]; Glenn[0.048]; flight[0.043]; Space[0.043]; John[0.043]; Apollo[0.042]; Apollo[0.042]; Apollo[0.042]; Shepard[0.042]; Shepard[0.042]; Columbia[0.042]; Alan[0.041]; shuttle[0.041]; orbit[0.041]; orbit[0.041]; orbit[0.041]; mission[0.041]; spacecraft[0.040]; spacecraft[0.040]; Grissom[0.040]; space[0.040]; space[0.040]; space[0.040]; Aldrin[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hubble Space Telescope\u001b[39m ==> ENTITY: \u001b[32mHubble Space Telescope\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.234:0.234[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHubble[0.044]; spacecraft[0.044]; spacecraft[0.044]; shuttle[0.042]; shuttle[0.042]; orbit[0.042]; orbit[0.042]; orbit[0.042]; launch[0.042]; Apollo[0.042]; Apollo[0.042]; Apollo[0.042]; moon[0.042]; space[0.041]; space[0.041]; space[0.041]; space[0.041]; mission[0.041]; mission[0.041]; launched[0.041]; launched[0.041]; rocket[0.040]; Discovery[0.040]; Earth[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Discovery ==> ENTITY: \u001b[32mSpace Shuttle Discovery\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.184:0.184[0]; log p(e|m)= -0.870:-0.870[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nshuttle[0.048]; Atlantis[0.045]; Space[0.044]; mission[0.043]; mission[0.043]; space[0.043]; space[0.043]; space[0.043]; released[0.042]; record[0.041]; record[0.041]; rocket[0.041]; Hubble[0.041]; Hubble[0.041]; orbit[0.041]; orbit[0.041]; Glenn[0.040]; American[0.040]; American[0.040]; launched[0.040]; launched[0.040]; Shannon[0.040]; docks[0.040]; John[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = space shuttle\u001b[39m ==> ENTITY: \u001b[32mSpace Shuttle\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.240:0.240[0]; log p(e|m)= -0.006:-0.006[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nshuttle[0.047]; Space[0.044]; Atlantis[0.044]; spacecraft[0.043]; spacecraft[0.043]; flight[0.043]; rocket[0.043]; space[0.042]; space[0.042]; Challenger[0.042]; mission[0.042]; mission[0.042]; launch[0.041]; Columbia[0.041]; Apollo[0.040]; Apollo[0.040]; Apollo[0.040]; orbit[0.040]; orbit[0.040]; orbit[0.040]; Hubble[0.040]; Hubble[0.040]; April[0.040]; April[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Neil Armstrong\u001b[39m ==> ENTITY: \u001b[32mNeil Armstrong\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nApollo[0.044]; Apollo[0.044]; Apollo[0.044]; Glenn[0.043]; Glenn[0.043]; John[0.043]; moon[0.043]; space[0.043]; space[0.043]; space[0.043]; flight[0.043]; Aldrin[0.042]; Space[0.042]; launch[0.041]; shuttle[0.041]; spacecraft[0.041]; spacecraft[0.041]; Alan[0.040]; rocket[0.040]; Roger[0.040]; mission[0.040]; Columbia[0.039]; Shepard[0.039]; Shepard[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Apollo 1\u001b[39m ==> ENTITY: \u001b[32mApollo 1\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.213:0.213[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGlenn[0.047]; Glenn[0.047]; shuttle[0.043]; Space[0.043]; space[0.042]; space[0.042]; space[0.042]; flight[0.042]; Apollo[0.042]; Apollo[0.042]; John[0.041]; Challenger[0.041]; spacecraft[0.041]; spacecraft[0.041]; Shepard[0.041]; Shepard[0.041]; Columbia[0.041]; mission[0.040]; Roger[0.040]; orbit[0.040]; orbit[0.040]; orbit[0.040]; Grissom[0.040]; launch[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlantis ==> ENTITY: \u001b[32mSpace Shuttle Atlantis\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.290:0.290[0]; log p(e|m)= -1.336:-1.336[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nshuttle[0.046]; shuttle[0.046]; released[0.043]; Space[0.043]; mission[0.043]; mission[0.043]; Discovery[0.043]; flight[0.042]; spacecraft[0.042]; Apollo[0.042]; Soyuz[0.041]; space[0.041]; space[0.041]; space[0.041]; space[0.041]; space[0.041]; Challenger[0.041]; Hubble[0.040]; Hubble[0.040]; world[0.040]; orbit[0.040]; orbit[0.040]; orbit[0.040]; Columbia[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36morbit\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mOrbital spaceflight\u001b[39m <---> \u001b[32mOrbit\u001b[39m\t\nSCORES: global= 0.257:0.240[\u001b[31m0.017\u001b[39m]; local()= 0.230:0.162[\u001b[31m0.068\u001b[39m]; log p(e|m)= -1.374:-0.256[\u001b[32m1.118\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nspacecraft[0.045]; spacecraft[0.045]; orbit[0.045]; space[0.044]; space[0.044]; space[0.044]; launch[0.043]; Glenn[0.043]; Glenn[0.043]; flight[0.043]; Space[0.042]; shuttle[0.042]; moon[0.042]; Soyuz[0.041]; Apollo[0.041]; Apollo[0.041]; Apollo[0.041]; Columbia[0.039]; Hubble[0.039]; Shepard[0.039]; Shepard[0.039]; Challenger[0.039]; John[0.038]; seven[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Norman Thagard\u001b[39m ==> ENTITY: \u001b[32mNorman Thagard\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.218:0.218[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGlenn[0.052]; shuttle[0.049]; shuttle[0.049]; flight[0.049]; Space[0.049]; John[0.049]; spacecraft[0.048]; spacecraft[0.048]; Apollo[0.048]; Apollo[0.048]; Apollo[0.048]; Shannon[0.047]; mission[0.047]; mission[0.047]; Discovery[0.047]; Challenger[0.046]; Atlantis[0.046]; American[0.046]; American[0.046]; American[0.046]; American[0.046]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mmoon\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mApollo 14\u001b[39m <---> \u001b[32mMoon\u001b[39m\t\nSCORES: global= 0.259:0.244[\u001b[31m0.016\u001b[39m]; local()= 0.190:0.151[\u001b[31m0.039\u001b[39m]; log p(e|m)= -1.609:-0.361[\u001b[32m1.248\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nApollo[0.044]; Apollo[0.044]; Apollo[0.044]; spacecraft[0.044]; spacecraft[0.044]; Space[0.044]; shuttle[0.043]; space[0.043]; space[0.043]; space[0.043]; Glenn[0.041]; Glenn[0.041]; Aldrin[0.041]; launch[0.041]; flight[0.041]; rocket[0.041]; orbit[0.040]; orbit[0.040]; orbit[0.040]; mission[0.040]; Atlantis[0.040]; April[0.040]; April[0.040]; days[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = spacecraft ==> ENTITY: \u001b[32mSpacecraft\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.225:0.225[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nspace[0.045]; space[0.045]; space[0.045]; Space[0.044]; shuttle[0.044]; launch[0.043]; spacecraft[0.043]; orbit[0.042]; orbit[0.042]; orbit[0.042]; Apollo[0.042]; Apollo[0.042]; Apollo[0.042]; flight[0.042]; mission[0.041]; Hubble[0.040]; Hubble[0.040]; Challenger[0.040]; Soyuz[0.040]; Columbia[0.040]; moon[0.040]; seven[0.039]; pad[0.039]; launched[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sally Ride\u001b[39m ==> ENTITY: \u001b[32mSally Ride\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.215:0.215[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGlenn[0.046]; Glenn[0.046]; Space[0.046]; shuttle[0.044]; shuttle[0.044]; Roger[0.044]; Shannon[0.043]; space[0.043]; space[0.043]; space[0.043]; space[0.043]; Challenger[0.043]; John[0.043]; mission[0.043]; mission[0.043]; Apollo[0.043]; Apollo[0.043]; Apollo[0.043]; spacecraft[0.043]; spacecraft[0.043]; orbit[0.043]; orbit[0.043]; orbit[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shannon Lucid\u001b[39m ==> ENTITY: \u001b[32mShannon Lucid\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.226:0.226[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nflight[0.045]; Glenn[0.044]; mission[0.044]; mission[0.044]; shuttle[0.043]; shuttle[0.043]; Space[0.043]; space[0.042]; space[0.042]; space[0.042]; space[0.042]; space[0.042]; orbit[0.041]; orbit[0.041]; orbit[0.041]; Atlantis[0.041]; Discovery[0.041]; seven[0.040]; days[0.040]; Hubble[0.040]; Hubble[0.040]; Columbia[0.040]; John[0.040]; station[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Columbia ==> ENTITY: \u001b[32mSpace Shuttle Columbia\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.218:0.218[0]; log p(e|m)= -2.419:-2.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nshuttle[0.045]; shuttle[0.045]; Space[0.044]; flight[0.043]; spacecraft[0.043]; spacecraft[0.043]; Challenger[0.043]; space[0.042]; space[0.042]; space[0.042]; mission[0.042]; launch[0.041]; rocket[0.041]; released[0.041]; orbit[0.040]; orbit[0.040]; orbit[0.040]; Neil[0.040]; American[0.040]; American[0.040]; American[0.040]; American[0.040]; American[0.040]; American[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Apollo 11\u001b[39m ==> ENTITY: \u001b[32mApollo 11\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.216:0.216[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nApollo[0.044]; Apollo[0.044]; spacecraft[0.043]; spacecraft[0.043]; Glenn[0.043]; Glenn[0.043]; Space[0.043]; space[0.043]; space[0.043]; space[0.043]; moon[0.042]; mission[0.041]; flight[0.041]; shuttle[0.041]; Aldrin[0.041]; launch[0.041]; Neil[0.040]; rocket[0.040]; orbit[0.040]; orbit[0.040]; orbit[0.040]; Challenger[0.040]; Shepard[0.039]; Shepard[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alan Shepard\u001b[39m ==> ENTITY: \u001b[32mAlan Shepard\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGlenn[0.047]; Glenn[0.047]; space[0.043]; space[0.043]; space[0.043]; John[0.043]; Shepard[0.042]; shuttle[0.042]; moon[0.042]; Columbia[0.041]; Roger[0.041]; Apollo[0.041]; Apollo[0.041]; Apollo[0.041]; American[0.040]; American[0.040]; American[0.040]; American[0.040]; Ride[0.040]; Challenger[0.040]; spacecraft[0.040]; spacecraft[0.040]; Neil[0.040]; Aldrin[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gus Grissom\u001b[39m ==> ENTITY: \u001b[32mGus Grissom\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGlenn[0.048]; Glenn[0.048]; flight[0.043]; Space[0.043]; John[0.042]; shuttle[0.042]; Shepard[0.042]; Shepard[0.042]; Columbia[0.041]; Roger[0.041]; Apollo[0.041]; Apollo[0.041]; Apollo[0.041]; spacecraft[0.041]; spacecraft[0.041]; Challenger[0.041]; orbit[0.041]; orbit[0.041]; orbit[0.041]; American[0.040]; American[0.040]; American[0.040]; American[0.040]; Alan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = rocket ==> ENTITY: \u001b[32mRocket\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.158:-0.158[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nshuttle[0.049]; shuttle[0.049]; Space[0.049]; launched[0.049]; launched[0.049]; spacecraft[0.049]; Discovery[0.048]; space[0.048]; space[0.048]; space[0.048]; space[0.048]; space[0.048]; orbit[0.047]; orbit[0.047]; orbit[0.047]; Apollo[0.047]; Apollo[0.047]; Soyuz[0.047]; flight[0.047]; Atlantis[0.045]; Challenger[0.045]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mChallenger\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSpace Shuttle Challenger disaster\u001b[39m <---> \u001b[32mSpace Shuttle Challenger\u001b[39m\t\nSCORES: global= 0.273:0.272[\u001b[31m0.001\u001b[39m]; local()= 0.238:0.232[\u001b[31m0.006\u001b[39m]; log p(e|m)= -1.565:-0.624[\u001b[32m0.942\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nshuttle[0.046]; shuttle[0.046]; Space[0.044]; space[0.044]; space[0.044]; space[0.044]; space[0.044]; spacecraft[0.044]; spacecraft[0.044]; launch[0.043]; orbit[0.043]; orbit[0.043]; orbit[0.043]; mission[0.043]; mission[0.043]; flight[0.043]; Atlantis[0.043]; Columbia[0.042]; Apollo[0.042]; Apollo[0.042]; Apollo[0.042]; rocket[0.042]; Soyuz[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Buzz Aldrin\u001b[39m ==> ENTITY: \u001b[32mBuzz Aldrin\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.209:0.209[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGlenn[0.045]; Glenn[0.045]; launch[0.043]; Space[0.043]; Apollo[0.043]; Apollo[0.043]; Apollo[0.043]; John[0.042]; flight[0.042]; space[0.042]; space[0.042]; space[0.042]; shuttle[0.041]; Neil[0.041]; Alan[0.041]; spacecraft[0.041]; spacecraft[0.041]; Shepard[0.041]; Shepard[0.041]; mission[0.041]; moon[0.040]; Columbia[0.040]; released[0.040]; Roger[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Story Musgrave\u001b[39m ==> ENTITY: \u001b[32mStory Musgrave\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGlenn[0.045]; shuttle[0.044]; mission[0.044]; mission[0.044]; John[0.043]; Discovery[0.043]; Space[0.043]; Thagard[0.042]; rocket[0.042]; Shannon[0.042]; space[0.041]; space[0.041]; space[0.041]; orbit[0.041]; orbit[0.041]; Hubble[0.041]; Hubble[0.041]; American[0.041]; American[0.041]; Challenger[0.040]; April[0.040]; Atlantis[0.039]; March[0.039]; June[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Edward White ==> ENTITY: \u001b[32mEdward Higgins White\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.210:0.210[0]; log p(e|m)= -0.086:-0.086[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGlenn[0.048]; Glenn[0.048]; flight[0.045]; Space[0.045]; shuttle[0.045]; John[0.044]; Roger[0.043]; Apollo[0.043]; Apollo[0.043]; Apollo[0.043]; American[0.043]; American[0.043]; American[0.043]; American[0.043]; orbit[0.043]; orbit[0.043]; orbit[0.043]; Alan[0.042]; space[0.042]; space[0.042]; space[0.042]; Challenger[0.042]; June[0.042]; \t\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19981109_1089.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Africa ==> ENTITY: \u001b[32mSouth Africa\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.263:-0.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.049]; batsman[0.047]; South[0.046]; Cricket[0.046]; Cricket[0.046]; players[0.045]; players[0.045]; players[0.045]; Africa[0.044]; tour[0.043]; West[0.043]; West[0.043]; West[0.043]; West[0.043]; Team[0.042]; captain[0.042]; captain[0.042]; Wednesday[0.042]; Monday[0.042]; Adams[0.041]; Lara[0.041]; Lara[0.041]; involving[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = West Indies Cricket Board\u001b[39m ==> ENTITY: \u001b[32mWest Indies Cricket Board\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCricket[0.044]; cricket[0.044]; cricket[0.044]; reinstated[0.043]; offered[0.043]; WICB[0.043]; Wednesday[0.042]; vice[0.042]; Monday[0.042]; Monday[0.042]; players[0.041]; players[0.041]; manager[0.041]; Board[0.041]; member[0.041]; Indies[0.041]; Indies[0.041]; Indies[0.041]; tour[0.041]; tour[0.041]; president[0.041]; negotiated[0.041]; Team[0.041]; meet[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Johannesburg ==> ENTITY: \u001b[32mJohannesburg\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.166:0.166[0]; log p(e|m)= -0.095:-0.095[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.044]; South[0.044]; South[0.044]; cricket[0.044]; Cricket[0.043]; tour[0.043]; tour[0.043]; African[0.042]; African[0.042]; African[0.042]; Sunday[0.042]; West[0.042]; West[0.042]; sports[0.041]; Monday[0.041]; players[0.041]; Friday[0.041]; weekend[0.041]; sponsorship[0.040]; sponsorship[0.040]; runs[0.040]; deal[0.040]; began[0.039]; sitting[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = wicketkeeper\u001b[39m ==> ENTITY: \u001b[32mWicket-keeper\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.050]; Cricket[0.046]; Clive[0.043]; South[0.043]; South[0.043]; South[0.043]; David[0.043]; Richardson[0.042]; African[0.041]; African[0.041]; African[0.041]; Adams[0.041]; runs[0.041]; sitting[0.041]; players[0.041]; players[0.041]; players[0.041]; Bacher[0.040]; WICB[0.040]; sports[0.040]; Ali[0.040]; Johannesburg[0.040]; Monday[0.040]; Monday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ali Bacher\u001b[39m ==> ENTITY: \u001b[32mAli Bacher\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.045]; Cricket[0.044]; Cricket[0.044]; African[0.044]; African[0.044]; South[0.043]; South[0.043]; South[0.043]; runs[0.042]; sports[0.042]; batsman[0.041]; David[0.041]; captain[0.041]; captain[0.041]; reinstated[0.041]; Africa[0.041]; Johannesburg[0.041]; players[0.041]; players[0.041]; players[0.041]; vice[0.040]; tour[0.040]; tour[0.040]; Adams[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Carl Hooper\u001b[39m ==> ENTITY: \u001b[32mCarl Hooper\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.048]; cricket[0.046]; cricket[0.046]; Cricket[0.044]; Cricket[0.044]; Lara[0.043]; Lara[0.043]; captain[0.043]; captain[0.043]; Hooper[0.042]; South[0.041]; South[0.041]; South[0.041]; Adams[0.041]; Indies[0.040]; Indies[0.040]; Indies[0.040]; Indies[0.040]; Indies[0.040]; African[0.040]; players[0.039]; players[0.039]; players[0.039]; Jimmy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = South African cricket\u001b[39m ==> ENTITY: \u001b[32mSouth Africa national cricket team\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.050]; Cricket[0.049]; Cricket[0.049]; wicketkeeper[0.048]; Johannesburg[0.046]; South[0.046]; South[0.046]; Sunday[0.045]; captain[0.045]; captain[0.045]; Wednesday[0.045]; Monday[0.045]; Monday[0.045]; African[0.045]; Adams[0.044]; Clive[0.044]; Lara[0.044]; Lara[0.044]; tour[0.044]; tour[0.044]; Friday[0.044]; runs[0.044]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWest Indies\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mWest Indies cricket team\u001b[39m <---> \u001b[32mCaribbean\u001b[39m\t\nSCORES: global= 0.258:0.231[\u001b[31m0.027\u001b[39m]; local()= 0.153:0.079[\u001b[31m0.073\u001b[39m]; log p(e|m)= -1.103:-0.637[\u001b[32m0.466\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.049]; West[0.048]; West[0.048]; batsman[0.047]; Cricket[0.045]; Lara[0.044]; Lara[0.044]; Indies[0.044]; Indies[0.044]; South[0.044]; South[0.044]; tour[0.044]; Wednesday[0.043]; captain[0.042]; captain[0.042]; Monday[0.042]; Team[0.041]; Hooper[0.041]; Hooper[0.041]; Adams[0.041]; players[0.040]; players[0.040]; players[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = batsman\u001b[39m ==> ENTITY: \u001b[32mBatting (cricket)\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.055]; cricket[0.055]; Cricket[0.049]; Cricket[0.049]; South[0.044]; South[0.044]; South[0.044]; players[0.043]; players[0.043]; players[0.043]; Lara[0.042]; Lara[0.042]; captain[0.042]; captain[0.042]; African[0.042]; better[0.041]; Wednesday[0.041]; Monday[0.040]; Monday[0.040]; Bacher[0.040]; Adams[0.040]; Jimmy[0.040]; Carl[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = sports marketing\u001b[39m ==> ENTITY: \u001b[32mSports marketing\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsponsorship[0.050]; sponsorship[0.050]; firm[0.044]; players[0.043]; players[0.043]; offered[0.042]; Bacher[0.042]; deal[0.042]; directly[0.041]; pay[0.041]; help[0.041]; manager[0.041]; announcement[0.041]; Cricket[0.040]; head[0.040]; wicketkeeper[0.040]; weekend[0.040]; Pat[0.040]; sw[0.040]; David[0.040]; sidelines[0.040]; WICB[0.039]; tour[0.039]; tour[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Clive Lloyd\u001b[39m ==> ENTITY: \u001b[32mClive Lloyd\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.047]; Cricket[0.046]; wicketkeeper[0.045]; Lara[0.045]; David[0.043]; captain[0.043]; Adams[0.042]; WICB[0.042]; Indies[0.041]; Indies[0.041]; South[0.041]; South[0.041]; South[0.041]; South[0.041]; players[0.040]; players[0.040]; players[0.040]; West[0.040]; West[0.040]; runs[0.040]; Bacher[0.040]; Garner[0.040]; Jimmy[0.040]; Richardson[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jimmy Adams\u001b[39m ==> ENTITY: \u001b[32mJimmy Adams\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.047]; cricket[0.044]; cricket[0.044]; wicketkeeper[0.043]; Cricket[0.043]; Cricket[0.043]; Brian[0.042]; Clive[0.042]; captain[0.042]; captain[0.042]; Lara[0.041]; Lara[0.041]; manager[0.041]; Indies[0.041]; Indies[0.041]; Indies[0.041]; Indies[0.041]; Indies[0.041]; South[0.040]; South[0.040]; South[0.040]; South[0.040]; Carl[0.040]; Wednesday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = cricket ==> ENTITY: \u001b[32mCricket\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.076:-0.076[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCricket[0.050]; Cricket[0.050]; batsman[0.049]; players[0.045]; players[0.045]; players[0.045]; South[0.044]; South[0.044]; captain[0.043]; captain[0.043]; Wednesday[0.043]; tour[0.042]; Indies[0.042]; Indies[0.042]; Indies[0.042]; Indies[0.042]; Brian[0.042]; Monday[0.042]; Adams[0.042]; Lara[0.041]; Lara[0.041]; boards[0.041]; Team[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Joel Garner\u001b[39m ==> ENTITY: \u001b[32mJoel Garner\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.047]; Cricket[0.045]; David[0.044]; Lara[0.044]; wicketkeeper[0.043]; Jimmy[0.042]; Richardson[0.042]; Clive[0.042]; runs[0.042]; players[0.042]; players[0.042]; players[0.042]; Adams[0.041]; Pat[0.041]; Bacher[0.041]; Lloyd[0.041]; manager[0.040]; Hooper[0.040]; WICB[0.040]; sitting[0.040]; better[0.040]; better[0.040]; Indies[0.040]; Indies[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Brian Lara\u001b[39m ==> ENTITY: \u001b[32mBrian Lara\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.049]; cricket[0.049]; cricket[0.049]; Lara[0.048]; Cricket[0.047]; Cricket[0.047]; Adams[0.043]; Indies[0.043]; Indies[0.043]; Indies[0.043]; Indies[0.043]; captain[0.042]; captain[0.042]; players[0.042]; players[0.042]; players[0.042]; came[0.041]; came[0.041]; superstar[0.041]; Bacher[0.041]; refused[0.041]; Carl[0.041]; tour[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = sponsorship\u001b[39m ==> ENTITY: \u001b[32mSponsor (commercial)\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsponsorship[0.053]; marketing[0.049]; sports[0.049]; deal[0.044]; Cricket[0.044]; help[0.043]; Monday[0.043]; Monday[0.043]; Sunday[0.043]; announcement[0.043]; players[0.043]; players[0.043]; players[0.043]; weekend[0.042]; Friday[0.042]; firm[0.042]; offered[0.042]; David[0.042]; president[0.042]; member[0.042]; cricket[0.042]; involved[0.041]; involved[0.041]; \t\n\r [======>.......................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 117/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19990120_0179.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rembrandt\u001b[39m ==> ENTITY: \u001b[32mRembrandt\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npainting[0.047]; art[0.046]; house[0.043]; Sotheby[0.042]; Sotheby[0.042]; Susan[0.042]; memorabilia[0.042]; jewelry[0.042]; collectibles[0.042]; antiques[0.042]; million[0.041]; million[0.041]; York[0.041]; launch[0.041]; image[0.041]; auction[0.041]; auction[0.041]; Tuesday[0.041]; initially[0.040]; expensive[0.040]; angles[0.040]; site[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = collectibles\u001b[39m ==> ENTITY: \u001b[32mCollectable\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.180:0.180[0]; log p(e|m)= -0.010:-0.010[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmemorabilia[0.047]; memorabilia[0.047]; antiques[0.044]; items[0.044]; items[0.044]; items[0.044]; jewelry[0.043]; produced[0.042]; popular[0.042]; art[0.041]; online[0.041]; online[0.041]; Internet[0.041]; Internet[0.041]; Internet[0.041]; computer[0.041]; expensive[0.040]; image[0.040]; feature[0.040]; sports[0.040]; launch[0.040]; sites[0.039]; sites[0.039]; item[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = online auction\u001b[39m ==> ENTITY: \u001b[32mOnline auction\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.275:0.275[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsell[0.044]; auction[0.043]; auction[0.043]; auction[0.043]; auction[0.043]; auction[0.043]; auction[0.043]; merchandise[0.043]; merchandise[0.043]; online[0.042]; online[0.042]; items[0.042]; items[0.042]; items[0.042]; Internet[0.041]; Internet[0.041]; bid[0.041]; collectibles[0.041]; memorabilia[0.040]; memorabilia[0.040]; bids[0.040]; virtual[0.040]; item[0.039]; expensive[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = antiques\u001b[39m ==> ENTITY: \u001b[32mAntique\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.203:0.203[0]; log p(e|m)= -0.001:-0.001[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmemorabilia[0.046]; memorabilia[0.046]; collectibles[0.044]; items[0.043]; items[0.043]; items[0.043]; public[0.043]; jewelry[0.043]; art[0.043]; auction[0.041]; auction[0.041]; auction[0.041]; auction[0.041]; major[0.041]; popular[0.041]; computer[0.040]; expensive[0.040]; garden[0.040]; said[0.040]; said[0.040]; item[0.040]; painting[0.040]; million[0.039]; million[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = baseball ==> ENTITY: \u001b[32mBaseball\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.085:0.085[0]; log p(e|m)= -0.070:-0.070[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsports[0.052]; popular[0.043]; York[0.043]; major[0.043]; memorabilia[0.043]; memorabilia[0.043]; million[0.042]; numbers[0.042]; said[0.041]; said[0.041]; said[0.041]; merchandise[0.041]; merchandise[0.041]; July[0.041]; groups[0.041]; bids[0.041]; posted[0.041]; feature[0.041]; fuzzy[0.040]; improves[0.040]; accepted[0.040]; angles[0.040]; multiple[0.040]; bid[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Beanie Babies\u001b[39m ==> ENTITY: \u001b[32mBeanie Babies\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.117:0.117[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlaunch[0.045]; memorabilia[0.045]; collectibles[0.045]; jewelry[0.043]; buy[0.043]; computer[0.042]; expensive[0.042]; com[0.042]; sports[0.042]; virtual[0.042]; fuzzy[0.041]; items[0.041]; items[0.041]; items[0.041]; said[0.041]; million[0.041]; million[0.041]; Internet[0.041]; Internet[0.041]; initially[0.040]; Announces[0.040]; multiple[0.040]; tickets[0.040]; imaging[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.061:0.061[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsaid[0.043]; said[0.043]; Tuesday[0.043]; million[0.043]; million[0.043]; site[0.042]; public[0.042]; YORK[0.042]; com[0.042]; bid[0.042]; posted[0.042]; online[0.042]; house[0.041]; major[0.041]; Home[0.041]; Susan[0.041]; block[0.041]; AP[0.041]; Solomon[0.041]; launch[0.041]; spend[0.041]; clear[0.041]; initially[0.041]; views[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = charge card\u001b[39m ==> ENTITY: \u001b[32mCharge card\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \naccepted[0.045]; customers[0.044]; merchandise[0.044]; merchandise[0.044]; item[0.044]; advance[0.043]; consumer[0.042]; consumer[0.042]; feature[0.042]; online[0.042]; online[0.042]; online[0.042]; sell[0.041]; having[0.041]; numbers[0.041]; dealers[0.041]; extremely[0.040]; head[0.040]; auction[0.040]; auction[0.040]; auction[0.040]; auction[0.040]; auction[0.040]; fraud[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = sports memorabilia ==> ENTITY: \u001b[32mSports memorabilia\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.159:0.159[0]; log p(e|m)= -0.112:-0.112[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmemorabilia[0.049]; baseball[0.047]; collectibles[0.045]; items[0.044]; items[0.044]; items[0.044]; feature[0.041]; value[0.041]; bid[0.041]; popular[0.041]; computer[0.041]; jewelry[0.041]; produced[0.041]; launch[0.040]; item[0.040]; million[0.040]; million[0.040]; expensive[0.040]; York[0.040]; image[0.040]; online[0.040]; online[0.040]; buy[0.040]; virtual[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sotheby ==> ENTITY: \u001b[32mSotheby's\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.218:0.218[0]; log p(e|m)= -0.014:-0.014[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncollectibles[0.048]; auction[0.046]; auction[0.046]; Sotheby[0.046]; Sotheby[0.046]; memorabilia[0.045]; antiques[0.045]; million[0.045]; million[0.045]; art[0.044]; Rembrandt[0.044]; painting[0.043]; jewelry[0.043]; bid[0.042]; York[0.042]; house[0.042]; expensive[0.041]; items[0.041]; items[0.041]; items[0.041]; Announces[0.041]; venture[0.041]; initially[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Christie ==> ENTITY: \u001b[32mChristie's\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.224:0.224[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSotheby[0.046]; Sotheby[0.046]; auction[0.045]; auction[0.045]; auction[0.045]; auction[0.045]; auction[0.045]; art[0.043]; memorabilia[0.043]; house[0.042]; house[0.042]; sell[0.041]; considering[0.041]; produced[0.041]; York[0.040]; accepted[0.040]; July[0.040]; shoddy[0.040]; bids[0.040]; site[0.038]; said[0.038]; said[0.038]; said[0.038]; venture[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = auction house ==> ENTITY: \u001b[32mAuction\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.208:0.208[0]; log p(e|m)= -0.023:-0.023[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nauction[0.050]; auction[0.050]; memorabilia[0.045]; antiques[0.045]; bid[0.044]; collectibles[0.044]; buy[0.043]; value[0.043]; expensive[0.043]; computer[0.043]; items[0.043]; items[0.043]; items[0.043]; million[0.042]; million[0.042]; Internet[0.042]; Internet[0.042]; virtual[0.042]; art[0.042]; online[0.042]; said[0.042]; said[0.042]; item[0.042]; \t\n\r [=======>......................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 129/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19980824_0827.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.161:0.161[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.053]; Israel[0.053]; Israeli[0.048]; Israeli[0.048]; Netanyahu[0.043]; international[0.043]; personnel[0.042]; security[0.042]; security[0.042]; David[0.040]; terrorism[0.040]; following[0.040]; officials[0.039]; response[0.039]; threats[0.039]; threats[0.039]; said[0.039]; said[0.039]; Monday[0.039]; militants[0.039]; Islamic[0.038]; adviser[0.038]; measures[0.038]; Minister[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mterrorism\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIslamic terrorism\u001b[39m <---> \u001b[32mTerrorism\u001b[39m\t\nSCORES: global= 0.255:0.248[\u001b[31m0.006\u001b[39m]; local()= 0.190:0.147[\u001b[31m0.043\u001b[39m]; log p(e|m)= -1.336:-0.182[\u001b[32m1.154\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.049]; Israel[0.049]; Israel[0.049]; Israeli[0.048]; Israeli[0.048]; Islamic[0.044]; Islamic[0.044]; Islamic[0.044]; international[0.043]; militant[0.043]; Muslim[0.042]; security[0.042]; security[0.042]; Afghanistan[0.042]; Saudi[0.041]; threats[0.041]; threats[0.041]; Osama[0.041]; linked[0.041]; militants[0.041]; militants[0.041]; bombings[0.041]; attack[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36madministration\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPresidency of George W. Bush\u001b[39m <---> \u001b[32mAcademic administration\u001b[39m\t\nSCORES: global= 0.248:0.217[\u001b[31m0.030\u001b[39m]; local()= 0.087:0.090[\u001b[32m0.003\u001b[39m]; log p(e|m)= -2.112:-2.733[\u001b[31m0.621\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUniversity[0.047]; University[0.047]; month[0.043]; security[0.043]; Monday[0.042]; private[0.042]; Connecticut[0.042]; Connecticut[0.042]; targeting[0.042]; Israel[0.042]; agreed[0.041]; hired[0.041]; Afghanistan[0.041]; Afghanistan[0.041]; team[0.041]; day[0.041]; operator[0.040]; need[0.040]; games[0.040]; sites[0.040]; said[0.040]; said[0.040]; said[0.040]; linked[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Muslim ==> ENTITY: \u001b[32mMuslim\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIslamic[0.048]; Islamic[0.048]; Afghanistan[0.044]; Afghanistan[0.044]; security[0.043]; militant[0.042]; Israel[0.042]; Israel[0.042]; Osama[0.042]; militants[0.042]; Saudi[0.042]; Sudan[0.041]; Sudan[0.041]; included[0.041]; leading[0.040]; Monday[0.040]; Africa[0.040]; deemed[0.040]; Israeli[0.040]; Israeli[0.040]; terrorism[0.040]; said[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saudi\u001b[39m ==> ENTITY: \u001b[32mSaudi Arabia\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOsama[0.047]; Afghanistan[0.044]; Israel[0.043]; Israel[0.043]; Sudan[0.043]; team[0.043]; Islamic[0.042]; Islamic[0.042]; security[0.042]; games[0.042]; militants[0.042]; Israeli[0.042]; Israeli[0.042]; Muslim[0.041]; month[0.041]; terrorism[0.041]; week[0.041]; Laden[0.040]; Laden[0.040]; air[0.040]; air[0.040]; airliners[0.040]; demanded[0.040]; militant[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mpre-emptive strikes\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mOperation Chengiz Khan\u001b[39m <---> \u001b[32mPreemptive war\u001b[39m\t\nSCORES: global= 0.289:0.266[\u001b[31m0.022\u001b[39m]; local()= 0.194:0.151[\u001b[31m0.044\u001b[39m]; log p(e|m)= 0.000:-0.255[\u001b[31m0.255\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsraeli[0.045]; Israeli[0.045]; Israeli[0.045]; Israel[0.045]; Israel[0.045]; Israel[0.045]; strikes[0.043]; strikes[0.043]; personnel[0.042]; targeting[0.041]; attack[0.040]; security[0.040]; security[0.040]; Afghanistan[0.040]; retaliate[0.040]; militants[0.040]; militants[0.040]; targets[0.040]; targets[0.040]; Monday[0.040]; close[0.040]; following[0.040]; air[0.040]; air[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Benjamin Netanyahu\u001b[39m ==> ENTITY: \u001b[32mBenjamin Netanyahu\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.052]; Israel[0.052]; Israel[0.052]; Israeli[0.048]; Israeli[0.048]; Prime[0.042]; David[0.042]; adviser[0.040]; said[0.040]; said[0.040]; international[0.040]; Minister[0.040]; security[0.039]; security[0.039]; week[0.039]; told[0.039]; senior[0.039]; officials[0.039]; response[0.039]; Press[0.039]; Monday[0.038]; terrorism[0.038]; following[0.038]; Asked[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = strategic ==> ENTITY: \u001b[32mStrategy\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.251:-0.251[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; security[0.044]; specific[0.044]; strikes[0.043]; strikes[0.043]; strikes[0.043]; team[0.043]; Afghanistan[0.042]; Afghanistan[0.042]; air[0.042]; air[0.042]; teams[0.042]; administration[0.042]; targeting[0.041]; Monday[0.041]; month[0.041]; targets[0.041]; airliners[0.041]; week[0.040]; Israel[0.040]; day[0.040]; retaliate[0.040]; Israeli[0.040]; Israeli[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = embassies ==> ENTITY: \u001b[32mDiplomatic mission\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.063:0.063[0]; log p(e|m)= -0.030:-0.030[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOsama[0.048]; security[0.046]; Saudi[0.046]; Afghanistan[0.046]; terrorism[0.045]; bombings[0.045]; Sudan[0.044]; games[0.044]; militant[0.043]; Israel[0.043]; Israel[0.043]; month[0.043]; administration[0.043]; University[0.043]; University[0.043]; Laden[0.043]; Laden[0.043]; militants[0.043]; sites[0.042]; use[0.042]; Africa[0.042]; week[0.042]; strategic[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = air strikes ==> ENTITY: \u001b[32mAirstrike\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.172:0.172[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.045]; Israeli[0.044]; Israeli[0.044]; Israeli[0.044]; Monday[0.043]; Monday[0.043]; strikes[0.043]; strikes[0.043]; attack[0.041]; air[0.041]; targets[0.041]; targets[0.041]; targeting[0.041]; militant[0.041]; Afghanistan[0.041]; militants[0.041]; militants[0.041]; close[0.040]; security[0.040]; emptive[0.040]; month[0.040]; Sunday[0.040]; officials[0.040]; response[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = security ==> ENTITY: \u001b[32mSecurity\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.146:0.146[0]; log p(e|m)= -0.846:-0.846[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecurity[0.049]; Israeli[0.045]; Israel[0.045]; Israel[0.045]; Israel[0.045]; airports[0.044]; airport[0.043]; personnel[0.043]; threats[0.042]; threats[0.042]; attack[0.041]; specific[0.041]; use[0.041]; included[0.041]; international[0.041]; officials[0.040]; terrorism[0.040]; militants[0.040]; airlines[0.040]; measures[0.039]; necessary[0.039]; deemed[0.039]; Press[0.038]; following[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Islamic militants\u001b[39m ==> ENTITY: \u001b[32mIslamic terrorism\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nterrorism[0.045]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Afghanistan[0.044]; Osama[0.043]; security[0.043]; security[0.043]; personnel[0.042]; Israeli[0.042]; Israeli[0.042]; Sudan[0.042]; officials[0.041]; international[0.041]; Monday[0.040]; included[0.040]; threats[0.040]; threats[0.040]; senior[0.040]; emptive[0.040]; attack[0.040]; said[0.040]; said[0.040]; week[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = East Africa ==> ENTITY: \u001b[32mEast Africa\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.091:0.091[0]; log p(e|m)= -0.164:-0.164[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOsama[0.045]; Sudan[0.044]; team[0.044]; security[0.043]; Saudi[0.043]; Afghanistan[0.043]; terrorism[0.043]; bombings[0.042]; bombings[0.042]; Muslim[0.042]; teams[0.042]; Islamic[0.042]; Islamic[0.042]; week[0.041]; militant[0.041]; embassies[0.040]; month[0.040]; tour[0.040]; tour[0.040]; administration[0.040]; Israel[0.040]; Israel[0.040]; Laden[0.040]; Laden[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sudan ==> ENTITY: \u001b[32mSudan\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.285:-0.285[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrica[0.051]; Saudi[0.050]; Afghanistan[0.049]; Israel[0.047]; Israel[0.047]; Israeli[0.045]; Israeli[0.045]; Israeli[0.045]; Muslim[0.044]; Osama[0.043]; militants[0.043]; militants[0.043]; Islamic[0.043]; Islamic[0.043]; Islamic[0.043]; militant[0.042]; terrorism[0.041]; bombings[0.041]; team[0.040]; Netanyahu[0.040]; embassies[0.040]; security[0.039]; games[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = David Bar-Illan\u001b[39m ==> ENTITY: \u001b[32mDavid Bar-Illan\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsraeli[0.046]; Israeli[0.046]; Israel[0.045]; Israel[0.045]; Israel[0.045]; international[0.044]; Press[0.043]; Illan[0.042]; month[0.042]; adviser[0.042]; Bar[0.041]; Netanyahu[0.041]; Monday[0.041]; Sunday[0.041]; senior[0.040]; week[0.040]; said[0.040]; said[0.040]; Benjamin[0.040]; leading[0.040]; East[0.039]; Prime[0.039]; told[0.039]; personnel[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = security guards\u001b[39m ==> ENTITY: \u001b[32mSecurity guard\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.098:0.098[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUniversity[0.043]; University[0.043]; bodyguards[0.043]; hired[0.043]; tour[0.043]; tour[0.043]; strategic[0.042]; team[0.042]; private[0.042]; said[0.041]; said[0.041]; said[0.041]; games[0.041]; basketball[0.041]; sites[0.041]; month[0.041]; Nels[0.041]; dl[0.041]; teams[0.041]; seen[0.041]; targeting[0.041]; accompany[0.041]; week[0.041]; day[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mactivist\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mYouth activism\u001b[39m <---> \u001b[32mActivism\u001b[39m\t\nSCORES: global= 0.248:0.242[\u001b[31m0.006\u001b[39m]; local()= 0.092:0.068[\u001b[31m0.024\u001b[39m]; log p(e|m)= -1.100:-0.195[\u001b[32m0.905\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.044]; Israel[0.044]; month[0.043]; security[0.043]; terrorism[0.042]; militant[0.042]; embassies[0.042]; want[0.042]; Israeli[0.042]; Israeli[0.042]; Afghanistan[0.042]; Afghanistan[0.042]; ruling[0.041]; Islamic[0.041]; Islamic[0.041]; day[0.041]; Press[0.041]; retaliate[0.041]; sites[0.041]; Africa[0.041]; Muslim[0.041]; fight[0.041]; Sudan[0.041]; Sudan[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = international airport ==> ENTITY: \u001b[32mInternational airport\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.237:-0.237[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairports[0.048]; personnel[0.046]; airlines[0.043]; included[0.043]; adding[0.043]; security[0.043]; security[0.043]; Monday[0.043]; week[0.042]; officials[0.042]; following[0.042]; air[0.042]; said[0.041]; said[0.041]; senior[0.041]; necessary[0.040]; measures[0.040]; Islamic[0.040]; militants[0.040]; pre[0.040]; Benjamin[0.039]; Israel[0.039]; Israel[0.039]; Israel[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Afghanistan ==> ENTITY: \u001b[32mAfghanistan\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.223:-0.223[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.046]; terrorism[0.045]; Muslim[0.044]; security[0.044]; Islamic[0.043]; Islamic[0.043]; Islamic[0.043]; administration[0.043]; Sudan[0.043]; militants[0.042]; militants[0.042]; Osama[0.041]; militant[0.041]; games[0.041]; Israel[0.041]; Israel[0.041]; month[0.040]; adviser[0.040]; tour[0.040]; Africa[0.040]; strategic[0.040]; basketball[0.039]; bombings[0.039]; day[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = tour operator\u001b[39m ==> ENTITY: \u001b[32mTour operator\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.005:-0.005[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntour[0.044]; private[0.044]; day[0.044]; University[0.043]; University[0.043]; leading[0.043]; sites[0.043]; strategic[0.042]; said[0.042]; said[0.042]; said[0.042]; month[0.042]; demanded[0.041]; basketball[0.041]; Nels[0.041]; dl[0.041]; seen[0.041]; need[0.041]; agreed[0.041]; arrived[0.041]; friendly[0.040]; security[0.040]; air[0.040]; embassies[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.045:0.045[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npersonnel[0.046]; week[0.044]; international[0.044]; Afghanistan[0.043]; David[0.043]; Press[0.042]; use[0.042]; following[0.042]; included[0.042]; specific[0.042]; security[0.041]; security[0.041]; Israel[0.041]; Israel[0.041]; Israel[0.041]; deemed[0.041]; ruling[0.041]; want[0.041]; Monday[0.041]; pre[0.041]; officials[0.040]; air[0.040]; don[0.040]; emerged[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = University of Connecticut\u001b[39m ==> ENTITY: \u001b[32mUniversity of Connecticut\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.119:0.119[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbasketball[0.049]; Connecticut[0.049]; University[0.049]; team[0.049]; teams[0.047]; games[0.046]; East[0.043]; Sunday[0.041]; week[0.041]; Monday[0.041]; strategic[0.040]; leading[0.040]; close[0.040]; hired[0.040]; month[0.040]; Nels[0.039]; dl[0.039]; day[0.038]; air[0.038]; air[0.038]; seen[0.038]; administration[0.038]; guards[0.038]; obviously[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Osama bin Laden\u001b[39m ==> ENTITY: \u001b[32mOsama bin Laden\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfghanistan[0.045]; games[0.043]; Laden[0.043]; security[0.043]; terrorism[0.042]; Saudi[0.042]; use[0.042]; Islamic[0.042]; Islamic[0.042]; Monday[0.042]; Monday[0.042]; militant[0.042]; Muslim[0.042]; administration[0.041]; militants[0.041]; militants[0.041]; targeting[0.041]; University[0.041]; University[0.041]; Israel[0.041]; Israel[0.041]; week[0.041]; linked[0.041]; month[0.041]; \t\n\r [========>.....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 152/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19981022_0710.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = nuclear power ==> ENTITY: \u001b[32mNuclear power\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.230:0.230[0]; log p(e|m)= -0.053:-0.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnuclear[0.045]; nuclear[0.045]; nuclear[0.045]; environmental[0.042]; environmental[0.042]; environmental[0.042]; Government[0.042]; build[0.042]; hydroelectric[0.042]; power[0.041]; power[0.041]; power[0.041]; power[0.041]; country[0.041]; country[0.041]; plants[0.041]; plants[0.041]; plants[0.041]; campaigners[0.040]; government[0.040]; government[0.040]; government[0.040]; government[0.040]; industry[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = power stations ==> ENTITY: \u001b[32mPower station\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.231:0.231[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nelectricity[0.047]; hydroelectric[0.047]; plants[0.045]; plants[0.045]; plants[0.045]; energy[0.045]; power[0.045]; power[0.045]; power[0.045]; power[0.045]; power[0.045]; waste[0.043]; shutting[0.043]; build[0.043]; Energy[0.043]; campaigners[0.042]; company[0.041]; related[0.041]; infrastructure[0.041]; nuclear[0.041]; nuclear[0.041]; nuclear[0.041]; nuclear[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = tax breaks ==> ENTITY: \u001b[32mTax break\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.121:-0.121[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfederal[0.048]; support[0.045]; related[0.045]; Federal[0.044]; government[0.043]; infrastructure[0.043]; upgrades[0.042]; officials[0.042]; governing[0.042]; national[0.042]; company[0.042]; domestic[0.042]; wants[0.041]; laws[0.041]; Energy[0.041]; liberalization[0.041]; encourage[0.040]; energy[0.040]; creation[0.040]; discussion[0.040]; said[0.039]; said[0.039]; said[0.039]; possibility[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = cabinet ==> ENTITY: \u001b[32mCabinet (government)\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.162:0.162[0]; log p(e|m)= -0.781:-0.781[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncabinet[0.045]; cabinet[0.045]; cabinet[0.045]; government[0.043]; government[0.043]; government[0.043]; Government[0.043]; Minister[0.042]; governing[0.042]; national[0.041]; work[0.041]; decision[0.041]; statement[0.041]; statement[0.041]; Federal[0.041]; federal[0.041]; federal[0.041]; representatives[0.040]; officials[0.040]; officials[0.040]; plants[0.040]; plants[0.040]; agreed[0.040]; domestic[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = power industry ==> ENTITY: \u001b[32mElectric power industry\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.178:0.178[0]; log p(e|m)= -2.733:-2.733[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nenergy[0.046]; Energy[0.046]; hydroelectric[0.045]; waste[0.045]; power[0.045]; power[0.045]; power[0.045]; power[0.045]; stations[0.045]; plants[0.043]; plants[0.043]; plants[0.043]; country[0.043]; country[0.043]; nuclear[0.042]; nuclear[0.042]; nuclear[0.042]; nuclear[0.042]; environmental[0.042]; environmental[0.042]; environmental[0.042]; governments[0.042]; Government[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Moritz Leuenberger\u001b[39m ==> ENTITY: \u001b[32mMoritz Leuenberger\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncabinet[0.046]; cabinet[0.046]; cabinet[0.046]; cabinet[0.046]; Minister[0.044]; Government[0.044]; Federal[0.044]; liberalization[0.044]; officials[0.043]; officials[0.043]; discuss[0.043]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; governing[0.043]; representatives[0.042]; government[0.042]; government[0.042]; government[0.042]; decision[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swiss ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.093:0.093[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.047]; country[0.047]; Swiss[0.046]; groups[0.043]; government[0.043]; government[0.043]; government[0.043]; lack[0.042]; governments[0.042]; representatives[0.042]; said[0.042]; planning[0.041]; criticized[0.041]; work[0.041]; eventual[0.040]; Government[0.040]; discuss[0.040]; Thursday[0.040]; support[0.040]; timeframe[0.040]; environmental[0.040]; environmental[0.040]; statement[0.040]; announced[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = electricity ==> ENTITY: \u001b[32mElectricity\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.191:0.191[0]; log p(e|m)= -0.109:-0.109[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nelectricity[0.049]; energy[0.048]; infrastructure[0.046]; Energy[0.046]; power[0.045]; power[0.045]; national[0.044]; government[0.044]; government[0.044]; government[0.044]; hydroelectric[0.043]; hydroelectric[0.043]; nuclear[0.043]; company[0.042]; domestic[0.042]; federal[0.042]; federal[0.042]; plants[0.042]; plants[0.042]; plants[0.042]; build[0.042]; stations[0.041]; creation[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = nuclear waste ==> ENTITY: \u001b[32mRadioactive waste\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.179:0.179[0]; log p(e|m)= -0.035:-0.035[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnuclear[0.044]; nuclear[0.044]; nuclear[0.044]; federal[0.043]; federal[0.043]; country[0.042]; country[0.042]; related[0.042]; Energy[0.041]; build[0.041]; concerns[0.041]; plants[0.041]; plants[0.041]; plants[0.041]; officials[0.041]; officials[0.041]; environmental[0.041]; environmental[0.041]; environmental[0.041]; hydroelectric[0.041]; industry[0.041]; governments[0.041]; saying[0.040]; energy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = hydroelectric plants ==> ENTITY: \u001b[32mHydroelectricity\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.177:0.177[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npower[0.043]; power[0.043]; power[0.043]; power[0.043]; power[0.043]; environmental[0.042]; environmental[0.042]; environmental[0.042]; industry[0.042]; produce[0.042]; plants[0.042]; plants[0.042]; stations[0.042]; waste[0.042]; country[0.042]; country[0.042]; build[0.041]; local[0.040]; government[0.040]; government[0.040]; government[0.040]; government[0.040]; governments[0.040]; lack[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = infrastructure ==> ENTITY: \u001b[32mInfrastructure\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.163:0.163[0]; log p(e|m)= -0.050:-0.050[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nenergy[0.052]; related[0.052]; national[0.051]; government[0.050]; government[0.050]; environmental[0.048]; domestic[0.047]; officials[0.047]; upgrades[0.047]; Federal[0.047]; electricity[0.047]; electricity[0.047]; federal[0.047]; federal[0.047]; creation[0.046]; support[0.046]; Energy[0.046]; plants[0.045]; plants[0.045]; plants[0.045]; paper[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = liberalization ==> ENTITY: \u001b[32mLiberalization\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.149:0.149[0]; log p(e|m)= -0.100:-0.100[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngovernment[0.046]; government[0.046]; domestic[0.045]; market[0.044]; market[0.044]; infrastructure[0.044]; national[0.044]; governing[0.042]; creation[0.042]; support[0.042]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; company[0.040]; discussion[0.040]; tax[0.040]; envisages[0.040]; officials[0.040]; timeframe[0.040]; changing[0.040]; parties[0.040]; laws[0.040]; \t\n\r [=========>....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 166/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19980930_0284.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Impressionist\u001b[39m ==> ENTITY: \u001b[32mImpressionism\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nart[0.046]; Impressionist[0.045]; Monet[0.045]; Monet[0.045]; painting[0.044]; paintings[0.043]; paintings[0.043]; Claude[0.042]; museum[0.042]; works[0.042]; work[0.041]; curator[0.041]; curator[0.041]; Museum[0.041]; described[0.041]; according[0.040]; said[0.040]; appears[0.040]; Alfred[0.040]; staged[0.040]; Arts[0.040]; time[0.040]; Fine[0.039]; small[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Monet\u001b[39m ==> ENTITY: \u001b[32mClaude Monet\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMonet[0.049]; art[0.044]; Claude[0.044]; painting[0.044]; Laurent[0.044]; paintings[0.043]; paintings[0.043]; paintings[0.043]; Impressionist[0.043]; Impressionist[0.043]; works[0.042]; Jean[0.042]; museum[0.041]; connection[0.040]; du[0.040]; days[0.040]; Saint[0.040]; work[0.040]; having[0.040]; Museum[0.040]; time[0.039]; staged[0.039]; Alfred[0.039]; judicial[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saint Laurent du Var\u001b[39m ==> ENTITY: \u001b[32mSaint-Laurent-du-Var\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.113:0.113[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsmall[0.045]; small[0.045]; Nice[0.045]; Riviera[0.043]; Arts[0.043]; Dieppe[0.043]; Alfred[0.042]; according[0.042]; said[0.042]; said[0.042]; said[0.042]; museum[0.041]; museum[0.041]; museum[0.041]; museum[0.041]; library[0.041]; home[0.041]; condition[0.041]; port[0.041]; Claude[0.041]; Jean[0.040]; Museum[0.040]; drove[0.040]; people[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Riviera ==> ENTITY: \u001b[32mFrench Riviera\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.130:0.130[0]; log p(e|m)= -1.839:-1.839[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nboat[0.045]; Wednesday[0.044]; car[0.043]; Alley[0.043]; Nice[0.042]; sail[0.042]; home[0.042]; port[0.042]; small[0.042]; small[0.042]; Laurent[0.041]; shut[0.041]; heist[0.041]; du[0.041]; museum[0.041]; museum[0.041]; museum[0.041]; museum[0.041]; according[0.041]; Jean[0.041]; Cliffs[0.041]; customary[0.041]; suspected[0.040]; anonymity[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = heist ==> ENTITY: \u001b[32mRobbery\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.138:0.138[0]; log p(e|m)= -0.569:-0.569[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrobbery[0.047]; theft[0.045]; staged[0.043]; Alley[0.043]; stolen[0.042]; according[0.042]; alleged[0.042]; work[0.042]; arrested[0.042]; arrested[0.042]; boat[0.042]; Arts[0.041]; time[0.041]; said[0.041]; police[0.041]; police[0.041]; Wednesday[0.041]; Alfred[0.041]; Police[0.040]; connection[0.040]; days[0.040]; armed[0.040]; described[0.040]; suspect[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alfred Sisley\u001b[39m ==> ENTITY: \u001b[32mAlfred Sisley\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.231:0.231[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMonet[0.050]; Monet[0.050]; Claude[0.048]; Jean[0.047]; Laurent[0.047]; paintings[0.046]; paintings[0.046]; Impressionist[0.046]; painting[0.045]; works[0.045]; art[0.045]; Sisley[0.045]; Museum[0.045]; museum[0.044]; museum[0.044]; museum[0.044]; museum[0.044]; work[0.044]; Riviera[0.043]; du[0.043]; Nice[0.043]; Arts[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nice ==> ENTITY: \u001b[32mNice\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.140:0.140[0]; log p(e|m)= -0.260:-0.260[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndu[0.044]; Laurent[0.043]; according[0.043]; Claude[0.043]; Jean[0.042]; Wednesday[0.042]; said[0.042]; said[0.042]; said[0.042]; Riviera[0.042]; Saint[0.042]; home[0.042]; museum[0.041]; museum[0.041]; took[0.041]; staged[0.041]; men[0.041]; ago[0.041]; time[0.041]; Monet[0.041]; Monet[0.041]; appears[0.041]; people[0.041]; days[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = curator ==> ENTITY: \u001b[32mCurator\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.020:-0.020[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncurator[0.047]; Museum[0.046]; museum[0.045]; art[0.043]; spoke[0.042]; Arts[0.042]; works[0.042]; theft[0.042]; judicial[0.042]; stolen[0.041]; said[0.041]; time[0.041]; Wednesday[0.041]; paintings[0.041]; paintings[0.041]; Claude[0.041]; small[0.040]; valuable[0.040]; according[0.040]; Impressionist[0.040]; Impressionist[0.040]; people[0.040]; sources[0.040]; sources[0.040]; \t\n\r [==========>...................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 174/727 \r\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bank of Thailand\u001b[39m ==> ENTITY: \u001b[32mBank of Thailand\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nViravaidya[0.046]; Thai[0.045]; Thai[0.045]; job[0.044]; job[0.044]; finance[0.044]; salary[0.043]; plan[0.043]; banks[0.042]; timely[0.042]; SET[0.041]; set[0.041]; executives[0.041]; executives[0.041]; continued[0.040]; better[0.040]; country[0.040]; guys[0.040]; corporate[0.040]; issue[0.040]; watchdog[0.040]; hard[0.040]; report[0.040]; propose[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = banks ==> ENTITY: \u001b[32mBank\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBank[0.049]; finance[0.047]; corporate[0.046]; companies[0.045]; companies[0.045]; checks[0.044]; watchdog[0.044]; listed[0.044]; country[0.044]; public[0.043]; possible[0.043]; Thailand[0.043]; failed[0.042]; company[0.042]; company[0.042]; report[0.042]; play[0.042]; know[0.042]; know[0.042]; issue[0.042]; newspapers[0.042]; fully[0.042]; timely[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Thai people\u001b[39m ==> ENTITY: \u001b[32mThai people\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.129:0.129[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThailand[0.050]; Thailand[0.050]; Thai[0.049]; country[0.047]; people[0.044]; Mechai[0.043]; Khun[0.043]; report[0.041]; finance[0.041]; lifestyle[0.041]; lifestyle[0.041]; play[0.040]; Bank[0.040]; TV[0.040]; like[0.040]; like[0.040]; media[0.039]; media[0.039]; media[0.039]; ordinary[0.039]; seminar[0.039]; guys[0.039]; issues[0.039]; plan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Viravaidya\u001b[39m ==> ENTITY: \u001b[32mMechai Viravaidya\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.113:0.113[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThailand[0.051]; support[0.047]; media[0.045]; company[0.045]; company[0.045]; advise[0.045]; corporate[0.045]; companies[0.044]; companies[0.044]; people[0.044]; abuses[0.043]; possible[0.043]; SET[0.043]; finance[0.043]; request[0.042]; executives[0.042]; failed[0.042]; lifestyle[0.042]; monitor[0.042]; ENGLISH[0.041]; evidence[0.041]; like[0.041]; watchdog[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = lifestyle ==> ENTITY: \u001b[32mLifestyle (sociology)\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.365:-0.365[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlifestyle[0.051]; media[0.045]; media[0.045]; corporate[0.044]; country[0.043]; newspapers[0.042]; TV[0.042]; people[0.042]; people[0.042]; ethics[0.042]; finance[0.041]; socially[0.041]; executives[0.041]; executives[0.041]; discover[0.041]; Viravaidya[0.041]; agree[0.040]; like[0.040]; like[0.040]; job[0.040]; job[0.040]; attract[0.040]; companies[0.040]; companies[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = newspapers ==> ENTITY: \u001b[32mNewspaper\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.016:-0.016[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.050]; media[0.045]; media[0.045]; media[0.045]; executives[0.043]; executives[0.043]; companies[0.042]; TV[0.042]; Thailand[0.041]; Thailand[0.041]; public[0.041]; report[0.041]; like[0.041]; job[0.040]; job[0.040]; continued[0.040]; stations[0.040]; shares[0.040]; Thai[0.040]; Thai[0.040]; hard[0.040]; financial[0.040]; cover[0.040]; socially[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = finance ==> ENTITY: \u001b[32mFinance\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.140:0.140[0]; log p(e|m)= -0.045:-0.045[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncorporate[0.049]; companies[0.048]; companies[0.048]; Bank[0.046]; banks[0.046]; company[0.045]; company[0.045]; public[0.044]; job[0.043]; job[0.043]; timely[0.043]; support[0.043]; executives[0.043]; executives[0.043]; country[0.042]; Thailand[0.042]; plan[0.042]; possible[0.041]; checks[0.041]; continued[0.041]; inefficiency[0.041]; listed[0.040]; watchdog[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = corporate ethics\u001b[39m ==> ENTITY: \u001b[32mBusiness ethics\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.127:0.127[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncompanies[0.045]; companies[0.045]; finance[0.044]; media[0.043]; media[0.043]; job[0.042]; executives[0.042]; company[0.042]; company[0.042]; support[0.041]; monitor[0.041]; guys[0.041]; possible[0.041]; better[0.041]; timely[0.041]; SET[0.041]; punished[0.041]; people[0.041]; shares[0.041]; evidence[0.041]; fully[0.041]; listed[0.040]; lifestyle[0.040]; Viravaidya[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = TV stations ==> ENTITY: \u001b[32mTelevision channel\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.103:0.103[0]; log p(e|m)= -1.194:-1.194[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmedia[0.044]; media[0.044]; media[0.044]; newspapers[0.044]; Thailand[0.044]; Thailand[0.044]; country[0.043]; Thai[0.043]; Thai[0.043]; Telephone[0.043]; public[0.041]; cover[0.041]; additional[0.041]; play[0.041]; Bank[0.041]; propose[0.041]; companies[0.041]; financial[0.040]; seminar[0.040]; guys[0.040]; plan[0.040]; executives[0.040]; executives[0.040]; finance[0.040]; \t\n\r [==========>...................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 186/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19981113_0729.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = airport ==> ENTITY: \u001b[32mAirport\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.095:0.095[0]; log p(e|m)= -0.121:-0.121[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairport[0.047]; Airport[0.046]; Airport[0.046]; said[0.043]; said[0.043]; Aviation[0.043]; International[0.043]; opened[0.043]; officials[0.042]; Thursday[0.042]; operation[0.042]; southern[0.042]; operations[0.041]; schedule[0.040]; Friday[0.040]; fledgling[0.040]; open[0.040]; Palestinian[0.040]; Palestinian[0.040]; Palestinian[0.040]; Palestinians[0.040]; Authority[0.040]; operating[0.039]; held[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = passport ==> ENTITY: \u001b[32mPassport\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.067:-0.067[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nchecks[0.046]; security[0.044]; luggage[0.044]; employees[0.043]; customs[0.043]; week[0.043]; officials[0.042]; procedures[0.042]; personnel[0.042]; need[0.042]; plane[0.042]; Thursday[0.041]; date[0.041]; arrived[0.041]; Authority[0.041]; operating[0.041]; passengers[0.041]; passengers[0.041]; lack[0.040]; Israelis[0.040]; Israelis[0.040]; departure[0.040]; Israeli[0.040]; set[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cabinet ==> ENTITY: \u001b[32mCabinet of Israel\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.184:0.184[0]; log p(e|m)= -3.381:-3.381[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.051]; Israel[0.051]; Israeli[0.048]; Israeli[0.048]; Israeli[0.048]; Palestinian[0.046]; Palestinian[0.046]; Palestinian[0.046]; Palestinian[0.046]; Palestinian[0.046]; Palestinian[0.046]; Israelis[0.045]; Minister[0.044]; Netanyahu[0.044]; week[0.043]; week[0.043]; Gaza[0.043]; officials[0.043]; officials[0.043]; Palestinians[0.043]; duties[0.043]; Prime[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yasser Arafat International Airport\u001b[39m ==> ENTITY: \u001b[32mYasser Arafat International Airport\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.252:0.252[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPalestinians[0.044]; Gaza[0.044]; Israel[0.044]; Israel[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Israelis[0.043]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; airport[0.040]; airport[0.040]; airport[0.040]; Fayez[0.040]; Strip[0.039]; Netanyahu[0.039]; open[0.039]; set[0.039]; security[0.038]; Zeidan[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = customs ==> ENTITY: \u001b[32mCustoms\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.498:-0.498[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nchecks[0.043]; inspection[0.043]; personnel[0.043]; luggage[0.043]; officials[0.043]; security[0.042]; duties[0.042]; employees[0.042]; Authority[0.042]; procedures[0.042]; airport[0.042]; airport[0.042]; airport[0.042]; airport[0.042]; Aviation[0.042]; operation[0.042]; passport[0.042]; set[0.041]; arrived[0.040]; Civil[0.040]; passengers[0.040]; passengers[0.040]; operating[0.040]; plane[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = ratification ==> ENTITY: \u001b[32mRatification\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.120:0.120[0]; log p(e|m)= -0.168:-0.168[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nofficials[0.044]; officials[0.044]; accord[0.043]; southern[0.042]; agreement[0.042]; Fayez[0.042]; held[0.042]; independence[0.042]; Cabinet[0.041]; date[0.041]; exercise[0.041]; Israel[0.041]; Israel[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; Thursday[0.041]; Authority[0.041]; Authority[0.041]; security[0.041]; personnel[0.041]; set[0.041]; duties[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Palestinian ==> ENTITY: \u001b[32mPalestinian people\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.161:0.161[0]; log p(e|m)= -0.389:-0.389[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.049]; Israel[0.049]; Palestinian[0.047]; Palestinian[0.047]; Israeli[0.046]; Israeli[0.046]; Gaza[0.045]; Palestinians[0.045]; Yasser[0.042]; Cabinet[0.041]; southern[0.041]; Netanyahu[0.040]; Arafat[0.040]; week[0.039]; week[0.039]; agreement[0.039]; peace[0.039]; weeks[0.039]; Authority[0.038]; operation[0.038]; Civil[0.038]; officials[0.038]; International[0.038]; await[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Civil Aviation ==> ENTITY: \u001b[32mCivil aviation\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.152:0.152[0]; log p(e|m)= -0.272:-0.272[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCabinet[0.048]; Minister[0.045]; Authority[0.044]; operations[0.043]; International[0.043]; Airport[0.043]; Airport[0.043]; airport[0.043]; airport[0.043]; operation[0.042]; operating[0.041]; Thursday[0.041]; fledgling[0.041]; officials[0.041]; officials[0.041]; agreement[0.041]; Friday[0.040]; week[0.040]; week[0.040]; ratification[0.040]; independence[0.040]; start[0.040]; held[0.040]; Prime[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Benjamin Netanyahu\u001b[39m ==> ENTITY: \u001b[32mBenjamin Netanyahu\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.049]; Israel[0.049]; Israeli[0.045]; Israeli[0.045]; Israeli[0.045]; Cabinet[0.044]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; Israelis[0.043]; Palestinians[0.040]; Prime[0.040]; Gaza[0.039]; said[0.038]; said[0.038]; said[0.038]; said[0.038]; Fayez[0.037]; Minister[0.037]; security[0.037]; International[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gaza Strip\u001b[39m ==> ENTITY: \u001b[32mGaza Strip\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.221:0.221[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.046]; Israel[0.046]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Israeli[0.044]; Israeli[0.044]; Israeli[0.044]; Palestinians[0.043]; Israelis[0.043]; Fayez[0.041]; southern[0.040]; security[0.039]; Cabinet[0.039]; operation[0.039]; operations[0.038]; Authority[0.038]; Authority[0.038]; Netanyahu[0.038]; Zeidan[0.038]; Yasser[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = independence ==> ENTITY: \u001b[32mIndependence\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.779:-0.779[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.049]; Israel[0.049]; Israeli[0.046]; Israeli[0.046]; Israeli[0.046]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinians[0.043]; Gaza[0.042]; agreement[0.040]; southern[0.040]; accord[0.039]; peace[0.039]; Cabinet[0.039]; Netanyahu[0.039]; week[0.039]; week[0.039]; held[0.038]; operations[0.038]; weeks[0.038]; fledgling[0.038]; Authority[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Civil Aviation Authority ==> ENTITY: \u001b[32mNational aviation authority\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.408:-0.408[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npersonnel[0.047]; Israel[0.044]; operations[0.043]; Airport[0.043]; airport[0.043]; airport[0.043]; airport[0.043]; airport[0.043]; Israeli[0.042]; Israeli[0.042]; Israeli[0.042]; security[0.042]; operation[0.042]; procedures[0.041]; employees[0.041]; International[0.041]; Palestinian[0.040]; Palestinian[0.040]; Palestinian[0.040]; Palestinian[0.040]; inspection[0.040]; plane[0.040]; operating[0.040]; Cabinet[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIsraeli\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIsrael\u001b[39m <---> \u001b[32mIsraelis\u001b[39m\t\nSCORES: global= 0.251:0.239[\u001b[31m0.012\u001b[39m]; local()= 0.208:0.217[\u001b[32m0.010\u001b[39m]; log p(e|m)= -0.348:-1.749[\u001b[31m1.401\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.051]; Israel[0.051]; Israeli[0.046]; Israeli[0.046]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinians[0.043]; Gaza[0.042]; Netanyahu[0.041]; operation[0.040]; International[0.038]; Cabinet[0.038]; Strip[0.038]; southern[0.038]; Yasser[0.038]; peace[0.038]; Fayez[0.038]; Thursday[0.038]; Arafat[0.037]; operations[0.037]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIsraelis\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIsrael\u001b[39m <---> \u001b[32mIsraelis\u001b[39m\t\nSCORES: global= 0.258:0.254[\u001b[31m0.005\u001b[39m]; local()= 0.195:0.204[\u001b[32m0.009\u001b[39m]; log p(e|m)= -0.546:-0.009[\u001b[32m0.537\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.056]; Israeli[0.052]; Israeli[0.052]; Israelis[0.050]; Palestinian[0.048]; Palestinian[0.048]; Palestinian[0.048]; Palestinian[0.048]; personnel[0.045]; Gaza[0.045]; security[0.045]; plane[0.043]; Netanyahu[0.043]; Zeidan[0.042]; Zeidan[0.042]; Zeidan[0.042]; operation[0.042]; Cabinet[0.042]; southern[0.042]; officials[0.041]; week[0.041]; week[0.041]; \t\n\r [===========>..................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 201/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19981001_0866.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hungarian Jewish\u001b[39m ==> ENTITY: \u001b[32mHistory of the Jews in Hungary\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJewish[0.048]; Jewish[0.048]; Jewish[0.048]; Hungarian[0.047]; Hungarian[0.047]; paid[0.041]; forints[0.041]; forints[0.041]; MTI[0.041]; confiscations[0.041]; annually[0.040]; schools[0.040]; reported[0.040]; compensation[0.040]; compensation[0.040]; communist[0.040]; million[0.040]; million[0.040]; million[0.040]; communists[0.040]; mainly[0.040]; money[0.039]; belonged[0.039]; Federation[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = MTI ==> ENTITY: \u001b[32mMagyar Távirati Iroda\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.580:-0.580[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnews[0.046]; service[0.043]; Hungarian[0.043]; Hungarian[0.043]; Hungarian[0.043]; reported[0.043]; forints[0.042]; forints[0.042]; director[0.042]; services[0.042]; financial[0.042]; agency[0.042]; funding[0.041]; current[0.041]; million[0.041]; million[0.041]; million[0.041]; religious[0.041]; executive[0.041]; Thursday[0.040]; schools[0.040]; government[0.040]; government[0.040]; annually[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = the Holocaust ==> ENTITY: \u001b[32mThe Holocaust\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.011:-0.011[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJewish[0.049]; Jewish[0.049]; Jews[0.048]; Hungary[0.043]; religious[0.042]; community[0.042]; community[0.042]; communists[0.042]; communists[0.042]; Hungarian[0.042]; mainly[0.041]; government[0.041]; service[0.041]; social[0.041]; said[0.040]; buildings[0.040]; Federation[0.040]; paid[0.040]; spared[0.040]; services[0.039]; million[0.039]; million[0.039]; million[0.039]; actual[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = social welfare ==> ENTITY: \u001b[32mWelfare\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.158:0.158[0]; log p(e|m)= -0.160:-0.160[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nservices[0.047]; financial[0.047]; funding[0.047]; education[0.046]; health[0.046]; government[0.045]; current[0.045]; collective[0.045]; service[0.044]; compensation[0.044]; era[0.043]; percentage[0.043]; state[0.043]; communities[0.043]; actual[0.042]; paid[0.042]; collectively[0.042]; value[0.042]; annually[0.042]; money[0.041]; assessed[0.041]; basis[0.041]; Hungary[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vatican ==> ENTITY: \u001b[32mHoly See\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.096:0.096[0]; log p(e|m)= -0.681:-0.681[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nreligious[0.045]; Hungary[0.045]; government[0.043]; similar[0.042]; June[0.042]; Federation[0.042]; service[0.042]; Hungarian[0.042]; signed[0.042]; said[0.042]; Jewish[0.042]; social[0.041]; agreement[0.041]; assumed[0.041]; paid[0.041]; spared[0.041]; current[0.041]; communists[0.041]; services[0.041]; kind[0.041]; mainly[0.041]; million[0.040]; million[0.040]; Jews[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = communist ==> ENTITY: \u001b[32mCommunism\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.083:-0.083[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncommunists[0.045]; Jewish[0.044]; Jewish[0.044]; Jewish[0.044]; Jewish[0.044]; social[0.043]; government[0.042]; government[0.042]; Hungarian[0.042]; Hungarian[0.042]; Hungarian[0.042]; Federation[0.042]; religious[0.042]; collective[0.041]; era[0.041]; agree[0.041]; real[0.040]; state[0.040]; current[0.040]; service[0.040]; confiscations[0.039]; news[0.039]; mainly[0.039]; collectively[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = forints\u001b[39m ==> ENTITY: \u001b[32mHungarian forint\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nforints[0.045]; Hungarian[0.044]; Hungarian[0.044]; Hungarian[0.044]; paid[0.044]; money[0.044]; Hungary[0.044]; million[0.043]; million[0.043]; million[0.043]; value[0.042]; property[0.041]; percentage[0.041]; Jews[0.040]; signed[0.040]; annually[0.040]; current[0.040]; government[0.040]; government[0.040]; estate[0.040]; mainly[0.040]; MTI[0.040]; financial[0.039]; June[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = nationalization ==> ENTITY: \u001b[32mNationalization\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.019:-0.019[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfinancial[0.046]; government[0.045]; government[0.045]; religious[0.045]; social[0.044]; services[0.044]; compensation[0.044]; compensation[0.044]; current[0.044]; service[0.044]; Thursday[0.043]; mainly[0.043]; paid[0.043]; collectively[0.043]; funding[0.043]; state[0.043]; Hungary[0.043]; collective[0.043]; signed[0.043]; million[0.042]; million[0.042]; million[0.042]; agency[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hungarian ==> ENTITY: \u001b[32mHungary\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.116:0.116[0]; log p(e|m)= -1.033:-1.033[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHungarian[0.046]; Hungarian[0.046]; Jewish[0.043]; Jewish[0.043]; Jewish[0.043]; Jewish[0.043]; communist[0.042]; communities[0.042]; communities[0.042]; communists[0.042]; Federation[0.042]; government[0.041]; government[0.041]; percentage[0.041]; state[0.041]; million[0.041]; million[0.041]; million[0.041]; confiscations[0.040]; community[0.040]; forints[0.040]; forints[0.040]; reported[0.040]; signed[0.040]; \t\n\r [============>.................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 210/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19980603_0791.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = breast cancer\u001b[39m ==> ENTITY: \u001b[32mBreast cancer\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.125:0.125[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbrain[0.047]; brain[0.047]; medical[0.044]; surgery[0.043]; surgery[0.043]; surgery[0.043]; care[0.042]; proper[0.042]; suffering[0.042]; cancerous[0.042]; cancerous[0.042]; diseases[0.042]; doctors[0.041]; tissues[0.041]; tissues[0.041]; Radio[0.041]; Radio[0.041]; people[0.040]; growth[0.040]; established[0.040]; dying[0.039]; conjunction[0.039]; Hospital[0.039]; change[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = tissues\u001b[39m ==> ENTITY: \u001b[32mTissue (biology)\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntissues[0.053]; brain[0.051]; brain[0.051]; brain[0.051]; brain[0.051]; Venous[0.049]; blood[0.049]; capillaries[0.049]; treatment[0.048]; Surgery[0.047]; Surgery[0.047]; deep[0.046]; patients[0.046]; Malformation[0.046]; clotting[0.046]; treating[0.045]; medical[0.045]; organs[0.045]; cancerous[0.045]; cancerous[0.045]; inside[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Malformation\u001b[39m ==> ENTITY: \u001b[32mCongenital disorder\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.223:0.223[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbrain[0.045]; brain[0.045]; Venous[0.044]; Surgery[0.043]; surgery[0.042]; surgery[0.042]; surgery[0.042]; surgery[0.042]; surgery[0.042]; causes[0.042]; tissues[0.041]; tissues[0.041]; blood[0.041]; blood[0.041]; diseases[0.041]; patients[0.041]; patients[0.041]; patients[0.041]; treatment[0.041]; treatments[0.040]; loss[0.040]; hospital[0.039]; hospital[0.039]; capillaries[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = three dimensional ==> ENTITY: \u001b[32mThree-dimensional space\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.059:0.059[0]; log p(e|m)= -0.766:-0.766[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVenous[0.046]; Surgery[0.045]; Surgery[0.045]; patients[0.043]; Malformation[0.043]; embedded[0.042]; world[0.041]; technology[0.041]; involves[0.041]; capillaries[0.041]; medical[0.041]; technique[0.041]; technique[0.041]; surgery[0.041]; surgery[0.041]; surgery[0.041]; surgery[0.041]; surgery[0.041]; sensitive[0.041]; Hospital[0.040]; method[0.040]; clotting[0.040]; doctors[0.040]; computers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Thailand ==> ENTITY: \u001b[32mThailand\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.066:0.066[0]; log p(e|m)= -0.241:-0.241[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBangkok[0.050]; THAILAND[0.048]; people[0.047]; world[0.046]; conjunction[0.046]; proper[0.044]; growth[0.044]; sensitive[0.043]; breast[0.043]; lack[0.043]; modern[0.043]; involves[0.043]; Radio[0.042]; Radio[0.042]; presently[0.042]; medical[0.042]; difficult[0.042]; risky[0.042]; Institute[0.042]; doctors[0.042]; remove[0.041]; removal[0.041]; equipment[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mclotting\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mThrombus\u001b[39m <---> \u001b[32mCoagulation\u001b[39m\t\nSCORES: global= 0.272:0.261[\u001b[31m0.012\u001b[39m]; local()= 0.233:0.220[\u001b[31m0.013\u001b[39m]; log p(e|m)= -1.435:-0.285[\u001b[32m1.150\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVenous[0.044]; blood[0.044]; blood[0.044]; brain[0.043]; brain[0.043]; treatment[0.043]; causes[0.042]; surgery[0.042]; surgery[0.042]; surgery[0.042]; surgery[0.042]; surgery[0.042]; treatments[0.042]; Surgery[0.041]; patients[0.041]; patients[0.041]; patients[0.041]; treating[0.040]; tissues[0.040]; tissues[0.040]; Malformation[0.040]; capillaries[0.040]; veins[0.040]; diseases[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = stereotaxic surgery\u001b[39m ==> ENTITY: \u001b[32mStereotactic surgery\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.238:0.238[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsurgery[0.044]; surgery[0.044]; surgery[0.044]; surgery[0.044]; surgery[0.044]; surgery[0.044]; Venous[0.043]; medical[0.043]; treatment[0.043]; Surgery[0.042]; treatments[0.042]; brain[0.041]; brain[0.041]; brain[0.041]; brain[0.041]; patients[0.041]; Malformation[0.041]; cancer[0.040]; doctors[0.040]; treating[0.040]; method[0.039]; tissues[0.039]; tissues[0.039]; organs[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = brain tumours\u001b[39m ==> ENTITY: \u001b[32mBrain tumor\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncancer[0.048]; brain[0.047]; Surgery[0.046]; Surgery[0.046]; suffering[0.045]; surgery[0.045]; surgery[0.045]; surgery[0.045]; medical[0.042]; doctors[0.041]; Hospital[0.041]; cancerous[0.041]; care[0.040]; haemorrhage[0.040]; dying[0.040]; tissues[0.040]; breast[0.040]; diseases[0.040]; fatal[0.040]; radiation[0.039]; Accelerator[0.039]; Institute[0.039]; remove[0.038]; removal[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bangkok\u001b[39m ==> ENTITY: \u001b[32mBangkok\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.060:0.060[0]; log p(e|m)= -0.001:-0.001[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThailand[0.051]; people[0.045]; THAILAND[0.044]; world[0.043]; medical[0.042]; near[0.042]; presently[0.042]; established[0.041]; technology[0.041]; equipment[0.041]; radio[0.041]; radio[0.041]; deep[0.041]; involves[0.041]; modern[0.041]; sensitive[0.041]; suffering[0.041]; proper[0.041]; difficult[0.041]; order[0.040]; lack[0.040]; require[0.040]; Institute[0.040]; dying[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = capillaries\u001b[39m ==> ENTITY: \u001b[32mCapillary\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbrain[0.045]; brain[0.045]; blood[0.045]; blood[0.045]; Venous[0.044]; tissues[0.043]; tissues[0.043]; Surgery[0.043]; veins[0.043]; deep[0.042]; patients[0.041]; patients[0.041]; patients[0.041]; Malformation[0.041]; causes[0.041]; clotting[0.041]; treatment[0.040]; treating[0.039]; hospital[0.039]; hospital[0.039]; loss[0.039]; normal[0.039]; inside[0.039]; contact[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mradiation\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRadiation therapy\u001b[39m <---> \u001b[32mRadiation\u001b[39m\t\nSCORES: global= 0.257:0.245[\u001b[31m0.012\u001b[39m]; local()= 0.220:0.158[\u001b[31m0.061\u001b[39m]; log p(e|m)= -1.487:-0.485[\u001b[32m1.003\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncancer[0.045]; treatment[0.044]; treatments[0.043]; medical[0.043]; radio[0.043]; radio[0.043]; radio[0.043]; surgery[0.042]; surgery[0.042]; surgery[0.042]; surgery[0.042]; surgery[0.042]; surgery[0.042]; patients[0.041]; patients[0.041]; suffering[0.041]; brain[0.041]; brain[0.041]; brain[0.041]; brain[0.041]; doctors[0.040]; sensitive[0.040]; cancerous[0.040]; cancerous[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = cancerous ==> ENTITY: \u001b[32mCancer\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.192:0.192[0]; log p(e|m)= -0.061:-0.061[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncancer[0.045]; brain[0.044]; brain[0.044]; brain[0.044]; brain[0.044]; treatment[0.042]; Venous[0.042]; tissues[0.042]; tissues[0.042]; cancerous[0.042]; Surgery[0.041]; Surgery[0.041]; surgery[0.041]; surgery[0.041]; surgery[0.041]; surgery[0.041]; surgery[0.041]; blood[0.040]; medical[0.040]; diseases[0.040]; diseases[0.040]; doctors[0.040]; treating[0.040]; patients[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = haemorrhage ==> ENTITY: \u001b[32mBleeding\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.037:-0.037[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbrain[0.044]; brain[0.044]; brain[0.044]; surgery[0.044]; surgery[0.044]; surgery[0.044]; Surgery[0.044]; Surgery[0.044]; care[0.043]; suffering[0.043]; medical[0.042]; tissues[0.042]; tissues[0.042]; cancer[0.042]; diseases[0.041]; doctors[0.041]; deep[0.040]; fatal[0.040]; breast[0.040]; Hospital[0.039]; cancerous[0.039]; cancerous[0.039]; dying[0.039]; lack[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Linear Accelerator\u001b[39m ==> ENTITY: \u001b[32mLinear particle accelerator\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVenous[0.046]; Surgery[0.045]; Surgery[0.045]; patients[0.043]; patients[0.043]; Malformation[0.043]; medical[0.042]; radiation[0.042]; hospital[0.041]; Institute[0.041]; treatment[0.041]; capillaries[0.041]; technology[0.041]; surgery[0.041]; surgery[0.041]; surgery[0.041]; surgery[0.041]; surgery[0.041]; surgery[0.041]; treatments[0.041]; Hospital[0.040]; clotting[0.040]; doctors[0.040]; successful[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Radio Surgery\u001b[39m ==> ENTITY: \u001b[32mRadiosurgery\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsurgery[0.045]; surgery[0.045]; surgery[0.045]; surgery[0.045]; surgery[0.045]; cancer[0.044]; medical[0.044]; Surgery[0.043]; doctors[0.042]; technology[0.041]; brain[0.041]; brain[0.041]; brain[0.041]; Hospital[0.041]; radiation[0.041]; cancerous[0.040]; cancerous[0.040]; care[0.040]; tissues[0.040]; tissues[0.040]; technique[0.040]; method[0.039]; sensitive[0.038]; growth[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = conscious ==> ENTITY: \u001b[32mConsciousness\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.075:-0.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbrain[0.046]; brain[0.046]; Venous[0.046]; Surgery[0.044]; patients[0.043]; patients[0.043]; patients[0.043]; Malformation[0.043]; world[0.041]; hospital[0.041]; hospital[0.041]; causes[0.041]; capillaries[0.041]; surgery[0.040]; surgery[0.040]; surgery[0.040]; surgery[0.040]; surgery[0.040]; surgery[0.040]; surgery[0.040]; involves[0.040]; order[0.040]; deep[0.040]; information[0.040]; \t\n\r [=============>................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 226/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19980610_0111.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = organized crime ==> ENTITY: \u001b[32mOrganized crime\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.036:0.036[0]; log p(e|m)= -0.012:-0.012[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.047]; Russian[0.047]; Yuri[0.044]; Russia[0.043]; Boris[0.042]; company[0.042]; agencies[0.042]; Berezovsky[0.042]; involved[0.042]; region[0.042]; business[0.041]; local[0.041]; according[0.041]; separate[0.040]; Sibneft[0.040]; Sibneft[0.040]; Sibneft[0.040]; businessman[0.040]; reports[0.040]; mogul[0.040]; building[0.040]; appeared[0.040]; Businessman[0.040]; controversial[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Boris Berezovsky ==> ENTITY: \u001b[32mBoris Berezovsky (businessman)\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.137:0.137[0]; log p(e|m)= -2.703:-2.703[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYuri[0.055]; Russian[0.054]; Russian[0.054]; Russia[0.051]; Russia[0.051]; Russia[0.051]; Sibneft[0.051]; Sibneft[0.051]; Sibneft[0.051]; Siberian[0.046]; Siberia[0.045]; Kuznetsky[0.042]; Kemerovo[0.041]; Kemerovo[0.041]; mobsters[0.040]; mogul[0.040]; saying[0.040]; news[0.039]; news[0.039]; motives[0.039]; fraud[0.039]; quoted[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = embezzlement\u001b[39m ==> ENTITY: \u001b[32mEmbezzlement\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.120:0.120[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfraud[0.051]; charges[0.049]; accused[0.047]; Kemerovo[0.045]; involved[0.044]; possible[0.044]; arrested[0.044]; linked[0.044]; trial[0.044]; Kuznetsky[0.043]; Zinin[0.043]; tax[0.042]; awaiting[0.042]; custody[0.042]; fight[0.042]; hundreds[0.042]; selling[0.042]; mob[0.042]; number[0.042]; Officials[0.041]; contract[0.041]; murder[0.041]; murder[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Leninsk-Kuznetsky\u001b[39m ==> ENTITY: \u001b[32mLeninsk-Kuznetsky (city)\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.106:0.106[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntown[0.046]; city[0.045]; Kemerovo[0.044]; regions[0.043]; main[0.043]; separate[0.043]; mining[0.042]; Zinin[0.042]; mayor[0.042]; solved[0.041]; Russia[0.041]; Russia[0.041]; involved[0.041]; coal[0.041]; coal[0.041]; launched[0.041]; custody[0.041]; Sibneft[0.040]; Sibneft[0.040]; Officials[0.040]; linked[0.040]; Gennady[0.040]; products[0.040]; contract[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36moil products\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPetroleum\u001b[39m <---> \u001b[32mOil\u001b[39m\t\nSCORES: global= 0.278:0.276[\u001b[31m0.003\u001b[39m]; local()= 0.143:0.131[\u001b[31m0.012\u001b[39m]; log p(e|m)= -0.629:-0.916[\u001b[31m0.287\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \noil[0.051]; companies[0.043]; mining[0.043]; region[0.043]; businesses[0.043]; business[0.042]; nationwide[0.042]; coal[0.042]; coal[0.042]; largest[0.042]; Kemerovo[0.041]; Kemerovo[0.041]; company[0.041]; according[0.041]; possible[0.041]; linked[0.041]; profitable[0.040]; regions[0.040]; building[0.040]; agencies[0.040]; locally[0.040]; small[0.040]; small[0.040]; Officials[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.159:0.159[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.048]; Russia[0.048]; Russian[0.046]; Russian[0.046]; Siberia[0.045]; building[0.044]; region[0.043]; contract[0.041]; regions[0.041]; agencies[0.041]; Yuri[0.041]; Siberian[0.041]; Boris[0.041]; controversial[0.040]; news[0.040]; news[0.040]; launched[0.040]; central[0.040]; hundreds[0.039]; main[0.039]; number[0.039]; town[0.039]; town[0.039]; involved[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = mayor ==> ENTITY: \u001b[32mMayor\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.045:0.045[0]; log p(e|m)= -0.152:-0.152[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMayor[0.043]; city[0.043]; wouldn[0.043]; Boris[0.043]; possible[0.042]; killer[0.042]; town[0.042]; tax[0.042]; charges[0.042]; number[0.042]; arrested[0.042]; Police[0.042]; business[0.041]; murder[0.041]; murder[0.041]; trial[0.041]; killings[0.041]; launched[0.041]; fraud[0.041]; Russia[0.041]; Russia[0.041]; Gennady[0.041]; accused[0.041]; Officials[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = coal ==> ENTITY: \u001b[32mCoal\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.053:-0.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncoal[0.049]; mining[0.045]; region[0.044]; businesses[0.044]; companies[0.042]; agencies[0.042]; town[0.042]; town[0.042]; building[0.042]; business[0.041]; oil[0.041]; oil[0.041]; regions[0.041]; small[0.041]; small[0.041]; company[0.041]; central[0.041]; number[0.040]; local[0.040]; linked[0.040]; city[0.040]; profitable[0.040]; locally[0.040]; organized[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mRussian\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRussia\u001b[39m <---> \u001b[32mRussians\u001b[39m\t\nSCORES: global= 0.252:0.226[\u001b[31m0.026\u001b[39m]; local()= 0.127:0.113[\u001b[31m0.014\u001b[39m]; log p(e|m)= -1.053:-2.235[\u001b[31m1.182\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.049]; Russia[0.049]; Siberia[0.045]; Yuri[0.044]; crime[0.044]; Boris[0.044]; region[0.044]; Berezovsky[0.042]; Siberian[0.041]; agencies[0.041]; work[0.041]; according[0.040]; building[0.040]; organized[0.040]; business[0.040]; central[0.040]; companies[0.040]; appeared[0.040]; local[0.040]; businessman[0.040]; Wednesday[0.039]; shot[0.039]; quoted[0.039]; unidentified[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kemerovo\u001b[39m ==> ENTITY: \u001b[32mKemerovo\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKemerovo[0.057]; region[0.045]; Siberia[0.044]; regions[0.043]; Siberian[0.042]; mining[0.042]; Russia[0.042]; Russia[0.042]; central[0.041]; agencies[0.041]; building[0.041]; coal[0.041]; coal[0.041]; main[0.040]; Sibneft[0.040]; Sibneft[0.040]; Sibneft[0.040]; town[0.040]; town[0.040]; Russian[0.040]; Russian[0.040]; largest[0.040]; company[0.040]; companies[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = tax fraud ==> ENTITY: \u001b[32mTax evasion\u001b[39m\t\nSCORES: global= 0.241:0.241[0]; local()= 0.149:0.149[0]; log p(e|m)= -2.120:-2.120[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncharges[0.048]; embezzlement[0.046]; accused[0.045]; trial[0.044]; linked[0.043]; possible[0.043]; involved[0.042]; businesses[0.042]; number[0.042]; mayor[0.042]; companies[0.041]; fight[0.041]; based[0.041]; selling[0.041]; control[0.040]; contract[0.040]; mobsters[0.040]; arrested[0.040]; separate[0.040]; awaiting[0.040]; Mayor[0.040]; launched[0.040]; custody[0.040]; nationwide[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = coal-mining ==> ENTITY: \u001b[32mCoal mining\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.142:0.142[0]; log p(e|m)= -0.041:-0.041[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncoal[0.045]; businesses[0.044]; region[0.044]; companies[0.043]; town[0.043]; town[0.043]; building[0.043]; business[0.042]; regions[0.042]; largest[0.042]; company[0.041]; number[0.041]; oil[0.041]; oil[0.041]; Kemerovo[0.041]; Kemerovo[0.041]; local[0.041]; city[0.041]; locally[0.041]; profitable[0.040]; contract[0.040]; Company[0.040]; possible[0.040]; producers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = mogul ==> ENTITY: \u001b[32mBusiness magnate\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.096:0.096[0]; log p(e|m)= -0.195:-0.195[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbusinessman[0.045]; business[0.045]; region[0.044]; Businessman[0.043]; controversial[0.042]; small[0.042]; small[0.042]; news[0.042]; news[0.042]; Russian[0.041]; Russian[0.041]; building[0.041]; companies[0.041]; central[0.041]; largest[0.041]; hundreds[0.041]; businesses[0.041]; based[0.041]; according[0.041]; Wednesday[0.041]; arrested[0.041]; mining[0.041]; launched[0.040]; regions[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gennady Konyakhin\u001b[39m ==> ENTITY: \u001b[32mGennady Konyakhin\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nembezzlement[0.046]; arrested[0.046]; involved[0.046]; mayor[0.046]; accused[0.046]; Mayor[0.044]; fraud[0.044]; charges[0.044]; Kemerovo[0.044]; Kuznetsky[0.044]; city[0.044]; custody[0.044]; trial[0.043]; Russia[0.043]; Russia[0.043]; Sibneft[0.043]; solved[0.042]; linked[0.042]; mining[0.042]; Officials[0.042]; town[0.041]; launched[0.041]; hundreds[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sibneft\u001b[39m ==> ENTITY: \u001b[32mGazprom Neft\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.044]; Russia[0.044]; Russian[0.044]; Russian[0.044]; Sibneft[0.043]; Sibneft[0.043]; Berezovsky[0.043]; oil[0.042]; oil[0.042]; company[0.042]; Kemerovo[0.042]; Kemerovo[0.042]; Siberia[0.041]; Siberian[0.041]; companies[0.041]; business[0.041]; Boris[0.041]; controlled[0.041]; launched[0.040]; Company[0.040]; Zinin[0.040]; Zinin[0.040]; Zinin[0.040]; largest[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = contract murder\u001b[39m ==> ENTITY: \u001b[32mContract killing\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.081:0.081[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmobsters[0.046]; linked[0.045]; killer[0.044]; mob[0.044]; Kemerovo[0.043]; killings[0.043]; involved[0.043]; murder[0.042]; fight[0.042]; accused[0.042]; charges[0.042]; Kuznetsky[0.041]; Zinin[0.041]; Zinin[0.041]; number[0.041]; business[0.040]; trial[0.040]; nationwide[0.040]; small[0.040]; small[0.040]; possible[0.040]; saying[0.040]; Russian[0.040]; separate[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Siberia ==> ENTITY: \u001b[32mSiberia\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.053:-0.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nregion[0.045]; Russian[0.045]; Russian[0.045]; Russia[0.044]; Russia[0.044]; Siberian[0.044]; regions[0.044]; Yuri[0.042]; separate[0.041]; town[0.041]; town[0.041]; killed[0.041]; killed[0.041]; Boris[0.041]; oil[0.040]; oil[0.040]; central[0.040]; Wednesday[0.040]; work[0.040]; mining[0.040]; fled[0.040]; killing[0.040]; largest[0.040]; controlled[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = trial ==> ENTITY: \u001b[32mTrial\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.107:0.107[0]; log p(e|m)= -0.327:-0.327[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncharges[0.047]; murder[0.044]; murder[0.044]; accused[0.043]; awaiting[0.043]; town[0.043]; custody[0.043]; linked[0.042]; nationwide[0.042]; arrested[0.042]; killings[0.041]; fight[0.041]; products[0.041]; fled[0.041]; launched[0.041]; possible[0.041]; mob[0.041]; city[0.040]; number[0.040]; Police[0.040]; fraud[0.040]; small[0.040]; small[0.040]; control[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mmobsters\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAmerican Mafia\u001b[39m <---> \u001b[32mMobsters\u001b[39m\t\nSCORES: global= 0.260:0.237[\u001b[31m0.023\u001b[39m]; local()= 0.073:0.064[\u001b[31m0.009\u001b[39m]; log p(e|m)= 0.000:-0.868[\u001b[31m0.868\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmob[0.047]; linked[0.045]; killer[0.043]; Kemerovo[0.042]; involved[0.042]; charges[0.042]; agencies[0.042]; local[0.042]; fight[0.042]; accused[0.042]; controversial[0.041]; business[0.041]; trial[0.041]; producers[0.041]; killings[0.041]; profitable[0.041]; small[0.041]; small[0.041]; mogul[0.041]; Kuznetsky[0.041]; fraud[0.040]; saying[0.040]; Zinin[0.040]; Zinin[0.040]; \t\n\r [==============>...............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 246/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19980614_0031.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = auction ==> ENTITY: \u001b[32mAuction\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.168:0.168[0]; log p(e|m)= -0.039:-0.039[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nauction[0.050]; items[0.046]; companies[0.046]; million[0.045]; firms[0.044]; firms[0.044]; related[0.044]; failed[0.044]; art[0.044]; art[0.044]; art[0.044]; required[0.043]; financial[0.043]; offices[0.043]; initial[0.043]; government[0.043]; bidders[0.043]; liquidate[0.042]; set[0.042]; assets[0.042]; assets[0.042]; reported[0.042]; Authority[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = baht\u001b[39m ==> ENTITY: \u001b[32mThai baht\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThailand[0.050]; Thailand[0.050]; Thai[0.048]; Thai[0.048]; million[0.045]; day[0.045]; pay[0.044]; billion[0.043]; financial[0.043]; debts[0.043]; economy[0.042]; set[0.042]; expected[0.042]; expected[0.042]; government[0.042]; prices[0.041]; finance[0.041]; finance[0.041]; program[0.041]; saying[0.041]; reported[0.041]; Financial[0.041]; offices[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bangkok Post\u001b[39m ==> ENTITY: \u001b[32mBangkok Post\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.105:0.105[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThailand[0.049]; Bangkok[0.048]; Thai[0.047]; Thawee[0.042]; baht[0.042]; baht[0.042]; British[0.042]; saying[0.042]; government[0.042]; quoted[0.041]; Sirikit[0.041]; program[0.041]; finance[0.041]; Center[0.040]; closings[0.040]; Chuan[0.040]; Minister[0.040]; International[0.040]; related[0.040]; million[0.040]; Prime[0.040]; bid[0.040]; depressed[0.040]; Monetary[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Thai economy\u001b[39m ==> ENTITY: \u001b[32mEconomy of Thailand\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.216:0.216[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThailand[0.046]; financial[0.045]; government[0.045]; finance[0.044]; billion[0.043]; Thai[0.042]; prices[0.042]; related[0.042]; expected[0.042]; expected[0.042]; companies[0.041]; companies[0.041]; Bangkok[0.041]; million[0.041]; million[0.041]; higher[0.041]; debts[0.041]; firms[0.040]; baht[0.040]; baht[0.040]; baht[0.040]; International[0.040]; British[0.040]; Financial[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chuan Leekpai\u001b[39m ==> ENTITY: \u001b[32mChuan Leekpai\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThai[0.045]; Thawee[0.045]; Sirikit[0.044]; economy[0.044]; Bangkok[0.044]; Bangkok[0.044]; Minister[0.043]; Prime[0.043]; baht[0.043]; baht[0.043]; debts[0.042]; bid[0.041]; program[0.041]; held[0.041]; Center[0.041]; house[0.041]; depressed[0.041]; higher[0.040]; high[0.040]; Monetary[0.040]; late[0.040]; pay[0.039]; expected[0.039]; saying[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = financial firms\u001b[39m ==> ENTITY: \u001b[32mFinancial institution\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFinancial[0.049]; firms[0.046]; companies[0.046]; offices[0.045]; required[0.045]; assets[0.045]; assets[0.045]; finance[0.045]; finance[0.045]; International[0.044]; related[0.043]; agency[0.043]; government[0.043]; economy[0.043]; Authority[0.042]; program[0.042]; million[0.042]; Thailand[0.042]; Thailand[0.042]; billion[0.041]; ended[0.041]; Monetary[0.041]; Fund[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = International Monetary Fund\u001b[39m ==> ENTITY: \u001b[32mInternational Monetary Fund\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfinancial[0.047]; finance[0.046]; Financial[0.045]; billion[0.045]; prices[0.045]; economy[0.044]; government[0.044]; million[0.044]; million[0.044]; agency[0.044]; program[0.043]; assets[0.043]; assets[0.043]; bid[0.043]; expected[0.042]; expected[0.042]; firms[0.042]; Authority[0.042]; higher[0.042]; debts[0.042]; sold[0.042]; companies[0.042]; companies[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = liquidate ==> ENTITY: \u001b[32mLiquidation\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.210:0.210[0]; log p(e|m)= -0.029:-0.029[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfinancial[0.046]; companies[0.043]; companies[0.043]; assets[0.043]; assets[0.043]; firms[0.043]; firms[0.043]; debts[0.043]; Financial[0.042]; International[0.041]; sold[0.041]; related[0.041]; economy[0.041]; Fund[0.041]; required[0.041]; auction[0.041]; auction[0.041]; auction[0.041]; Authority[0.041]; expected[0.040]; expected[0.040]; billion[0.040]; prices[0.040]; pay[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Christies\u001b[39m ==> ENTITY: \u001b[32mChristie's\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nauction[0.046]; auction[0.046]; auction[0.046]; art[0.044]; sold[0.044]; painting[0.044]; gallery[0.043]; house[0.043]; million[0.043]; British[0.043]; bid[0.041]; untitled[0.041]; proceeds[0.041]; donated[0.041]; late[0.040]; failed[0.040]; Queen[0.040]; Sirikit[0.040]; sketches[0.040]; program[0.039]; pay[0.039]; Post[0.039]; Center[0.039]; held[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Thailand ==> ENTITY: \u001b[32mThailand\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.107:0.107[0]; log p(e|m)= -0.241:-0.241[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThailand[0.055]; Thai[0.048]; ended[0.046]; weekend[0.042]; financial[0.042]; baht[0.041]; day[0.041]; set[0.041]; government[0.041]; International[0.041]; Sunday[0.041]; Sunday[0.041]; firms[0.041]; firms[0.041]; snapped[0.040]; media[0.040]; collapsed[0.040]; finance[0.040]; finance[0.040]; companies[0.040]; Financial[0.040]; Restructuring[0.040]; related[0.040]; Fund[0.039]; \t\n\r [===============>..............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 257/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19980611_0774.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Budapest ==> ENTITY: \u001b[32mBudapest\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.104:0.104[0]; log p(e|m)= -0.030:-0.030[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncombat[0.045]; Hungarian[0.045]; April[0.044]; East[0.043]; Thursday[0.043]; met[0.043]; European[0.042]; team[0.042]; France[0.042]; Held[0.041]; International[0.041]; International[0.041]; closed[0.041]; meeting[0.041]; Tamas[0.041]; traffic[0.041]; traffic[0.041]; traffic[0.041]; official[0.041]; said[0.041]; said[0.041]; police[0.041]; police[0.041]; police[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mHungarian\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHungary\u001b[39m <---> \u001b[32mHungarians\u001b[39m\t\nSCORES: global= 0.253:0.237[\u001b[31m0.016\u001b[39m]; local()= 0.127:0.104[\u001b[31m0.024\u001b[39m]; log p(e|m)= -1.033:-1.570[\u001b[31m0.537\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHungary[0.048]; Hungary[0.048]; king[0.045]; European[0.043]; national[0.043]; countries[0.042]; France[0.042]; terms[0.042]; Budapest[0.041]; communism[0.041]; total[0.041]; million[0.041]; million[0.041]; million[0.041]; personal[0.041]; said[0.040]; said[0.040]; said[0.040]; cases[0.040]; cases[0.040]; forints[0.040]; forints[0.040]; involving[0.040]; Organization[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = auctioneers\u001b[39m ==> ENTITY: \u001b[32mAuction\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.116:0.116[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmillion[0.043]; million[0.043]; worth[0.043]; art[0.043]; art[0.043]; art[0.043]; art[0.043]; market[0.043]; network[0.042]; national[0.042]; involving[0.042]; special[0.041]; museums[0.041]; work[0.041]; country[0.041]; country[0.041]; data[0.041]; rp[0.041]; dealers[0.041]; stolen[0.041]; stolen[0.041]; stolen[0.041]; forints[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = communism ==> ENTITY: \u001b[32mCommunism\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.079:-0.079[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.047]; Hungary[0.044]; Hungary[0.044]; European[0.044]; national[0.043]; group[0.043]; terms[0.043]; largely[0.042]; Hungarian[0.042]; present[0.042]; work[0.042]; April[0.042]; France[0.041]; Organization[0.041]; official[0.041]; member[0.040]; personal[0.040]; total[0.040]; help[0.040]; working[0.040]; market[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = East European\u001b[39m ==> ENTITY: \u001b[32mEastern Europe\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.113:0.113[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCol[0.047]; countries[0.046]; Hungary[0.045]; communism[0.044]; Hungarian[0.043]; Organization[0.042]; make[0.042]; Tamas[0.042]; counterparts[0.041]; France[0.041]; experience[0.041]; terms[0.041]; group[0.041]; total[0.041]; said[0.041]; said[0.041]; day[0.040]; cases[0.040]; cases[0.040]; million[0.040]; million[0.040]; million[0.040]; Budapest[0.040]; set[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = forints\u001b[39m ==> ENTITY: \u001b[32mHungarian forint\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCol[0.045]; forints[0.045]; Hungarian[0.044]; worth[0.044]; Hungary[0.044]; Hungary[0.044]; million[0.043]; million[0.043]; million[0.043]; value[0.042]; total[0.041]; stolen[0.041]; stolen[0.041]; stolen[0.041]; stolen[0.041]; country[0.041]; country[0.041]; market[0.040]; Tamas[0.040]; national[0.040]; terms[0.040]; involving[0.039]; growth[0.039]; European[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Interpol ==> ENTITY: \u001b[32mInterpol\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.419:-0.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInternational[0.044]; International[0.044]; official[0.043]; working[0.043]; said[0.043]; said[0.043]; group[0.042]; seek[0.042]; seek[0.042]; Organization[0.042]; Simon[0.042]; April[0.042]; Thursday[0.042]; set[0.042]; countries[0.041]; allowed[0.041]; make[0.041]; officers[0.041]; officers[0.041]; European[0.041]; member[0.041]; ways[0.040]; personal[0.040]; king[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = data base\u001b[39m ==> ENTITY: \u001b[32mDatabase\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.057:0.057[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncataloging[0.047]; network[0.045]; involving[0.044]; terms[0.044]; national[0.042]; rp[0.042]; value[0.042]; said[0.041]; work[0.041]; museums[0.041]; forints[0.041]; forints[0.041]; million[0.041]; million[0.041]; million[0.041]; contacts[0.041]; market[0.041]; largely[0.041]; director[0.041]; traffic[0.041]; establishing[0.040]; worth[0.040]; total[0.040]; cases[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.096:0.096[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.046]; countries[0.045]; European[0.045]; king[0.044]; Hungary[0.043]; communism[0.042]; combat[0.042]; total[0.042]; involving[0.042]; group[0.042]; said[0.041]; said[0.041]; said[0.041]; terms[0.041]; special[0.041]; million[0.041]; million[0.041]; million[0.041]; Organization[0.041]; East[0.040]; day[0.040]; personal[0.040]; official[0.040]; International[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = illegal traffic\u001b[39m ==> ENTITY: \u001b[32mSmuggling\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.109:0.109[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCol[0.049]; illegal[0.048]; illegal[0.048]; Illegal[0.044]; help[0.044]; Tamas[0.044]; countries[0.044]; industry[0.043]; working[0.043]; personal[0.043]; allowed[0.043]; combat[0.043]; officers[0.042]; officers[0.042]; traffic[0.042]; traffic[0.042]; traffic[0.042]; communism[0.042]; East[0.042]; European[0.042]; police[0.042]; police[0.042]; police[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = International Police\u001b[39m ==> ENTITY: \u001b[32mUnited Nations Police\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.126:0.126[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInterpol[0.048]; Col[0.046]; combat[0.044]; Police[0.044]; officers[0.043]; officers[0.043]; police[0.042]; police[0.042]; police[0.042]; seek[0.042]; seek[0.042]; Organization[0.042]; official[0.041]; Tamas[0.041]; countries[0.041]; International[0.040]; European[0.040]; member[0.040]; April[0.040]; team[0.039]; communism[0.039]; meeting[0.039]; help[0.039]; working[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = art dealers\u001b[39m ==> ENTITY: \u001b[32mArt dealer\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmuseums[0.047]; art[0.046]; art[0.046]; art[0.046]; work[0.043]; worth[0.042]; auctioneers[0.042]; Hungary[0.041]; million[0.041]; million[0.041]; million[0.041]; French[0.041]; stolen[0.041]; stolen[0.041]; stolen[0.041]; cataloging[0.041]; rp[0.040]; appear[0.040]; involving[0.040]; Jean[0.040]; contacts[0.040]; forints[0.040]; forints[0.040]; market[0.040]; \t\n\r [================>.............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 271/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19980709_0263.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Prime Minister ==> ENTITY: \u001b[32mPrime minister\u001b[39m\t\nSCORES: global= 0.239:0.239[0]; local()= 0.094:0.094[0]; log p(e|m)= -1.435:-1.435[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nminister[0.045]; said[0.043]; Van[0.043]; public[0.043]; Vietnam[0.043]; Vietnam[0.043]; Khai[0.043]; Khai[0.043]; Phan[0.042]; Thursday[0.042]; leaders[0.041]; accusing[0.041]; instructions[0.041]; ordered[0.041]; demands[0.041]; criticisms[0.041]; self[0.041]; prime[0.041]; bureaucracy[0.041]; particularly[0.041]; measures[0.041]; based[0.040]; state[0.040]; state[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mhydroelectric\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHydropower\u001b[39m <---> \u001b[32mHydroelectricity\u001b[39m\t\nSCORES: global= 0.268:0.263[\u001b[31m0.005\u001b[39m]; local()= 0.188:0.189[\u001b[32m0.001\u001b[39m]; log p(e|m)= 0.000:-0.021[\u001b[31m0.021\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwater[0.046]; electricity[0.045]; reservoir[0.044]; plant[0.044]; plants[0.044]; power[0.043]; power[0.043]; country[0.043]; construction[0.043]; thermal[0.042]; levels[0.041]; levels[0.041]; output[0.041]; chronically[0.041]; drought[0.041]; electric[0.041]; projects[0.041]; projects[0.041]; company[0.040]; largest[0.039]; critical[0.039]; century[0.039]; costs[0.039]; feed[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = project accounting\u001b[39m ==> ENTITY: \u001b[32mProject accounting\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncosts[0.046]; costs[0.046]; projects[0.045]; projects[0.045]; measures[0.045]; Investment[0.044]; Army[0.044]; individual[0.044]; number[0.044]; work[0.044]; work[0.044]; process[0.043]; process[0.043]; based[0.043]; assessment[0.043]; Dau[0.043]; construction[0.043]; media[0.042]; corrective[0.042]; company[0.042]; output[0.042]; task[0.042]; rotating[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vietnam ==> ENTITY: \u001b[32mVietnam\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.390:-0.390[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVietnam[0.051]; Vietnam[0.051]; Army[0.048]; country[0.046]; foreign[0.043]; Khai[0.041]; Khai[0.041]; Phan[0.041]; reported[0.041]; said[0.040]; leaders[0.040]; demands[0.040]; criticized[0.040]; Anh[0.040]; accusing[0.040]; measures[0.040]; Le[0.040]; particularly[0.040]; Thursday[0.040]; President[0.039]; Duc[0.039]; March[0.039]; help[0.039]; individual[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Le Duc Anh\u001b[39m ==> ENTITY: \u001b[32mLê Đức Anh\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.045]; Vietnam[0.044]; Vietnam[0.044]; Vietnam[0.044]; Khai[0.044]; Army[0.043]; leaders[0.043]; Dau[0.042]; country[0.042]; imposed[0.041]; March[0.041]; accusing[0.041]; Tu[0.041]; rotating[0.041]; foreign[0.041]; said[0.040]; criticized[0.040]; reported[0.040]; reported[0.040]; problems[0.040]; EVN[0.040]; EVN[0.040]; led[0.040]; Thursday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = monopoly ==> ENTITY: \u001b[32mMonopoly\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.135:-0.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmonopoly[0.047]; costs[0.045]; public[0.044]; overcharging[0.043]; newspaper[0.043]; measures[0.043]; excessive[0.042]; state[0.042]; state[0.042]; individual[0.042]; problems[0.041]; power[0.041]; accusing[0.041]; electricity[0.041]; abuse[0.041]; work[0.041]; utility[0.041]; demands[0.041]; EVN[0.041]; self[0.040]; bureaucracy[0.040]; criticisms[0.040]; spark[0.040]; ordered[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = electricity ==> ENTITY: \u001b[32mElectricity\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.109:-0.109[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nutility[0.049]; power[0.045]; state[0.044]; state[0.044]; public[0.044]; Electricity[0.044]; EVN[0.044]; waste[0.043]; costs[0.042]; work[0.042]; process[0.041]; monopoly[0.041]; monopoly[0.041]; help[0.040]; particularly[0.040]; project[0.040]; prime[0.040]; excessive[0.040]; overcharging[0.040]; driving[0.040]; individual[0.040]; abuse[0.039]; improvements[0.039]; measures[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = self-criticisms\u001b[39m ==> ENTITY: \u001b[32mSelf-criticism\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.122:0.122[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncriticized[0.045]; Vietnam[0.043]; Vietnam[0.043]; Vietnam[0.043]; Vietnam[0.043]; Army[0.043]; ordered[0.042]; Dau[0.042]; leaders[0.042]; Khai[0.042]; Khai[0.042]; corrective[0.042]; excessive[0.042]; individual[0.041]; affect[0.041]; accusing[0.041]; abuse[0.041]; problems[0.041]; chronically[0.040]; complaints[0.040]; public[0.040]; media[0.040]; country[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Phan Van Khai\u001b[39m ==> ENTITY: \u001b[32mPhan Văn Khải\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.109:0.109[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKhai[0.047]; Vietnam[0.046]; Vietnam[0.046]; minister[0.046]; prime[0.045]; Prime[0.045]; Minister[0.044]; Dau[0.043]; problems[0.041]; EVN[0.041]; accusing[0.041]; said[0.040]; Tu[0.040]; Electricity[0.040]; Thursday[0.040]; corrective[0.040]; tender[0.040]; leaders[0.040]; discontent[0.040]; cut[0.039]; self[0.039]; accounting[0.039]; Investment[0.039]; overcharging[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = drought ==> ENTITY: \u001b[32mDrought\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.049:-0.049[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwater[0.049]; country[0.045]; reservoir[0.044]; levels[0.043]; levels[0.043]; century[0.042]; earlier[0.042]; imposed[0.042]; plants[0.042]; output[0.042]; costs[0.041]; electricity[0.041]; worst[0.041]; short[0.041]; plant[0.041]; chronically[0.041]; construction[0.041]; cuts[0.041]; thermal[0.040]; work[0.040]; bogged[0.040]; number[0.040]; hydroelectric[0.040]; foreign[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = bureaucracy ==> ENTITY: \u001b[32mBureaucracy\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.052:-0.052[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nprocess[0.045]; costs[0.044]; public[0.043]; state[0.043]; state[0.043]; foreign[0.043]; measures[0.042]; work[0.042]; reputation[0.042]; monopoly[0.041]; monopoly[0.041]; said[0.041]; President[0.041]; particularly[0.041]; task[0.041]; leaders[0.041]; criticisms[0.041]; accusing[0.041]; based[0.041]; demands[0.041]; minister[0.040]; help[0.040]; Prime[0.040]; waste[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = reservoir ==> ENTITY: \u001b[32mReservoir\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.078:-0.078[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwater[0.048]; hydroelectric[0.046]; plant[0.044]; electricity[0.044]; drought[0.044]; levels[0.043]; levels[0.043]; thermal[0.043]; construction[0.042]; plants[0.042]; projects[0.041]; projects[0.041]; power[0.041]; power[0.041]; company[0.040]; EVN[0.040]; country[0.040]; key[0.040]; output[0.040]; century[0.040]; rotating[0.039]; electric[0.039]; raise[0.039]; largest[0.039]; \t\n\r [=================>............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 285/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19980625_1136.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Boeing\u001b[39m ==> ENTITY: \u001b[32mBoeing\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \naircraft[0.048]; jet[0.045]; Air[0.044]; plane[0.044]; plane[0.044]; takeoff[0.044]; international[0.043]; international[0.043]; airport[0.043]; airport[0.043]; airport[0.043]; airliner[0.042]; Plane[0.042]; month[0.041]; passenger[0.040]; baggage[0.040]; tower[0.039]; Thursday[0.039]; luggage[0.039]; largest[0.039]; Tuesday[0.039]; large[0.039]; said[0.039]; clear[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lagos ==> ENTITY: \u001b[32mLagos\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLagos[0.050]; Nigerian[0.047]; Nigeria[0.046]; officials[0.044]; month[0.043]; international[0.042]; international[0.042]; personnel[0.042]; past[0.041]; escaped[0.041]; airport[0.041]; airport[0.041]; airport[0.041]; airport[0.041]; reported[0.041]; went[0.040]; Thursday[0.040]; jet[0.040]; airliner[0.040]; tower[0.040]; compartment[0.040]; robbed[0.040]; sent[0.040]; large[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = violent crime ==> ENTITY: \u001b[32mViolent crime\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.103:0.103[0]; log p(e|m)= -0.075:-0.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.046]; robberies[0.045]; rate[0.045]; high[0.044]; involving[0.043]; notoriously[0.042]; routine[0.042]; vehicle[0.042]; populous[0.042]; military[0.042]; security[0.042]; large[0.042]; gam[0.041]; suspected[0.041]; maintenance[0.041]; said[0.041]; shoot[0.041]; despite[0.040]; armed[0.040]; breakdown[0.040]; officials[0.040]; law[0.040]; million[0.040]; incident[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cameroon Airlines\u001b[39m ==> ENTITY: \u001b[32mCameroon Airlines\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmaintenance[0.044]; jet[0.043]; airliner[0.043]; involving[0.043]; Africa[0.043]; order[0.043]; high[0.042]; million[0.042]; plane[0.042]; plane[0.042]; plane[0.042]; plane[0.042]; orders[0.042]; landing[0.041]; airport[0.041]; airport[0.041]; officials[0.041]; officials[0.041]; country[0.041]; security[0.041]; security[0.041]; military[0.040]; tarmac[0.040]; Nigeria[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = tarmac ==> ENTITY: \u001b[32mTarmac\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.024:-0.024[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmaintenance[0.046]; airport[0.046]; airport[0.046]; jet[0.043]; landing[0.043]; plane[0.042]; plane[0.042]; plane[0.042]; plane[0.042]; vehicle[0.042]; country[0.042]; Airlines[0.041]; security[0.041]; security[0.041]; passenger[0.041]; high[0.041]; personnel[0.041]; gear[0.040]; large[0.040]; million[0.040]; officials[0.040]; officials[0.040]; waiting[0.040]; military[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = distress call\u001b[39m ==> ENTITY: \u001b[32mDistress signal\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsent[0.050]; landing[0.050]; plane[0.048]; plane[0.048]; plane[0.048]; plane[0.048]; plane[0.048]; involving[0.048]; takeoff[0.048]; international[0.048]; international[0.048]; aircraft[0.047]; said[0.047]; said[0.047]; said[0.047]; said[0.047]; tower[0.047]; personnel[0.047]; permission[0.046]; reported[0.046]; Tuesday[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nigeria ==> ENTITY: \u001b[32mNigeria\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.315:-0.315[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.049]; country[0.049]; Cameroon[0.047]; Africa[0.045]; security[0.042]; security[0.042]; week[0.042]; came[0.042]; involving[0.041]; orders[0.041]; armed[0.041]; populous[0.041]; order[0.040]; maintenance[0.040]; despite[0.040]; high[0.040]; said[0.040]; said[0.040]; said[0.040]; worth[0.040]; officials[0.040]; gear[0.040]; suspected[0.039]; duty[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Air Afrique\u001b[39m ==> ENTITY: \u001b[32mAir Afrique\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBoeing[0.044]; largest[0.044]; aircraft[0.044]; international[0.044]; international[0.044]; plane[0.043]; plane[0.043]; jet[0.043]; airliner[0.043]; month[0.042]; airport[0.042]; airport[0.042]; airport[0.042]; takeoff[0.041]; Plane[0.041]; Thursday[0.041]; Tuesday[0.040]; said[0.040]; reported[0.039]; tower[0.039]; compartment[0.039]; luggage[0.039]; passenger[0.039]; large[0.039]; \t\n\r [=================>............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 296/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19990519_0141.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = cyclone ==> ENTITY: \u001b[32mCyclone\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.154:0.154[0]; log p(e|m)= -0.360:-0.360[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstorm[0.047]; storm[0.047]; hurricanes[0.046]; region[0.045]; clouds[0.044]; according[0.043]; Storm[0.043]; month[0.043]; northern[0.043]; cloud[0.041]; April[0.041]; miles[0.041]; Texas[0.041]; water[0.041]; size[0.041]; today[0.040]; center[0.040]; ice[0.040]; immense[0.039]; mile[0.039]; space[0.039]; counterclockwise[0.039]; polar[0.039]; spiraled[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMartian\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMars\u001b[39m <---> \u001b[32mMartian\u001b[39m\t\nSCORES: global= 0.261:0.236[\u001b[31m0.024\u001b[39m]; local()= 0.189:0.140[\u001b[31m0.048\u001b[39m]; log p(e|m)= -1.273:-0.559[\u001b[32m0.714\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMars[0.045]; Mars[0.045]; Mars[0.045]; Space[0.044]; Space[0.044]; space[0.044]; Earth[0.043]; Martian[0.042]; planet[0.042]; planet[0.042]; hemisphere[0.042]; polar[0.041]; Science[0.041]; Hubble[0.040]; telescope[0.040]; region[0.040]; ice[0.040]; ice[0.040]; wide[0.040]; wide[0.040]; resembled[0.040]; northern[0.040]; northern[0.040]; larger[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mice sheet\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mGreenland ice sheet\u001b[39m <---> \u001b[32mIce sheet\u001b[39m\t\nSCORES: global= 0.256:0.249[\u001b[31m0.007\u001b[39m]; local()= 0.186:0.203[\u001b[32m0.018\u001b[39m]; log p(e|m)= 0.000:-0.166[\u001b[31m0.166\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nice[0.060]; water[0.055]; summer[0.054]; center[0.054]; north[0.053]; miles[0.053]; nearly[0.053]; warming[0.053]; process[0.052]; clouds[0.052]; northern[0.052]; wide[0.052]; wide[0.052]; wide[0.052]; eastward[0.051]; mile[0.051]; Earth[0.051]; Earth[0.051]; hemisphere[0.051]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = ice ==> ENTITY: \u001b[32mIce\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.178:0.178[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nice[0.062]; polar[0.054]; region[0.052]; water[0.052]; clouds[0.052]; miles[0.050]; north[0.049]; according[0.049]; red[0.049]; northern[0.048]; northern[0.048]; wide[0.048]; wide[0.048]; summer[0.048]; size[0.048]; Earth[0.048]; mile[0.048]; long[0.048]; shape[0.048]; caps[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Space Telescope Science Institute\u001b[39m ==> ENTITY: \u001b[32mSpace Telescope Science Institute\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.122:0.122[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHubble[0.044]; Telescope[0.044]; telescope[0.044]; Space[0.044]; April[0.043]; today[0.043]; space[0.042]; according[0.042]; astronomers[0.042]; announced[0.042]; dioxide[0.041]; center[0.041]; polar[0.041]; Bands[0.041]; Mars[0.041]; Mars[0.041]; Mars[0.041]; wide[0.041]; wide[0.041]; month[0.041]; said[0.041]; Earth[0.040]; Astronomers[0.039]; times[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Martian north pole\u001b[39m ==> ENTITY: \u001b[32mPlanum Boreum\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMars[0.044]; Mars[0.044]; ice[0.043]; ice[0.043]; storm[0.042]; storm[0.042]; storm[0.042]; storm[0.042]; storm[0.042]; storm[0.042]; storm[0.042]; storm[0.042]; nearly[0.042]; eastward[0.042]; water[0.041]; Martian[0.041]; clouds[0.041]; Viking[0.041]; miles[0.041]; summer[0.041]; caps[0.040]; northern[0.040]; planetary[0.040]; mile[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = hurricanes\u001b[39m ==> ENTITY: \u001b[32mTropical cyclone\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstorm[0.047]; storm[0.047]; storm[0.047]; storm[0.047]; storm[0.047]; nearly[0.042]; region[0.041]; according[0.041]; clouds[0.041]; water[0.041]; month[0.041]; miles[0.041]; mile[0.040]; said[0.040]; said[0.040]; April[0.040]; April[0.040]; cyclone[0.040]; seasonal[0.040]; summer[0.040]; north[0.039]; occurred[0.039]; northern[0.039]; northern[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Astronomers ==> ENTITY: \u001b[32mAstronomer\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.084:-0.084[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScience[0.045]; April[0.044]; telescope[0.043]; Space[0.043]; Space[0.043]; space[0.043]; month[0.043]; Mars[0.042]; Mars[0.042]; according[0.042]; Institute[0.041]; Telescope[0.041]; Telescope[0.041]; polar[0.041]; today[0.041]; Bands[0.041]; wide[0.041]; wide[0.041]; size[0.040]; Hubble[0.040]; planet[0.040]; planet[0.040]; Texas[0.040]; times[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mars ==> ENTITY: \u001b[32mMars\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.189:0.189[0]; log p(e|m)= -0.393:-0.393[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMars[0.046]; Mars[0.046]; space[0.045]; Earth[0.044]; Martian[0.044]; planet[0.043]; planet[0.043]; polar[0.042]; northern[0.042]; northern[0.042]; announced[0.041]; Space[0.041]; Space[0.041]; month[0.041]; according[0.041]; size[0.041]; telescope[0.040]; April[0.040]; center[0.040]; Hubble[0.040]; shape[0.040]; clouds[0.039]; region[0.039]; hemisphere[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Viking Orbiter\u001b[39m ==> ENTITY: \u001b[32mViking program\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nspacecraft[0.047]; Mars[0.047]; Mars[0.047]; Martian[0.045]; Earth[0.044]; Earth[0.044]; Hubble[0.042]; Hubble[0.042]; nearly[0.042]; planet[0.041]; planetary[0.041]; clouds[0.040]; ice[0.040]; ice[0.040]; field[0.040]; dioxide[0.040]; long[0.040]; Bands[0.040]; astronomers[0.040]; astronomers[0.040]; showed[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Earth ==> ENTITY: \u001b[32mEarth\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.177:0.177[0]; log p(e|m)= -0.078:-0.078[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEarth[0.045]; water[0.043]; planet[0.043]; planet[0.043]; planet[0.043]; space[0.042]; Mars[0.042]; Mars[0.042]; Mars[0.042]; ice[0.042]; ice[0.042]; planetary[0.042]; times[0.042]; times[0.042]; closest[0.042]; Martian[0.041]; Martian[0.041]; nearly[0.041]; larger[0.041]; clouds[0.040]; size[0.040]; cloud[0.040]; taking[0.040]; polar[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hubble Space Telescope\u001b[39m ==> ENTITY: \u001b[32mHubble Space Telescope\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntelescope[0.047]; Telescope[0.046]; Mars[0.044]; Mars[0.044]; Mars[0.044]; astronomers[0.043]; space[0.043]; Space[0.043]; size[0.042]; Astronomers[0.042]; Earth[0.041]; month[0.041]; today[0.041]; April[0.041]; dioxide[0.040]; polar[0.040]; Bands[0.040]; planet[0.040]; planet[0.040]; according[0.040]; clouds[0.040]; taking[0.040]; cloud[0.039]; Science[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = phenomenon ==> ENTITY: \u001b[32mPhenomenon\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.100:0.100[0]; log p(e|m)= -0.145:-0.145[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstorm[0.044]; storm[0.044]; storm[0.044]; storm[0.044]; storm[0.044]; planetary[0.044]; picture[0.043]; appeared[0.042]; Earth[0.042]; planet[0.042]; camera[0.041]; camera[0.041]; pictures[0.041]; pictures[0.041]; photograph[0.041]; wide[0.041]; second[0.040]; short[0.040]; hours[0.040]; approach[0.040]; said[0.040]; said[0.040]; showed[0.040]; April[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Texas ==> ENTITY: \u001b[32mTexas\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.077:0.077[0]; log p(e|m)= -0.161:-0.161[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmile[0.045]; miles[0.044]; size[0.044]; region[0.044]; Science[0.044]; month[0.043]; northern[0.043]; water[0.042]; according[0.042]; Institute[0.042]; center[0.042]; times[0.041]; Space[0.041]; Space[0.041]; announced[0.041]; Bands[0.041]; today[0.041]; AP[0.041]; April[0.040]; long[0.040]; Telescope[0.040]; Telescope[0.040]; WASHINGTON[0.040]; red[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = evaporated ==> ENTITY: \u001b[32mEvaporation\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.037:-0.037[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwater[0.052]; ice[0.045]; ice[0.045]; clouds[0.042]; seasonal[0.042]; nearly[0.042]; slightly[0.042]; dioxide[0.042]; Viking[0.042]; picture[0.041]; long[0.041]; Earth[0.041]; Earth[0.041]; carbon[0.041]; Bands[0.040]; counterclockwise[0.040]; storm[0.040]; storm[0.040]; storm[0.040]; storm[0.040]; storm[0.040]; storm[0.040]; storm[0.040]; cloud[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = space telescope\u001b[39m ==> ENTITY: \u001b[32mSpace observatory\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpace[0.046]; Space[0.046]; Mars[0.044]; Mars[0.044]; Mars[0.044]; Telescope[0.043]; Telescope[0.043]; Hubble[0.043]; astronomers[0.042]; Astronomers[0.042]; Science[0.042]; polar[0.041]; Earth[0.041]; today[0.041]; clouds[0.041]; size[0.040]; miles[0.040]; according[0.040]; cloud[0.040]; Martian[0.039]; April[0.039]; month[0.039]; wide[0.039]; wide[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = carbon dioxide ==> ENTITY: \u001b[32mCarbon dioxide\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.023:-0.023[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nice[0.046]; ice[0.046]; water[0.046]; Earth[0.044]; Earth[0.044]; planet[0.044]; planet[0.044]; clouds[0.042]; cloud[0.042]; warming[0.042]; planetary[0.042]; nearly[0.042]; process[0.040]; hurricanes[0.040]; Mars[0.040]; Mars[0.040]; seasonal[0.040]; long[0.040]; field[0.040]; times[0.039]; Martian[0.039]; Martian[0.039]; caps[0.039]; Officials[0.039]; \t\n\r [==================>...........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 314/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW20000303_0067.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gulf Coast ==> ENTITY: \u001b[32mGulf Coast of the United States\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLouisiana[0.047]; storm[0.045]; Carolinas[0.045]; northward[0.045]; Texas[0.044]; Georgia[0.044]; seven[0.043]; Tennessee[0.043]; Northeast[0.042]; areas[0.042]; Oklahoma[0.042]; International[0.041]; Worth[0.041]; Thursday[0.041]; Valley[0.041]; expected[0.041]; thunderstorms[0.040]; Maine[0.040]; Mountains[0.039]; Fort[0.039]; Dallas[0.039]; Ohio[0.039]; roared[0.039]; Lightning[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = temperature ==> ENTITY: \u001b[32mTemperature\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.036:-0.036[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nconditions[0.044]; expected[0.044]; freezing[0.044]; degrees[0.043]; Atlantic[0.043]; Plains[0.042]; Plains[0.042]; Plains[0.042]; forecast[0.042]; forecast[0.042]; forecast[0.042]; mid[0.042]; winds[0.042]; readings[0.041]; low[0.041]; continental[0.041]; Gusty[0.041]; minus[0.041]; wind[0.040]; Northeast[0.040]; reach[0.040]; reported[0.040]; possible[0.040]; chill[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rockies ==> ENTITY: \u001b[32mRocky Mountains\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.159:0.159[0]; log p(e|m)= -0.367:-0.367[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMountains[0.045]; continental[0.043]; Plains[0.043]; Plains[0.043]; Plains[0.043]; Southwest[0.042]; areas[0.042]; Valley[0.042]; communities[0.042]; Rockies[0.042]; seven[0.042]; Northeast[0.041]; Northeast[0.041]; hitting[0.041]; West[0.041]; States[0.041]; Fort[0.041]; states[0.041]; Southeast[0.041]; Southeast[0.041]; Great[0.041]; Midwest[0.041]; Grand[0.040]; Northwest[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cotulla\u001b[39m ==> ENTITY: \u001b[32mCotulla, Texas\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.064:0.064[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTexas[0.045]; Southwest[0.044]; Minn[0.043]; communities[0.043]; Northwest[0.043]; Southeast[0.042]; Southeast[0.042]; Plains[0.042]; Plains[0.042]; Plains[0.042]; Midwest[0.042]; Lakes[0.042]; Rockies[0.041]; Rockies[0.041]; Thursday[0.041]; Grand[0.041]; West[0.041]; States[0.041]; today[0.041]; today[0.041]; reported[0.041]; Northeast[0.040]; Highs[0.040]; Washington[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Northeast ==> ENTITY: \u001b[32mNortheastern United States\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.182:0.182[0]; log p(e|m)= -0.830:-0.830[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNortheast[0.047]; Carolinas[0.044]; Atlantic[0.044]; Midwest[0.043]; Pennsylvania[0.043]; flights[0.043]; Southeast[0.043]; mid[0.042]; Maine[0.042]; states[0.042]; tonight[0.041]; Ohio[0.041]; areas[0.041]; Northwest[0.041]; Thursday[0.041]; Airport[0.041]; Mountains[0.041]; expected[0.040]; expected[0.040]; seven[0.040]; Plains[0.040]; Plains[0.040]; West[0.040]; Coast[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = hail ==> ENTITY: \u001b[32mHail\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.215:0.215[0]; log p(e|m)= -0.012:-0.012[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nthunderstorms[0.045]; hail[0.045]; storm[0.044]; Weather[0.043]; rain[0.043]; rain[0.043]; inches[0.043]; inches[0.043]; Severe[0.043]; winds[0.042]; freezing[0.042]; wind[0.041]; wind[0.041]; wind[0.041]; Rain[0.040]; areas[0.040]; forecast[0.040]; forecast[0.040]; forecast[0.040]; forecast[0.040]; conditions[0.040]; heavy[0.040]; expected[0.040]; expected[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Great Lakes ==> ENTITY: \u001b[32mGreat Lakes\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.100:-0.100[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncontinental[0.044]; Midwest[0.044]; Northeast[0.044]; Northeast[0.044]; states[0.043]; Maine[0.042]; communities[0.042]; Pennsylvania[0.042]; Southeast[0.042]; Southeast[0.042]; Atlantic[0.042]; States[0.041]; Washington[0.041]; Northwest[0.041]; coasts[0.041]; Plains[0.041]; Plains[0.041]; Plains[0.041]; West[0.041]; Southwest[0.041]; mid[0.041]; Adirondack[0.041]; seven[0.040]; Texas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rain ==> ENTITY: \u001b[32mRain\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.201:0.201[0]; log p(e|m)= -1.492:-1.492[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrain[0.045]; rain[0.045]; Weather[0.044]; storm[0.043]; inches[0.043]; inches[0.043]; areas[0.042]; thunderstorms[0.042]; hail[0.042]; hail[0.042]; forecast[0.041]; forecast[0.041]; forecast[0.041]; heavy[0.041]; wind[0.041]; wind[0.041]; wind[0.041]; expected[0.041]; expected[0.041]; seven[0.041]; Severe[0.040]; Mountains[0.040]; winds[0.040]; northward[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = wind-chill\u001b[39m ==> ENTITY: \u001b[32mWind chill\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndegrees[0.049]; freezing[0.047]; temperature[0.047]; conditions[0.047]; wind[0.046]; wind[0.046]; expected[0.043]; expected[0.043]; mph[0.043]; storm[0.043]; rain[0.042]; rain[0.042]; forecast[0.042]; forecast[0.042]; forecast[0.042]; forecast[0.042]; inches[0.042]; inches[0.042]; heavy[0.042]; winds[0.042]; Northwest[0.042]; Southeast[0.042]; Southeast[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington (state)\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.124:0.124[0]; log p(e|m)= -1.945:-1.945[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMaine[0.048]; Mountains[0.043]; states[0.043]; Northeast[0.043]; Northeast[0.043]; West[0.043]; Adirondack[0.043]; Valley[0.042]; Oregon[0.042]; States[0.042]; communities[0.041]; Lakes[0.041]; Southwest[0.041]; Pennsylvania[0.041]; Southeast[0.040]; Southeast[0.040]; Fort[0.040]; Atlantic[0.040]; continental[0.040]; Plains[0.040]; Plains[0.040]; Plains[0.040]; United[0.040]; England[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Green Mountains\u001b[39m ==> ENTITY: \u001b[32mGreen Mountains\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMaine[0.045]; Northeast[0.043]; Northeast[0.043]; Pennsylvania[0.043]; Adirondack[0.043]; Carolinas[0.043]; Plains[0.042]; Plains[0.042]; Plains[0.042]; Tennessee[0.042]; areas[0.042]; West[0.042]; Valley[0.041]; Ohio[0.041]; Northwest[0.041]; Lakes[0.041]; Southeast[0.041]; Southeast[0.041]; Atlantic[0.041]; Georgia[0.040]; England[0.040]; freezing[0.040]; mid[0.040]; Southwest[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ohio Valley ==> ENTITY: \u001b[32mOhio River\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.968:-0.968[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennessee[0.046]; Oklahoma[0.045]; Northeast[0.044]; Georgia[0.043]; storm[0.043]; Carolinas[0.042]; Thursday[0.042]; Pennsylvania[0.042]; Louisiana[0.042]; tonight[0.042]; Southeast[0.042]; Texas[0.042]; freezing[0.042]; West[0.041]; Washington[0.041]; Coast[0.041]; Plains[0.041]; northward[0.040]; areas[0.040]; rain[0.040]; rain[0.040]; heavy[0.040]; thunderstorms[0.040]; Fort[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Alice ==> ENTITY: \u001b[32mAlice, Texas\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.076:0.076[0]; log p(e|m)= -2.797:-2.797[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWest[0.045]; Texas[0.044]; communities[0.043]; states[0.043]; today[0.042]; today[0.042]; continental[0.042]; Cotulla[0.042]; highest[0.042]; Southwest[0.042]; States[0.042]; Washington[0.042]; Grand[0.041]; Southeast[0.041]; Southeast[0.041]; Plains[0.041]; Plains[0.041]; Plains[0.041]; reported[0.041]; low[0.041]; Northeast[0.041]; Great[0.041]; Northwest[0.040]; Lakes[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = the Carolinas ==> ENTITY: \u001b[32mThe Carolinas\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.175:0.175[0]; log p(e|m)= -0.050:-0.050[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNortheast[0.045]; Tennessee[0.044]; Thursday[0.043]; northward[0.043]; Louisiana[0.043]; Georgia[0.043]; Lightning[0.043]; storm[0.042]; Gulf[0.042]; Maine[0.042]; Texas[0.041]; tonight[0.041]; Oklahoma[0.041]; Pennsylvania[0.041]; Coast[0.041]; areas[0.041]; Ohio[0.040]; seven[0.040]; expected[0.040]; expected[0.040]; Dallas[0.040]; thunderstorms[0.040]; hail[0.040]; hail[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Southeast ==> ENTITY: \u001b[32mSoutheastern United States\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.160:0.160[0]; log p(e|m)= -1.514:-1.514[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoutheast[0.045]; Northeast[0.044]; Northeast[0.044]; West[0.042]; Southwest[0.042]; coasts[0.042]; continental[0.042]; Oklahoma[0.042]; seven[0.041]; areas[0.041]; Midwest[0.041]; Plains[0.041]; Plains[0.041]; Plains[0.041]; Mountains[0.041]; communities[0.041]; Atlantic[0.041]; states[0.041]; Lakes[0.041]; mid[0.041]; Maine[0.041]; Ohio[0.041]; Washington[0.041]; delayed[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Louisiana ==> ENTITY: \u001b[32mLouisiana\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.094:0.094[0]; log p(e|m)= -0.178:-0.178[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstorm[0.047]; Texas[0.046]; Tennessee[0.045]; Oklahoma[0.045]; Gulf[0.045]; Pennsylvania[0.044]; Georgia[0.044]; Maine[0.044]; Fort[0.044]; seven[0.044]; Northeast[0.044]; Ohio[0.043]; Thursday[0.043]; expected[0.043]; Carolinas[0.043]; Weather[0.043]; Coast[0.042]; areas[0.042]; large[0.042]; Nation[0.042]; Dallas[0.042]; delayed[0.042]; tonight[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Northwest ==> ENTITY: \u001b[32mPacific Northwest\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.144:0.144[0]; log p(e|m)= -2.813:-2.813[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoutheast[0.047]; Southeast[0.047]; West[0.045]; Southwest[0.045]; Northeast[0.045]; Northeast[0.045]; Midwest[0.044]; continental[0.044]; states[0.044]; States[0.044]; Oregon[0.044]; seven[0.043]; Airport[0.043]; Thursday[0.043]; mid[0.043]; delayed[0.043]; communities[0.042]; Washington[0.042]; Maine[0.042]; United[0.042]; Atlantic[0.042]; International[0.041]; Mountains[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = freezing ==> ENTITY: \u001b[32mFreezing\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.129:0.129[0]; log p(e|m)= -0.147:-0.147[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntemperature[0.046]; conditions[0.046]; low[0.044]; delayed[0.043]; amounts[0.043]; continental[0.042]; degrees[0.042]; chill[0.042]; communities[0.042]; heavy[0.042]; inches[0.041]; inches[0.041]; minus[0.041]; storm[0.041]; Northwest[0.041]; Southeast[0.041]; Southeast[0.041]; mid[0.041]; hail[0.041]; hail[0.041]; Lighter[0.040]; wind[0.040]; wind[0.040]; wind[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dallas-Fort Worth International Airport\u001b[39m ==> ENTITY: \u001b[32mDallas/Fort Worth International Airport\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.094:0.094[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWashington[0.045]; Thursday[0.045]; flights[0.045]; Texas[0.043]; Northwest[0.043]; Southeast[0.043]; mid[0.042]; today[0.042]; today[0.042]; Midwest[0.042]; delayed[0.042]; Lakes[0.042]; West[0.041]; Rockies[0.041]; Rockies[0.041]; Oklahoma[0.041]; seven[0.041]; Northeast[0.040]; Northeast[0.040]; expected[0.040]; expected[0.040]; Highs[0.040]; tonight[0.040]; Severe[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lightning ==> ENTITY: \u001b[32mLightning\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.545:-0.545[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstorm[0.043]; Thursday[0.043]; wind[0.043]; wind[0.043]; wind[0.043]; Dallas[0.043]; tonight[0.043]; Fort[0.042]; freezing[0.042]; thunderstorms[0.042]; Weather[0.042]; hours[0.041]; hail[0.041]; hail[0.041]; hitting[0.041]; heavy[0.041]; seven[0.041]; Coast[0.041]; rain[0.041]; rain[0.041]; winds[0.041]; Adirondack[0.041]; delayed[0.040]; Rain[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oregon ==> ENTITY: \u001b[32mOregon\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.113:0.113[0]; log p(e|m)= -0.286:-0.286[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncommunities[0.045]; areas[0.044]; Washington[0.043]; states[0.043]; Mountains[0.042]; Northeast[0.042]; Northeast[0.042]; Maine[0.042]; Oklahoma[0.042]; Southeast[0.042]; Southeast[0.042]; Southwest[0.041]; States[0.041]; Plains[0.041]; Plains[0.041]; Plains[0.041]; Pennsylvania[0.041]; Valley[0.041]; West[0.041]; seven[0.041]; Texas[0.041]; Texas[0.041]; Fort[0.040]; Airport[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pennsylvania ==> ENTITY: \u001b[32mPennsylvania\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.118:-0.118[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNortheast[0.044]; Northeast[0.044]; Carolinas[0.043]; Thursday[0.043]; Thursday[0.043]; Washington[0.043]; wind[0.043]; wind[0.043]; wind[0.043]; states[0.042]; Ohio[0.042]; Maine[0.041]; England[0.041]; West[0.041]; Southeast[0.041]; Southeast[0.041]; Coast[0.041]; States[0.040]; Atlantic[0.040]; areas[0.040]; Texas[0.040]; Valley[0.040]; coasts[0.040]; Tennessee[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Plains ==> ENTITY: \u001b[32mGreat Plains\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.181:0.181[0]; log p(e|m)= -1.772:-1.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMountains[0.048]; continental[0.047]; Plains[0.046]; Plains[0.046]; areas[0.045]; Southwest[0.045]; West[0.044]; Cotulla[0.044]; communities[0.043]; Pennsylvania[0.043]; Valley[0.043]; Texas[0.043]; Texas[0.043]; Midwest[0.043]; Northwest[0.043]; states[0.042]; States[0.042]; Northeast[0.042]; Northeast[0.042]; coasts[0.042]; mid[0.042]; Oklahoma[0.042]; Maine[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Adirondack ==> ENTITY: \u001b[32mAdirondack Mountains\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.960:-0.960[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMountains[0.047]; West[0.046]; Maine[0.046]; Washington[0.045]; seven[0.044]; Atlantic[0.044]; Valley[0.044]; areas[0.044]; Northeast[0.044]; Northeast[0.044]; Plains[0.043]; Plains[0.043]; Plains[0.043]; large[0.043]; Lightning[0.043]; conditions[0.043]; Pennsylvania[0.043]; Great[0.043]; Carolinas[0.042]; Southwest[0.042]; Coast[0.042]; states[0.042]; England[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = wind ==> ENTITY: \u001b[32mWind\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.179:0.179[0]; log p(e|m)= -0.434:-0.434[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwind[0.048]; wind[0.048]; freezing[0.047]; Weather[0.046]; conditions[0.046]; chill[0.045]; storm[0.045]; winds[0.044]; northward[0.044]; heavy[0.043]; large[0.043]; mph[0.042]; Gulf[0.042]; thunderstorms[0.042]; Maine[0.042]; expected[0.042]; expected[0.042]; areas[0.042]; amounts[0.042]; hail[0.042]; hail[0.042]; rain[0.042]; rain[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = mid-Atlantic states\u001b[39m ==> ENTITY: \u001b[32mMid-Atlantic states\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.213:0.213[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNortheast[0.048]; Northeast[0.048]; Southeast[0.044]; Southeast[0.044]; Midwest[0.043]; continental[0.042]; Pennsylvania[0.042]; States[0.042]; Plains[0.042]; Plains[0.042]; Plains[0.042]; expected[0.042]; Southwest[0.041]; communities[0.041]; England[0.041]; coasts[0.040]; Minn[0.040]; Maine[0.040]; Thursday[0.040]; Northwest[0.040]; Texas[0.039]; reported[0.039]; low[0.039]; inches[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Severe thunderstorms\u001b[39m ==> ENTITY: \u001b[32mThunderstorm\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstorm[0.047]; Weather[0.046]; wind[0.044]; wind[0.044]; rain[0.043]; rain[0.043]; Rain[0.042]; forecast[0.042]; forecast[0.042]; expected[0.042]; northward[0.041]; heavy[0.041]; Northeast[0.041]; inches[0.041]; inches[0.041]; hail[0.041]; hail[0.041]; tonight[0.041]; areas[0.040]; Thursday[0.040]; hours[0.040]; Mountains[0.040]; Carolinas[0.040]; roared[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Georgia ==> ENTITY: \u001b[32mGeorgia (U.S. state)\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.701:-0.701[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennessee[0.046]; Carolinas[0.045]; Oklahoma[0.044]; Northeast[0.044]; Texas[0.043]; tonight[0.043]; seven[0.043]; Louisiana[0.042]; Ohio[0.042]; Thursday[0.042]; Dallas[0.042]; Maine[0.041]; hitting[0.041]; Pennsylvania[0.041]; Coast[0.040]; Mountains[0.040]; forecast[0.040]; forecast[0.040]; forecast[0.040]; delayed[0.040]; large[0.040]; possible[0.040]; possible[0.040]; Fort[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = the West ==> ENTITY: \u001b[32mWestern United States\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.138:0.138[0]; log p(e|m)= -1.650:-1.650[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMidwest[0.048]; Northeast[0.046]; Northeast[0.046]; Southeast[0.044]; Southeast[0.044]; Ohio[0.044]; continental[0.044]; states[0.044]; Southwest[0.043]; mid[0.043]; Pennsylvania[0.043]; Mountains[0.043]; Plains[0.043]; Plains[0.043]; Plains[0.043]; Oklahoma[0.043]; England[0.042]; communities[0.042]; Maine[0.042]; States[0.042]; Great[0.042]; today[0.042]; today[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Midwest ==> ENTITY: \u001b[32mMidwestern United States\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.163:0.163[0]; log p(e|m)= -0.031:-0.031[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncontinental[0.046]; West[0.046]; Northeast[0.045]; Northeast[0.045]; states[0.044]; Southeast[0.044]; Southeast[0.044]; Dallas[0.044]; seven[0.044]; communities[0.044]; Pennsylvania[0.043]; States[0.043]; Southwest[0.043]; Washington[0.043]; mid[0.043]; Maine[0.043]; Northwest[0.043]; expected[0.042]; Plains[0.042]; Plains[0.042]; Plains[0.042]; England[0.042]; Atlantic[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Maine ==> ENTITY: \u001b[32mMaine\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPennsylvania[0.045]; Atlantic[0.045]; England[0.044]; Northeast[0.043]; Northeast[0.043]; Coast[0.043]; Carolinas[0.042]; Tennessee[0.042]; Louisiana[0.042]; states[0.042]; Oregon[0.041]; Plains[0.041]; Plains[0.041]; Plains[0.041]; States[0.041]; areas[0.041]; Adirondack[0.041]; West[0.041]; Georgia[0.040]; expected[0.040]; expected[0.040]; Mountains[0.040]; Oklahoma[0.040]; Washington[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oklahoma ==> ENTITY: \u001b[32mOklahoma\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.352:-0.352[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTexas[0.046]; Plains[0.044]; Tennessee[0.043]; West[0.043]; Southeast[0.043]; Louisiana[0.042]; Northeast[0.042]; Dallas[0.042]; areas[0.042]; Pennsylvania[0.042]; Valley[0.042]; northward[0.041]; Thursday[0.041]; Ohio[0.041]; Georgia[0.041]; tonight[0.041]; Carolinas[0.041]; seven[0.041]; Washington[0.041]; Nation[0.040]; Mountains[0.040]; Fort[0.040]; Rain[0.040]; coasts[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tennessee ==> ENTITY: \u001b[32mTennessee\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.114:0.114[0]; log p(e|m)= -0.280:-0.280[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTexas[0.045]; Georgia[0.045]; Thursday[0.044]; Oklahoma[0.044]; Carolinas[0.043]; Dallas[0.043]; Fort[0.042]; Pennsylvania[0.042]; Ohio[0.042]; Northeast[0.042]; Louisiana[0.042]; seven[0.042]; Valley[0.041]; areas[0.041]; Gulf[0.041]; expected[0.040]; expected[0.040]; Mountains[0.040]; tonight[0.040]; storm[0.040]; northward[0.040]; large[0.040]; Maine[0.039]; roared[0.039]; \t\n\r [=====================>........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 349/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19980930_0522.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.168:0.168[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraqi[0.048]; Islamic[0.043]; Iran[0.043]; Iran[0.043]; Iran[0.043]; international[0.043]; Iranian[0.042]; border[0.042]; western[0.042]; Republic[0.041]; Saleh[0.041]; Saleh[0.041]; said[0.041]; official[0.041]; seven[0.041]; ties[0.041]; Mohammed[0.041]; trade[0.040]; trade[0.040]; trade[0.040]; saying[0.040]; Agency[0.040]; Mehdi[0.040]; Mohammad[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tehran\u001b[39m ==> ENTITY: \u001b[32mTehran\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.046]; Iran[0.046]; Iran[0.046]; Iran[0.046]; Iran[0.046]; Iranian[0.044]; Iranian[0.044]; Tehran[0.042]; Mehdi[0.042]; Kermanshah[0.041]; western[0.041]; province[0.040]; Iraq[0.040]; Iraq[0.040]; Iraq[0.040]; Iraq[0.040]; official[0.040]; Islamic[0.039]; saying[0.039]; border[0.039]; people[0.039]; Agency[0.039]; Mohammad[0.039]; discussions[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Taha Yassin Ramadan\u001b[39m ==> ENTITY: \u001b[32mTaha Yassin Ramadan\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraqi[0.046]; Iraqi[0.046]; Iraq[0.046]; Iraq[0.046]; Iraq[0.046]; President[0.043]; Baghdad[0.043]; December[0.042]; Vice[0.042]; Iranian[0.042]; vj[0.041]; war[0.040]; war[0.040]; war[0.040]; prisoners[0.040]; prisoners[0.040]; Iran[0.040]; Iran[0.040]; Iran[0.040]; Islamic[0.040]; Conference[0.040]; ranking[0.039]; infiltrators[0.039]; Organization[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = eight-year war\u001b[39m ==> ENTITY: \u001b[32mIran–Iraq War\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.046]; Iraq[0.046]; Iraq[0.046]; Iraq[0.046]; Iranian[0.043]; Iran[0.043]; Iran[0.043]; Iran[0.043]; Iran[0.043]; Iran[0.043]; Iraqi[0.043]; Iraqi[0.043]; war[0.042]; war[0.042]; Baghdad[0.040]; Tehran[0.040]; Tehran[0.040]; Tehran[0.040]; Gulf[0.039]; War[0.038]; saying[0.038]; countries[0.038]; December[0.038]; border[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = 1991 Gulf War\u001b[39m ==> ENTITY: \u001b[32mGulf War\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.048]; Iraq[0.048]; Iraq[0.048]; war[0.044]; war[0.044]; war[0.044]; Iraqi[0.044]; Iraqi[0.044]; Baghdad[0.042]; Organization[0.041]; Iran[0.040]; Iran[0.040]; vj[0.040]; Iranian[0.040]; Conference[0.040]; December[0.040]; infiltrators[0.040]; visit[0.039]; helping[0.039]; Islamic[0.039]; accuses[0.039]; wounded[0.038]; ti[0.038]; cease[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.120:0.120[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.044]; Iraq[0.044]; Iraq[0.044]; Iraq[0.044]; saying[0.043]; countries[0.043]; Iraqi[0.042]; Iraqi[0.042]; Iran[0.042]; Iran[0.042]; Iran[0.042]; Iran[0.042]; Iran[0.042]; Organization[0.041]; groups[0.041]; Iranian[0.040]; International[0.040]; Baghdad[0.040]; President[0.040]; Tehran[0.040]; Tehran[0.040]; Tehran[0.040]; agency[0.040]; borders[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = trade fair\u001b[39m ==> ENTITY: \u001b[32mTrade fair\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.010:-0.010[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninternational[0.046]; trade[0.045]; trade[0.045]; commercial[0.044]; Trade[0.044]; news[0.043]; attend[0.042]; attend[0.042]; hold[0.042]; said[0.042]; fair[0.041]; official[0.041]; discuss[0.041]; Wednesday[0.041]; traveled[0.040]; seven[0.040]; Kermanshah[0.040]; Madari[0.040]; Minister[0.040]; agency[0.040]; News[0.040]; Republic[0.040]; delegation[0.040]; Agency[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = delegation ==> ENTITY: \u001b[32mDelegation\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.119:0.119[0]; log p(e|m)= -0.662:-0.662[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninternational[0.047]; countries[0.044]; member[0.043]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; discuss[0.042]; International[0.042]; discussions[0.042]; Republic[0.042]; official[0.042]; ties[0.042]; attend[0.042]; attend[0.042]; Iranian[0.041]; province[0.041]; seven[0.041]; hold[0.040]; agency[0.040]; Iran[0.040]; Iran[0.040]; Iran[0.040]; Iran[0.040]; Kermanshah[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Organization of the Islamic Conference\u001b[39m ==> ENTITY: \u001b[32mOrganisation of Islamic Cooperation\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmeeting[0.045]; Iran[0.043]; Iran[0.043]; Gulf[0.043]; vj[0.042]; December[0.042]; Iranian[0.042]; highest[0.042]; ranking[0.042]; Ramadan[0.042]; Iraq[0.042]; Iraq[0.042]; President[0.042]; attended[0.041]; groups[0.041]; visit[0.041]; Tehran[0.041]; Tehran[0.041]; Iraqi[0.041]; Iraqi[0.041]; brokered[0.041]; million[0.040]; wounded[0.040]; ti[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = prisoners of war\u001b[39m ==> ENTITY: \u001b[32mPrisoner of war\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.103:0.103[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nprisoners[0.045]; Iraq[0.044]; Iraq[0.044]; Iraq[0.044]; Iraq[0.044]; Iraqi[0.044]; Iraqi[0.044]; war[0.042]; war[0.042]; War[0.042]; Baghdad[0.042]; million[0.041]; said[0.041]; Islamic[0.040]; despite[0.040]; people[0.040]; fought[0.040]; killed[0.040]; Mohammad[0.040]; countries[0.040]; traditional[0.040]; Organization[0.040]; helping[0.040]; saying[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = commercial ==> ENTITY: \u001b[32mCommerce\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.699:-0.699[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrade[0.048]; trade[0.048]; business[0.046]; countries[0.044]; international[0.044]; Republic[0.042]; million[0.041]; hold[0.041]; western[0.041]; holding[0.041]; despite[0.041]; traditional[0.041]; International[0.041]; need[0.040]; land[0.040]; agency[0.040]; border[0.040]; people[0.040]; Trade[0.040]; Trade[0.040]; Islamic[0.040]; province[0.040]; news[0.040]; Agency[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iranian ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.216:0.216[0]; log p(e|m)= -0.571:-0.571[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.046]; Iran[0.046]; Iran[0.046]; Mehdi[0.045]; Iraq[0.044]; Iraq[0.044]; Tehran[0.043]; Islamic[0.042]; western[0.042]; Kermanshah[0.042]; Mohammad[0.042]; Mohammed[0.041]; Iraqi[0.041]; countries[0.041]; ties[0.040]; international[0.040]; traveled[0.039]; Agency[0.039]; said[0.039]; attend[0.039]; attend[0.039]; province[0.039]; official[0.039]; Republic[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baghdad\u001b[39m ==> ENTITY: \u001b[32mBaghdad\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.221:0.221[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.050]; Iraq[0.050]; Iraq[0.050]; Iraq[0.050]; Iraqi[0.046]; Iraqi[0.046]; Islamic[0.041]; war[0.040]; war[0.040]; war[0.040]; Iranian[0.040]; Iran[0.040]; Iran[0.040]; Iran[0.040]; Yassin[0.040]; Taha[0.039]; wounded[0.039]; Tehran[0.039]; Tehran[0.039]; Tehran[0.039]; meeting[0.038]; visit[0.038]; million[0.038]; Gulf[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kermanshah province\u001b[39m ==> ENTITY: \u001b[32mKermanshah Province\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraqi[0.045]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; Iranian[0.044]; western[0.043]; Mohammad[0.043]; Iran[0.043]; Iran[0.043]; Iran[0.043]; Iran[0.043]; Iran[0.043]; Tehran[0.042]; Mehdi[0.042]; Islamic[0.041]; border[0.040]; Mohammed[0.039]; official[0.038]; war[0.038]; war[0.038]; Madari[0.038]; reported[0.038]; Minister[0.037]; \t\n\r [======================>.......................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 365/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19981010_0354.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = William Cohen\u001b[39m ==> ENTITY: \u001b[32mWilliam Cohen\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSecretary[0.045]; Defense[0.045]; Cohen[0.045]; Cohen[0.045]; said[0.043]; said[0.043]; nation[0.043]; States[0.043]; reporters[0.042]; defense[0.042]; defense[0.042]; defense[0.042]; ballistic[0.041]; speaking[0.041]; Iraq[0.041]; tour[0.041]; allies[0.040]; missile[0.040]; weapons[0.040]; Iran[0.040]; Arab[0.039]; Oman[0.039]; proliferation[0.039]; Kuwait[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Qatar ==> ENTITY: \u001b[32mQatar\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.262:0.262[0]; log p(e|m)= -0.355:-0.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKuwait[0.049]; Bahrain[0.049]; Arab[0.047]; Oman[0.046]; countries[0.045]; countries[0.045]; countries[0.045]; Saudi[0.043]; Emirates[0.043]; second[0.043]; Iran[0.043]; Iran[0.043]; Iran[0.043]; region[0.043]; Arabia[0.043]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; Gulf[0.041]; Gulf[0.041]; Gulf[0.041]; Gulf[0.041]; Gulf[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United Arab Emirates ==> ENTITY: \u001b[32mUnited Arab Emirates\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.254:0.254[0]; log p(e|m)= -0.250:-0.250[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nQatar[0.047]; Kuwait[0.047]; Bahrain[0.046]; Oman[0.044]; Iran[0.042]; Iran[0.042]; Iran[0.042]; countries[0.042]; countries[0.042]; countries[0.042]; Saudi[0.041]; Iraq[0.041]; Iraq[0.041]; Iraq[0.041]; Gulf[0.040]; Gulf[0.040]; Gulf[0.040]; Gulf[0.040]; Gulf[0.040]; Gulf[0.040]; major[0.040]; development[0.040]; Arabia[0.040]; second[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.217:0.217[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKuwait[0.048]; Qatar[0.047]; Arab[0.045]; region[0.043]; Bahrain[0.043]; Arabia[0.042]; countries[0.042]; countries[0.042]; countries[0.042]; Saudi[0.041]; Iran[0.041]; Oman[0.041]; missile[0.041]; missile[0.041]; Gulf[0.040]; Gulf[0.040]; Gulf[0.040]; Gulf[0.040]; Gulf[0.040]; Emirates[0.040]; weapons[0.040]; allies[0.040]; allies[0.040]; Defense[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oman ==> ENTITY: \u001b[32mOman\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.295:0.295[0]; log p(e|m)= -0.390:-0.390[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKuwait[0.047]; Bahrain[0.045]; Qatar[0.044]; region[0.043]; region[0.043]; Arab[0.042]; Gulf[0.042]; Gulf[0.042]; Gulf[0.042]; Gulf[0.042]; Gulf[0.042]; Iraq[0.041]; Iraq[0.041]; Iraq[0.041]; countries[0.041]; countries[0.041]; countries[0.041]; Arabia[0.040]; second[0.040]; Iran[0.040]; Iran[0.040]; Iran[0.040]; Saudi[0.040]; Emirates[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gulf ==> ENTITY: \u001b[32mPersian Gulf\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.199:0.199[0]; log p(e|m)= -2.163:-2.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.046]; Gulf[0.046]; Gulf[0.046]; Gulf[0.046]; Arab[0.046]; Bahrain[0.045]; countries[0.044]; countries[0.044]; Qatar[0.043]; Emirates[0.043]; nation[0.041]; States[0.041]; Iran[0.041]; Defense[0.040]; said[0.039]; said[0.039]; development[0.039]; includes[0.039]; allies[0.039]; missile[0.039]; weapons[0.038]; Secretary[0.038]; acquire[0.038]; long[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = ballistic missile defense system ==> ENTITY: \u001b[32mUnited States national missile defense\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.188:0.188[0]; log p(e|m)= -1.100:-1.100[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmissile[0.047]; Defense[0.046]; Missile[0.045]; States[0.044]; systems[0.043]; weapons[0.043]; nation[0.041]; countries[0.041]; countries[0.041]; development[0.041]; defense[0.041]; defense[0.041]; allies[0.041]; United[0.041]; United[0.041]; range[0.041]; technology[0.041]; technology[0.041]; proliferation[0.040]; Iran[0.040]; Iraq[0.040]; believe[0.040]; long[0.039]; includes[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Saudi Arabia ==> ENTITY: \u001b[32mSaudi Arabia\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.264:0.264[0]; log p(e|m)= -0.255:-0.255[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKuwait[0.052]; Qatar[0.049]; Arab[0.049]; Bahrain[0.048]; Iraq[0.046]; Iraq[0.046]; Iraq[0.046]; Iraq[0.046]; Oman[0.045]; countries[0.045]; countries[0.045]; countries[0.045]; forces[0.044]; second[0.044]; Gulf[0.044]; Gulf[0.044]; Gulf[0.044]; Gulf[0.044]; Gulf[0.044]; region[0.043]; region[0.043]; Emirates[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bahrain ==> ENTITY: \u001b[32mBahrain\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.270:0.270[0]; log p(e|m)= -0.423:-0.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKuwait[0.048]; Qatar[0.048]; Arab[0.044]; Oman[0.044]; Saudi[0.042]; Iran[0.042]; Iran[0.042]; Gulf[0.042]; Gulf[0.042]; Gulf[0.042]; Gulf[0.042]; Gulf[0.042]; Iraq[0.041]; Iraq[0.041]; Emirates[0.041]; second[0.041]; Arabia[0.040]; region[0.040]; countries[0.040]; countries[0.040]; countries[0.040]; allies[0.039]; allies[0.039]; Secretary[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.064:0.064[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDefense[0.048]; Iraq[0.046]; William[0.046]; States[0.045]; second[0.045]; nation[0.045]; missile[0.044]; missile[0.044]; tour[0.044]; Kuwait[0.044]; Missile[0.044]; weapons[0.043]; ballistic[0.042]; leaving[0.042]; includes[0.042]; countries[0.042]; countries[0.042]; defense[0.042]; defense[0.042]; defense[0.042]; speaking[0.042]; systems[0.042]; globally[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.052]; Iraq[0.052]; Iraq[0.052]; countries[0.050]; Iran[0.050]; Iran[0.050]; threats[0.048]; threats[0.048]; States[0.048]; citing[0.047]; protection[0.046]; said[0.046]; said[0.046]; said[0.046]; forces[0.046]; efforts[0.046]; Saudi[0.046]; permissible[0.046]; proliferation[0.045]; United[0.045]; sanctions[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iran ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.171:0.171[0]; log p(e|m)= -0.138:-0.138[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nQatar[0.047]; Bahrain[0.044]; Kuwait[0.044]; Arab[0.044]; Iraq[0.044]; Oman[0.043]; missile[0.042]; missile[0.042]; proliferation[0.042]; region[0.042]; weapons[0.041]; Saudi[0.041]; Emirates[0.041]; Arabia[0.041]; countries[0.040]; countries[0.040]; allies[0.040]; allies[0.040]; said[0.040]; said[0.040]; said[0.040]; second[0.040]; nation[0.040]; ballistic[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = sanctions ==> ENTITY: \u001b[32mInternational sanctions\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.163:0.163[0]; log p(e|m)= -0.629:-0.629[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.046]; Iraq[0.046]; Iraq[0.046]; Iran[0.045]; Iran[0.045]; countries[0.044]; States[0.044]; protection[0.042]; region[0.042]; region[0.042]; citing[0.041]; efforts[0.041]; Gulf[0.041]; Gulf[0.041]; allies[0.041]; globally[0.040]; permissible[0.040]; Saudi[0.039]; proliferation[0.039]; said[0.039]; said[0.039]; said[0.039]; technology[0.039]; technology[0.039]; \t\n\r [======================>.......................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 378/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19990526_0131.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Anchorage ==> ENTITY: \u001b[32mAnchorage, Alaska\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.167:0.167[0]; log p(e|m)= -0.046:-0.046[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhometown[0.046]; Alaska[0.045]; Alaska[0.045]; Alaska[0.045]; southwest[0.044]; Anchorage[0.044]; rural[0.044]; area[0.043]; area[0.043]; area[0.043]; miles[0.042]; miles[0.042]; National[0.040]; near[0.040]; Wednesday[0.040]; Wednesday[0.040]; State[0.040]; Tuesday[0.040]; Kenai[0.040]; AP[0.039]; wooded[0.039]; trail[0.039]; Soldotna[0.039]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = seismic ==> ENTITY: \u001b[32mSeismology\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.096:0.096[0]; log p(e|m)= -0.104:-0.104[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnear[0.044]; near[0.044]; working[0.044]; warning[0.044]; shell[0.043]; oil[0.042]; threat[0.042]; Kenai[0.042]; said[0.042]; said[0.042]; continuing[0.042]; Alaska[0.042]; single[0.042]; real[0.042]; real[0.042]; body[0.041]; rifle[0.041]; field[0.041]; involved[0.041]; traces[0.040]; casings[0.040]; suggest[0.040]; wounded[0.040]; head[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Soldotna\u001b[39m ==> ENTITY: \u001b[32mSoldotna, Alaska\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmiles[0.044]; miles[0.044]; area[0.044]; area[0.044]; area[0.044]; Anchorage[0.043]; Alaska[0.043]; Alaska[0.043]; Alaska[0.043]; Kenai[0.042]; southwest[0.042]; state[0.042]; near[0.042]; federal[0.042]; National[0.040]; AP[0.040]; single[0.040]; Wednesday[0.040]; Wednesday[0.040]; State[0.040]; State[0.040]; Tuesday[0.039]; rural[0.039]; day[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Swanson River\u001b[39m ==> ENTITY: \u001b[32mSwanson River\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.076:0.076[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \noil[0.045]; near[0.045]; near[0.045]; Alaska[0.044]; Kenai[0.044]; field[0.043]; pose[0.042]; rifle[0.042]; wounded[0.042]; head[0.041]; bear[0.041]; bear[0.041]; casings[0.041]; said[0.041]; said[0.041]; spent[0.041]; blood[0.041]; Bartley[0.040]; caliber[0.040]; real[0.040]; real[0.040]; continuing[0.040]; Troopers[0.040]; seismic[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bear ==> ENTITY: \u001b[32mBear\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.749:-0.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbear[0.048]; area[0.044]; area[0.044]; area[0.044]; trail[0.044]; near[0.043]; southwest[0.043]; miles[0.042]; miles[0.042]; wooded[0.042]; man[0.041]; Alaska[0.041]; Alaska[0.041]; Alaska[0.041]; Wednesday[0.041]; Wednesday[0.041]; Tuesday[0.041]; rural[0.041]; Wildlife[0.040]; horse[0.040]; said[0.040]; heavily[0.040]; Kenai[0.040]; single[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alaska State Troopers\u001b[39m ==> ENTITY: \u001b[32mAlaska State Troopers\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlaska[0.046]; Alaska[0.046]; state[0.046]; Bruce[0.045]; Kenneth[0.044]; area[0.044]; area[0.044]; area[0.044]; wildlife[0.044]; fatal[0.043]; Soldotna[0.043]; rural[0.043]; Kenai[0.043]; Wildlife[0.043]; Wildlife[0.043]; federal[0.043]; State[0.043]; officials[0.043]; Anchorage[0.042]; trail[0.042]; hometown[0.042]; National[0.041]; Division[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = oil field ==> ENTITY: \u001b[32mOil field\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.020:-0.020[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseismic[0.047]; near[0.047]; near[0.047]; Alaska[0.047]; River[0.044]; continuing[0.043]; Luis[0.042]; said[0.041]; said[0.041]; working[0.041]; Kenai[0.041]; single[0.041]; suggest[0.040]; casings[0.040]; threat[0.040]; shot[0.040]; Cortes[0.040]; Cortes[0.040]; bear[0.040]; bear[0.040]; spent[0.040]; shots[0.039]; head[0.039]; shell[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = rifle\u001b[39m ==> ENTITY: \u001b[32mRifle\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.009:-0.009[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncasings[0.047]; caliber[0.045]; shell[0.045]; shot[0.043]; working[0.043]; shots[0.043]; said[0.042]; said[0.042]; said[0.042]; field[0.042]; looking[0.042]; single[0.041]; single[0.041]; attack[0.041]; attack[0.041]; wound[0.041]; bear[0.040]; bear[0.040]; bear[0.040]; state[0.040]; know[0.040]; Cates[0.040]; Cates[0.040]; Cates[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Alaska ==> ENTITY: \u001b[32mAlaska\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.158:0.158[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlaska[0.050]; Alaska[0.050]; area[0.043]; area[0.043]; area[0.043]; Anchorage[0.042]; miles[0.042]; miles[0.042]; Troopers[0.042]; near[0.041]; Kenai[0.041]; fatal[0.041]; hometown[0.041]; Wednesday[0.041]; Wednesday[0.041]; southwest[0.040]; National[0.040]; Wildlife[0.040]; Tuesday[0.040]; day[0.040]; rural[0.040]; State[0.040]; Soldotna[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kenai ==> ENTITY: \u001b[32mKenai, Alaska\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.087:0.087[0]; log p(e|m)= -0.078:-0.078[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlaska[0.049]; River[0.047]; near[0.046]; near[0.046]; oil[0.042]; single[0.042]; crew[0.042]; fatal[0.042]; bear[0.042]; bear[0.042]; bite[0.041]; said[0.041]; said[0.041]; seismic[0.041]; Feb[0.040]; shots[0.040]; spent[0.040]; field[0.040]; blood[0.040]; rifle[0.040]; continuing[0.040]; mauling[0.040]; Swanson[0.040]; body[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kenai National Wildlife Refuge\u001b[39m ==> ENTITY: \u001b[32mKenai National Wildlife Refuge\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwildlife[0.051]; Wildlife[0.049]; wooded[0.048]; Alaska[0.047]; Alaska[0.047]; Alaska[0.047]; Conservation[0.046]; trail[0.046]; bear[0.045]; bear[0.045]; bear[0.045]; miles[0.045]; miles[0.045]; Soldotna[0.045]; State[0.044]; State[0.044]; rural[0.044]; southwest[0.043]; federal[0.043]; near[0.043]; Anchorage[0.043]; Bear[0.043]; \t\n\r [=======================>......................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 390/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19980811_0512.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = toxic ==> ENTITY: \u001b[32mToxicity\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.194:0.194[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npoison[0.046]; chemical[0.046]; poisoning[0.044]; arsenic[0.044]; substance[0.043]; compound[0.043]; preservative[0.043]; water[0.042]; water[0.042]; affects[0.042]; sodium[0.042]; nausea[0.042]; cyanide[0.042]; numbness[0.042]; nervous[0.041]; suffering[0.041]; sickened[0.040]; identify[0.040]; said[0.040]; added[0.039]; tea[0.039]; tea[0.039]; tea[0.039]; hospitalized[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tokyo ==> ENTITY: \u001b[32mTokyo\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.093:-0.093[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.048]; Japan[0.048]; central[0.045]; Japanese[0.044]; Friday[0.043]; Tuesday[0.043]; Monday[0.042]; Monday[0.042]; Niigata[0.042]; southwestern[0.042]; morning[0.042]; Masashi[0.041]; air[0.040]; kilometers[0.040]; employees[0.040]; miles[0.040]; northwest[0.040]; attacks[0.040]; filled[0.040]; affects[0.040]; company[0.040]; company[0.040]; break[0.040]; suffered[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mliver disorders\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mLiver\u001b[39m <---> \u001b[32mLiver disease\u001b[39m\t\nSCORES: global= 0.276:0.276[\u001b[31m0.000\u001b[39m]; local()= 0.136:0.146[\u001b[32m0.010\u001b[39m]; log p(e|m)= -0.178:-1.814[\u001b[31m1.636\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnerve[0.048]; suffering[0.045]; mild[0.044]; numbness[0.043]; nausea[0.043]; sodium[0.043]; hospitalized[0.043]; suffered[0.042]; condition[0.042]; poisoning[0.042]; poisoning[0.042]; arsenic[0.042]; tests[0.042]; Hospital[0.041]; azide[0.041]; chemical[0.040]; attacks[0.040]; attack[0.040]; festival[0.040]; Toma[0.040]; Hirose[0.040]; traces[0.039]; air[0.039]; case[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = copycat ==> ENTITY: \u001b[32mCopycat crime\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.065:0.065[0]; log p(e|m)= -0.646:-0.646[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nattacks[0.045]; people[0.045]; affects[0.044]; Lab[0.044]; case[0.044]; revealed[0.043]; nerve[0.043]; Hospital[0.042]; compound[0.042]; wary[0.041]; azide[0.041]; victims[0.041]; attack[0.041]; tests[0.041]; fears[0.041]; March[0.041]; festival[0.040]; suffering[0.040]; Toma[0.040]; Hirose[0.040]; gave[0.040]; cult[0.040]; disorders[0.040]; carried[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = nausea\u001b[39m ==> ENTITY: \u001b[32mNausea\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.187:0.187[0]; log p(e|m)= -0.009:-0.009[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnumbness[0.049]; liver[0.048]; mild[0.048]; poisoning[0.046]; poisoning[0.046]; disorders[0.045]; nervous[0.045]; affects[0.044]; substance[0.044]; sodium[0.044]; sodium[0.044]; suffering[0.043]; curry[0.042]; water[0.042]; water[0.042]; hospitalized[0.042]; condition[0.041]; sickened[0.041]; night[0.041]; chemical[0.041]; chemical[0.041]; compound[0.041]; central[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = wood preservative ==> ENTITY: \u001b[32mWood preservation\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.955:-0.955[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nproducts[0.047]; chemical[0.045]; chemical[0.045]; compound[0.044]; industrial[0.043]; water[0.043]; water[0.043]; toxic[0.042]; company[0.042]; company[0.042]; said[0.041]; said[0.041]; arsenic[0.041]; azide[0.041]; substance[0.041]; affects[0.040]; southwestern[0.040]; tea[0.040]; tea[0.040]; tea[0.040]; tea[0.040]; sodium[0.040]; sodium[0.040]; central[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMarch 1995\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31m1994–95 NBA season\u001b[39m <---> \u001b[32m1995\u001b[39m\t\nSCORES: global= 0.259:0.243[\u001b[31m0.016\u001b[39m]; local()= 0.015:0.008[\u001b[31m0.007\u001b[39m]; log p(e|m)= 0.000:-0.272[\u001b[31m0.272\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCity[0.045]; Tuesday[0.044]; weeks[0.044]; Hospital[0.043]; suffered[0.043]; azide[0.043]; cult[0.042]; festival[0.042]; Toma[0.042]; Hirose[0.042]; said[0.042]; said[0.042]; disorders[0.042]; wary[0.041]; air[0.041]; official[0.041]; liver[0.041]; revealed[0.041]; copycat[0.041]; incidents[0.040]; particularly[0.040]; Japan[0.040]; Japan[0.040]; Japan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = sodium azide\u001b[39m ==> ENTITY: \u001b[32mSodium azide\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnerve[0.046]; cyanide[0.045]; chemical[0.045]; chemical[0.045]; poisoning[0.043]; poisoning[0.043]; traces[0.042]; tests[0.042]; Hospital[0.042]; arsenic[0.041]; industrial[0.041]; products[0.041]; shock[0.041]; laced[0.041]; bag[0.041]; preservative[0.041]; gas[0.040]; festival[0.040]; Toma[0.040]; Hirose[0.040]; liver[0.040]; disorders[0.040]; water[0.040]; mild[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = numbness ==> ENTITY: \u001b[32mParesthesia\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.172:0.172[0]; log p(e|m)= -0.069:-0.069[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmild[0.048]; disorders[0.048]; liver[0.047]; suffering[0.046]; nausea[0.046]; suffered[0.045]; condition[0.045]; nervous[0.045]; sodium[0.043]; sodium[0.043]; hospitalized[0.043]; shock[0.043]; affects[0.043]; attacks[0.042]; tests[0.042]; identify[0.042]; azide[0.042]; toxic[0.042]; substance[0.041]; festival[0.041]; mechanisms[0.041]; Toma[0.041]; poison[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = tea ==> ENTITY: \u001b[32mTea\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.170:0.170[0]; log p(e|m)= -0.135:-0.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntea[0.051]; tea[0.051]; sodium[0.044]; brewed[0.044]; substance[0.043]; preservative[0.043]; nausea[0.042]; toxic[0.041]; poisoning[0.041]; morning[0.041]; pot[0.041]; pot[0.041]; chemical[0.041]; Japanese[0.041]; drank[0.040]; affects[0.040]; water[0.040]; water[0.040]; night[0.040]; compound[0.040]; wood[0.040]; poison[0.039]; Monday[0.038]; Monday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = central nervous system\u001b[39m ==> ENTITY: \u001b[32mCentral nervous system\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.003:-0.003[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \naffects[0.046]; chemical[0.045]; toxic[0.044]; substance[0.044]; compound[0.043]; poison[0.043]; poisoning[0.043]; sodium[0.043]; nausea[0.043]; arsenic[0.043]; numbness[0.042]; attacks[0.042]; identify[0.041]; tests[0.041]; preservative[0.040]; shock[0.040]; suffering[0.040]; laced[0.040]; filled[0.040]; cyanide[0.040]; case[0.039]; case[0.039]; festival[0.039]; added[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = air-bag\u001b[39m ==> ENTITY: \u001b[32mAirbag\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.060:0.060[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHospital[0.044]; azide[0.043]; festival[0.042]; Toma[0.042]; incidents[0.042]; Hirose[0.042]; shock[0.042]; gas[0.042]; disorders[0.042]; tests[0.042]; products[0.042]; hospitalized[0.042]; liver[0.042]; particularly[0.042]; preservative[0.041]; numbness[0.041]; mechanisms[0.041]; added[0.041]; stable[0.041]; said[0.041]; said[0.041]; Niigata[0.041]; Niigata[0.041]; company[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mcurry\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mCurry powder\u001b[39m <---> \u001b[32mCurry\u001b[39m\t\nSCORES: global= 0.253:0.244[\u001b[31m0.008\u001b[39m]; local()= 0.115:0.131[\u001b[32m0.016\u001b[39m]; log p(e|m)= 0.000:-0.233[\u001b[31m0.233\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntea[0.046]; tea[0.046]; tea[0.046]; case[0.044]; northwest[0.044]; Japanese[0.043]; Japan[0.043]; Japan[0.043]; mild[0.042]; sodium[0.041]; sodium[0.041]; liver[0.041]; brewed[0.041]; southwestern[0.041]; bag[0.040]; mechanisms[0.040]; central[0.040]; pot[0.040]; pot[0.040]; laced[0.040]; miles[0.040]; compound[0.040]; nausea[0.040]; affects[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNiigata\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNiigata Prefecture\u001b[39m <---> \u001b[32mNiigata, Niigata\u001b[39m\t\nSCORES: global= 0.255:0.253[\u001b[31m0.002\u001b[39m]; local()= 0.142:0.136[\u001b[31m0.005\u001b[39m]; log p(e|m)= -0.627:-0.519[\u001b[32m0.108\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncase[0.045]; case[0.045]; Japan[0.045]; Japan[0.045]; Tokyo[0.044]; Japanese[0.044]; central[0.043]; southwestern[0.043]; tests[0.042]; Monday[0.042]; Monday[0.042]; Friday[0.041]; Tuesday[0.041]; northwest[0.041]; people[0.041]; kilometers[0.040]; revealed[0.040]; miles[0.040]; Masashi[0.040]; attacks[0.040]; filled[0.040]; toxic[0.040]; water[0.040]; water[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Aum Shinri Kyo\u001b[39m ==> ENTITY: \u001b[32mAum Shinrikyo\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nattacks[0.045]; cult[0.045]; attack[0.043]; copycat[0.043]; Hospital[0.042]; Japan[0.042]; Japan[0.042]; nerve[0.042]; Tokyo[0.042]; March[0.042]; azide[0.041]; religious[0.041]; Tuesday[0.041]; subway[0.041]; revealed[0.041]; subways[0.041]; festival[0.041]; Toma[0.041]; killed[0.041]; killed[0.041]; Hirose[0.041]; disorders[0.040]; chemical[0.040]; incidents[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Niigata City\u001b[39m ==> ENTITY: \u001b[32mNiigata, Niigata\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.115:0.115[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTokyo[0.046]; Japan[0.045]; Japan[0.045]; Japan[0.045]; Tuesday[0.043]; Hospital[0.042]; Yasuo[0.042]; southwestern[0.042]; subways[0.042]; March[0.041]; subway[0.041]; attacks[0.041]; azide[0.041]; attack[0.041]; official[0.041]; festival[0.041]; Masashi[0.041]; Toma[0.041]; Hirose[0.040]; disorders[0.040]; said[0.040]; said[0.040]; air[0.040]; liver[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = cyanide ==> ENTITY: \u001b[32mCyanide\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.165:0.165[0]; log p(e|m)= -0.089:-0.089[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nchemical[0.045]; chemical[0.045]; compound[0.044]; toxic[0.044]; arsenic[0.043]; nerve[0.042]; laced[0.042]; azide[0.042]; industrial[0.042]; nausea[0.042]; sodium[0.041]; sodium[0.041]; poisoning[0.041]; numbness[0.041]; water[0.041]; water[0.041]; affects[0.041]; shock[0.041]; hospitalized[0.041]; drank[0.040]; liver[0.040]; products[0.040]; tests[0.040]; traces[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = police ==> ENTITY: \u001b[32mPolice\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.104:0.104[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPolice[0.046]; Police[0.046]; case[0.044]; case[0.044]; Tuesday[0.043]; Monday[0.043]; Monday[0.043]; attacks[0.043]; identify[0.042]; Friday[0.042]; employees[0.041]; weeks[0.041]; tests[0.041]; gave[0.041]; workers[0.041]; central[0.040]; night[0.040]; northwest[0.040]; morning[0.040]; Lab[0.040]; people[0.040]; said[0.040]; copycat[0.040]; killed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = nerve gas\u001b[39m ==> ENTITY: \u001b[32mNerve agent\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.135:0.135[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nchemical[0.049]; attacks[0.044]; Hospital[0.043]; cyanide[0.042]; attack[0.042]; tests[0.042]; azide[0.042]; Japan[0.042]; Japan[0.042]; poisoning[0.041]; poisoning[0.041]; festival[0.041]; suffered[0.041]; Toma[0.041]; copycat[0.041]; Hirose[0.041]; disorders[0.041]; liver[0.041]; air[0.041]; laced[0.041]; revealed[0.041]; mild[0.041]; carried[0.040]; victims[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = arsenic ==> ENTITY: \u001b[32mArsenic\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.038:-0.038[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwater[0.046]; water[0.046]; cyanide[0.044]; toxic[0.043]; liver[0.043]; compound[0.043]; chemical[0.042]; chemical[0.042]; poisoning[0.042]; affects[0.042]; nerve[0.041]; disorders[0.041]; nausea[0.041]; sodium[0.041]; sodium[0.041]; preservative[0.041]; products[0.041]; tea[0.040]; tea[0.040]; tea[0.040]; suffering[0.040]; tests[0.040]; mechanisms[0.040]; azide[0.040]; \t\n\r [=========================>....................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 411/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19990203_0315.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = The J. Paul Getty Museum\u001b[39m ==> ENTITY: \u001b[32mJ. Paul Getty Museum\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGetty[0.047]; collection[0.045]; collection[0.045]; Museum[0.045]; museum[0.044]; art[0.043]; exhibit[0.043]; sculpture[0.042]; according[0.042]; site[0.042]; works[0.042]; antiquities[0.041]; loaned[0.041]; ancient[0.041]; LOS[0.040]; objects[0.040]; archaeological[0.040]; Monday[0.040]; pieces[0.040]; private[0.040]; Italian[0.040]; Italy[0.039]; Italy[0.039]; Stolen[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = B.C. ==> ENTITY: \u001b[32mAnno Domini\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.728:-0.728[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nancient[0.049]; century[0.045]; century[0.045]; Roman[0.045]; archaeological[0.044]; Mithras[0.042]; Monday[0.042]; Paul[0.042]; week[0.041]; according[0.041]; officials[0.041]; field[0.041]; return[0.041]; excavated[0.041]; claimed[0.041]; athlete[0.041]; museum[0.040]; museum[0.040]; head[0.040]; antiquities[0.040]; returning[0.040]; apparently[0.040]; reasons[0.040]; sculpture[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = looting ==> ENTITY: \u001b[32mLooting\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.052:-0.052[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmuseum[0.045]; museum[0.045]; Roman[0.043]; Getty[0.043]; area[0.043]; billion[0.042]; archaeological[0.042]; rampant[0.042]; pieces[0.042]; art[0.042]; officials[0.042]; private[0.041]; stolen[0.041]; experts[0.041]; collection[0.041]; security[0.041]; worth[0.041]; earlier[0.041]; sites[0.041]; Greek[0.041]; apparently[0.040]; scientific[0.040]; Italian[0.040]; Italian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Roman ==> ENTITY: \u001b[32mAncient Rome\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.180:0.180[0]; log p(e|m)= -1.024:-1.024[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \narchaeological[0.045]; archaeological[0.045]; god[0.045]; area[0.043]; excavated[0.043]; century[0.043]; century[0.043]; sculpture[0.043]; Mithras[0.042]; museum[0.041]; museum[0.041]; museum[0.041]; sites[0.041]; antiquities[0.041]; site[0.041]; art[0.041]; Greek[0.041]; sculpted[0.041]; excavation[0.041]; hill[0.040]; works[0.040]; Italian[0.040]; Italian[0.040]; according[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = excavated\u001b[39m ==> ENTITY: \u001b[32mExcavation (archaeology)\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \narchaeological[0.056]; excavation[0.053]; ancient[0.051]; site[0.047]; antiquities[0.046]; scientific[0.046]; museum[0.046]; museum[0.046]; Roman[0.045]; century[0.045]; century[0.045]; private[0.044]; Museum[0.044]; Museum[0.044]; concern[0.044]; work[0.044]; field[0.043]; according[0.043]; exhibit[0.043]; apparently[0.043]; experts[0.042]; rampant[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = collection ==> ENTITY: \u001b[32mCollection (artwork)\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.175:0.175[0]; log p(e|m)= -1.162:-1.162[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmuseum[0.046]; Museum[0.045]; Museum[0.045]; collection[0.045]; art[0.044]; exhibit[0.044]; objects[0.044]; items[0.044]; pieces[0.042]; century[0.041]; century[0.041]; Italy[0.041]; Italy[0.041]; antiquities[0.041]; sculpture[0.040]; works[0.040]; ancient[0.040]; Roman[0.040]; archaeological[0.040]; according[0.040]; private[0.039]; Getty[0.039]; Getty[0.039]; athlete[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mGreek\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mGreek language\u001b[39m <---> \u001b[32mAncient Greece\u001b[39m\t\nSCORES: global= 0.240:0.233[\u001b[31m0.007\u001b[39m]; local()= 0.122:0.173[\u001b[32m0.051\u001b[39m]; log p(e|m)= -0.625:-2.830[\u001b[31m2.205\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.045]; Roman[0.044]; sculpture[0.044]; objects[0.043]; art[0.043]; archaeological[0.043]; archaeological[0.043]; century[0.042]; century[0.042]; god[0.042]; works[0.042]; collections[0.041]; work[0.041]; Mithras[0.041]; antiquities[0.041]; sculpted[0.041]; Italian[0.041]; Italian[0.041]; pieces[0.040]; pieces[0.040]; excavated[0.040]; exhibit[0.039]; collection[0.039]; collection[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = archaeological site ==> ENTITY: \u001b[32mArchaeological site\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.138:0.138[0]; log p(e|m)= -0.045:-0.045[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nancient[0.045]; excavated[0.045]; objects[0.044]; excavation[0.043]; apparently[0.043]; scientific[0.042]; sculpture[0.042]; museum[0.042]; museum[0.042]; pieces[0.042]; pieces[0.042]; antiquities[0.041]; Museum[0.041]; Museum[0.041]; Roman[0.041]; according[0.041]; say[0.041]; collection[0.040]; collection[0.040]; sculpted[0.040]; century[0.040]; century[0.040]; exhibit[0.040]; torso[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Los Angeles ==> ENTITY: \u001b[32mLos Angeles\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.076:0.076[0]; log p(e|m)= -0.101:-0.101[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \naccording[0.044]; week[0.044]; Monday[0.044]; museum[0.043]; Paul[0.043]; Museum[0.043]; Museum[0.043]; site[0.042]; Getty[0.042]; Getty[0.042]; illegally[0.042]; loaned[0.042]; AP[0.041]; collection[0.041]; flown[0.041]; art[0.041]; exhibit[0.041]; century[0.041]; 2nd[0.041]; returning[0.041]; pieces[0.040]; excavated[0.040]; works[0.040]; statement[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.111:0.111[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.056]; Italian[0.048]; week[0.043]; sculpture[0.042]; art[0.042]; sculpted[0.042]; objects[0.042]; pieces[0.041]; 2nd[0.041]; 2nd[0.041]; works[0.041]; ancient[0.041]; cup[0.041]; Getty[0.041]; Getty[0.041]; Monday[0.041]; statement[0.040]; exhibit[0.040]; issued[0.040]; loaned[0.040]; according[0.039]; flown[0.039]; athlete[0.039]; determining[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = thefts\u001b[39m ==> ENTITY: \u001b[32mTheft\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.097:0.097[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstolen[0.047]; copy[0.044]; rampant[0.044]; security[0.043]; looting[0.043]; information[0.042]; worth[0.042]; storeroom[0.042]; work[0.041]; field[0.041]; athlete[0.041]; say[0.041]; apparently[0.041]; head[0.041]; works[0.041]; 2nd[0.041]; 2nd[0.041]; torso[0.041]; reasons[0.041]; return[0.041]; private[0.041]; area[0.040]; claimed[0.040]; consultation[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = torso\u001b[39m ==> ENTITY: \u001b[32mTrunk (anatomy)\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.070:0.070[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhead[0.047]; LOS[0.044]; sculpted[0.043]; work[0.043]; pieces[0.043]; pieces[0.043]; objects[0.043]; works[0.042]; works[0.042]; cotta[0.042]; field[0.042]; athlete[0.042]; apparently[0.041]; 2nd[0.041]; 2nd[0.041]; items[0.041]; concern[0.041]; cup[0.041]; Museum[0.040]; Museum[0.040]; consultation[0.040]; ANGELES[0.040]; according[0.040]; statement[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = antiquities ==> ENTITY: \u001b[32mAntiquities\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.197:0.197[0]; log p(e|m)= -0.124:-0.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nancient[0.049]; archaeological[0.047]; Roman[0.044]; museum[0.043]; museum[0.043]; Greek[0.043]; pieces[0.042]; pieces[0.042]; excavated[0.042]; collection[0.041]; collection[0.041]; century[0.041]; century[0.041]; excavation[0.041]; objects[0.041]; scientific[0.041]; art[0.041]; Museum[0.040]; Museum[0.040]; exhibit[0.040]; sculpture[0.040]; site[0.039]; say[0.039]; worth[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBrentwood\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBrentwood, Essex\u001b[39m <---> \u001b[32mBrentwood, Los Angeles\u001b[39m\t\nSCORES: global= 0.244:0.243[\u001b[31m0.001\u001b[39m]; local()= 0.094:0.084[\u001b[31m0.010\u001b[39m]; log p(e|m)= -2.002:-1.754[\u001b[32m0.248\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \narea[0.050]; located[0.045]; field[0.042]; dollar[0.042]; museum[0.042]; museum[0.042]; Getty[0.042]; officials[0.042]; say[0.042]; growing[0.042]; based[0.042]; copy[0.041]; apparently[0.041]; work[0.041]; rampant[0.041]; opened[0.041]; worth[0.041]; information[0.041]; works[0.041]; return[0.040]; security[0.040]; hill[0.040]; claimed[0.040]; century[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = works of art\u001b[39m ==> ENTITY: \u001b[32mWork of art\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.117:0.117[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nantiquities[0.046]; sculpture[0.045]; collection[0.045]; collection[0.045]; museum[0.044]; pieces[0.043]; objects[0.043]; exhibit[0.043]; Getty[0.042]; Getty[0.042]; Museum[0.042]; Museum[0.042]; archaeological[0.041]; Italy[0.041]; Italy[0.041]; loaned[0.041]; items[0.040]; excavated[0.040]; ancient[0.040]; stolen[0.039]; stolen[0.039]; stolen[0.039]; Stolen[0.039]; LOS[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mithras\u001b[39m ==> ENTITY: \u001b[32mMithraic mysteries\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRoman[0.048]; god[0.048]; ancient[0.048]; century[0.044]; century[0.044]; archaeological[0.042]; Greek[0.042]; works[0.041]; works[0.041]; sculpture[0.041]; scientific[0.041]; terra[0.041]; LOS[0.040]; week[0.040]; art[0.040]; art[0.040]; excavated[0.040]; museum[0.040]; museum[0.040]; objects[0.040]; Getty[0.040]; Getty[0.040]; Getty[0.040]; antiquities[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = sculpture ==> ENTITY: \u001b[32mSculpture\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.177:0.177[0]; log p(e|m)= -0.173:-0.173[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nancient[0.046]; art[0.045]; sculpted[0.045]; Roman[0.044]; exhibit[0.043]; museum[0.043]; Italian[0.042]; objects[0.042]; works[0.042]; pieces[0.041]; Getty[0.041]; Getty[0.041]; Italy[0.041]; Italy[0.041]; collection[0.041]; collection[0.041]; torso[0.040]; century[0.040]; century[0.040]; antiquities[0.040]; work[0.040]; Museum[0.040]; Museum[0.040]; Paul[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = 2nd century\u001b[39m ==> ENTITY: \u001b[32m2nd century\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.094:0.094[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncentury[0.062]; 2nd[0.048]; Roman[0.048]; ancient[0.047]; Mithras[0.044]; god[0.043]; Paul[0.043]; collection[0.042]; collection[0.042]; according[0.042]; museum[0.042]; museum[0.042]; International[0.042]; collections[0.042]; excavated[0.042]; work[0.042]; archaeological[0.041]; archaeological[0.041]; issued[0.041]; copy[0.041]; experts[0.041]; return[0.041]; loaned[0.041]; \t\n\r [==========================>...................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 429/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW20000312_0050.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = civil war ==> ENTITY: \u001b[32mCivil war\u001b[39m\t\nSCORES: global= 0.237:0.237[0]; local()= 0.124:0.124[0]; log p(e|m)= -0.976:-0.976[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNicaragua[0.046]; Nicaragua[0.046]; Nicaragua[0.046]; Nicaragua[0.046]; Nicaragua[0.046]; government[0.045]; country[0.044]; war[0.042]; war[0.042]; rebels[0.041]; Sandinistas[0.041]; Contra[0.041]; Agencia[0.041]; Nueva[0.040]; helping[0.040]; helped[0.040]; Sandinista[0.040]; turbulent[0.040]; agency[0.039]; local[0.039]; working[0.039]; joined[0.039]; Managua[0.039]; Managua[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = volcanic eruptions\u001b[39m ==> ENTITY: \u001b[32mTypes of volcanic eruptions\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.109:0.109[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nvolcanoes[0.052]; south[0.043]; miles[0.043]; reported[0.043]; reporting[0.042]; University[0.042]; floods[0.042]; town[0.042]; wave[0.042]; Nueva[0.042]; civil[0.041]; civil[0.041]; country[0.041]; total[0.041]; conditions[0.041]; created[0.041]; news[0.041]; turbulent[0.041]; local[0.040]; long[0.040]; government[0.040]; Central[0.040]; agency[0.040]; Hurricane[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = diabetes\u001b[39m ==> ENTITY: \u001b[32mDiabetes mellitus\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.097:0.097[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncancer[0.048]; John[0.046]; illnesses[0.045]; wife[0.044]; suffered[0.043]; south[0.043]; Paul[0.043]; pressure[0.042]; Mexico[0.042]; volcanic[0.042]; America[0.041]; maintain[0.041]; died[0.041]; Central[0.040]; miles[0.040]; said[0.040]; conditions[0.040]; Mitch[0.040]; series[0.040]; reporting[0.040]; born[0.040]; buried[0.040]; Nicaragua[0.040]; total[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pope John Paul II\u001b[39m ==> ENTITY: \u001b[32mPope John Paul II\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.102:0.102[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npapal[0.048]; news[0.046]; National[0.045]; Sunday[0.044]; office[0.044]; country[0.044]; AP[0.044]; AP[0.044]; Press[0.044]; Press[0.044]; long[0.044]; America[0.043]; visit[0.043]; visit[0.043]; Saturday[0.043]; Mexico[0.043]; managed[0.043]; town[0.042]; broadcasts[0.042]; short[0.042]; created[0.042]; total[0.042]; south[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sandinista\u001b[39m ==> ENTITY: \u001b[32mSandinista National Liberation Front\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNicaragua[0.046]; Nicaragua[0.046]; Nicaragua[0.046]; Nicaragua[0.046]; Nicaragua[0.046]; Sandinistas[0.043]; government[0.043]; Radio[0.042]; Contra[0.042]; country[0.041]; Mexico[0.041]; helping[0.040]; Managua[0.040]; Managua[0.040]; John[0.040]; Aguilar[0.040]; Paul[0.040]; Agencia[0.040]; civil[0.039]; civil[0.039]; Garcia[0.039]; Garcia[0.039]; Garcia[0.039]; Garcia[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = cancer ==> ENTITY: \u001b[32mCancer\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.023:-0.023[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndiabetes[0.048]; suffered[0.046]; illnesses[0.046]; died[0.044]; John[0.043]; total[0.042]; National[0.042]; Paul[0.042]; wife[0.042]; Mexico[0.041]; America[0.041]; Sunday[0.041]; Saturday[0.041]; said[0.041]; born[0.041]; visit[0.040]; survived[0.040]; created[0.040]; AP[0.040]; reporting[0.040]; University[0.040]; difficult[0.040]; Central[0.040]; series[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hurricane Mitch\u001b[39m ==> ENTITY: \u001b[32mHurricane Mitch\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNicaragua[0.046]; Nicaragua[0.046]; Nicaragua[0.046]; total[0.044]; country[0.044]; Mexico[0.042]; University[0.042]; south[0.042]; helping[0.042]; miles[0.041]; Central[0.041]; America[0.041]; local[0.041]; suffered[0.041]; wave[0.041]; reported[0.040]; loss[0.040]; government[0.040]; Ana[0.040]; conditions[0.040]; rebels[0.040]; reporting[0.040]; agency[0.040]; town[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nicaragua ==> ENTITY: \u001b[32mNicaragua\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.177:0.177[0]; log p(e|m)= -0.247:-0.247[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNicaragua[0.050]; Nicaragua[0.050]; Nicaragua[0.050]; Nicaragua[0.050]; Contra[0.044]; Sandinistas[0.042]; Sandinista[0.042]; government[0.041]; Garcia[0.041]; Garcia[0.041]; Garcia[0.041]; Managua[0.041]; Managua[0.041]; Agencia[0.040]; Nueva[0.040]; civil[0.040]; war[0.040]; war[0.040]; emergence[0.038]; agency[0.038]; visit[0.038]; helped[0.038]; joined[0.038]; working[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexico ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.183:-0.183[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.046]; Nicaragua[0.045]; Nicaragua[0.045]; Autonomous[0.044]; America[0.043]; Hernandez[0.042]; Garcia[0.041]; Garcia[0.041]; civil[0.041]; south[0.041]; Central[0.041]; town[0.041]; Ana[0.041]; Managua[0.041]; war[0.041]; war[0.041]; war[0.041]; National[0.041]; Aguilar[0.041]; government[0.041]; Contra[0.040]; effort[0.040]; Leonor[0.040]; Eloy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Contra ==> ENTITY: \u001b[32mContras\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.145:0.145[0]; log p(e|m)= -1.309:-1.309[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNicaragua[0.046]; Nicaragua[0.046]; Nicaragua[0.046]; Nicaragua[0.046]; government[0.043]; reporting[0.042]; Mexico[0.042]; Sandinista[0.042]; Sandinistas[0.042]; reported[0.041]; said[0.041]; America[0.041]; Agencia[0.041]; country[0.041]; war[0.040]; war[0.040]; war[0.040]; war[0.040]; agency[0.040]; director[0.040]; chief[0.040]; bureau[0.040]; effort[0.039]; joined[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Managua\u001b[39m ==> ENTITY: \u001b[32mManagua\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.137:0.137[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nManagua[0.051]; Nicaragua[0.047]; Nicaragua[0.047]; Nicaragua[0.047]; Nicaragua[0.047]; Nicaragua[0.047]; Sandinistas[0.042]; country[0.042]; Nueva[0.041]; long[0.041]; reported[0.041]; Noticias[0.041]; Garcia[0.040]; Garcia[0.040]; Garcia[0.040]; Sandinista[0.040]; papal[0.040]; government[0.040]; short[0.039]; Contra[0.038]; Agencia[0.038]; news[0.038]; local[0.038]; Rodolfo[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Central America ==> ENTITY: \u001b[32mCentral America\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.102:0.102[0]; log p(e|m)= -0.059:-0.059[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNicaragua[0.051]; country[0.047]; Mexico[0.046]; government[0.043]; south[0.043]; Managua[0.042]; civil[0.042]; Sandinista[0.041]; Autonomous[0.041]; Contra[0.041]; Ana[0.041]; volcanic[0.041]; Leonor[0.041]; miles[0.041]; Sandinistas[0.040]; Aguilar[0.040]; Eloy[0.040]; Garcia[0.040]; Garcia[0.040]; turbulent[0.040]; National[0.040]; covered[0.039]; said[0.039]; created[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = short-wave\u001b[39m ==> ENTITY: \u001b[32mShortwave radio\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.105:0.105[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbroadcasts[0.053]; Radio[0.049]; country[0.047]; local[0.045]; news[0.045]; Nicaragua[0.043]; Nicaragua[0.043]; Nicaragua[0.043]; Nicaragua[0.043]; Nicaragua[0.043]; rebels[0.043]; reporting[0.042]; reported[0.042]; visit[0.042]; visit[0.042]; agency[0.042]; Dies[0.042]; AP[0.042]; AP[0.042]; AP[0.042]; power[0.042]; working[0.042]; floods[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = National Autonomous University of Nicaragua\u001b[39m ==> ENTITY: \u001b[32mNational Autonomous University of Nicaragua\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.082:0.082[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nManagua[0.047]; Mexico[0.045]; country[0.044]; south[0.044]; attended[0.043]; Ana[0.042]; Central[0.042]; Leonor[0.042]; miles[0.041]; AP[0.041]; diabetes[0.041]; total[0.041]; Press[0.041]; civil[0.041]; Garcia[0.041]; Garcia[0.041]; said[0.041]; bureau[0.040]; chief[0.040]; illnesses[0.040]; America[0.040]; Mitch[0.040]; town[0.040]; Sunday[0.040]; \t\n\r [===========================>..................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 444/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19981215_1083.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zoo ==> ENTITY: \u001b[32mZoo\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.145:0.145[0]; log p(e|m)= -1.044:-1.044[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nelephants[0.049]; elephants[0.049]; elephants[0.049]; elephants[0.049]; Zoo[0.049]; zoo[0.049]; zoos[0.048]; wild[0.044]; Africa[0.043]; month[0.042]; South[0.042]; South[0.042]; African[0.042]; African[0.042]; Tuesday[0.041]; August[0.040]; director[0.040]; forbidding[0.040]; group[0.040]; early[0.040]; baby[0.040]; baby[0.040]; herds[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.174:0.174[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.049]; Germany[0.049]; Switzerland[0.047]; German[0.044]; eastern[0.044]; Cologne[0.043]; Juergen[0.043]; ruled[0.041]; week[0.041]; Dresden[0.041]; Dresden[0.041]; Erfurt[0.041]; month[0.041]; month[0.041]; forbidding[0.040]; August[0.040]; face[0.040]; early[0.039]; South[0.039]; South[0.039]; said[0.039]; said[0.039]; Africa[0.039]; filed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = put to sleep\u001b[39m ==> ENTITY: \u001b[32mAnimal euthanasia\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.103:0.103[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nanimals[0.048]; animals[0.048]; animals[0.048]; zoos[0.045]; elephants[0.044]; elephants[0.044]; elephants[0.044]; school[0.044]; addition[0.044]; Animals[0.044]; week[0.043]; baby[0.043]; steps[0.043]; Erfurt[0.042]; abuse[0.042]; abused[0.042]; Tuesday[0.042]; temporary[0.042]; aet[0.042]; Norbert[0.042]; come[0.042]; cultural[0.042]; prevent[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Botswana ==> ENTITY: \u001b[32mBotswana\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.159:0.159[0]; log p(e|m)= -0.371:-0.371[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrica[0.045]; export[0.045]; South[0.045]; South[0.045]; African[0.044]; African[0.044]; National[0.042]; Switzerland[0.041]; granted[0.041]; Tuesday[0.041]; eastern[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; farm[0.040]; farm[0.040]; August[0.040]; month[0.040]; month[0.040]; court[0.040]; court[0.040]; court[0.040]; group[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cologne ==> ENTITY: \u001b[32mCologne\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.080:0.080[0]; log p(e|m)= -0.082:-0.082[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.050]; Germany[0.050]; Germany[0.050]; German[0.049]; Erfurt[0.047]; Switzerland[0.045]; Juergen[0.044]; zoo[0.044]; Dresden[0.043]; Dresden[0.043]; Zoo[0.043]; Zoo[0.043]; zoos[0.043]; elephants[0.042]; elephants[0.042]; elephants[0.042]; elephants[0.042]; elephants[0.042]; animals[0.040]; Trittin[0.039]; Trittin[0.039]; Africa[0.038]; wild[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Saxony ==> ENTITY: \u001b[32mSaxony\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.149:0.149[0]; log p(e|m)= -0.146:-0.146[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Switzerland[0.044]; Otto[0.043]; granted[0.043]; state[0.043]; Dresden[0.042]; Erfurt[0.042]; Norbert[0.042]; parliament[0.041]; write[0.041]; lawmaker[0.041]; said[0.040]; minister[0.040]; protest[0.040]; protest[0.040]; federal[0.040]; addition[0.039]; South[0.039]; cultural[0.039]; approved[0.039]; temporary[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South African ==> ENTITY: \u001b[32mSouth Africa\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrican[0.049]; South[0.046]; Africa[0.045]; Tuesday[0.042]; Switzerland[0.042]; said[0.042]; said[0.042]; National[0.042]; Botswana[0.042]; month[0.041]; month[0.041]; week[0.041]; eastern[0.041]; granted[0.041]; August[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Minister[0.040]; went[0.040]; demanded[0.040]; approved[0.039]; early[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dresden ==> ENTITY: \u001b[32mDresden\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.201:-0.201[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDresden[0.046]; August[0.045]; Germany[0.045]; eastern[0.045]; German[0.044]; ruled[0.043]; Juergen[0.042]; Cologne[0.042]; Tuesday[0.042]; month[0.042]; South[0.042]; South[0.042]; sent[0.041]; said[0.041]; Erfurt[0.041]; went[0.041]; Trittin[0.040]; farm[0.040]; order[0.040]; young[0.040]; group[0.040]; allow[0.039]; director[0.039]; forbidding[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = elephants ==> ENTITY: \u001b[32mElephant\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.190:0.190[0]; log p(e|m)= -0.154:-0.154[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nelephants[0.051]; elephants[0.051]; elephants[0.051]; Zoo[0.046]; Zoo[0.046]; zoo[0.046]; animals[0.045]; wild[0.045]; zoos[0.045]; African[0.043]; African[0.043]; herds[0.043]; baby[0.042]; baby[0.042]; said[0.041]; Africa[0.041]; South[0.040]; South[0.040]; National[0.040]; farm[0.040]; farm[0.040]; allow[0.039]; sent[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Erfurt\u001b[39m ==> ENTITY: \u001b[32mErfurt\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.045]; Germany[0.045]; Germany[0.045]; Cologne[0.044]; August[0.044]; German[0.044]; Dresden[0.043]; Dresden[0.043]; Juergen[0.043]; administrative[0.042]; eastern[0.041]; early[0.041]; went[0.040]; month[0.040]; month[0.040]; Switzerland[0.040]; Tuesday[0.040]; said[0.040]; said[0.040]; hopes[0.040]; sent[0.040]; temporary[0.040]; director[0.040]; order[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Switzerland ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; week[0.043]; month[0.043]; month[0.043]; federal[0.042]; state[0.042]; judge[0.042]; judge[0.042]; National[0.041]; August[0.041]; Tuesday[0.041]; South[0.041]; South[0.041]; South[0.041]; eastern[0.040]; group[0.040]; addition[0.040]; said[0.040]; filed[0.040]; court[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Africa ==> ENTITY: \u001b[32mAfrica\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrican[0.046]; African[0.046]; South[0.045]; South[0.045]; eastern[0.044]; month[0.043]; month[0.043]; National[0.042]; ruled[0.041]; sent[0.041]; Germany[0.041]; elephants[0.041]; elephants[0.041]; elephants[0.041]; elephants[0.041]; hopes[0.041]; German[0.040]; granted[0.040]; August[0.040]; captured[0.040]; group[0.040]; zoos[0.040]; said[0.040]; Tuesday[0.040]; \t\n\r [============================>.................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 459/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19981130_0743.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.056]; Spanish[0.049]; Portugal[0.049]; Portugal[0.049]; Portugal[0.049]; Jose[0.047]; Antonio[0.046]; countries[0.045]; Aznar[0.045]; Monday[0.045]; Albufeira[0.044]; historic[0.044]; Iberian[0.044]; governments[0.044]; peninsula[0.044]; south[0.044]; said[0.043]; said[0.043]; supplies[0.043]; reach[0.042]; called[0.042]; reached[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = volume ==> ENTITY: \u001b[32mVolume\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.101:0.101[0]; log p(e|m)= -0.257:-0.257[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwide[0.048]; country[0.043]; ranging[0.042]; links[0.042]; countries[0.042]; countries[0.042]; countries[0.042]; peninsula[0.042]; peninsula[0.042]; constant[0.042]; south[0.042]; water[0.042]; water[0.042]; water[0.042]; reach[0.041]; miles[0.041]; kilometers[0.041]; historic[0.041]; neighboring[0.041]; said[0.041]; said[0.041]; governments[0.040]; governments[0.040]; common[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mdrug trafficking\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMexican Drug War\u001b[39m <---> \u001b[32mIllegal drug trade\u001b[39m\t\nSCORES: global= 0.262:0.254[\u001b[31m0.007\u001b[39m]; local()= 0.163:0.131[\u001b[31m0.033\u001b[39m]; log p(e|m)= 0.000:-0.031[\u001b[31m0.031\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.047]; border[0.044]; border[0.044]; countries[0.044]; illegal[0.043]; governments[0.043]; neighboring[0.043]; campaign[0.042]; provided[0.042]; immigration[0.042]; links[0.042]; continued[0.042]; nations[0.041]; threat[0.041]; cooperation[0.041]; mainly[0.041]; ties[0.040]; reiterated[0.040]; curb[0.040]; Cabinet[0.040]; funds[0.040]; agreed[0.039]; projects[0.039]; sources[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jose Maria Aznar\u001b[39m ==> ENTITY: \u001b[32mJosé María Aznar\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.048]; Antonio[0.044]; Spanish[0.044]; Spanish[0.044]; reiterated[0.043]; country[0.042]; Portugal[0.042]; Portugal[0.042]; Portugal[0.042]; countries[0.041]; countries[0.041]; countries[0.041]; European[0.041]; European[0.041]; Minister[0.041]; governments[0.041]; governments[0.041]; political[0.041]; said[0.041]; said[0.041]; Prime[0.040]; Guterres[0.040]; summit[0.040]; summit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Antonio Guterres\u001b[39m ==> ENTITY: \u001b[32mAntónio Guterres\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npolitical[0.044]; Minister[0.044]; Portugal[0.043]; Portugal[0.043]; Portugal[0.043]; governments[0.043]; Spain[0.042]; Spain[0.042]; said[0.042]; said[0.042]; Lisbon[0.042]; Jose[0.042]; Prime[0.041]; countries[0.041]; countries[0.041]; countries[0.041]; management[0.040]; European[0.040]; relations[0.040]; volume[0.040]; resources[0.040]; worked[0.040]; ties[0.040]; links[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Water supply\u001b[39m ==> ENTITY: \u001b[32mWater supply\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwater[0.049]; supply[0.043]; supply[0.043]; south[0.043]; governments[0.043]; governments[0.043]; projects[0.042]; Spanish[0.042]; Spanish[0.042]; rivers[0.041]; rivers[0.041]; Jose[0.041]; funds[0.041]; road[0.041]; provided[0.041]; plan[0.041]; development[0.041]; peninsula[0.040]; peninsula[0.040]; miles[0.040]; country[0.040]; volume[0.040]; historic[0.040]; southwest[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portugal ==> ENTITY: \u001b[32mPortugal\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.197:0.197[0]; log p(e|m)= -0.309:-0.309[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.049]; Spain[0.049]; Portugal[0.048]; Portugal[0.048]; Spanish[0.046]; countries[0.043]; Iberian[0.043]; historic[0.042]; Lisbon[0.042]; governments[0.041]; Albufeira[0.041]; Jose[0.041]; peninsula[0.041]; modern[0.040]; south[0.039]; Maria[0.039]; Guterres[0.039]; Antonio[0.039]; Monday[0.038]; resources[0.038]; said[0.038]; said[0.038]; Aznar[0.038]; common[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iberian peninsula\u001b[39m ==> ENTITY: \u001b[32mIberian Peninsula\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.049]; Spain[0.049]; Spanish[0.046]; Iberian[0.046]; Portugal[0.043]; Portugal[0.043]; Portugal[0.043]; countries[0.042]; countries[0.042]; modern[0.042]; historic[0.042]; south[0.042]; Jose[0.041]; Antonio[0.040]; reached[0.040]; miles[0.040]; kilometers[0.039]; called[0.039]; reach[0.039]; wide[0.039]; Monday[0.039]; common[0.039]; rivers[0.039]; wrapping[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mhydroelectric\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHydropower\u001b[39m <---> \u001b[32mHydroelectricity\u001b[39m\t\nSCORES: global= 0.271:0.269[\u001b[31m0.001\u001b[39m]; local()= 0.184:0.177[\u001b[31m0.007\u001b[39m]; log p(e|m)= 0.000:-0.021[\u001b[31m0.021\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwater[0.046]; supply[0.045]; supply[0.045]; supply[0.045]; Water[0.043]; country[0.043]; rivers[0.043]; rivers[0.043]; links[0.041]; links[0.041]; neighboring[0.041]; development[0.041]; projects[0.041]; countries[0.041]; countries[0.041]; countries[0.041]; countries[0.041]; sources[0.040]; produced[0.040]; management[0.040]; plan[0.040]; plan[0.040]; nations[0.040]; build[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lisbon ==> ENTITY: \u001b[32mLisbon\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.098:-0.098[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPortugal[0.048]; Portugal[0.048]; Portugal[0.048]; Portugal[0.048]; Spain[0.046]; Spain[0.046]; Albufeira[0.044]; historic[0.043]; Jose[0.043]; miles[0.043]; plan[0.043]; governments[0.043]; plans[0.043]; south[0.042]; Water[0.042]; Antonio[0.042]; agreement[0.042]; agreement[0.042]; agreement[0.042]; agreement[0.042]; border[0.041]; southwest[0.041]; management[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = water resources ==> ENTITY: \u001b[32mWater resources\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.185:0.185[0]; log p(e|m)= -0.022:-0.022[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwater[0.050]; water[0.050]; rivers[0.046]; governments[0.044]; supply[0.044]; management[0.042]; efficient[0.042]; reached[0.041]; ranging[0.041]; monitor[0.041]; countries[0.041]; countries[0.041]; wide[0.041]; supplies[0.041]; historic[0.040]; plan[0.040]; modern[0.040]; miles[0.040]; common[0.040]; volume[0.040]; south[0.040]; reach[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = European Union\u001b[39m ==> ENTITY: \u001b[32mEuropean Union\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.006:-0.006[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuropean[0.046]; cooperation[0.043]; countries[0.043]; countries[0.043]; countries[0.043]; trade[0.043]; governments[0.042]; border[0.042]; border[0.042]; mainly[0.042]; development[0.042]; ties[0.041]; country[0.041]; nations[0.041]; summit[0.041]; decision[0.041]; relations[0.041]; agreement[0.040]; neighboring[0.040]; links[0.040]; links[0.040]; agreed[0.040]; supply[0.040]; supply[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Prime Minister ==> ENTITY: \u001b[32mPrime minister\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.112:0.112[0]; log p(e|m)= -1.435:-1.435[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpanish[0.045]; Spanish[0.045]; Aznar[0.045]; Spain[0.045]; Spain[0.045]; Premier[0.045]; political[0.045]; Jose[0.045]; premiers[0.045]; European[0.044]; Antonio[0.044]; said[0.043]; said[0.043]; Monday[0.042]; countries[0.042]; countries[0.042]; countries[0.042]; Portugal[0.042]; Portugal[0.042]; Portugal[0.042]; governments[0.042]; plans[0.042]; day[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = illegal immigration ==> ENTITY: \u001b[32mIllegal immigration\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuropean[0.046]; country[0.045]; border[0.045]; border[0.045]; campaign[0.044]; countries[0.044]; trafficking[0.044]; neighboring[0.043]; governments[0.042]; Portugal[0.042]; funds[0.041]; mainly[0.041]; cooperation[0.041]; reiterated[0.040]; increase[0.040]; drug[0.040]; continued[0.040]; wealthier[0.040]; Union[0.040]; nations[0.040]; political[0.040]; provided[0.040]; plan[0.040]; threat[0.039]; \t\n\r [============================>.................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 473/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19990827_0184.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = terminal ==> ENTITY: \u001b[32mAirport terminal\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.145:0.145[0]; log p(e|m)= -1.435:-1.435[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nterminal[0.048]; Terminal[0.046]; Airport[0.046]; passengers[0.042]; flights[0.042]; flights[0.042]; officials[0.042]; shut[0.042]; normal[0.042]; delayed[0.042]; delayed[0.042]; carrier[0.041]; passenger[0.041]; schedule[0.041]; crowd[0.041]; airline[0.040]; security[0.040]; past[0.040]; Airlines[0.040]; United[0.040]; United[0.040]; nationwide[0.040]; today[0.040]; shutdown[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = O'Hare International Airport\u001b[39m ==> ENTITY: \u001b[32mO'Hare International Airport\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nflights[0.044]; flights[0.044]; canceled[0.043]; canceled[0.043]; Chicago[0.043]; Airlines[0.043]; delayed[0.043]; delayed[0.043]; Terminal[0.042]; terminal[0.042]; terminal[0.042]; airline[0.042]; Hare[0.042]; area[0.041]; United[0.041]; United[0.041]; Thursday[0.041]; carrier[0.041]; passenger[0.041]; concourse[0.040]; passengers[0.040]; past[0.040]; baggage[0.040]; schedule[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36msecurity\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSecurity\u001b[39m <---> \u001b[32mAirport security\u001b[39m\t\nSCORES: global= 0.248:0.231[\u001b[31m0.017\u001b[39m]; local()= 0.103:0.150[\u001b[32m0.047\u001b[39m]; log p(e|m)= -0.846:-6.215[\u001b[31m5.368\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSecurity[0.046]; Airport[0.046]; officers[0.045]; police[0.043]; checkpoint[0.043]; officials[0.043]; terminal[0.042]; terminal[0.042]; airline[0.042]; nationwide[0.042]; passenger[0.042]; passengers[0.041]; flights[0.041]; flights[0.041]; Terminal[0.041]; authorities[0.041]; numerous[0.041]; International[0.041]; Airlines[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; today[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = baggage claim\u001b[39m ==> ENTITY: \u001b[32mBaggage claim\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nterminal[0.048]; terminal[0.048]; concourse[0.046]; concourse[0.046]; Airport[0.044]; flights[0.042]; flights[0.042]; flights[0.042]; flights[0.042]; passengers[0.042]; carrier[0.042]; airline[0.041]; area[0.041]; leaving[0.041]; gates[0.040]; schedule[0.040]; landed[0.040]; adding[0.040]; Thursday[0.039]; cleared[0.039]; Kristina[0.039]; canceled[0.039]; canceled[0.039]; International[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = concourse\u001b[39m ==> ENTITY: \u001b[32mConcourse\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nconcourse[0.053]; terminal[0.048]; terminal[0.048]; area[0.043]; today[0.042]; gates[0.042]; Airport[0.041]; flights[0.041]; flights[0.041]; flights[0.041]; flights[0.041]; schedule[0.041]; early[0.040]; numerous[0.040]; shutdown[0.040]; Thursday[0.040]; way[0.040]; staircase[0.040]; staircase[0.040]; baggage[0.040]; leading[0.040]; canceled[0.040]; canceled[0.040]; Chicago[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = United Airlines\u001b[39m ==> ENTITY: \u001b[32mUnited Airlines\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairline[0.047]; flights[0.046]; flights[0.046]; carrier[0.046]; Airport[0.044]; United[0.043]; Thursday[0.043]; canceled[0.042]; canceled[0.042]; spokeswoman[0.042]; today[0.041]; passengers[0.041]; Terminal[0.041]; delayed[0.041]; delayed[0.041]; International[0.040]; passenger[0.040]; CHICAGO[0.040]; officials[0.040]; past[0.040]; nationwide[0.040]; schedule[0.039]; terminal[0.039]; terminal[0.039]; \t\n\r [=============================>................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 479/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19981106_0920.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mhydrogen\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHydrogen vehicle\u001b[39m <---> \u001b[32mHydrogen\u001b[39m\t\nSCORES: global= 0.251:0.247[\u001b[31m0.004\u001b[39m]; local()= 0.182:0.125[\u001b[31m0.057\u001b[39m]; log p(e|m)= -1.317:-0.068[\u001b[32m1.248\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhydrogen[0.046]; technology[0.045]; plant[0.045]; plant[0.045]; plant[0.045]; plant[0.045]; electricity[0.044]; use[0.044]; power[0.044]; power[0.044]; produce[0.043]; Energy[0.043]; clean[0.043]; clean[0.043]; generate[0.043]; Hydro[0.043]; company[0.042]; company[0.042]; company[0.042]; company[0.042]; climate[0.042]; burning[0.042]; burning[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = natural gas ==> ENTITY: \u001b[32mNatural gas\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.257:0.257[0]; log p(e|m)= -0.019:-0.019[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngas[0.047]; gas[0.047]; gas[0.047]; oil[0.043]; offshore[0.042]; petroleum[0.042]; hydrogen[0.042]; hydrogen[0.042]; hydrogen[0.042]; plant[0.041]; plant[0.041]; plant[0.041]; plant[0.041]; hydroelectric[0.041]; plants[0.041]; plants[0.041]; electricity[0.040]; electricity[0.040]; electricity[0.040]; production[0.040]; produce[0.040]; produce[0.040]; Energy[0.040]; water[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oslo ==> ENTITY: \u001b[32mOslo\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.107:0.107[0]; log p(e|m)= -0.036:-0.036[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.049]; miles[0.044]; country[0.044]; company[0.043]; west[0.043]; Norsk[0.043]; build[0.043]; second[0.043]; percent[0.042]; kilometers[0.041]; vast[0.041]; water[0.041]; said[0.041]; world[0.041]; offshore[0.041]; Hydro[0.040]; island[0.040]; air[0.040]; gas[0.040]; gas[0.040]; gas[0.040]; plans[0.040]; comes[0.040]; lower[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = petroleum ==> ENTITY: \u001b[32mPetroleum\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.220:0.220[0]; log p(e|m)= -0.015:-0.015[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \noil[0.047]; gas[0.045]; gas[0.045]; gas[0.045]; gas[0.045]; offshore[0.043]; production[0.041]; company[0.041]; company[0.041]; company[0.041]; company[0.041]; construction[0.041]; plant[0.041]; plant[0.041]; plant[0.041]; plant[0.041]; plants[0.041]; plants[0.041]; fields[0.041]; technology[0.041]; use[0.040]; major[0.040]; largest[0.040]; government[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = state-owned company\u001b[39m ==> ENTITY: \u001b[32mGovernment-owned corporation\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngovernment[0.045]; Norsk[0.044]; Norsk[0.044]; company[0.043]; company[0.043]; company[0.043]; Hydro[0.043]; Hydro[0.043]; agencies[0.043]; Energy[0.042]; Resources[0.042]; petroleum[0.042]; plans[0.041]; plans[0.041]; Directorate[0.041]; major[0.040]; construction[0.040]; power[0.040]; power[0.040]; build[0.040]; build[0.040]; electricity[0.040]; electricity[0.040]; gas[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mcarbon dioxide\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mGreenhouse gas\u001b[39m <---> \u001b[32mCarbon dioxide\u001b[39m\t\nSCORES: global= 0.260:0.259[\u001b[31m0.001\u001b[39m]; local()= 0.225:0.229[\u001b[32m0.004\u001b[39m]; log p(e|m)= -0.691:-0.023[\u001b[32m0.668\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nemissions[0.044]; nitrogen[0.044]; carbon[0.044]; water[0.043]; amounts[0.043]; gas[0.042]; gas[0.042]; gas[0.042]; gas[0.042]; use[0.042]; percent[0.041]; plants[0.041]; plants[0.041]; hydrogen[0.041]; hydrogen[0.041]; hydrogen[0.041]; plant[0.041]; plant[0.041]; plant[0.041]; plant[0.041]; plant[0.041]; electricity[0.041]; electricity[0.041]; petroleum[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Norsk Hydro ASA\u001b[39m ==> ENTITY: \u001b[32mNorsk Hydro\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.218:0.218[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncompany[0.045]; company[0.045]; company[0.045]; company[0.045]; Norwegian[0.044]; Norwegian[0.044]; construction[0.042]; Energy[0.042]; electricity[0.042]; owned[0.041]; says[0.041]; plant[0.041]; plant[0.041]; plant[0.041]; plant[0.041]; production[0.041]; announced[0.040]; announced[0.040]; government[0.040]; Directorate[0.040]; Water[0.040]; said[0.040]; build[0.040]; build[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = electricity ==> ENTITY: \u001b[32mElectricity\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.220:0.220[0]; log p(e|m)= -0.109:-0.109[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nelectricity[0.046]; gas[0.044]; Energy[0.044]; power[0.043]; power[0.043]; construction[0.042]; plant[0.042]; plant[0.042]; plant[0.042]; plant[0.042]; plant[0.042]; state[0.042]; generate[0.041]; Hydro[0.041]; emissions[0.041]; company[0.041]; company[0.041]; company[0.041]; company[0.041]; technology[0.041]; Water[0.040]; agencies[0.040]; use[0.040]; climate[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.050]; Norsk[0.048]; Norsk[0.048]; second[0.048]; west[0.047]; Oslo[0.047]; world[0.046]; vast[0.046]; island[0.046]; early[0.045]; major[0.045]; percent[0.045]; offshore[0.044]; miles[0.044]; announced[0.044]; comes[0.044]; allow[0.044]; company[0.044]; company[0.044]; company[0.044]; kilometers[0.043]; use[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Norwegian Water Resources and Energy Directorate\u001b[39m ==> ENTITY: \u001b[32mNorwegian Water Resources and Energy Directorate\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorsk[0.047]; Norsk[0.047]; Norwegian[0.047]; electricity[0.046]; electricity[0.046]; agencies[0.045]; petroleum[0.044]; production[0.044]; gas[0.043]; company[0.043]; company[0.043]; company[0.043]; company[0.043]; ASA[0.042]; announced[0.042]; announced[0.042]; approvals[0.042]; government[0.042]; construction[0.042]; use[0.042]; state[0.042]; produce[0.041]; approval[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = power plant ==> ENTITY: \u001b[32mPower station\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.212:0.212[0]; log p(e|m)= -0.080:-0.080[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplant[0.046]; plant[0.046]; plant[0.046]; plant[0.046]; electricity[0.044]; electricity[0.044]; Hydro[0.042]; build[0.042]; build[0.042]; power[0.042]; Water[0.041]; construction[0.041]; owned[0.041]; Energy[0.041]; company[0.040]; company[0.040]; company[0.040]; company[0.040]; Norsk[0.040]; plans[0.039]; plans[0.039]; announced[0.039]; announced[0.039]; generate[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = hydroelectric plants ==> ENTITY: \u001b[32mHydroelectricity\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.194:0.194[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwater[0.045]; electricity[0.045]; electricity[0.045]; plant[0.043]; plant[0.043]; plant[0.043]; Norsk[0.042]; Norsk[0.042]; petroleum[0.042]; Hydro[0.041]; Hydro[0.041]; produce[0.041]; plants[0.041]; fields[0.041]; gas[0.041]; gas[0.041]; gas[0.041]; gas[0.041]; use[0.041]; generate[0.041]; country[0.041]; oil[0.040]; build[0.040]; company[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36memissions\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mGreenhouse gas\u001b[39m <---> \u001b[32mAir pollution\u001b[39m\t\nSCORES: global= 0.253:0.249[\u001b[31m0.004\u001b[39m]; local()= 0.181:0.166[\u001b[31m0.015\u001b[39m]; log p(e|m)= -1.317:-1.402[\u001b[31m0.086\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclimate[0.046]; gas[0.045]; electricity[0.042]; electricity[0.042]; state[0.042]; use[0.042]; plant[0.042]; plant[0.042]; plant[0.042]; plant[0.042]; plant[0.042]; government[0.042]; Energy[0.041]; allow[0.041]; petroleum[0.041]; concern[0.041]; company[0.041]; company[0.041]; company[0.041]; company[0.041]; natural[0.040]; burning[0.040]; burning[0.040]; generate[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = nitrogen ==> ENTITY: \u001b[32mNitrogen\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.016:-0.016[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwater[0.044]; plants[0.044]; plants[0.044]; plant[0.043]; plant[0.043]; hydrogen[0.043]; air[0.042]; petroleum[0.042]; natural[0.042]; natural[0.042]; natural[0.042]; dioxide[0.042]; dioxide[0.042]; produce[0.041]; amounts[0.041]; carbon[0.041]; carbon[0.041]; gas[0.041]; gas[0.041]; gas[0.041]; gas[0.041]; oil[0.041]; emissions[0.040]; fields[0.040]; \t\n\r [==============================>...............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 493/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19981109_0152.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iran ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.179:0.179[0]; log p(e|m)= -0.138:-0.138[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.049]; Iran[0.049]; Iran[0.049]; Iran[0.049]; Iran[0.049]; saying[0.043]; Kharrazi[0.042]; government[0.041]; reported[0.041]; Europe[0.040]; ambassador[0.040]; quoted[0.040]; said[0.040]; organizations[0.040]; Foreign[0.039]; International[0.039]; assistance[0.039]; Radio[0.039]; protest[0.039]; protest[0.039]; internal[0.039]; Monday[0.038]; global[0.038]; English[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Prague ==> ENTITY: \u001b[32mPrague\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.053:-0.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPrague[0.048]; Germany[0.043]; Czech[0.043]; Czech[0.043]; Soviet[0.042]; Tuesday[0.042]; Munich[0.042]; set[0.042]; Europe[0.042]; Europe[0.042]; Friday[0.041]; Monday[0.041]; English[0.041]; August[0.041]; ambassador[0.041]; began[0.041]; began[0.041]; protest[0.041]; protest[0.041]; sponsored[0.041]; Radio[0.041]; Radio[0.041]; Iraq[0.040]; Iraq[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Communist ==> ENTITY: \u001b[32mCommunism\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.393:-0.393[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.046]; communism[0.045]; Europe[0.045]; countries[0.045]; American[0.044]; government[0.043]; economic[0.043]; political[0.043]; Germany[0.042]; Czech[0.042]; Prague[0.041]; Prague[0.041]; cooperation[0.041]; following[0.040]; Iraq[0.040]; Iraq[0.040]; service[0.040]; August[0.040]; moved[0.040]; uncensored[0.040]; West[0.040]; news[0.040]; Iran[0.039]; Iran[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Soviet ==> ENTITY: \u001b[32mSoviet Union\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.140:0.140[0]; log p(e|m)= -0.098:-0.098[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.043]; Iran[0.043]; communism[0.043]; American[0.043]; Communist[0.043]; economic[0.042]; War[0.042]; countries[0.042]; Iraq[0.042]; Iraq[0.042]; Europe[0.042]; language[0.041]; Germany[0.041]; program[0.041]; government[0.041]; government[0.041]; ambassador[0.041]; West[0.041]; cooperation[0.041]; Prague[0.041]; Prague[0.041]; Prague[0.041]; following[0.041]; August[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.137:0.137[0]; log p(e|m)= -1.207:-1.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlanguage[0.046]; language[0.046]; American[0.044]; English[0.044]; Prague[0.043]; Prague[0.043]; Czech[0.043]; Germany[0.043]; Soviet[0.043]; Europe[0.042]; Europe[0.042]; ambassador[0.041]; internal[0.041]; service[0.041]; radio[0.040]; radio[0.040]; radio[0.040]; countries[0.040]; Radio[0.040]; Radio[0.040]; economic[0.040]; assistance[0.039]; August[0.039]; broadcasts[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = ICJ\u001b[39m ==> ENTITY: \u001b[32mInternational Court of Justice\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCourt[0.059]; tribunal[0.059]; saying[0.057]; Hague[0.057]; Hague[0.057]; Justice[0.056]; Iraq[0.055]; Iraq[0.055]; trial[0.055]; Foreign[0.055]; International[0.055]; political[0.055]; Europe[0.054]; Europe[0.054]; government[0.054]; government[0.054]; Netherlands[0.054]; filed[0.054]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iron Curtain\u001b[39m ==> ENTITY: \u001b[32mIron Curtain\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.047]; Europe[0.044]; Communist[0.043]; countries[0.043]; Germany[0.043]; American[0.043]; communism[0.042]; West[0.042]; cooperation[0.042]; Prague[0.041]; Prague[0.041]; War[0.041]; economic[0.041]; East[0.041]; broadcast[0.041]; government[0.041]; government[0.041]; political[0.041]; period[0.041]; Czech[0.041]; Czech[0.041]; service[0.041]; directed[0.040]; level[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.174:0.174[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.052]; government[0.044]; government[0.044]; cooperation[0.043]; countries[0.043]; American[0.042]; Iran[0.042]; Iran[0.042]; Iran[0.042]; Iran[0.042]; Iran[0.042]; War[0.041]; economic[0.041]; said[0.041]; Foreign[0.040]; assistance[0.040]; West[0.040]; saying[0.040]; protest[0.040]; Kharrazi[0.040]; following[0.040]; Europe[0.040]; political[0.040]; Germany[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.059:0.059[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlanguage[0.049]; English[0.048]; International[0.046]; Europe[0.044]; Court[0.044]; Foreign[0.042]; reported[0.042]; ambassador[0.042]; Radio[0.042]; radio[0.041]; radio[0.041]; Czech[0.041]; Netherlands[0.040]; saying[0.040]; say[0.040]; say[0.040]; set[0.040]; filed[0.040]; global[0.040]; Tuesday[0.040]; Monday[0.039]; tribunal[0.039]; internal[0.039]; complain[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Radio Free Europe\u001b[39m ==> ENTITY: \u001b[32mRadio Free Europe/Radio Liberty\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbroadcasts[0.045]; quoted[0.043]; said[0.043]; radio[0.043]; radio[0.043]; radio[0.043]; ambassador[0.042]; recalled[0.042]; Prague[0.042]; Prague[0.042]; Czech[0.042]; set[0.042]; American[0.042]; internal[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; International[0.041]; Monday[0.040]; Tuesday[0.040]; reported[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cold War\u001b[39m ==> ENTITY: \u001b[32mCold War\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.046]; countries[0.045]; Iraq[0.044]; Iraq[0.044]; economic[0.043]; American[0.043]; Iran[0.043]; Iran[0.043]; Communist[0.043]; communism[0.043]; Europe[0.042]; cooperation[0.042]; period[0.041]; government[0.041]; political[0.041]; Germany[0.040]; following[0.040]; West[0.040]; collapse[0.040]; service[0.040]; program[0.039]; weeks[0.039]; Prague[0.039]; Prague[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Farsi ==> ENTITY: \u001b[32mPersian language\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.171:-0.171[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlanguage[0.050]; language[0.050]; English[0.048]; broadcast[0.044]; Iran[0.044]; Iran[0.044]; Iran[0.044]; Iran[0.044]; Iran[0.044]; broadcasts[0.044]; American[0.043]; Iraq[0.043]; Iraq[0.043]; radio[0.042]; station[0.042]; station[0.042]; say[0.042]; say[0.042]; countries[0.041]; Radio[0.041]; West[0.041]; cooperation[0.041]; daily[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = the Netherlands ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.103:0.103[0]; log p(e|m)= -0.176:-0.176[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; Europe[0.046]; Europe[0.046]; day[0.043]; approved[0.042]; saying[0.042]; government[0.041]; government[0.041]; English[0.041]; Munich[0.041]; service[0.041]; ambassador[0.041]; August[0.041]; minute[0.041]; weeks[0.041]; affairs[0.041]; International[0.041]; level[0.041]; Hague[0.040]; Hague[0.040]; internal[0.040]; Czech[0.040]; Czech[0.040]; Tuesday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = the Hague ==> ENTITY: \u001b[32mThe Hague\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.070:-0.070[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.048]; Hague[0.045]; tribunal[0.044]; ambassador[0.044]; ICJ[0.043]; Court[0.042]; government[0.042]; Europe[0.042]; interfere[0.041]; political[0.041]; Monday[0.041]; filed[0.041]; Tuesday[0.041]; International[0.041]; saying[0.041]; cooperation[0.041]; economic[0.041]; protest[0.041]; protest[0.041]; Free[0.040]; organizations[0.040]; affairs[0.040]; Justice[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Czech government\u001b[39m ==> ENTITY: \u001b[32mPolitics of the Czech Republic\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.122:0.122[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nICJ[0.044]; government[0.044]; Czech[0.044]; political[0.044]; ambassador[0.043]; cooperation[0.043]; Prague[0.043]; Prague[0.043]; Prague[0.043]; Communist[0.042]; countries[0.042]; program[0.041]; organizations[0.041]; Germany[0.041]; Munich[0.041]; communism[0.041]; Cold[0.041]; Europe[0.040]; said[0.040]; Curtain[0.040]; protest[0.040]; Soviet[0.040]; Netherlands[0.040]; broadcasts[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kamal Kharrazi\u001b[39m ==> ENTITY: \u001b[32mKamal Kharazi\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.045]; Iran[0.045]; Iran[0.045]; Iran[0.045]; Iran[0.045]; Iran[0.045]; Iraq[0.045]; Foreign[0.043]; American[0.042]; ambassador[0.042]; affairs[0.042]; International[0.041]; Monday[0.041]; Minister[0.040]; weeks[0.040]; Tuesday[0.040]; reported[0.040]; Europe[0.040]; cooperation[0.039]; Daily[0.039]; English[0.039]; language[0.039]; political[0.039]; government[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Munich, Germany\u001b[39m ==> ENTITY: \u001b[32mMunich\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.122:0.122[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurope[0.044]; Prague[0.044]; Prague[0.044]; Prague[0.044]; American[0.043]; Czech[0.043]; Czech[0.043]; headquarters[0.042]; War[0.042]; communism[0.042]; service[0.041]; government[0.041]; government[0.041]; August[0.041]; Friday[0.041]; began[0.041]; began[0.041]; sponsored[0.041]; ambassador[0.041]; program[0.040]; Soviet[0.040]; following[0.040]; earlier[0.040]; economic[0.040]; \t\n\r [===============================>..............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 510/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19980713_0449.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = planes ==> ENTITY: \u001b[32mFixed-wing aircraft\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.126:0.126[0]; log p(e|m)= -1.058:-1.058[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nflying[0.047]; Germany[0.046]; takeoff[0.045]; landing[0.044]; German[0.044]; German[0.044]; passengers[0.043]; example[0.042]; carrier[0.042]; Cologne[0.041]; flights[0.041]; Monday[0.041]; international[0.041]; Frankfurt[0.041]; Frankfurt[0.041]; hub[0.040]; closer[0.040]; flag[0.040]; work[0.040]; main[0.040]; hopes[0.039]; instead[0.039]; short[0.039]; services[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cologne ==> ENTITY: \u001b[32mCologne\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.165:0.165[0]; log p(e|m)= -0.082:-0.082[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCologne[0.056]; Germany[0.046]; Nuremberg[0.045]; German[0.045]; German[0.045]; Saarbruecken[0.044]; Wuerzburg[0.043]; Frankfurt[0.042]; Frankfurt[0.042]; Frankfurt[0.042]; Duesseldorf[0.042]; Juergen[0.040]; Bonn[0.040]; Lufthansa[0.039]; Lufthansa[0.039]; Lufthansa[0.039]; Lufthansa[0.039]; Lufthansa[0.039]; Lufthansa[0.039]; Lufthansa[0.039]; Deutsche[0.038]; trains[0.037]; train[0.037]; train[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = SAS ==> ENTITY: \u001b[32mScandinavian Airlines\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.188:0.188[0]; log p(e|m)= -2.364:-2.364[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairline[0.045]; flights[0.044]; flights[0.044]; flight[0.044]; use[0.043]; service[0.043]; service[0.043]; service[0.043]; Lufthansa[0.042]; Lufthansa[0.042]; Lufthansa[0.042]; Lufthansa[0.042]; schedules[0.042]; security[0.042]; international[0.041]; partners[0.041]; airport[0.041]; landing[0.040]; slots[0.040]; September[0.040]; plane[0.040]; time[0.039]; flying[0.039]; example[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = landing slots\u001b[39m ==> ENTITY: \u001b[32mLanding slot\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nflights[0.046]; airline[0.045]; flight[0.045]; carrier[0.043]; hub[0.042]; schedules[0.042]; Lufthansa[0.042]; Lufthansa[0.042]; Lufthansa[0.042]; Lufthansa[0.042]; Lufthansa[0.042]; Lufthansa[0.042]; takeoff[0.041]; planes[0.041]; Wuerzburg[0.041]; Frankfurt[0.041]; Frankfurt[0.041]; Frankfurt[0.041]; Deutsche[0.041]; Saarbruecken[0.041]; Juergen[0.040]; service[0.040]; service[0.040]; service[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = main hub ==> ENTITY: \u001b[32mAirline hub\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.171:0.171[0]; log p(e|m)= -0.644:-0.644[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nflights[0.050]; international[0.048]; domestic[0.047]; service[0.047]; service[0.047]; carrier[0.046]; flight[0.046]; schedules[0.046]; travel[0.045]; travel[0.045]; services[0.045]; Deutsche[0.045]; Saarbruecken[0.045]; planes[0.045]; station[0.045]; Frankfurt[0.044]; Frankfurt[0.044]; Frankfurt[0.044]; Juergen[0.044]; Monday[0.044]; cities[0.044]; Bahn[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Deutsche Bahn\u001b[39m ==> ENTITY: \u001b[32mDeutsche Bahn\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nservice[0.044]; service[0.044]; service[0.044]; Frankfurt[0.043]; Frankfurt[0.043]; Frankfurt[0.043]; railway[0.042]; railway[0.042]; Lufthansa[0.042]; Lufthansa[0.042]; Lufthansa[0.042]; Lufthansa[0.042]; Lufthansa[0.042]; Lufthansa[0.042]; trains[0.042]; services[0.041]; hub[0.041]; Wuerzburg[0.041]; Germany[0.040]; Saarbruecken[0.040]; train[0.040]; train[0.040]; Duesseldorf[0.040]; Juergen[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = German railway\u001b[39m ==> ENTITY: \u001b[32mRail transport in Germany\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrains[0.046]; railway[0.045]; Germany[0.043]; Lufthansa[0.043]; Lufthansa[0.043]; Lufthansa[0.043]; Lufthansa[0.043]; train[0.043]; Frankfurt[0.042]; Frankfurt[0.042]; Juergen[0.042]; planes[0.042]; services[0.042]; travel[0.041]; travel[0.041]; closer[0.041]; carrier[0.041]; German[0.040]; schedules[0.040]; hub[0.040]; main[0.040]; Cologne[0.040]; cooperation[0.040]; domestic[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.095:0.095[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninternational[0.047]; German[0.045]; German[0.045]; Cologne[0.044]; goal[0.044]; Frankfurt[0.043]; Frankfurt[0.043]; Juergen[0.042]; Monday[0.042]; hopes[0.042]; domestic[0.041]; lucrative[0.041]; travel[0.040]; travel[0.040]; boss[0.040]; flag[0.040]; Weber[0.040]; short[0.040]; cooperation[0.040]; planes[0.040]; signed[0.040]; schedules[0.040]; example[0.040]; carrier[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBonn\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mCologne Bonn Airport\u001b[39m <---> \u001b[32mBonn\u001b[39m\t\nSCORES: global= 0.268:0.263[\u001b[31m0.005\u001b[39m]; local()= 0.245:0.111[\u001b[31m0.134\u001b[39m]; log p(e|m)= -2.198:0.000[\u001b[32m2.198\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairport[0.043]; Frankfurt[0.043]; Frankfurt[0.043]; Frankfurt[0.043]; hub[0.042]; Lufthansa[0.042]; Lufthansa[0.042]; Lufthansa[0.042]; Lufthansa[0.042]; Lufthansa[0.042]; railway[0.042]; railway[0.042]; flights[0.042]; flights[0.042]; train[0.041]; train[0.041]; train[0.041]; Saarbruecken[0.041]; Cologne[0.041]; Cologne[0.041]; takeoff[0.040]; airline[0.040]; Wuerzburg[0.040]; flight[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Frankfurt ==> ENTITY: \u001b[32mFrankfurt\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.169:0.169[0]; log p(e|m)= -0.070:-0.070[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrankfurt[0.045]; Frankfurt[0.045]; Germany[0.043]; railway[0.043]; railway[0.043]; Bahn[0.042]; Cologne[0.042]; Cologne[0.042]; trains[0.042]; train[0.041]; train[0.041]; service[0.041]; service[0.041]; Saarbruecken[0.041]; travel[0.041]; travel[0.041]; hub[0.041]; Lufthansa[0.041]; Lufthansa[0.041]; Lufthansa[0.041]; Lufthansa[0.041]; Lufthansa[0.041]; Lufthansa[0.041]; passengers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = trains ==> ENTITY: \u001b[32mTrain\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.198:-0.198[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrain[0.053]; railway[0.050]; railway[0.050]; international[0.049]; schedules[0.049]; travel[0.048]; travel[0.048]; passengers[0.048]; services[0.048]; Germany[0.048]; hub[0.048]; domestic[0.047]; carrier[0.047]; shift[0.046]; slots[0.046]; flights[0.046]; Monday[0.046]; work[0.046]; deal[0.045]; flying[0.045]; lucrative[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wuerzburg\u001b[39m ==> ENTITY: \u001b[32mWürzburg\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.129:0.129[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaarbruecken[0.044]; Cologne[0.043]; Cologne[0.043]; Frankfurt[0.043]; Frankfurt[0.043]; Frankfurt[0.043]; Bonn[0.043]; Deutsche[0.042]; Deutsche[0.042]; Juergen[0.041]; said[0.041]; said[0.041]; said[0.041]; Nuremberg[0.041]; Bahn[0.041]; Bahn[0.041]; Lufthansa[0.041]; Lufthansa[0.041]; Lufthansa[0.041]; Lufthansa[0.041]; Lufthansa[0.041]; Johannes[0.041]; international[0.040]; Duesseldorf[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lufthansa\u001b[39m ==> ENTITY: \u001b[32mLufthansa\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.255:0.255[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLufthansa[0.047]; Lufthansa[0.047]; Lufthansa[0.047]; Lufthansa[0.047]; Lufthansa[0.047]; carrier[0.044]; flights[0.043]; Frankfurt[0.042]; Frankfurt[0.042]; planes[0.041]; takeoff[0.041]; hub[0.040]; passengers[0.040]; flying[0.040]; international[0.040]; Germany[0.040]; German[0.039]; German[0.039]; Deutsche[0.039]; luggage[0.039]; Monday[0.039]; services[0.039]; Juergen[0.039]; Bahn[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United ==> ENTITY: \u001b[32mUnited Airlines\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.201:0.201[0]; log p(e|m)= -1.580:-1.580[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairline[0.046]; flight[0.046]; flights[0.045]; flights[0.045]; international[0.044]; flying[0.043]; service[0.042]; service[0.042]; service[0.042]; airport[0.041]; takeoff[0.041]; schedules[0.040]; passengers[0.040]; lucrative[0.040]; Lufthansa[0.040]; Lufthansa[0.040]; Lufthansa[0.040]; Lufthansa[0.040]; partners[0.040]; deal[0.040]; plane[0.040]; landing[0.040]; cities[0.040]; group[0.039]; \t\n\r [================================>.............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 524/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19981026_0096.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = apartheid ==> ENTITY: \u001b[32mApartheid\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.142:0.142[0]; log p(e|m)= -0.073:-0.073[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrican[0.044]; South[0.044]; South[0.044]; governments[0.044]; condemn[0.042]; white[0.042]; white[0.042]; Reconciliation[0.042]; services[0.042]; security[0.042]; Africa[0.042]; Congress[0.042]; rights[0.041]; accused[0.041]; acknowledged[0.041]; release[0.041]; National[0.041]; previously[0.041]; said[0.041]; report[0.040]; report[0.040]; ANC[0.040]; ANC[0.040]; torture[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = white rule\u001b[39m ==> ENTITY: \u001b[32mDominant minority\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrights[0.044]; rights[0.044]; acknowledged[0.043]; led[0.043]; National[0.043]; security[0.043]; African[0.043]; African[0.043]; white[0.042]; governments[0.042]; Reconciliation[0.041]; Africa[0.041]; spokesman[0.041]; power[0.041]; apartheid[0.041]; apartheid[0.041]; said[0.041]; said[0.041]; document[0.041]; Report[0.041]; state[0.041]; previously[0.041]; condemns[0.040]; Monday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Truth and Reconciliation Commission ==> ENTITY: \u001b[32mTruth and Reconciliation Commission (South Africa)\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.449:-0.449[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncommission[0.044]; acknowledged[0.044]; abuses[0.043]; abuses[0.043]; governments[0.043]; truth[0.042]; security[0.042]; preliminary[0.042]; responsible[0.041]; report[0.041]; report[0.041]; report[0.041]; panel[0.041]; panel[0.041]; attacks[0.041]; rights[0.041]; rights[0.041]; African[0.041]; African[0.041]; National[0.041]; condemn[0.041]; apartheid[0.041]; apartheid[0.041]; condemns[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = bomb ==> ENTITY: \u001b[32mBomb\u001b[39m\t\nSCORES: global= 0.241:0.241[0]; local()= 0.073:0.073[0]; log p(e|m)= -0.208:-0.208[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecurity[0.045]; attacks[0.045]; condemn[0.043]; Thursday[0.043]; Monday[0.043]; said[0.043]; said[0.043]; previously[0.042]; services[0.042]; National[0.042]; acknowledged[0.042]; document[0.041]; report[0.041]; report[0.041]; Reconciliation[0.041]; condemns[0.041]; investigation[0.041]; place[0.041]; accused[0.041]; African[0.040]; African[0.040]; South[0.040]; South[0.040]; told[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Africa ==> ENTITY: \u001b[32mSouth Africa\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.087:0.087[0]; log p(e|m)= -0.263:-0.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrican[0.047]; final[0.044]; rights[0.044]; Monday[0.043]; Reconciliation[0.043]; Thursday[0.043]; acknowledged[0.043]; National[0.042]; previously[0.042]; apartheid[0.042]; place[0.041]; Commission[0.041]; expected[0.041]; Report[0.041]; panel[0.041]; panel[0.041]; abuses[0.041]; ANC[0.041]; ANC[0.041]; broadcaster[0.040]; release[0.040]; Truth[0.040]; told[0.040]; right[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = human rights\u001b[39m ==> ENTITY: \u001b[32mHuman rights\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nabuses[0.045]; abuses[0.045]; torture[0.043]; security[0.043]; violations[0.043]; violations[0.043]; governments[0.043]; report[0.043]; report[0.043]; acknowledged[0.043]; condemn[0.042]; attacks[0.042]; Africa[0.041]; release[0.041]; Report[0.041]; services[0.040]; brutal[0.040]; National[0.040]; state[0.040]; document[0.040]; Reconciliation[0.040]; said[0.040]; said[0.040]; accused[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mcommissioners\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEuropean Commission\u001b[39m <---> \u001b[32mCommissioner\u001b[39m\t\nSCORES: global= 0.250:0.242[\u001b[31m0.008\u001b[39m]; local()= 0.154:0.116[\u001b[31m0.038\u001b[39m]; log p(e|m)= -1.884:-0.184[\u001b[32m1.700\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCommission[0.046]; request[0.044]; commission[0.043]; letter[0.043]; document[0.043]; report[0.042]; report[0.042]; South[0.042]; refused[0.042]; meeting[0.042]; month[0.042]; sent[0.042]; border[0.041]; spokesman[0.041]; officials[0.041]; earlier[0.041]; military[0.040]; panel[0.040]; intention[0.040]; said[0.040]; said[0.040]; says[0.040]; responsible[0.040]; news[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = broadcaster ==> ENTITY: \u001b[32mBroadcasting\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.116:0.116[0]; log p(e|m)= -1.378:-1.378[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nradio[0.050]; Broadcasting[0.046]; acknowledged[0.044]; services[0.043]; told[0.043]; Monday[0.042]; Thursday[0.042]; National[0.042]; previously[0.042]; said[0.041]; said[0.041]; panel[0.041]; panel[0.041]; report[0.040]; report[0.040]; rights[0.040]; Report[0.040]; era[0.040]; era[0.040]; South[0.040]; South[0.040]; final[0.040]; expect[0.040]; release[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = African National Congress\u001b[39m ==> ENTITY: \u001b[32mAfrican National Congress\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nANC[0.046]; ANC[0.046]; African[0.046]; South[0.045]; South[0.045]; apartheid[0.044]; Africa[0.043]; accused[0.042]; rights[0.042]; white[0.041]; white[0.041]; release[0.041]; acknowledged[0.040]; Reconciliation[0.040]; rule[0.040]; condemn[0.040]; abuses[0.040]; abuses[0.040]; governments[0.040]; previously[0.040]; services[0.040]; struggle[0.040]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = land mines ==> ENTITY: \u001b[32mLand mine\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.067:0.067[0]; log p(e|m)= -0.019:-0.019[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.044]; rights[0.043]; rights[0.043]; camps[0.042]; party[0.042]; Angola[0.042]; gross[0.042]; gross[0.042]; avoid[0.042]; panel[0.042]; border[0.042]; Commission[0.041]; commission[0.041]; planting[0.041]; intention[0.041]; violations[0.041]; violations[0.041]; farms[0.041]; advance[0.041]; likely[0.041]; report[0.041]; report[0.041]; month[0.041]; fall[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mtorture\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHuman rights in Zimbabwe\u001b[39m <---> \u001b[32mTorture\u001b[39m\t\nSCORES: global= 0.253:0.244[\u001b[31m0.009\u001b[39m]; local()= 0.138:0.070[\u001b[31m0.068\u001b[39m]; log p(e|m)= -1.386:-0.069[\u001b[32m1.317\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrights[0.043]; apartheid[0.043]; abuses[0.043]; abuses[0.043]; security[0.042]; ANC[0.042]; ANC[0.042]; ANC[0.042]; document[0.042]; governments[0.042]; Reconciliation[0.042]; National[0.042]; acknowledged[0.041]; South[0.041]; South[0.041]; condemns[0.041]; attacks[0.041]; violations[0.041]; violations[0.041]; report[0.041]; report[0.041]; Commission[0.040]; accused[0.040]; condemn[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Angola ==> ENTITY: \u001b[32mAngola\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.068:0.068[0]; log p(e|m)= -0.393:-0.393[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.048]; border[0.044]; month[0.043]; camps[0.043]; sent[0.042]; said[0.042]; abuses[0.042]; officials[0.042]; advance[0.042]; appearing[0.041]; believed[0.041]; rights[0.041]; nervous[0.041]; request[0.041]; commission[0.041]; earlier[0.041]; bombings[0.041]; accusations[0.041]; immediately[0.041]; Commission[0.041]; informing[0.041]; spokesman[0.041]; land[0.040]; farms[0.040]; \t\n\r [================================>.............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 537/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19981210_0433.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.085:0.085[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStates[0.045]; American[0.045]; Britain[0.045]; groups[0.044]; groups[0.044]; United[0.043]; world[0.042]; Arab[0.042]; said[0.042]; said[0.042]; Islamic[0.042]; British[0.042]; month[0.041]; Thursday[0.041]; Israel[0.041]; million[0.041]; London[0.040]; receiving[0.040]; bid[0.040]; officials[0.040]; opposes[0.040]; opposes[0.040]; hosted[0.040]; Iranian[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Arab ==> ENTITY: \u001b[32mArab people\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.225:0.225[0]; log p(e|m)= -2.048:-2.048[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.049]; Palestinian[0.047]; American[0.047]; Islamic[0.046]; Syrian[0.045]; Syria[0.044]; Syria[0.044]; Hassan[0.044]; Hussein[0.043]; Hussein[0.043]; Britain[0.043]; country[0.043]; Damascus[0.043]; world[0.042]; world[0.042]; Iranian[0.042]; groups[0.042]; groups[0.042]; British[0.042]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Islamic ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.178:0.178[0]; log p(e|m)= -0.212:-0.212[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArab[0.047]; Iranian[0.046]; Syria[0.045]; Syria[0.045]; Syria[0.045]; Iran[0.045]; Syrian[0.045]; Israel[0.044]; country[0.044]; month[0.043]; Hassan[0.043]; Damascus[0.043]; Palestinian[0.043]; groups[0.043]; groups[0.043]; opposition[0.042]; Iraqi[0.042]; Iraqi[0.042]; Iraqi[0.042]; Iraqi[0.042]; Britain[0.042]; Hussein[0.042]; Hussein[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Abdul-Halim Khaddam\u001b[39m ==> ENTITY: \u001b[32mAbdul Halim Khaddam\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.241:0.241[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.046]; Syria[0.046]; Syrian[0.046]; Khaddam[0.045]; Khaddam[0.045]; Damascus[0.045]; Israel[0.043]; Arab[0.042]; Iranian[0.042]; Hussein[0.041]; Hussein[0.041]; Hassan[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; Islamic[0.040]; said[0.040]; said[0.040]; said[0.040]; Saddam[0.039]; Saddam[0.039]; leader[0.039]; month[0.038]; President[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hafez Assad\u001b[39m ==> ENTITY: \u001b[32mHafez al-Assad\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyrian[0.047]; Syrian[0.047]; Syria[0.047]; Syria[0.047]; Arab[0.044]; Palestinian[0.043]; Palestinian[0.043]; Khaddam[0.042]; IRMA[0.041]; Iranian[0.041]; Iranian[0.041]; Iranian[0.041]; Iraq[0.041]; Iraqi[0.041]; Iraqi[0.041]; said[0.040]; said[0.040]; President[0.040]; official[0.039]; ally[0.039]; visit[0.039]; Iran[0.039]; Iran[0.039]; Iran[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hassan Habibi\u001b[39m ==> ENTITY: \u001b[32mHassan Habibi\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.107:0.107[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIranian[0.046]; Iranian[0.046]; Iran[0.045]; Habibi[0.043]; IRMA[0.042]; Syria[0.042]; Syria[0.042]; Syria[0.042]; Islamic[0.042]; President[0.042]; Israel[0.041]; Syrian[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; Arab[0.041]; British[0.041]; Hussein[0.040]; Hussein[0.040]; month[0.040]; Britain[0.040]; country[0.040]; Damascus[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syria ==> ENTITY: \u001b[32mSyria\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.230:0.230[0]; log p(e|m)= -0.261:-0.261[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.049]; Israel[0.049]; Syrian[0.046]; Arab[0.046]; Damascus[0.043]; Iranian[0.043]; Islamic[0.042]; Iraqi[0.042]; Iraqi[0.042]; Iraqi[0.042]; Saddam[0.041]; Saddam[0.041]; Hussein[0.041]; Hussein[0.041]; month[0.041]; Britain[0.040]; Hassan[0.040]; British[0.039]; States[0.039]; States[0.039]; officials[0.039]; Khaddam[0.038]; Khaddam[0.038]; Thursday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.206:0.206[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArab[0.046]; Iraqi[0.046]; Iraqi[0.046]; Syrian[0.045]; Syria[0.044]; Syria[0.044]; British[0.043]; Islamic[0.043]; Iranian[0.042]; Hussein[0.042]; Hussein[0.042]; Britain[0.042]; American[0.042]; Saddam[0.041]; Saddam[0.041]; Hassan[0.041]; Israel[0.041]; Damascus[0.040]; Khaddam[0.039]; Khaddam[0.039]; granted[0.038]; overthrow[0.038]; overthrow[0.038]; world[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iran-Iraq war\u001b[39m ==> ENTITY: \u001b[32mIran–Iraq War\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.046]; Iranian[0.045]; Iranian[0.045]; Iranian[0.045]; Iran[0.045]; Iraqi[0.045]; Syrian[0.044]; Syrian[0.044]; Arab[0.043]; IRMA[0.042]; Palestinian[0.040]; Kharrazi[0.040]; Foreign[0.040]; foreign[0.040]; said[0.040]; official[0.040]; President[0.040]; Assad[0.040]; Hafez[0.039]; cooperation[0.039]; biannually[0.039]; visit[0.039]; trade[0.039]; supported[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Damascus airport\u001b[39m ==> ENTITY: \u001b[32mDamascus International Airport\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.046]; Syria[0.046]; Syria[0.046]; Syrian[0.044]; Arab[0.043]; IRMA[0.042]; Hussein[0.042]; Hussein[0.042]; Israel[0.042]; British[0.042]; Britain[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; London[0.041]; Palestinian[0.041]; Thursday[0.040]; month[0.040]; Hassan[0.040]; Saddam[0.039]; Saddam[0.039]; American[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.157:0.157[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArab[0.047]; Syria[0.046]; Syria[0.046]; Palestinian[0.046]; Britain[0.043]; Syrian[0.043]; Iranian[0.043]; cooperation[0.042]; President[0.042]; American[0.042]; country[0.041]; States[0.041]; States[0.041]; Damascus[0.041]; United[0.041]; United[0.041]; British[0.040]; month[0.040]; officials[0.039]; Thursday[0.039]; Iraqi[0.039]; Iraqi[0.039]; Iraqi[0.039]; Iraqi[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.051]; London[0.045]; American[0.044]; States[0.043]; States[0.043]; Thursday[0.043]; month[0.042]; world[0.042]; granted[0.041]; opposes[0.041]; opposes[0.041]; million[0.041]; Israel[0.041]; Syria[0.041]; Syria[0.041]; Arab[0.040]; United[0.040]; United[0.040]; Islamic[0.040]; serve[0.040]; promote[0.040]; groups[0.040]; groups[0.040]; officials[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.093:0.093[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.048]; British[0.047]; committee[0.044]; meeting[0.044]; month[0.044]; news[0.044]; groups[0.043]; groups[0.043]; world[0.043]; world[0.043]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; Committee[0.043]; meetings[0.043]; Thursday[0.043]; country[0.042]; leader[0.042]; issues[0.042]; issues[0.042]; focus[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saddam Hussein\u001b[39m ==> ENTITY: \u001b[32mSaddam Hussein\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaddam[0.046]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Syria[0.045]; Syria[0.045]; Hussein[0.044]; Iranian[0.044]; Syrian[0.043]; Israel[0.042]; Arab[0.042]; American[0.042]; Damascus[0.040]; Hassan[0.040]; Britain[0.040]; British[0.040]; said[0.040]; said[0.040]; Islamic[0.039]; Thursday[0.039]; President[0.039]; overthrow[0.038]; overthrow[0.038]; month[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Palestinian ==> ENTITY: \u001b[32mPalestinian people\u001b[39m\t\nSCORES: global= 0.241:0.241[0]; local()= 0.180:0.180[0]; log p(e|m)= -0.389:-0.389[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.051]; Palestinian[0.047]; Arab[0.046]; Syria[0.046]; Syrian[0.044]; Syrian[0.044]; Islamic[0.043]; country[0.041]; Damascus[0.041]; Hassan[0.041]; Assad[0.041]; British[0.041]; Iranian[0.040]; Iranian[0.040]; Iranian[0.040]; Iranian[0.040]; American[0.040]; States[0.040]; Iran[0.040]; groups[0.039]; groups[0.039]; countries[0.039]; Hafez[0.039]; official[0.039]; \t\n\r [=================================>............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 553/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19981113_0500.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = airport ==> ENTITY: \u001b[32mAirport\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.095:0.095[0]; log p(e|m)= -0.121:-0.121[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairport[0.047]; Airport[0.046]; Airport[0.046]; said[0.043]; said[0.043]; Aviation[0.043]; International[0.043]; opened[0.043]; officials[0.042]; Thursday[0.042]; operation[0.042]; southern[0.042]; operations[0.041]; schedule[0.040]; Friday[0.040]; fledgling[0.040]; open[0.040]; Palestinian[0.040]; Palestinian[0.040]; Palestinian[0.040]; Palestinians[0.040]; Authority[0.040]; operating[0.039]; held[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIsraeli\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIsrael\u001b[39m <---> \u001b[32mIsraelis\u001b[39m\t\nSCORES: global= 0.250:0.239[\u001b[31m0.011\u001b[39m]; local()= 0.208:0.217[\u001b[32m0.010\u001b[39m]; log p(e|m)= -0.348:-1.749[\u001b[31m1.401\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.051]; Israel[0.051]; Israeli[0.046]; Israeli[0.046]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinians[0.043]; Gaza[0.042]; Netanyahu[0.041]; operation[0.040]; International[0.038]; Cabinet[0.038]; Strip[0.038]; southern[0.038]; Yasser[0.038]; peace[0.038]; Fayez[0.038]; Thursday[0.038]; Arafat[0.037]; operations[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = bus ==> ENTITY: \u001b[32mBus\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.028:0.028[0]; log p(e|m)= -0.322:-0.322[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsraeli[0.049]; Israel[0.048]; Israelis[0.045]; Israelis[0.045]; Cabinet[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; employees[0.044]; schedule[0.044]; security[0.044]; officials[0.043]; operating[0.043]; lack[0.042]; set[0.042]; personnel[0.042]; Authority[0.042]; week[0.042]; inspection[0.042]; date[0.042]; covered[0.042]; operation[0.041]; held[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Civil Aviation Authority ==> ENTITY: \u001b[32mNational aviation authority\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.408:-0.408[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npersonnel[0.047]; Israel[0.044]; operations[0.043]; Airport[0.043]; airport[0.043]; airport[0.043]; airport[0.043]; airport[0.043]; Israeli[0.042]; Israeli[0.042]; Israeli[0.042]; security[0.042]; operation[0.042]; procedures[0.041]; employees[0.041]; International[0.041]; Palestinian[0.040]; Palestinian[0.040]; Palestinian[0.040]; Palestinian[0.040]; inspection[0.040]; plane[0.040]; operating[0.040]; Cabinet[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIsraelis\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIsrael\u001b[39m <---> \u001b[32mIsraelis\u001b[39m\t\nSCORES: global= 0.258:0.254[\u001b[31m0.004\u001b[39m]; local()= 0.195:0.204[\u001b[32m0.009\u001b[39m]; log p(e|m)= -0.546:-0.009[\u001b[32m0.537\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.056]; Israeli[0.052]; Israeli[0.052]; Israelis[0.050]; Palestinian[0.048]; Palestinian[0.048]; Palestinian[0.048]; Palestinian[0.048]; personnel[0.045]; Gaza[0.045]; security[0.045]; plane[0.043]; Netanyahu[0.043]; Zeidan[0.042]; Zeidan[0.042]; Zeidan[0.042]; operation[0.042]; Cabinet[0.042]; southern[0.042]; officials[0.041]; week[0.041]; week[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Civil Aviation ==> ENTITY: \u001b[32mCivil aviation\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.152:0.152[0]; log p(e|m)= -0.272:-0.272[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCabinet[0.048]; Minister[0.045]; Authority[0.044]; operations[0.043]; International[0.043]; Airport[0.043]; Airport[0.043]; airport[0.043]; airport[0.043]; operation[0.042]; operating[0.041]; Thursday[0.041]; fledgling[0.041]; officials[0.041]; officials[0.041]; agreement[0.041]; Friday[0.040]; week[0.040]; week[0.040]; ratification[0.040]; independence[0.040]; start[0.040]; held[0.040]; Prime[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gaza Strip\u001b[39m ==> ENTITY: \u001b[32mGaza Strip\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.221:0.221[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.046]; Israel[0.046]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Israeli[0.044]; Israeli[0.044]; Israeli[0.044]; Palestinians[0.043]; Israelis[0.043]; Fayez[0.041]; southern[0.040]; security[0.039]; Cabinet[0.039]; operation[0.039]; operations[0.038]; Authority[0.038]; Authority[0.038]; Netanyahu[0.038]; Zeidan[0.038]; Yasser[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yasser Arafat International Airport\u001b[39m ==> ENTITY: \u001b[32mYasser Arafat International Airport\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.252:0.252[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPalestinians[0.044]; Gaza[0.044]; Israel[0.044]; Israel[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Israelis[0.043]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; airport[0.040]; airport[0.040]; airport[0.040]; Fayez[0.040]; Strip[0.039]; Netanyahu[0.039]; open[0.039]; set[0.039]; security[0.038]; Zeidan[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = security ==> ENTITY: \u001b[32mSecurity\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.095:0.095[0]; log p(e|m)= -0.846:-0.846[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsraeli[0.045]; Israeli[0.045]; Israel[0.044]; procedures[0.043]; airport[0.042]; airport[0.042]; airport[0.042]; airport[0.042]; Airport[0.042]; luggage[0.042]; personnel[0.042]; checks[0.042]; Gaza[0.041]; Palestinian[0.041]; Palestinian[0.041]; Palestinian[0.041]; Palestinian[0.041]; employees[0.041]; inspection[0.040]; Israelis[0.040]; Israelis[0.040]; operation[0.040]; officials[0.039]; passport[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Palestinian ==> ENTITY: \u001b[32mPalestinian people\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.161:0.161[0]; log p(e|m)= -0.389:-0.389[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.049]; Israel[0.049]; Palestinian[0.047]; Palestinian[0.047]; Israeli[0.046]; Israeli[0.046]; Gaza[0.045]; Palestinians[0.045]; Yasser[0.042]; Cabinet[0.041]; southern[0.041]; Netanyahu[0.040]; Arafat[0.040]; week[0.039]; week[0.039]; agreement[0.039]; peace[0.039]; weeks[0.039]; Authority[0.038]; operation[0.038]; Civil[0.038]; officials[0.038]; International[0.038]; await[0.038]; \t\n\r [==================================>...........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 564/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19980816_0994.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Auckland Art Gallery\u001b[39m ==> ENTITY: \u001b[32mAuckland Art Gallery\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAuckland[0.046]; Gallery[0.045]; Art[0.045]; Zealand[0.044]; Zealand[0.044]; James[0.043]; million[0.043]; million[0.043]; painting[0.043]; painting[0.043]; painting[0.043]; works[0.042]; Senior[0.041]; artist[0.041]; century[0.040]; expensive[0.040]; worth[0.040]; ago[0.040]; said[0.040]; said[0.040]; said[0.040]; valued[0.039]; helmet[0.039]; understood[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = shotgun ==> ENTITY: \u001b[32mShotgun\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.089:-0.089[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nshotgun[0.052]; firearm[0.045]; Gary[0.045]; motorcycle[0.044]; armed[0.044]; door[0.044]; days[0.044]; European[0.044]; black[0.044]; century[0.043]; James[0.043]; gunman[0.043]; know[0.043]; expensive[0.043]; grabbed[0.042]; man[0.042]; man[0.042]; man[0.042]; man[0.042]; man[0.042]; displayed[0.042]; crowbar[0.042]; ran[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAlbert Park\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAlbert Park, Victoria\u001b[39m <---> \u001b[32mAlbert Park, Auckland\u001b[39m\t\nSCORES: global= 0.245:0.237[\u001b[31m0.008\u001b[39m]; local()= 0.091:0.102[\u001b[32m0.011\u001b[39m]; log p(e|m)= -1.100:-1.945[\u001b[31m0.845\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnearby[0.047]; entrance[0.046]; wall[0.043]; Museum[0.043]; man[0.042]; man[0.042]; public[0.042]; highway[0.042]; remove[0.042]; ground[0.042]; believed[0.042]; appeared[0.041]; main[0.041]; displayed[0.041]; ran[0.041]; Lower[0.041]; member[0.041]; Wellesley[0.040]; gallery[0.040]; going[0.040]; said[0.040]; said[0.040]; backing[0.040]; pursued[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = firearm\u001b[39m ==> ENTITY: \u001b[32mFirearm\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.002:-0.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nshotgun[0.050]; shotgun[0.050]; ground[0.045]; armed[0.045]; gloves[0.044]; sawed[0.044]; robber[0.044]; frame[0.044]; valued[0.043]; dropped[0.043]; displayed[0.043]; helmet[0.042]; believed[0.042]; appeared[0.042]; black[0.042]; million[0.042]; robbery[0.042]; European[0.042]; public[0.042]; door[0.042]; stolen[0.042]; grabbed[0.042]; member[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = crowbar\u001b[39m ==> ENTITY: \u001b[32mCrowbar (tool)\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nground[0.046]; crowbar[0.045]; frame[0.044]; door[0.043]; gloves[0.042]; visor[0.042]; sawed[0.042]; wall[0.042]; stolen[0.041]; remove[0.041]; dropped[0.041]; shotgun[0.041]; shotgun[0.041]; armed[0.041]; James[0.041]; main[0.041]; main[0.041]; waving[0.041]; man[0.041]; man[0.041]; man[0.041]; man[0.041]; helmet[0.040]; daring[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Auckland ==> ENTITY: \u001b[32mAuckland\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.071:0.071[0]; log p(e|m)= -0.297:-0.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAuckland[0.048]; Zealand[0.044]; James[0.043]; went[0.043]; century[0.043]; Gary[0.042]; dropped[0.042]; displayed[0.042]; Lower[0.042]; Monday[0.041]; going[0.041]; wearing[0.041]; Wellesley[0.041]; ago[0.041]; works[0.041]; main[0.041]; ground[0.041]; staff[0.041]; appeared[0.040]; helmet[0.040]; clad[0.040]; Allcock[0.040]; Museum[0.040]; visor[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Zealand ==> ENTITY: \u001b[32mNew Zealand\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.086:0.086[0]; log p(e|m)= -0.238:-0.238[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAuckland[0.047]; James[0.046]; Zealand[0.045]; understood[0.043]; European[0.043]; Monday[0.043]; Gary[0.042]; century[0.042]; said[0.041]; said[0.041]; said[0.041]; ago[0.041]; million[0.041]; million[0.041]; Aug[0.041]; Allcock[0.041]; born[0.041]; Senior[0.040]; man[0.040]; man[0.040]; popular[0.040]; person[0.040]; Joseph[0.040]; days[0.039]; \t\n\r [===================================>..........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 572/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19980808_0196.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.083:0.083[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnation[0.049]; States[0.048]; United[0.045]; said[0.045]; said[0.045]; television[0.045]; Saturday[0.044]; Wednesday[0.044]; decision[0.044]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; ties[0.043]; late[0.043]; thought[0.042]; thought[0.042]; says[0.042]; allies[0.042]; government[0.042]; mark[0.042]; Iranian[0.042]; campaign[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ayatollah Ruhollah Khomeini\u001b[39m ==> ENTITY: \u001b[32mRuhollah Khomeini\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIranian[0.044]; speech[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; Iran[0.043]; Iran[0.043]; Iran[0.043]; Kuwait[0.042]; Hussein[0.041]; government[0.041]; Saddam[0.041]; Saddam[0.041]; Saddam[0.041]; Saddam[0.041]; Saddam[0.041]; President[0.041]; decreed[0.040]; States[0.040]; States[0.040]; nation[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = patriotic ==> ENTITY: \u001b[32mPatriotism\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.060:-0.060[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsongs[0.046]; speech[0.045]; National[0.043]; imposed[0.043]; referring[0.043]; today[0.042]; decreed[0.042]; celebrated[0.042]; played[0.042]; artillery[0.042]; Saturday[0.042]; television[0.041]; salute[0.041]; invasion[0.041]; president[0.041]; spoke[0.041]; taste[0.041]; said[0.041]; said[0.041]; aggression[0.041]; imposing[0.040]; delivered[0.040]; day[0.040]; black[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = embargo ==> ENTITY: \u001b[32mEmbargo\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.170:0.170[0]; log p(e|m)= -0.361:-0.361[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsanctions[0.047]; government[0.046]; States[0.046]; imposed[0.046]; nation[0.044]; Iran[0.044]; Iran[0.044]; Iran[0.044]; United[0.043]; United[0.043]; war[0.043]; war[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; allies[0.043]; Kuwait[0.043]; freeze[0.042]; referring[0.042]; unfair[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mallies\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAllies\u001b[39m <---> \u001b[32mAllies of World War II\u001b[39m\t\nSCORES: global= 0.246:0.245[\u001b[31m0.000\u001b[39m]; local()= 0.159:0.151[\u001b[31m0.009\u001b[39m]; log p(e|m)= -1.121:-0.906[\u001b[32m0.215\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwar[0.045]; war[0.045]; cooperation[0.044]; said[0.044]; said[0.044]; said[0.044]; said[0.044]; States[0.044]; States[0.044]; Iranian[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; government[0.043]; referring[0.043]; Wednesday[0.043]; says[0.043]; invasion[0.042]; urged[0.042]; nation[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36msanctions\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSanctions against Iraq\u001b[39m <---> \u001b[32mInternational sanctions\u001b[39m\t\nSCORES: global= 0.258:0.256[\u001b[31m0.002\u001b[39m]; local()= 0.196:0.183[\u001b[31m0.013\u001b[39m]; log p(e|m)= -1.262:-0.629[\u001b[32m0.633\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.044]; Iraq[0.044]; Iraq[0.044]; Iraq[0.044]; Iraq[0.044]; imposed[0.043]; Kuwait[0.043]; embargo[0.042]; States[0.042]; government[0.042]; National[0.041]; Wednesday[0.041]; war[0.041]; war[0.041]; United[0.040]; United[0.040]; Iran[0.040]; Iran[0.040]; Iran[0.040]; Saddam[0.040]; Saddam[0.040]; Saddam[0.040]; Saddam[0.040]; Saddam[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saddam Hussein\u001b[39m ==> ENTITY: \u001b[32mSaddam Hussein\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.047]; Iraq[0.047]; Iraq[0.047]; Iraq[0.047]; Saddam[0.046]; Saddam[0.046]; Saddam[0.046]; Iranian[0.043]; Iran[0.043]; Iran[0.043]; Iran[0.043]; war[0.042]; Ayatollah[0.040]; government[0.040]; Khomeini[0.040]; said[0.039]; said[0.039]; Wednesday[0.039]; allies[0.038]; President[0.038]; States[0.038]; States[0.038]; campaign[0.037]; cooperation[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iran-Iraq war\u001b[39m ==> ENTITY: \u001b[32mIran–Iraq War\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.047]; Iraq[0.047]; Iraq[0.047]; Iraq[0.047]; Iranian[0.044]; Iran[0.044]; Iran[0.044]; war[0.042]; Saddam[0.041]; Saddam[0.041]; Saddam[0.041]; Saddam[0.041]; Saddam[0.041]; Saddam[0.041]; Hussein[0.041]; Khomeini[0.040]; government[0.040]; referring[0.039]; said[0.039]; said[0.039]; said[0.039]; said[0.039]; President[0.039]; allies[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kuwait ==> ENTITY: \u001b[32mKuwait\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.169:0.169[0]; log p(e|m)= -0.337:-0.337[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.047]; Iraq[0.047]; Iraq[0.047]; Iraq[0.047]; Iran[0.044]; Iranian[0.043]; government[0.043]; States[0.042]; war[0.042]; cooperation[0.041]; played[0.041]; embargo[0.041]; Baghdad[0.040]; invasion[0.040]; sanctions[0.040]; allies[0.040]; National[0.040]; victory[0.040]; failed[0.040]; relations[0.039]; normalize[0.039]; day[0.039]; United[0.039]; United[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iranian government\u001b[39m ==> ENTITY: \u001b[32mPolitics of Iran\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.205:0.205[0]; log p(e|m)= -0.006:-0.006[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.055]; Iran[0.055]; Iran[0.055]; Iran[0.055]; Ayatollah[0.051]; television[0.050]; Kuwait[0.050]; Hussein[0.050]; Ruhollah[0.049]; Iraq[0.049]; Iraq[0.049]; Iraq[0.049]; Iraq[0.049]; Iraq[0.049]; Iraq[0.049]; Khomeini[0.048]; sanctions[0.047]; speech[0.047]; president[0.047]; allies[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.241:0.241[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.050]; Iraq[0.050]; Iraq[0.050]; Iraq[0.050]; Saddam[0.046]; Saddam[0.046]; Saddam[0.046]; Saddam[0.046]; Saddam[0.046]; Hussein[0.043]; government[0.043]; war[0.043]; war[0.043]; cooperation[0.042]; Iran[0.041]; Iran[0.041]; Iran[0.041]; Iranian[0.040]; allies[0.040]; said[0.039]; said[0.039]; said[0.039]; relations[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = artillery ==> ENTITY: \u001b[32mArtillery\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.102:0.102[0]; log p(e|m)= -0.046:-0.046[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nguns[0.045]; National[0.044]; war[0.043]; early[0.042]; today[0.042]; said[0.042]; said[0.042]; boomed[0.042]; salute[0.042]; Iraq[0.042]; Iraq[0.042]; delivered[0.042]; Kuwait[0.042]; 7th[0.041]; day[0.041]; late[0.041]; patriotic[0.041]; exactly[0.041]; Saturday[0.041]; invasion[0.041]; imposed[0.041]; suit[0.040]; played[0.040]; referring[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.121:0.121[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.044]; Iraq[0.044]; Iraq[0.044]; Iraq[0.044]; Iraq[0.044]; Iraq[0.044]; inspectors[0.042]; Iran[0.042]; Iran[0.042]; Iran[0.042]; today[0.041]; Saddam[0.041]; Saddam[0.041]; Saddam[0.041]; Saddam[0.041]; Saddam[0.041]; Saddam[0.041]; Kuwait[0.041]; referring[0.041]; States[0.040]; States[0.040]; government[0.040]; Iranian[0.040]; President[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baghdad\u001b[39m ==> ENTITY: \u001b[32mBaghdad\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.048]; Iraq[0.048]; Kuwait[0.046]; played[0.044]; Saddam[0.044]; Saddam[0.044]; Saddam[0.044]; songs[0.043]; Ayatollah[0.042]; artillery[0.041]; invasion[0.041]; Saturday[0.041]; Iran[0.041]; Khomeini[0.040]; National[0.040]; war[0.040]; late[0.040]; today[0.039]; radio[0.039]; early[0.039]; Ruhollah[0.039]; day[0.039]; referring[0.038]; imposed[0.038]; \t\n\r [====================================>.........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 586/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19980627_0596.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = journalist ==> ENTITY: \u001b[32mJournalist\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.118:-0.118[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \njournalist[0.047]; magazine[0.046]; news[0.046]; newspapers[0.045]; television[0.045]; radio[0.044]; award[0.044]; University[0.043]; magazines[0.043]; columnist[0.042]; Mainstream[0.042]; saying[0.041]; weekly[0.041]; appeared[0.040]; graduated[0.040]; leading[0.040]; India[0.039]; edited[0.039]; Indian[0.039]; Indian[0.039]; Politics[0.039]; said[0.039]; decades[0.039]; Saturday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Calcutta University\u001b[39m ==> ENTITY: \u001b[32mUniversity of Calcutta\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.044]; India[0.044]; India[0.044]; College[0.044]; graduated[0.043]; Calcutta[0.043]; University[0.043]; University[0.043]; Indian[0.043]; studied[0.043]; Oxford[0.041]; Indira[0.041]; award[0.041]; hospital[0.041]; Assam[0.040]; Prasar[0.040]; Chakravarty[0.040]; Chakravarty[0.040]; Chakravarty[0.040]; Chakravarty[0.040]; Chakravarty[0.040]; history[0.040]; taught[0.040]; emergency[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Prasar Bharati\u001b[39m ==> ENTITY: \u001b[32mPrasar Bharati\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.045]; India[0.045]; College[0.045]; corporation[0.044]; radio[0.043]; television[0.043]; news[0.042]; University[0.042]; University[0.042]; University[0.042]; hospital[0.041]; newspapers[0.041]; governing[0.041]; government[0.041]; government[0.041]; government[0.041]; appointed[0.041]; magazines[0.041]; autonomous[0.040]; award[0.040]; Indian[0.040]; Indian[0.040]; establishment[0.040]; imposed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = journalism ==> ENTITY: \u001b[32mJournalism\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.162:0.162[0]; log p(e|m)= -0.026:-0.026[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnewspapers[0.044]; newspapers[0.044]; news[0.043]; television[0.042]; journalist[0.042]; University[0.042]; University[0.042]; University[0.042]; magazine[0.042]; award[0.042]; government[0.042]; government[0.042]; government[0.042]; political[0.041]; imposed[0.041]; Mainstream[0.041]; Mainstream[0.041]; radio[0.041]; weekly[0.041]; weekly[0.041]; weekly[0.041]; Oxford[0.041]; edited[0.041]; establishment[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = India ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.111:0.111[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndian[0.046]; Indian[0.046]; Nikhil[0.043]; Nikhil[0.043]; Chakravarty[0.042]; Chakravarty[0.042]; Chakravarty[0.042]; Chakravarty[0.042]; Chakravarty[0.042]; television[0.042]; Bharati[0.042]; magazine[0.042]; government[0.042]; government[0.042]; government[0.042]; award[0.041]; establishment[0.041]; decades[0.041]; Saturday[0.040]; Prasar[0.040]; magazines[0.039]; leading[0.039]; appeared[0.039]; family[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Politics ==> ENTITY: \u001b[32mPolitics\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.136:0.136[0]; log p(e|m)= -1.839:-1.839[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \naward[0.044]; political[0.044]; journalist[0.043]; journalist[0.043]; history[0.043]; magazine[0.043]; columnist[0.043]; edited[0.043]; governing[0.042]; leading[0.041]; England[0.041]; newspapers[0.041]; studied[0.041]; news[0.041]; Oxford[0.041]; magazines[0.041]; establishment[0.041]; Mainstream[0.041]; saying[0.041]; weekly[0.040]; declined[0.040]; University[0.040]; University[0.040]; University[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Communist Party of India\u001b[39m ==> ENTITY: \u001b[32mCommunist Party of India\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.046]; India[0.046]; government[0.044]; government[0.044]; government[0.044]; Assam[0.044]; Indira[0.043]; CPI[0.042]; Gandhi[0.042]; Prasar[0.042]; member[0.042]; Bharati[0.041]; political[0.040]; Chakravarty[0.040]; Chakravarty[0.040]; Chakravarty[0.040]; Chakravarty[0.040]; autonomous[0.040]; Calcutta[0.040]; Calcutta[0.040]; joined[0.040]; brought[0.040]; decades[0.039]; establishment[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = corporation ==> ENTITY: \u001b[32mCorporation\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.073:0.073[0]; log p(e|m)= -0.049:-0.049[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngovernment[0.044]; government[0.044]; government[0.044]; Indian[0.043]; Indian[0.043]; University[0.042]; University[0.042]; University[0.042]; India[0.042]; India[0.042]; Calcutta[0.042]; Calcutta[0.042]; studied[0.041]; England[0.041]; news[0.041]; member[0.041]; history[0.041]; taught[0.041]; journalist[0.041]; journalist[0.041]; governing[0.041]; television[0.041]; newspapers[0.040]; journalism[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = brain cancer\u001b[39m ==> ENTITY: \u001b[32mBrain tumor\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.119:0.119[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCollege[0.049]; suffering[0.048]; died[0.046]; died[0.046]; hospital[0.045]; University[0.045]; University[0.045]; Prasar[0.044]; control[0.043]; declined[0.043]; said[0.043]; television[0.042]; studied[0.042]; family[0.042]; news[0.042]; weekly[0.042]; appeared[0.042]; identified[0.042]; Saturday[0.042]; establishment[0.042]; started[0.042]; leading[0.042]; graduated[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Indira Gandhi\u001b[39m ==> ENTITY: \u001b[32mIndira Gandhi\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPrime[0.046]; India[0.045]; India[0.045]; India[0.045]; Minister[0.044]; Prasar[0.042]; state[0.042]; Calcutta[0.042]; Calcutta[0.042]; Assam[0.042]; Bharati[0.042]; November[0.041]; government[0.041]; government[0.041]; award[0.041]; Party[0.041]; member[0.041]; imposed[0.040]; political[0.039]; Chakravarty[0.039]; Chakravarty[0.039]; Chakravarty[0.039]; CPI[0.039]; chief[0.039]; \t\n\r [====================================>.........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 598/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19981109_0464.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = the Hague ==> ENTITY: \u001b[32mThe Hague\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.070:-0.070[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.048]; ambassador[0.044]; ICJ[0.043]; Court[0.043]; government[0.042]; Europe[0.042]; interfere[0.042]; political[0.042]; Monday[0.042]; filed[0.042]; Tuesday[0.042]; International[0.041]; saying[0.041]; cooperation[0.041]; economic[0.041]; protest[0.041]; protest[0.041]; Free[0.041]; organizations[0.041]; affairs[0.041]; Justice[0.040]; said[0.040]; Foreign[0.040]; Prague[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Prague ==> ENTITY: \u001b[32mPrague\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.053:-0.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPrague[0.048]; Germany[0.043]; Czech[0.043]; Czech[0.043]; Soviet[0.042]; Tuesday[0.042]; Munich[0.042]; set[0.042]; Europe[0.042]; Europe[0.042]; Friday[0.041]; Monday[0.041]; English[0.041]; August[0.041]; ambassador[0.041]; began[0.041]; began[0.041]; protest[0.041]; protest[0.041]; sponsored[0.041]; Radio[0.041]; Radio[0.041]; Iraq[0.040]; Iraq[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iron Curtain\u001b[39m ==> ENTITY: \u001b[32mIron Curtain\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.047]; Europe[0.044]; Communist[0.043]; countries[0.043]; Germany[0.043]; American[0.043]; communism[0.042]; West[0.042]; cooperation[0.042]; Prague[0.041]; Prague[0.041]; War[0.041]; economic[0.041]; vj[0.041]; East[0.041]; broadcast[0.041]; government[0.041]; government[0.041]; political[0.041]; period[0.041]; Czech[0.041]; Czech[0.041]; service[0.041]; directed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.140:0.140[0]; log p(e|m)= -1.207:-1.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlanguage[0.046]; language[0.046]; American[0.044]; English[0.044]; Prague[0.043]; Prague[0.043]; Czech[0.043]; Germany[0.043]; Soviet[0.043]; Europe[0.042]; Europe[0.042]; ambassador[0.042]; internal[0.041]; service[0.041]; radio[0.040]; radio[0.040]; countries[0.040]; Radio[0.040]; Radio[0.040]; economic[0.040]; assistance[0.039]; August[0.039]; broadcasts[0.039]; Munich[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = ICJ\u001b[39m ==> ENTITY: \u001b[32mInternational Court of Justice\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCourt[0.045]; saying[0.044]; Hague[0.043]; Justice[0.042]; Iraq[0.042]; Iraq[0.042]; trial[0.042]; Foreign[0.042]; International[0.042]; political[0.042]; Europe[0.041]; Europe[0.041]; government[0.041]; government[0.041]; Netherlands[0.041]; filed[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; ambassador[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cold War\u001b[39m ==> ENTITY: \u001b[32mCold War\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.046]; countries[0.045]; Iraq[0.044]; Iraq[0.044]; economic[0.043]; American[0.043]; Iran[0.043]; Iran[0.043]; Communist[0.043]; communism[0.043]; Europe[0.042]; cooperation[0.042]; period[0.041]; government[0.041]; political[0.041]; Germany[0.040]; following[0.040]; West[0.040]; collapse[0.040]; service[0.040]; program[0.039]; weeks[0.039]; Prague[0.039]; Prague[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.174:0.174[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.052]; government[0.044]; government[0.044]; cooperation[0.043]; countries[0.043]; American[0.042]; Iran[0.042]; Iran[0.042]; Iran[0.042]; Iran[0.042]; Iran[0.042]; War[0.041]; economic[0.041]; said[0.041]; Foreign[0.040]; assistance[0.040]; West[0.040]; saying[0.040]; protest[0.040]; Kharrazi[0.040]; following[0.040]; Europe[0.040]; political[0.040]; Germany[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iran ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.171:0.171[0]; log p(e|m)= -0.138:-0.138[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.052]; Iran[0.052]; Iran[0.052]; Iran[0.052]; saying[0.045]; Kharrazi[0.045]; government[0.043]; reported[0.043]; Europe[0.043]; ambassador[0.043]; quoted[0.042]; said[0.042]; organizations[0.042]; Foreign[0.041]; International[0.041]; assistance[0.041]; Radio[0.041]; protest[0.041]; protest[0.041]; internal[0.041]; Monday[0.040]; global[0.040]; English[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Foreign Minister ==> ENTITY: \u001b[32mForeign minister\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.125:0.125[0]; log p(e|m)= -1.877:-1.877[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.044]; Iran[0.044]; Iran[0.044]; Iran[0.044]; Iran[0.044]; ambassador[0.044]; government[0.044]; cooperation[0.043]; Kharrazi[0.043]; affairs[0.042]; assistance[0.042]; American[0.041]; Netherlands[0.041]; said[0.041]; International[0.041]; economic[0.041]; saying[0.040]; reported[0.040]; recalled[0.040]; Europe[0.040]; political[0.040]; internal[0.040]; global[0.040]; quoted[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Communist ==> ENTITY: \u001b[32mCommunism\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.393:-0.393[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.046]; communism[0.045]; Europe[0.045]; countries[0.045]; American[0.044]; government[0.043]; economic[0.043]; political[0.043]; Germany[0.042]; Czech[0.042]; Prague[0.041]; Prague[0.041]; cooperation[0.041]; following[0.040]; Iraq[0.040]; Iraq[0.040]; service[0.040]; August[0.040]; moved[0.040]; uncensored[0.040]; West[0.040]; news[0.040]; Iran[0.039]; Iran[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Farsi ==> ENTITY: \u001b[32mPersian language\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.171:-0.171[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlanguage[0.050]; language[0.050]; English[0.048]; broadcast[0.044]; Iran[0.044]; Iran[0.044]; Iran[0.044]; Iran[0.044]; Iran[0.044]; broadcasts[0.044]; American[0.043]; Iraq[0.043]; Iraq[0.043]; radio[0.042]; station[0.042]; station[0.042]; say[0.042]; say[0.042]; countries[0.041]; Radio[0.041]; West[0.041]; cooperation[0.041]; daily[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = the Netherlands ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.104:0.104[0]; log p(e|m)= -0.176:-0.176[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; Europe[0.046]; Europe[0.046]; day[0.043]; approved[0.042]; saying[0.042]; government[0.041]; government[0.041]; English[0.041]; Munich[0.041]; service[0.041]; ambassador[0.041]; August[0.041]; minute[0.041]; weeks[0.041]; affairs[0.041]; International[0.041]; level[0.041]; Hague[0.040]; internal[0.040]; Czech[0.040]; Czech[0.040]; Tuesday[0.040]; broadcast[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Radio Free Europe\u001b[39m ==> ENTITY: \u001b[32mRadio Free Europe/Radio Liberty\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbroadcasts[0.046]; quoted[0.043]; said[0.043]; radio[0.043]; radio[0.043]; ambassador[0.042]; recalled[0.042]; Prague[0.042]; Prague[0.042]; Czech[0.042]; set[0.042]; American[0.042]; internal[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; International[0.041]; Monday[0.041]; Tuesday[0.040]; reported[0.040]; affairs[0.040]; political[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Munich, Germany\u001b[39m ==> ENTITY: \u001b[32mMunich\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.107:0.107[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurope[0.044]; Prague[0.044]; Prague[0.044]; Prague[0.044]; American[0.043]; Czech[0.043]; Czech[0.043]; headquarters[0.042]; War[0.042]; communism[0.042]; service[0.041]; government[0.041]; government[0.041]; August[0.041]; vj[0.041]; Friday[0.041]; began[0.041]; began[0.041]; sponsored[0.041]; ambassador[0.041]; program[0.040]; Soviet[0.040]; following[0.040]; earlier[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kamal Kharrazi\u001b[39m ==> ENTITY: \u001b[32mKamal Kharazi\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.046]; Iran[0.046]; Iran[0.046]; Iran[0.046]; Iran[0.046]; Iraq[0.045]; Foreign[0.044]; American[0.042]; ambassador[0.042]; affairs[0.042]; International[0.041]; Monday[0.041]; Minister[0.041]; weeks[0.041]; Tuesday[0.040]; reported[0.040]; Europe[0.040]; cooperation[0.039]; Daily[0.039]; English[0.039]; language[0.039]; political[0.039]; government[0.039]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Soviet ==> ENTITY: \u001b[32mSoviet Union\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.098:-0.098[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncommunism[0.044]; Communist[0.043]; War[0.043]; August[0.042]; language[0.042]; Germany[0.042]; program[0.042]; American[0.042]; ambassador[0.042]; Europe[0.042]; Prague[0.042]; Prague[0.042]; Prague[0.042]; following[0.041]; West[0.041]; protest[0.041]; Iraq[0.041]; Iraq[0.041]; cooperation[0.041]; Iran[0.041]; Iran[0.041]; Czech[0.041]; Czech[0.041]; Farsi[0.041]; \t\n\r [=====================================>........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 614/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19981120_1056.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = P.W. Botha\u001b[39m ==> ENTITY: \u001b[32mP. W. Botha\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.046]; National[0.045]; South[0.045]; apartheid[0.044]; apartheid[0.044]; African[0.044]; Botha[0.043]; Botha[0.043]; Botha[0.043]; Reconciliation[0.042]; Africa[0.042]; Commission[0.041]; Congress[0.041]; issued[0.041]; evaluated[0.040]; month[0.040]; Johannesburg[0.040]; trial[0.040]; trial[0.040]; Comission[0.040]; said[0.040]; said[0.040]; Jan[0.040]; landmark[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Truth and Reconciliation Commission (South Africa)\u001b[39m ==> ENTITY: \u001b[32mTruth and Reconciliation Commission (South Africa)\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncommission[0.044]; investigated[0.043]; trial[0.043]; trial[0.043]; trial[0.043]; report[0.042]; African[0.042]; National[0.042]; apartheid[0.042]; apartheid[0.042]; apartheid[0.042]; committed[0.042]; committed[0.042]; prosecute[0.042]; issued[0.041]; amnesty[0.041]; evidence[0.041]; Comission[0.041]; abuses[0.041]; abuses[0.041]; abuses[0.041]; Truth[0.040]; charges[0.040]; Friday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = contempt of court\u001b[39m ==> ENTITY: \u001b[32mContempt of court\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.190:0.190[0]; log p(e|m)= -0.005:-0.005[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nordered[0.046]; issued[0.044]; trial[0.044]; refusing[0.044]; appealed[0.043]; jail[0.043]; convicted[0.042]; decision[0.042]; sentence[0.042]; sought[0.042]; commission[0.041]; commission[0.041]; testify[0.041]; people[0.041]; month[0.041]; abuses[0.041]; grant[0.041]; Comission[0.040]; fine[0.040]; suspended[0.040]; National[0.040]; committed[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = amnesty ==> ENTITY: \u001b[32mAmnesty\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.052:-0.052[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \namnesty[0.045]; appealed[0.043]; commission[0.043]; commission[0.043]; Reconciliation[0.043]; abuses[0.042]; abuses[0.042]; convicted[0.042]; decision[0.042]; Congress[0.042]; sought[0.041]; committed[0.041]; issued[0.041]; Truth[0.041]; Truth[0.041]; group[0.041]; said[0.041]; said[0.041]; report[0.041]; anti[0.041]; Comission[0.041]; National[0.041]; Commission[0.040]; ordered[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = rand ==> ENTITY: \u001b[32mSouth African rand\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.096:0.096[0]; log p(e|m)= -0.060:-0.060[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrican[0.045]; grant[0.044]; Johannesburg[0.044]; fully[0.043]; National[0.043]; building[0.042]; aos[0.042]; sapa[0.042]; said[0.041]; housing[0.041]; commission[0.041]; commission[0.041]; group[0.041]; committed[0.041]; head[0.041]; apartheid[0.041]; apartheid[0.041]; thousands[0.041]; London[0.041]; suspended[0.041]; state[0.040]; trial[0.040]; decision[0.040]; headquarters[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = white rule\u001b[39m ==> ENTITY: \u001b[32mDominant minority\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngroup[0.045]; National[0.044]; issued[0.043]; African[0.043]; landmark[0.043]; insufficient[0.043]; Reconciliation[0.042]; Africa[0.042]; anti[0.042]; apartheid[0.042]; apartheid[0.042]; apartheid[0.042]; said[0.042]; said[0.042]; brought[0.041]; committed[0.041]; evaluated[0.041]; South[0.041]; Jan[0.040]; report[0.040]; head[0.040]; era[0.040]; President[0.040]; building[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Johannesburg ==> ENTITY: \u001b[32mJohannesburg\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.151:0.151[0]; log p(e|m)= -0.095:-0.095[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.046]; African[0.044]; landmark[0.043]; housing[0.043]; Botha[0.042]; Botha[0.042]; Botha[0.042]; Botha[0.042]; Botha[0.042]; Botha[0.042]; building[0.042]; apartheid[0.041]; apartheid[0.041]; apartheid[0.041]; apartheid[0.041]; white[0.041]; headquarters[0.041]; London[0.041]; Africa[0.041]; rand[0.041]; state[0.040]; month[0.040]; National[0.040]; currently[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = African National Congress\u001b[39m ==> ENTITY: \u001b[32mAfrican National Congress\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.049]; apartheid[0.048]; apartheid[0.048]; apartheid[0.048]; apartheid[0.048]; President[0.048]; Johannesburg[0.047]; Africa[0.046]; white[0.045]; group[0.044]; committed[0.044]; committed[0.044]; Botha[0.044]; Botha[0.044]; Botha[0.044]; Botha[0.044]; Botha[0.044]; Botha[0.044]; Reconciliation[0.044]; rule[0.044]; charges[0.044]; convicted[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = prosecutor ==> ENTITY: \u001b[32mProsecutor\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.027:-0.027[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nevidence[0.047]; charges[0.046]; trial[0.045]; trial[0.045]; prosecute[0.043]; issued[0.043]; investigated[0.042]; said[0.042]; said[0.042]; South[0.042]; insufficient[0.042]; month[0.042]; Prosecutor[0.041]; Prosecutor[0.041]; ordered[0.041]; Friday[0.041]; President[0.040]; evaluated[0.040]; ultimately[0.040]; brought[0.040]; National[0.039]; collected[0.039]; Commission[0.039]; Comission[0.039]; \t\n\r [======================================>.......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 623/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19980624_0607.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iran ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.190:0.190[0]; log p(e|m)= -0.138:-0.138[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.048]; Iran[0.048]; Iranian[0.046]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; Iraqi[0.042]; Iraqi[0.042]; official[0.041]; war[0.041]; Najafi[0.041]; Najafi[0.041]; Najafi[0.041]; Agency[0.040]; Baghdad[0.040]; Baghdad[0.040]; Baghdad[0.040]; issue[0.039]; reported[0.039]; officials[0.039]; week[0.039]; week[0.039]; visiting[0.039]; talks[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baghdad\u001b[39m ==> ENTITY: \u001b[32mBaghdad\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.260:0.260[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.048]; Iraq[0.048]; Iraq[0.048]; Iraq[0.048]; Iraq[0.048]; Iraqi[0.044]; Iraqi[0.044]; Baghdad[0.044]; Baghdad[0.044]; al[0.041]; al[0.041]; official[0.040]; war[0.039]; Al[0.039]; Iranian[0.039]; Iranian[0.039]; Iranian[0.039]; Iran[0.039]; Iran[0.039]; Iran[0.039]; officials[0.038]; Najafi[0.038]; Najafi[0.038]; Najafi[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = International Committee of the Red Cross\u001b[39m ==> ENTITY: \u001b[32mInternational Committee of the Red Cross\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; civil[0.043]; prisoners[0.042]; released[0.042]; released[0.042]; week[0.042]; law[0.042]; repatriation[0.041]; POWs[0.041]; POWs[0.041]; POWs[0.041]; POWs[0.041]; POWs[0.041]; POWs[0.041]; detainees[0.041]; war[0.041]; Iraqi[0.041]; Iraqi[0.041]; classified[0.041]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = repatriation ==> ENTITY: \u001b[32mRepatriation\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.142:0.142[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPOWs[0.043]; POWs[0.043]; POWs[0.043]; POWs[0.043]; POWs[0.043]; POWs[0.043]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; angered[0.042]; prisoners[0.042]; Iranians[0.042]; Iranian[0.041]; Iranian[0.041]; Iranian[0.041]; Iranian[0.041]; News[0.041]; Agency[0.041]; people[0.041]; million[0.041]; Committee[0.041]; claims[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iran-Iraq war\u001b[39m ==> ENTITY: \u001b[32mIran–Iraq War\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.053]; Iraq[0.053]; Iraq[0.053]; Iraq[0.053]; Iranian[0.049]; Iranian[0.049]; Iranian[0.049]; Iran[0.049]; Iran[0.049]; Iraqi[0.049]; Iraqi[0.049]; Iranians[0.049]; Iraqis[0.047]; Tehran[0.046]; classified[0.045]; reportedly[0.044]; said[0.044]; claims[0.043]; wounded[0.043]; civil[0.042]; says[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = prisoners of war\u001b[39m ==> ENTITY: \u001b[32mPrisoner of war\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.104:0.104[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPOWs[0.048]; Iraq[0.046]; Iraq[0.046]; Iraq[0.046]; Iraq[0.046]; Iraq[0.046]; Iraqi[0.045]; Iraqi[0.045]; POW[0.044]; POW[0.044]; Al[0.043]; Wednesday[0.043]; Wednesday[0.043]; Baghdad[0.043]; Baghdad[0.043]; Baghdad[0.043]; al[0.042]; al[0.042]; held[0.042]; commission[0.041]; officials[0.041]; visiting[0.041]; arrived[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tehran\u001b[39m ==> ENTITY: \u001b[32mTehran\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIranian[0.044]; Iranian[0.044]; Iranian[0.044]; Iranian[0.044]; Iranian[0.044]; Iran[0.042]; Iran[0.042]; Iran[0.042]; Iran[0.042]; Iranians[0.042]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; official[0.041]; freed[0.041]; Najafi[0.040]; Najafi[0.040]; Najafi[0.040]; detainees[0.040]; following[0.040]; prisoners[0.040]; prisoners[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iraqi News Agency\u001b[39m ==> ENTITY: \u001b[32mNational Iraqi News Agency\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.045]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; Baghdad[0.044]; Baghdad[0.044]; Baghdad[0.044]; Iraqi[0.044]; Iraqi[0.044]; al[0.042]; al[0.042]; agency[0.040]; senior[0.039]; official[0.039]; Iranian[0.039]; Iranian[0.039]; Iranian[0.039]; Iranian[0.039]; Tehran[0.038]; war[0.038]; head[0.038]; Al[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.229:0.229[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.055]; Iraq[0.055]; Iraqi[0.050]; Iraqi[0.050]; Baghdad[0.048]; Baghdad[0.048]; Baghdad[0.048]; war[0.046]; Iran[0.044]; Iran[0.044]; Iran[0.044]; Iranian[0.044]; al[0.043]; al[0.043]; officials[0.043]; official[0.043]; Abdullah[0.043]; following[0.042]; Al[0.042]; Agency[0.042]; delegation[0.041]; News[0.041]; \t\n\r [======================================>.......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 632/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19980624_0436.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iran ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.190:0.190[0]; log p(e|m)= -0.138:-0.138[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.048]; Iran[0.048]; Iranian[0.046]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; Iraqi[0.042]; Iraqi[0.042]; official[0.041]; war[0.041]; Najafi[0.041]; Najafi[0.041]; Najafi[0.041]; Agency[0.040]; Baghdad[0.040]; Baghdad[0.040]; Baghdad[0.040]; issue[0.039]; reported[0.039]; officials[0.039]; week[0.039]; week[0.039]; visiting[0.039]; talks[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baghdad\u001b[39m ==> ENTITY: \u001b[32mBaghdad\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.260:0.260[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.048]; Iraq[0.048]; Iraq[0.048]; Iraq[0.048]; Iraq[0.048]; Iraqi[0.044]; Iraqi[0.044]; Baghdad[0.044]; Baghdad[0.044]; al[0.041]; al[0.041]; official[0.040]; war[0.039]; Al[0.039]; Iranian[0.039]; Iranian[0.039]; Iranian[0.039]; Iran[0.039]; Iran[0.039]; Iran[0.039]; officials[0.038]; Najafi[0.038]; Najafi[0.038]; Najafi[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iran-Iraq war\u001b[39m ==> ENTITY: \u001b[32mIran–Iraq War\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.051]; Iraq[0.051]; Iraq[0.051]; Iraq[0.051]; Iranian[0.047]; Iranian[0.047]; Iranian[0.047]; Iran[0.047]; Iran[0.047]; Iraqi[0.047]; Iraqi[0.047]; Iranians[0.047]; Iraqis[0.045]; Tehran[0.044]; classified[0.043]; reportedly[0.043]; said[0.042]; lb[0.042]; claims[0.041]; wounded[0.041]; civil[0.041]; says[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = detainees ==> ENTITY: \u001b[32mDetention (imprisonment)\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.081:0.081[0]; log p(e|m)= -0.138:-0.138[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nprisoners[0.049]; Iraq[0.048]; Iraq[0.048]; Iraq[0.048]; Iraq[0.048]; Iraq[0.048]; released[0.048]; released[0.048]; classified[0.048]; Iranian[0.048]; Iranian[0.048]; Iranian[0.048]; Iranian[0.048]; repatriation[0.047]; Iraqi[0.047]; Iraqi[0.047]; Iraqi[0.047]; details[0.047]; Iran[0.047]; Iran[0.047]; Iran[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tehran\u001b[39m ==> ENTITY: \u001b[32mTehran\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIranian[0.044]; Iranian[0.044]; Iranian[0.044]; Iranian[0.044]; Iranian[0.044]; Iran[0.042]; Iran[0.042]; Iran[0.042]; Iran[0.042]; Iranians[0.042]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; official[0.041]; freed[0.041]; Najafi[0.040]; Najafi[0.040]; Najafi[0.040]; detainees[0.040]; following[0.040]; prisoners[0.040]; prisoners[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = repatriation ==> ENTITY: \u001b[32mRepatriation\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.142:0.142[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPOWs[0.043]; POWs[0.043]; POWs[0.043]; POWs[0.043]; POWs[0.043]; POWs[0.043]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; angered[0.042]; prisoners[0.042]; Iranians[0.042]; Iranian[0.041]; Iranian[0.041]; Iranian[0.041]; Iranian[0.041]; News[0.041]; Agency[0.041]; people[0.041]; million[0.041]; Committee[0.041]; claims[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = prisoners of war\u001b[39m ==> ENTITY: \u001b[32mPrisoner of war\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.104:0.104[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPOWs[0.048]; Iraq[0.046]; Iraq[0.046]; Iraq[0.046]; Iraq[0.046]; Iraq[0.046]; Iraqi[0.045]; Iraqi[0.045]; POW[0.044]; POW[0.044]; Al[0.043]; Wednesday[0.043]; Wednesday[0.043]; Baghdad[0.043]; Baghdad[0.043]; Baghdad[0.043]; al[0.042]; al[0.042]; held[0.042]; commission[0.041]; officials[0.041]; visiting[0.041]; arrived[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iraqi News Agency\u001b[39m ==> ENTITY: \u001b[32mNational Iraqi News Agency\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.045]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; Baghdad[0.044]; Baghdad[0.044]; Baghdad[0.044]; Iraqi[0.044]; Iraqi[0.044]; al[0.042]; al[0.042]; agency[0.040]; senior[0.039]; official[0.039]; Iranian[0.039]; Iranian[0.039]; Iranian[0.039]; Iranian[0.039]; Tehran[0.038]; war[0.038]; head[0.038]; Al[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = delegation ==> ENTITY: \u001b[32mDelegation\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.662:-0.662[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncommission[0.043]; visiting[0.043]; pact[0.043]; agreement[0.043]; Al[0.043]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; official[0.042]; officials[0.042]; Iranian[0.041]; Iranian[0.041]; Iranian[0.041]; signed[0.041]; gave[0.041]; sign[0.041]; agency[0.041]; Iran[0.040]; Iran[0.040]; Iran[0.040]; Iran[0.040]; held[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = International Committee of the Red Cross\u001b[39m ==> ENTITY: \u001b[32mInternational Committee of the Red Cross\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; civil[0.043]; prisoners[0.042]; released[0.042]; released[0.042]; week[0.042]; law[0.042]; repatriation[0.041]; POWs[0.041]; POWs[0.041]; POWs[0.041]; POWs[0.041]; POWs[0.041]; POWs[0.041]; detainees[0.041]; war[0.041]; Iraqi[0.041]; Iraqi[0.041]; classified[0.041]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.229:0.229[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.055]; Iraq[0.055]; Iraqi[0.050]; Iraqi[0.050]; Baghdad[0.048]; Baghdad[0.048]; Baghdad[0.048]; war[0.046]; Iran[0.044]; Iran[0.044]; Iran[0.044]; Iranian[0.044]; al[0.043]; al[0.043]; officials[0.043]; official[0.043]; Abdullah[0.043]; following[0.042]; Al[0.042]; Agency[0.042]; delegation[0.041]; News[0.041]; \t\n\r [=======================================>......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 643/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19980620_0458.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = prisoner of war\u001b[39m ==> ENTITY: \u001b[32mPrisoner of war\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.103:0.103[0]; log p(e|m)= -0.007:-0.007[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPOWs[0.047]; POWs[0.047]; POW[0.044]; POW[0.044]; information[0.043]; Iraq[0.043]; bid[0.041]; states[0.041]; thousands[0.041]; needed[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; official[0.041]; talks[0.040]; Foreign[0.040]; issue[0.040]; remains[0.040]; remains[0.040]; News[0.040]; said[0.040]; glean[0.040]; closure[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iraqi News Agency\u001b[39m ==> ENTITY: \u001b[32mNational Iraqi News Agency\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.047]; Iraq[0.047]; Iraq[0.047]; Iraqi[0.046]; Iraqi[0.046]; Iraqi[0.046]; Sahhaf[0.044]; al[0.043]; al[0.043]; official[0.041]; War[0.041]; Iranian[0.040]; Iranian[0.040]; Iranian[0.040]; INA[0.040]; normalizing[0.040]; Mohammed[0.039]; war[0.039]; head[0.039]; thousands[0.039]; Abdullah[0.039]; closure[0.039]; Minister[0.039]; negotiator[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = International Committee of the Red Cross\u001b[39m ==> ENTITY: \u001b[32mInternational Committee of the Red Cross\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; ICRC[0.043]; civil[0.042]; camps[0.042]; prisoners[0.042]; organization[0.042]; released[0.042]; released[0.042]; law[0.041]; POWs[0.041]; POWs[0.041]; POWs[0.041]; POWs[0.041]; POWs[0.041]; detainees[0.041]; Agency[0.041]; visits[0.041]; countries[0.041]; Iraqi[0.041]; Iraqi[0.041]; classified[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iran-Iraq War\u001b[39m ==> ENTITY: \u001b[32mIran–Iraq War\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.047]; Iraq[0.047]; Iraq[0.047]; Iraq[0.047]; Iranian[0.044]; Iranian[0.044]; Iran[0.044]; Iran[0.044]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iranians[0.043]; Baghdad[0.041]; War[0.039]; ICRC[0.039]; al[0.039]; al[0.039]; countries[0.039]; organization[0.039]; said[0.038]; Sahhaf[0.038]; official[0.038]; civil[0.038]; captured[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baghdad\u001b[39m ==> ENTITY: \u001b[32mBaghdad\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.050]; Iraq[0.050]; Iraq[0.050]; Iraq[0.050]; Iraqi[0.046]; al[0.043]; Iranians[0.041]; Sahhaf[0.041]; Iranian[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; Najafi[0.040]; released[0.040]; released[0.040]; Abdullah[0.039]; million[0.039]; countries[0.038]; POWs[0.038]; POWs[0.038]; POWs[0.038]; POWs[0.038]; POWs[0.038]; captured[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Missing-in-action\u001b[39m ==> ENTITY: \u001b[32mMissing in action\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPOWs[0.045]; POWs[0.045]; POWs[0.045]; POWs[0.045]; war[0.044]; remains[0.043]; remains[0.043]; missing[0.043]; POW[0.042]; POW[0.042]; Iraq[0.041]; Iraq[0.041]; Iraq[0.041]; ICRC[0.041]; War[0.041]; said[0.041]; reported[0.040]; Bessler[0.040]; INA[0.040]; released[0.040]; captured[0.040]; normalizing[0.040]; camps[0.039]; prisoners[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mohammed Saeed al-Sahhaf\u001b[39m ==> ENTITY: \u001b[32mMuhammad Saeed al-Sahhaf\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.045]; Iraq[0.045]; Iraq[0.045]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; official[0.044]; Baghdad[0.042]; information[0.042]; al[0.041]; ICRC[0.041]; Iranian[0.041]; Iranian[0.041]; Iranian[0.041]; Foreign[0.040]; war[0.040]; News[0.040]; said[0.040]; Cabinet[0.040]; told[0.040]; Iran[0.039]; Iran[0.039]; INA[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Associated Press\u001b[39m ==> ENTITY: \u001b[32mAssociated Press\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.064:0.064[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.050]; Iraq[0.050]; Iraq[0.050]; Iraq[0.050]; Iraqi[0.043]; Baghdad[0.042]; War[0.041]; War[0.041]; released[0.041]; released[0.041]; told[0.041]; regular[0.040]; people[0.040]; countries[0.040]; classified[0.040]; Committee[0.039]; detainees[0.039]; organization[0.039]; POWs[0.039]; POWs[0.039]; POWs[0.039]; POWs[0.039]; POWs[0.039]; million[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iran ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.171:0.171[0]; log p(e|m)= -0.138:-0.138[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIranian[0.048]; Iranian[0.048]; Iraq[0.047]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; official[0.042]; war[0.042]; Najafi[0.042]; Agency[0.041]; issue[0.041]; reported[0.041]; Saeed[0.040]; said[0.040]; Cabinet[0.040]; talks[0.040]; told[0.040]; bid[0.039]; News[0.039]; Foreign[0.039]; Abdullah[0.039]; al[0.039]; al[0.039]; \t\n\r [========================================>.....] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 653/727 \r\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = allocation\u001b[39m ==> ENTITY: \u001b[32mResource allocation\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \neconomic[0.049]; allocation[0.045]; policy[0.044]; desired[0.043]; given[0.042]; given[0.042]; poor[0.042]; general[0.041]; projects[0.041]; projects[0.041]; projects[0.041]; projects[0.041]; projects[0.041]; agencies[0.041]; paper[0.041]; closely[0.041]; provided[0.041]; added[0.041]; benefit[0.040]; progress[0.040]; said[0.040]; said[0.040]; said[0.040]; officials[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Negri Sembilan\u001b[39m ==> ENTITY: \u001b[32mNegeri Sembilan\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKuala[0.046]; Lukut[0.044]; Annuar[0.044]; Annuar[0.044]; district[0.044]; Nawawi[0.043]; proposed[0.042]; senior[0.042]; general[0.042]; Hassan[0.042]; Government[0.042]; Firdaus[0.041]; allocation[0.041]; added[0.041]; added[0.041]; Rahman[0.041]; officials[0.041]; agencies[0.040]; said[0.040]; director[0.040]; Port[0.040]; deputy[0.040]; Abdul[0.040]; day[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = ministry ==> ENTITY: \u001b[32mMinistry (government department)\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.167:0.167[0]; log p(e|m)= -1.130:-1.130[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGovernment[0.048]; Government[0.048]; Federal[0.048]; ministry[0.045]; ministry[0.045]; minister[0.045]; minister[0.045]; secretary[0.044]; Minister[0.043]; projects[0.043]; projects[0.043]; projects[0.043]; projects[0.043]; projects[0.043]; policy[0.042]; State[0.042]; parliamentary[0.042]; implementation[0.042]; implementation[0.042]; allocation[0.042]; allocation[0.042]; agencies[0.042]; officials[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = implementation ==> ENTITY: \u001b[32mImplementation\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.373:-0.373[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nimplementation[0.048]; desired[0.046]; agencies[0.045]; projects[0.045]; projects[0.045]; projects[0.045]; projects[0.045]; provided[0.044]; participants[0.044]; allocation[0.043]; policy[0.043]; successful[0.043]; results[0.043]; general[0.043]; task[0.043]; ensure[0.043]; ensure[0.043]; closely[0.042]; benefit[0.042]; Programme[0.042]; progress[0.042]; said[0.042]; said[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Port Dickson\u001b[39m ==> ENTITY: \u001b[32mPort Dickson\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLukut[0.049]; Kuala[0.049]; Annuar[0.045]; Annuar[0.045]; Sembilan[0.045]; Nawawi[0.045]; senior[0.044]; district[0.044]; Government[0.043]; proposed[0.043]; project[0.043]; projects[0.043]; projects[0.043]; projects[0.043]; projects[0.043]; Firdaus[0.042]; Harun[0.042]; visiting[0.042]; long[0.042]; ministry[0.041]; ministry[0.041]; today[0.041]; allocation[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mpoor\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAct for the Relief of the Poor 1601\u001b[39m <---> \u001b[32mPoverty\u001b[39m\t\nSCORES: global= 0.246:0.245[\u001b[31m0.001\u001b[39m]; local()= 0.159:0.095[\u001b[31m0.064\u001b[39m]; log p(e|m)= -0.693:-0.205[\u001b[32m0.489\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npolicy[0.053]; economic[0.053]; Government[0.052]; Government[0.052]; agencies[0.051]; Federal[0.051]; Rural[0.051]; Poor[0.050]; allocation[0.050]; allocation[0.050]; responsible[0.050]; responsible[0.050]; provided[0.049]; Programme[0.049]; benefit[0.048]; success[0.048]; given[0.048]; given[0.048]; Development[0.048]; Development[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Malaysia ==> ENTITY: \u001b[32mMalaysia\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.081:0.081[0]; log p(e|m)= -0.202:-0.202[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGovernment[0.044]; Datuk[0.043]; yesterday[0.043]; Annuar[0.043]; economic[0.042]; officers[0.042]; State[0.042]; want[0.042]; progress[0.042]; district[0.042]; Rural[0.041]; said[0.041]; said[0.041]; Development[0.041]; Development[0.041]; given[0.041]; poor[0.041]; personally[0.041]; Programme[0.041]; agencies[0.041]; responsible[0.041]; responsible[0.041]; closely[0.041]; successful[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Government agencies ==> ENTITY: \u001b[32mGovernment agency\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.083:-0.083[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFederal[0.049]; Minister[0.043]; Development[0.043]; Development[0.043]; responsible[0.043]; responsible[0.043]; State[0.043]; implementation[0.043]; implementation[0.043]; economic[0.042]; given[0.042]; ministry[0.042]; allocation[0.042]; Rural[0.042]; monitor[0.041]; undertake[0.040]; provided[0.040]; task[0.040]; said[0.040]; said[0.040]; yesterday[0.039]; benefit[0.039]; personally[0.039]; Programme[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = studies ==> ENTITY: \u001b[32mResearch\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.106:0.106[0]; log p(e|m)= -1.452:-1.452[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nadded[0.046]; said[0.044]; conduct[0.044]; agencies[0.044]; given[0.043]; ensure[0.042]; policy[0.042]; general[0.042]; day[0.042]; projects[0.041]; projects[0.041]; projects[0.041]; officials[0.041]; visiting[0.041]; visit[0.041]; visit[0.041]; project[0.041]; today[0.041]; director[0.041]; long[0.041]; depth[0.041]; implementing[0.040]; ministry[0.040]; proposed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = minister ==> ENTITY: \u001b[32mMinister (government)\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.148:0.148[0]; log p(e|m)= -1.324:-1.324[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nministry[0.046]; ministry[0.046]; ministry[0.046]; Government[0.046]; minister[0.044]; Minister[0.043]; policy[0.043]; parliamentary[0.042]; secretary[0.042]; yesterday[0.041]; Port[0.041]; general[0.041]; said[0.041]; said[0.041]; said[0.041]; deputy[0.040]; today[0.040]; allocation[0.040]; allocation[0.040]; responsible[0.040]; responsible[0.040]; given[0.040]; visiting[0.040]; conduct[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rural ==> ENTITY: \u001b[32mRural area\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.074:0.074[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \neconomic[0.046]; Government[0.045]; district[0.043]; policy[0.043]; poor[0.042]; agencies[0.042]; ensure[0.042]; ensure[0.042]; According[0.042]; minister[0.042]; minister[0.042]; implementation[0.041]; implementation[0.041]; allocation[0.041]; Development[0.041]; Development[0.041]; Minister[0.041]; projects[0.041]; projects[0.041]; projects[0.041]; projects[0.041]; said[0.041]; said[0.041]; Federal[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = mussel\u001b[39m ==> ENTITY: \u001b[32mMussel\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.029:0.029[0]; log p(e|m)= -0.007:-0.007[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nviability[0.044]; senior[0.044]; said[0.042]; said[0.042]; Port[0.042]; Kuala[0.042]; long[0.042]; rearing[0.042]; general[0.042]; Lukut[0.042]; day[0.042]; project[0.041]; poor[0.041]; Annuar[0.041]; Annuar[0.041]; desired[0.041]; district[0.041]; district[0.041]; proposed[0.041]; ministry[0.041]; ministry[0.041]; ministry[0.041]; responsible[0.041]; Sembilan[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = economic ==> ENTITY: \u001b[32mEconomy\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.118:0.118[0]; log p(e|m)= -0.648:-0.648[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGovernment[0.046]; allocation[0.046]; agencies[0.045]; implementation[0.045]; implementation[0.045]; provided[0.044]; progress[0.044]; Programme[0.044]; poor[0.043]; projects[0.043]; projects[0.043]; projects[0.043]; projects[0.043]; Rural[0.043]; given[0.043]; added[0.043]; closely[0.043]; said[0.042]; said[0.042]; Development[0.042]; Development[0.042]; Federal[0.042]; undertake[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lukut\u001b[39m ==> ENTITY: \u001b[32mLukut\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKuala[0.049]; Sembilan[0.046]; district[0.044]; district[0.044]; Annuar[0.044]; Annuar[0.044]; parliamentary[0.043]; Nawawi[0.043]; Port[0.042]; Dickson[0.042]; senior[0.041]; said[0.041]; said[0.041]; Harun[0.040]; added[0.040]; added[0.040]; general[0.040]; visiting[0.040]; happening[0.040]; proposed[0.040]; Firdaus[0.039]; ministry[0.039]; ministry[0.039]; deputy[0.039]; \t\n\r [=========================================>....] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 669/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19981109_0140.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Radio Free Europe\u001b[39m ==> ENTITY: \u001b[32mRadio Free Europe/Radio Liberty\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbroadcasts[0.045]; quoted[0.043]; said[0.043]; radio[0.043]; radio[0.043]; radio[0.043]; ambassador[0.042]; recalled[0.042]; Prague[0.042]; Prague[0.042]; Czech[0.042]; set[0.042]; American[0.042]; internal[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; International[0.041]; Monday[0.040]; Tuesday[0.040]; reported[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.137:0.137[0]; log p(e|m)= -1.207:-1.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlanguage[0.046]; language[0.046]; American[0.044]; English[0.044]; Prague[0.043]; Prague[0.043]; Czech[0.043]; Germany[0.043]; Soviet[0.043]; Europe[0.042]; Europe[0.042]; ambassador[0.041]; internal[0.041]; service[0.041]; radio[0.040]; radio[0.040]; radio[0.040]; countries[0.040]; Radio[0.040]; Radio[0.040]; economic[0.040]; assistance[0.039]; August[0.039]; broadcasts[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cold War\u001b[39m ==> ENTITY: \u001b[32mCold War\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.046]; countries[0.045]; Iraq[0.044]; Iraq[0.044]; economic[0.043]; American[0.043]; Iran[0.043]; Iran[0.043]; Communist[0.043]; communism[0.043]; Europe[0.042]; cooperation[0.042]; period[0.041]; government[0.041]; political[0.041]; Germany[0.040]; following[0.040]; West[0.040]; collapse[0.040]; service[0.040]; program[0.039]; weeks[0.039]; Prague[0.039]; Prague[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = censored ==> ENTITY: \u001b[32mCensorship\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.102:0.102[0]; log p(e|m)= -0.261:-0.261[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbroadcast[0.045]; broadcasts[0.044]; government[0.043]; government[0.043]; Germany[0.043]; Iran[0.042]; Iran[0.042]; Iran[0.042]; news[0.042]; protest[0.042]; service[0.042]; countries[0.042]; political[0.042]; language[0.041]; radio[0.041]; Radio[0.041]; Soviet[0.041]; economic[0.041]; American[0.040]; Europe[0.040]; said[0.040]; Farsi[0.040]; Munich[0.040]; communism[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = economic ==> ENTITY: \u001b[32mEconomy\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.138:0.138[0]; log p(e|m)= -0.648:-0.648[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.047]; government[0.046]; government[0.046]; global[0.045]; political[0.043]; countries[0.043]; Europe[0.042]; Europe[0.042]; level[0.042]; organizations[0.041]; approach[0.041]; English[0.041]; Germany[0.041]; saying[0.041]; Foreign[0.040]; service[0.040]; period[0.040]; said[0.040]; Iraq[0.040]; Iraq[0.040]; program[0.040]; assistance[0.040]; Communist[0.040]; Netherlands[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = ambassador ==> ENTITY: \u001b[32mAmbassador\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.107:0.107[0]; log p(e|m)= -0.027:-0.027[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.046]; cooperation[0.046]; Europe[0.045]; Europe[0.045]; Foreign[0.044]; broadcast[0.044]; Iran[0.044]; Iran[0.044]; Iran[0.044]; Iran[0.044]; Iran[0.044]; Iran[0.044]; Iran[0.044]; Iran[0.044]; Iraq[0.044]; Iraq[0.044]; organizations[0.043]; economic[0.043]; Germany[0.042]; Netherlands[0.042]; International[0.042]; saying[0.041]; global[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Communist ==> ENTITY: \u001b[32mCommunism\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.393:-0.393[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.046]; communism[0.045]; Europe[0.045]; countries[0.044]; American[0.044]; government[0.043]; economic[0.043]; political[0.043]; Germany[0.042]; Czech[0.042]; Prague[0.041]; Prague[0.041]; cooperation[0.041]; following[0.040]; Iraq[0.040]; Iraq[0.040]; censored[0.040]; service[0.040]; August[0.040]; moved[0.040]; West[0.040]; news[0.040]; Iran[0.039]; Iran[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Soviet ==> ENTITY: \u001b[32mSoviet Union\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.098:-0.098[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncommunism[0.044]; Communist[0.043]; War[0.043]; August[0.042]; language[0.042]; Germany[0.042]; program[0.042]; American[0.042]; ambassador[0.042]; Europe[0.042]; Prague[0.042]; Prague[0.042]; Prague[0.042]; following[0.041]; West[0.041]; protest[0.041]; Iraq[0.041]; Iraq[0.041]; cooperation[0.041]; Iran[0.041]; Iran[0.041]; Czech[0.041]; Czech[0.041]; Farsi[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = political ==> ENTITY: \u001b[32mPolitics\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.103:0.103[0]; log p(e|m)= -0.387:-0.387[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurope[0.044]; Europe[0.044]; countries[0.043]; ambassador[0.043]; economic[0.043]; Iraq[0.042]; Iraq[0.042]; American[0.042]; Iran[0.042]; Iran[0.042]; Iran[0.042]; Iran[0.042]; Iran[0.042]; Iran[0.042]; organizations[0.041]; West[0.041]; service[0.041]; government[0.041]; government[0.041]; affairs[0.041]; cooperation[0.040]; global[0.040]; approved[0.040]; news[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = the Hague ==> ENTITY: \u001b[32mThe Hague\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.070:-0.070[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.048]; Hague[0.045]; tribunal[0.044]; ambassador[0.044]; ICJ[0.043]; Court[0.042]; government[0.042]; Europe[0.042]; interfere[0.041]; political[0.041]; Monday[0.041]; filed[0.041]; Tuesday[0.041]; International[0.041]; saying[0.041]; cooperation[0.041]; economic[0.041]; protest[0.041]; protest[0.041]; Free[0.040]; organizations[0.040]; affairs[0.040]; Justice[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = ICJ\u001b[39m ==> ENTITY: \u001b[32mInternational Court of Justice\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCourt[0.059]; tribunal[0.059]; saying[0.057]; Hague[0.057]; Hague[0.057]; Justice[0.056]; Iraq[0.055]; Iraq[0.055]; trial[0.055]; Foreign[0.055]; International[0.055]; political[0.055]; Europe[0.054]; Europe[0.054]; government[0.054]; government[0.054]; Netherlands[0.054]; filed[0.054]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iron Curtain\u001b[39m ==> ENTITY: \u001b[32mIron Curtain\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.047]; Europe[0.044]; Communist[0.043]; countries[0.043]; Germany[0.043]; American[0.043]; communism[0.042]; West[0.042]; cooperation[0.042]; Prague[0.041]; Prague[0.041]; War[0.041]; economic[0.041]; vj[0.041]; East[0.041]; broadcast[0.041]; government[0.041]; government[0.041]; political[0.041]; period[0.041]; Czech[0.041]; Czech[0.041]; service[0.041]; directed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.174:0.174[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.052]; government[0.044]; government[0.044]; cooperation[0.043]; countries[0.043]; American[0.042]; Iran[0.042]; Iran[0.042]; Iran[0.042]; Iran[0.042]; Iran[0.042]; War[0.041]; economic[0.041]; said[0.041]; Foreign[0.040]; assistance[0.040]; West[0.040]; saying[0.040]; protest[0.040]; Kharrazi[0.040]; following[0.040]; Europe[0.040]; political[0.040]; Germany[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Prague ==> ENTITY: \u001b[32mPrague\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.053:-0.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPrague[0.048]; Germany[0.043]; Czech[0.043]; Czech[0.043]; Soviet[0.042]; Tuesday[0.042]; Munich[0.042]; set[0.042]; Europe[0.042]; Europe[0.042]; Friday[0.041]; Monday[0.041]; English[0.041]; August[0.041]; ambassador[0.041]; began[0.041]; began[0.041]; protest[0.041]; protest[0.041]; sponsored[0.041]; Radio[0.041]; Radio[0.041]; Iraq[0.040]; Iraq[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West ==> ENTITY: \u001b[32mWestern world\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.128:0.128[0]; log p(e|m)= -1.635:-1.635[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.050]; language[0.048]; Europe[0.045]; countries[0.044]; American[0.044]; East[0.044]; Munich[0.043]; Soviet[0.042]; Czech[0.042]; Farsi[0.041]; communism[0.041]; economic[0.040]; Friday[0.040]; August[0.040]; moved[0.040]; earlier[0.040]; political[0.040]; following[0.040]; Iraq[0.040]; Iraq[0.040]; day[0.040]; began[0.039]; began[0.039]; weeks[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kamal Kharrazi\u001b[39m ==> ENTITY: \u001b[32mKamal Kharazi\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.045]; Iran[0.045]; Iran[0.045]; Iran[0.045]; Iran[0.045]; Iran[0.045]; Iraq[0.045]; Foreign[0.043]; American[0.042]; ambassador[0.042]; affairs[0.042]; International[0.041]; Monday[0.041]; Minister[0.040]; weeks[0.040]; Tuesday[0.040]; reported[0.040]; Europe[0.040]; cooperation[0.039]; Daily[0.039]; English[0.039]; language[0.039]; political[0.039]; government[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Farsi ==> ENTITY: \u001b[32mPersian language\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.171:-0.171[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlanguage[0.050]; language[0.050]; English[0.048]; broadcast[0.044]; Iran[0.044]; Iran[0.044]; Iran[0.044]; Iran[0.044]; Iran[0.044]; broadcasts[0.044]; American[0.043]; Iraq[0.043]; Iraq[0.043]; radio[0.042]; station[0.042]; station[0.042]; say[0.042]; say[0.042]; countries[0.041]; Radio[0.041]; West[0.041]; cooperation[0.041]; daily[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = the Netherlands ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.103:0.103[0]; log p(e|m)= -0.176:-0.176[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; Europe[0.046]; Europe[0.046]; day[0.043]; approved[0.042]; saying[0.042]; government[0.041]; government[0.041]; English[0.041]; Munich[0.041]; service[0.041]; ambassador[0.041]; August[0.041]; minute[0.041]; weeks[0.041]; affairs[0.041]; International[0.041]; level[0.041]; Hague[0.040]; Hague[0.040]; internal[0.040]; Czech[0.040]; Czech[0.040]; Tuesday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Munich, Germany\u001b[39m ==> ENTITY: \u001b[32mMunich\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.107:0.107[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurope[0.044]; Prague[0.044]; Prague[0.044]; Prague[0.044]; American[0.043]; Czech[0.043]; Czech[0.043]; headquarters[0.042]; War[0.042]; communism[0.042]; service[0.041]; government[0.041]; government[0.041]; August[0.041]; vj[0.041]; Friday[0.041]; began[0.041]; began[0.041]; sponsored[0.041]; ambassador[0.041]; program[0.040]; Soviet[0.040]; following[0.040]; earlier[0.040]; \t\n\r [==========================================>...] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 689/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19980615_0417.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = damages ==> ENTITY: \u001b[32mDamages\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.026:-0.026[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndamages[0.050]; libel[0.046]; libel[0.046]; actress[0.044]; suit[0.043]; Monday[0.043]; Actress[0.042]; Actress[0.042]; substantial[0.042]; alleging[0.042]; filed[0.041]; Sunday[0.041]; offenses[0.041]; reported[0.041]; Brooke[0.040]; Brooke[0.040]; apology[0.040]; apology[0.040]; continued[0.040]; public[0.040]; accepted[0.039]; false[0.039]; Airport[0.039]; began[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mdrug\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mCrack cocaine\u001b[39m <---> \u001b[32mDrug\u001b[39m\t\nSCORES: global= 0.247:0.234[\u001b[31m0.013\u001b[39m]; local()= 0.095:0.059[\u001b[31m0.036\u001b[39m]; log p(e|m)= -1.790:-1.211[\u001b[32m0.579\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npublic[0.043]; luggage[0.043]; alleging[0.043]; damages[0.042]; damages[0.042]; said[0.042]; substantial[0.042]; continued[0.042]; reported[0.042]; began[0.042]; offenses[0.042]; false[0.042]; false[0.042]; searched[0.041]; connection[0.041]; Monday[0.041]; libel[0.041]; libel[0.041]; questioned[0.041]; quizzed[0.041]; police[0.041]; police[0.041]; published[0.041]; accepts[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = substance abuse ==> ENTITY: \u001b[32mSubstance abuse\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.124:0.124[0]; log p(e|m)= -0.014:-0.014[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndrug[0.050]; drug[0.050]; actress[0.046]; actress[0.046]; abuse[0.044]; personal[0.042]; programs[0.041]; programs[0.041]; time[0.041]; needed[0.041]; Miss[0.041]; Miss[0.041]; amounts[0.040]; Monday[0.040]; committed[0.040]; Jonathan[0.040]; Claire[0.040]; spent[0.040]; suit[0.039]; money[0.039]; stopped[0.039]; family[0.039]; anti[0.039]; anti[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = French police\u001b[39m ==> ENTITY: \u001b[32mNational Police (France)\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.071:0.071[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npolice[0.046]; offenses[0.043]; drug[0.043]; drug[0.043]; abuse[0.043]; hours[0.043]; journalists[0.043]; public[0.042]; programs[0.042]; Festival[0.041]; anti[0.041]; anti[0.041]; searched[0.041]; accepted[0.041]; supporting[0.041]; said[0.041]; conceded[0.041]; Film[0.041]; luggage[0.041]; filed[0.040]; committed[0.040]; suit[0.040]; needed[0.040]; Nice[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nice Airport\u001b[39m ==> ENTITY: \u001b[32mNice Côte d'Azur Airport\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.108:0.108[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairport[0.049]; French[0.044]; Cannes[0.043]; luggage[0.042]; public[0.042]; flight[0.042]; amounts[0.042]; programs[0.042]; Festival[0.042]; hours[0.042]; said[0.041]; anti[0.041]; Film[0.041]; damages[0.041]; damages[0.041]; reported[0.041]; exclusive[0.041]; Associated[0.040]; page[0.040]; page[0.040]; page[0.040]; Monday[0.040]; time[0.040]; delayed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Associated Newspapers Ltd\u001b[39m ==> ENTITY: \u001b[32mDMG Media\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnewspaper[0.047]; libel[0.045]; libel[0.045]; Mail[0.045]; Mail[0.045]; Mail[0.045]; Mail[0.045]; publisher[0.044]; alleging[0.044]; exclusive[0.043]; reported[0.043]; retraction[0.043]; retraction[0.043]; published[0.043]; accepts[0.043]; damages[0.042]; damages[0.042]; false[0.042]; false[0.042]; apology[0.041]; apology[0.041]; apology[0.041]; Monday[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Brooke Shields\u001b[39m ==> ENTITY: \u001b[32mBrooke Shields\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.119:0.119[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nactress[0.050]; Brooke[0.046]; Actress[0.046]; Actress[0.046]; Miss[0.043]; Shields[0.043]; Shields[0.043]; Monday[0.042]; began[0.042]; Film[0.041]; exclusive[0.041]; Sunday[0.041]; Sunday[0.041]; Festival[0.041]; accepted[0.040]; drug[0.040]; accepts[0.040]; publisher[0.039]; reported[0.039]; apology[0.039]; apology[0.039]; delayed[0.039]; page[0.039]; page[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = The Mail on Sunday\u001b[39m ==> ENTITY: \u001b[32mThe Mail on Sunday\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnewspaper[0.050]; published[0.048]; libel[0.046]; libel[0.046]; Mail[0.046]; Mail[0.046]; Mail[0.046]; editor[0.046]; reported[0.046]; publisher[0.046]; apology[0.045]; apology[0.045]; apology[0.045]; article[0.045]; retraction[0.045]; retraction[0.045]; alleging[0.045]; Sunday[0.044]; Sunday[0.044]; story[0.044]; Newspapers[0.044]; flight[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cannes Film Festival ==> ENTITY: \u001b[32mCannes Film Festival\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.085:0.085[0]; log p(e|m)= -0.046:-0.046[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSunday[0.044]; Sunday[0.044]; Sunday[0.044]; actress[0.044]; Monday[0.043]; accepted[0.043]; Miss[0.043]; Miss[0.043]; French[0.042]; began[0.042]; published[0.042]; Actress[0.041]; Actress[0.041]; accepts[0.041]; exclusive[0.041]; journalists[0.041]; article[0.041]; delayed[0.040]; page[0.040]; page[0.040]; page[0.040]; story[0.040]; libel[0.040]; libel[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = retraction ==> ENTITY: \u001b[32mRetraction\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.162:0.162[0]; log p(e|m)= -0.792:-0.792[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nretraction[0.049]; article[0.046]; article[0.046]; apology[0.043]; published[0.042]; newspaper[0.042]; newspaper[0.042]; page[0.042]; page[0.042]; page[0.042]; page[0.042]; journalists[0.042]; left[0.041]; editor[0.041]; editor[0.041]; story[0.041]; connection[0.041]; publisher[0.040]; reported[0.040]; alleging[0.040]; suit[0.040]; paper[0.039]; paper[0.039]; totally[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = libel\u001b[39m ==> ENTITY: \u001b[32mDefamation\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.152:0.152[0]; log p(e|m)= -0.010:-0.010[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlibel[0.050]; damages[0.046]; damages[0.046]; alleging[0.044]; filed[0.043]; newspaper[0.043]; apology[0.042]; apology[0.042]; suit[0.042]; Newspapers[0.042]; public[0.041]; reported[0.041]; substantial[0.041]; false[0.041]; article[0.040]; page[0.040]; page[0.040]; Mail[0.040]; exclusive[0.040]; questioned[0.040]; delayed[0.039]; French[0.039]; police[0.039]; police[0.039]; \t\n\r [===========================================>..] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 700/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19981109_1172.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Indies ==> ENTITY: \u001b[32mWest Indies cricket team\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.143:0.143[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCaribbean[0.048]; batsman[0.045]; Lara[0.045]; Lara[0.045]; Lara[0.045]; Cricket[0.044]; South[0.043]; captain[0.042]; captain[0.042]; captain[0.042]; Indies[0.042]; Wednesday[0.041]; tour[0.041]; tour[0.041]; players[0.040]; players[0.040]; ahead[0.040]; West[0.040]; going[0.040]; Adams[0.039]; Africa[0.039]; travel[0.039]; came[0.038]; vice[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = treasurer ==> ENTITY: \u001b[32mTreasurer\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.062:-0.062[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBoard[0.045]; reinstated[0.044]; reinstated[0.044]; David[0.043]; vice[0.043]; Adams[0.043]; Clive[0.042]; Richardson[0.042]; Lloyd[0.041]; announcement[0.041]; Walsh[0.041]; Wednesday[0.041]; Monday[0.041]; Monday[0.041]; member[0.041]; offered[0.041]; manager[0.041]; Agency[0.041]; Caribbean[0.041]; firm[0.041]; Indies[0.040]; Indies[0.040]; Indies[0.040]; Cricket[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = marketing ==> ENTITY: \u001b[32mMarketing\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.100:0.100[0]; log p(e|m)= -0.022:-0.022[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsports[0.044]; help[0.044]; firm[0.043]; South[0.042]; South[0.042]; South[0.042]; sponsorship[0.042]; sponsorship[0.042]; African[0.042]; African[0.042]; African[0.042]; announcement[0.042]; directly[0.041]; pay[0.041]; deal[0.041]; players[0.041]; players[0.041]; players[0.041]; West[0.041]; David[0.041]; involved[0.041]; involved[0.041]; manager[0.040]; began[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = sponsorship\u001b[39m ==> ENTITY: \u001b[32mSponsor (commercial)\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsponsorship[0.055]; marketing[0.051]; sports[0.051]; deal[0.046]; help[0.045]; Team[0.045]; Monday[0.045]; Monday[0.045]; Sunday[0.045]; announcement[0.045]; players[0.045]; players[0.045]; players[0.045]; weekend[0.044]; Friday[0.044]; firm[0.044]; offered[0.044]; David[0.044]; member[0.044]; cricket[0.043]; involved[0.043]; involved[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = negotiated ==> ENTITY: \u001b[32mNegotiation\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.130:0.130[0]; log p(e|m)= -1.650:-1.650[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDavid[0.051]; Joel[0.045]; involved[0.045]; involved[0.045]; association[0.044]; deal[0.044]; Clive[0.043]; West[0.043]; West[0.043]; West[0.043]; demanding[0.043]; reinstated[0.043]; reinstated[0.043]; players[0.043]; players[0.043]; players[0.043]; Ali[0.043]; pay[0.043]; settlement[0.042]; talks[0.042]; talks[0.042]; insisted[0.042]; offered[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = West Indies Cricket Board\u001b[39m ==> ENTITY: \u001b[32mWest Indies Cricket Board\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCaribbean[0.046]; cricket[0.044]; reinstated[0.043]; reinstated[0.043]; Wednesday[0.042]; vice[0.042]; Monday[0.042]; Monday[0.042]; players[0.042]; players[0.042]; players[0.042]; manager[0.041]; Indies[0.041]; Indies[0.041]; tour[0.041]; tour[0.041]; tour[0.041]; agreement[0.041]; negotiated[0.041]; Team[0.041]; insisted[0.040]; captain[0.040]; captain[0.040]; captain[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = wicketkeeper\u001b[39m ==> ENTITY: \u001b[32mWicket-keeper\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.126:0.126[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.050]; Clive[0.043]; South[0.043]; South[0.043]; South[0.043]; David[0.043]; Lara[0.042]; Richardson[0.042]; African[0.041]; African[0.041]; African[0.041]; Adams[0.041]; runs[0.041]; sitting[0.041]; players[0.041]; players[0.041]; players[0.041]; Jimmy[0.041]; Bacher[0.041]; WICB[0.040]; sports[0.040]; Ali[0.040]; Johannesburg[0.040]; Hooper[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ali Bacher\u001b[39m ==> ENTITY: \u001b[32mAli Bacher\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.045]; Cricket[0.044]; African[0.044]; African[0.044]; South[0.043]; South[0.043]; South[0.043]; runs[0.042]; sports[0.042]; batsman[0.041]; David[0.041]; captain[0.041]; captain[0.041]; reinstated[0.041]; reinstated[0.041]; Africa[0.041]; Johannesburg[0.041]; players[0.041]; players[0.041]; players[0.041]; vice[0.040]; tour[0.040]; tour[0.040]; tour[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Courtney Walsh\u001b[39m ==> ENTITY: \u001b[32mCourtney Walsh\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.052]; Lara[0.048]; Lara[0.048]; Lara[0.048]; Cricket[0.047]; Caribbean[0.047]; ahead[0.046]; Indies[0.046]; Indies[0.046]; Jimmy[0.045]; Adams[0.045]; Wednesday[0.044]; better[0.044]; South[0.044]; going[0.044]; tour[0.044]; tour[0.044]; superstar[0.044]; Hooper[0.044]; Hooper[0.044]; Hooper[0.044]; players[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lara ==> ENTITY: \u001b[32mBrian Lara\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.157:0.157[0]; log p(e|m)= -2.674:-2.674[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.046]; Lara[0.045]; Lara[0.045]; players[0.044]; players[0.044]; players[0.044]; Cricket[0.044]; News[0.043]; Wednesday[0.042]; Monday[0.042]; captain[0.041]; captain[0.041]; captain[0.041]; told[0.040]; Team[0.040]; Adams[0.040]; Caribbean[0.040]; South[0.040]; South[0.040]; Indies[0.040]; Indies[0.040]; fired[0.040]; insisted[0.039]; ahead[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = David Richardson ==> ENTITY: \u001b[32mDave Richardson\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.110:0.110[0]; log p(e|m)= -2.830:-2.830[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClive[0.047]; cricket[0.045]; Monday[0.043]; Monday[0.043]; players[0.043]; players[0.043]; players[0.043]; Friday[0.043]; sports[0.042]; Joel[0.042]; manager[0.042]; wicketkeeper[0.042]; Lloyd[0.041]; Sunday[0.041]; South[0.041]; South[0.041]; South[0.041]; Jimmy[0.040]; Adams[0.040]; sponsorship[0.040]; sponsorship[0.040]; firm[0.040]; Hooper[0.040]; WICB[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = batsman\u001b[39m ==> ENTITY: \u001b[32mBatting (cricket)\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.054]; Cricket[0.048]; South[0.043]; South[0.043]; players[0.042]; players[0.042]; players[0.042]; Clive[0.041]; Lara[0.041]; Lara[0.041]; Lara[0.041]; captain[0.041]; captain[0.041]; captain[0.041]; African[0.041]; Sunday[0.040]; better[0.040]; Wednesday[0.040]; Caribbean[0.040]; ahead[0.040]; Monday[0.039]; Monday[0.039]; Bacher[0.039]; Adams[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Johannesburg ==> ENTITY: \u001b[32mJohannesburg\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.168:0.168[0]; log p(e|m)= -0.095:-0.095[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.044]; South[0.044]; South[0.044]; cricket[0.044]; tour[0.043]; tour[0.043]; African[0.042]; African[0.042]; African[0.042]; Sunday[0.042]; West[0.042]; sports[0.041]; Monday[0.041]; Monday[0.041]; players[0.041]; players[0.041]; players[0.041]; Friday[0.041]; weekend[0.041]; sponsorship[0.040]; sponsorship[0.040]; runs[0.040]; deal[0.040]; began[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Caribbean News Agency\u001b[39m ==> ENTITY: \u001b[32mCaribbean News Agency\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.051:0.051[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBacher[0.044]; vice[0.044]; Monday[0.043]; Wednesday[0.043]; head[0.042]; head[0.042]; head[0.042]; agreement[0.042]; association[0.041]; crisis[0.041]; LONDON[0.041]; Africa[0.041]; Courtney[0.041]; Cricket[0.041]; Nov[0.041]; behalf[0.041]; members[0.041]; Team[0.041]; tour[0.041]; tour[0.041]; South[0.041]; batsman[0.041]; Board[0.041]; leaked[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Clive Lloyd\u001b[39m ==> ENTITY: \u001b[32mClive Lloyd\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.047]; cricket[0.047]; wicketkeeper[0.045]; Lara[0.044]; Lara[0.044]; David[0.043]; captain[0.043]; captain[0.043]; Adams[0.042]; WICB[0.042]; Indies[0.041]; South[0.040]; South[0.040]; South[0.040]; South[0.040]; players[0.040]; players[0.040]; players[0.040]; West[0.040]; runs[0.040]; Bacher[0.040]; Garner[0.040]; Jimmy[0.040]; Richardson[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hooper ==> ENTITY: \u001b[32mCarl Hooper\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.167:0.167[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.047]; Cricket[0.043]; Lara[0.043]; Lara[0.043]; Lara[0.043]; captain[0.042]; captain[0.042]; captain[0.042]; players[0.042]; players[0.042]; players[0.042]; Hooper[0.042]; Hooper[0.042]; going[0.042]; Adams[0.041]; settlement[0.041]; South[0.041]; South[0.041]; Wednesday[0.041]; ahead[0.040]; Monday[0.040]; Indies[0.040]; Indies[0.040]; African[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Joel Garner\u001b[39m ==> ENTITY: \u001b[32mJoel Garner\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.122:0.122[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.047]; David[0.044]; Lara[0.044]; wicketkeeper[0.043]; Jimmy[0.042]; captain[0.042]; captain[0.042]; Richardson[0.042]; Clive[0.042]; runs[0.042]; players[0.042]; players[0.042]; players[0.042]; Adams[0.041]; Bacher[0.041]; Lloyd[0.041]; vice[0.040]; manager[0.040]; Hooper[0.040]; Hooper[0.040]; WICB[0.040]; sitting[0.040]; better[0.040]; better[0.040]; \t\n\r [============================================>.] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 718/727 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW19981022_0630.htm ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Interpol ==> ENTITY: \u001b[32mInterpol\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.169:0.169[0]; log p(e|m)= -0.419:-0.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInterpol[0.047]; Interpol[0.047]; Interpol[0.047]; Interpol[0.047]; International[0.042]; international[0.042]; crime[0.042]; crime[0.042]; crime[0.042]; said[0.041]; said[0.041]; said[0.041]; Organization[0.041]; opening[0.041]; Thursday[0.040]; told[0.040]; criminal[0.040]; wants[0.040]; revamp[0.040]; pace[0.040]; Criminal[0.040]; organized[0.039]; organization[0.039]; organization[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = organized crime ==> ENTITY: \u001b[32mOrganized crime\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.012:-0.012[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapanese[0.046]; crime[0.045]; crime[0.045]; crime[0.045]; increased[0.045]; criminality[0.043]; drug[0.043]; syndicates[0.043]; criminal[0.042]; international[0.042]; general[0.041]; called[0.041]; discuss[0.040]; discuss[0.040]; money[0.040]; expected[0.040]; Criminal[0.040]; wants[0.040]; laundering[0.040]; organization[0.040]; organization[0.040]; officials[0.040]; officials[0.040]; terrorism[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mevidence\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mScientific method\u001b[39m <---> \u001b[32mEvidence\u001b[39m\t\nSCORES: global= 0.247:0.245[\u001b[31m0.002\u001b[39m]; local()= 0.072:0.100[\u001b[32m0.027\u001b[39m]; log p(e|m)= 0.000:-0.503[\u001b[31m0.503\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nprosecution[0.047]; police[0.045]; police[0.045]; police[0.045]; Criminal[0.045]; drug[0.045]; crime[0.044]; crime[0.044]; require[0.044]; examining[0.044]; states[0.044]; international[0.044]; gather[0.043]; risks[0.043]; forms[0.043]; development[0.042]; official[0.042]; Police[0.042]; January[0.042]; failure[0.042]; assembly[0.042]; assembly[0.042]; needs[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = money laundering ==> ENTITY: \u001b[32mMoney laundering\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.157:0.157[0]; log p(e|m)= -0.026:-0.026[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nprosecution[0.044]; crime[0.044]; crime[0.044]; international[0.043]; criminality[0.043]; member[0.043]; illegal[0.042]; January[0.042]; told[0.042]; drug[0.041]; officials[0.041]; officials[0.041]; Criminal[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; International[0.041]; session[0.041]; trade[0.041]; terrorism[0.041]; evidence[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = crime syndicates ==> ENTITY: \u001b[32mOrganized crime\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.135:0.135[0]; log p(e|m)= -0.546:-0.546[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncrime[0.047]; crime[0.047]; criminal[0.043]; wants[0.042]; international[0.042]; Criminal[0.042]; fight[0.042]; organized[0.042]; Organization[0.042]; discussed[0.042]; organization[0.041]; organization[0.041]; greater[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; Interpol[0.041]; Interpol[0.041]; Interpol[0.041]; Interpol[0.041]; Interpol[0.041]; Japanese[0.040]; adopted[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = illegal drug trade ==> ENTITY: \u001b[32mIllegal drug trade\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.029:-0.029[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncrime[0.048]; crime[0.048]; criminality[0.044]; criminal[0.043]; terrorism[0.043]; states[0.042]; discuss[0.041]; discuss[0.041]; international[0.041]; increased[0.041]; laundering[0.041]; strategic[0.041]; officials[0.041]; officials[0.041]; money[0.041]; Organization[0.041]; organization[0.040]; discussed[0.040]; Interpol[0.040]; Interpol[0.040]; Interpol[0.040]; Interpol[0.040]; Interpol[0.040]; greater[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = terrorism ==> ENTITY: \u001b[32mTerrorism\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.120:0.120[0]; log p(e|m)= -0.182:-0.182[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninternational[0.061]; criminal[0.059]; Criminal[0.057]; sophisticated[0.057]; crime[0.056]; crime[0.056]; illegal[0.056]; states[0.056]; official[0.055]; organization[0.055]; organization[0.055]; Organization[0.055]; group[0.055]; general[0.054]; criminality[0.054]; cooperation[0.053]; trade[0.053]; member[0.053]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = strategy ==> ENTITY: \u001b[32mStrategy\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.135:0.135[0]; log p(e|m)= -0.569:-0.569[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstrategy[0.045]; needs[0.044]; organization[0.044]; organization[0.044]; environment[0.043]; sophisticated[0.043]; said[0.042]; said[0.042]; said[0.042]; general[0.042]; greater[0.042]; international[0.041]; Thursday[0.041]; session[0.041]; told[0.041]; risks[0.041]; effectively[0.040]; changed[0.040]; Organization[0.040]; revamp[0.040]; fight[0.040]; president[0.040]; modernize[0.039]; modernize[0.039]; \t\n\r [=============================================>] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b Tot: 7m17s | Step: 23h59m\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 727/727 \n\r==> time to test 1 sample = 5.9228586199523ms\t\n==> \u001b[31mAQUAINT\u001b[39m AQUAINT ; EPOCH = 307: Micro recall = 88.03% ; Micro F1 = \u001b[31m89.51%\u001b[39m\t\n ===> entity frequency stats :\t\nfreq = 11-20 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\nfreq = 2-5 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\nfreq = 50+ : num = 681 ; correctly classified = 635 ; perc = 93.25\t\nfreq = 1 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\nfreq = 21-50 : num = 3 ; correctly classified = 3 ; perc = 100.00\t\nfreq = 0 : num = 1 ; correctly classified = 1 ; perc = 100.00\t\nfreq = 6-10 : num = 1 ; correctly classified = 1 ; perc = 100.00\t\n\t\n ===> entity p(e|m) stats :\t\np(e|m) = <=0.001 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\np(e|m) = 0.01-0.03 : num = 1 ; correctly classified = 1 ; perc = 100.00\t\np(e|m) = 0.003-0.01 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\np(e|m) = 0.3+ : num = 634 ; correctly classified = 601 ; perc = 94.79\t\np(e|m) = 0.1-0.3 : num = 38 ; correctly classified = 28 ; perc = 73.68\t\np(e|m) = 0.03-0.1 : num = 12 ; correctly classified = 10 ; perc = 83.33\t\np(e|m) = 0.001-0.003 : num = 1 ; correctly classified = 0 ; perc = 0.00\t\n num_mentions_w/o_gold_ent_in_candidates = 41 total num mentions in dataset = 727\t\n percentage_mentions_w/o_gold_ent_in_candidates = 5.64%; percentage_mentions_solved : both_pem_ours = 79.78%; only_pem_not_ours = 3.44%; only_ours_not_pem = 8.25%; not_ours_not_pem = 2.89%\t\n\n===> MSNBC; num mentions = 656\t\n [..............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 0ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 0/656 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 16454435 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Los Angeles ==> ENTITY: \u001b[32mLos Angeles\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= -0.003:-0.003[0]; log p(e|m)= -0.101:-0.101[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngineers[0.045]; Engineers[0.045]; Planet[0.044]; onboard[0.044]; onboard[0.044]; Earth[0.043]; Propulsion[0.043]; NASA[0.042]; help[0.042]; mission[0.042]; agency[0.042]; John[0.041]; Jet[0.041]; project[0.041]; Laboratory[0.041]; landing[0.040]; time[0.040]; roll[0.040]; charge[0.040]; Mars[0.040]; Mars[0.040]; smarter[0.040]; said[0.040]; aging[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Earth ==> ENTITY: \u001b[32mEarth\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.078:-0.078[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEarth[0.046]; rock[0.045]; rock[0.045]; Mars[0.043]; help[0.043]; missions[0.042]; clouds[0.042]; time[0.042]; time[0.042]; save[0.041]; Spirit[0.041]; Spirit[0.041]; piece[0.041]; NASA[0.041]; allowing[0.041]; previously[0.041]; images[0.041]; upgrades[0.041]; use[0.040]; day[0.040]; landed[0.040]; robot[0.040]; decide[0.040]; decide[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Opportunity\u001b[39m ==> ENTITY: \u001b[32mOpportunity (rover)\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMars[0.046]; rovers[0.046]; rovers[0.046]; crater[0.045]; Opportunity[0.045]; Martian[0.044]; geologic[0.043]; scientists[0.043]; Spirit[0.042]; Spirit[0.042]; rocks[0.041]; exploration[0.041]; water[0.041]; months[0.040]; soil[0.040]; Jan[0.040]; evidence[0.040]; missions[0.040]; sized[0.039]; problem[0.039]; near[0.039]; possible[0.039]; landed[0.039]; long[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mars ==> ENTITY: \u001b[32mMars\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.393:-0.393[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmissions[0.045]; Martian[0.045]; crater[0.044]; robots[0.044]; robots[0.044]; rovers[0.043]; rovers[0.043]; sized[0.043]; longevity[0.042]; time[0.042]; time[0.042]; Spirit[0.041]; Spirit[0.041]; allowing[0.041]; scientists[0.041]; landed[0.041]; Opportunity[0.041]; Opportunity[0.041]; originally[0.040]; planned[0.040]; near[0.040]; rocks[0.040]; admit[0.040]; die[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mars ==> ENTITY: \u001b[32mMars\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.163:0.163[0]; log p(e|m)= -0.393:-0.393[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEarth[0.046]; missions[0.046]; missions[0.046]; robot[0.045]; robots[0.045]; rovers[0.044]; rovers[0.044]; sized[0.044]; day[0.044]; sending[0.044]; images[0.044]; longevity[0.043]; time[0.043]; time[0.043]; Spirit[0.042]; Spirit[0.042]; robotic[0.042]; high[0.042]; allowing[0.042]; scientists[0.042]; scientists[0.042]; scientists[0.042]; landed[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NASA\u001b[39m ==> ENTITY: \u001b[32mNASA\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngineers[0.044]; Engineers[0.044]; robotic[0.043]; Propulsion[0.043]; mission[0.043]; John[0.043]; Mars[0.041]; Mars[0.041]; scientists[0.041]; scientists[0.041]; rovers[0.041]; rovers[0.041]; rovers[0.041]; rovers[0.041]; rovers[0.041]; rovers[0.041]; rovers[0.041]; Laboratory[0.041]; flight[0.041]; project[0.041]; agency[0.041]; landing[0.041]; getting[0.040]; images[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spirit ==> ENTITY: \u001b[32mSpirit (rover)\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.177:0.177[0]; log p(e|m)= -1.252:-1.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMars[0.047]; rovers[0.047]; rovers[0.047]; Martian[0.044]; Opportunity[0.044]; Opportunity[0.044]; crater[0.043]; Spirit[0.043]; scientists[0.042]; geologic[0.042]; geologic[0.042]; rocks[0.042]; exploration[0.041]; ways[0.041]; originally[0.040]; soil[0.040]; problem[0.040]; missions[0.040]; previously[0.039]; Callas[0.039]; fourth[0.039]; possible[0.039]; landed[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spirit ==> ENTITY: \u001b[32mSpirit (rover)\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.193:0.193[0]; log p(e|m)= -1.252:-1.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMars[0.046]; rovers[0.046]; rovers[0.046]; Opportunity[0.043]; Opportunity[0.043]; Earth[0.043]; Spirit[0.042]; scientists[0.042]; scientists[0.042]; scientists[0.042]; geologic[0.042]; exploration[0.041]; rock[0.041]; rock[0.041]; images[0.041]; robotic[0.041]; help[0.040]; dust[0.040]; originally[0.040]; devils[0.040]; problem[0.040]; missions[0.040]; missions[0.040]; clouds[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mars rovers\u001b[39m ==> ENTITY: \u001b[32mMars rover\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrovers[0.047]; rovers[0.047]; rovers[0.047]; NASA[0.047]; Mars[0.045]; flight[0.044]; mission[0.043]; onboard[0.043]; onboard[0.043]; landing[0.042]; Laboratory[0.041]; Engineers[0.040]; Engineers[0.040]; LOS[0.040]; agency[0.040]; time[0.040]; Propulsion[0.040]; Callas[0.039]; computers[0.039]; computers[0.039]; help[0.039]; twin[0.038]; project[0.038]; roll[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Red Planet ==> ENTITY: \u001b[32mMars\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.169:0.169[0]; log p(e|m)= -1.363:-1.363[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMars[0.048]; Mars[0.048]; Earth[0.046]; NASA[0.046]; mission[0.044]; images[0.043]; John[0.043]; rovers[0.043]; rovers[0.043]; rovers[0.043]; rovers[0.043]; rovers[0.043]; rovers[0.043]; rovers[0.043]; landing[0.043]; help[0.042]; transmit[0.042]; scientists[0.042]; aimed[0.042]; clouds[0.042]; dust[0.042]; robotic[0.042]; LOS[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMartian\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMars\u001b[39m <---> \u001b[32mMartian\u001b[39m\t\nSCORES: global= 0.260:0.231[\u001b[31m0.029\u001b[39m]; local()= 0.142:0.115[\u001b[31m0.027\u001b[39m]; log p(e|m)= -1.273:-0.559[\u001b[32m0.714\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncrater[0.046]; Mars[0.046]; geologic[0.044]; geologic[0.044]; rocks[0.044]; Opportunity[0.043]; Opportunity[0.043]; missions[0.043]; rovers[0.042]; rovers[0.042]; robots[0.042]; robots[0.042]; Spirit[0.041]; Spirit[0.041]; scientists[0.041]; months[0.040]; massive[0.040]; sized[0.040]; samples[0.040]; ridge[0.040]; possible[0.040]; long[0.039]; soil[0.039]; previously[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Opportunity\u001b[39m ==> ENTITY: \u001b[32mOpportunity (rover)\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMars[0.048]; rovers[0.048]; rovers[0.048]; Opportunity[0.046]; geologic[0.045]; Earth[0.045]; scientists[0.044]; scientists[0.044]; scientists[0.044]; Spirit[0.044]; Spirit[0.044]; exploration[0.043]; dust[0.043]; images[0.042]; robotic[0.042]; months[0.042]; Jan[0.041]; missions[0.041]; missions[0.041]; future[0.041]; clouds[0.041]; sized[0.041]; problem[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Earth ==> ENTITY: \u001b[32mEarth\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.078:-0.078[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEarth[0.047]; rock[0.047]; rock[0.047]; Mars[0.045]; help[0.044]; help[0.044]; Planet[0.044]; onboard[0.043]; clouds[0.043]; time[0.043]; time[0.043]; save[0.043]; piece[0.043]; humans[0.043]; NASA[0.043]; John[0.043]; mission[0.043]; images[0.042]; upgrades[0.042]; Red[0.042]; use[0.042]; day[0.042]; robot[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mars rovers\u001b[39m ==> ENTITY: \u001b[32mMars rover\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrovers[0.046]; rovers[0.046]; rovers[0.046]; rovers[0.046]; rovers[0.046]; NASA[0.046]; Mars[0.044]; flight[0.043]; mission[0.042]; onboard[0.042]; onboard[0.042]; landing[0.041]; Earth[0.041]; dust[0.040]; Laboratory[0.040]; charge[0.040]; ability[0.039]; Engineers[0.039]; Engineers[0.039]; LOS[0.039]; agency[0.039]; time[0.039]; Propulsion[0.039]; Callas[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spirit ==> ENTITY: \u001b[32mSpirit (rover)\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.205:0.205[0]; log p(e|m)= -1.252:-1.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMars[0.047]; Mars[0.047]; rovers[0.047]; rovers[0.047]; Martian[0.045]; Opportunity[0.044]; Opportunity[0.044]; Opportunity[0.044]; Earth[0.044]; crater[0.044]; Spirit[0.043]; Spirit[0.043]; scientists[0.043]; scientists[0.043]; geologic[0.042]; geologic[0.042]; rocks[0.042]; exploration[0.042]; rock[0.042]; help[0.041]; originally[0.041]; soil[0.040]; problem[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jet Propulsion Laboratory\u001b[39m ==> ENTITY: \u001b[32mJet Propulsion Laboratory\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNASA[0.049]; scientists[0.044]; scientists[0.044]; Engineers[0.044]; Engineers[0.044]; Mars[0.043]; Mars[0.043]; mission[0.042]; LOS[0.041]; Earth[0.041]; Earth[0.041]; robotic[0.041]; agency[0.041]; rovers[0.040]; rovers[0.040]; rovers[0.040]; rovers[0.040]; rovers[0.040]; rovers[0.040]; rovers[0.040]; tech[0.040]; flight[0.040]; onboard[0.040]; onboard[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Opportunity\u001b[39m ==> ENTITY: \u001b[32mOpportunity (rover)\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.221:0.221[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMars[0.047]; Mars[0.047]; rovers[0.047]; rovers[0.047]; crater[0.046]; Opportunity[0.045]; Opportunity[0.045]; Martian[0.044]; geologic[0.044]; geologic[0.044]; Earth[0.044]; scientists[0.043]; scientists[0.043]; Spirit[0.043]; Spirit[0.043]; Spirit[0.043]; rocks[0.042]; exploration[0.042]; water[0.041]; months[0.041]; soil[0.041]; Jan[0.040]; evidence[0.040]; \t\n\r [>.............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 3s530ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 17/656 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 16442287 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = New Year's resolution\u001b[39m ==> ENTITY: \u001b[32mNew Year's resolution\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nresolutions[0.044]; resolutions[0.044]; healthy[0.043]; healthy[0.043]; healthy[0.043]; action[0.042]; make[0.042]; Exercise[0.042]; Resolutions[0.042]; Healthy[0.042]; fasting[0.042]; doctor[0.042]; sooner[0.041]; sooner[0.041]; visit[0.041]; list[0.041]; better[0.041]; numbers[0.041]; foods[0.041]; regularly[0.041]; Make[0.041]; priority[0.040]; stand[0.040]; recommendations[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = HDL-cholesterol\u001b[39m ==> ENTITY: \u001b[32mHigh-density lipoprotein\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.223:0.223[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncholesterol[0.046]; cholesterol[0.046]; triglyceride[0.044]; blood[0.043]; blood[0.043]; blood[0.043]; Cholesterol[0.043]; Cholesterol[0.043]; LDL[0.042]; healthy[0.042]; levels[0.042]; Triglycerides[0.042]; Clinical[0.041]; dL[0.040]; dL[0.040]; dL[0.040]; dL[0.040]; dL[0.040]; study[0.040]; Healthy[0.039]; compounds[0.039]; antioxidants[0.039]; higher[0.039]; diet[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Raspberries ==> ENTITY: \u001b[32mRaspberry\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.154:0.154[0]; log p(e|m)= -0.451:-0.451[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nvarieties[0.047]; brown[0.046]; Blueberries[0.044]; amaranth[0.043]; white[0.043]; Strawberries[0.043]; Walnuts[0.042]; Clinical[0.042]; dL[0.042]; dL[0.042]; naturally[0.042]; foods[0.041]; foods[0.041]; rice[0.041]; Artichokes[0.041]; starch[0.040]; like[0.040]; Instead[0.040]; Cloves[0.040]; Cranberries[0.040]; millet[0.040]; Nutrition[0.040]; plant[0.040]; healthy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Strawberries ==> ENTITY: \u001b[32mStrawberry\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.129:0.129[0]; log p(e|m)= -0.104:-0.104[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nvarieties[0.049]; amaranth[0.045]; Blueberries[0.043]; Clinical[0.042]; dL[0.042]; dL[0.042]; dL[0.042]; rice[0.042]; plant[0.042]; foods[0.041]; foods[0.041]; type[0.041]; type[0.041]; brown[0.041]; naturally[0.041]; Walnuts[0.041]; Cranberries[0.041]; Raspberries[0.041]; like[0.040]; Nutrition[0.040]; white[0.040]; Artichokes[0.040]; millet[0.040]; Cloves[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Triglycerides\u001b[39m ==> ENTITY: \u001b[32mTriglyceride\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncholesterol[0.051]; cholesterol[0.051]; triglyceride[0.048]; blood[0.047]; blood[0.047]; blood[0.047]; Cholesterol[0.046]; Cholesterol[0.046]; Cholesterol[0.046]; diet[0.045]; levels[0.044]; sugar[0.044]; LDL[0.044]; compounds[0.044]; antioxidants[0.044]; Clinical[0.044]; dL[0.044]; dL[0.044]; dL[0.044]; dL[0.044]; dL[0.044]; healthy[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Year ==> ENTITY: \u001b[32mNew Year\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.095:0.095[0]; log p(e|m)= -0.082:-0.082[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfasting[0.047]; resolutions[0.043]; resolutions[0.043]; sugar[0.043]; healthy[0.042]; healthy[0.042]; healthy[0.042]; resolution[0.042]; sooner[0.042]; sooner[0.042]; action[0.042]; make[0.042]; Exercise[0.042]; Resolutions[0.041]; Healthy[0.041]; visit[0.041]; list[0.041]; naturally[0.041]; sleep[0.041]; foods[0.041]; better[0.040]; numbers[0.040]; body[0.040]; regularly[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Whole Wheat\u001b[39m ==> ENTITY: \u001b[32mWhole grain\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.246:0.246[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwheat[0.047]; wheat[0.047]; grain[0.045]; grain[0.045]; grain[0.045]; varieties[0.043]; healthy[0.043]; fat[0.042]; eat[0.042]; eat[0.042]; amaranth[0.041]; pasta[0.041]; bread[0.041]; bread[0.041]; bread[0.041]; bread[0.041]; bread[0.041]; bread[0.041]; millet[0.041]; ingredients[0.041]; type[0.039]; type[0.039]; Grain[0.039]; different[0.037]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mArtichokes\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mJerusalem artichoke\u001b[39m <---> \u001b[32mArtichoke\u001b[39m\t\nSCORES: global= 0.280:0.274[\u001b[31m0.006\u001b[39m]; local()= 0.189:0.161[\u001b[31m0.028\u001b[39m]; log p(e|m)= -1.280:-0.384[\u001b[32m0.896\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \namaranth[0.046]; varieties[0.045]; Walnuts[0.044]; Blueberries[0.043]; Clinical[0.042]; dL[0.042]; dL[0.042]; wheat[0.042]; pasta[0.042]; starch[0.042]; Strawberries[0.042]; foods[0.042]; foods[0.042]; grain[0.041]; grain[0.041]; grains[0.041]; rice[0.041]; millet[0.041]; diet[0.040]; plant[0.040]; Nutrition[0.040]; Raspberries[0.040]; Food[0.040]; Cloves[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = blood sugars\u001b[39m ==> ENTITY: \u001b[32mBlood sugar\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.216:0.216[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncholesterol[0.044]; cholesterol[0.044]; blood[0.044]; blood[0.044]; blood[0.044]; levels[0.042]; diet[0.042]; Cholesterol[0.042]; Cholesterol[0.042]; Cholesterol[0.042]; triglyceride[0.042]; Triglycerides[0.041]; foods[0.041]; foods[0.041]; Clinical[0.041]; dL[0.041]; dL[0.041]; dL[0.041]; dL[0.041]; dL[0.041]; Exercise[0.041]; sugar[0.040]; LDL[0.040]; help[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Joy Bauer\u001b[39m ==> ENTITY: \u001b[32mJoy Bauer\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnutritionist[0.046]; Nutritionist[0.045]; Healthy[0.044]; healthy[0.044]; healthy[0.044]; healthy[0.044]; Joy[0.043]; Joy[0.043]; doctor[0.043]; sleep[0.042]; Exercise[0.042]; cholesterol[0.041]; created[0.040]; regularly[0.040]; learn[0.040]; Bauer[0.040]; blood[0.040]; blood[0.040]; broken[0.040]; triglyceride[0.040]; Don[0.040]; visit[0.040]; Know[0.039]; priority[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Joy Bauer\u001b[39m ==> ENTITY: \u001b[32mJoy Bauer\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnutritionist[0.050]; Nutritionist[0.049]; Healthy[0.048]; healthy[0.047]; healthy[0.047]; healthy[0.047]; Joy[0.046]; Joy[0.046]; mmHg[0.046]; doctor[0.046]; Total[0.046]; sleep[0.045]; Exercise[0.045]; know[0.044]; Optimal[0.044]; cholesterol[0.044]; cholesterol[0.044]; corrections[0.043]; created[0.043]; regularly[0.043]; learn[0.043]; Bauer[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cranberries ==> ENTITY: \u001b[32mCranberry\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.298:-0.298[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBlueberries[0.050]; varieties[0.050]; brown[0.047]; Clinical[0.047]; dL[0.047]; dL[0.047]; white[0.046]; amaranth[0.046]; Walnuts[0.046]; Strawberries[0.046]; foods[0.045]; foods[0.045]; Nutrition[0.044]; Raspberries[0.044]; Artichokes[0.044]; starch[0.044]; pasta[0.044]; wheat[0.044]; wheat[0.044]; plant[0.044]; rice[0.043]; millet[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = American Journal of Clinical Nutrition\u001b[39m ==> ENTITY: \u001b[32mThe American Journal of Clinical Nutrition\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.099:0.099[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstudy[0.044]; dL[0.044]; dL[0.044]; dL[0.044]; dL[0.044]; dL[0.044]; cholesterol[0.044]; harmful[0.042]; compounds[0.042]; antioxidants[0.042]; published[0.041]; Cholesterol[0.041]; Cholesterol[0.041]; Cholesterol[0.041]; diet[0.040]; mg[0.040]; mg[0.040]; mg[0.040]; mg[0.040]; mg[0.040]; daily[0.040]; Triglycerides[0.040]; mmHg[0.040]; healthy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = LDL-cholesterol\u001b[39m ==> ENTITY: \u001b[32mLow-density lipoprotein\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.237:0.237[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncholesterol[0.047]; cholesterol[0.047]; triglyceride[0.044]; Cholesterol[0.043]; Cholesterol[0.043]; blood[0.042]; blood[0.042]; blood[0.042]; Triglycerides[0.042]; HDL[0.042]; levels[0.042]; antioxidants[0.041]; compounds[0.041]; healthy[0.041]; healthy[0.041]; study[0.040]; Clinical[0.040]; dL[0.040]; dL[0.040]; dL[0.040]; dL[0.040]; dL[0.040]; harmful[0.039]; higher[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Joy Bauer\u001b[39m ==> ENTITY: \u001b[32mJoy Bauer\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.096:0.096[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnutrition[0.047]; expert[0.046]; healthy[0.045]; pounds[0.043]; www[0.042]; possible[0.042]; spoon[0.042]; researchers[0.041]; times[0.041]; women[0.041]; women[0.041]; women[0.041]; women[0.041]; day[0.041]; day[0.041]; sound[0.041]; additional[0.041]; gained[0.041]; com[0.041]; information[0.041]; eat[0.040]; eating[0.040]; given[0.040]; visit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Blueberries\u001b[39m ==> ENTITY: \u001b[32mBlueberry\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.141:0.141[0]; log p(e|m)= -0.002:-0.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nvarieties[0.046]; amaranth[0.045]; brown[0.044]; Walnuts[0.043]; Clinical[0.043]; dL[0.043]; dL[0.043]; white[0.042]; naturally[0.042]; confused[0.041]; compounds[0.041]; Cranberries[0.041]; diet[0.041]; plant[0.041]; Strawberries[0.041]; like[0.041]; type[0.041]; type[0.041]; foods[0.040]; foods[0.040]; Nutrition[0.040]; healthy[0.040]; Raspberries[0.040]; starch[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nutritionist\u001b[39m ==> ENTITY: \u001b[32mNutritionist\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndoctor[0.049]; nutritionist[0.048]; cholesterol[0.044]; healthy[0.044]; healthy[0.044]; healthy[0.044]; Healthy[0.044]; mmHg[0.042]; regularly[0.041]; foods[0.041]; levels[0.041]; sleep[0.040]; body[0.040]; Exercise[0.040]; know[0.040]; fasting[0.040]; Optimal[0.040]; blood[0.040]; blood[0.040]; blood[0.040]; list[0.040]; sugar[0.039]; corrections[0.039]; recommendations[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Blackberries ==> ENTITY: \u001b[32mBlackberry\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.110:0.110[0]; log p(e|m)= -0.538:-0.538[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nvarieties[0.048]; amaranth[0.044]; brown[0.044]; Clinical[0.043]; dL[0.043]; dL[0.043]; dL[0.043]; Walnuts[0.043]; Blueberries[0.043]; white[0.042]; Artichokes[0.041]; Strawberries[0.041]; Nutrition[0.041]; buy[0.040]; wheat[0.040]; Cloves[0.040]; Raspberries[0.040]; rice[0.040]; millet[0.040]; mg[0.040]; mg[0.040]; mg[0.040]; Cranberries[0.040]; Triglycerides[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Joy\u001b[39m ==> ENTITY: \u001b[32mJoy Bauer\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.099:0.099[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndL[0.046]; dL[0.046]; nutritionist[0.045]; Nutritionist[0.044]; Healthy[0.043]; healthy[0.043]; healthy[0.043]; healthy[0.043]; mg[0.042]; mg[0.042]; Joy[0.042]; Joy[0.042]; mmHg[0.042]; doctor[0.041]; Total[0.041]; sleep[0.040]; Exercise[0.040]; Cholesterol[0.040]; Cholesterol[0.040]; know[0.040]; LDL[0.040]; Optimal[0.040]; cholesterol[0.039]; cholesterol[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Whole Grain\u001b[39m ==> ENTITY: \u001b[32mWhole grain\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.241:0.241[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhealthy[0.044]; fat[0.043]; eat[0.043]; eat[0.043]; pasta[0.043]; bread[0.042]; bread[0.042]; bread[0.042]; bread[0.042]; bread[0.042]; bread[0.042]; ingredients[0.042]; grain[0.042]; grain[0.042]; wheat[0.042]; wheat[0.042]; amaranth[0.041]; varieties[0.041]; type[0.040]; type[0.040]; Wheat[0.040]; different[0.039]; Food[0.039]; like[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Walnuts ==> ENTITY: \u001b[32mWalnut\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.120:0.120[0]; log p(e|m)= -0.030:-0.030[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nvarieties[0.047]; amaranth[0.047]; brown[0.046]; Clinical[0.044]; dL[0.044]; dL[0.044]; dL[0.044]; American[0.044]; white[0.044]; Blueberries[0.044]; wheat[0.043]; Cloves[0.043]; naturally[0.043]; pasta[0.043]; millet[0.043]; healthy[0.042]; like[0.042]; Nutrition[0.042]; bread[0.042]; bread[0.042]; bread[0.042]; bread[0.042]; Artichokes[0.042]; \t\n\r [=>............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 3s228ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 38/656 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 16453733 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rosie O'Donnell\u001b[39m ==> ENTITY: \u001b[32mRosie O'Donnell\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhost[0.046]; comedian[0.045]; Rosie[0.045]; View[0.043]; View[0.043]; claimed[0.042]; Walters[0.042]; Walters[0.042]; Walters[0.042]; Walters[0.042]; Barbara[0.042]; Barbara[0.042]; daytime[0.042]; ABC[0.041]; Donnell[0.040]; Donnell[0.040]; Donnell[0.040]; claim[0.040]; creator[0.040]; Trump[0.040]; Trump[0.040]; Trump[0.040]; Donald[0.040]; chat[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tara Conner\u001b[39m ==> ENTITY: \u001b[32mTara Conner\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConner[0.044]; USA[0.044]; USA[0.044]; USA[0.044]; Miss[0.043]; Miss[0.043]; Miss[0.043]; Miss[0.043]; month[0.042]; View[0.041]; interview[0.041]; won[0.041]; Walters[0.041]; Walters[0.041]; crown[0.041]; Trump[0.040]; Trump[0.040]; Trump[0.040]; Trump[0.040]; Trump[0.040]; twice[0.040]; title[0.040]; announced[0.040]; conference[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Walters\u001b[39m ==> ENTITY: \u001b[32mBarbara Walters\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninterview[0.058]; ABC[0.056]; Walters[0.054]; Walters[0.054]; Walters[0.054]; Walters[0.054]; View[0.053]; asked[0.053]; told[0.052]; week[0.052]; announced[0.052]; Barbara[0.052]; Barbara[0.052]; personal[0.052]; month[0.051]; Rosie[0.051]; Trump[0.050]; Trump[0.050]; Trump[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Trump\u001b[39m ==> ENTITY: \u001b[32mDonald Trump\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nApprentice[0.051]; Trump[0.048]; Trump[0.048]; Trump[0.048]; Trump[0.048]; NBC[0.045]; host[0.044]; interviews[0.042]; announced[0.041]; mogul[0.041]; month[0.040]; September[0.039]; ratings[0.039]; Rosie[0.039]; View[0.039]; America[0.039]; includes[0.039]; joined[0.039]; working[0.038]; Miss[0.038]; Miss[0.038]; Miss[0.038]; Miss[0.038]; twice[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = O'Donnell\u001b[39m ==> ENTITY: \u001b[32mRosie O'Donnell\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhost[0.048]; comedian[0.047]; Rosie[0.046]; Rosie[0.046]; personal[0.044]; View[0.044]; View[0.044]; View[0.044]; claimed[0.043]; Walters[0.043]; Walters[0.043]; Walters[0.043]; Walters[0.043]; Walters[0.043]; Walters[0.043]; Barbara[0.043]; daytime[0.043]; interview[0.042]; asked[0.042]; ABC[0.042]; ABC[0.042]; filed[0.042]; filed[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = ABC ==> ENTITY: \u001b[32mAmerican Broadcasting Company\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.102:0.102[0]; log p(e|m)= -0.334:-0.334[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhost[0.045]; daytime[0.044]; Wednesday[0.044]; ratings[0.043]; comedian[0.043]; creator[0.043]; week[0.043]; got[0.042]; Walters[0.041]; Walters[0.041]; Walters[0.041]; Walters[0.041]; Walters[0.041]; Walters[0.041]; Trump[0.041]; Trump[0.041]; Trump[0.041]; took[0.041]; Donald[0.041]; Donald[0.041]; View[0.041]; View[0.041]; View[0.041]; brought[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = O'Donnell\u001b[39m ==> ENTITY: \u001b[32mRosie O'Donnell\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRosie[0.048]; personal[0.046]; View[0.045]; claimed[0.045]; Walters[0.044]; Walters[0.044]; Barbara[0.044]; Barbara[0.044]; interview[0.043]; September[0.043]; filed[0.043]; filed[0.043]; America[0.043]; won[0.043]; news[0.043]; Conner[0.043]; Conner[0.043]; Miss[0.043]; Miss[0.043]; Miss[0.043]; Miss[0.043]; announced[0.042]; month[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rosie O'Donnell\u001b[39m ==> ENTITY: \u001b[32mRosie O'Donnell\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhost[0.046]; comedian[0.045]; Rosie[0.045]; View[0.043]; View[0.043]; claimed[0.042]; Walters[0.042]; Walters[0.042]; Walters[0.042]; Walters[0.042]; Walters[0.042]; Walters[0.042]; Barbara[0.042]; Barbara[0.042]; daytime[0.042]; ABC[0.041]; Donnell[0.040]; Donnell[0.040]; Donnell[0.040]; claim[0.040]; creator[0.040]; Trump[0.040]; Trump[0.040]; Trump[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = The View\u001b[39m ==> ENTITY: \u001b[32mThe View (U.S. TV series)\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nABC[0.047]; ABC[0.047]; interview[0.045]; daytime[0.044]; week[0.044]; View[0.043]; Walters[0.042]; Walters[0.042]; Walters[0.042]; Walters[0.042]; Walters[0.042]; Wednesday[0.041]; told[0.041]; told[0.041]; Barbara[0.040]; Barbara[0.040]; Donald[0.040]; Donald[0.040]; people[0.040]; time[0.039]; Trump[0.039]; Trump[0.039]; Trump[0.039]; chat[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = America ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.054:0.054[0]; log p(e|m)= -0.785:-0.785[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUSA[0.043]; USA[0.043]; USA[0.043]; month[0.043]; Donnell[0.043]; Donnell[0.043]; interviews[0.042]; September[0.042]; host[0.042]; includes[0.042]; news[0.042]; divorced[0.042]; calling[0.042]; said[0.042]; announced[0.041]; outspoken[0.041]; Teen[0.041]; conference[0.041]; old[0.040]; Organization[0.040]; media[0.040]; moral[0.040]; twice[0.040]; joined[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Trump\u001b[39m ==> ENTITY: \u001b[32mDonald Trump\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nApprentice[0.051]; Trump[0.049]; Trump[0.049]; Trump[0.049]; NBC[0.046]; host[0.045]; interviews[0.042]; announced[0.041]; mogul[0.041]; month[0.041]; Sunday[0.040]; September[0.040]; View[0.039]; America[0.039]; includes[0.039]; joined[0.039]; Miss[0.039]; Miss[0.039]; Miss[0.039]; Miss[0.039]; twice[0.039]; USA[0.039]; USA[0.039]; USA[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Barbara\u001b[39m ==> ENTITY: \u001b[32mBarbara Walters\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninterview[0.047]; ABC[0.045]; Walters[0.043]; Walters[0.043]; View[0.043]; asked[0.043]; told[0.042]; announced[0.042]; Barbara[0.042]; personal[0.042]; month[0.041]; Rosie[0.041]; Trump[0.041]; Trump[0.041]; Trump[0.041]; Trump[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; Press[0.040]; people[0.040]; Donald[0.040]; Donald[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Walters\u001b[39m ==> ENTITY: \u001b[32mBarbara Walters\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninterview[0.053]; ABC[0.050]; ABC[0.050]; Walters[0.048]; Walters[0.048]; Walters[0.048]; Walters[0.048]; Walters[0.048]; View[0.048]; View[0.048]; asked[0.048]; daytime[0.047]; told[0.047]; week[0.047]; personal[0.046]; Wednesday[0.046]; Rosie[0.046]; Trump[0.045]; Trump[0.045]; Trump[0.045]; Trump[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Trump\u001b[39m ==> ENTITY: \u001b[32mDonald Trump\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTrump[0.049]; Trump[0.049]; host[0.045]; ABC[0.044]; ABC[0.044]; comedian[0.043]; Wednesday[0.042]; Donald[0.042]; Donald[0.042]; week[0.042]; bankrupt[0.041]; creator[0.041]; took[0.040]; open[0.040]; ratings[0.040]; daytime[0.040]; Rosie[0.040]; Rosie[0.040]; View[0.040]; View[0.040]; View[0.040]; chat[0.039]; didn[0.039]; returns[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Miss Universe Organization\u001b[39m ==> ENTITY: \u001b[32mMiss Universe Organization\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUSA[0.044]; USA[0.044]; USA[0.044]; Miss[0.044]; Miss[0.044]; Miss[0.044]; Trump[0.043]; Trump[0.043]; Trump[0.043]; Trump[0.043]; Teen[0.042]; month[0.042]; Apprentice[0.042]; NBC[0.042]; host[0.041]; Walters[0.041]; America[0.040]; Rosie[0.040]; won[0.040]; ratings[0.040]; season[0.040]; includes[0.039]; conference[0.039]; September[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Donald Trump\u001b[39m ==> ENTITY: \u001b[32mDonald Trump\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTrump[0.059]; Trump[0.059]; host[0.054]; ABC[0.053]; comedian[0.051]; Wednesday[0.050]; Donald[0.050]; week[0.050]; creator[0.049]; took[0.048]; open[0.048]; ratings[0.048]; daytime[0.048]; Rosie[0.048]; Rosie[0.048]; View[0.048]; View[0.048]; chat[0.047]; didn[0.047]; returns[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = The Associated Press\u001b[39m ==> ENTITY: \u001b[32mAssociated Press\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nweek[0.048]; interview[0.046]; told[0.046]; USA[0.046]; month[0.046]; said[0.044]; said[0.044]; said[0.044]; said[0.044]; said[0.044]; ABC[0.044]; announced[0.044]; won[0.043]; time[0.043]; personal[0.043]; quote[0.042]; claimed[0.042]; asked[0.042]; appreciated[0.042]; times[0.042]; Tara[0.042]; ratings[0.041]; retraction[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Miss USA\u001b[39m ==> ENTITY: \u001b[32mMiss USA\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMiss[0.043]; Miss[0.043]; Miss[0.043]; Conner[0.043]; Conner[0.043]; USA[0.043]; USA[0.043]; Trump[0.042]; Trump[0.042]; Trump[0.042]; Trump[0.042]; host[0.042]; title[0.042]; crown[0.042]; won[0.042]; Teen[0.041]; month[0.041]; twice[0.041]; America[0.041]; Universe[0.040]; View[0.040]; claimed[0.040]; Apprentice[0.040]; moderator[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Trump\u001b[39m ==> ENTITY: \u001b[32mDonald Trump\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTrump[0.052]; Trump[0.052]; host[0.048]; ABC[0.047]; comedian[0.045]; Wednesday[0.044]; Donald[0.044]; week[0.044]; creator[0.043]; ratings[0.042]; daytime[0.042]; Rosie[0.042]; Rosie[0.042]; View[0.042]; View[0.042]; chat[0.042]; didn[0.041]; Walters[0.041]; Walters[0.041]; Walters[0.041]; Walters[0.041]; wanted[0.041]; prove[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Trump\u001b[39m ==> ENTITY: \u001b[32mDonald Trump\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTrump[0.049]; Trump[0.049]; Trump[0.049]; Trump[0.049]; business[0.043]; bankruptcy[0.042]; bankruptcy[0.042]; announced[0.041]; mogul[0.041]; month[0.041]; interview[0.040]; September[0.040]; bankruptcies[0.040]; ratings[0.040]; Rosie[0.040]; View[0.040]; America[0.040]; includes[0.039]; joined[0.039]; working[0.039]; Miss[0.039]; Miss[0.039]; Miss[0.039]; Miss[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = ABC ==> ENTITY: \u001b[32mAmerican Broadcasting Company\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.082:0.082[0]; log p(e|m)= -0.334:-0.334[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nweek[0.044]; got[0.043]; interview[0.043]; time[0.042]; ratings[0.042]; ratings[0.042]; View[0.042]; Trump[0.042]; Trump[0.042]; Trump[0.042]; took[0.042]; won[0.042]; Walters[0.042]; Walters[0.042]; Walters[0.042]; Walters[0.042]; Walters[0.042]; Donald[0.041]; Donald[0.041]; times[0.041]; returns[0.041]; personal[0.040]; Press[0.040]; brought[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Walters\u001b[39m ==> ENTITY: \u001b[32mBarbara Walters\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhost[0.045]; ABC[0.044]; ABC[0.044]; Walters[0.042]; Walters[0.042]; Walters[0.042]; Walters[0.042]; Walters[0.042]; Walters[0.042]; View[0.042]; View[0.042]; View[0.042]; comedian[0.042]; asked[0.042]; daytime[0.041]; told[0.041]; week[0.041]; Barbara[0.041]; Barbara[0.041]; personal[0.040]; Wednesday[0.040]; wanted[0.040]; Rosie[0.040]; Rosie[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = O'Donnell\u001b[39m ==> ENTITY: \u001b[32mRosie O'Donnell\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhost[0.048]; View[0.044]; NBC[0.044]; media[0.042]; September[0.042]; America[0.042]; Apprentice[0.042]; interviews[0.041]; news[0.041]; Conner[0.041]; Conner[0.041]; Miss[0.041]; Miss[0.041]; Miss[0.041]; Miss[0.041]; announced[0.041]; month[0.041]; Trump[0.041]; Trump[0.041]; Trump[0.041]; outspoken[0.041]; divorced[0.040]; includes[0.040]; joined[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Walters\u001b[39m ==> ENTITY: \u001b[32mBarbara Walters\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhost[0.049]; ABC[0.048]; Walters[0.046]; Walters[0.046]; Walters[0.046]; Walters[0.046]; Walters[0.046]; View[0.046]; View[0.046]; comedian[0.046]; daytime[0.046]; told[0.045]; week[0.045]; Barbara[0.045]; Barbara[0.045]; Wednesday[0.044]; wanted[0.044]; Rosie[0.044]; Rosie[0.044]; Trump[0.044]; Trump[0.044]; Trump[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Barbara\u001b[39m ==> ENTITY: \u001b[32mBarbara Walters\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninterview[0.049]; ABC[0.047]; Walters[0.045]; Walters[0.045]; View[0.045]; View[0.045]; asked[0.044]; told[0.044]; announced[0.044]; Barbara[0.043]; personal[0.043]; month[0.043]; Rosie[0.042]; Trump[0.042]; Trump[0.042]; Trump[0.042]; Trump[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; Press[0.042]; people[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NBC ==> ENTITY: \u001b[32mNBC\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.064:-0.064[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhost[0.047]; Sunday[0.046]; season[0.045]; Apprentice[0.045]; includes[0.044]; America[0.043]; USA[0.043]; USA[0.043]; media[0.042]; interviews[0.042]; news[0.042]; twice[0.042]; View[0.042]; September[0.041]; conference[0.041]; joined[0.041]; Conner[0.040]; calling[0.040]; Trump[0.039]; Trump[0.039]; Trump[0.039]; Miss[0.039]; Miss[0.039]; Miss[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = The Donald\u001b[39m ==> ENTITY: \u001b[32mDonald Trump\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTrump[0.051]; Trump[0.051]; Trump[0.051]; Trump[0.051]; business[0.043]; interview[0.041]; ABC[0.041]; ABC[0.041]; View[0.041]; View[0.041]; Donald[0.041]; Donald[0.041]; ratings[0.040]; ratings[0.040]; personal[0.040]; casino[0.040]; week[0.039]; Walters[0.039]; Walters[0.039]; Walters[0.039]; Walters[0.039]; Walters[0.039]; Walters[0.039]; asked[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Miss Teen USA\u001b[39m ==> ENTITY: \u001b[32mMiss Teen USA\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUSA[0.044]; USA[0.044]; Miss[0.044]; Miss[0.044]; Miss[0.044]; title[0.043]; crown[0.043]; host[0.043]; won[0.042]; America[0.041]; Conner[0.041]; Conner[0.041]; month[0.041]; interviews[0.041]; conference[0.041]; View[0.041]; Trump[0.041]; Trump[0.041]; Trump[0.041]; Trump[0.041]; twice[0.040]; time[0.040]; NBC[0.040]; ratings[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = The View\u001b[39m ==> ENTITY: \u001b[32mThe View (U.S. TV series)\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhost[0.049]; NBC[0.049]; interviews[0.046]; Apprentice[0.045]; month[0.045]; Rosie[0.044]; work[0.044]; America[0.044]; September[0.044]; announced[0.043]; Sunday[0.043]; news[0.043]; drinking[0.043]; told[0.043]; joined[0.042]; Barbara[0.042]; Barbara[0.042]; time[0.042]; twice[0.042]; Trump[0.041]; Trump[0.041]; Trump[0.041]; Trump[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Trump\u001b[39m ==> ENTITY: \u001b[32mDonald Trump\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTrump[0.049]; Trump[0.049]; Trump[0.049]; ABC[0.044]; business[0.043]; Donald[0.042]; Donald[0.042]; bankruptcy[0.041]; bankruptcy[0.041]; announced[0.041]; week[0.041]; bankrupt[0.041]; month[0.041]; interview[0.040]; took[0.040]; open[0.040]; bankruptcies[0.040]; ratings[0.040]; Rosie[0.040]; View[0.040]; working[0.039]; Miss[0.039]; USA[0.039]; like[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Walters\u001b[39m ==> ENTITY: \u001b[32mBarbara Walters\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninterview[0.052]; ABC[0.050]; ABC[0.050]; Walters[0.048]; Walters[0.048]; Walters[0.048]; Walters[0.048]; Walters[0.048]; Walters[0.048]; View[0.048]; View[0.048]; comedian[0.048]; asked[0.047]; daytime[0.047]; told[0.047]; week[0.047]; Barbara[0.046]; personal[0.046]; Wednesday[0.046]; wanted[0.045]; Rosie[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = The View\u001b[39m ==> ENTITY: \u001b[32mThe View (U.S. TV series)\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.206:0.206[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nABC[0.051]; host[0.050]; daytime[0.047]; week[0.047]; View[0.046]; View[0.046]; comedian[0.046]; Walters[0.046]; Walters[0.046]; Walters[0.046]; Walters[0.046]; Walters[0.046]; Walters[0.046]; Rosie[0.045]; Rosie[0.045]; Wednesday[0.045]; told[0.044]; Barbara[0.043]; Barbara[0.043]; Donald[0.043]; Donald[0.043]; creator[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Miss USA\u001b[39m ==> ENTITY: \u001b[32mMiss USA\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMiss[0.043]; Miss[0.043]; Miss[0.043]; Conner[0.042]; Conner[0.042]; USA[0.042]; USA[0.042]; Trump[0.042]; Trump[0.042]; Trump[0.042]; Trump[0.042]; Trump[0.042]; Walters[0.042]; Walters[0.042]; title[0.042]; crown[0.042]; won[0.041]; Teen[0.041]; month[0.041]; twice[0.041]; America[0.040]; Universe[0.040]; View[0.040]; interview[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Barbara Walters\u001b[39m ==> ENTITY: \u001b[32mBarbara Walters\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhost[0.046]; ABC[0.045]; Walters[0.043]; Walters[0.043]; Walters[0.043]; View[0.043]; View[0.043]; comedian[0.043]; daytime[0.042]; told[0.042]; Barbara[0.042]; Wednesday[0.041]; wanted[0.041]; Rosie[0.041]; Rosie[0.041]; Trump[0.041]; Trump[0.041]; Trump[0.041]; said[0.040]; said[0.040]; said[0.040]; chat[0.040]; Donald[0.040]; denies[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.082:0.082[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwanted[0.044]; host[0.043]; week[0.043]; ABC[0.042]; vacation[0.042]; vacation[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; comedian[0.042]; View[0.042]; View[0.042]; didn[0.041]; returns[0.041]; denies[0.041]; told[0.041]; Wednesday[0.041]; Trump[0.041]; Trump[0.041]; Trump[0.041]; daytime[0.041]; creator[0.041]; Donald[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Donald Trump\u001b[39m ==> ENTITY: \u001b[32mDonald Trump\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTrump[0.057]; ABC[0.051]; business[0.051]; Donald[0.049]; bankruptcy[0.048]; bankruptcy[0.048]; week[0.048]; bankrupt[0.048]; interview[0.047]; took[0.047]; open[0.047]; bankruptcies[0.047]; ratings[0.047]; ratings[0.047]; Rosie[0.046]; View[0.046]; working[0.046]; like[0.045]; personal[0.045]; casino[0.045]; returns[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = The Apprentice ==> ENTITY: \u001b[32mThe Apprentice (U.S. TV series)\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.161:0.161[0]; log p(e|m)= -0.701:-0.701[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNBC[0.047]; host[0.046]; Trump[0.045]; Trump[0.045]; Trump[0.045]; interviews[0.044]; season[0.043]; America[0.042]; USA[0.041]; USA[0.041]; joined[0.041]; mogul[0.041]; twice[0.040]; Miss[0.040]; Miss[0.040]; Miss[0.040]; includes[0.040]; divorced[0.040]; Conner[0.040]; Donnell[0.040]; loser[0.040]; September[0.040]; Sunday[0.040]; View[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rosie\u001b[39m ==> ENTITY: \u001b[32mRosie O'Donnell\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npersonal[0.048]; View[0.048]; claimed[0.047]; Walters[0.046]; Walters[0.046]; Barbara[0.046]; Barbara[0.046]; interview[0.046]; September[0.045]; filed[0.045]; filed[0.045]; won[0.045]; news[0.045]; Conner[0.045]; Conner[0.045]; Miss[0.045]; Miss[0.045]; Miss[0.045]; Miss[0.045]; Donnell[0.045]; announced[0.045]; month[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = O'Donnell\u001b[39m ==> ENTITY: \u001b[32mRosie O'Donnell\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npersonal[0.045]; View[0.045]; View[0.045]; claimed[0.044]; claimed[0.044]; Walters[0.044]; Walters[0.044]; Walters[0.044]; Walters[0.044]; Walters[0.044]; Walters[0.044]; Barbara[0.044]; Barbara[0.044]; daytime[0.044]; interview[0.043]; asked[0.043]; ABC[0.043]; ABC[0.043]; filed[0.043]; filed[0.043]; Donnell[0.042]; Donnell[0.042]; creator[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = O'Donnell\u001b[39m ==> ENTITY: \u001b[32mRosie O'Donnell\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhost[0.046]; comedian[0.045]; Rosie[0.045]; Rosie[0.045]; View[0.042]; View[0.042]; View[0.042]; claimed[0.042]; Walters[0.041]; Walters[0.041]; Walters[0.041]; Walters[0.041]; Walters[0.041]; Walters[0.041]; Barbara[0.041]; Barbara[0.041]; daytime[0.041]; asked[0.040]; ABC[0.040]; ABC[0.040]; Donnell[0.040]; Donnell[0.040]; Donnell[0.040]; Donnell[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Trump\u001b[39m ==> ENTITY: \u001b[32mDonald Trump\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nApprentice[0.053]; Trump[0.051]; Trump[0.051]; Trump[0.051]; NBC[0.047]; host[0.047]; interviews[0.044]; announced[0.043]; mogul[0.043]; month[0.042]; Sunday[0.042]; September[0.041]; Rosie[0.041]; View[0.041]; America[0.041]; includes[0.041]; joined[0.041]; working[0.040]; Miss[0.040]; Miss[0.040]; Miss[0.040]; Miss[0.040]; twice[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Conner\u001b[39m ==> ENTITY: \u001b[32mTara Conner\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConner[0.044]; USA[0.044]; USA[0.044]; USA[0.044]; Miss[0.043]; Miss[0.043]; Miss[0.043]; Miss[0.043]; month[0.042]; host[0.042]; View[0.041]; won[0.041]; crown[0.041]; Apprentice[0.041]; Trump[0.041]; Trump[0.041]; Trump[0.041]; Trump[0.041]; twice[0.040]; title[0.040]; announced[0.040]; conference[0.040]; NBC[0.040]; Sunday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Walters\u001b[39m ==> ENTITY: \u001b[32mBarbara Walters\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninterview[0.047]; ABC[0.045]; Walters[0.043]; Walters[0.043]; Walters[0.043]; View[0.043]; asked[0.043]; told[0.042]; announced[0.042]; Barbara[0.041]; Barbara[0.041]; personal[0.041]; month[0.041]; Rosie[0.041]; Trump[0.040]; Trump[0.040]; Trump[0.040]; Trump[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; Press[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Barbara Walters\u001b[39m ==> ENTITY: \u001b[32mBarbara Walters\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhost[0.045]; ABC[0.045]; Walters[0.043]; Walters[0.043]; Walters[0.043]; Walters[0.043]; View[0.043]; View[0.043]; comedian[0.043]; daytime[0.042]; told[0.042]; week[0.042]; Barbara[0.041]; Wednesday[0.041]; wanted[0.041]; Rosie[0.040]; Rosie[0.040]; Trump[0.040]; Trump[0.040]; Trump[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\r [====>.........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 2s291ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 82/656 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 16455207 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Miami Sol\u001b[39m ==> ENTITY: \u001b[32mMiami Sol\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.046]; season[0.046]; season[0.046]; WNBA[0.045]; WNBA[0.045]; team[0.044]; team[0.044]; team[0.044]; seasons[0.042]; franchise[0.042]; NBA[0.041]; NBA[0.041]; games[0.041]; Heat[0.040]; Heat[0.040]; coached[0.040]; final[0.040]; stint[0.039]; game[0.039]; finishing[0.039]; expansion[0.038]; coach[0.038]; coach[0.038]; coach[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = WNBA\u001b[39m ==> ENTITY: \u001b[32mWomen's National Basketball Association\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.223:0.223[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWNBA[0.048]; NBA[0.045]; NBA[0.045]; team[0.045]; team[0.045]; team[0.045]; games[0.043]; franchise[0.042]; game[0.041]; seasons[0.041]; Wednesday[0.041]; Wednesday[0.041]; coached[0.040]; Heat[0.040]; Heat[0.040]; final[0.040]; coach[0.040]; coach[0.040]; coach[0.040]; Miami[0.040]; season[0.040]; season[0.040]; season[0.040]; Detroit[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Riley\u001b[39m ==> ENTITY: \u001b[32mPat Riley\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.057]; championship[0.054]; championship[0.054]; games[0.054]; season[0.051]; season[0.051]; season[0.051]; Neal[0.051]; coach[0.050]; Gundy[0.049]; Wade[0.048]; Stan[0.048]; Heat[0.048]; Heat[0.048]; Miami[0.048]; bench[0.048]; offseason[0.048]; miss[0.048]; Wednesday[0.047]; Clippers[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = O'Neal\u001b[39m ==> ENTITY: \u001b[32mShaquille O'Neal\u001b[39m\t\nSCORES: global= 0.313:0.313[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.044]; games[0.043]; Wade[0.043]; game[0.043]; game[0.043]; finals[0.043]; Showtime[0.042]; playoffs[0.042]; championship[0.042]; Lakers[0.042]; Wednesday[0.042]; Wednesday[0.042]; MVP[0.042]; Phoenix[0.042]; career[0.042]; record[0.041]; marquee[0.041]; roster[0.041]; miss[0.041]; season[0.041]; season[0.041]; season[0.041]; Neal[0.040]; preseason[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Riley\u001b[39m ==> ENTITY: \u001b[32mPat Riley\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.049]; playoffs[0.047]; championship[0.046]; games[0.046]; finals[0.046]; seasons[0.045]; coaches[0.044]; season[0.044]; season[0.044]; game[0.044]; game[0.044]; Neal[0.043]; Neal[0.043]; MVP[0.043]; Larry[0.043]; coach[0.043]; career[0.042]; Lakers[0.042]; Wade[0.041]; title[0.041]; missed[0.041]; miss[0.040]; roster[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Riley\u001b[39m ==> ENTITY: \u001b[32mPat Riley\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSurgery[0.048]; season[0.048]; season[0.048]; Neal[0.048]; coach[0.047]; coach[0.047]; assistant[0.046]; Wade[0.046]; Heat[0.045]; Heat[0.045]; Miami[0.045]; Miami[0.045]; Wednesday[0.045]; Wednesday[0.045]; Clippers[0.044]; Riley[0.044]; Riley[0.044]; surgery[0.044]; surgery[0.044]; absence[0.044]; record[0.044]; said[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Riley\u001b[39m ==> ENTITY: \u001b[32mPat Riley\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.044]; season[0.044]; season[0.044]; coached[0.044]; coach[0.043]; coach[0.043]; Gundy[0.043]; Stan[0.042]; Heat[0.041]; Heat[0.041]; Heat[0.041]; bench[0.041]; offseason[0.041]; head[0.041]; week[0.041]; Riley[0.040]; Riley[0.040]; Riley[0.040]; Riley[0.040]; Riley[0.040]; Riley[0.040]; surgery[0.040]; surgery[0.040]; resigned[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mourning\u001b[39m ==> ENTITY: \u001b[32mAlonzo Mourning\u001b[39m\t\nSCORES: global= 0.313:0.313[0]; local()= 0.218:0.218[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNBA[0.047]; team[0.047]; games[0.045]; season[0.043]; season[0.043]; season[0.043]; franchise[0.043]; Mourning[0.042]; Riley[0.041]; Riley[0.041]; Riley[0.041]; Riley[0.041]; Riley[0.041]; Heat[0.041]; Heat[0.041]; center[0.041]; Alonzo[0.040]; Pistons[0.040]; coach[0.040]; coach[0.040]; coach[0.040]; injury[0.040]; record[0.040]; record[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dwyane Wade\u001b[39m ==> ENTITY: \u001b[32mDwyane Wade\u001b[39m\t\nSCORES: global= 0.311:0.311[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.046]; Surgery[0.046]; Shaquille[0.045]; games[0.045]; Williams[0.045]; Heat[0.044]; Heat[0.044]; Gundy[0.044]; Miami[0.044]; Neal[0.043]; miss[0.043]; Riley[0.043]; Riley[0.043]; Riley[0.043]; Riley[0.043]; offseason[0.043]; record[0.043]; Jason[0.043]; bench[0.043]; championship[0.042]; championship[0.042]; season[0.042]; season[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Riley\u001b[39m ==> ENTITY: \u001b[32mPat Riley\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.048]; team[0.048]; team[0.048]; games[0.045]; seasons[0.044]; NBA[0.044]; season[0.043]; season[0.043]; game[0.043]; coached[0.042]; coach[0.042]; final[0.041]; James[0.041]; Heat[0.040]; Miami[0.040]; Wednesday[0.039]; Wednesday[0.039]; Shaq[0.039]; Riley[0.039]; Riley[0.039]; WNBA[0.039]; WNBA[0.039]; record[0.039]; record[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Heat ==> ENTITY: \u001b[32mMiami Heat\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.228:0.228[0]; log p(e|m)= -1.470:-1.470[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNBA[0.046]; games[0.045]; team[0.045]; team[0.045]; Miami[0.044]; seasons[0.043]; final[0.043]; season[0.042]; season[0.042]; Wednesday[0.042]; Wednesday[0.042]; game[0.042]; record[0.042]; record[0.042]; James[0.041]; coached[0.040]; overall[0.040]; Shaq[0.040]; remaining[0.039]; WNBA[0.039]; WNBA[0.039]; Posey[0.039]; goal[0.039]; body[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Riley\u001b[39m ==> ENTITY: \u001b[32mPat Riley\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.049]; playoffs[0.046]; championship[0.046]; games[0.045]; finals[0.045]; seasons[0.045]; coaches[0.044]; season[0.043]; season[0.043]; season[0.043]; game[0.043]; game[0.043]; Neal[0.043]; MVP[0.043]; Larry[0.042]; coach[0.042]; coach[0.042]; career[0.042]; Phil[0.042]; Lakers[0.042]; Lakers[0.042]; Wade[0.041]; title[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Riley\u001b[39m ==> ENTITY: \u001b[32mPat Riley\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.049]; playoffs[0.046]; championship[0.046]; games[0.045]; finals[0.045]; seasons[0.045]; coaches[0.044]; season[0.043]; season[0.043]; season[0.043]; game[0.043]; game[0.043]; Neal[0.043]; MVP[0.043]; Larry[0.042]; coach[0.042]; coach[0.042]; career[0.042]; Phil[0.042]; Lakers[0.042]; Lakers[0.042]; Wade[0.041]; title[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Riley\u001b[39m ==> ENTITY: \u001b[32mPat Riley\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.044]; season[0.044]; season[0.044]; coached[0.044]; coach[0.043]; coach[0.043]; Gundy[0.043]; Stan[0.042]; Heat[0.041]; Heat[0.041]; Heat[0.041]; bench[0.041]; offseason[0.041]; head[0.041]; week[0.041]; Riley[0.040]; Riley[0.040]; Riley[0.040]; Riley[0.040]; Riley[0.040]; Riley[0.040]; surgery[0.040]; surgery[0.040]; resigned[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Los Angeles Clippers\u001b[39m ==> ENTITY: \u001b[32mLos Angeles Clippers\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.048]; season[0.048]; games[0.046]; Miami[0.045]; team[0.044]; Wednesday[0.043]; Wednesday[0.043]; Williams[0.042]; Wade[0.041]; Heat[0.041]; Heat[0.041]; Neal[0.041]; December[0.041]; Friday[0.040]; record[0.040]; championship[0.040]; championship[0.040]; basis[0.040]; bench[0.040]; shortly[0.039]; miss[0.039]; assistant[0.039]; Dwyane[0.039]; loss[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rothstein\u001b[39m ==> ENTITY: \u001b[32mRon Rothstein\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNBA[0.045]; NBA[0.045]; coach[0.044]; coach[0.044]; coach[0.044]; coached[0.044]; coached[0.044]; team[0.043]; seasons[0.042]; stint[0.042]; Miami[0.042]; head[0.041]; head[0.041]; franchise[0.041]; WNBA[0.041]; Pistons[0.041]; Heat[0.040]; Heat[0.040]; games[0.040]; record[0.040]; record[0.040]; game[0.039]; Rothstein[0.039]; knee[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Antoine Walker\u001b[39m ==> ENTITY: \u001b[32mAntoine Walker\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.225:0.225[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNBA[0.045]; James[0.045]; Miami[0.043]; team[0.043]; team[0.043]; season[0.042]; season[0.042]; Posey[0.042]; games[0.042]; preseason[0.042]; seasons[0.042]; Shaq[0.041]; game[0.041]; Wednesday[0.041]; Wednesday[0.041]; starting[0.041]; Dwyane[0.041]; Heat[0.041]; Heat[0.041]; Riley[0.041]; Riley[0.041]; WNBA[0.041]; WNBA[0.041]; struggled[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = James Posey\u001b[39m ==> ENTITY: \u001b[32mJames Posey\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNBA[0.046]; game[0.044]; Shaq[0.043]; preseason[0.043]; Dwyane[0.043]; starting[0.042]; season[0.042]; season[0.042]; team[0.042]; team[0.042]; Miami[0.042]; seasons[0.042]; WNBA[0.041]; WNBA[0.041]; struggled[0.041]; Walker[0.041]; defensive[0.041]; offensive[0.040]; Heat[0.040]; Heat[0.040]; Wednesday[0.040]; Wednesday[0.040]; forwards[0.040]; coached[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Riley\u001b[39m ==> ENTITY: \u001b[32mPat Riley\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.057]; championship[0.054]; championship[0.054]; games[0.054]; season[0.051]; season[0.051]; season[0.051]; Neal[0.051]; coach[0.050]; Gundy[0.049]; Wade[0.048]; Stan[0.048]; Heat[0.048]; Heat[0.048]; Miami[0.048]; bench[0.048]; offseason[0.048]; miss[0.048]; Wednesday[0.047]; Clippers[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Riley\u001b[39m ==> ENTITY: \u001b[32mPat Riley\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.049]; playoffs[0.047]; championship[0.046]; games[0.046]; seasons[0.045]; coaches[0.044]; season[0.044]; season[0.044]; game[0.044]; game[0.044]; Neal[0.044]; Larry[0.043]; coach[0.043]; coach[0.043]; career[0.043]; Phil[0.043]; Lakers[0.042]; Lakers[0.042]; title[0.041]; missed[0.041]; missed[0.041]; Jackson[0.041]; roster[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Los Angeles Lakers\u001b[39m ==> ENTITY: \u001b[32mLos Angeles Lakers\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.215:0.215[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; playoffs[0.044]; Lakers[0.044]; team[0.043]; missed[0.043]; missed[0.043]; seasons[0.043]; Wednesday[0.043]; championship[0.042]; Pistons[0.042]; season[0.042]; season[0.042]; loss[0.041]; Phil[0.041]; Neal[0.041]; fifth[0.041]; Jackson[0.040]; record[0.040]; seventh[0.040]; roster[0.040]; month[0.040]; Larry[0.040]; career[0.039]; run[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mChicago\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mChicago Bulls\u001b[39m <---> \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.250:0.243[\u001b[31m0.007\u001b[39m]; local()= 0.221:0.062[\u001b[31m0.159\u001b[39m]; log p(e|m)= -4.510:-0.207[\u001b[32m4.303\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNBA[0.047]; team[0.045]; season[0.044]; season[0.044]; season[0.044]; franchise[0.044]; offseason[0.044]; record[0.042]; coached[0.042]; stint[0.041]; Pistons[0.041]; week[0.041]; Detroit[0.041]; coach[0.041]; coach[0.041]; Heat[0.041]; Heat[0.041]; returned[0.040]; Stan[0.040]; September[0.040]; bench[0.040]; December[0.039]; Mourning[0.039]; Mourning[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Detroit Pistons\u001b[39m ==> ENTITY: \u001b[32mDetroit Pistons\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.046]; season[0.046]; NBA[0.044]; NBA[0.044]; team[0.044]; games[0.044]; franchise[0.044]; seasons[0.043]; game[0.042]; final[0.042]; Riley[0.041]; Riley[0.041]; Riley[0.041]; Riley[0.041]; Heat[0.041]; Heat[0.041]; Miami[0.041]; Riles[0.040]; WNBA[0.040]; WNBA[0.040]; coached[0.039]; coached[0.039]; Chicago[0.039]; finishing[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mJason Williams\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mJay Williams (basketball)\u001b[39m <---> \u001b[32mJason Williams (basketball, born 1975)\u001b[39m\t\nSCORES: global= 0.258:0.247[\u001b[31m0.011\u001b[39m]; local()= 0.190:0.180[\u001b[31m0.010\u001b[39m]; log p(e|m)= -0.955:-2.749[\u001b[31m1.794\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.051]; season[0.049]; season[0.049]; Wade[0.048]; team[0.048]; record[0.047]; offseason[0.046]; Neal[0.046]; Miami[0.045]; deciding[0.045]; loss[0.045]; Shaquille[0.045]; championship[0.044]; championship[0.044]; knee[0.044]; coach[0.044]; coach[0.044]; bench[0.044]; miss[0.044]; Wednesday[0.044]; Clippers[0.043]; returned[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Riley\u001b[39m ==> ENTITY: \u001b[32mPat Riley\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.049]; team[0.049]; games[0.046]; season[0.044]; season[0.044]; Neal[0.044]; preseason[0.042]; James[0.042]; Heat[0.041]; Heat[0.041]; Wednesday[0.041]; Wednesday[0.041]; Shaq[0.040]; Riley[0.040]; starting[0.040]; record[0.040]; struggled[0.040]; said[0.040]; said[0.040]; said[0.040]; offensive[0.040]; Posey[0.040]; met[0.039]; effort[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shaq\u001b[39m ==> ENTITY: \u001b[32mShaquille O'Neal\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.044]; team[0.044]; James[0.044]; games[0.043]; defensive[0.043]; overall[0.042]; Walker[0.042]; Wednesday[0.042]; Wednesday[0.042]; soon[0.042]; record[0.042]; Heat[0.041]; Heat[0.041]; defense[0.041]; Dwyane[0.041]; offensive[0.041]; Posey[0.041]; preseason[0.041]; season[0.041]; season[0.041]; Neal[0.041]; Riley[0.040]; Riley[0.040]; looking[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ron Rothstein\u001b[39m ==> ENTITY: \u001b[32mRon Rothstein\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nassistant[0.047]; coach[0.045]; coach[0.045]; Surgery[0.045]; interim[0.043]; Miami[0.043]; Miami[0.043]; Wednesday[0.042]; Wednesday[0.042]; Wade[0.042]; surgery[0.041]; surgery[0.041]; Heat[0.041]; Heat[0.041]; Friday[0.040]; record[0.040]; absence[0.040]; knee[0.040]; knee[0.040]; knee[0.040]; season[0.040]; season[0.040]; hip[0.040]; hip[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Larry Brown ==> ENTITY: \u001b[32mLarry Brown (basketball)\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.459:-0.459[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseasons[0.046]; coach[0.044]; coach[0.044]; coaches[0.044]; games[0.044]; season[0.044]; season[0.044]; playoffs[0.043]; career[0.042]; Jackson[0.041]; missed[0.041]; missed[0.041]; championship[0.041]; Wednesday[0.041]; roster[0.041]; surgery[0.041]; Phil[0.041]; Los[0.040]; era[0.040]; Pistons[0.040]; replaced[0.040]; record[0.040]; loss[0.039]; seventh[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pistons\u001b[39m ==> ENTITY: \u001b[32mDetroit Pistons\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayoffs[0.049]; games[0.048]; season[0.046]; season[0.046]; Lakers[0.045]; Lakers[0.045]; championship[0.045]; seasons[0.045]; missed[0.045]; missed[0.045]; roster[0.044]; Wednesday[0.043]; coaches[0.042]; Jackson[0.041]; Larry[0.041]; record[0.041]; Brown[0.041]; career[0.041]; long[0.041]; hip[0.041]; hip[0.041]; loss[0.041]; following[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMiami\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMiami Heat\u001b[39m <---> \u001b[32mMiami\u001b[39m\t\nSCORES: global= 0.255:0.245[\u001b[31m0.010\u001b[39m]; local()= 0.151:0.074[\u001b[31m0.077\u001b[39m]; log p(e|m)= -3.730:-0.354[\u001b[32m3.376\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMiami[0.049]; season[0.047]; Wednesday[0.046]; Friday[0.043]; Heat[0.043]; assistant[0.043]; coach[0.043]; coach[0.043]; spot[0.041]; began[0.041]; leave[0.041]; Pat[0.041]; Riley[0.040]; Riley[0.040]; Riley[0.040]; left[0.040]; surgery[0.040]; surgery[0.040]; knee[0.040]; knee[0.040]; knee[0.040]; Ron[0.040]; scheduled[0.040]; displaced[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Riley\u001b[39m ==> ENTITY: \u001b[32mPat Riley\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.049]; NBA[0.045]; season[0.044]; season[0.044]; franchise[0.044]; coached[0.043]; coach[0.043]; coach[0.043]; stint[0.043]; Gundy[0.042]; Stan[0.041]; Heat[0.041]; Heat[0.041]; offseason[0.041]; head[0.040]; head[0.040]; week[0.040]; Riley[0.040]; Riley[0.040]; Riley[0.040]; Riley[0.040]; Riley[0.040]; surgery[0.040]; surgery[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Heat ==> ENTITY: \u001b[32mMiami Heat\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.226:0.226[0]; log p(e|m)= -1.470:-1.470[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; team[0.046]; Miami[0.046]; Wade[0.045]; season[0.045]; season[0.045]; Neal[0.044]; Wednesday[0.044]; Wednesday[0.044]; record[0.044]; Heat[0.044]; championship[0.044]; championship[0.044]; Jason[0.043]; Clippers[0.043]; Williams[0.042]; spot[0.042]; Shaquille[0.042]; Friday[0.042]; miss[0.042]; loss[0.041]; run[0.041]; Stan[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Riley\u001b[39m ==> ENTITY: \u001b[32mPat Riley\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.049]; games[0.046]; seasons[0.046]; NBA[0.045]; NBA[0.045]; season[0.044]; season[0.044]; season[0.044]; game[0.044]; franchise[0.044]; coached[0.044]; coached[0.044]; coach[0.043]; coach[0.043]; coach[0.043]; stint[0.043]; final[0.043]; Heat[0.041]; Heat[0.041]; Heat[0.041]; Miami[0.041]; head[0.041]; head[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dwyane\u001b[39m ==> ENTITY: \u001b[32mDwyane Wade\u001b[39m\t\nSCORES: global= 0.313:0.313[0]; local()= 0.208:0.208[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nShaq[0.045]; James[0.045]; team[0.044]; team[0.044]; Posey[0.044]; games[0.043]; Heat[0.043]; Heat[0.043]; Neal[0.042]; Riley[0.042]; Riley[0.042]; record[0.041]; starting[0.041]; preseason[0.041]; forward[0.041]; Walker[0.040]; offensive[0.040]; season[0.040]; season[0.040]; effort[0.040]; struggled[0.040]; overall[0.040]; Plus[0.039]; defensive[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Heat ==> ENTITY: \u001b[32mMiami Heat\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.222:0.222[0]; log p(e|m)= -1.470:-1.470[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; team[0.044]; team[0.044]; team[0.044]; season[0.043]; season[0.043]; Wednesday[0.042]; Wednesday[0.042]; Wednesday[0.042]; game[0.042]; game[0.042]; Wade[0.042]; preseason[0.042]; finals[0.041]; MVP[0.041]; James[0.040]; supposed[0.040]; think[0.040]; Neal[0.040]; Friday[0.040]; everybody[0.040]; everybody[0.040]; frustrated[0.040]; record[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Heat ==> ENTITY: \u001b[32mMiami Heat\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.231:0.231[0]; log p(e|m)= -1.470:-1.470[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; team[0.046]; Miami[0.046]; Wade[0.045]; season[0.045]; season[0.045]; season[0.045]; Neal[0.044]; Wednesday[0.044]; record[0.044]; Heat[0.044]; championship[0.044]; championship[0.044]; Jason[0.043]; Clippers[0.043]; offseason[0.043]; Williams[0.042]; Shaquille[0.042]; miss[0.042]; loss[0.041]; run[0.041]; Stan[0.041]; reportedly[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lakers ==> ENTITY: \u001b[32mLos Angeles Lakers\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.206:0.206[0]; log p(e|m)= -0.061:-0.061[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; Lakers[0.045]; playoffs[0.045]; seasons[0.043]; Wednesday[0.043]; missed[0.042]; missed[0.042]; season[0.042]; season[0.042]; record[0.042]; Pistons[0.041]; Los[0.041]; championship[0.041]; career[0.041]; Jackson[0.041]; fifth[0.041]; seventh[0.040]; Phil[0.040]; roster[0.040]; Larry[0.040]; coaches[0.040]; month[0.040]; run[0.040]; marquee[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Phil Jackson\u001b[39m ==> ENTITY: \u001b[32mPhil Jackson\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseasons[0.046]; playoffs[0.045]; games[0.044]; coaches[0.044]; coach[0.043]; coach[0.043]; championship[0.043]; career[0.042]; season[0.042]; season[0.042]; Larry[0.042]; Lakers[0.041]; Lakers[0.041]; missed[0.041]; missed[0.041]; seventh[0.040]; Brown[0.040]; following[0.040]; fifth[0.040]; Wednesday[0.040]; month[0.040]; roster[0.040]; surgery[0.039]; guys[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Miami ==> ENTITY: \u001b[32mMiami Heat\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.237:0.237[0]; log p(e|m)= -3.730:-3.730[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; Miami[0.046]; team[0.045]; season[0.044]; season[0.044]; Wednesday[0.043]; Wednesday[0.043]; Wade[0.043]; championship[0.042]; championship[0.042]; Williams[0.041]; Clippers[0.041]; Neal[0.041]; record[0.041]; Friday[0.041]; Heat[0.041]; Heat[0.041]; Heat[0.041]; loss[0.040]; assistant[0.040]; coach[0.040]; coach[0.040]; coach[0.040]; began[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Heat ==> ENTITY: \u001b[32mMiami Heat\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.227:0.227[0]; log p(e|m)= -1.470:-1.470[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNBA[0.047]; team[0.045]; franchise[0.044]; season[0.043]; season[0.043]; record[0.043]; record[0.043]; Heat[0.043]; Pistons[0.042]; offseason[0.041]; week[0.041]; Detroit[0.041]; final[0.041]; stint[0.041]; coached[0.041]; asked[0.040]; Chicago[0.040]; Mourning[0.040]; Mourning[0.040]; apparently[0.040]; loss[0.040]; center[0.040]; following[0.040]; reportedly[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pat Riley\u001b[39m ==> ENTITY: \u001b[32mPat Riley\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSurgery[0.045]; season[0.045]; coach[0.044]; coach[0.044]; assistant[0.044]; Heat[0.042]; Miami[0.042]; Wednesday[0.042]; Riley[0.041]; Riley[0.041]; surgery[0.041]; surgery[0.041]; absence[0.041]; knee[0.041]; knee[0.041]; knee[0.041]; Ron[0.041]; began[0.040]; interim[0.040]; hip[0.040]; hip[0.040]; Friday[0.040]; taking[0.040]; going[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rothstein\u001b[39m ==> ENTITY: \u001b[32mRon Rothstein\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNBA[0.045]; NBA[0.045]; coach[0.044]; coach[0.044]; coach[0.044]; coached[0.043]; team[0.043]; team[0.043]; team[0.043]; seasons[0.042]; stint[0.042]; Coaching[0.041]; Miami[0.041]; head[0.041]; head[0.041]; franchise[0.040]; Wednesday[0.040]; Wednesday[0.040]; WNBA[0.040]; WNBA[0.040]; Pistons[0.040]; Heat[0.040]; Heat[0.040]; games[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Riley\u001b[39m ==> ENTITY: \u001b[32mPat Riley\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.052]; championship[0.049]; championship[0.049]; games[0.049]; Surgery[0.047]; season[0.047]; season[0.047]; season[0.047]; Neal[0.046]; coach[0.046]; coach[0.046]; Gundy[0.045]; Wade[0.044]; Stan[0.044]; Heat[0.043]; Heat[0.043]; Miami[0.043]; bench[0.043]; offseason[0.043]; miss[0.043]; Wednesday[0.043]; Clippers[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = O'Neal\u001b[39m ==> ENTITY: \u001b[32mShaquille O'Neal\u001b[39m\t\nSCORES: global= 0.313:0.313[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nShaq[0.046]; team[0.044]; games[0.044]; Wade[0.044]; game[0.044]; game[0.044]; finals[0.043]; Wednesday[0.042]; MVP[0.042]; Phoenix[0.042]; Heat[0.042]; Dwyane[0.041]; miss[0.041]; season[0.041]; season[0.041]; Neal[0.041]; preseason[0.041]; Friday[0.041]; straight[0.040]; offensive[0.040]; overall[0.039]; sprained[0.039]; forward[0.039]; defensive[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wade\u001b[39m ==> ENTITY: \u001b[32mDwyane Wade\u001b[39m\t\nSCORES: global= 0.312:0.312[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.044]; MVP[0.044]; games[0.043]; Heat[0.043]; game[0.043]; game[0.043]; finals[0.043]; Neal[0.042]; Neal[0.042]; miss[0.042]; Riley[0.042]; Riley[0.042]; Riley[0.042]; sprained[0.041]; record[0.041]; starting[0.041]; career[0.041]; preseason[0.041]; forward[0.041]; season[0.040]; season[0.040]; season[0.040]; effort[0.040]; Phoenix[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Stan Van Gundy\u001b[39m ==> ENTITY: \u001b[32mStan Van Gundy\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.045]; coach[0.045]; games[0.043]; offseason[0.043]; Wade[0.042]; Heat[0.042]; Heat[0.042]; Heat[0.042]; season[0.042]; season[0.042]; season[0.042]; championship[0.042]; championship[0.042]; resigned[0.042]; Riley[0.041]; Riley[0.041]; Riley[0.041]; Riley[0.041]; Riley[0.041]; Riley[0.041]; Riley[0.041]; surgery[0.041]; surgery[0.041]; Clippers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Riley\u001b[39m ==> ENTITY: \u001b[32mPat Riley\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayoffs[0.048]; championship[0.047]; games[0.047]; seasons[0.046]; coaches[0.045]; season[0.045]; season[0.045]; Larry[0.044]; coach[0.044]; coach[0.044]; career[0.044]; Phil[0.043]; Lakers[0.043]; Lakers[0.043]; title[0.042]; missed[0.042]; missed[0.042]; Jackson[0.041]; roster[0.041]; Wednesday[0.041]; head[0.041]; marquee[0.041]; dealt[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alonzo Mourning\u001b[39m ==> ENTITY: \u001b[32mAlonzo Mourning\u001b[39m\t\nSCORES: global= 0.313:0.313[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNBA[0.049]; team[0.049]; games[0.047]; offseason[0.045]; season[0.045]; season[0.045]; franchise[0.044]; Mourning[0.044]; Riley[0.043]; Riley[0.043]; Riley[0.043]; Riley[0.043]; Riley[0.043]; Riley[0.043]; Heat[0.042]; Heat[0.042]; center[0.042]; surgery[0.042]; surgery[0.042]; Pistons[0.042]; coach[0.041]; coach[0.041]; injury[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shaquille O'Neal\u001b[39m ==> ENTITY: \u001b[32mShaquille O'Neal\u001b[39m\t\nSCORES: global= 0.311:0.311[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSurgery[0.045]; team[0.044]; games[0.043]; Wade[0.043]; Miami[0.043]; Gundy[0.043]; Williams[0.043]; championship[0.042]; championship[0.042]; Wednesday[0.042]; Heat[0.041]; Heat[0.041]; Dwyane[0.041]; record[0.041]; miss[0.041]; season[0.041]; season[0.041]; surgery[0.041]; surgery[0.041]; Friday[0.041]; spot[0.040]; plays[0.040]; knee[0.040]; knee[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = WNBA\u001b[39m ==> ENTITY: \u001b[32mWomen's National Basketball Association\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.223:0.223[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWNBA[0.048]; NBA[0.045]; NBA[0.045]; team[0.045]; team[0.045]; team[0.045]; games[0.043]; franchise[0.042]; game[0.041]; seasons[0.041]; Wednesday[0.041]; Wednesday[0.041]; coached[0.040]; Heat[0.040]; Heat[0.040]; final[0.040]; coach[0.040]; coach[0.040]; coach[0.040]; Miami[0.040]; season[0.040]; season[0.040]; season[0.040]; Detroit[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Riley\u001b[39m ==> ENTITY: \u001b[32mPat Riley\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSurgery[0.045]; season[0.045]; coach[0.044]; coach[0.044]; assistant[0.044]; Heat[0.042]; Miami[0.042]; Wednesday[0.042]; Riley[0.041]; Riley[0.041]; surgery[0.041]; surgery[0.041]; absence[0.041]; knee[0.041]; knee[0.041]; knee[0.041]; Ron[0.041]; began[0.040]; interim[0.040]; hip[0.040]; hip[0.040]; Friday[0.040]; taking[0.040]; going[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Miami Heat\u001b[39m ==> ENTITY: \u001b[32mMiami Heat\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.049]; Miami[0.046]; Wednesday[0.044]; Wednesday[0.044]; record[0.044]; MIAMI[0.043]; left[0.043]; spot[0.042]; Friday[0.042]; loss[0.041]; break[0.041]; coach[0.040]; coach[0.040]; assistant[0.040]; long[0.040]; Riley[0.040]; Riley[0.040]; Riley[0.040]; began[0.040]; absence[0.040]; right[0.040]; right[0.040]; hip[0.040]; hip[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = MVP ==> ENTITY: \u001b[32mMost Valuable Player\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.206:0.206[0]; log p(e|m)= -1.076:-1.076[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npreseason[0.046]; team[0.046]; games[0.045]; game[0.045]; game[0.045]; season[0.044]; season[0.044]; season[0.044]; overall[0.042]; Wade[0.042]; finals[0.041]; record[0.041]; roster[0.041]; Phoenix[0.040]; starting[0.040]; career[0.040]; Neal[0.040]; Neal[0.040]; straight[0.039]; run[0.039]; Western[0.039]; Wednesday[0.039]; Wednesday[0.039]; effort[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPhoenix\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPhoenix Suns\u001b[39m <---> \u001b[32mPhoenix, Arizona\u001b[39m\t\nSCORES: global= 0.261:0.239[\u001b[31m0.022\u001b[39m]; local()= 0.229:0.086[\u001b[31m0.144\u001b[39m]; log p(e|m)= -3.474:-0.909[\u001b[32m2.565\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; team[0.045]; playoffs[0.043]; preseason[0.043]; game[0.043]; game[0.043]; season[0.043]; season[0.043]; season[0.043]; MVP[0.042]; Wednesday[0.042]; Wednesday[0.042]; roster[0.041]; career[0.041]; Wade[0.041]; coach[0.040]; finals[0.040]; championship[0.040]; Neal[0.040]; Neal[0.040]; Friday[0.040]; Western[0.040]; starting[0.040]; fifth[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Riley\u001b[39m ==> ENTITY: \u001b[32mPat Riley\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.049]; championship[0.046]; championship[0.046]; games[0.046]; season[0.044]; season[0.044]; coach[0.043]; Gundy[0.042]; Stan[0.041]; Heat[0.041]; Heat[0.041]; bench[0.041]; offseason[0.041]; miss[0.040]; week[0.040]; Riley[0.040]; Riley[0.040]; Riley[0.040]; Riley[0.040]; Riley[0.040]; Riley[0.040]; Riley[0.040]; surgery[0.040]; surgery[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Riley\u001b[39m ==> ENTITY: \u001b[32mPat Riley\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.048]; games[0.045]; NBA[0.044]; season[0.043]; season[0.043]; season[0.043]; franchise[0.043]; coached[0.043]; coach[0.042]; coach[0.042]; coach[0.042]; stint[0.042]; final[0.042]; Heat[0.040]; Heat[0.040]; offseason[0.040]; head[0.040]; head[0.040]; week[0.039]; Riley[0.039]; Riley[0.039]; Riley[0.039]; Riley[0.039]; Riley[0.039]; \t\n\r [========>.....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s987ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 138/656 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 16443053 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.073:0.073[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmetropolitan[0.044]; pledged[0.044]; nation[0.043]; country[0.043]; Democrats[0.043]; acknowledged[0.043]; pledges[0.043]; communities[0.043]; promised[0.043]; said[0.043]; government[0.042]; Wednesday[0.042]; Michael[0.041]; address[0.041]; advanced[0.041]; Secretary[0.041]; cost[0.040]; areas[0.040]; say[0.040]; job[0.040]; news[0.040]; security[0.040]; security[0.040]; chief[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Charles Schumer\u001b[39m ==> ENTITY: \u001b[32mChuck Schumer\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.126:0.126[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDemocrats[0.046]; York[0.045]; York[0.045]; said[0.045]; said[0.045]; Congress[0.045]; emergency[0.045]; emergency[0.045]; emergency[0.045]; week[0.044]; Sen[0.044]; leadership[0.043]; term[0.043]; ear[0.043]; majority[0.043]; says[0.042]; Center[0.042]; pay[0.042]; promised[0.042]; cost[0.042]; report[0.042]; major[0.042]; planning[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.065:0.065[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \narea[0.045]; cities[0.044]; Washington[0.044]; American[0.043]; Cleveland[0.043]; communities[0.043]; Columbus[0.042]; Paul[0.042]; includes[0.042]; suburbs[0.042]; Ohio[0.042]; week[0.042]; areas[0.041]; territories[0.041]; large[0.041]; County[0.041]; Laramie[0.041]; regions[0.041]; state[0.040]; Minneapolis[0.040]; agencies[0.040]; agencies[0.040]; percent[0.040]; percent[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Trade Center ==> ENTITY: \u001b[32mWorld Trade Center\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.098:-0.098[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntowers[0.047]; buildings[0.044]; York[0.043]; York[0.043]; surrounding[0.043]; attacks[0.042]; Charles[0.042]; crumbling[0.041]; included[0.041]; report[0.041]; problems[0.041]; talk[0.041]; talk[0.041]; operating[0.041]; said[0.041]; said[0.041]; disaster[0.041]; cost[0.041]; major[0.041]; major[0.041]; based[0.041]; systems[0.040]; systems[0.040]; gaps[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Diego ==> ENTITY: \u001b[32mSan Diego\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.188:-0.188[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \narea[0.045]; officials[0.043]; personnel[0.043]; County[0.043]; communities[0.043]; communities[0.043]; American[0.042]; state[0.042]; Washington[0.042]; agencies[0.042]; cities[0.042]; includes[0.041]; federal[0.041]; police[0.041]; report[0.041]; report[0.041]; report[0.041]; report[0.041]; La[0.041]; Cleveland[0.040]; highest[0.040]; areas[0.040]; Columbus[0.040]; Chicago[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = The Associated Press\u001b[39m ==> ENTITY: \u001b[32mAssociated Press\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.110:0.110[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWednesday[0.044]; Tuesday[0.044]; medical[0.043]; emergency[0.042]; emergency[0.042]; emergency[0.042]; personnel[0.042]; report[0.042]; report[0.042]; report[0.042]; said[0.042]; said[0.042]; obtained[0.042]; country[0.042]; released[0.041]; seamless[0.041]; overall[0.041]; provided[0.041]; communications[0.041]; rivalries[0.041]; cultural[0.040]; cultural[0.040]; highest[0.040]; think[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.S. territories\u001b[39m ==> ENTITY: \u001b[32mTerritories of the United States\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.119:0.119[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstate[0.045]; federal[0.044]; Washington[0.043]; Mandan[0.043]; Sioux[0.043]; Ohio[0.042]; agencies[0.042]; Samoa[0.042]; County[0.042]; American[0.042]; area[0.042]; Cleveland[0.041]; communities[0.041]; Columbus[0.041]; small[0.041]; Laramie[0.041]; specifically[0.041]; large[0.041]; formalized[0.041]; regions[0.040]; cities[0.040]; includes[0.040]; officials[0.040]; areas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Congress ==> ENTITY: \u001b[32mUnited States Congress\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.146:0.146[0]; log p(e|m)= -0.335:-0.335[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngovernment[0.045]; Wednesday[0.044]; Wednesday[0.044]; nation[0.044]; Tuesday[0.043]; obtained[0.043]; country[0.043]; secretary[0.043]; agencies[0.042]; acknowledged[0.042]; gave[0.042]; Secretary[0.041]; Press[0.041]; effort[0.041]; highest[0.041]; briefing[0.041]; said[0.040]; said[0.040]; provided[0.040]; report[0.040]; report[0.040]; WASHINGTON[0.040]; acknowledges[0.040]; promised[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = S.D.\u001b[39m ==> ENTITY: \u001b[32mSouth Dakota\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.102:0.102[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMandan[0.045]; state[0.045]; personnel[0.043]; federal[0.043]; Laramie[0.042]; Sioux[0.042]; agencies[0.042]; medical[0.042]; County[0.042]; emergency[0.042]; emergency[0.042]; Ohio[0.041]; American[0.041]; Washington[0.041]; overview[0.041]; battle[0.041]; area[0.041]; cities[0.041]; Wyo[0.041]; highest[0.041]; leadership[0.041]; week[0.041]; officials[0.040]; finds[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = N.D. ==> ENTITY: \u001b[32mNorth Dakota\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.047:0.047[0]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMandan[0.046]; limited[0.045]; limited[0.045]; County[0.043]; emergency[0.043]; emergency[0.043]; Emergency[0.042]; advanced[0.042]; cities[0.042]; Tammy[0.042]; coordinator[0.042]; jurisdictions[0.041]; Management[0.041]; intermediate[0.041]; study[0.041]; study[0.041]; judged[0.040]; personnel[0.040]; fall[0.040]; Office[0.040]; early[0.040]; stage[0.040]; grades[0.040]; grades[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mandan ==> ENTITY: \u001b[32mMandan, North Dakota\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.101:0.101[0]; log p(e|m)= -0.997:-0.997[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \narea[0.045]; communities[0.043]; says[0.043]; County[0.043]; state[0.042]; officials[0.042]; Sioux[0.042]; studied[0.042]; American[0.042]; areas[0.042]; cities[0.042]; week[0.042]; Washington[0.041]; territories[0.041]; suburbs[0.041]; use[0.041]; Ohio[0.041]; specifically[0.041]; study[0.041]; think[0.041]; Laramie[0.041]; finds[0.041]; place[0.041]; federal[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Columbus ==> ENTITY: \u001b[32mColumbus, Ohio\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.073:0.073[0]; log p(e|m)= -1.076:-1.076[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOhio[0.045]; Cleveland[0.045]; area[0.045]; cities[0.044]; Washington[0.044]; County[0.043]; Diego[0.042]; American[0.042]; suburbs[0.042]; communities[0.042]; communities[0.042]; state[0.041]; small[0.041]; territories[0.041]; San[0.041]; officials[0.040]; regions[0.040]; areas[0.040]; cultural[0.040]; surveyed[0.040]; La[0.040]; said[0.040]; overall[0.040]; agencies[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chertoff\u001b[39m ==> ENTITY: \u001b[32mMichael Chertoff\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.115:0.115[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChertoff[0.046]; Laramie[0.043]; Sioux[0.043]; federal[0.043]; medical[0.042]; Washington[0.042]; Wednesday[0.042]; Tuesday[0.042]; emergency[0.042]; emergency[0.042]; emergency[0.042]; said[0.042]; said[0.042]; terrorist[0.042]; longstanding[0.041]; officials[0.041]; Wyo[0.041]; Sept[0.041]; Sept[0.041]; agencies[0.041]; agencies[0.041]; Paul[0.040]; seamless[0.040]; personnel[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.066:0.066[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsays[0.044]; York[0.044]; towers[0.043]; cities[0.042]; cities[0.042]; cities[0.042]; World[0.042]; buildings[0.042]; surrounding[0.042]; revealed[0.042]; major[0.042]; major[0.042]; Charles[0.041]; communities[0.041]; term[0.041]; said[0.041]; crumbling[0.041]; Center[0.041]; scores[0.041]; local[0.041]; local[0.041]; based[0.041]; agencies[0.041]; agencies[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chertoff\u001b[39m ==> ENTITY: \u001b[32mMichael Chertoff\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChertoff[0.045]; Chertoff[0.045]; Security[0.044]; Secretary[0.044]; Homeland[0.044]; Michael[0.043]; acknowledged[0.043]; chief[0.041]; Wednesday[0.041]; Wednesday[0.041]; Tuesday[0.041]; emergency[0.041]; emergency[0.041]; emergency[0.041]; said[0.041]; said[0.041]; terrorist[0.041]; job[0.041]; communications[0.040]; communications[0.040]; Sept[0.040]; Sept[0.040]; agencies[0.040]; agencies[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cleveland ==> ENTITY: \u001b[32mCleveland\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.078:0.078[0]; log p(e|m)= -0.573:-0.573[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWashington[0.045]; County[0.044]; Ohio[0.044]; American[0.044]; area[0.043]; cities[0.043]; Columbus[0.042]; areas[0.042]; Congress[0.042]; Paul[0.041]; Chicago[0.041]; communities[0.041]; overall[0.041]; federal[0.041]; week[0.041]; regions[0.041]; state[0.041]; firefighters[0.041]; suburbs[0.041]; showed[0.040]; officials[0.040]; agencies[0.040]; agencies[0.040]; likely[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.078:0.078[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npersonnel[0.044]; major[0.044]; major[0.044]; reached[0.043]; World[0.043]; early[0.042]; climbing[0.042]; report[0.041]; cities[0.041]; cities[0.041]; cities[0.041]; scores[0.041]; radio[0.041]; Charles[0.041]; use[0.041]; based[0.041]; included[0.041]; jurisdictions[0.041]; categories[0.041]; categories[0.041]; categories[0.041]; category[0.041]; surrounding[0.041]; Thirteen[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Homeland Security ==> ENTITY: \u001b[32mUnited States Department of Homeland Security\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.171:0.171[0]; log p(e|m)= -0.989:-0.989[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHomeland[0.049]; agencies[0.045]; security[0.045]; security[0.045]; Chertoff[0.043]; Chertoff[0.043]; Chertoff[0.043]; Secretary[0.043]; acknowledged[0.042]; communications[0.042]; communications[0.042]; communications[0.042]; job[0.041]; address[0.040]; government[0.040]; Sept[0.040]; secretary[0.040]; Michael[0.040]; attacks[0.040]; news[0.039]; systems[0.039]; released[0.039]; briefing[0.039]; chief[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Congress ==> ENTITY: \u001b[32mUnited States Congress\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.084:0.084[0]; log p(e|m)= -0.335:-0.335[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngovernance[0.043]; American[0.043]; agencies[0.043]; agencies[0.043]; majority[0.043]; report[0.042]; report[0.042]; Sen[0.042]; includes[0.042]; territories[0.042]; formalized[0.042]; leadership[0.042]; term[0.042]; policies[0.042]; overview[0.041]; large[0.041]; regions[0.041]; place[0.041]; said[0.041]; said[0.041]; went[0.041]; promised[0.040]; week[0.040]; La[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Laramie County\u001b[39m ==> ENTITY: \u001b[32mLaramie County, Wyoming\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.093:0.093[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMandan[0.045]; state[0.044]; area[0.044]; Sioux[0.043]; medical[0.043]; federal[0.043]; emergency[0.042]; emergency[0.042]; suburbs[0.042]; American[0.042]; areas[0.042]; overview[0.041]; Washington[0.041]; Wyo[0.041]; leadership[0.041]; territories[0.041]; seamless[0.041]; highest[0.041]; small[0.040]; communities[0.040]; lowest[0.040]; personnel[0.040]; longstanding[0.040]; cities[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = D.C. ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.158:-0.158[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWashington[0.048]; state[0.047]; federal[0.044]; territories[0.044]; scores[0.043]; American[0.043]; formalized[0.042]; overall[0.042]; use[0.042]; regions[0.041]; highest[0.041]; agencies[0.041]; Cleveland[0.041]; place[0.041]; County[0.041]; lowest[0.040]; officials[0.040]; equipment[0.040]; longstanding[0.040]; includes[0.040]; Chicago[0.040]; planning[0.040]; Columbus[0.039]; lagged[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.070:0.070[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.046]; cost[0.045]; released[0.045]; billion[0.043]; highest[0.043]; rivalries[0.043]; news[0.042]; acknowledged[0.042]; Michael[0.042]; report[0.042]; report[0.042]; Press[0.042]; longstanding[0.041]; obtained[0.041]; Longstanding[0.041]; far[0.041]; results[0.040]; Wednesday[0.040]; advanced[0.040]; Portions[0.040]; Tuesday[0.040]; different[0.040]; continues[0.040]; determined[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = N.D. ==> ENTITY: \u001b[32mNorth Dakota\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.090:0.090[0]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstate[0.045]; Mandan[0.045]; studied[0.044]; use[0.043]; Laramie[0.042]; Sioux[0.042]; Washington[0.042]; federal[0.042]; County[0.042]; Ohio[0.042]; emergency[0.041]; emergency[0.041]; emergency[0.041]; highest[0.041]; cities[0.041]; territories[0.041]; specifically[0.041]; overview[0.041]; Wyo[0.041]; leadership[0.040]; finds[0.040]; challenge[0.040]; challenge[0.040]; formalized[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ohio ==> ENTITY: \u001b[32mOhio\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.191:-0.191[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCounty[0.046]; area[0.045]; Cleveland[0.044]; Washington[0.043]; state[0.043]; Columbus[0.042]; fuel[0.042]; officials[0.042]; cities[0.042]; communities[0.041]; communities[0.041]; federal[0.041]; areas[0.041]; report[0.041]; report[0.041]; report[0.041]; report[0.041]; planning[0.041]; American[0.041]; said[0.040]; studied[0.040]; overall[0.040]; emergency[0.040]; emergency[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.082:0.082[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsays[0.044]; York[0.043]; major[0.043]; major[0.043]; towers[0.043]; cities[0.042]; cities[0.042]; cities[0.042]; World[0.042]; buildings[0.042]; surrounding[0.042]; revealed[0.041]; local[0.041]; local[0.041]; Charles[0.041]; communities[0.041]; place[0.041]; Chicago[0.041]; said[0.041]; crumbling[0.041]; early[0.041]; Center[0.041]; scores[0.041]; based[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.059:0.059[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstage[0.045]; surrounding[0.043]; use[0.043]; based[0.043]; included[0.043]; limited[0.042]; limited[0.042]; major[0.042]; said[0.042]; said[0.042]; said[0.042]; early[0.042]; report[0.041]; cities[0.041]; cities[0.041]; assumption[0.041]; methodology[0.041]; governments[0.041]; Office[0.041]; place[0.040]; statement[0.040]; crisis[0.040]; strongly[0.040]; procedures[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Baton Rouge ==> ENTITY: \u001b[32mBaton Rouge, Louisiana\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.076:0.076[0]; log p(e|m)= -0.065:-0.065[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \narea[0.046]; cities[0.045]; state[0.044]; communities[0.042]; Paul[0.042]; place[0.042]; Washington[0.042]; County[0.042]; federal[0.042]; culture[0.041]; crisis[0.041]; includes[0.041]; lowest[0.041]; agencies[0.041]; agencies[0.041]; promised[0.041]; regions[0.041]; areas[0.041]; personnel[0.041]; suburbs[0.041]; officials[0.041]; highest[0.041]; add[0.041]; Cleveland[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Minneapolis-St. Paul\u001b[39m ==> ENTITY: \u001b[32mMinneapolis–Saint Paul\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.110:0.110[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncities[0.044]; area[0.044]; Washington[0.044]; Columbus[0.044]; equipment[0.043]; ratings[0.042]; suburbs[0.042]; Cleveland[0.042]; overall[0.042]; Chicago[0.042]; federal[0.042]; use[0.042]; officials[0.041]; longstanding[0.041]; emergency[0.041]; emergency[0.041]; emergency[0.041]; personnel[0.041]; communities[0.040]; communities[0.040]; includes[0.040]; needed[0.040]; American[0.040]; state[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chertoff\u001b[39m ==> ENTITY: \u001b[32mMichael Chertoff\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChertoff[0.046]; Chertoff[0.046]; medical[0.043]; Wednesday[0.042]; Tuesday[0.042]; emergency[0.042]; emergency[0.042]; emergency[0.042]; said[0.042]; said[0.042]; terrorist[0.042]; longstanding[0.041]; communications[0.041]; Sept[0.041]; Sept[0.041]; agencies[0.041]; agencies[0.041]; seamless[0.041]; personnel[0.041]; report[0.040]; report[0.040]; report[0.040]; rivalries[0.040]; Congress[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Democrats ==> ENTITY: \u001b[32mDemocratic Party (United States)\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.083:0.083[0]; log p(e|m)= -0.180:-0.180[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.048]; majority[0.048]; leadership[0.045]; policies[0.045]; territories[0.044]; Cleveland[0.044]; Ohio[0.044]; Charles[0.043]; lagged[0.043]; pay[0.043]; County[0.043]; surveyed[0.043]; Congress[0.043]; Columbus[0.043]; Samoa[0.043]; term[0.042]; large[0.042]; said[0.042]; said[0.042]; went[0.042]; major[0.042]; cost[0.042]; problems[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chertoff\u001b[39m ==> ENTITY: \u001b[32mMichael Chertoff\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChertoff[0.045]; Security[0.044]; Secretary[0.044]; Homeland[0.044]; Homeland[0.044]; security[0.044]; security[0.044]; Michael[0.043]; acknowledged[0.043]; secretary[0.042]; chief[0.041]; Wednesday[0.041]; emergency[0.041]; said[0.041]; job[0.041]; acknowledges[0.040]; communications[0.040]; communications[0.040]; agencies[0.040]; briefing[0.040]; address[0.039]; WASHINGTON[0.039]; government[0.039]; news[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Michael Chertoff\u001b[39m ==> ENTITY: \u001b[32mMichael Chertoff\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChertoff[0.045]; Chertoff[0.045]; Security[0.044]; Secretary[0.044]; Homeland[0.044]; Homeland[0.044]; security[0.043]; security[0.043]; acknowledged[0.042]; secretary[0.041]; chief[0.041]; Wednesday[0.041]; emergency[0.041]; emergency[0.041]; said[0.041]; job[0.040]; acknowledges[0.040]; communications[0.040]; communications[0.040]; communications[0.040]; Sept[0.040]; agencies[0.040]; briefing[0.039]; address[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wyo.\u001b[39m ==> ENTITY: \u001b[32mWyoming\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.082:0.082[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMandan[0.045]; state[0.044]; Laramie[0.043]; Sioux[0.043]; federal[0.043]; medical[0.043]; area[0.042]; emergency[0.042]; emergency[0.042]; Washington[0.042]; American[0.042]; overview[0.042]; leadership[0.041]; cities[0.041]; County[0.041]; territories[0.041]; areas[0.041]; seamless[0.041]; communities[0.041]; Ohio[0.040]; finds[0.040]; week[0.040]; includes[0.040]; small[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = American Samoa ==> ENTITY: \u001b[32mAmerican Samoa\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.093:0.093[0]; log p(e|m)= -0.370:-0.370[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nweek[0.044]; territories[0.043]; state[0.043]; percent[0.042]; percent[0.042]; equipment[0.042]; agencies[0.042]; agencies[0.042]; cities[0.042]; Washington[0.042]; small[0.042]; use[0.042]; add[0.042]; culture[0.042]; large[0.041]; surveyed[0.041]; highest[0.041]; communities[0.041]; formalized[0.041]; regions[0.041]; battle[0.041]; lowest[0.041]; place[0.041]; planning[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sioux Falls\u001b[39m ==> ENTITY: \u001b[32mSioux Falls, South Dakota\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.084:0.084[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.045]; Washington[0.044]; Chicago[0.044]; Paul[0.044]; overall[0.043]; helping[0.043]; Columbus[0.043]; scores[0.042]; state[0.042]; area[0.042]; County[0.041]; different[0.041]; week[0.041]; cities[0.041]; American[0.041]; Minneapolis[0.041]; add[0.041]; place[0.040]; lowest[0.040]; needed[0.040]; personnel[0.040]; equipment[0.040]; majority[0.040]; suburbs[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.066:0.066[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCounty[0.049]; state[0.049]; studied[0.049]; area[0.048]; American[0.047]; suburbs[0.046]; communities[0.046]; communities[0.046]; federal[0.046]; cities[0.046]; Cleveland[0.045]; firefighters[0.045]; medical[0.044]; Ohio[0.044]; Sioux[0.044]; territories[0.044]; areas[0.044]; study[0.044]; Chicago[0.044]; percent[0.044]; percent[0.044]; lowest[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mandan ==> ENTITY: \u001b[32mMandan, North Dakota\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.043:0.043[0]; log p(e|m)= -0.997:-0.997[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCounty[0.044]; areas[0.043]; cities[0.043]; Office[0.043]; emergency[0.043]; emergency[0.043]; Emergency[0.042]; feel[0.042]; study[0.042]; study[0.042]; fall[0.042]; coordinator[0.042]; said[0.041]; said[0.041]; said[0.041]; Management[0.041]; early[0.041]; intermediate[0.041]; report[0.041]; low[0.041]; going[0.041]; funds[0.041]; personnel[0.040]; inconsistent[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMorton County\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMorton County, Kansas\u001b[39m <---> \u001b[32mMorton County, North Dakota\u001b[39m\t\nSCORES: global= 0.279:0.278[\u001b[31m0.001\u001b[39m]; local()= 0.057:0.050[\u001b[31m0.007\u001b[39m]; log p(e|m)= -0.135:0.000[\u001b[32m0.135\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \njurisdictions[0.049]; Mandan[0.048]; areas[0.044]; emergency[0.044]; emergency[0.044]; Emergency[0.044]; Lapp[0.044]; cities[0.044]; limited[0.043]; limited[0.043]; coordinator[0.043]; category[0.043]; Management[0.043]; intermediate[0.043]; funds[0.042]; personnel[0.042]; fall[0.042]; lowest[0.042]; Office[0.042]; grades[0.042]; grades[0.042]; local[0.042]; early[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = La.\u001b[39m ==> ENTITY: \u001b[32mLouisiana\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.054:0.054[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMandan[0.045]; cities[0.043]; federal[0.043]; territories[0.043]; Laramie[0.042]; Sioux[0.042]; state[0.042]; said[0.042]; said[0.042]; Paul[0.042]; regions[0.042]; emergency[0.041]; emergency[0.041]; emergency[0.041]; battle[0.041]; area[0.041]; overview[0.041]; says[0.041]; Wyo[0.041]; talk[0.041]; leadership[0.041]; highest[0.040]; includes[0.040]; American[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Democrats ==> ENTITY: \u001b[32mDemocratic Party (United States)\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.090:0.090[0]; log p(e|m)= -0.180:-0.180[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnation[0.044]; country[0.044]; government[0.043]; effort[0.043]; gave[0.043]; Michael[0.042]; acknowledged[0.042]; control[0.042]; address[0.042]; survey[0.042]; Congress[0.042]; job[0.041]; secretary[0.041]; pledged[0.041]; security[0.041]; security[0.041]; Tuesday[0.041]; said[0.041]; said[0.041]; provided[0.041]; highest[0.041]; cost[0.041]; Secretary[0.040]; obtained[0.040]; \t\n\r [===========>..................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s846ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 178/656 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 16442342 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sago Mine accident\u001b[39m ==> ENTITY: \u001b[32mSago Mine disaster\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsafety[0.045]; miners[0.044]; miners[0.044]; safe[0.043]; Tuesday[0.043]; Tuesday[0.043]; Sago[0.043]; Buckhannon[0.043]; Virginia[0.042]; inside[0.042]; near[0.042]; West[0.042]; McCloy[0.042]; nation[0.041]; disaster[0.041]; mines[0.041]; families[0.040]; trapped[0.040]; say[0.040]; need[0.040]; breath[0.040]; alive[0.040]; holding[0.039]; Randal[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Virginia ==> ENTITY: \u001b[32mWest Virginia\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.300:-0.300[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVirginia[0.048]; inside[0.046]; ground[0.046]; nation[0.045]; West[0.045]; families[0.045]; holding[0.044]; won[0.044]; Tuesday[0.044]; Tuesday[0.044]; began[0.043]; near[0.043]; Buckhannon[0.043]; lone[0.043]; men[0.042]; men[0.042]; Va[0.042]; Va[0.042]; joined[0.042]; day[0.042]; deep[0.042]; remember[0.041]; turned[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sago Mine\u001b[39m ==> ENTITY: \u001b[32mSago Mine disaster\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsafety[0.045]; safety[0.045]; safety[0.045]; safety[0.045]; Mines[0.044]; Mines[0.044]; miners[0.044]; miners[0.044]; miners[0.044]; miners[0.044]; safe[0.044]; Health[0.043]; Sago[0.043]; Sago[0.043]; emergency[0.043]; emergency[0.043]; Virginia[0.043]; Virginia[0.043]; Administration[0.042]; Safety[0.042]; federal[0.042]; federal[0.042]; federal[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Randal McCloy\u001b[39m ==> ENTITY: \u001b[32mSago Mine disaster\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsafety[0.044]; safety[0.044]; miners[0.043]; miners[0.043]; miners[0.043]; safe[0.042]; Tuesday[0.042]; Tuesday[0.042]; Sago[0.042]; Sago[0.042]; Sago[0.042]; Buckhannon[0.042]; Virginia[0.041]; Virginia[0.041]; inside[0.041]; near[0.041]; federal[0.041]; West[0.041]; West[0.041]; nation[0.040]; disaster[0.040]; says[0.040]; mines[0.040]; mines[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Buckhannon\u001b[39m ==> ENTITY: \u001b[32mBuckhannon, West Virginia\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVirginia[0.045]; Virginia[0.045]; West[0.044]; West[0.044]; nation[0.044]; families[0.043]; near[0.043]; federal[0.042]; disaster[0.042]; ago[0.041]; ago[0.041]; passed[0.041]; Tuesday[0.041]; Tuesday[0.041]; began[0.041]; Va[0.041]; Va[0.041]; say[0.041]; Hamner[0.041]; trapped[0.040]; miners[0.040]; miners[0.040]; mines[0.040]; mines[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sago ==> ENTITY: \u001b[32mSago Mine disaster\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.162:0.162[0]; log p(e|m)= -1.423:-1.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSago[0.048]; West[0.048]; Virginia[0.044]; safety[0.043]; safety[0.043]; safety[0.043]; safety[0.043]; nation[0.042]; miners[0.041]; miners[0.041]; miners[0.041]; emergency[0.041]; emergency[0.041]; underground[0.041]; safe[0.041]; federal[0.041]; federal[0.041]; Administration[0.040]; tragedy[0.040]; Safety[0.040]; union[0.039]; McAteer[0.039]; McAteer[0.039]; coal[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sago ==> ENTITY: \u001b[32mSago Mine disaster\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.154:0.154[0]; log p(e|m)= -1.423:-1.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSago[0.046]; Sago[0.046]; West[0.045]; West[0.045]; West[0.045]; near[0.045]; Virginia[0.042]; Virginia[0.042]; Virginia[0.042]; Hamner[0.041]; families[0.041]; Buckhannon[0.041]; safety[0.041]; safety[0.041]; disaster[0.040]; nation[0.040]; miners[0.040]; miners[0.040]; miners[0.040]; miners[0.040]; emergency[0.040]; safe[0.039]; ground[0.039]; Va[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.046:0.046[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVirginia[0.045]; nation[0.045]; West[0.044]; began[0.044]; ago[0.044]; ago[0.044]; laws[0.043]; Tuesday[0.042]; Tuesday[0.042]; joined[0.042]; need[0.042]; think[0.042]; say[0.042]; inside[0.041]; near[0.041]; mines[0.040]; deep[0.040]; men[0.040]; remember[0.040]; breath[0.040]; disaster[0.040]; turned[0.040]; safe[0.039]; holding[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = United Mine Workers of America\u001b[39m ==> ENTITY: \u001b[32mUnited Mine Workers\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.121:0.121[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nunion[0.048]; Mines[0.045]; miners[0.043]; emergency[0.043]; emergency[0.043]; coal[0.043]; coal[0.043]; coal[0.043]; Dennis[0.042]; nation[0.042]; representing[0.041]; inspectors[0.041]; required[0.041]; required[0.041]; government[0.041]; safety[0.041]; safety[0.041]; safety[0.041]; tougher[0.040]; rooms[0.040]; procedures[0.040]; record[0.040]; times[0.040]; violations[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Virginia ==> ENTITY: \u001b[32mWest Virginia\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.096:0.096[0]; log p(e|m)= -0.300:-0.300[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteams[0.047]; Virginia[0.046]; Virginia[0.046]; Tuesday[0.044]; West[0.043]; West[0.043]; families[0.042]; won[0.042]; near[0.041]; Buckhannon[0.041]; ran[0.041]; emergency[0.041]; Health[0.041]; lone[0.040]; federal[0.040]; began[0.040]; men[0.040]; men[0.040]; inside[0.040]; tracking[0.040]; Administration[0.040]; required[0.040]; required[0.040]; safe[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Buckhannon\u001b[39m ==> ENTITY: \u001b[32mBuckhannon, West Virginia\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVirginia[0.048]; Buckhannon[0.047]; West[0.046]; nation[0.046]; families[0.045]; near[0.045]; disaster[0.044]; ago[0.043]; ago[0.043]; Tuesday[0.043]; Tuesday[0.043]; began[0.043]; Va[0.043]; Va[0.043]; say[0.043]; Hamner[0.043]; trapped[0.042]; miners[0.042]; miners[0.042]; mines[0.042]; turned[0.042]; deep[0.042]; breath[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mine Safety and Health Administration\u001b[39m ==> ENTITY: \u001b[32mMine Safety and Health Administration\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.217:0.217[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nviolations[0.050]; procedures[0.050]; safety[0.049]; safety[0.049]; safety[0.049]; safety[0.049]; inspectors[0.049]; Mines[0.048]; Mines[0.048]; federal[0.047]; federal[0.047]; federal[0.047]; mines[0.047]; required[0.047]; required[0.047]; required[0.047]; fines[0.046]; laws[0.046]; notify[0.046]; emergency[0.046]; emergency[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sago Mine\u001b[39m ==> ENTITY: \u001b[32mSago Mine disaster\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsafety[0.045]; safety[0.045]; miners[0.043]; miners[0.043]; safe[0.043]; Tuesday[0.042]; Tuesday[0.042]; Sago[0.042]; Sago[0.042]; Buckhannon[0.042]; Virginia[0.042]; Virginia[0.042]; inside[0.042]; near[0.041]; federal[0.041]; West[0.041]; West[0.041]; McCloy[0.041]; nation[0.040]; disaster[0.040]; says[0.040]; mines[0.040]; families[0.040]; trapped[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Virginia ==> ENTITY: \u001b[32mWest Virginia\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.063:0.063[0]; log p(e|m)= -0.300:-0.300[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteams[0.047]; laws[0.046]; Virginia[0.046]; law[0.045]; West[0.043]; won[0.042]; government[0.041]; government[0.041]; government[0.041]; protect[0.041]; Administration[0.041]; federal[0.041]; federal[0.041]; federal[0.041]; ran[0.041]; Health[0.041]; provide[0.041]; safe[0.040]; safety[0.040]; safety[0.040]; safety[0.040]; safety[0.040]; communications[0.040]; led[0.040]; \t\n\r [============>.................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s838ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 194/656 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 16451112 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nardelli\u001b[39m ==> ENTITY: \u001b[32mRobert Nardelli\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.137:0.137[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; college[0.044]; job[0.043]; CEO[0.042]; declined[0.042]; December[0.041]; deal[0.041]; company[0.041]; company[0.041]; company[0.041]; won[0.041]; board[0.041]; senior[0.041]; executive[0.040]; appointment[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; returned[0.040]; football[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Home Depot\u001b[39m ==> ENTITY: \u001b[32mThe Home Depot\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncollege[0.049]; company[0.048]; company[0.048]; company[0.048]; declined[0.046]; Home[0.046]; Home[0.046]; Home[0.046]; senior[0.045]; deal[0.045]; CEO[0.045]; big[0.045]; job[0.045]; football[0.045]; Wednesday[0.045]; York[0.044]; replace[0.044]; appointment[0.044]; said[0.044]; said[0.044]; said[0.044]; said[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Home Depot\u001b[39m ==> ENTITY: \u001b[32mThe Home Depot\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstores[0.045]; stores[0.045]; American[0.044]; executives[0.043]; executives[0.043]; companies[0.042]; Home[0.042]; Home[0.042]; Home[0.042]; Home[0.042]; Home[0.042]; senior[0.042]; senior[0.042]; CEO[0.041]; employees[0.041]; employees[0.041]; Tuesday[0.041]; roughly[0.041]; Canada[0.041]; customer[0.040]; million[0.040]; million[0.040]; million[0.040]; Mexico[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Los Angeles ==> ENTITY: \u001b[32mLos Angeles\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.101:-0.101[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWednesday[0.045]; Wednesday[0.045]; recently[0.043]; thought[0.043]; thought[0.043]; interview[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; news[0.041]; week[0.041]; near[0.041]; told[0.041]; board[0.041]; earlier[0.041]; leaving[0.041]; leaving[0.041]; office[0.041]; took[0.040]; AP[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Home Depot\u001b[39m ==> ENTITY: \u001b[32mThe Home Depot\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.115:0.115[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHome[0.049]; Home[0.049]; Home[0.049]; Home[0.049]; Home[0.049]; recently[0.049]; Wednesday[0.048]; key[0.048]; leadership[0.047]; adjusted[0.047]; interview[0.047]; Bob[0.047]; said[0.047]; said[0.047]; said[0.047]; said[0.047]; said[0.047]; said[0.047]; said[0.047]; time[0.047]; time[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nardelli\u001b[39m ==> ENTITY: \u001b[32mRobert Nardelli\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Wednesday[0.043]; pay[0.042]; pay[0.042]; CEO[0.042]; resigned[0.042]; million[0.042]; hefty[0.041]; company[0.041]; company[0.041]; chief[0.041]; chief[0.041]; board[0.041]; performance[0.041]; receive[0.041]; chairman[0.040]; chairman[0.040]; executive[0.040]; executive[0.040]; package[0.040]; Services[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.072:0.072[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nvice[0.053]; won[0.053]; official[0.053]; Canada[0.052]; Mexico[0.052]; June[0.051]; roughly[0.050]; chairman[0.050]; seven[0.050]; service[0.049]; hiring[0.049]; supports[0.049]; Tuesday[0.049]; includes[0.049]; replaced[0.049]; States[0.048]; replace[0.048]; million[0.048]; million[0.048]; million[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Home Depot\u001b[39m ==> ENTITY: \u001b[32mThe Home Depot\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstores[0.047]; company[0.046]; company[0.046]; Home[0.044]; Home[0.044]; Home[0.044]; Home[0.044]; Home[0.044]; Home[0.044]; CEO[0.044]; job[0.043]; operates[0.043]; helped[0.043]; leadership[0.043]; million[0.043]; Bob[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; approximately[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nardelli\u001b[39m ==> ENTITY: \u001b[32mRobert Nardelli\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCollege[0.054]; Nardelli[0.054]; job[0.051]; past[0.051]; pay[0.050]; CEO[0.050]; University[0.050]; million[0.050]; James[0.050]; Business[0.050]; second[0.050]; company[0.049]; company[0.049]; earned[0.049]; board[0.049]; strategy[0.049]; disclosed[0.049]; State[0.049]; leadership[0.048]; leadership[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nardelli\u001b[39m ==> ENTITY: \u001b[32mRobert Nardelli\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNardelli[0.051]; Nardelli[0.051]; Nardelli[0.051]; college[0.050]; hiring[0.048]; employees[0.048]; employees[0.048]; June[0.048]; CEO[0.048]; million[0.047]; deal[0.047]; company[0.047]; company[0.047]; Canada[0.047]; won[0.047]; board[0.046]; vice[0.046]; chairman[0.046]; service[0.046]; package[0.046]; package[0.046]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mFrank\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFrank Blake\u001b[39m <---> \u001b[32mBarney Frank\u001b[39m\t\nSCORES: global= 0.275:0.265[\u001b[31m0.010\u001b[39m]; local()= 0.124:0.132[\u001b[32m0.008\u001b[39m]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nServices[0.044]; legislation[0.043]; Frank[0.043]; CEO[0.042]; Rep[0.042]; Committee[0.042]; companies[0.042]; House[0.042]; deferred[0.042]; Financial[0.042]; executives[0.041]; executives[0.041]; Nardelli[0.041]; Nardelli[0.041]; company[0.041]; company[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; helm[0.041]; roughly[0.041]; roughly[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = General Electric Co.\u001b[39m ==> ENTITY: \u001b[32mGeneral Electric\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.090:0.090[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncompany[0.053]; company[0.053]; company[0.053]; college[0.052]; job[0.048]; senior[0.048]; stock[0.048]; stock[0.048]; stock[0.048]; football[0.047]; CEO[0.047]; appointment[0.047]; York[0.046]; executive[0.046]; deal[0.046]; care[0.045]; shares[0.045]; available[0.045]; Marcus[0.045]; Marcus[0.045]; Marcus[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nardelli\u001b[39m ==> ENTITY: \u001b[32mRobert Nardelli\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNardelli[0.045]; Nardelli[0.045]; college[0.044]; Wednesday[0.043]; job[0.043]; CEO[0.042]; declined[0.042]; December[0.042]; deal[0.042]; company[0.041]; company[0.041]; company[0.041]; won[0.041]; senior[0.041]; executive[0.041]; week[0.041]; appointment[0.041]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; returned[0.040]; football[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nardelli\u001b[39m ==> ENTITY: \u001b[32mRobert Nardelli\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Wednesday[0.043]; recently[0.042]; past[0.042]; earlier[0.042]; despite[0.041]; board[0.041]; key[0.040]; leadership[0.040]; AP[0.040]; Hill[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; percent[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Home Depot\u001b[39m ==> ENTITY: \u001b[32mThe Home Depot\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHome[0.047]; Home[0.047]; Home[0.047]; Home[0.047]; Home[0.047]; recently[0.047]; helped[0.046]; Wednesday[0.045]; key[0.045]; leadership[0.045]; interview[0.045]; Bob[0.045]; said[0.045]; said[0.045]; said[0.045]; said[0.045]; said[0.045]; said[0.045]; said[0.045]; said[0.045]; time[0.045]; time[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Home Depot\u001b[39m ==> ENTITY: \u001b[32mThe Home Depot\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstores[0.047]; stores[0.047]; company[0.046]; hiring[0.045]; Home[0.044]; Home[0.044]; Home[0.044]; Home[0.044]; Home[0.044]; deal[0.043]; big[0.043]; employees[0.043]; employees[0.043]; Tuesday[0.043]; roughly[0.043]; Canada[0.043]; customer[0.042]; million[0.042]; million[0.042]; million[0.042]; replace[0.042]; Mexico[0.042]; awards[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Blake\u001b[39m ==> ENTITY: \u001b[32mFrank Blake\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCollege[0.045]; Business[0.045]; CEO[0.043]; increased[0.043]; Nardelli[0.042]; Nardelli[0.042]; job[0.042]; University[0.042]; revenue[0.042]; leadership[0.042]; Home[0.041]; Home[0.041]; Home[0.041]; Home[0.041]; million[0.041]; State[0.041]; director[0.041]; public[0.041]; Bob[0.041]; Depot[0.041]; Depot[0.041]; Depot[0.041]; Depot[0.041]; earned[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Georgia State University\u001b[39m ==> ENTITY: \u001b[32mGeorgia State University\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.093:0.093[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCollege[0.050]; Center[0.049]; Business[0.047]; public[0.046]; run[0.046]; increased[0.046]; James[0.046]; director[0.046]; considered[0.046]; Leadership[0.045]; million[0.045]; Robinson[0.045]; sign[0.045]; job[0.045]; Jones[0.045]; second[0.045]; experts[0.045]; consider[0.044]; leadership[0.044]; board[0.044]; Edward[0.044]; earned[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nardelli\u001b[39m ==> ENTITY: \u001b[32mRobert Nardelli\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nexecutives[0.046]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Wednesday[0.043]; American[0.042]; pay[0.042]; pay[0.042]; resigned[0.041]; million[0.041]; companies[0.041]; hefty[0.041]; company[0.041]; company[0.041]; chief[0.041]; board[0.041]; performance[0.041]; receive[0.040]; senior[0.040]; chairman[0.040]; chairman[0.040]; executive[0.040]; package[0.040]; Services[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Blake\u001b[39m ==> ENTITY: \u001b[32mFrank Blake\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncollege[0.055]; vice[0.055]; Blake[0.055]; CEO[0.054]; Nardelli[0.053]; Nardelli[0.053]; Nardelli[0.053]; Nardelli[0.053]; hiring[0.052]; chairman[0.052]; Home[0.052]; Home[0.052]; Home[0.052]; Home[0.052]; Home[0.052]; Home[0.052]; effective[0.052]; million[0.051]; June[0.051]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Blake\u001b[39m ==> ENTITY: \u001b[32mFrank Blake\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncollege[0.053]; Blake[0.052]; CEO[0.052]; executive[0.051]; Nardelli[0.050]; Nardelli[0.050]; Nardelli[0.050]; Nardelli[0.050]; Nardelli[0.050]; job[0.050]; declined[0.050]; Home[0.049]; Home[0.049]; Home[0.049]; Home[0.049]; Home[0.049]; appointment[0.049]; rose[0.049]; secretary[0.049]; senior[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Home Depot\u001b[39m ==> ENTITY: \u001b[32mThe Home Depot\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.117:0.117[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncollege[0.049]; company[0.048]; company[0.048]; company[0.048]; declined[0.046]; Home[0.046]; Home[0.046]; Home[0.046]; senior[0.045]; deal[0.045]; CEO[0.045]; big[0.045]; job[0.045]; football[0.045]; Wednesday[0.044]; York[0.044]; adjusted[0.044]; news[0.044]; appointment[0.044]; said[0.044]; said[0.044]; said[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Home Depot\u001b[39m ==> ENTITY: \u001b[32mThe Home Depot\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstores[0.045]; stores[0.045]; hiring[0.043]; executives[0.043]; executives[0.043]; Home[0.042]; Home[0.042]; Home[0.042]; Home[0.042]; Home[0.042]; senior[0.042]; senior[0.042]; CEO[0.042]; employees[0.041]; employees[0.041]; Tuesday[0.041]; roughly[0.041]; Canada[0.041]; customer[0.041]; million[0.041]; million[0.041]; million[0.041]; replace[0.040]; Mexico[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marcus\u001b[39m ==> ENTITY: \u001b[32mBernard Marcus\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMarcus[0.045]; helped[0.045]; board[0.043]; board[0.043]; Bob[0.043]; leadership[0.042]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; director[0.041]; director[0.041]; Nardelli[0.041]; Nardelli[0.041]; Nardelli[0.041]; Nardelli[0.041]; Nardelli[0.041]; Nardelli[0.041]; Nardelli[0.041]; office[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nardelli\u001b[39m ==> ENTITY: \u001b[32mRobert Nardelli\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Wednesday[0.043]; past[0.042]; earlier[0.042]; company[0.041]; despite[0.041]; board[0.041]; board[0.041]; key[0.041]; leadership[0.040]; AP[0.040]; Hill[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nardelli\u001b[39m ==> ENTITY: \u001b[32mRobert Nardelli\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNardelli[0.052]; Nardelli[0.052]; Tuesday[0.049]; hiring[0.049]; employees[0.048]; employees[0.048]; June[0.048]; million[0.048]; million[0.048]; million[0.048]; deal[0.047]; company[0.047]; Canada[0.047]; won[0.047]; board[0.047]; vice[0.046]; chairman[0.046]; service[0.046]; package[0.046]; package[0.046]; payment[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bob\u001b[39m ==> ENTITY: \u001b[32mRobert Nardelli\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCollege[0.045]; Nardelli[0.045]; Nardelli[0.045]; job[0.043]; pay[0.042]; CEO[0.042]; University[0.042]; million[0.042]; James[0.042]; Business[0.042]; second[0.041]; company[0.041]; company[0.041]; earned[0.041]; board[0.041]; strategy[0.041]; strategy[0.041]; disclosed[0.041]; State[0.041]; leadership[0.041]; package[0.040]; Home[0.040]; Home[0.040]; analysts[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marcus\u001b[39m ==> ENTITY: \u001b[32mBernard Marcus\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfounder[0.044]; college[0.044]; Marcus[0.044]; Marcus[0.044]; Marcus[0.044]; chairman[0.043]; CEO[0.043]; board[0.042]; hiring[0.042]; job[0.042]; company[0.041]; company[0.041]; senior[0.041]; Bernie[0.041]; Blake[0.041]; Blake[0.041]; football[0.040]; replace[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; Frank[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Home Depot\u001b[39m ==> ENTITY: \u001b[32mThe Home Depot\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.112:0.112[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.044]; executives[0.044]; executives[0.044]; companies[0.043]; Home[0.042]; Home[0.042]; senior[0.042]; senior[0.042]; CEO[0.042]; Tuesday[0.041]; roughly[0.041]; roughly[0.041]; million[0.041]; million[0.041]; million[0.041]; million[0.041]; message[0.041]; awards[0.041]; said[0.041]; said[0.041]; said[0.041]; Nardelli[0.041]; Nardelli[0.041]; total[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Blake\u001b[39m ==> ENTITY: \u001b[32mFrank Blake\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncollege[0.044]; Blake[0.044]; CEO[0.043]; executive[0.042]; Nardelli[0.042]; Nardelli[0.042]; Nardelli[0.042]; Nardelli[0.042]; Nardelli[0.042]; job[0.042]; declined[0.042]; Home[0.041]; Home[0.041]; Home[0.041]; Home[0.041]; appointment[0.041]; recently[0.041]; rose[0.041]; secretary[0.041]; senior[0.041]; Depot[0.041]; Depot[0.041]; Depot[0.041]; Depot[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nardelli\u001b[39m ==> ENTITY: \u001b[32mRobert Nardelli\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; college[0.044]; Wednesday[0.043]; job[0.043]; recently[0.043]; CEO[0.042]; declined[0.042]; December[0.041]; deal[0.041]; company[0.041]; company[0.041]; senior[0.041]; executive[0.040]; week[0.040]; appointment[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; returned[0.040]; football[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = The Home Depot Inc.\u001b[39m ==> ENTITY: \u001b[32mThe Home Depot\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nretailer[0.046]; company[0.044]; company[0.044]; store[0.044]; chain[0.044]; Atlanta[0.043]; Home[0.042]; Home[0.042]; CEO[0.042]; ATLANTA[0.041]; Wednesday[0.041]; million[0.041]; Bob[0.041]; Nardelli[0.040]; Nardelli[0.040]; Nardelli[0.040]; Nardelli[0.040]; Nardelli[0.040]; hefty[0.040]; pay[0.040]; pay[0.040]; executive[0.040]; executive[0.040]; largest[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nardelli\u001b[39m ==> ENTITY: \u001b[32mRobert Nardelli\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Wednesday[0.043]; past[0.042]; earlier[0.042]; company[0.041]; board[0.041]; leadership[0.041]; AP[0.041]; Hill[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; percent[0.040]; discussion[0.040]; increased[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marcus\u001b[39m ==> ENTITY: \u001b[32mBernard Marcus\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.137:0.137[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMarcus[0.045]; helped[0.045]; board[0.043]; Bob[0.043]; company[0.043]; company[0.043]; leadership[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; approximately[0.041]; Nardelli[0.041]; Nardelli[0.041]; Nardelli[0.041]; Nardelli[0.041]; Nardelli[0.041]; Nardelli[0.041]; office[0.041]; leader[0.041]; idea[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nardelli\u001b[39m ==> ENTITY: \u001b[32mRobert Nardelli\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Wednesday[0.043]; pay[0.042]; pay[0.042]; CEO[0.042]; resigned[0.042]; hefty[0.042]; company[0.041]; chief[0.041]; chief[0.041]; board[0.041]; performance[0.041]; chairman[0.041]; executive[0.041]; executive[0.041]; Home[0.040]; Home[0.040]; Home[0.040]; analysts[0.040]; stock[0.040]; ATLANTA[0.040]; abruptly[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Frank Blake\u001b[39m ==> ENTITY: \u001b[32mFrank Blake\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nvice[0.052]; Blake[0.052]; deferred[0.051]; Nardelli[0.051]; Nardelli[0.051]; Nardelli[0.051]; hiring[0.050]; chairman[0.050]; Home[0.050]; Home[0.050]; Home[0.050]; Home[0.050]; Home[0.050]; Home[0.050]; effective[0.049]; effective[0.049]; million[0.049]; million[0.049]; million[0.049]; June[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Barney Frank\u001b[39m ==> ENTITY: \u001b[32mBarney Frank\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlegislation[0.044]; Frank[0.043]; Rep[0.043]; Committee[0.043]; companies[0.043]; House[0.042]; Financial[0.042]; company[0.042]; company[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; Services[0.041]; roughly[0.041]; lawmakers[0.041]; American[0.041]; executives[0.041]; executives[0.041]; senior[0.041]; senior[0.041]; payment[0.041]; chairman[0.040]; chairman[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Home Depot\u001b[39m ==> ENTITY: \u001b[32mThe Home Depot\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstores[0.045]; stores[0.045]; college[0.044]; company[0.044]; company[0.044]; hiring[0.043]; Home[0.041]; Home[0.041]; Home[0.041]; Home[0.041]; Home[0.041]; senior[0.041]; deal[0.041]; CEO[0.041]; big[0.041]; job[0.041]; employees[0.041]; employees[0.041]; football[0.041]; Canada[0.040]; customer[0.040]; replace[0.040]; appointment[0.040]; Mexico[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Home Depot\u001b[39m ==> ENTITY: \u001b[32mThe Home Depot\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.102:0.102[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncollege[0.054]; company[0.053]; declined[0.051]; Home[0.050]; Home[0.050]; Home[0.050]; recently[0.050]; senior[0.050]; CEO[0.050]; job[0.050]; football[0.049]; Wednesday[0.049]; key[0.049]; York[0.049]; adjusted[0.049]; news[0.049]; appointment[0.049]; said[0.049]; said[0.049]; said[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = New York Stock Exchange\u001b[39m ==> ENTITY: \u001b[32mNew York Stock Exchange\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstock[0.049]; company[0.047]; shares[0.046]; college[0.046]; Wednesday[0.046]; Wednesday[0.046]; declined[0.044]; office[0.044]; investors[0.043]; week[0.043]; senior[0.043]; close[0.043]; said[0.042]; said[0.042]; said[0.042]; recently[0.042]; December[0.042]; football[0.042]; leaving[0.042]; rose[0.042]; return[0.042]; immediately[0.042]; immediately[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.081:0.081[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.048]; Mexico[0.046]; seven[0.044]; China[0.044]; times[0.044]; Blake[0.044]; Blake[0.044]; Tuesday[0.041]; awards[0.041]; total[0.041]; June[0.041]; decision[0.041]; timing[0.041]; currently[0.041]; deferred[0.041]; set[0.040]; individual[0.040]; remains[0.040]; official[0.040]; board[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Home Depot\u001b[39m ==> ENTITY: \u001b[32mThe Home Depot\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstores[0.045]; company[0.044]; company[0.044]; Home[0.042]; Home[0.042]; Home[0.042]; Home[0.042]; Home[0.042]; Home[0.042]; Home[0.042]; operates[0.041]; helped[0.041]; Wednesday[0.041]; leadership[0.041]; interview[0.041]; Bob[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; approximately[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nardelli\u001b[39m ==> ENTITY: \u001b[32mRobert Nardelli\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; past[0.042]; pay[0.042]; earlier[0.042]; company[0.041]; company[0.041]; board[0.041]; leadership[0.041]; AP[0.041]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; percent[0.040]; discussion[0.040]; increased[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlanta ==> ENTITY: \u001b[32mAtlanta\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.055:0.055[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWednesday[0.045]; American[0.043]; roughly[0.043]; home[0.043]; Bob[0.043]; ATLANTA[0.043]; million[0.042]; based[0.042]; largest[0.042]; executive[0.041]; executive[0.041]; Services[0.041]; legislation[0.041]; package[0.041]; chief[0.041]; chief[0.041]; Depot[0.041]; sends[0.041]; receive[0.041]; Financial[0.041]; Home[0.041]; House[0.041]; pay[0.040]; pay[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Home Depot\u001b[39m ==> ENTITY: \u001b[32mThe Home Depot\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstores[0.049]; stores[0.049]; company[0.048]; company[0.048]; hiring[0.047]; Home[0.045]; Home[0.045]; Home[0.045]; Home[0.045]; Home[0.045]; deal[0.045]; big[0.045]; employees[0.045]; employees[0.045]; Tuesday[0.044]; roughly[0.044]; Canada[0.044]; customer[0.044]; million[0.044]; million[0.044]; replace[0.044]; Mexico[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlanta ==> ENTITY: \u001b[32mAtlanta\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.066:0.066[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAtlanta[0.057]; Wednesday[0.054]; Bob[0.051]; home[0.051]; based[0.050]; executive[0.050]; executive[0.050]; chief[0.050]; chief[0.050]; Depot[0.049]; Depot[0.049]; receive[0.049]; Home[0.049]; Home[0.049]; Home[0.049]; largest[0.049]; helm[0.049]; pay[0.048]; pay[0.048]; chairman[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marcus\u001b[39m ==> ENTITY: \u001b[32mBernard Marcus\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfounder[0.044]; college[0.044]; Marcus[0.044]; Marcus[0.044]; Marcus[0.044]; chairman[0.043]; CEO[0.043]; board[0.042]; Jerry[0.042]; hiring[0.042]; job[0.042]; company[0.041]; company[0.041]; company[0.041]; executive[0.041]; senior[0.041]; Bernie[0.041]; Blake[0.041]; Blake[0.041]; Blake[0.041]; Blake[0.041]; football[0.040]; replace[0.040]; appointment[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nardelli\u001b[39m ==> ENTITY: \u001b[32mRobert Nardelli\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nexecutives[0.047]; executives[0.047]; American[0.043]; Tuesday[0.043]; pay[0.042]; employees[0.042]; employees[0.042]; June[0.042]; CEO[0.042]; million[0.042]; million[0.042]; million[0.042]; companies[0.042]; annual[0.041]; Canada[0.041]; senior[0.041]; senior[0.041]; chairman[0.040]; service[0.040]; package[0.040]; package[0.040]; package[0.040]; meeting[0.040]; Services[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Home Depot\u001b[39m ==> ENTITY: \u001b[32mThe Home Depot\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.115:0.115[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncompany[0.045]; declined[0.043]; Home[0.043]; Home[0.043]; Home[0.043]; recently[0.043]; senior[0.042]; job[0.042]; Wednesday[0.042]; Wednesday[0.042]; key[0.041]; York[0.041]; adjusted[0.041]; news[0.041]; interview[0.041]; appointment[0.041]; said[0.041]; said[0.041]; Nardelli[0.041]; Nardelli[0.041]; Nardelli[0.041]; Nardelli[0.041]; Nardelli[0.041]; week[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nardelli\u001b[39m ==> ENTITY: \u001b[32mRobert Nardelli\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Wednesday[0.043]; Wednesday[0.043]; recently[0.043]; earlier[0.042]; declined[0.042]; company[0.041]; despite[0.041]; board[0.041]; key[0.041]; senior[0.041]; executive[0.040]; AP[0.040]; week[0.040]; appointment[0.040]; Hill[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; returned[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.063:0.063[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nroughly[0.044]; awards[0.043]; total[0.043]; States[0.042]; Tuesday[0.042]; China[0.042]; million[0.042]; million[0.042]; million[0.042]; seven[0.042]; times[0.042]; Blake[0.042]; Blake[0.042]; Home[0.041]; Home[0.041]; Home[0.041]; Home[0.041]; Home[0.041]; Home[0.041]; timing[0.041]; decision[0.041]; currently[0.041]; June[0.041]; largest[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Home Depot\u001b[39m ==> ENTITY: \u001b[32mThe Home Depot\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.137:0.137[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstores[0.045]; company[0.044]; company[0.044]; Home[0.042]; Home[0.042]; Home[0.042]; Home[0.042]; Home[0.042]; Home[0.042]; Home[0.042]; operates[0.042]; helped[0.041]; leadership[0.041]; million[0.041]; interview[0.041]; Bob[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; approximately[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bob\u001b[39m ==> ENTITY: \u001b[32mRobert Nardelli\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; past[0.042]; pay[0.042]; million[0.041]; company[0.041]; company[0.041]; earned[0.041]; board[0.041]; leadership[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; percent[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bernie Marcus\u001b[39m ==> ENTITY: \u001b[32mBernard Marcus\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.142:0.142[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfounder[0.045]; Marcus[0.045]; Marcus[0.045]; chairman[0.044]; board[0.043]; hiring[0.043]; company[0.042]; company[0.042]; Blake[0.042]; Blake[0.042]; replace[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; Frank[0.041]; Nardelli[0.040]; Nardelli[0.040]; Nardelli[0.040]; times[0.040]; supports[0.040]; million[0.039]; million[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = House Financial Services Committee\u001b[39m ==> ENTITY: \u001b[32mUnited States House Committee on Financial Services\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nchairman[0.044]; chairman[0.044]; legislation[0.044]; executive[0.043]; resigned[0.043]; Frank[0.043]; Frank[0.043]; senior[0.042]; senior[0.042]; Rep[0.042]; lawmakers[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; Wednesday[0.041]; American[0.041]; awards[0.041]; incoming[0.041]; push[0.040]; meeting[0.040]; helm[0.040]; CEO[0.040]; decried[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marcus\u001b[39m ==> ENTITY: \u001b[32mBernard Marcus\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncollege[0.045]; Marcus[0.044]; Marcus[0.044]; CEO[0.043]; board[0.042]; Jerry[0.042]; hiring[0.042]; job[0.042]; company[0.042]; company[0.042]; company[0.042]; executive[0.042]; senior[0.042]; Blake[0.041]; Blake[0.041]; Blake[0.041]; football[0.041]; replace[0.040]; appointment[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Home Depot\u001b[39m ==> ENTITY: \u001b[32mThe Home Depot\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nretailer[0.046]; company[0.045]; store[0.044]; chain[0.044]; Home[0.042]; Home[0.042]; CEO[0.042]; ATLANTA[0.042]; world[0.041]; Bob[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; Nardelli[0.041]; Nardelli[0.041]; Nardelli[0.041]; Nardelli[0.041]; hefty[0.041]; didn[0.041]; pay[0.041]; pay[0.041]; executive[0.041]; executive[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bob Nardelli\u001b[39m ==> ENTITY: \u001b[32mRobert Nardelli\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Wednesday[0.043]; pay[0.042]; pay[0.042]; CEO[0.042]; resigned[0.042]; hefty[0.042]; company[0.041]; company[0.041]; chief[0.041]; chief[0.041]; board[0.041]; performance[0.041]; receive[0.041]; chairman[0.040]; executive[0.040]; executive[0.040]; package[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Edward Jones ==> ENTITY: \u001b[32mEdward Jones Investments\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.839:-0.839[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJames[0.045]; Business[0.044]; State[0.043]; certain[0.043]; College[0.043]; University[0.042]; credibility[0.042]; Center[0.042]; added[0.042]; company[0.042]; job[0.042]; pay[0.042]; criticism[0.041]; Robinson[0.041]; Blake[0.041]; Blake[0.041]; Bob[0.041]; considered[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; articulate[0.041]; demeanor[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = The Associated Press\u001b[39m ==> ENTITY: \u001b[32mAssociated Press\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nweek[0.045]; AP[0.045]; news[0.045]; Wednesday[0.044]; Wednesday[0.044]; interview[0.044]; told[0.044]; recently[0.042]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; adjusted[0.041]; York[0.041]; earlier[0.041]; Asked[0.040]; Asked[0.040]; declined[0.040]; despite[0.040]; available[0.039]; return[0.039]; messages[0.039]; messages[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nardelli\u001b[39m ==> ENTITY: \u001b[32mRobert Nardelli\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Wednesday[0.043]; Wednesday[0.043]; recently[0.043]; earlier[0.042]; despite[0.041]; board[0.041]; key[0.041]; AP[0.040]; week[0.040]; Hill[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; returned[0.040]; percent[0.040]; percent[0.040]; stock[0.040]; adjusted[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexico ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.059:0.059[0]; log p(e|m)= -0.183:-0.183[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.047]; Tuesday[0.046]; roughly[0.045]; Blake[0.045]; Blake[0.045]; States[0.045]; service[0.044]; United[0.044]; June[0.044]; official[0.043]; replace[0.043]; vice[0.043]; resignation[0.043]; China[0.043]; Asked[0.043]; hiring[0.043]; seven[0.043]; set[0.042]; timing[0.042]; capacity[0.042]; times[0.042]; currently[0.042]; immediately[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = The Home Depot\u001b[39m ==> ENTITY: \u001b[32mThe Home Depot\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstores[0.045]; company[0.044]; company[0.044]; Home[0.042]; Home[0.042]; Home[0.042]; Home[0.042]; Home[0.042]; Home[0.042]; Home[0.042]; CEO[0.042]; operates[0.041]; helped[0.041]; leadership[0.041]; million[0.041]; Bob[0.041]; approximately[0.041]; Nardelli[0.040]; Nardelli[0.040]; Nardelli[0.040]; Nardelli[0.040]; Nardelli[0.040]; Nardelli[0.040]; Nardelli[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nardelli\u001b[39m ==> ENTITY: \u001b[32mRobert Nardelli\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; Nardelli[0.045]; past[0.042]; pay[0.042]; CEO[0.042]; million[0.042]; company[0.041]; company[0.041]; earned[0.041]; board[0.041]; disclosed[0.041]; leadership[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; Home[0.040]; analysts[0.040]; percent[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nardelli\u001b[39m ==> ENTITY: \u001b[32mRobert Nardelli\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNardelli[0.054]; Nardelli[0.054]; Nardelli[0.054]; Tuesday[0.051]; hiring[0.051]; employees[0.050]; employees[0.050]; June[0.050]; million[0.050]; deal[0.049]; company[0.049]; company[0.049]; Canada[0.049]; won[0.049]; board[0.049]; vice[0.049]; chairman[0.048]; service[0.048]; package[0.048]; package[0.048]; \t\n\r [=================>............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s587ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 259/656 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 16447201 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ganges ==> ENTITY: \u001b[32mGanges\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.175:0.175[0]; log p(e|m)= -0.035:-0.035[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGanges[0.047]; river[0.047]; river[0.047]; Hindus[0.044]; Hindus[0.044]; Kumbh[0.043]; waters[0.042]; India[0.041]; India[0.041]; India[0.041]; sins[0.041]; cleanse[0.041]; bathing[0.041]; sin[0.041]; northern[0.041]; earth[0.041]; near[0.041]; children[0.040]; day[0.040]; people[0.040]; braved[0.039]; women[0.039]; festival[0.039]; festival[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Garuda ==> ENTITY: \u001b[32mGaruda\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.173:0.173[0]; log p(e|m)= -0.043:-0.043[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngod[0.048]; flight[0.045]; winged[0.044]; Kumbh[0.044]; Kumbh[0.044]; Kumbh[0.044]; Kumbh[0.044]; Garuda[0.044]; Vishnu[0.044]; divine[0.044]; divine[0.044]; Hindu[0.043]; mortal[0.043]; beings[0.043]; pilgrims[0.043]; pilgrims[0.043]; demons[0.042]; Maha[0.042]; Maha[0.042]; Maha[0.042]; human[0.042]; nectar[0.041]; Indian[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kumbh\u001b[39m ==> ENTITY: \u001b[32mKumbh Mela\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKumbh[0.049]; Kumbh[0.049]; Mela[0.044]; Hindu[0.043]; Ganges[0.043]; holy[0.042]; Indian[0.042]; Indian[0.042]; Indian[0.042]; Allahabad[0.041]; Allahabad[0.041]; Shakuntala[0.041]; god[0.040]; traveled[0.040]; God[0.040]; God[0.040]; waters[0.040]; waters[0.040]; Vishnu[0.040]; time[0.040]; divine[0.040]; mile[0.039]; school[0.039]; despite[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Maha Kumbh Mela\u001b[39m ==> ENTITY: \u001b[32mKumbh Mela\u001b[39m\t\nSCORES: global= 0.311:0.311[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKumbh[0.050]; Kumbh[0.050]; Mela[0.045]; pilgrims[0.043]; pilgrims[0.043]; Hindu[0.043]; festival[0.042]; festival[0.042]; Allahabad[0.041]; celebrations[0.041]; god[0.041]; celebration[0.040]; days[0.040]; gathering[0.040]; Indian[0.040]; moon[0.040]; Vishnu[0.040]; time[0.040]; time[0.040]; million[0.040]; million[0.040]; divine[0.040]; divine[0.040]; mile[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hindu scriptures\u001b[39m ==> ENTITY: \u001b[32mHindu texts\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsacred[0.046]; mythical[0.044]; gods[0.044]; verses[0.044]; Hindus[0.043]; God[0.043]; holy[0.043]; holy[0.043]; auspicious[0.042]; India[0.041]; India[0.041]; Kumbh[0.041]; Kumbh[0.041]; Ganges[0.041]; Ganges[0.041]; Ganges[0.041]; sin[0.040]; afterlife[0.040]; Yamuna[0.040]; people[0.040]; Maha[0.040]; day[0.040]; day[0.040]; day[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = India ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.164:0.164[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.051]; India[0.051]; Hindu[0.046]; Hindus[0.043]; Hindus[0.043]; Allahabad[0.043]; abroad[0.043]; day[0.042]; day[0.042]; day[0.042]; northern[0.041]; Wednesday[0.041]; week[0.040]; week[0.040]; areas[0.040]; people[0.040]; event[0.039]; half[0.039]; million[0.039]; million[0.039]; Pilgrims[0.039]; auspicious[0.039]; saffron[0.038]; Nearly[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Garuda ==> ENTITY: \u001b[32mGaruda\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.173:0.173[0]; log p(e|m)= -0.043:-0.043[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngod[0.045]; flight[0.043]; winged[0.042]; Kumbh[0.042]; Kumbh[0.042]; Kumbh[0.042]; Kumbh[0.042]; Kumbh[0.042]; Garuda[0.042]; Vishnu[0.042]; Rama[0.042]; divine[0.042]; divine[0.042]; God[0.042]; God[0.042]; Hindu[0.041]; mortal[0.041]; pilgrims[0.041]; demons[0.040]; Maha[0.040]; Maha[0.040]; Maha[0.040]; Devi[0.040]; nectar[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hindus ==> ENTITY: \u001b[32mHindu\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.170:0.170[0]; log p(e|m)= -0.229:-0.229[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHindus[0.048]; day[0.046]; Kumbh[0.045]; India[0.044]; India[0.044]; India[0.044]; celebration[0.043]; Mela[0.042]; festival[0.041]; festival[0.041]; week[0.041]; week[0.041]; holy[0.041]; people[0.041]; abroad[0.041]; Allahabad[0.040]; women[0.040]; Ganges[0.040]; Ganges[0.040]; Ganges[0.040]; Pilgrims[0.039]; sun[0.039]; mythical[0.039]; Festival[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Madhya Pradesh\u001b[39m ==> ENTITY: \u001b[32mMadhya Pradesh\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndian[0.044]; Indian[0.044]; Indian[0.044]; Kumbh[0.043]; Kumbh[0.043]; Kumbh[0.043]; Kumbh[0.043]; Allahabad[0.043]; eastern[0.043]; Bengal[0.043]; central[0.042]; state[0.042]; state[0.042]; sacred[0.042]; Mela[0.041]; Mela[0.041]; Kumar[0.041]; Ganges[0.041]; Shakuntala[0.040]; holy[0.040]; cherished[0.039]; Devi[0.039]; despite[0.039]; Rama[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Allahabad\u001b[39m ==> ENTITY: \u001b[32mAllahabad\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAllahabad[0.046]; state[0.045]; Hindu[0.044]; Indian[0.044]; Hindi[0.043]; Pradesh[0.043]; Uttar[0.042]; heartland[0.042]; Kumbh[0.042]; Kumbh[0.042]; Kumbh[0.042]; Kumbh[0.042]; Kumbh[0.042]; Maha[0.041]; Maha[0.041]; Maha[0.041]; Mela[0.041]; Mela[0.041]; Vishnu[0.041]; flight[0.040]; days[0.040]; Devi[0.040]; make[0.040]; holy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ganges ==> ENTITY: \u001b[32mGanges\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.175:0.175[0]; log p(e|m)= -0.035:-0.035[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGanges[0.046]; river[0.046]; gods[0.044]; sacred[0.044]; Yamuna[0.043]; God[0.043]; Hindus[0.042]; scriptures[0.042]; Kumbh[0.042]; Kumbh[0.042]; Indian[0.041]; waters[0.041]; Shakuntala[0.041]; Hindu[0.041]; mythical[0.041]; holy[0.041]; holy[0.041]; afterlife[0.040]; bathing[0.040]; bathing[0.040]; eastern[0.040]; sin[0.040]; Bengal[0.040]; West[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ganges ==> ENTITY: \u001b[32mGanges\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.210:0.210[0]; log p(e|m)= -0.035:-0.035[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGanges[0.047]; Ganges[0.047]; river[0.047]; river[0.047]; gods[0.046]; Yamuna[0.045]; Hindus[0.044]; Hindus[0.044]; scriptures[0.043]; Kumbh[0.043]; Kumbh[0.043]; waters[0.043]; Hindu[0.042]; mythical[0.042]; holy[0.042]; holy[0.042]; India[0.042]; India[0.042]; India[0.042]; sins[0.042]; cleanse[0.042]; bathing[0.041]; bathing[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Maha Kumbh\u001b[39m ==> ENTITY: \u001b[32mKumbh Mela\u001b[39m\t\nSCORES: global= 0.312:0.312[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKumbh[0.049]; Kumbh[0.049]; Mela[0.044]; Mela[0.044]; pilgrims[0.043]; pilgrims[0.043]; Hindu[0.043]; festival[0.042]; festival[0.042]; Allahabad[0.041]; celebrations[0.041]; god[0.040]; celebration[0.040]; days[0.040]; gathering[0.040]; moon[0.040]; Vishnu[0.040]; time[0.040]; million[0.040]; million[0.040]; divine[0.040]; divine[0.040]; drew[0.039]; area[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ganges river\u001b[39m ==> ENTITY: \u001b[32mGanges\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGanges[0.050]; Ganges[0.050]; river[0.050]; river[0.050]; Yamuna[0.047]; Hindus[0.046]; Hindus[0.046]; scriptures[0.046]; Kumbh[0.046]; Hindu[0.045]; mythical[0.044]; holy[0.044]; sins[0.044]; waters[0.044]; cleanse[0.044]; bathing[0.044]; sin[0.043]; northern[0.043]; earth[0.043]; Allahabad[0.043]; near[0.043]; children[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Allahabad\u001b[39m ==> ENTITY: \u001b[32mAllahabad\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAllahabad[0.044]; state[0.043]; state[0.043]; state[0.043]; Hindu[0.043]; Indian[0.043]; Indian[0.043]; Indian[0.043]; central[0.043]; eastern[0.042]; Hindi[0.042]; Pradesh[0.042]; Pradesh[0.042]; Ganges[0.041]; traveled[0.041]; Uttar[0.041]; heartland[0.041]; Kumbh[0.040]; Kumbh[0.040]; Kumbh[0.040]; Madhya[0.040]; Bengal[0.040]; West[0.040]; Mela[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Maha Kumbh Mela\u001b[39m ==> ENTITY: \u001b[32mKumbh Mela\u001b[39m\t\nSCORES: global= 0.312:0.312[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKumbh[0.050]; Kumbh[0.050]; Mela[0.045]; pilgrims[0.044]; pilgrims[0.044]; Hindu[0.043]; festival[0.042]; festival[0.042]; Allahabad[0.041]; celebrations[0.041]; god[0.041]; celebration[0.041]; days[0.040]; gathering[0.040]; moon[0.040]; Vishnu[0.040]; time[0.040]; million[0.040]; million[0.040]; divine[0.040]; divine[0.040]; drew[0.039]; area[0.039]; Thousands[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pilgrims ==> ENTITY: \u001b[32mPilgrim\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.061:-0.061[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nholy[0.046]; people[0.045]; Hindus[0.044]; Hindus[0.044]; day[0.044]; children[0.043]; Wednesday[0.042]; week[0.042]; week[0.042]; throng[0.042]; near[0.041]; Kumbh[0.041]; bathing[0.041]; northern[0.041]; festival[0.041]; festival[0.041]; abroad[0.040]; believe[0.040]; million[0.040]; million[0.040]; Pitcher[0.040]; sins[0.040]; Ganges[0.040]; Ganges[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hindu god\u001b[39m ==> ENTITY: \u001b[32mHindu deities\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.213:0.213[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndivine[0.049]; divine[0.049]; mortal[0.048]; Vishnu[0.048]; Indian[0.046]; Kumbh[0.046]; Kumbh[0.046]; Kumbh[0.046]; Kumbh[0.046]; Kumbh[0.046]; God[0.046]; God[0.046]; winged[0.045]; beings[0.045]; Devi[0.045]; demons[0.045]; Rama[0.045]; Garuda[0.044]; Garuda[0.044]; Hindi[0.043]; human[0.043]; considered[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ganges ==> ENTITY: \u001b[32mGanges\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.170:0.170[0]; log p(e|m)= -0.035:-0.035[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsacred[0.044]; divine[0.044]; Bengal[0.043]; God[0.043]; God[0.043]; bathe[0.042]; Kumbh[0.042]; Kumbh[0.042]; Kumbh[0.042]; Kumbh[0.042]; mile[0.042]; Indian[0.042]; Indian[0.042]; Indian[0.042]; waters[0.041]; waters[0.041]; Shakuntala[0.041]; eastern[0.041]; Rama[0.041]; holy[0.041]; Allahabad[0.040]; central[0.040]; West[0.040]; Devi[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Allahabad\u001b[39m ==> ENTITY: \u001b[32mAllahabad\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.046]; India[0.046]; India[0.046]; Hindu[0.045]; Yamuna[0.045]; city[0.044]; Hindus[0.044]; Hindus[0.044]; northern[0.044]; areas[0.044]; Kumbh[0.044]; Kumbh[0.044]; confluence[0.043]; Ganges[0.043]; Ganges[0.043]; Ganges[0.043]; falls[0.042]; river[0.042]; river[0.042]; attainment[0.042]; million[0.041]; million[0.041]; Maha[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = northern India\u001b[39m ==> ENTITY: \u001b[32mNorth India\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.051]; India[0.051]; Hindu[0.046]; Hindus[0.044]; Hindus[0.044]; Allahabad[0.043]; abroad[0.042]; saffron[0.042]; Kumbh[0.041]; Yamuna[0.041]; city[0.040]; auspicious[0.040]; Ganges[0.040]; Ganges[0.040]; Ganges[0.040]; near[0.040]; Mela[0.040]; mythical[0.040]; holy[0.039]; holy[0.039]; largest[0.039]; day[0.039]; day[0.039]; day[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hindi\u001b[39m ==> ENTITY: \u001b[32mHindi\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.097:0.097[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHindu[0.050]; Indian[0.048]; Vishnu[0.043]; god[0.042]; Devi[0.041]; divine[0.041]; divine[0.041]; divine[0.041]; holy[0.041]; Kumbh[0.041]; Kumbh[0.041]; Kumbh[0.041]; Kumbh[0.041]; Kumbh[0.041]; time[0.041]; time[0.041]; sun[0.041]; days[0.041]; state[0.041]; Allahabad[0.040]; Allahabad[0.040]; pilgrims[0.040]; heartland[0.040]; mortal[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kumbh Mela\u001b[39m ==> ENTITY: \u001b[32mKumbh Mela\u001b[39m\t\nSCORES: global= 0.311:0.311[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKumbh[0.049]; Kumbh[0.049]; Hindus[0.045]; Mela[0.044]; Mela[0.044]; Ganges[0.043]; Ganges[0.043]; holy[0.042]; sacred[0.042]; festival[0.042]; festivals[0.041]; celebrated[0.041]; traveled[0.040]; God[0.040]; God[0.040]; God[0.040]; Indian[0.040]; Indian[0.040]; divine[0.039]; school[0.039]; Shakuntala[0.039]; gods[0.039]; Festival[0.039]; Pradesh[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Guinness Book of Records\u001b[39m ==> ENTITY: \u001b[32mGuinness World Records\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.059:0.059[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntime[0.046]; single[0.045]; Mela[0.044]; Mela[0.044]; considered[0.044]; million[0.044]; million[0.044]; flight[0.044]; festival[0.044]; festival[0.044]; camps[0.044]; Garuda[0.043]; Garuda[0.043]; lasted[0.043]; said[0.043]; said[0.043]; human[0.043]; largest[0.043]; celebration[0.043]; Vishnu[0.043]; days[0.042]; dubbed[0.042]; Uttar[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kumbh\u001b[39m ==> ENTITY: \u001b[32mKumbh Mela\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKumbh[0.050]; Kumbh[0.050]; Mela[0.044]; Hindu[0.043]; Ganges[0.043]; holy[0.042]; Indian[0.042]; Indian[0.042]; Allahabad[0.041]; Allahabad[0.041]; Shakuntala[0.041]; god[0.041]; traveled[0.041]; days[0.040]; God[0.040]; God[0.040]; waters[0.040]; waters[0.040]; Vishnu[0.040]; time[0.040]; divine[0.040]; divine[0.040]; divine[0.040]; mile[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Maha Kumbh Mela\u001b[39m ==> ENTITY: \u001b[32mKumbh Mela\u001b[39m\t\nSCORES: global= 0.312:0.312[0]; local()= 0.215:0.215[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKumbh[0.049]; Hindus[0.045]; Mela[0.044]; Mela[0.044]; day[0.043]; day[0.043]; Hindu[0.042]; Ganges[0.042]; Ganges[0.042]; auspicious[0.042]; holy[0.042]; holy[0.042]; sacred[0.041]; event[0.041]; festival[0.041]; scriptures[0.040]; festivals[0.040]; Allahabad[0.040]; celebrated[0.040]; gathered[0.040]; earth[0.039]; God[0.039]; Indian[0.039]; Yamuna[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ardh Kumbh Mela\u001b[39m ==> ENTITY: \u001b[32mKumbh Mela\u001b[39m\t\nSCORES: global= 0.313:0.313[0]; local()= 0.213:0.213[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHindus[0.045]; Hindus[0.045]; day[0.043]; day[0.043]; day[0.043]; Hindu[0.043]; Ganges[0.043]; Ganges[0.043]; Ganges[0.043]; auspicious[0.042]; holy[0.042]; holy[0.042]; event[0.041]; festival[0.041]; festival[0.041]; festival[0.041]; people[0.041]; scriptures[0.041]; Allahabad[0.040]; India[0.040]; India[0.040]; India[0.040]; gathered[0.040]; celebration[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Vishnu\u001b[39m ==> ENTITY: \u001b[32mVishnu\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.005:-0.005[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngod[0.047]; divine[0.046]; divine[0.046]; mortal[0.044]; God[0.044]; Hindu[0.043]; Devi[0.042]; demons[0.042]; beings[0.042]; Indian[0.041]; Allahabad[0.041]; Allahabad[0.041]; Hindi[0.041]; immortality[0.041]; Rama[0.041]; Kumbh[0.040]; Kumbh[0.040]; Kumbh[0.040]; Kumbh[0.040]; Kumbh[0.040]; Book[0.040]; days[0.039]; pilgrims[0.039]; winged[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hindus ==> ENTITY: \u001b[32mHindu\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.164:0.164[0]; log p(e|m)= -0.229:-0.229[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHindus[0.049]; day[0.046]; Kumbh[0.045]; India[0.045]; India[0.045]; India[0.045]; celebration[0.044]; Mela[0.043]; festival[0.041]; festival[0.041]; week[0.041]; week[0.041]; people[0.041]; abroad[0.041]; Ganges[0.040]; Ganges[0.040]; Pilgrims[0.040]; Festival[0.040]; gatherings[0.039]; expected[0.039]; Ardh[0.039]; largest[0.038]; northern[0.038]; bathing[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Allahabad\u001b[39m ==> ENTITY: \u001b[32mAllahabad\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.142:0.142[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnorthern[0.046]; India[0.045]; India[0.045]; India[0.045]; Yamuna[0.043]; Hindus[0.043]; Hindus[0.043]; confluence[0.042]; Ganges[0.042]; Ganges[0.042]; Ganges[0.042]; river[0.041]; river[0.041]; river[0.041]; Kumbh[0.041]; near[0.040]; million[0.040]; million[0.040]; day[0.040]; Mela[0.040]; people[0.040]; week[0.039]; week[0.039]; holy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kumbh Mela\u001b[39m ==> ENTITY: \u001b[32mKumbh Mela\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKumbh[0.050]; Kumbh[0.050]; Kumbh[0.050]; Mela[0.045]; Ganges[0.044]; holy[0.043]; Allahabad[0.041]; traveled[0.041]; God[0.041]; God[0.041]; Indian[0.040]; Indian[0.040]; Indian[0.040]; time[0.040]; divine[0.040]; mile[0.040]; school[0.040]; Shakuntala[0.040]; despite[0.039]; walk[0.039]; walk[0.039]; Pradesh[0.039]; Devi[0.039]; Madhya[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Uttar Pradesh\u001b[39m ==> ENTITY: \u001b[32mUttar Pradesh\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHindu[0.045]; Kumbh[0.044]; Kumbh[0.044]; Kumbh[0.044]; Kumbh[0.044]; Kumbh[0.044]; Allahabad[0.043]; Allahabad[0.043]; Indian[0.043]; pilgrims[0.043]; Hindi[0.042]; holy[0.042]; Vishnu[0.041]; Mela[0.041]; Mela[0.041]; heartland[0.040]; Devi[0.040]; state[0.039]; Maha[0.039]; Maha[0.039]; Maha[0.039]; considered[0.039]; time[0.039]; time[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = West Bengal\u001b[39m ==> ENTITY: \u001b[32mWest Bengal\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndian[0.045]; Indian[0.045]; Hindus[0.044]; Kumbh[0.043]; Kumbh[0.043]; Kumbh[0.043]; sacred[0.043]; Ghosh[0.042]; eastern[0.042]; Pradesh[0.042]; Devi[0.041]; Allahabad[0.041]; like[0.041]; state[0.041]; state[0.041]; young[0.041]; central[0.041]; Madhya[0.041]; Ganges[0.040]; Ganges[0.040]; festivals[0.040]; celebrated[0.040]; Like[0.040]; divine[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yamuna\u001b[39m ==> ENTITY: \u001b[32mYamuna\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGanges[0.045]; Ganges[0.045]; Ganges[0.045]; river[0.043]; river[0.043]; Allahabad[0.043]; scriptures[0.042]; bathing[0.042]; bathing[0.042]; India[0.042]; India[0.042]; India[0.042]; gods[0.042]; Hindus[0.041]; Hindus[0.041]; near[0.041]; Hindu[0.040]; children[0.040]; sang[0.040]; Kumbh[0.040]; Kumbh[0.040]; sin[0.040]; sin[0.040]; mythical[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Indian army\u001b[39m ==> ENTITY: \u001b[32mIndian Army\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.084:0.084[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndian[0.049]; security[0.045]; central[0.044]; soldier[0.043]; Allahabad[0.042]; Allahabad[0.042]; Hindu[0.042]; days[0.042]; Madhya[0.042]; Mela[0.041]; Mela[0.041]; despite[0.041]; single[0.041]; missed[0.041]; state[0.041]; state[0.041]; Garuda[0.041]; Garuda[0.041]; considered[0.041]; determined[0.041]; Tight[0.040]; position[0.040]; celebrations[0.040]; Vishnu[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = India ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.152:0.152[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.052]; India[0.052]; Hindus[0.044]; Hindus[0.044]; Allahabad[0.044]; abroad[0.043]; day[0.043]; northern[0.042]; Wednesday[0.041]; week[0.041]; week[0.041]; half[0.040]; million[0.040]; million[0.040]; largest[0.040]; billed[0.040]; people[0.040]; Pilgrims[0.039]; saffron[0.039]; Nearly[0.039]; expected[0.039]; Ganges[0.039]; Ganges[0.039]; Ganges[0.039]; \t\n\r [===================>..........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s421ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 295/656 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 16452612 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = President George W. Bush\u001b[39m ==> ENTITY: \u001b[32mGeorge W. Bush\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.099:0.099[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBush[0.047]; Iraq[0.044]; Iraq[0.044]; sense[0.044]; Presidential[0.043]; American[0.043]; war[0.042]; election[0.042]; nation[0.041]; president[0.041]; president[0.041]; surge[0.041]; Americans[0.041]; type[0.041]; business[0.041]; country[0.041]; country[0.041]; does[0.040]; voters[0.040]; voters[0.040]; figure[0.040]; work[0.040]; certitude[0.040]; WASHINGTON[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Schumer\u001b[39m ==> ENTITY: \u001b[32mChuck Schumer\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.089:0.089[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSchumer[0.048]; York[0.044]; money[0.043]; Charles[0.043]; president[0.043]; Sen[0.043]; American[0.042]; American[0.042]; party[0.042]; proposals[0.042]; really[0.042]; opening[0.042]; address[0.041]; cultural[0.041]; congenital[0.041]; student[0.041]; forthcoming[0.040]; direction[0.040]; family[0.040]; family[0.040]; family[0.040]; family[0.040]; understands[0.040]; understands[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pfeiffer\u001b[39m ==> ENTITY: \u001b[32mDaniel Pfeiffer\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinton[0.044]; John[0.044]; Kerry[0.044]; Indiana[0.044]; campaign[0.043]; Washington[0.043]; Pfeiffer[0.042]; Pfeiffer[0.042]; Pfeiffer[0.042]; Bayh[0.042]; Bayh[0.042]; senator[0.041]; Bill[0.041]; Americans[0.041]; Americans[0.041]; world[0.041]; separate[0.041]; Bush[0.040]; Bush[0.040]; country[0.040]; Dan[0.040]; governor[0.040]; opening[0.040]; Vietnam[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.093:0.093[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWashington[0.047]; Washington[0.047]; Washington[0.047]; Senate[0.045]; studied[0.044]; career[0.042]; American[0.042]; voters[0.041]; senator[0.041]; governor[0.041]; office[0.041]; Indiana[0.040]; run[0.040]; run[0.040]; promised[0.040]; said[0.040]; president[0.040]; Voters[0.040]; Voters[0.040]; elections[0.040]; people[0.040]; Bayh[0.040]; Bayh[0.040]; Bayh[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nixon\u001b[39m ==> ENTITY: \u001b[32mRichard Nixon\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.122:0.122[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNixon[0.046]; Richard[0.044]; American[0.044]; Americans[0.043]; Americans[0.043]; spending[0.043]; Presidential[0.042]; office[0.042]; Washington[0.042]; Washington[0.042]; Washington[0.042]; Bush[0.042]; mid[0.041]; Carter[0.041]; comes[0.041]; voters[0.040]; Ford[0.040]; Ford[0.040]; going[0.040]; example[0.040]; Iraq[0.040]; pardoned[0.040]; leadership[0.040]; candidate[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pfeiffer\u001b[39m ==> ENTITY: \u001b[32mDaniel Pfeiffer\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.047]; Clinton[0.047]; John[0.046]; Kerry[0.046]; campaign[0.045]; Pfeiffer[0.045]; Pfeiffer[0.045]; understands[0.044]; Bill[0.043]; world[0.043]; separate[0.043]; Bush[0.043]; Bush[0.043]; governmental[0.043]; really[0.042]; opening[0.042]; Charles[0.042]; Vietnam[0.042]; family[0.042]; family[0.042]; family[0.042]; says[0.042]; cultural[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gerald Ford\u001b[39m ==> ENTITY: \u001b[32mGerald Ford\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNixon[0.046]; Nixon[0.046]; days[0.046]; Presidential[0.044]; Carter[0.043]; Ford[0.042]; American[0.042]; Washington[0.042]; Washington[0.042]; Washington[0.042]; Richard[0.042]; worked[0.041]; office[0.041]; Bush[0.041]; work[0.041]; evidence[0.041]; promised[0.040]; career[0.040]; mid[0.040]; leadership[0.040]; spending[0.040]; details[0.039]; old[0.039]; Jimmy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.080:0.080[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWashington[0.050]; Washington[0.050]; Senate[0.047]; studied[0.046]; election[0.045]; voters[0.044]; senator[0.044]; country[0.043]; governor[0.043]; office[0.043]; Indiana[0.043]; said[0.043]; said[0.043]; president[0.043]; Voters[0.043]; elections[0.043]; Bayh[0.042]; Bayh[0.042]; Bayh[0.042]; says[0.041]; arguing[0.041]; Vietnam[0.041]; strategists[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = John Kerry\u001b[39m ==> ENTITY: \u001b[32mJohn Kerry\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncampaign[0.048]; Clinton[0.045]; voters[0.044]; election[0.043]; Bush[0.043]; Bush[0.043]; Bill[0.042]; won[0.042]; Voters[0.042]; Vietnam[0.041]; referring[0.041]; country[0.041]; said[0.041]; said[0.041]; said[0.041]; Americans[0.041]; arguing[0.041]; party[0.040]; says[0.040]; elections[0.040]; issues[0.040]; Bayh[0.040]; culture[0.040]; understands[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indiana ==> ENTITY: \u001b[32mIndiana\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.091:0.091[0]; log p(e|m)= -0.272:-0.272[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBayh[0.044]; Bayh[0.044]; Bayh[0.044]; Senate[0.043]; studied[0.043]; American[0.043]; office[0.043]; president[0.043]; governor[0.042]; Washington[0.042]; Washington[0.042]; Washington[0.042]; senator[0.041]; run[0.041]; run[0.041]; Dan[0.041]; library[0.041]; worked[0.040]; time[0.040]; country[0.040]; voters[0.040]; spending[0.040]; change[0.040]; elections[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.077:0.077[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWashington[0.048]; Washington[0.048]; studied[0.044]; career[0.043]; American[0.043]; voters[0.042]; Presidential[0.042]; senator[0.042]; student[0.042]; incumbent[0.042]; office[0.041]; run[0.041]; troops[0.041]; promised[0.041]; president[0.041]; elections[0.041]; people[0.040]; candidate[0.040]; library[0.040]; Richard[0.040]; spending[0.040]; Voters[0.040]; Voters[0.040]; Voters[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Richard Nixon\u001b[39m ==> ENTITY: \u001b[32mRichard Nixon\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.116:0.116[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNixon[0.046]; American[0.044]; Americans[0.043]; Americans[0.043]; spending[0.043]; war[0.042]; Presidential[0.042]; office[0.042]; Washington[0.042]; Washington[0.042]; Washington[0.042]; Bush[0.042]; mid[0.042]; Carter[0.041]; comes[0.041]; voters[0.041]; Ford[0.040]; Ford[0.040]; going[0.040]; example[0.040]; Iraq[0.040]; Iraq[0.040]; pardoned[0.040]; leadership[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vietnam ==> ENTITY: \u001b[32mVietnam War\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.093:0.093[0]; log p(e|m)= -2.354:-2.354[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.048]; John[0.045]; Dan[0.043]; won[0.043]; Kerry[0.043]; Americans[0.043]; Washington[0.043]; campaign[0.042]; Bill[0.042]; lack[0.042]; referring[0.041]; opening[0.041]; era[0.041]; says[0.041]; family[0.040]; family[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; Bush[0.040]; Bush[0.040]; money[0.040]; Clinton[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.146:0.146[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.054]; troops[0.047]; war[0.046]; country[0.044]; country[0.044]; Bush[0.043]; Bush[0.043]; surge[0.043]; Americans[0.041]; George[0.041]; nation[0.040]; President[0.040]; likely[0.040]; elections[0.040]; figure[0.039]; based[0.039]; career[0.039]; failure[0.039]; leader[0.039]; leader[0.039]; necessary[0.039]; leaders[0.039]; time[0.039]; time[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Charles Schumer\u001b[39m ==> ENTITY: \u001b[32mChuck Schumer\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.115:0.115[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSchumer[0.051]; Bill[0.049]; Clinton[0.048]; York[0.047]; said[0.047]; said[0.047]; money[0.046]; president[0.046]; Sen[0.046]; American[0.045]; American[0.045]; party[0.045]; proposals[0.045]; really[0.044]; opening[0.044]; address[0.044]; concerns[0.044]; cultural[0.044]; congenital[0.043]; student[0.043]; forthcoming[0.043]; direction[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dan Pfeiffer\u001b[39m ==> ENTITY: \u001b[32mDaniel Pfeiffer\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJohn[0.044]; president[0.044]; Indiana[0.044]; campaign[0.043]; Washington[0.043]; Washington[0.043]; Washington[0.043]; Pfeiffer[0.043]; Pfeiffer[0.043]; Bayh[0.042]; Bayh[0.042]; Bayh[0.042]; senator[0.041]; Americans[0.041]; Americans[0.041]; world[0.041]; separate[0.041]; Bush[0.041]; country[0.040]; governor[0.040]; Vietnam[0.040]; change[0.040]; says[0.040]; doesn[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.094:0.094[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWashington[0.047]; Washington[0.047]; Washington[0.047]; Senate[0.045]; studied[0.044]; career[0.042]; American[0.042]; voters[0.042]; senator[0.041]; incumbent[0.041]; governor[0.041]; office[0.041]; Indiana[0.041]; promised[0.040]; said[0.040]; president[0.040]; Voters[0.040]; Voters[0.040]; people[0.040]; Bayh[0.040]; Bayh[0.040]; candidate[0.039]; Richard[0.039]; spending[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ford\u001b[39m ==> ENTITY: \u001b[32mGerald Ford\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.137:0.137[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNixon[0.046]; Nixon[0.046]; days[0.046]; Presidential[0.044]; Carter[0.043]; Ford[0.042]; American[0.042]; Washington[0.042]; Washington[0.042]; Washington[0.042]; Richard[0.042]; worked[0.042]; office[0.041]; Bush[0.041]; work[0.041]; evidence[0.041]; promised[0.040]; career[0.040]; mid[0.040]; leadership[0.040]; spending[0.040]; details[0.039]; Jimmy[0.039]; pardoned[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bush\u001b[39m ==> ENTITY: \u001b[32mGeorge W. Bush\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.107:0.107[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBush[0.047]; sense[0.044]; Clinton[0.044]; Kerry[0.043]; campaign[0.043]; John[0.043]; arguing[0.042]; election[0.042]; obsessions[0.042]; said[0.042]; said[0.042]; said[0.042]; Americans[0.041]; life[0.041]; Bill[0.041]; country[0.041]; referring[0.041]; culture[0.040]; voters[0.040]; work[0.040]; party[0.040]; won[0.040]; elections[0.040]; thing[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Senate ==> ENTITY: \u001b[32mUnited States Senate\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.115:0.115[0]; log p(e|m)= -1.277:-1.277[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsenator[0.046]; governor[0.045]; office[0.044]; president[0.043]; Washington[0.043]; Washington[0.043]; Washington[0.043]; country[0.043]; Dan[0.042]; Bayh[0.041]; Bayh[0.041]; Bayh[0.041]; elections[0.041]; Indiana[0.041]; voters[0.041]; Sen[0.041]; run[0.040]; run[0.040]; Voters[0.040]; Voters[0.040]; said[0.040]; case[0.040]; time[0.040]; arguing[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pfeiffer\u001b[39m ==> ENTITY: \u001b[32mDaniel Pfeiffer\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinton[0.045]; John[0.044]; Kerry[0.044]; campaign[0.043]; Washington[0.043]; Pfeiffer[0.043]; Pfeiffer[0.043]; Pfeiffer[0.043]; Bayh[0.042]; understands[0.042]; Bill[0.041]; Americans[0.041]; world[0.041]; separate[0.041]; Bush[0.041]; Bush[0.041]; country[0.041]; Dan[0.041]; really[0.040]; opening[0.040]; Vietnam[0.040]; family[0.040]; family[0.040]; change[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Evan Bayh\u001b[39m ==> ENTITY: \u001b[32mEvan Bayh\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBayh[0.046]; Bayh[0.046]; senator[0.045]; Senate[0.043]; Indiana[0.043]; Washington[0.042]; Washington[0.042]; Washington[0.042]; Washington[0.042]; Nixon[0.042]; governor[0.042]; office[0.041]; spending[0.041]; American[0.041]; president[0.041]; candidate[0.041]; Sen[0.041]; said[0.040]; run[0.040]; run[0.040]; Dan[0.039]; Carter[0.039]; pardoned[0.039]; promised[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jimmy Carter\u001b[39m ==> ENTITY: \u001b[32mJimmy Carter\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.118:0.118[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNixon[0.046]; Nixon[0.046]; Bush[0.046]; Iraq[0.045]; Richard[0.045]; career[0.044]; Washington[0.044]; Washington[0.044]; Washington[0.044]; American[0.044]; Presidential[0.044]; president[0.043]; troops[0.043]; pardoned[0.043]; evidence[0.042]; office[0.042]; necessary[0.042]; want[0.042]; Gerald[0.042]; look[0.042]; thinking[0.042]; going[0.042]; time[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew York\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew York City\u001b[39m <---> \u001b[32mNew York\u001b[39m\t\nSCORES: global= 0.256:0.252[\u001b[31m0.004\u001b[39m]; local()= 0.069:0.058[\u001b[31m0.011\u001b[39m]; log p(e|m)= -1.790:-0.289[\u001b[32m1.501\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseries[0.046]; forthcoming[0.044]; president[0.043]; book[0.043]; work[0.042]; American[0.042]; American[0.042]; hard[0.042]; time[0.042]; Bill[0.042]; Charles[0.041]; understands[0.041]; understands[0.041]; said[0.041]; said[0.041]; class[0.041]; class[0.041]; class[0.041]; party[0.041]; address[0.041]; prove[0.040]; Positively[0.040]; really[0.040]; leaves[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.072:0.072[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGeorge[0.045]; election[0.045]; President[0.044]; American[0.044]; voters[0.043]; voters[0.043]; nation[0.043]; country[0.043]; country[0.043]; politicians[0.043]; president[0.042]; president[0.042]; leaders[0.041]; based[0.041]; Bush[0.040]; strategists[0.040]; time[0.040]; time[0.040]; passed[0.040]; business[0.039]; want[0.039]; want[0.039]; talking[0.039]; laid[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bush\u001b[39m ==> ENTITY: \u001b[32mGeorge W. Bush\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.115:0.115[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBush[0.046]; President[0.045]; Iraq[0.044]; Iraq[0.044]; George[0.043]; sense[0.043]; Presidential[0.043]; war[0.042]; Nixon[0.041]; Nixon[0.041]; nation[0.041]; Richard[0.041]; surge[0.041]; Americans[0.041]; Americans[0.041]; type[0.041]; life[0.041]; pure[0.041]; Carter[0.040]; country[0.040]; country[0.040]; Gerald[0.040]; candidate[0.040]; does[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bayh\u001b[39m ==> ENTITY: \u001b[32mEvan Bayh\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBayh[0.045]; Bayh[0.045]; Kerry[0.045]; senator[0.044]; John[0.044]; Bush[0.043]; Bush[0.043]; campaign[0.043]; Senate[0.043]; Indiana[0.043]; Washington[0.041]; Washington[0.041]; governor[0.041]; election[0.041]; office[0.040]; president[0.040]; Sen[0.040]; Sen[0.040]; said[0.040]; said[0.040]; run[0.039]; run[0.039]; says[0.039]; Dan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bill Clinton\u001b[39m ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.120:0.120[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nelection[0.044]; Bush[0.044]; Bush[0.044]; John[0.044]; book[0.043]; campaign[0.043]; American[0.043]; Americans[0.042]; Kerry[0.042]; voters[0.042]; referring[0.041]; Charles[0.041]; country[0.041]; Bushes[0.041]; said[0.041]; said[0.041]; said[0.041]; upbeat[0.041]; time[0.041]; answer[0.040]; selling[0.040]; says[0.040]; lack[0.040]; Know[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bayh\u001b[39m ==> ENTITY: \u001b[32mEvan Bayh\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBayh[0.047]; Bayh[0.047]; senator[0.046]; Senate[0.044]; Indiana[0.044]; Washington[0.043]; Washington[0.043]; Washington[0.043]; governor[0.042]; election[0.042]; office[0.042]; spending[0.041]; president[0.041]; Sen[0.041]; said[0.041]; run[0.040]; run[0.040]; says[0.040]; Dan[0.040]; voters[0.040]; Voters[0.039]; Americans[0.038]; Americans[0.038]; resume[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAmerican people\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUnited States\u001b[39m <---> \u001b[32mAmericans\u001b[39m\t\nSCORES: global= 0.262:0.247[\u001b[31m0.015\u001b[39m]; local()= 0.086:0.075[\u001b[31m0.012\u001b[39m]; log p(e|m)= -0.518:-0.183[\u001b[32m0.335\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmericans[0.045]; Richard[0.044]; Presidential[0.044]; voters[0.043]; Washington[0.043]; Washington[0.043]; Washington[0.043]; candidate[0.043]; Indiana[0.043]; president[0.042]; senator[0.042]; career[0.042]; mid[0.042]; series[0.041]; elections[0.041]; Jimmy[0.041]; office[0.040]; run[0.040]; strategists[0.040]; chose[0.040]; Gerald[0.040]; worked[0.040]; want[0.040]; time[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bush campaign\u001b[39m ==> ENTITY: \u001b[32mGeorge W. Bush presidential campaign, 2004\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKerry[0.047]; Clinton[0.046]; John[0.046]; Bush[0.044]; election[0.043]; Bill[0.043]; voters[0.042]; won[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; Dan[0.040]; Americans[0.040]; Vietnam[0.040]; says[0.040]; elections[0.040]; era[0.040]; era[0.040]; arguing[0.040]; Bayh[0.040]; really[0.040]; worked[0.040]; opening[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.141:0.141[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.054]; troops[0.047]; war[0.046]; country[0.044]; Bush[0.043]; Bush[0.043]; surge[0.043]; Washington[0.042]; Americans[0.041]; Americans[0.041]; George[0.041]; nation[0.040]; President[0.040]; promised[0.040]; likely[0.040]; elections[0.040]; based[0.039]; career[0.039]; failure[0.039]; leader[0.039]; necessary[0.039]; leaders[0.039]; time[0.039]; time[0.039]; \t\n\r [=====================>........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s284ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 327/656 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 3683270 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = New York Stock Exchange\u001b[39m ==> ENTITY: \u001b[32mNew York Stock Exchange\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.213:0.213[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstock[0.051]; shares[0.048]; shares[0.048]; shares[0.048]; Tuesday[0.047]; Nasdaq[0.047]; month[0.046]; month[0.046]; Market[0.045]; investors[0.045]; investors[0.045]; Stock[0.045]; Trading[0.045]; stocks[0.044]; stocks[0.044]; day[0.044]; day[0.044]; commodity[0.044]; share[0.044]; indexes[0.044]; declining[0.044]; session[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = FTSE 100\u001b[39m ==> ENTITY: \u001b[32mFTSE 100 Index\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncompanies[0.048]; Britain[0.047]; stock[0.047]; Corp[0.044]; tumbled[0.044]; Overseas[0.044]; DAX[0.043]; Russell[0.043]; percent[0.043]; percent[0.043]; percent[0.043]; percent[0.043]; percent[0.043]; percent[0.043]; percent[0.043]; percent[0.043]; percent[0.043]; percent[0.043]; sales[0.043]; index[0.042]; index[0.042]; Sunwear[0.042]; Pacific[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Exxon Mobil Corp.\u001b[39m ==> ENTITY: \u001b[32mExxonMobil\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncompanies[0.047]; companies[0.047]; Oil[0.045]; oil[0.044]; Corp[0.042]; Sunwear[0.042]; Pacific[0.042]; higher[0.041]; Monday[0.041]; Target[0.041]; Japan[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; California[0.041]; spurred[0.040]; Exchange[0.040]; Overseas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.074:0.074[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.047]; Britain[0.045]; Japan[0.045]; lost[0.043]; short[0.042]; sales[0.041]; slipped[0.041]; rose[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; fell[0.041]; fell[0.041]; fell[0.041]; fell[0.041]; fell[0.041]; fell[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oklahoma City ==> ENTITY: \u001b[32mOklahoma City\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.093:0.093[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMonday[0.047]; October[0.042]; October[0.042]; sign[0.042]; nation[0.042]; level[0.042]; points[0.042]; Jones[0.042]; expected[0.042]; based[0.042]; days[0.042]; low[0.042]; Department[0.042]; report[0.041]; composite[0.041]; said[0.041]; said[0.041]; said[0.041]; highest[0.041]; issue[0.041]; industrial[0.041]; Advisory[0.041]; close[0.040]; Oct[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.048:0.048[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstock[0.049]; Landesman[0.045]; Corp[0.045]; Management[0.045]; months[0.044]; October[0.044]; October[0.044]; Monday[0.043]; market[0.043]; market[0.043]; market[0.043]; reported[0.043]; senior[0.043]; spending[0.043]; spending[0.043]; surged[0.043]; said[0.043]; said[0.043]; said[0.043]; investors[0.043]; chief[0.043]; Investment[0.042]; manager[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.093:0.093[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.046]; Germany[0.046]; Pacific[0.045]; France[0.043]; Russell[0.043]; United[0.043]; States[0.043]; short[0.042]; Overseas[0.042]; slipped[0.042]; tumbled[0.042]; California[0.041]; companies[0.041]; fell[0.041]; fell[0.041]; fell[0.041]; fell[0.041]; fell[0.041]; fell[0.041]; lost[0.041]; Mobil[0.040]; rose[0.040]; rose[0.040]; rose[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saks Inc.\u001b[39m ==> ENTITY: \u001b[32mSaks, Inc.\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nretail[0.046]; ING[0.044]; Target[0.043]; retailers[0.042]; Corp[0.042]; biggest[0.042]; senior[0.042]; Monday[0.042]; stock[0.042]; shares[0.042]; operating[0.042]; quarter[0.042]; Management[0.042]; shopping[0.041]; manager[0.041]; Depot[0.041]; reported[0.041]; share[0.041]; season[0.041]; Investment[0.041]; portfolio[0.040]; earnings[0.040]; surged[0.040]; lack[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NYSE\u001b[39m ==> ENTITY: \u001b[32mNew York Stock Exchange\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nExchange[0.049]; market[0.048]; shares[0.048]; Nasdaq[0.047]; Mercantile[0.047]; ended[0.046]; gain[0.045]; stocks[0.045]; stocks[0.045]; analysts[0.045]; month[0.045]; climbed[0.045]; Monday[0.045]; Monday[0.045]; close[0.045]; report[0.045]; March[0.044]; billion[0.044]; jumped[0.044]; index[0.044]; index[0.044]; retail[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Target ==> ENTITY: \u001b[32mTarget Corporation\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.092:0.092[0]; log p(e|m)= -0.147:-0.147[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsales[0.047]; Monday[0.044]; smaller[0.044]; Saks[0.044]; companies[0.044]; companies[0.044]; Home[0.044]; Bond[0.044]; Sunwear[0.044]; Corp[0.044]; Corp[0.044]; Russell[0.043]; Britain[0.043]; Depot[0.043]; tumbled[0.043]; note[0.043]; Platinum[0.042]; Overseas[0.042]; slipped[0.042]; slipped[0.042]; short[0.042]; jumped[0.042]; lower[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dow ==> ENTITY: \u001b[32mDow Jones Industrial Average\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.248:0.248[0]; log p(e|m)= -1.519:-1.519[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngain[0.045]; Dow[0.043]; Dow[0.043]; Dow[0.043]; stocks[0.042]; month[0.042]; index[0.042]; index[0.042]; market[0.042]; low[0.042]; close[0.041]; rose[0.041]; rose[0.041]; rose[0.041]; rose[0.041]; rose[0.041]; Index[0.041]; composite[0.041]; ended[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Thomson Reuters\u001b[39m ==> ENTITY: \u001b[32mThomson Reuters\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbased[0.043]; investment[0.043]; ING[0.043]; Investment[0.043]; market[0.042]; market[0.042]; report[0.042]; report[0.042]; saying[0.042]; Group[0.042]; said[0.041]; said[0.041]; said[0.041]; Department[0.041]; October[0.041]; October[0.041]; October[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; Monday[0.041]; nation[0.041]; retail[0.041]; investors[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = DAX\u001b[39m ==> ENTITY: \u001b[32mDAX\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncompanies[0.046]; FTSE[0.044]; stock[0.044]; index[0.043]; index[0.043]; Nikkei[0.043]; California[0.042]; Britain[0.042]; Germany[0.042]; tumbled[0.041]; Russell[0.041]; France[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; short[0.040]; rose[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Target Corp.\u001b[39m ==> ENTITY: \u001b[32mTarget Corporation\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.121:0.121[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nretail[0.049]; retailers[0.047]; shopping[0.044]; ING[0.044]; sales[0.044]; senior[0.042]; consumer[0.041]; biggest[0.041]; operating[0.041]; Saks[0.041]; quarter[0.041]; Management[0.041]; season[0.041]; goods[0.041]; low[0.041]; Home[0.041]; expected[0.041]; manager[0.040]; holiday[0.040]; reported[0.040]; recessionary[0.040]; focused[0.040]; Monday[0.040]; volume[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.057:0.057[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.058]; Britain[0.057]; lost[0.055]; CAC[0.054]; Japan[0.054]; slipped[0.053]; fell[0.052]; fell[0.052]; fell[0.052]; fell[0.052]; fell[0.052]; fell[0.052]; Russell[0.052]; sales[0.052]; short[0.051]; tumbled[0.051]; Overseas[0.051]; rose[0.050]; average[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nasdaq Stock Market\u001b[39m ==> ENTITY: \u001b[32mNASDAQ\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstock[0.049]; Exchange[0.044]; shares[0.044]; shares[0.044]; shares[0.044]; investors[0.043]; investors[0.043]; declining[0.042]; Stock[0.042]; Trading[0.042]; share[0.041]; Tuesday[0.041]; low[0.040]; day[0.040]; day[0.040]; Stocks[0.040]; Stocks[0.040]; Stocks[0.040]; month[0.040]; month[0.040]; stocks[0.040]; stocks[0.040]; volume[0.040]; indexes[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nikkei stock average\u001b[39m ==> ENTITY: \u001b[32mNikkei 225\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntumbled[0.053]; index[0.053]; index[0.053]; Japan[0.052]; companies[0.051]; short[0.050]; Overseas[0.050]; rose[0.050]; rose[0.050]; rose[0.050]; DAX[0.049]; fell[0.049]; fell[0.049]; fell[0.049]; fell[0.049]; fell[0.049]; fell[0.049]; FTSE[0.049]; Corp[0.049]; Corp[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dow ==> ENTITY: \u001b[32mDow Jones Industrial Average\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.250:0.250[0]; log p(e|m)= -1.519:-1.519[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngain[0.051]; gains[0.049]; Dow[0.049]; Dow[0.049]; Dow[0.049]; stocks[0.048]; stocks[0.048]; companies[0.048]; month[0.048]; index[0.048]; index[0.048]; market[0.048]; low[0.047]; close[0.047]; rose[0.047]; rose[0.047]; rose[0.047]; rose[0.047]; climbed[0.046]; composite[0.046]; ended[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Producer Price Index\u001b[39m ==> ENTITY: \u001b[32mProducer price index\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninflation[0.044]; inflation[0.044]; inflation[0.044]; consumer[0.044]; wholesale[0.043]; level[0.043]; rates[0.043]; market[0.043]; market[0.043]; Labor[0.042]; output[0.041]; measures[0.041]; retail[0.041]; economy[0.041]; economy[0.041]; economy[0.041]; based[0.041]; report[0.041]; report[0.041]; economists[0.040]; economists[0.040]; government[0.040]; production[0.040]; average[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = ING Investment Management\u001b[39m ==> ENTITY: \u001b[32mING Group\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCorp[0.044]; retail[0.043]; senior[0.043]; portfolio[0.043]; government[0.042]; Home[0.042]; stock[0.042]; season[0.042]; investors[0.042]; biggest[0.041]; biggest[0.041]; market[0.041]; market[0.041]; market[0.041]; market[0.041]; nation[0.041]; Stocks[0.041]; Stocks[0.041]; earnings[0.041]; operating[0.041]; recessionary[0.041]; quarter[0.041]; Investors[0.041]; Investors[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Standard & Poor's 500\u001b[39m ==> ENTITY: \u001b[32mS&P 500\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.217:0.217[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nindex[0.044]; index[0.044]; investment[0.043]; average[0.043]; broader[0.043]; market[0.042]; market[0.042]; gain[0.042]; analysts[0.042]; NYSE[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; stocks[0.041]; shares[0.041]; smaller[0.041]; rose[0.041]; rose[0.041]; rose[0.041]; rose[0.041]; outpaced[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = CAC-40\u001b[39m ==> ENTITY: \u001b[32mCAC 40\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFTSE[0.044]; stock[0.043]; DAX[0.043]; index[0.043]; index[0.043]; companies[0.043]; percent[0.043]; percent[0.043]; percent[0.043]; percent[0.043]; percent[0.043]; percent[0.043]; percent[0.043]; percent[0.043]; percent[0.043]; average[0.042]; France[0.041]; Nikkei[0.041]; Sunwear[0.040]; tumbled[0.040]; Pacific[0.040]; rose[0.039]; Germany[0.039]; Russell[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Home Depot\u001b[39m ==> ENTITY: \u001b[32mThe Home Depot\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.088:0.088[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTarget[0.047]; companies[0.044]; companies[0.044]; sales[0.044]; DAX[0.043]; Corp[0.042]; Corp[0.042]; Monday[0.042]; York[0.042]; smaller[0.042]; Sunwear[0.042]; late[0.042]; Pacific[0.041]; Russell[0.041]; Britain[0.041]; Germany[0.041]; spurred[0.040]; Exxon[0.040]; tumbled[0.040]; stock[0.040]; Saks[0.040]; Mobil[0.040]; Mercantile[0.040]; note[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dow Jones industrial average\u001b[39m ==> ENTITY: \u001b[32mDow Jones Industrial Average\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.242:0.242[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngain[0.045]; Dow[0.044]; Dow[0.044]; index[0.042]; index[0.042]; market[0.042]; low[0.042]; close[0.042]; rose[0.042]; rose[0.042]; rose[0.042]; rose[0.042]; rose[0.042]; Index[0.041]; composite[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; rise[0.041]; million[0.041]; stocks[0.040]; highest[0.040]; investment[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = The Fed\u001b[39m ==> ENTITY: \u001b[32mFederal Reserve System\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nING[0.045]; government[0.044]; economy[0.044]; economy[0.044]; economy[0.044]; Department[0.044]; Management[0.044]; nation[0.044]; investment[0.044]; report[0.044]; report[0.044]; months[0.043]; spending[0.043]; Price[0.043]; senior[0.043]; issue[0.043]; market[0.043]; market[0.043]; market[0.043]; recovery[0.042]; welcome[0.042]; measures[0.042]; portfolio[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBritain\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUnited Kingdom\u001b[39m <---> \u001b[32mGreat Britain\u001b[39m\t\nSCORES: global= 0.260:0.246[\u001b[31m0.014\u001b[39m]; local()= 0.077:0.063[\u001b[31m0.013\u001b[39m]; log p(e|m)= -0.611:-1.698[\u001b[31m1.088\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.045]; Germany[0.044]; States[0.044]; short[0.043]; Japan[0.043]; companies[0.043]; Russell[0.043]; Home[0.042]; lost[0.041]; Overseas[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; slipped[0.041]; California[0.040]; Pacific[0.040]; United[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.026:0.026[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstock[0.047]; shares[0.044]; shares[0.044]; Exchange[0.043]; Nasdaq[0.043]; month[0.043]; month[0.043]; York[0.043]; slightly[0.042]; Tuesday[0.041]; Market[0.041]; investors[0.041]; news[0.041]; Stock[0.041]; Stock[0.041]; stocks[0.041]; day[0.041]; day[0.041]; dollar[0.040]; dollar[0.040]; erratic[0.040]; commodity[0.040]; indexes[0.040]; declining[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.059:0.059[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndollar[0.045]; dollar[0.045]; dollar[0.045]; month[0.044]; month[0.044]; Record[0.044]; Home[0.042]; volume[0.042]; reaching[0.042]; reported[0.042]; reports[0.042]; news[0.041]; lifted[0.041]; expected[0.041]; day[0.041]; share[0.041]; stock[0.040]; biggest[0.040]; lifting[0.040]; lifting[0.040]; days[0.040]; advance[0.040]; industrial[0.040]; Tuesday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = New York Mercantile Exchange\u001b[39m ==> ENTITY: \u001b[32mNew York Mercantile Exchange\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmarket[0.044]; volume[0.043]; million[0.042]; Oil[0.042]; benchmark[0.042]; cents[0.042]; cents[0.042]; cents[0.042]; cents[0.042]; cents[0.042]; barrel[0.042]; oil[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; stocks[0.041]; stocks[0.041]; stocks[0.041]; companies[0.041]; crude[0.041]; prices[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nasdaq\u001b[39m ==> ENTITY: \u001b[32mNASDAQ\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmarket[0.045]; shares[0.044]; NYSE[0.044]; jumped[0.043]; ended[0.042]; Group[0.042]; gain[0.042]; investment[0.042]; rose[0.041]; rose[0.041]; rose[0.041]; rose[0.041]; rose[0.041]; close[0.041]; Monday[0.041]; Monday[0.041]; million[0.041]; sales[0.041]; analysts[0.041]; low[0.041]; low[0.041]; retail[0.041]; March[0.040]; helped[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Labor Department\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Labor\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlevel[0.044]; October[0.043]; October[0.043]; October[0.043]; report[0.043]; report[0.043]; Advisory[0.042]; government[0.042]; City[0.042]; ended[0.041]; Monday[0.041]; nation[0.041]; rose[0.041]; rose[0.041]; rose[0.041]; rates[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; economy[0.041]; economy[0.041]; economy[0.041]; wholesale[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Saks ==> ENTITY: \u001b[32mSaks, Inc.\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.135:0.135[0]; log p(e|m)= -1.671:-1.671[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTarget[0.044]; sales[0.044]; Mercantile[0.043]; average[0.043]; States[0.042]; Corp[0.042]; Corp[0.042]; Monday[0.042]; stock[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; Britain[0.041]; Exchange[0.041]; smaller[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Home Depot Inc.\u001b[39m ==> ENTITY: \u001b[32mThe Home Depot\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.118:0.118[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTarget[0.047]; retailers[0.044]; ING[0.044]; shopping[0.044]; spending[0.043]; senior[0.043]; biggest[0.042]; manager[0.042]; Corp[0.042]; Monday[0.042]; focused[0.041]; consumer[0.041]; Management[0.041]; season[0.041]; operating[0.041]; ahead[0.041]; expected[0.041]; said[0.041]; said[0.041]; quarter[0.040]; past[0.040]; low[0.040]; goods[0.040]; recessionary[0.040]; \t\n\r [========================>.....................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s128ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 363/656 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 16444023 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Diaz\u001b[39m ==> ENTITY: \u001b[32mCameron Diaz\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nactress[0.049]; Timberlake[0.047]; Timberlake[0.047]; Timberlake[0.047]; singer[0.043]; singer[0.043]; reportedly[0.042]; Justin[0.042]; musical[0.042]; Diaz[0.041]; Diaz[0.041]; Diaz[0.041]; magazine[0.041]; magazine[0.041]; magazine[0.041]; Vail[0.040]; Night[0.040]; Cameron[0.040]; quits[0.040]; Colo[0.039]; says[0.039]; friends[0.039]; couple[0.039]; according[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = N' Sync\u001b[39m ==> ENTITY: \u001b[32m'N Sync\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimberlake[0.046]; Timberlake[0.046]; Timberlake[0.046]; Timberlake[0.046]; Timberlake[0.046]; singer[0.044]; singer[0.044]; Spears[0.042]; Spears[0.042]; reportedly[0.042]; music[0.041]; Britney[0.041]; performance[0.041]; Justin[0.040]; Vail[0.040]; musical[0.040]; magazine[0.039]; magazine[0.039]; magazine[0.039]; magazine[0.039]; Colo[0.039]; Star[0.039]; Star[0.039]; Kids[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Christmas ==> ENTITY: \u001b[32mChristmas\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.078:0.078[0]; log p(e|m)= -0.060:-0.060[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmagazine[0.044]; magazine[0.044]; magazine[0.044]; magazine[0.044]; Justin[0.043]; shortly[0.042]; singer[0.042]; singer[0.042]; late[0.042]; Sync[0.042]; Sync[0.042]; Saturday[0.042]; according[0.041]; musical[0.041]; Timberlake[0.041]; Timberlake[0.041]; Timberlake[0.041]; Timberlake[0.041]; Timberlake[0.041]; near[0.041]; Star[0.041]; Star[0.041]; reports[0.041]; Night[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Star magazine\u001b[39m ==> ENTITY: \u001b[32mStar (magazine)\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmagazine[0.058]; magazine[0.058]; magazine[0.058]; Star[0.049]; reports[0.048]; dating[0.048]; told[0.047]; reported[0.047]; Vail[0.046]; reportedly[0.046]; says[0.046]; quotes[0.046]; Colo[0.045]; family[0.045]; family[0.045]; Kids[0.045]; couple[0.045]; Awards[0.045]; musical[0.045]; source[0.044]; according[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Timberlake\u001b[39m ==> ENTITY: \u001b[32mJustin Timberlake\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimberlake[0.050]; Timberlake[0.050]; singer[0.045]; music[0.043]; Spears[0.043]; Spears[0.043]; magazine[0.042]; magazine[0.042]; magazine[0.042]; Live[0.041]; late[0.041]; actress[0.041]; Sync[0.040]; Britney[0.040]; musical[0.040]; Awards[0.040]; started[0.040]; started[0.040]; Night[0.040]; told[0.040]; began[0.040]; performance[0.040]; Saturday[0.040]; shortly[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Diaz\u001b[39m ==> ENTITY: \u001b[32mCameron Diaz\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nactress[0.048]; Timberlake[0.046]; Timberlake[0.046]; Timberlake[0.046]; Timberlake[0.046]; Lou[0.042]; singer[0.042]; singer[0.042]; reportedly[0.041]; Justin[0.041]; musical[0.041]; Finn[0.041]; Finn[0.041]; Diaz[0.040]; Diaz[0.040]; Diaz[0.040]; magazine[0.040]; magazine[0.040]; magazine[0.040]; magazine[0.040]; dating[0.040]; dating[0.040]; Vail[0.040]; Veronica[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Colo.\u001b[39m ==> ENTITY: \u001b[32mColorado\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.066:0.066[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVail[0.044]; late[0.044]; spent[0.043]; Kids[0.043]; Awards[0.042]; family[0.042]; family[0.042]; began[0.042]; musical[0.042]; according[0.042]; Sync[0.042]; Sync[0.042]; near[0.042]; says[0.042]; dating[0.042]; dating[0.042]; Live[0.041]; Saturday[0.041]; seen[0.040]; couple[0.040]; Dec[0.040]; Christmas[0.040]; Memphis[0.040]; Britney[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Star ==> ENTITY: \u001b[32mStar (magazine)\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.161:0.161[0]; log p(e|m)= -3.016:-3.016[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmagazine[0.049]; magazine[0.049]; magazine[0.049]; magazine[0.049]; actress[0.044]; Star[0.041]; family[0.041]; family[0.041]; reports[0.040]; dating[0.040]; Timberlake[0.040]; Timberlake[0.040]; Timberlake[0.040]; Timberlake[0.040]; Timberlake[0.040]; told[0.040]; Veronica[0.040]; Night[0.040]; Vail[0.040]; reported[0.040]; musical[0.040]; Sync[0.039]; Sync[0.039]; according[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Spears\u001b[39m ==> ENTITY: \u001b[32mBritney Spears\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimberlake[0.047]; Timberlake[0.047]; Timberlake[0.047]; singer[0.045]; Spears[0.044]; Britney[0.043]; according[0.043]; magazine[0.043]; magazine[0.043]; music[0.042]; musical[0.041]; Live[0.041]; performance[0.041]; actress[0.040]; reported[0.040]; says[0.040]; called[0.040]; briefly[0.039]; began[0.039]; impresario[0.039]; Night[0.039]; seen[0.039]; Awards[0.039]; told[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = N' Sync\u001b[39m ==> ENTITY: \u001b[32m'N Sync\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimberlake[0.047]; Timberlake[0.047]; Timberlake[0.047]; Timberlake[0.047]; singer[0.045]; singer[0.045]; reportedly[0.043]; performance[0.042]; Justin[0.041]; Vail[0.041]; musical[0.041]; magazine[0.040]; magazine[0.040]; magazine[0.040]; Colo[0.040]; Star[0.040]; Star[0.040]; Night[0.040]; Sync[0.039]; says[0.039]; Memphis[0.039]; Live[0.039]; Saturday[0.039]; Cameron[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Veronica Finn\u001b[39m ==> ENTITY: \u001b[32mInnosense\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimberlake[0.046]; Timberlake[0.046]; Timberlake[0.046]; singer[0.045]; Spears[0.045]; Spears[0.045]; magazine[0.044]; magazine[0.044]; magazine[0.044]; started[0.044]; started[0.044]; Britney[0.043]; Vail[0.043]; Pearlman[0.043]; says[0.042]; impresario[0.042]; briefly[0.042]; began[0.042]; Diaz[0.042]; Diaz[0.042]; Colo[0.042]; girl[0.042]; Kids[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Memphis ==> ENTITY: \u001b[32mMemphis, Tennessee\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.107:0.107[0]; log p(e|m)= -0.646:-0.646[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsays[0.049]; seeing[0.049]; seeing[0.049]; singer[0.048]; singer[0.048]; musical[0.047]; near[0.046]; told[0.046]; late[0.045]; Saturday[0.045]; reportedly[0.045]; according[0.044]; according[0.044]; began[0.044]; dating[0.044]; dating[0.044]; spent[0.044]; Night[0.044]; performance[0.044]; quits[0.044]; Justin[0.044]; shortly[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Diaz\u001b[39m ==> ENTITY: \u001b[32mCameron Diaz\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.213:0.213[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nactress[0.048]; Timberlake[0.046]; Timberlake[0.046]; Timberlake[0.046]; Timberlake[0.046]; Timberlake[0.046]; singer[0.042]; singer[0.042]; reportedly[0.041]; Justin[0.041]; musical[0.041]; Finn[0.040]; Diaz[0.040]; Diaz[0.040]; Diaz[0.040]; magazine[0.040]; magazine[0.040]; magazine[0.040]; magazine[0.040]; dating[0.040]; Vail[0.040]; Veronica[0.040]; Night[0.039]; Cameron[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Timberlake\u001b[39m ==> ENTITY: \u001b[32mJustin Timberlake\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.227:0.227[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimberlake[0.051]; Timberlake[0.051]; Timberlake[0.051]; singer[0.045]; singer[0.045]; music[0.044]; Spears[0.044]; Spears[0.044]; magazine[0.043]; magazine[0.043]; magazine[0.043]; magazine[0.043]; reportedly[0.042]; Live[0.042]; late[0.042]; actress[0.041]; Sync[0.041]; Sync[0.041]; Justin[0.041]; Britney[0.041]; musical[0.041]; reports[0.041]; Awards[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Innosense\u001b[39m ==> ENTITY: \u001b[32mInnosense\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimberlake[0.044]; Timberlake[0.044]; singer[0.044]; Spears[0.043]; Spears[0.043]; magazine[0.043]; magazine[0.043]; started[0.042]; started[0.042]; Britney[0.042]; Pearlman[0.041]; says[0.041]; impresario[0.041]; briefly[0.041]; began[0.041]; Diaz[0.041]; girl[0.041]; Kids[0.041]; according[0.040]; group[0.040]; Awards[0.040]; music[0.040]; Lou[0.040]; told[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saturday Night Live\u001b[39m ==> ENTITY: \u001b[32mSaturday Night Live\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.003:-0.003[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlate[0.044]; Timberlake[0.043]; Timberlake[0.043]; Timberlake[0.043]; Timberlake[0.043]; Timberlake[0.043]; quits[0.043]; reportedly[0.043]; musical[0.042]; performance[0.042]; Awards[0.042]; actress[0.041]; began[0.041]; briefly[0.041]; singer[0.041]; singer[0.041]; shortly[0.041]; dating[0.040]; dating[0.040]; break[0.040]; Lou[0.040]; Cameron[0.040]; Christmas[0.040]; seen[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vail ==> ENTITY: \u001b[32mVail, Colorado\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.392:-0.392[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaturday[0.048]; late[0.047]; actress[0.046]; magazine[0.046]; magazine[0.046]; magazine[0.046]; magazine[0.046]; according[0.046]; family[0.046]; family[0.046]; singer[0.046]; singer[0.046]; near[0.045]; seeing[0.045]; seeing[0.045]; started[0.045]; started[0.045]; met[0.044]; began[0.044]; Justin[0.044]; spent[0.044]; quits[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Timberlake\u001b[39m ==> ENTITY: \u001b[32mJustin Timberlake\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimberlake[0.050]; Timberlake[0.050]; singer[0.045]; singer[0.045]; magazine[0.042]; magazine[0.042]; magazine[0.042]; reportedly[0.042]; Live[0.041]; according[0.041]; actress[0.041]; Sync[0.040]; Sync[0.040]; Justin[0.040]; musical[0.040]; According[0.040]; reports[0.040]; Night[0.040]; told[0.040]; performance[0.040]; Saturday[0.040]; Star[0.040]; Star[0.040]; Cameron[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Justin Timberlake\u001b[39m ==> ENTITY: \u001b[32mJustin Timberlake\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimberlake[0.050]; Timberlake[0.050]; Timberlake[0.050]; singer[0.044]; singer[0.044]; magazine[0.042]; magazine[0.042]; magazine[0.042]; magazine[0.042]; reportedly[0.041]; Live[0.041]; according[0.040]; actress[0.040]; Sync[0.040]; Sync[0.040]; musical[0.040]; According[0.040]; Awards[0.040]; reports[0.039]; started[0.039]; Night[0.039]; told[0.039]; performance[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Britney Spears\u001b[39m ==> ENTITY: \u001b[32mBritney Spears\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimberlake[0.047]; Timberlake[0.047]; Timberlake[0.047]; singer[0.045]; Spears[0.044]; according[0.043]; magazine[0.043]; magazine[0.043]; music[0.042]; musical[0.042]; Live[0.041]; performance[0.041]; actress[0.041]; reported[0.040]; says[0.040]; called[0.040]; briefly[0.040]; began[0.040]; impresario[0.040]; Night[0.039]; seen[0.039]; Awards[0.039]; told[0.039]; Choice[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kids' Choice Awards\u001b[39m ==> ENTITY: \u001b[32mNickelodeon Kids' Choice Awards\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.142:0.142[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimberlake[0.044]; Timberlake[0.044]; Timberlake[0.044]; Timberlake[0.044]; Cameron[0.042]; Star[0.042]; Star[0.042]; magazine[0.042]; magazine[0.042]; magazine[0.042]; magazine[0.042]; Lou[0.041]; Night[0.041]; says[0.041]; Saturday[0.041]; actress[0.041]; musical[0.041]; Spears[0.041]; Spears[0.041]; Live[0.041]; late[0.040]; Finn[0.040]; Finn[0.040]; Dec[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Timberlake\u001b[39m ==> ENTITY: \u001b[32mJustin Timberlake\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.234:0.234[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimberlake[0.049]; Timberlake[0.049]; Timberlake[0.049]; Timberlake[0.049]; singer[0.043]; singer[0.043]; Spears[0.042]; Spears[0.042]; magazine[0.041]; magazine[0.041]; magazine[0.041]; magazine[0.041]; reportedly[0.040]; Live[0.040]; late[0.040]; according[0.040]; actress[0.039]; Sync[0.039]; Sync[0.039]; Justin[0.039]; Britney[0.039]; musical[0.039]; According[0.039]; Awards[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lou Pearlman\u001b[39m ==> ENTITY: \u001b[32mLou Pearlman\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nimpresario[0.045]; singer[0.044]; magazine[0.044]; told[0.043]; began[0.042]; Timberlake[0.042]; Timberlake[0.042]; shortly[0.042]; late[0.042]; musical[0.042]; says[0.042]; called[0.042]; music[0.041]; Awards[0.041]; reported[0.041]; Kids[0.041]; Live[0.041]; group[0.041]; introduced[0.041]; dating[0.040]; dating[0.040]; briefly[0.040]; started[0.040]; started[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cameron Diaz\u001b[39m ==> ENTITY: \u001b[32mCameron Diaz\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nactress[0.048]; Timberlake[0.046]; Timberlake[0.046]; Timberlake[0.046]; Timberlake[0.046]; Timberlake[0.046]; singer[0.042]; singer[0.042]; reportedly[0.041]; Justin[0.041]; musical[0.041]; Diaz[0.040]; Diaz[0.040]; Diaz[0.040]; magazine[0.040]; magazine[0.040]; magazine[0.040]; magazine[0.040]; dating[0.040]; Vail[0.040]; Night[0.039]; quits[0.039]; Colo[0.039]; Kids[0.039]; \t\n\r [==========================>...................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s48ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 388/656 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 13259309 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mTimes\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mThe Times\u001b[39m <---> \u001b[32mThe New York Times\u001b[39m\t\nSCORES: global= 0.244:0.236[\u001b[31m0.008\u001b[39m]; local()= 0.156:0.151[\u001b[31m0.005\u001b[39m]; log p(e|m)= -0.252:-2.397[\u001b[31m2.145\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninterview[0.046]; Times[0.046]; newspaper[0.046]; reported[0.044]; told[0.043]; video[0.042]; video[0.042]; national[0.042]; cell[0.042]; Wednesday[0.041]; Wednesday[0.041]; death[0.041]; adviser[0.041]; adviser[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; intelligence[0.041]; AP[0.040]; appeared[0.040]; Jazeera[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = al-Rubaie\u001b[39m ==> ENTITY: \u001b[32mMowaffak al-Rubaie\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRubaie[0.046]; Iraqi[0.044]; Maliki[0.044]; Maliki[0.044]; al[0.043]; al[0.043]; al[0.043]; Saddam[0.041]; Saddam[0.041]; Muqtada[0.041]; security[0.041]; security[0.041]; security[0.041]; Hussein[0.041]; Al[0.041]; Al[0.041]; video[0.041]; video[0.041]; Najaf[0.040]; adviser[0.040]; adviser[0.040]; military[0.040]; Jazeera[0.040]; national[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saddam\u001b[39m ==> ENTITY: \u001b[32mSaddam Hussein\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaddam[0.046]; Saddam[0.046]; Saddam[0.046]; Iraqi[0.045]; Iraqi[0.045]; military[0.043]; Jazeera[0.042]; saying[0.042]; report[0.041]; forces[0.041]; Al[0.040]; Al[0.040]; execution[0.040]; execution[0.040]; execution[0.040]; execution[0.040]; execution[0.040]; al[0.040]; video[0.040]; Maliki[0.040]; Maliki[0.040]; said[0.040]; said[0.040]; witnesses[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sunnis ==> ENTITY: \u001b[32mSunni Islam\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.036:-0.036[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nShiite[0.048]; forces[0.046]; Iraq[0.046]; Mahdi[0.046]; al[0.046]; al[0.046]; military[0.045]; military[0.045]; militant[0.045]; government[0.044]; targeted[0.043]; Army[0.043]; Saddam[0.042]; Saddam[0.042]; Saddam[0.042]; Maliki[0.042]; believed[0.042]; American[0.042]; Arabiya[0.042]; Al[0.041]; Al[0.041]; official[0.041]; militia[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Al-Sadr\u001b[39m ==> ENTITY: \u001b[32mMuqtada al-Sadr\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitia[0.046]; Mahdi[0.045]; Iraq[0.045]; Iraqi[0.045]; Sadr[0.045]; forces[0.045]; Shiite[0.044]; Maliki[0.044]; military[0.044]; military[0.044]; militant[0.043]; Muqtada[0.043]; Muqtada[0.043]; Muqtada[0.043]; Muqtada[0.043]; official[0.043]; al[0.043]; al[0.043]; government[0.043]; Army[0.042]; Saddam[0.042]; Saddam[0.042]; Saddam[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saddam Hussein\u001b[39m ==> ENTITY: \u001b[32mSaddam Hussein\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.047]; Iraq[0.047]; Saddam[0.046]; Saddam[0.046]; Iraqi[0.045]; Hussein[0.044]; Rubaie[0.043]; security[0.042]; security[0.042]; Al[0.040]; execution[0.040]; execution[0.040]; execution[0.040]; execution[0.040]; al[0.040]; video[0.040]; Maliki[0.040]; said[0.040]; said[0.040]; told[0.039]; Wednesday[0.039]; News[0.039]; official[0.039]; media[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. military ==> ENTITY: \u001b[32mUnited States Armed Forces\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.015:-0.015[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraqi[0.045]; execution[0.044]; execution[0.044]; execution[0.044]; execution[0.044]; security[0.043]; ordered[0.042]; national[0.042]; case[0.042]; charge[0.042]; government[0.041]; announced[0.041]; Maliki[0.041]; Maliki[0.041]; States[0.041]; officials[0.041]; gallows[0.040]; Rubaie[0.040]; Rubaie[0.040]; hanged[0.040]; spokesman[0.040]; official[0.040]; report[0.040]; responsibility[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Web site ==> ENTITY: \u001b[32mWebsite\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.120:0.120[0]; log p(e|m)= -0.017:-0.017[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nvideo[0.048]; video[0.048]; video[0.048]; Press[0.042]; Mr[0.042]; Mr[0.042]; government[0.042]; national[0.041]; security[0.041]; security[0.041]; adviser[0.041]; adviser[0.041]; Wednesday[0.041]; Wednesday[0.041]; phone[0.041]; phone[0.041]; official[0.041]; newspaper[0.041]; pictures[0.041]; secretary[0.040]; interview[0.040]; mobile[0.040]; men[0.040]; believed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saddam\u001b[39m ==> ENTITY: \u001b[32mSaddam Hussein\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaddam[0.046]; Saddam[0.046]; Saddam[0.046]; Saddam[0.046]; Iraqi[0.044]; Iraqi[0.044]; Hussein[0.043]; Rubaie[0.043]; military[0.042]; Jazeera[0.042]; saying[0.041]; security[0.041]; forces[0.040]; Al[0.040]; Al[0.040]; execution[0.040]; execution[0.040]; execution[0.040]; execution[0.040]; al[0.040]; al[0.040]; video[0.040]; Maliki[0.039]; Maliki[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = AP ==> ENTITY: \u001b[32mAssociated Press\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.687:-0.687[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnational[0.045]; Wednesday[0.043]; Wednesday[0.043]; newspaper[0.043]; Times[0.043]; Times[0.043]; phones[0.042]; told[0.042]; phone[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; reported[0.041]; security[0.041]; security[0.041]; security[0.041]; video[0.041]; video[0.041]; believed[0.041]; believed[0.041]; mobile[0.041]; satellite[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = al-Maliki\u001b[39m ==> ENTITY: \u001b[32mNouri al-Maliki\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMaliki[0.049]; military[0.048]; Rubaie[0.047]; Rubaie[0.047]; al[0.047]; al[0.047]; al[0.047]; Iraqi[0.046]; adviser[0.045]; adviser[0.045]; Hussein[0.045]; Najaf[0.045]; Saddam[0.045]; Saddam[0.045]; Muqtada[0.045]; security[0.045]; security[0.045]; security[0.045]; secretary[0.044]; national[0.044]; Al[0.044]; believed[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saddam\u001b[39m ==> ENTITY: \u001b[32mSaddam Hussein\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.047]; Saddam[0.047]; Saddam[0.047]; Iraqi[0.045]; military[0.043]; military[0.043]; military[0.043]; American[0.042]; forces[0.041]; government[0.041]; Al[0.040]; according[0.040]; according[0.040]; execution[0.040]; execution[0.040]; al[0.040]; al[0.040]; video[0.040]; Maliki[0.040]; targeted[0.040]; reporters[0.040]; said[0.040]; believed[0.040]; Mahdi[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = al-Maliki\u001b[39m ==> ENTITY: \u001b[32mNouri al-Maliki\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.045]; military[0.043]; military[0.043]; military[0.043]; Mahdi[0.043]; al[0.042]; Shiite[0.042]; official[0.042]; Iraqi[0.042]; government[0.042]; militia[0.041]; Sadr[0.041]; Sadr[0.041]; militant[0.041]; Saddam[0.041]; Saddam[0.041]; Saddam[0.041]; Muqtada[0.041]; Muqtada[0.041]; Muqtada[0.041]; Muqtada[0.041]; forces[0.041]; Army[0.040]; Al[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.066:0.066[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.048]; national[0.045]; taking[0.044]; taking[0.044]; taking[0.044]; day[0.043]; said[0.042]; television[0.042]; Saturday[0.042]; Wednesday[0.042]; moments[0.041]; Tuesday[0.041]; saying[0.041]; sent[0.041]; Associated[0.040]; Associated[0.040]; late[0.040]; official[0.040]; video[0.040]; video[0.040]; officials[0.040]; saw[0.040]; denied[0.040]; charge[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mowaffak al-Rubaie\u001b[39m ==> ENTITY: \u001b[32mMowaffak al-Rubaie\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.045]; Iraq[0.045]; Iraqi[0.044]; Iraqi[0.044]; Maliki[0.044]; Saddam[0.042]; Saddam[0.042]; Saddam[0.042]; Saddam[0.042]; security[0.042]; security[0.042]; Hussein[0.041]; Hussein[0.041]; Al[0.041]; video[0.041]; video[0.041]; forces[0.041]; adviser[0.041]; adviser[0.041]; adviser[0.041]; adviser[0.041]; leader[0.040]; national[0.040]; national[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saddam\u001b[39m ==> ENTITY: \u001b[32mSaddam Hussein\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaddam[0.047]; Hussein[0.045]; Rubaie[0.044]; Rubaie[0.044]; Jazeera[0.043]; security[0.042]; security[0.042]; satellite[0.041]; government[0.041]; Al[0.041]; execution[0.041]; execution[0.041]; execution[0.041]; execution[0.041]; al[0.041]; al[0.041]; al[0.041]; al[0.041]; al[0.041]; video[0.041]; video[0.041]; video[0.041]; Maliki[0.041]; Maliki[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Najaf\u001b[39m ==> ENTITY: \u001b[32mNajaf\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nal[0.049]; al[0.049]; al[0.049]; al[0.049]; Iraqi[0.049]; Hussein[0.048]; Al[0.048]; Muqtada[0.046]; Rubaie[0.046]; Rubaie[0.046]; believed[0.045]; Maliki[0.045]; Maliki[0.045]; Saddam[0.045]; Saddam[0.045]; military[0.043]; video[0.043]; facility[0.042]; security[0.042]; security[0.042]; security[0.042]; national[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saddam\u001b[39m ==> ENTITY: \u001b[32mSaddam Hussein\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraqi[0.052]; Rubaie[0.050]; Rubaie[0.050]; military[0.049]; security[0.048]; security[0.048]; security[0.048]; Arabiya[0.047]; intelligence[0.047]; Al[0.047]; Al[0.047]; according[0.047]; execution[0.047]; execution[0.047]; execution[0.047]; al[0.047]; al[0.047]; al[0.047]; video[0.047]; Maliki[0.046]; Maliki[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Associated Press\u001b[39m ==> ENTITY: \u001b[32mAssociated Press\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nreport[0.047]; national[0.045]; team[0.045]; military[0.044]; Iraqi[0.044]; Iraqi[0.044]; media[0.042]; video[0.042]; official[0.042]; Wednesday[0.042]; Tuesday[0.042]; forces[0.041]; saying[0.041]; television[0.041]; Jazeera[0.041]; security[0.041]; announced[0.040]; pictures[0.040]; late[0.040]; Saturday[0.040]; charge[0.040]; quoted[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.065:0.065[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArmy[0.051]; military[0.046]; military[0.046]; military[0.046]; American[0.045]; Iraq[0.042]; according[0.042]; according[0.042]; William[0.042]; video[0.042]; reference[0.041]; words[0.041]; decision[0.041]; decision[0.041]; asking[0.041]; charge[0.040]; Wednesday[0.039]; handled[0.039]; person[0.039]; heard[0.039]; Caldwell[0.039]; Caldwell[0.039]; believed[0.038]; transferred[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = William Caldwell ==> ENTITY: \u001b[32mWilliam B. Caldwell\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.807:-0.807[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArmy[0.048]; military[0.046]; military[0.046]; military[0.046]; Caldwell[0.044]; Iraq[0.043]; charge[0.043]; forces[0.042]; Iraqi[0.042]; American[0.042]; father[0.041]; prison[0.041]; Wednesday[0.040]; transferred[0.040]; official[0.040]; spokesman[0.040]; said[0.040]; said[0.040]; key[0.040]; added[0.040]; getting[0.039]; Maj[0.039]; responsible[0.039]; believed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = al-Rubaie\u001b[39m ==> ENTITY: \u001b[32mMowaffak al-Rubaie\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRubaie[0.047]; Iraqi[0.044]; Maliki[0.044]; Maliki[0.044]; al[0.044]; al[0.044]; al[0.044]; al[0.044]; Saddam[0.042]; Saddam[0.042]; security[0.042]; Al[0.041]; video[0.041]; video[0.041]; government[0.041]; adviser[0.041]; adviser[0.041]; military[0.040]; saying[0.040]; national[0.040]; Nouri[0.040]; Minister[0.039]; official[0.039]; official[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = New York Times\u001b[39m ==> ENTITY: \u001b[32mThe New York Times\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.117:0.117[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimes[0.048]; newspaper[0.048]; video[0.045]; video[0.045]; video[0.045]; interview[0.045]; Press[0.044]; Web[0.044]; reported[0.044]; national[0.043]; Wednesday[0.043]; Wednesday[0.043]; prosecutor[0.043]; security[0.042]; security[0.042]; adviser[0.042]; adviser[0.042]; appeared[0.042]; officials[0.042]; officials[0.042]; government[0.042]; site[0.042]; official[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Al-Jazeera\u001b[39m ==> ENTITY: \u001b[32mAl Jazeera\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntelevision[0.048]; military[0.047]; Al[0.046]; al[0.045]; video[0.045]; announced[0.044]; media[0.044]; filmed[0.044]; Press[0.044]; quoted[0.044]; Iraqi[0.043]; Iraqi[0.043]; official[0.043]; prime[0.043]; saying[0.043]; team[0.042]; United[0.042]; Wednesday[0.042]; forces[0.042]; Saddam[0.042]; Saddam[0.042]; Saddam[0.042]; Saddam[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Al-Arabiya\u001b[39m ==> ENTITY: \u001b[32mAl Arabiya\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntelevision[0.049]; official[0.048]; al[0.047]; Shiite[0.046]; Al[0.046]; Al[0.046]; Mahdi[0.046]; according[0.046]; according[0.046]; video[0.046]; video[0.046]; Iraqi[0.046]; MSNBC[0.045]; Saddam[0.045]; Saddam[0.045]; reported[0.044]; Army[0.044]; targeted[0.044]; officials[0.044]; officials[0.044]; officials[0.044]; officials[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = MSNBC\u001b[39m ==> ENTITY: \u001b[32mMSNBC\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.107:0.107[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntelevision[0.051]; video[0.048]; video[0.048]; reported[0.046]; according[0.045]; according[0.045]; team[0.044]; key[0.043]; American[0.043]; said[0.043]; said[0.043]; pictures[0.043]; car[0.043]; targeted[0.042]; leaked[0.042]; member[0.042]; facility[0.041]; responsible[0.041]; official[0.041]; phones[0.041]; phones[0.041]; response[0.041]; taking[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NBC News\u001b[39m ==> ENTITY: \u001b[32mNBC News\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.103:0.103[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfilmed[0.047]; prime[0.047]; filming[0.046]; video[0.044]; Wednesday[0.044]; national[0.042]; national[0.042]; adviser[0.042]; adviser[0.042]; adviser[0.042]; adviser[0.042]; recording[0.041]; told[0.041]; announcement[0.041]; official[0.041]; officials[0.040]; Iraq[0.040]; Iraq[0.040]; individuals[0.040]; individuals[0.040]; case[0.040]; security[0.040]; security[0.040]; came[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. military ==> ENTITY: \u001b[32mUnited States Armed Forces\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.126:0.126[0]; log p(e|m)= -0.015:-0.015[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArmy[0.049]; military[0.049]; military[0.049]; Iraq[0.048]; American[0.045]; forces[0.045]; Iraqi[0.044]; charge[0.041]; government[0.041]; William[0.040]; Maliki[0.040]; according[0.040]; spokesman[0.040]; key[0.040]; official[0.039]; responsible[0.039]; militia[0.039]; decision[0.039]; decision[0.039]; Caldwell[0.039]; Caldwell[0.039]; response[0.039]; Wednesday[0.038]; control[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. forces ==> ENTITY: \u001b[32mUnited States Armed Forces\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.138:0.138[0]; log p(e|m)= -0.104:-0.104[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArmy[0.049]; military[0.048]; military[0.048]; military[0.048]; military[0.048]; Iraq[0.048]; American[0.045]; Iraqi[0.044]; government[0.040]; charge[0.040]; William[0.040]; Maliki[0.040]; spokesman[0.039]; targeted[0.039]; responsible[0.039]; Maj[0.039]; key[0.039]; decision[0.039]; decision[0.039]; Caldwell[0.038]; Caldwell[0.038]; Caldwell[0.038]; response[0.038]; Wednesday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baghdad\u001b[39m ==> ENTITY: \u001b[32mBaghdad\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.051]; Iraq[0.051]; Iraqi[0.046]; Rubaie[0.044]; Hussein[0.043]; Hussein[0.043]; al[0.043]; Saddam[0.042]; Saddam[0.042]; Saddam[0.042]; official[0.042]; security[0.042]; security[0.042]; Maliki[0.041]; Al[0.041]; officials[0.040]; announcement[0.039]; Wednesday[0.039]; recording[0.039]; national[0.039]; national[0.039]; arrests[0.038]; process[0.038]; phone[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mowaffak al-Rubaie\u001b[39m ==> ENTITY: \u001b[32mMowaffak al-Rubaie\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRubaie[0.059]; Iraqi[0.055]; Maliki[0.055]; al[0.055]; al[0.055]; al[0.055]; Saddam[0.053]; Saddam[0.053]; security[0.052]; Al[0.052]; video[0.052]; video[0.052]; video[0.052]; government[0.051]; adviser[0.051]; military[0.051]; national[0.050]; official[0.049]; official[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saddam\u001b[39m ==> ENTITY: \u001b[32mSaddam Hussein\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.225:0.225[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.046]; Iraq[0.046]; Saddam[0.045]; Saddam[0.045]; Saddam[0.045]; Saddam[0.045]; Iraqi[0.044]; Iraqi[0.044]; Hussein[0.043]; Hussein[0.043]; Rubaie[0.042]; Jazeera[0.041]; security[0.040]; security[0.040]; forces[0.039]; Al[0.039]; Al[0.039]; execution[0.039]; execution[0.039]; execution[0.039]; execution[0.039]; al[0.039]; video[0.039]; video[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hussein\u001b[39m ==> ENTITY: \u001b[32mSaddam Hussein\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaddam[0.051]; Iraqi[0.050]; Rubaie[0.048]; Rubaie[0.048]; Jazeera[0.047]; security[0.046]; security[0.046]; satellite[0.045]; Al[0.044]; Al[0.044]; execution[0.044]; execution[0.044]; execution[0.044]; al[0.044]; al[0.044]; al[0.044]; al[0.044]; al[0.044]; video[0.044]; video[0.044]; Maliki[0.044]; Maliki[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saddam\u001b[39m ==> ENTITY: \u001b[32mSaddam Hussein\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaddam[0.048]; military[0.044]; American[0.043]; security[0.043]; Arabiya[0.042]; Al[0.041]; Al[0.041]; according[0.041]; according[0.041]; execution[0.041]; execution[0.041]; al[0.041]; al[0.041]; video[0.041]; video[0.041]; Maliki[0.041]; targeted[0.041]; reporters[0.041]; said[0.041]; reported[0.041]; believed[0.041]; believed[0.041]; Mahdi[0.041]; bodyguards[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Muqtada\u001b[39m ==> ENTITY: \u001b[32mMuqtada al-Sadr\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitia[0.045]; Mahdi[0.044]; Sadr[0.043]; Sadr[0.043]; forces[0.043]; Shiite[0.043]; Maliki[0.043]; military[0.042]; militant[0.042]; Muqtada[0.042]; Muqtada[0.042]; Muqtada[0.042]; official[0.042]; al[0.041]; al[0.041]; Army[0.041]; Saddam[0.041]; Saddam[0.041]; cleric[0.040]; cleric[0.040]; officials[0.040]; officials[0.040]; charge[0.040]; Sunnis[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nouri al-Maliki\u001b[39m ==> ENTITY: \u001b[32mNouri al-Maliki\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMaliki[0.045]; military[0.044]; Rubaie[0.043]; al[0.043]; official[0.042]; Iraqi[0.042]; Iraqi[0.042]; adviser[0.042]; adviser[0.042]; adviser[0.042]; leader[0.042]; announced[0.041]; Jazeera[0.041]; Saddam[0.041]; Saddam[0.041]; Saddam[0.041]; Prime[0.041]; security[0.041]; forces[0.041]; minister[0.041]; national[0.040]; Al[0.040]; Al[0.040]; quoted[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.182:0.182[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraqi[0.047]; Saddam[0.046]; Saddam[0.046]; forces[0.044]; military[0.044]; military[0.044]; military[0.044]; military[0.044]; Shiite[0.042]; Army[0.042]; government[0.042]; American[0.041]; squad[0.041]; Sunnis[0.041]; al[0.040]; Maliki[0.040]; said[0.040]; said[0.040]; said[0.040]; told[0.039]; key[0.039]; Mahdi[0.039]; response[0.039]; Al[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Caldwell ==> ENTITY: \u001b[32mWilliam B. Caldwell\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.165:0.165[0]; log p(e|m)= -0.807:-0.807[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArmy[0.047]; military[0.045]; military[0.045]; military[0.045]; military[0.045]; Caldwell[0.043]; Caldwell[0.043]; Iraq[0.043]; charge[0.043]; William[0.042]; forces[0.041]; Iraqi[0.041]; American[0.041]; father[0.041]; prison[0.040]; Wednesday[0.040]; transferred[0.040]; spokesman[0.040]; said[0.039]; said[0.039]; said[0.039]; key[0.039]; added[0.039]; squad[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saddam Hussein\u001b[39m ==> ENTITY: \u001b[32mSaddam Hussein\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.046]; Iraq[0.046]; Saddam[0.046]; Saddam[0.046]; Saddam[0.046]; Iraqi[0.044]; Iraqi[0.044]; Hussein[0.043]; Rubaie[0.043]; security[0.041]; security[0.041]; forces[0.040]; Al[0.040]; execution[0.040]; execution[0.040]; execution[0.040]; execution[0.040]; al[0.040]; video[0.040]; video[0.040]; Maliki[0.039]; said[0.039]; said[0.039]; witnesses[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saddam\u001b[39m ==> ENTITY: \u001b[32mSaddam Hussein\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaddam[0.049]; Saddam[0.049]; Iraqi[0.047]; Rubaie[0.045]; Rubaie[0.045]; military[0.044]; Jazeera[0.044]; security[0.043]; report[0.043]; satellite[0.042]; government[0.042]; Al[0.042]; execution[0.042]; execution[0.042]; execution[0.042]; execution[0.042]; execution[0.042]; execution[0.042]; execution[0.042]; al[0.042]; al[0.042]; al[0.042]; al[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.178:0.178[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.053]; Iraqi[0.048]; Saddam[0.048]; Saddam[0.048]; Saddam[0.048]; Hussein[0.045]; Hussein[0.045]; security[0.044]; security[0.044]; national[0.043]; national[0.043]; Rubaie[0.042]; al[0.042]; Maliki[0.042]; said[0.041]; official[0.041]; told[0.041]; Al[0.040]; execution[0.040]; execution[0.040]; execution[0.040]; execution[0.040]; says[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mowaffak al-Rubaie\u001b[39m ==> ENTITY: \u001b[32mMowaffak al-Rubaie\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.208:0.208[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRubaie[0.046]; Iraqi[0.043]; Maliki[0.043]; Maliki[0.043]; al[0.043]; al[0.043]; al[0.043]; al[0.043]; Saddam[0.041]; Saddam[0.041]; Muqtada[0.041]; security[0.041]; security[0.041]; Hussein[0.041]; Al[0.041]; Al[0.041]; video[0.041]; video[0.041]; Najaf[0.040]; adviser[0.040]; adviser[0.040]; military[0.040]; Jazeera[0.040]; national[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Internet ==> ENTITY: \u001b[32mInternet\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.140:0.140[0]; log p(e|m)= -0.037:-0.037[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nvideo[0.048]; audio[0.046]; recording[0.045]; television[0.045]; media[0.044]; phone[0.043]; mobile[0.043]; saying[0.042]; Press[0.042]; announced[0.041]; authorized[0.041]; surfaced[0.041]; States[0.041]; Wednesday[0.040]; charge[0.040]; pictures[0.040]; official[0.040]; Tuesday[0.040]; announcement[0.040]; late[0.040]; said[0.040]; said[0.040]; talk[0.039]; prime[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Muqtada\u001b[39m ==> ENTITY: \u001b[32mMuqtada al-Sadr\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitia[0.045]; Mahdi[0.044]; Sadr[0.043]; Sadr[0.043]; forces[0.043]; Shiite[0.043]; Maliki[0.043]; military[0.042]; militant[0.042]; Muqtada[0.042]; Muqtada[0.042]; Muqtada[0.042]; official[0.042]; al[0.041]; al[0.041]; Army[0.041]; Saddam[0.041]; Saddam[0.041]; cleric[0.040]; cleric[0.040]; officials[0.040]; officials[0.040]; charge[0.040]; Sunnis[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Muqtada\u001b[39m ==> ENTITY: \u001b[32mMuqtada al-Sadr\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitia[0.045]; Mahdi[0.044]; Sadr[0.043]; Sadr[0.043]; forces[0.043]; Shiite[0.043]; Maliki[0.043]; military[0.042]; militant[0.042]; Muqtada[0.042]; Muqtada[0.042]; Muqtada[0.042]; official[0.042]; al[0.041]; al[0.041]; Army[0.041]; Saddam[0.041]; Saddam[0.041]; cleric[0.040]; cleric[0.040]; officials[0.040]; officials[0.040]; charge[0.040]; Sunnis[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Al-Maliki\u001b[39m ==> ENTITY: \u001b[32mNouri al-Maliki\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.216:0.216[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMaliki[0.045]; military[0.043]; Rubaie[0.043]; Rubaie[0.043]; al[0.042]; al[0.042]; al[0.042]; official[0.042]; Iraqi[0.042]; Iraqi[0.042]; adviser[0.041]; adviser[0.041]; adviser[0.041]; leader[0.041]; announced[0.041]; Jazeera[0.041]; Saddam[0.041]; Saddam[0.041]; Saddam[0.041]; Prime[0.041]; Nouri[0.041]; security[0.041]; forces[0.041]; minister[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saddam\u001b[39m ==> ENTITY: \u001b[32mSaddam Hussein\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.048]; Saddam[0.047]; Iraqi[0.046]; military[0.043]; military[0.043]; military[0.043]; military[0.043]; forces[0.041]; government[0.041]; Al[0.041]; execution[0.041]; al[0.041]; Maliki[0.041]; targeted[0.040]; reporters[0.040]; said[0.040]; said[0.040]; said[0.040]; believed[0.040]; Mahdi[0.040]; told[0.040]; Wednesday[0.040]; Sunnis[0.040]; killed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = US military ==> ENTITY: \u001b[32mUnited States Armed Forces\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.062:-0.062[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.050]; military[0.050]; military[0.050]; Iraq[0.045]; charge[0.042]; added[0.042]; William[0.042]; government[0.041]; said[0.041]; said[0.041]; said[0.041]; forces[0.041]; key[0.041]; Iraqi[0.040]; response[0.040]; targeted[0.040]; Wednesday[0.039]; custody[0.039]; transferred[0.039]; Caldwell[0.039]; Caldwell[0.039]; Caldwell[0.039]; spokesman[0.039]; execution[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = The Associated Press\u001b[39m ==> ENTITY: \u001b[32mAssociated Press\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnational[0.045]; Times[0.044]; Wednesday[0.044]; Wednesday[0.044]; interview[0.043]; told[0.043]; reported[0.043]; video[0.042]; video[0.042]; video[0.042]; official[0.042]; accusing[0.041]; report[0.041]; taping[0.041]; said[0.041]; said[0.041]; said[0.041]; York[0.041]; military[0.041]; security[0.040]; officials[0.040]; officials[0.040]; leaked[0.040]; denied[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iraqi forces\u001b[39m ==> ENTITY: \u001b[32mIraqi Army\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraqi[0.048]; security[0.047]; announced[0.042]; Hussein[0.042]; national[0.042]; officials[0.042]; Saddam[0.042]; Saddam[0.042]; Saddam[0.042]; Saddam[0.042]; Saddam[0.042]; Maliki[0.041]; Maliki[0.041]; Rubaie[0.041]; Al[0.041]; Al[0.041]; Jazeera[0.041]; al[0.041]; al[0.041]; official[0.040]; suspected[0.040]; Tuesday[0.040]; anonymity[0.040]; authorized[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Maliki\u001b[39m ==> ENTITY: \u001b[32mNouri al-Maliki\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.221:0.221[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nal[0.054]; al[0.054]; al[0.054]; al[0.054]; al[0.054]; Maliki[0.052]; Al[0.050]; Al[0.050]; Hussein[0.050]; Rubaie[0.050]; Rubaie[0.050]; Iraqi[0.049]; adviser[0.048]; adviser[0.048]; Jazeera[0.048]; security[0.047]; security[0.047]; secretary[0.047]; believed[0.047]; Najaf[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Al-Maliki\u001b[39m ==> ENTITY: \u001b[32mNouri al-Maliki\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.048]; Iraq[0.048]; Rubaie[0.045]; al[0.045]; official[0.044]; Iraqi[0.044]; adviser[0.044]; adviser[0.044]; adviser[0.044]; Hussein[0.044]; Hussein[0.044]; Saddam[0.043]; Saddam[0.043]; Saddam[0.043]; security[0.043]; security[0.043]; national[0.042]; national[0.042]; announcement[0.042]; said[0.042]; Wednesday[0.041]; officials[0.041]; told[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.178:0.178[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.053]; Iraqi[0.048]; Saddam[0.048]; Saddam[0.048]; Saddam[0.048]; Hussein[0.045]; Hussein[0.045]; security[0.044]; security[0.044]; national[0.043]; national[0.043]; Rubaie[0.042]; al[0.042]; Maliki[0.042]; said[0.041]; official[0.041]; told[0.041]; Al[0.040]; execution[0.040]; execution[0.040]; execution[0.040]; execution[0.040]; says[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Muqtada al-Sadr\u001b[39m ==> ENTITY: \u001b[32mMuqtada al-Sadr\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitia[0.045]; Mahdi[0.044]; Iraq[0.044]; Sadr[0.043]; forces[0.043]; Shiite[0.043]; Maliki[0.043]; military[0.042]; militant[0.042]; Muqtada[0.042]; Muqtada[0.042]; Muqtada[0.042]; official[0.042]; al[0.041]; government[0.041]; Army[0.041]; Saddam[0.041]; Saddam[0.041]; cleric[0.040]; cleric[0.040]; officials[0.040]; officials[0.040]; charge[0.040]; Sunnis[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mahdi Army\u001b[39m ==> ENTITY: \u001b[32mMahdi Army\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nShiite[0.045]; forces[0.045]; Iraq[0.044]; Sadr[0.043]; Sadr[0.043]; militia[0.043]; Maliki[0.043]; Sunnis[0.042]; Muqtada[0.042]; Muqtada[0.042]; Muqtada[0.042]; Muqtada[0.042]; military[0.041]; al[0.041]; al[0.041]; militant[0.041]; official[0.040]; Saddam[0.040]; Saddam[0.040]; Saddam[0.040]; government[0.040]; cleric[0.040]; cleric[0.040]; targeted[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saddam\u001b[39m ==> ENTITY: \u001b[32mSaddam Hussein\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaddam[0.046]; Saddam[0.046]; Saddam[0.046]; Iraqi[0.045]; Rubaie[0.043]; Rubaie[0.043]; military[0.042]; Jazeera[0.042]; saying[0.041]; security[0.041]; report[0.041]; forces[0.040]; Al[0.040]; Al[0.040]; execution[0.040]; execution[0.040]; execution[0.040]; execution[0.040]; al[0.040]; al[0.040]; al[0.040]; al[0.040]; video[0.040]; video[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Caldwell ==> ENTITY: \u001b[32mWilliam B. Caldwell\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.807:-0.807[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.047]; military[0.047]; military[0.047]; Caldwell[0.045]; Iraq[0.044]; charge[0.044]; forces[0.043]; Iraqi[0.043]; prison[0.042]; transferred[0.041]; said[0.041]; said[0.041]; said[0.041]; added[0.041]; squad[0.041]; getting[0.040]; thanked[0.039]; differently[0.039]; differently[0.039]; police[0.039]; police[0.039]; execution[0.039]; interpreter[0.039]; government[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Al-Jazeera\u001b[39m ==> ENTITY: \u001b[32mAl Jazeera\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninterview[0.046]; al[0.045]; al[0.045]; al[0.045]; al[0.045]; al[0.045]; told[0.044]; told[0.044]; satellite[0.044]; video[0.044]; video[0.044]; video[0.044]; Rubaie[0.044]; Rubaie[0.044]; Press[0.043]; official[0.043]; reported[0.042]; security[0.042]; security[0.042]; Times[0.042]; Times[0.042]; Wednesday[0.042]; Wednesday[0.042]; \t\n\r [==============================>...............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 812ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 446/656 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 16444287 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.057:0.057[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncase[0.045]; hoping[0.043]; days[0.043]; days[0.043]; held[0.043]; wanted[0.043]; group[0.043]; Thursday[0.042]; called[0.042]; away[0.042]; surprise[0.042]; sent[0.042]; German[0.041]; home[0.041]; descent[0.041]; descent[0.041]; Vegas[0.041]; Vegas[0.041]; Vegas[0.041]; civil[0.041]; authorities[0.040]; turned[0.040]; Muslim[0.040]; Muslim[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Phoenix ==> ENTITY: \u001b[32mPhoenix, Arizona\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.085:0.085[0]; log p(e|m)= -0.909:-0.909[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirport[0.044]; Airport[0.044]; area[0.043]; South[0.042]; flight[0.042]; flight[0.042]; Airlines[0.042]; reported[0.042]; McCarran[0.042]; Thursday[0.042]; Protection[0.042]; according[0.042]; month[0.041]; Tuesday[0.041]; Airways[0.041]; American[0.041]; American[0.041]; October[0.041]; involving[0.041]; families[0.041]; terrorism[0.040]; war[0.040]; family[0.040]; civil[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = FBI\u001b[39m ==> ENTITY: \u001b[32mFederal Bureau of Investigation\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.128:0.128[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFBI[0.053]; agents[0.044]; Shehadeh[0.044]; Shehadeh[0.044]; Shehadeh[0.044]; officials[0.042]; asked[0.041]; questioned[0.041]; handcuffed[0.041]; Officials[0.041]; Officials[0.041]; passport[0.041]; jail[0.041]; Tuesday[0.041]; family[0.041]; family[0.041]; declined[0.040]; interview[0.040]; fake[0.040]; denied[0.040]; denied[0.040]; members[0.040]; Customs[0.040]; home[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = California ==> ENTITY: \u001b[32mCalifornia\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.100:0.100[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCalifornia[0.049]; country[0.046]; American[0.043]; American[0.043]; case[0.042]; case[0.042]; civil[0.042]; Calif[0.042]; Tuesday[0.042]; October[0.042]; Bakersfield[0.041]; enter[0.041]; Council[0.041]; Thursday[0.041]; according[0.041]; record[0.041]; discrimination[0.041]; similar[0.041]; criminal[0.040]; waiver[0.040]; entry[0.040]; involving[0.040]; home[0.040]; series[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.077:0.077[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.046]; American[0.046]; California[0.044]; California[0.044]; case[0.043]; news[0.043]; enter[0.042]; Tuesday[0.041]; days[0.041]; days[0.041]; held[0.041]; couldn[0.041]; group[0.041]; home[0.041]; civil[0.041]; Thursday[0.041]; entry[0.041]; denied[0.041]; hoping[0.041]; meet[0.040]; called[0.040]; sent[0.040]; away[0.040]; Instead[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew Year\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew Year's Day\u001b[39m <---> \u001b[32mNew Year\u001b[39m\t\nSCORES: global= 0.249:0.247[\u001b[31m0.001\u001b[39m]; local()= 0.122:0.155[\u001b[32m0.033\u001b[39m]; log p(e|m)= -2.263:-0.082[\u001b[32m2.181\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArabic[0.050]; celebrating[0.050]; Thursday[0.049]; Tuesday[0.049]; Iraq[0.049]; anniversary[0.049]; visit[0.048]; families[0.048]; message[0.048]; send[0.048]; Muslims[0.048]; planned[0.047]; California[0.047]; wedding[0.047]; interview[0.047]; hours[0.047]; phone[0.046]; couple[0.046]; Central[0.046]; American[0.046]; second[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Calif.\u001b[39m ==> ENTITY: \u001b[32mCalifornia\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCalifornia[0.052]; California[0.052]; American[0.045]; Bakersfield[0.044]; Thursday[0.042]; home[0.042]; Frankfurt[0.042]; flew[0.041]; descent[0.041]; descent[0.041]; group[0.041]; family[0.041]; born[0.041]; meet[0.040]; surprise[0.040]; treatment[0.040]; away[0.040]; visit[0.039]; authorities[0.039]; days[0.039]; days[0.039]; drive[0.039]; holiday[0.039]; wanted[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = San Francisco International Airport\u001b[39m ==> ENTITY: \u001b[32mSan Francisco International Airport\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirport[0.047]; series[0.045]; October[0.044]; flight[0.044]; flight[0.044]; Airlines[0.042]; Thursday[0.042]; involving[0.042]; McCarran[0.042]; Tuesday[0.042]; month[0.042]; International[0.041]; Phoenix[0.041]; passenger[0.041]; follows[0.041]; planned[0.041]; passengers[0.040]; second[0.040]; visa[0.040]; afternoon[0.040]; enter[0.040]; waiting[0.040]; Airways[0.040]; coordinator[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Condor Airlines\u001b[39m ==> ENTITY: \u001b[32mCondor Flugdienst\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirport[0.045]; Airport[0.045]; flight[0.044]; flight[0.044]; passenger[0.043]; International[0.043]; International[0.043]; Airways[0.043]; McCarran[0.042]; Alzenau[0.041]; October[0.041]; month[0.041]; coordinator[0.041]; Central[0.040]; planned[0.040]; class[0.040]; direct[0.040]; countries[0.040]; American[0.040]; American[0.040]; stamps[0.040]; Thursday[0.040]; German[0.040]; area[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Council on American-Islamic Relations\u001b[39m ==> ENTITY: \u001b[32mCouncil on American–Islamic Relations\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCAIR[0.050]; Islamic[0.044]; Muslim[0.044]; according[0.043]; terrorism[0.043]; Muslims[0.043]; said[0.042]; said[0.042]; said[0.042]; spokeswoman[0.041]; International[0.041]; rights[0.041]; denied[0.041]; denied[0.041]; denied[0.041]; reported[0.041]; send[0.040]; Shehadeh[0.040]; Shehadeh[0.040]; Shehadeh[0.040]; visa[0.040]; issue[0.040]; criticize[0.040]; told[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Las Vegas ==> ENTITY: \u001b[32mLas Vegas\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.146:0.146[0]; log p(e|m)= -0.496:-0.496[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVegas[0.049]; Vegas[0.049]; held[0.043]; Tuesday[0.043]; Las[0.043]; Las[0.043]; California[0.042]; California[0.042]; country[0.042]; Thursday[0.042]; Bakersfield[0.041]; American[0.041]; descent[0.041]; descent[0.041]; gotten[0.041]; home[0.040]; drive[0.040]; hoping[0.040]; Calif[0.040]; wanted[0.040]; wife[0.040]; meet[0.039]; holding[0.039]; holding[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = North Las Vegas\u001b[39m ==> ENTITY: \u001b[32mNorth Las Vegas, Nevada\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.106:0.106[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTuesday[0.048]; Las[0.048]; office[0.048]; Vegas[0.048]; officials[0.047]; Officials[0.047]; Officials[0.047]; airport[0.047]; airport[0.047]; home[0.046]; visa[0.045]; Alzenau[0.045]; small[0.045]; hours[0.044]; Customs[0.044]; police[0.044]; additional[0.044]; jail[0.043]; Immigration[0.043]; transported[0.043]; FBI[0.043]; FBI[0.043]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSyrian\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSyria\u001b[39m <---> \u001b[32mSyrian people\u001b[39m\t\nSCORES: global= 0.261:0.250[\u001b[31m0.011\u001b[39m]; local()= 0.162:0.169[\u001b[32m0.007\u001b[39m]; log p(e|m)= -0.609:-1.501[\u001b[31m0.892\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyrian[0.050]; American[0.046]; born[0.046]; Muslim[0.045]; Muslim[0.045]; Shehadeh[0.043]; Shehadeh[0.043]; Majed[0.043]; sent[0.042]; descent[0.041]; descent[0.041]; family[0.041]; civil[0.041]; detained[0.040]; group[0.040]; hoping[0.040]; holding[0.040]; Thursday[0.040]; days[0.039]; days[0.039]; home[0.039]; visit[0.039]; German[0.039]; discrimination[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.S. war\u001b[39m ==> ENTITY: \u001b[32mUnited States military casualties of war\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \naccording[0.046]; October[0.045]; reported[0.045]; Iraq[0.045]; American[0.043]; American[0.043]; month[0.042]; South[0.042]; civil[0.042]; Overall[0.042]; terrorism[0.042]; Thursday[0.041]; cases[0.041]; incidents[0.041]; California[0.041]; visa[0.040]; involving[0.040]; coordinator[0.040]; Central[0.040]; direct[0.040]; Africa[0.040]; International[0.040]; International[0.040]; Muslim[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.063:0.063[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWashington[0.047]; North[0.047]; declined[0.044]; additional[0.044]; car[0.044]; paperwork[0.044]; list[0.044]; waiver[0.043]; officials[0.043]; ticket[0.043]; ticket[0.043]; Tuesday[0.043]; Tuesday[0.043]; office[0.043]; entry[0.043]; right[0.043]; hours[0.043]; Officials[0.042]; Officials[0.042]; German[0.042]; confirmed[0.042]; home[0.042]; got[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSyrian\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSyria\u001b[39m <---> \u001b[32mSyrian people\u001b[39m\t\nSCORES: global= 0.261:0.250[\u001b[31m0.011\u001b[39m]; local()= 0.162:0.171[\u001b[32m0.008\u001b[39m]; log p(e|m)= -0.609:-1.501[\u001b[31m0.892\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyrian[0.050]; American[0.046]; born[0.046]; Muslim[0.045]; Muslim[0.045]; Majed[0.043]; sent[0.043]; family[0.042]; descent[0.041]; descent[0.041]; Shehadeh[0.041]; Shehadeh[0.041]; civil[0.041]; detained[0.041]; group[0.040]; hoping[0.040]; holding[0.040]; Thursday[0.040]; home[0.039]; visit[0.039]; German[0.039]; discrimination[0.039]; businessman[0.039]; businessman[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Central California\u001b[39m ==> ENTITY: \u001b[32mCentral California\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.080:0.080[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \narea[0.048]; civil[0.044]; cases[0.044]; McCarran[0.043]; Las[0.042]; Airport[0.042]; airport[0.042]; families[0.042]; coordinator[0.041]; Alzenau[0.041]; American[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; questioned[0.041]; Airlines[0.041]; Washington[0.041]; kept[0.041]; Tuesday[0.040]; second[0.040]; planned[0.040]; asked[0.040]; Thursday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = US Airways\u001b[39m ==> ENTITY: \u001b[32mUS Airways\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirport[0.050]; Airport[0.050]; Airlines[0.050]; American[0.044]; American[0.044]; flight[0.044]; flight[0.044]; International[0.044]; International[0.044]; Phoenix[0.044]; area[0.043]; Thursday[0.043]; Tuesday[0.043]; McCarran[0.042]; October[0.042]; month[0.041]; spokeswoman[0.041]; planned[0.041]; passengers[0.041]; visa[0.041]; passenger[0.041]; second[0.041]; coordinator[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = California ==> ENTITY: \u001b[32mCalifornia\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.080:0.080[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCalifornia[0.050]; country[0.046]; American[0.043]; case[0.043]; case[0.043]; civil[0.043]; Calif[0.042]; Tuesday[0.042]; Bakersfield[0.042]; enter[0.041]; Thursday[0.041]; record[0.041]; discrimination[0.041]; similar[0.041]; criminal[0.041]; waiver[0.041]; entry[0.041]; involving[0.040]; home[0.040]; series[0.040]; denied[0.040]; denied[0.040]; issue[0.040]; rights[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Las Vegas ==> ENTITY: \u001b[32mLas Vegas\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.141:0.141[0]; log p(e|m)= -0.496:-0.496[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVegas[0.049]; Vegas[0.049]; held[0.043]; Las[0.043]; Las[0.043]; California[0.042]; California[0.042]; country[0.042]; Thursday[0.042]; Bakersfield[0.042]; American[0.041]; descent[0.041]; descent[0.041]; gotten[0.041]; home[0.041]; drive[0.040]; hoping[0.040]; Calif[0.040]; wanted[0.040]; wife[0.040]; meet[0.040]; holding[0.040]; holding[0.040]; bar[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = McCarran International Airport\u001b[39m ==> ENTITY: \u001b[32mMcCarran International Airport\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirport[0.047]; flight[0.045]; flight[0.045]; area[0.044]; Phoenix[0.044]; Vegas[0.043]; Airlines[0.042]; Thursday[0.042]; October[0.042]; Las[0.042]; Tuesday[0.042]; planned[0.041]; South[0.041]; California[0.041]; Airways[0.041]; month[0.041]; Central[0.040]; American[0.040]; Minneapolis[0.040]; entry[0.040]; International[0.040]; coordinator[0.040]; Alzenau[0.040]; home[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIslamic scholar\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIslamic studies\u001b[39m <---> \u001b[32mUlama\u001b[39m\t\nSCORES: global= 0.282:0.275[\u001b[31m0.007\u001b[39m]; local()= 0.184:0.188[\u001b[32m0.004\u001b[39m]; log p(e|m)= 0.000:-0.007[\u001b[31m0.007\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIslamic[0.048]; imams[0.046]; Muslim[0.045]; Muslims[0.043]; according[0.042]; American[0.042]; Shehadeh[0.042]; Shehadeh[0.042]; Shehadeh[0.042]; Shaikh[0.042]; class[0.041]; country[0.041]; said[0.041]; said[0.041]; said[0.041]; Council[0.041]; month[0.040]; send[0.040]; exam[0.040]; civil[0.040]; Iraq[0.040]; visa[0.040]; South[0.040]; Relations[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.S. Immigration and Customs Enforcement\u001b[39m ==> ENTITY: \u001b[32mU.S. Immigration and Customs Enforcement\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFBI[0.046]; FBI[0.046]; agents[0.046]; officials[0.044]; office[0.044]; visa[0.042]; entry[0.042]; Border[0.042]; additional[0.042]; Officials[0.042]; Officials[0.042]; Protection[0.041]; Alzenau[0.041]; military[0.041]; questioned[0.041]; paperwork[0.041]; Tuesday[0.041]; grants[0.040]; North[0.040]; declined[0.040]; enter[0.040]; waiver[0.040]; police[0.040]; list[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = FBI\u001b[39m ==> ENTITY: \u001b[32mFederal Bureau of Investigation\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFBI[0.053]; agents[0.044]; office[0.044]; paperwork[0.042]; denied[0.042]; denied[0.042]; aide[0.041]; officials[0.041]; contacted[0.041]; handcuffed[0.041]; jail[0.041]; Tuesday[0.041]; family[0.041]; family[0.041]; family[0.041]; list[0.041]; declined[0.041]; lot[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.160:0.160[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwar[0.046]; Islamic[0.044]; Islamic[0.044]; American[0.044]; American[0.044]; Muslim[0.043]; terrorism[0.042]; second[0.042]; Shehadeh[0.042]; Shehadeh[0.042]; civil[0.041]; Muslims[0.041]; imams[0.041]; send[0.041]; gave[0.041]; month[0.041]; according[0.040]; Council[0.040]; International[0.040]; International[0.040]; October[0.040]; Thursday[0.040]; reported[0.040]; involving[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.S. Customs and Border Protection\u001b[39m ==> ENTITY: \u001b[32mU.S. Customs and Border Protection\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.132:0.132[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirport[0.046]; authorities[0.045]; entry[0.045]; entry[0.045]; passenger[0.045]; International[0.044]; flight[0.044]; American[0.044]; American[0.044]; passengers[0.044]; visa[0.044]; Frankfurt[0.043]; Airways[0.043]; Phoenix[0.043]; according[0.043]; month[0.043]; detention[0.042]; criminal[0.042]; involving[0.042]; spokeswoman[0.042]; waiver[0.042]; reported[0.042]; terrorism[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Frankfurt ==> ENTITY: \u001b[32mFrankfurt\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.070:-0.070[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGerman[0.045]; country[0.044]; Thursday[0.044]; American[0.043]; Tuesday[0.043]; drive[0.042]; held[0.042]; home[0.041]; meet[0.041]; said[0.041]; said[0.041]; entry[0.041]; civil[0.041]; authorities[0.041]; flew[0.041]; record[0.041]; visit[0.041]; born[0.041]; days[0.041]; days[0.041]; enter[0.041]; confirmed[0.041]; spokeswoman[0.041]; family[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alzenau\u001b[39m ==> ENTITY: \u001b[32mAlzenau\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.003:-0.003[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nvillage[0.045]; area[0.045]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; additional[0.042]; German[0.042]; German[0.042]; Central[0.042]; gave[0.042]; visa[0.042]; Bavarian[0.041]; visit[0.041]; visited[0.041]; hours[0.041]; small[0.041]; North[0.041]; citizens[0.040]; countries[0.040]; paperwork[0.040]; planned[0.040]; stamps[0.040]; car[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Muslim ==> ENTITY: \u001b[32mMuslim\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMuslim[0.052]; Majed[0.045]; Shehadeh[0.044]; Shehadeh[0.044]; civil[0.043]; family[0.042]; Syrian[0.042]; Syrian[0.042]; descent[0.042]; descent[0.042]; Las[0.041]; Las[0.041]; Las[0.041]; visit[0.041]; group[0.040]; sent[0.040]; authorities[0.040]; days[0.040]; days[0.040]; wanted[0.040]; businessman[0.040]; businessman[0.040]; home[0.039]; held[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Las Vegas ==> ENTITY: \u001b[32mLas Vegas\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.142:0.142[0]; log p(e|m)= -0.496:-0.496[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVegas[0.049]; Vegas[0.049]; held[0.043]; Las[0.043]; Las[0.043]; California[0.043]; California[0.043]; Thursday[0.042]; Bakersfield[0.042]; American[0.041]; descent[0.041]; descent[0.041]; gotten[0.041]; home[0.041]; drive[0.041]; hoping[0.040]; Calif[0.040]; wanted[0.040]; wife[0.040]; meet[0.040]; holding[0.040]; family[0.039]; days[0.039]; days[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Minneapolis ==> ENTITY: \u001b[32mMinneapolis\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.013:-0.013[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirport[0.046]; Airport[0.046]; Airlines[0.044]; flight[0.043]; flight[0.043]; Council[0.043]; Thursday[0.042]; passenger[0.042]; Tuesday[0.042]; South[0.042]; Airways[0.042]; passengers[0.041]; families[0.041]; Phoenix[0.041]; cases[0.041]; area[0.041]; McCarran[0.040]; citizens[0.040]; according[0.040]; October[0.040]; involving[0.039]; American[0.039]; American[0.039]; second[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Las Vegas ==> ENTITY: \u001b[32mLas Vegas\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.496:-0.496[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \narea[0.046]; Airport[0.046]; McCarran[0.045]; flight[0.044]; flight[0.044]; Central[0.043]; Phoenix[0.043]; California[0.042]; Airlines[0.042]; Washington[0.041]; Thursday[0.041]; month[0.041]; kept[0.041]; wedding[0.041]; Tuesday[0.041]; planned[0.041]; American[0.041]; families[0.041]; Airways[0.040]; home[0.040]; International[0.040]; couple[0.039]; Minneapolis[0.039]; wife[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Africa ==> ENTITY: \u001b[32mSouth Africa\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.263:-0.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.045]; series[0.045]; case[0.043]; second[0.043]; rights[0.042]; Tuesday[0.042]; denied[0.042]; denied[0.042]; denied[0.042]; involving[0.042]; civil[0.042]; month[0.042]; American[0.041]; American[0.041]; discuss[0.041]; Thursday[0.041]; International[0.041]; International[0.041]; incidents[0.041]; October[0.041]; similar[0.040]; Relations[0.040]; record[0.040]; issue[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bakersfield ==> ENTITY: \u001b[32mBakersfield, California\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.095:0.095[0]; log p(e|m)= -0.076:-0.076[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCalifornia[0.044]; California[0.044]; country[0.044]; American[0.043]; discrimination[0.042]; Tuesday[0.042]; wanted[0.042]; Thursday[0.042]; home[0.042]; descent[0.042]; descent[0.042]; called[0.041]; record[0.041]; denied[0.041]; denied[0.041]; group[0.041]; similar[0.041]; Calif[0.041]; flew[0.041]; involving[0.041]; gotten[0.041]; Muslim[0.041]; Muslim[0.041]; away[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Las Vegas ==> ENTITY: \u001b[32mLas Vegas\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.496:-0.496[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVegas[0.050]; Las[0.050]; police[0.042]; North[0.042]; Washington[0.042]; Tuesday[0.042]; family[0.042]; family[0.042]; family[0.042]; travel[0.041]; travel[0.041]; list[0.041]; wife[0.041]; wouldn[0.041]; airport[0.041]; airport[0.041]; officials[0.041]; got[0.040]; office[0.040]; citizens[0.040]; Feinstein[0.040]; Feinstein[0.040]; additional[0.040]; entry[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = CAIR\u001b[39m ==> ENTITY: \u001b[32mCouncil on American–Islamic Relations\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIslamic[0.044]; Islamic[0.044]; Muslim[0.044]; Council[0.043]; according[0.043]; area[0.043]; American[0.043]; American[0.043]; Muslims[0.043]; said[0.042]; said[0.042]; October[0.042]; International[0.041]; International[0.041]; rights[0.041]; denied[0.041]; Relations[0.041]; reported[0.041]; send[0.040]; asked[0.040]; Shehadeh[0.040]; Shehadeh[0.040]; direct[0.040]; criticize[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Fresno ==> ENTITY: \u001b[32mFresno, California\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.100:0.100[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCalifornia[0.045]; California[0.045]; Bakersfield[0.044]; home[0.043]; sent[0.043]; Calif[0.042]; wanted[0.042]; called[0.042]; American[0.042]; descent[0.042]; descent[0.042]; Thursday[0.042]; Las[0.041]; Las[0.041]; Las[0.041]; flew[0.040]; drive[0.040]; family[0.040]; turned[0.040]; days[0.040]; days[0.040]; visit[0.040]; Muslim[0.040]; Muslim[0.040]; \t\n\r [================================>.............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 664ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 484/656 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 16384904 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mayo Clinic\u001b[39m ==> ENTITY: \u001b[32mMayo Clinic\u001b[39m\t\nSCORES: global= 0.234:0.234[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinic[0.045]; cancer[0.044]; study[0.044]; study[0.044]; study[0.044]; health[0.044]; liver[0.043]; Mayo[0.043]; limit[0.042]; heart[0.042]; says[0.042]; family[0.041]; recent[0.041]; annual[0.041]; day[0.041]; day[0.041]; disease[0.041]; karaoke[0.040]; pregnancy[0.040]; goes[0.040]; alcohol[0.040]; alcohol[0.040]; alcohol[0.040]; cirrhosis[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Stony Brook University\u001b[39m ==> ENTITY: \u001b[32mStony Brook University\u001b[39m\t\nSCORES: global= 0.221:0.221[0]; local()= 0.057:0.057[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUniversity[0.050]; Clinic[0.045]; colorectal[0.044]; colorectal[0.044]; Goteborg[0.043]; York[0.043]; conducted[0.043]; study[0.042]; study[0.042]; study[0.042]; doctors[0.042]; doctors[0.042]; week[0.041]; week[0.041]; gain[0.041]; women[0.041]; women[0.041]; Mayo[0.040]; liver[0.040]; liver[0.040]; disease[0.039]; day[0.039]; cut[0.039]; planning[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = University of Missouri\u001b[39m ==> ENTITY: \u001b[32mUniversity of Missouri\u001b[39m\t\nSCORES: global= 0.237:0.237[0]; local()= 0.090:0.090[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUniversity[0.050]; academic[0.045]; State[0.044]; medical[0.043]; general[0.043]; time[0.043]; Ohio[0.042]; esophagus[0.042]; high[0.041]; started[0.041]; began[0.041]; breast[0.041]; improved[0.041]; increase[0.040]; health[0.040]; avoid[0.040]; especially[0.040]; hosted[0.040]; boost[0.040]; swizzle[0.040]; longer[0.040]; longer[0.040]; liver[0.040]; liver[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.049:0.049[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOhio[0.044]; University[0.043]; University[0.043]; University[0.043]; lower[0.043]; lower[0.043]; low[0.043]; week[0.043]; Live[0.042]; day[0.042]; began[0.042]; water[0.041]; family[0.041]; significantly[0.041]; success[0.041]; started[0.041]; daily[0.041]; increases[0.040]; drinking[0.040]; drinking[0.040]; non[0.040]; longer[0.040]; people[0.040]; moderately[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.083:0.083[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGoteborg[0.045]; week[0.044]; week[0.044]; workers[0.043]; order[0.042]; developing[0.042]; developing[0.042]; lower[0.042]; lower[0.042]; potential[0.042]; group[0.042]; health[0.041]; started[0.041]; people[0.041]; gain[0.041]; medical[0.041]; mention[0.041]; company[0.041]; hosted[0.041]; family[0.040]; cut[0.040]; like[0.040]; weight[0.040]; believe[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hodgkin's lymphoma\u001b[39m ==> ENTITY: \u001b[32mHodgkin's lymphoma\u001b[39m\t\nSCORES: global= 0.211:0.211[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncancer[0.047]; cancer[0.047]; doctors[0.042]; Goteborg[0.042]; began[0.042]; brain[0.042]; arthritis[0.042]; heart[0.042]; University[0.042]; University[0.042]; University[0.042]; medical[0.042]; disease[0.042]; liver[0.042]; academic[0.041]; State[0.041]; started[0.041]; day[0.040]; Missouri[0.040]; dementia[0.040]; esophagus[0.040]; showed[0.040]; showed[0.040]; researchers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mayo Clinic\u001b[39m ==> ENTITY: \u001b[32mMayo Clinic\u001b[39m\t\nSCORES: global= 0.235:0.235[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndoctors[0.047]; Clinic[0.044]; colorectal[0.043]; cancer[0.043]; cancer[0.043]; study[0.043]; study[0.043]; study[0.043]; study[0.043]; University[0.043]; health[0.043]; liver[0.042]; Mayo[0.041]; limit[0.041]; week[0.041]; heart[0.041]; says[0.041]; recent[0.040]; day[0.040]; day[0.040]; disease[0.040]; Brook[0.039]; pregnancy[0.039]; drinking[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Goteborg University\u001b[39m ==> ENTITY: \u001b[32mUniversity of Gothenburg\u001b[39m\t\nSCORES: global= 0.219:0.219[0]; local()= 0.098:0.098[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.047]; University[0.047]; colorectal[0.045]; colorectal[0.045]; medical[0.043]; researchers[0.043]; study[0.042]; studies[0.042]; doctors[0.042]; doctors[0.042]; esophagus[0.041]; conducted[0.041]; rheumatoid[0.041]; week[0.040]; week[0.040]; journals[0.040]; swizzle[0.040]; liver[0.040]; health[0.040]; opportunities[0.040]; group[0.040]; began[0.039]; disease[0.039]; arthritis[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.044:0.044[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsays[0.046]; week[0.043]; week[0.043]; read[0.043]; mention[0.043]; chilling[0.042]; low[0.042]; skinny[0.042]; planning[0.042]; pregnancy[0.041]; kissing[0.041]; University[0.041]; University[0.041]; day[0.041]; Sweden[0.041]; group[0.041]; gain[0.041]; potential[0.041]; finding[0.040]; family[0.040]; OK[0.040]; heart[0.040]; believe[0.040]; category[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ohio State University\u001b[39m ==> ENTITY: \u001b[32mOhio State University\u001b[39m\t\nSCORES: global= 0.234:0.234[0]; local()= 0.080:0.080[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUniversity[0.053]; academic[0.048]; Missouri[0.044]; medical[0.044]; lower[0.044]; lower[0.044]; online[0.044]; high[0.044]; levels[0.043]; time[0.043]; boost[0.043]; esophagus[0.043]; laboratory[0.043]; day[0.042]; rheumatoid[0.042]; began[0.042]; increase[0.042]; breast[0.042]; stack[0.042]; started[0.042]; brain[0.042]; studies[0.042]; non[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Santa Baby\u001b[39m ==> ENTITY: \u001b[32mSanta Baby\u001b[39m\t\nSCORES: global= 0.216:0.216[0]; local()= 0.033:0.033[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinic[0.046]; version[0.044]; medical[0.043]; drive[0.043]; gift[0.043]; holiday[0.043]; holiday[0.043]; karaoke[0.043]; family[0.042]; annual[0.042]; list[0.041]; fruitcake[0.041]; recent[0.041]; time[0.041]; time[0.041]; cirrhosis[0.041]; writer[0.040]; liver[0.040]; health[0.040]; day[0.040]; Unfortunately[0.040]; usually[0.040]; intake[0.040]; ve[0.040]; \t\n\r [=================================>............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 664ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 495/656 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 16451635 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Beijing ==> ENTITY: \u001b[32mBeijing\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.042:-0.042[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.049]; China[0.049]; China[0.049]; China[0.049]; record[0.043]; record[0.043]; southwestern[0.042]; Wednesday[0.041]; previous[0.041]; Xinhua[0.041]; facilities[0.040]; Administration[0.040]; total[0.040]; pandas[0.040]; pandas[0.040]; pandas[0.040]; pandas[0.040]; pandas[0.040]; panda[0.039]; nearly[0.039]; number[0.039]; giant[0.039]; Cao[0.039]; mini[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sichuan\u001b[39m ==> ENTITY: \u001b[32mSichuan\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.248:0.248[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nprovince[0.046]; province[0.046]; Sichuan[0.046]; Chengdu[0.046]; southwestern[0.045]; southwestern[0.045]; China[0.045]; China[0.045]; China[0.045]; China[0.045]; Chongqing[0.044]; Shaanxi[0.043]; western[0.043]; state[0.042]; pandas[0.042]; pandas[0.042]; pandas[0.042]; pandas[0.042]; pandas[0.042]; pandas[0.042]; pandas[0.042]; Wolong[0.042]; Xinhua[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chongqing\u001b[39m ==> ENTITY: \u001b[32mChongqing\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncity[0.055]; China[0.053]; China[0.053]; China[0.053]; Chengdu[0.053]; Shaanxi[0.052]; Sichuan[0.052]; Sichuan[0.052]; Xinhua[0.051]; nearly[0.050]; province[0.050]; province[0.050]; southwestern[0.048]; southwestern[0.048]; facilities[0.048]; world[0.048]; centers[0.047]; western[0.047]; Wolong[0.047]; Cao[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.208:0.208[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.055]; China[0.055]; China[0.055]; China[0.055]; nearly[0.048]; city[0.047]; western[0.047]; southwestern[0.046]; southwestern[0.046]; record[0.046]; record[0.046]; Chengdu[0.046]; Sichuan[0.046]; Sichuan[0.046]; Xinhua[0.045]; province[0.045]; province[0.045]; Chongqing[0.045]; total[0.045]; Cao[0.045]; number[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.179:0.179[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.049]; China[0.049]; nearly[0.043]; city[0.043]; western[0.042]; southwestern[0.042]; southwestern[0.042]; record[0.042]; low[0.042]; Chengdu[0.042]; Sichuan[0.041]; Sichuan[0.041]; Xinhua[0.041]; province[0.041]; province[0.041]; Chongqing[0.041]; time[0.041]; lower[0.040]; pandas[0.040]; pandas[0.040]; pandas[0.040]; pandas[0.040]; pandas[0.040]; world[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.149:0.149[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.050]; China[0.050]; China[0.050]; Cao[0.043]; record[0.042]; record[0.042]; Xinhua[0.042]; total[0.041]; number[0.041]; pandas[0.040]; pandas[0.040]; pandas[0.040]; pandas[0.040]; pandas[0.040]; state[0.040]; panda[0.040]; previous[0.040]; centers[0.040]; Wednesday[0.040]; media[0.039]; Administration[0.039]; said[0.039]; Current[0.039]; State[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = State Forestry Administration\u001b[39m ==> ENTITY: \u001b[32mState Forestry Administration of the People's Republic of China\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.125:0.125[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.045]; China[0.045]; China[0.045]; China[0.045]; Agency[0.043]; Protection[0.043]; endangered[0.042]; Current[0.042]; Sichuan[0.042]; Center[0.041]; zoo[0.041]; Xinhua[0.041]; spokesman[0.041]; Wednesday[0.041]; state[0.041]; quoted[0.041]; province[0.040]; facilities[0.040]; said[0.040]; mini[0.040]; Chongqing[0.040]; Research[0.040]; Research[0.040]; Base[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sichuan\u001b[39m ==> ENTITY: \u001b[32mSichuan\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.226:0.226[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nprovince[0.044]; province[0.044]; Sichuan[0.044]; Chengdu[0.044]; southwestern[0.043]; southwestern[0.043]; China[0.043]; China[0.043]; China[0.043]; Chongqing[0.043]; Shaanxi[0.042]; western[0.042]; pandas[0.041]; pandas[0.041]; pandas[0.041]; pandas[0.041]; pandas[0.041]; Wolong[0.041]; panda[0.040]; panda[0.040]; time[0.040]; centers[0.039]; city[0.038]; Agency[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.053]; China[0.053]; China[0.053]; China[0.053]; nearly[0.046]; city[0.045]; western[0.045]; southwestern[0.044]; southwestern[0.044]; record[0.044]; record[0.044]; low[0.044]; Chengdu[0.044]; Sichuan[0.044]; Sichuan[0.044]; Xinhua[0.043]; province[0.043]; province[0.043]; Chongqing[0.043]; total[0.043]; Cao[0.043]; number[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.165:0.165[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.050]; China[0.050]; China[0.050]; nearly[0.043]; Cao[0.042]; southwestern[0.042]; record[0.042]; record[0.042]; Xinhua[0.041]; Sichuan[0.041]; province[0.041]; total[0.040]; number[0.040]; pandas[0.040]; pandas[0.040]; pandas[0.040]; pandas[0.040]; pandas[0.040]; state[0.040]; facilities[0.040]; panda[0.040]; previous[0.039]; centers[0.039]; Wednesday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shaanxi\u001b[39m ==> ENTITY: \u001b[32mShaanxi\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.213:0.213[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsouthwestern[0.046]; southwestern[0.046]; China[0.045]; China[0.045]; China[0.045]; Chongqing[0.044]; Sichuan[0.044]; Sichuan[0.044]; province[0.044]; province[0.044]; western[0.043]; Chengdu[0.043]; Wolong[0.040]; city[0.040]; nearly[0.039]; pandas[0.039]; pandas[0.039]; pandas[0.039]; pandas[0.039]; pandas[0.039]; low[0.039]; panda[0.039]; panda[0.039]; giant[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Xinhua News Agency\u001b[39m ==> ENTITY: \u001b[32mXinhua News Agency\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.048]; China[0.048]; China[0.048]; China[0.048]; China[0.048]; media[0.047]; quoted[0.045]; saying[0.045]; said[0.043]; State[0.042]; Wednesday[0.042]; Chengdu[0.042]; Sichuan[0.042]; Sichuan[0.042]; Chongqing[0.041]; Forestry[0.041]; Administration[0.041]; Current[0.041]; state[0.041]; Center[0.041]; zoo[0.041]; spokesman[0.041]; world[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Giant pandas\u001b[39m ==> ENTITY: \u001b[32mGiant panda\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.223:0.223[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npanda[0.047]; panda[0.047]; pandas[0.047]; pandas[0.047]; pandas[0.047]; pandas[0.047]; zoos[0.044]; zoo[0.042]; wild[0.042]; animals[0.040]; cubs[0.040]; cubs[0.040]; breeding[0.040]; China[0.040]; China[0.040]; China[0.040]; Giant[0.040]; Wolong[0.039]; Panda[0.039]; rarest[0.039]; giant[0.038]; Sichuan[0.038]; Sichuan[0.038]; living[0.038]; \t\n\r [==================================>...........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 618ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 508/656 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 16451212 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ala.\u001b[39m ==> ENTITY: \u001b[32mAlabama\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.057:0.057[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlabama[0.049]; Health[0.044]; Health[0.044]; medical[0.044]; medical[0.044]; medical[0.044]; BIRMINGHAM[0.043]; kids[0.042]; authority[0.041]; really[0.041]; health[0.041]; care[0.041]; care[0.041]; said[0.040]; financial[0.040]; time[0.040]; drop[0.040]; directed[0.040]; Rohling[0.040]; Rohling[0.040]; Rohling[0.040]; Rohling[0.040]; quit[0.040]; quit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Congress ==> ENTITY: \u001b[32mUnited States Congress\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.335:-0.335[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPennsylvania[0.045]; enacted[0.045]; state[0.044]; American[0.043]; states[0.042]; states[0.042]; participate[0.042]; Alabama[0.042]; funding[0.042]; health[0.042]; health[0.042]; health[0.042]; include[0.042]; expanded[0.041]; nearly[0.041]; money[0.041]; individual[0.041]; provide[0.041]; private[0.041]; companies[0.040]; program[0.040]; program[0.040]; program[0.040]; program[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.053:0.053[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntold[0.044]; family[0.043]; percent[0.043]; private[0.043]; private[0.043]; free[0.043]; example[0.042]; medical[0.042]; asthma[0.042]; said[0.042]; dollars[0.042]; jobs[0.041]; working[0.041]; thousands[0.041]; months[0.041]; doctor[0.041]; support[0.041]; epilepsy[0.041]; epilepsy[0.041]; mentally[0.041]; homes[0.041]; problem[0.041]; times[0.040]; received[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.087:0.087[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstate[0.049]; nation[0.047]; States[0.046]; State[0.046]; said[0.046]; said[0.046]; uninsured[0.045]; uninsured[0.045]; uninsured[0.045]; uninsured[0.045]; uninsured[0.045]; debates[0.045]; United[0.045]; SCHIP[0.045]; private[0.045]; children[0.045]; children[0.045]; children[0.045]; children[0.045]; Health[0.045]; order[0.044]; based[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Medicaid\u001b[39m ==> ENTITY: \u001b[32mMedicaid\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.273:0.273[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMedicaid[0.046]; uninsured[0.043]; uninsured[0.043]; uninsured[0.043]; health[0.042]; health[0.042]; health[0.042]; health[0.042]; health[0.042]; health[0.042]; insurance[0.042]; insurance[0.042]; insurance[0.042]; insurance[0.042]; Health[0.042]; medical[0.042]; subsidized[0.041]; care[0.040]; care[0.040]; care[0.040]; care[0.040]; Uninsured[0.040]; private[0.040]; employer[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Highmark Blue Cross Blue Shield\u001b[39m ==> ENTITY: \u001b[32mHighmark\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHighmark[0.047]; Pennsylvania[0.044]; insurance[0.043]; insurance[0.043]; insurance[0.043]; insurance[0.043]; Nursing[0.043]; Pittsburgh[0.042]; University[0.042]; health[0.042]; health[0.042]; health[0.042]; health[0.042]; Western[0.041]; state[0.041]; state[0.041]; recent[0.040]; expanded[0.040]; companies[0.040]; states[0.040]; School[0.040]; include[0.040]; Kids[0.040]; screenings[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = University of Alabama at Birmingham\u001b[39m ==> ENTITY: \u001b[32mUniversity of Alabama at Birmingham\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.102:0.102[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNursing[0.044]; recruit[0.044]; program[0.043]; program[0.043]; School[0.042]; special[0.042]; came[0.042]; money[0.042]; Highmark[0.042]; began[0.042]; doctor[0.042]; helped[0.041]; Western[0.041]; student[0.041]; screenings[0.041]; presentation[0.041]; needs[0.041]; American[0.041]; said[0.041]; ear[0.041]; nurse[0.041]; help[0.041]; job[0.041]; schools[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = State Children's Health Insurance Program\u001b[39m ==> ENTITY: \u001b[32mState Children's Health Insurance Program\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.255:0.255[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSCHIP[0.045]; uninsured[0.045]; uninsured[0.045]; uninsured[0.045]; uninsured[0.045]; uninsured[0.045]; uninsured[0.045]; Medicaid[0.044]; insurance[0.042]; reauthorizing[0.042]; increase[0.042]; coverage[0.041]; program[0.040]; employer[0.040]; health[0.040]; health[0.040]; health[0.040]; health[0.040]; Congress[0.040]; provisions[0.039]; number[0.039]; Uninsured[0.039]; Kaiser[0.039]; million[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.068:0.068[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmillion[0.046]; order[0.046]; number[0.046]; paid[0.044]; Washington[0.044]; States[0.043]; nation[0.043]; Commission[0.042]; based[0.042]; example[0.042]; provisions[0.041]; financial[0.041]; set[0.041]; increase[0.041]; Blue[0.041]; Blue[0.041]; risen[0.040]; patchwork[0.040]; private[0.040]; provider[0.040]; United[0.039]; Alabama[0.039]; Jennifer[0.039]; getting[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = SCHIP\u001b[39m ==> ENTITY: \u001b[32mState Children's Health Insurance Program\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.230:0.230[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMedicaid[0.045]; insurance[0.043]; insurance[0.043]; insurance[0.043]; insurance[0.043]; insurance[0.043]; insurance[0.043]; expanded[0.042]; income[0.042]; enacted[0.042]; states[0.042]; states[0.042]; program[0.041]; program[0.041]; program[0.041]; program[0.041]; funding[0.041]; health[0.041]; health[0.041]; health[0.041]; insured[0.041]; Congress[0.040]; families[0.040]; private[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. government ==> ENTITY: \u001b[32mFederal government of the United States\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.118:0.118[0]; log p(e|m)= -0.022:-0.022[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfinancial[0.045]; Health[0.043]; Health[0.043]; housing[0.043]; medical[0.043]; medical[0.043]; medical[0.043]; free[0.042]; job[0.042]; official[0.042]; authority[0.041]; jobs[0.041]; kids[0.041]; said[0.041]; health[0.041]; health[0.041]; health[0.041]; incomes[0.040]; help[0.040]; help[0.040]; help[0.040]; help[0.040]; quit[0.040]; quit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Western Pennsylvania\u001b[39m ==> ENTITY: \u001b[32mWestern Pennsylvania\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUniversity[0.049]; Pittsburgh[0.046]; state[0.043]; state[0.043]; American[0.043]; states[0.042]; Alabama[0.042]; Alabama[0.042]; School[0.042]; schools[0.041]; level[0.041]; program[0.041]; program[0.041]; program[0.041]; program[0.041]; Highmark[0.041]; Highmark[0.041]; began[0.041]; Birmingham[0.040]; said[0.040]; companies[0.040]; student[0.040]; expanded[0.040]; recent[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Alabama ==> ENTITY: \u001b[32mAlabama\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.068:0.068[0]; log p(e|m)= -0.324:-0.324[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJody[0.047]; Washington[0.043]; state[0.043]; state[0.043]; State[0.043]; work[0.043]; based[0.042]; program[0.042]; miss[0.042]; Dana[0.042]; Dana[0.042]; said[0.041]; address[0.041]; Kyle[0.041]; walked[0.041]; programs[0.041]; bring[0.041]; family[0.041]; foster[0.040]; Child[0.040]; leaving[0.040]; sign[0.040]; Tolbert[0.040]; example[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.080:0.080[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnation[0.047]; Washington[0.044]; number[0.044]; said[0.043]; said[0.043]; said[0.043]; think[0.042]; based[0.042]; example[0.042]; million[0.042]; state[0.041]; set[0.041]; order[0.041]; getting[0.041]; paid[0.041]; private[0.040]; free[0.040]; really[0.040]; increase[0.040]; risen[0.040]; greater[0.040]; State[0.040]; Alabama[0.040]; patchwork[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = School of Nursing ==> ENTITY: \u001b[32mNursing school\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.169:0.169[0]; log p(e|m)= -1.682:-1.682[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUniversity[0.046]; nurse[0.046]; private[0.044]; private[0.044]; program[0.043]; program[0.043]; Pennsylvania[0.043]; doctor[0.042]; student[0.041]; health[0.041]; health[0.041]; health[0.041]; health[0.041]; health[0.041]; health[0.041]; Birmingham[0.041]; schools[0.041]; need[0.041]; months[0.041]; child[0.040]; support[0.040]; tours[0.040]; needs[0.040]; screenings[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Alabama ==> ENTITY: \u001b[32mAlabama\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.081:0.081[0]; log p(e|m)= -0.324:-0.324[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsettlement[0.043]; state[0.043]; nearly[0.043]; qualify[0.043]; families[0.043]; Pittsburgh[0.042]; American[0.042]; states[0.042]; states[0.042]; single[0.042]; work[0.042]; Pennsylvania[0.042]; earn[0.041]; earn[0.041]; money[0.041]; came[0.041]; Denise[0.041]; matching[0.041]; expanded[0.041]; eligibility[0.041]; times[0.041]; Low[0.041]; began[0.041]; provide[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.S. Congress\u001b[39m ==> ENTITY: \u001b[32mUnited States Congress\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.115:0.115[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWashington[0.047]; state[0.045]; States[0.044]; nation[0.043]; said[0.042]; said[0.042]; United[0.042]; Health[0.042]; Insurance[0.042]; increase[0.042]; Alabama[0.041]; Medicaid[0.041]; Commission[0.041]; State[0.041]; number[0.041]; cost[0.041]; uninsured[0.040]; uninsured[0.040]; uninsured[0.040]; uninsured[0.040]; uninsured[0.040]; uninsured[0.040]; provisions[0.040]; work[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Birmingham ==> ENTITY: \u001b[32mBirmingham, Alabama\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.073:0.073[0]; log p(e|m)= -1.814:-1.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npoverty[0.046]; poverty[0.046]; families[0.045]; Alabama[0.045]; incomes[0.044]; children[0.042]; children[0.042]; children[0.042]; said[0.042]; watched[0.041]; job[0.041]; help[0.041]; help[0.041]; help[0.041]; time[0.041]; housing[0.041]; jobs[0.041]; making[0.040]; drive[0.040]; quit[0.040]; quit[0.040]; drop[0.040]; got[0.040]; says[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Alabama ==> ENTITY: \u001b[32mAlabama\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.070:0.070[0]; log p(e|m)= -0.324:-0.324[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlabama[0.049]; BIRMINGHAM[0.043]; help[0.043]; help[0.043]; help[0.043]; help[0.043]; families[0.043]; time[0.043]; job[0.042]; got[0.041]; Rohling[0.041]; Rohling[0.041]; Rohling[0.041]; Rohling[0.041]; said[0.041]; hardship[0.041]; government[0.040]; set[0.040]; reached[0.040]; Blue[0.040]; Blue[0.040]; jobs[0.040]; quit[0.040]; quit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Highmark\u001b[39m ==> ENTITY: \u001b[32mHighmark\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHighmark[0.047]; Pennsylvania[0.044]; insurance[0.043]; insurance[0.043]; insurance[0.043]; insurance[0.043]; insurance[0.043]; Pittsburgh[0.043]; health[0.042]; health[0.042]; health[0.042]; Medicaid[0.042]; Cross[0.041]; Western[0.041]; state[0.041]; expanded[0.040]; companies[0.040]; Shield[0.040]; states[0.040]; states[0.040]; include[0.040]; Kids[0.040]; American[0.040]; Grabner[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pittsburgh ==> ENTITY: \u001b[32mPittsburgh\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.084:0.084[0]; log p(e|m)= -0.298:-0.298[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUniversity[0.049]; Pennsylvania[0.047]; American[0.042]; School[0.042]; Alabama[0.042]; Alabama[0.042]; Highmark[0.042]; Highmark[0.042]; states[0.041]; state[0.041]; state[0.041]; expanded[0.041]; Grabner[0.041]; program[0.041]; program[0.041]; program[0.041]; program[0.041]; schools[0.041]; Western[0.041]; help[0.041]; Birmingham[0.041]; began[0.040]; recruit[0.040]; Bear[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = SCHIP\u001b[39m ==> ENTITY: \u001b[32mState Children's Health Insurance Program\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.253:0.253[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nuninsured[0.045]; uninsured[0.045]; uninsured[0.045]; uninsured[0.045]; uninsured[0.045]; uninsured[0.045]; Medicaid[0.044]; insurance[0.042]; reauthorizing[0.042]; Insurance[0.042]; increase[0.042]; Health[0.042]; coverage[0.041]; program[0.040]; health[0.040]; health[0.040]; health[0.040]; health[0.040]; Congress[0.040]; provisions[0.039]; number[0.039]; Uninsured[0.039]; Kaiser[0.039]; million[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Medicaid\u001b[39m ==> ENTITY: \u001b[32mMedicaid\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhealth[0.044]; insurance[0.044]; insurance[0.044]; insurance[0.044]; insurance[0.044]; funding[0.043]; program[0.042]; program[0.042]; program[0.042]; program[0.042]; eligibility[0.042]; SCHIP[0.042]; private[0.041]; insured[0.041]; expanded[0.041]; states[0.041]; states[0.041]; qualify[0.041]; state[0.041]; enacted[0.041]; tobacco[0.040]; affordability[0.040]; families[0.039]; Highmark[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBlue Cross and Blue Shield of Alabama\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBlue Cross Blue Shield Association\u001b[39m <---> \u001b[32mBlue Cross and Blue Shield of Alabama\u001b[39m\t\nSCORES: global= 0.292:0.246[\u001b[31m0.046\u001b[39m]; local()= 0.204:0.175[\u001b[31m0.028\u001b[39m]; log p(e|m)= -0.016:-4.135[\u001b[31m4.119\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfinancial[0.043]; provider[0.043]; Health[0.043]; Alabama[0.043]; Medicaid[0.043]; insurance[0.043]; insurance[0.043]; insurance[0.043]; insurance[0.043]; medical[0.042]; medical[0.042]; private[0.041]; health[0.041]; health[0.041]; health[0.041]; health[0.041]; health[0.041]; health[0.041]; privately[0.041]; million[0.040]; care[0.040]; care[0.040]; care[0.040]; care[0.040]; \t\n\r [====================================>.........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 510ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 533/656 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 16454203 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.074:0.074[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.046]; London[0.046]; Heathrow[0.045]; Heathrow[0.045]; Heathrow[0.045]; Heathrow[0.045]; Airport[0.044]; Airport[0.044]; British[0.043]; planes[0.043]; flights[0.041]; Terminal[0.040]; Terminal[0.040]; Terminal[0.040]; airline[0.040]; airline[0.040]; airline[0.040]; said[0.040]; said[0.040]; said[0.040]; passengers[0.039]; Wednesday[0.039]; Wednesday[0.039]; late[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = BA ==> ENTITY: \u001b[32mBritish Airways\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.236:0.236[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairline[0.045]; airline[0.045]; airline[0.045]; airline[0.045]; flights[0.042]; British[0.042]; Airways[0.042]; Heathrow[0.042]; Heathrow[0.042]; Heathrow[0.042]; Heathrow[0.042]; Heathrow[0.042]; Heathrow[0.042]; season[0.041]; BA[0.041]; BA[0.041]; London[0.040]; London[0.040]; planes[0.040]; Airport[0.040]; Airport[0.040]; received[0.040]; LONDON[0.039]; passengers[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = BA ==> ENTITY: \u001b[32mBritish Airways\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.235:0.235[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairline[0.050]; airline[0.050]; airline[0.050]; flights[0.046]; British[0.046]; Airways[0.046]; Heathrow[0.046]; Heathrow[0.046]; Heathrow[0.046]; Heathrow[0.046]; Heathrow[0.046]; BA[0.045]; BA[0.045]; London[0.044]; London[0.044]; planes[0.044]; Airport[0.044]; Airport[0.044]; received[0.044]; LONDON[0.043]; passengers[0.043]; spokesman[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = BA ==> ENTITY: \u001b[32mBritish Airways\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.236:0.236[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairline[0.046]; airline[0.046]; airline[0.046]; airline[0.046]; flights[0.042]; British[0.042]; Airways[0.042]; Heathrow[0.042]; Heathrow[0.042]; Heathrow[0.042]; Heathrow[0.042]; Heathrow[0.042]; season[0.042]; BA[0.042]; BA[0.042]; BA[0.042]; London[0.041]; planes[0.040]; Airport[0.040]; passengers[0.039]; spokesman[0.039]; spokesman[0.039]; spokesman[0.039]; leading[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = London's Heathrow Airport\u001b[39m ==> ENTITY: \u001b[32mLondon Heathrow Airport\u001b[39m\t\nSCORES: global= 0.313:0.313[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHeathrow[0.046]; Heathrow[0.046]; Heathrow[0.046]; Airport[0.045]; London[0.045]; flights[0.042]; airline[0.042]; airline[0.042]; airline[0.042]; British[0.042]; freight[0.041]; Terminal[0.041]; Terminal[0.041]; Terminal[0.041]; planes[0.041]; travelers[0.040]; passengers[0.040]; luggage[0.040]; luggage[0.040]; Airways[0.039]; BA[0.039]; BA[0.039]; Wednesday[0.039]; Wednesday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Heathrow\u001b[39m ==> ENTITY: \u001b[32mLondon Heathrow Airport\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.232:0.232[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHeathrow[0.046]; Heathrow[0.046]; Heathrow[0.046]; Heathrow[0.046]; Airport[0.045]; Airport[0.045]; London[0.045]; London[0.045]; Terminal[0.045]; Terminal[0.045]; Terminal[0.045]; Terminal[0.045]; flights[0.043]; passengers[0.043]; British[0.042]; airline[0.041]; airline[0.041]; airline[0.041]; planes[0.041]; officials[0.040]; travelers[0.040]; luggage[0.040]; luggage[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Heathrow\u001b[39m ==> ENTITY: \u001b[32mLondon Heathrow Airport\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHeathrow[0.045]; Heathrow[0.045]; Heathrow[0.045]; Terminal[0.044]; Terminal[0.044]; Terminal[0.044]; Terminal[0.044]; Terminal[0.044]; Terminal[0.044]; passengers[0.042]; Terminals[0.042]; British[0.041]; airline[0.040]; airline[0.040]; airline[0.040]; planes[0.040]; planned[0.040]; baggage[0.039]; baggage[0.039]; baggage[0.039]; baggage[0.039]; Airways[0.039]; Wednesday[0.039]; day[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Heathrow\u001b[39m ==> ENTITY: \u001b[32mLondon Heathrow Airport\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.238:0.238[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHeathrow[0.046]; Heathrow[0.046]; Heathrow[0.046]; Heathrow[0.046]; Airport[0.045]; Airport[0.045]; London[0.045]; London[0.045]; Terminal[0.045]; Terminal[0.045]; Terminal[0.045]; Terminal[0.045]; Terminal[0.045]; flights[0.043]; passengers[0.043]; British[0.042]; airline[0.041]; airline[0.041]; airline[0.041]; airline[0.041]; planes[0.041]; officials[0.040]; travelers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = London's Heathrow Airport\u001b[39m ==> ENTITY: \u001b[32mLondon Heathrow Airport\u001b[39m\t\nSCORES: global= 0.313:0.313[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHeathrow[0.046]; Heathrow[0.046]; Heathrow[0.046]; Airport[0.045]; London[0.045]; flights[0.042]; airline[0.042]; airline[0.042]; airline[0.042]; British[0.041]; freight[0.041]; Terminal[0.041]; Terminal[0.041]; Terminal[0.041]; Terminal[0.041]; planes[0.041]; travelers[0.040]; passengers[0.040]; luggage[0.040]; luggage[0.040]; Airways[0.039]; BA[0.039]; BA[0.039]; BA[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Heathrow\u001b[39m ==> ENTITY: \u001b[32mLondon Heathrow Airport\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHeathrow[0.051]; Terminal[0.050]; Terminal[0.050]; Terminal[0.050]; passengers[0.047]; Terminals[0.047]; airline[0.046]; airline[0.046]; planes[0.046]; planned[0.045]; baggage[0.044]; baggage[0.044]; baggage[0.044]; baggage[0.044]; day[0.044]; freight[0.043]; bags[0.043]; bags[0.043]; bags[0.043]; million[0.043]; backlog[0.043]; deliver[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = British Airways\u001b[39m ==> ENTITY: \u001b[32mBritish Airways\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.223:0.223[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairline[0.048]; airline[0.048]; airline[0.048]; flights[0.044]; Heathrow[0.044]; Heathrow[0.044]; Heathrow[0.044]; Heathrow[0.044]; Heathrow[0.044]; BA[0.044]; BA[0.044]; BA[0.044]; London[0.043]; London[0.043]; passengers[0.043]; developed[0.042]; planes[0.042]; Airport[0.042]; Airport[0.042]; LONDON[0.042]; luggage[0.042]; luggage[0.042]; glitch[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = BA ==> ENTITY: \u001b[32mBritish Airways\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.168:0.168[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairline[0.048]; airline[0.048]; Heathrow[0.044]; Heathrow[0.044]; season[0.044]; BA[0.044]; passengers[0.041]; spokesman[0.041]; spokesman[0.041]; spokesman[0.041]; leading[0.041]; Terminal[0.041]; Terminal[0.041]; Terminal[0.041]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; baggage[0.040]; baggage[0.040]; baggage[0.040]; baggage[0.040]; working[0.039]; \t\n\r [=====================================>........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 461ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 546/656 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 16444229 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Houston ==> ENTITY: \u001b[32mHouston\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.101:0.101[0]; log p(e|m)= -0.323:-0.323[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.051]; Houston[0.051]; Texas[0.046]; Tuesday[0.044]; George[0.043]; Bownds[0.041]; charged[0.041]; charged[0.041]; airport[0.041]; Airport[0.041]; undisclosed[0.041]; locations[0.041]; weekend[0.040]; trash[0.040]; firm[0.040]; picked[0.040]; store[0.040]; store[0.040]; Rick[0.040]; Bush[0.040]; Police[0.040]; Police[0.040]; contractor[0.039]; discovered[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Houston ==> ENTITY: \u001b[32mHouston\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.104:0.104[0]; log p(e|m)= -0.323:-0.323[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.051]; Texas[0.046]; County[0.044]; Tuesday[0.044]; George[0.043]; week[0.042]; Harris[0.042]; Bownds[0.042]; charged[0.042]; charged[0.042]; HOUSTON[0.041]; undisclosed[0.041]; Police[0.040]; Police[0.040]; Daniel[0.040]; held[0.040]; weekend[0.040]; trash[0.040]; Carlos[0.040]; picked[0.040]; Rick[0.040]; Bush[0.040]; locations[0.040]; contractor[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.084:0.084[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAviation[0.046]; flights[0.045]; week[0.044]; said[0.043]; said[0.043]; said[0.043]; Tuesday[0.043]; Tuesday[0.043]; office[0.042]; airlines[0.042]; worked[0.042]; States[0.042]; message[0.041]; baggage[0.041]; returned[0.041]; night[0.041]; United[0.040]; held[0.040]; company[0.040]; company[0.040]; company[0.040]; Group[0.040]; connecting[0.039]; hired[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = George Bush Intercontinental Airport\u001b[39m ==> ENTITY: \u001b[32mGeorge Bush Intercontinental Airport\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairport[0.049]; Houston[0.046]; Houston[0.046]; Tuesday[0.043]; Texas[0.043]; firm[0.043]; Bownds[0.042]; contractor[0.042]; locations[0.042]; handling[0.041]; handling[0.041]; baggage[0.041]; baggage[0.041]; employees[0.041]; HOUSTON[0.040]; week[0.040]; undisclosed[0.040]; Capt[0.040]; luggage[0.040]; luggage[0.040]; luggage[0.040]; luggage[0.040]; luggage[0.040]; Dec[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.068:0.068[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nweek[0.044]; Ricardo[0.044]; men[0.042]; men[0.042]; Tuesday[0.042]; Tuesday[0.042]; worked[0.042]; Carlos[0.042]; Group[0.042]; held[0.042]; returned[0.042]; London[0.042]; Daniel[0.041]; night[0.041]; Osorio[0.041]; Osorio[0.041]; trying[0.041]; Perez[0.041]; Perez[0.041]; figure[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Police ==> ENTITY: \u001b[32mPolice\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.114:0.114[0]; log p(e|m)= -1.019:-1.019[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npolice[0.046]; Police[0.045]; charged[0.043]; charged[0.043]; George[0.043]; criminal[0.043]; Rick[0.042]; accused[0.042]; week[0.042]; employees[0.041]; undisclosed[0.041]; contractor[0.041]; handling[0.041]; handling[0.041]; Tuesday[0.041]; Daniel[0.041]; Airport[0.040]; said[0.040]; said[0.040]; said[0.040]; thefts[0.040]; thefts[0.040]; held[0.040]; Carlos[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Texas ==> ENTITY: \u001b[32mTexas\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.060:0.060[0]; log p(e|m)= -0.161:-0.161[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.048]; George[0.046]; appeared[0.044]; Tuesday[0.043]; HOUSTON[0.043]; locations[0.042]; picked[0.042]; Bush[0.042]; handling[0.042]; handling[0.042]; stolen[0.041]; stolen[0.041]; firm[0.041]; undisclosed[0.041]; Airport[0.041]; discovered[0.041]; charged[0.041]; charged[0.041]; Workers[0.040]; thefts[0.040]; thefts[0.040]; said[0.040]; said[0.040]; contractor[0.040]; \t\n\r [=====================================>........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 436ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 553/656 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 16417540 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mTide\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAlabama Crimson Tide football\u001b[39m <---> \u001b[32mAlabama Crimson Tide\u001b[39m\t\nSCORES: global= 0.284:0.266[\u001b[31m0.018\u001b[39m]; local()= 0.174:0.114[\u001b[31m0.060\u001b[39m]; log p(e|m)= -2.071:-2.079[\u001b[31m0.008\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlabama[0.045]; LSU[0.044]; Tuscaloosa[0.044]; Bowden[0.043]; Florida[0.043]; recruiting[0.043]; Seminoles[0.042]; coaching[0.042]; Bowl[0.042]; LaRon[0.041]; Saban[0.041]; Saban[0.041]; Saban[0.041]; Saban[0.041]; players[0.041]; players[0.041]; assistant[0.041]; university[0.041]; university[0.041]; played[0.041]; Bobby[0.041]; coach[0.040]; coach[0.040]; hiring[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Florida State ==> ENTITY: \u001b[32mFlorida State University\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.173:0.173[0]; log p(e|m)= -0.924:-0.924[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUniversity[0.044]; players[0.044]; players[0.044]; Seminoles[0.044]; Bowden[0.043]; college[0.043]; LSU[0.043]; coaching[0.042]; played[0.042]; Alabama[0.042]; coach[0.042]; coach[0.042]; academic[0.041]; academic[0.041]; academic[0.041]; academic[0.041]; Bowl[0.041]; recruiting[0.041]; defensive[0.041]; Tide[0.040]; assistant[0.040]; university[0.040]; university[0.040]; Bobby[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saban\u001b[39m ==> ENTITY: \u001b[32mNick Saban\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.045]; Saban[0.043]; Saban[0.043]; Alabama[0.043]; Alabama[0.043]; Alabama[0.043]; Alabama[0.043]; LSU[0.043]; game[0.043]; seasons[0.042]; football[0.042]; Mike[0.041]; college[0.041]; league[0.041]; divisional[0.041]; teams[0.041]; win[0.041]; win[0.041]; Miami[0.041]; Coach[0.040]; championship[0.040]; championship[0.040]; play[0.040]; play[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saban\u001b[39m ==> ENTITY: \u001b[32mNick Saban\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaban[0.044]; Saban[0.044]; Saban[0.044]; Saban[0.044]; LSU[0.044]; LSU[0.044]; Michigan[0.042]; coaching[0.042]; Florida[0.042]; Bowden[0.042]; college[0.041]; Tuscaloosa[0.041]; coach[0.041]; coach[0.041]; recruiting[0.041]; LaRon[0.041]; Bowl[0.041]; Seminoles[0.041]; assistant[0.041]; staff[0.041]; job[0.040]; hired[0.040]; players[0.040]; players[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mTuscaloosa\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUniversity of Alabama\u001b[39m <---> \u001b[32mTuscaloosa, Alabama\u001b[39m\t\nSCORES: global= 0.265:0.257[\u001b[31m0.008\u001b[39m]; local()= 0.112:0.067[\u001b[31m0.045\u001b[39m]; log p(e|m)= -2.079:-0.067[\u001b[32m2.012\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncollege[0.044]; coaching[0.043]; LSU[0.043]; LSU[0.043]; Michigan[0.042]; recruiting[0.042]; final[0.042]; played[0.042]; Bobby[0.042]; assistant[0.042]; coach[0.042]; coach[0.042]; Terry[0.042]; Terry[0.042]; Kevin[0.041]; players[0.041]; players[0.041]; according[0.041]; time[0.041]; landed[0.041]; equipment[0.041]; arrived[0.041]; crowd[0.040]; Bowl[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Miami Dolphins ==> ENTITY: \u001b[32mMiami Dolphins\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfans[0.046]; fans[0.046]; football[0.046]; championship[0.046]; coach[0.046]; coach[0.046]; won[0.044]; Thursday[0.044]; leave[0.044]; Saban[0.044]; Saban[0.044]; Saban[0.044]; Nick[0.043]; late[0.043]; past[0.043]; national[0.042]; resume[0.042]; led[0.042]; taking[0.041]; Alabama[0.041]; Alabama[0.041]; wanted[0.041]; cut[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAlabama\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAlabama Crimson Tide football\u001b[39m <---> \u001b[32mUniversity of Alabama\u001b[39m\t\nSCORES: global= 0.251:0.238[\u001b[31m0.013\u001b[39m]; local()= 0.187:0.139[\u001b[31m0.048\u001b[39m]; log p(e|m)= -3.244:-3.058[\u001b[32m0.187\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlabama[0.044]; Alabama[0.044]; team[0.044]; football[0.043]; LSU[0.043]; game[0.043]; teams[0.042]; teams[0.042]; Tide[0.042]; Tide[0.042]; Division[0.042]; college[0.042]; championship[0.042]; championship[0.042]; league[0.041]; seasons[0.041]; championships[0.041]; national[0.041]; national[0.041]; national[0.041]; national[0.041]; athletic[0.040]; Miami[0.040]; won[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saban\u001b[39m ==> ENTITY: \u001b[32mNick Saban\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.045]; Saban[0.044]; Alabama[0.044]; Alabama[0.044]; Alabama[0.044]; Alabama[0.044]; LSU[0.044]; seasons[0.043]; Mike[0.042]; league[0.041]; divisional[0.041]; teams[0.041]; teams[0.041]; win[0.041]; Miami[0.041]; John[0.041]; Coach[0.040]; championship[0.040]; faculty[0.040]; play[0.040]; hire[0.040]; national[0.040]; national[0.040]; national[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mal Moore\u001b[39m ==> ENTITY: \u001b[32mMal Moore\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nathletic[0.046]; football[0.044]; coach[0.043]; LSU[0.043]; national[0.043]; national[0.043]; national[0.043]; college[0.043]; Alabama[0.042]; Alabama[0.042]; Alabama[0.042]; divisional[0.042]; director[0.041]; championships[0.041]; tremendous[0.041]; league[0.041]; Mike[0.041]; program[0.040]; past[0.040]; game[0.040]; championship[0.040]; Saban[0.039]; Saban[0.039]; Saban[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kevin Steele\u001b[39m ==> ENTITY: \u001b[32mKevin Steele\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLSU[0.045]; coaching[0.045]; college[0.045]; LaRon[0.044]; coach[0.044]; coach[0.044]; assistant[0.044]; played[0.044]; university[0.044]; university[0.044]; Alabama[0.044]; recruiting[0.043]; Seminoles[0.043]; players[0.043]; players[0.043]; defensive[0.043]; staff[0.043]; University[0.043]; Quarterback[0.042]; job[0.042]; prep[0.042]; receiver[0.042]; Tuscaloosa[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSeminoles\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFlorida State University\u001b[39m <---> \u001b[32mFlorida State Seminoles football\u001b[39m\t\nSCORES: global= 0.258:0.256[\u001b[31m0.002\u001b[39m]; local()= 0.177:0.147[\u001b[31m0.030\u001b[39m]; log p(e|m)= -1.917:-1.351[\u001b[32m0.566\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFlorida[0.045]; University[0.044]; players[0.044]; players[0.044]; Bowden[0.043]; college[0.043]; LSU[0.042]; coaching[0.042]; played[0.042]; Alabama[0.042]; coach[0.042]; coach[0.042]; State[0.041]; academic[0.041]; academic[0.041]; academic[0.041]; Bowl[0.041]; recruiting[0.041]; defensive[0.041]; assistant[0.040]; university[0.040]; university[0.040]; family[0.040]; Bobby[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saban\u001b[39m ==> ENTITY: \u001b[32mNick Saban\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaban[0.044]; Saban[0.044]; Saban[0.044]; Saban[0.044]; LSU[0.044]; LSU[0.044]; Michigan[0.043]; coaching[0.042]; Bowden[0.042]; college[0.042]; Tuscaloosa[0.042]; coach[0.042]; LaRon[0.042]; Bowl[0.041]; Georgia[0.041]; job[0.041]; players[0.040]; players[0.040]; Bobby[0.040]; played[0.040]; Quarterback[0.040]; State[0.040]; prep[0.040]; victory[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mLSU\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mLSU Tigers football\u001b[39m <---> \u001b[32mLouisiana State University\u001b[39m\t\nSCORES: global= 0.257:0.245[\u001b[31m0.012\u001b[39m]; local()= 0.164:0.112[\u001b[31m0.052\u001b[39m]; log p(e|m)= -1.448:-0.839[\u001b[32m0.609\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLSU[0.046]; coach[0.045]; players[0.044]; players[0.044]; coaching[0.043]; college[0.042]; Tuscaloosa[0.042]; final[0.042]; Bowl[0.042]; State[0.042]; played[0.041]; receiver[0.041]; Landry[0.041]; victory[0.041]; Saban[0.041]; Saban[0.041]; Saban[0.041]; Saban[0.041]; Saban[0.041]; Georgia[0.041]; Doucet[0.041]; crowd[0.041]; Wednesday[0.040]; Wednesday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mike DuBose\u001b[39m ==> ENTITY: \u001b[32mMike DuBose\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlabama[0.043]; Alabama[0.043]; Alabama[0.043]; Alabama[0.043]; LSU[0.043]; team[0.043]; game[0.042]; league[0.042]; championship[0.042]; college[0.042]; seasons[0.042]; football[0.042]; divisional[0.042]; teams[0.041]; teams[0.041]; win[0.041]; win[0.041]; Coach[0.041]; came[0.041]; Tide[0.040]; Tide[0.040]; Saban[0.040]; Saban[0.040]; Saban[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bear Bryant\u001b[39m ==> ENTITY: \u001b[32mBear Bryant\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlabama[0.044]; Alabama[0.044]; football[0.044]; football[0.044]; championship[0.043]; championship[0.043]; college[0.042]; college[0.042]; national[0.042]; national[0.042]; Saban[0.042]; Saban[0.042]; Saban[0.042]; won[0.042]; coach[0.042]; coach[0.042]; athletic[0.041]; titles[0.040]; title[0.040]; Tide[0.040]; Tide[0.040]; Moore[0.040]; program[0.039]; late[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Moore\u001b[39m ==> ENTITY: \u001b[32mMal Moore\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.046]; coached[0.044]; seasons[0.043]; national[0.043]; national[0.043]; teams[0.043]; teams[0.043]; Alabama[0.043]; Alabama[0.043]; Alabama[0.043]; Coach[0.042]; faculty[0.042]; Gene[0.042]; championships[0.041]; league[0.041]; Mike[0.041]; Stallings[0.040]; committee[0.040]; John[0.040]; state[0.040]; state[0.040]; state[0.040]; championship[0.040]; history[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bobby Bowden\u001b[39m ==> ENTITY: \u001b[32mBobby Bowden\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFlorida[0.043]; coaching[0.043]; coach[0.043]; coach[0.043]; Alabama[0.043]; recruiting[0.042]; LSU[0.042]; Saban[0.042]; Saban[0.042]; Saban[0.042]; Saban[0.042]; Saban[0.042]; college[0.042]; LaRon[0.041]; assistant[0.041]; Seminoles[0.041]; Tuscaloosa[0.041]; players[0.041]; players[0.041]; university[0.041]; university[0.041]; played[0.040]; University[0.040]; Quarterback[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saban\u001b[39m ==> ENTITY: \u001b[32mNick Saban\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaban[0.046]; Saban[0.046]; Saban[0.046]; Saban[0.046]; Alabama[0.045]; LSU[0.045]; coaching[0.044]; Florida[0.043]; Bowden[0.043]; college[0.043]; Tuscaloosa[0.043]; coach[0.043]; coach[0.043]; recruiting[0.043]; LaRon[0.043]; Bowl[0.043]; Seminoles[0.043]; assistant[0.042]; staff[0.042]; university[0.042]; job[0.042]; hire[0.042]; hired[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAlabama\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAlabama Crimson Tide football\u001b[39m <---> \u001b[32mUniversity of Alabama\u001b[39m\t\nSCORES: global= 0.254:0.241[\u001b[31m0.013\u001b[39m]; local()= 0.191:0.144[\u001b[31m0.047\u001b[39m]; log p(e|m)= -3.244:-3.058[\u001b[32m0.187\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlabama[0.044]; Alabama[0.044]; Alabama[0.044]; team[0.044]; football[0.043]; LSU[0.043]; game[0.042]; teams[0.042]; teams[0.042]; Tide[0.042]; Tide[0.042]; college[0.041]; championship[0.041]; championship[0.041]; league[0.041]; seasons[0.041]; championships[0.041]; national[0.040]; national[0.040]; national[0.040]; national[0.040]; athletic[0.040]; Miami[0.040]; won[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Michigan State ==> ENTITY: \u001b[32mMichigan State Spartans football\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.127:0.127[0]; log p(e|m)= -1.394:-1.394[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncoaching[0.047]; coach[0.045]; players[0.044]; players[0.044]; college[0.043]; LSU[0.042]; LSU[0.042]; final[0.042]; receiver[0.042]; home[0.042]; home[0.042]; Georgia[0.041]; played[0.041]; Bowl[0.041]; missed[0.041]; victory[0.041]; Landry[0.041]; Saban[0.040]; Saban[0.040]; Saban[0.040]; Saban[0.040]; Saban[0.040]; JaMarcus[0.040]; Quarterback[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saban\u001b[39m ==> ENTITY: \u001b[32mNick Saban\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaban[0.045]; Saban[0.045]; Saban[0.045]; Saban[0.045]; LSU[0.045]; Michigan[0.044]; coaching[0.044]; college[0.043]; Tuscaloosa[0.043]; coach[0.043]; Georgia[0.042]; job[0.042]; State[0.041]; final[0.041]; Wednesday[0.040]; missed[0.040]; Terry[0.040]; Terry[0.040]; hurt[0.039]; said[0.039]; said[0.039]; said[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mTide\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAlabama Crimson Tide football\u001b[39m <---> \u001b[32mAlabama Crimson Tide\u001b[39m\t\nSCORES: global= 0.280:0.262[\u001b[31m0.018\u001b[39m]; local()= 0.198:0.190[\u001b[31m0.009\u001b[39m]; log p(e|m)= -2.071:-2.079[\u001b[31m0.008\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlabama[0.044]; Alabama[0.044]; Alabama[0.044]; Alabama[0.044]; team[0.044]; football[0.043]; LSU[0.043]; game[0.043]; seasons[0.042]; teams[0.042]; teams[0.042]; championships[0.041]; league[0.041]; time[0.041]; Miami[0.041]; Saban[0.040]; Saban[0.040]; Saban[0.040]; national[0.040]; national[0.040]; national[0.040]; championship[0.040]; sixth[0.040]; won[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAlabama\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAlabama Crimson Tide football\u001b[39m <---> \u001b[32mUniversity of Alabama\u001b[39m\t\nSCORES: global= 0.254:0.243[\u001b[31m0.012\u001b[39m]; local()= 0.170:0.136[\u001b[31m0.034\u001b[39m]; log p(e|m)= -3.244:-3.058[\u001b[32m0.187\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlabama[0.045]; team[0.045]; teams[0.043]; teams[0.043]; Tide[0.043]; championship[0.043]; seasons[0.043]; national[0.042]; national[0.042]; John[0.042]; Saban[0.041]; coached[0.041]; Richard[0.041]; academic[0.041]; academic[0.041]; Bryant[0.041]; university[0.041]; play[0.041]; played[0.041]; sixth[0.040]; time[0.040]; Moore[0.040]; Moore[0.040]; win[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mTide\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAlabama Crimson Tide football\u001b[39m <---> \u001b[32mAlabama Crimson Tide\u001b[39m\t\nSCORES: global= 0.280:0.262[\u001b[31m0.017\u001b[39m]; local()= 0.183:0.175[\u001b[31m0.008\u001b[39m]; log p(e|m)= -2.071:-2.079[\u001b[31m0.008\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlabama[0.045]; Alabama[0.045]; Alabama[0.045]; football[0.044]; LSU[0.044]; game[0.043]; championships[0.042]; league[0.042]; athletic[0.041]; Miami[0.041]; Saban[0.041]; Saban[0.041]; Saban[0.041]; national[0.041]; national[0.041]; national[0.041]; national[0.041]; championship[0.041]; won[0.040]; won[0.040]; Bryant[0.040]; play[0.040]; college[0.040]; Tide[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Quarterback\u001b[39m ==> ENTITY: \u001b[32mQuarterback\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.001:-0.001[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndefensive[0.048]; LSU[0.047]; LSU[0.047]; college[0.047]; Michigan[0.046]; LaRon[0.046]; coaching[0.046]; players[0.046]; players[0.046]; university[0.046]; Florida[0.045]; key[0.045]; Bowl[0.045]; Seminoles[0.045]; Kevin[0.045]; receiver[0.045]; University[0.045]; Alabama[0.044]; played[0.044]; prep[0.044]; academic[0.044]; academic[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = JaMarcus Russell\u001b[39m ==> ENTITY: \u001b[32mJaMarcus Russell\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.118:0.118[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLSU[0.044]; LSU[0.044]; college[0.044]; LaRon[0.043]; players[0.043]; players[0.043]; university[0.043]; Seminoles[0.042]; receiver[0.042]; University[0.042]; Quarterback[0.041]; Landry[0.041]; missed[0.041]; defensive[0.041]; prep[0.041]; academic[0.041]; academic[0.041]; coaching[0.041]; played[0.041]; State[0.040]; State[0.040]; recruiting[0.040]; Kevin[0.040]; Doucet[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAlabama\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAlabama Crimson Tide football\u001b[39m <---> \u001b[32mUniversity of Alabama\u001b[39m\t\nSCORES: global= 0.248:0.236[\u001b[31m0.012\u001b[39m]; local()= 0.149:0.113[\u001b[31m0.036\u001b[39m]; log p(e|m)= -3.244:-3.058[\u001b[32m0.187\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfootball[0.046]; Tide[0.044]; college[0.044]; college[0.044]; championship[0.044]; Miami[0.043]; won[0.043]; Saban[0.042]; Saban[0.042]; Bryant[0.042]; Crimson[0.042]; Dolphins[0.041]; coach[0.041]; coach[0.041]; led[0.041]; fans[0.041]; Nick[0.041]; Thursday[0.041]; late[0.040]; wanted[0.040]; country[0.040]; hear[0.039]; program[0.039]; Bear[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lake Burton\u001b[39m ==> ENTITY: \u001b[32mLake Burton (Georgia)\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.064:0.064[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLSU[0.047]; Michigan[0.046]; north[0.045]; Georgia[0.044]; coaching[0.044]; State[0.043]; coach[0.042]; missed[0.042]; Terry[0.041]; Saban[0.041]; Saban[0.041]; Saban[0.041]; Saban[0.041]; Wednesday[0.041]; time[0.041]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; home[0.040]; home[0.040]; stop[0.040]; stop[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Georgia ==> ENTITY: \u001b[32mGeorgia (U.S. state)\u001b[39m\t\nSCORES: global= 0.239:0.239[0]; local()= 0.056:0.056[0]; log p(e|m)= -0.701:-0.701[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncoaching[0.047]; coach[0.046]; LSU[0.045]; final[0.043]; missed[0.043]; Saban[0.042]; Saban[0.042]; Saban[0.042]; Saban[0.042]; hurt[0.042]; traveled[0.042]; Michigan[0.042]; State[0.042]; north[0.041]; Wednesday[0.040]; cheering[0.040]; retiring[0.040]; home[0.040]; home[0.040]; time[0.040]; Terry[0.040]; reception[0.040]; family[0.039]; didn[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ala.\u001b[39m ==> ENTITY: \u001b[32mAlabama\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.056:0.056[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlabama[0.049]; college[0.046]; college[0.046]; Crimson[0.043]; Thursday[0.042]; football[0.042]; Dolphins[0.042]; hear[0.042]; country[0.042]; coach[0.042]; Tide[0.042]; citing[0.041]; roots[0.041]; Miami[0.041]; wanted[0.040]; places[0.040]; Saban[0.040]; Saban[0.040]; said[0.040]; said[0.040]; affect[0.040]; couldn[0.040]; opportunity[0.039]; quit[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gene Stallings\u001b[39m ==> ENTITY: \u001b[32mGene Stallings\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.046]; game[0.044]; football[0.044]; LSU[0.042]; league[0.042]; teams[0.042]; teams[0.042]; college[0.042]; championship[0.042]; seasons[0.042]; divisional[0.042]; play[0.041]; play[0.041]; Alabama[0.041]; Alabama[0.041]; Alabama[0.041]; Alabama[0.041]; Coach[0.041]; Mike[0.041]; coached[0.041]; championships[0.040]; played[0.040]; Bryant[0.040]; Miami[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saban\u001b[39m ==> ENTITY: \u001b[32mNick Saban\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaban[0.045]; Alabama[0.045]; game[0.045]; football[0.043]; football[0.043]; college[0.043]; college[0.043]; coach[0.042]; Dolphins[0.042]; win[0.042]; Miami[0.042]; championship[0.041]; championship[0.041]; national[0.041]; national[0.041]; Tide[0.041]; Tide[0.041]; athletic[0.041]; won[0.040]; Bryant[0.040]; Moore[0.040]; title[0.039]; Crimson[0.039]; Thursday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saban\u001b[39m ==> ENTITY: \u001b[32mNick Saban\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaban[0.044]; Saban[0.044]; Alabama[0.044]; LSU[0.044]; coaching[0.042]; Florida[0.042]; Bowden[0.042]; coach[0.042]; coach[0.042]; recruiting[0.042]; LaRon[0.042]; Seminoles[0.041]; assistant[0.041]; John[0.041]; staff[0.041]; university[0.041]; university[0.041]; faculty[0.041]; hire[0.041]; hired[0.041]; players[0.040]; Tide[0.040]; Bobby[0.040]; University[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Early Doucet\u001b[39m ==> ENTITY: \u001b[32mEarly Doucet\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nreceiver[0.044]; LSU[0.044]; LSU[0.044]; college[0.043]; players[0.043]; players[0.043]; LaRon[0.043]; university[0.042]; Seminoles[0.042]; Bowl[0.042]; defensive[0.042]; missed[0.041]; University[0.041]; recruiting[0.041]; Quarterback[0.041]; prep[0.041]; played[0.041]; academic[0.041]; Alabama[0.040]; coaching[0.040]; Kevin[0.040]; State[0.040]; State[0.040]; according[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = LaRon Landry\u001b[39m ==> ENTITY: \u001b[32mLaRon Landry\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLSU[0.046]; LSU[0.046]; college[0.046]; players[0.045]; players[0.045]; safety[0.045]; Seminoles[0.044]; Doucet[0.044]; missed[0.044]; defensive[0.044]; University[0.043]; Quarterback[0.043]; prep[0.043]; receiver[0.043]; academic[0.043]; JaMarcus[0.043]; coaching[0.043]; State[0.042]; State[0.042]; time[0.042]; agent[0.042]; Notre[0.042]; Bowl[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mLSU\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mLSU Tigers football\u001b[39m <---> \u001b[32mLouisiana State University\u001b[39m\t\nSCORES: global= 0.253:0.242[\u001b[31m0.011\u001b[39m]; local()= 0.183:0.152[\u001b[31m0.032\u001b[39m]; log p(e|m)= -1.448:-0.839[\u001b[32m0.609\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.044]; teams[0.044]; teams[0.044]; football[0.044]; athletic[0.043]; league[0.043]; game[0.043]; Alabama[0.042]; Alabama[0.042]; Alabama[0.042]; Alabama[0.042]; championship[0.042]; championship[0.042]; championships[0.042]; seasons[0.041]; national[0.040]; national[0.040]; national[0.040]; national[0.040]; play[0.040]; play[0.040]; college[0.040]; win[0.040]; win[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saban\u001b[39m ==> ENTITY: \u001b[32mNick Saban\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaban[0.048]; Saban[0.048]; LSU[0.047]; Michigan[0.046]; coaching[0.046]; coach[0.045]; Georgia[0.044]; State[0.043]; final[0.042]; Terry[0.042]; said[0.040]; said[0.040]; said[0.040]; manager[0.040]; time[0.040]; retiring[0.039]; crowd[0.039]; cheering[0.039]; home[0.039]; home[0.039]; equipment[0.038]; landed[0.038]; agent[0.038]; huge[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCrimson Tide\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAlabama Crimson Tide football\u001b[39m <---> \u001b[32mAlabama Crimson Tide\u001b[39m\t\nSCORES: global= 0.271:0.250[\u001b[31m0.020\u001b[39m]; local()= 0.169:0.146[\u001b[31m0.023\u001b[39m]; log p(e|m)= -1.181:-1.191[\u001b[31m0.010\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlabama[0.045]; Alabama[0.045]; football[0.045]; national[0.043]; national[0.043]; athletic[0.042]; coach[0.042]; coach[0.042]; Miami[0.042]; Saban[0.041]; Saban[0.041]; Saban[0.041]; championship[0.041]; championship[0.041]; won[0.041]; college[0.041]; college[0.041]; Bryant[0.041]; titles[0.041]; Dolphins[0.040]; fans[0.040]; fans[0.040]; program[0.040]; Nick[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAlabama\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAlabama Crimson Tide football\u001b[39m <---> \u001b[32mUniversity of Alabama\u001b[39m\t\nSCORES: global= 0.255:0.242[\u001b[31m0.013\u001b[39m]; local()= 0.196:0.147[\u001b[31m0.049\u001b[39m]; log p(e|m)= -3.244:-3.058[\u001b[32m0.187\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlabama[0.044]; Alabama[0.044]; Alabama[0.044]; team[0.044]; football[0.043]; LSU[0.043]; game[0.043]; teams[0.042]; teams[0.042]; Tide[0.042]; college[0.042]; championship[0.042]; league[0.041]; seasons[0.041]; championships[0.041]; national[0.041]; national[0.041]; national[0.041]; Miami[0.040]; won[0.040]; Saban[0.040]; Saban[0.040]; Saban[0.040]; coached[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nick Saban\u001b[39m ==> ENTITY: \u001b[32mNick Saban\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaban[0.047]; Alabama[0.047]; football[0.045]; college[0.044]; college[0.044]; coach[0.044]; Dolphins[0.044]; Miami[0.044]; championship[0.043]; Tide[0.043]; Crimson[0.041]; Thursday[0.041]; fans[0.041]; cut[0.040]; said[0.040]; said[0.040]; wanted[0.040]; obviously[0.040]; citing[0.040]; quit[0.039]; pay[0.039]; leave[0.039]; couldn[0.038]; reason[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = University of Alabama\u001b[39m ==> ENTITY: \u001b[32mUniversity of Alabama\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLSU[0.044]; players[0.043]; Bowden[0.043]; Florida[0.043]; recruiting[0.043]; Tide[0.042]; coaching[0.042]; Seminoles[0.042]; coach[0.042]; coach[0.042]; Bowl[0.042]; university[0.041]; university[0.041]; Saban[0.041]; Saban[0.041]; Saban[0.041]; assistant[0.041]; victory[0.041]; academic[0.041]; academic[0.041]; academic[0.041]; academic[0.041]; faculty[0.041]; played[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saban\u001b[39m ==> ENTITY: \u001b[32mNick Saban\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaban[0.049]; Saban[0.049]; Alabama[0.049]; Alabama[0.049]; Alabama[0.049]; Alabama[0.049]; LSU[0.049]; game[0.049]; seasons[0.048]; football[0.047]; Mike[0.047]; college[0.047]; coach[0.047]; league[0.047]; divisional[0.047]; teams[0.046]; win[0.046]; Miami[0.046]; championship[0.046]; play[0.045]; play[0.045]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mLSU\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mLSU Tigers football\u001b[39m <---> \u001b[32mLouisiana State University\u001b[39m\t\nSCORES: global= 0.256:0.245[\u001b[31m0.011\u001b[39m]; local()= 0.158:0.120[\u001b[31m0.038\u001b[39m]; log p(e|m)= -1.448:-0.839[\u001b[32m0.609\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLSU[0.052]; players[0.049]; players[0.049]; coach[0.049]; coach[0.049]; Alabama[0.049]; University[0.048]; receiver[0.047]; college[0.047]; Tuscaloosa[0.047]; coaching[0.047]; Bowl[0.047]; Tide[0.047]; assistant[0.047]; Michigan[0.047]; played[0.047]; Seminoles[0.046]; Florida[0.046]; Landry[0.046]; Bobby[0.046]; victory[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sugar Bowl ==> ENTITY: \u001b[32mSugar Bowl\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.141:0.141[0]; log p(e|m)= -0.329:-0.329[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLSU[0.045]; LSU[0.045]; receiver[0.043]; Alabama[0.043]; recruiting[0.042]; Seminoles[0.042]; defensive[0.042]; Kevin[0.042]; Florida[0.042]; players[0.042]; players[0.042]; key[0.041]; missed[0.041]; college[0.041]; Tide[0.041]; coaching[0.041]; Tuscaloosa[0.041]; coach[0.041]; coach[0.041]; Bowden[0.041]; played[0.041]; Wednesday[0.040]; Wednesday[0.040]; Michigan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = SEC ==> ENTITY: \u001b[32mSoutheastern Conference\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.219:0.219[0]; log p(e|m)= -0.435:-0.435[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.049]; teams[0.047]; teams[0.047]; league[0.047]; football[0.047]; game[0.046]; championship[0.046]; championship[0.046]; LSU[0.045]; won[0.045]; Alabama[0.045]; Alabama[0.045]; Alabama[0.045]; Alabama[0.045]; divisional[0.045]; seasons[0.044]; win[0.044]; win[0.044]; national[0.044]; national[0.044]; national[0.044]; national[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Moore\u001b[39m ==> ENTITY: \u001b[32mMal Moore\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.049]; coached[0.047]; LSU[0.047]; seasons[0.046]; national[0.046]; national[0.046]; national[0.046]; teams[0.046]; teams[0.046]; Alabama[0.046]; Alabama[0.046]; Alabama[0.046]; Alabama[0.046]; divisional[0.046]; Coach[0.045]; Gene[0.045]; championships[0.045]; league[0.044]; Mike[0.044]; Stallings[0.043]; committee[0.043]; game[0.043]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAlabama\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAlabama Crimson Tide football\u001b[39m <---> \u001b[32mAlabama\u001b[39m\t\nSCORES: global= 0.255:0.231[\u001b[31m0.024\u001b[39m]; local()= 0.168:0.035[\u001b[31m0.133\u001b[39m]; log p(e|m)= -3.244:-0.324[\u001b[32m2.920\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlabama[0.045]; football[0.044]; LSU[0.044]; game[0.043]; Tide[0.042]; Tide[0.042]; college[0.042]; championship[0.042]; championships[0.042]; national[0.041]; national[0.041]; national[0.041]; athletic[0.041]; Miami[0.041]; Miami[0.041]; won[0.041]; Saban[0.041]; Saban[0.041]; Saban[0.041]; reported[0.041]; divisional[0.040]; Bryant[0.040]; play[0.040]; Crimson[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saban\u001b[39m ==> ENTITY: \u001b[32mNick Saban\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaban[0.044]; Saban[0.044]; Alabama[0.044]; LSU[0.044]; coaching[0.042]; Florida[0.042]; Bowden[0.042]; coach[0.042]; coach[0.042]; recruiting[0.042]; LaRon[0.042]; Bowl[0.041]; Seminoles[0.041]; assistant[0.041]; John[0.041]; staff[0.041]; university[0.041]; university[0.041]; faculty[0.041]; hire[0.041]; hired[0.041]; players[0.040]; players[0.040]; Tide[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mTuscaloosa\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUniversity of Alabama\u001b[39m <---> \u001b[32mTuscaloosa, Alabama\u001b[39m\t\nSCORES: global= 0.265:0.257[\u001b[31m0.007\u001b[39m]; local()= 0.106:0.065[\u001b[31m0.041\u001b[39m]; log p(e|m)= -2.079:-0.067[\u001b[32m2.012\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlabama[0.047]; college[0.046]; college[0.046]; football[0.046]; coach[0.043]; championship[0.042]; Thursday[0.042]; fans[0.042]; Saban[0.041]; Saban[0.041]; Tide[0.041]; said[0.041]; said[0.041]; country[0.041]; Miami[0.041]; Dolphins[0.041]; Nick[0.040]; Crimson[0.040]; places[0.040]; people[0.040]; wanted[0.040]; citing[0.040]; leave[0.039]; taking[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saban\u001b[39m ==> ENTITY: \u001b[32mNick Saban\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaban[0.046]; Alabama[0.046]; football[0.044]; college[0.044]; college[0.044]; coach[0.044]; coach[0.044]; Dolphins[0.044]; Miami[0.043]; championship[0.043]; Tide[0.042]; Nick[0.042]; Crimson[0.040]; Thursday[0.040]; late[0.040]; fans[0.040]; led[0.040]; cut[0.039]; said[0.039]; said[0.039]; wanted[0.039]; obviously[0.039]; program[0.039]; citing[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Notre Dame ==> ENTITY: \u001b[32mNotre Dame Fighting Irish football\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.146:0.146[0]; log p(e|m)= -1.378:-1.378[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrecruiting[0.044]; players[0.043]; players[0.043]; college[0.043]; coaching[0.043]; Michigan[0.042]; receiver[0.042]; defensive[0.042]; academic[0.042]; LSU[0.042]; LSU[0.042]; victory[0.041]; Kevin[0.041]; coach[0.041]; coach[0.041]; played[0.041]; huge[0.041]; Bowl[0.041]; missed[0.041]; Quarterback[0.041]; Florida[0.041]; State[0.041]; State[0.041]; Bowden[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Baton Rouge ==> ENTITY: \u001b[32mBaton Rouge, Louisiana\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.096:0.096[0]; log p(e|m)= -0.065:-0.065[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayers[0.046]; players[0.046]; LSU[0.046]; LSU[0.046]; Tuscaloosa[0.045]; crowd[0.045]; played[0.044]; Bobby[0.044]; cheering[0.044]; Florida[0.043]; victory[0.043]; night[0.043]; Michigan[0.043]; Kevin[0.043]; Landry[0.043]; staff[0.043]; Bowl[0.042]; Wednesday[0.042]; Wednesday[0.042]; reception[0.042]; according[0.042]; Doucet[0.042]; traveled[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bryant\u001b[39m ==> ENTITY: \u001b[32mBear Bryant\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.043]; Alabama[0.043]; Alabama[0.043]; Alabama[0.043]; Alabama[0.043]; seasons[0.043]; game[0.043]; championship[0.042]; LSU[0.042]; championships[0.042]; national[0.041]; national[0.041]; national[0.041]; league[0.041]; divisional[0.041]; Coach[0.041]; coached[0.041]; Saban[0.041]; Saban[0.041]; won[0.041]; teams[0.041]; teams[0.041]; faculty[0.040]; history[0.040]; \t\n\r [=========================================>....] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 202ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 607/656 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 16447720 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Grand Rapids ==> ENTITY: \u001b[32mGrand Rapids, Michigan\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.285:-0.285[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWednesday[0.045]; downtown[0.044]; Grand[0.044]; land[0.044]; Rapids[0.043]; nearly[0.043]; roads[0.042]; month[0.042]; Lansing[0.042]; Ford[0.041]; Ford[0.041]; Ford[0.041]; Ford[0.041]; Ford[0.041]; lined[0.041]; family[0.041]; night[0.041]; skies[0.041]; patrol[0.040]; said[0.040]; said[0.040]; said[0.040]; nation[0.040]; close[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lansing ==> ENTITY: \u001b[32mLansing, Michigan\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.120:0.120[0]; log p(e|m)= -0.446:-0.446[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhometown[0.046]; downtown[0.044]; Grand[0.044]; Grand[0.044]; Grand[0.044]; Wednesday[0.043]; early[0.042]; Church[0.041]; church[0.041]; nearly[0.041]; traveled[0.041]; Episcopal[0.041]; Ford[0.041]; Ford[0.041]; Ford[0.041]; Ford[0.041]; Ford[0.041]; person[0.041]; pay[0.041]; Richard[0.040]; Philip[0.040]; library[0.040]; predecessor[0.040]; noon[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Donald Rumsfeld\u001b[39m ==> ENTITY: \u001b[32mDonald Rumsfeld\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRumsfeld[0.045]; Rumsfeld[0.045]; Rumsfeld[0.045]; Rumsfeld[0.045]; President[0.044]; President[0.044]; American[0.042]; Bush[0.042]; recently[0.042]; said[0.041]; said[0.041]; defense[0.041]; defense[0.041]; included[0.041]; decisions[0.040]; needed[0.040]; secretary[0.040]; secretary[0.040]; inside[0.040]; nation[0.040]; nation[0.040]; Cheney[0.039]; Carter[0.039]; Earlier[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ford ==> ENTITY: \u001b[32mGerald Ford\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.114:0.114[0]; log p(e|m)= -0.405:-0.405[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNixon[0.047]; Wednesday[0.044]; mourners[0.044]; funeral[0.043]; funeral[0.043]; Ford[0.043]; said[0.043]; honor[0.043]; president[0.043]; Richard[0.043]; national[0.042]; outside[0.042]; home[0.041]; Michigan[0.041]; Michigan[0.041]; night[0.040]; stood[0.040]; procession[0.040]; resigned[0.040]; University[0.039]; wore[0.039]; museum[0.039]; museum[0.039]; Smith[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Richard Nixon\u001b[39m ==> ENTITY: \u001b[32mRichard Nixon\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.090:0.090[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfootball[0.047]; funeral[0.045]; funeral[0.045]; played[0.045]; Wednesday[0.043]; national[0.043]; resigned[0.043]; teams[0.042]; said[0.042]; president[0.042]; honor[0.041]; Ford[0.041]; Ford[0.041]; died[0.041]; mourners[0.041]; letter[0.041]; hours[0.040]; stood[0.040]; University[0.040]; tens[0.040]; night[0.040]; Smith[0.040]; outside[0.039]; home[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCalifornia desert\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPalm Desert, California\u001b[39m <---> \u001b[32mDeserts of California\u001b[39m\t\nSCORES: global= 0.283:0.248[\u001b[31m0.036\u001b[39m]; local()= 0.122:0.028[\u001b[31m0.094\u001b[39m]; log p(e|m)= -0.955:-1.465[\u001b[31m0.511\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhome[0.046]; hometown[0.045]; hometown[0.045]; American[0.043]; night[0.043]; nation[0.042]; GRAND[0.042]; Tuesday[0.042]; Wednesday[0.042]; Wednesday[0.042]; country[0.042]; included[0.041]; aircraft[0.041]; Betty[0.041]; graveside[0.041]; sunset[0.041]; flyover[0.041]; burial[0.041]; Gerald[0.040]; Gerald[0.040]; husband[0.040]; boyhood[0.040]; museum[0.040]; museum[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rumsfeld\u001b[39m ==> ENTITY: \u001b[32mDonald Rumsfeld\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRumsfeld[0.049]; Rumsfeld[0.049]; Rumsfeld[0.049]; Rumsfeld[0.049]; President[0.047]; Bush[0.046]; recently[0.045]; said[0.045]; said[0.045]; said[0.045]; month[0.045]; defense[0.045]; defense[0.045]; defense[0.045]; Donald[0.044]; decisions[0.044]; decision[0.044]; needed[0.044]; secretary[0.044]; secretary[0.044]; veteran[0.044]; inside[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ford ==> ENTITY: \u001b[32mGerald Ford\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.405:-0.405[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.045]; motorcade[0.043]; Carter[0.043]; Carter[0.043]; Carter[0.043]; mourners[0.042]; month[0.042]; Ford[0.042]; Ford[0.042]; Ford[0.042]; Ford[0.042]; dignitaries[0.042]; service[0.042]; casket[0.041]; couple[0.041]; Bush[0.041]; family[0.041]; recently[0.041]; predecessor[0.040]; wife[0.040]; nation[0.040]; nation[0.040]; nation[0.040]; traveled[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = American flag\u001b[39m ==> ENTITY: \u001b[32mFlag of the United States\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nofficial[0.043]; museum[0.043]; museum[0.043]; nation[0.043]; draped[0.043]; salute[0.043]; honor[0.042]; included[0.042]; casket[0.042]; casket[0.042]; Dick[0.042]; inside[0.042]; spanned[0.042]; country[0.041]; public[0.041]; viewing[0.041]; boyhood[0.041]; Wednesday[0.041]; Wednesday[0.041]; California[0.041]; decades[0.041]; make[0.041]; remembrance[0.041]; recalled[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ford ==> ENTITY: \u001b[32mGerald Ford\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.405:-0.405[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWednesday[0.043]; motorcade[0.043]; nearly[0.042]; mourners[0.042]; mourners[0.042]; mourners[0.042]; Ford[0.042]; Ford[0.042]; Ford[0.042]; Ford[0.042]; respects[0.042]; presidential[0.042]; presidential[0.042]; said[0.042]; honor[0.042]; Richard[0.041]; early[0.041]; casket[0.041]; casket[0.041]; casket[0.041]; outside[0.041]; family[0.041]; family[0.041]; flag[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ford ==> ENTITY: \u001b[32mGerald Ford\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.405:-0.405[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWednesday[0.043]; nearly[0.042]; mourners[0.042]; mourners[0.042]; funeral[0.042]; Ford[0.042]; Ford[0.042]; Ford[0.042]; respects[0.042]; presidential[0.042]; presidential[0.042]; said[0.042]; said[0.042]; honor[0.042]; Richard[0.042]; early[0.041]; casket[0.041]; casket[0.041]; outside[0.041]; family[0.041]; family[0.041]; flag[0.041]; returned[0.040]; prayers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rancho Mirage\u001b[39m ==> ENTITY: \u001b[32mRancho Mirage, California\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRichard[0.044]; University[0.044]; died[0.044]; played[0.044]; night[0.043]; Wednesday[0.043]; football[0.042]; alma[0.042]; championship[0.042]; home[0.042]; Ford[0.041]; Ford[0.041]; museum[0.041]; museum[0.041]; honor[0.041]; funeral[0.041]; funeral[0.041]; mourners[0.041]; sweat[0.041]; teams[0.041]; Dec[0.040]; said[0.040]; national[0.040]; Calif[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Calif.\u001b[39m ==> ENTITY: \u001b[32mCalifornia\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.064:0.064[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfootball[0.048]; teams[0.046]; played[0.044]; championship[0.044]; Michigan[0.044]; Michigan[0.044]; national[0.044]; University[0.043]; Smith[0.043]; alma[0.041]; Rancho[0.041]; Wednesday[0.041]; home[0.041]; night[0.041]; Richard[0.041]; resigned[0.040]; honor[0.040]; Ford[0.039]; Ford[0.039]; said[0.039]; museum[0.039]; wore[0.039]; Nixon[0.039]; Dec[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rumsfeld\u001b[39m ==> ENTITY: \u001b[32mDonald Rumsfeld\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRumsfeld[0.049]; Rumsfeld[0.049]; Rumsfeld[0.049]; Rumsfeld[0.049]; President[0.047]; Bush[0.046]; recently[0.045]; said[0.045]; said[0.045]; said[0.045]; month[0.045]; defense[0.045]; defense[0.045]; defense[0.045]; Donald[0.044]; personal[0.044]; decision[0.044]; needed[0.044]; secretary[0.044]; secretary[0.044]; veteran[0.044]; inside[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vice President ==> ENTITY: \u001b[32mVice President of the United States\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.152:0.152[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwife[0.049]; presidential[0.047]; husband[0.047]; president[0.046]; president[0.046]; Carter[0.045]; Donald[0.044]; American[0.044]; lady[0.043]; included[0.043]; Cheney[0.043]; Dick[0.043]; Ford[0.042]; Ford[0.042]; Ford[0.042]; Ford[0.042]; Ford[0.042]; Ford[0.042]; service[0.042]; Tuesday[0.042]; Wednesday[0.042]; Wednesday[0.042]; country[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gerald R. Ford\u001b[39m ==> ENTITY: \u001b[32mGerald Ford\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \naircraft[0.045]; President[0.044]; Navy[0.044]; Navy[0.044]; carrier[0.043]; nation[0.042]; nation[0.042]; nation[0.042]; Carter[0.042]; Carter[0.042]; presidential[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; Ford[0.041]; Ford[0.041]; Ford[0.041]; Ford[0.041]; Bush[0.040]; flag[0.040]; motorcade[0.040]; decades[0.040]; leader[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Grand Rapids ==> ENTITY: \u001b[32mGrand Rapids, Michigan\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.285:-0.285[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhometown[0.044]; Grand[0.044]; Grand[0.044]; blue[0.044]; downtown[0.043]; Rapids[0.043]; Rapids[0.043]; nearly[0.043]; family[0.042]; family[0.042]; early[0.041]; Lansing[0.041]; Wednesday[0.041]; library[0.041]; Church[0.041]; flag[0.041]; night[0.040]; night[0.040]; supporters[0.040]; Richard[0.040]; person[0.040]; outside[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Congress ==> ENTITY: \u001b[32mUnited States Congress\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.091:0.091[0]; log p(e|m)= -0.335:-0.335[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npolitical[0.044]; Wednesday[0.044]; Richard[0.043]; honor[0.043]; represented[0.042]; Smith[0.042]; Smith[0.042]; file[0.042]; flag[0.042]; Norton[0.042]; presidential[0.042]; Philip[0.042]; returned[0.042]; returning[0.041]; period[0.041]; early[0.041]; allies[0.041]; nearly[0.041]; supporters[0.041]; historian[0.041]; hometown[0.041]; funeral[0.041]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ford ==> ENTITY: \u001b[32mGerald Ford\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.405:-0.405[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.044]; President[0.044]; Navy[0.044]; nation[0.042]; nation[0.042]; Carter[0.042]; Betty[0.042]; service[0.042]; funeral[0.042]; Ford[0.042]; Ford[0.042]; Ford[0.042]; Ford[0.042]; dignitaries[0.042]; said[0.041]; said[0.041]; honor[0.041]; American[0.041]; casket[0.041]; casket[0.041]; aircraft[0.041]; couple[0.040]; couple[0.040]; Bush[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = President Bush\u001b[39m ==> ENTITY: \u001b[32mGeorge W. Bush\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.120:0.120[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrecently[0.047]; honor[0.047]; decision[0.046]; Rumsfeld[0.046]; Rumsfeld[0.046]; Rumsfeld[0.046]; Rumsfeld[0.046]; Rumsfeld[0.046]; Carter[0.046]; Carter[0.046]; Carter[0.046]; month[0.046]; public[0.045]; personal[0.045]; Donald[0.045]; aircraft[0.044]; service[0.044]; leader[0.044]; said[0.044]; said[0.044]; said[0.044]; said[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ford ==> ENTITY: \u001b[32mGerald Ford\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.405:-0.405[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.044]; Navy[0.044]; Navy[0.044]; Wednesday[0.042]; motorcade[0.042]; nation[0.042]; nation[0.042]; nation[0.042]; Carter[0.042]; Carter[0.042]; nearly[0.042]; mourners[0.041]; mourners[0.041]; month[0.041]; Ford[0.041]; Ford[0.041]; Ford[0.041]; Ford[0.041]; respects[0.041]; presidential[0.041]; carrier[0.041]; said[0.041]; said[0.041]; said[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ford ==> ENTITY: \u001b[32mGerald Ford\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.405:-0.405[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.044]; Wednesday[0.042]; Wednesday[0.042]; nation[0.042]; Tuesday[0.042]; Carter[0.042]; Betty[0.042]; service[0.042]; service[0.042]; funeral[0.042]; Ford[0.042]; Ford[0.042]; Ford[0.042]; Ford[0.042]; dignitaries[0.042]; presidential[0.042]; honor[0.041]; president[0.041]; American[0.041]; casket[0.041]; casket[0.041]; aircraft[0.041]; couple[0.041]; couple[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Richard Norton Smith\u001b[39m ==> ENTITY: \u001b[32mRichard Norton Smith\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.122:0.122[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhonor[0.044]; University[0.044]; historian[0.043]; Ford[0.043]; Ford[0.043]; Ford[0.043]; Ford[0.043]; Ford[0.043]; director[0.042]; Smith[0.042]; Congress[0.042]; Rapids[0.041]; Rapids[0.041]; museum[0.041]; museum[0.041]; national[0.041]; presidential[0.041]; author[0.041]; Michigan[0.041]; football[0.040]; tens[0.040]; library[0.040]; Philip[0.040]; alma[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ford ==> ENTITY: \u001b[32mGerald Ford\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.114:0.114[0]; log p(e|m)= -0.405:-0.405[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmourners[0.044]; mourners[0.044]; funeral[0.044]; Ford[0.044]; Ford[0.044]; Ford[0.044]; Ford[0.044]; respects[0.044]; presidential[0.044]; said[0.044]; said[0.044]; honor[0.044]; Richard[0.043]; early[0.043]; casket[0.043]; casket[0.043]; national[0.043]; outside[0.043]; family[0.043]; family[0.043]; flag[0.043]; returned[0.042]; prayers[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gerald R. Ford\u001b[39m ==> ENTITY: \u001b[32mGerald Ford\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \naircraft[0.046]; President[0.045]; nation[0.043]; Tuesday[0.043]; service[0.043]; Wednesday[0.042]; Wednesday[0.042]; presidential[0.042]; Ford[0.042]; Ford[0.042]; Ford[0.042]; president[0.042]; president[0.042]; American[0.042]; flag[0.041]; country[0.041]; boyhood[0.041]; ex[0.040]; Michigan[0.040]; Vice[0.040]; flyover[0.040]; included[0.040]; Dick[0.039]; home[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Michigan ==> ENTITY: \u001b[32mMichigan\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.113:0.113[0]; log p(e|m)= -0.248:-0.248[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhometown[0.046]; hometown[0.046]; president[0.046]; president[0.046]; presidential[0.044]; Tuesday[0.044]; Ford[0.043]; Ford[0.043]; Ford[0.043]; Wednesday[0.043]; Wednesday[0.043]; Gerald[0.043]; Gerald[0.043]; boyhood[0.043]; country[0.043]; California[0.043]; nation[0.043]; night[0.042]; Mich[0.042]; caps[0.042]; capped[0.042]; included[0.042]; home[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ford ==> ENTITY: \u001b[32mGerald Ford\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.405:-0.405[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.044]; Navy[0.044]; Navy[0.044]; nation[0.042]; nation[0.042]; nation[0.042]; Carter[0.042]; Carter[0.042]; Carter[0.042]; month[0.042]; service[0.041]; Ford[0.041]; Ford[0.041]; Ford[0.041]; Ford[0.041]; Ford[0.041]; dignitaries[0.041]; carrier[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; casket[0.041]; aircraft[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dick Cheney\u001b[39m ==> ENTITY: \u001b[32mDick Cheney\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.113:0.113[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \naircraft[0.049]; President[0.045]; Wednesday[0.043]; Wednesday[0.043]; presidential[0.043]; Vice[0.043]; Tuesday[0.042]; Rumsfeld[0.042]; Rumsfeld[0.042]; president[0.041]; included[0.041]; official[0.041]; honor[0.041]; flyover[0.041]; nation[0.041]; rest[0.041]; service[0.041]; service[0.041]; American[0.040]; gun[0.040]; Earlier[0.040]; Carter[0.040]; church[0.040]; Jimmy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Carter\u001b[39m ==> ENTITY: \u001b[32mJimmy Carter\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.137:0.137[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.045]; Carter[0.044]; recently[0.043]; Bush[0.043]; month[0.042]; presidential[0.042]; said[0.042]; said[0.042]; said[0.042]; nation[0.042]; nation[0.042]; Navy[0.042]; Navy[0.042]; want[0.041]; Gerald[0.041]; veteran[0.041]; nearly[0.041]; pay[0.040]; Ford[0.040]; Ford[0.040]; Ford[0.040]; Ford[0.040]; Ford[0.040]; described[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ford ==> ENTITY: \u001b[32mGerald Ford\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.111:0.111[0]; log p(e|m)= -0.405:-0.405[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmourners[0.043]; mourners[0.043]; funeral[0.043]; Ford[0.043]; Ford[0.043]; Ford[0.043]; Ford[0.043]; presidential[0.042]; said[0.042]; said[0.042]; honor[0.042]; Richard[0.042]; early[0.042]; casket[0.042]; casket[0.042]; national[0.041]; outside[0.041]; family[0.041]; family[0.041]; flag[0.041]; returned[0.040]; prayers[0.040]; Grand[0.040]; Grand[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Grand Rapids ==> ENTITY: \u001b[32mGrand Rapids, Michigan\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.125:0.125[0]; log p(e|m)= -0.285:-0.285[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayed[0.045]; Michigan[0.044]; Michigan[0.044]; national[0.044]; blue[0.044]; hometown[0.043]; Grand[0.043]; Rapids[0.042]; teams[0.042]; championship[0.042]; University[0.041]; football[0.041]; family[0.041]; family[0.041]; early[0.041]; Lansing[0.041]; Wednesday[0.041]; night[0.041]; night[0.041]; library[0.040]; Church[0.040]; Richard[0.039]; person[0.039]; outside[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jimmy Carter\u001b[39m ==> ENTITY: \u001b[32mJimmy Carter\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.121:0.121[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.045]; President[0.045]; recently[0.043]; Bush[0.043]; American[0.043]; Donald[0.043]; said[0.042]; said[0.042]; nation[0.042]; nation[0.042]; honor[0.042]; Jack[0.042]; Navy[0.042]; service[0.041]; Dick[0.041]; dark[0.041]; legend[0.041]; Ford[0.040]; Ford[0.040]; Ford[0.040]; Ford[0.040]; Ford[0.040]; Ford[0.040]; considering[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ford ==> ENTITY: \u001b[32mGerald Ford\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.135:0.135[0]; log p(e|m)= -0.405:-0.405[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWednesday[0.052]; Wednesday[0.052]; Tuesday[0.051]; days[0.043]; days[0.043]; President[0.043]; Betty[0.041]; Ford[0.041]; Ford[0.041]; Ford[0.041]; Ford[0.041]; presidential[0.040]; service[0.040]; president[0.040]; president[0.040]; American[0.040]; casket[0.040]; salute[0.039]; night[0.039]; wife[0.039]; graveside[0.039]; boyhood[0.039]; eulogies[0.039]; nation[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Watergate\u001b[39m ==> ENTITY: \u001b[32mWatergate scandal\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.046]; Carter[0.043]; Carter[0.043]; Carter[0.043]; personal[0.042]; Jimmy[0.042]; month[0.042]; recalled[0.042]; Jack[0.042]; Ford[0.041]; Ford[0.041]; Ford[0.041]; Ford[0.041]; Ford[0.041]; Ford[0.041]; Bush[0.041]; naming[0.041]; veteran[0.041]; recently[0.041]; leader[0.041]; predecessor[0.041]; Donald[0.041]; Gerald[0.041]; public[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Grand Rapids ==> ENTITY: \u001b[32mGrand Rapids, Michigan\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.285:-0.285[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhometown[0.045]; hometown[0.045]; Michigan[0.045]; Tuesday[0.043]; home[0.042]; American[0.042]; Wednesday[0.042]; Wednesday[0.042]; boyhood[0.042]; Mich[0.041]; Dick[0.041]; night[0.041]; country[0.041]; Betty[0.041]; nation[0.041]; caps[0.041]; Ford[0.041]; Ford[0.041]; Ford[0.041]; Ford[0.041]; lady[0.040]; hour[0.040]; capital[0.040]; service[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rosalynn\u001b[39m ==> ENTITY: \u001b[32mRosalynn Carter\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwife[0.044]; President[0.044]; presidential[0.043]; Carter[0.043]; Carter[0.043]; personal[0.043]; nation[0.042]; nation[0.042]; Rapids[0.042]; loved[0.041]; motorcade[0.041]; Bush[0.041]; returning[0.041]; Ford[0.041]; Ford[0.041]; Ford[0.041]; Ford[0.041]; Ford[0.041]; said[0.041]; said[0.041]; said[0.041]; remembered[0.041]; Wednesday[0.041]; month[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ford ==> ENTITY: \u001b[32mGerald Ford\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.119:0.119[0]; log p(e|m)= -0.405:-0.405[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.044]; President[0.044]; nation[0.042]; nation[0.042]; Carter[0.042]; Betty[0.042]; service[0.042]; service[0.042]; funeral[0.042]; Ford[0.042]; Ford[0.042]; Ford[0.042]; Ford[0.042]; dignitaries[0.042]; said[0.041]; honor[0.041]; American[0.041]; casket[0.041]; casket[0.041]; aircraft[0.041]; couple[0.041]; couple[0.041]; Bush[0.041]; salute[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Smith\u001b[39m ==> ENTITY: \u001b[32mRichard Norton Smith\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhonor[0.045]; University[0.045]; Richard[0.045]; Richard[0.045]; historian[0.044]; Ford[0.044]; Ford[0.044]; Ford[0.044]; Ford[0.044]; Ford[0.044]; director[0.044]; Nixon[0.043]; Smith[0.043]; Congress[0.043]; president[0.043]; Rapids[0.043]; Rapids[0.043]; museum[0.042]; museum[0.042]; museum[0.042]; national[0.042]; presidential[0.042]; author[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ford ==> ENTITY: \u001b[32mGerald Ford\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.405:-0.405[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNixon[0.046]; Wednesday[0.043]; mourners[0.043]; mourners[0.043]; funeral[0.042]; funeral[0.042]; Ford[0.042]; Ford[0.042]; Ford[0.042]; presidential[0.042]; honor[0.042]; president[0.042]; Richard[0.042]; Richard[0.042]; early[0.042]; national[0.041]; outside[0.041]; family[0.041]; returned[0.040]; home[0.040]; said[0.040]; Grand[0.040]; Michigan[0.040]; Michigan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rumsfeld\u001b[39m ==> ENTITY: \u001b[32mDonald Rumsfeld\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRumsfeld[0.045]; Rumsfeld[0.045]; Rumsfeld[0.045]; Rumsfeld[0.045]; President[0.044]; American[0.042]; Bush[0.042]; recently[0.042]; said[0.041]; said[0.041]; defense[0.041]; defense[0.041]; Donald[0.041]; decisions[0.041]; decision[0.040]; needed[0.040]; secretary[0.040]; secretary[0.040]; veteran[0.040]; inside[0.040]; nation[0.040]; nation[0.040]; naming[0.040]; Cheney[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rumsfeld\u001b[39m ==> ENTITY: \u001b[32mDonald Rumsfeld\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRumsfeld[0.049]; Rumsfeld[0.049]; Rumsfeld[0.049]; Rumsfeld[0.049]; President[0.047]; Bush[0.046]; recently[0.045]; said[0.045]; said[0.045]; said[0.045]; said[0.045]; month[0.045]; defense[0.045]; defense[0.045]; defense[0.045]; Donald[0.044]; personal[0.044]; decision[0.044]; needed[0.044]; secretary[0.044]; secretary[0.044]; veteran[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Betty Ford\u001b[39m ==> ENTITY: \u001b[32mBetty Ford\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFord[0.044]; Ford[0.044]; Ford[0.044]; Ford[0.044]; Ford[0.044]; wife[0.042]; Donald[0.042]; American[0.042]; home[0.042]; Tuesday[0.042]; decades[0.041]; honor[0.041]; husband[0.041]; Dick[0.041]; night[0.041]; President[0.041]; Wednesday[0.041]; Wednesday[0.041]; nation[0.041]; days[0.040]; president[0.040]; recalled[0.040]; included[0.040]; GRAND[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ford ==> ENTITY: \u001b[32mGerald Ford\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.405:-0.405[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNavy[0.044]; Wednesday[0.042]; motorcade[0.042]; nation[0.042]; Carter[0.042]; Carter[0.042]; nearly[0.042]; mourners[0.042]; mourners[0.042]; month[0.042]; Ford[0.042]; Ford[0.042]; Ford[0.042]; Ford[0.042]; respects[0.041]; presidential[0.041]; carrier[0.041]; said[0.041]; said[0.041]; said[0.041]; honor[0.041]; casket[0.041]; casket[0.041]; casket[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gerald Ford\u001b[39m ==> ENTITY: \u001b[32mGerald Ford\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.110:0.110[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWednesday[0.053]; Wednesday[0.053]; Tuesday[0.052]; days[0.044]; days[0.044]; nation[0.042]; service[0.041]; presidential[0.041]; Ford[0.041]; Ford[0.041]; president[0.041]; president[0.041]; night[0.040]; graveside[0.040]; country[0.039]; sunset[0.039]; Michigan[0.039]; boyhood[0.039]; included[0.039]; capital[0.038]; grounds[0.038]; California[0.038]; hometown[0.038]; hometown[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mich.\u001b[39m ==> ENTITY: \u001b[32mMichigan\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.118:0.118[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMichigan[0.050]; hometown[0.047]; hometown[0.047]; Tuesday[0.045]; Wednesday[0.044]; Wednesday[0.044]; included[0.043]; country[0.043]; boyhood[0.043]; American[0.043]; nation[0.043]; Ford[0.043]; Ford[0.043]; Ford[0.043]; Ford[0.043]; presidential[0.043]; night[0.043]; President[0.042]; Dick[0.042]; California[0.042]; grounds[0.042]; home[0.042]; GRAND[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = University of Michigan\u001b[39m ==> ENTITY: \u001b[32mUniversity of Michigan\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.080:0.080[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfootball[0.046]; teams[0.045]; Ford[0.045]; Ford[0.045]; Ford[0.045]; Ford[0.045]; Michigan[0.044]; championship[0.044]; national[0.044]; played[0.042]; honor[0.041]; hometown[0.041]; Richard[0.041]; Richard[0.041]; Wednesday[0.040]; night[0.040]; Smith[0.040]; Smith[0.040]; Nixon[0.039]; Rapids[0.039]; times[0.039]; early[0.039]; Grand[0.039]; resigned[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Michigan ==> ENTITY: \u001b[32mMichigan\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.083:0.083[0]; log p(e|m)= -0.248:-0.248[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUniversity[0.048]; Michigan[0.047]; football[0.046]; teams[0.046]; championship[0.045]; national[0.044]; president[0.043]; played[0.043]; Richard[0.042]; Ford[0.041]; Ford[0.041]; Wednesday[0.041]; Smith[0.040]; honor[0.040]; Nixon[0.040]; night[0.040]; pew[0.040]; times[0.040]; home[0.039]; museum[0.039]; museum[0.039]; outside[0.039]; line[0.039]; resigned[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jack Nicklaus\u001b[39m ==> ENTITY: \u001b[32mJack Nicklaus\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.129:0.129[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngolfing[0.051]; Donald[0.047]; Jimmy[0.043]; recently[0.043]; month[0.042]; couple[0.042]; couple[0.042]; man[0.042]; legend[0.042]; honor[0.041]; Navy[0.041]; Navy[0.041]; attended[0.041]; Ford[0.040]; Ford[0.040]; Ford[0.040]; Ford[0.040]; Ford[0.040]; special[0.040]; decades[0.040]; decades[0.040]; healed[0.040]; considering[0.040]; needed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Navy ==> ENTITY: \u001b[32mUnited States Navy\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.111:0.111[0]; log p(e|m)= -1.061:-1.061[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNavy[0.057]; President[0.049]; service[0.048]; month[0.048]; nation[0.048]; nation[0.048]; nation[0.048]; patrol[0.048]; aircraft[0.048]; special[0.047]; defense[0.047]; defense[0.047]; Carter[0.047]; Carter[0.047]; Carter[0.047]; recently[0.047]; carrier[0.047]; Donald[0.046]; high[0.046]; considering[0.046]; predecessor[0.046]; \t\n\r [=============================================>] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b Tot: 2s715ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 656/656 \n\r==> time to test 1 sample = 4.1393506817701ms\t\n==> \u001b[31mMSNBC\u001b[39m MSNBC ; EPOCH = 307: Micro recall = 93.29% ; Micro F1 = \u001b[31m93.65%\u001b[39m\t\n ===> entity frequency stats :\t\nfreq = 11-20 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\nfreq = 2-5 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\nfreq = 50+ : num = 640 ; correctly classified = 608 ; perc = 95.00\t\nfreq = 1 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\nfreq = 21-50 : num = 1 ; correctly classified = 0 ; perc = 0.00\t\nfreq = 0 : num = 5 ; correctly classified = 4 ; perc = 80.00\t\nfreq = 6-10 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\n\t\n ===> entity p(e|m) stats :\t\np(e|m) = <=0.001 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\np(e|m) = 0.01-0.03 : num = 2 ; correctly classified = 1 ; perc = 50.00\t\np(e|m) = 0.003-0.01 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\np(e|m) = 0.3+ : num = 590 ; correctly classified = 572 ; perc = 96.95\t\np(e|m) = 0.1-0.3 : num = 41 ; correctly classified = 33 ; perc = 80.49\t\np(e|m) = 0.03-0.1 : num = 13 ; correctly classified = 6 ; perc = 46.15\t\np(e|m) = 0.001-0.003 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\n num_mentions_w/o_gold_ent_in_candidates = 10 total num mentions in dataset = 656\t\n percentage_mentions_w/o_gold_ent_in_candidates = 1.52%; percentage_mentions_solved : both_pem_ours = 87.04%; only_pem_not_ours = 2.44%; only_ours_not_pem = 6.25%; not_ours_not_pem = 2.74%\t\n\n===> ACE04; num mentions = 257\t\n [..............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 0ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 0/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : NYT20001123.1511.0062 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Fox Sports ==> ENTITY: \u001b[32mFox Sports (United States)\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.166:-0.166[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntelevision[0.047]; Wednesday[0.044]; Wednesday[0.044]; Wednesday[0.044]; ABC[0.043]; commercials[0.043]; commercials[0.043]; tournament[0.043]; week[0.043]; media[0.043]; Live[0.043]; featured[0.042]; Saturday[0.042]; day[0.041]; marketing[0.040]; Carey[0.040]; World[0.040]; Florida[0.040]; studio[0.040]; York[0.039]; Bank[0.039]; sponsored[0.039]; Drew[0.039]; sitcoms[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Volvo ==> ENTITY: \u001b[32mVolvo\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.160:-0.160[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncompany[0.050]; automotive[0.046]; drivers[0.045]; owned[0.045]; spinoff[0.045]; network[0.044]; inside[0.044]; commercials[0.044]; commercials[0.044]; buyers[0.044]; Communications[0.043]; Technologies[0.043]; introduce[0.043]; introduce[0.043]; Rover[0.043]; recent[0.042]; red[0.042]; Communication[0.042]; ad[0.042]; campaigns[0.042]; appealing[0.041]; Avaya[0.041]; played[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Times Square\u001b[39m ==> ENTITY: \u001b[32mTimes Square\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.112:0.112[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGreg[0.047]; Jack[0.047]; photograph[0.046]; magazine[0.045]; bus[0.045]; James[0.045]; television[0.045]; facility[0.044]; featured[0.044]; Roundabout[0.044]; commercials[0.043]; Theater[0.043]; Theater[0.043]; describing[0.043]; Levi[0.042]; described[0.042]; Playbills[0.042]; produced[0.042]; day[0.042]; posters[0.042]; moment[0.042]; label[0.042]; Tropicana[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Helen Hunt\u001b[39m ==> ENTITY: \u001b[32mHelen Hunt\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.091:0.091[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.047]; James[0.045]; featured[0.043]; Theater[0.043]; Theater[0.043]; Jack[0.043]; gets[0.042]; produced[0.042]; day[0.042]; article[0.041]; describing[0.041]; Nicholson[0.041]; Playbills[0.041]; recent[0.041]; packed[0.041]; Company[0.041]; consulted[0.040]; issue[0.040]; Greg[0.040]; Digest[0.040]; Digest[0.040]; good[0.040]; great[0.040]; began[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = MSN Network\u001b[39m ==> ENTITY: \u001b[32mMSN\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.111:0.111[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmagazine[0.046]; featured[0.045]; spinoff[0.044]; television[0.044]; television[0.044]; commercials[0.043]; commercials[0.043]; ad[0.042]; Avaya[0.042]; Microsoft[0.042]; Digest[0.041]; Digest[0.041]; Sprint[0.041]; Sprint[0.041]; automotive[0.040]; Communication[0.040]; company[0.040]; Reader[0.040]; ya[0.040]; Arthur[0.040]; advertisers[0.040]; recent[0.040]; appealing[0.040]; tells[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = CBS ==> ENTITY: \u001b[32mCBS\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.138:0.138[0]; log p(e|m)= -0.108:-0.108[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbroadcast[0.051]; broadcast[0.051]; played[0.043]; episode[0.043]; episode[0.043]; shows[0.043]; network[0.042]; Bowl[0.042]; Bowl[0.042]; featuring[0.042]; John[0.041]; longtime[0.041]; promotional[0.040]; Houston[0.040]; star[0.040]; entertainment[0.040]; Survivor[0.040]; Survivor[0.040]; Brooklyn[0.040]; recent[0.040]; columnist[0.039]; calling[0.039]; sell[0.039]; City[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reader's Digest magazine\u001b[39m ==> ENTITY: \u001b[32mReader's Digest\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.105:0.105[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDigest[0.048]; featured[0.047]; American[0.047]; Times[0.047]; commercials[0.046]; television[0.045]; ad[0.044]; ad[0.044]; Reader[0.044]; MSN[0.044]; Empire[0.043]; advertisers[0.043]; tells[0.042]; James[0.042]; Company[0.042]; automotive[0.042]; campaigns[0.042]; Square[0.041]; photograph[0.041]; Playbills[0.041]; network[0.041]; placebo[0.041]; pocket[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = James L. Brooks\u001b[39m ==> ENTITY: \u001b[32mJames L. Brooks\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.093:0.093[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfeatured[0.045]; produced[0.044]; Playbills[0.044]; Hunt[0.043]; Jack[0.043]; gets[0.042]; coincidence[0.042]; Greg[0.042]; Theater[0.042]; Theater[0.042]; Company[0.042]; consulted[0.042]; calling[0.042]; moment[0.041]; Nicholson[0.041]; Reader[0.041]; Reader[0.041]; trends[0.041]; American[0.041]; Roundabout[0.040]; day[0.040]; Digest[0.040]; Digest[0.040]; Helen[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = PepsiCo\u001b[39m ==> ENTITY: \u001b[32mPepsiCo\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCoffee[0.045]; juice[0.045]; ad[0.044]; ad[0.044]; ad[0.044]; Pure[0.044]; American[0.043]; Tropicana[0.043]; coffee[0.043]; Soup[0.042]; recent[0.042]; produced[0.041]; Cream[0.041]; slogan[0.041]; peanuts[0.041]; orange[0.040]; division[0.040]; Company[0.040]; box[0.040]; magazine[0.040]; line[0.040]; Premium[0.040]; began[0.040]; day[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sun Microsystems\u001b[39m ==> ENTITY: \u001b[32mSun Microsystems\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.014:0.014[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSports[0.043]; recent[0.043]; employee[0.043]; commercial[0.043]; late[0.042]; advantage[0.042]; sponsored[0.042]; advertisement[0.042]; refer[0.042]; tournament[0.042]; ear[0.042]; customer[0.042]; box[0.041]; Scotties[0.041]; Scotties[0.041]; BIG[0.041]; Thanksgiving[0.041]; Thanksgiving[0.041]; Investments[0.041]; television[0.041]; Doesn[0.041]; featured[0.041]; World[0.041]; Citibank[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Microsoft\u001b[39m ==> ENTITY: \u001b[32mMicrosoft\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.077:0.077[0]; log p(e|m)= -0.007:-0.007[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnetwork[0.044]; disparate[0.044]; ad[0.044]; research[0.043]; featured[0.043]; commercials[0.043]; commercials[0.043]; company[0.043]; MSN[0.043]; information[0.043]; zeitgeist[0.042]; zeitgeist[0.042]; Avaya[0.041]; television[0.040]; television[0.040]; Digest[0.040]; Reader[0.040]; recent[0.040]; led[0.040]; led[0.040]; Blue[0.040]; Blue[0.040]; introduce[0.040]; introduce[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Roundabout Theater Company\u001b[39m ==> ENTITY: \u001b[32mRoundabout Theatre Company\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.112:0.112[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTheater[0.051]; featured[0.046]; James[0.045]; Jack[0.044]; Hunt[0.043]; Playbills[0.043]; American[0.043]; Brooks[0.042]; produced[0.042]; Nicholson[0.042]; Helen[0.041]; Kinnear[0.041]; Square[0.041]; Campbell[0.040]; Campbell[0.040]; television[0.040]; Times[0.040]; pocket[0.040]; trends[0.040]; Greg[0.039]; division[0.039]; Soup[0.039]; ushers[0.039]; Cream[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jack Nicholson\u001b[39m ==> ENTITY: \u001b[32mJack Nicholson\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.093:0.093[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJames[0.047]; Hunt[0.045]; Playbills[0.044]; featured[0.044]; Greg[0.043]; Theater[0.043]; Theater[0.043]; American[0.043]; Helen[0.042]; produced[0.041]; describing[0.041]; magazine[0.041]; Company[0.041]; Brooks[0.041]; gets[0.041]; described[0.040]; trends[0.040]; Campbell[0.040]; Campbell[0.040]; division[0.040]; pair[0.040]; headline[0.040]; ushers[0.039]; opposite[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brooklyn ==> ENTITY: \u001b[32mBrooklyn\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.029:0.029[0]; log p(e|m)= -0.159:-0.159[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBowl[0.046]; Bowl[0.046]; opera[0.044]; chorus[0.044]; featuring[0.043]; serenades[0.043]; Opera[0.042]; fantastic[0.042]; shows[0.042]; someday[0.042]; mailing[0.041]; lot[0.041]; long[0.041]; headlines[0.041]; entertainment[0.041]; dressed[0.041]; Super[0.041]; Super[0.041]; Great[0.040]; broadcast[0.040]; broadcast[0.040]; Grand[0.040]; tell[0.040]; cattle[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Levi ==> ENTITY: \u001b[32mLevi Strauss & Co.\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.129:0.129[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \njeans[0.048]; commercials[0.045]; commercials[0.045]; American[0.044]; magazine[0.044]; roads[0.043]; featured[0.043]; ad[0.043]; Volvo[0.043]; label[0.041]; facility[0.041]; campaigns[0.041]; television[0.041]; automotive[0.040]; pants[0.040]; pair[0.040]; trouble[0.040]; word[0.040]; time[0.040]; time[0.040]; introduce[0.040]; introduce[0.040]; worn[0.039]; Cross[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = John Corbett ==> ENTITY: \u001b[32mJohn Corbett (actor)\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.092:0.092[0]; log p(e|m)= -0.071:-0.071[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nepisode[0.043]; episode[0.043]; CBS[0.043]; PCS[0.043]; Arthur[0.043]; radio[0.043]; star[0.042]; longtime[0.042]; longtime[0.042]; played[0.042]; television[0.042]; entertainment[0.042]; shows[0.042]; Avaya[0.041]; featuring[0.041]; cattle[0.041]; Aidan[0.041]; Grand[0.041]; introduce[0.041]; company[0.041]; Sprint[0.041]; Sprint[0.041]; recent[0.041]; Aiden[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mReader's Digest Association\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mReader's Digest\u001b[39m <---> \u001b[32mReader's Digest Association\u001b[39m\t\nSCORES: global= 0.267:0.262[\u001b[31m0.005\u001b[39m]; local()= 0.127:0.100[\u001b[31m0.027\u001b[39m]; log p(e|m)= -2.397:-0.095[\u001b[32m2.301\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmagazine[0.053]; Digest[0.045]; featured[0.045]; American[0.044]; Times[0.044]; commercials[0.043]; television[0.042]; Reader[0.042]; ad[0.042]; ad[0.042]; division[0.041]; MSN[0.041]; produced[0.041]; Empire[0.040]; advertisers[0.040]; PepsiCo[0.040]; tells[0.040]; James[0.040]; Company[0.040]; network[0.040]; consumers[0.040]; Greg[0.039]; Square[0.039]; photograph[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Greg Kinnear\u001b[39m ==> ENTITY: \u001b[32mGreg Kinnear\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.106:0.106[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJack[0.045]; James[0.045]; Hunt[0.044]; featured[0.043]; Theater[0.043]; Theater[0.043]; gets[0.042]; Helen[0.042]; recent[0.041]; Company[0.041]; Playbills[0.041]; Nicholson[0.041]; began[0.041]; magazine[0.041]; magazine[0.041]; opposite[0.041]; American[0.041]; Campbell[0.041]; Campbell[0.041]; Reader[0.041]; Reader[0.041]; Real[0.040]; trends[0.040]; Brooks[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = ABC ==> ENTITY: \u001b[32mAmerican Broadcasting Company\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.334:-0.334[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFox[0.052]; studio[0.049]; Sports[0.048]; commercials[0.047]; Hollywood[0.047]; sitcoms[0.047]; featured[0.046]; World[0.046]; movie[0.046]; Wednesday[0.045]; Wednesday[0.045]; Wednesday[0.045]; week[0.045]; Spin[0.045]; News[0.044]; tournament[0.044]; title[0.044]; City[0.043]; time[0.043]; Carey[0.043]; day[0.043]; Drew[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fidelity Investments\u001b[39m ==> ENTITY: \u001b[32mFidelity Investments\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.091:0.091[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBank[0.046]; campaign[0.044]; employee[0.044]; sponsored[0.044]; Citibank[0.043]; Sports[0.042]; commercial[0.042]; commercials[0.042]; advertisement[0.042]; customer[0.042]; City[0.041]; World[0.041]; tournament[0.041]; ear[0.041]; Scotties[0.041]; Scotties[0.041]; BIG[0.041]; Wednesday[0.041]; ad[0.041]; recent[0.040]; telephone[0.040]; Doesn[0.040]; Microsystems[0.040]; newspapers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Land Rover\u001b[39m ==> ENTITY: \u001b[32mLand Rover\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.127:0.127[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVolvo[0.046]; company[0.044]; research[0.043]; Shield[0.043]; owned[0.042]; automotive[0.042]; commercials[0.042]; commercials[0.042]; spinoff[0.042]; Blue[0.042]; Blue[0.042]; campaigns[0.042]; drivers[0.042]; disclose[0.041]; buyers[0.041]; tests[0.041]; inside[0.041]; Empire[0.041]; network[0.040]; time[0.040]; time[0.040]; time[0.040]; recent[0.040]; featured[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Empire Blue Cross Blue Shield ==> ENTITY: \u001b[32mBlue Cross Blue Shield Association\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.060:0.060[0]; log p(e|m)= -0.080:-0.080[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncompany[0.047]; Association[0.044]; disclose[0.043]; Avaya[0.042]; MSN[0.042]; offer[0.042]; network[0.042]; research[0.042]; Sprint[0.042]; recent[0.042]; commercials[0.041]; commercials[0.041]; spinoff[0.041]; introduce[0.041]; introduce[0.041]; automotive[0.041]; Communication[0.041]; Communications[0.041]; ya[0.041]; ad[0.041]; campaigns[0.040]; placebo[0.040]; pocket[0.040]; tab[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Houston Grand Opera\u001b[39m ==> ENTITY: \u001b[32mHouston Grand Opera\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.062:0.062[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nopera[0.048]; Arthur[0.046]; played[0.044]; PCS[0.043]; chorus[0.043]; John[0.042]; City[0.042]; Brooklyn[0.042]; neck[0.042]; Avaya[0.041]; shows[0.041]; cattle[0.041]; featuring[0.041]; star[0.041]; Sprint[0.041]; Sprint[0.041]; hear[0.040]; time[0.040]; serenades[0.040]; headlines[0.040]; spinoff[0.040]; Communication[0.040]; recent[0.040]; character[0.040]; \t\n\r [===>..........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 918ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 27/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 20001115_AFP_ARB.0210.eng ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.056:0.056[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFlorida[0.072]; Wednesday[0.072]; count[0.071]; today[0.071]; state[0.070]; counties[0.069]; votes[0.066]; refused[0.066]; hand[0.065]; Court[0.065]; authorities[0.064]; stop[0.063]; Supreme[0.062]; Tallahassee[0.062]; AFP[0.061]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = AFP ==> ENTITY: \u001b[32mAgence France-Presse\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.022:0.022[0]; log p(e|m)= -0.660:-0.660[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \napplication[0.066]; Wednesday[0.064]; authorities[0.063]; state[0.063]; today[0.061]; Court[0.060]; States[0.059]; votes[0.059]; counties[0.059]; refused[0.059]; stop[0.058]; United[0.058]; count[0.056]; Supreme[0.055]; Florida[0.054]; hand[0.053]; Tallahassee[0.053]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Florida ==> ENTITY: \u001b[32mFlorida\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.064:0.064[0]; log p(e|m)= -0.154:-0.154[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncount[0.069]; Tallahassee[0.065]; state[0.062]; counties[0.062]; Wednesday[0.061]; States[0.061]; today[0.059]; votes[0.059]; stop[0.058]; application[0.058]; authorities[0.057]; AFP[0.057]; Supreme[0.056]; United[0.054]; Court[0.054]; hand[0.054]; refused[0.054]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tallahassee ==> ENTITY: \u001b[32mTallahassee, Florida\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.076:0.076[0]; log p(e|m)= -0.024:-0.024[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFlorida[0.067]; state[0.063]; Court[0.062]; Wednesday[0.062]; United[0.061]; States[0.060]; today[0.060]; counties[0.060]; authorities[0.059]; stop[0.059]; refused[0.058]; application[0.057]; hand[0.056]; count[0.055]; votes[0.054]; Supreme[0.053]; AFP[0.053]; \t\n\r [====>.........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s47ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 32/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 20001115_AFP_ARB.0060.eng ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = AFP ==> ENTITY: \u001b[32mAgence France-Presse\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.107:0.107[0]; log p(e|m)= -0.660:-0.660[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nofficial[0.045]; official[0.045]; American[0.045]; American[0.045]; American[0.045]; does[0.043]; does[0.043]; confirmed[0.042]; high[0.042]; level[0.042]; capability[0.042]; announced[0.042]; include[0.041]; announcement[0.041]; Wednesday[0.041]; today[0.041]; contain[0.040]; Russian[0.040]; Russian[0.040]; wake[0.040]; Bill[0.040]; President[0.040]; President[0.040]; meeting[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sultanate of Brunei\u001b[39m ==> ENTITY: \u001b[32mBrunei\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.125:0.125[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAsia[0.049]; Begawan[0.046]; Seri[0.045]; strategic[0.045]; Bandar[0.044]; contain[0.044]; current[0.044]; States[0.044]; important[0.043]; defensive[0.043]; proposed[0.043]; proposed[0.043]; number[0.043]; number[0.043]; presidents[0.043]; Pacific[0.043]; capital[0.043]; paid[0.043]; require[0.042]; APEC[0.042]; said[0.042]; include[0.042]; disparities[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iran ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.103:0.103[0]; log p(e|m)= -0.138:-0.138[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nofficial[0.044]; official[0.044]; administration[0.044]; threats[0.043]; discussions[0.043]; discussions[0.043]; Russia[0.042]; Russia[0.042]; issue[0.042]; American[0.042]; American[0.042]; said[0.041]; said[0.041]; said[0.041]; Bush[0.041]; sensitive[0.041]; strategic[0.041]; response[0.041]; States[0.041]; States[0.041]; States[0.041]; proposed[0.040]; possible[0.040]; continue[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bandar Seri Begawan\u001b[39m ==> ENTITY: \u001b[32mBandar Seri Begawan\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.072:0.072[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninclude[0.045]; today[0.044]; capability[0.044]; high[0.043]; examination[0.043]; require[0.043]; Wednesday[0.042]; official[0.042]; official[0.042]; AFP[0.042]; level[0.042]; disparities[0.042]; announced[0.041]; Vladimir[0.041]; meeting[0.041]; great[0.041]; paid[0.041]; reducing[0.041]; proposals[0.040]; attention[0.040]; confirmed[0.040]; announcement[0.040]; President[0.040]; President[0.040]; \t\n\r [=====>........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s228ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 36/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 20001115_AFP_ARB.0072.eng ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = AFP ==> ENTITY: \u001b[32mAgence France-Presse\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.158:0.158[0]; log p(e|m)= -0.660:-0.660[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninterview[0.045]; official[0.044]; newspaper[0.044]; Iraq[0.043]; Iraq[0.043]; French[0.042]; French[0.042]; French[0.042]; French[0.042]; France[0.042]; international[0.042]; decision[0.041]; decision[0.041]; decision[0.041]; said[0.041]; said[0.041]; payments[0.041]; Wednesday[0.041]; today[0.041]; today[0.041]; diplomat[0.040]; Baghdad[0.040]; Baghdad[0.040]; welcomed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.135:0.135[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.052]; Baghdad[0.046]; Baghdad[0.046]; Ambassador[0.044]; countries[0.043]; diplomat[0.043]; international[0.042]; al[0.042]; relations[0.041]; relations[0.041]; France[0.041]; economic[0.041]; said[0.041]; said[0.041]; official[0.041]; European[0.041]; welcomed[0.041]; trade[0.040]; Interests[0.040]; AFP[0.039]; Europeans[0.039]; paid[0.039]; Wednesday[0.039]; interview[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baghdad\u001b[39m ==> ENTITY: \u001b[32mBaghdad\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.049]; Iraq[0.049]; Baghdad[0.049]; French[0.047]; French[0.047]; French[0.047]; French[0.047]; al[0.046]; France[0.045]; official[0.045]; Wednesday[0.042]; today[0.041]; today[0.041]; trade[0.041]; international[0.041]; diplomat[0.041]; AFP[0.040]; edition[0.040]; said[0.040]; said[0.040]; Ambassador[0.040]; euros[0.040]; economic[0.040]; \t\n\r [=====>........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s434ms | Step: 6ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 39/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 20001115_AFP_ARB.0013.eng ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAFP\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustralian Federal Police\u001b[39m <---> \u001b[32mAgence France-Presse\u001b[39m\t\nSCORES: global= 0.240:0.239[\u001b[31m0.001\u001b[39m]; local()= 0.037:0.062[\u001b[32m0.025\u001b[39m]; log p(e|m)= -2.375:-0.660[\u001b[32m1.715\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntold[0.045]; Economic[0.044]; Wednesday[0.043]; President[0.043]; President[0.043]; group[0.043]; American[0.042]; presidential[0.042]; forum[0.042]; money[0.042]; today[0.041]; maintain[0.041]; Governor[0.041]; Bill[0.041]; Clinton[0.041]; Clinton[0.041]; Gore[0.041]; elections[0.041]; elections[0.041]; Democrats[0.041]; issues[0.040]; Vice[0.040]; State[0.040]; Republicans[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = State of Florida\u001b[39m ==> ENTITY: \u001b[32mFlorida\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.038:0.038[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGovernor[0.044]; Bill[0.044]; Begawan[0.044]; order[0.043]; President[0.043]; President[0.043]; States[0.042]; States[0.042]; today[0.042]; Wednesday[0.042]; elections[0.042]; elections[0.042]; committed[0.042]; presidential[0.041]; American[0.041]; American[0.041]; integration[0.041]; George[0.041]; following[0.041]; Asia[0.041]; Pacific[0.040]; maintain[0.040]; markets[0.040]; Economic[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brunei ==> ENTITY: \u001b[32mBrunei\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.387:-0.387[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAsia[0.045]; States[0.043]; States[0.043]; States[0.043]; Cooperation[0.043]; Wednesday[0.043]; today[0.043]; Begawan[0.042]; Tuesday[0.042]; State[0.042]; Seri[0.042]; following[0.042]; President[0.041]; President[0.041]; world[0.041]; world[0.041]; group[0.041]; Bandar[0.041]; Governor[0.041]; order[0.040]; achieve[0.040]; final[0.040]; Economic[0.040]; trade[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bandar Seri Begawan\u001b[39m ==> ENTITY: \u001b[32mBandar Seri Begawan\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.102:0.102[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrunei[0.051]; Asia[0.044]; Governor[0.044]; State[0.044]; today[0.044]; Wednesday[0.042]; Al[0.042]; Pacific[0.042]; AFP[0.042]; forum[0.042]; markets[0.042]; money[0.042]; Economic[0.041]; presidential[0.041]; maintain[0.040]; businessmen[0.040]; Florida[0.040]; President[0.040]; President[0.040]; States[0.040]; need[0.040]; group[0.040]; American[0.039]; told[0.039]; \t\n\r [======>.......................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s551ms | Step: 7ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 43/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 20001015_AFP_ARB.0229.eng ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = AFP ==> ENTITY: \u001b[32mAgence France-Presse\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.660:-0.660[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.045]; yesterday[0.043]; yesterday[0.043]; operation[0.043]; Jordanian[0.042]; spokesman[0.042]; announced[0.042]; members[0.042]; Saudi[0.042]; Saudi[0.042]; Saudi[0.042]; Saudi[0.042]; Saudi[0.042]; said[0.042]; space[0.041]; International[0.041]; local[0.041]; London[0.041]; London[0.041]; London[0.041]; Baghdad[0.040]; Baghdad[0.040]; Sunday[0.040]; Sunday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baghdad\u001b[39m ==> ENTITY: \u001b[32mBaghdad\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBaghdad[0.047]; Jordanian[0.045]; Riyadh[0.043]; Riyadh[0.043]; Saudi[0.042]; Saudi[0.042]; Saudi[0.042]; Saudi[0.042]; Saudi[0.042]; Airport[0.042]; Airways[0.042]; Airways[0.042]; Saudis[0.041]; Pakistanis[0.041]; British[0.041]; local[0.041]; members[0.040]; operation[0.040]; capital[0.040]; announced[0.040]; American[0.039]; Sunday[0.039]; Sunday[0.039]; Yemenis[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Riyadh\u001b[39m ==> ENTITY: \u001b[32mRiyadh\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRiyadh[0.046]; Saudi[0.043]; Saudi[0.043]; Saudi[0.043]; Saudi[0.043]; Saudi[0.043]; Airport[0.043]; Jordanian[0.043]; Saudis[0.042]; Boeing[0.042]; Boeing[0.042]; Yemenis[0.042]; American[0.041]; air[0.041]; Pakistanis[0.041]; London[0.040]; London[0.040]; London[0.040]; British[0.040]; hijackers[0.040]; Airways[0.040]; Airways[0.040]; yesterday[0.040]; yesterday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = RIYADH INTERNATIONAL AIRPORT\u001b[39m ==> ENTITY: \u001b[32mKing Khalid International Airport\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRiyadh[0.045]; Saudi[0.044]; Saudi[0.044]; Saudi[0.044]; Saudi[0.044]; Saudi[0.044]; Jordanian[0.043]; Boeing[0.043]; Boeing[0.043]; air[0.042]; Saudis[0.042]; Airways[0.041]; Airways[0.041]; Omani[0.041]; capital[0.041]; announced[0.041]; space[0.040]; operation[0.040]; plane[0.040]; approximately[0.040]; British[0.040]; Yemenis[0.039]; passengers[0.039]; passengers[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.072:0.072[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.049]; London[0.049]; Airport[0.046]; operation[0.044]; said[0.042]; passengers[0.041]; passengers[0.041]; Boeing[0.041]; Boeing[0.041]; Sunday[0.041]; Sunday[0.041]; announced[0.041]; hours[0.041]; yesterday[0.041]; yesterday[0.041]; Airways[0.040]; Airways[0.040]; air[0.040]; began[0.040]; Saturday[0.040]; Saturday[0.040]; members[0.040]; time[0.040]; Jordanian[0.039]; \t\n\r [=======>......................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s622ms | Step: 7ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 48/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 20001115_AFP_ARB.0065.eng ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mHolland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNetherlands\u001b[39m <---> \u001b[32mHolland\u001b[39m\t\nSCORES: global= 0.251:0.247[\u001b[31m0.004\u001b[39m]; local()= 0.085:0.098[\u001b[32m0.013\u001b[39m]; log p(e|m)= -2.781:-0.631[\u001b[32m2.150\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHolland[0.051]; Zeist[0.046]; Zeist[0.046]; Richard[0.043]; having[0.043]; lived[0.043]; Keen[0.043]; Wednesday[0.042]; time[0.042]; serving[0.041]; charge[0.041]; Scottish[0.041]; arose[0.041]; faced[0.041]; life[0.040]; Friday[0.040]; Today[0.040]; continue[0.040]; deal[0.039]; law[0.039]; Camp[0.039]; Camp[0.039]; court[0.039]; Union[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Soviet Union ==> ENTITY: \u001b[32mSoviet Union\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.069:0.069[0]; log p(e|m)= -0.076:-0.076[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHolland[0.044]; Holland[0.044]; Sweden[0.044]; time[0.044]; professional[0.043]; serving[0.042]; considered[0.042]; months[0.042]; detailed[0.042]; terrorist[0.042]; force[0.042]; faced[0.041]; Zeist[0.041]; Zeist[0.041]; having[0.041]; involvement[0.041]; period[0.041]; followed[0.041]; Scottish[0.040]; Scottish[0.040]; deal[0.040]; Stockholm[0.040]; defense[0.040]; defense[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Malta ==> ENTITY: \u001b[32mMalta\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.084:0.084[0]; log p(e|m)= -0.392:-0.392[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.046]; principle[0.045]; Scotland[0.045]; case[0.044]; December[0.043]; people[0.042]; apply[0.042]; far[0.041]; sending[0.041]; defense[0.041]; short[0.041]; said[0.041]; Scottish[0.041]; situation[0.041]; attempt[0.041]; village[0.041]; session[0.041]; judges[0.041]; Libyans[0.041]; president[0.041]; appears[0.041]; important[0.040]; plane[0.040]; Khalifa[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Stockholm ==> ENTITY: \u001b[32mStockholm\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.100:0.100[0]; log p(e|m)= -0.104:-0.104[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.048]; Sweden[0.048]; Richard[0.043]; council[0.043]; lived[0.042]; case[0.042]; months[0.042]; acts[0.042]; Friday[0.041]; hearing[0.041]; period[0.041]; court[0.041]; court[0.041]; court[0.041]; court[0.041]; court[0.041]; court[0.041]; incidents[0.041]; reasons[0.041]; Copenhagen[0.040]; apply[0.040]; means[0.040]; problem[0.040]; testimony[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Copenhagen ==> ENTITY: \u001b[32mCopenhagen\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.076:0.076[0]; log p(e|m)= -0.119:-0.119[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.047]; Sweden[0.047]; Stockholm[0.046]; Richard[0.044]; considered[0.043]; having[0.042]; involvement[0.042]; detailed[0.041]; Scottish[0.041]; Scottish[0.041]; case[0.041]; reasons[0.041]; answer[0.041]; means[0.041]; Scotland[0.041]; Amsterdam[0.041]; far[0.041]; insists[0.040]; precedents[0.040]; attempting[0.040]; question[0.040]; question[0.040]; question[0.040]; apply[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = AFP ==> ENTITY: \u001b[32mAgence France-Presse\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.080:0.080[0]; log p(e|m)= -0.660:-0.660[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncharge[0.045]; subject[0.044]; detailed[0.043]; case[0.043]; Wednesday[0.043]; Palestinian[0.043]; continue[0.042]; Union[0.042]; Today[0.042]; defense[0.042]; serving[0.042]; time[0.041]; Friday[0.041]; refused[0.041]; lawyer[0.041]; court[0.041]; faced[0.041]; deal[0.041]; having[0.041]; law[0.041]; Richard[0.041]; refusing[0.041]; Lockerbie[0.040]; hearing[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Amsterdam ==> ENTITY: \u001b[32mAmsterdam\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.075:0.075[0]; log p(e|m)= -0.064:-0.064[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRichard[0.044]; asked[0.044]; Sweden[0.044]; Sweden[0.044]; lawyer[0.043]; lawyer[0.043]; Lockerbie[0.043]; Lockerbie[0.043]; Stockholm[0.042]; answer[0.042]; Union[0.041]; lived[0.041]; old[0.041]; Friday[0.041]; Scotland[0.041]; case[0.041]; detailed[0.041]; insists[0.041]; Scottish[0.040]; Scottish[0.040]; responding[0.040]; reasons[0.040]; life[0.040]; Copenhagen[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.111:0.111[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.053]; Scotland[0.045]; Stockholm[0.044]; professional[0.043]; force[0.043]; considered[0.042]; asked[0.042]; having[0.042]; months[0.041]; means[0.041]; period[0.041]; defense[0.041]; defense[0.041]; Union[0.041]; lived[0.040]; followed[0.040]; arose[0.040]; judges[0.040]; judges[0.040]; reasons[0.040]; Copenhagen[0.040]; aimed[0.040]; Friday[0.040]; refused[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCamp Zeist\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mScottish Court in the Netherlands\u001b[39m <---> \u001b[32mZeist\u001b[39m\t\nSCORES: global= 0.267:0.252[\u001b[31m0.015\u001b[39m]; local()= 0.149:0.099[\u001b[31m0.050\u001b[39m]; log p(e|m)= 0.000:-2.408[\u001b[31m2.408\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZeist[0.052]; case[0.046]; court[0.045]; Lockerbie[0.044]; hearing[0.044]; Scottish[0.043]; sentence[0.043]; judges[0.043]; Talib[0.042]; Holland[0.042]; Holland[0.042]; witness[0.041]; Friday[0.041]; charge[0.041]; Wednesday[0.040]; serving[0.040]; testimony[0.040]; law[0.039]; Camp[0.039]; defense[0.039]; refused[0.039]; arose[0.039]; subject[0.038]; AFP[0.038]; \t\n\r [=========>....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s458ms | Step: 7ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 59/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : NYT20001129.2040.0383 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fla.\u001b[39m ==> ENTITY: \u001b[32mFlorida\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.004:0.004[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFlorida[0.056]; Florida[0.056]; UR[0.053]; UR[0.053]; appeal[0.051]; appeal[0.051]; court[0.051]; state[0.050]; Court[0.050]; Court[0.050]; Court[0.050]; Service[0.049]; Appeal[0.048]; filed[0.048]; election[0.047]; ART[0.047]; LR[0.047]; LR[0.047]; LR[0.047]; ADV[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = District Court of Appeal\u001b[39m ==> ENTITY: \u001b[32mFlorida District Courts of Appeal\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCourt[0.044]; Court[0.044]; UR[0.044]; UR[0.044]; court[0.043]; court[0.043]; Judge[0.042]; legal[0.042]; lawyers[0.042]; lawyers[0.042]; appeal[0.042]; appeal[0.042]; Lawyers[0.041]; Supreme[0.041]; Supreme[0.041]; Circuit[0.041]; order[0.041]; filed[0.041]; state[0.040]; disputed[0.040]; lawsuit[0.040]; Florida[0.040]; Florida[0.040]; highest[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Electoral College ==> ENTITY: \u001b[32mElectoral College (United States)\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.165:-0.165[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nelectors[0.044]; election[0.043]; election[0.043]; Gore[0.043]; Gore[0.043]; Gore[0.043]; Gore[0.043]; Gore[0.043]; Gore[0.043]; office[0.042]; won[0.042]; Florida[0.042]; counting[0.041]; counting[0.041]; Friday[0.041]; state[0.041]; count[0.041]; Wednesday[0.041]; Kendall[0.041]; Saturday[0.040]; disputed[0.040]; ballots[0.040]; ballots[0.040]; ballots[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tallahassee ==> ENTITY: \u001b[32mTallahassee, Florida\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.020:0.020[0]; log p(e|m)= -0.024:-0.024[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUR[0.047]; UR[0.047]; Florida[0.044]; Florida[0.044]; ART[0.041]; LR[0.041]; LR[0.041]; LR[0.041]; ADV[0.041]; QL[0.041]; QL[0.041]; QL[0.041]; QL[0.041]; QL[0.041]; late[0.041]; state[0.041]; Wednesday[0.041]; Wednesday[0.041]; York[0.041]; Thursday[0.041]; began[0.040]; begin[0.040]; time[0.040]; early[0.040]; \t\n\r [==========>...................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s497ms | Step: 7ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 64/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : NYT20001106.1705.0187 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tenn.\u001b[39m ==> ENTITY: \u001b[32mTennessee\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.038:0.038[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennessee[0.048]; UR[0.046]; UR[0.046]; Arkansas[0.044]; Washington[0.043]; states[0.042]; months[0.042]; issued[0.042]; ART[0.041]; LR[0.041]; LR[0.041]; ADV[0.041]; QL[0.041]; QL[0.041]; QL[0.041]; QL[0.041]; George[0.041]; Monday[0.040]; presidency[0.040]; EDS[0.040]; Gov[0.040]; nearly[0.040]; Iowa[0.039]; early[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Arkansas ==> ENTITY: \u001b[32mArkansas\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.102:0.102[0]; log p(e|m)= -0.320:-0.320[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.046]; Tennessee[0.046]; Texas[0.044]; trail[0.044]; Monday[0.043]; Washington[0.043]; nearly[0.042]; states[0.042]; ahead[0.042]; final[0.042]; start[0.041]; country[0.041]; country[0.041]; George[0.041]; symbolism[0.041]; White[0.041]; Clinton[0.041]; campaign[0.040]; clear[0.040]; governor[0.040]; divides[0.040]; months[0.040]; placed[0.040]; pledging[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = D.C. ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.158:-0.158[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWashington[0.047]; Washington[0.047]; Washington[0.047]; state[0.047]; government[0.045]; federal[0.044]; federal[0.044]; country[0.043]; win[0.042]; win[0.042]; American[0.042]; governor[0.041]; Tuesday[0.040]; carry[0.040]; certain[0.040]; political[0.040]; added[0.039]; home[0.039]; Tennessee[0.039]; Tennessee[0.039]; come[0.039]; decades[0.038]; spending[0.038]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Michigan ==> ENTITY: \u001b[32mMichigan\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.052:0.052[0]; log p(e|m)= -0.248:-0.248[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nvotes[0.044]; states[0.043]; president[0.043]; president[0.043]; president[0.043]; campaign[0.043]; Tennessee[0.042]; vice[0.042]; win[0.042]; campaigning[0.042]; victory[0.041]; victory[0.041]; visiting[0.041]; voters[0.041]; Iowa[0.041]; Bush[0.041]; Bush[0.041]; Bush[0.041]; Bush[0.041]; Bush[0.041]; Wisconsin[0.041]; Monday[0.041]; Election[0.041]; Pennsylvania[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iowa ==> ENTITY: \u001b[32mIowa\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.355:-0.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstates[0.047]; voters[0.046]; votes[0.046]; campaign[0.045]; Wisconsin[0.044]; Michigan[0.044]; Arkansas[0.044]; Tennessee[0.044]; appeal[0.044]; Gore[0.043]; battlegrounds[0.043]; supporters[0.043]; partisan[0.043]; Bush[0.043]; Bush[0.043]; Bush[0.043]; Bush[0.043]; nation[0.043]; nation[0.043]; nation[0.043]; Pennsylvania[0.043]; Monday[0.042]; Election[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pennsylvania ==> ENTITY: \u001b[32mPennsylvania\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.047:0.047[0]; log p(e|m)= -0.118:-0.118[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncampaign[0.046]; votes[0.045]; voters[0.045]; tossups[0.044]; campaigning[0.043]; supporters[0.043]; win[0.043]; sought[0.041]; Election[0.041]; president[0.041]; president[0.041]; president[0.041]; Monday[0.041]; states[0.041]; Bush[0.041]; Bush[0.041]; Bush[0.041]; Bush[0.041]; Bush[0.041]; Tennessee[0.040]; nation[0.040]; key[0.040]; victory[0.040]; victory[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.102:0.102[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.045]; George[0.045]; campaign[0.044]; President[0.043]; President[0.043]; rally[0.042]; country[0.042]; governor[0.042]; states[0.042]; nearly[0.042]; final[0.042]; Clinton[0.042]; pledging[0.041]; Tennessee[0.041]; early[0.041]; Monday[0.041]; major[0.041]; start[0.040]; Gov[0.040]; planned[0.040]; home[0.040]; Service[0.040]; sent[0.040]; left[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wisconsin ==> ENTITY: \u001b[32mWisconsin\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncampaign[0.055]; Michigan[0.054]; Iowa[0.054]; Tennessee[0.053]; schedule[0.053]; Bush[0.053]; Bush[0.053]; Bush[0.053]; Bush[0.053]; Election[0.053]; tossups[0.053]; Arkansas[0.053]; nation[0.053]; nation[0.053]; nation[0.053]; states[0.052]; Monday[0.051]; Day[0.051]; Pennsylvania[0.051]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = White House\u001b[39m ==> ENTITY: \u001b[32mWhite House\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.006:-0.006[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBush[0.045]; Bush[0.045]; Bush[0.045]; President[0.044]; President[0.044]; campaign[0.043]; Clinton[0.042]; Tennessee[0.042]; president[0.041]; president[0.041]; president[0.041]; nation[0.041]; nation[0.041]; nation[0.041]; country[0.041]; country[0.041]; decision[0.041]; states[0.041]; Americans[0.040]; major[0.040]; trail[0.040]; said[0.040]; people[0.039]; people[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Green Bay ==> ENTITY: \u001b[32mGreen Bay, Wisconsin\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.598:-0.598[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncenter[0.051]; visiting[0.050]; Monday[0.050]; win[0.049]; American[0.049]; point[0.048]; come[0.048]; tell[0.048]; tell[0.048]; believe[0.048]; business[0.048]; going[0.047]; predicted[0.047]; confidence[0.047]; convention[0.046]; want[0.046]; victory[0.046]; Chattanooga[0.046]; comes[0.046]; pleading[0.046]; make[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chattanooga ==> ENTITY: \u001b[32mChattanooga, Tennessee\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.032:0.032[0]; log p(e|m)= -0.168:-0.168[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUR[0.046]; UR[0.046]; Tennessee[0.046]; Monday[0.043]; rally[0.042]; season[0.042]; Washington[0.042]; Arkansas[0.041]; ART[0.041]; LR[0.041]; LR[0.041]; early[0.041]; ADV[0.041]; QL[0.041]; QL[0.041]; QL[0.041]; QL[0.041]; sent[0.041]; campaign[0.041]; George[0.041]; Davenport[0.040]; nearly[0.040]; EDS[0.040]; final[0.040]; \t\n\r [============>.................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s467ms | Step: 8ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 75/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 20000815_AFP_ARB.0071.eng ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = AFP ==> ENTITY: \u001b[32mAgence France-Presse\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.126:0.126[0]; log p(e|m)= -0.660:-0.660[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nreports[0.045]; official[0.044]; According[0.044]; Lebanese[0.043]; Lebanese[0.043]; Lebanese[0.043]; shared[0.042]; operations[0.042]; international[0.041]; confirmed[0.041]; high[0.041]; authorities[0.041]; Lebanon[0.041]; force[0.041]; source[0.041]; Beirut[0.041]; investigate[0.041]; command[0.041]; contacted[0.041]; Israel[0.041]; Israel[0.041]; indicating[0.041]; added[0.040]; territory[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanon ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.189:0.189[0]; log p(e|m)= -0.409:-0.409[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBeirut[0.048]; Lebanese[0.048]; Lebanese[0.048]; Lebanese[0.048]; Israel[0.047]; Israel[0.047]; Israeli[0.044]; international[0.042]; countries[0.042]; sovereignty[0.041]; territory[0.041]; official[0.041]; border[0.040]; force[0.040]; peacekeeping[0.040]; peacekeeping[0.040]; south[0.039]; According[0.039]; building[0.039]; high[0.038]; operations[0.038]; said[0.037]; indicating[0.037]; added[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.178:0.178[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.054]; Israeli[0.049]; Lebanon[0.047]; countries[0.045]; international[0.044]; Lebanese[0.043]; Lebanese[0.043]; Lebanese[0.043]; Lebanese[0.043]; Beirut[0.041]; official[0.041]; today[0.040]; operations[0.040]; Tuesday[0.040]; territory[0.039]; said[0.039]; added[0.039]; sovereignty[0.039]; confirmed[0.039]; force[0.039]; border[0.039]; asked[0.039]; asked[0.039]; indicating[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Beirut ==> ENTITY: \u001b[32mBeirut\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.047:-0.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.051]; Israel[0.048]; Israel[0.048]; Lebanese[0.048]; Lebanese[0.048]; Lebanese[0.048]; force[0.042]; international[0.041]; territory[0.041]; border[0.041]; official[0.040]; peacekeeping[0.040]; peacekeeping[0.040]; south[0.040]; AFP[0.039]; sovereignty[0.039]; authorities[0.038]; operations[0.038]; According[0.038]; carrying[0.038]; added[0.038]; command[0.038]; confirmed[0.038]; countries[0.038]; \t\n\r [=============>................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s500ms | Step: 8ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 79/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : VOA20001220.2000.0060 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = America ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.082:0.082[0]; log p(e|m)= -0.785:-0.785[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNeil[0.049]; America[0.046]; David[0.046]; George[0.044]; Paul[0.043]; Mel[0.043]; California[0.043]; abroad[0.041]; Don[0.041]; Evans[0.041]; global[0.041]; Texas[0.040]; says[0.040]; Secretary[0.040]; Secretary[0.040]; Secretary[0.040]; Secretary[0.040]; Secretary[0.040]; Thursday[0.040]; Governor[0.040]; Governor[0.040]; Wednesday[0.040]; Florida[0.040]; business[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Texas ==> ENTITY: \u001b[32mTexas\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.059:0.059[0]; log p(e|m)= -0.161:-0.161[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.044]; political[0.043]; America[0.043]; America[0.043]; priority[0.043]; Florida[0.043]; Thursday[0.043]; businesses[0.042]; Mel[0.042]; Bush[0.042]; Bush[0.042]; Bush[0.042]; place[0.041]; says[0.041]; advocate[0.041]; Secretary[0.041]; close[0.041]; open[0.041]; free[0.041]; free[0.041]; Elect[0.040]; Governor[0.040]; Governor[0.040]; named[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = VOA News\u001b[39m ==> ENTITY: \u001b[32mVoice of America\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.098:0.098[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerica[0.048]; America[0.048]; David[0.047]; abroad[0.047]; Wednesday[0.047]; Agriculture[0.046]; Agriculture[0.046]; Al[0.046]; free[0.046]; free[0.046]; Free[0.046]; Coler[0.045]; George[0.045]; promotion[0.045]; appointments[0.044]; Neil[0.044]; California[0.044]; Paul[0.044]; Koha[0.044]; Commerce[0.044]; President[0.044]; associate[0.043]; \t\n\r [=============>................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s500ms | Step: 8ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 82/257 \r\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Comcast\u001b[39m ==> ENTITY: \u001b[32mComcast\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncable[0.053]; deal[0.050]; deal[0.050]; limited[0.050]; media[0.049]; Warner[0.048]; Warner[0.048]; announced[0.048]; America[0.047]; Online[0.047]; company[0.047]; company[0.047]; combined[0.047]; combined[0.047]; executives[0.046]; advertisers[0.046]; said[0.046]; said[0.046]; said[0.046]; said[0.046]; negotiated[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Roger McNamee\u001b[39m ==> ENTITY: \u001b[32mRoger McNamee\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncompany[0.045]; announced[0.045]; Capital[0.043]; Capital[0.043]; media[0.043]; media[0.043]; John[0.042]; Tom[0.042]; bought[0.042]; companies[0.042]; companies[0.042]; sold[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; shares[0.041]; AOL[0.040]; AOL[0.040]; portfolio[0.040]; merged[0.040]; Integral[0.039]; combo[0.039]; manager[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Capital Cities/ABC\u001b[39m ==> ENTITY: \u001b[32mCapital Cities Communications\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.121:0.121[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDisney[0.048]; Disney[0.048]; company[0.042]; company[0.042]; media[0.042]; media[0.042]; distribution[0.042]; announced[0.042]; bought[0.042]; Strategically[0.041]; Warner[0.041]; Warner[0.041]; Warner[0.041]; recent[0.041]; merged[0.041]; Capital[0.041]; companies[0.041]; companies[0.041]; sold[0.041]; Integral[0.041]; AOL[0.040]; AOL[0.040]; struggling[0.040]; combo[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Merrill Lynch\u001b[39m ==> ENTITY: \u001b[32mMerrill Lynch\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCapital[0.045]; Capital[0.045]; Capital[0.045]; announced[0.045]; company[0.044]; company[0.044]; companies[0.044]; companies[0.044]; companies[0.044]; portfolio[0.044]; million[0.044]; fund[0.044]; shares[0.044]; shares[0.044]; deals[0.043]; merger[0.042]; merger[0.042]; largest[0.042]; bought[0.042]; John[0.041]; shareholder[0.041]; profit[0.041]; Washington[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Time Warner\u001b[39m ==> ENTITY: \u001b[32mTime Warner\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNews[0.045]; Warner[0.044]; Warner[0.044]; Warner[0.044]; Warner[0.044]; Warner[0.044]; Online[0.044]; Online[0.044]; AOL[0.043]; AOL[0.043]; AOL[0.043]; AOL[0.043]; America[0.043]; America[0.043]; media[0.043]; Stephen[0.043]; company[0.043]; company[0.043]; Time[0.043]; Time[0.043]; Time[0.043]; Time[0.043]; Time[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sanford C. Bernstein\u001b[39m ==> ENTITY: \u001b[32mSanford Bernstein\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.142:0.142[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncompanies[0.044]; company[0.044]; company[0.044]; Strategically[0.043]; bought[0.043]; Capital[0.042]; Capital[0.042]; Integral[0.042]; history[0.042]; combo[0.042]; merged[0.042]; marketing[0.042]; financially[0.041]; shareholder[0.041]; media[0.041]; recent[0.041]; mergers[0.041]; sold[0.041]; strategically[0.041]; really[0.041]; month[0.040]; widely[0.040]; merger[0.040]; merger[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NBC ==> ENTITY: \u001b[32mNBC\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.082:0.082[0]; log p(e|m)= -0.064:-0.064[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncable[0.045]; million[0.044]; Warner[0.042]; Warner[0.042]; Warner[0.042]; deals[0.042]; AOL[0.042]; AOL[0.042]; channels[0.042]; distribution[0.042]; America[0.042]; announced[0.042]; sell[0.042]; media[0.042]; Disney[0.042]; Disney[0.042]; Washington[0.041]; watch[0.041]; company[0.041]; company[0.041]; day[0.041]; assistant[0.040]; sold[0.040]; McNamee[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Time ==> ENTITY: \u001b[32mTime Inc.\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.159:0.159[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nannounced[0.044]; Warner[0.043]; Warner[0.043]; Warner[0.043]; Time[0.043]; Time[0.043]; media[0.042]; media[0.042]; recent[0.042]; AOL[0.042]; AOL[0.042]; companies[0.041]; companies[0.041]; history[0.041]; company[0.041]; company[0.041]; ABC[0.041]; Internet[0.041]; sense[0.041]; really[0.041]; advertising[0.040]; distribution[0.040]; day[0.040]; figure[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Boston ==> ENTITY: \u001b[32mBoston\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.072:0.072[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSteve[0.045]; Larry[0.044]; Stephen[0.043]; Street[0.043]; company[0.043]; company[0.043]; company[0.043]; America[0.042]; time[0.042]; longtime[0.042]; major[0.042]; month[0.042]; probably[0.041]; deal[0.041]; deal[0.041]; Time[0.041]; Time[0.041]; Time[0.041]; Time[0.041]; Time[0.041]; like[0.041]; like[0.041]; stock[0.040]; fallen[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.065:0.065[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJohn[0.045]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; America[0.043]; people[0.043]; begun[0.042]; Roger[0.042]; assistant[0.041]; million[0.041]; trying[0.041]; multiple[0.041]; Warner[0.041]; Warner[0.041]; Warner[0.041]; big[0.041]; big[0.041]; Lynch[0.041]; day[0.041]; uncertainty[0.041]; reason[0.040]; work[0.040]; consumer[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Disney ==> ENTITY: \u001b[32mThe Walt Disney Company\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.247:-0.247[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDisney[0.048]; ABC[0.045]; Warner[0.044]; Warner[0.044]; Warner[0.044]; company[0.042]; company[0.042]; announced[0.042]; deal[0.042]; Roger[0.041]; sold[0.041]; Tom[0.041]; created[0.041]; recent[0.041]; figure[0.040]; Time[0.040]; Time[0.040]; Time[0.040]; AOL[0.040]; AOL[0.040]; merged[0.040]; deals[0.040]; bought[0.040]; history[0.040]; \t\n\r [================>.............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s311ms | Step: 8ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 96/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : chtb_165.eng ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Shanghai ==> ENTITY: \u001b[32mShanghai\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.157:0.157[0]; log p(e|m)= -0.101:-0.101[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nShanghai[0.049]; Shanghai[0.049]; Shanghai[0.049]; China[0.047]; trading[0.044]; Japan[0.044]; Japan[0.044]; Tokyo[0.043]; Lujiazui[0.043]; Xinhua[0.043]; Kong[0.043]; Recently[0.042]; Pudong[0.042]; Pudong[0.042]; Zhou[0.042]; HSBC[0.042]; HSBC[0.042]; Hong[0.042]; RMB[0.042]; RMB[0.042]; Shipping[0.042]; Mansion[0.041]; Industrial[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tokyo Mitsubishi Bank\u001b[39m ==> ENTITY: \u001b[32mThe Bank of Tokyo-Mitsubishi UFJ\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.236:0.236[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBank[0.048]; Bank[0.048]; Bank[0.048]; Bank[0.048]; Bank[0.048]; Citibank[0.047]; bank[0.046]; Japan[0.045]; Japan[0.045]; Japan[0.045]; Japan[0.045]; Financial[0.045]; financial[0.045]; financial[0.045]; HSBC[0.044]; HSBC[0.044]; institutions[0.044]; banks[0.044]; banks[0.044]; banks[0.044]; banks[0.044]; April[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Citibank\u001b[39m ==> ENTITY: \u001b[32mCitibank\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.254:0.254[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBank[0.046]; Bank[0.046]; Bank[0.046]; Bank[0.046]; Bank[0.046]; Bank[0.046]; HSBC[0.044]; HSBC[0.044]; financial[0.043]; Financial[0.043]; bank[0.042]; branch[0.042]; business[0.040]; business[0.040]; business[0.040]; Japan[0.039]; Japan[0.039]; Japan[0.039]; Japan[0.039]; banks[0.039]; banks[0.039]; banks[0.039]; announced[0.038]; trading[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = US ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.042:0.042[0]; log p(e|m)= -0.488:-0.488[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfinancial[0.050]; policy[0.049]; China[0.049]; foreign[0.048]; foreign[0.048]; foreign[0.048]; bank[0.048]; business[0.048]; business[0.048]; banks[0.048]; banks[0.048]; banks[0.048]; Japan[0.047]; Japan[0.047]; Japan[0.047]; Japan[0.047]; capital[0.046]; capital[0.046]; capital[0.046]; trading[0.046]; News[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hong Kong ==> ENTITY: \u001b[32mHong Kong\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.167:0.167[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.046]; Japan[0.045]; Japan[0.045]; Japan[0.045]; Japan[0.045]; Shanghai[0.044]; Shanghai[0.044]; Shanghai[0.044]; Shanghai[0.044]; Shanghai[0.044]; Shanghai[0.044]; Shanghai[0.044]; HSBC[0.043]; HSBC[0.043]; Dai[0.042]; RMB[0.042]; RMB[0.042]; RMB[0.042]; Pudong[0.042]; Pudong[0.042]; Pudong[0.042]; Pudong[0.042]; Zhou[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = HSBC\u001b[39m ==> ENTITY: \u001b[32mHSBC\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.238:0.238[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHSBC[0.049]; Bank[0.047]; Bank[0.047]; Bank[0.047]; Bank[0.047]; Citibank[0.046]; bank[0.043]; branch[0.041]; Shanghai[0.041]; Shanghai[0.041]; Shanghai[0.041]; Shanghai[0.041]; Shanghai[0.041]; Shanghai[0.041]; financial[0.040]; business[0.040]; business[0.040]; banks[0.039]; China[0.039]; capital[0.038]; capital[0.038]; Lujiazui[0.038]; Pudong[0.038]; Pudong[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pudong\u001b[39m ==> ENTITY: \u001b[32mPudong\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nShanghai[0.046]; Shanghai[0.046]; Shanghai[0.046]; Shanghai[0.046]; Shanghai[0.046]; Shanghai[0.046]; Shanghai[0.046]; Pudong[0.042]; Pudong[0.042]; Pudong[0.042]; China[0.042]; Lujiazui[0.042]; Lujiazui[0.042]; Tokyo[0.041]; Industrial[0.039]; district[0.039]; capital[0.038]; capital[0.038]; Japan[0.038]; Japan[0.038]; Japan[0.038]; Japan[0.038]; April[0.038]; Xinhua[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.149:0.149[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.055]; Japan[0.055]; Japan[0.055]; Tokyo[0.048]; China[0.047]; formally[0.045]; policy[0.044]; Shanghai[0.044]; Shanghai[0.044]; Shanghai[0.044]; Shanghai[0.044]; Shanghai[0.044]; Shanghai[0.044]; Shanghai[0.044]; foreign[0.044]; foreign[0.044]; foreign[0.044]; foreign[0.044]; Agency[0.043]; Citibank[0.043]; announced[0.043]; Recently[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Xinhua News Agency\u001b[39m ==> ENTITY: \u001b[32mXinhua News Agency\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.048]; Shanghai[0.044]; Shanghai[0.044]; Shanghai[0.044]; Shanghai[0.044]; Zhou[0.043]; foreign[0.042]; foreign[0.042]; April[0.042]; Hong[0.041]; reporter[0.041]; financial[0.041]; Kong[0.041]; 3rd[0.040]; capital[0.040]; capital[0.040]; Pudong[0.040]; Pudong[0.040]; district[0.040]; Lujiazui[0.040]; Japan[0.040]; Japan[0.040]; business[0.040]; business[0.040]; \t\n\r [==================>...........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s198ms | Step: 7ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 107/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 20001115_AFP_ARB.0089.eng ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = AFP ==> ENTITY: \u001b[32mAgence France-Presse\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.083:0.083[0]; log p(e|m)= -0.660:-0.660[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.048]; child[0.043]; Lebanon[0.043]; Lebanon[0.043]; National[0.043]; army[0.043]; Israeli[0.042]; Beirut[0.042]; pictures[0.042]; announcement[0.042]; form[0.042]; children[0.041]; Wednesday[0.041]; today[0.041]; Palestinian[0.041]; Palestinian[0.041]; Palestinian[0.041]; Palestinian[0.041]; Council[0.041]; Algeria[0.040]; independence[0.040]; commemorated[0.040]; held[0.040]; early[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Algeria ==> ENTITY: \u001b[32mAlgeria\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.169:0.169[0]; log p(e|m)= -0.337:-0.337[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.048]; Mohamed[0.044]; Lebanon[0.044]; Lebanon[0.044]; Lebanon[0.044]; military[0.044]; southern[0.043]; Saida[0.042]; Israeli[0.042]; south[0.042]; army[0.042]; independence[0.041]; refugee[0.041]; al[0.041]; main[0.041]; National[0.040]; Beirut[0.040]; cities[0.040]; October[0.040]; refugees[0.040]; AFP[0.040]; camps[0.040]; early[0.039]; Wednesday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Saida ==> ENTITY: \u001b[32mSidon\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.159:0.159[0]; log p(e|m)= -0.028:-0.028[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaida[0.052]; Lebanon[0.051]; Algeria[0.049]; Beirut[0.048]; Israel[0.046]; al[0.046]; Mohamed[0.046]; southern[0.045]; main[0.045]; Yasser[0.044]; started[0.044]; young[0.044]; American[0.044]; Khalid[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; cities[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mohamed Al-Dura\u001b[39m ==> ENTITY: \u001b[32mMuhammad al-Durrah incident\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsraeli[0.047]; Israeli[0.047]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinians[0.042]; Palestinians[0.042]; Beirut[0.042]; Lebanon[0.041]; Lebanon[0.041]; Lebanon[0.041]; pictures[0.040]; AFP[0.040]; Yasser[0.040]; Fatah[0.039]; Arafat[0.039]; army[0.039]; outside[0.038]; near[0.038]; near[0.038]; reported[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanon ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.209:0.209[0]; log p(e|m)= -0.409:-0.409[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBeirut[0.047]; Israel[0.046]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; Palestinian[0.042]; Palestinian[0.042]; Palestinian[0.042]; Palestinian[0.042]; Palestinian[0.042]; Palestinian[0.042]; Algeria[0.041]; military[0.041]; southern[0.041]; Fatah[0.041]; Fatah[0.041]; Intifada[0.040]; Intifada[0.040]; Jerusalem[0.040]; Palestinians[0.040]; Palestinians[0.040]; refugee[0.040]; al[0.040]; France[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Palestinian National Council ==> ENTITY: \u001b[32mPalestinian National Council\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.184:0.184[0]; log p(e|m)= -0.013:-0.013[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPalestinian[0.047]; Palestinian[0.047]; Palestinian[0.047]; Lebanon[0.046]; Lebanon[0.046]; Lebanon[0.046]; Beirut[0.046]; Palestinians[0.045]; Palestinians[0.045]; Israeli[0.044]; al[0.042]; Algeria[0.039]; Mohamed[0.039]; Saida[0.039]; announcement[0.039]; held[0.039]; reported[0.038]; refugee[0.038]; army[0.038]; refugees[0.038]; military[0.038]; today[0.038]; camps[0.038]; camp[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Beirut ==> ENTITY: \u001b[32mBeirut\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.175:0.175[0]; log p(e|m)= -0.047:-0.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.048]; Lebanon[0.048]; Lebanon[0.048]; Israeli[0.046]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; military[0.043]; Palestinians[0.041]; Palestinians[0.041]; Fatah[0.041]; southern[0.041]; army[0.040]; al[0.039]; Yasser[0.039]; Arafat[0.039]; October[0.038]; Saida[0.038]; main[0.038]; Wednesday[0.038]; movement[0.037]; early[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jerusalem ==> ENTITY: \u001b[32mJerusalem\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.193:0.193[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJerusalem[0.048]; Israel[0.048]; Lebanon[0.046]; Israeli[0.044]; Israeli[0.044]; Israeli[0.044]; Israeli[0.044]; occupied[0.044]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; holy[0.043]; Palestinians[0.042]; settler[0.042]; settlers[0.042]; outside[0.041]; occupying[0.041]; Arafat[0.041]; forces[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanon ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.176:0.176[0]; log p(e|m)= -0.409:-0.409[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.053]; Beirut[0.048]; Israeli[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Algeria[0.043]; military[0.042]; Palestinians[0.042]; refugees[0.041]; al[0.041]; National[0.041]; camps[0.041]; army[0.040]; Council[0.040]; symbolic[0.039]; Wednesday[0.039]; commemorated[0.039]; south[0.038]; thousand[0.038]; today[0.038]; Mohamed[0.038]; held[0.038]; \t\n\r [===================>..........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s97ms | Step: 7ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 117/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 20001115_AFP_ARB.0061.eng ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = AFP ==> ENTITY: \u001b[32mAgence France-Presse\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.100:0.100[0]; log p(e|m)= -0.660:-0.660[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSecurity[0.058]; Security[0.058]; issued[0.056]; issued[0.056]; President[0.056]; President[0.056]; high[0.056]; complete[0.056]; source[0.056]; source[0.056]; level[0.055]; today[0.055]; armed[0.055]; said[0.054]; said[0.054]; Wednesday[0.054]; led[0.054]; led[0.054]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mGaza\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mGaza Strip\u001b[39m <---> \u001b[32mGaza\u001b[39m\t\nSCORES: global= 0.251:0.246[\u001b[31m0.005\u001b[39m]; local()= 0.184:0.167[\u001b[31m0.017\u001b[39m]; log p(e|m)= -1.374:-0.304[\u001b[32m1.071\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPalestinian[0.048]; Palestinian[0.048]; Palestinian[0.048]; Palestinian[0.048]; Palestinian[0.048]; Palestinian[0.048]; civilians[0.045]; area[0.042]; Yasser[0.040]; Yasser[0.040]; armed[0.039]; Wednesday[0.039]; said[0.039]; said[0.039]; Territories[0.039]; Arafat[0.039]; Arafat[0.039]; issued[0.039]; issued[0.039]; Security[0.039]; Security[0.039]; today[0.038]; orders[0.038]; orders[0.038]; \t\n\r [====================>.........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s108ms | Step: 8ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 119/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : NYT20001217.2241.0165 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = La Guardia ==> ENTITY: \u001b[32mLaGuardia Airport\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.096:0.096[0]; log p(e|m)= -1.302:-1.302[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUR[0.045]; UR[0.045]; airport[0.045]; airport[0.045]; flight[0.043]; flight[0.043]; York[0.043]; airports[0.042]; airports[0.042]; delays[0.042]; flights[0.041]; late[0.041]; nonstop[0.041]; leg[0.041]; delay[0.041]; Airlines[0.041]; major[0.040]; LR[0.040]; LR[0.040]; ADV[0.040]; Chicago[0.040]; QL[0.040]; QL[0.040]; QL[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = John Wayne Airport\u001b[39m ==> ENTITY: \u001b[32mJohn Wayne Airport\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirport[0.052]; Orange[0.048]; Burbank[0.047]; JFK[0.047]; airports[0.046]; delayed[0.046]; Monday[0.046]; Calif[0.046]; International[0.045]; delays[0.045]; delays[0.045]; Friday[0.045]; earlier[0.045]; Angeles[0.044]; County[0.044]; County[0.044]; leaving[0.044]; cancellations[0.044]; schedules[0.044]; travel[0.044]; Los[0.044]; carriers[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.067:0.067[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWashington[0.050]; pledged[0.044]; Department[0.043]; Department[0.043]; federal[0.043]; renting[0.043]; DC[0.043]; Protection[0.042]; mile[0.042]; pay[0.042]; serve[0.041]; increase[0.041]; flight[0.041]; flight[0.041]; monthly[0.041]; reimbursement[0.041]; driving[0.041]; Consumer[0.040]; Consumer[0.040]; Transportation[0.040]; Transportation[0.040]; Travel[0.040]; consumer[0.040]; car[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Orange County ==> ENTITY: \u001b[32mOrange County, California\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.090:0.090[0]; log p(e|m)= -0.449:-0.449[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirport[0.046]; Airport[0.046]; County[0.046]; York[0.045]; Westchester[0.044]; Wayne[0.042]; Burbank[0.042]; JFK[0.041]; Monday[0.041]; month[0.041]; schedules[0.041]; airports[0.041]; earlier[0.041]; delayed[0.041]; Los[0.041]; cancellations[0.041]; Friday[0.041]; Calif[0.040]; Saturday[0.040]; International[0.040]; delays[0.040]; delays[0.040]; days[0.040]; John[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Los Angeles International\u001b[39m ==> ENTITY: \u001b[32mLos Angeles International Airport\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirport[0.046]; Airport[0.046]; JFK[0.045]; Monday[0.044]; Burbank[0.044]; Orange[0.043]; Friday[0.042]; delayed[0.042]; earlier[0.041]; leaving[0.041]; month[0.041]; delays[0.041]; delays[0.041]; airports[0.041]; Saturday[0.041]; Westchester[0.041]; week[0.040]; Wayne[0.040]; schedules[0.040]; airline[0.040]; airline[0.040]; Calif[0.040]; day[0.040]; day[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Calif\u001b[39m ==> ENTITY: \u001b[32mCalifornia\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.080:0.080[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInternational[0.045]; business[0.045]; week[0.044]; month[0.043]; Burbank[0.043]; Monday[0.042]; traveled[0.042]; Travel[0.042]; secure[0.042]; schedules[0.041]; earlier[0.041]; Orange[0.041]; Angeles[0.041]; Friday[0.041]; Los[0.041]; John[0.041]; day[0.041]; day[0.041]; major[0.041]; Wayne[0.041]; number[0.040]; number[0.040]; www[0.040]; smaller[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chicago O'Hare\u001b[39m ==> ENTITY: \u001b[32mO'Hare International Airport\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairport[0.046]; airport[0.046]; UR[0.045]; UR[0.045]; airports[0.044]; airports[0.044]; flights[0.043]; delays[0.042]; Airlines[0.042]; flight[0.042]; flight[0.042]; nonstop[0.041]; allow[0.041]; Transportation[0.040]; times[0.040]; delay[0.040]; airlines[0.040]; leg[0.040]; late[0.040]; ART[0.040]; LR[0.040]; LR[0.040]; LR[0.040]; ADV[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South ==> ENTITY: \u001b[32mSouthern United States\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.081:0.081[0]; log p(e|m)= -1.332:-1.332[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.048]; record[0.045]; York[0.044]; summer[0.043]; late[0.042]; weather[0.042]; major[0.041]; time[0.041]; time[0.041]; time[0.041]; time[0.041]; time[0.041]; winter[0.041]; purchases[0.041]; records[0.041]; lower[0.041]; domestic[0.041]; connecting[0.041]; fewer[0.041]; fewer[0.041]; leg[0.040]; spring[0.040]; chance[0.040]; miss[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Westchester County Airport\u001b[39m ==> ENTITY: \u001b[32mWestchester County Airport\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirport[0.046]; County[0.044]; airports[0.043]; Monday[0.043]; JFK[0.043]; week[0.043]; delayed[0.042]; Wayne[0.042]; day[0.042]; day[0.042]; York[0.042]; Friday[0.042]; Orange[0.042]; delays[0.041]; delays[0.041]; delays[0.041]; schedules[0.041]; cancellations[0.041]; cancellations[0.041]; Saturday[0.040]; times[0.040]; times[0.040]; days[0.040]; Burbank[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burbank ==> ENTITY: \u001b[32mBob Hope Airport\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.141:0.141[0]; log p(e|m)= -2.718:-2.718[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirport[0.046]; Airport[0.046]; JFK[0.044]; County[0.043]; County[0.043]; Monday[0.042]; Friday[0.042]; delays[0.041]; delays[0.041]; day[0.041]; day[0.041]; Los[0.041]; time[0.041]; consider[0.041]; ticket[0.041]; John[0.041]; Wayne[0.041]; month[0.041]; Orange[0.041]; schedules[0.041]; Calif[0.041]; week[0.041]; airports[0.040]; delayed[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew York\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew York\u001b[39m <---> \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.243:0.243[\u001b[31m0.000\u001b[39m]; local()= 0.029:0.034[\u001b[32m0.005\u001b[39m]; log p(e|m)= -0.289:-1.790[\u001b[31m1.501\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nflight[0.045]; flight[0.045]; airport[0.045]; airport[0.045]; Times[0.044]; airports[0.042]; airports[0.042]; flights[0.042]; NYT[0.042]; NYT[0.042]; News[0.042]; nonstop[0.041]; Airlines[0.041]; book[0.041]; book[0.041]; delays[0.041]; late[0.040]; connecting[0.040]; Chicago[0.040]; major[0.040]; South[0.040]; domestic[0.040]; delay[0.040]; congested[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = JFK ==> ENTITY: \u001b[32mJohn F. Kennedy International Airport\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.182:0.182[0]; log p(e|m)= -1.221:-1.221[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirport[0.046]; Airport[0.046]; John[0.043]; airports[0.042]; delays[0.042]; delays[0.042]; delays[0.042]; delayed[0.042]; York[0.042]; Monday[0.042]; leaving[0.041]; cancellations[0.041]; cancellations[0.041]; Friday[0.041]; Burbank[0.041]; use[0.041]; airline[0.041]; airline[0.041]; Westchester[0.041]; schedules[0.040]; travel[0.040]; travel[0.040]; week[0.040]; flying[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Transportation Department\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Transportation\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.101:0.101[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleg[0.044]; delays[0.044]; records[0.043]; month[0.043]; requires[0.043]; Airlines[0.042]; allow[0.042]; Travel[0.042]; delayed[0.041]; airlines[0.041]; Monday[0.041]; week[0.041]; consideration[0.041]; airline[0.041]; season[0.041]; winter[0.041]; business[0.041]; miss[0.041]; www[0.041]; record[0.041]; record[0.041]; airport[0.041]; airport[0.041]; summer[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Federal Aviation Administration\u001b[39m ==> ENTITY: \u001b[32mFederal Aviation Administration\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nflight[0.044]; operations[0.044]; airport[0.043]; airports[0.043]; airlines[0.043]; airline[0.042]; airline[0.042]; airline[0.042]; airline[0.042]; carriers[0.042]; delays[0.042]; delays[0.042]; carry[0.042]; flights[0.042]; Use[0.041]; delay[0.041]; Ship[0.041]; Automated[0.041]; fly[0.041]; provide[0.040]; status[0.040]; information[0.040]; information[0.040]; services[0.040]; \t\n\r [======================>.......................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 940ms | Step: 7ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 134/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 20001115_AFP_ARB.0184.eng ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = AFP ==> ENTITY: \u001b[32mAgence France-Presse\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.660:-0.660[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecurity[0.045]; security[0.045]; Military[0.045]; reported[0.043]; reported[0.043]; officials[0.043]; officials[0.043]; saying[0.043]; President[0.042]; army[0.042]; procedure[0.042]; political[0.041]; Israeli[0.041]; radio[0.041]; radio[0.041]; cabinet[0.041]; cabinet[0.041]; escalation[0.040]; Wednesday[0.040]; Palestinian[0.040]; continue[0.040]; senior[0.039]; meeting[0.039]; meeting[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Bank ==> ENTITY: \u001b[32mWest Bank\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.174:0.174[0]; log p(e|m)= -0.054:-0.054[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsraeli[0.047]; Israelis[0.047]; Gaza[0.046]; Barak[0.045]; Barak[0.045]; West[0.045]; Bank[0.044]; Intifada[0.043]; Fatah[0.043]; Fatah[0.043]; Arafat[0.041]; security[0.040]; security[0.040]; military[0.040]; cabinet[0.040]; civilians[0.040]; political[0.039]; Strip[0.039]; forces[0.039]; forces[0.039]; officials[0.039]; policy[0.039]; reach[0.039]; movement[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jerusalem ==> ENTITY: \u001b[32mJerusalem\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsraeli[0.047]; Ehud[0.046]; Palestinian[0.046]; Barak[0.045]; Palestinians[0.045]; Wednesday[0.042]; officials[0.042]; officials[0.042]; Yasser[0.041]; wants[0.041]; Arafat[0.041]; reported[0.041]; reported[0.041]; days[0.041]; close[0.041]; saying[0.041]; followed[0.040]; met[0.040]; series[0.040]; desire[0.040]; security[0.040]; security[0.040]; army[0.040]; past[0.040]; \t\n\r [=======================>......................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 946ms | Step: 7ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 137/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : NYT20001124.2050.0257 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Asia ==> ENTITY: \u001b[32mAsia\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.091:0.091[0]; log p(e|m)= -0.150:-0.150[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAsian[0.047]; Times[0.045]; News[0.044]; economic[0.043]; month[0.043]; month[0.043]; financial[0.043]; news[0.043]; reported[0.042]; turmoil[0.042]; warned[0.042]; predicted[0.042]; performed[0.042]; International[0.041]; conference[0.041]; markets[0.041]; sort[0.040]; hit[0.040]; Friday[0.040]; economies[0.040]; economies[0.040]; projection[0.040]; growing[0.040]; Hong[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Malaysia ==> ENTITY: \u001b[32mMalaysia\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.218:0.218[0]; log p(e|m)= -0.202:-0.202[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMalaysia[0.050]; Indonesia[0.046]; Indonesia[0.046]; Asian[0.043]; Asian[0.043]; Asia[0.042]; Philippines[0.042]; Philippines[0.042]; Korea[0.041]; Korea[0.041]; Korea[0.041]; beating[0.041]; Taiwan[0.041]; Taiwan[0.041]; quarters[0.040]; Chen[0.040]; countries[0.040]; countries[0.040]; countries[0.040]; countries[0.040]; countries[0.040]; quarter[0.039]; week[0.039]; world[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hong Kong General Chamber of Commerce\u001b[39m ==> ENTITY: \u001b[32mHong Kong General Chamber of Commerce\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbusiness[0.046]; Hong[0.045]; Hong[0.045]; Hong[0.045]; China[0.045]; economy[0.045]; Kong[0.043]; Kong[0.043]; Kong[0.043]; growth[0.043]; growth[0.043]; growth[0.043]; Exports[0.043]; economies[0.043]; economies[0.043]; Asia[0.043]; export[0.043]; Asian[0.043]; Asian[0.043]; mainland[0.043]; October[0.043]; crunch[0.042]; particularly[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Korea ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.191:0.191[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.051]; Korea[0.051]; Singapore[0.045]; beating[0.045]; South[0.045]; Asian[0.044]; Asian[0.044]; Taiwan[0.044]; Taiwan[0.044]; Malaysia[0.043]; Malaysia[0.043]; countries[0.042]; countries[0.042]; countries[0.042]; countries[0.042]; countries[0.042]; week[0.042]; Asia[0.042]; Philippines[0.042]; Philippines[0.042]; Chen[0.041]; president[0.041]; president[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.231:0.231[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMalaysia[0.046]; Asian[0.046]; Asian[0.046]; Asian[0.046]; Asia[0.045]; countries[0.044]; countries[0.044]; countries[0.044]; countries[0.044]; Philippines[0.044]; Philippines[0.044]; Korea[0.042]; Korea[0.042]; world[0.041]; Taiwan[0.041]; Taiwan[0.041]; economic[0.039]; instability[0.039]; economies[0.038]; South[0.038]; week[0.038]; turmoil[0.037]; conference[0.037]; weeks[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMalaysia[0.048]; Asian[0.047]; countries[0.044]; countries[0.044]; Korea[0.044]; Korea[0.044]; Korea[0.044]; Indonesia[0.043]; shipping[0.043]; China[0.042]; mainland[0.042]; analysts[0.041]; quarters[0.040]; economists[0.040]; second[0.040]; rose[0.040]; economic[0.040]; economic[0.040]; business[0.040]; goods[0.040]; increase[0.040]; surged[0.039]; beating[0.039]; nearly[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hong Kong ==> ENTITY: \u001b[32mHong Kong\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.093:0.093[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKong[0.046]; Asia[0.045]; Hong[0.045]; digit[0.044]; Asian[0.044]; month[0.043]; month[0.043]; rates[0.042]; International[0.042]; conference[0.041]; economic[0.041]; Service[0.041]; hit[0.041]; LR[0.040]; LR[0.040]; LR[0.040]; percent[0.040]; percent[0.040]; percent[0.040]; Friday[0.040]; hard[0.040]; financial[0.040]; double[0.040]; growth[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Philippines ==> ENTITY: \u001b[32mPhilippines\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.179:0.179[0]; log p(e|m)= -0.180:-0.180[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhilippines[0.052]; Malaysia[0.045]; Asian[0.045]; Asian[0.045]; Asian[0.045]; Indonesia[0.042]; Asia[0.042]; Korea[0.042]; Korea[0.042]; countries[0.041]; countries[0.041]; countries[0.041]; countries[0.041]; Taiwan[0.040]; Taiwan[0.040]; Estrada[0.040]; weeks[0.040]; International[0.040]; week[0.040]; world[0.040]; president[0.039]; president[0.039]; economic[0.039]; conference[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBloomberg News\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBloomberg Television\u001b[39m <---> \u001b[32mBloomberg News\u001b[39m\t\nSCORES: global= 0.288:0.273[\u001b[31m0.015\u001b[39m]; local()= 0.131:0.141[\u001b[32m0.010\u001b[39m]; log p(e|m)= 0.000:-2.033[\u001b[31m2.033\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nreported[0.045]; business[0.045]; Tuesday[0.044]; analysts[0.043]; acknowledged[0.042]; consumer[0.042]; economic[0.042]; China[0.042]; Asian[0.042]; Consumer[0.042]; nearly[0.041]; chief[0.041]; said[0.041]; particularly[0.041]; Singapore[0.041]; retail[0.041]; countries[0.041]; spending[0.041]; referring[0.041]; rose[0.041]; began[0.041]; quarter[0.041]; quarter[0.041]; quarter[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chen Shui-bian\u001b[39m ==> ENTITY: \u001b[32mChen Shui-bian\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTaiwan[0.050]; Taiwan[0.050]; acknowledged[0.045]; vote[0.045]; president[0.045]; president[0.045]; recent[0.044]; Tuesday[0.044]; week[0.043]; weeks[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; economic[0.042]; economic[0.042]; Korea[0.042]; Korea[0.042]; Korea[0.042]; Malaysia[0.042]; referring[0.042]; Asian[0.042]; Asian[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = International Monetary Fund\u001b[39m ==> ENTITY: \u001b[32mInternational Monetary Fund\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.125:0.125[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUR[0.045]; UR[0.045]; economic[0.044]; financial[0.043]; economies[0.042]; economies[0.042]; IMF[0.042]; Asian[0.042]; Asian[0.042]; Asian[0.042]; rates[0.042]; rates[0.042]; countries[0.042]; countries[0.042]; markets[0.041]; prices[0.041]; notably[0.040]; Asia[0.040]; Asia[0.040]; fund[0.040]; price[0.040]; soaring[0.040]; LR[0.040]; LR[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.063:0.063[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnearly[0.045]; second[0.044]; surveyed[0.044]; Korea[0.043]; South[0.043]; acknowledged[0.042]; China[0.042]; Malaysia[0.042]; said[0.042]; beating[0.041]; quarter[0.041]; quarter[0.041]; quarter[0.041]; Ian[0.041]; quarters[0.041]; Tuesday[0.041]; countries[0.041]; countries[0.041]; Asian[0.041]; Asian[0.041]; Singapore[0.041]; percentage[0.041]; particularly[0.041]; spending[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Taiwan ==> ENTITY: \u001b[32mTaiwan\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.208:0.208[0]; log p(e|m)= -0.078:-0.078[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTaiwan[0.052]; Philippines[0.048]; Philippines[0.048]; Korea[0.047]; Korea[0.047]; countries[0.046]; countries[0.046]; countries[0.046]; countries[0.046]; Asian[0.046]; Asian[0.046]; Asian[0.046]; Malaysia[0.046]; economic[0.045]; Asia[0.044]; Indonesia[0.044]; Chen[0.043]; political[0.043]; political[0.043]; recent[0.042]; referring[0.042]; instability[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmainland[0.046]; Malaysia[0.044]; Asian[0.043]; Asian[0.043]; countries[0.042]; Kong[0.042]; Kong[0.042]; Kong[0.042]; Kong[0.042]; Kong[0.042]; Hong[0.042]; Hong[0.042]; Hong[0.042]; Hong[0.042]; Hong[0.042]; Singapore[0.042]; region[0.042]; Indonesia[0.041]; second[0.041]; beating[0.041]; domestic[0.040]; domestic[0.040]; began[0.040]; General[0.039]; \t\n\r [==========================>...................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 820ms | Step: 7ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 151/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : NYT20001109.1946.0315 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ralph Nader\u001b[39m ==> ENTITY: \u001b[32mRalph Nader\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncampaign[0.046]; campaign[0.046]; Nader[0.046]; Nader[0.046]; presidential[0.045]; America[0.044]; Party[0.043]; country[0.042]; concerned[0.041]; website[0.041]; Green[0.041]; Green[0.041]; bad[0.041]; official[0.041]; Ralph[0.040]; politics[0.040]; member[0.040]; instance[0.040]; WASHINGTON[0.040]; org[0.040]; posted[0.039]; Service[0.039]; Thursday[0.039]; regrets[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = White House\u001b[39m ==> ENTITY: \u001b[32mWhite House\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.006:-0.006[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGeorge[0.047]; Bush[0.047]; Bush[0.047]; President[0.046]; interview[0.043]; Thursday[0.042]; Thursday[0.042]; position[0.042]; support[0.042]; country[0.042]; remarks[0.042]; Gore[0.040]; Gore[0.040]; Gore[0.040]; expressed[0.040]; House[0.040]; White[0.040]; member[0.040]; member[0.040]; Instead[0.040]; Ralph[0.040]; Ralph[0.040]; Vice[0.040]; party[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gainesville ==> ENTITY: \u001b[32mGainesville, Florida\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.658:-0.658[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFlorida[0.046]; Florida[0.046]; Florida[0.046]; University[0.044]; record[0.043]; Tuesday[0.043]; near[0.042]; scare[0.042]; won[0.042]; supporters[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; guilty[0.041]; rights[0.041]; worry[0.040]; fewer[0.040]; women[0.040]; Bush[0.040]; Club[0.040]; commercial[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portland ==> ENTITY: \u001b[32mPortland, Oregon\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.082:0.082[0]; log p(e|m)= -0.664:-0.664[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWednesday[0.045]; won[0.044]; state[0.044]; Thursday[0.044]; Oregon[0.043]; close[0.042]; presidency[0.042]; presidency[0.042]; winning[0.042]; percent[0.042]; percent[0.042]; percent[0.042]; despite[0.041]; prior[0.041]; Green[0.041]; said[0.041]; said[0.041]; said[0.041]; going[0.041]; 7th[0.041]; president[0.040]; federal[0.040]; funds[0.040]; probably[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sierra Club\u001b[39m ==> ENTITY: \u001b[32mSierra Club\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.103:0.103[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsupporters[0.044]; environmentalists[0.044]; state[0.043]; chairwoman[0.043]; University[0.043]; Everglades[0.042]; Everglades[0.042]; coordinator[0.042]; Gainesville[0.042]; near[0.042]; president[0.041]; Gore[0.041]; Gore[0.041]; Gore[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; Green[0.041]; rights[0.041]; receive[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = America ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.063:0.063[0]; log p(e|m)= -0.785:-0.785[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.048]; Ralph[0.045]; Ralph[0.045]; Ralph[0.045]; Ralph[0.045]; writer[0.044]; happened[0.044]; chat[0.044]; bad[0.044]; feel[0.044]; feeling[0.043]; man[0.043]; man[0.043]; sounded[0.043]; named[0.042]; Thursday[0.042]; Thursday[0.042]; today[0.042]; movement[0.042]; movement[0.042]; seen[0.042]; come[0.042]; interview[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oregon ==> ENTITY: \u001b[32mOregon\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.101:0.101[0]; log p(e|m)= -0.286:-0.286[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstate[0.051]; state[0.051]; percent[0.049]; percent[0.049]; John[0.049]; president[0.049]; Portland[0.049]; voters[0.048]; House[0.048]; vote[0.047]; votes[0.047]; said[0.047]; said[0.047]; said[0.047]; Wednesday[0.047]; Thursday[0.046]; Thursday[0.046]; White[0.046]; Ralph[0.046]; won[0.046]; likely[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Florida ==> ENTITY: \u001b[32mFlorida\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.056:0.056[0]; log p(e|m)= -0.154:-0.154[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJohn[0.045]; trailed[0.043]; Thursday[0.043]; Thursday[0.043]; run[0.043]; despite[0.042]; goal[0.042]; position[0.042]; prior[0.042]; White[0.042]; White[0.042]; defeat[0.041]; shared[0.041]; state[0.041]; major[0.041]; George[0.041]; votes[0.041]; votes[0.041]; votes[0.041]; likely[0.041]; said[0.041]; House[0.040]; House[0.040]; 7th[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = University of Florida\u001b[39m ==> ENTITY: \u001b[32mUniversity of Florida\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.112:0.112[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFlorida[0.049]; Florida[0.049]; Gainesville[0.046]; won[0.045]; record[0.044]; Tuesday[0.043]; women[0.042]; fewer[0.041]; chapter[0.041]; felt[0.041]; votes[0.040]; abysmal[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; Patricia[0.040]; Newell[0.040]; president[0.040]; knew[0.040]; far[0.039]; commercial[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.049:0.049[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncampaign[0.046]; campaign[0.046]; presidential[0.044]; country[0.044]; America[0.043]; Thursday[0.042]; room[0.042]; member[0.042]; Service[0.042]; LR[0.041]; LR[0.041]; News[0.041]; instance[0.041]; official[0.041]; come[0.041]; Times[0.041]; Nader[0.040]; Nader[0.040]; Nader[0.040]; politics[0.040]; Ralph[0.040]; Ralph[0.040]; duped[0.040]; named[0.040]; \t\n\r [===========================>..................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 724ms | Step: 7ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 162/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : chtb_267.eng ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Beijing ==> ENTITY: \u001b[32mBeijing\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.126:0.126[0]; log p(e|m)= -0.042:-0.042[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBeijing[0.054]; China[0.053]; capacity[0.045]; September[0.044]; Xinhua[0.044]; Xinhua[0.044]; Wu[0.044]; operation[0.044]; built[0.044]; build[0.043]; Jiangsu[0.043]; International[0.043]; today[0.042]; signed[0.041]; 1st[0.041]; large[0.041]; reporters[0.041]; recently[0.041]; Province[0.041]; successful[0.041]; Weir[0.041]; Department[0.041]; Department[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = US ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.050:0.050[0]; log p(e|m)= -0.488:-0.488[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.044]; recently[0.043]; build[0.042]; Department[0.042]; Department[0.042]; operation[0.042]; Industry[0.042]; coastal[0.042]; power[0.042]; power[0.042]; plants[0.042]; today[0.042]; Electric[0.041]; Electric[0.041]; News[0.041]; News[0.041]; capacity[0.041]; built[0.041]; Dan[0.041]; plant[0.041]; plant[0.041]; September[0.041]; regions[0.041]; similar[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.118:0.118[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.048]; Taiwan[0.045]; China[0.044]; China[0.044]; China[0.044]; international[0.044]; national[0.043]; national[0.043]; Asian[0.042]; John[0.041]; Brown[0.041]; Group[0.041]; Group[0.041]; Group[0.041]; began[0.040]; coastal[0.040]; coastal[0.040]; visited[0.040]; regions[0.040]; regions[0.040]; signed[0.040]; South[0.040]; material[0.040]; protection[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rudong\u001b[39m ==> ENTITY: \u001b[32mRudong County\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJiangsu[0.047]; Jiangsu[0.047]; Chinese[0.046]; coastal[0.046]; coastal[0.046]; China[0.046]; dollars[0.045]; large[0.044]; situated[0.044]; regions[0.043]; regions[0.043]; Company[0.043]; pollution[0.042]; Luo[0.042]; plant[0.042]; plant[0.042]; plant[0.042]; specialized[0.042]; according[0.042]; annual[0.042]; globally[0.042]; kilowatts[0.042]; kilowatts[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Korea ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnational[0.046]; national[0.046]; Japan[0.046]; international[0.045]; China[0.044]; China[0.044]; China[0.044]; Asian[0.044]; Taiwan[0.044]; Group[0.041]; Group[0.041]; Group[0.041]; terms[0.040]; reached[0.040]; Weir[0.040]; Weir[0.040]; Weir[0.040]; Weir[0.040]; gross[0.040]; gross[0.040]; Jiangsu[0.039]; experience[0.039]; held[0.039]; financial[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Weir Group\u001b[39m ==> ENTITY: \u001b[32mWeir Group\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \noperation[0.045]; plant[0.043]; plant[0.043]; International[0.043]; gas[0.043]; build[0.042]; today[0.042]; built[0.042]; Department[0.042]; Department[0.042]; fuel[0.042]; signed[0.042]; capacity[0.042]; 1st[0.041]; Coal[0.041]; News[0.041]; News[0.041]; plants[0.041]; water[0.040]; oil[0.040]; traditional[0.040]; Agency[0.040]; Agency[0.040]; successful[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBeijing[0.048]; Beijing[0.048]; September[0.043]; operation[0.043]; recently[0.043]; Agency[0.042]; Agency[0.042]; Wu[0.042]; implementation[0.042]; today[0.042]; Xinhua[0.042]; Xinhua[0.042]; regions[0.042]; International[0.042]; Jiangsu[0.041]; Province[0.041]; joint[0.041]; Department[0.040]; Department[0.040]; Industry[0.040]; director[0.040]; advantages[0.039]; Group[0.039]; 1st[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jiangsu Province\u001b[39m ==> ENTITY: \u001b[32mJiangsu\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.115:0.115[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.047]; Wu[0.047]; Beijing[0.046]; Beijing[0.046]; Xinhua[0.044]; Xinhua[0.044]; coastal[0.043]; regions[0.042]; traditional[0.041]; built[0.041]; 1st[0.041]; operation[0.041]; recently[0.041]; September[0.041]; today[0.040]; plant[0.040]; plant[0.040]; director[0.040]; capacity[0.040]; build[0.040]; liquefied[0.039]; liquefied[0.039]; implementation[0.039]; Industry[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Taiwan ==> ENTITY: \u001b[32mTaiwan\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.161:0.161[0]; log p(e|m)= -0.078:-0.078[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.048]; China[0.048]; China[0.048]; Korea[0.044]; Japan[0.044]; international[0.043]; Asian[0.043]; national[0.043]; national[0.043]; coastal[0.042]; coastal[0.042]; market[0.041]; market[0.041]; development[0.040]; regions[0.039]; regions[0.039]; natural[0.039]; natural[0.039]; natural[0.039]; natural[0.039]; natural[0.039]; said[0.039]; Jiangsu[0.039]; agreement[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Xinhua News Agency\u001b[39m ==> ENTITY: \u001b[32mXinhua News Agency\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.047]; Beijing[0.047]; Beijing[0.047]; Xinhua[0.046]; News[0.045]; Agency[0.044]; Wu[0.043]; reporters[0.043]; today[0.042]; September[0.041]; International[0.041]; Department[0.041]; Department[0.041]; 1st[0.040]; recently[0.040]; reporter[0.040]; expressed[0.040]; operation[0.039]; Industry[0.039]; Ministry[0.039]; director[0.039]; implementation[0.039]; liquefied[0.039]; capacity[0.039]; \t\n\r [=============================>................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 646ms | Step: 7ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 172/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : APW20001001.2021.0521 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Egypt ==> ENTITY: \u001b[32mEgypt\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.168:0.168[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.050]; Syrian[0.047]; Mideast[0.046]; Egyptian[0.045]; Egyptian[0.045]; Palestinian[0.045]; fourth[0.045]; Iraq[0.044]; Amr[0.044]; Assad[0.043]; Assad[0.043]; West[0.043]; Gaza[0.043]; Jerusalem[0.043]; territories[0.042]; Foreign[0.042]; abroad[0.042]; Bashar[0.042]; Bank[0.041]; policy[0.041]; foreign[0.041]; talks[0.041]; talks[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syria ==> ENTITY: \u001b[32mSyria\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.213:0.213[0]; log p(e|m)= -0.261:-0.261[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEgypt[0.048]; Syrian[0.046]; Iraq[0.045]; Palestinian[0.044]; Mideast[0.043]; Gaza[0.043]; Egyptian[0.042]; Egyptian[0.042]; Palestinians[0.042]; Bank[0.041]; abroad[0.041]; Assad[0.041]; Assad[0.041]; Amr[0.041]; foreign[0.041]; Bashar[0.041]; Foreign[0.040]; territories[0.040]; West[0.040]; policy[0.040]; Jerusalem[0.040]; trip[0.039]; peace[0.039]; Sunday[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mGaza Strip\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mGaza\u001b[39m <---> \u001b[32mGaza Strip\u001b[39m\t\nSCORES: global= 0.264:0.263[\u001b[31m0.001\u001b[39m]; local()= 0.179:0.158[\u001b[31m0.020\u001b[39m]; log p(e|m)= -1.097:0.000[\u001b[32m1.097\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPalestinian[0.047]; Palestinians[0.046]; Egypt[0.045]; Syria[0.044]; Bank[0.043]; West[0.043]; Syrian[0.043]; Jerusalem[0.043]; Egyptian[0.043]; Egyptian[0.043]; Amr[0.042]; Iraq[0.041]; Mideast[0.041]; Moussa[0.041]; territories[0.040]; repercussions[0.040]; Hafez[0.040]; policy[0.040]; Assad[0.040]; Assad[0.040]; Assad[0.040]; Assad[0.040]; Assad[0.040]; Assad[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Bank ==> ENTITY: \u001b[32mWest Bank\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.182:0.182[0]; log p(e|m)= -0.054:-0.054[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPalestinian[0.047]; Egypt[0.047]; Syria[0.046]; Palestinians[0.045]; Jerusalem[0.045]; Gaza[0.045]; Syrian[0.044]; territories[0.043]; Egyptian[0.042]; Egyptian[0.042]; Amr[0.042]; Mideast[0.042]; Iraq[0.041]; Moussa[0.039]; Assad[0.039]; Assad[0.039]; Assad[0.039]; Assad[0.039]; Assad[0.039]; Assad[0.039]; peace[0.039]; peace[0.039]; agreement[0.039]; Hafez[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jerusalem ==> ENTITY: \u001b[32mJerusalem\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.165:0.165[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEgypt[0.047]; Palestinian[0.046]; Bashar[0.045]; Bashar[0.045]; West[0.045]; Palestinians[0.045]; Egyptian[0.044]; Egyptian[0.044]; Syrian[0.044]; Bank[0.044]; Gaza[0.044]; Syria[0.044]; Iraq[0.044]; territories[0.043]; day[0.043]; Amr[0.042]; Moussa[0.042]; Hafez[0.042]; Sunday[0.042]; Mideast[0.042]; trip[0.041]; visit[0.041]; visit[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = AP ==> ENTITY: \u001b[32mAssociated Press\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.051:0.051[0]; log p(e|m)= -0.687:-0.687[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nday[0.046]; Sunday[0.044]; fourth[0.043]; said[0.043]; Syrian[0.042]; Assad[0.042]; Assad[0.042]; July[0.042]; leader[0.042]; Syria[0.042]; Egypt[0.042]; president[0.041]; president[0.041]; West[0.041]; peace[0.041]; Foreign[0.041]; abroad[0.041]; foreign[0.041]; Egyptian[0.041]; Egyptian[0.041]; policy[0.040]; Iraq[0.040]; Mubarak[0.040]; Mubarak[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.157:0.157[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyrian[0.046]; Syria[0.045]; Egypt[0.044]; Egyptian[0.043]; Egyptian[0.043]; Amr[0.042]; Palestinian[0.042]; Mideast[0.042]; abroad[0.042]; Assad[0.042]; Assad[0.042]; Assad[0.042]; Assad[0.042]; Assad[0.042]; Palestinians[0.041]; Hafez[0.041]; Jerusalem[0.041]; foreign[0.041]; Bashar[0.041]; policy[0.040]; Gaza[0.040]; territories[0.040]; peace[0.039]; peace[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPresident\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPresident\u001b[39m <---> \u001b[32mPresident of Egypt\u001b[39m\t\nSCORES: global= 0.238:0.231[\u001b[31m0.007\u001b[39m]; local()= 0.088:0.197[\u001b[32m0.108\u001b[39m]; log p(e|m)= -1.704:-4.828[\u001b[31m3.125\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEgypt[0.045]; Egyptian[0.043]; Egyptian[0.043]; Palestinian[0.043]; Amr[0.043]; Hosni[0.042]; president[0.042]; president[0.042]; Mubarak[0.042]; Mubarak[0.042]; Mubarak[0.042]; Jerusalem[0.042]; Syria[0.042]; Assad[0.042]; Assad[0.042]; Assad[0.042]; Syrian[0.042]; Bashar[0.041]; Iraq[0.040]; visit[0.040]; Bank[0.040]; Palestinians[0.040]; Gaza[0.039]; Foreign[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cairo ==> ENTITY: \u001b[32mCairo\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.180:0.180[0]; log p(e|m)= -0.152:-0.152[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEgypt[0.046]; Amr[0.045]; Egyptian[0.044]; Egyptian[0.044]; Bashar[0.043]; Syrian[0.043]; Syria[0.042]; Mideast[0.042]; President[0.042]; peace[0.042]; said[0.041]; abroad[0.041]; visit[0.041]; territories[0.041]; Palestinian[0.041]; Iraq[0.041]; policy[0.040]; Hosni[0.040]; Assad[0.040]; Assad[0.040]; July[0.040]; trip[0.040]; talks[0.040]; talks[0.040]; \t\n\r [===============================>..............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 582ms | Step: 7ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 181/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : PRI20001128.2000.0055 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWashington\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mWashington (state)\u001b[39m <---> \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.259:0.259[\u001b[31m0.000\u001b[39m]; local()= 0.079:0.052[\u001b[31m0.027\u001b[39m]; log p(e|m)= -1.945:-1.749[\u001b[32m0.196\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstate[0.048]; local[0.047]; County[0.047]; election[0.046]; election[0.046]; according[0.046]; State[0.046]; Democrats[0.045]; presidential[0.045]; said[0.044]; order[0.043]; order[0.043]; Miami[0.042]; counted[0.042]; comes[0.042]; ballots[0.041]; ballots[0.041]; ballots[0.041]; recount[0.041]; recount[0.041]; news[0.041]; Florida[0.041]; officials[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Democrats ==> ENTITY: \u001b[32mDemocratic Party (United States)\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.180:-0.180[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nvotes[0.046]; Democratic[0.046]; state[0.046]; Florida[0.046]; election[0.045]; election[0.045]; election[0.045]; Tallahassee[0.044]; presidential[0.044]; presidential[0.044]; Democrats[0.043]; Washington[0.043]; Steve[0.043]; Steve[0.043]; proposal[0.043]; County[0.042]; County[0.042]; said[0.042]; officials[0.042]; officials[0.042]; Miami[0.042]; Miami[0.042]; Al[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Seattle ==> ENTITY: \u001b[32mSeattle\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.075:0.075[0]; log p(e|m)= -0.105:-0.105[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWashington[0.047]; State[0.043]; State[0.043]; State[0.043]; point[0.043]; Florida[0.043]; Florida[0.043]; tight[0.042]; reporting[0.042]; issues[0.041]; said[0.041]; margin[0.041]; Gordon[0.041]; Gordon[0.041]; fewer[0.041]; victory[0.041]; Director[0.041]; days[0.040]; timetable[0.040]; member[0.040]; probably[0.040]; Steven[0.040]; organized[0.040]; contest[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington State ==> ENTITY: \u001b[32mWashington (state)\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.074:0.074[0]; log p(e|m)= -1.924:-1.924[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nvotes[0.044]; votes[0.044]; votes[0.044]; Senator[0.044]; State[0.043]; Republican[0.043]; Republican[0.043]; election[0.043]; Seattle[0.042]; Democrat[0.042]; Senate[0.042]; Saturday[0.042]; incumbent[0.041]; Democrats[0.041]; Republicans[0.041]; ballots[0.041]; said[0.041]; Elections[0.041]; major[0.040]; race[0.040]; Florida[0.039]; Florida[0.039]; recount[0.039]; leads[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Miami Dade County\u001b[39m ==> ENTITY: \u001b[32mMiami-Dade County, Florida\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.116:0.116[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCounty[0.048]; Florida[0.048]; officials[0.047]; officials[0.047]; Beach[0.045]; State[0.045]; state[0.044]; said[0.043]; according[0.043]; considered[0.043]; total[0.043]; Steve[0.043]; Steve[0.043]; tonight[0.042]; Palm[0.042]; judge[0.042]; judge[0.042]; judge[0.042]; begin[0.042]; Judge[0.042]; Corey[0.042]; Corey[0.042]; Washington[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Al Gore\u001b[39m ==> ENTITY: \u001b[32mAl Gore\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.108:0.108[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npresidential[0.047]; Steve[0.044]; Steve[0.044]; election[0.044]; Democratic[0.043]; recount[0.043]; judge[0.042]; judge[0.042]; judge[0.042]; Judge[0.042]; ballots[0.042]; ballots[0.042]; ballots[0.042]; ballots[0.042]; votes[0.041]; said[0.040]; counting[0.040]; according[0.040]; apparently[0.040]; apparently[0.040]; issues[0.040]; Republicans[0.040]; deciding[0.039]; deciding[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Palm Beach County\u001b[39m ==> ENTITY: \u001b[32mPalm Beach County, Florida\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.115:0.115[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMiami[0.045]; Miami[0.045]; County[0.044]; state[0.043]; begin[0.043]; Dade[0.043]; officials[0.042]; officials[0.042]; tonight[0.042]; according[0.042]; Tallahassee[0.042]; presidential[0.041]; presidential[0.041]; add[0.041]; considered[0.041]; said[0.041]; total[0.041]; week[0.041]; final[0.040]; local[0.040]; local[0.040]; votes[0.040]; Judge[0.040]; counting[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NPR NEWS\u001b[39m ==> ENTITY: \u001b[32mNPR\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.084:0.084[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNPR[0.051]; Washington[0.047]; according[0.046]; Steve[0.044]; Steve[0.044]; State[0.044]; local[0.042]; Florida[0.042]; Miami[0.042]; said[0.042]; tonight[0.042]; apparently[0.041]; Dade[0.041]; Corey[0.041]; Corey[0.041]; actually[0.040]; officials[0.040]; state[0.039]; Sauls[0.039]; came[0.039]; hearing[0.039]; County[0.039]; ordered[0.039]; Flintoff[0.039]; \t\n\r [================================>.............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 522ms | Step: 7ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 189/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 20001115_AFP_ARB.0030.eng ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = AFP ==> ENTITY: \u001b[32mAgence France-Presse\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.086:0.086[0]; log p(e|m)= -0.660:-0.660[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nreported[0.046]; medical[0.046]; army[0.045]; army[0.045]; source[0.044]; today[0.044]; soldiers[0.044]; soldiers[0.044]; Israeli[0.044]; Israeli[0.044]; Israeli[0.044]; Egypt[0.044]; children[0.043]; said[0.043]; said[0.043]; Wednesday[0.043]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; began[0.042]; Medical[0.042]; head[0.042]; broke[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Egypt ==> ENTITY: \u001b[32mEgypt\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \narmy[0.045]; army[0.045]; army[0.045]; army[0.045]; military[0.044]; Israeli[0.042]; Israeli[0.042]; Israeli[0.042]; Israeli[0.042]; Israeli[0.042]; Israeli[0.042]; second[0.042]; Gaza[0.041]; Gaza[0.041]; Gaza[0.041]; soldiers[0.041]; soldiers[0.041]; southern[0.040]; southern[0.040]; Palestinian[0.039]; Palestinian[0.039]; Palestinian[0.039]; Palestinian[0.039]; authorities[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mGaza\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mGaza Strip\u001b[39m <---> \u001b[32mGaza\u001b[39m\t\nSCORES: global= 0.251:0.246[\u001b[31m0.004\u001b[39m]; local()= 0.204:0.201[\u001b[31m0.003\u001b[39m]; log p(e|m)= -1.374:-0.304[\u001b[32m1.071\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGaza[0.047]; Palestinian[0.046]; Palestinian[0.046]; Palestinian[0.046]; Israeli[0.046]; Israeli[0.046]; Israeli[0.046]; Palestinians[0.044]; Palestinians[0.044]; Rafah[0.044]; Egypt[0.043]; southern[0.042]; border[0.041]; army[0.039]; army[0.039]; near[0.039]; said[0.038]; said[0.038]; Strip[0.038]; town[0.038]; reported[0.038]; soldiers[0.038]; Wednesday[0.038]; bulldozing[0.037]; \t\n\r [=================================>............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 498ms | Step: 7ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 193/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : PRI20001122.2000.0320 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Corey Flintoff\u001b[39m ==> ENTITY: \u001b[32mCorey Flintoff\u001b[39m\t\nSCORES: global= 0.224:0.224[0]; local()= 0.042:0.042[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWashington[0.046]; NPR[0.046]; Northern[0.045]; Minister[0.044]; Prime[0.043]; said[0.043]; said[0.043]; today[0.043]; today[0.043]; injured[0.042]; faction[0.041]; Yasser[0.041]; soldiers[0.041]; Hadera[0.041]; deaths[0.041]; news[0.040]; people[0.040]; day[0.040]; Ehud[0.040]; headed[0.040]; bombing[0.040]; members[0.040]; town[0.039]; avenge[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NPR NEWS\u001b[39m ==> ENTITY: \u001b[32mNPR\u001b[39m\t\nSCORES: global= 0.234:0.234[0]; local()= 0.026:0.026[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWashington[0.048]; Northern[0.045]; today[0.043]; today[0.043]; said[0.043]; said[0.043]; Minister[0.043]; Prime[0.043]; Israeli[0.042]; Israeli[0.042]; Israeli[0.042]; Corey[0.042]; day[0.041]; injured[0.041]; members[0.041]; faction[0.041]; soldiers[0.040]; deaths[0.040]; Flintoff[0.040]; Palestinian[0.040]; people[0.039]; Hadera[0.039]; car[0.039]; bombing[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.042:0.042[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntown[0.046]; people[0.044]; Corey[0.043]; soldiers[0.043]; deaths[0.043]; today[0.043]; today[0.043]; news[0.043]; said[0.043]; said[0.043]; Barak[0.043]; day[0.042]; Northern[0.042]; NPR[0.041]; car[0.041]; members[0.040]; bombing[0.040]; killed[0.040]; killed[0.040]; injured[0.040]; headed[0.040]; retaliate[0.039]; Prime[0.039]; Hadera[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yasser Arafat\u001b[39m ==> ENTITY: \u001b[32mYasser Arafat\u001b[39m\t\nSCORES: global= 0.234:0.234[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsraeli[0.050]; Israeli[0.050]; Israeli[0.050]; Palestinian[0.049]; Barak[0.048]; Hadera[0.045]; Ehud[0.044]; bombing[0.041]; said[0.041]; said[0.041]; soldiers[0.040]; faction[0.040]; Washington[0.040]; Northern[0.040]; members[0.039]; Prime[0.039]; day[0.039]; killed[0.038]; killed[0.038]; news[0.038]; Minister[0.038]; today[0.038]; today[0.038]; retaliate[0.038]; \t\n\r [==================================>...........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 494ms | Step: 8ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 197/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : PRI20001031.2000.1824 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.053:0.053[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \naircraft[0.047]; flight[0.046]; BBC[0.044]; Flight[0.043]; plane[0.043]; plane[0.043]; flying[0.043]; runway[0.042]; shortly[0.042]; said[0.042]; Airlines[0.042]; Airlines[0.042]; passengers[0.041]; passengers[0.041]; David[0.041]; passenger[0.041]; people[0.041]; News[0.040]; American[0.040]; reports[0.040]; World[0.039]; carrying[0.039]; split[0.039]; decided[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Taiwan ==> ENTITY: \u001b[32mTaiwan\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.078:-0.078[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.047]; Singapore[0.047]; people[0.045]; said[0.043]; World[0.043]; American[0.042]; reports[0.042]; split[0.042]; David[0.042]; Airlines[0.042]; Airlines[0.042]; flight[0.041]; crashed[0.041]; crashed[0.041]; remain[0.041]; called[0.040]; London[0.040]; aircraft[0.040]; shortly[0.040]; decided[0.040]; flying[0.040]; carrying[0.039]; turned[0.039]; Flight[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Los Angeles ==> ENTITY: \u001b[32mLos Angeles\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.050:0.050[0]; log p(e|m)= -0.101:-0.101[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nflight[0.046]; shortly[0.044]; Flight[0.043]; plane[0.043]; plane[0.043]; runway[0.043]; American[0.043]; David[0.042]; Airlines[0.042]; Airlines[0.042]; board[0.042]; London[0.042]; leaving[0.042]; flying[0.041]; aircraft[0.041]; Singapore[0.041]; Singapore[0.041]; World[0.040]; crew[0.040]; passenger[0.040]; said[0.040]; Taiwan[0.039]; News[0.039]; times[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = BBC News\u001b[39m ==> ENTITY: \u001b[32mBBC News\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.090:0.090[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nreports[0.050]; London[0.048]; World[0.047]; David[0.046]; American[0.045]; shortly[0.044]; said[0.044]; people[0.042]; Singapore[0.041]; Singapore[0.041]; described[0.041]; flight[0.041]; flying[0.040]; Airlines[0.040]; Airlines[0.040]; crew[0.040]; remain[0.040]; decided[0.040]; Taiwan[0.039]; Flight[0.039]; passengers[0.039]; passengers[0.039]; split[0.038]; board[0.038]; \t\n\r [==================================>...........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 466ms | Step: 8ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 201/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 20001115_AFP_ARB.0212.eng ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Madeleine Albright\u001b[39m ==> ENTITY: \u001b[32mMadeleine Albright\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSecretary[0.046]; President[0.045]; Clinton[0.044]; American[0.043]; university[0.043]; university[0.043]; Richard[0.043]; administration[0.043]; Washington[0.043]; University[0.042]; State[0.041]; State[0.041]; Department[0.041]; week[0.041]; Bill[0.041]; America[0.040]; Yale[0.040]; envoy[0.040]; Ross[0.040]; diplomat[0.040]; Levin[0.040]; January[0.040]; said[0.040]; Chancellor[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Richard Levin\u001b[39m ==> ENTITY: \u001b[32mRick Levin\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUniversity[0.045]; graduate[0.045]; Yale[0.044]; Yale[0.044]; university[0.044]; university[0.044]; announced[0.043]; American[0.042]; announce[0.042]; ranking[0.041]; President[0.041]; Connecticut[0.041]; prestigious[0.041]; Chancellor[0.041]; Dennis[0.041]; Bill[0.040]; State[0.040]; State[0.040]; administration[0.040]; administration[0.040]; resigning[0.040]; January[0.040]; July[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Time magazine\u001b[39m ==> ENTITY: \u001b[32mTime (magazine)\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.075:0.075[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.047]; personal[0.046]; university[0.046]; university[0.046]; prestigious[0.045]; week[0.044]; ranking[0.044]; Levin[0.044]; Ross[0.044]; said[0.044]; correspondent[0.043]; current[0.043]; State[0.043]; days[0.043]; Dennis[0.043]; Yale[0.042]; comes[0.042]; term[0.042]; term[0.042]; Bill[0.042]; January[0.042]; Chancellor[0.042]; Affairs[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yale University\u001b[39m ==> ENTITY: \u001b[32mYale University\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.129:0.129[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConnecticut[0.047]; prestigious[0.046]; university[0.045]; university[0.045]; American[0.045]; Richard[0.045]; America[0.044]; Washington[0.044]; announced[0.044]; President[0.043]; East[0.043]; Middle[0.043]; January[0.043]; week[0.043]; Bill[0.043]; July[0.043]; Center[0.042]; Dennis[0.042]; Chancellor[0.042]; Ross[0.042]; Department[0.042]; State[0.042]; State[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Middle East\u001b[39m ==> ENTITY: \u001b[32mMiddle East\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.081:0.081[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.047]; administration[0.043]; administration[0.043]; University[0.042]; Studies[0.042]; graduate[0.042]; university[0.042]; university[0.042]; President[0.042]; Globalization[0.042]; ambassador[0.042]; appointed[0.041]; announced[0.041]; diplomat[0.041]; post[0.041]; Levin[0.041]; Yale[0.041]; Yale[0.041]; northeast[0.041]; Soviet[0.041]; Department[0.041]; resigning[0.041]; January[0.041]; prestigious[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = America ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.088:0.088[0]; log p(e|m)= -0.785:-0.785[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRichard[0.046]; American[0.046]; Washington[0.044]; Connecticut[0.044]; Bill[0.043]; January[0.043]; President[0.042]; Stewart[0.042]; said[0.042]; Department[0.041]; July[0.041]; state[0.041]; administration[0.041]; resigning[0.041]; second[0.041]; second[0.041]; second[0.041]; decided[0.041]; announce[0.041]; Clinton[0.040]; charge[0.040]; diplomat[0.040]; days[0.040]; northeast[0.040]; \t\n\r [====================================>.........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 410ms | Step: 8ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 208/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 20001015_AFP_ARB.0053.eng ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = AFP ==> ENTITY: \u001b[32mAgence France-Presse\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.101:0.101[0]; log p(e|m)= -0.660:-0.660[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBureau[0.046]; Bureau[0.046]; statement[0.044]; yesterday[0.043]; saying[0.043]; released[0.043]; Beirut[0.042]; Federation[0.042]; Federation[0.042]; Saudi[0.041]; Saudi[0.041]; Saudi[0.041]; Saudi[0.041]; Sports[0.041]; Sports[0.041]; president[0.041]; president[0.041]; today[0.041]; post[0.041]; coach[0.040]; Youth[0.039]; Youth[0.039]; Ben[0.039]; assistant[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Emirates ==> ENTITY: \u001b[32mUnited Arab Emirates\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.203:0.203[0]; log p(e|m)= -2.718:-2.718[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEmirates[0.052]; soccer[0.048]; Lebanon[0.047]; Saudi[0.047]; Saudi[0.047]; Saudi[0.047]; Saudi[0.047]; Kuwaiti[0.046]; Championship[0.046]; Championship[0.046]; Cup[0.044]; Cup[0.044]; fourth[0.044]; Gulf[0.044]; Gulf[0.044]; team[0.044]; team[0.044]; team[0.044]; team[0.044]; team[0.044]; second[0.044]; October[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanon ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.083:0.083[0]; log p(e|m)= -0.409:-0.409[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEmirates[0.043]; Emirates[0.043]; Cup[0.043]; Cup[0.043]; finals[0.043]; fourth[0.043]; team[0.043]; team[0.043]; team[0.043]; team[0.043]; team[0.043]; series[0.043]; soccer[0.042]; Kuwaiti[0.042]; region[0.042]; championship[0.041]; France[0.041]; win[0.041]; Soccer[0.040]; place[0.040]; Championship[0.040]; Championship[0.040]; Federation[0.040]; Federation[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bangkok\u001b[39m ==> ENTITY: \u001b[32mBangkok\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.001:-0.001[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChampionship[0.044]; Championship[0.044]; February[0.043]; place[0.043]; place[0.043]; Cup[0.043]; team[0.043]; team[0.043]; team[0.043]; July[0.043]; series[0.042]; fourth[0.042]; win[0.041]; second[0.041]; Asian[0.041]; Asian[0.041]; hosted[0.041]; title[0.041]; especially[0.040]; took[0.040]; region[0.040]; Games[0.039]; late[0.039]; holds[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexico ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.102:0.102[0]; log p(e|m)= -0.183:-0.183[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.045]; Cup[0.045]; soccer[0.045]; team[0.044]; team[0.044]; team[0.044]; team[0.044]; Championship[0.043]; Championship[0.043]; finals[0.042]; second[0.042]; fourth[0.041]; region[0.041]; France[0.041]; place[0.040]; place[0.040]; Games[0.040]; win[0.040]; level[0.040]; regarding[0.040]; Machala[0.039]; Machala[0.039]; Machala[0.039]; Machala[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Beirut ==> ENTITY: \u001b[32mBeirut\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.047:-0.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNasser[0.044]; Jawhar[0.044]; Arabia[0.044]; Saudi[0.043]; Saudi[0.043]; Saudi[0.043]; Saudi[0.043]; Ben[0.042]; released[0.042]; team[0.042]; team[0.042]; saying[0.042]; Fahd[0.042]; statement[0.041]; AFP[0.041]; post[0.041]; match[0.041]; Sultan[0.041]; today[0.041]; Sunday[0.041]; president[0.040]; president[0.040]; Czech[0.040]; yesterday[0.040]; \t\n\r [=====================================>........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 360ms | Step: 8ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 214/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : VOA20001020.2100.1853 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mUSS Cole\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUSS Cole bombing\u001b[39m <---> \u001b[32mUSS Cole (DDG-67)\u001b[39m\t\nSCORES: global= 0.273:0.262[\u001b[31m0.012\u001b[39m]; local()= 0.121:0.137[\u001b[32m0.016\u001b[39m]; log p(e|m)= -0.154:-0.396[\u001b[31m0.242\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nship[0.047]; week[0.046]; Yemen[0.045]; navy[0.044]; navy[0.044]; Friday[0.043]; blast[0.043]; explosives[0.043]; incident[0.043]; explosion[0.043]; near[0.042]; boat[0.042]; refueled[0.042]; carrying[0.042]; killing[0.041]; statement[0.041]; able[0.040]; set[0.040]; given[0.039]; says[0.039]; revised[0.038]; small[0.038]; accounting[0.038]; raising[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Yemen ==> ENTITY: \u001b[32mYemen\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.070:0.070[0]; log p(e|m)= -0.272:-0.272[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nweek[0.045]; blast[0.044]; set[0.044]; Friday[0.044]; statement[0.043]; explosion[0.043]; given[0.043]; revised[0.043]; near[0.042]; ship[0.042]; accounting[0.041]; carrying[0.041]; explosives[0.041]; small[0.041]; navy[0.041]; navy[0.041]; incident[0.041]; killing[0.040]; says[0.040]; raising[0.040]; able[0.040]; USS[0.040]; ripped[0.040]; boat[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mnavy\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNavy\u001b[39m <---> \u001b[32mUnited States Navy\u001b[39m\t\nSCORES: global= 0.248:0.239[\u001b[31m0.009\u001b[39m]; local()= 0.087:0.070[\u001b[31m0.017\u001b[39m]; log p(e|m)= -0.707:-1.864[\u001b[31m1.157\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nship[0.050]; navy[0.049]; USS[0.047]; boat[0.045]; able[0.044]; small[0.042]; carrying[0.042]; week[0.042]; given[0.042]; Cole[0.041]; refueled[0.041]; explosives[0.041]; Yemen[0.041]; statement[0.040]; explosion[0.040]; says[0.040]; Friday[0.040]; revised[0.040]; incident[0.040]; near[0.039]; accounting[0.039]; set[0.039]; ripped[0.038]; raising[0.038]; \t\n\r [=====================================>........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 338ms | Step: 8ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 217/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 20000715_AFP_ARB.0072.eng ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = AFP ==> ENTITY: \u001b[32mAgence France-Presse\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.660:-0.660[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.046]; military[0.046]; journalists[0.044]; official[0.044]; official[0.044]; official[0.044]; told[0.043]; missiles[0.041]; missiles[0.041]; army[0.041]; high[0.041]; Iran[0.041]; Iran[0.041]; level[0.041]; capability[0.041]; capability[0.041]; Israel[0.040]; Israel[0.040]; Israel[0.040]; Israeli[0.040]; said[0.040]; today[0.040]; space[0.040]; equip[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iran ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.138:-0.138[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.047]; Israel[0.047]; Israel[0.047]; Iran[0.047]; Israeli[0.044]; military[0.044]; military[0.044]; Tehran[0.043]; nuclear[0.042]; nuclear[0.042]; nuclear[0.042]; nuclear[0.042]; official[0.041]; official[0.041]; official[0.041]; missiles[0.040]; missiles[0.040]; announced[0.039]; Shahab[0.039]; said[0.038]; army[0.038]; start[0.038]; today[0.038]; today[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.074:0.074[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.055]; reached[0.053]; military[0.053]; reach[0.051]; completed[0.051]; reaching[0.050]; program[0.050]; announced[0.050]; Saturday[0.050]; space[0.049]; start[0.049]; kilometers[0.049]; kilometers[0.049]; ground[0.049]; ground[0.049]; ground[0.049]; ground[0.049]; strategic[0.049]; strategic[0.049]; North[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jerusalem ==> ENTITY: \u001b[32mJerusalem\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israel[0.050]; Israel[0.050]; Israeli[0.047]; set[0.043]; today[0.041]; army[0.041]; military[0.041]; military[0.041]; start[0.040]; said[0.040]; official[0.040]; official[0.040]; believes[0.040]; begin[0.040]; Saturday[0.040]; told[0.040]; position[0.040]; position[0.040]; wished[0.039]; purposes[0.039]; purposes[0.039]; remain[0.039]; space[0.039]; \t\n\r [======================================>.......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 295ms | Step: 8ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 222/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : PRI20001201.2000.1828 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.060:0.060[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbuilding[0.048]; George[0.046]; BBC[0.046]; held[0.046]; election[0.045]; crucial[0.044]; stately[0.044]; Democrat[0.044]; Court[0.043]; Court[0.043]; Court[0.043]; High[0.043]; World[0.043]; Judith[0.043]; Spencer[0.043]; Federal[0.043]; lawyers[0.042]; News[0.042]; Ruth[0.042]; Republican[0.042]; Bater[0.042]; argued[0.042]; results[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.070:0.070[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncounties[0.046]; Miami[0.044]; said[0.043]; State[0.043]; State[0.043]; State[0.043]; early[0.043]; legislature[0.042]; Florida[0.042]; Florida[0.042]; Dade[0.041]; line[0.041]; does[0.041]; rejected[0.041]; Beach[0.041]; suggested[0.040]; deadline[0.040]; Day[0.040]; date[0.040]; days[0.040]; count[0.040]; World[0.040]; know[0.040]; Al[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = US Supreme Court\u001b[39m ==> ENTITY: \u001b[32mSupreme Court of the United States\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nvalidity[0.044]; Court[0.044]; Court[0.044]; Court[0.044]; legal[0.044]; Republican[0.043]; Federal[0.042]; argued[0.042]; Supreme[0.042]; Supreme[0.042]; state[0.041]; presidential[0.041]; election[0.041]; decide[0.041]; lawyers[0.041]; Florida[0.041]; Florida[0.041]; George[0.040]; hearing[0.040]; Gore[0.040]; Bush[0.040]; Bush[0.040]; Democrat[0.040]; held[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sandra Day O'Connor\u001b[39m ==> ENTITY: \u001b[32mSandra Day O'Connor\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nvalidity[0.053]; disputed[0.051]; court[0.050]; court[0.050]; court[0.050]; rejected[0.050]; Court[0.050]; Court[0.050]; Court[0.050]; Court[0.050]; Court[0.050]; Court[0.050]; decision[0.050]; Justice[0.049]; ballots[0.049]; argued[0.049]; George[0.049]; Ginsburg[0.049]; appeal[0.049]; rule[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Florida Supreme Court\u001b[39m ==> ENTITY: \u001b[32mSupreme Court of Florida\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncourt[0.045]; court[0.045]; court[0.045]; Court[0.045]; Court[0.045]; Court[0.045]; Court[0.045]; Court[0.045]; decision[0.045]; hearing[0.044]; Justice[0.044]; legal[0.044]; argued[0.043]; election[0.042]; Florida[0.042]; Florida[0.042]; decide[0.042]; legislature[0.042]; lawyers[0.042]; Supreme[0.042]; Supreme[0.042]; Supreme[0.042]; Supreme[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = BBC News\u001b[39m ==> ENTITY: \u001b[32mBBC News\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.101:0.101[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nreports[0.052]; London[0.050]; World[0.049]; George[0.045]; Judith[0.044]; Inside[0.044]; results[0.044]; election[0.043]; hearing[0.043]; argued[0.043]; held[0.043]; lawyers[0.042]; legal[0.042]; dispute[0.042]; crucial[0.042]; Ruth[0.042]; Bush[0.042]; Bush[0.042]; High[0.042]; Spencer[0.042]; building[0.042]; counts[0.041]; wrangling[0.041]; \t\n\r [=======================================>......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 236ms | Step: 8ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 229/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 20001115_AFP_ARB.0093.eng ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = AFP ==> ENTITY: \u001b[32mAgence France-Presse\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.100:0.100[0]; log p(e|m)= -0.660:-0.660[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nissued[0.043]; President[0.043]; President[0.043]; release[0.043]; release[0.043]; American[0.043]; spokesman[0.043]; today[0.043]; announcement[0.042]; Authority[0.042]; form[0.042]; said[0.042]; Wednesday[0.041]; Palestinian[0.041]; Palestinian[0.041]; considered[0.041]; James[0.041]; control[0.041]; territories[0.041]; order[0.041]; States[0.040]; ceasefire[0.040]; deemed[0.040]; terms[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.059:0.059[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.047]; said[0.044]; James[0.044]; Wednesday[0.043]; today[0.043]; territories[0.043]; considered[0.043]; White[0.042]; visiting[0.042]; terms[0.042]; order[0.042]; positively[0.041]; release[0.041]; release[0.041]; welcome[0.041]; Bill[0.041]; announcement[0.041]; constitute[0.040]; positive[0.040]; Brunei[0.040]; Palestinian[0.040]; Palestinian[0.040]; deemed[0.040]; form[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = White House\u001b[39m ==> ENTITY: \u001b[32mWhite House\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.006:-0.006[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.045]; American[0.045]; President[0.044]; President[0.044]; President[0.044]; Clinton[0.043]; Clinton[0.043]; Clinton[0.043]; Economic[0.042]; James[0.042]; territories[0.042]; visiting[0.042]; States[0.041]; announcement[0.041]; Wednesday[0.040]; release[0.040]; release[0.040]; officials[0.040]; issued[0.040]; terms[0.040]; Bill[0.040]; Bill[0.040]; agreement[0.040]; region[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bandar Seri Begawan\u001b[39m ==> ENTITY: \u001b[32mBandar Seri Begawan\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrunei[0.049]; Sultanate[0.045]; capital[0.045]; Begawan[0.045]; Seri[0.044]; Bandar[0.043]; today[0.042]; territories[0.042]; spokesman[0.041]; Authority[0.041]; House[0.041]; Wednesday[0.041]; considered[0.041]; visiting[0.041]; AFP[0.041]; course[0.041]; ceasefire[0.040]; said[0.040]; welcome[0.040]; agreement[0.040]; Sheikh[0.040]; Yasser[0.039]; Sharm[0.039]; constitute[0.039]; \t\n\r [========================================>.....] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 195ms | Step: 8ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 234/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : NYT20001125.1558.0117 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mich.\u001b[39m ==> ENTITY: \u001b[32mMichigan\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.069:0.069[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMichigan[0.048]; Michigan[0.048]; Michigan[0.048]; UR[0.046]; season[0.045]; season[0.045]; season[0.045]; like[0.042]; legal[0.042]; state[0.042]; practice[0.041]; LR[0.041]; QL[0.041]; states[0.040]; York[0.040]; PARK[0.040]; safety[0.040]; education[0.038]; relatively[0.038]; Service[0.038]; opened[0.038]; News[0.038]; Nov[0.038]; Nov[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pennsylvania ==> ENTITY: \u001b[32mPennsylvania\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.118:-0.118[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstate[0.045]; states[0.045]; York[0.044]; said[0.042]; Michigan[0.042]; Michigan[0.042]; Michigan[0.042]; Michigan[0.042]; Michigan[0.042]; Michigan[0.042]; efforts[0.042]; seven[0.042]; Department[0.042]; officials[0.041]; officials[0.041]; began[0.041]; relatively[0.041]; education[0.041]; expect[0.041]; practice[0.040]; practice[0.040]; practice[0.040]; law[0.040]; ground[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew York\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew York City\u001b[39m <---> \u001b[32mNew York\u001b[39m\t\nSCORES: global= 0.259:0.258[\u001b[31m0.001\u001b[39m]; local()= 0.063:0.085[\u001b[32m0.022\u001b[39m]; log p(e|m)= -1.790:-0.289[\u001b[32m1.501\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPennsylvania[0.045]; state[0.044]; Department[0.043]; long[0.043]; began[0.043]; said[0.042]; seven[0.042]; administrator[0.042]; continues[0.041]; appears[0.041]; kind[0.041]; exercised[0.041]; officials[0.041]; officials[0.041]; states[0.041]; shooting[0.041]; shooting[0.041]; shooting[0.041]; Michigan[0.041]; Michigan[0.041]; Michigan[0.041]; Michigan[0.041]; Michigan[0.041]; Michigan[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Detroit ==> ENTITY: \u001b[32mDetroit\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.049:0.049[0]; log p(e|m)= -0.217:-0.217[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWilliams[0.048]; Williams[0.048]; kind[0.048]; west[0.047]; division[0.047]; field[0.046]; Rob[0.046]; recent[0.046]; practice[0.045]; morning[0.045]; miles[0.045]; ground[0.045]; ground[0.045]; procedures[0.045]; Art[0.045]; tied[0.044]; hard[0.044]; following[0.044]; guard[0.044]; haul[0.044]; shots[0.044]; thump[0.044]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mDepartment of Natural Resources\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMinnesota Department of Natural Resources\u001b[39m <---> \u001b[32mMichigan Department of Natural Resources\u001b[39m\t\nSCORES: global= 0.246:0.242[\u001b[31m0.004\u001b[39m]; local()= 0.186:0.176[\u001b[31m0.010\u001b[39m]; log p(e|m)= -0.906:-2.163[\u001b[31m1.256\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMichigan[0.043]; Michigan[0.043]; Michigan[0.043]; Michigan[0.043]; Michigan[0.043]; Michigan[0.043]; officials[0.043]; hunting[0.043]; hunting[0.043]; hunting[0.043]; hunters[0.042]; hunters[0.042]; hunters[0.042]; hunters[0.042]; hunters[0.042]; enforcement[0.042]; learn[0.041]; division[0.041]; trees[0.040]; ground[0.040]; woods[0.040]; states[0.039]; deer[0.039]; administrator[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Grosse Pointe Park\u001b[39m ==> ENTITY: \u001b[32mGrosse Pointe Park, Michigan\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.042:0.042[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUR[0.052]; LR[0.046]; QL[0.046]; Michigan[0.044]; Michigan[0.044]; Michigan[0.044]; state[0.044]; season[0.044]; season[0.044]; season[0.044]; shooting[0.043]; education[0.043]; relatively[0.043]; states[0.042]; heavy[0.042]; climb[0.042]; Firearm[0.042]; safety[0.042]; deer[0.042]; deer[0.042]; busiest[0.042]; men[0.042]; opened[0.042]; \t\n\r [=========================================>....] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 145ms | Step: 8ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 240/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 20001115_AFP_ARB.0217.eng ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAFP\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustralian Federal Police\u001b[39m <---> \u001b[32mAgence France-Presse\u001b[39m\t\nSCORES: global= 0.244:0.243[\u001b[31m0.001\u001b[39m]; local()= 0.107:0.104[\u001b[31m0.002\u001b[39m]; log p(e|m)= -2.375:-0.660[\u001b[32m1.715\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nofficial[0.046]; official[0.046]; announced[0.045]; President[0.043]; today[0.043]; decision[0.043]; said[0.042]; forum[0.042]; Wednesday[0.042]; Economic[0.041]; peace[0.041]; responsible[0.041]; agreed[0.041]; World[0.041]; current[0.041]; countries[0.041]; meeting[0.041]; signing[0.041]; hold[0.040]; hold[0.040]; Minister[0.040]; Asia[0.040]; solution[0.039]; treaty[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.050]; Russian[0.047]; Vladimir[0.045]; Putin[0.042]; War[0.042]; Islands[0.042]; World[0.042]; treaty[0.042]; countries[0.041]; peace[0.041]; official[0.041]; official[0.041]; Japan[0.041]; President[0.041]; signing[0.041]; Pacific[0.040]; agreed[0.040]; Asia[0.040]; Wednesday[0.040]; hindered[0.040]; decision[0.040]; dispute[0.040]; east[0.040]; today[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brunei ==> ENTITY: \u001b[32mBrunei\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.179:0.179[0]; log p(e|m)= -0.387:-0.387[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nofficial[0.045]; official[0.045]; countries[0.044]; Japan[0.043]; Japan[0.043]; Japan[0.043]; Japan[0.043]; islands[0.043]; Asia[0.043]; Islands[0.042]; Wednesday[0.041]; Russian[0.041]; Russian[0.041]; Russian[0.041]; meet[0.040]; Japanese[0.040]; Japanese[0.040]; Japanese[0.040]; Japanese[0.040]; today[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bandar Seri Begawan\u001b[39m ==> ENTITY: \u001b[32mBandar Seri Begawan\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.102:0.102[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrunei[0.049]; Asia[0.043]; World[0.043]; II[0.043]; today[0.042]; Japan[0.042]; countries[0.042]; current[0.042]; Wednesday[0.041]; Pacific[0.041]; official[0.041]; official[0.041]; hold[0.041]; hold[0.041]; AFP[0.041]; Yoshiro[0.041]; Islands[0.041]; forum[0.041]; Minister[0.041]; announced[0.040]; Prime[0.040]; Economic[0.040]; Vladimir[0.040]; said[0.040]; \t\n\r [==========================================>...] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 102ms | Step: 8ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 245/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : VOA20001129.2000.0364 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Supreme Court ==> ENTITY: \u001b[32mSupreme Court of the United States\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.149:0.149[0]; log p(e|m)= -0.994:-0.994[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCourts[0.047]; courts[0.046]; Court[0.046]; case[0.044]; case[0.044]; law[0.044]; constitutional[0.044]; Federal[0.043]; states[0.042]; Supreme[0.041]; state[0.041]; President[0.041]; vote[0.041]; Washington[0.041]; unprecedented[0.040]; University[0.040]; Florida[0.040]; elections[0.039]; issue[0.039]; week[0.039]; count[0.039]; took[0.039]; reject[0.039]; instance[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Baltimore ==> ENTITY: \u001b[32mBaltimore\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.200:-0.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWashington[0.051]; University[0.051]; Florida[0.047]; School[0.047]; Lawrence[0.046]; Tribe[0.046]; states[0.046]; week[0.046]; Joe[0.046]; state[0.045]; professor[0.045]; Federal[0.045]; took[0.045]; unprecedented[0.045]; hold[0.045]; kind[0.044]; Courts[0.044]; Hopkins[0.044]; Professor[0.044]; Law[0.043]; Harvard[0.043]; law[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gore ==> ENTITY: \u001b[32mAl Gore\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.102:0.102[0]; log p(e|m)= -1.204:-1.204[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncampaign[0.045]; Gore[0.045]; states[0.043]; want[0.043]; like[0.043]; election[0.043]; kind[0.043]; kind[0.043]; Bush[0.042]; Court[0.041]; fact[0.041]; probably[0.041]; probably[0.041]; ask[0.041]; try[0.041]; try[0.041]; particular[0.041]; says[0.041]; arguing[0.040]; Lawrence[0.040]; law[0.040]; make[0.040]; issue[0.040]; issues[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Congress ==> ENTITY: \u001b[32mUnited States Congress\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.107:0.107[0]; log p(e|m)= -0.335:-0.335[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.045]; legislature[0.044]; legislature[0.044]; Congress[0.043]; state[0.043]; George[0.042]; constitutional[0.042]; order[0.042]; week[0.042]; law[0.042]; Friday[0.042]; decide[0.041]; decide[0.041]; electors[0.041]; electors[0.041]; issue[0.041]; specific[0.041]; Republican[0.041]; accept[0.041]; briefs[0.040]; ultimately[0.040]; Vice[0.040]; certain[0.040]; slates[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lawrence Tribe\u001b[39m ==> ENTITY: \u001b[32mLaurence Tribe\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCourts[0.045]; Law[0.045]; professor[0.043]; constitutional[0.043]; dismiss[0.042]; court[0.042]; court[0.042]; Professor[0.042]; Harvard[0.042]; Court[0.042]; University[0.042]; case[0.042]; case[0.042]; case[0.042]; case[0.042]; case[0.042]; law[0.041]; arguing[0.041]; Supreme[0.041]; issues[0.040]; general[0.040]; unprecedented[0.040]; School[0.040]; recognizes[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Johns Hopkins University\u001b[39m ==> ENTITY: \u001b[32mJohns Hopkins University\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.074:0.074[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWashington[0.047]; Harvard[0.045]; Law[0.044]; President[0.044]; Baltimore[0.044]; professor[0.043]; Professor[0.043]; kind[0.043]; Lawrence[0.042]; Florida[0.041]; unseemly[0.041]; School[0.041]; unprecedented[0.041]; week[0.041]; felt[0.040]; Federal[0.040]; hold[0.040]; Tribe[0.040]; Joe[0.040]; instance[0.040]; like[0.040]; probably[0.039]; Courts[0.039]; states[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.072:0.072[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUniversity[0.050]; state[0.046]; states[0.046]; professor[0.045]; President[0.044]; Baltimore[0.044]; vote[0.043]; elections[0.042]; virtually[0.041]; unprecedented[0.041]; Federal[0.041]; says[0.040]; week[0.040]; News[0.040]; constitutional[0.040]; responsibility[0.040]; instance[0.040]; hour[0.040]; Courts[0.040]; Florida[0.040]; VOA[0.040]; probably[0.040]; law[0.039]; think[0.039]; \t\n\r [============================================>.] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 25ms | Step: 8ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 254/257 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : VOA20001223.2000.0139 ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Slobodan Milosevic\u001b[39m ==> ENTITY: \u001b[32mSlobodan Milošević\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYugoslav[0.045]; Yugoslav[0.045]; Vojislav[0.044]; Serbia[0.044]; Milosevic[0.043]; opposition[0.043]; coalition[0.042]; Serbian[0.042]; ousted[0.042]; allies[0.042]; allies[0.042]; President[0.042]; President[0.042]; President[0.042]; uprising[0.041]; Party[0.041]; today[0.040]; elections[0.040]; October[0.040]; Socialist[0.040]; parliamentary[0.039]; vote[0.039]; vote[0.039]; Mr[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSocialist Party\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSocialist Party (France)\u001b[39m <---> \u001b[32mSocialist Party of Serbia\u001b[39m\t\nSCORES: global= 0.243:0.237[\u001b[31m0.005\u001b[39m]; local()= 0.101:0.164[\u001b[32m0.064\u001b[39m]; log p(e|m)= -1.336:-4.510[\u001b[31m3.174\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nparliamentary[0.047]; elections[0.046]; Vojislav[0.044]; coalition[0.044]; Serbia[0.043]; Yugoslav[0.043]; Yugoslav[0.043]; Serbian[0.042]; Slobodan[0.042]; opposition[0.042]; vote[0.041]; vote[0.041]; October[0.041]; victory[0.041]; Reformist[0.041]; Democratic[0.041]; President[0.040]; President[0.040]; President[0.040]; ousted[0.040]; allies[0.040]; allies[0.040]; won[0.039]; won[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = DEMOCRATIC OPPOSITION\u001b[39m ==> ENTITY: \u001b[32mDemocratic Opposition of Serbia\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVojislav[0.047]; coalition[0.046]; ousted[0.045]; Party[0.045]; Serbia[0.045]; parliamentary[0.045]; Slobodan[0.044]; won[0.043]; won[0.043]; elections[0.043]; Serbian[0.043]; Yugoslav[0.043]; Yugoslav[0.043]; Socialist[0.043]; allies[0.043]; allies[0.043]; President[0.042]; President[0.042]; President[0.042]; uprising[0.042]; Mr[0.042]; Milosevic[0.042]; Milosevic[0.042]; \t\n\r [=============================================>] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b Tot: 2s207ms | Step: 8ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 257/257 \n\r==> time to test 1 sample = 8.5906416525637ms\t\n==> \u001b[31mACE04\u001b[39m ACE04 ; EPOCH = 307: Micro recall = 84.05% ; Micro F1 = \u001b[31m86.92%\u001b[39m\t\n ===> entity frequency stats :\t\nfreq = 11-20 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\nfreq = 2-5 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\nfreq = 50+ : num = 233 ; correctly classified = 216 ; perc = 92.70\t\nfreq = 1 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\nfreq = 21-50 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\nfreq = 0 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\nfreq = 6-10 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\n\t\n ===> entity p(e|m) stats :\t\np(e|m) = <=0.001 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\np(e|m) = 0.01-0.03 : num = 3 ; correctly classified = 2 ; perc = 66.67\t\np(e|m) = 0.003-0.01 : num = 1 ; correctly classified = 0 ; perc = 0.00\t\np(e|m) = 0.3+ : num = 209 ; correctly classified = 200 ; perc = 95.69\t\np(e|m) = 0.1-0.3 : num = 17 ; correctly classified = 12 ; perc = 70.59\t\np(e|m) = 0.03-0.1 : num = 3 ; correctly classified = 2 ; perc = 66.67\t\np(e|m) = 0.001-0.003 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\n num_mentions_w/o_gold_ent_in_candidates = 24 total num mentions in dataset = 257\t\n percentage_mentions_w/o_gold_ent_in_candidates = 9.34%; percentage_mentions_solved : both_pem_ours = 80.93%; only_pem_not_ours = 3.50%; only_ours_not_pem = 3.11%; not_ours_not_pem = 3.11%\t\n\u001b[34mcoref mention = peter nicol replaces original mention = nicol\u001b[39m ; DOC = 1211testb\t\n\u001b[34mcoref mention = rodney eyles replaces original mention = eyles\u001b[39m ; DOC = 1211testb\t\n\u001b[34mcoref mention = massimo cuttitta replaces original mention = cuttitta\u001b[39m ; DOC = 1164testb\t\n\u001b[34mcoref mention = marcello cuttitta replaces original mention = cuttitta\u001b[39m ; DOC = 1164testb\t\n\u001b[34mcoref mention = massimo cuttitta replaces original mention = cuttitta\u001b[39m ; DOC = 1164testb\t\n\u001b[34mcoref mention = marcello cuttitta replaces original mention = cuttitta\u001b[39m ; DOC = 1164testb\t\n\u001b[34mcoref mention = massimo cuttitta replaces original mention = cuttitta\u001b[39m ; DOC = 1164testb\t\n\u001b[34mcoref mention = marcello cuttitta replaces original mention = cuttitta\u001b[39m ; DOC = 1164testb\t\n\u001b[34mcoref mention = florence masnada replaces original mention = florence\u001b[39m ; DOC = 1349testb\t\n\u001b[34mcoref mention = florence masnada replaces original mention = florence\u001b[39m ; DOC = 1349testb\t\n\u001b[34mcoref mention = florence masnada replaces original mention = masnada\u001b[39m ; DOC = 1349testb\t\n\u001b[34mcoref mention = florence masnada replaces original mention = masnada\u001b[39m ; DOC = 1349testb\t\n\u001b[34mcoref mention = winston peters replaces original mention = peters\u001b[39m ; DOC = 1330testb\t\n\u001b[34mcoref mention = winston peters replaces original mention = peters\u001b[39m ; DOC = 1330testb\t\n\u001b[34mcoref mention = winston peters replaces original mention = peters\u001b[39m ; DOC = 1330testb\t\n\u001b[34mcoref mention = david campese replaces original mention = campese\u001b[39m ; DOC = 1367testb\t\n\u001b[34mcoref mention = david campese replaces original mention = campese\u001b[39m ; DOC = 1367testb\t\n\u001b[34mcoref mention = david campese replaces original mention = campese\u001b[39m ; DOC = 1367testb\t\n\u001b[34mcoref mention = emil constantinescu replaces original mention = constantinescu\u001b[39m ; DOC = 1214testb\t\n\u001b[34mcoref mention = takuya takagi replaces original mention = takagi\u001b[39m ; DOC = 1165testb\t\n\u001b[34mcoref mention = kazuyoshi miura replaces original mention = miura\u001b[39m ; DOC = 1165testb\t\n\u001b[34mcoref mention = garry galley replaces original mention = galley\u001b[39m ; DOC = 1387testb\t\n\u001b[34mcoref mention = pavel bure replaces original mention = bure\u001b[39m ; DOC = 1387testb\t\n\u001b[34mcoref mention = pavel bure replaces original mention = bure\u001b[39m ; DOC = 1387testb\t\n\u001b[34mcoref mention = pavel bure replaces original mention = bure\u001b[39m ; DOC = 1387testb\t\n\u001b[34mcoref mention = garry galley replaces original mention = galley\u001b[39m ; DOC = 1387testb\t\n\u001b[34mcoref mention = pavel bure replaces original mention = bure\u001b[39m ; DOC = 1387testb\t\n\u001b[34mcoref mention = garry galley replaces original mention = galley\u001b[39m ; DOC = 1387testb\t\n\u001b[34mcoref mention = pavel bure replaces original mention = bure\u001b[39m ; DOC = 1387testb\t\n\u001b[34mcoref mention = boris yeltsin replaces original mention = yeltsin\u001b[39m ; DOC = 1254testb\t\n\u001b[34mcoref mention = boris yeltsin replaces original mention = yeltsin\u001b[39m ; DOC = 1254testb\t\n\u001b[34mcoref mention = boris yeltsin replaces original mention = yeltsin\u001b[39m ; DOC = 1254testb\t\n\u001b[34mcoref mention = boris yeltsin replaces original mention = yeltsin\u001b[39m ; DOC = 1254testb\t\n\u001b[34mcoref mention = boris yeltsin replaces original mention = yeltsin\u001b[39m ; DOC = 1254testb\t\n\u001b[34mcoref mention = svetlana paramygina replaces original mention = paramygina\u001b[39m ; DOC = 1351testb\t\n\u001b[34mcoref mention = simone greiner-petter-memm replaces original mention = greiner-petter-memm\u001b[39m ; DOC = 1351testb\t\n\u001b[34mcoref mention = vadim sashurin replaces original mention = sashurin\u001b[39m ; DOC = 1351testb\t\n\u001b[34mcoref mention = sven fischer replaces original mention = fischer\u001b[39m ; DOC = 1351testb\t\n\u001b[34mcoref mention = petra behle replaces original mention = behle\u001b[39m ; DOC = 1351testb\t\n\u001b[34mcoref mention = matthew burke replaces original mention = burke\u001b[39m ; DOC = 1375testb\t\n\u001b[34mcoref mention = matthew burke replaces original mention = burke\u001b[39m ; DOC = 1375testb\t\n\u001b[34mcoref mention = alan greenspan replaces original mention = greenspan\u001b[39m ; DOC = 1258testb\t\n\u001b[34mcoref mention = alan greenspan replaces original mention = greenspan\u001b[39m ; DOC = 1258testb\t\n\u001b[34mcoref mention = alan greenspan replaces original mention = greenspan\u001b[39m ; DOC = 1258testb\t\n\u001b[34mcoref mention = john major replaces original mention = major\u001b[39m ; DOC = 1226testb\t\n\u001b[34mcoref mention = john gorst replaces original mention = gorst\u001b[39m ; DOC = 1226testb\t\n\u001b[34mcoref mention = john gorst replaces original mention = gorst\u001b[39m ; DOC = 1226testb\t\n\u001b[34mcoref mention = john major replaces original mention = major\u001b[39m ; DOC = 1226testb\t\n\u001b[34mcoref mention = john gorst replaces original mention = gorst\u001b[39m ; DOC = 1226testb\t\n\u001b[34mcoref mention = john gorst replaces original mention = gorst\u001b[39m ; DOC = 1226testb\t\n\u001b[34mcoref mention = john major replaces original mention = major\u001b[39m ; DOC = 1226testb\t\n\u001b[34mcoref mention = john gorst replaces original mention = gorst\u001b[39m ; DOC = 1226testb\t\n\u001b[34mcoref mention = george weah replaces original mention = weah\u001b[39m ; DOC = 1206testb\t\n\u001b[34mcoref mention = jorge costa replaces original mention = costa\u001b[39m ; DOC = 1206testb\t\n\u001b[34mcoref mention = george weah replaces original mention = weah\u001b[39m ; DOC = 1206testb\t\n\u001b[34mcoref mention = jorge costa replaces original mention = costa\u001b[39m ; DOC = 1206testb\t\n\u001b[34mcoref mention = jorge costa replaces original mention = costa\u001b[39m ; DOC = 1206testb\t\n\u001b[34mcoref mention = jorge costa replaces original mention = costa\u001b[39m ; DOC = 1206testb\t\n\u001b[34mcoref mention = jorge costa replaces original mention = costa\u001b[39m ; DOC = 1206testb\t\n\u001b[34mcoref mention = george weah replaces original mention = weah\u001b[39m ; DOC = 1206testb\t\n\u001b[34mcoref mention = george weah replaces original mention = weah\u001b[39m ; DOC = 1206testb\t\n\u001b[34mcoref mention = george weah replaces original mention = weah\u001b[39m ; DOC = 1206testb\t\n\u001b[34mcoref mention = george weah replaces original mention = weah\u001b[39m ; DOC = 1206testb\t\n\u001b[34mcoref mention = laurie harper replaces original mention = harper\u001b[39m ; DOC = 1378testb\t\n\u001b[34mcoref mention = john major replaces original mention = major\u001b[39m ; DOC = 1301testb\t\n\u001b[34mcoref mention = john major replaces original mention = major\u001b[39m ; DOC = 1301testb\t\n\u001b[34mcoref mention = kenneth clarke replaces original mention = clarke\u001b[39m ; DOC = 1301testb\t\n\u001b[34mcoref mention = john major replaces original mention = major\u001b[39m ; DOC = 1301testb\t\n\u001b[34mcoref mention = kenneth clarke replaces original mention = clarke\u001b[39m ; DOC = 1301testb\t\n\u001b[34mcoref mention = john major replaces original mention = major\u001b[39m ; DOC = 1301testb\t\n\u001b[34mcoref mention = john major replaces original mention = major\u001b[39m ; DOC = 1301testb\t\n\u001b[34mcoref mention = kenneth clarke replaces original mention = clarke\u001b[39m ; DOC = 1301testb\t\n\u001b[34mcoref mention = vaclav havel replaces original mention = havel\u001b[39m ; DOC = 1218testb\t\n\u001b[34mcoref mention = vaclav havel replaces original mention = havel\u001b[39m ; DOC = 1218testb\t\n\u001b[34mcoref mention = vaclav havel replaces original mention = havel\u001b[39m ; DOC = 1218testb\t\n\u001b[34mcoref mention = vaclav havel replaces original mention = havel\u001b[39m ; DOC = 1218testb\t\n\u001b[34mcoref mention = vaclav havel replaces original mention = havel\u001b[39m ; DOC = 1218testb\t\n\u001b[34mcoref mention = jani soininen replaces original mention = soininen\u001b[39m ; DOC = 1358testb\t\n\u001b[34mcoref mention = andreas goldberger replaces original mention = goldberger\u001b[39m ; DOC = 1358testb\t\n\u001b[34mcoref mention = takanobu okabe replaces original mention = okabe\u001b[39m ; DOC = 1358testb\t\n\u001b[34mcoref mention = dieter thoma replaces original mention = thoma\u001b[39m ; DOC = 1358testb\t\n\u001b[34mcoref mention = ari-pekka nikkola replaces original mention = nikkola\u001b[39m ; DOC = 1358testb\t\n\u001b[34mcoref mention = kazuyoshi funaki replaces original mention = funaki\u001b[39m ; DOC = 1358testb\t\n\u001b[34mcoref mention = kristian brenden replaces original mention = brenden\u001b[39m ; DOC = 1358testb\t\n\u001b[34mcoref mention = kenneth clarke replaces original mention = clarke\u001b[39m ; DOC = 1326testb\t\n\u001b[34mcoref mention = tony marlow replaces original mention = marlow\u001b[39m ; DOC = 1326testb\t\n\u001b[34mcoref mention = kenneth clarke replaces original mention = clarke\u001b[39m ; DOC = 1326testb\t\n\u001b[34mcoref mention = john major replaces original mention = major\u001b[39m ; DOC = 1326testb\t\n\u001b[34mcoref mention = kenneth clarke replaces original mention = clarke\u001b[39m ; DOC = 1326testb\t\n\u001b[34mcoref mention = kenneth clarke replaces original mention = clarke\u001b[39m ; DOC = 1326testb\t\n\u001b[34mcoref mention = kenneth clarke replaces original mention = clarke\u001b[39m ; DOC = 1326testb\t\n\u001b[34mcoref mention = tony marlow replaces original mention = marlow\u001b[39m ; DOC = 1326testb\t\n\u001b[34mcoref mention = patrick vieira replaces original mention = vieira\u001b[39m ; DOC = 1371testb\t\n\u001b[34mcoref mention = paul justin replaces original mention = justin\u001b[39m ; DOC = 1188testb\t\n\u001b[34mcoref mention = jim harbaugh replaces original mention = harbaugh\u001b[39m ; DOC = 1188testb\t\n\u001b[34mcoref mention = paul justin replaces original mention = justin\u001b[39m ; DOC = 1188testb\t\n\u001b[34mcoref mention = marshall faulk replaces original mention = faulk\u001b[39m ; DOC = 1188testb\t\n\u001b[34mcoref mention = marshall faulk replaces original mention = faulk\u001b[39m ; DOC = 1188testb\t\n\u001b[34mcoref mention = sali berisha replaces original mention = berisha\u001b[39m ; DOC = 1249testb\t\n\u001b[34mcoref mention = brian shimer replaces original mention = shimer\u001b[39m ; DOC = 1353testb\t\n\u001b[34mcoref mention = jose ramos horta replaces original mention = ramos horta\u001b[39m ; DOC = 1264testb\t\n\u001b[34mcoref mention = carlos belo replaces original mention = belo\u001b[39m ; DOC = 1264testb\t\n\u001b[34mcoref mention = ali alatas replaces original mention = alatas\u001b[39m ; DOC = 1264testb\t\n\u001b[34mcoref mention = jose ramos horta replaces original mention = ramos horta\u001b[39m ; DOC = 1264testb\t\n\u001b[34mcoref mention = ali alatas replaces original mention = alatas\u001b[39m ; DOC = 1264testb\t\n\u001b[34mcoref mention = helmut kohl replaces original mention = kohl\u001b[39m ; DOC = 1264testb\t\n\u001b[34mcoref mention = carlos belo replaces original mention = belo\u001b[39m ; DOC = 1264testb\t\n\u001b[34mcoref mention = carlos belo replaces original mention = belo\u001b[39m ; DOC = 1264testb\t\n\u001b[34mcoref mention = carlos belo replaces original mention = belo\u001b[39m ; DOC = 1264testb\t\n\u001b[34mcoref mention = carlos belo replaces original mention = belo\u001b[39m ; DOC = 1264testb\t\n\u001b[34mcoref mention = jose ramos horta replaces original mention = ramos horta\u001b[39m ; DOC = 1264testb\t\n\u001b[34mcoref mention = carlos belo replaces original mention = belo\u001b[39m ; DOC = 1264testb\t\n\u001b[34mcoref mention = ali alatas replaces original mention = alatas\u001b[39m ; DOC = 1264testb\t\n\u001b[34mcoref mention = jose ramos horta replaces original mention = ramos horta\u001b[39m ; DOC = 1264testb\t\n\u001b[34mcoref mention = carlos belo replaces original mention = belo\u001b[39m ; DOC = 1264testb\t\n\u001b[34mcoref mention = ali alatas replaces original mention = alatas\u001b[39m ; DOC = 1264testb\t\n\u001b[34mcoref mention = maurice baril replaces original mention = baril\u001b[39m ; DOC = 1241testb\t\n\u001b[34mcoref mention = maurice baril replaces original mention = baril\u001b[39m ; DOC = 1241testb\t\n\u001b[34mcoref mention = maurice baril replaces original mention = baril\u001b[39m ; DOC = 1241testb\t\n\u001b[34mcoref mention = doug young replaces original mention = young\u001b[39m ; DOC = 1241testb\t\n\u001b[34mcoref mention = bob woolmer replaces original mention = woolmer\u001b[39m ; DOC = 1356testb\t\n\u001b[34mcoref mention = clarence woolmer replaces original mention = woolmer\u001b[39m ; DOC = 1356testb\t\n\u001b[34mcoref mention = bob woolmer replaces original mention = woolmer\u001b[39m ; DOC = 1356testb\t\n\u001b[34mcoref mention = clarence woolmer replaces original mention = woolmer\u001b[39m ; DOC = 1356testb\t\n\u001b[34mcoref mention = bob woolmer replaces original mention = woolmer\u001b[39m ; DOC = 1356testb\t\n\u001b[34mcoref mention = clarence woolmer replaces original mention = woolmer\u001b[39m ; DOC = 1356testb\t\n\u001b[34mcoref mention = bob woolmer replaces original mention = woolmer\u001b[39m ; DOC = 1356testb\t\n\u001b[34mcoref mention = clarence woolmer replaces original mention = woolmer\u001b[39m ; DOC = 1356testb\t\n\u001b[34mcoref mention = bob woolmer replaces original mention = woolmer\u001b[39m ; DOC = 1356testb\t\n\u001b[34mcoref mention = clarence woolmer replaces original mention = woolmer\u001b[39m ; DOC = 1356testb\t\n\u001b[34mcoref mention = bob woolmer replaces original mention = woolmer\u001b[39m ; DOC = 1356testb\t\n\u001b[34mcoref mention = clarence woolmer replaces original mention = woolmer\u001b[39m ; DOC = 1356testb\t\n\u001b[34mcoref mention = bob woolmer replaces original mention = woolmer\u001b[39m ; DOC = 1356testb\t\n\u001b[34mcoref mention = clarence woolmer replaces original mention = woolmer\u001b[39m ; DOC = 1356testb\t\n\u001b[34mcoref mention = greg blewett replaces original mention = blewett\u001b[39m ; DOC = 1177testb\t\n\u001b[34mcoref mention = brian lara replaces original mention = lara\u001b[39m ; DOC = 1177testb\t\n\u001b[34mcoref mention = brian lara replaces original mention = lara\u001b[39m ; DOC = 1177testb\t\n\u001b[34mcoref mention = brian lara replaces original mention = lara\u001b[39m ; DOC = 1177testb\t\n\u001b[34mcoref mention = brian lara replaces original mention = lara\u001b[39m ; DOC = 1177testb\t\n\u001b[34mcoref mention = brian lara replaces original mention = lara\u001b[39m ; DOC = 1177testb\t\n\u001b[34mcoref mention = shivnarine chanderpaul replaces original mention = chanderpaul\u001b[39m ; DOC = 1177testb\t\n\u001b[34mcoref mention = aung san suu kyi replaces original mention = aung san\u001b[39m ; DOC = 1267testb\t\n\u001b[34mcoref mention = aung san suu kyi replaces original mention = suu kyi\u001b[39m ; DOC = 1267testb\t\n\u001b[34mcoref mention = aung san suu kyi replaces original mention = suu kyi\u001b[39m ; DOC = 1267testb\t\n\u001b[34mcoref mention = aung san suu kyi replaces original mention = suu kyi\u001b[39m ; DOC = 1267testb\t\n\u001b[34mcoref mention = peter nicol replaces original mention = nicol\u001b[39m ; DOC = 1210testb\t\n\u001b[34mcoref mention = rodney eyles replaces original mention = eyles\u001b[39m ; DOC = 1210testb\t\n\u001b[34mcoref mention = peter nicol replaces original mention = nicol\u001b[39m ; DOC = 1210testb\t\n\u001b[34mcoref mention = rodney eyles replaces original mention = eyles\u001b[39m ; DOC = 1210testb\t\n\u001b[34mcoref mention = dariusz rosati replaces original mention = rosati\u001b[39m ; DOC = 1216testb\t\n\u001b[34mcoref mention = dariusz rosati replaces original mention = rosati\u001b[39m ; DOC = 1216testb\t\n\u001b[34mcoref mention = dariusz rosati replaces original mention = rosati\u001b[39m ; DOC = 1216testb\t\n\u001b[34mcoref mention = dariusz rosati replaces original mention = rosati\u001b[39m ; DOC = 1216testb\t\n\u001b[34mcoref mention = dariusz rosati replaces original mention = rosati\u001b[39m ; DOC = 1216testb\t\n\u001b[34mcoref mention = shimon peres replaces original mention = peres\u001b[39m ; DOC = 1271testb\t\n\u001b[34mcoref mention = nicholas burns replaces original mention = burns\u001b[39m ; DOC = 1310testb\t\n\u001b[34mcoref mention = nicholas burns replaces original mention = burns\u001b[39m ; DOC = 1310testb\t\n\u001b[34mcoref mention = gianfranco fini replaces original mention = fini\u001b[39m ; DOC = 1290testb\t\n\u001b[34mcoref mention = alessandra mussolini replaces original mention = mussolini\u001b[39m ; DOC = 1290testb\t\n\u001b[34mcoref mention = benito mussolini replaces original mention = mussolini\u001b[39m ; DOC = 1290testb\t\n\u001b[34mcoref mention = alessandra mussolini replaces original mention = mussolini\u001b[39m ; DOC = 1290testb\t\n\u001b[34mcoref mention = benito mussolini replaces original mention = mussolini\u001b[39m ; DOC = 1290testb\t\n\u001b[34mcoref mention = alessandra mussolini replaces original mention = mussolini\u001b[39m ; DOC = 1290testb\t\n\u001b[34mcoref mention = benito mussolini replaces original mention = mussolini\u001b[39m ; DOC = 1290testb\t\n\u001b[34mcoref mention = alessandra mussolini replaces original mention = mussolini\u001b[39m ; DOC = 1290testb\t\n\u001b[34mcoref mention = benito mussolini replaces original mention = mussolini\u001b[39m ; DOC = 1290testb\t\n\u001b[34mcoref mention = davor suker replaces original mention = suker\u001b[39m ; DOC = 1390testb\t\n\u001b[34mcoref mention = predrag mijatovic replaces original mention = mijatovic\u001b[39m ; DOC = 1390testb\t\n\u001b[34mcoref mention = jim bolger replaces original mention = bolger\u001b[39m ; DOC = 1329testb\t\n\u001b[34mcoref mention = jim bolger replaces original mention = bolger\u001b[39m ; DOC = 1329testb\t\n\u001b[34mcoref mention = oscar luigi scalfaro replaces original mention = scalfaro\u001b[39m ; DOC = 1315testb\t\n\u001b[34mcoref mention = oscar luigi scalfaro replaces original mention = scalfaro\u001b[39m ; DOC = 1315testb\t\n\u001b[34mcoref mention = oscar luigi scalfaro replaces original mention = scalfaro\u001b[39m ; DOC = 1315testb\t\n\u001b[34mcoref mention = axel schulz replaces original mention = schulz\u001b[39m ; DOC = 1388testb\t\n\u001b[34mcoref mention = wasim akram replaces original mention = wasim\u001b[39m ; DOC = 1168testb\t\n\u001b[34mcoref mention = waqar younis replaces original mention = waqar\u001b[39m ; DOC = 1168testb\t\n\u001b[34mcoref mention = waqar younis replaces original mention = waqar\u001b[39m ; DOC = 1168testb\t\n\u001b[34mcoref mention = wasim akram replaces original mention = wasim\u001b[39m ; DOC = 1168testb\t\n\u001b[34mcoref mention = saqlain mushtaq replaces original mention = saqlain\u001b[39m ; DOC = 1168testb\t\n\u001b[34mcoref mention = shahid afridi replaces original mention = afridi\u001b[39m ; DOC = 1168testb\t\n\u001b[34mcoref mention = saqlain mushtaq replaces original mention = saqlain\u001b[39m ; DOC = 1168testb\t\n\u001b[34mcoref mention = wasim akram replaces original mention = wasim\u001b[39m ; DOC = 1168testb\t\n\u001b[34mcoref mention = shahid afridi replaces original mention = afridi\u001b[39m ; DOC = 1168testb\t\n\u001b[34mcoref mention = clive lloyd replaces original mention = lloyd\u001b[39m ; DOC = 1182testb\t\n\u001b[34mcoref mention = brian lara replaces original mention = lara\u001b[39m ; DOC = 1182testb\t\n\u001b[34mcoref mention = brian lara replaces original mention = lara\u001b[39m ; DOC = 1182testb\t\n\u001b[34mcoref mention = brian lara replaces original mention = lara\u001b[39m ; DOC = 1182testb\t\n\u001b[34mcoref mention = brian lara replaces original mention = lara\u001b[39m ; DOC = 1182testb\t\n\u001b[34mcoref mention = brian lara replaces original mention = lara\u001b[39m ; DOC = 1182testb\t\n\u001b[34mcoref mention = ian healy replaces original mention = healy\u001b[39m ; DOC = 1182testb\t\n\u001b[34mcoref mention = brian lara replaces original mention = lara\u001b[39m ; DOC = 1182testb\t\n\u001b[34mcoref mention = brian lara replaces original mention = lara\u001b[39m ; DOC = 1182testb\t\n\u001b[34mcoref mention = jack charlton replaces original mention = charlton\u001b[39m ; DOC = 1393testb\t\n\u001b[34mcoref mention = jack charlton replaces original mention = charlton\u001b[39m ; DOC = 1393testb\t\n\u001b[34mcoref mention = jack charlton replaces original mention = charlton\u001b[39m ; DOC = 1393testb\t\n\u001b[34mcoref mention = jack charlton replaces original mention = charlton\u001b[39m ; DOC = 1393testb\t\n\u001b[34mcoref mention = jack charlton replaces original mention = charlton\u001b[39m ; DOC = 1393testb\t\n\u001b[34mcoref mention = david campese replaces original mention = campese\u001b[39m ; DOC = 1374testb\t\n\u001b[34mcoref mention = david campese replaces original mention = campese\u001b[39m ; DOC = 1374testb\t\n\u001b[34mcoref mention = aleksander kwasniewski replaces original mention = kwasniewski\u001b[39m ; DOC = 1250testb\t\n\u001b[34mcoref mention = dariusz rosati replaces original mention = rosati\u001b[39m ; DOC = 1250testb\t\n\u001b[34mcoref mention = aleksander kwasniewski replaces original mention = kwasniewski\u001b[39m ; DOC = 1250testb\t\n\u001b[34mcoref mention = orlando pace replaces original mention = pace\u001b[39m ; DOC = 1192testb\t\n\u001b[34mcoref mention = orlando pace replaces original mention = pace\u001b[39m ; DOC = 1192testb\t\n\u001b[34mcoref mention = orlando pace replaces original mention = pace\u001b[39m ; DOC = 1192testb\t\n\u001b[34mcoref mention = orlando pace replaces original mention = pace\u001b[39m ; DOC = 1192testb\t\n\u001b[34mcoref mention = vaclav havel replaces original mention = havel\u001b[39m ; DOC = 1213testb\t\n\u001b[34mcoref mention = vaclav havel replaces original mention = havel\u001b[39m ; DOC = 1213testb\t\n\u001b[34mcoref mention = vaclav havel replaces original mention = havel\u001b[39m ; DOC = 1213testb\t\n\u001b[34mcoref mention = madeleine albright replaces original mention = albright\u001b[39m ; DOC = 1213testb\t\n\u001b[34mcoref mention = madeleine albright replaces original mention = albright\u001b[39m ; DOC = 1213testb\t\n\u001b[34mcoref mention = madeleine albright replaces original mention = albright\u001b[39m ; DOC = 1213testb\t\n\u001b[34mcoref mention = madeleine albright replaces original mention = albright\u001b[39m ; DOC = 1213testb\t\n\u001b[34mcoref mention = madeleine albright replaces original mention = albright\u001b[39m ; DOC = 1213testb\t\n\u001b[34mcoref mention = vaclav havel replaces original mention = havel\u001b[39m ; DOC = 1213testb\t\n\u001b[34mcoref mention = vaclav havel replaces original mention = havel\u001b[39m ; DOC = 1213testb\t\n\u001b[34mcoref mention = madeleine albright replaces original mention = albright\u001b[39m ; DOC = 1213testb\t\n\u001b[34mcoref mention = madeleine albright replaces original mention = albright\u001b[39m ; DOC = 1213testb\t\n\u001b[34mcoref mention = jorge costa replaces original mention = costa\u001b[39m ; DOC = 1201testb\t\n\u001b[34mcoref mention = rui costa replaces original mention = costa\u001b[39m ; DOC = 1201testb\t\n\u001b[34mcoref mention = bob halverson replaces original mention = halverson\u001b[39m ; DOC = 1331testb\t\n\u001b[34mcoref mention = bob halverson replaces original mention = halverson\u001b[39m ; DOC = 1331testb\t\n\u001b[34mcoref mention = john langmore replaces original mention = langmore\u001b[39m ; DOC = 1331testb\t\n\u001b[34mcoref mention = john vance langmore replaces original mention = langmore\u001b[39m ; DOC = 1331testb\t\n\u001b[34mcoref mention = john langmore replaces original mention = langmore\u001b[39m ; DOC = 1331testb\t\n\u001b[34mcoref mention = john vance langmore replaces original mention = langmore\u001b[39m ; DOC = 1331testb\t\n\u001b[34mcoref mention = abdou diouf replaces original mention = diouf\u001b[39m ; DOC = 1244testb\t\n\u001b[34mcoref mention = abdou diouf replaces original mention = diouf\u001b[39m ; DOC = 1244testb\t\n\u001b[34mcoref mention = jacques chirac replaces original mention = chirac\u001b[39m ; DOC = 1245testb\t\n\u001b[34mcoref mention = madeleine albright replaces original mention = albright\u001b[39m ; DOC = 1253testb\t\n\u001b[34mcoref mention = bill clinton replaces original mention = clinton\u001b[39m ; DOC = 1253testb\t\n\u001b[34mcoref mention = madeleine albright replaces original mention = albright\u001b[39m ; DOC = 1253testb\t\n\u001b[34mcoref mention = madeleine albright replaces original mention = albright\u001b[39m ; DOC = 1253testb\t\n\u001b[34mcoref mention = igor ivanov replaces original mention = ivanov\u001b[39m ; DOC = 1253testb\t\n\u001b[34mcoref mention = boris yeltsin replaces original mention = yeltsin\u001b[39m ; DOC = 1253testb\t\n\u001b[34mcoref mention = bill clinton replaces original mention = clinton\u001b[39m ; DOC = 1253testb\t\n\u001b[34mcoref mention = benjamin netanyahu replaces original mention = netanyahu\u001b[39m ; DOC = 1303testb\t\n\u001b[34mcoref mention = yasser arafat replaces original mention = arafat\u001b[39m ; DOC = 1303testb\t\n\u001b[34mcoref mention = yasser arafat replaces original mention = arafat\u001b[39m ; DOC = 1303testb\t\n\u001b[34mcoref mention = yasser arafat replaces original mention = arafat\u001b[39m ; DOC = 1303testb\t\n\u001b[34mcoref mention = benjamin netanyahu replaces original mention = netanyahu\u001b[39m ; DOC = 1303testb\t\n\u001b[34mcoref mention = benjamin netanyahu replaces original mention = netanyahu\u001b[39m ; DOC = 1303testb\t\n\u001b[34mcoref mention = benjamin netanyahu replaces original mention = netanyahu\u001b[39m ; DOC = 1303testb\t\n\u001b[34mcoref mention = hosni mubarak replaces original mention = mubarak\u001b[39m ; DOC = 1303testb\t\n\u001b[34mcoref mention = hosni mubarak replaces original mention = mubarak\u001b[39m ; DOC = 1303testb\t\n\u001b[34mcoref mention = yasser arafat replaces original mention = arafat\u001b[39m ; DOC = 1303testb\t\n\u001b[34mcoref mention = yasser arafat replaces original mention = arafat\u001b[39m ; DOC = 1303testb\t\n\u001b[34mcoref mention = yasser arafat replaces original mention = arafat\u001b[39m ; DOC = 1303testb\t\n\u001b[34mcoref mention = yasser arafat replaces original mention = arafat\u001b[39m ; DOC = 1303testb\t\n\u001b[34mcoref mention = hosni mubarak replaces original mention = mubarak\u001b[39m ; DOC = 1303testb\t\n\u001b[34mcoref mention = david campese replaces original mention = campese\u001b[39m ; DOC = 1173testb\t\n\u001b[34mcoref mention = david campese replaces original mention = campese\u001b[39m ; DOC = 1173testb\t\n\u001b[34mcoref mention = lee bowyer replaces original mention = bowyer\u001b[39m ; DOC = 1171testb\t\n\u001b[34mcoref mention = lee bowyer replaces original mention = bowyer\u001b[39m ; DOC = 1171testb\t\n\u001b[34mcoref mention = lee bowyer replaces original mention = bowyer\u001b[39m ; DOC = 1171testb\t\n\u001b[34mcoref mention = lee bowyer replaces original mention = bowyer\u001b[39m ; DOC = 1171testb\t\n\u001b[34mcoref mention = boutros boutros-ghali replaces original mention = boutros-ghali\u001b[39m ; DOC = 1288testb\t\n\u001b[34mcoref mention = boutros boutros-ghali replaces original mention = boutros-ghali\u001b[39m ; DOC = 1288testb\t\n\u001b[34mcoref mention = josef zieleniec replaces original mention = zieleniec\u001b[39m ; DOC = 1215testb\t\n\u001b[34mcoref mention = josef zieleniec replaces original mention = zieleniec\u001b[39m ; DOC = 1215testb\t\n\n===> aida-B; num mentions = 4485\t\n [..............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 0ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 0/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1373testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wolverhampton ==> ENTITY: \u001b[32mWolverhampton Wanderers F.C.\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.203:0.203[0]; log p(e|m)= -2.273:-2.273[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.043]; League[0.043]; Charlton[0.042]; Southampton[0.042]; matches[0.042]; Sheffield[0.042]; Sheffield[0.042]; Coventry[0.042]; Blackburn[0.042]; Leicester[0.042]; league[0.042]; league[0.042]; Birmingham[0.042]; Bristol[0.042]; Bradford[0.042]; Middlesbrough[0.042]; Sunderland[0.042]; Leeds[0.041]; Oxford[0.041]; Manchester[0.041]; Bolton[0.041]; Cambridge[0.041]; Oldham[0.041]; Huddersfield[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Manchester City\u001b[39m ==> ENTITY: \u001b[32mManchester City F.C.\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.229:0.229[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; Southampton[0.044]; Charlton[0.043]; league[0.043]; league[0.043]; Hull[0.043]; Blackburn[0.043]; Portsmouth[0.042]; Sunderland[0.042]; Coventry[0.042]; Middlesbrough[0.042]; Sheffield[0.041]; Sheffield[0.041]; Birmingham[0.041]; matches[0.041]; Division[0.041]; Leeds[0.041]; Bolton[0.041]; Liverpool[0.040]; Soccer[0.040]; Bristol[0.040]; Norwich[0.040]; City[0.040]; Ipswich[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Stockport ==> ENTITY: \u001b[32mStockport County F.C.\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.226:0.226[0]; log p(e|m)= -2.526:-2.526[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.046]; Preston[0.045]; Cardiff[0.045]; York[0.045]; Sheffield[0.044]; Bradford[0.044]; Walsall[0.044]; Chesterfield[0.044]; Chester[0.044]; Rochdale[0.043]; Watford[0.043]; Scunthorpe[0.043]; Bristol[0.043]; Burnley[0.043]; Gillingham[0.043]; Brentford[0.043]; Wrexham[0.043]; Carlisle[0.043]; Luton[0.043]; Portsmouth[0.043]; Enfield[0.043]; Crewe[0.043]; Oldham[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sheffield Wednesday\u001b[39m ==> ENTITY: \u001b[32mSheffield Wednesday F.C.\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; Sheffield[0.045]; league[0.044]; league[0.044]; Cup[0.044]; Cup[0.044]; matches[0.042]; Coventry[0.042]; Leeds[0.042]; Soccer[0.042]; Southampton[0.041]; Charlton[0.041]; Blackburn[0.041]; Division[0.041]; Birmingham[0.041]; Sunderland[0.041]; Manchester[0.040]; Middlesbrough[0.040]; Reading[0.040]; Wolverhampton[0.040]; Southend[0.040]; Huddersfield[0.040]; Tranmere[0.040]; Barnsley[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barnet ==> ENTITY: \u001b[32mBarnet F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.219:0.219[0]; log p(e|m)= -0.679:-0.679[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.046]; Southampton[0.046]; Charlton[0.044]; Stevenage[0.044]; Wycombe[0.044]; Watford[0.044]; Walsall[0.044]; Brentford[0.044]; Hull[0.044]; Norwich[0.043]; Oxford[0.043]; Swindon[0.043]; Rochdale[0.043]; Stockport[0.043]; Enfield[0.043]; Gillingham[0.043]; Burnley[0.043]; Southend[0.043]; Oldham[0.043]; Leeds[0.043]; Luton[0.043]; Carlisle[0.043]; Barnsley[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wycombe ==> ENTITY: \u001b[32mWycombe Wanderers F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.226:0.226[0]; log p(e|m)= -1.146:-1.146[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthampton[0.044]; Preston[0.044]; Hull[0.042]; Reading[0.042]; Brentford[0.042]; Watford[0.042]; Charlton[0.042]; Walsall[0.042]; Stockport[0.042]; Carlisle[0.042]; Enfield[0.042]; Sheffield[0.042]; Sheffield[0.042]; Chesterfield[0.041]; Swindon[0.041]; Leeds[0.041]; Oxford[0.041]; Barnsley[0.041]; Norwich[0.041]; Rochdale[0.041]; Oldham[0.041]; Bristol[0.041]; Cardiff[0.041]; Birmingham[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Derby ==> ENTITY: \u001b[32mDerby County F.C.\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.220:0.220[0]; log p(e|m)= -3.689:-3.689[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; Southampton[0.043]; league[0.043]; league[0.043]; Charlton[0.043]; Birmingham[0.042]; Sheffield[0.042]; Sheffield[0.042]; Coventry[0.042]; Blackburn[0.042]; matches[0.042]; Middlesbrough[0.042]; Oldham[0.041]; Sunderland[0.041]; Bradford[0.041]; Wolverhampton[0.041]; Manchester[0.041]; Reading[0.041]; Leeds[0.041]; Swindon[0.041]; Park[0.041]; Portsmouth[0.040]; Norwich[0.040]; Barnsley[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sunderland ==> ENTITY: \u001b[32mSunderland A.F.C.\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.215:0.215[0]; log p(e|m)= -0.396:-0.396[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.043]; league[0.043]; League[0.043]; Charlton[0.043]; Southampton[0.042]; Coventry[0.042]; Middlesbrough[0.042]; Sheffield[0.042]; Sheffield[0.042]; Ipswich[0.042]; Blackburn[0.042]; Grimsby[0.042]; Barnsley[0.042]; Manchester[0.042]; Leeds[0.041]; Southend[0.041]; Bristol[0.041]; matches[0.041]; Norwich[0.041]; Blackpool[0.041]; Barnet[0.040]; Tranmere[0.040]; Wolverhampton[0.040]; Birmingham[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Peterborough ==> ENTITY: \u001b[32mPeterborough United F.C.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.219:0.219[0]; log p(e|m)= -3.170:-3.170[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Hull[0.045]; Bristol[0.043]; Enfield[0.042]; Sheffield[0.042]; Charlton[0.042]; Portsmouth[0.042]; Walsall[0.042]; Ipswich[0.042]; Wolverhampton[0.042]; Swindon[0.041]; Manchester[0.041]; Stockport[0.041]; Rochdale[0.041]; Norwich[0.041]; Carlisle[0.041]; Watford[0.041]; second[0.041]; Bradford[0.041]; Brentford[0.041]; Oldham[0.041]; Luton[0.041]; Chesterfield[0.041]; Gillingham[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rochdale ==> ENTITY: \u001b[32mRochdale A.F.C.\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.225:0.225[0]; log p(e|m)= -0.952:-0.952[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Chesterfield[0.043]; Hull[0.043]; Stockport[0.043]; Walsall[0.042]; Carlisle[0.042]; Peterborough[0.042]; Brentford[0.042]; Enfield[0.042]; Scunthorpe[0.042]; Darlington[0.042]; Bristol[0.042]; Queens[0.041]; Sheffield[0.041]; Burnley[0.041]; Watford[0.041]; Gillingham[0.041]; Chester[0.041]; Wycombe[0.041]; Luton[0.041]; City[0.041]; Park[0.041]; Town[0.041]; Town[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swindon ==> ENTITY: \u001b[32mSwindon Town F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.244:0.244[0]; log p(e|m)= -2.489:-2.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; Hull[0.044]; Southampton[0.043]; Coventry[0.042]; Bristol[0.042]; league[0.042]; league[0.042]; Charlton[0.042]; Sheffield[0.042]; Sheffield[0.042]; Blackburn[0.042]; Sunderland[0.041]; Darlington[0.041]; Reading[0.041]; Leeds[0.041]; Southend[0.041]; Middlesbrough[0.041]; Enfield[0.041]; Chesterfield[0.041]; matches[0.041]; Peterborough[0.041]; Portsmouth[0.041]; Oldham[0.041]; Barnsley[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leicester ==> ENTITY: \u001b[32mLeicester City F.C.\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.210:0.210[0]; log p(e|m)= -3.219:-3.219[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; Coventry[0.044]; Oxford[0.043]; Birmingham[0.043]; league[0.042]; league[0.042]; Wolverhampton[0.042]; Charlton[0.042]; Norwich[0.042]; Southampton[0.042]; Blackburn[0.042]; matches[0.042]; Bradford[0.042]; Sunderland[0.041]; Sheffield[0.041]; Sheffield[0.041]; Swindon[0.041]; Leeds[0.041]; Middlesbrough[0.041]; Manchester[0.041]; Reading[0.041]; City[0.040]; Ipswich[0.040]; London[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bradford ==> ENTITY: \u001b[32mBradford City A.F.C.\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.256:0.256[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; league[0.043]; league[0.043]; Hull[0.043]; Leeds[0.043]; Coventry[0.042]; Southampton[0.042]; Sheffield[0.042]; Sheffield[0.042]; Charlton[0.042]; matches[0.042]; Bristol[0.041]; Sunderland[0.041]; Middlesbrough[0.041]; Huddersfield[0.041]; Blackburn[0.041]; Stockport[0.041]; Manchester[0.041]; Reading[0.041]; Chesterfield[0.041]; City[0.041]; City[0.041]; Barnsley[0.040]; Leicester[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oldham ==> ENTITY: \u001b[32mOldham Athletic A.F.C.\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.222:0.222[0]; log p(e|m)= -2.865:-2.865[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; league[0.046]; League[0.044]; Bradford[0.043]; Hull[0.043]; Huddersfield[0.043]; Leeds[0.042]; Charlton[0.042]; Southampton[0.042]; Enfield[0.041]; Bristol[0.041]; Coventry[0.041]; Blackburn[0.041]; Chesterfield[0.041]; Peterborough[0.041]; Stockport[0.041]; matches[0.041]; Carlisle[0.041]; Sunderland[0.040]; Sheffield[0.040]; Sheffield[0.040]; Southend[0.040]; Barnsley[0.040]; Middlesbrough[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brentford ==> ENTITY: \u001b[32mBrentford F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.215:0.215[0]; log p(e|m)= -0.361:-0.361[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Hull[0.044]; Watford[0.043]; Sheffield[0.043]; Walsall[0.042]; Gillingham[0.042]; Burnley[0.042]; Stockport[0.042]; Cardiff[0.042]; Tranmere[0.042]; Scunthorpe[0.042]; Luton[0.041]; second[0.041]; Rochdale[0.041]; Wycombe[0.041]; Blackpool[0.041]; Portsmouth[0.041]; Bolton[0.041]; Bristol[0.041]; Peterborough[0.041]; Chesterfield[0.041]; City[0.040]; Crewe[0.040]; Barnet[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Blackpool ==> ENTITY: \u001b[32mBlackpool F.C.\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.211:0.211[0]; log p(e|m)= -0.853:-0.853[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.043]; Southend[0.043]; Preston[0.043]; Wycombe[0.043]; Leeds[0.042]; Rochdale[0.042]; Walsall[0.042]; Watford[0.042]; Huddersfield[0.042]; Wrexham[0.042]; Southampton[0.042]; Swindon[0.042]; Stockport[0.041]; Charlton[0.041]; Sheffield[0.041]; Sheffield[0.041]; Scunthorpe[0.041]; Gillingham[0.041]; Brentford[0.041]; Barnsley[0.041]; Peterborough[0.041]; Bradford[0.041]; Cardiff[0.041]; Oldham[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hednesford ==> ENTITY: \u001b[32mHednesford Town F.C.\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.244:0.244[0]; log p(e|m)= -1.022:-1.022[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.044]; Stockport[0.043]; Hull[0.043]; Walsall[0.042]; Peterborough[0.042]; Southampton[0.042]; Bradford[0.042]; Watford[0.042]; Rochdale[0.042]; Darlington[0.042]; Sheffield[0.041]; Sheffield[0.041]; Swindon[0.041]; Ipswich[0.041]; Wycombe[0.041]; Barnsley[0.041]; Charlton[0.041]; Brentford[0.041]; Bristol[0.041]; Birmingham[0.041]; Carlisle[0.041]; Chesterfield[0.041]; Southend[0.041]; Scunthorpe[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wimbledon ==> ENTITY: \u001b[32mWimbledon F.C.\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.237:0.237[0]; log p(e|m)= -1.890:-1.890[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; matches[0.043]; Charlton[0.043]; Southampton[0.043]; league[0.042]; league[0.042]; Sheffield[0.042]; Sheffield[0.042]; Coventry[0.042]; Sunderland[0.042]; London[0.042]; Swindon[0.041]; Portsmouth[0.041]; Birmingham[0.041]; Middlesbrough[0.041]; Manchester[0.041]; Blackburn[0.041]; Leeds[0.041]; Reading[0.041]; Barnsley[0.041]; Wycombe[0.041]; Cup[0.041]; Cup[0.041]; Bristol[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hull ==> ENTITY: \u001b[32mHull City A.F.C.\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.229:0.229[0]; log p(e|m)= -3.772:-3.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.044]; Huddersfield[0.044]; Sheffield[0.043]; Walsall[0.042]; Cardiff[0.042]; Rochdale[0.042]; Manchester[0.042]; Watford[0.042]; Burnley[0.042]; Bristol[0.042]; Bradford[0.041]; Norwich[0.041]; Stockport[0.041]; Portsmouth[0.041]; Scunthorpe[0.041]; Carlisle[0.041]; Chesterfield[0.041]; Peterborough[0.041]; City[0.041]; City[0.041]; Brentford[0.041]; Swindon[0.041]; Tranmere[0.041]; Reading[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Crewe ==> ENTITY: \u001b[32mCrewe Alexandra F.C.\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.205:0.205[0]; log p(e|m)= -2.830:-2.830[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.044]; Wolverhampton[0.043]; Hull[0.043]; Watford[0.043]; Bristol[0.043]; Norwich[0.042]; Sheffield[0.042]; Enfield[0.042]; Portsmouth[0.042]; Chesterfield[0.042]; Walsall[0.042]; Carlisle[0.042]; second[0.042]; Manchester[0.041]; Ipswich[0.041]; Burnley[0.041]; Stockport[0.041]; Wycombe[0.041]; Peterborough[0.041]; Oldham[0.041]; Scunthorpe[0.041]; Brentford[0.041]; Rochdale[0.041]; Gillingham[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Everton ==> ENTITY: \u001b[32mEverton F.C.\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.208:0.208[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; league[0.044]; league[0.044]; Charlton[0.043]; Blackburn[0.043]; Birmingham[0.042]; Manchester[0.042]; Coventry[0.042]; Southampton[0.042]; matches[0.042]; Sheffield[0.041]; Sheffield[0.041]; Middlesbrough[0.041]; Soccer[0.041]; Leeds[0.041]; Sunderland[0.041]; Ipswich[0.041]; Cup[0.041]; Cup[0.041]; Park[0.040]; Oxford[0.040]; Aston[0.040]; City[0.040]; Portsmouth[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Crystal Palace ==> ENTITY: \u001b[32mCrystal Palace F.C.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.213:0.213[0]; log p(e|m)= -0.511:-0.511[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthampton[0.043]; Hull[0.043]; League[0.043]; Coventry[0.043]; Charlton[0.043]; Sheffield[0.042]; Sheffield[0.042]; Sunderland[0.042]; league[0.042]; league[0.042]; Birmingham[0.042]; Manchester[0.042]; Blackburn[0.041]; Middlesbrough[0.041]; Bolton[0.041]; Portsmouth[0.041]; Leeds[0.041]; Ipswich[0.041]; matches[0.041]; Park[0.041]; Southend[0.041]; Norwich[0.041]; Bristol[0.040]; Barnsley[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Queens Park Rangers\u001b[39m ==> ENTITY: \u001b[32mQueens Park Rangers F.C.\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; League[0.045]; Southampton[0.043]; Charlton[0.043]; Blackburn[0.043]; Hull[0.042]; league[0.042]; league[0.042]; Sheffield[0.042]; Sheffield[0.042]; Coventry[0.042]; matches[0.041]; Manchester[0.041]; Leeds[0.041]; Rochdale[0.041]; Chesterfield[0.041]; Middlesbrough[0.041]; Luton[0.041]; Sunderland[0.041]; Blackpool[0.041]; Southend[0.040]; Barnsley[0.040]; Stockport[0.040]; Birmingham[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Reading ==> ENTITY: \u001b[32mReading F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.203:0.203[0]; log p(e|m)= -1.201:-1.201[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.044]; league[0.043]; league[0.043]; Bristol[0.043]; Hull[0.043]; Southampton[0.043]; Sheffield[0.042]; Sheffield[0.042]; Charlton[0.042]; Coventry[0.042]; Blackburn[0.042]; Wolverhampton[0.042]; Sunderland[0.041]; Oxford[0.041]; English[0.041]; Enfield[0.041]; Peterborough[0.041]; Bolton[0.041]; Swindon[0.041]; matches[0.041]; Carlisle[0.041]; Division[0.041]; Birmingham[0.041]; Manchester[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burnley ==> ENTITY: \u001b[32mBurnley F.C.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.190:0.190[0]; log p(e|m)= -0.992:-0.992[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.046]; Chesterfield[0.045]; County[0.044]; Hull[0.043]; Watford[0.043]; Stockport[0.043]; Walsall[0.042]; Gillingham[0.042]; Carlisle[0.042]; Brentford[0.042]; Scunthorpe[0.041]; Darlington[0.041]; Portsmouth[0.041]; Rochdale[0.041]; Bristol[0.041]; Town[0.041]; Town[0.041]; Tranmere[0.040]; Peterborough[0.040]; Wycombe[0.040]; Wrexham[0.040]; City[0.040]; Enfield[0.040]; Mansfield[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Middlesbrough ==> ENTITY: \u001b[32mMiddlesbrough F.C.\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.208:0.208[0]; log p(e|m)= -0.518:-0.518[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; Charlton[0.043]; league[0.043]; league[0.043]; Blackburn[0.042]; Southampton[0.042]; Manchester[0.042]; matches[0.042]; Bristol[0.042]; Sheffield[0.042]; Sheffield[0.042]; Park[0.042]; Bradford[0.042]; Sunderland[0.042]; Soccer[0.041]; Leeds[0.041]; Birmingham[0.041]; United[0.041]; Coventry[0.041]; City[0.041]; City[0.041]; Portsmouth[0.040]; second[0.040]; Bolton[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bristol City\u001b[39m ==> ENTITY: \u001b[32mBristol City F.C.\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthampton[0.044]; Preston[0.043]; Hull[0.043]; Manchester[0.042]; Cup[0.042]; Watford[0.042]; Cardiff[0.042]; second[0.042]; Charlton[0.042]; Leeds[0.042]; City[0.042]; Walsall[0.041]; Norwich[0.041]; Portsmouth[0.041]; Darlington[0.041]; Reading[0.041]; Gillingham[0.041]; Sheffield[0.041]; Middlesbrough[0.041]; Chesterfield[0.041]; Stockport[0.041]; Brentford[0.041]; Ipswich[0.041]; Sunderland[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = St Albans ==> ENTITY: \u001b[32mSt Albans City F.C.\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.178:0.178[0]; log p(e|m)= -2.976:-2.976[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCambridge[0.044]; Southampton[0.043]; Preston[0.043]; Bristol[0.043]; Brentford[0.043]; Rochdale[0.043]; Norwich[0.042]; York[0.042]; Birmingham[0.042]; Chester[0.042]; Bradford[0.042]; Oxford[0.041]; Stockport[0.041]; Watford[0.041]; Enfield[0.041]; Charlton[0.041]; Peterborough[0.041]; Hull[0.041]; Walsall[0.041]; Huddersfield[0.041]; Stevenage[0.041]; Swindon[0.041]; Boston[0.041]; Wycombe[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barnsley ==> ENTITY: \u001b[32mBarnsley F.C.\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.204:0.204[0]; log p(e|m)= -0.713:-0.713[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwindon[0.044]; Wycombe[0.043]; Stoke[0.043]; Middlesbrough[0.043]; Ipswich[0.042]; Huddersfield[0.042]; Southend[0.042]; Sunderland[0.042]; Hednesford[0.042]; Leeds[0.042]; Charlton[0.042]; Wolverhampton[0.042]; Coventry[0.042]; Blackpool[0.042]; Liverpool[0.041]; Grimsby[0.041]; Southampton[0.041]; Tranmere[0.041]; Barnet[0.041]; Bradford[0.041]; Blackburn[0.041]; Sheffield[0.041]; Sheffield[0.041]; Tottenham[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Leyton Orient\u001b[39m ==> ENTITY: \u001b[32mLeyton Orient F.C.\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.225:0.225[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Hull[0.043]; Cardiff[0.042]; Stockport[0.042]; Norwich[0.042]; Walsall[0.042]; Chesterfield[0.042]; Watford[0.042]; Sheffield[0.042]; Brentford[0.042]; Oldham[0.042]; Burnley[0.042]; Manchester[0.042]; Portsmouth[0.042]; Rochdale[0.041]; Ipswich[0.041]; Gillingham[0.041]; Reading[0.041]; Bristol[0.041]; Carlisle[0.041]; Blackpool[0.041]; Scunthorpe[0.041]; Bolton[0.041]; Huddersfield[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Luton ==> ENTITY: \u001b[32mLuton Town F.C.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.221:0.221[0]; log p(e|m)= -2.551:-2.551[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.046]; Hull[0.044]; Walsall[0.042]; Watford[0.042]; Stockport[0.042]; Chesterfield[0.042]; Gillingham[0.042]; Sheffield[0.042]; Burnley[0.042]; Enfield[0.042]; Scunthorpe[0.042]; Brentford[0.042]; Manchester[0.041]; Bristol[0.041]; Norwich[0.041]; Wolverhampton[0.041]; Darlington[0.041]; Portsmouth[0.041]; Crewe[0.041]; Park[0.041]; Cardiff[0.041]; Blackpool[0.041]; Rochdale[0.041]; Oldham[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ashford Town\u001b[39m ==> ENTITY: \u001b[32mAshford United F.C.\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.043]; Chesterfield[0.043]; Stockport[0.043]; Wycombe[0.043]; Rochdale[0.043]; Darlington[0.042]; Town[0.042]; Notts[0.042]; Peterborough[0.042]; Hull[0.042]; Walsall[0.042]; Watford[0.042]; Barnet[0.041]; Mansfield[0.041]; Scarborough[0.041]; Brentford[0.041]; Chester[0.041]; second[0.041]; Cup[0.041]; Enfield[0.041]; Stevenage[0.041]; City[0.041]; Scunthorpe[0.041]; Bristol[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oxford ==> ENTITY: \u001b[32mOxford United F.C.\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.230:0.230[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCambridge[0.044]; Hull[0.043]; Southampton[0.043]; Coventry[0.043]; League[0.043]; league[0.042]; league[0.042]; Sheffield[0.042]; Sheffield[0.042]; Charlton[0.042]; Wolverhampton[0.042]; Peterborough[0.041]; Birmingham[0.041]; Blackburn[0.041]; City[0.041]; City[0.041]; Reading[0.041]; Sunderland[0.041]; Bristol[0.041]; Leeds[0.041]; Carlisle[0.041]; matches[0.040]; Norwich[0.040]; Swindon[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Blackburn ==> ENTITY: \u001b[32mBlackburn Rovers F.C.\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.205:0.205[0]; log p(e|m)= -2.733:-2.733[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; Charlton[0.043]; Sheffield[0.043]; Sheffield[0.043]; Park[0.043]; Sunderland[0.043]; league[0.043]; league[0.043]; Southampton[0.042]; matches[0.042]; Middlesbrough[0.042]; Manchester[0.042]; Leeds[0.042]; Coventry[0.041]; Soccer[0.041]; Birmingham[0.041]; Derby[0.041]; Bolton[0.041]; Bradford[0.041]; Wolverhampton[0.040]; Norwich[0.040]; Barnsley[0.040]; Reading[0.040]; Swindon[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Port Vale\u001b[39m ==> ENTITY: \u001b[32mPort Vale F.C.\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.044]; Preston[0.044]; Southampton[0.042]; Chesterfield[0.042]; Coventry[0.042]; league[0.042]; league[0.042]; Sheffield[0.042]; Sheffield[0.042]; Chester[0.042]; Bristol[0.042]; Southend[0.041]; Charlton[0.041]; Darlington[0.041]; Blackburn[0.041]; Peterborough[0.041]; Middlesbrough[0.041]; Stockport[0.041]; Barnsley[0.041]; Portsmouth[0.041]; Sunderland[0.041]; City[0.041]; City[0.041]; Norwich[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sheffield United\u001b[39m ==> ENTITY: \u001b[32mSheffield United F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.234:0.234[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Sheffield[0.043]; Hull[0.043]; Coventry[0.043]; Southampton[0.042]; Charlton[0.042]; Birmingham[0.042]; Walsall[0.042]; Sunderland[0.042]; Leeds[0.042]; Blackburn[0.042]; league[0.041]; league[0.041]; Burnley[0.041]; Carlisle[0.041]; Middlesbrough[0.041]; Stockport[0.041]; Southend[0.041]; Manchester[0.041]; matches[0.041]; Barnsley[0.041]; Brentford[0.041]; Oldham[0.041]; Norwich[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Arsenal ==> ENTITY: \u001b[32mArsenal F.C.\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.194:0.194[0]; log p(e|m)= -0.158:-0.158[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; league[0.045]; league[0.045]; matches[0.044]; Manchester[0.042]; Chelsea[0.042]; Charlton[0.042]; Sheffield[0.042]; Sheffield[0.042]; Soccer[0.041]; Birmingham[0.041]; Southampton[0.041]; Middlesbrough[0.041]; Coventry[0.041]; Liverpool[0.041]; Tottenham[0.041]; Blackburn[0.041]; Sunderland[0.041]; Villa[0.041]; Leeds[0.041]; Cup[0.041]; City[0.040]; Portsmouth[0.040]; Park[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scunthorpe ==> ENTITY: \u001b[32mScunthorpe United F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.214:0.214[0]; log p(e|m)= -1.510:-1.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.046]; Preston[0.044]; Stockport[0.043]; Cardiff[0.042]; Walsall[0.042]; Burnley[0.042]; Watford[0.042]; Bristol[0.042]; Rochdale[0.042]; Chesterfield[0.042]; Carlisle[0.041]; Peterborough[0.041]; Brentford[0.041]; Darlington[0.041]; Enfield[0.041]; Blackpool[0.041]; Luton[0.041]; Gillingham[0.041]; Wrexham[0.041]; second[0.041]; Crewe[0.041]; Bolton[0.040]; Wycombe[0.040]; Cup[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cambridge United\u001b[39m ==> ENTITY: \u001b[32mCambridge United F.C.\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.251:0.251[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.045]; Southampton[0.043]; Preston[0.043]; Charlton[0.043]; Manchester[0.042]; Norwich[0.042]; Watford[0.042]; Bristol[0.042]; Sheffield[0.042]; Peterborough[0.042]; Portsmouth[0.042]; Cardiff[0.041]; Sunderland[0.041]; Burnley[0.041]; Birmingham[0.041]; Division[0.041]; Carlisle[0.041]; Scunthorpe[0.041]; Walsall[0.041]; Southend[0.041]; Luton[0.041]; Gillingham[0.041]; Darlington[0.041]; Stockport[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Woking ==> ENTITY: \u001b[32mWoking F.C.\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.224:0.224[0]; log p(e|m)= -1.523:-1.523[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Sheffield[0.043]; Watford[0.042]; Birmingham[0.042]; Bristol[0.042]; Enfield[0.042]; Stockport[0.042]; second[0.042]; Peterborough[0.042]; Swindon[0.042]; Walsall[0.041]; Rochdale[0.041]; Southend[0.041]; Bradford[0.041]; Darlington[0.041]; Barnet[0.041]; Charlton[0.041]; Hull[0.041]; Burnley[0.041]; Ipswich[0.041]; Oxford[0.041]; Cardiff[0.041]; Portsmouth[0.041]; Chesterfield[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.087:0.087[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.044]; league[0.044]; league[0.044]; League[0.043]; Manchester[0.043]; Charlton[0.042]; Sheffield[0.042]; Birmingham[0.042]; Southampton[0.042]; Middlesbrough[0.042]; Coventry[0.042]; Blackburn[0.042]; Sunderland[0.041]; Soccer[0.041]; Leeds[0.041]; City[0.041]; Wolverhampton[0.041]; Cup[0.041]; Liverpool[0.041]; Tottenham[0.041]; Oxford[0.040]; Norwich[0.040]; Park[0.040]; Arsenal[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Southend ==> ENTITY: \u001b[32mSouthend United F.C.\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.252:0.252[0]; log p(e|m)= -1.772:-1.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthampton[0.044]; Charlton[0.043]; Coventry[0.043]; League[0.043]; Sheffield[0.043]; Sheffield[0.043]; Bristol[0.042]; Chesterfield[0.042]; Blackburn[0.042]; Oldham[0.042]; Leeds[0.041]; Reading[0.041]; Sunderland[0.041]; Norwich[0.041]; Portsmouth[0.041]; league[0.041]; league[0.041]; matches[0.041]; Manchester[0.041]; Carlisle[0.041]; Barnsley[0.041]; Middlesbrough[0.041]; City[0.041]; City[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mansfield ==> ENTITY: \u001b[32mMansfield Town F.C.\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.249:0.249[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.044]; Hull[0.043]; Stockport[0.043]; Carlisle[0.043]; Reading[0.043]; Walsall[0.042]; Bradford[0.042]; Sheffield[0.042]; Watford[0.042]; Oldham[0.042]; Peterborough[0.042]; Chesterfield[0.041]; Rochdale[0.041]; Brentford[0.041]; Bristol[0.041]; Portsmouth[0.041]; Wycombe[0.041]; Cardiff[0.041]; Burnley[0.041]; Town[0.041]; Town[0.041]; Scunthorpe[0.041]; Gillingham[0.041]; Chester[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Notts County\u001b[39m ==> ENTITY: \u001b[32mNotts County F.C.\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.232:0.232[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Hull[0.044]; Walsall[0.043]; Sheffield[0.043]; Chesterfield[0.042]; Burnley[0.042]; Watford[0.042]; Gillingham[0.042]; Scunthorpe[0.042]; Stockport[0.042]; Portsmouth[0.042]; Brentford[0.042]; Peterborough[0.042]; Luton[0.041]; Rochdale[0.041]; Oldham[0.041]; Darlington[0.041]; Town[0.041]; Town[0.041]; second[0.041]; Carlisle[0.041]; Cardiff[0.041]; Wycombe[0.041]; Reading[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Coventry ==> ENTITY: \u001b[32mCoventry City F.C.\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.212:0.212[0]; log p(e|m)= -3.124:-3.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.044]; Leeds[0.043]; Blackburn[0.043]; League[0.043]; Southampton[0.043]; Leicester[0.042]; Sheffield[0.042]; Sheffield[0.042]; Huddersfield[0.042]; Birmingham[0.042]; Manchester[0.042]; league[0.042]; league[0.042]; matches[0.041]; City[0.041]; Middlesbrough[0.041]; Sunderland[0.041]; Bolton[0.041]; Portsmouth[0.041]; Swindon[0.041]; Southend[0.040]; Oxford[0.040]; Barnsley[0.040]; Bradford[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Huddersfield ==> ENTITY: \u001b[32mHuddersfield Town F.C.\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.219:0.219[0]; log p(e|m)= -2.283:-2.283[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.044]; Hull[0.043]; Leeds[0.043]; League[0.043]; Bradford[0.042]; Sheffield[0.042]; Sheffield[0.042]; Southampton[0.042]; Wolverhampton[0.042]; league[0.042]; league[0.042]; Coventry[0.042]; Charlton[0.041]; Leicester[0.041]; Bristol[0.041]; Sunderland[0.041]; Chesterfield[0.041]; Manchester[0.041]; second[0.041]; Birmingham[0.041]; Park[0.040]; Middlesbrough[0.040]; Blackburn[0.040]; Oldham[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gillingham ==> ENTITY: \u001b[32mGillingham F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.214:0.214[0]; log p(e|m)= -0.257:-0.257[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; County[0.044]; Hull[0.043]; Brentford[0.043]; Watford[0.043]; Walsall[0.042]; Stockport[0.042]; Peterborough[0.042]; Enfield[0.042]; City[0.042]; Chesterfield[0.042]; Rochdale[0.042]; Town[0.041]; Town[0.041]; Scunthorpe[0.041]; Burnley[0.041]; Wycombe[0.041]; Carlisle[0.041]; Bristol[0.041]; Cardiff[0.041]; Darlington[0.041]; second[0.041]; Blackpool[0.040]; Luton[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portsmouth ==> ENTITY: \u001b[32mPortsmouth F.C.\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.235:0.235[0]; log p(e|m)= -1.231:-1.231[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.044]; Southampton[0.044]; Charlton[0.044]; Hull[0.043]; Watford[0.043]; Blackburn[0.043]; Sunderland[0.042]; Coventry[0.042]; Middlesbrough[0.042]; Bristol[0.042]; league[0.042]; league[0.042]; Burnley[0.041]; Manchester[0.041]; Birmingham[0.041]; Palace[0.041]; Norwich[0.041]; Sheffield[0.041]; Sheffield[0.041]; matches[0.041]; Ipswich[0.041]; York[0.041]; Bolton[0.041]; Brentford[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Stoke ==> ENTITY: \u001b[32mStoke City F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.231:0.231[0]; log p(e|m)= -0.341:-0.341[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.044]; Portsmouth[0.043]; Southampton[0.043]; Hull[0.043]; Coventry[0.042]; league[0.042]; Southend[0.042]; Charlton[0.042]; Wolverhampton[0.042]; Sunderland[0.042]; Sheffield[0.042]; Sheffield[0.042]; Bristol[0.042]; Manchester[0.041]; Norwich[0.041]; Watford[0.041]; Blackburn[0.041]; City[0.041]; City[0.041]; Birmingham[0.041]; Stockport[0.041]; Ipswich[0.041]; Enfield[0.041]; Middlesbrough[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leeds ==> ENTITY: \u001b[32mLeeds United A.F.C.\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.209:0.209[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBradford[0.044]; Bristol[0.043]; Oxford[0.043]; League[0.043]; Huddersfield[0.042]; Wolverhampton[0.042]; Birmingham[0.042]; Blackburn[0.042]; Sheffield[0.042]; Sheffield[0.042]; Coventry[0.042]; Charlton[0.042]; Southampton[0.042]; league[0.041]; league[0.041]; Manchester[0.041]; Southend[0.041]; Middlesbrough[0.041]; Oldham[0.041]; Sunderland[0.041]; Swindon[0.041]; Norwich[0.041]; matches[0.040]; Leicester[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Carlisle ==> ENTITY: \u001b[32mCarlisle United F.C.\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.201:0.201[0]; log p(e|m)= -3.612:-3.612[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.044]; Bristol[0.043]; Hull[0.043]; Bradford[0.043]; Rochdale[0.042]; Peterborough[0.042]; Manchester[0.042]; second[0.042]; Stockport[0.042]; Sheffield[0.042]; Cardiff[0.041]; Walsall[0.041]; Reading[0.041]; Enfield[0.041]; Norwich[0.041]; York[0.041]; Charlton[0.041]; Luton[0.041]; Watford[0.041]; City[0.041]; City[0.041]; Sunderland[0.041]; Oldham[0.041]; Park[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Darlington ==> ENTITY: \u001b[32mDarlington F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.215:0.215[0]; log p(e|m)= -1.871:-1.871[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.044]; Hull[0.043]; Wolverhampton[0.043]; Bristol[0.043]; Chesterfield[0.043]; Charlton[0.042]; Ipswich[0.042]; Stockport[0.042]; Sheffield[0.042]; Peterborough[0.041]; Barnsley[0.041]; Scunthorpe[0.041]; York[0.041]; Enfield[0.041]; Portsmouth[0.041]; Gillingham[0.041]; Carlisle[0.041]; Huddersfield[0.041]; Swindon[0.041]; Watford[0.041]; Walsall[0.041]; County[0.041]; Cardiff[0.041]; Burnley[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chelsea ==> ENTITY: \u001b[32mChelsea F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.202:0.202[0]; log p(e|m)= -0.518:-0.518[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.045]; league[0.043]; league[0.043]; Manchester[0.043]; Portsmouth[0.042]; City[0.042]; Charlton[0.042]; London[0.042]; matches[0.042]; Sheffield[0.042]; Sheffield[0.042]; Queens[0.041]; Soccer[0.041]; Division[0.041]; Southampton[0.041]; Blackburn[0.041]; Coventry[0.041]; Middlesbrough[0.041]; Leeds[0.041]; Liverpool[0.041]; Birmingham[0.041]; Oldham[0.040]; Sunderland[0.040]; Tottenham[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Birmingham ==> ENTITY: \u001b[32mBirmingham City F.C.\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.190:0.190[0]; log p(e|m)= -3.101:-3.101[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOxford[0.043]; Charlton[0.043]; Cambridge[0.043]; Wolverhampton[0.043]; League[0.042]; Bristol[0.042]; league[0.042]; league[0.042]; Southampton[0.042]; Sheffield[0.042]; Sheffield[0.042]; Coventry[0.042]; Blackburn[0.042]; Manchester[0.041]; matches[0.041]; Sunderland[0.041]; Leeds[0.041]; Middlesbrough[0.041]; Carlisle[0.041]; City[0.041]; City[0.041]; London[0.041]; Norwich[0.041]; Blackpool[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Preston ==> ENTITY: \u001b[32mPreston North End F.C.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.185:0.185[0]; log p(e|m)= -3.194:-3.194[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.045]; Carlisle[0.043]; Watford[0.043]; Sheffield[0.042]; Walsall[0.042]; Bristol[0.042]; Burnley[0.042]; Stockport[0.042]; Rochdale[0.042]; Chesterfield[0.042]; City[0.042]; Enfield[0.041]; Brentford[0.041]; Gillingham[0.041]; Scunthorpe[0.041]; Darlington[0.041]; second[0.041]; Peterborough[0.041]; Scarborough[0.041]; Cardiff[0.041]; Bolton[0.041]; Portsmouth[0.041]; Ashford[0.041]; Blackpool[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sudbury Town ==> ENTITY: \u001b[32mSudbury Town F.C.\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.184:0.184[0]; log p(e|m)= -0.104:-0.104[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPeterborough[0.045]; Rochdale[0.044]; Town[0.044]; Cup[0.043]; Wycombe[0.043]; Enfield[0.042]; Stockport[0.042]; Walsall[0.042]; Scarborough[0.042]; Notts[0.042]; Stevenage[0.042]; Preston[0.042]; Brentford[0.041]; Barnet[0.041]; West[0.041]; United[0.041]; United[0.041]; Gillingham[0.041]; Luton[0.041]; second[0.041]; York[0.040]; Scunthorpe[0.040]; Albans[0.040]; Hednesford[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norwich ==> ENTITY: \u001b[32mNorwich City F.C.\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.215:0.215[0]; log p(e|m)= -2.733:-2.733[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthampton[0.043]; Charlton[0.043]; League[0.042]; league[0.042]; league[0.042]; Hull[0.042]; Coventry[0.042]; Wolverhampton[0.042]; Blackburn[0.042]; Carlisle[0.042]; Sheffield[0.042]; Sheffield[0.042]; Sunderland[0.042]; Birmingham[0.042]; matches[0.041]; Portsmouth[0.041]; Manchester[0.041]; Enfield[0.041]; Bristol[0.041]; Middlesbrough[0.041]; Chesterfield[0.041]; Peterborough[0.041]; Barnsley[0.041]; Ipswich[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Liverpool ==> ENTITY: \u001b[32mLiverpool F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.180:0.180[0]; log p(e|m)= -1.171:-1.171[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.045]; league[0.043]; league[0.043]; Manchester[0.043]; matches[0.043]; Blackburn[0.042]; Oldham[0.042]; Middlesbrough[0.042]; Birmingham[0.042]; Bradford[0.042]; Charlton[0.042]; Sunderland[0.042]; Leeds[0.042]; Southampton[0.041]; Tottenham[0.041]; Sheffield[0.041]; Sheffield[0.041]; Soccer[0.041]; London[0.041]; Division[0.041]; Huddersfield[0.041]; second[0.041]; Wycombe[0.041]; Coventry[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tranmere ==> ENTITY: \u001b[32mTranmere Rovers F.C.\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.251:0.251[0]; log p(e|m)= -0.600:-0.600[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.044]; Hull[0.043]; Charlton[0.043]; Southampton[0.043]; Sunderland[0.042]; Watford[0.042]; Blackburn[0.042]; matches[0.042]; Burnley[0.042]; Coventry[0.042]; Middlesbrough[0.042]; Stockport[0.042]; Walsall[0.041]; league[0.041]; Leeds[0.041]; Portsmouth[0.041]; Southend[0.041]; Brentford[0.041]; Barnsley[0.041]; Sheffield[0.041]; Sheffield[0.041]; Carlisle[0.041]; Manchester[0.041]; Birmingham[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Bromwich ==> ENTITY: \u001b[32mWest Bromwich Albion F.C.\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.229:0.229[0]; log p(e|m)= -1.501:-1.501[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.044]; league[0.043]; Charlton[0.043]; Blackburn[0.043]; Watford[0.043]; Birmingham[0.042]; Southampton[0.042]; Hull[0.042]; Wolverhampton[0.042]; Sheffield[0.042]; Sheffield[0.042]; Coventry[0.041]; Middlesbrough[0.041]; City[0.041]; City[0.041]; Bolton[0.041]; Bradford[0.041]; Enfield[0.041]; Sunderland[0.041]; Oxford[0.041]; Bristol[0.041]; Manchester[0.041]; Aston[0.041]; Portsmouth[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Southampton ==> ENTITY: \u001b[32mSouthampton F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.184:0.184[0]; log p(e|m)= -1.149:-1.149[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPortsmouth[0.046]; Coventry[0.045]; Norwich[0.044]; Ipswich[0.044]; Grimsby[0.044]; Bristol[0.043]; Southend[0.043]; Sunderland[0.042]; Cambridge[0.042]; Liverpool[0.042]; Manchester[0.042]; Queens[0.041]; Middlesbrough[0.041]; Barnsley[0.041]; Charlton[0.041]; Blackpool[0.041]; Oxford[0.040]; League[0.040]; Chelsea[0.040]; Barnet[0.040]; Blackburn[0.040]; Wolverhampton[0.040]; Stoke[0.040]; Birmingham[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chester ==> ENTITY: \u001b[32mChester City F.C.\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.220:0.220[0]; log p(e|m)= -3.219:-3.219[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Hull[0.044]; Charlton[0.042]; Stockport[0.042]; Watford[0.042]; Walsall[0.042]; Carlisle[0.042]; Bristol[0.042]; Gillingham[0.041]; Peterborough[0.041]; Town[0.041]; Town[0.041]; Brentford[0.041]; Portsmouth[0.041]; Southend[0.041]; Burnley[0.041]; Norwich[0.041]; Oldham[0.041]; Chesterfield[0.041]; Grimsby[0.041]; Darlington[0.041]; Sheffield[0.041]; Swindon[0.041]; Scunthorpe[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Grimsby ==> ENTITY: \u001b[32mGrimsby Town F.C.\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.234:0.234[0]; log p(e|m)= -2.604:-2.604[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthampton[0.044]; Bristol[0.044]; Coventry[0.043]; Sheffield[0.042]; Sheffield[0.042]; Ipswich[0.042]; league[0.042]; league[0.042]; Southend[0.042]; League[0.042]; Leeds[0.042]; Swindon[0.042]; Sunderland[0.042]; Norwich[0.041]; Portsmouth[0.041]; Charlton[0.041]; Chesterfield[0.041]; Huddersfield[0.041]; Middlesbrough[0.041]; Barnsley[0.041]; Darlington[0.041]; Oldham[0.041]; Manchester[0.041]; Bradford[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Walsall ==> ENTITY: \u001b[32mWalsall F.C.\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.193:0.193[0]; log p(e|m)= -0.620:-0.620[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Hull[0.043]; Enfield[0.042]; Rochdale[0.042]; Stockport[0.042]; Watford[0.042]; second[0.042]; Bristol[0.042]; Brentford[0.042]; Luton[0.042]; City[0.042]; Sheffield[0.042]; Peterborough[0.041]; Hednesford[0.041]; Burnley[0.041]; Darlington[0.041]; Scunthorpe[0.041]; Gillingham[0.041]; Carlisle[0.041]; Cardiff[0.041]; Stevenage[0.041]; Crewe[0.041]; Wycombe[0.041]; Wrexham[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Watford ==> ENTITY: \u001b[32mWatford F.C.\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.960:-0.960[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Hull[0.043]; Brentford[0.043]; Cardiff[0.042]; Burnley[0.042]; Portsmouth[0.042]; Walsall[0.042]; Bristol[0.042]; Stockport[0.042]; Enfield[0.042]; Chesterfield[0.042]; Luton[0.041]; Carlisle[0.041]; Blackpool[0.041]; Scunthorpe[0.041]; Bolton[0.041]; Rochdale[0.041]; Gillingham[0.041]; Wycombe[0.041]; Crewe[0.041]; Peterborough[0.041]; City[0.041]; second[0.040]; Tranmere[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tottenham ==> ENTITY: \u001b[32mTottenham Hotspur F.C.\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.221:0.221[0]; log p(e|m)= -0.997:-0.997[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; Charlton[0.045]; league[0.045]; league[0.045]; matches[0.045]; Southampton[0.044]; Manchester[0.044]; Blackburn[0.044]; Soccer[0.044]; Birmingham[0.043]; Sheffield[0.043]; Sheffield[0.043]; London[0.043]; Sunderland[0.043]; Bolton[0.043]; Coventry[0.043]; Middlesbrough[0.043]; Reading[0.043]; Portsmouth[0.042]; Leeds[0.042]; Park[0.042]; City[0.042]; Liverpool[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ipswich ==> ENTITY: \u001b[32mIpswich Town F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.229:0.229[0]; log p(e|m)= -2.688:-2.688[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.043]; League[0.043]; Coventry[0.043]; Charlton[0.043]; Southampton[0.043]; Bristol[0.042]; Sunderland[0.042]; league[0.042]; league[0.042]; Sheffield[0.042]; Sheffield[0.042]; Manchester[0.042]; Blackburn[0.042]; Enfield[0.042]; Middlesbrough[0.041]; Bolton[0.041]; Portsmouth[0.041]; Birmingham[0.041]; Norwich[0.041]; Leeds[0.041]; matches[0.041]; Barnsley[0.040]; Reading[0.040]; Southend[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = English ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.132:0.132[0]; log p(e|m)= -1.355:-1.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEnglish[0.046]; League[0.045]; league[0.044]; league[0.044]; Leeds[0.043]; matches[0.043]; Coventry[0.042]; Division[0.042]; Oxford[0.042]; Birmingham[0.041]; Swindon[0.041]; Southampton[0.041]; Sheffield[0.041]; Huddersfield[0.041]; Barnsley[0.041]; Leicester[0.041]; Southend[0.041]; Oldham[0.041]; Norwich[0.041]; Bradford[0.040]; Cup[0.040]; Ipswich[0.040]; City[0.040]; Sunderland[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Premier league\u001b[39m ==> ENTITY: \u001b[32mPremier League\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; Southampton[0.043]; Sunderland[0.043]; league[0.043]; Charlton[0.043]; Birmingham[0.043]; Middlesbrough[0.042]; matches[0.042]; Blackburn[0.042]; Soccer[0.042]; Manchester[0.042]; Sheffield[0.042]; Sheffield[0.042]; Coventry[0.041]; Leeds[0.041]; City[0.041]; Reading[0.041]; Portsmouth[0.041]; Tottenham[0.041]; Barnsley[0.041]; Liverpool[0.041]; Stoke[0.040]; Southend[0.040]; Norwich[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Stevenage ==> ENTITY: \u001b[32mStevenage F.C.\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.254:0.254[0]; log p(e|m)= -0.607:-0.607[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Hull[0.043]; Stockport[0.043]; Chesterfield[0.043]; Bradford[0.042]; Oldham[0.042]; Walsall[0.042]; Sheffield[0.042]; Enfield[0.042]; Watford[0.042]; Bristol[0.042]; Carlisle[0.042]; Rochdale[0.041]; Portsmouth[0.041]; Burnley[0.041]; Brentford[0.041]; Darlington[0.041]; Wycombe[0.041]; Manchester[0.041]; Reading[0.041]; Norwich[0.041]; Ipswich[0.041]; Wolverhampton[0.041]; Huddersfield[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bolton ==> ENTITY: \u001b[32mBolton Wanderers F.C.\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.191:0.191[0]; log p(e|m)= -2.283:-2.283[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStockport[0.044]; Preston[0.043]; Charlton[0.043]; Blackburn[0.042]; Hull[0.042]; Manchester[0.042]; Mansfield[0.042]; Bradford[0.042]; Sunderland[0.042]; Watford[0.042]; Southampton[0.042]; Coventry[0.042]; York[0.041]; Oldham[0.041]; Leeds[0.041]; Leicester[0.041]; Luton[0.041]; Middlesbrough[0.041]; Sheffield[0.041]; Sheffield[0.041]; Burnley[0.041]; Bristol[0.041]; Enfield[0.041]; Walsall[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = F.A. Challenge Cup\u001b[39m ==> ENTITY: \u001b[32mFA Cup\u001b[39m\t\nSCORES: global= 0.313:0.313[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.044]; Hull[0.043]; Coventry[0.042]; Southampton[0.042]; Watford[0.042]; Sheffield[0.042]; Sheffield[0.042]; Charlton[0.042]; Cardiff[0.042]; Bristol[0.042]; second[0.042]; Leeds[0.042]; Portsmouth[0.042]; Peterborough[0.041]; Rochdale[0.041]; Southend[0.041]; Burnley[0.041]; Manchester[0.041]; Birmingham[0.041]; Walsall[0.041]; Sunderland[0.041]; Swindon[0.041]; Brentford[0.041]; Wycombe[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Aston Villa\u001b[39m ==> ENTITY: \u001b[32mAston Villa F.C.\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.222:0.222[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; league[0.043]; league[0.043]; Charlton[0.043]; Southampton[0.043]; matches[0.042]; Blackburn[0.042]; Middlesbrough[0.042]; Birmingham[0.042]; Sunderland[0.042]; Coventry[0.042]; Manchester[0.042]; Ipswich[0.041]; Sheffield[0.041]; Sheffield[0.041]; Park[0.041]; Soccer[0.041]; Leeds[0.041]; Bolton[0.041]; City[0.040]; City[0.040]; Portsmouth[0.040]; Barnsley[0.040]; Blackpool[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chesterfield ==> ENTITY: \u001b[32mChesterfield F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.202:0.202[0]; log p(e|m)= -1.269:-1.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.044]; Hull[0.043]; Birmingham[0.043]; Carlisle[0.043]; Stockport[0.042]; Oxford[0.042]; Bradford[0.042]; Bristol[0.042]; Wolverhampton[0.042]; Walsall[0.042]; Oldham[0.042]; second[0.042]; Sheffield[0.041]; Norwich[0.041]; York[0.041]; Rochdale[0.041]; Portsmouth[0.041]; Watford[0.041]; Burnley[0.041]; Manchester[0.041]; Charlton[0.041]; Swindon[0.041]; Southend[0.041]; Darlington[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scarborough ==> ENTITY: \u001b[32mScarborough F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.219:0.219[0]; log p(e|m)= -1.814:-1.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPeterborough[0.043]; Swindon[0.043]; Preston[0.043]; Rochdale[0.043]; Walsall[0.043]; Bristol[0.043]; Carlisle[0.042]; Sheffield[0.042]; Birmingham[0.042]; Stockport[0.042]; Ipswich[0.042]; Portsmouth[0.042]; Hull[0.041]; Darlington[0.041]; Huddersfield[0.041]; Norwich[0.041]; Cardiff[0.041]; Watford[0.041]; Grimsby[0.041]; Wrexham[0.041]; Oxford[0.041]; Luton[0.041]; Blackpool[0.041]; Wolverhampton[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Boreham Wood\u001b[39m ==> ENTITY: \u001b[32mBoreham Wood F.C.\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.044]; Enfield[0.043]; Preston[0.043]; Peterborough[0.043]; Sheffield[0.042]; Hull[0.042]; Chester[0.042]; Rochdale[0.042]; Darlington[0.042]; Scarborough[0.042]; Bristol[0.042]; City[0.042]; City[0.042]; Stockport[0.041]; Chesterfield[0.041]; Wolverhampton[0.041]; Park[0.041]; Luton[0.041]; Watford[0.041]; Wycombe[0.041]; Bradford[0.041]; Walsall[0.041]; Stevenage[0.041]; Cup[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Charlton ==> ENTITY: \u001b[32mCharlton Athletic F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.224:0.224[0]; log p(e|m)= -2.104:-2.104[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthampton[0.043]; League[0.043]; Coventry[0.043]; Sheffield[0.042]; Sheffield[0.042]; league[0.042]; league[0.042]; Blackburn[0.042]; Norwich[0.042]; Manchester[0.042]; Sunderland[0.042]; Portsmouth[0.042]; Middlesbrough[0.042]; Enfield[0.042]; Bolton[0.041]; Bristol[0.041]; Reading[0.041]; Southend[0.041]; Leeds[0.041]; matches[0.041]; Blackpool[0.041]; City[0.041]; City[0.041]; Birmingham[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wrexham ==> ENTITY: \u001b[32mWrexham F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.205:0.205[0]; log p(e|m)= -0.882:-0.882[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.046]; Hull[0.044]; Cardiff[0.043]; Stockport[0.043]; Walsall[0.043]; Chesterfield[0.042]; Bristol[0.042]; Peterborough[0.042]; Watford[0.042]; Rochdale[0.042]; Chester[0.041]; Gillingham[0.041]; second[0.041]; Scunthorpe[0.041]; Luton[0.041]; Darlington[0.041]; Wycombe[0.041]; Carlisle[0.041]; Brentford[0.041]; Burnley[0.041]; Crewe[0.040]; County[0.040]; Town[0.040]; Town[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cardiff ==> ENTITY: \u001b[32mCardiff City F.C.\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.210:0.210[0]; log p(e|m)= -3.474:-3.474[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.046]; Hull[0.044]; Bristol[0.043]; Carlisle[0.043]; Watford[0.043]; Walsall[0.042]; Stockport[0.042]; Cup[0.042]; Burnley[0.042]; Chesterfield[0.042]; Brentford[0.042]; Wrexham[0.041]; Rochdale[0.041]; Scunthorpe[0.041]; Luton[0.041]; Blackpool[0.041]; second[0.041]; Wycombe[0.041]; Crewe[0.041]; Peterborough[0.041]; Darlington[0.041]; City[0.040]; Gillingham[0.040]; Enfield[0.040]; \t\n\r [..............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s750ms | Step: 2ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 83/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1270testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tokyo ==> ENTITY: \u001b[32mTokyo\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.093:-0.093[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.047]; Japan[0.047]; Japan[0.047]; Japan[0.047]; appearance[0.043]; yen[0.043]; Thursday[0.042]; market[0.042]; Friday[0.042]; Friday[0.042]; Yasuo[0.041]; seeking[0.041]; central[0.041]; economic[0.040]; nation[0.040]; Eisuke[0.040]; bank[0.040]; authorities[0.040]; authorities[0.040]; late[0.040]; levels[0.039]; seen[0.039]; dollar[0.039]; dollar[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = BOJ\u001b[39m ==> ENTITY: \u001b[32mBank of Japan\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBoj[0.047]; Boj[0.047]; currency[0.043]; Japan[0.043]; dollar[0.042]; dollar[0.042]; yen[0.042]; yen[0.042]; yen[0.042]; yen[0.042]; yen[0.042]; yen[0.042]; yen[0.042]; yen[0.042]; Japanese[0.041]; rate[0.041]; exchange[0.040]; economy[0.040]; economy[0.040]; central[0.040]; Banking[0.040]; Keizai[0.040]; forex[0.039]; bank[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.060:0.060[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndollar[0.044]; dollar[0.044]; dollar[0.044]; dollar[0.044]; dollar[0.044]; dollar[0.044]; dollar[0.044]; dollar[0.044]; States[0.042]; level[0.042]; acceptable[0.041]; current[0.040]; Robert[0.040]; Bank[0.040]; Bank[0.040]; shows[0.040]; Switzerland[0.040]; certain[0.040]; Rubin[0.040]; Rubin[0.040]; ranking[0.040]; does[0.040]; high[0.040]; necessarily[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.S. Treasury\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of the Treasury\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTreasury[0.047]; Bank[0.045]; late[0.043]; government[0.042]; States[0.042]; Robert[0.042]; Finance[0.042]; Secretary[0.041]; dollar[0.041]; dollar[0.041]; dollar[0.041]; dollar[0.041]; dollar[0.041]; dollar[0.041]; dollar[0.041]; official[0.041]; fall[0.041]; Rubin[0.041]; Rubin[0.041]; dealer[0.041]; economic[0.041]; Thursday[0.040]; Eisuke[0.040]; does[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = BOJ\u001b[39m ==> ENTITY: \u001b[32mBank of Japan\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBoj[0.049]; Boj[0.049]; dollar[0.043]; yen[0.043]; yen[0.043]; financial[0.043]; Japanese[0.042]; Tokyo[0.042]; economy[0.041]; economy[0.041]; markets[0.041]; markets[0.041]; markets[0.041]; Senior[0.041]; Credit[0.041]; consumption[0.040]; consumption[0.040]; senior[0.040]; market[0.040]; Reuters[0.040]; April[0.040]; officials[0.040]; Friday[0.040]; official[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.053]; Japan[0.053]; Japan[0.053]; Tokyo[0.046]; government[0.045]; central[0.044]; Secretary[0.044]; official[0.044]; appearance[0.044]; nation[0.043]; governor[0.042]; Robert[0.042]; Ministry[0.042]; rare[0.042]; Yasuo[0.041]; Bank[0.041]; late[0.041]; Rubin[0.040]; Eisuke[0.040]; influential[0.040]; widely[0.040]; yen[0.040]; yen[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTokyo[0.046]; Switzerland[0.046]; president[0.043]; governor[0.043]; International[0.042]; stronger[0.041]; stronger[0.041]; Bank[0.041]; level[0.041]; level[0.041]; hit[0.041]; following[0.041]; Takao[0.041]; November[0.041]; ministry[0.041]; recent[0.041]; believed[0.041]; prominence[0.041]; yen[0.041]; yen[0.041]; yen[0.041]; yen[0.041]; yen[0.041]; yen[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Credit Suisse\u001b[39m ==> ENTITY: \u001b[32mCredit Suisse\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.122:0.122[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfinancial[0.049]; bank[0.048]; Banking[0.047]; investors[0.046]; Senior[0.045]; senior[0.044]; April[0.044]; market[0.044]; Yasuhito[0.043]; Toyo[0.043]; Trust[0.043]; tax[0.043]; tax[0.043]; manager[0.042]; Friday[0.042]; Japan[0.042]; dollar[0.042]; markets[0.042]; markets[0.042]; markets[0.042]; Reuters[0.042]; forex[0.042]; sell[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSumitomo\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSumitomo Group\u001b[39m <---> \u001b[32mThe Sumitomo Bank\u001b[39m\t\nSCORES: global= 0.258:0.253[\u001b[31m0.005\u001b[39m]; local()= 0.107:0.169[\u001b[32m0.062\u001b[39m]; log p(e|m)= -0.020:-1.599[\u001b[31m1.579\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSumitomo[0.057]; Takao[0.046]; Japan[0.046]; Sakakibara[0.045]; Sakakibara[0.045]; Sakakibara[0.045]; Sakakibara[0.045]; Tokyo[0.045]; yen[0.045]; yen[0.045]; yen[0.045]; Bank[0.041]; Bank[0.041]; high[0.038]; president[0.038]; International[0.037]; dollar[0.037]; dollar[0.037]; dollar[0.037]; dollar[0.037]; dollar[0.037]; dollar[0.037]; dollar[0.037]; dollar[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rubin ==> ENTITY: \u001b[32mRobert Rubin\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.152:0.152[0]; log p(e|m)= -1.423:-1.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRubin[0.081]; Robert[0.061]; Hank[0.052]; thinks[0.052]; vice[0.051]; economic[0.051]; United[0.051]; high[0.050]; Bank[0.050]; Bank[0.050]; president[0.050]; Treasury[0.050]; chief[0.050]; does[0.050]; mean[0.050]; Secretary[0.050]; Finance[0.050]; lower[0.050]; ranking[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Finance Ministry ==> ENTITY: \u001b[32mMinistry of Finance (Japan)\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.084:-0.084[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBank[0.048]; Bank[0.048]; government[0.046]; bank[0.046]; Secretary[0.045]; Japan[0.044]; Japan[0.044]; central[0.044]; economic[0.044]; official[0.043]; Treasury[0.043]; Thursday[0.043]; nation[0.043]; current[0.042]; stance[0.042]; chief[0.042]; chief[0.042]; Yasuo[0.042]; dollar[0.042]; dollar[0.042]; dollar[0.042]; dollar[0.042]; dollar[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.081:0.081[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapanese[0.050]; Tokyo[0.048]; Taisuke[0.047]; Tanaka[0.045]; foreign[0.044]; lead[0.044]; lead[0.044]; past[0.044]; officials[0.044]; public[0.043]; Yasuhito[0.043]; Keizai[0.043]; does[0.042]; does[0.042]; does[0.042]; Shimbun[0.042]; central[0.042]; manager[0.042]; issue[0.042]; Kawashima[0.042]; Senior[0.042]; shows[0.042]; chief[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.142:0.142[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.052]; Japan[0.052]; Japan[0.052]; Tokyo[0.045]; central[0.044]; appearance[0.043]; nation[0.042]; governor[0.042]; rare[0.041]; Yasuo[0.041]; Bank[0.040]; late[0.040]; Eisuke[0.039]; current[0.039]; influential[0.039]; yen[0.039]; Friday[0.039]; Friday[0.039]; ranking[0.039]; Thursday[0.039]; leading[0.039]; chief[0.039]; Sakakibara[0.039]; seeking[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.100:0.100[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfinancial[0.043]; told[0.043]; market[0.043]; views[0.043]; views[0.043]; spending[0.042]; markets[0.042]; markets[0.042]; markets[0.042]; percent[0.042]; dollar[0.042]; negative[0.042]; negative[0.042]; negative[0.042]; official[0.042]; said[0.041]; said[0.041]; stocks[0.041]; economy[0.041]; economy[0.041]; April[0.041]; Credit[0.040]; public[0.040]; separately[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = BOJ\u001b[39m ==> ENTITY: \u001b[32mBank of Japan\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBoj[0.050]; currency[0.045]; currency[0.045]; Japan[0.045]; dollar[0.044]; dollar[0.044]; yen[0.044]; yen[0.044]; yen[0.044]; yen[0.044]; yen[0.044]; yen[0.044]; rate[0.043]; exchange[0.043]; economy[0.043]; Finance[0.042]; Banking[0.042]; Keizai[0.042]; Dealers[0.042]; forex[0.042]; inflation[0.041]; import[0.041]; import[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rubin ==> ENTITY: \u001b[32mRobert Rubin\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.148:0.148[0]; log p(e|m)= -1.423:-1.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRubin[0.073]; Robert[0.056]; Hank[0.048]; thinks[0.047]; vice[0.047]; level[0.046]; United[0.046]; high[0.046]; Bank[0.046]; Bank[0.046]; president[0.046]; Treasury[0.046]; chief[0.046]; does[0.046]; mean[0.045]; Secretary[0.045]; current[0.045]; Finance[0.045]; lower[0.045]; lower[0.045]; ranking[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.060:0.060[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndollar[0.045]; dollar[0.045]; dollar[0.045]; dollar[0.045]; dollar[0.045]; dollar[0.045]; certain[0.042]; States[0.042]; late[0.042]; television[0.042]; Robert[0.041]; Bank[0.041]; Bank[0.041]; shows[0.041]; followed[0.041]; necessarily[0.040]; Rubin[0.040]; Rubin[0.040]; ranking[0.040]; does[0.040]; widely[0.040]; saying[0.040]; remark[0.040]; mean[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tokyo ==> ENTITY: \u001b[32mTokyo\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.064:0.064[0]; log p(e|m)= -0.093:-0.093[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.046]; Japanese[0.043]; yen[0.043]; yen[0.043]; yen[0.043]; yen[0.043]; yen[0.043]; financial[0.043]; Taisuke[0.042]; market[0.042]; Friday[0.042]; central[0.042]; markets[0.041]; markets[0.041]; markets[0.041]; economy[0.041]; economy[0.041]; economy[0.041]; April[0.041]; foreign[0.040]; investors[0.040]; Banking[0.040]; stocks[0.040]; bank[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Union Bank of Switzerland\u001b[39m ==> ENTITY: \u001b[32mUnion Bank of Switzerland\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBank[0.046]; International[0.045]; Finance[0.044]; dollar[0.042]; dollar[0.042]; dollar[0.042]; dollar[0.042]; dollar[0.042]; dollar[0.042]; dollar[0.042]; market[0.042]; market[0.042]; currency[0.041]; Dealers[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; Treasury[0.041]; November[0.041]; Robert[0.041]; hit[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japanese ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.086:0.086[0]; log p(e|m)= -0.992:-0.992[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.052]; Tokyo[0.045]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; levels[0.043]; foreign[0.043]; Taisuke[0.042]; concerned[0.041]; negative[0.041]; negative[0.041]; negative[0.041]; economy[0.040]; economy[0.040]; economy[0.040]; past[0.040]; fact[0.040]; current[0.040]; weaker[0.040]; central[0.040]; Keizai[0.040]; Tanaka[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.068:0.068[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.047]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; took[0.043]; Union[0.042]; certain[0.042]; level[0.041]; Robert[0.041]; International[0.041]; strong[0.041]; strong[0.041]; participants[0.041]; think[0.041]; shows[0.041]; Bank[0.040]; Bank[0.040]; necessarily[0.040]; thought[0.040]; Asked[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.052]; Japan[0.052]; Japan[0.052]; Tokyo[0.045]; central[0.044]; nation[0.043]; governor[0.042]; rare[0.041]; Yasuo[0.041]; Bank[0.040]; current[0.040]; influential[0.040]; yen[0.039]; Friday[0.039]; Friday[0.039]; leading[0.039]; chief[0.039]; seeking[0.039]; immediate[0.039]; authorities[0.039]; authorities[0.039]; seen[0.039]; bureaucrat[0.039]; saying[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bank of Japan\u001b[39m ==> ENTITY: \u001b[32mBank of Japan\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.044]; Japan[0.044]; Japan[0.044]; dollar[0.043]; dollar[0.043]; dollar[0.043]; dollar[0.043]; yen[0.043]; yen[0.043]; economic[0.043]; Tokyo[0.042]; central[0.042]; Finance[0.041]; currencies[0.041]; bank[0.041]; government[0.040]; Yasuo[0.040]; governor[0.040]; high[0.040]; nation[0.040]; market[0.040]; market[0.040]; late[0.039]; dealers[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nihon Keizai Shimbun\u001b[39m ==> ENTITY: \u001b[32mNihon Keizai Shimbun\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapanese[0.045]; Japan[0.044]; yen[0.043]; yen[0.043]; yen[0.043]; yen[0.043]; yen[0.043]; yen[0.043]; yen[0.043]; yen[0.043]; dollar[0.042]; dollar[0.042]; rate[0.041]; Banking[0.041]; Yasuhito[0.041]; Toyo[0.041]; Mr[0.040]; foreign[0.040]; Yen[0.040]; exchange[0.040]; stocks[0.040]; recent[0.039]; currency[0.039]; forex[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = BOJ\u001b[39m ==> ENTITY: \u001b[32mBank of Japan\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBoj[0.048]; Boj[0.048]; Japan[0.043]; dollar[0.043]; yen[0.042]; yen[0.042]; yen[0.042]; yen[0.042]; yen[0.042]; financial[0.042]; Japanese[0.041]; Tokyo[0.041]; economy[0.041]; economy[0.041]; economy[0.041]; central[0.041]; markets[0.041]; markets[0.041]; markets[0.041]; Banking[0.040]; Senior[0.040]; forex[0.040]; bank[0.040]; Credit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tokyo ==> ENTITY: \u001b[32mTokyo\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.068:0.068[0]; log p(e|m)= -0.093:-0.093[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.057]; yen[0.052]; yen[0.052]; yen[0.052]; market[0.051]; market[0.051]; economy[0.050]; Finance[0.050]; International[0.050]; level[0.049]; Sumitomo[0.049]; Sumitomo[0.049]; November[0.049]; Switzerland[0.049]; high[0.048]; Dealers[0.048]; shows[0.048]; ministry[0.048]; sharply[0.048]; does[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sumitomo Bank\u001b[39m ==> ENTITY: \u001b[32mThe Sumitomo Bank\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBank[0.060]; high[0.056]; Finance[0.055]; yen[0.055]; yen[0.055]; yen[0.055]; Sumitomo[0.054]; late[0.053]; Tokyo[0.052]; level[0.052]; Thursday[0.052]; government[0.051]; market[0.050]; lower[0.050]; current[0.050]; Treasury[0.050]; dealer[0.050]; followed[0.049]; official[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = BOJ\u001b[39m ==> ENTITY: \u001b[32mBank of Japan\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBoj[0.047]; Boj[0.047]; Boj[0.047]; Japan[0.043]; dollar[0.042]; yen[0.042]; yen[0.042]; yen[0.042]; yen[0.042]; yen[0.042]; financial[0.041]; Japanese[0.041]; Tokyo[0.040]; economy[0.040]; economy[0.040]; economy[0.040]; central[0.040]; markets[0.040]; markets[0.040]; markets[0.040]; Banking[0.040]; Senior[0.040]; Keizai[0.040]; forex[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = BOJ\u001b[39m ==> ENTITY: \u001b[32mBank of Japan\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.044]; Japan[0.044]; Japan[0.044]; Japan[0.044]; dollar[0.043]; dollar[0.043]; dollar[0.043]; dollar[0.043]; Bank[0.043]; yen[0.043]; yen[0.043]; economic[0.043]; Tokyo[0.041]; central[0.041]; Finance[0.041]; currencies[0.041]; bank[0.040]; government[0.040]; daily[0.040]; market[0.040]; market[0.040]; Yasuo[0.040]; authorities[0.040]; authorities[0.040]; \t\n\r [>.............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s821ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 113/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1211testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = India ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.034:0.034[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.049]; tournament[0.048]; England[0.048]; England[0.048]; Squash[0.046]; Bombay[0.046]; squash[0.044]; Mahindra[0.044]; Mahindra[0.044]; Cairns[0.043]; Saturday[0.043]; International[0.042]; International[0.042]; Scotland[0.042]; Eyles[0.042]; Eyles[0.042]; Final[0.041]; Semifinal[0.041]; semifinals[0.041]; beat[0.041]; beat[0.041]; Friday[0.041]; Parke[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.084:0.084[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.048]; England[0.048]; India[0.046]; tournament[0.045]; Scotland[0.044]; Mark[0.043]; Peter[0.042]; Cairns[0.042]; semifinals[0.042]; Saturday[0.041]; squash[0.041]; Squash[0.040]; Parke[0.040]; Rodney[0.040]; International[0.040]; International[0.040]; Nicol[0.040]; Nicol[0.040]; Friday[0.040]; Final[0.040]; beat[0.039]; beat[0.039]; Eyles[0.039]; Eyles[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mark Cairns ==> ENTITY: \u001b[32mMark Cairns (squash player)\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.104:0.104[0]; log p(e|m)= -0.183:-0.183[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSquash[0.047]; tournament[0.046]; Australia[0.046]; Scotland[0.045]; England[0.045]; England[0.045]; Simon[0.045]; Semifinal[0.044]; squash[0.044]; Peter[0.044]; semifinals[0.043]; International[0.043]; International[0.043]; India[0.043]; Bombay[0.043]; Parke[0.042]; Nicol[0.042]; Nicol[0.042]; Friday[0.042]; Mahindra[0.042]; Mahindra[0.042]; Saturday[0.041]; Rodney[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEngland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEngland cricket team\u001b[39m <---> \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.241:0.239[\u001b[31m0.002\u001b[39m]; local()= 0.127:0.062[\u001b[31m0.065\u001b[39m]; log p(e|m)= -3.297:-0.190[\u001b[32m3.107\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.051]; Australia[0.048]; Scotland[0.046]; India[0.045]; tournament[0.043]; Mark[0.042]; Cairns[0.042]; Saturday[0.041]; Squash[0.041]; Eyles[0.041]; Eyles[0.041]; Simon[0.041]; Peter[0.041]; Final[0.041]; Parke[0.041]; Rodney[0.040]; International[0.040]; International[0.040]; Nicol[0.040]; Nicol[0.040]; beat[0.039]; beat[0.039]; squash[0.039]; Friday[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEngland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEngland cricket team\u001b[39m <---> \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.241:0.239[\u001b[31m0.002\u001b[39m]; local()= 0.127:0.062[\u001b[31m0.065\u001b[39m]; log p(e|m)= -3.297:-0.190[\u001b[32m3.107\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.051]; Australia[0.048]; Scotland[0.046]; India[0.045]; tournament[0.043]; Mark[0.042]; Cairns[0.042]; Saturday[0.041]; Squash[0.041]; Eyles[0.041]; Eyles[0.041]; Simon[0.041]; Peter[0.041]; Final[0.041]; Parke[0.041]; Rodney[0.040]; International[0.040]; International[0.040]; Nicol[0.040]; Nicol[0.040]; beat[0.039]; beat[0.039]; squash[0.039]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bombay ==> ENTITY: \u001b[32mMumbai\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.050:0.050[0]; log p(e|m)= -0.242:-0.242[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.052]; tournament[0.045]; England[0.045]; England[0.045]; Australia[0.044]; Squash[0.043]; International[0.043]; International[0.043]; beat[0.041]; beat[0.041]; Mahindra[0.041]; Mahindra[0.041]; Scotland[0.041]; squash[0.041]; semifinals[0.040]; Parke[0.040]; Saturday[0.040]; Cairns[0.039]; Rodney[0.039]; Eyles[0.039]; Eyles[0.039]; Mark[0.039]; Friday[0.039]; Final[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scotland ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.240:0.240[0]; local()= 0.050:0.050[0]; log p(e|m)= -0.229:-0.229[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.050]; England[0.050]; Australia[0.046]; tournament[0.045]; India[0.044]; Saturday[0.042]; Squash[0.041]; semifinals[0.041]; International[0.041]; International[0.041]; Cairns[0.041]; Nicol[0.041]; Nicol[0.041]; Final[0.041]; Mark[0.040]; Peter[0.040]; beat[0.040]; beat[0.040]; Parke[0.040]; Friday[0.040]; Rodney[0.039]; squash[0.039]; Simon[0.039]; Semifinal[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nicol\u001b[39m ==> ENTITY: \u001b[32mPeter Nicol\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.129:0.129[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.047]; Squash[0.046]; squash[0.046]; Scotland[0.044]; Semifinal[0.043]; Nicol[0.043]; Australia[0.043]; England[0.043]; England[0.043]; semifinals[0.042]; Simon[0.042]; Mark[0.041]; Peter[0.041]; International[0.041]; International[0.041]; beat[0.040]; beat[0.040]; Cairns[0.040]; India[0.039]; Friday[0.039]; Saturday[0.039]; Mahindra[0.039]; Mahindra[0.039]; Final[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rodney Eyles\u001b[39m ==> ENTITY: \u001b[32mRodney Eyles\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.046]; Squash[0.045]; tournament[0.044]; Semifinal[0.043]; semifinals[0.043]; India[0.043]; Nicol[0.042]; Nicol[0.042]; Final[0.042]; squash[0.042]; Eyles[0.042]; Peter[0.041]; International[0.041]; International[0.041]; England[0.041]; England[0.041]; Mark[0.041]; Cairns[0.041]; Simon[0.040]; Scotland[0.040]; Mahindra[0.039]; Mahindra[0.039]; Parke[0.039]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Peter Nicol\u001b[39m ==> ENTITY: \u001b[32mPeter Nicol\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.047]; Squash[0.046]; squash[0.046]; Scotland[0.044]; Semifinal[0.044]; Nicol[0.043]; Australia[0.043]; England[0.043]; England[0.043]; semifinals[0.043]; Simon[0.042]; Mark[0.041]; International[0.041]; International[0.041]; beat[0.040]; beat[0.040]; Cairns[0.040]; India[0.040]; Friday[0.039]; Saturday[0.039]; Mahindra[0.039]; Mahindra[0.039]; Final[0.039]; Parke[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Eyles\u001b[39m ==> ENTITY: \u001b[32mRodney Eyles\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.129:0.129[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.046]; Squash[0.045]; tournament[0.044]; Semifinal[0.043]; semifinals[0.043]; Rodney[0.043]; India[0.043]; Nicol[0.042]; Nicol[0.042]; Final[0.042]; squash[0.042]; Eyles[0.041]; Peter[0.041]; International[0.041]; International[0.041]; England[0.041]; England[0.041]; Mark[0.041]; Cairns[0.041]; Simon[0.040]; Scotland[0.040]; Mahindra[0.039]; Mahindra[0.039]; Parke[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Simon Parke\u001b[39m ==> ENTITY: \u001b[32mSimon Parke\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.098:0.098[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.049]; England[0.049]; Squash[0.048]; tournament[0.046]; Semifinal[0.046]; semifinals[0.045]; squash[0.045]; Australia[0.044]; Saturday[0.044]; Scotland[0.043]; Final[0.043]; Cairns[0.042]; International[0.042]; International[0.042]; Mark[0.042]; Nicol[0.042]; Nicol[0.042]; Friday[0.042]; Rodney[0.041]; Results[0.041]; Results[0.041]; Mahindra[0.041]; Mahindra[0.041]; \t\n\r [>.............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s986ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 125/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1386testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew York\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew York\u001b[39m <---> \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.236:0.235[\u001b[31m0.001\u001b[39m]; local()= 0.084:0.089[\u001b[32m0.006\u001b[39m]; log p(e|m)= -0.289:-1.790[\u001b[31m1.501\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.047]; games[0.047]; Philadelphia[0.046]; Rangers[0.045]; League[0.044]; Pittsburgh[0.043]; Toronto[0.043]; Washington[0.043]; Dallas[0.042]; Buffalo[0.042]; Anaheim[0.041]; Friday[0.041]; Friday[0.041]; Montreal[0.041]; National[0.041]; Chicago[0.041]; Ottawa[0.040]; Colorado[0.040]; home[0.040]; Louis[0.040]; Hockey[0.038]; Hockey[0.038]; Caps[0.038]; Edmonton[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pittsburgh ==> ENTITY: \u001b[32mPittsburgh Penguins\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.221:0.221[0]; log p(e|m)= -3.689:-3.689[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.047]; Hockey[0.047]; games[0.047]; team[0.047]; Anaheim[0.047]; Philadelphia[0.046]; Ottawa[0.045]; Buffalo[0.044]; League[0.044]; Edmonton[0.044]; Montreal[0.044]; Toronto[0.043]; Caps[0.042]; Rangers[0.042]; Ice[0.042]; Washington[0.042]; Louis[0.042]; Colorado[0.041]; Dallas[0.041]; National[0.041]; York[0.041]; Chicago[0.040]; home[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ottawa ==> ENTITY: \u001b[32mOttawa Senators\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.222:0.222[0]; log p(e|m)= -3.689:-3.689[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.046]; Hockey[0.046]; Hockey[0.046]; games[0.045]; Anaheim[0.044]; Edmonton[0.044]; Toronto[0.044]; Buffalo[0.043]; Caps[0.043]; Montreal[0.043]; League[0.042]; Rangers[0.042]; Pittsburgh[0.042]; Philadelphia[0.041]; Ice[0.040]; Louis[0.040]; Washington[0.039]; Friday[0.039]; Friday[0.039]; Chicago[0.039]; home[0.038]; York[0.038]; National[0.038]; Colorado[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = St Louis ==> ENTITY: \u001b[32mSt. Louis Blues\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.232:0.232[0]; log p(e|m)= -2.564:-2.564[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.046]; games[0.045]; Hockey[0.044]; Hockey[0.044]; Edmonton[0.044]; Ottawa[0.043]; Buffalo[0.043]; Montreal[0.043]; team[0.043]; Toronto[0.042]; League[0.042]; Philadelphia[0.042]; Pittsburgh[0.042]; Caps[0.041]; Colorado[0.041]; Washington[0.040]; Rangers[0.040]; Chicago[0.040]; Dallas[0.040]; Friday[0.039]; Friday[0.039]; York[0.039]; National[0.038]; Ice[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington Capitals\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.227:0.227[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; Hockey[0.045]; Hockey[0.045]; Anaheim[0.044]; team[0.044]; Ottawa[0.044]; Philadelphia[0.044]; Caps[0.043]; Edmonton[0.043]; Montreal[0.043]; Buffalo[0.043]; Toronto[0.042]; Pittsburgh[0.042]; Rangers[0.041]; Dallas[0.041]; League[0.041]; Chicago[0.040]; Colorado[0.039]; Ice[0.039]; National[0.039]; York[0.038]; Friday[0.038]; Friday[0.038]; Louis[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Buffalo ==> ENTITY: \u001b[32mBuffalo Sabres\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.218:0.218[0]; log p(e|m)= -3.540:-3.540[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; Anaheim[0.045]; Hockey[0.045]; Hockey[0.045]; team[0.044]; Edmonton[0.043]; Ottawa[0.043]; League[0.043]; National[0.042]; Montreal[0.042]; Toronto[0.042]; Caps[0.042]; Pittsburgh[0.042]; Dallas[0.041]; Philadelphia[0.041]; York[0.040]; Rangers[0.040]; Washington[0.040]; Colorado[0.040]; Ice[0.039]; Chicago[0.039]; Friday[0.039]; Friday[0.039]; home[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colorado ==> ENTITY: \u001b[32mColorado Avalanche\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.223:0.223[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; Anaheim[0.045]; Edmonton[0.044]; team[0.044]; League[0.044]; Hockey[0.043]; Hockey[0.043]; Buffalo[0.043]; Montreal[0.043]; Ottawa[0.043]; Caps[0.043]; Toronto[0.042]; Philadelphia[0.042]; Pittsburgh[0.041]; Rangers[0.040]; Dallas[0.040]; home[0.040]; Louis[0.040]; National[0.040]; Washington[0.040]; Ice[0.040]; Chicago[0.039]; Friday[0.038]; Friday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Montreal ==> ENTITY: \u001b[32mMontreal Canadiens\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.225:0.225[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.045]; Hockey[0.045]; games[0.045]; team[0.045]; Anaheim[0.044]; Ottawa[0.044]; Toronto[0.043]; Edmonton[0.043]; League[0.043]; Caps[0.043]; Philadelphia[0.042]; Louis[0.042]; Buffalo[0.041]; Pittsburgh[0.041]; Rangers[0.041]; Friday[0.040]; Friday[0.040]; Ice[0.039]; York[0.039]; Washington[0.039]; Chicago[0.039]; Dallas[0.039]; National[0.039]; Colorado[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NHL ==> ENTITY: \u001b[32mNational Hockey League\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.230:0.230[0]; log p(e|m)= -0.036:-0.036[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.046]; Hockey[0.046]; team[0.046]; games[0.045]; Anaheim[0.044]; Ottawa[0.044]; Edmonton[0.043]; Montreal[0.043]; League[0.043]; Toronto[0.043]; Caps[0.042]; Buffalo[0.042]; Philadelphia[0.042]; Rangers[0.042]; Pittsburgh[0.042]; National[0.040]; Ice[0.040]; York[0.039]; Colorado[0.039]; Washington[0.039]; Dallas[0.038]; Friday[0.038]; Friday[0.038]; Chicago[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NY RANGERS\u001b[39m ==> ENTITY: \u001b[32mNew York Rangers\u001b[39m\t\nSCORES: global= 0.318:0.318[0]; local()= 0.233:0.233[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.046]; Hockey[0.046]; games[0.046]; Anaheim[0.045]; team[0.045]; Montreal[0.044]; Ottawa[0.044]; Edmonton[0.043]; Buffalo[0.043]; Toronto[0.042]; Caps[0.042]; Philadelphia[0.042]; Pittsburgh[0.042]; League[0.042]; Ice[0.040]; Louis[0.040]; Colorado[0.039]; Chicago[0.039]; Dallas[0.039]; Friday[0.039]; Friday[0.039]; York[0.039]; Washington[0.039]; home[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Edmonton ==> ENTITY: \u001b[32mEdmonton Oilers\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.215:0.215[0]; log p(e|m)= -3.270:-3.270[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; Ottawa[0.046]; Hockey[0.045]; Hockey[0.045]; Anaheim[0.045]; team[0.044]; League[0.043]; Buffalo[0.043]; Toronto[0.043]; Montreal[0.043]; Caps[0.041]; Philadelphia[0.041]; Pittsburgh[0.041]; Rangers[0.041]; Louis[0.040]; home[0.040]; Ice[0.040]; Colorado[0.039]; Dallas[0.039]; National[0.039]; Chicago[0.039]; York[0.039]; Friday[0.038]; Friday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago Blackhawks\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.238:0.238[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.046]; team[0.045]; games[0.045]; League[0.045]; Hockey[0.044]; Hockey[0.044]; Ottawa[0.044]; Edmonton[0.043]; Montreal[0.043]; Caps[0.043]; Buffalo[0.043]; Toronto[0.042]; Philadelphia[0.041]; Rangers[0.041]; Pittsburgh[0.041]; Ice[0.040]; Louis[0.040]; Colorado[0.040]; Washington[0.040]; Dallas[0.039]; Friday[0.038]; Friday[0.038]; National[0.038]; York[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Anaheim ==> ENTITY: \u001b[32mAnaheim Ducks\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.228:0.228[0]; log p(e|m)= -2.189:-2.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.048]; Hockey[0.045]; Hockey[0.045]; Edmonton[0.044]; team[0.043]; Montreal[0.043]; Caps[0.043]; Buffalo[0.043]; League[0.043]; Ottawa[0.043]; Toronto[0.042]; Pittsburgh[0.042]; Philadelphia[0.042]; Rangers[0.041]; Colorado[0.041]; Ice[0.041]; Dallas[0.040]; Chicago[0.040]; Louis[0.039]; Friday[0.039]; Friday[0.039]; Washington[0.039]; National[0.038]; York[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = National Hockey League\u001b[39m ==> ENTITY: \u001b[32mNational Hockey League\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.222:0.222[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.047]; team[0.047]; games[0.047]; Anaheim[0.046]; Ottawa[0.045]; Montreal[0.044]; Edmonton[0.044]; Caps[0.043]; Buffalo[0.043]; Rangers[0.043]; Toronto[0.042]; Philadelphia[0.042]; Pittsburgh[0.042]; Ice[0.041]; Dallas[0.040]; Colorado[0.040]; York[0.040]; Louis[0.039]; Chicago[0.039]; Washington[0.039]; Friday[0.039]; Friday[0.039]; home[0.037]; Ot[0.034]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Philadelphia ==> ENTITY: \u001b[32mPhiladelphia Flyers\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.232:0.232[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.048]; games[0.046]; Anaheim[0.045]; Hockey[0.045]; Hockey[0.045]; Montreal[0.043]; Ottawa[0.043]; Buffalo[0.043]; Edmonton[0.043]; Toronto[0.042]; League[0.042]; Pittsburgh[0.042]; Caps[0.042]; Washington[0.041]; Rangers[0.040]; Ice[0.040]; York[0.039]; Dallas[0.039]; Louis[0.039]; Chicago[0.039]; National[0.039]; Colorado[0.039]; Friday[0.039]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Toronto ==> ENTITY: \u001b[32mToronto Maple Leafs\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.212:0.212[0]; log p(e|m)= -4.075:-4.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.047]; Ottawa[0.046]; Hockey[0.045]; Hockey[0.045]; Caps[0.045]; games[0.044]; Anaheim[0.044]; League[0.043]; Edmonton[0.043]; Montreal[0.043]; Rangers[0.043]; Buffalo[0.042]; Philadelphia[0.041]; Washington[0.041]; Pittsburgh[0.040]; Louis[0.040]; Ice[0.039]; Dallas[0.039]; National[0.038]; Friday[0.038]; Friday[0.038]; Colorado[0.038]; York[0.038]; Chicago[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dallas ==> ENTITY: \u001b[32mDallas Stars\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.226:0.226[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; Anaheim[0.045]; Hockey[0.045]; Hockey[0.045]; team[0.044]; Washington[0.044]; Edmonton[0.043]; Buffalo[0.043]; Ottawa[0.043]; Toronto[0.042]; Montreal[0.042]; League[0.042]; Caps[0.042]; Philadelphia[0.041]; Rangers[0.041]; Pittsburgh[0.041]; York[0.041]; Colorado[0.040]; Ice[0.039]; Chicago[0.039]; Friday[0.039]; Friday[0.039]; National[0.039]; home[0.038]; \t\n\r [>.............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 16s224ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 142/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1200testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMunich\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mGrand Slam Cup\u001b[39m <---> \u001b[32mMunich\u001b[39m\t\nSCORES: global= 0.256:0.249[\u001b[31m0.007\u001b[39m]; local()= 0.201:0.080[\u001b[31m0.121\u001b[39m]; log p(e|m)= -3.689:-0.119[\u001b[32m3.570\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntennis[0.048]; tournament[0.047]; Tennis[0.046]; Germany[0.044]; Slam[0.044]; Slam[0.044]; final[0.044]; Cup[0.044]; Cup[0.044]; Ivanisevic[0.042]; Kafelnikov[0.042]; Grand[0.042]; Grand[0.042]; Final[0.041]; Results[0.040]; Mark[0.039]; Friday[0.039]; Australia[0.039]; Woodforde[0.038]; results[0.038]; Goran[0.038]; beat[0.038]; beat[0.038]; million[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.239:0.239[0]; local()= 0.066:0.066[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.046]; Cup[0.046]; tournament[0.046]; tennis[0.045]; Tennis[0.044]; Woodforde[0.044]; Croatia[0.044]; final[0.044]; Mark[0.043]; Germany[0.043]; Russia[0.042]; Kafelnikov[0.041]; million[0.040]; Friday[0.040]; Final[0.040]; Slam[0.040]; Slam[0.040]; Ivanisevic[0.040]; beat[0.040]; beat[0.040]; Jim[0.039]; Goran[0.038]; Grand[0.038]; Grand[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yevgeny Kafelnikov\u001b[39m ==> ENTITY: \u001b[32mYevgeny Kafelnikov\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntennis[0.049]; beat[0.046]; beat[0.046]; tournament[0.046]; Tennis[0.045]; Ivanisevic[0.043]; Woodforde[0.043]; Mark[0.042]; Slam[0.042]; Slam[0.042]; Goran[0.042]; final[0.042]; Cup[0.041]; Cup[0.041]; Russia[0.041]; Grand[0.040]; Grand[0.040]; Quarter[0.040]; Quarter[0.040]; Croatia[0.039]; Germany[0.039]; Australia[0.038]; Munich[0.038]; Final[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Croatia ==> ENTITY: \u001b[32mCroatia\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.096:0.096[0]; log p(e|m)= -0.350:-0.350[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntennis[0.046]; Tennis[0.045]; tournament[0.045]; Cup[0.044]; Cup[0.044]; final[0.044]; Ivanisevic[0.043]; Germany[0.043]; Final[0.042]; results[0.042]; Slam[0.041]; Slam[0.041]; Australia[0.041]; Goran[0.041]; Russia[0.041]; beat[0.041]; beat[0.041]; Kafelnikov[0.041]; Results[0.040]; Woodforde[0.039]; Jim[0.039]; Mark[0.039]; Grand[0.039]; Grand[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.238:0.238[0]; local()= 0.068:0.068[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMark[0.046]; Croatia[0.046]; Cup[0.046]; Cup[0.046]; tournament[0.045]; final[0.043]; Germany[0.043]; Final[0.042]; Australia[0.042]; Tennis[0.042]; Jim[0.041]; beat[0.041]; beat[0.041]; Yevgeny[0.041]; Slam[0.041]; Slam[0.041]; tennis[0.040]; million[0.040]; Friday[0.039]; Grand[0.039]; Grand[0.039]; Goran[0.039]; Kafelnikov[0.039]; Munich[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Grand Slam Cup\u001b[39m ==> ENTITY: \u001b[32mGrand Slam Cup\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntennis[0.049]; tournament[0.049]; Tennis[0.047]; Slam[0.045]; final[0.045]; Cup[0.045]; Ivanisevic[0.044]; Kafelnikov[0.043]; Grand[0.043]; Final[0.042]; Germany[0.041]; Mark[0.040]; Friday[0.040]; Quarter[0.040]; Quarter[0.040]; Munich[0.039]; Woodforde[0.039]; Goran[0.039]; beat[0.039]; beat[0.039]; million[0.039]; Croatia[0.039]; Australia[0.038]; Results[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Grand Slam Cup\u001b[39m ==> ENTITY: \u001b[32mGrand Slam Cup\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntennis[0.049]; tournament[0.049]; Tennis[0.047]; Slam[0.045]; final[0.045]; Cup[0.045]; Ivanisevic[0.044]; Kafelnikov[0.043]; Grand[0.043]; Final[0.042]; Germany[0.041]; Mark[0.040]; Friday[0.040]; Quarter[0.040]; Quarter[0.040]; Munich[0.039]; Woodforde[0.039]; Goran[0.039]; beat[0.039]; beat[0.039]; million[0.039]; Croatia[0.039]; Australia[0.038]; Results[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jim Courier\u001b[39m ==> ENTITY: \u001b[32mJim Courier\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntennis[0.050]; Tennis[0.048]; tournament[0.047]; Kafelnikov[0.046]; Ivanisevic[0.045]; beat[0.043]; beat[0.043]; Slam[0.043]; Slam[0.043]; final[0.042]; Mark[0.042]; Goran[0.041]; Quarter[0.040]; Quarter[0.040]; Woodforde[0.040]; Friday[0.039]; Yevgeny[0.039]; Croatia[0.039]; Munich[0.039]; Cup[0.038]; Cup[0.038]; Grand[0.038]; Grand[0.038]; Germany[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mark Woodforde\u001b[39m ==> ENTITY: \u001b[32mMark Woodforde\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntennis[0.049]; Tennis[0.046]; Ivanisevic[0.044]; tournament[0.044]; Kafelnikov[0.044]; Goran[0.044]; Slam[0.043]; Slam[0.043]; final[0.042]; Cup[0.042]; Cup[0.042]; Grand[0.041]; Grand[0.041]; Quarter[0.041]; Quarter[0.041]; Australia[0.041]; beat[0.040]; beat[0.040]; Croatia[0.039]; Jim[0.039]; Munich[0.039]; Friday[0.039]; Courier[0.039]; Germany[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.241:0.241[0]; local()= 0.052:0.052[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.049]; Croatia[0.045]; final[0.044]; Cup[0.044]; Cup[0.044]; beat[0.043]; beat[0.043]; Russia[0.043]; Munich[0.043]; Final[0.042]; Australia[0.041]; results[0.041]; Goran[0.041]; Mark[0.040]; tennis[0.040]; Grand[0.040]; Grand[0.040]; Tennis[0.040]; Slam[0.039]; Slam[0.039]; Friday[0.039]; Results[0.039]; Jim[0.039]; Ivanisevic[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.043:0.043[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfinal[0.046]; tournament[0.046]; Russia[0.044]; Final[0.044]; Cup[0.043]; Cup[0.043]; Australia[0.043]; Germany[0.042]; beat[0.042]; beat[0.042]; Tennis[0.042]; Slam[0.042]; Slam[0.042]; million[0.041]; tennis[0.041]; Munich[0.041]; Croatia[0.040]; Mark[0.040]; results[0.040]; Jim[0.040]; Friday[0.040]; Grand[0.039]; Grand[0.039]; Results[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Goran Ivanisevic\u001b[39m ==> ENTITY: \u001b[32mGoran Ivanišević\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntennis[0.050]; tournament[0.048]; Tennis[0.045]; final[0.045]; Mark[0.044]; Slam[0.043]; Slam[0.043]; Kafelnikov[0.042]; Cup[0.041]; Cup[0.041]; beat[0.041]; beat[0.041]; Courier[0.041]; Woodforde[0.040]; Quarter[0.040]; Quarter[0.040]; Grand[0.040]; Grand[0.040]; Jim[0.039]; Friday[0.039]; Croatia[0.039]; Munich[0.039]; Germany[0.038]; Australia[0.038]; \t\n\r [>.............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 18s433ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 154/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1204testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Madrid ==> ENTITY: \u001b[32mMadrid\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMadrid[0.047]; Madrid[0.047]; Soccer[0.044]; games[0.044]; games[0.044]; Spanish[0.043]; Spanish[0.043]; Coruna[0.043]; Bilbao[0.042]; played[0.042]; won[0.042]; Barcelona[0.041]; Deportivo[0.041]; lost[0.041]; Valladolid[0.040]; Athletic[0.040]; Santander[0.040]; Atletico[0.040]; ahead[0.040]; Sociedad[0.040]; Real[0.039]; Real[0.039]; Real[0.039]; division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Athletic Bilbao\u001b[39m ==> ENTITY: \u001b[32mAthletic Bilbao\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.227:0.227[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.045]; Soccer[0.044]; Valencia[0.043]; Deportivo[0.043]; games[0.043]; games[0.043]; Sociedad[0.042]; Atletico[0.042]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Coruna[0.042]; Spanish[0.042]; Spanish[0.042]; Barcelona[0.041]; Espanyol[0.041]; Sporting[0.040]; division[0.040]; Valladolid[0.040]; Standings[0.040]; Standings[0.040]; Betis[0.040]; Santander[0.040]; Gijon[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hercules ==> ENTITY: \u001b[32mHércules CF\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.220:0.220[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.071]; Zaragoza[0.068]; Valencia[0.068]; Sevilla[0.066]; Espanyol[0.066]; Tenerife[0.064]; Gijon[0.063]; Rayo[0.063]; Celta[0.062]; Santander[0.062]; Extremadura[0.062]; Sporting[0.060]; Compostela[0.058]; Oviedo[0.056]; Vigo[0.056]; Racing[0.054]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oviedo ==> ENTITY: \u001b[32mReal Oviedo\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.194:0.194[0]; log p(e|m)= -2.254:-2.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.048]; Valencia[0.045]; Madrid[0.044]; Bilbao[0.043]; Atletico[0.043]; Zaragoza[0.043]; Espanyol[0.042]; Tenerife[0.042]; Sevilla[0.042]; Sociedad[0.042]; Extremadura[0.041]; Celta[0.041]; Sporting[0.041]; Rayo[0.041]; Gijon[0.041]; Betis[0.041]; Valladolid[0.041]; Compostela[0.040]; Athletic[0.040]; Real[0.040]; Real[0.040]; Santander[0.040]; Vigo[0.040]; Hercules[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Celta Vigo\u001b[39m ==> ENTITY: \u001b[32mCelta de Vigo\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.213:0.213[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.046]; Atletico[0.044]; Valencia[0.044]; Deportivo[0.044]; Sociedad[0.043]; Zaragoza[0.043]; Espanyol[0.043]; Coruna[0.043]; Bilbao[0.043]; Sevilla[0.042]; Madrid[0.042]; Madrid[0.042]; Barcelona[0.041]; Tenerife[0.041]; Betis[0.041]; Gijon[0.041]; Sporting[0.040]; Athletic[0.040]; Real[0.040]; Real[0.040]; Real[0.040]; Valladolid[0.040]; Rayo[0.039]; Santander[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Sociedad\u001b[39m ==> ENTITY: \u001b[32mReal Sociedad\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.046]; Soccer[0.044]; Deportivo[0.043]; games[0.042]; games[0.042]; Valencia[0.042]; Spanish[0.042]; Spanish[0.042]; ahead[0.042]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Atletico[0.041]; Coruna[0.041]; division[0.041]; Zaragoza[0.041]; Sevilla[0.041]; played[0.041]; Real[0.041]; Real[0.041]; Barcelona[0.041]; Sporting[0.040]; lost[0.040]; Espanyol[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Betis\u001b[39m ==> ENTITY: \u001b[32mReal Betis\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.241:0.241[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.047]; Valencia[0.044]; Deportivo[0.043]; Atletico[0.043]; Sociedad[0.043]; games[0.042]; games[0.042]; Espanyol[0.042]; Bilbao[0.042]; Soccer[0.042]; Valladolid[0.042]; division[0.041]; Coruna[0.041]; Celta[0.041]; Madrid[0.041]; Madrid[0.041]; Madrid[0.041]; Barcelona[0.040]; Rayo[0.040]; Spanish[0.040]; Spanish[0.040]; won[0.040]; Athletic[0.040]; Santander[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zaragoza ==> ENTITY: \u001b[32mReal Zaragoza\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.157:0.157[0]; log p(e|m)= -1.609:-1.609[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nValladolid[0.060]; Sevilla[0.058]; Valencia[0.058]; Celta[0.058]; Espanyol[0.057]; Sociedad[0.057]; Gijon[0.056]; Vallecano[0.055]; Santander[0.053]; Extremadura[0.053]; Compostela[0.052]; Rayo[0.052]; Vigo[0.051]; Tenerife[0.051]; Oviedo[0.050]; Hercules[0.046]; Sporting[0.046]; Real[0.045]; Racing[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tenerife ==> ENTITY: \u001b[32mCD Tenerife\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.188:0.188[0]; log p(e|m)= -1.565:-1.565[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaragoza[0.046]; Vallecano[0.045]; Santander[0.045]; Extremadura[0.044]; Gijon[0.044]; Madrid[0.043]; Sociedad[0.043]; Valladolid[0.043]; Deportivo[0.043]; Bilbao[0.042]; Valencia[0.042]; Atletico[0.042]; Compostela[0.041]; Espanyol[0.041]; Celta[0.041]; Vigo[0.041]; Betis[0.041]; Sevilla[0.041]; Oviedo[0.041]; Rayo[0.040]; Coruna[0.040]; Athletic[0.038]; Barcelona[0.038]; Sporting[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Valladolid ==> ENTITY: \u001b[32mReal Valladolid\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.202:0.202[0]; log p(e|m)= -1.650:-1.650[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.045]; Madrid[0.043]; Madrid[0.043]; Madrid[0.043]; Soccer[0.043]; Spanish[0.042]; Spanish[0.042]; Deportivo[0.042]; Bilbao[0.042]; Sociedad[0.042]; games[0.042]; games[0.042]; Zaragoza[0.042]; Atletico[0.042]; Compostela[0.042]; Espanyol[0.041]; Valencia[0.041]; Santander[0.041]; Sevilla[0.040]; Celta[0.040]; Rayo[0.040]; Coruna[0.040]; Betis[0.040]; Gijon[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spanish ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.125:0.125[0]; log p(e|m)= -1.241:-1.241[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpanish[0.051]; games[0.044]; games[0.044]; Soccer[0.044]; won[0.043]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Division[0.042]; played[0.041]; Standings[0.041]; Standings[0.041]; Barcelona[0.041]; division[0.041]; points[0.040]; goals[0.040]; Atletico[0.040]; Real[0.040]; Real[0.040]; Real[0.040]; Deportivo[0.040]; weekend[0.040]; Bilbao[0.040]; Athletic[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Madrid\u001b[39m ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.046]; Vallecano[0.045]; Madrid[0.043]; Madrid[0.043]; games[0.043]; games[0.043]; Deportivo[0.042]; Valencia[0.042]; Barcelona[0.042]; Atletico[0.042]; Coruna[0.042]; Division[0.041]; Real[0.041]; Real[0.041]; ahead[0.041]; Athletic[0.041]; played[0.040]; won[0.040]; points[0.040]; Spanish[0.040]; Spanish[0.040]; division[0.040]; Betis[0.040]; Sociedad[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Espanyol\u001b[39m ==> ENTITY: \u001b[32mRCD Espanyol\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.046]; Sociedad[0.044]; Valencia[0.044]; Zaragoza[0.044]; Madrid[0.043]; Atletico[0.043]; Deportivo[0.043]; Coruna[0.043]; Celta[0.042]; Extremadura[0.042]; Sevilla[0.042]; Valladolid[0.042]; Sporting[0.042]; Gijon[0.041]; Vigo[0.041]; Athletic[0.041]; Betis[0.041]; Bilbao[0.040]; Tenerife[0.040]; Real[0.040]; Real[0.040]; Santander[0.039]; Rayo[0.039]; Oviedo[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Racing Santander\u001b[39m ==> ENTITY: \u001b[32mRacing de Santander\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.229:0.229[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.046]; Valencia[0.043]; Deportivo[0.043]; Zaragoza[0.043]; Sociedad[0.043]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; games[0.042]; games[0.042]; Coruna[0.042]; Bilbao[0.042]; Sevilla[0.042]; Atletico[0.042]; Espanyol[0.041]; Athletic[0.041]; Betis[0.040]; Barcelona[0.040]; Spanish[0.040]; goals[0.040]; Real[0.040]; Real[0.040]; Real[0.040]; division[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Atletico Madrid\u001b[39m ==> ENTITY: \u001b[32mAtlético Madrid\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.226:0.226[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.048]; Soccer[0.046]; Valencia[0.045]; Madrid[0.045]; Madrid[0.045]; games[0.045]; games[0.045]; Coruna[0.045]; Deportivo[0.045]; Barcelona[0.044]; Sociedad[0.043]; Espanyol[0.043]; Sporting[0.042]; Spanish[0.042]; Spanish[0.042]; Bilbao[0.042]; Betis[0.042]; Athletic[0.042]; Gijon[0.042]; won[0.041]; goals[0.041]; division[0.041]; Valladolid[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Compostela ==> ENTITY: \u001b[32mSD Compostela\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.118:0.118[0]; log p(e|m)= -0.875:-0.875[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.069]; Tenerife[0.066]; Extremadura[0.065]; Zaragoza[0.065]; Valencia[0.064]; Rayo[0.064]; Santander[0.064]; Oviedo[0.064]; Valladolid[0.063]; Gijon[0.061]; Vigo[0.061]; Hercules[0.061]; Sevilla[0.060]; Espanyol[0.059]; Celta[0.058]; Sporting[0.058]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSporting\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSporting Clube de Portugal\u001b[39m <---> \u001b[32mSporting de Gijón\u001b[39m\t\nSCORES: global= 0.274:0.262[\u001b[31m0.012\u001b[39m]; local()= 0.152:0.198[\u001b[32m0.047\u001b[39m]; log p(e|m)= -0.121:-1.858[\u001b[31m1.737\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.049]; Valencia[0.047]; Atletico[0.046]; Athletic[0.046]; Sociedad[0.045]; Zaragoza[0.045]; Gijon[0.045]; Espanyol[0.044]; Bilbao[0.044]; Madrid[0.044]; Sevilla[0.044]; Tenerife[0.043]; Compostela[0.043]; Valladolid[0.043]; Celta[0.043]; Hercules[0.042]; Extremadura[0.042]; Vigo[0.042]; Real[0.042]; Oviedo[0.041]; Rayo[0.041]; Santander[0.041]; Racing[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spanish ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.131:0.131[0]; log p(e|m)= -1.241:-1.241[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpanish[0.051]; games[0.044]; games[0.044]; Soccer[0.044]; Vallecano[0.044]; won[0.043]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Division[0.042]; played[0.041]; Standings[0.041]; Standings[0.041]; Barcelona[0.041]; division[0.040]; points[0.040]; goals[0.040]; Santander[0.040]; Atletico[0.040]; Real[0.040]; Real[0.040]; Real[0.040]; Deportivo[0.039]; weekend[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rayo Vallecano\u001b[39m ==> ENTITY: \u001b[32mRayo Vallecano\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; games[0.044]; division[0.044]; Valladolid[0.043]; Valencia[0.043]; Madrid[0.043]; Madrid[0.043]; Sevilla[0.042]; Deportivo[0.042]; Zaragoza[0.042]; Spanish[0.042]; Gijon[0.041]; Tenerife[0.041]; Sociedad[0.041]; won[0.041]; Atletico[0.041]; Oviedo[0.041]; Espanyol[0.041]; goals[0.041]; Coruna[0.041]; Betis[0.041]; Athletic[0.041]; ahead[0.040]; weekend[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Extremadura ==> ENTITY: \u001b[32mCF Extremadura\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.167:0.167[0]; log p(e|m)= -1.211:-1.211[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.078]; Tenerife[0.074]; Rayo[0.074]; Valencia[0.073]; Zaragoza[0.073]; Espanyol[0.072]; Hercules[0.071]; Oviedo[0.071]; Compostela[0.071]; Celta[0.070]; Gijon[0.069]; Sevilla[0.069]; Sporting[0.068]; Vigo[0.068]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barcelona ==> ENTITY: \u001b[32mFC Barcelona\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.188:0.188[0]; log p(e|m)= -1.709:-1.709[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.045]; Vallecano[0.044]; Valencia[0.044]; Madrid[0.043]; Madrid[0.043]; Madrid[0.043]; games[0.043]; games[0.043]; Bilbao[0.043]; won[0.042]; Coruna[0.041]; ahead[0.041]; Spanish[0.041]; Spanish[0.041]; Deportivo[0.041]; Atletico[0.041]; Racing[0.041]; Rayo[0.041]; played[0.040]; Division[0.040]; division[0.040]; Athletic[0.040]; points[0.040]; Santander[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sevilla ==> ENTITY: \u001b[32mSevilla FC\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.559:-0.559[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.062]; Compostela[0.058]; Zaragoza[0.058]; Valladolid[0.058]; Valencia[0.058]; Gijon[0.058]; Espanyol[0.058]; Sociedad[0.057]; Extremadura[0.056]; Celta[0.055]; Santander[0.054]; Oviedo[0.054]; Tenerife[0.054]; Hercules[0.053]; Sporting[0.053]; Rayo[0.053]; Vigo[0.052]; Racing[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Valencia ==> ENTITY: \u001b[32mValencia CF\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.209:0.209[0]; log p(e|m)= -1.604:-1.604[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaragoza[0.051]; Vallecano[0.049]; Madrid[0.048]; Madrid[0.048]; Bilbao[0.047]; Gijon[0.046]; Santander[0.046]; Atletico[0.045]; Valladolid[0.045]; games[0.045]; games[0.045]; Oviedo[0.045]; Extremadura[0.045]; Deportivo[0.044]; Coruna[0.044]; Barcelona[0.044]; Sevilla[0.044]; Compostela[0.044]; Betis[0.044]; Tenerife[0.044]; Sociedad[0.043]; Espanyol[0.043]; \t\n\r [>.............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 17s802ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 177/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1265testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hainan\u001b[39m ==> ENTITY: \u001b[32mHainan\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHainan[0.046]; China[0.046]; China[0.046]; Beijing[0.045]; province[0.044]; Wenchang[0.043]; Wenchang[0.043]; southern[0.043]; Xinhua[0.042]; Xinhua[0.042]; Xinhua[0.042]; yuan[0.042]; yuan[0.042]; ships[0.041]; ships[0.041]; vessels[0.040]; vessels[0.040]; million[0.039]; million[0.039]; news[0.039]; Friday[0.039]; foreign[0.039]; foreign[0.039]; foreign[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.142:0.142[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.053]; Beijing[0.050]; southern[0.046]; Hainan[0.045]; Hainan[0.045]; Xinhua[0.044]; Xinhua[0.044]; Xinhua[0.044]; province[0.044]; yuan[0.043]; yuan[0.043]; foreign[0.043]; foreign[0.043]; foreign[0.043]; berth[0.042]; built[0.042]; opened[0.041]; city[0.041]; ships[0.041]; ships[0.041]; Wenchang[0.041]; Wenchang[0.041]; cabinet[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wenchang\u001b[39m ==> ENTITY: \u001b[32mWenchang\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.046]; China[0.046]; Wenchang[0.045]; Hainan[0.045]; Hainan[0.045]; Beijing[0.044]; Xinhua[0.042]; Xinhua[0.042]; Xinhua[0.042]; yuan[0.042]; yuan[0.042]; southern[0.041]; province[0.041]; Friday[0.041]; State[0.040]; city[0.040]; Council[0.040]; cabinet[0.040]; million[0.040]; million[0.040]; port[0.040]; port[0.040]; port[0.040]; port[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Xinhua\u001b[39m ==> ENTITY: \u001b[32mXinhua News Agency\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nXinhua[0.048]; Xinhua[0.048]; city[0.046]; Beijing[0.045]; province[0.045]; China[0.045]; China[0.045]; Hainan[0.043]; Hainan[0.043]; southern[0.043]; yuan[0.042]; yuan[0.042]; news[0.042]; Wenchang[0.040]; Wenchang[0.040]; agency[0.040]; Council[0.039]; said[0.038]; said[0.038]; foreign[0.037]; foreign[0.037]; foreign[0.037]; opened[0.037]; Friday[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wenchang\u001b[39m ==> ENTITY: \u001b[32mWenchang\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.046]; China[0.046]; Wenchang[0.045]; Hainan[0.045]; Hainan[0.045]; Beijing[0.044]; Xinhua[0.042]; Xinhua[0.042]; Xinhua[0.042]; yuan[0.042]; yuan[0.042]; southern[0.041]; province[0.041]; Friday[0.041]; State[0.040]; city[0.040]; Council[0.040]; cabinet[0.040]; million[0.040]; million[0.040]; port[0.040]; port[0.040]; port[0.040]; port[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Beijing ==> ENTITY: \u001b[32mBeijing\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.042:-0.042[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.051]; China[0.051]; southern[0.045]; Hainan[0.045]; Hainan[0.045]; Friday[0.044]; yuan[0.044]; yuan[0.044]; opened[0.043]; Xinhua[0.043]; Xinhua[0.043]; Xinhua[0.043]; million[0.043]; million[0.043]; built[0.042]; city[0.042]; province[0.042]; Wenchang[0.042]; Wenchang[0.042]; gave[0.041]; foreign[0.041]; foreign[0.041]; foreign[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Xinhua\u001b[39m ==> ENTITY: \u001b[32mXinhua News Agency\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nXinhua[0.048]; Xinhua[0.048]; city[0.046]; Beijing[0.045]; province[0.045]; China[0.045]; China[0.045]; Hainan[0.043]; Hainan[0.043]; southern[0.043]; yuan[0.042]; yuan[0.042]; news[0.042]; Wenchang[0.040]; Wenchang[0.040]; agency[0.040]; Council[0.039]; said[0.038]; said[0.038]; foreign[0.037]; foreign[0.037]; foreign[0.037]; opened[0.037]; Friday[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hainan\u001b[39m ==> ENTITY: \u001b[32mHainan\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHainan[0.046]; China[0.046]; China[0.046]; Beijing[0.045]; province[0.044]; Wenchang[0.043]; Wenchang[0.043]; southern[0.043]; Xinhua[0.042]; Xinhua[0.042]; Xinhua[0.042]; yuan[0.042]; yuan[0.042]; ships[0.041]; ships[0.041]; vessels[0.040]; vessels[0.040]; million[0.039]; million[0.039]; news[0.039]; Friday[0.039]; foreign[0.039]; foreign[0.039]; foreign[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Xinhua\u001b[39m ==> ENTITY: \u001b[32mXinhua News Agency\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nXinhua[0.048]; Xinhua[0.048]; city[0.046]; Beijing[0.045]; province[0.045]; China[0.045]; China[0.045]; Hainan[0.043]; Hainan[0.043]; southern[0.043]; yuan[0.042]; yuan[0.042]; news[0.042]; Wenchang[0.040]; Wenchang[0.040]; agency[0.040]; Council[0.039]; said[0.038]; said[0.038]; foreign[0.037]; foreign[0.037]; foreign[0.037]; opened[0.037]; Friday[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.142:0.142[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.053]; Beijing[0.050]; southern[0.046]; Hainan[0.045]; Hainan[0.045]; Xinhua[0.044]; Xinhua[0.044]; Xinhua[0.044]; province[0.044]; yuan[0.043]; yuan[0.043]; foreign[0.043]; foreign[0.043]; foreign[0.043]; berth[0.042]; built[0.042]; opened[0.041]; city[0.041]; ships[0.041]; ships[0.041]; Wenchang[0.041]; Wenchang[0.041]; cabinet[0.041]; \t\n\r [>.............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 18s532ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 187/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1224testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.039:0.039[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsaid[0.046]; market[0.046]; day[0.046]; Thursday[0.046]; Volume[0.045]; Friday[0.045]; voted[0.045]; discussed[0.044]; lower[0.044]; lower[0.044]; long[0.044]; Spain[0.043]; products[0.043]; products[0.043]; line[0.042]; Trades[0.042]; quotes[0.042]; high[0.042]; workers[0.042]; gains[0.042]; inquiries[0.042]; Lebanon[0.042]; moves[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.045]; France[0.043]; lost[0.043]; lost[0.043]; American[0.043]; Mediterranean[0.043]; Lebanon[0.043]; week[0.042]; following[0.042]; transatlantic[0.042]; suffered[0.041]; earlier[0.041]; January[0.041]; Syria[0.041]; day[0.041]; said[0.041]; nervous[0.041]; remained[0.041]; long[0.041]; lower[0.040]; jeopardy[0.040]; chance[0.040]; discussed[0.040]; erased[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italian ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.090:0.090[0]; log p(e|m)= -0.759:-0.759[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.049]; Europe[0.048]; expected[0.044]; said[0.042]; home[0.042]; work[0.042]; week[0.041]; week[0.041]; week[0.041]; mid[0.041]; lower[0.041]; units[0.041]; prime[0.041]; transatlantic[0.041]; workers[0.041]; tight[0.041]; north[0.041]; earlier[0.041]; open[0.040]; open[0.040]; fell[0.040]; low[0.040]; continued[0.040]; voted[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rotterdam ==> ENTITY: \u001b[32mRotterdam\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.145:-0.145[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurope[0.045]; units[0.044]; home[0.043]; ground[0.043]; following[0.043]; north[0.043]; week[0.042]; week[0.042]; week[0.042]; earlier[0.042]; mid[0.042]; pressure[0.042]; January[0.041]; lower[0.041]; transatlantic[0.041]; chance[0.041]; heading[0.041]; low[0.041]; sharply[0.040]; cracked[0.040]; high[0.040]; voted[0.040]; said[0.040]; work[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = India ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.065:0.065[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nday[0.044]; second[0.044]; market[0.043]; market[0.043]; American[0.043]; January[0.042]; remains[0.042]; week[0.042]; following[0.042]; pricing[0.042]; France[0.042]; expected[0.042]; percent[0.042]; Thursday[0.041]; delivery[0.041]; north[0.041]; basis[0.041]; speed[0.041]; Spain[0.041]; Friday[0.041]; chance[0.040]; Lebanon[0.040]; home[0.040]; sharply[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanon ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.409:-0.409[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.054]; second[0.046]; Mediterranean[0.045]; France[0.045]; American[0.045]; India[0.045]; Spain[0.044]; week[0.044]; earlier[0.043]; Thursday[0.043]; day[0.043]; expected[0.043]; January[0.043]; sharply[0.042]; discussed[0.042]; chance[0.042]; north[0.042]; High[0.042]; high[0.042]; high[0.042]; long[0.042]; line[0.042]; suffered[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mElf\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mTotal S.A.\u001b[39m <---> \u001b[32mElf Aquitaine\u001b[39m\t\nSCORES: global= 0.261:0.258[\u001b[31m0.002\u001b[39m]; local()= 0.128:0.120[\u001b[31m0.008\u001b[39m]; log p(e|m)= -2.976:-1.766[\u001b[32m1.210\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nElf[0.045]; oil[0.045]; refinery[0.044]; Europe[0.043]; lost[0.043]; north[0.042]; prime[0.042]; American[0.042]; fuel[0.042]; fuel[0.042]; week[0.042]; week[0.042]; week[0.042]; mid[0.042]; said[0.041]; open[0.041]; open[0.041]; threatening[0.041]; continued[0.040]; work[0.040]; home[0.040]; workers[0.040]; following[0.040]; jeopardy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = American ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.051:0.051[0]; log p(e|m)= -0.130:-0.130[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmid[0.043]; workers[0.043]; transatlantic[0.043]; January[0.042]; Europe[0.042]; low[0.042]; low[0.042]; Lebanon[0.042]; units[0.042]; fuel[0.042]; fuel[0.042]; fuel[0.042]; second[0.042]; week[0.042]; week[0.042]; Italian[0.041]; pricing[0.041]; sharply[0.041]; work[0.041]; high[0.041]; India[0.041]; earlier[0.041]; market[0.041]; lower[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mElf\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mTotal S.A.\u001b[39m <---> \u001b[32mElf Aquitaine\u001b[39m\t\nSCORES: global= 0.261:0.256[\u001b[31m0.005\u001b[39m]; local()= 0.114:0.077[\u001b[31m0.037\u001b[39m]; log p(e|m)= -2.976:-1.766[\u001b[32m1.210\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nElf[0.045]; oil[0.044]; oil[0.044]; oil[0.044]; oil[0.044]; refinery[0.044]; Gas[0.043]; Thursday[0.042]; market[0.042]; workers[0.042]; Friday[0.041]; screen[0.041]; screen[0.041]; erased[0.041]; products[0.041]; products[0.041]; nervous[0.041]; said[0.040]; Lebanon[0.040]; Syria[0.040]; gains[0.040]; day[0.040]; Mediterranean[0.040]; long[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Genoa ==> ENTITY: \u001b[32mGenoa\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.092:0.092[0]; log p(e|m)= -0.402:-0.402[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMediterranean[0.047]; Spain[0.044]; long[0.044]; second[0.044]; January[0.043]; line[0.043]; France[0.043]; lost[0.042]; lost[0.042]; high[0.041]; high[0.041]; lower[0.041]; lower[0.041]; steady[0.041]; remained[0.041]; remains[0.041]; suffered[0.040]; low[0.040]; Lebanon[0.040]; speed[0.040]; diesel[0.040]; diesel[0.040]; day[0.040]; Thursday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.056:0.056[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nweek[0.045]; week[0.045]; week[0.045]; American[0.045]; units[0.043]; following[0.043]; Europe[0.043]; earlier[0.042]; expected[0.042]; fell[0.042]; mid[0.041]; cif[0.041]; pegged[0.041]; continued[0.041]; material[0.041]; material[0.041]; Low[0.040]; prime[0.040]; low[0.040]; lower[0.040]; fuel[0.040]; fuel[0.040]; home[0.039]; oil[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.S\u001b[39m ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.024:0.024[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.044]; American[0.044]; fuel[0.043]; fuel[0.043]; fuel[0.043]; week[0.043]; week[0.043]; week[0.043]; diesel[0.042]; diesel[0.042]; Europe[0.042]; pricing[0.041]; oil[0.041]; oil[0.041]; January[0.041]; market[0.041]; prices[0.041]; prices[0.041]; prices[0.041]; low[0.041]; said[0.041]; earlier[0.041]; Fuel[0.041]; lower[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.096:0.096[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.046]; second[0.045]; American[0.044]; India[0.044]; week[0.043]; Lebanon[0.042]; January[0.042]; following[0.042]; said[0.041]; remains[0.041]; low[0.041]; expected[0.041]; suffered[0.041]; north[0.041]; Mediterranean[0.041]; heading[0.041]; purchase[0.041]; remained[0.041]; Thursday[0.041]; day[0.041]; percent[0.040]; chance[0.040]; earlier[0.040]; transatlantic[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mElf\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mTotal S.A.\u001b[39m <---> \u001b[32mElf Aquitaine\u001b[39m\t\nSCORES: global= 0.261:0.258[\u001b[31m0.003\u001b[39m]; local()= 0.130:0.117[\u001b[31m0.013\u001b[39m]; log p(e|m)= -2.976:-1.766[\u001b[32m1.210\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nElf[0.046]; oil[0.045]; refinery[0.045]; Europe[0.043]; prime[0.042]; American[0.042]; fuel[0.042]; fuel[0.042]; week[0.042]; week[0.042]; week[0.042]; mid[0.042]; said[0.041]; open[0.041]; open[0.041]; continued[0.041]; work[0.041]; home[0.040]; workers[0.040]; following[0.040]; jeopardy[0.040]; supplies[0.040]; material[0.040]; expected[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mElf\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mTotal S.A.\u001b[39m <---> \u001b[32mElf Aquitaine\u001b[39m\t\nSCORES: global= 0.262:0.257[\u001b[31m0.005\u001b[39m]; local()= 0.131:0.093[\u001b[31m0.038\u001b[39m]; log p(e|m)= -2.976:-1.766[\u001b[32m1.210\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nElf[0.044]; oil[0.044]; oil[0.044]; oil[0.044]; oil[0.044]; refinery[0.043]; India[0.043]; France[0.043]; Gas[0.043]; diesel[0.042]; Thursday[0.041]; market[0.041]; workers[0.041]; Friday[0.041]; Spain[0.041]; screen[0.041]; screen[0.041]; erased[0.040]; products[0.040]; products[0.040]; nervous[0.040]; said[0.040]; Lebanon[0.040]; second[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.067:0.067[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.044]; week[0.043]; week[0.043]; week[0.043]; earlier[0.043]; units[0.043]; ground[0.043]; second[0.042]; mid[0.042]; Europe[0.042]; expected[0.042]; following[0.041]; January[0.041]; cif[0.041]; high[0.041]; delivery[0.041]; pegged[0.041]; fell[0.041]; speed[0.040]; prime[0.040]; material[0.040]; material[0.040]; Low[0.040]; sharply[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syria ==> ENTITY: \u001b[32mSyria\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.152:0.152[0]; log p(e|m)= -0.261:-0.261[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.057]; Mediterranean[0.048]; France[0.047]; week[0.047]; north[0.046]; Spain[0.046]; Thursday[0.046]; Friday[0.046]; January[0.045]; India[0.045]; said[0.045]; day[0.045]; American[0.045]; strike[0.045]; strike[0.045]; expected[0.044]; sharply[0.044]; earlier[0.043]; discussed[0.043]; ground[0.043]; second[0.043]; long[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Europe ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsharply[0.045]; market[0.044]; transatlantic[0.044]; week[0.044]; week[0.044]; week[0.044]; American[0.043]; mid[0.042]; following[0.042]; January[0.042]; continued[0.042]; expected[0.041]; supplies[0.041]; weaker[0.041]; workers[0.041]; fell[0.040]; talked[0.040]; earlier[0.040]; Italian[0.040]; prices[0.040]; prices[0.040]; prices[0.040]; low[0.040]; said[0.040]; \t\n\r [=>............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 18s609ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 205/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1342testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tokyo ==> ENTITY: \u001b[32mTokyo\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.093:-0.093[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.047]; company[0.043]; company[0.043]; company[0.043]; company[0.043]; Friday[0.042]; business[0.042]; business[0.042]; business[0.042]; business[0.042]; firms[0.042]; following[0.042]; Corp[0.041]; international[0.041]; international[0.041]; international[0.041]; Nippon[0.041]; told[0.040]; start[0.040]; start[0.040]; government[0.040]; government[0.040]; official[0.040]; official[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.129:0.129[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnews[0.046]; companies[0.043]; conference[0.043]; business[0.042]; business[0.042]; business[0.042]; business[0.042]; company[0.042]; company[0.042]; company[0.042]; company[0.042]; company[0.042]; told[0.042]; told[0.042]; international[0.041]; international[0.041]; international[0.041]; firms[0.041]; firms[0.041]; firms[0.041]; operate[0.041]; latest[0.040]; local[0.040]; local[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.056]; government[0.049]; government[0.049]; government[0.049]; international[0.048]; international[0.048]; official[0.045]; official[0.045]; official[0.045]; latest[0.045]; specify[0.045]; officials[0.044]; Hisao[0.044]; plans[0.044]; local[0.044]; local[0.044]; decision[0.043]; Minister[0.043]; timing[0.043]; added[0.043]; allow[0.043]; planned[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.056]; government[0.049]; government[0.049]; government[0.049]; international[0.048]; international[0.048]; official[0.045]; official[0.045]; official[0.045]; latest[0.045]; specify[0.045]; officials[0.044]; Hisao[0.044]; plans[0.044]; local[0.044]; local[0.044]; decision[0.043]; Minister[0.043]; timing[0.043]; added[0.043]; allow[0.043]; planned[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.094:0.094[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTokyo[0.046]; government[0.046]; government[0.046]; international[0.045]; international[0.045]; international[0.045]; start[0.044]; start[0.044]; company[0.044]; company[0.044]; company[0.044]; company[0.044]; business[0.043]; business[0.043]; business[0.043]; official[0.042]; official[0.042]; latest[0.042]; possible[0.042]; possible[0.042]; hopes[0.042]; hopes[0.042]; Nippon[0.041]; \t\n\r [=>............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 19s265ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 210/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1257testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Budapest ==> ENTITY: \u001b[32mBudapest\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.030:-0.030[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBudapest[0.046]; Hungary[0.045]; Hungarian[0.044]; December[0.044]; December[0.044]; Thursday[0.042]; Friday[0.042]; security[0.042]; overnight[0.041]; overnight[0.041]; overnight[0.041]; social[0.041]; Bank[0.041]; bank[0.041]; dropped[0.041]; percent[0.041]; percent[0.041]; closed[0.040]; closed[0.040]; opened[0.040]; moved[0.040]; Sandor[0.040]; contribution[0.040]; deadline[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hungarian ==> ENTITY: \u001b[32mHungary\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.098:0.098[0]; log p(e|m)= -1.033:-1.033[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHungary[0.052]; Budapest[0.045]; Budapest[0.045]; percent[0.045]; percent[0.045]; opened[0.044]; tightened[0.044]; large[0.044]; said[0.044]; said[0.044]; social[0.044]; Bank[0.043]; dropped[0.043]; contribution[0.043]; preparing[0.043]; money[0.042]; money[0.042]; taken[0.042]; borrowing[0.042]; Sandor[0.042]; substantial[0.042]; Friday[0.042]; moved[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Budapest ==> ENTITY: \u001b[32mBudapest\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.111:0.111[0]; log p(e|m)= -0.030:-0.030[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBudapest[0.046]; Budapest[0.046]; Hungary[0.045]; Hungarian[0.043]; December[0.043]; December[0.043]; Thursday[0.042]; Friday[0.042]; security[0.041]; security[0.041]; Peto[0.041]; overnight[0.041]; overnight[0.041]; overnight[0.041]; social[0.041]; social[0.041]; Bank[0.041]; bank[0.040]; dropped[0.040]; week[0.040]; percent[0.040]; percent[0.040]; closed[0.040]; closed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Budapest ==> ENTITY: \u001b[32mBudapest\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.092:0.092[0]; log p(e|m)= -0.030:-0.030[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBudapest[0.051]; December[0.048]; Thursday[0.047]; security[0.046]; Peto[0.046]; overnight[0.046]; overnight[0.046]; social[0.046]; Bank[0.045]; bank[0.045]; dropped[0.045]; week[0.045]; percent[0.045]; percent[0.045]; closed[0.045]; opened[0.045]; moved[0.045]; Sandor[0.044]; Sandor[0.044]; contribution[0.044]; deadline[0.044]; payments[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hungary ==> ENTITY: \u001b[32mHungary\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.082:0.082[0]; log p(e|m)= -0.326:-0.326[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHungarian[0.046]; security[0.045]; Budapest[0.044]; Budapest[0.044]; contribution[0.043]; liquidity[0.042]; December[0.042]; December[0.042]; closed[0.042]; closed[0.042]; market[0.042]; market[0.042]; tightened[0.041]; dropped[0.041]; Bank[0.041]; Friday[0.040]; dealers[0.040]; finished[0.040]; overnight[0.040]; overnight[0.040]; tax[0.040]; tax[0.040]; said[0.040]; said[0.040]; \t\n\r [=>............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 19s898ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 215/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1164testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCuttitta\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMassimo Cuttitta\u001b[39m <---> \u001b[32mMarcello Cuttitta\u001b[39m\t\nSCORES: global= 0.284:0.281[\u001b[31m0.003\u001b[39m]; local()= 0.213:0.161[\u001b[31m0.052\u001b[39m]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.049]; Italy[0.048]; Italy[0.048]; England[0.044]; England[0.044]; Murrayfield[0.043]; Twickenham[0.043]; Scotland[0.042]; team[0.041]; squad[0.041]; Cuttitta[0.041]; Cuttitta[0.041]; Marcello[0.041]; Corrado[0.041]; play[0.040]; month[0.040]; week[0.040]; retiring[0.040]; Stefano[0.040]; wing[0.040]; recalled[0.039]; recalled[0.039]; game[0.039]; following[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Francesco Mazzariol\u001b[39m ==> ENTITY: \u001b[32mFrancesco Mazzariol\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.221:0.221[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.047]; Troncon[0.046]; team[0.045]; player[0.044]; Cup[0.043]; Cup[0.043]; Mazzucato[0.043]; Gianluca[0.042]; Paolo[0.042]; comeback[0.042]; Alessandro[0.042]; Alessandro[0.042]; England[0.041]; Massimo[0.040]; Massimo[0.040]; national[0.040]; World[0.040]; World[0.040]; Carlo[0.040]; Carlo[0.040]; Orlandi[0.040]; Squad[0.039]; Marcello[0.039]; time[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = 1995 World Cup ==> ENTITY: \u001b[32m1995 Rugby World Cup\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.191:0.191[0]; log p(e|m)= -0.014:-0.014[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; England[0.047]; England[0.047]; squad[0.045]; team[0.045]; team[0.045]; Scotland[0.045]; player[0.045]; play[0.044]; game[0.044]; national[0.044]; Italy[0.044]; Cup[0.043]; time[0.043]; World[0.042]; Twickenham[0.042]; week[0.042]; Murrayfield[0.042]; comeback[0.041]; row[0.041]; Saturday[0.041]; selection[0.041]; pool[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCuttitta\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMassimo Cuttitta\u001b[39m <---> \u001b[32mMarcello Cuttitta\u001b[39m\t\nSCORES: global= 0.284:0.281[\u001b[31m0.003\u001b[39m]; local()= 0.237:0.171[\u001b[31m0.066\u001b[39m]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.048]; Italy[0.046]; Italy[0.046]; Italy[0.046]; England[0.042]; England[0.042]; England[0.042]; Murrayfield[0.042]; Twickenham[0.042]; Cup[0.041]; Scotland[0.041]; player[0.041]; team[0.040]; team[0.040]; squad[0.040]; Cuttitta[0.040]; Cuttitta[0.040]; Cuttitta[0.040]; Marcello[0.040]; Corrado[0.040]; took[0.040]; play[0.039]; month[0.039]; national[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marcello Cuttitta\u001b[39m ==> ENTITY: \u001b[32mMarcello Cuttitta\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.048]; Italy[0.045]; Italy[0.045]; Cuttitta[0.043]; Cuttitta[0.043]; Cuttitta[0.043]; team[0.042]; team[0.042]; squad[0.042]; England[0.042]; England[0.042]; national[0.042]; Murrayfield[0.042]; Twickenham[0.041]; retiring[0.041]; wing[0.041]; retirement[0.040]; recalled[0.040]; recalled[0.040]; month[0.040]; Scotland[0.040]; play[0.039]; Corrado[0.039]; week[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marcello Cuttitta\u001b[39m ==> ENTITY: \u001b[32mMarcello Cuttitta\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.047]; player[0.045]; Troncon[0.045]; Cuttitta[0.045]; Cuttitta[0.045]; Gianluca[0.044]; team[0.044]; England[0.044]; England[0.044]; national[0.044]; Cup[0.044]; Cup[0.044]; Francesco[0.043]; Massimo[0.043]; Massimo[0.043]; Mazzucato[0.043]; Alessandro[0.042]; Alessandro[0.042]; ago[0.042]; World[0.042]; World[0.042]; retirement[0.042]; Paolo[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nicola Mazzucato\u001b[39m ==> ENTITY: \u001b[32mNicola Mazzucato\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTroncon[0.044]; Cup[0.043]; player[0.043]; Alessandro[0.043]; Alessandro[0.043]; Paolo[0.043]; Orlandi[0.043]; Andrea[0.042]; Andrea[0.042]; Francesco[0.042]; Carlo[0.042]; Carlo[0.042]; Massimo[0.042]; Massimo[0.042]; Gianluca[0.041]; Marcello[0.041]; Walter[0.041]; Vaccari[0.041]; World[0.041]; Diego[0.040]; Cuttitta[0.040]; Cuttitta[0.040]; Orazio[0.040]; ended[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland national rugby union team\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.192:0.192[0]; log p(e|m)= -3.772:-3.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.048]; England[0.048]; team[0.045]; team[0.045]; Twickenham[0.044]; squad[0.044]; player[0.042]; game[0.042]; Cup[0.042]; Cup[0.042]; month[0.041]; play[0.041]; week[0.041]; comeback[0.040]; national[0.040]; Troncon[0.040]; Italy[0.040]; George[0.040]; time[0.040]; row[0.039]; Saturday[0.039]; months[0.039]; dropped[0.039]; dropped[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Twickenham ==> ENTITY: \u001b[32mTwickenham Stadium\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.192:0.192[0]; log p(e|m)= -0.594:-0.594[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; England[0.047]; England[0.047]; Rugby[0.047]; team[0.044]; team[0.044]; Scotland[0.043]; game[0.042]; squad[0.042]; player[0.041]; week[0.041]; Cup[0.041]; Cup[0.041]; Murrayfield[0.040]; play[0.040]; month[0.040]; Saturday[0.040]; row[0.039]; comeback[0.039]; Italy[0.039]; Italy[0.039]; Italy[0.039]; national[0.039]; months[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scotland ==> ENTITY: \u001b[32mScotland national rugby union team\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.188:0.188[0]; log p(e|m)= -3.381:-3.381[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.047]; England[0.047]; England[0.047]; Murrayfield[0.044]; team[0.044]; team[0.044]; squad[0.043]; Twickenham[0.043]; game[0.042]; Italy[0.041]; Italy[0.041]; Cup[0.041]; play[0.041]; national[0.041]; week[0.041]; George[0.040]; Saturday[0.040]; following[0.040]; recalled[0.040]; recalled[0.040]; month[0.039]; World[0.039]; Union[0.039]; selection[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alessandro Troncon\u001b[39m ==> ENTITY: \u001b[32mAlessandro Troncon\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayer[0.046]; Italy[0.045]; England[0.044]; pool[0.044]; team[0.043]; Alessandro[0.043]; time[0.043]; Gianluca[0.042]; Javier[0.042]; World[0.042]; World[0.042]; Diego[0.041]; national[0.041]; comeback[0.041]; Paolo[0.041]; return[0.040]; Walter[0.040]; Francesco[0.040]; Arancio[0.040]; Cup[0.040]; Cup[0.040]; Carlo[0.040]; Carlo[0.040]; Castellani[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mItaly\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mItaly\u001b[39m <---> \u001b[32mItaly national rugby union team\u001b[39m\t\nSCORES: global= 0.245:0.237[\u001b[31m0.008\u001b[39m]; local()= 0.138:0.180[\u001b[32m0.042\u001b[39m]; log p(e|m)= -0.206:-4.135[\u001b[31m3.929\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.052]; Rugby[0.046]; England[0.045]; England[0.045]; team[0.045]; squad[0.044]; Scotland[0.044]; Rome[0.043]; Murrayfield[0.043]; game[0.042]; Twickenham[0.041]; play[0.040]; week[0.040]; following[0.040]; month[0.040]; announced[0.040]; Stefano[0.039]; Saturday[0.039]; Corrado[0.039]; friendly[0.039]; Friday[0.039]; row[0.039]; selection[0.039]; named[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ivan Francescato\u001b[39m ==> ENTITY: \u001b[32mIvan Francescato\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.046]; game[0.044]; national[0.043]; Troncon[0.042]; team[0.042]; recalled[0.042]; player[0.042]; comeback[0.042]; Alessandro[0.042]; Alessandro[0.042]; Francesco[0.042]; retirement[0.041]; England[0.041]; England[0.041]; Massimo[0.041]; Massimo[0.041]; Gianluca[0.041]; World[0.041]; World[0.041]; Paolo[0.041]; Diego[0.040]; Marcello[0.040]; Carlo[0.040]; Carlo[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCuttitta\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMassimo Cuttitta\u001b[39m <---> \u001b[32mMarcello Cuttitta\u001b[39m\t\nSCORES: global= 0.284:0.281[\u001b[31m0.003\u001b[39m]; local()= 0.237:0.170[\u001b[31m0.067\u001b[39m]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.047]; Italy[0.047]; Troncon[0.043]; England[0.043]; England[0.043]; England[0.043]; Murrayfield[0.042]; Francesco[0.042]; Twickenham[0.042]; Cup[0.041]; Cup[0.041]; Scotland[0.041]; player[0.041]; team[0.041]; team[0.041]; squad[0.041]; Cuttitta[0.041]; Cuttitta[0.041]; Cuttitta[0.041]; Alessandro[0.041]; Paolo[0.040]; Marcello[0.040]; Marcello[0.040]; Corrado[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Murrayfield\u001b[39m ==> ENTITY: \u001b[32mMurrayfield Stadium\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.193:0.193[0]; log p(e|m)= -0.007:-0.007[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.047]; Rugby[0.046]; team[0.046]; team[0.046]; England[0.044]; England[0.044]; Twickenham[0.043]; Cup[0.043]; game[0.042]; play[0.041]; squad[0.041]; week[0.041]; Saturday[0.040]; row[0.040]; Italy[0.040]; Italy[0.040]; national[0.040]; recalled[0.040]; recalled[0.040]; month[0.040]; dropped[0.040]; World[0.039]; following[0.039]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alessandro Moscardi\u001b[39m ==> ENTITY: \u001b[32mAlessandro Moscardi\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTroncon[0.048]; Mazzucato[0.045]; Alessandro[0.045]; Cup[0.044]; Massimo[0.043]; Massimo[0.043]; Paolo[0.043]; Gianluca[0.043]; Francesco[0.042]; Vaccari[0.041]; Marcello[0.041]; comeback[0.041]; Cuttitta[0.041]; Cuttitta[0.041]; Squad[0.040]; Carlo[0.040]; Carlo[0.040]; Guidi[0.040]; Franco[0.040]; Nicola[0.040]; Orazio[0.040]; Diego[0.040]; Arancio[0.040]; Giovanelli[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWorld Cup\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFIFA World Cup\u001b[39m <---> \u001b[32mRugby World Cup\u001b[39m\t\nSCORES: global= 0.242:0.232[\u001b[31m0.010\u001b[39m]; local()= 0.159:0.161[\u001b[32m0.002\u001b[39m]; log p(e|m)= -1.002:-3.058[\u001b[31m2.055\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.047]; squad[0.045]; England[0.045]; England[0.045]; England[0.045]; national[0.044]; player[0.043]; team[0.043]; team[0.043]; Cup[0.043]; World[0.043]; game[0.041]; week[0.041]; Twickenham[0.040]; month[0.040]; pool[0.040]; time[0.040]; months[0.039]; Leandro[0.039]; Diego[0.039]; return[0.039]; comeback[0.039]; Squad[0.039]; row[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Carlo Checchinato\u001b[39m ==> ENTITY: \u001b[32mCarlo Checchinato\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.051]; Paolo[0.049]; Troncon[0.048]; player[0.047]; Carlo[0.047]; England[0.047]; Nicola[0.046]; Alessandro[0.046]; Alessandro[0.046]; Gianluca[0.046]; Mazzucato[0.046]; Francesco[0.046]; Massimo[0.046]; Massimo[0.046]; time[0.045]; Marcello[0.044]; Orazio[0.043]; Andrea[0.043]; Andrea[0.043]; months[0.043]; Javier[0.043]; Arancio[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mItaly\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mItaly\u001b[39m <---> \u001b[32mItaly national rugby union team\u001b[39m\t\nSCORES: global= 0.245:0.236[\u001b[31m0.009\u001b[39m]; local()= 0.146:0.171[\u001b[32m0.024\u001b[39m]; log p(e|m)= -0.206:-4.135[\u001b[31m3.929\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.045]; England[0.045]; England[0.045]; team[0.045]; team[0.045]; squad[0.044]; Troncon[0.042]; national[0.042]; player[0.042]; game[0.041]; Twickenham[0.041]; Massimo[0.041]; Paolo[0.041]; Alessandro[0.041]; Cup[0.041]; Cup[0.041]; Francesco[0.040]; play[0.040]; week[0.040]; following[0.040]; month[0.040]; Squad[0.040]; announced[0.040]; announced[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland national rugby union team\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.188:0.188[0]; log p(e|m)= -3.772:-3.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; England[0.047]; Rugby[0.047]; squad[0.044]; team[0.044]; team[0.044]; Scotland[0.043]; Twickenham[0.043]; player[0.043]; play[0.042]; game[0.041]; Murrayfield[0.041]; Cup[0.041]; Cup[0.041]; national[0.040]; selection[0.040]; week[0.040]; named[0.040]; month[0.039]; comeback[0.039]; Italy[0.039]; Italy[0.039]; Italy[0.039]; George[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rugby Union\u001b[39m ==> ENTITY: \u001b[32mRugby union\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.049]; England[0.049]; Scotland[0.046]; Twickenham[0.045]; squad[0.045]; team[0.044]; Murrayfield[0.044]; play[0.043]; Italy[0.043]; Italy[0.043]; week[0.041]; month[0.041]; Saturday[0.040]; selection[0.040]; row[0.040]; recalled[0.039]; recalled[0.039]; named[0.039]; retiring[0.039]; George[0.039]; following[0.039]; friendly[0.038]; wing[0.038]; Friday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rome ==> ENTITY: \u001b[32mRome\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.050]; Italy[0.050]; England[0.043]; England[0.043]; week[0.042]; month[0.042]; George[0.042]; team[0.041]; dropped[0.041]; Stefano[0.041]; national[0.041]; Saturday[0.041]; Friday[0.041]; man[0.041]; said[0.041]; said[0.041]; beaten[0.040]; retiring[0.040]; play[0.040]; following[0.040]; friendly[0.040]; Marcello[0.040]; Corrado[0.040]; game[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Diego Dominguez\u001b[39m ==> ENTITY: \u001b[32mDiego Domínguez\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTroncon[0.046]; Italy[0.045]; game[0.045]; England[0.044]; England[0.044]; player[0.044]; team[0.043]; Javier[0.043]; comeback[0.043]; return[0.043]; Vaccari[0.042]; Orlandi[0.042]; national[0.041]; time[0.040]; ended[0.040]; Paolo[0.040]; Gianluca[0.040]; dropped[0.040]; announced[0.040]; Carlo[0.040]; Carlo[0.040]; months[0.039]; right[0.039]; Arancio[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paolo Vaccari\u001b[39m ==> ENTITY: \u001b[32mPaolo Vaccari\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.047]; Troncon[0.044]; Gianluca[0.043]; team[0.043]; player[0.043]; England[0.043]; England[0.043]; Alessandro[0.042]; Alessandro[0.042]; Cup[0.042]; Cup[0.042]; Massimo[0.042]; Massimo[0.042]; Francesco[0.042]; Corrado[0.042]; Mazzucato[0.041]; national[0.040]; game[0.040]; time[0.040]; Squad[0.040]; recalled[0.040]; comeback[0.040]; Orlandi[0.039]; Marcello[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland national rugby union team\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.182:0.182[0]; log p(e|m)= -3.772:-3.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.048]; England[0.048]; squad[0.044]; team[0.044]; team[0.044]; Scotland[0.043]; Twickenham[0.043]; player[0.043]; play[0.042]; game[0.041]; Murrayfield[0.041]; Cup[0.041]; Cup[0.041]; national[0.041]; selection[0.040]; week[0.040]; named[0.040]; month[0.040]; comeback[0.040]; Italy[0.039]; Italy[0.039]; Italy[0.039]; George[0.039]; time[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Massimo Cuttitta\u001b[39m ==> ENTITY: \u001b[32mMassimo Cuttitta\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTroncon[0.045]; England[0.045]; Gianluca[0.045]; Francesco[0.044]; Cup[0.043]; Mazzucato[0.043]; player[0.043]; Massimo[0.043]; Alessandro[0.042]; Alessandro[0.042]; Paolo[0.042]; Cuttitta[0.042]; Marcello[0.041]; Carlo[0.041]; Carlo[0.041]; time[0.041]; comeback[0.041]; Nicola[0.040]; months[0.040]; ago[0.040]; Orlandi[0.039]; Arancio[0.039]; Orazio[0.039]; Squad[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mItaly\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mItaly\u001b[39m <---> \u001b[32mItaly national rugby union team\u001b[39m\t\nSCORES: global= 0.245:0.236[\u001b[31m0.009\u001b[39m]; local()= 0.146:0.182[\u001b[32m0.036\u001b[39m]; log p(e|m)= -0.206:-4.135[\u001b[31m3.929\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.051]; Rugby[0.045]; England[0.045]; England[0.045]; team[0.044]; team[0.044]; squad[0.043]; Scotland[0.043]; Rome[0.042]; Murrayfield[0.042]; national[0.041]; game[0.041]; Twickenham[0.041]; play[0.040]; week[0.040]; following[0.040]; month[0.040]; announced[0.039]; announced[0.039]; Stefano[0.039]; Saturday[0.039]; Corrado[0.039]; friendly[0.038]; Friday[0.038]; \t\n\r [=>............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 19s174ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 242/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1285testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.092:0.092[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMichael[0.046]; American[0.044]; American[0.044]; recorded[0.044]; Friday[0.044]; told[0.043]; told[0.043]; million[0.042]; million[0.042]; says[0.041]; United[0.041]; United[0.041]; meeting[0.041]; annual[0.041]; Animal[0.040]; delivered[0.040]; Ohio[0.040]; States[0.040]; States[0.040]; Columbus[0.040]; half[0.040]; half[0.040]; Society[0.040]; Medical[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.047]; American[0.047]; United[0.045]; States[0.045]; Michael[0.044]; said[0.044]; said[0.044]; Chicago[0.044]; Columbus[0.043]; half[0.043]; half[0.043]; Ohio[0.043]; jump[0.043]; Friday[0.043]; percent[0.043]; million[0.043]; million[0.043]; recorded[0.042]; territories[0.042]; suffered[0.042]; reported[0.042]; says[0.042]; Association[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Doberman\u001b[39m ==> ENTITY: \u001b[32mDoberman Pinscher\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.236:0.236[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndog[0.045]; dog[0.045]; dog[0.045]; dog[0.045]; dogs[0.045]; dogs[0.045]; Rottweilers[0.045]; Rottweilers[0.045]; breed[0.044]; breeds[0.043]; breeds[0.043]; spaniels[0.042]; cocker[0.042]; Dalmatians[0.042]; neutered[0.041]; pet[0.040]; stray[0.039]; pets[0.039]; chows[0.038]; aggressiveness[0.038]; particular[0.037]; Today[0.037]; German[0.037]; neck[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = German shepherds\u001b[39m ==> ENTITY: \u001b[32mGerman Shepherd\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.215:0.215[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndogs[0.046]; dogs[0.046]; dog[0.045]; dog[0.045]; dog[0.045]; dog[0.045]; breed[0.044]; Doberman[0.044]; Rottweilers[0.043]; Rottweilers[0.043]; neutered[0.042]; Dalmatians[0.042]; spaniels[0.042]; breeds[0.042]; breeds[0.042]; stray[0.040]; pet[0.040]; pets[0.039]; cocker[0.039]; medical[0.038]; neck[0.038]; Animal[0.037]; chow[0.037]; particular[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = American Veterinary Medical Association\u001b[39m ==> ENTITY: \u001b[32mAmerican Veterinary Medical Association\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmedical[0.050]; Society[0.049]; animal[0.048]; Humane[0.047]; Association[0.047]; Animal[0.046]; Animal[0.046]; breed[0.046]; National[0.045]; dogs[0.045]; dogs[0.045]; veterinarian[0.045]; Glencoe[0.044]; doctors[0.044]; American[0.044]; Hospital[0.044]; neutered[0.044]; meeting[0.044]; vet[0.044]; States[0.044]; States[0.044]; kids[0.043]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mcocker spaniels\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAmerican Cocker Spaniel\u001b[39m <---> \u001b[32mCocker Spaniel\u001b[39m\t\nSCORES: global= 0.291:0.275[\u001b[31m0.016\u001b[39m]; local()= 0.180:0.191[\u001b[32m0.011\u001b[39m]; log p(e|m)= 0.000:-0.263[\u001b[31m0.263\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbreed[0.046]; breeds[0.045]; breeds[0.045]; dogs[0.045]; dogs[0.045]; Rottweilers[0.044]; Rottweilers[0.044]; dog[0.044]; dog[0.044]; dog[0.044]; dog[0.044]; Doberman[0.042]; pet[0.041]; pets[0.041]; neutered[0.041]; medical[0.039]; particular[0.039]; Dalmatians[0.039]; chows[0.039]; neck[0.038]; male[0.038]; children[0.038]; children[0.038]; suffered[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.085:0.085[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.046]; American[0.046]; United[0.044]; States[0.043]; Columbus[0.042]; Michael[0.042]; half[0.042]; half[0.042]; Ohio[0.042]; Friday[0.042]; million[0.041]; million[0.041]; Chicago[0.041]; recorded[0.041]; suffered[0.041]; says[0.040]; Association[0.040]; respect[0.040]; jump[0.040]; annual[0.040]; estimates[0.040]; teaching[0.040]; meeting[0.040]; Society[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ohio ==> ENTITY: \u001b[32mOhio\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.092:0.092[0]; log p(e|m)= -0.191:-0.191[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nColumbus[0.044]; estimated[0.044]; American[0.043]; American[0.043]; recorded[0.043]; National[0.043]; annual[0.042]; Michael[0.042]; said[0.042]; said[0.042]; said[0.042]; States[0.042]; States[0.042]; million[0.041]; million[0.041]; president[0.041]; Friday[0.041]; teaching[0.041]; territories[0.041]; gender[0.040]; estimates[0.040]; Medical[0.040]; Don[0.040]; needed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Columbus ==> ENTITY: \u001b[32mColumbus, Ohio\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.087:0.087[0]; log p(e|m)= -1.076:-1.076[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOhio[0.048]; American[0.044]; American[0.044]; million[0.044]; million[0.044]; Friday[0.044]; estimated[0.044]; States[0.044]; States[0.044]; territories[0.044]; reported[0.043]; reported[0.043]; Michael[0.043]; half[0.043]; half[0.043]; National[0.043]; said[0.043]; said[0.043]; said[0.043]; percent[0.043]; percent[0.043]; percent[0.043]; medical[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dalmatians ==> ENTITY: \u001b[32mDalmatian (dog)\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.197:0.197[0]; log p(e|m)= -0.017:-0.017[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndog[0.045]; dog[0.045]; dog[0.045]; dog[0.045]; breed[0.044]; dogs[0.044]; dogs[0.044]; spaniels[0.044]; breeds[0.043]; breeds[0.043]; cocker[0.043]; Doberman[0.042]; pet[0.042]; Rottweilers[0.041]; Rottweilers[0.041]; pets[0.041]; territories[0.041]; stray[0.039]; neutered[0.039]; medical[0.038]; chow[0.038]; shepherds[0.038]; neck[0.038]; particular[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = American ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.075:0.075[0]; log p(e|m)= -0.130:-0.130[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.051]; Ohio[0.045]; Michael[0.045]; fellow[0.044]; States[0.043]; States[0.043]; Columbus[0.043]; million[0.042]; million[0.042]; United[0.042]; United[0.042]; Chicago[0.042]; says[0.041]; half[0.041]; half[0.041]; Friday[0.040]; recorded[0.040]; Cornwell[0.040]; children[0.039]; children[0.039]; annual[0.039]; provoked[0.039]; provoked[0.039]; Half[0.039]; \t\n\r [=>............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 19s625ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 253/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1349testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.291:0.291[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.053]; Germany[0.053]; Germany[0.053]; Germany[0.053]; Germany[0.053]; Germany[0.053]; Austria[0.051]; Austria[0.051]; Austria[0.051]; Austria[0.051]; Austria[0.051]; Switzerland[0.050]; Switzerland[0.050]; Italy[0.048]; Italy[0.048]; Italy[0.048]; Italy[0.048]; Canada[0.047]; Sweden[0.047]; Sweden[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.282:0.282[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.051]; Italy[0.051]; Italy[0.051]; Cup[0.042]; Switzerland[0.041]; Switzerland[0.041]; Switzerland[0.041]; Slovenia[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; Austria[0.040]; Austria[0.040]; Austria[0.040]; Austria[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Sweden[0.038]; Martina[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.335:0.335[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.055]; Italy[0.055]; Italy[0.055]; Italy[0.055]; Italy[0.055]; Italy[0.055]; Switzerland[0.045]; Switzerland[0.045]; Switzerland[0.045]; Slovenia[0.045]; France[0.045]; France[0.045]; France[0.045]; France[0.045]; France[0.045]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.214:0.214[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Switzerland[0.043]; Sweden[0.041]; Italy[0.041]; Italy[0.041]; Canada[0.041]; France[0.040]; France[0.040]; France[0.040]; standings[0.040]; standings[0.040]; Cup[0.040]; Cup[0.040]; Cup[0.040]; Saturday[0.040]; Saturday[0.040]; Norway[0.039]; women[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Isolde Kostner\u001b[39m ==> ENTITY: \u001b[32mIsolde Kostner\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.126:0.126[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSeizinger[0.043]; Cup[0.043]; Hilde[0.042]; Hilde[0.042]; Gerg[0.042]; Gerg[0.042]; Gerg[0.042]; Cavagnoud[0.042]; Super[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Martina[0.042]; races[0.041]; World[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; standings[0.041]; standings[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.326:0.326[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.045]; Austria[0.045]; Austria[0.045]; Austria[0.045]; Switzerland[0.043]; Switzerland[0.043]; Switzerland[0.043]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Slovenia[0.041]; Sweden[0.040]; Sweden[0.040]; Cup[0.040]; Italy[0.039]; Italy[0.039]; Italy[0.039]; Italy[0.039]; Italy[0.039]; Italy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.162:0.162[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.046]; Sweden[0.046]; Switzerland[0.043]; World[0.043]; World[0.043]; World[0.043]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Cup[0.041]; Cup[0.041]; Cup[0.041]; Canada[0.040]; Italy[0.040]; Italy[0.040]; race[0.040]; Women[0.040]; Women[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.295:0.295[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.059]; Italy[0.059]; Italy[0.059]; Italy[0.059]; Switzerland[0.048]; Switzerland[0.048]; Slovenia[0.048]; standings[0.048]; standings[0.048]; France[0.048]; France[0.048]; France[0.048]; France[0.048]; France[0.048]; Austria[0.047]; Austria[0.047]; Austria[0.047]; Austria[0.047]; Austria[0.047]; points[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.110:0.110[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.045]; Switzerland[0.044]; Sweden[0.042]; Sweden[0.042]; Russia[0.042]; Russia[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Colorado[0.041]; World[0.040]; World[0.040]; World[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.285:0.285[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.045]; France[0.045]; France[0.045]; France[0.045]; France[0.045]; France[0.045]; Switzerland[0.042]; Switzerland[0.042]; Switzerland[0.042]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Martina[0.040]; Martina[0.040]; Austria[0.040]; Austria[0.040]; Austria[0.040]; Austria[0.040]; Austria[0.040]; Austria[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kate Pace Lindsay\u001b[39m ==> ENTITY: \u001b[32mKate Pace\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.047]; downhill[0.045]; downhill[0.045]; Downhill[0.045]; Seizinger[0.044]; Seizinger[0.044]; Cup[0.044]; Hilde[0.044]; Gerg[0.044]; Gerg[0.044]; Kostner[0.044]; Kostner[0.044]; Standings[0.043]; Ingeborg[0.043]; Martina[0.043]; Regina[0.043]; Picabo[0.043]; World[0.043]; Deborah[0.043]; race[0.042]; Carole[0.042]; Carole[0.042]; races[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.096:0.096[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.044]; Switzerland[0.044]; Switzerland[0.044]; Russia[0.043]; Cup[0.043]; Sweden[0.042]; Norway[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Slovenia[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; France[0.040]; France[0.040]; France[0.040]; France[0.040]; standings[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hilde Gerg\u001b[39m ==> ENTITY: \u001b[32mHilde Gerg\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGerg[0.045]; Super[0.045]; Seizinger[0.045]; Cup[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; standings[0.044]; standings[0.044]; Martina[0.043]; Martina[0.043]; Kostner[0.043]; Kostner[0.043]; Sweden[0.043]; Zurbriggen[0.042]; Zurbriggen[0.042]; Renate[0.042]; Pernilla[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.284:0.284[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.066]; Italy[0.066]; Italy[0.066]; Italy[0.066]; Cup[0.054]; Switzerland[0.054]; Slovenia[0.053]; standings[0.053]; standings[0.053]; France[0.053]; France[0.053]; France[0.053]; Austria[0.052]; Austria[0.052]; Austria[0.052]; Austria[0.052]; Austria[0.052]; points[0.052]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alexandra Meissnitzer\u001b[39m ==> ENTITY: \u001b[32mAlexandra Meissnitzer\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.045]; standings[0.043]; standings[0.043]; Switzerland[0.043]; Switzerland[0.043]; Switzerland[0.043]; Norway[0.042]; Gerg[0.042]; Super[0.042]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Seizinger[0.041]; Sweden[0.041]; Sweden[0.041]; Hilde[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Kostner[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Martina Ertl\u001b[39m ==> ENTITY: \u001b[32mMartina Ertl-Renz\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.045]; Austria[0.045]; Austria[0.045]; Austria[0.045]; Austria[0.045]; Ertl[0.045]; races[0.044]; downhill[0.044]; Martina[0.043]; World[0.043]; Kostner[0.043]; Kostner[0.043]; Seizinger[0.043]; Seizinger[0.043]; Cup[0.043]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Hilde[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Carole Montillet\u001b[39m ==> ENTITY: \u001b[32mCarole Montillet\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndownhill[0.044]; Women[0.043]; Women[0.043]; Alpine[0.043]; Downhill[0.043]; Skiing[0.043]; Seizinger[0.043]; race[0.042]; Gerg[0.042]; standings[0.042]; Hilde[0.042]; Cavagnoud[0.042]; Ingeborg[0.041]; Vail[0.041]; women[0.041]; Regina[0.041]; Cup[0.041]; Cup[0.041]; Colorado[0.041]; Saturday[0.041]; Zurbriggen[0.040]; Standings[0.040]; Standings[0.040]; Renate[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.292:0.292[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Switzerland[0.043]; Switzerland[0.043]; Switzerland[0.043]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Regine Cavagnoud\u001b[39m ==> ENTITY: \u001b[32mRégine Cavagnoud\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndownhill[0.044]; standings[0.044]; Cup[0.043]; Seizinger[0.042]; races[0.042]; Gerg[0.042]; Gerg[0.042]; Saturday[0.042]; Hilde[0.042]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Ingeborg[0.041]; Zurbriggen[0.041]; Zurbriggen[0.041]; Switzerland[0.041]; Switzerland[0.041]; Martina[0.041]; Regina[0.041]; World[0.040]; points[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.196:0.196[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.063]; Russia[0.063]; Russia[0.063]; Slovenia[0.059]; Switzerland[0.058]; Switzerland[0.058]; Switzerland[0.058]; Austria[0.058]; Austria[0.058]; Austria[0.058]; Austria[0.058]; Austria[0.058]; Austria[0.058]; Austria[0.058]; Austria[0.058]; points[0.058]; Sweden[0.058]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alexandra Meissnitzer\u001b[39m ==> ENTITY: \u001b[32mAlexandra Meissnitzer\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.215:0.215[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.046]; Cup[0.044]; Cup[0.044]; Cup[0.044]; Downhill[0.043]; race[0.042]; standings[0.042]; standings[0.042]; Switzerland[0.042]; Norway[0.041]; races[0.041]; Gerg[0.041]; downhill[0.041]; downhill[0.041]; Skiing[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Seizinger[0.041]; Seizinger[0.041]; Sweden[0.040]; Standings[0.040]; Standings[0.040]; Hilde[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.047]; Russia[0.047]; Russia[0.047]; Russia[0.047]; Canada[0.044]; Slovenia[0.044]; Switzerland[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; points[0.043]; Sweden[0.043]; Sweden[0.043]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.323:0.323[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.053]; Austria[0.053]; Austria[0.053]; Switzerland[0.052]; Switzerland[0.052]; Switzerland[0.052]; Germany[0.051]; Germany[0.051]; Germany[0.051]; Germany[0.051]; Germany[0.051]; Germany[0.051]; Slovenia[0.049]; Sweden[0.048]; Cup[0.048]; Italy[0.047]; Italy[0.047]; Italy[0.047]; Italy[0.047]; Norway[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.046]; Canada[0.044]; Switzerland[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Sweden[0.042]; Sweden[0.042]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Cup[0.041]; Cup[0.041]; Cup[0.041]; Norway[0.041]; Overall[0.040]; Italy[0.040]; Italy[0.040]; Standings[0.040]; Standings[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Zealand ==> ENTITY: \u001b[32mNew Zealand\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.083:0.083[0]; log p(e|m)= -0.238:-0.238[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.053]; Austria[0.053]; Austria[0.053]; Austria[0.053]; Austria[0.053]; Austria[0.053]; Switzerland[0.053]; Switzerland[0.053]; Switzerland[0.053]; races[0.053]; Slovenia[0.052]; World[0.052]; Italy[0.052]; Italy[0.052]; Italy[0.052]; Italy[0.052]; Italy[0.052]; Italy[0.052]; Cup[0.052]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Miriam Vogt\u001b[39m ==> ENTITY: \u001b[32mMiriam Vogt\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Switzerland[0.043]; Switzerland[0.043]; Seizinger[0.042]; Seizinger[0.042]; Downhill[0.042]; Renate[0.041]; Renate[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Hilde[0.041]; Vail[0.041]; Skiing[0.041]; Gerg[0.041]; Gerg[0.041]; Ingeborg[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.308:0.308[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.054]; Italy[0.054]; Italy[0.054]; Italy[0.054]; Italy[0.054]; Cup[0.044]; Switzerland[0.044]; Switzerland[0.044]; Switzerland[0.044]; Slovenia[0.043]; Slovenia[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.276:0.276[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.046]; France[0.046]; France[0.046]; France[0.046]; Switzerland[0.043]; Switzerland[0.043]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Martina[0.041]; Austria[0.040]; Austria[0.040]; Austria[0.040]; Austria[0.040]; Austria[0.040]; Austria[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.222:0.222[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Austria[0.046]; Austria[0.046]; Austria[0.046]; Switzerland[0.045]; Sweden[0.043]; Italy[0.043]; Italy[0.043]; Canada[0.043]; France[0.042]; France[0.042]; standings[0.042]; Cup[0.042]; Cup[0.042]; Saturday[0.042]; Norway[0.041]; Russia[0.041]; Russia[0.041]; World[0.041]; World[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Isolde Kostner\u001b[39m ==> ENTITY: \u001b[32mIsolde Kostner\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSeizinger[0.043]; Cup[0.043]; Hilde[0.042]; Gerg[0.042]; Super[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Martina[0.042]; Martina[0.042]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; standings[0.041]; standings[0.041]; points[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Renate Goetschl\u001b[39m ==> ENTITY: \u001b[32mRenate Götschl\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndownhill[0.047]; standings[0.045]; standings[0.045]; Cup[0.045]; Kostner[0.044]; Kostner[0.044]; races[0.044]; Saturday[0.044]; Seizinger[0.044]; Seizinger[0.044]; Switzerland[0.044]; Gerg[0.043]; Gerg[0.043]; Gerg[0.043]; women[0.043]; Hilde[0.043]; Hilde[0.043]; points[0.042]; Martina[0.042]; Ertl[0.042]; Canada[0.042]; Zurbriggen[0.042]; Super[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colorado ==> ENTITY: \u001b[32mColorado\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.032:0.032[0]; log p(e|m)= -0.269:-0.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.044]; Alpine[0.044]; Switzerland[0.043]; Vail[0.043]; Norway[0.043]; Heidi[0.042]; Carole[0.042]; Russia[0.042]; Pernilla[0.042]; Stefanie[0.042]; Skiing[0.042]; Picabo[0.042]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Helen[0.041]; Italy[0.040]; Italy[0.040]; Saturday[0.040]; World[0.040]; World[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.267:0.267[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.046]; France[0.046]; France[0.046]; France[0.046]; Switzerland[0.043]; Switzerland[0.043]; Switzerland[0.043]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Austria[0.040]; Austria[0.040]; Austria[0.040]; Austria[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Cup[0.040]; Sweden[0.039]; Slovenia[0.039]; Norway[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Florence\u001b[39m ==> ENTITY: \u001b[32mFlorence Masnada\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.049]; downhill[0.048]; downhill[0.048]; Kostner[0.047]; Seizinger[0.046]; Seizinger[0.046]; Cavagnoud[0.045]; Switzerland[0.045]; Zurbriggen[0.045]; Hilde[0.045]; France[0.045]; France[0.045]; France[0.045]; Italy[0.045]; Italy[0.045]; Gerg[0.045]; Gerg[0.045]; Ingeborg[0.045]; Downhill[0.045]; Norway[0.044]; Vail[0.044]; Regina[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.373:0.373[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.058]; Austria[0.058]; Austria[0.058]; Austria[0.058]; Austria[0.058]; Switzerland[0.056]; Switzerland[0.056]; Switzerland[0.056]; Germany[0.055]; Germany[0.055]; Germany[0.055]; Germany[0.055]; Germany[0.055]; Germany[0.055]; Germany[0.055]; Germany[0.055]; Germany[0.055]; Slovenia[0.052]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.278:0.278[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Switzerland[0.042]; Switzerland[0.042]; Sweden[0.041]; Sweden[0.041]; Italy[0.040]; Italy[0.040]; Italy[0.040]; France[0.039]; France[0.039]; France[0.039]; Saturday[0.039]; Saturday[0.039]; Canada[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.293:0.293[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.060]; France[0.060]; France[0.060]; France[0.060]; Switzerland[0.056]; Switzerland[0.056]; Switzerland[0.056]; Italy[0.054]; Italy[0.054]; Italy[0.054]; Italy[0.054]; Italy[0.054]; Italy[0.054]; Austria[0.053]; Austria[0.053]; Austria[0.053]; Austria[0.053]; Austria[0.053]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Katja Seizinger\u001b[39m ==> ENTITY: \u001b[32mKatja Seizinger\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nraces[0.046]; Cup[0.042]; Kostner[0.042]; Kostner[0.042]; standings[0.042]; Zurbriggen[0.042]; Gerg[0.042]; Gerg[0.042]; points[0.042]; women[0.042]; Hilde[0.041]; downhill[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Ingeborg[0.041]; Martina[0.041]; Picabo[0.041]; Norway[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Martina Ertl\u001b[39m ==> ENTITY: \u001b[32mMartina Ertl-Renz\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.217:0.217[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Ertl[0.044]; Martina[0.042]; Kostner[0.041]; Seizinger[0.041]; Cup[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Norway[0.041]; Hilde[0.040]; Sweden[0.040]; Sweden[0.040]; Gerg[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Carole Montillet\u001b[39m ==> ENTITY: \u001b[32mCarole Montillet\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndownhill[0.050]; Seizinger[0.049]; Seizinger[0.049]; Gerg[0.048]; Gerg[0.048]; Gerg[0.048]; standings[0.048]; standings[0.048]; Hilde[0.048]; Hilde[0.048]; Cavagnoud[0.048]; races[0.048]; Martina[0.047]; women[0.047]; Cup[0.047]; Saturday[0.047]; Super[0.046]; Zurbriggen[0.046]; Renate[0.046]; Wachter[0.046]; Stefanie[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.238:0.238[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.046]; Austria[0.046]; Switzerland[0.044]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Canada[0.043]; Norway[0.042]; Sweden[0.041]; Cup[0.041]; Cup[0.041]; Women[0.041]; Women[0.041]; World[0.041]; World[0.041]; Italy[0.040]; Italy[0.040]; France[0.039]; France[0.039]; France[0.039]; standings[0.039]; Regina[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.296:0.296[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Switzerland[0.041]; Switzerland[0.041]; Switzerland[0.041]; Italy[0.040]; Italy[0.040]; Italy[0.040]; Italy[0.040]; Italy[0.040]; Italy[0.040]; Italy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Miriam Vogt\u001b[39m ==> ENTITY: \u001b[32mMiriam Vogt\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Switzerland[0.043]; Switzerland[0.043]; Switzerland[0.043]; Vogt[0.043]; Seizinger[0.042]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Hilde[0.041]; Gerg[0.041]; Martina[0.040]; Kostner[0.040]; Miriam[0.040]; Pernilla[0.039]; Bibiana[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Switzerland ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.048]; Cup[0.047]; Cup[0.047]; Cup[0.047]; Sweden[0.047]; Austria[0.046]; Austria[0.046]; Austria[0.046]; Alpine[0.045]; Italy[0.045]; Italy[0.045]; Women[0.045]; Women[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Russia[0.045]; Russia[0.045]; women[0.044]; Norway[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.202:0.202[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.047]; Norway[0.044]; Switzerland[0.043]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; women[0.042]; Slovenia[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Canada[0.041]; Cup[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Anita Wachter\u001b[39m ==> ENTITY: \u001b[32mAnita Wachter\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Switzerland[0.042]; Switzerland[0.042]; Switzerland[0.042]; Seizinger[0.042]; Norway[0.041]; Pernilla[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Sweden[0.041]; Sweden[0.041]; Gerg[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Isolde Kostner\u001b[39m ==> ENTITY: \u001b[32mIsolde Kostner\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.048]; Seizinger[0.046]; Seizinger[0.046]; Cup[0.046]; Cup[0.046]; Cup[0.046]; Downhill[0.046]; Hilde[0.046]; race[0.046]; Gerg[0.045]; Cavagnoud[0.045]; Skiing[0.045]; Ingeborg[0.045]; Italy[0.045]; Italy[0.045]; Vail[0.045]; Regina[0.045]; races[0.045]; World[0.045]; World[0.045]; World[0.045]; Colorado[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Florence\u001b[39m ==> ENTITY: \u001b[32mFlorence Masnada\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.222:0.222[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFlorence[0.047]; downhill[0.046]; Kostner[0.045]; Kostner[0.045]; Seizinger[0.044]; Seizinger[0.044]; Cavagnoud[0.043]; Switzerland[0.043]; Zurbriggen[0.043]; Hilde[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Gerg[0.043]; Gerg[0.043]; Gerg[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.223:0.223[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.047]; France[0.047]; Switzerland[0.044]; Italy[0.043]; Italy[0.043]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Cup[0.041]; Cup[0.041]; Cup[0.041]; standings[0.041]; standings[0.041]; Canada[0.040]; Sweden[0.040]; Sweden[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.191:0.191[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.045]; Russia[0.045]; Russia[0.045]; Slovenia[0.042]; Switzerland[0.042]; Switzerland[0.042]; Switzerland[0.042]; Switzerland[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Sweden[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hilde Gerg\u001b[39m ==> ENTITY: \u001b[32mHilde Gerg\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDownhill[0.044]; Gerg[0.043]; Cavagnoud[0.042]; races[0.042]; Seizinger[0.042]; Seizinger[0.042]; Standings[0.042]; Cup[0.042]; Cup[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; standings[0.041]; standings[0.041]; downhill[0.041]; downhill[0.041]; Ingeborg[0.041]; Picabo[0.041]; World[0.041]; World[0.041]; Regina[0.041]; Canada[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.260:0.260[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Switzerland[0.042]; Switzerland[0.042]; Switzerland[0.042]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; Sweden[0.041]; Slovenia[0.040]; Norway[0.040]; Cup[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Katja Seizinger\u001b[39m ==> ENTITY: \u001b[32mKatja Seizinger\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrace[0.045]; Standings[0.043]; Standings[0.043]; Cup[0.042]; Cup[0.042]; Kostner[0.042]; standings[0.042]; Alpine[0.042]; Zurbriggen[0.042]; Gerg[0.042]; Downhill[0.042]; points[0.042]; Hilde[0.041]; downhill[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Ingeborg[0.041]; Vail[0.041]; Picabo[0.041]; Regina[0.041]; Norway[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sabina Panzanini\u001b[39m ==> ENTITY: \u001b[32mSabina Panzanini\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstandings[0.045]; standings[0.045]; races[0.044]; Cup[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Seizinger[0.044]; Seizinger[0.044]; Switzerland[0.044]; Switzerland[0.044]; Switzerland[0.044]; Hilde[0.043]; World[0.043]; Gerg[0.043]; Gerg[0.043]; downhill[0.042]; Martina[0.042]; Martina[0.042]; Alexandra[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.284:0.284[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.048]; Germany[0.048]; Germany[0.048]; Germany[0.048]; Germany[0.048]; Germany[0.048]; Germany[0.048]; Austria[0.046]; Austria[0.046]; Austria[0.046]; Austria[0.046]; Austria[0.046]; Switzerland[0.046]; Sweden[0.044]; Sweden[0.044]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Canada[0.043]; Slovenia[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Katharina Gutensohn\u001b[39m ==> ENTITY: \u001b[32mKatharina Gutensohn\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Pernilla[0.042]; Switzerland[0.042]; Switzerland[0.042]; Switzerland[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Seizinger[0.041]; Cup[0.041]; Hilde[0.040]; Gerg[0.040]; Kostner[0.040]; Sweden[0.040]; Sweden[0.040]; Martina[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Masnada\u001b[39m ==> ENTITY: \u001b[32mFlorence Masnada\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.231:0.231[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFlorence[0.047]; Florence[0.047]; downhill[0.046]; Kostner[0.045]; Kostner[0.045]; Seizinger[0.043]; Seizinger[0.043]; Cavagnoud[0.043]; Switzerland[0.043]; Zurbriggen[0.043]; Hilde[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Gerg[0.043]; Gerg[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.310:0.310[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.056]; France[0.056]; France[0.056]; France[0.056]; France[0.056]; France[0.056]; Switzerland[0.052]; Switzerland[0.052]; Switzerland[0.052]; Switzerland[0.052]; Italy[0.051]; Italy[0.051]; Italy[0.051]; Italy[0.051]; Italy[0.051]; Italy[0.051]; Italy[0.051]; Martina[0.050]; Martina[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.288:0.288[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Switzerland[0.042]; Switzerland[0.042]; Italy[0.040]; Italy[0.040]; Italy[0.040]; Canada[0.040]; Sweden[0.040]; Sweden[0.040]; France[0.039]; France[0.039]; France[0.039]; France[0.039]; France[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.159:0.159[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.052]; Switzerland[0.049]; Switzerland[0.049]; Switzerland[0.049]; Germany[0.048]; Germany[0.048]; Germany[0.048]; Germany[0.048]; Germany[0.048]; Austria[0.048]; Austria[0.048]; Austria[0.048]; Austria[0.048]; Slovenia[0.047]; Cup[0.047]; Italy[0.046]; Italy[0.046]; Italy[0.046]; Italy[0.046]; standings[0.045]; Russia[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Miriam Vogt\u001b[39m ==> ENTITY: \u001b[32mMiriam Vogt\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.215:0.215[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Switzerland[0.044]; Switzerland[0.044]; Switzerland[0.044]; Vogt[0.044]; Seizinger[0.044]; Seizinger[0.044]; Renate[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Hilde[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.308:0.308[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.045]; Austria[0.045]; Austria[0.045]; Austria[0.045]; Switzerland[0.043]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; women[0.041]; Slovenia[0.041]; Sweden[0.040]; Sweden[0.040]; Cup[0.040]; World[0.040]; Italy[0.039]; Italy[0.039]; Italy[0.039]; Italy[0.039]; Italy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Renate Goetschl\u001b[39m ==> ENTITY: \u001b[32mRenate Götschl\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndownhill[0.045]; Alpine[0.044]; standings[0.043]; Downhill[0.043]; Cup[0.043]; Cup[0.043]; Kostner[0.042]; Saturday[0.042]; Seizinger[0.042]; Switzerland[0.042]; Gerg[0.041]; Picabo[0.041]; Women[0.041]; Women[0.041]; Skiing[0.041]; race[0.041]; Hilde[0.041]; points[0.040]; Standings[0.040]; Standings[0.040]; Norway[0.040]; Ingeborg[0.040]; Vail[0.040]; Regina[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.310:0.310[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.062]; Italy[0.062]; Italy[0.062]; Italy[0.062]; Cup[0.051]; Switzerland[0.051]; Switzerland[0.051]; Switzerland[0.051]; Slovenia[0.051]; France[0.050]; France[0.050]; France[0.050]; France[0.050]; France[0.050]; Austria[0.049]; Austria[0.049]; Austria[0.049]; Austria[0.049]; Austria[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Deborah Compagnoni\u001b[39m ==> ENTITY: \u001b[32mDeborah Compagnoni\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.044]; Italy[0.044]; Italy[0.044]; Italy[0.044]; Italy[0.044]; Seizinger[0.042]; Seizinger[0.042]; downhill[0.042]; races[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Cup[0.041]; Hilde[0.041]; Gerg[0.041]; Gerg[0.041]; Martina[0.041]; Martina[0.041]; Super[0.041]; World[0.040]; Kostner[0.039]; Kostner[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.342:0.342[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.060]; Italy[0.060]; Italy[0.060]; Italy[0.060]; Italy[0.060]; Italy[0.060]; Cup[0.050]; Switzerland[0.049]; Switzerland[0.049]; Switzerland[0.049]; Slovenia[0.049]; Slovenia[0.049]; France[0.049]; France[0.049]; France[0.049]; France[0.049]; France[0.049]; France[0.049]; France[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.171:0.171[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.049]; Switzerland[0.048]; Austria[0.047]; Austria[0.047]; Austria[0.047]; women[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Italy[0.045]; Italy[0.045]; Canada[0.045]; Cup[0.044]; Cup[0.044]; Cup[0.044]; Women[0.044]; Women[0.044]; Downhill[0.043]; Russia[0.043]; Russia[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.043]; Sweden[0.043]; Cup[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; races[0.042]; World[0.042]; Switzerland[0.042]; Switzerland[0.042]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Slovenia ==> ENTITY: \u001b[32mSlovenia\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.233:0.233[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Norway[0.042]; Switzerland[0.042]; Switzerland[0.042]; Switzerland[0.042]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Cup[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; standings[0.040]; points[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pernilla Wiberg\u001b[39m ==> ENTITY: \u001b[32mPernilla Wiberg\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZurbriggen[0.054]; Zurbriggen[0.054]; Gerg[0.054]; Gerg[0.054]; Seizinger[0.054]; Sweden[0.054]; Sweden[0.054]; Wiberg[0.053]; Hilde[0.053]; standings[0.052]; Pernilla[0.052]; Martina[0.052]; Martina[0.052]; Kostner[0.051]; Kostner[0.051]; Switzerland[0.051]; Switzerland[0.051]; Switzerland[0.051]; Renate[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Heidi Zurbriggen\u001b[39m ==> ENTITY: \u001b[32mHeidi Zurbriggen\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.044]; Switzerland[0.043]; races[0.043]; Pernilla[0.043]; Pernilla[0.043]; Seizinger[0.042]; Martina[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Hilde[0.042]; Hilde[0.042]; Claudia[0.042]; Gerg[0.041]; Gerg[0.041]; Gerg[0.041]; Sweden[0.041]; Sweden[0.041]; World[0.041]; Cavagnoud[0.040]; Anita[0.040]; Renate[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.283:0.283[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.048]; Germany[0.048]; Germany[0.048]; Germany[0.048]; Germany[0.048]; Germany[0.048]; Austria[0.046]; Austria[0.046]; Austria[0.046]; Austria[0.046]; Austria[0.046]; Switzerland[0.046]; Switzerland[0.046]; Switzerland[0.046]; Sweden[0.044]; Sweden[0.044]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Norway[0.043]; Slovenia[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.100:0.100[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.049]; Switzerland[0.048]; Sweden[0.046]; Sweden[0.046]; Russia[0.046]; Russia[0.046]; Austria[0.046]; Austria[0.046]; Austria[0.046]; Italy[0.045]; Italy[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Colorado[0.044]; Cup[0.044]; Cup[0.044]; Cup[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.278:0.278[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Switzerland[0.044]; Switzerland[0.044]; Switzerland[0.044]; Sweden[0.042]; Sweden[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Slovenia[0.041]; Canada[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.279:0.279[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.045]; Austria[0.045]; Austria[0.045]; Switzerland[0.044]; Switzerland[0.044]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; women[0.042]; Norway[0.040]; Italy[0.040]; Italy[0.040]; Italy[0.040]; Sweden[0.040]; Sweden[0.040]; points[0.039]; points[0.039]; Women[0.039]; Women[0.039]; Skiing[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.363:0.363[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.058]; Austria[0.058]; Austria[0.058]; Austria[0.058]; Austria[0.058]; Switzerland[0.056]; Switzerland[0.056]; Switzerland[0.056]; Germany[0.055]; Germany[0.055]; Germany[0.055]; Germany[0.055]; Germany[0.055]; Germany[0.055]; Germany[0.055]; Germany[0.055]; Slovenia[0.053]; Slovenia[0.053]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.045]; Russia[0.045]; Russia[0.045]; Canada[0.043]; Switzerland[0.042]; Switzerland[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Sweden[0.041]; Sweden[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Cup[0.040]; Norway[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Florence Masnada\u001b[39m ==> ENTITY: \u001b[32mFlorence Masnada\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.235:0.235[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFlorence[0.046]; Kostner[0.043]; Seizinger[0.042]; Switzerland[0.042]; Switzerland[0.042]; Switzerland[0.042]; Switzerland[0.042]; Zurbriggen[0.041]; Zurbriggen[0.041]; Hilde[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Switzerland ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.276:0.276[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.049]; France[0.049]; France[0.049]; France[0.049]; France[0.049]; Italy[0.048]; Italy[0.048]; Italy[0.048]; Italy[0.048]; Italy[0.048]; Slovenia[0.047]; Sweden[0.047]; Sweden[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Switzerland ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.283:0.283[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.046]; Switzerland[0.046]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Slovenia[0.041]; Sweden[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Austria[0.040]; Austria[0.040]; Austria[0.040]; Austria[0.040]; Russia[0.040]; Russia[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alexandra Meissnitzer\u001b[39m ==> ENTITY: \u001b[32mAlexandra Meissnitzer\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.051]; standings[0.049]; standings[0.049]; Switzerland[0.048]; Switzerland[0.048]; Switzerland[0.048]; races[0.048]; Gerg[0.048]; Gerg[0.048]; Super[0.048]; downhill[0.047]; Austria[0.047]; Austria[0.047]; Austria[0.047]; Austria[0.047]; Austria[0.047]; Seizinger[0.047]; Seizinger[0.047]; Sweden[0.046]; Sweden[0.046]; Hilde[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Switzerland ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.279:0.279[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.060]; Switzerland[0.060]; France[0.056]; France[0.056]; France[0.056]; France[0.056]; France[0.056]; Italy[0.055]; Italy[0.055]; Italy[0.055]; Italy[0.055]; Italy[0.055]; Italy[0.055]; Italy[0.055]; Slovenia[0.054]; Slovenia[0.054]; Sweden[0.054]; Sweden[0.054]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Heidi Zurbriggen\u001b[39m ==> ENTITY: \u001b[32mHeidi Zurbriggen\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.044]; Cup[0.044]; Cup[0.044]; race[0.043]; Switzerland[0.043]; Pernilla[0.042]; Seizinger[0.042]; Downhill[0.042]; Alpine[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Hilde[0.041]; Gerg[0.041]; Ingeborg[0.041]; Helen[0.041]; Vail[0.041]; Sweden[0.041]; Regina[0.041]; World[0.041]; World[0.041]; World[0.041]; Skiing[0.041]; Picabo[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Cup ==> ENTITY: \u001b[32mFIS Alpine Ski World Cup\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.233:0.233[0]; log p(e|m)= -2.765:-2.765[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.045]; Sweden[0.045]; Norway[0.044]; Switzerland[0.043]; Switzerland[0.043]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Canada[0.041]; downhill[0.041]; Slovenia[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; races[0.040]; France[0.040]; France[0.040]; France[0.040]; France[0.040]; France[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Katja Seizinger\u001b[39m ==> ENTITY: \u001b[32mKatja Seizinger\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.045]; Kostner[0.045]; standings[0.045]; standings[0.045]; Zurbriggen[0.045]; Gerg[0.044]; points[0.044]; Hilde[0.044]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Martina[0.043]; Martina[0.043]; Norway[0.043]; Renate[0.042]; Switzerland[0.042]; Switzerland[0.042]; Switzerland[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.316:0.316[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.045]; Austria[0.045]; Austria[0.045]; Switzerland[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; women[0.041]; Slovenia[0.041]; Italy[0.040]; Italy[0.040]; Italy[0.040]; Italy[0.040]; Italy[0.040]; Italy[0.040]; Sweden[0.039]; Sweden[0.039]; points[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.188:0.188[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.045]; Switzerland[0.044]; Switzerland[0.044]; Switzerland[0.044]; Cup[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Slovenia[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; France[0.038]; France[0.038]; France[0.038]; France[0.038]; Ertl[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.316:0.316[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.056]; Italy[0.056]; Italy[0.056]; Italy[0.056]; Italy[0.056]; Cup[0.046]; Switzerland[0.046]; Switzerland[0.046]; Switzerland[0.046]; Slovenia[0.045]; France[0.045]; France[0.045]; France[0.045]; France[0.045]; France[0.045]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.193:0.193[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.048]; France[0.048]; Switzerland[0.045]; Italy[0.043]; Italy[0.043]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Cup[0.041]; Cup[0.041]; standings[0.041]; Canada[0.041]; Sweden[0.040]; Women[0.040]; Women[0.040]; Norway[0.039]; women[0.039]; Russia[0.039]; Russia[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.291:0.291[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.052]; Germany[0.052]; Germany[0.052]; Germany[0.052]; Germany[0.052]; Germany[0.052]; Austria[0.050]; Austria[0.050]; Austria[0.050]; Austria[0.050]; Austria[0.050]; Switzerland[0.049]; Switzerland[0.049]; Switzerland[0.049]; Italy[0.048]; Italy[0.048]; Italy[0.048]; Italy[0.048]; Italy[0.048]; Italy[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.274:0.274[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Switzerland[0.044]; Sweden[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Norway[0.041]; Canada[0.041]; Slovenia[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Carole Montillet\u001b[39m ==> ENTITY: \u001b[32mCarole Montillet\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.108:0.108[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSeizinger[0.044]; Gerg[0.044]; standings[0.043]; standings[0.043]; Hilde[0.043]; Martina[0.043]; Martina[0.043]; Cup[0.042]; Super[0.042]; Zurbriggen[0.041]; Wachter[0.041]; Stefanie[0.041]; Stefanie[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; Kostner[0.041]; Madlen[0.041]; Katharina[0.041]; Norway[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.260:0.260[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.046]; Austria[0.046]; Switzerland[0.045]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Canada[0.044]; Norway[0.042]; Sweden[0.041]; Italy[0.041]; Italy[0.041]; points[0.040]; standings[0.039]; standings[0.039]; Regina[0.039]; Skiing[0.039]; France[0.039]; France[0.039]; France[0.039]; Alpine[0.039]; Pernilla[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.280:0.280[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.046]; France[0.046]; France[0.046]; France[0.046]; Switzerland[0.043]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Martina[0.041]; Austria[0.040]; Austria[0.040]; Austria[0.040]; Austria[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.146:0.146[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSlovenia[0.049]; Switzerland[0.049]; Switzerland[0.049]; Switzerland[0.049]; Austria[0.049]; Austria[0.049]; Austria[0.049]; Austria[0.049]; Sweden[0.048]; Germany[0.048]; Germany[0.048]; Germany[0.048]; Germany[0.048]; Cup[0.047]; Norway[0.047]; Italy[0.046]; Italy[0.046]; Italy[0.046]; Italy[0.046]; States[0.046]; standings[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Masnada\u001b[39m ==> ENTITY: \u001b[32mFlorence Masnada\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFlorence[0.047]; Alpine[0.047]; downhill[0.046]; downhill[0.046]; Kostner[0.045]; Seizinger[0.043]; Seizinger[0.043]; Cavagnoud[0.043]; Switzerland[0.043]; Zurbriggen[0.043]; Hilde[0.043]; France[0.043]; France[0.043]; France[0.043]; Italy[0.043]; Italy[0.043]; Gerg[0.043]; Gerg[0.043]; Ingeborg[0.042]; Downhill[0.042]; Norway[0.042]; Vail[0.042]; Regina[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.232:0.232[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.056]; Switzerland[0.052]; Switzerland[0.052]; Switzerland[0.052]; Austria[0.050]; Austria[0.050]; Austria[0.050]; Austria[0.050]; Austria[0.050]; Austria[0.050]; Austria[0.050]; Slovenia[0.049]; Germany[0.049]; Germany[0.049]; Germany[0.049]; Germany[0.049]; Germany[0.049]; Germany[0.049]; Germany[0.049]; Germany[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Cup ==> ENTITY: \u001b[32mFIS Alpine Ski World Cup\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.236:0.236[0]; log p(e|m)= -2.765:-2.765[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.047]; Norway[0.046]; Alpine[0.046]; Downhill[0.046]; World[0.045]; Switzerland[0.045]; Cup[0.045]; Austria[0.044]; Austria[0.044]; Austria[0.044]; downhill[0.043]; Skiing[0.043]; Italy[0.043]; Italy[0.043]; France[0.042]; France[0.042]; Gerg[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; race[0.042]; standings[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.263:0.263[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.051]; Italy[0.051]; Cup[0.042]; Cup[0.042]; Switzerland[0.042]; Switzerland[0.042]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; women[0.040]; Canada[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.284:0.284[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.048]; France[0.048]; France[0.048]; France[0.048]; Switzerland[0.045]; Switzerland[0.045]; Switzerland[0.045]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Cup[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pernilla Wiberg\u001b[39m ==> ENTITY: \u001b[32mPernilla Wiberg\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndownhill[0.049]; Norway[0.047]; races[0.047]; Zurbriggen[0.047]; Gerg[0.046]; Gerg[0.046]; Seizinger[0.046]; Sweden[0.046]; Wiberg[0.046]; Hilde[0.046]; Ingeborg[0.045]; standings[0.045]; standings[0.045]; Pernilla[0.045]; Martina[0.045]; women[0.045]; Cup[0.045]; Kostner[0.044]; Kostner[0.044]; super[0.044]; Switzerland[0.044]; Renate[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Heidi Zurbriggen\u001b[39m ==> ENTITY: \u001b[32mHeidi Zurbriggen\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.048]; Switzerland[0.047]; Switzerland[0.047]; Switzerland[0.047]; Pernilla[0.046]; Seizinger[0.046]; Martina[0.046]; Martina[0.046]; Austria[0.045]; Austria[0.045]; Austria[0.045]; Austria[0.045]; Austria[0.045]; Austria[0.045]; Hilde[0.045]; Claudia[0.045]; Gerg[0.045]; Sweden[0.044]; Sweden[0.044]; Marianne[0.044]; Anita[0.044]; Carole[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.047]; Russia[0.047]; Russia[0.047]; Russia[0.047]; Canada[0.044]; Slovenia[0.044]; Switzerland[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; points[0.043]; Sweden[0.043]; Sweden[0.043]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.330:0.330[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.057]; Germany[0.057]; Germany[0.057]; Germany[0.057]; Germany[0.057]; Germany[0.057]; Germany[0.057]; Austria[0.055]; Austria[0.055]; Austria[0.055]; Austria[0.055]; Austria[0.055]; Austria[0.055]; Austria[0.055]; Austria[0.055]; Switzerland[0.054]; Switzerland[0.054]; Switzerland[0.054]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pernilla Wiberg\u001b[39m ==> ENTITY: \u001b[32mPernilla Wiberg\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndownhill[0.045]; Alpine[0.044]; Downhill[0.043]; Norway[0.043]; race[0.042]; Zurbriggen[0.042]; Gerg[0.042]; Skiing[0.042]; Seizinger[0.042]; Sweden[0.042]; Hilde[0.041]; Women[0.041]; Women[0.041]; Ingeborg[0.041]; standings[0.041]; standings[0.041]; Vail[0.041]; women[0.041]; Regina[0.041]; Cup[0.041]; Cup[0.041]; Cup[0.041]; Colorado[0.040]; Kostner[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.324:0.324[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.051]; Austria[0.051]; Austria[0.051]; Switzerland[0.049]; Switzerland[0.049]; Switzerland[0.049]; Germany[0.049]; Germany[0.049]; Germany[0.049]; Germany[0.049]; Germany[0.049]; Germany[0.049]; Slovenia[0.047]; Norway[0.045]; Italy[0.045]; Italy[0.045]; Italy[0.045]; Italy[0.045]; Sweden[0.045]; Sweden[0.045]; points[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vail ==> ENTITY: \u001b[32mVail, Colorado\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.392:-0.392[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.051]; Downhill[0.049]; Skiing[0.048]; Saturday[0.048]; Colorado[0.047]; race[0.047]; downhill[0.047]; points[0.046]; Norway[0.046]; Regina[0.045]; Picabo[0.045]; Switzerland[0.045]; Women[0.044]; Women[0.044]; World[0.044]; World[0.044]; Cup[0.044]; Cup[0.044]; Helen[0.044]; Sweden[0.043]; standings[0.043]; Carole[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Slovenia ==> ENTITY: \u001b[32mSlovenia\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.262:0.262[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.043]; Sweden[0.043]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Switzerland[0.042]; Switzerland[0.042]; Switzerland[0.042]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Anita Wachter\u001b[39m ==> ENTITY: \u001b[32mAnita Wachter\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.047]; Austria[0.046]; Austria[0.046]; Austria[0.046]; Austria[0.046]; Austria[0.046]; races[0.046]; Switzerland[0.046]; downhill[0.046]; Seizinger[0.046]; Pernilla[0.045]; Pernilla[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Sweden[0.045]; Sweden[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Switzerland ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.280:0.280[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.063]; Switzerland[0.063]; France[0.059]; France[0.059]; France[0.059]; France[0.059]; France[0.059]; Italy[0.058]; Italy[0.058]; Italy[0.058]; Italy[0.058]; Italy[0.058]; Italy[0.058]; Slovenia[0.057]; Slovenia[0.057]; Sweden[0.057]; Sweden[0.057]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.214:0.214[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.052]; Switzerland[0.042]; standings[0.042]; standings[0.042]; France[0.042]; France[0.042]; France[0.042]; Standings[0.041]; Standings[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; points[0.041]; points[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Cup[0.041]; Cup[0.041]; Cup[0.041]; Canada[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.272:0.272[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.064]; France[0.064]; France[0.064]; France[0.064]; Switzerland[0.060]; Italy[0.058]; Italy[0.058]; Italy[0.058]; Italy[0.058]; Italy[0.058]; Martina[0.057]; Martina[0.057]; Austria[0.056]; Austria[0.056]; Austria[0.056]; Austria[0.056]; Austria[0.056]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Picabo Street\u001b[39m ==> ENTITY: \u001b[32mPicabo Street\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSkiing[0.043]; Seizinger[0.043]; Seizinger[0.043]; races[0.042]; Downhill[0.042]; Hilde[0.042]; Gerg[0.042]; Gerg[0.042]; downhill[0.042]; downhill[0.042]; Alpine[0.042]; race[0.042]; Cup[0.042]; Cup[0.042]; Cup[0.042]; Ingeborg[0.042]; Vail[0.041]; Regina[0.041]; World[0.041]; World[0.041]; World[0.041]; Saturday[0.041]; Saturday[0.041]; Canada[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.291:0.291[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Switzerland[0.041]; Switzerland[0.041]; Switzerland[0.041]; Italy[0.040]; Italy[0.040]; Italy[0.040]; Italy[0.040]; Italy[0.040]; Italy[0.040]; Italy[0.040]; \t\n\r [==>...........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 17s451ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 367/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1330testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wellington ==> ENTITY: \u001b[32mWellington\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.926:-0.926[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWellington[0.054]; Zealand[0.048]; Friday[0.047]; Friday[0.047]; National[0.047]; National[0.047]; National[0.047]; needed[0.045]; level[0.045]; level[0.045]; Peters[0.045]; Peters[0.045]; Peters[0.045]; Peters[0.045]; good[0.044]; similar[0.044]; similar[0.044]; Clark[0.044]; stage[0.044]; said[0.043]; said[0.043]; minutes[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NZ First\u001b[39m ==> ENTITY: \u001b[32mNew Zealand First\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPeters[0.043]; Peters[0.043]; Peters[0.043]; Peters[0.043]; National[0.043]; National[0.043]; National[0.043]; leader[0.043]; leader[0.043]; coalition[0.042]; Wellington[0.042]; Wellington[0.042]; meeting[0.041]; Labour[0.041]; Labour[0.041]; Labour[0.041]; parties[0.041]; Winston[0.041]; Nz[0.041]; completion[0.041]; completion[0.041]; Zealand[0.041]; left[0.040]; left[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Peters\u001b[39m ==> ENTITY: \u001b[32mWinston Peters\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNational[0.047]; National[0.047]; National[0.047]; leader[0.047]; leader[0.047]; Labour[0.046]; Labour[0.046]; Labour[0.046]; Wellington[0.046]; Wellington[0.046]; coalition[0.045]; meeting[0.045]; Peters[0.045]; Peters[0.045]; Peters[0.045]; Zealand[0.045]; Nz[0.045]; Nz[0.045]; completion[0.045]; completion[0.045]; said[0.043]; said[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = National ==> ENTITY: \u001b[32mNew Zealand National Party\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.158:0.158[0]; log p(e|m)= -1.877:-1.877[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNational[0.044]; National[0.044]; leader[0.044]; leader[0.044]; Wellington[0.043]; Wellington[0.043]; coalition[0.043]; Peters[0.042]; Peters[0.042]; Peters[0.042]; Peters[0.042]; Labour[0.042]; Labour[0.042]; Labour[0.042]; parties[0.041]; Zealand[0.041]; Friday[0.041]; Friday[0.041]; level[0.040]; level[0.040]; Clark[0.040]; Winston[0.040]; meeting[0.039]; needed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Labour ==> ENTITY: \u001b[32mNew Zealand Labour Party\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.172:0.172[0]; log p(e|m)= -2.120:-2.120[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLabour[0.045]; Labour[0.045]; leader[0.044]; leader[0.044]; level[0.043]; level[0.043]; coalition[0.042]; National[0.042]; National[0.042]; National[0.042]; Wellington[0.042]; Wellington[0.042]; Clark[0.041]; meeting[0.041]; Zealand[0.040]; Helen[0.040]; said[0.040]; said[0.040]; Peters[0.040]; Peters[0.040]; Peters[0.040]; Peters[0.040]; Winston[0.040]; parties[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Labour ==> ENTITY: \u001b[32mNew Zealand Labour Party\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.172:0.172[0]; log p(e|m)= -2.120:-2.120[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLabour[0.045]; Labour[0.045]; leader[0.044]; leader[0.044]; level[0.043]; level[0.043]; coalition[0.042]; National[0.042]; National[0.042]; National[0.042]; Wellington[0.042]; Wellington[0.042]; Clark[0.041]; meeting[0.041]; Zealand[0.040]; Helen[0.040]; said[0.040]; said[0.040]; Peters[0.040]; Peters[0.040]; Peters[0.040]; Peters[0.040]; Winston[0.040]; parties[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Labour ==> ENTITY: \u001b[32mNew Zealand Labour Party\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.172:0.172[0]; log p(e|m)= -2.120:-2.120[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLabour[0.045]; Labour[0.045]; leader[0.044]; leader[0.044]; level[0.043]; level[0.043]; coalition[0.042]; National[0.042]; National[0.042]; National[0.042]; Wellington[0.042]; Wellington[0.042]; Clark[0.041]; meeting[0.041]; Zealand[0.040]; Helen[0.040]; said[0.040]; said[0.040]; Peters[0.040]; Peters[0.040]; Peters[0.040]; Peters[0.040]; Winston[0.040]; parties[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Peters\u001b[39m ==> ENTITY: \u001b[32mWinston Peters\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNational[0.047]; National[0.047]; National[0.047]; leader[0.047]; leader[0.047]; Labour[0.046]; Labour[0.046]; Labour[0.046]; Wellington[0.046]; Wellington[0.046]; coalition[0.045]; meeting[0.045]; Peters[0.045]; Peters[0.045]; Peters[0.045]; Zealand[0.045]; Nz[0.045]; Nz[0.045]; completion[0.045]; completion[0.045]; said[0.043]; said[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Winston Peters\u001b[39m ==> ENTITY: \u001b[32mWinston Peters\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNational[0.047]; National[0.047]; National[0.047]; leader[0.047]; leader[0.047]; Labour[0.046]; Labour[0.046]; Labour[0.046]; Wellington[0.046]; Wellington[0.046]; coalition[0.045]; meeting[0.045]; Peters[0.045]; Peters[0.045]; Peters[0.045]; Zealand[0.045]; Nz[0.045]; Nz[0.045]; completion[0.045]; completion[0.045]; said[0.043]; said[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = National ==> ENTITY: \u001b[32mNew Zealand National Party\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.161:0.161[0]; log p(e|m)= -1.877:-1.877[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNational[0.044]; National[0.044]; leader[0.044]; leader[0.044]; Wellington[0.043]; Wellington[0.043]; coalition[0.043]; Peters[0.042]; Peters[0.042]; Peters[0.042]; Peters[0.042]; Labour[0.042]; Labour[0.042]; Labour[0.042]; Zealand[0.041]; parties[0.041]; level[0.041]; level[0.041]; Clark[0.041]; meeting[0.040]; Winston[0.040]; needed[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Helen Clark\u001b[39m ==> ENTITY: \u001b[32mHelen Clark\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNational[0.043]; National[0.043]; National[0.043]; speaking[0.043]; Zealand[0.042]; Labour[0.042]; Labour[0.042]; Labour[0.042]; meeting[0.042]; Wellington[0.042]; Wellington[0.042]; leader[0.042]; leader[0.042]; coalition[0.042]; said[0.042]; said[0.042]; speak[0.041]; Nz[0.041]; Nz[0.041]; completion[0.041]; completion[0.041]; told[0.041]; Winston[0.040]; parties[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Peters\u001b[39m ==> ENTITY: \u001b[32mWinston Peters\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNational[0.047]; National[0.047]; National[0.047]; leader[0.047]; leader[0.047]; Labour[0.046]; Labour[0.046]; Labour[0.046]; Wellington[0.046]; Wellington[0.046]; coalition[0.045]; meeting[0.045]; Peters[0.045]; Peters[0.045]; Peters[0.045]; Zealand[0.045]; Nz[0.045]; Nz[0.045]; completion[0.045]; completion[0.045]; said[0.043]; said[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = New Zealand First\u001b[39m ==> ENTITY: \u001b[32mNew Zealand First\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPeters[0.043]; Peters[0.043]; Peters[0.043]; Peters[0.043]; National[0.043]; National[0.043]; National[0.043]; leader[0.043]; leader[0.043]; coalition[0.042]; Wellington[0.042]; Wellington[0.042]; meeting[0.041]; Labour[0.041]; Labour[0.041]; Labour[0.041]; parties[0.041]; Winston[0.041]; Nz[0.041]; Nz[0.041]; completion[0.041]; completion[0.041]; left[0.040]; left[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.098:0.098[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntold[0.045]; said[0.043]; said[0.043]; needed[0.042]; National[0.042]; National[0.042]; National[0.042]; says[0.042]; negotiators[0.042]; coalition[0.041]; spend[0.041]; Friday[0.041]; Friday[0.041]; speak[0.041]; level[0.041]; level[0.041]; speaking[0.041]; good[0.041]; parties[0.041]; similar[0.041]; similar[0.041]; leader[0.041]; leader[0.041]; meeting[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = National ==> ENTITY: \u001b[32mNew Zealand National Party\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.158:0.158[0]; log p(e|m)= -1.877:-1.877[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNational[0.044]; National[0.044]; leader[0.044]; leader[0.044]; Wellington[0.043]; Wellington[0.043]; coalition[0.043]; Peters[0.042]; Peters[0.042]; Peters[0.042]; Peters[0.042]; Labour[0.042]; Labour[0.042]; Labour[0.042]; parties[0.041]; Zealand[0.041]; Friday[0.041]; Friday[0.041]; level[0.040]; level[0.040]; Clark[0.040]; Winston[0.040]; meeting[0.039]; needed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wellington ==> ENTITY: \u001b[32mWellington\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.926:-0.926[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWellington[0.054]; Zealand[0.048]; Friday[0.047]; Friday[0.047]; National[0.047]; National[0.047]; National[0.047]; needed[0.045]; level[0.045]; level[0.045]; Peters[0.045]; Peters[0.045]; Peters[0.045]; Peters[0.045]; good[0.044]; similar[0.044]; similar[0.044]; Clark[0.044]; stage[0.044]; said[0.043]; said[0.043]; minutes[0.043]; \t\n\r [==>...........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 18s176ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 383/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1190testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Baltimore ==> ENTITY: \u001b[32mBaltimore Ravens\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.219:0.219[0]; log p(e|m)= -4.075:-4.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJacksonville[0.045]; Carolina[0.043]; Carolina[0.043]; Houston[0.043]; Cincinnati[0.043]; Pittsburgh[0.043]; Washington[0.043]; Buffalo[0.043]; Giants[0.042]; Oakland[0.042]; Jets[0.042]; Miami[0.042]; Detroit[0.041]; Detroit[0.041]; Dallas[0.041]; Tampa[0.041]; Tampa[0.041]; Sunday[0.040]; Arizona[0.040]; Monday[0.040]; England[0.040]; Denver[0.040]; Atlanta[0.040]; Atlanta[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cincinnati ==> ENTITY: \u001b[32mCincinnati Bengals\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.212:0.212[0]; log p(e|m)= -3.963:-3.963[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJacksonville[0.046]; Giants[0.045]; Oakland[0.045]; Buffalo[0.045]; Houston[0.045]; Carolina[0.044]; Carolina[0.044]; Jets[0.044]; Pittsburgh[0.044]; Miami[0.044]; Baltimore[0.044]; Monday[0.044]; Sunday[0.043]; Arizona[0.043]; Louis[0.043]; Louis[0.043]; Dallas[0.043]; Denver[0.042]; Washington[0.042]; Detroit[0.042]; Detroit[0.042]; Minnesota[0.042]; Minnesota[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kansas City ==> ENTITY: \u001b[32mKansas City Chiefs\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.193:0.193[0]; log p(e|m)= -3.170:-3.170[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.043]; Jacksonville[0.043]; Football[0.043]; Football[0.043]; Baltimore[0.042]; Philadelphia[0.042]; Oakland[0.042]; Division[0.042]; Division[0.042]; Division[0.042]; Division[0.042]; Division[0.042]; Cincinnati[0.042]; Dallas[0.042]; Buffalo[0.042]; Miami[0.042]; Giants[0.041]; Jets[0.041]; Conference[0.041]; Conference[0.041]; Pittsburgh[0.041]; Indianapolis[0.041]; Washington[0.040]; Arizona[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dallas ==> ENTITY: \u001b[32mDallas Cowboys\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.207:0.207[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.045]; Washington[0.044]; Jacksonville[0.044]; Buffalo[0.043]; Giants[0.043]; Baltimore[0.042]; Carolina[0.042]; Carolina[0.042]; Miami[0.042]; Jets[0.042]; Pittsburgh[0.042]; Oakland[0.041]; City[0.041]; Cincinnati[0.041]; Detroit[0.041]; Detroit[0.041]; Tampa[0.041]; Tampa[0.041]; Playoff[0.041]; Sunday[0.041]; Monday[0.041]; Atlanta[0.040]; Atlanta[0.040]; England[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Arizona ==> ENTITY: \u001b[32mArizona Cardinals\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.199:0.199[0]; log p(e|m)= -4.343:-4.343[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJacksonville[0.044]; Carolina[0.043]; Carolina[0.043]; Giants[0.043]; Houston[0.043]; Jets[0.043]; Oakland[0.042]; Baltimore[0.042]; Dallas[0.042]; Pittsburgh[0.042]; Cincinnati[0.042]; Buffalo[0.042]; Miami[0.041]; Atlanta[0.041]; Atlanta[0.041]; Division[0.041]; Division[0.041]; Monday[0.041]; Tampa[0.040]; Tampa[0.040]; Clinched[0.040]; Clinched[0.040]; Detroit[0.040]; Detroit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Buffalo ==> ENTITY: \u001b[32mBuffalo Bills\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.177:0.177[0]; log p(e|m)= -4.017:-4.017[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJacksonville[0.044]; National[0.044]; Houston[0.044]; won[0.044]; game[0.044]; Jets[0.044]; Division[0.044]; Division[0.044]; Division[0.044]; Baltimore[0.044]; York[0.044]; Pittsburgh[0.043]; Football[0.043]; Football[0.043]; Football[0.043]; Oakland[0.043]; Cincinnati[0.043]; City[0.043]; Miami[0.043]; League[0.043]; Central[0.043]; Denver[0.043]; Indianapolis[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tampa Bay ==> ENTITY: \u001b[32mTampa Bay Buccaneers\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.226:0.226[0]; log p(e|m)= -2.071:-2.071[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJacksonville[0.045]; Carolina[0.043]; Carolina[0.043]; Tampa[0.043]; Oakland[0.043]; Baltimore[0.042]; Philadelphia[0.042]; Buffalo[0.042]; Houston[0.042]; Football[0.041]; Miami[0.041]; Cincinnati[0.041]; Giants[0.041]; Giants[0.041]; Atlanta[0.041]; Atlanta[0.041]; Pittsburgh[0.041]; Dallas[0.041]; Dallas[0.041]; Jets[0.041]; Detroit[0.041]; Playoff[0.040]; Arizona[0.040]; Arizona[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jacksonville ==> ENTITY: \u001b[32mJacksonville Jaguars\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.194:0.194[0]; log p(e|m)= -3.194:-3.194[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNational[0.050]; National[0.050]; Houston[0.042]; League[0.042]; Dallas[0.042]; Oakland[0.041]; Giants[0.041]; standings[0.041]; Philadelphia[0.041]; Cincinnati[0.041]; Miami[0.041]; Buffalo[0.041]; Baltimore[0.041]; Football[0.041]; Football[0.041]; Football[0.041]; Football[0.041]; Pittsburgh[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Jets[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Denver ==> ENTITY: \u001b[32mDenver Broncos\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.192:0.192[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.044]; Miami[0.043]; Jacksonville[0.043]; Baltimore[0.043]; Carolina[0.042]; Carolina[0.042]; Dallas[0.042]; Cincinnati[0.042]; Jets[0.042]; Oakland[0.042]; Giants[0.042]; Pittsburgh[0.042]; Buffalo[0.042]; Central[0.041]; Washington[0.041]; Arizona[0.041]; Sunday[0.041]; Atlanta[0.041]; Atlanta[0.041]; Monday[0.041]; Chicago[0.041]; Chicago[0.041]; England[0.040]; Seattle[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Green Bay ==> ENTITY: \u001b[32mGreen Bay Packers\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.173:0.173[0]; log p(e|m)= -2.033:-2.033[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBay[0.043]; Bay[0.043]; Bay[0.043]; Jacksonville[0.043]; Giants[0.042]; Carolina[0.042]; Carolina[0.042]; Division[0.042]; Division[0.042]; Division[0.042]; Jets[0.042]; Baltimore[0.042]; Houston[0.042]; Dallas[0.041]; Buffalo[0.041]; Oakland[0.041]; Tampa[0.041]; Tampa[0.041]; Pittsburgh[0.041]; Playoff[0.041]; Cincinnati[0.041]; Sunday[0.040]; Miami[0.040]; Washington[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NY JETS\u001b[39m ==> ENTITY: \u001b[32mNew York Jets\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGiants[0.047]; Carolina[0.046]; Carolina[0.046]; Jacksonville[0.045]; Baltimore[0.045]; Buffalo[0.045]; Pittsburgh[0.044]; Oakland[0.044]; Houston[0.044]; Cincinnati[0.044]; Sunday[0.044]; England[0.043]; Miami[0.043]; Monday[0.043]; Dallas[0.043]; Clinched[0.042]; Clinched[0.042]; Arizona[0.042]; Denver[0.042]; Tampa[0.042]; Tampa[0.042]; Detroit[0.042]; Detroit[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New England ==> ENTITY: \u001b[32mNew England Patriots\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.215:0.215[0]; log p(e|m)= -3.324:-3.324[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBuffalo[0.044]; Jacksonville[0.043]; Carolina[0.043]; Carolina[0.043]; Jets[0.043]; Giants[0.043]; Baltimore[0.042]; Pittsburgh[0.042]; Oakland[0.042]; Houston[0.042]; Tampa[0.042]; Tampa[0.042]; Detroit[0.041]; Detroit[0.041]; Miami[0.041]; Atlanta[0.041]; Atlanta[0.041]; Dallas[0.041]; Sunday[0.041]; Cincinnati[0.041]; Washington[0.040]; Denver[0.040]; Playoff[0.040]; Monday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Philadelphia ==> ENTITY: \u001b[32mPhiladelphia Eagles\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.189:0.189[0]; log p(e|m)= -4.510:-4.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGiants[0.044]; Jacksonville[0.044]; Carolina[0.043]; Houston[0.043]; Cincinnati[0.043]; Baltimore[0.043]; Pittsburgh[0.043]; Football[0.043]; Washington[0.043]; Oakland[0.042]; Conference[0.042]; Dallas[0.042]; Arizona[0.041]; Detroit[0.041]; Atlanta[0.041]; Denver[0.041]; National[0.040]; Chicago[0.040]; Tampa[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Miami ==> ENTITY: \u001b[32mMiami Dolphins\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.192:0.192[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJacksonville[0.044]; Jets[0.043]; Conference[0.043]; Conference[0.043]; Baltimore[0.042]; game[0.042]; Houston[0.042]; League[0.042]; Buffalo[0.042]; Dallas[0.042]; Thursday[0.042]; Thursday[0.042]; Football[0.042]; Football[0.042]; Football[0.042]; Football[0.042]; Oakland[0.041]; Cincinnati[0.041]; Pittsburgh[0.041]; Kansas[0.041]; Game[0.041]; Indianapolis[0.041]; City[0.041]; won[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Francisco ==> ENTITY: \u001b[32mSan Francisco 49ers\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.187:0.187[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.043]; Football[0.043]; Giants[0.043]; Giants[0.043]; Bay[0.042]; Bay[0.042]; Bay[0.042]; Bay[0.042]; Houston[0.042]; Oakland[0.042]; Oakland[0.042]; Jacksonville[0.042]; Clinched[0.042]; Clinched[0.042]; Carolina[0.041]; Carolina[0.041]; Jets[0.041]; San[0.041]; San[0.041]; Miami[0.041]; Dallas[0.041]; Dallas[0.041]; Arizona[0.041]; Arizona[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jacksonville ==> ENTITY: \u001b[32mJacksonville Jaguars\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.227:0.227[0]; log p(e|m)= -3.194:-3.194[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCarolina[0.045]; Carolina[0.045]; Houston[0.043]; Dallas[0.043]; Oakland[0.042]; Giants[0.042]; Cincinnati[0.042]; Miami[0.042]; Buffalo[0.042]; Baltimore[0.042]; Jets[0.041]; Pittsburgh[0.041]; Atlanta[0.041]; Atlanta[0.041]; Arizona[0.041]; Clinched[0.041]; Clinched[0.041]; Tampa[0.041]; Tampa[0.041]; Detroit[0.040]; Detroit[0.040]; Denver[0.040]; Sunday[0.040]; Playoff[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Diego ==> ENTITY: \u001b[32mSan Diego Chargers\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.191:0.191[0]; log p(e|m)= -3.817:-3.817[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJacksonville[0.043]; Conference[0.043]; Conference[0.043]; Houston[0.043]; Oakland[0.043]; Bay[0.043]; Bay[0.043]; Giants[0.042]; Football[0.042]; Football[0.042]; Jets[0.042]; Buffalo[0.042]; Cincinnati[0.042]; Philadelphia[0.042]; Dallas[0.042]; Miami[0.041]; Baltimore[0.041]; Pittsburgh[0.041]; Indianapolis[0.041]; Denver[0.041]; Arizona[0.040]; Washington[0.040]; National[0.040]; City[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Buffalo ==> ENTITY: \u001b[32mBuffalo Bills\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.191:0.191[0]; log p(e|m)= -4.017:-4.017[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOrleans[0.044]; Orleans[0.044]; Cincinnati[0.043]; Giants[0.043]; Jacksonville[0.043]; Houston[0.043]; Jets[0.042]; Baltimore[0.042]; Pittsburgh[0.042]; Carolina[0.042]; Carolina[0.042]; Dallas[0.042]; Oakland[0.042]; City[0.041]; Miami[0.041]; Atlanta[0.041]; Atlanta[0.041]; Minnesota[0.041]; Denver[0.041]; Sunday[0.041]; Arizona[0.041]; Tampa[0.040]; Tampa[0.040]; Washington[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Seattle ==> ENTITY: \u001b[32mSeattle Seahawks\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.201:0.201[0]; log p(e|m)= -3.650:-3.650[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJacksonville[0.046]; Washington[0.045]; Carolina[0.044]; Carolina[0.044]; Houston[0.044]; Miami[0.044]; Oakland[0.044]; Cincinnati[0.044]; Baltimore[0.044]; Pittsburgh[0.044]; Jets[0.044]; Giants[0.043]; Western[0.043]; Buffalo[0.043]; Dallas[0.043]; Sunday[0.043]; Monday[0.043]; Division[0.043]; Division[0.043]; Denver[0.042]; Arizona[0.042]; Tampa[0.042]; Tampa[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Arizona ==> ENTITY: \u001b[32mArizona Cardinals\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.181:0.181[0]; log p(e|m)= -4.343:-4.343[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJacksonville[0.044]; Carolina[0.043]; Giants[0.043]; Houston[0.043]; National[0.043]; Oakland[0.043]; Philadelphia[0.042]; Baltimore[0.042]; Dallas[0.042]; Pittsburgh[0.042]; Cincinnati[0.042]; Football[0.042]; Atlanta[0.041]; Conference[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Tampa[0.041]; Clinched[0.040]; Clinched[0.040]; Detroit[0.040]; Kansas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Carolina ==> ENTITY: \u001b[32mCarolina Panthers\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.217:0.217[0]; log p(e|m)= -2.283:-2.283[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPlayoff[0.049]; Carolina[0.046]; Jacksonville[0.046]; Giants[0.045]; Giants[0.045]; Jets[0.044]; Football[0.044]; Oakland[0.044]; Houston[0.043]; Baltimore[0.043]; Buffalo[0.043]; Pittsburgh[0.043]; Sunday[0.043]; Francisco[0.043]; Francisco[0.043]; Tampa[0.042]; Tampa[0.042]; Miami[0.042]; Monday[0.042]; Philadelphia[0.042]; Cincinnati[0.042]; Atlanta[0.042]; Atlanta[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cincinnati ==> ENTITY: \u001b[32mCincinnati Bengals\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.184:0.184[0]; log p(e|m)= -3.963:-3.963[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConference[0.044]; Conference[0.044]; game[0.043]; Division[0.043]; Division[0.043]; Division[0.043]; Division[0.043]; Division[0.043]; Football[0.042]; Football[0.042]; Football[0.042]; Football[0.042]; Giants[0.042]; Jacksonville[0.041]; Oakland[0.041]; Philadelphia[0.041]; won[0.041]; Houston[0.041]; League[0.041]; Jets[0.040]; Pittsburgh[0.040]; Miami[0.040]; Baltimore[0.040]; Buffalo[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NY GIANTS ==> ENTITY: \u001b[32mNew York Giants\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.189:0.189[0]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJets[0.045]; Baltimore[0.044]; Jacksonville[0.044]; Carolina[0.043]; Carolina[0.043]; Buffalo[0.043]; Oakland[0.043]; Cincinnati[0.043]; Houston[0.043]; Pittsburgh[0.043]; Miami[0.042]; Arizona[0.042]; Dallas[0.041]; Denver[0.041]; Sunday[0.041]; Detroit[0.041]; Detroit[0.041]; Monday[0.040]; England[0.040]; Atlanta[0.040]; Atlanta[0.040]; Minnesota[0.040]; Minnesota[0.040]; Kansas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Miami ==> ENTITY: \u001b[32mMiami Dolphins\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.216:0.216[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJacksonville[0.044]; Jets[0.044]; Giants[0.043]; Carolina[0.043]; Carolina[0.043]; Baltimore[0.043]; Houston[0.042]; Buffalo[0.042]; Tampa[0.042]; Tampa[0.042]; Dallas[0.042]; Monday[0.042]; Oakland[0.041]; Sunday[0.041]; Cincinnati[0.041]; Pittsburgh[0.041]; Kansas[0.041]; City[0.041]; Detroit[0.040]; Detroit[0.040]; Atlanta[0.040]; Atlanta[0.040]; Arizona[0.040]; Denver[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New England ==> ENTITY: \u001b[32mNew England Patriots\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.170:0.170[0]; log p(e|m)= -3.324:-3.324[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJacksonville[0.043]; League[0.043]; Division[0.043]; Division[0.043]; Division[0.043]; Jets[0.043]; Baltimore[0.042]; Pittsburgh[0.042]; Buffalo[0.042]; Oakland[0.042]; tied[0.042]; game[0.042]; Standings[0.041]; Houston[0.041]; Football[0.041]; Football[0.041]; Football[0.041]; Miami[0.041]; National[0.041]; Indianapolis[0.041]; Cincinnati[0.041]; standings[0.040]; Eastern[0.040]; York[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Carolina ==> ENTITY: \u001b[32mCarolina Panthers\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.191:0.191[0]; log p(e|m)= -2.283:-2.283[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCarolina[0.048]; San[0.045]; San[0.045]; San[0.045]; Jacksonville[0.043]; Giants[0.042]; Jets[0.042]; Miami[0.042]; Oakland[0.041]; Diego[0.041]; Houston[0.041]; Baltimore[0.041]; Buffalo[0.041]; Francisco[0.041]; Francisco[0.041]; Pittsburgh[0.041]; Sunday[0.041]; Tampa[0.040]; Tampa[0.040]; Monday[0.040]; Cincinnati[0.040]; Atlanta[0.040]; Atlanta[0.040]; Arizona[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Francisco ==> ENTITY: \u001b[32mSan Francisco 49ers\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.179:0.179[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGiants[0.043]; Bay[0.043]; Bay[0.043]; Bay[0.043]; Houston[0.043]; Oakland[0.042]; Jacksonville[0.042]; Clinched[0.042]; Clinched[0.042]; Carolina[0.042]; Carolina[0.042]; Jets[0.042]; San[0.041]; San[0.041]; Miami[0.041]; Dallas[0.041]; Arizona[0.041]; Diego[0.041]; Francisco[0.041]; Baltimore[0.041]; Cincinnati[0.041]; Washington[0.041]; Monday[0.041]; Buffalo[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew York\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew York City\u001b[39m <---> \u001b[32mNew York\u001b[39m\t\nSCORES: global= 0.246:0.240[\u001b[31m0.006\u001b[39m]; local()= 0.066:0.056[\u001b[31m0.009\u001b[39m]; log p(e|m)= -1.790:-0.289[\u001b[32m1.501\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.045]; game[0.045]; Thursday[0.044]; Thursday[0.044]; Miami[0.043]; Conference[0.043]; won[0.043]; Game[0.042]; Houston[0.042]; Baltimore[0.041]; National[0.041]; England[0.041]; Football[0.041]; Football[0.041]; Football[0.041]; Buffalo[0.041]; Central[0.041]; American[0.041]; American[0.041]; lost[0.041]; Pittsburgh[0.040]; Division[0.040]; Division[0.040]; Cincinnati[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NY GIANTS ==> ENTITY: \u001b[32mNew York Giants\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.158:0.158[0]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBaltimore[0.044]; Jacksonville[0.044]; Philadelphia[0.044]; Football[0.044]; Carolina[0.044]; Oakland[0.043]; Cincinnati[0.043]; Houston[0.043]; Arizona[0.042]; Dallas[0.042]; Denver[0.042]; Sunday[0.042]; Detroit[0.041]; Eastern[0.041]; Atlanta[0.041]; Minnesota[0.040]; Kansas[0.040]; Washington[0.040]; Tampa[0.040]; National[0.040]; Louis[0.040]; Louis[0.040]; Chicago[0.040]; Chicago[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NY JETS\u001b[39m ==> ENTITY: \u001b[32mNew York Jets\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJacksonville[0.046]; Baltimore[0.046]; Buffalo[0.045]; game[0.045]; Pittsburgh[0.044]; Oakland[0.044]; Houston[0.044]; Cincinnati[0.044]; Football[0.044]; Football[0.044]; Football[0.044]; Football[0.044]; York[0.044]; England[0.044]; Miami[0.043]; Indianapolis[0.043]; Dallas[0.043]; League[0.042]; Thursday[0.042]; Thursday[0.042]; Denver[0.042]; tied[0.042]; Washington[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlanta ==> ENTITY: \u001b[32mAtlanta Falcons\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.216:0.216[0]; log p(e|m)= -3.963:-3.963[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.046]; Jacksonville[0.046]; Carolina[0.045]; Carolina[0.045]; Atlanta[0.045]; Cincinnati[0.045]; Giants[0.044]; Giants[0.044]; Houston[0.044]; Pittsburgh[0.044]; Miami[0.044]; Diego[0.044]; Baltimore[0.043]; Oakland[0.043]; Jets[0.043]; Buffalo[0.043]; Detroit[0.043]; Detroit[0.043]; Dallas[0.042]; Dallas[0.042]; Washington[0.042]; Washington[0.042]; Western[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlanta ==> ENTITY: \u001b[32mAtlanta Falcons\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.199:0.199[0]; log p(e|m)= -3.963:-3.963[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJacksonville[0.046]; Carolina[0.045]; Carolina[0.045]; Atlanta[0.045]; Cincinnati[0.045]; Giants[0.044]; Houston[0.044]; Pittsburgh[0.044]; Miami[0.044]; Diego[0.044]; Baltimore[0.044]; Oakland[0.043]; Jets[0.043]; Buffalo[0.043]; Detroit[0.043]; Detroit[0.043]; Dallas[0.043]; Washington[0.042]; Western[0.042]; Francisco[0.042]; Francisco[0.042]; Chicago[0.041]; Chicago[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Orleans ==> ENTITY: \u001b[32mNew Orleans Saints\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.204:0.204[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.044]; Jacksonville[0.044]; Carolina[0.043]; Carolina[0.043]; Miami[0.043]; Baltimore[0.042]; Dallas[0.042]; Atlanta[0.042]; Atlanta[0.042]; Detroit[0.042]; Detroit[0.042]; Giants[0.041]; Tampa[0.041]; Tampa[0.041]; Orleans[0.041]; Division[0.041]; Division[0.041]; Monday[0.041]; Jets[0.041]; Oakland[0.041]; Buffalo[0.041]; Cincinnati[0.041]; Sunday[0.040]; Washington[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oakland ==> ENTITY: \u001b[32mOakland Raiders\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.177:0.177[0]; log p(e|m)= -3.912:-3.912[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPittsburgh[0.044]; Jacksonville[0.044]; Baltimore[0.043]; Football[0.043]; Houston[0.043]; Giants[0.043]; Cincinnati[0.042]; Jets[0.042]; Philadelphia[0.042]; Dallas[0.042]; Detroit[0.041]; Miami[0.041]; Bay[0.041]; Bay[0.041]; Central[0.041]; Central[0.041]; Buffalo[0.041]; Indianapolis[0.041]; Arizona[0.041]; City[0.040]; Denver[0.040]; Washington[0.040]; Seattle[0.040]; Tampa[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Minnesota ==> ENTITY: \u001b[32mMinnesota Vikings\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.200:0.200[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGiants[0.043]; Jacksonville[0.043]; Oakland[0.043]; Jets[0.043]; Detroit[0.043]; Baltimore[0.043]; Cincinnati[0.043]; Buffalo[0.042]; Houston[0.042]; Monday[0.042]; Carolina[0.042]; Carolina[0.042]; Pittsburgh[0.042]; Playoff[0.042]; Miami[0.042]; Dallas[0.041]; Sunday[0.041]; Washington[0.040]; Denver[0.040]; Atlanta[0.040]; Atlanta[0.040]; Division[0.040]; Division[0.040]; Tampa[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Detroit ==> ENTITY: \u001b[32mDetroit Lions\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.207:0.207[0]; log p(e|m)= -4.343:-4.343[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGiants[0.044]; Cincinnati[0.044]; Buffalo[0.043]; Jacksonville[0.043]; Playoff[0.043]; Oakland[0.043]; Jets[0.042]; Carolina[0.042]; Carolina[0.042]; Pittsburgh[0.042]; Houston[0.042]; Baltimore[0.041]; Miami[0.041]; Dallas[0.041]; Tampa[0.041]; December[0.041]; December[0.041]; Division[0.041]; Division[0.041]; Monday[0.041]; Arizona[0.041]; Sunday[0.040]; Atlanta[0.040]; Atlanta[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indianapolis ==> ENTITY: \u001b[32mIndianapolis Colts\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.181:0.181[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJacksonville[0.046]; game[0.046]; Baltimore[0.043]; won[0.043]; Cincinnati[0.043]; Houston[0.043]; Jets[0.042]; Pittsburgh[0.042]; Oakland[0.042]; Buffalo[0.041]; Miami[0.041]; Football[0.041]; Football[0.041]; Football[0.041]; Football[0.041]; City[0.041]; tied[0.041]; Thursday[0.041]; Thursday[0.041]; League[0.040]; Denver[0.040]; lost[0.040]; American[0.040]; American[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Minnesota ==> ENTITY: \u001b[32mMinnesota Vikings\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.204:0.204[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGiants[0.047]; Giants[0.047]; Jacksonville[0.047]; Football[0.047]; Oakland[0.047]; Detroit[0.046]; Baltimore[0.046]; Philadelphia[0.046]; Cincinnati[0.046]; Houston[0.046]; Carolina[0.046]; National[0.045]; Conference[0.045]; Miami[0.045]; Dallas[0.045]; Sunday[0.044]; Central[0.044]; Washington[0.044]; Denver[0.044]; Denver[0.044]; Atlanta[0.044]; Atlanta[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = National Football League\u001b[39m ==> ENTITY: \u001b[32mNational Football League\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwon[0.046]; game[0.045]; Jacksonville[0.045]; Football[0.044]; Football[0.044]; Cincinnati[0.044]; Baltimore[0.044]; Pittsburgh[0.044]; Houston[0.044]; Division[0.044]; Division[0.044]; Division[0.044]; Buffalo[0.043]; Jets[0.043]; standings[0.043]; Miami[0.043]; Standings[0.043]; Indianapolis[0.042]; York[0.042]; Thursday[0.042]; Thursday[0.042]; Conference[0.042]; England[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Orleans ==> ENTITY: \u001b[32mNew Orleans Saints\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.203:0.203[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.044]; Jacksonville[0.044]; Carolina[0.043]; Carolina[0.043]; Miami[0.043]; Philadelphia[0.042]; Baltimore[0.042]; Dallas[0.042]; Atlanta[0.041]; Atlanta[0.041]; Detroit[0.041]; Detroit[0.041]; Giants[0.041]; Giants[0.041]; Tampa[0.041]; Tampa[0.041]; Orleans[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Monday[0.041]; Jets[0.041]; Oakland[0.041]; Buffalo[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Baltimore ==> ENTITY: \u001b[32mBaltimore Ravens\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.205:0.205[0]; log p(e|m)= -4.075:-4.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJacksonville[0.045]; Houston[0.043]; Cincinnati[0.043]; Pittsburgh[0.043]; Philadelphia[0.043]; League[0.042]; Washington[0.042]; won[0.042]; Indianapolis[0.042]; Buffalo[0.042]; Giants[0.042]; Oakland[0.042]; Jets[0.042]; Miami[0.042]; National[0.041]; National[0.041]; game[0.041]; Dallas[0.040]; Football[0.040]; Football[0.040]; Football[0.040]; England[0.040]; Arizona[0.040]; Denver[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Diego ==> ENTITY: \u001b[32mSan Diego Chargers\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.196:0.196[0]; log p(e|m)= -3.817:-3.817[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJacksonville[0.043]; Houston[0.043]; Oakland[0.043]; Bay[0.043]; Bay[0.043]; Bay[0.043]; Giants[0.042]; December[0.042]; December[0.042]; Carolina[0.042]; Carolina[0.042]; Jets[0.042]; Pittsburgh[0.042]; Buffalo[0.042]; Cincinnati[0.042]; Monday[0.041]; Dallas[0.041]; Miami[0.041]; Baltimore[0.041]; Denver[0.040]; Tampa[0.040]; Tampa[0.040]; Sunday[0.040]; Arizona[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pittsburgh ==> ENTITY: \u001b[32mPittsburgh Steelers\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.193:0.193[0]; log p(e|m)= -3.963:-3.963[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCarolina[0.044]; Carolina[0.044]; Jacksonville[0.044]; Cincinnati[0.043]; Buffalo[0.043]; Jets[0.042]; Playoff[0.042]; Giants[0.042]; Sunday[0.042]; Baltimore[0.042]; Oakland[0.041]; Houston[0.041]; Detroit[0.041]; Detroit[0.041]; Atlanta[0.041]; Atlanta[0.041]; Washington[0.041]; Tampa[0.041]; Tampa[0.041]; Miami[0.040]; Clinched[0.040]; Clinched[0.040]; Denver[0.040]; Arizona[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Seattle ==> ENTITY: \u001b[32mSeattle Seahawks\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.183:0.183[0]; log p(e|m)= -3.650:-3.650[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJacksonville[0.044]; Conference[0.044]; Washington[0.043]; Football[0.042]; Houston[0.042]; Miami[0.042]; Cincinnati[0.042]; Oakland[0.042]; Baltimore[0.042]; Pittsburgh[0.042]; Jets[0.042]; Giants[0.042]; Philadelphia[0.041]; Western[0.041]; Western[0.041]; Buffalo[0.041]; Dallas[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Denver[0.041]; Arizona[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kansas City ==> ENTITY: \u001b[32mKansas City Chiefs\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.210:0.210[0]; log p(e|m)= -3.170:-3.170[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJacksonville[0.046]; Houston[0.046]; Baltimore[0.045]; Buffalo[0.045]; Oakland[0.045]; Jets[0.045]; Carolina[0.044]; Carolina[0.044]; Cincinnati[0.044]; Dallas[0.044]; Miami[0.044]; Giants[0.044]; Pittsburgh[0.043]; Monday[0.043]; Washington[0.043]; Arizona[0.042]; Denver[0.042]; England[0.042]; Detroit[0.042]; Sunday[0.042]; Atlanta[0.041]; Atlanta[0.041]; Chicago[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oakland ==> ENTITY: \u001b[32mOakland Raiders\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.191:0.191[0]; log p(e|m)= -3.912:-3.912[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPittsburgh[0.045]; Jacksonville[0.044]; Carolina[0.043]; Carolina[0.043]; Houston[0.043]; Giants[0.043]; Cincinnati[0.043]; Jets[0.043]; Baltimore[0.042]; Dallas[0.042]; Detroit[0.042]; Miami[0.042]; Buffalo[0.041]; Arizona[0.041]; Bay[0.041]; Bay[0.041]; City[0.041]; Denver[0.041]; Washington[0.040]; Atlanta[0.040]; Atlanta[0.040]; Seattle[0.040]; Sunday[0.040]; Monday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago Bears\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.189:0.189[0]; log p(e|m)= -4.828:-4.828[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGiants[0.043]; Giants[0.043]; Jacksonville[0.043]; Carolina[0.043]; Chicago[0.043]; Football[0.042]; Philadelphia[0.042]; Playoff[0.042]; Conference[0.042]; Baltimore[0.042]; Miami[0.042]; Pittsburgh[0.042]; Houston[0.042]; Oakland[0.041]; Cincinnati[0.041]; Detroit[0.041]; Dallas[0.041]; Dallas[0.041]; Louis[0.041]; Louis[0.041]; Tampa[0.041]; Tampa[0.041]; Washington[0.040]; Washington[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Houston ==> ENTITY: \u001b[32mHouston Texans\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.205:0.205[0]; log p(e|m)= -4.828:-4.828[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; Jacksonville[0.043]; Conference[0.043]; Conference[0.043]; Dallas[0.043]; Football[0.042]; Football[0.042]; Football[0.042]; Football[0.042]; game[0.042]; Baltimore[0.042]; Cincinnati[0.042]; Giants[0.042]; Philadelphia[0.041]; Jets[0.041]; Oakland[0.041]; Buffalo[0.041]; won[0.041]; Indianapolis[0.041]; Pittsburgh[0.041]; Miami[0.041]; Thursday[0.040]; Thursday[0.040]; City[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago Bears\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.205:0.205[0]; log p(e|m)= -4.828:-4.828[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGiants[0.043]; Giants[0.043]; Buffalo[0.043]; Jacksonville[0.043]; Carolina[0.043]; Carolina[0.043]; Chicago[0.043]; Jets[0.042]; Playoff[0.042]; Baltimore[0.042]; Miami[0.042]; Pittsburgh[0.042]; Houston[0.042]; Oakland[0.041]; Cincinnati[0.041]; Detroit[0.041]; Detroit[0.041]; Dallas[0.041]; Louis[0.040]; Louis[0.040]; Tampa[0.040]; Tampa[0.040]; Washington[0.040]; Monday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NFL ==> ENTITY: \u001b[32mNational Football League\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.171:0.171[0]; log p(e|m)= -0.019:-0.019[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.047]; won[0.044]; game[0.043]; Football[0.043]; Football[0.043]; Football[0.043]; Pittsburgh[0.042]; Cincinnati[0.042]; Jacksonville[0.042]; Houston[0.042]; Division[0.042]; Division[0.042]; Buffalo[0.041]; Jets[0.041]; Miami[0.041]; National[0.041]; Indianapolis[0.041]; York[0.040]; Thursday[0.040]; Thursday[0.040]; American[0.040]; American[0.040]; England[0.040]; Conference[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAmerican\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAmerican football\u001b[39m <---> \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.245:0.244[\u001b[31m0.001\u001b[39m]; local()= 0.159:0.064[\u001b[31m0.096\u001b[39m]; log p(e|m)= -4.962:-0.130[\u001b[32m4.832\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.046]; Football[0.045]; Football[0.045]; Football[0.045]; League[0.045]; game[0.043]; won[0.043]; Cincinnati[0.042]; Jacksonville[0.042]; Baltimore[0.042]; England[0.042]; Houston[0.042]; National[0.041]; Pittsburgh[0.041]; Indianapolis[0.041]; Buffalo[0.040]; Standings[0.040]; Jets[0.040]; Miami[0.040]; City[0.040]; York[0.039]; standings[0.039]; Thursday[0.039]; Thursday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tampa Bay ==> ENTITY: \u001b[32mTampa Bay Buccaneers\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.226:0.226[0]; log p(e|m)= -2.071:-2.071[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJacksonville[0.045]; Carolina[0.044]; Carolina[0.044]; Tampa[0.043]; Oakland[0.043]; Baltimore[0.043]; Buffalo[0.042]; Houston[0.042]; Giants[0.042]; Miami[0.042]; Cincinnati[0.042]; Atlanta[0.041]; Atlanta[0.041]; Pittsburgh[0.041]; Jets[0.041]; Dallas[0.041]; Detroit[0.041]; Detroit[0.041]; Playoff[0.040]; Arizona[0.040]; Sunday[0.040]; Monday[0.040]; Bay[0.040]; Bay[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dallas ==> ENTITY: \u001b[32mDallas Cowboys\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.196:0.196[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.045]; Washington[0.044]; Conference[0.044]; Jacksonville[0.044]; Giants[0.043]; Baltimore[0.043]; Carolina[0.042]; Miami[0.042]; Football[0.042]; Jets[0.042]; Philadelphia[0.042]; Pittsburgh[0.042]; Oakland[0.041]; City[0.041]; Cincinnati[0.041]; Detroit[0.041]; Tampa[0.041]; Central[0.041]; Central[0.041]; Atlanta[0.040]; Arizona[0.040]; Denver[0.040]; Chicago[0.040]; Minnesota[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Detroit ==> ENTITY: \u001b[32mDetroit Lions\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.194:0.194[0]; log p(e|m)= -4.343:-4.343[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGiants[0.043]; Giants[0.043]; Cincinnati[0.043]; Conference[0.043]; Philadelphia[0.043]; Jacksonville[0.043]; Playoff[0.042]; Oakland[0.042]; Jets[0.042]; Carolina[0.042]; Football[0.042]; Pittsburgh[0.042]; Houston[0.042]; Baltimore[0.041]; Miami[0.041]; Dallas[0.041]; Dallas[0.041]; Tampa[0.041]; Tampa[0.041]; December[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; \t\n\r [===>..........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 17s387ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 440/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1367testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Campese\u001b[39m ==> ENTITY: \u001b[32mDavid Campese\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWallaby[0.046]; rugby[0.046]; Campese[0.045]; Campese[0.045]; Australia[0.045]; England[0.043]; Rugby[0.043]; Lynagh[0.043]; Michael[0.043]; winger[0.043]; David[0.042]; jersey[0.042]; game[0.041]; played[0.040]; match[0.040]; play[0.040]; said[0.040]; obviously[0.040]; Saracens[0.039]; think[0.039]; Pienaar[0.038]; opportunity[0.038]; final[0.038]; club[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.047]; England[0.045]; England[0.045]; England[0.045]; club[0.045]; match[0.043]; clubs[0.043]; final[0.043]; played[0.042]; game[0.042]; play[0.042]; play[0.042]; Rugby[0.041]; David[0.041]; Wales[0.040]; contract[0.040]; Michael[0.040]; Saturday[0.040]; currently[0.039]; South[0.039]; obviously[0.039]; home[0.039]; target[0.039]; rugby[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Campese\u001b[39m ==> ENTITY: \u001b[32mDavid Campese\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.215:0.215[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWallaby[0.046]; Wallaby[0.046]; Campese[0.045]; Campese[0.045]; England[0.043]; England[0.043]; Lynagh[0.043]; Michael[0.042]; winger[0.042]; Wales[0.042]; David[0.042]; jersey[0.041]; jersey[0.041]; Jones[0.041]; captain[0.041]; game[0.041]; Nick[0.040]; believes[0.040]; played[0.040]; got[0.040]; match[0.040]; play[0.040]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New South Wales ==> ENTITY: \u001b[32mNew South Wales rugby league team\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.195:0.195[0]; log p(e|m)= -2.397:-2.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.046]; club[0.045]; Wallaby[0.044]; Wallaby[0.044]; jersey[0.044]; jersey[0.044]; match[0.043]; game[0.042]; played[0.042]; clubs[0.042]; England[0.042]; England[0.042]; Campese[0.042]; Campese[0.042]; Campese[0.042]; play[0.042]; Super[0.041]; captain[0.041]; Saracens[0.040]; final[0.039]; contract[0.039]; currently[0.038]; Jones[0.038]; Saturday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Saracens ==> ENTITY: \u001b[32mSaracens F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.203:0.203[0]; log p(e|m)= -0.567:-0.567[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.049]; rugby[0.049]; England[0.045]; England[0.045]; England[0.045]; club[0.043]; played[0.041]; Wallaby[0.041]; Wallaby[0.041]; final[0.041]; clubs[0.041]; Super[0.041]; people[0.041]; jersey[0.040]; Saturday[0.040]; match[0.040]; Wales[0.040]; game[0.040]; London[0.040]; London[0.040]; contract[0.040]; captain[0.040]; play[0.040]; play[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAustralia\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustralia national rugby union team\u001b[39m <---> \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.246:0.238[\u001b[31m0.008\u001b[39m]; local()= 0.220:0.095[\u001b[31m0.125\u001b[39m]; log p(e|m)= -4.200:-0.172[\u001b[32m4.028\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.049]; Rugby[0.048]; England[0.045]; Wallaby[0.045]; Wallaby[0.045]; Wales[0.044]; match[0.042]; Campese[0.041]; Campese[0.041]; Campese[0.041]; jersey[0.041]; Super[0.041]; captain[0.041]; played[0.040]; Saracens[0.040]; South[0.040]; game[0.040]; club[0.040]; final[0.039]; play[0.039]; play[0.039]; winger[0.039]; David[0.039]; contract[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wallaby ==> ENTITY: \u001b[32mAustralia national rugby union team\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.210:0.210[0]; log p(e|m)= -0.562:-0.562[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.046]; Wallaby[0.046]; Wales[0.045]; England[0.044]; England[0.044]; match[0.043]; jersey[0.042]; jersey[0.042]; Super[0.042]; played[0.041]; Saracens[0.041]; game[0.041]; Campese[0.041]; Campese[0.041]; Campese[0.041]; captain[0.041]; South[0.041]; club[0.040]; final[0.040]; play[0.040]; contract[0.039]; winger[0.039]; clubs[0.039]; Lynagh[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Francois Pienaar\u001b[39m ==> ENTITY: \u001b[32mFrancois Pienaar\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.045]; England[0.045]; England[0.045]; Rugby[0.044]; rugby[0.044]; Wales[0.043]; Australia[0.043]; match[0.042]; jersey[0.042]; jersey[0.042]; played[0.042]; Saracens[0.041]; game[0.041]; South[0.041]; play[0.041]; play[0.041]; David[0.041]; David[0.041]; delighted[0.040]; Michael[0.040]; captain[0.040]; Wallaby[0.040]; Wallaby[0.040]; believes[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rugby Union\u001b[39m ==> ENTITY: \u001b[32mRugby union\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.054]; England[0.047]; Australia[0.045]; Wallaby[0.044]; club[0.043]; clubs[0.043]; Campese[0.043]; Campese[0.043]; Campese[0.043]; played[0.043]; match[0.042]; Saracens[0.042]; play[0.042]; jersey[0.041]; game[0.041]; final[0.039]; Lynagh[0.039]; Saturday[0.039]; winger[0.039]; David[0.038]; London[0.038]; London[0.038]; Michael[0.038]; currently[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = David Campese\u001b[39m ==> ENTITY: \u001b[32mDavid Campese\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWallaby[0.047]; rugby[0.046]; Campese[0.046]; Campese[0.046]; Australia[0.045]; England[0.044]; Rugby[0.044]; Lynagh[0.043]; Michael[0.043]; winger[0.043]; jersey[0.042]; game[0.041]; played[0.041]; match[0.040]; play[0.040]; said[0.040]; obviously[0.040]; Saracens[0.039]; think[0.039]; Pienaar[0.039]; opportunity[0.038]; final[0.038]; club[0.038]; chasing[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nick Farr-Jones\u001b[39m ==> ENTITY: \u001b[32mNick Farr-Jones\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWales[0.046]; Australia[0.046]; England[0.045]; England[0.045]; Michael[0.043]; David[0.042]; match[0.042]; Campese[0.042]; Campese[0.042]; Campese[0.042]; Wallaby[0.042]; Wallaby[0.042]; played[0.041]; Lynagh[0.041]; South[0.040]; captain[0.040]; believes[0.040]; game[0.040]; play[0.040]; final[0.040]; Jones[0.040]; Saracens[0.039]; Pienaar[0.039]; think[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Michael Lynagh\u001b[39m ==> ENTITY: \u001b[32mMichael Lynagh\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.046]; Wallaby[0.044]; Wallaby[0.044]; Rugby[0.044]; Wales[0.042]; Campese[0.042]; Campese[0.042]; Campese[0.042]; Campese[0.042]; Campese[0.042]; England[0.041]; England[0.041]; England[0.041]; Australia[0.041]; Super[0.041]; game[0.041]; Saracens[0.041]; played[0.040]; jersey[0.040]; jersey[0.040]; David[0.040]; David[0.040]; club[0.040]; club[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wallaby ==> ENTITY: \u001b[32mAustralia national rugby union team\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.229:0.229[0]; log p(e|m)= -0.562:-0.562[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.049]; Rugby[0.048]; Australia[0.045]; Wallaby[0.045]; Wales[0.044]; England[0.043]; England[0.043]; match[0.042]; jersey[0.041]; Super[0.041]; played[0.041]; Saracens[0.041]; game[0.040]; Campese[0.040]; Campese[0.040]; Campese[0.040]; Campese[0.040]; captain[0.040]; South[0.040]; club[0.040]; final[0.039]; play[0.039]; play[0.039]; contract[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.090:0.090[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.050]; England[0.048]; club[0.048]; match[0.046]; clubs[0.046]; final[0.046]; played[0.045]; game[0.045]; play[0.044]; Rugby[0.043]; David[0.043]; Michael[0.042]; Saturday[0.042]; currently[0.042]; obviously[0.042]; target[0.041]; rugby[0.041]; Australia[0.041]; Union[0.041]; opportunity[0.041]; winger[0.041]; consider[0.040]; opportunities[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Philippe Sella\u001b[39m ==> ENTITY: \u001b[32mPhilippe Sella\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.044]; Rugby[0.044]; Lynagh[0.044]; Wallaby[0.043]; Wallaby[0.043]; Campese[0.042]; Campese[0.042]; Campese[0.042]; Campese[0.042]; Campese[0.042]; jersey[0.042]; jersey[0.042]; club[0.042]; club[0.042]; play[0.041]; play[0.041]; played[0.041]; game[0.041]; Saracens[0.041]; Michael[0.040]; Pienaar[0.040]; Nick[0.040]; captain[0.040]; winger[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = David Campese\u001b[39m ==> ENTITY: \u001b[32mDavid Campese\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWallaby[0.047]; Campese[0.046]; Campese[0.046]; England[0.044]; England[0.044]; Lynagh[0.044]; Michael[0.043]; Wales[0.043]; jersey[0.042]; Jones[0.042]; captain[0.042]; Nick[0.041]; believes[0.041]; got[0.041]; play[0.041]; said[0.040]; said[0.040]; obviously[0.040]; Saracens[0.040]; think[0.039]; South[0.039]; Pienaar[0.039]; opportunity[0.039]; delighted[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Campese\u001b[39m ==> ENTITY: \u001b[32mDavid Campese\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.226:0.226[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWallaby[0.045]; Wallaby[0.045]; rugby[0.044]; Campese[0.044]; Campese[0.044]; Campese[0.044]; Campese[0.044]; Australia[0.043]; England[0.042]; England[0.042]; England[0.042]; Lynagh[0.041]; Michael[0.041]; winger[0.041]; Wales[0.041]; David[0.041]; David[0.041]; jersey[0.040]; jersey[0.040]; Jones[0.040]; captain[0.040]; game[0.040]; Nick[0.039]; believes[0.039]; \t\n\r [===>..........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 17s294ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 460/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1383testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = LA CLIPPERS\u001b[39m ==> ENTITY: \u001b[32mLos Angeles Clippers\u001b[39m\t\nSCORES: global= 0.317:0.317[0]; local()= 0.228:0.228[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.046]; Miami[0.046]; Houston[0.046]; Houston[0.046]; Clippers[0.044]; Phoenix[0.044]; Toronto[0.044]; Toronto[0.044]; Denver[0.044]; Denver[0.044]; Dallas[0.043]; Dallas[0.043]; Charlotte[0.043]; Charlotte[0.043]; Atlanta[0.043]; Chicago[0.043]; Washington[0.042]; Lakers[0.042]; Sacramento[0.042]; Conference[0.042]; Detroit[0.042]; Portland[0.042]; December[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Golden State ==> ENTITY: \u001b[32mGolden State Warriors\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.187:0.187[0]; log p(e|m)= -0.742:-0.742[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.044]; Portland[0.044]; Lakers[0.044]; Clippers[0.043]; Clippers[0.043]; Division[0.043]; Division[0.043]; Houston[0.042]; Houston[0.042]; Conference[0.042]; Miami[0.042]; Western[0.041]; Charlotte[0.041]; Charlotte[0.041]; San[0.041]; Phoenix[0.041]; Pacific[0.041]; Sacramento[0.041]; Dallas[0.041]; Dallas[0.041]; Washington[0.040]; Toronto[0.040]; Toronto[0.040]; Detroit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Detroit ==> ENTITY: \u001b[32mDetroit Pistons\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.216:0.216[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; games[0.047]; Basketball[0.044]; Basketball[0.044]; Cleveland[0.043]; teams[0.042]; Conference[0.042]; Conference[0.042]; Philadelphia[0.042]; Vancouver[0.042]; Toronto[0.042]; Boston[0.041]; Houston[0.041]; played[0.041]; Charlotte[0.041]; Miami[0.040]; Dallas[0.040]; Friday[0.040]; Friday[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Orlando[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vancouver ==> ENTITY: \u001b[32mMemphis Grizzlies\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.205:0.205[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVancouver[0.049]; Toronto[0.045]; Charlotte[0.043]; Houston[0.043]; Miami[0.043]; Clippers[0.042]; Clippers[0.042]; Phoenix[0.042]; Detroit[0.042]; Philadelphia[0.042]; Washington[0.042]; Seattle[0.041]; Seattle[0.041]; Portland[0.041]; Atlanta[0.041]; Chicago[0.041]; Lakers[0.041]; Dallas[0.040]; Denver[0.040]; Denver[0.040]; December[0.040]; Pacific[0.040]; Utah[0.040]; Sacramento[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dallas ==> ENTITY: \u001b[32mDallas Mavericks\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.189:0.189[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.046]; Washington[0.045]; Toronto[0.043]; Miami[0.043]; Vancouver[0.043]; Vancouver[0.043]; Philadelphia[0.043]; York[0.042]; Clippers[0.042]; Clippers[0.042]; Detroit[0.042]; Phoenix[0.042]; Charlotte[0.042]; Lakers[0.041]; Atlanta[0.041]; Denver[0.041]; Denver[0.041]; Chicago[0.041]; Saturday[0.040]; Utah[0.040]; December[0.040]; Golden[0.040]; Sacramento[0.040]; Division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York Knicks\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.211:0.211[0]; log p(e|m)= -4.828:-4.828[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.047]; Basketball[0.047]; games[0.044]; games[0.044]; Philadelphia[0.043]; Cleveland[0.043]; Boston[0.042]; teams[0.042]; Miami[0.042]; Conference[0.041]; Conference[0.041]; Friday[0.041]; Friday[0.041]; played[0.041]; Houston[0.041]; Washington[0.041]; Toronto[0.041]; York[0.040]; Charlotte[0.040]; Orlando[0.040]; Detroit[0.040]; National[0.039]; Jersey[0.039]; Central[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Charlotte ==> ENTITY: \u001b[32mNew Orleans Pelicans\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.236:0.236[0]; log p(e|m)= -4.828:-4.828[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.044]; games[0.044]; games[0.044]; Miami[0.043]; Conference[0.043]; Conference[0.043]; Cleveland[0.042]; teams[0.042]; Philadelphia[0.042]; Atlanta[0.042]; Orlando[0.042]; Dallas[0.042]; Division[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Central[0.041]; Boston[0.041]; Detroit[0.041]; Clippers[0.041]; Toronto[0.040]; Washington[0.040]; Chicago[0.040]; Friday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Antonio ==> ENTITY: \u001b[32mSan Antonio Spurs\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.215:0.215[0]; log p(e|m)= -2.865:-2.865[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConference[0.044]; Houston[0.044]; Houston[0.044]; Dallas[0.043]; Dallas[0.043]; Cleveland[0.043]; Philadelphia[0.042]; Charlotte[0.042]; Charlotte[0.042]; Miami[0.042]; Phoenix[0.042]; Detroit[0.042]; Detroit[0.042]; Boston[0.041]; Lakers[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Atlanta[0.040]; Atlanta[0.040]; Toronto[0.040]; Toronto[0.040]; Clippers[0.040]; Clippers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Boston ==> ENTITY: \u001b[32mBoston Celtics\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.213:0.213[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; games[0.046]; York[0.045]; York[0.045]; Basketball[0.044]; Basketball[0.044]; Cleveland[0.043]; Philadelphia[0.043]; teams[0.042]; Toronto[0.042]; Detroit[0.041]; played[0.041]; Houston[0.041]; Miami[0.041]; Conference[0.040]; Conference[0.040]; won[0.040]; Washington[0.040]; Charlotte[0.040]; Orlando[0.039]; Western[0.039]; Chicago[0.039]; Atlanta[0.039]; Dallas[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Phoenix ==> ENTITY: \u001b[32mPhoenix Suns\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.205:0.205[0]; log p(e|m)= -3.474:-3.474[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMiami[0.044]; Toronto[0.043]; Conference[0.043]; Houston[0.043]; Houston[0.043]; Vancouver[0.043]; Vancouver[0.043]; Philadelphia[0.043]; Dallas[0.042]; Dallas[0.042]; Atlanta[0.041]; Detroit[0.041]; Lakers[0.041]; Clippers[0.041]; Clippers[0.041]; Charlotte[0.041]; Chicago[0.041]; Denver[0.041]; Denver[0.041]; Washington[0.041]; Utah[0.040]; Utah[0.040]; Western[0.040]; Saturday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Philadelphia ==> ENTITY: \u001b[32mPhiladelphia 76ers\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.202:0.202[0]; log p(e|m)= -3.912:-3.912[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.044]; Clippers[0.044]; Clippers[0.044]; Miami[0.043]; Washington[0.043]; Lakers[0.042]; Dallas[0.042]; Charlotte[0.042]; Detroit[0.042]; Atlanta[0.042]; York[0.042]; Denver[0.042]; Denver[0.042]; Phoenix[0.041]; Toronto[0.041]; Milwaukee[0.041]; Chicago[0.041]; Division[0.041]; Portland[0.041]; Saturday[0.040]; Jersey[0.040]; Sacramento[0.040]; December[0.040]; Golden[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Houston ==> ENTITY: \u001b[32mHouston Rockets\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.192:0.192[0]; log p(e|m)= -3.817:-3.817[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDallas[0.045]; Philadelphia[0.044]; Miami[0.043]; Clippers[0.043]; Clippers[0.043]; Phoenix[0.042]; Lakers[0.042]; Washington[0.042]; Atlanta[0.042]; Denver[0.042]; Denver[0.042]; Toronto[0.042]; Saturday[0.042]; Charlotte[0.041]; December[0.041]; Golden[0.041]; Detroit[0.041]; Chicago[0.041]; State[0.041]; Utah[0.040]; Division[0.040]; Sacramento[0.040]; Seattle[0.040]; Seattle[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Toronto ==> ENTITY: \u001b[32mToronto Raptors\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.191:0.191[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; Vancouver[0.044]; Cleveland[0.043]; Philadelphia[0.043]; Boston[0.043]; Detroit[0.042]; Houston[0.042]; won[0.042]; Atlanta[0.042]; Orlando[0.042]; Charlotte[0.041]; York[0.041]; Clippers[0.041]; Miami[0.041]; Conference[0.041]; Conference[0.041]; Western[0.041]; Indiana[0.041]; Dallas[0.041]; Washington[0.041]; Chicago[0.040]; Atlantic[0.040]; Minnesota[0.040]; Lakers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington Wizards\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.226:0.226[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; games[0.045]; Basketball[0.044]; Basketball[0.044]; Philadelphia[0.044]; Miami[0.043]; Cleveland[0.043]; teams[0.043]; Toronto[0.042]; Houston[0.042]; Orlando[0.042]; Boston[0.042]; Atlanta[0.041]; played[0.041]; Charlotte[0.041]; won[0.041]; Division[0.040]; Division[0.040]; Division[0.040]; Chicago[0.040]; Conference[0.040]; Conference[0.040]; Detroit[0.040]; National[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vancouver ==> ENTITY: \u001b[32mMemphis Grizzlies\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.218:0.218[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVancouver[0.050]; Toronto[0.046]; Toronto[0.046]; Cleveland[0.045]; Charlotte[0.044]; Charlotte[0.044]; Houston[0.044]; Houston[0.044]; Miami[0.044]; Clippers[0.043]; Clippers[0.043]; Phoenix[0.043]; Detroit[0.043]; Detroit[0.043]; Philadelphia[0.043]; Washington[0.043]; Portland[0.042]; Conference[0.042]; Atlanta[0.042]; Atlanta[0.042]; Chicago[0.042]; Chicago[0.042]; Lakers[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Toronto ==> ENTITY: \u001b[32mToronto Raptors\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.183:0.183[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVancouver[0.045]; Vancouver[0.045]; Philadelphia[0.043]; Detroit[0.043]; Houston[0.043]; Houston[0.043]; Atlanta[0.042]; Charlotte[0.042]; York[0.042]; Clippers[0.042]; Clippers[0.042]; Miami[0.042]; Phoenix[0.041]; Dallas[0.041]; Dallas[0.041]; Washington[0.041]; Chicago[0.041]; Saturday[0.040]; Minnesota[0.040]; Lakers[0.040]; Jersey[0.040]; Division[0.040]; December[0.040]; Milwaukee[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlanta ==> ENTITY: \u001b[32mAtlanta Hawks\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.219:0.219[0]; log p(e|m)= -3.650:-3.650[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.045]; Houston[0.044]; Miami[0.044]; Charlotte[0.043]; Dallas[0.042]; Dallas[0.042]; Clippers[0.042]; Clippers[0.042]; Detroit[0.042]; Washington[0.042]; Lakers[0.042]; Toronto[0.042]; Chicago[0.041]; Milwaukee[0.041]; Division[0.041]; Phoenix[0.041]; Saturday[0.041]; Denver[0.041]; Denver[0.041]; Antonio[0.041]; Minnesota[0.040]; York[0.040]; Sacramento[0.040]; Jersey[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cleveland ==> ENTITY: \u001b[32mCleveland Cavaliers\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.206:0.206[0]; log p(e|m)= -3.730:-3.730[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.047]; Basketball[0.047]; games[0.047]; games[0.047]; Miami[0.045]; Philadelphia[0.045]; Washington[0.044]; Houston[0.044]; Boston[0.044]; teams[0.043]; Dallas[0.043]; Toronto[0.042]; Charlotte[0.042]; Atlanta[0.042]; Milwaukee[0.042]; York[0.042]; York[0.042]; played[0.042]; Indiana[0.042]; Orlando[0.042]; Detroit[0.042]; Lakers[0.042]; Minnesota[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Utah ==> ENTITY: \u001b[32mUtah Jazz\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.184:0.184[0]; log p(e|m)= -3.576:-3.576[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhoenix[0.044]; Charlotte[0.043]; Houston[0.043]; Vancouver[0.043]; Vancouver[0.043]; Atlanta[0.043]; Clippers[0.043]; Clippers[0.043]; Lakers[0.043]; Dallas[0.042]; Division[0.042]; Portland[0.042]; Miami[0.042]; Denver[0.041]; Philadelphia[0.041]; Saturday[0.041]; Washington[0.041]; Toronto[0.041]; Sacramento[0.041]; State[0.041]; Detroit[0.040]; Chicago[0.040]; York[0.039]; Golden[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Denver ==> ENTITY: \u001b[32mDenver Nuggets\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.216:0.216[0]; log p(e|m)= -3.270:-3.270[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.044]; Philadelphia[0.044]; Miami[0.044]; Phoenix[0.043]; Charlotte[0.043]; Dallas[0.043]; Clippers[0.042]; Clippers[0.042]; Washington[0.042]; Utah[0.042]; Toronto[0.042]; Atlanta[0.042]; Chicago[0.041]; Sacramento[0.041]; Division[0.041]; Portland[0.041]; Detroit[0.041]; Lakers[0.041]; Saturday[0.040]; Milwaukee[0.040]; Seattle[0.040]; Seattle[0.040]; December[0.040]; State[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Houston ==> ENTITY: \u001b[32mHouston Rockets\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.200:0.200[0]; log p(e|m)= -3.817:-3.817[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConference[0.045]; Dallas[0.045]; Cleveland[0.044]; Philadelphia[0.043]; Orlando[0.043]; Clippers[0.042]; Boston[0.042]; Phoenix[0.042]; Lakers[0.042]; Washington[0.042]; Atlanta[0.041]; Denver[0.041]; Toronto[0.041]; Toronto[0.041]; Saturday[0.041]; Charlotte[0.041]; December[0.041]; Western[0.041]; Golden[0.041]; Detroit[0.041]; Chicago[0.040]; State[0.040]; Central[0.040]; Indiana[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pacific ==> ENTITY: \u001b[32mPacific Division (NBA)\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.181:0.181[0]; log p(e|m)= -3.194:-3.194[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLakers[0.044]; Vancouver[0.044]; Vancouver[0.044]; Division[0.043]; Division[0.043]; Western[0.042]; Phoenix[0.042]; Dallas[0.042]; Dallas[0.042]; Clippers[0.042]; Clippers[0.042]; Conference[0.042]; Saturday[0.041]; Toronto[0.041]; Toronto[0.041]; Houston[0.041]; Houston[0.041]; Detroit[0.041]; Detroit[0.041]; Denver[0.041]; Denver[0.041]; Cleveland[0.040]; Chicago[0.040]; Chicago[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Central Division ==> ENTITY: \u001b[32mCentral Division (NBA)\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.236:0.236[0]; log p(e|m)= -1.575:-1.575[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteams[0.045]; games[0.044]; games[0.044]; Division[0.044]; Division[0.044]; Basketball[0.043]; Basketball[0.043]; Cleveland[0.043]; Conference[0.043]; Conference[0.043]; won[0.042]; Philadelphia[0.042]; Boston[0.042]; Western[0.041]; Midwest[0.040]; Detroit[0.040]; Dallas[0.040]; Houston[0.040]; Eastern[0.040]; played[0.040]; Standings[0.039]; Standings[0.039]; Denver[0.039]; Atlanta[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = LA CLIPPERS\u001b[39m ==> ENTITY: \u001b[32mLos Angeles Clippers\u001b[39m\t\nSCORES: global= 0.317:0.317[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.045]; Miami[0.044]; Houston[0.044]; Clippers[0.043]; Phoenix[0.043]; Toronto[0.042]; Denver[0.042]; Denver[0.042]; Dallas[0.042]; Dallas[0.042]; Charlotte[0.042]; Atlanta[0.042]; Chicago[0.041]; Washington[0.041]; Lakers[0.041]; Sacramento[0.041]; Detroit[0.041]; Portland[0.040]; December[0.040]; Utah[0.040]; Utah[0.040]; Milwaukee[0.040]; Seattle[0.040]; Seattle[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York Knicks\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.187:0.187[0]; log p(e|m)= -4.828:-4.828[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.045]; Miami[0.044]; Houston[0.043]; Washington[0.043]; Toronto[0.043]; Saturday[0.043]; Lakers[0.042]; Charlotte[0.042]; Clippers[0.042]; Clippers[0.042]; Dallas[0.042]; Dallas[0.042]; Detroit[0.042]; Jersey[0.041]; Phoenix[0.041]; Atlanta[0.041]; Division[0.041]; Denver[0.041]; Denver[0.041]; December[0.041]; Chicago[0.040]; State[0.040]; Minnesota[0.040]; Milwaukee[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Miami ==> ENTITY: \u001b[32mMiami Heat\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.227:0.227[0]; log p(e|m)= -3.730:-3.730[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; games[0.045]; Basketball[0.045]; Basketball[0.045]; teams[0.044]; Cleveland[0.043]; Conference[0.043]; Conference[0.043]; Orlando[0.042]; Philadelphia[0.042]; Boston[0.042]; Charlotte[0.042]; played[0.041]; won[0.040]; Toronto[0.040]; Detroit[0.040]; Atlanta[0.040]; Friday[0.040]; Friday[0.040]; Washington[0.040]; Chicago[0.040]; Indiana[0.040]; National[0.039]; Atlantic[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Charlotte ==> ENTITY: \u001b[32mNew Orleans Pelicans\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.213:0.213[0]; log p(e|m)= -4.828:-4.828[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.046]; Miami[0.044]; Philadelphia[0.044]; Atlanta[0.043]; Phoenix[0.043]; Dallas[0.043]; Division[0.042]; Detroit[0.042]; Clippers[0.042]; Clippers[0.042]; Toronto[0.041]; Washington[0.041]; Chicago[0.041]; Milwaukee[0.041]; Lakers[0.041]; December[0.041]; Denver[0.041]; Sacramento[0.041]; York[0.040]; Seattle[0.040]; Seattle[0.040]; Saturday[0.040]; Portland[0.040]; Jersey[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Seattle ==> ENTITY: \u001b[32mSeattle SuperSonics\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.205:0.205[0]; log p(e|m)= -3.650:-3.650[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSeattle[0.045]; Washington[0.044]; Philadelphia[0.044]; Toronto[0.043]; Houston[0.043]; Miami[0.043]; Clippers[0.042]; Clippers[0.042]; Portland[0.042]; Dallas[0.042]; Phoenix[0.041]; Vancouver[0.041]; Vancouver[0.041]; Division[0.041]; Denver[0.041]; Charlotte[0.041]; Chicago[0.041]; Atlanta[0.041]; Detroit[0.041]; Lakers[0.041]; York[0.040]; Saturday[0.040]; Utah[0.040]; Antonio[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew York\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew York Knicks\u001b[39m <---> \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.250:0.241[\u001b[31m0.009\u001b[39m]; local()= 0.204:0.092[\u001b[31m0.112\u001b[39m]; log p(e|m)= -4.828:-1.790[\u001b[32m3.039\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.049]; Basketball[0.049]; games[0.046]; games[0.046]; Philadelphia[0.045]; Cleveland[0.044]; Boston[0.044]; teams[0.044]; Miami[0.043]; Conference[0.043]; won[0.043]; played[0.043]; Friday[0.043]; Friday[0.043]; Washington[0.042]; York[0.042]; Orlando[0.042]; National[0.041]; Detroit[0.041]; Central[0.041]; lost[0.041]; Jersey[0.041]; Atlanta[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Utah ==> ENTITY: \u001b[32mUtah Jazz\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.198:0.198[0]; log p(e|m)= -3.576:-3.576[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConference[0.045]; Phoenix[0.043]; Charlotte[0.043]; Cleveland[0.043]; Houston[0.043]; Vancouver[0.043]; Atlanta[0.042]; Atlanta[0.042]; Clippers[0.042]; Clippers[0.042]; Lakers[0.042]; Dallas[0.042]; Division[0.042]; Division[0.042]; Division[0.042]; Portland[0.041]; Orlando[0.041]; Denver[0.041]; Philadelphia[0.041]; Saturday[0.041]; Washington[0.040]; Boston[0.040]; Toronto[0.040]; Toronto[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Philadelphia ==> ENTITY: \u001b[32mPhiladelphia 76ers\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.211:0.211[0]; log p(e|m)= -3.912:-3.912[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteams[0.050]; games[0.049]; games[0.049]; Basketball[0.048]; Basketball[0.048]; Cleveland[0.047]; Toronto[0.046]; played[0.045]; won[0.045]; Boston[0.045]; Miami[0.045]; Washington[0.045]; Orlando[0.044]; Games[0.044]; National[0.044]; Conference[0.044]; Conference[0.044]; York[0.044]; York[0.044]; Houston[0.044]; Jersey[0.043]; Charlotte[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Seattle ==> ENTITY: \u001b[32mSeattle SuperSonics\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.228:0.228[0]; log p(e|m)= -3.650:-3.650[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSeattle[0.046]; Conference[0.046]; Washington[0.045]; Cleveland[0.045]; Philadelphia[0.045]; Toronto[0.044]; Toronto[0.044]; Houston[0.044]; Houston[0.044]; Miami[0.044]; Western[0.043]; Clippers[0.043]; Clippers[0.043]; Portland[0.043]; Dallas[0.043]; Dallas[0.043]; Phoenix[0.042]; Vancouver[0.042]; Vancouver[0.042]; Division[0.042]; Division[0.042]; Denver[0.042]; Denver[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Milwaukee ==> ENTITY: \u001b[32mMilwaukee Bucks\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.205:0.205[0]; log p(e|m)= -3.124:-3.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.046]; Houston[0.044]; Detroit[0.043]; Miami[0.043]; Atlanta[0.043]; Clippers[0.042]; Clippers[0.042]; Dallas[0.042]; Dallas[0.042]; Washington[0.042]; Toronto[0.042]; Charlotte[0.042]; Lakers[0.041]; York[0.041]; Chicago[0.041]; Phoenix[0.041]; Minnesota[0.041]; Denver[0.041]; Denver[0.041]; Saturday[0.041]; Division[0.040]; Sacramento[0.040]; Utah[0.040]; Jersey[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington Wizards\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.213:0.213[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.045]; Miami[0.044]; Toronto[0.043]; Houston[0.043]; Clippers[0.042]; Clippers[0.042]; Atlanta[0.042]; Denver[0.042]; Denver[0.042]; Vancouver[0.042]; Vancouver[0.042]; Seattle[0.042]; Seattle[0.042]; Dallas[0.042]; Dallas[0.042]; Charlotte[0.042]; Chicago[0.041]; Phoenix[0.041]; Detroit[0.041]; Minnesota[0.041]; Lakers[0.040]; Portland[0.040]; Sacramento[0.040]; Milwaukee[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlanta ==> ENTITY: \u001b[32mAtlanta Hawks\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.231:0.231[0]; log p(e|m)= -3.650:-3.650[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.045]; games[0.045]; games[0.045]; Philadelphia[0.044]; Cleveland[0.044]; Houston[0.043]; Miami[0.042]; teams[0.042]; Boston[0.042]; Conference[0.042]; Conference[0.042]; Detroit[0.041]; Charlotte[0.041]; Dallas[0.041]; won[0.040]; Orlando[0.040]; Washington[0.040]; played[0.040]; Standings[0.040]; Standings[0.040]; Games[0.040]; Lakers[0.040]; Toronto[0.040]; Western[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Orlando ==> ENTITY: \u001b[32mOrlando Magic\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.220:0.220[0]; log p(e|m)= -3.079:-3.079[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; games[0.046]; Basketball[0.045]; Basketball[0.045]; Cleveland[0.042]; Miami[0.042]; Conference[0.042]; Conference[0.042]; teams[0.042]; won[0.042]; Houston[0.042]; Philadelphia[0.041]; Boston[0.041]; Atlanta[0.041]; Detroit[0.041]; Charlotte[0.041]; Toronto[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; played[0.040]; Atlantic[0.040]; Friday[0.039]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Minnesota ==> ENTITY: \u001b[32mMinnesota Timberwolves\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.187:0.187[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.045]; Detroit[0.043]; Detroit[0.043]; Philadelphia[0.043]; Houston[0.043]; Charlotte[0.042]; Phoenix[0.042]; Conference[0.042]; Boston[0.042]; Dallas[0.041]; Clippers[0.041]; Clippers[0.041]; Saturday[0.041]; Central[0.041]; Washington[0.041]; Washington[0.041]; Toronto[0.041]; Toronto[0.041]; Denver[0.041]; Atlanta[0.041]; Atlanta[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = National Basketball Association\u001b[39m ==> ENTITY: \u001b[32mNational Basketball Association\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.047]; Cleveland[0.044]; games[0.044]; games[0.044]; teams[0.044]; Miami[0.044]; Philadelphia[0.043]; Boston[0.043]; Conference[0.042]; Detroit[0.041]; Orlando[0.041]; played[0.041]; Atlanta[0.041]; won[0.041]; Friday[0.041]; Friday[0.041]; Indiana[0.041]; Chicago[0.040]; percentage[0.040]; Division[0.040]; Division[0.040]; Games[0.039]; Washington[0.039]; Milwaukee[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = LA LAKERS\u001b[39m ==> ENTITY: \u001b[32mLos Angeles Lakers\u001b[39m\t\nSCORES: global= 0.317:0.317[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConference[0.045]; Cleveland[0.044]; Philadelphia[0.043]; Houston[0.043]; Houston[0.043]; Miami[0.042]; Western[0.042]; Dallas[0.042]; Dallas[0.042]; Clippers[0.042]; Clippers[0.042]; Toronto[0.041]; Toronto[0.041]; Phoenix[0.041]; Charlotte[0.041]; Charlotte[0.041]; Denver[0.041]; Denver[0.041]; Detroit[0.041]; Washington[0.040]; Division[0.040]; Division[0.040]; Chicago[0.040]; Sacramento[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago Bulls\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.200:0.200[0]; log p(e|m)= -4.510:-4.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; games[0.046]; Basketball[0.044]; Basketball[0.044]; Cleveland[0.043]; teams[0.042]; Vancouver[0.042]; Toronto[0.042]; won[0.042]; Philadelphia[0.041]; Conference[0.041]; Conference[0.041]; Boston[0.041]; Houston[0.041]; Miami[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Detroit[0.040]; played[0.040]; Orlando[0.040]; Washington[0.040]; Charlotte[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Detroit ==> ENTITY: \u001b[32mDetroit Pistons\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.182:0.182[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.044]; Vancouver[0.044]; Vancouver[0.044]; Toronto[0.044]; Charlotte[0.043]; Miami[0.042]; December[0.042]; Houston[0.042]; Division[0.042]; Phoenix[0.042]; Atlanta[0.042]; Chicago[0.041]; Washington[0.041]; Lakers[0.041]; Dallas[0.041]; Dallas[0.041]; Clippers[0.041]; Clippers[0.041]; Denver[0.041]; Denver[0.041]; Saturday[0.041]; York[0.040]; Milwaukee[0.040]; Jersey[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Jersey ==> ENTITY: \u001b[32mBrooklyn Nets\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.198:0.198[0]; log p(e|m)= -3.912:-3.912[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.045]; York[0.044]; Toronto[0.043]; Clippers[0.043]; Clippers[0.043]; Houston[0.043]; Vancouver[0.042]; Vancouver[0.042]; Miami[0.042]; Charlotte[0.042]; Lakers[0.042]; Detroit[0.042]; Atlanta[0.042]; Saturday[0.041]; Division[0.041]; Dallas[0.041]; Dallas[0.041]; Phoenix[0.041]; Denver[0.040]; Denver[0.040]; Washington[0.040]; Chicago[0.040]; December[0.040]; Utah[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Milwaukee ==> ENTITY: \u001b[32mMilwaukee Bucks\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.214:0.214[0]; log p(e|m)= -3.124:-3.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.044]; Basketball[0.044]; games[0.044]; games[0.044]; Cleveland[0.044]; Boston[0.042]; Houston[0.042]; teams[0.042]; Conference[0.042]; Conference[0.042]; Detroit[0.042]; Miami[0.041]; Atlanta[0.041]; Central[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; played[0.041]; lost[0.041]; Orlando[0.040]; Clippers[0.040]; Dallas[0.040]; Washington[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dallas ==> ENTITY: \u001b[32mDallas Mavericks\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.195:0.195[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.045]; Washington[0.044]; Conference[0.044]; Toronto[0.043]; Toronto[0.043]; Miami[0.043]; Vancouver[0.042]; Philadelphia[0.042]; Cleveland[0.042]; Boston[0.042]; York[0.041]; Clippers[0.041]; Clippers[0.041]; Detroit[0.041]; Detroit[0.041]; Phoenix[0.041]; Charlotte[0.041]; Central[0.041]; Lakers[0.041]; Atlanta[0.040]; Atlanta[0.040]; Indiana[0.040]; Denver[0.040]; Chicago[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Jersey ==> ENTITY: \u001b[32mBrooklyn Nets\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.220:0.220[0]; log p(e|m)= -3.912:-3.912[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; games[0.047]; Basketball[0.046]; Basketball[0.046]; Philadelphia[0.044]; played[0.044]; Conference[0.044]; Conference[0.044]; teams[0.043]; York[0.043]; York[0.043]; Toronto[0.043]; Boston[0.043]; Cleveland[0.043]; Houston[0.043]; won[0.043]; Atlantic[0.042]; Miami[0.042]; Charlotte[0.042]; Central[0.042]; Detroit[0.041]; Orlando[0.041]; Atlanta[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portland ==> ENTITY: \u001b[32mPortland Trail Blazers\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.215:0.215[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConference[0.043]; Philadelphia[0.043]; Washington[0.043]; Houston[0.043]; Houston[0.043]; Miami[0.043]; Toronto[0.042]; Toronto[0.042]; Clippers[0.042]; Clippers[0.042]; Seattle[0.041]; Seattle[0.041]; Dallas[0.041]; Dallas[0.041]; Atlanta[0.041]; Atlanta[0.041]; Phoenix[0.041]; Division[0.041]; Division[0.041]; Charlotte[0.041]; Charlotte[0.041]; Denver[0.041]; Denver[0.041]; Detroit[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NBA\u001b[39m ==> ENTITY: \u001b[32mNational Basketball Association\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.046]; Basketball[0.046]; games[0.045]; games[0.045]; Conference[0.044]; teams[0.044]; Division[0.043]; Division[0.043]; Cleveland[0.042]; won[0.042]; Boston[0.041]; Miami[0.041]; Philadelphia[0.041]; played[0.041]; Orlando[0.041]; Eastern[0.040]; National[0.040]; Detroit[0.040]; Friday[0.040]; Friday[0.040]; Games[0.040]; Central[0.040]; Atlantic[0.039]; Washington[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlantic ==> ENTITY: \u001b[32mAtlantic Division (NBA)\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.237:0.237[0]; log p(e|m)= -3.101:-3.101[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; games[0.045]; teams[0.045]; Basketball[0.044]; Basketball[0.044]; won[0.044]; Division[0.043]; Division[0.043]; Boston[0.043]; Philadelphia[0.043]; Conference[0.042]; Conference[0.042]; Detroit[0.041]; Cleveland[0.041]; Toronto[0.041]; lost[0.040]; Atlanta[0.040]; played[0.040]; Eastern[0.039]; Jersey[0.039]; Miami[0.039]; Western[0.039]; Charlotte[0.039]; National[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Miami ==> ENTITY: \u001b[32mMiami Heat\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.194:0.194[0]; log p(e|m)= -3.730:-3.730[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.044]; Philadelphia[0.044]; Charlotte[0.043]; Dallas[0.043]; Dallas[0.043]; State[0.042]; Toronto[0.042]; Phoenix[0.042]; Detroit[0.042]; Atlanta[0.042]; Lakers[0.042]; Clippers[0.042]; Clippers[0.042]; Denver[0.041]; Denver[0.041]; Washington[0.041]; Saturday[0.041]; Chicago[0.041]; Seattle[0.041]; Seattle[0.041]; Division[0.040]; York[0.040]; December[0.040]; Milwaukee[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago Bulls\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.177:0.177[0]; log p(e|m)= -4.510:-4.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVancouver[0.044]; Vancouver[0.044]; Toronto[0.044]; Philadelphia[0.044]; Houston[0.043]; Miami[0.043]; Detroit[0.042]; Washington[0.042]; Lakers[0.042]; Phoenix[0.042]; Clippers[0.042]; Clippers[0.042]; Charlotte[0.041]; Dallas[0.041]; Saturday[0.041]; York[0.041]; Atlanta[0.041]; Denver[0.041]; Denver[0.041]; Milwaukee[0.040]; Division[0.040]; Utah[0.040]; Sacramento[0.040]; Portland[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indiana ==> ENTITY: \u001b[32mIndiana Pacers\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.224:0.224[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.046]; games[0.044]; games[0.044]; Cleveland[0.043]; Philadelphia[0.043]; Charlotte[0.043]; Conference[0.043]; Conference[0.043]; Miami[0.042]; Houston[0.041]; Toronto[0.041]; Boston[0.041]; played[0.041]; Orlando[0.041]; teams[0.041]; Clippers[0.041]; Dallas[0.040]; Washington[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Detroit[0.040]; Central[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Denver ==> ENTITY: \u001b[32mDenver Nuggets\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.220:0.220[0]; log p(e|m)= -3.270:-3.270[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.043]; Houston[0.043]; Philadelphia[0.043]; Cleveland[0.043]; Miami[0.043]; Conference[0.043]; Phoenix[0.042]; Charlotte[0.042]; Dallas[0.042]; Dallas[0.042]; Boston[0.042]; Clippers[0.041]; Clippers[0.041]; Central[0.041]; Washington[0.041]; Utah[0.041]; Toronto[0.041]; Toronto[0.041]; Atlanta[0.041]; Atlanta[0.041]; Indiana[0.041]; Chicago[0.040]; Chicago[0.040]; Sacramento[0.040]; \t\n\r [====>.........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 16s558ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 513/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1214testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = UDMR\u001b[39m ==> ENTITY: \u001b[32mDemocratic Union of Hungarians in Romania\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncoalition[0.044]; coalition[0.044]; Social[0.042]; Romania[0.042]; Romania[0.042]; Romania[0.042]; Romania[0.042]; government[0.042]; government[0.042]; parties[0.042]; support[0.042]; leader[0.042]; President[0.041]; Hungarian[0.041]; parliamentary[0.041]; news[0.041]; partners[0.041]; wants[0.041]; Marko[0.041]; committed[0.041]; Democratic[0.041]; Democratic[0.041]; Democratic[0.041]; Newsroom[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marko Bela\u001b[39m ==> ENTITY: \u001b[32mBéla Markó\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRomania[0.045]; Romania[0.045]; Romania[0.045]; Romania[0.045]; leader[0.044]; Hungarian[0.042]; parliamentary[0.042]; President[0.042]; Democratic[0.041]; Democratic[0.041]; Democratic[0.041]; Bucharest[0.041]; Bucharest[0.041]; Emil[0.041]; Union[0.041]; Union[0.041]; economic[0.040]; Newsroom[0.040]; Constantinescu[0.040]; Constantinescu[0.040]; Sibiu[0.039]; government[0.039]; government[0.039]; Timisoara[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Timisoara\u001b[39m ==> ENTITY: \u001b[32mTimișoara\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRomania[0.045]; Romania[0.045]; Romania[0.045]; Bucharest[0.043]; Bucharest[0.043]; Oradea[0.043]; President[0.042]; government[0.041]; government[0.041]; ceremony[0.041]; ceremony[0.041]; attended[0.041]; signed[0.041]; Hungarian[0.041]; Marko[0.041]; support[0.041]; Sibiu[0.041]; economic[0.041]; international[0.041]; Radio[0.041]; Newsroom[0.040]; committed[0.040]; revival[0.040]; Social[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = UDMR\u001b[39m ==> ENTITY: \u001b[32mDemocratic Union of Hungarians in Romania\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncoalition[0.044]; coalition[0.044]; Social[0.042]; Romania[0.042]; Romania[0.042]; Romania[0.042]; Romania[0.042]; government[0.042]; government[0.042]; parties[0.042]; support[0.042]; leader[0.042]; President[0.041]; Hungarian[0.041]; parliamentary[0.041]; news[0.041]; partners[0.041]; wants[0.041]; Marko[0.041]; committed[0.041]; Democratic[0.041]; Democratic[0.041]; Democratic[0.041]; Newsroom[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Arad ==> ENTITY: \u001b[32mArad, Romania\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.119:0.119[0]; log p(e|m)= -0.683:-0.683[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairport[0.046]; Oradea[0.045]; airports[0.044]; Romania[0.043]; Romania[0.043]; attended[0.043]; international[0.043]; Bucharest[0.042]; Hungarian[0.042]; domestic[0.042]; economic[0.041]; Marko[0.041]; Sibiu[0.041]; Timisoara[0.041]; news[0.041]; Bela[0.040]; wants[0.040]; closed[0.040]; ceremony[0.040]; ceremony[0.040]; Emil[0.040]; economy[0.040]; Social[0.040]; partners[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bucharest ==> ENTITY: \u001b[32mBucharest\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.018:-0.018[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRomania[0.045]; Romania[0.045]; Romania[0.045]; Romania[0.045]; signed[0.044]; agreement[0.043]; attended[0.042]; coalition[0.041]; coalition[0.041]; government[0.041]; government[0.041]; Radio[0.041]; Radio[0.041]; support[0.041]; economic[0.041]; Hungarian[0.041]; union[0.041]; international[0.040]; President[0.040]; committed[0.040]; ceremony[0.040]; ceremony[0.040]; partners[0.040]; Convention[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Romania ==> ENTITY: \u001b[32mRomania\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.212:-0.212[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRomania[0.048]; Romania[0.048]; Romania[0.048]; President[0.044]; economic[0.044]; international[0.043]; domestic[0.042]; Bucharest[0.041]; Bucharest[0.041]; Timisoara[0.041]; government[0.041]; government[0.041]; support[0.041]; ceremony[0.040]; ceremony[0.040]; Hungarian[0.040]; reform[0.040]; reform[0.040]; parliamentary[0.040]; economy[0.039]; coalition[0.039]; coalition[0.039]; Union[0.039]; Union[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mOradea\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mOradea\u001b[39m <---> \u001b[32mOradea International Airport\u001b[39m\t\nSCORES: global= 0.288:0.258[\u001b[31m0.031\u001b[39m]; local()= 0.101:0.094[\u001b[31m0.007\u001b[39m]; log p(e|m)= 0.000:-1.094[\u001b[31m1.094\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBucharest[0.044]; Social[0.044]; airport[0.044]; Romania[0.043]; Romania[0.043]; partners[0.043]; airports[0.043]; Sibiu[0.042]; Hungarian[0.042]; Newsroom[0.042]; parties[0.042]; Timisoara[0.042]; international[0.041]; Arad[0.041]; said[0.041]; said[0.041]; Emil[0.041]; support[0.040]; wants[0.040]; Constantinescu[0.040]; Constantinescu[0.040]; signed[0.040]; agreement[0.040]; Bela[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Romania ==> ENTITY: \u001b[32mRomania\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.212:-0.212[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRomania[0.048]; Romania[0.048]; Romania[0.048]; economic[0.044]; international[0.043]; President[0.042]; domestic[0.042]; Bucharest[0.042]; Bucharest[0.042]; Timisoara[0.042]; government[0.041]; government[0.041]; support[0.041]; ceremony[0.040]; ceremony[0.040]; Hungarian[0.040]; reform[0.040]; reform[0.040]; economy[0.039]; coalition[0.039]; coalition[0.039]; Union[0.039]; Union[0.039]; union[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Emil Constantinescu\u001b[39m ==> ENTITY: \u001b[32mEmil Constantinescu\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.044]; Romania[0.043]; Romania[0.043]; Romania[0.043]; Romania[0.043]; economic[0.043]; government[0.042]; government[0.042]; leader[0.042]; support[0.042]; coalition[0.042]; coalition[0.042]; Bucharest[0.042]; Bucharest[0.042]; Social[0.042]; economy[0.042]; Constantinescu[0.041]; Convention[0.041]; said[0.040]; said[0.040]; attended[0.040]; Hungarian[0.040]; ceremony[0.040]; ceremony[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSibiu\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSibiu\u001b[39m <---> \u001b[32mSibiu International Airport\u001b[39m\t\nSCORES: global= 0.284:0.243[\u001b[31m0.041\u001b[39m]; local()= 0.119:0.120[\u001b[32m0.001\u001b[39m]; log p(e|m)= 0.000:-1.604[\u001b[31m1.604\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairport[0.046]; airports[0.046]; Romania[0.045]; Romania[0.045]; economic[0.045]; Oradea[0.045]; international[0.045]; Social[0.045]; social[0.044]; Bucharest[0.043]; domestic[0.043]; Hungarian[0.043]; Bela[0.043]; closed[0.043]; Timisoara[0.043]; attended[0.043]; said[0.042]; said[0.042]; support[0.042]; Newsroom[0.042]; partners[0.042]; agreement[0.042]; economy[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Constantinescu\u001b[39m ==> ENTITY: \u001b[32mEmil Constantinescu\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.044]; Romania[0.043]; Romania[0.043]; Romania[0.043]; Romania[0.043]; economic[0.043]; government[0.042]; government[0.042]; leader[0.042]; support[0.042]; coalition[0.042]; coalition[0.042]; Bucharest[0.042]; Bucharest[0.042]; Social[0.042]; economy[0.042]; Constantinescu[0.041]; Convention[0.041]; said[0.040]; said[0.040]; attended[0.040]; Hungarian[0.040]; ceremony[0.040]; ceremony[0.040]; \t\n\r [====>.........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 16s788ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 525/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1184testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gong Zhichao\u001b[39m ==> ENTITY: \u001b[32mGong Zhichao\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWang[0.047]; Semifinals[0.046]; Semifinals[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; Gong[0.045]; Taiwan[0.045]; Audina[0.045]; Malaysia[0.045]; Malaysia[0.045]; singles[0.044]; Chen[0.044]; Ye[0.044]; Ye[0.044]; Han[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.262:0.262[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTaiwan[0.063]; China[0.063]; China[0.063]; China[0.063]; China[0.063]; China[0.063]; China[0.063]; China[0.063]; China[0.063]; Malaysia[0.057]; Audina[0.056]; Denmark[0.054]; Denmark[0.054]; Denmark[0.054]; Denmark[0.054]; Han[0.053]; Hu[0.053]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.193:0.193[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTaiwan[0.047]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; badminton[0.043]; Malaysia[0.042]; Malaysia[0.042]; matches[0.042]; finals[0.041]; Denmark[0.040]; Denmark[0.040]; Denmark[0.040]; Denmark[0.040]; Badminton[0.039]; Indonesia[0.039]; Indonesia[0.039]; Indonesia[0.039]; Indonesia[0.039]; Indonesia[0.039]; Indonesia[0.039]; Gong[0.039]; Chen[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Susi Susanti\u001b[39m ==> ENTITY: \u001b[32mSusi Susanti\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAudina[0.053]; Semifinals[0.053]; Semifinals[0.053]; Women[0.051]; China[0.050]; China[0.050]; China[0.050]; China[0.050]; China[0.050]; China[0.050]; China[0.050]; China[0.050]; singles[0.050]; Indonesia[0.049]; Indonesia[0.049]; Indonesia[0.049]; Indonesia[0.049]; Indonesia[0.049]; Indonesia[0.049]; Allan[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.250:0.250[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.047]; Indonesia[0.047]; Indonesia[0.047]; Indonesia[0.047]; Indonesia[0.047]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; Audina[0.041]; Taiwan[0.040]; Denmark[0.039]; Denmark[0.039]; Wijaya[0.038]; singles[0.037]; Kusuma[0.036]; beat[0.036]; beat[0.036]; beat[0.036]; beat[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dong Jiong\u001b[39m ==> ENTITY: \u001b[32mDong Jiong\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.044]; Denmark[0.044]; Denmark[0.044]; Denmark[0.044]; China[0.043]; China[0.043]; China[0.043]; China[0.043]; badminton[0.043]; Malaysia[0.043]; Malaysia[0.043]; Badminton[0.042]; Semifinals[0.042]; matches[0.041]; Taiwan[0.040]; Lauridsen[0.040]; Chen[0.040]; Erik[0.040]; singles[0.039]; Nielsen[0.039]; Stuer[0.039]; World[0.039]; World[0.039]; Men[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Denmark ==> ENTITY: \u001b[32mDenmark\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.268:-0.268[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.058]; Denmark[0.058]; Denmark[0.058]; Denmark[0.058]; badminton[0.056]; matches[0.052]; Dong[0.049]; Audina[0.048]; Malaysia[0.048]; Malaysia[0.048]; Taiwan[0.048]; finals[0.047]; Indonesia[0.046]; Indonesia[0.046]; Indonesia[0.046]; Indonesia[0.046]; Indonesia[0.046]; Indonesia[0.046]; Indonesia[0.046]; Indonesia[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Taiwan ==> ENTITY: \u001b[32mTaiwan\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.257:0.257[0]; log p(e|m)= -0.078:-0.078[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.045]; China[0.045]; China[0.045]; China[0.045]; China[0.045]; China[0.045]; China[0.045]; China[0.045]; China[0.045]; Chen[0.041]; Malaysia[0.041]; Malaysia[0.041]; Zhichao[0.040]; Indonesia[0.039]; Indonesia[0.039]; Indonesia[0.039]; Indonesia[0.039]; Indonesia[0.039]; Indonesia[0.039]; Indonesia[0.039]; Indonesia[0.039]; Indonesia[0.039]; Indonesia[0.039]; Hu[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ong Ewe Hock\u001b[39m ==> ENTITY: \u001b[32mOng Ewe Hock\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBadminton[0.047]; badminton[0.046]; Semifinals[0.044]; Malaysia[0.044]; Malaysia[0.044]; Jun[0.042]; Jun[0.042]; Lauridsen[0.042]; Fung[0.041]; Saturday[0.041]; Friday[0.041]; Stuer[0.041]; Chen[0.041]; Dong[0.041]; Erik[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; finals[0.040]; Taiwan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mia Audina\u001b[39m ==> ENTITY: \u001b[32mMia Audina\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsingles[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; Wang[0.041]; Semifinals[0.041]; Semifinals[0.041]; Malaysia[0.041]; Taiwan[0.041]; Chen[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.325:0.325[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.045]; Indonesia[0.045]; Indonesia[0.045]; Indonesia[0.045]; Indonesia[0.045]; Indonesia[0.045]; Indonesia[0.045]; Indonesia[0.045]; Malaysia[0.043]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; Audina[0.039]; Taiwan[0.038]; Denmark[0.037]; Denmark[0.037]; Denmark[0.037]; Denmark[0.037]; Hermawan[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Camilla Martin\u001b[39m ==> ENTITY: \u001b[32mCamilla Martin\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMartin[0.045]; Denmark[0.045]; Denmark[0.045]; Semifinals[0.043]; Semifinals[0.043]; Wang[0.042]; Audina[0.042]; Han[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; singles[0.041]; Camilla[0.041]; Chen[0.041]; Gong[0.041]; Gong[0.041]; Taiwan[0.040]; Women[0.040]; Zhichao[0.040]; Jun[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Denmark ==> ENTITY: \u001b[32mDenmark\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.067:0.067[0]; log p(e|m)= -0.268:-0.268[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.052]; Audina[0.043]; Taiwan[0.043]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; China[0.041]; China[0.041]; China[0.041]; China[0.041]; China[0.041]; China[0.041]; China[0.041]; Han[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; singles[0.041]; Gong[0.040]; Gong[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.332:0.332[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.051]; Indonesia[0.051]; Indonesia[0.051]; Indonesia[0.051]; Indonesia[0.051]; Indonesia[0.051]; Indonesia[0.051]; Malaysia[0.049]; Malaysia[0.049]; China[0.047]; China[0.047]; China[0.047]; China[0.047]; China[0.047]; China[0.047]; China[0.047]; China[0.047]; badminton[0.045]; Audina[0.044]; Taiwan[0.043]; World[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Thomas Stuer-Lauridsen\u001b[39m ==> ENTITY: \u001b[32mThomas Stuer-Lauridsen\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbadminton[0.048]; Semifinals[0.045]; Denmark[0.045]; Denmark[0.045]; Denmark[0.045]; Denmark[0.045]; Badminton[0.043]; Lauridsen[0.043]; Men[0.042]; Jun[0.041]; Erik[0.041]; finals[0.041]; matches[0.040]; Malaysia[0.040]; Malaysia[0.040]; Wijaya[0.040]; Wijaya[0.040]; singles[0.040]; Friday[0.040]; Hansen[0.039]; World[0.039]; World[0.039]; Budi[0.039]; Budi[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.354:0.354[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.048]; Indonesia[0.048]; Indonesia[0.048]; Indonesia[0.048]; Indonesia[0.048]; Indonesia[0.048]; Indonesia[0.048]; Indonesia[0.048]; Indonesia[0.048]; Malaysia[0.046]; Malaysia[0.046]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; Audina[0.041]; Taiwan[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ye Zhaoying\u001b[39m ==> ENTITY: \u001b[32mYe Zhaoying\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAudina[0.048]; Malaysia[0.047]; Semifinals[0.047]; Semifinals[0.047]; Denmark[0.046]; Denmark[0.046]; Denmark[0.046]; Denmark[0.046]; China[0.045]; China[0.045]; China[0.045]; China[0.045]; China[0.045]; China[0.045]; China[0.045]; China[0.045]; Wang[0.045]; Zhichao[0.045]; Taiwan[0.044]; Gong[0.044]; Gong[0.044]; Erik[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.224:0.224[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTaiwan[0.049]; China[0.048]; China[0.048]; China[0.048]; China[0.048]; China[0.048]; China[0.048]; China[0.048]; Audina[0.043]; Denmark[0.042]; Denmark[0.042]; Denmark[0.042]; Denmark[0.042]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Wang[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rashid Sidek\u001b[39m ==> ENTITY: \u001b[32mRashid Sidek\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMalaysia[0.047]; Malaysia[0.047]; badminton[0.046]; Semifinals[0.045]; finals[0.045]; beat[0.044]; beat[0.044]; beat[0.044]; beat[0.044]; beat[0.044]; beat[0.044]; beat[0.044]; beat[0.044]; Badminton[0.043]; Lauridsen[0.043]; Audina[0.043]; Stuer[0.042]; Erik[0.042]; Dong[0.041]; Ong[0.041]; singles[0.041]; singles[0.041]; Hock[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ye Zhaoying\u001b[39m ==> ENTITY: \u001b[32mYe Zhaoying\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAudina[0.047]; Semifinals[0.046]; Denmark[0.045]; Denmark[0.045]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; Wang[0.044]; Zhichao[0.044]; Taiwan[0.043]; Gong[0.043]; Gong[0.043]; Indonesia[0.042]; Indonesia[0.042]; Indonesia[0.042]; Indonesia[0.042]; Indonesia[0.042]; Indonesia[0.042]; Ye[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.207:0.207[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTaiwan[0.047]; China[0.047]; China[0.047]; China[0.047]; China[0.047]; China[0.047]; China[0.047]; Audina[0.042]; Denmark[0.041]; Denmark[0.041]; Han[0.040]; Indonesia[0.039]; Indonesia[0.039]; Indonesia[0.039]; Indonesia[0.039]; Indonesia[0.039]; Indonesia[0.039]; Wang[0.039]; Gong[0.039]; Gong[0.039]; Chen[0.039]; Zhichao[0.039]; singles[0.038]; Jun[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Poul-Erik Hoyer-Larsen\u001b[39m ==> ENTITY: \u001b[32mPoul-Erik Høyer Larsen\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.047]; Denmark[0.047]; Denmark[0.047]; Denmark[0.047]; Denmark[0.047]; badminton[0.046]; Semifinals[0.045]; Audina[0.044]; Dong[0.043]; Lauridsen[0.043]; Stuer[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; Malaysia[0.042]; Malaysia[0.042]; Larsen[0.042]; finals[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sun Jun ==> ENTITY: \u001b[32mSun Jun (badminton)\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.166:0.166[0]; log p(e|m)= -0.137:-0.137[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.054]; Denmark[0.054]; Denmark[0.054]; Denmark[0.054]; Denmark[0.054]; Semifinals[0.053]; Semifinals[0.053]; Chen[0.053]; Zhichao[0.053]; singles[0.053]; Gong[0.052]; Gong[0.052]; Wang[0.052]; Audina[0.052]; Sun[0.052]; Erik[0.051]; Hu[0.051]; Lauridsen[0.051]; Jun[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Malaysia ==> ENTITY: \u001b[32mMalaysia\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.228:0.228[0]; log p(e|m)= -0.202:-0.202[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMalaysia[0.048]; Indonesia[0.044]; Indonesia[0.044]; Indonesia[0.044]; Indonesia[0.044]; Indonesia[0.044]; Indonesia[0.044]; badminton[0.043]; Dong[0.041]; matches[0.041]; Denmark[0.041]; Denmark[0.041]; Denmark[0.041]; Denmark[0.041]; China[0.041]; China[0.041]; China[0.041]; China[0.041]; China[0.041]; Badminton[0.041]; World[0.039]; World[0.039]; Taiwan[0.039]; finals[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.316:0.316[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.045]; Indonesia[0.045]; Indonesia[0.045]; Indonesia[0.045]; Indonesia[0.045]; Indonesia[0.045]; Indonesia[0.045]; Indonesia[0.045]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; Audina[0.039]; Taiwan[0.039]; Denmark[0.038]; Denmark[0.038]; Denmark[0.038]; Denmark[0.038]; Wijaya[0.036]; Santoso[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Camilla Martin\u001b[39m ==> ENTITY: \u001b[32mCamilla Martin\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMartin[0.044]; Denmark[0.044]; Denmark[0.044]; Denmark[0.044]; Denmark[0.044]; Semifinals[0.043]; Semifinals[0.043]; Wang[0.042]; Audina[0.041]; Han[0.041]; China[0.041]; China[0.041]; China[0.041]; China[0.041]; China[0.041]; China[0.041]; China[0.041]; China[0.041]; singles[0.040]; Camilla[0.040]; Chen[0.040]; Gong[0.040]; Gong[0.040]; Taiwan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Denmark ==> ENTITY: \u001b[32mDenmark\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.268:-0.268[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.059]; Denmark[0.059]; Denmark[0.059]; badminton[0.057]; matches[0.053]; Badminton[0.051]; Dong[0.049]; Malaysia[0.049]; Malaysia[0.049]; finals[0.048]; Indonesia[0.047]; Indonesia[0.047]; Indonesia[0.047]; Indonesia[0.047]; China[0.047]; China[0.047]; China[0.047]; China[0.047]; World[0.046]; World[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.202:0.202[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.049]; China[0.049]; China[0.049]; China[0.049]; China[0.049]; China[0.049]; Audina[0.044]; Denmark[0.043]; Denmark[0.043]; Han[0.042]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Wang[0.041]; Gong[0.041]; Gong[0.041]; Chen[0.041]; Zhichao[0.041]; singles[0.040]; Jun[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bali ==> ENTITY: \u001b[32mBali\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.167:0.167[0]; log p(e|m)= -0.104:-0.104[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.049]; Indonesia[0.049]; matches[0.047]; Malaysia[0.045]; Malaysia[0.045]; Wijaya[0.044]; Badminton[0.044]; finals[0.044]; badminton[0.043]; Denmark[0.043]; Denmark[0.043]; Denmark[0.043]; Hermawan[0.043]; China[0.042]; China[0.042]; China[0.042]; Hock[0.042]; Susanto[0.042]; singles[0.042]; Martin[0.042]; Indra[0.041]; Friday[0.041]; group[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Denmark ==> ENTITY: \u001b[32mDenmark\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.158:0.158[0]; log p(e|m)= -0.268:-0.268[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.065]; Denmark[0.065]; Denmark[0.065]; Denmark[0.065]; Dong[0.055]; Audina[0.054]; Malaysia[0.054]; Malaysia[0.054]; Taiwan[0.054]; Indonesia[0.052]; Indonesia[0.052]; Indonesia[0.052]; Indonesia[0.052]; Indonesia[0.052]; Indonesia[0.052]; Indonesia[0.052]; Indonesia[0.052]; Indonesia[0.052]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Denmark ==> ENTITY: \u001b[32mDenmark\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.268:-0.268[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.052]; Denmark[0.052]; Denmark[0.052]; badminton[0.050]; matches[0.046]; Badminton[0.045]; Dong[0.043]; Malaysia[0.043]; Malaysia[0.043]; Taiwan[0.042]; finals[0.042]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; China[0.041]; China[0.041]; China[0.041]; China[0.041]; China[0.041]; World[0.041]; World[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.262:0.262[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTaiwan[0.063]; China[0.062]; China[0.062]; China[0.062]; China[0.062]; China[0.062]; China[0.062]; China[0.062]; China[0.062]; Malaysia[0.057]; Malaysia[0.057]; Audina[0.056]; Denmark[0.054]; Denmark[0.054]; Denmark[0.054]; Denmark[0.054]; Denmark[0.054]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Denmark ==> ENTITY: \u001b[32mDenmark\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.268:-0.268[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.055]; Denmark[0.055]; Denmark[0.055]; Audina[0.046]; Taiwan[0.045]; Indonesia[0.044]; Indonesia[0.044]; Indonesia[0.044]; Indonesia[0.044]; Indonesia[0.044]; Indonesia[0.044]; Indonesia[0.044]; Indonesia[0.044]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; Han[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wang Chen ==> ENTITY: \u001b[32mWang Chen (badminton)\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.254:0.254[0]; log p(e|m)= -0.195:-0.195[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAudina[0.056]; China[0.054]; China[0.054]; China[0.054]; China[0.054]; China[0.054]; China[0.054]; China[0.054]; China[0.054]; Han[0.053]; singles[0.052]; Taiwan[0.052]; Zhichao[0.052]; Semifinals[0.052]; Semifinals[0.052]; Gong[0.051]; Gong[0.051]; Ye[0.050]; Ye[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Malaysia ==> ENTITY: \u001b[32mMalaysia\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.243:0.243[0]; log p(e|m)= -0.202:-0.202[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMalaysia[0.049]; Indonesia[0.045]; Indonesia[0.045]; Indonesia[0.045]; Indonesia[0.045]; Indonesia[0.045]; Indonesia[0.045]; Indonesia[0.045]; badminton[0.044]; Audina[0.043]; Dong[0.043]; matches[0.042]; Denmark[0.042]; Denmark[0.042]; Denmark[0.042]; Denmark[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; China[0.042]; Badminton[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.158:0.158[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTaiwan[0.048]; China[0.047]; China[0.047]; China[0.047]; badminton[0.044]; Malaysia[0.043]; Malaysia[0.043]; matches[0.043]; finals[0.041]; Denmark[0.041]; Denmark[0.041]; Denmark[0.041]; Denmark[0.041]; World[0.040]; World[0.040]; Badminton[0.040]; Indonesia[0.040]; Indonesia[0.040]; Indonesia[0.040]; Indonesia[0.040]; Chen[0.039]; Dong[0.039]; Hock[0.039]; Men[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.348:0.348[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.050]; Indonesia[0.050]; Indonesia[0.050]; Indonesia[0.050]; Indonesia[0.050]; Indonesia[0.050]; Indonesia[0.050]; Indonesia[0.050]; Malaysia[0.048]; Malaysia[0.048]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; Audina[0.043]; Taiwan[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Grand Prix ==> ENTITY: \u001b[32mWorld Badminton Grand Prix Finals\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.172:0.172[0]; log p(e|m)= -1.431:-1.431[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWorld[0.057]; badminton[0.050]; Grand[0.048]; Prix[0.046]; finals[0.046]; Badminton[0.046]; matches[0.046]; Denmark[0.045]; Denmark[0.045]; Denmark[0.045]; singles[0.045]; Malaysia[0.044]; Malaysia[0.044]; beat[0.044]; beat[0.044]; beat[0.044]; beat[0.044]; beat[0.044]; beat[0.044]; Group[0.043]; Group[0.043]; Group[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Allan Budi Kusuma\u001b[39m ==> ENTITY: \u001b[32mAlan Budikusuma\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAudina[0.066]; Semifinals[0.064]; Semifinals[0.064]; China[0.062]; China[0.062]; China[0.062]; China[0.062]; China[0.062]; China[0.062]; China[0.062]; China[0.062]; China[0.062]; Malaysia[0.062]; Malaysia[0.062]; Wang[0.060]; singles[0.060]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sun Jun ==> ENTITY: \u001b[32mSun Jun (badminton)\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.173:0.173[0]; log p(e|m)= -0.137:-0.137[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.050]; Denmark[0.050]; Denmark[0.050]; Denmark[0.050]; badminton[0.049]; Semifinals[0.048]; Chen[0.048]; Zhichao[0.048]; Dong[0.048]; singles[0.048]; singles[0.048]; matches[0.048]; Gong[0.047]; Sun[0.047]; Erik[0.046]; Hu[0.046]; finals[0.046]; Lauridsen[0.046]; Jun[0.046]; World[0.046]; World[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.226:0.226[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTaiwan[0.049]; China[0.048]; China[0.048]; China[0.048]; China[0.048]; China[0.048]; China[0.048]; China[0.048]; Audina[0.043]; Denmark[0.042]; Denmark[0.042]; Denmark[0.042]; Han[0.041]; Hu[0.041]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Wang[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Susi Susanti\u001b[39m ==> ENTITY: \u001b[32mSusi Susanti\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAudina[0.048]; Semifinals[0.048]; Semifinals[0.048]; Women[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; singles[0.045]; Erik[0.045]; Indonesia[0.044]; Indonesia[0.044]; Indonesia[0.044]; Indonesia[0.044]; Indonesia[0.044]; Indonesia[0.044]; Indonesia[0.044]; Allan[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hermawan Susanto\u001b[39m ==> ENTITY: \u001b[32mHermawan Susanto\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.232:0.232[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbadminton[0.046]; Audina[0.045]; Indonesia[0.045]; Indonesia[0.045]; Indonesia[0.045]; Indonesia[0.045]; Indonesia[0.045]; Indonesia[0.045]; Indonesia[0.045]; matches[0.044]; Malaysia[0.044]; Malaysia[0.044]; Semifinals[0.043]; Bali[0.043]; Wijaya[0.043]; Wijaya[0.043]; Denmark[0.042]; Denmark[0.042]; Denmark[0.042]; Denmark[0.042]; Sidek[0.042]; finals[0.042]; Badminton[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.255:0.255[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTaiwan[0.067]; China[0.066]; China[0.066]; China[0.066]; China[0.066]; China[0.066]; China[0.066]; China[0.066]; Malaysia[0.061]; Malaysia[0.061]; Audina[0.059]; Denmark[0.058]; Denmark[0.058]; Denmark[0.058]; Denmark[0.058]; Denmark[0.058]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.274:0.274[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.048]; Indonesia[0.048]; Indonesia[0.048]; Indonesia[0.048]; Malaysia[0.047]; Malaysia[0.047]; China[0.045]; China[0.045]; China[0.045]; China[0.045]; China[0.045]; badminton[0.043]; matches[0.043]; Taiwan[0.041]; World[0.041]; World[0.041]; Bali[0.041]; Denmark[0.040]; Denmark[0.040]; Denmark[0.040]; Denmark[0.040]; finals[0.040]; group[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.348:0.348[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.050]; Indonesia[0.050]; Indonesia[0.050]; Indonesia[0.050]; Indonesia[0.050]; Indonesia[0.050]; Indonesia[0.050]; Indonesia[0.050]; Malaysia[0.048]; Malaysia[0.048]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; Audina[0.043]; Taiwan[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Grand Prix ==> ENTITY: \u001b[32mWorld Badminton Grand Prix Finals\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.164:0.164[0]; log p(e|m)= -1.431:-1.431[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWorld[0.052]; badminton[0.046]; Grand[0.044]; Prix[0.043]; finals[0.043]; Badminton[0.043]; matches[0.042]; Denmark[0.042]; Denmark[0.042]; singles[0.041]; Malaysia[0.041]; Malaysia[0.041]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; Group[0.040]; Group[0.040]; China[0.040]; China[0.040]; China[0.040]; Martin[0.040]; Men[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.126:0.126[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.048]; China[0.048]; badminton[0.044]; Malaysia[0.044]; Malaysia[0.044]; matches[0.044]; finals[0.042]; Denmark[0.042]; Denmark[0.042]; Denmark[0.042]; Denmark[0.042]; World[0.041]; World[0.041]; Badminton[0.040]; Indonesia[0.040]; Indonesia[0.040]; Indonesia[0.040]; Indonesia[0.040]; Chen[0.040]; Dong[0.040]; Hock[0.039]; Men[0.039]; singles[0.039]; Thomas[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.282:0.282[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.046]; Indonesia[0.046]; Indonesia[0.046]; Indonesia[0.046]; Indonesia[0.046]; Indonesia[0.046]; China[0.043]; China[0.043]; China[0.043]; China[0.043]; China[0.043]; China[0.043]; China[0.043]; China[0.043]; Audina[0.040]; Taiwan[0.040]; Denmark[0.038]; Denmark[0.038]; Denmark[0.038]; Santoso[0.037]; Han[0.037]; Permadi[0.037]; Wijaya[0.036]; Zhichao[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.226:0.226[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTaiwan[0.047]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; Audina[0.041]; Denmark[0.040]; Denmark[0.040]; Han[0.039]; Hu[0.039]; Indonesia[0.039]; Indonesia[0.039]; Indonesia[0.039]; Indonesia[0.039]; Indonesia[0.039]; Indonesia[0.039]; Indonesia[0.039]; Wang[0.039]; Gong[0.039]; Gong[0.039]; Chen[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.316:0.316[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.047]; Indonesia[0.047]; Indonesia[0.047]; Indonesia[0.047]; Indonesia[0.047]; Indonesia[0.047]; Indonesia[0.047]; Malaysia[0.045]; Malaysia[0.045]; China[0.043]; China[0.043]; China[0.043]; China[0.043]; China[0.043]; China[0.043]; China[0.043]; badminton[0.041]; matches[0.041]; Audina[0.040]; Taiwan[0.040]; World[0.039]; World[0.039]; Bali[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fung Permadi\u001b[39m ==> ENTITY: \u001b[32mFung Permadi\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSemifinals[0.043]; Semifinals[0.043]; singles[0.043]; Malaysia[0.042]; Malaysia[0.042]; Chen[0.042]; Indonesia[0.042]; Indonesia[0.042]; Indonesia[0.042]; Indonesia[0.042]; Indonesia[0.042]; Indonesia[0.042]; Indonesia[0.042]; Indonesia[0.042]; Indonesia[0.042]; Indonesia[0.042]; Wang[0.042]; Allan[0.041]; Allan[0.041]; Lauridsen[0.041]; Stuer[0.041]; Erik[0.040]; Taiwan[0.040]; Hermawan[0.040]; \t\n\r [====>.........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 16s412ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 577/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1284testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tempe ==> ENTITY: \u001b[32mTempe, Arizona\u001b[39m\t\nSCORES: global= 0.240:0.240[0]; local()= 0.057:0.057[0]; log p(e|m)= -0.221:-0.221[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nproducts[0.045]; Friday[0.044]; million[0.043]; million[0.043]; expected[0.043]; sale[0.043]; distribution[0.043]; said[0.043]; closing[0.043]; annual[0.043]; close[0.043]; manufacture[0.043]; cash[0.041]; firms[0.041]; firms[0.041]; Traditions[0.040]; design[0.040]; subject[0.040]; customary[0.040]; conditions[0.040]; Performance[0.040]; Performance[0.040]; agreed[0.040]; Ariz[0.040]; \t\n\r [====>.........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 16s678ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 578/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1165testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syria ==> ENTITY: \u001b[32mSyria national football team\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.154:0.154[0]; log p(e|m)= -3.058:-3.058[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.052]; Syrian[0.049]; Arab[0.048]; Soccer[0.046]; Cup[0.046]; Hassan[0.042]; Emirates[0.042]; Al[0.042]; Asian[0.041]; Abbas[0.041]; group[0.040]; defensive[0.040]; game[0.040]; United[0.040]; second[0.040]; Japan[0.040]; Japan[0.040]; Salem[0.039]; area[0.039]; Bitar[0.039]; Friday[0.039]; long[0.038]; goalkeeper[0.038]; Win[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hiroaki Morishima\u001b[39m ==> ENTITY: \u001b[32mHiroaki Morishima\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nminute[0.047]; minute[0.047]; Japan[0.047]; Japan[0.047]; Naoki[0.043]; Naoki[0.043]; Hiroshi[0.043]; Takuya[0.042]; half[0.042]; Masayuki[0.042]; Kenichi[0.041]; Yamaguchi[0.040]; header[0.040]; header[0.040]; Miura[0.040]; Miura[0.040]; Miura[0.040]; Louay[0.040]; second[0.040]; Norio[0.040]; Yanagimoto[0.040]; Afash[0.040]; Mohammed[0.039]; started[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hiroshige Yanagimoto\u001b[39m ==> ENTITY: \u001b[32mHiroshige Yanagimoto\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.255:0.255[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNaoki[0.044]; Naoki[0.044]; Hiroshi[0.044]; Takuya[0.044]; Kenichi[0.043]; Kazuyoshi[0.043]; Kazuyoshi[0.043]; Masami[0.043]; Norio[0.043]; Masayuki[0.043]; Ihara[0.042]; Japan[0.041]; Japan[0.041]; Japan[0.041]; Hiroaki[0.041]; Nanami[0.041]; Miura[0.040]; Miura[0.040]; Miura[0.040]; Takagi[0.040]; Takagi[0.040]; Yamaguchi[0.040]; Omura[0.039]; Shimokawa[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Masayuki Okano\u001b[39m ==> ENTITY: \u001b[32mMasayuki Okano\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.050]; Japan[0.050]; minute[0.046]; minute[0.046]; Takuya[0.044]; Hiroshi[0.044]; Norio[0.044]; Naoki[0.043]; Naoki[0.043]; dived[0.043]; half[0.043]; Kazuyoshi[0.043]; Syria[0.042]; header[0.042]; header[0.042]; Yanagimoto[0.042]; Afash[0.042]; Miura[0.042]; Miura[0.042]; Masami[0.042]; Morishima[0.042]; saved[0.041]; Louay[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSyrian\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSyria\u001b[39m <---> \u001b[32mSyria national football team\u001b[39m\t\nSCORES: global= 0.252:0.248[\u001b[31m0.003\u001b[39m]; local()= 0.162:0.102[\u001b[31m0.060\u001b[39m]; log p(e|m)= -0.609:-3.170[\u001b[31m2.561\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.052]; Syrian[0.050]; Hassan[0.044]; Ammar[0.044]; area[0.043]; area[0.043]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Awad[0.042]; Abbas[0.041]; started[0.041]; rose[0.040]; second[0.040]; second[0.040]; Bitar[0.040]; Bitar[0.040]; Bitar[0.040]; Bitar[0.040]; Bitar[0.040]; Salem[0.040]; right[0.039]; right[0.039]; right[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSyrian\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSyria\u001b[39m <---> \u001b[32mSyria national football team\u001b[39m\t\nSCORES: global= 0.253:0.252[\u001b[31m0.001\u001b[39m]; local()= 0.197:0.161[\u001b[31m0.036\u001b[39m]; log p(e|m)= -0.609:-3.170[\u001b[31m2.561\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.049]; Syria[0.049]; Syria[0.049]; Syrian[0.047]; Arab[0.046]; Cup[0.044]; Hassan[0.042]; Soccer[0.041]; Al[0.041]; Ammar[0.041]; attack[0.041]; area[0.040]; area[0.040]; Japan[0.040]; Japan[0.040]; Japan[0.040]; Emirates[0.039]; Awad[0.039]; Abbas[0.039]; Asian[0.039]; siege[0.038]; United[0.038]; group[0.038]; Ain[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Takuya Takagi\u001b[39m ==> ENTITY: \u001b[32mTakuya Takagi\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.046]; Japan[0.046]; second[0.044]; low[0.043]; Hiroshi[0.043]; Kazuyoshi[0.042]; Kazuyoshi[0.042]; Yamaguchi[0.042]; Honda[0.042]; started[0.042]; Norio[0.041]; double[0.041]; Hiroaki[0.041]; Afash[0.041]; Masayuki[0.041]; Morishima[0.041]; Yanagimoto[0.040]; Naoki[0.040]; Naoki[0.040]; Kenichi[0.040]; header[0.040]; header[0.040]; denied[0.040]; Louay[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Asian Cup ==> ENTITY: \u001b[32mAFC Asian Cup\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.179:0.179[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngoals[0.047]; goal[0.047]; Syria[0.044]; Syria[0.044]; Syria[0.044]; Soccer[0.044]; Friday[0.043]; Japan[0.043]; Japan[0.043]; Emirates[0.041]; Goals[0.041]; Syrian[0.041]; second[0.041]; game[0.040]; Arab[0.040]; Hassan[0.040]; attack[0.040]; holders[0.040]; minute[0.039]; minute[0.039]; minute[0.039]; victory[0.039]; minutes[0.039]; minutes[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Naoki Soma\u001b[39m ==> ENTITY: \u001b[32mNaoki Soma\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.044]; Japan[0.044]; Japan[0.044]; Soma[0.044]; Naoki[0.043]; Takuya[0.043]; Hiroshi[0.042]; Kazuyoshi[0.042]; Kazuyoshi[0.042]; offside[0.042]; goal[0.041]; goal[0.041]; Kenichi[0.041]; Kenichi[0.041]; Norio[0.041]; Yamaguchi[0.040]; Masayuki[0.040]; Masami[0.040]; Morishima[0.040]; Yanagimoto[0.040]; second[0.040]; Motohiro[0.040]; Hiroaki[0.040]; Honda[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Motohiro Yamaguchi\u001b[39m ==> ENTITY: \u001b[32mMotohiro Yamaguchi\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.047]; Japan[0.047]; Naoki[0.044]; Naoki[0.044]; Miura[0.043]; Miura[0.043]; Miura[0.043]; Yanagimoto[0.043]; Masayuki[0.042]; Takuya[0.042]; Norio[0.042]; Kazuyoshi[0.041]; Kazuyoshi[0.041]; Hiroshi[0.041]; Hiroaki[0.040]; produced[0.040]; Nanami[0.040]; Masami[0.040]; offside[0.040]; Honda[0.040]; half[0.040]; Shimokawa[0.040]; header[0.039]; header[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Takagi\u001b[39m ==> ENTITY: \u001b[32mTakuya Takagi\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.045]; Japan[0.045]; Japan[0.045]; Takuya[0.044]; second[0.043]; seventh[0.043]; low[0.042]; Hiroshi[0.042]; long[0.042]; Kazuyoshi[0.041]; Kazuyoshi[0.041]; goal[0.041]; goal[0.041]; goal[0.041]; Yamaguchi[0.041]; Honda[0.041]; started[0.041]; Norio[0.040]; double[0.040]; ball[0.040]; Hiroaki[0.040]; offside[0.040]; Masayuki[0.040]; Morishima[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syria ==> ENTITY: \u001b[32mSyria national football team\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.144:0.144[0]; log p(e|m)= -3.058:-3.058[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.051]; Syrian[0.048]; Syrian[0.048]; Arab[0.048]; Cup[0.045]; shot[0.042]; Hassan[0.042]; Ammar[0.042]; seventh[0.041]; Emirates[0.041]; Asian[0.041]; Abbas[0.040]; second[0.040]; second[0.040]; defensive[0.040]; game[0.040]; Awad[0.039]; Japan[0.039]; Japan[0.039]; Japan[0.039]; Salem[0.039]; lead[0.039]; area[0.039]; area[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syria ==> ENTITY: \u001b[32mSyria national football team\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.100:0.100[0]; log p(e|m)= -3.058:-3.058[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nal[0.045]; Afash[0.044]; Japan[0.044]; Japan[0.044]; Hassan[0.044]; Ammar[0.044]; Ammar[0.044]; Abbas[0.043]; Mohammed[0.042]; Awad[0.042]; Zaher[0.041]; Tarek[0.041]; Louay[0.041]; Ali[0.041]; Nader[0.041]; Khaled[0.041]; Salem[0.041]; Helou[0.041]; Dib[0.040]; Bitar[0.040]; Bitar[0.040]; Bitar[0.040]; Bachar[0.040]; Srour[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syria ==> ENTITY: \u001b[32mSyria national football team\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.168:0.168[0]; log p(e|m)= -3.058:-3.058[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.053]; Syria[0.053]; Syrian[0.050]; Arab[0.049]; Soccer[0.047]; Cup[0.047]; Al[0.043]; Hassan[0.043]; Ammar[0.043]; Emirates[0.042]; Asian[0.042]; Abbas[0.042]; winner[0.041]; defensive[0.041]; game[0.041]; United[0.041]; second[0.041]; Awad[0.041]; Japan[0.040]; Japan[0.040]; Salem[0.040]; group[0.040]; area[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Naoki Soma\u001b[39m ==> ENTITY: \u001b[32mNaoki Soma\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.125:0.125[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.045]; Japan[0.045]; Soma[0.045]; Naoki[0.044]; Takuya[0.043]; Hiroshi[0.042]; Kazuyoshi[0.042]; Kazuyoshi[0.042]; offside[0.042]; goal[0.042]; Kenichi[0.042]; Norio[0.042]; Yamaguchi[0.041]; Masayuki[0.041]; Masami[0.041]; Morishima[0.041]; Yanagimoto[0.041]; second[0.041]; Motohiro[0.040]; Hiroaki[0.040]; Honda[0.040]; header[0.040]; header[0.040]; Louay[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hassan Abbas\u001b[39m ==> ENTITY: \u001b[32mHassan Abbas\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.047]; Afash[0.046]; Ammar[0.043]; Ammar[0.043]; Mohammed[0.043]; Ali[0.043]; al[0.042]; Khaled[0.042]; Salem[0.042]; Tarek[0.042]; Nader[0.042]; Japan[0.041]; Japan[0.041]; Zaher[0.041]; Awad[0.041]; started[0.041]; Dib[0.041]; Nihad[0.040]; Morishima[0.040]; Srour[0.040]; header[0.040]; Louay[0.040]; Hiroaki[0.039]; second[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Al Ain\u001b[39m ==> ENTITY: \u001b[32mAl Ain\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEmirates[0.046]; area[0.045]; Arab[0.044]; Soccer[0.044]; Cup[0.043]; Hassan[0.043]; Syria[0.042]; Syria[0.042]; second[0.042]; goals[0.042]; Ain[0.042]; managed[0.041]; United[0.041]; game[0.041]; goalkeeper[0.041]; Syrian[0.040]; Defender[0.040]; Asian[0.040]; winner[0.040]; long[0.040]; free[0.040]; defensive[0.040]; victory[0.040]; Friday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hiroshi Nanami\u001b[39m ==> ENTITY: \u001b[32mHiroshi Nanami\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.046]; Japan[0.046]; Naoki[0.045]; Naoki[0.045]; Kenichi[0.045]; Yanagimoto[0.044]; started[0.044]; Kazuyoshi[0.044]; Kazuyoshi[0.044]; second[0.044]; Yamaguchi[0.044]; Hiroaki[0.043]; Morishima[0.043]; Masayuki[0.043]; Honda[0.043]; header[0.042]; header[0.042]; Norio[0.042]; Takuya[0.042]; Louay[0.042]; Shimokawa[0.042]; Soma[0.042]; Soma[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mJapan\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mJapan\u001b[39m <---> \u001b[32mJapan national football team\u001b[39m\t\nSCORES: global= 0.250:0.246[\u001b[31m0.003\u001b[39m]; local()= 0.131:0.151[\u001b[32m0.020\u001b[39m]; log p(e|m)= -0.136:-3.730[\u001b[31m3.594\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.055]; Japan[0.055]; Honda[0.048]; second[0.047]; goal[0.046]; goal[0.046]; Kenichi[0.045]; Kenichi[0.045]; Hiroshi[0.045]; half[0.045]; Yamaguchi[0.045]; started[0.044]; drive[0.044]; Takuya[0.044]; Ihara[0.044]; produced[0.044]; siege[0.043]; penalty[0.043]; Hiroaki[0.043]; Naoki[0.043]; Naoki[0.043]; Motohiro[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan national football team\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.165:0.165[0]; log p(e|m)= -3.730:-3.730[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.049]; Japan[0.049]; second[0.045]; second[0.045]; game[0.044]; goal[0.043]; goal[0.043]; goal[0.043]; seventh[0.041]; penalty[0.040]; half[0.040]; started[0.040]; minute[0.040]; minute[0.040]; minute[0.040]; minute[0.040]; minute[0.040]; minute[0.040]; minute[0.040]; lead[0.040]; drive[0.039]; group[0.039]; minutes[0.039]; produced[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Takuya Takagi\u001b[39m ==> ENTITY: \u001b[32mTakuya Takagi\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.045]; Japan[0.045]; Japan[0.045]; game[0.044]; second[0.043]; seventh[0.043]; long[0.042]; goals[0.042]; Asian[0.042]; Emirates[0.042]; blunder[0.042]; goal[0.041]; Cup[0.041]; mistake[0.041]; defensive[0.041]; ball[0.041]; ball[0.041]; United[0.040]; minutes[0.040]; minutes[0.040]; victory[0.040]; Goals[0.040]; gave[0.040]; goalkeeper[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Masakiyo Maezono\u001b[39m ==> ENTITY: \u001b[32mMasakiyo Maezono\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.035:0.035[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.045]; Japan[0.045]; Miura[0.043]; Miura[0.043]; Miura[0.043]; Honda[0.042]; Morishima[0.042]; header[0.042]; header[0.042]; Hiroshi[0.042]; Louay[0.042]; Hiroaki[0.041]; Yanagimoto[0.041]; Shimokawa[0.041]; Ihara[0.041]; Afash[0.041]; minute[0.041]; minute[0.041]; minute[0.041]; minute[0.041]; Takuya[0.040]; blocked[0.040]; second[0.040]; Yamaguchi[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Miura\u001b[39m ==> ENTITY: \u001b[32mKazuyoshi Miura\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.209:0.209[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.056]; Japan[0.056]; Japan[0.056]; second[0.055]; goal[0.054]; goal[0.054]; goal[0.054]; dived[0.052]; Honda[0.052]; half[0.052]; Takuya[0.052]; Kazuyoshi[0.051]; Kazuyoshi[0.051]; seventh[0.051]; offside[0.051]; lead[0.051]; Miura[0.051]; Miura[0.051]; Hiroshi[0.051]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tarek Jabban\u001b[39m ==> ENTITY: \u001b[32mTarek Jabban\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.050]; Afash[0.045]; started[0.043]; Khaled[0.042]; Nihad[0.042]; al[0.042]; Abbas[0.042]; Mohammed[0.042]; Ali[0.042]; Ammar[0.041]; Ammar[0.041]; Hassan[0.041]; Nader[0.041]; Bitar[0.041]; Bitar[0.041]; Morishima[0.041]; Srour[0.041]; Awad[0.041]; header[0.040]; Salem[0.040]; Louay[0.040]; Hiroaki[0.040]; denied[0.040]; Yanagimoto[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Khaled Zaher\u001b[39m ==> ENTITY: \u001b[32mKhaled Al Zaher\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.047]; Afash[0.044]; Mohammed[0.044]; Ammar[0.043]; Ammar[0.043]; Tarek[0.043]; al[0.043]; Dib[0.043]; Hassan[0.043]; Salem[0.042]; Abbas[0.042]; Ali[0.041]; Taleb[0.041]; Nihad[0.041]; Nader[0.041]; Bitar[0.041]; Japan[0.040]; Bachar[0.040]; Awad[0.040]; Latif[0.040]; Morishima[0.040]; Helou[0.039]; Louay[0.039]; Hiroaki[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Masami Ihara\u001b[39m ==> ENTITY: \u001b[32mMasami Ihara\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNaoki[0.045]; Naoki[0.045]; Hiroshi[0.045]; Japan[0.044]; Japan[0.044]; Kazuyoshi[0.043]; Kazuyoshi[0.043]; Takuya[0.043]; Kenichi[0.042]; Norio[0.042]; Soma[0.042]; Soma[0.042]; Yanagimoto[0.041]; Masayuki[0.041]; Motohiro[0.040]; Morishima[0.040]; Honda[0.040]; Miura[0.040]; Miura[0.040]; Miura[0.040]; Yamaguchi[0.040]; header[0.040]; header[0.040]; Hiroaki[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kazuyoshi Miura\u001b[39m ==> ENTITY: \u001b[32mKazuyoshi Miura\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.206:0.206[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.053]; Japan[0.053]; Japan[0.053]; second[0.053]; goal[0.051]; goal[0.051]; goal[0.051]; dived[0.050]; Honda[0.050]; half[0.049]; Takuya[0.049]; Kazuyoshi[0.049]; seventh[0.049]; minutes[0.049]; offside[0.049]; lead[0.049]; Miura[0.048]; Miura[0.048]; Hiroshi[0.048]; long[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hassan Abbas\u001b[39m ==> ENTITY: \u001b[32mHassan Abbas\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.227:0.227[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.047]; Syrian[0.045]; Syrian[0.045]; Syria[0.045]; Syria[0.045]; Syria[0.045]; Emirates[0.043]; Soccer[0.042]; Al[0.042]; Arab[0.042]; Ammar[0.042]; game[0.041]; United[0.041]; goals[0.040]; victory[0.040]; Salem[0.040]; Nader[0.040]; penalty[0.040]; goal[0.040]; goal[0.040]; goal[0.040]; Japan[0.039]; Japan[0.039]; Japan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kazuyoshi Miura\u001b[39m ==> ENTITY: \u001b[32mKazuyoshi Miura\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.048]; Japan[0.048]; second[0.047]; dived[0.045]; Honda[0.044]; half[0.044]; Takuya[0.044]; Kazuyoshi[0.044]; Miura[0.043]; Miura[0.043]; Hiroshi[0.043]; minute[0.043]; minute[0.043]; low[0.043]; corner[0.043]; Naoki[0.042]; Naoki[0.042]; started[0.042]; Norio[0.042]; Kenichi[0.042]; brightly[0.042]; Yanagimoto[0.041]; Yamaguchi[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan national football team\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.161:0.161[0]; log p(e|m)= -3.730:-3.730[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.051]; second[0.046]; Soccer[0.045]; game[0.045]; Cup[0.044]; goals[0.043]; Asian[0.042]; minute[0.041]; minute[0.041]; victory[0.041]; Takuya[0.041]; winner[0.041]; group[0.041]; minutes[0.041]; minutes[0.041]; Emirates[0.040]; United[0.040]; Defender[0.040]; ball[0.040]; ball[0.040]; holders[0.039]; allowing[0.039]; Friday[0.039]; long[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United Arab Emirates ==> ENTITY: \u001b[32mUnited Arab Emirates\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.250:-0.250[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.045]; Syria[0.044]; Syria[0.044]; Syria[0.044]; Asian[0.044]; Soccer[0.043]; Syrian[0.043]; holders[0.042]; game[0.042]; second[0.042]; Al[0.042]; Hassan[0.042]; victory[0.041]; Japan[0.041]; Japan[0.041]; long[0.040]; Salem[0.040]; group[0.040]; goal[0.040]; free[0.040]; Friday[0.040]; blunder[0.040]; winner[0.040]; rose[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kenichi Shimokawa\u001b[39m ==> ENTITY: \u001b[32mKenichi Shimokawa\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.047]; Japan[0.047]; Japan[0.047]; Shimokawa[0.047]; Kenichi[0.046]; Naoki[0.046]; Naoki[0.046]; Takuya[0.044]; Norio[0.044]; Hiroshi[0.043]; Masami[0.043]; Masayuki[0.043]; Kazuyoshi[0.042]; Kazuyoshi[0.042]; Miura[0.042]; Miura[0.042]; Miura[0.042]; half[0.042]; Yamaguchi[0.041]; disallowed[0.041]; Hiroshige[0.041]; drive[0.041]; second[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yasuto Honda\u001b[39m ==> ENTITY: \u001b[32mYasuto Honda\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.048]; Japan[0.048]; Kazuyoshi[0.043]; Kazuyoshi[0.043]; Norio[0.043]; Takuya[0.043]; Hiroshi[0.043]; Naoki[0.042]; Naoki[0.042]; Yanagimoto[0.042]; Masayuki[0.041]; Afash[0.041]; started[0.041]; Yamaguchi[0.040]; header[0.040]; header[0.040]; second[0.040]; half[0.040]; Nanami[0.040]; minute[0.040]; minute[0.040]; minute[0.040]; minute[0.040]; Hiroaki[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mJapan\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mJapan\u001b[39m <---> \u001b[32mJapan national football team\u001b[39m\t\nSCORES: global= 0.245:0.245[\u001b[31m0.001\u001b[39m]; local()= 0.107:0.161[\u001b[32m0.055\u001b[39m]; log p(e|m)= -0.136:-3.730[\u001b[31m3.594\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.050]; Japan[0.050]; second[0.045]; goal[0.043]; goal[0.043]; goal[0.043]; seventh[0.041]; Kenichi[0.041]; Kenichi[0.041]; penalty[0.041]; Hiroshi[0.041]; half[0.041]; Yamaguchi[0.041]; started[0.040]; minute[0.040]; minute[0.040]; minute[0.040]; minute[0.040]; minute[0.040]; lead[0.040]; Takuya[0.040]; Ihara[0.040]; produced[0.040]; disallowed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kenichi Shimokawa\u001b[39m ==> ENTITY: \u001b[32mKenichi Shimokawa\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.046]; Japan[0.046]; Japan[0.046]; Shimokawa[0.046]; Kenichi[0.045]; Naoki[0.045]; game[0.043]; minutes[0.042]; Kazuyoshi[0.041]; Miura[0.041]; Miura[0.041]; half[0.041]; disallowed[0.040]; drive[0.040]; second[0.040]; second[0.040]; Takagi[0.040]; defensive[0.040]; ball[0.040]; ball[0.040]; started[0.040]; dived[0.039]; managed[0.039]; long[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Norio Omura\u001b[39m ==> ENTITY: \u001b[32mNorio Omura\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.048]; Japan[0.048]; goal[0.044]; Naoki[0.043]; Naoki[0.043]; minute[0.042]; minute[0.042]; minute[0.042]; minute[0.042]; Takuya[0.042]; Hiroshi[0.042]; Masayuki[0.042]; half[0.041]; Honda[0.040]; disallowed[0.040]; Kazuyoshi[0.040]; Kazuyoshi[0.040]; drive[0.040]; Yamaguchi[0.040]; second[0.040]; Miura[0.040]; Miura[0.040]; Miura[0.040]; started[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan national football team\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.166:0.166[0]; log p(e|m)= -3.730:-3.730[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.050]; second[0.045]; Soccer[0.045]; game[0.045]; Cup[0.044]; goal[0.044]; goals[0.042]; seventh[0.042]; Asian[0.041]; minute[0.041]; minute[0.041]; minute[0.041]; victory[0.041]; lead[0.040]; Takuya[0.040]; winner[0.040]; group[0.040]; minutes[0.040]; minutes[0.040]; Emirates[0.040]; United[0.040]; Defender[0.040]; ball[0.039]; ball[0.039]; \t\n\r [=====>........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 16s420ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 615/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1259testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rio De Janeiro\u001b[39m ==> ENTITY: \u001b[32mRio de Janeiro\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.113:0.113[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.048]; Brazil[0.048]; Surgery[0.046]; Brazilian[0.045]; Brazilians[0.043]; surgeries[0.042]; surgery[0.042]; surgery[0.042]; surgery[0.042]; July[0.041]; latest[0.041]; number[0.041]; number[0.041]; conference[0.041]; Simona[0.040]; international[0.040]; jumped[0.040]; boost[0.040]; president[0.040]; Society[0.039]; treatments[0.039]; operations[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazil ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.152:0.152[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.048]; Brazil[0.048]; Brazil[0.048]; match[0.045]; match[0.045]; match[0.045]; country[0.044]; Latin[0.044]; qualified[0.042]; famous[0.042]; America[0.042]; compete[0.041]; world[0.040]; decades[0.040]; People[0.040]; changes[0.039]; rate[0.039]; come[0.039]; create[0.039]; areas[0.038]; home[0.038]; economy[0.038]; races[0.038]; does[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazil ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.157:0.157[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.048]; Brazil[0.048]; Brazil[0.048]; match[0.044]; match[0.044]; match[0.044]; country[0.044]; Latin[0.043]; popular[0.042]; qualified[0.042]; famous[0.042]; America[0.041]; world[0.040]; changes[0.040]; decades[0.040]; Ivo[0.040]; tropical[0.039]; rate[0.039]; create[0.039]; does[0.038]; come[0.038]; areas[0.038]; home[0.038]; economy[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Latin America\u001b[39m ==> ENTITY: \u001b[32mLatin America\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.046]; Brazil[0.046]; Brazil[0.046]; Brazil[0.046]; trade[0.045]; policies[0.045]; world[0.043]; decades[0.042]; respect[0.042]; available[0.042]; increased[0.041]; monetary[0.041]; economy[0.040]; profound[0.040]; believe[0.040]; difficulties[0.040]; exchange[0.040]; statistics[0.040]; given[0.039]; famous[0.039]; changes[0.039]; operations[0.039]; operations[0.039]; attributes[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazil ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.163:0.163[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.048]; Brazil[0.048]; match[0.044]; match[0.044]; match[0.044]; country[0.044]; country[0.044]; Oswaldo[0.043]; Brazilians[0.042]; popular[0.042]; States[0.042]; famous[0.041]; cultures[0.041]; probably[0.041]; world[0.040]; qualified[0.040]; want[0.039]; want[0.039]; tropical[0.039]; United[0.039]; does[0.039]; create[0.039]; come[0.039]; areas[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazil ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.050]; Brazilian[0.047]; Rio[0.045]; Brazilians[0.044]; international[0.044]; conference[0.044]; Janeiro[0.043]; day[0.042]; Friday[0.042]; July[0.041]; number[0.041]; number[0.041]; men[0.041]; men[0.041]; president[0.041]; especially[0.040]; percent[0.040]; percent[0.040]; booming[0.039]; said[0.039]; said[0.039]; estimated[0.039]; increased[0.039]; wealth[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazil ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.149:0.149[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.055]; Brazil[0.055]; Brazilian[0.050]; Rio[0.049]; international[0.049]; highest[0.048]; men[0.048]; men[0.048]; Brazilians[0.048]; conference[0.048]; boost[0.047]; ranks[0.047]; Janeiro[0.046]; respect[0.046]; increased[0.046]; jumped[0.046]; day[0.045]; Friday[0.045]; percent[0.045]; percent[0.045]; July[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazilians ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.083:0.083[0]; log p(e|m)= -0.384:-0.384[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.047]; Brazilian[0.046]; liposuction[0.045]; liposuction[0.045]; country[0.043]; country[0.043]; buttocks[0.043]; breasts[0.042]; women[0.042]; women[0.042]; implants[0.042]; Oswaldo[0.042]; cultures[0.042]; treatment[0.041]; Claudia[0.040]; buttock[0.040]; breast[0.040]; breast[0.040]; rate[0.040]; percent[0.039]; areas[0.039]; legs[0.039]; tend[0.039]; popular[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazil ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.135:0.135[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.048]; Brazil[0.048]; policies[0.045]; Latin[0.045]; match[0.044]; match[0.044]; country[0.044]; trade[0.043]; America[0.041]; operations[0.041]; exchange[0.041]; economy[0.040]; respect[0.040]; need[0.040]; monetary[0.040]; world[0.040]; decades[0.040]; People[0.040]; qualified[0.039]; conferences[0.039]; number[0.039]; number[0.039]; number[0.039]; number[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazilians ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.384:-0.384[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.048]; Brazil[0.048]; Brazil[0.048]; Brazilian[0.047]; treatments[0.043]; number[0.042]; number[0.042]; especially[0.041]; international[0.041]; plastic[0.041]; plastic[0.041]; plastic[0.041]; conference[0.041]; estimated[0.040]; percent[0.040]; percent[0.040]; Society[0.040]; Rio[0.040]; ranks[0.040]; anti[0.039]; said[0.039]; said[0.039]; said[0.039]; increased[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.056:0.056[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.048]; match[0.048]; country[0.046]; Brazil[0.045]; women[0.043]; women[0.043]; probably[0.042]; said[0.042]; said[0.042]; said[0.042]; October[0.041]; Brazilians[0.041]; percent[0.041]; fell[0.040]; fell[0.040]; Oswaldo[0.040]; sought[0.040]; Brazilian[0.040]; cultures[0.040]; tend[0.040]; ideals[0.040]; increase[0.039]; Liz[0.039]; popular[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazil ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.110:0.110[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.047]; international[0.046]; policies[0.044]; Latin[0.044]; Brazilian[0.043]; trade[0.043]; highest[0.042]; America[0.042]; compete[0.042]; men[0.041]; men[0.041]; men[0.041]; Brazilians[0.041]; conference[0.041]; operations[0.040]; operations[0.040]; ranks[0.040]; exchange[0.040]; president[0.040]; economy[0.040]; respect[0.040]; need[0.040]; monetary[0.040]; increased[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazilian ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.104:0.104[0]; log p(e|m)= -0.273:-0.273[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazilians[0.047]; country[0.045]; cultures[0.045]; Oswaldo[0.044]; different[0.043]; probably[0.043]; tropical[0.043]; Claudia[0.042]; United[0.042]; said[0.042]; said[0.042]; women[0.042]; women[0.042]; States[0.041]; want[0.041]; want[0.041]; popular[0.041]; like[0.040]; rate[0.040]; percent[0.040]; October[0.040]; body[0.040]; immediately[0.040]; unlike[0.039]; \t\n\r [=====>........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 16s952ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 628/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1347testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tarpon Springs\u001b[39m ==> ENTITY: \u001b[32mTarpon Springs, Florida\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.098:0.098[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFlorida[0.047]; history[0.045]; million[0.043]; said[0.043]; Classic[0.042]; Classic[0.042]; Sunday[0.042]; JCPenney[0.042]; Resort[0.042]; event[0.042]; tournament[0.042]; Saturday[0.042]; time[0.041]; Pat[0.041]; played[0.041]; Andrews[0.041]; holes[0.041]; holes[0.041]; team[0.040]; Golf[0.040]; Officials[0.040]; rains[0.040]; washed[0.040]; special[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Donna Andrews\u001b[39m ==> ENTITY: \u001b[32mDonna Andrews (golfer)\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmateur[0.046]; played[0.045]; tournament[0.043]; team[0.043]; Golf[0.043]; players[0.042]; Woods[0.042]; Innisbrook[0.042]; champions[0.042]; champions[0.042]; Classic[0.042]; Classic[0.042]; history[0.042]; event[0.041]; Sunday[0.041]; Pat[0.040]; Resort[0.040]; start[0.040]; Florida[0.040]; JCPenney[0.040]; Kelli[0.040]; Saturday[0.040]; Daniel[0.040]; Tarpon[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Davis Love\u001b[39m ==> ENTITY: \u001b[32mDavis Love III\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmateur[0.048]; tournament[0.047]; Woods[0.045]; round[0.045]; round[0.045]; played[0.043]; par[0.042]; Tiger[0.042]; Scott[0.042]; players[0.041]; team[0.041]; event[0.041]; holes[0.040]; holes[0.040]; history[0.040]; final[0.040]; final[0.040]; Mike[0.040]; shot[0.040]; shot[0.040]; tied[0.040]; Kelli[0.039]; Hurst[0.039]; start[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mike Hulbert\u001b[39m ==> ENTITY: \u001b[32mMike Hulbert\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nround[0.045]; round[0.045]; round[0.045]; Golf[0.044]; Amateur[0.044]; Classic[0.044]; Classic[0.044]; tournament[0.044]; played[0.042]; players[0.042]; final[0.041]; final[0.041]; team[0.041]; Innisbrook[0.041]; event[0.040]; par[0.040]; Davis[0.040]; start[0.040]; Woods[0.040]; Round[0.040]; history[0.040]; tied[0.039]; Kuehne[0.039]; JCPenney[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pat Hurst\u001b[39m ==> ENTITY: \u001b[32mPat Hurst\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmateur[0.047]; Golf[0.045]; Classic[0.045]; Classic[0.045]; played[0.044]; tournament[0.044]; round[0.043]; round[0.043]; round[0.043]; Innisbrook[0.042]; players[0.041]; tied[0.040]; par[0.040]; Andrews[0.040]; history[0.040]; final[0.040]; final[0.040]; Woods[0.040]; JCPenney[0.040]; Kelli[0.040]; time[0.040]; Tiger[0.040]; Tarpon[0.040]; Davis[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.049:0.049[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.046]; final[0.044]; final[0.044]; players[0.044]; played[0.044]; tournament[0.044]; format[0.043]; event[0.043]; champions[0.043]; champions[0.043]; Amateur[0.042]; Love[0.042]; round[0.041]; round[0.041]; round[0.041]; Classic[0.041]; reigning[0.040]; Davis[0.040]; time[0.040]; million[0.039]; history[0.039]; alternate[0.039]; Scott[0.039]; start[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = JCPENNEY CLASSIC\u001b[39m ==> ENTITY: \u001b[32mJCPenney Classic\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.115:0.115[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.050]; JCPenney[0.046]; players[0.045]; Innisbrook[0.045]; team[0.044]; Classic[0.044]; tournament[0.044]; Florida[0.043]; played[0.043]; special[0.041]; Tarpon[0.041]; Hurst[0.040]; Andrews[0.040]; Resort[0.040]; pairs[0.040]; Springs[0.040]; Pat[0.040]; holes[0.039]; said[0.039]; Tours[0.039]; history[0.039]; reduced[0.039]; rains[0.039]; McCarron[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = PGA ==> ENTITY: \u001b[32mPGA Tour\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.216:0.216[0]; log p(e|m)= -1.766:-1.766[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.047]; tournament[0.047]; round[0.046]; round[0.046]; round[0.046]; Amateur[0.044]; Classic[0.042]; Classic[0.042]; players[0.042]; event[0.041]; Woods[0.041]; team[0.041]; played[0.041]; final[0.040]; final[0.040]; Innisbrook[0.040]; Andrews[0.040]; history[0.039]; Hurst[0.039]; holes[0.039]; holes[0.039]; Tours[0.039]; Sunday[0.039]; Pat[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kelli Kuehne\u001b[39m ==> ENTITY: \u001b[32mKelli Kuehne\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmateur[0.046]; tournament[0.045]; team[0.043]; played[0.043]; Classic[0.042]; final[0.042]; final[0.042]; players[0.042]; champions[0.042]; champions[0.042]; tied[0.042]; event[0.042]; round[0.041]; round[0.041]; round[0.041]; Tiger[0.041]; Woods[0.041]; Pat[0.041]; Hurst[0.041]; Scott[0.041]; JCPenney[0.040]; Mike[0.040]; shot[0.040]; shot[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Scott McCarron\u001b[39m ==> ENTITY: \u001b[32mScott McCarron\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.129:0.129[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.045]; Amateur[0.045]; tournament[0.043]; Kuehne[0.043]; Mike[0.043]; Andrews[0.042]; Woods[0.042]; round[0.042]; round[0.042]; round[0.042]; Classic[0.041]; Classic[0.041]; par[0.041]; Defending[0.041]; players[0.041]; played[0.041]; Tiger[0.041]; JCPenney[0.041]; Davis[0.041]; Kelli[0.041]; team[0.041]; Innisbrook[0.040]; Pat[0.040]; Tarpon[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = JCPenney Classic\u001b[39m ==> ENTITY: \u001b[32mJCPenney Classic\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.049]; players[0.045]; Innisbrook[0.044]; team[0.044]; Classic[0.044]; Amateur[0.044]; tournament[0.043]; Florida[0.043]; Woods[0.042]; played[0.042]; Kuehne[0.042]; special[0.041]; Kelli[0.041]; Tarpon[0.040]; Hurst[0.040]; Andrews[0.040]; Resort[0.040]; Tiger[0.040]; pairs[0.040]; Springs[0.040]; Pat[0.040]; holes[0.039]; holes[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tiger Woods\u001b[39m ==> ENTITY: \u001b[32mTiger Woods\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.046]; Amateur[0.046]; round[0.043]; round[0.043]; round[0.043]; players[0.043]; Innisbrook[0.043]; event[0.043]; final[0.042]; final[0.042]; team[0.042]; played[0.042]; Scott[0.041]; Classic[0.041]; Andrews[0.041]; million[0.040]; par[0.040]; Love[0.040]; Sunday[0.040]; lead[0.040]; alternate[0.040]; history[0.040]; reigning[0.039]; Davis[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mLPGA\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mLPGA\u001b[39m <---> \u001b[32mLPGA Championship\u001b[39m\t\nSCORES: global= 0.293:0.277[\u001b[31m0.016\u001b[39m]; local()= 0.202:0.165[\u001b[31m0.037\u001b[39m]; log p(e|m)= 0.000:-0.685[\u001b[31m0.685\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.048]; Amateur[0.047]; tournament[0.047]; Classic[0.045]; Classic[0.045]; round[0.045]; round[0.045]; round[0.045]; played[0.044]; Woods[0.044]; players[0.043]; event[0.043]; Andrews[0.043]; Innisbrook[0.043]; team[0.042]; Hurst[0.042]; Tiger[0.042]; time[0.042]; history[0.041]; final[0.041]; final[0.041]; start[0.041]; Sunday[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Beth Daniel\u001b[39m ==> ENTITY: \u001b[32mBeth Daniel\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.045]; Amateur[0.045]; players[0.045]; team[0.043]; played[0.043]; time[0.043]; history[0.043]; round[0.043]; round[0.043]; Kuehne[0.042]; tied[0.041]; start[0.041]; final[0.041]; final[0.041]; Woods[0.041]; lead[0.041]; Hurst[0.041]; Kelli[0.041]; Tiger[0.040]; Andrews[0.040]; Scott[0.040]; Resort[0.040]; event[0.040]; Davis[0.040]; \t\n\r [=====>........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 16s902ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 642/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1275testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = GMT\u001b[39m ==> ENTITY: \u001b[32mGreenwich Mean Time\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.085:0.085[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFriday[0.051]; Speaking[0.049]; John[0.048]; Department[0.047]; Glenn[0.047]; said[0.046]; said[0.046]; said[0.046]; spokesman[0.045]; official[0.045]; switch[0.045]; final[0.045]; quoted[0.045]; chair[0.044]; Corrections[0.044]; statement[0.044]; conversion[0.044]; saying[0.044]; officials[0.044]; Fla[0.044]; sent[0.044]; Arabic[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Islam ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.168:0.168[0]; log p(e|m)= -0.121:-0.121[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAllah[0.052]; prophet[0.048]; God[0.047]; God[0.047]; Muhammed[0.045]; spiritual[0.045]; Mohammed[0.045]; Abdullah[0.044]; saying[0.040]; added[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; sent[0.040]; Yuhanna[0.039]; family[0.039]; asked[0.039]; conversion[0.038]; conversion[0.038]; bear[0.038]; bear[0.038]; present[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.S. Supreme Court\u001b[39m ==> ENTITY: \u001b[32mSupreme Court of the United States\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncourt[0.045]; Court[0.044]; Court[0.044]; Court[0.044]; federal[0.043]; appeal[0.043]; Appeals[0.043]; sentence[0.042]; Supreme[0.042]; reversed[0.042]; Michael[0.041]; prison[0.041]; saying[0.041]; Frederick[0.041]; Circuit[0.041]; request[0.041]; denied[0.041]; execution[0.040]; Department[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mohammed ==> ENTITY: \u001b[32mMuhammad\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.269:-0.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArabic[0.045]; Allah[0.045]; prophet[0.045]; Muhammed[0.044]; Islam[0.044]; final[0.044]; Abdullah[0.043]; sent[0.042]; official[0.042]; God[0.042]; God[0.042]; asked[0.041]; file[0.041]; officials[0.041]; officials[0.041]; saying[0.040]; father[0.040]; Friday[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; added[0.040]; quoted[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Florida Supreme Court\u001b[39m ==> ENTITY: \u001b[32mSupreme Court of Florida\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nappeal[0.047]; Appeals[0.047]; court[0.046]; Court[0.046]; Court[0.046]; serving[0.045]; sentence[0.045]; sentence[0.045]; request[0.045]; federal[0.044]; Florida[0.043]; death[0.043]; Supreme[0.042]; prison[0.042]; said[0.042]; said[0.042]; said[0.042]; March[0.042]; file[0.042]; reversed[0.042]; Wednesday[0.041]; temporarily[0.041]; scheduled[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tallahassee ==> ENTITY: \u001b[32mTallahassee, Florida\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.091:0.091[0]; log p(e|m)= -0.024:-0.024[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFlorida[0.046]; Florida[0.046]; Florida[0.046]; final[0.043]; home[0.043]; Morris[0.043]; Glenn[0.043]; Department[0.043]; John[0.042]; Fla[0.041]; Friday[0.041]; Mills[0.041]; Mills[0.041]; Mills[0.041]; rural[0.040]; dies[0.040]; Eugene[0.040]; Lawhon[0.040]; Lawhon[0.040]; sent[0.040]; spokesman[0.040]; switch[0.040]; death[0.040]; citizen[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = God ==> ENTITY: \u001b[32mGod\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.199:0.199[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGod[0.051]; Allah[0.049]; prophet[0.046]; Arabic[0.045]; Muhammed[0.043]; Islam[0.043]; messenger[0.042]; Mohammed[0.041]; saying[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; asked[0.040]; Supreme[0.040]; added[0.040]; John[0.040]; person[0.040]; sent[0.040]; bear[0.039]; bear[0.039]; witness[0.039]; witness[0.039]; statement[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mU.S. Court of Appeals\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUnited States Court of Appeals for the Eleventh Circuit\u001b[39m <---> \u001b[32mUnited States courts of appeals\u001b[39m\t\nSCORES: global= 0.245:0.245[\u001b[31m0.000\u001b[39m]; local()= 0.200:0.167[\u001b[31m0.033\u001b[39m]; log p(e|m)= -1.378:-0.872[\u001b[32m0.506\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAppeals[0.046]; Court[0.046]; Court[0.046]; court[0.045]; federal[0.044]; appeal[0.044]; request[0.043]; Circuit[0.042]; reversed[0.041]; Morris[0.041]; March[0.041]; serving[0.041]; file[0.041]; denied[0.041]; Wednesday[0.040]; Supreme[0.040]; Supreme[0.040]; Michael[0.040]; Thursday[0.040]; Thursday[0.040]; Florida[0.040]; Florida[0.040]; temporarily[0.040]; offered[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = God ==> ENTITY: \u001b[32mGod\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.192:0.192[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGod[0.051]; Allah[0.049]; prophet[0.046]; Arabic[0.045]; Muhammed[0.044]; Islam[0.043]; messenger[0.042]; Mohammed[0.041]; saying[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; asked[0.040]; Supreme[0.040]; added[0.040]; person[0.040]; sent[0.040]; bear[0.039]; bear[0.039]; witness[0.039]; witness[0.039]; statement[0.039]; Abdullah[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Arabic ==> ENTITY: \u001b[32mArabic language\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.151:0.151[0]; log p(e|m)= -0.084:-0.084[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMohammed[0.045]; Allah[0.043]; saying[0.043]; Muhammed[0.043]; usually[0.043]; Abdullah[0.043]; added[0.042]; pronounced[0.042]; official[0.042]; sent[0.042]; prophet[0.042]; said[0.042]; said[0.042]; said[0.042]; quoted[0.041]; statement[0.041]; God[0.041]; God[0.041]; file[0.041]; Yuhanna[0.041]; John[0.040]; home[0.040]; record[0.040]; messenger[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlanta ==> ENTITY: \u001b[32mAtlanta\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.087:0.087[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWednesday[0.047]; Thursday[0.046]; Thursday[0.046]; tire[0.045]; Lester[0.045]; Lester[0.045]; Florida[0.044]; Florida[0.044]; postponed[0.044]; spent[0.043]; March[0.043]; Michael[0.043]; serving[0.043]; Morris[0.043]; scheduled[0.043]; officials[0.043]; temporarily[0.042]; indication[0.042]; special[0.042]; Frederick[0.042]; Frederick[0.042]; run[0.042]; taken[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Allah\u001b[39m ==> ENTITY: \u001b[32mAllah\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.207:0.207[0]; log p(e|m)= -0.008:-0.008[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGod[0.048]; God[0.048]; prophet[0.045]; Arabic[0.045]; Muhammed[0.043]; Islam[0.043]; official[0.042]; Mohammed[0.041]; saying[0.041]; minister[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; sent[0.041]; Abdullah[0.041]; Supreme[0.040]; father[0.040]; added[0.040]; asked[0.040]; special[0.040]; messenger[0.040]; bear[0.039]; bear[0.039]; \t\n\r [=====>........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 17s20ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 654/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1387testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canucks ==> ENTITY: \u001b[32mVancouver Canucks\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.209:0.209[0]; log p(e|m)= -0.143:-0.143[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.047]; Hockey[0.047]; Vancouver[0.046]; Vancouver[0.046]; League[0.045]; game[0.045]; game[0.045]; Buffalo[0.044]; Buffalo[0.044]; Canucks[0.044]; overtime[0.044]; Game[0.043]; Sabres[0.043]; penalty[0.042]; Ice[0.042]; win[0.042]; Bure[0.041]; Bure[0.041]; Bure[0.041]; Bure[0.041]; period[0.041]; concussion[0.041]; Pavel[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Buffalo Sabres\u001b[39m ==> ENTITY: \u001b[32mBuffalo Sabres\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.047]; Hockey[0.047]; game[0.044]; game[0.044]; Buffalo[0.044]; League[0.043]; win[0.042]; Vancouver[0.042]; Vancouver[0.042]; overtime[0.042]; period[0.041]; Friday[0.041]; Burke[0.040]; Game[0.040]; Canucks[0.040]; Canucks[0.040]; Bure[0.040]; Bure[0.040]; Bure[0.040]; Bure[0.040]; Bure[0.040]; return[0.040]; left[0.040]; left[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NHL ==> ENTITY: \u001b[32mNational Hockey League\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.204:0.204[0]; log p(e|m)= -0.036:-0.036[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.048]; games[0.048]; Ottawa[0.046]; League[0.045]; game[0.045]; game[0.045]; game[0.045]; Buffalo[0.044]; Buffalo[0.044]; Sabres[0.044]; Vancouver[0.043]; assists[0.043]; Burke[0.043]; Burke[0.043]; win[0.043]; period[0.042]; concussion[0.042]; Saturday[0.042]; National[0.042]; received[0.041]; Wednesday[0.041]; Wednesday[0.041]; goals[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Galley\u001b[39m ==> ENTITY: \u001b[32mGarry Galley\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; Hockey[0.045]; Hockey[0.045]; Buffalo[0.042]; Buffalo[0.042]; Vancouver[0.042]; Vancouver[0.042]; Game[0.042]; Burke[0.042]; Burke[0.042]; game[0.042]; game[0.042]; League[0.041]; Ice[0.041]; Canucks[0.041]; Canucks[0.041]; Sabres[0.041]; Galley[0.040]; Galley[0.040]; Galley[0.040]; assists[0.040]; star[0.040]; concussion[0.040]; forearm[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Garry Galley\u001b[39m ==> ENTITY: \u001b[32mGarry Galley\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.045]; Hockey[0.045]; Buffalo[0.043]; Buffalo[0.043]; Vancouver[0.043]; Vancouver[0.043]; Game[0.043]; Burke[0.042]; game[0.042]; game[0.042]; League[0.042]; Ice[0.042]; Canucks[0.041]; Canucks[0.041]; Sabres[0.041]; Galley[0.041]; Galley[0.041]; Galley[0.041]; star[0.041]; concussion[0.040]; forearm[0.040]; Wednesday[0.040]; Wednesday[0.040]; period[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pavel Bure\u001b[39m ==> ENTITY: \u001b[32mPavel Bure\u001b[39m\t\nSCORES: global= 0.311:0.311[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBure[0.044]; Bure[0.044]; Bure[0.044]; Hockey[0.044]; Hockey[0.044]; Vancouver[0.044]; Vancouver[0.044]; game[0.043]; game[0.043]; win[0.042]; Canucks[0.042]; Canucks[0.042]; star[0.041]; received[0.041]; concussion[0.041]; Ice[0.041]; Sabres[0.041]; Buffalo[0.040]; Buffalo[0.040]; return[0.040]; overtime[0.039]; forearm[0.039]; left[0.039]; left[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bure\u001b[39m ==> ENTITY: \u001b[32mPavel Bure\u001b[39m\t\nSCORES: global= 0.312:0.312[0]; local()= 0.218:0.218[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.048]; Bure[0.046]; Bure[0.046]; Bure[0.046]; Vancouver[0.046]; game[0.045]; game[0.045]; win[0.044]; assists[0.044]; Ottawa[0.043]; goals[0.043]; received[0.043]; concussion[0.043]; Sabres[0.043]; Buffalo[0.042]; Buffalo[0.042]; Brian[0.042]; return[0.042]; overtime[0.041]; forearm[0.041]; left[0.041]; left[0.041]; suspension[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bure\u001b[39m ==> ENTITY: \u001b[32mPavel Bure\u001b[39m\t\nSCORES: global= 0.312:0.312[0]; local()= 0.234:0.234[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; Bure[0.043]; Bure[0.043]; Bure[0.043]; Bure[0.043]; Bure[0.043]; Hockey[0.043]; Hockey[0.043]; Vancouver[0.043]; Vancouver[0.043]; Pavel[0.042]; game[0.042]; game[0.042]; game[0.042]; win[0.041]; Canucks[0.041]; Canucks[0.041]; assists[0.041]; Ottawa[0.040]; star[0.040]; goals[0.040]; received[0.040]; concussion[0.040]; Sabres[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bure\u001b[39m ==> ENTITY: \u001b[32mPavel Bure\u001b[39m\t\nSCORES: global= 0.312:0.312[0]; local()= 0.217:0.217[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBure[0.044]; Bure[0.044]; Bure[0.044]; Bure[0.044]; Hockey[0.044]; Hockey[0.044]; Vancouver[0.043]; Vancouver[0.043]; Pavel[0.042]; game[0.042]; game[0.042]; win[0.042]; Canucks[0.041]; Canucks[0.041]; star[0.041]; received[0.041]; concussion[0.040]; Ice[0.040]; Sabres[0.040]; Buffalo[0.040]; Buffalo[0.040]; Brian[0.040]; return[0.040]; overtime[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NHL ==> ENTITY: \u001b[32mNational Hockey League\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.207:0.207[0]; log p(e|m)= -0.036:-0.036[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.047]; Hockey[0.047]; League[0.043]; game[0.043]; game[0.043]; Vancouver[0.043]; Vancouver[0.043]; Buffalo[0.043]; Buffalo[0.043]; Sabres[0.042]; win[0.041]; Canucks[0.041]; Canucks[0.041]; Bure[0.041]; Bure[0.041]; Bure[0.041]; National[0.040]; Ice[0.040]; received[0.040]; Game[0.039]; return[0.039]; period[0.039]; Wednesday[0.039]; Wednesday[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew York\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew York\u001b[39m <---> \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.236:0.235[\u001b[31m0.001\u001b[39m]; local()= 0.071:0.073[\u001b[32m0.002\u001b[39m]; log p(e|m)= -0.289:-1.790[\u001b[31m1.501\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngame[0.047]; game[0.047]; League[0.045]; Wednesday[0.044]; Wednesday[0.044]; Game[0.044]; star[0.043]; win[0.043]; Buffalo[0.043]; Buffalo[0.043]; Friday[0.042]; overtime[0.042]; National[0.042]; return[0.041]; deliver[0.040]; hit[0.039]; received[0.039]; high[0.039]; suffered[0.039]; minor[0.039]; double[0.039]; Hockey[0.039]; Hockey[0.039]; period[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Galley\u001b[39m ==> ENTITY: \u001b[32mGarry Galley\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; Hockey[0.045]; Ottawa[0.043]; Buffalo[0.043]; Buffalo[0.043]; Vancouver[0.043]; Vancouver[0.043]; Burke[0.042]; Burke[0.042]; game[0.042]; game[0.042]; game[0.042]; League[0.041]; Canucks[0.041]; Sabres[0.041]; Galley[0.041]; Galley[0.041]; Galley[0.041]; assists[0.041]; star[0.040]; concussion[0.040]; forearm[0.040]; Wednesday[0.040]; Wednesday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vancouver ==> ENTITY: \u001b[32mVancouver Canucks\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.212:0.212[0]; log p(e|m)= -3.912:-3.912[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVancouver[0.054]; League[0.053]; Hockey[0.051]; Hockey[0.051]; game[0.048]; game[0.048]; Buffalo[0.048]; Buffalo[0.048]; Canucks[0.048]; Canucks[0.048]; Sabres[0.048]; win[0.047]; period[0.046]; return[0.046]; Wednesday[0.046]; Wednesday[0.046]; Friday[0.046]; Ice[0.045]; Burke[0.045]; Burke[0.045]; York[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = National Hockey League\u001b[39m ==> ENTITY: \u001b[32mNational Hockey League\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.208:0.208[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.049]; game[0.046]; game[0.046]; Buffalo[0.045]; Buffalo[0.045]; Sabres[0.045]; Vancouver[0.044]; Vancouver[0.044]; Canucks[0.043]; Canucks[0.043]; period[0.043]; Burke[0.043]; win[0.043]; concussion[0.043]; Ice[0.042]; suspension[0.042]; Game[0.042]; Bure[0.042]; Bure[0.042]; Bure[0.042]; Bure[0.042]; return[0.042]; overtime[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ottawa ==> ENTITY: \u001b[32mOttawa Senators\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.179:0.179[0]; log p(e|m)= -3.689:-3.689[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.049]; game[0.046]; game[0.046]; Burke[0.044]; Burke[0.044]; Bure[0.044]; Bure[0.044]; Bure[0.044]; home[0.042]; struggling[0.042]; Saturday[0.042]; goals[0.041]; assists[0.041]; Brian[0.041]; return[0.041]; said[0.040]; said[0.040]; opponent[0.039]; opponent[0.039]; left[0.039]; suspension[0.039]; miss[0.039]; hit[0.038]; handing[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brian Burke ==> ENTITY: \u001b[32mBrian Burke (ice hockey)\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.165:-0.165[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.049]; Ottawa[0.047]; Burke[0.046]; Vancouver[0.045]; games[0.044]; Sabres[0.044]; game[0.044]; game[0.044]; game[0.044]; Mr[0.044]; Mr[0.044]; Mr[0.044]; received[0.043]; Buffalo[0.043]; Buffalo[0.043]; Garry[0.043]; National[0.042]; period[0.042]; Wednesday[0.041]; Wednesday[0.041]; win[0.041]; concussion[0.041]; forearm[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bure\u001b[39m ==> ENTITY: \u001b[32mPavel Bure\u001b[39m\t\nSCORES: global= 0.312:0.312[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBure[0.044]; Bure[0.044]; Bure[0.044]; Hockey[0.044]; Hockey[0.044]; Vancouver[0.043]; Vancouver[0.043]; Pavel[0.042]; game[0.042]; game[0.042]; win[0.042]; Canucks[0.041]; Canucks[0.041]; star[0.041]; received[0.041]; concussion[0.041]; Ice[0.041]; Sabres[0.040]; Buffalo[0.040]; Buffalo[0.040]; return[0.040]; overtime[0.039]; left[0.039]; left[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Galley\u001b[39m ==> ENTITY: \u001b[32mGarry Galley\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.047]; Hockey[0.047]; Buffalo[0.044]; Buffalo[0.044]; Vancouver[0.044]; Vancouver[0.044]; Game[0.044]; Burke[0.044]; Burke[0.044]; game[0.044]; game[0.044]; League[0.043]; Ice[0.043]; Canucks[0.043]; Canucks[0.043]; Sabres[0.043]; Galley[0.042]; Galley[0.042]; Galley[0.042]; assists[0.042]; star[0.042]; concussion[0.042]; forearm[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Vancouver Canucks\u001b[39m ==> ENTITY: \u001b[32mVancouver Canucks\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.047]; Hockey[0.047]; Vancouver[0.046]; game[0.046]; game[0.046]; Canucks[0.046]; League[0.045]; Buffalo[0.044]; Buffalo[0.044]; period[0.043]; overtime[0.043]; left[0.043]; left[0.043]; Sabres[0.042]; Game[0.042]; Ice[0.042]; Wednesday[0.042]; Wednesday[0.042]; return[0.042]; received[0.041]; right[0.041]; concussion[0.041]; Friday[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bure\u001b[39m ==> ENTITY: \u001b[32mPavel Bure\u001b[39m\t\nSCORES: global= 0.311:0.311[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.050]; Bure[0.049]; Bure[0.049]; Vancouver[0.048]; game[0.047]; game[0.047]; win[0.047]; assists[0.046]; Ottawa[0.046]; goals[0.046]; concussion[0.045]; Buffalo[0.045]; Brian[0.045]; return[0.044]; overtime[0.044]; forearm[0.044]; left[0.044]; suspension[0.043]; Burke[0.043]; Burke[0.043]; miss[0.043]; suffered[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Buffalo ==> ENTITY: \u001b[32mBuffalo Sabres\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.210:0.210[0]; log p(e|m)= -3.540:-3.540[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.047]; Hockey[0.047]; Buffalo[0.046]; game[0.046]; game[0.046]; League[0.045]; National[0.044]; Vancouver[0.044]; Vancouver[0.044]; Sabres[0.044]; Bure[0.042]; Bure[0.042]; Bure[0.042]; Bure[0.042]; Bure[0.042]; Bure[0.042]; York[0.042]; Canucks[0.042]; Canucks[0.042]; win[0.042]; Game[0.042]; Burke[0.042]; Burke[0.042]; \t\n\r [=====>........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 16s872ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 675/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1254testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mKremlin\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPresident of Russia\u001b[39m <---> \u001b[32mMoscow Kremlin\u001b[39m\t\nSCORES: global= 0.262:0.242[\u001b[31m0.019\u001b[39m]; local()= 0.169:0.102[\u001b[31m0.066\u001b[39m]; log p(e|m)= -2.198:-0.761[\u001b[32m1.437\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.048]; Russia[0.048]; Moscow[0.047]; President[0.047]; Interfax[0.045]; country[0.045]; Yeltsin[0.045]; Yeltsin[0.045]; Yeltsin[0.045]; Boris[0.044]; parliament[0.044]; month[0.043]; plans[0.043]; plans[0.043]; residence[0.042]; agency[0.042]; said[0.042]; bypass[0.041]; Yegor[0.041]; news[0.041]; December[0.041]; December[0.041]; met[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yeltsin\u001b[39m ==> ENTITY: \u001b[32mBoris Yeltsin\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.046]; President[0.045]; Gorbachev[0.043]; Russia[0.043]; Russian[0.043]; Moscow[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Kremlin[0.042]; Kremlin[0.042]; Mikhail[0.042]; Interfax[0.041]; Boris[0.041]; parliament[0.040]; month[0.040]; Yegor[0.040]; surgery[0.040]; lbs[0.040]; met[0.039]; country[0.039]; Union[0.039]; news[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yeltsin\u001b[39m ==> ENTITY: \u001b[32mBoris Yeltsin\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.047]; Gorbachev[0.044]; Russia[0.044]; Yeltsin[0.044]; Yeltsin[0.044]; Yeltsin[0.044]; Kremlin[0.043]; Mikhail[0.043]; Renat[0.043]; Interfax[0.042]; Surgeon[0.042]; Tass[0.041]; parliament[0.041]; Yegor[0.041]; lbs[0.040]; met[0.040]; country[0.040]; Union[0.040]; news[0.040]; news[0.040]; November[0.040]; address[0.039]; television[0.039]; television[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.172:0.172[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.046]; Russian[0.045]; country[0.045]; Moscow[0.045]; Kremlin[0.043]; Kremlin[0.043]; Mikhail[0.042]; Interfax[0.042]; fifth[0.042]; Yegor[0.041]; Boris[0.041]; month[0.041]; marking[0.041]; nation[0.041]; Tass[0.041]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Gorbachev[0.040]; parliament[0.040]; President[0.040]; Union[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Interfax\u001b[39m ==> ENTITY: \u001b[32mInterfax\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnews[0.047]; Moscow[0.047]; Russian[0.046]; Russia[0.043]; agency[0.043]; Kremlin[0.043]; Kremlin[0.043]; Mikhail[0.042]; Soviet[0.041]; told[0.041]; told[0.041]; month[0.041]; surgery[0.041]; lbs[0.040]; Friday[0.040]; Friday[0.040]; December[0.040]; December[0.040]; December[0.040]; President[0.040]; Boris[0.040]; Gorbachev[0.040]; Yegor[0.040]; Today[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Itar-Tass\u001b[39m ==> ENTITY: \u001b[32mInformation Telegraph Agency of Russia\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nagency[0.046]; news[0.045]; Tass[0.045]; Russia[0.044]; Renat[0.044]; Surgeon[0.043]; Soviet[0.043]; Union[0.042]; lbs[0.042]; television[0.041]; television[0.041]; nation[0.041]; Kremlin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; day[0.040]; day[0.040]; Mikhail[0.040]; address[0.040]; Gorbachev[0.040]; kg[0.040]; December[0.040]; December[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yeltsin\u001b[39m ==> ENTITY: \u001b[32mBoris Yeltsin\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.046]; President[0.045]; Gorbachev[0.043]; Russia[0.043]; Russian[0.043]; Moscow[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Kremlin[0.042]; Mikhail[0.042]; Renat[0.041]; Interfax[0.041]; Surgeon[0.041]; Boris[0.040]; Tass[0.040]; parliament[0.040]; month[0.040]; Yegor[0.040]; surgery[0.040]; lbs[0.039]; met[0.039]; country[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yeltsin\u001b[39m ==> ENTITY: \u001b[32mBoris Yeltsin\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.047]; Russian[0.045]; Moscow[0.045]; Yeltsin[0.045]; Yeltsin[0.045]; Kremlin[0.044]; Interfax[0.043]; Boris[0.042]; parliament[0.042]; month[0.042]; Yegor[0.041]; surgery[0.041]; lbs[0.041]; met[0.041]; country[0.041]; news[0.040]; December[0.040]; December[0.040]; speaker[0.040]; kg[0.039]; cheeks[0.039]; said[0.039]; plans[0.039]; plans[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.142:0.142[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.048]; country[0.046]; Moscow[0.046]; Yegor[0.044]; Boris[0.043]; Kremlin[0.043]; Kremlin[0.043]; Interfax[0.043]; month[0.042]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; parliament[0.041]; President[0.041]; work[0.040]; lost[0.040]; return[0.040]; return[0.040]; December[0.040]; December[0.040]; ago[0.039]; working[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yeltsin\u001b[39m ==> ENTITY: \u001b[32mBoris Yeltsin\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.047]; Gorbachev[0.045]; Russia[0.045]; Yeltsin[0.045]; Yeltsin[0.045]; Kremlin[0.044]; Mikhail[0.043]; Renat[0.043]; Surgeon[0.043]; Tass[0.042]; lbs[0.041]; Union[0.040]; news[0.040]; November[0.040]; address[0.040]; television[0.040]; television[0.040]; resigned[0.040]; took[0.040]; took[0.040]; anniversary[0.040]; led[0.040]; December[0.039]; December[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mKremlin\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPresident of Russia\u001b[39m <---> \u001b[32mMoscow Kremlin\u001b[39m\t\nSCORES: global= 0.254:0.236[\u001b[31m0.018\u001b[39m]; local()= 0.178:0.116[\u001b[31m0.063\u001b[39m]; log p(e|m)= -2.198:-0.761[\u001b[32m1.437\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.046]; Russian[0.045]; Russia[0.045]; President[0.044]; Interfax[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Mikhail[0.042]; Gorbachev[0.042]; country[0.041]; Boris[0.041]; parliament[0.041]; nation[0.041]; Tass[0.040]; month[0.040]; major[0.040]; residence[0.040]; residence[0.040]; agency[0.040]; agency[0.040]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yegor Stroyev\u001b[39m ==> ENTITY: \u001b[32mYegor Stroyev\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.049]; Soviet[0.049]; Russian[0.049]; Boris[0.048]; Mikhail[0.048]; Kremlin[0.048]; Kremlin[0.048]; Moscow[0.047]; Union[0.046]; speaker[0.045]; Yeltsin[0.044]; Yeltsin[0.044]; Yeltsin[0.044]; Yeltsin[0.044]; Yeltsin[0.044]; November[0.044]; Gorbachev[0.044]; surgery[0.043]; parliament[0.043]; lbs[0.043]; Interfax[0.043]; President[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.142:0.142[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.046]; Russia[0.045]; fifth[0.044]; Boris[0.043]; Kremlin[0.043]; Yegor[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; country[0.042]; Friday[0.041]; Friday[0.041]; month[0.041]; Interfax[0.041]; December[0.041]; December[0.041]; lost[0.041]; bypass[0.040]; day[0.040]; agency[0.040]; plans[0.040]; plans[0.040]; ago[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mikhail Gorbachev\u001b[39m ==> ENTITY: \u001b[32mMikhail Gorbachev\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.049]; country[0.046]; Russia[0.046]; Yeltsin[0.045]; Yeltsin[0.045]; Yeltsin[0.045]; Yeltsin[0.045]; Kremlin[0.045]; Renat[0.044]; Interfax[0.044]; Surgeon[0.044]; nation[0.043]; Union[0.043]; Tass[0.043]; parliament[0.042]; lbs[0.042]; Yegor[0.042]; marking[0.042]; said[0.041]; news[0.041]; news[0.041]; November[0.041]; met[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Boris Yeltsin\u001b[39m ==> ENTITY: \u001b[32mBoris Yeltsin\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.046]; Russia[0.044]; Russian[0.044]; Moscow[0.044]; Yeltsin[0.044]; Yeltsin[0.044]; Yeltsin[0.044]; Kremlin[0.043]; Kremlin[0.043]; Interfax[0.042]; parliament[0.041]; month[0.041]; Yegor[0.041]; surgery[0.041]; lbs[0.041]; met[0.040]; country[0.040]; news[0.040]; took[0.039]; took[0.039]; anniversary[0.039]; December[0.039]; December[0.039]; speaker[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Soviet Union ==> ENTITY: \u001b[32mSoviet Union\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.076:-0.076[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.048]; Mikhail[0.045]; country[0.043]; fifth[0.043]; Kremlin[0.043]; Yegor[0.043]; day[0.042]; day[0.042]; lost[0.042]; nation[0.042]; took[0.041]; took[0.041]; Gorbachev[0.041]; times[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; led[0.041]; Renat[0.040]; Interfax[0.040]; operation[0.040]; operation[0.040]; major[0.040]; \t\n\r [======>.......................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 16s817ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 691/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1163testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina national football team\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.177:0.177[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.049]; China[0.049]; Japan[0.044]; Japan[0.044]; Soccer[0.043]; Uzbekistan[0.043]; Asian[0.043]; Cup[0.042]; win[0.042]; match[0.042]; match[0.042]; match[0.042]; championship[0.041]; second[0.041]; group[0.040]; Surprise[0.040]; began[0.040]; Emirates[0.040]; defeat[0.040]; advantage[0.039]; surprise[0.039]; missed[0.039]; Arab[0.039]; title[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mJapan\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mJapan national football team\u001b[39m <---> \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.252:0.245[\u001b[31m0.007\u001b[39m]; local()= 0.196:0.119[\u001b[31m0.077\u001b[39m]; log p(e|m)= -3.730:-0.136[\u001b[32m3.594\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.049]; Japan[0.049]; Korea[0.045]; matches[0.044]; game[0.043]; game[0.043]; Cup[0.042]; goal[0.042]; goal[0.042]; played[0.041]; play[0.041]; scored[0.041]; teams[0.041]; coach[0.041]; Fifa[0.040]; seventh[0.040]; World[0.040]; penalty[0.040]; Kuwait[0.040]; regain[0.040]; minute[0.039]; minute[0.039]; early[0.039]; lead[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syrian ==> ENTITY: \u001b[32mSyria national football team\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.178:0.178[0]; log p(e|m)= -3.170:-3.170[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.052]; Syria[0.052]; Syria[0.052]; Syrian[0.049]; Syrian[0.049]; Cup[0.046]; Hassan[0.044]; Uzbekistan[0.043]; area[0.042]; area[0.042]; area[0.042]; finals[0.042]; finals[0.042]; Japan[0.041]; Japan[0.041]; Abbas[0.041]; Asian[0.041]; Asian[0.041]; Games[0.040]; outside[0.040]; rose[0.040]; second[0.039]; playing[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Cup ==> ENTITY: \u001b[32mFIFA World Cup\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.152:0.152[0]; log p(e|m)= -1.002:-1.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.047]; teams[0.046]; Fifa[0.045]; played[0.043]; Korea[0.042]; play[0.042]; coach[0.042]; South[0.042]; game[0.042]; game[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; world[0.041]; hosts[0.041]; penalty[0.040]; scored[0.040]; title[0.040]; seventh[0.040]; ranked[0.040]; goal[0.040]; goal[0.040]; level[0.040]; early[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Soviet ==> ENTITY: \u001b[32mSoviet Union\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.067:0.067[0]; log p(e|m)= -0.098:-0.098[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplaying[0.044]; Games[0.043]; championship[0.043]; winning[0.043]; goalkeeper[0.043]; Cup[0.042]; group[0.042]; Uzbekistan[0.042]; Uzbekistan[0.042]; finals[0.042]; finals[0.042]; ago[0.041]; Oleg[0.041]; title[0.041]; republic[0.041]; goals[0.041]; goal[0.041]; win[0.041]; errors[0.041]; Igor[0.041]; Group[0.041]; missed[0.040]; second[0.040]; second[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Uzbekistan ==> ENTITY: \u001b[32mUzbekistan national football team\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.195:0.195[0]; log p(e|m)= -2.226:-2.226[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUzbekistan[0.050]; Games[0.048]; Asian[0.048]; Asian[0.048]; Asian[0.048]; China[0.044]; China[0.044]; China[0.044]; match[0.043]; match[0.043]; match[0.043]; second[0.043]; Uzbek[0.043]; Chinese[0.043]; Japan[0.042]; Japan[0.042]; finals[0.041]; finals[0.041]; Soccer[0.041]; playing[0.040]; Soviet[0.040]; championship[0.040]; Oleg[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan national football team\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.200:0.200[0]; log p(e|m)= -3.730:-3.730[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.048]; Japan[0.048]; Korea[0.045]; second[0.044]; matches[0.044]; game[0.043]; game[0.043]; Cup[0.042]; goal[0.042]; goal[0.042]; played[0.041]; play[0.041]; scored[0.041]; teams[0.041]; coach[0.040]; Fifa[0.040]; seventh[0.040]; World[0.040]; penalty[0.040]; Kuwait[0.040]; regain[0.039]; minute[0.039]; minute[0.039]; early[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syria ==> ENTITY: \u001b[32mSyria national football team\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.162:0.162[0]; log p(e|m)= -3.058:-3.058[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.051]; Syria[0.051]; Games[0.047]; Syrian[0.046]; Asian[0.046]; Asian[0.046]; Cup[0.044]; Uzbekistan[0.042]; finals[0.041]; finals[0.041]; Chinese[0.041]; republic[0.041]; Japan[0.040]; Japan[0.040]; Hassan[0.040]; Abbas[0.039]; second[0.038]; title[0.038]; meeting[0.038]; Soviet[0.038]; Salem[0.038]; outside[0.038]; area[0.037]; area[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan national football team\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.201:0.201[0]; log p(e|m)= -3.730:-3.730[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.050]; Japan[0.050]; Korea[0.047]; second[0.046]; matches[0.045]; game[0.045]; Cup[0.044]; goal[0.044]; goal[0.044]; goal[0.044]; played[0.043]; play[0.042]; scored[0.042]; scored[0.042]; coach[0.042]; Fifa[0.042]; seventh[0.042]; World[0.041]; penalty[0.041]; Kuwait[0.041]; regain[0.041]; opening[0.041]; early[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Uzbek ==> ENTITY: \u001b[32mUzbekistan national football team\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.210:0.210[0]; log p(e|m)= -2.180:-2.180[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUzbekistan[0.047]; Uzbekistan[0.047]; match[0.044]; match[0.044]; match[0.044]; Arab[0.042]; finals[0.042]; finals[0.042]; Soviet[0.042]; Soccer[0.042]; Asian[0.041]; Asian[0.041]; Asian[0.041]; second[0.041]; playing[0.041]; republic[0.041]; China[0.040]; China[0.040]; China[0.040]; win[0.040]; win[0.040]; Cup[0.040]; Cup[0.040]; goals[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = FIFA\u001b[39m ==> ENTITY: \u001b[32mFIFA\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.046]; game[0.046]; game[0.046]; Cup[0.045]; World[0.042]; played[0.042]; teams[0.042]; Korea[0.042]; play[0.042]; ranked[0.042]; title[0.041]; Japan[0.041]; Japan[0.041]; Japan[0.041]; minute[0.041]; level[0.041]; coach[0.040]; world[0.040]; South[0.040]; hosts[0.040]; penalty[0.040]; Indonesia[0.040]; goal[0.040]; goal[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syria ==> ENTITY: \u001b[32mSyria national football team\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.180:0.180[0]; log p(e|m)= -3.058:-3.058[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAsian[0.047]; Arab[0.047]; Cup[0.044]; China[0.044]; China[0.044]; China[0.044]; championship[0.043]; Uzbekistan[0.043]; Soccer[0.042]; Chinese[0.042]; Al[0.042]; Japan[0.041]; Japan[0.041]; Emirates[0.040]; Nadim[0.040]; match[0.040]; match[0.040]; match[0.040]; title[0.039]; win[0.039]; win[0.039]; began[0.039]; Group[0.039]; United[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syria ==> ENTITY: \u001b[32mSyria national football team\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.183:0.183[0]; log p(e|m)= -3.058:-3.058[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.053]; Syria[0.053]; Syrian[0.050]; Syrian[0.050]; Syrian[0.050]; Syrian[0.050]; Syrians[0.048]; Cup[0.047]; coach[0.044]; finals[0.044]; finals[0.044]; Uzbekistan[0.044]; Hassan[0.044]; seventh[0.043]; Asian[0.043]; Asian[0.043]; Abbas[0.042]; Games[0.042]; second[0.042]; defensive[0.041]; game[0.041]; title[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mUAE\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUnited Arab Emirates\u001b[39m <---> \u001b[32mUnited Arab Emirates national football team\u001b[39m\t\nSCORES: global= 0.254:0.248[\u001b[31m0.005\u001b[39m]; local()= 0.148:0.176[\u001b[32m0.028\u001b[39m]; log p(e|m)= -0.195:-2.283[\u001b[31m2.088\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKuwait[0.053]; matches[0.051]; play[0.049]; teams[0.047]; Cup[0.047]; Fifa[0.047]; played[0.046]; game[0.046]; game[0.046]; Syria[0.045]; Korea[0.045]; seventh[0.044]; World[0.044]; Indonesia[0.044]; hosts[0.043]; Group[0.043]; level[0.043]; balls[0.043]; Hosts[0.043]; Saturday[0.043]; given[0.043]; hard[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Al Ain\u001b[39m ==> ENTITY: \u001b[32mAl Ain\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEmirates[0.046]; desert[0.044]; Arab[0.044]; Soccer[0.044]; Cup[0.043]; Syria[0.042]; win[0.042]; second[0.042]; striker[0.042]; Ain[0.042]; match[0.042]; match[0.042]; match[0.042]; took[0.041]; Uzbekistan[0.041]; United[0.041]; championship[0.041]; advantage[0.040]; Nadim[0.040]; Asian[0.040]; missed[0.040]; defensive[0.040]; defeat[0.040]; net[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shu Kamo\u001b[39m ==> ENTITY: \u001b[32mShu Kamo\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.111:0.111[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncoach[0.044]; matches[0.044]; played[0.043]; Japan[0.043]; Japan[0.043]; Japan[0.043]; game[0.043]; game[0.043]; teams[0.043]; lucky[0.042]; level[0.042]; ball[0.041]; ranked[0.041]; Cup[0.041]; Korea[0.041]; header[0.041]; defensively[0.041]; minute[0.041]; minute[0.041]; balls[0.040]; Saturday[0.040]; minutes[0.040]; play[0.040]; Fifa[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Takuya Takagi\u001b[39m ==> ENTITY: \u001b[32mTakuya Takagi\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.045]; Japan[0.045]; second[0.043]; playing[0.043]; seventh[0.042]; finals[0.042]; finals[0.042]; long[0.042]; goals[0.042]; Asian[0.041]; Asian[0.041]; blunder[0.041]; opening[0.041]; goal[0.041]; goal[0.041]; injury[0.041]; Uzbekistan[0.041]; Games[0.041]; Cup[0.041]; scored[0.041]; ago[0.041]; defensive[0.041]; defensive[0.041]; win[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chinese ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.151:0.151[0]; log p(e|m)= -1.419:-1.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.050]; China[0.050]; Asian[0.048]; Asian[0.048]; Asian[0.048]; Japan[0.042]; Japan[0.042]; playing[0.042]; Games[0.041]; newcomers[0.041]; Uzbekistan[0.040]; Uzbekistan[0.040]; group[0.040]; area[0.040]; time[0.040]; time[0.040]; republic[0.039]; Uzbek[0.039]; second[0.039]; come[0.039]; Syria[0.039]; Syria[0.039]; outside[0.038]; began[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina national football team\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.194:0.194[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.047]; China[0.047]; Chinese[0.044]; win[0.044]; win[0.044]; Japan[0.043]; Japan[0.043]; Soccer[0.042]; match[0.041]; match[0.041]; match[0.041]; finals[0.041]; Uzbekistan[0.041]; championship[0.041]; Asian[0.041]; Asian[0.041]; Cup[0.041]; Cup[0.041]; playing[0.041]; time[0.040]; second[0.040]; advantage[0.039]; title[0.039]; Surprise[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syrian ==> ENTITY: \u001b[32mSyria national football team\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.176:0.176[0]; log p(e|m)= -3.170:-3.170[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.049]; Syria[0.049]; Syrian[0.047]; Syrian[0.047]; Syrian[0.047]; Syrians[0.044]; Kuwait[0.044]; Cup[0.043]; matches[0.042]; Hassan[0.041]; area[0.040]; area[0.040]; South[0.040]; teams[0.040]; Japan[0.039]; Japan[0.039]; Japan[0.039]; Fifa[0.039]; world[0.039]; Abbas[0.039]; World[0.039]; coach[0.039]; Korea[0.038]; siege[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Uzbekistan ==> ENTITY: \u001b[32mUzbekistan national football team\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.189:0.189[0]; log p(e|m)= -2.226:-2.226[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUzbekistan[0.050]; Games[0.047]; Asian[0.047]; Asian[0.047]; China[0.044]; match[0.043]; Uzbek[0.043]; Chinese[0.042]; time[0.042]; time[0.042]; playing[0.042]; Japan[0.041]; finals[0.041]; finals[0.041]; second[0.040]; Soviet[0.040]; Cup[0.040]; scored[0.039]; win[0.039]; goals[0.039]; missed[0.039]; injury[0.038]; Syria[0.038]; Syria[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kuwait ==> ENTITY: \u001b[32mKuwait national football team\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.173:0.173[0]; log p(e|m)= -2.364:-2.364[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.045]; matches[0.044]; Korea[0.043]; teams[0.043]; Fifa[0.043]; played[0.042]; Cup[0.042]; Syrian[0.042]; Syrian[0.042]; Syrian[0.042]; seventh[0.042]; Japan[0.041]; Japan[0.041]; Japan[0.041]; Syrians[0.041]; play[0.041]; game[0.041]; game[0.041]; title[0.041]; World[0.040]; siege[0.040]; hosts[0.040]; ranked[0.040]; penalty[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Korea ==> ENTITY: \u001b[32mSouth Korea national football team\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.184:0.184[0]; log p(e|m)= -2.847:-2.847[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseventh[0.046]; matches[0.045]; game[0.045]; game[0.045]; Japan[0.044]; Japan[0.044]; Japan[0.044]; teams[0.043]; play[0.042]; Cup[0.042]; played[0.042]; Kuwait[0.041]; scored[0.041]; Fifa[0.041]; title[0.040]; minute[0.040]; early[0.040]; coach[0.040]; goal[0.040]; level[0.039]; Group[0.039]; ranked[0.039]; Indonesia[0.039]; World[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hassan Abbas\u001b[39m ==> ENTITY: \u001b[32mHassan Abbas\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyrian[0.045]; Syrian[0.045]; Syrian[0.045]; Syrian[0.045]; Syria[0.044]; Syria[0.044]; Syria[0.044]; played[0.043]; finals[0.042]; finals[0.042]; Syrians[0.042]; title[0.041]; game[0.041]; coach[0.041]; Games[0.040]; goals[0.040]; Salem[0.040]; Uzbekistan[0.040]; Kamo[0.040]; Nader[0.039]; penalty[0.039]; goal[0.039]; goal[0.039]; goal[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Asian Cup ==> ENTITY: \u001b[32mAFC Asian Cup\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.202:0.202[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAsian[0.044]; China[0.044]; China[0.044]; finals[0.043]; finals[0.043]; Uzbekistan[0.043]; Uzbekistan[0.043]; Japan[0.043]; Friday[0.042]; match[0.042]; match[0.042]; match[0.042]; second[0.042]; second[0.042]; Chinese[0.041]; win[0.041]; winning[0.041]; Games[0.040]; injury[0.040]; playing[0.040]; opening[0.040]; missed[0.039]; goal[0.039]; winner[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina national football team\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.214:0.214[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.046]; China[0.046]; Chinese[0.044]; win[0.043]; win[0.043]; Games[0.043]; Japan[0.042]; Japan[0.042]; Soccer[0.041]; winning[0.041]; match[0.041]; match[0.041]; match[0.041]; finals[0.041]; finals[0.041]; Uzbekistan[0.041]; Uzbekistan[0.041]; championship[0.041]; Asian[0.040]; Asian[0.040]; Asian[0.040]; Cup[0.040]; Cup[0.040]; playing[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syrian ==> ENTITY: \u001b[32mSyria national football team\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.179:0.179[0]; log p(e|m)= -3.170:-3.170[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.049]; Syria[0.049]; Syrian[0.047]; Syrian[0.047]; Syrians[0.045]; Kuwait[0.044]; Cup[0.043]; matches[0.042]; Hassan[0.041]; early[0.041]; area[0.040]; area[0.040]; South[0.040]; teams[0.040]; Japan[0.040]; Japan[0.040]; Japan[0.040]; Fifa[0.039]; world[0.039]; Abbas[0.039]; World[0.039]; Korea[0.039]; coach[0.039]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan national football team\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.186:0.186[0]; log p(e|m)= -3.730:-3.730[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.049]; second[0.044]; Soccer[0.044]; match[0.044]; match[0.044]; match[0.044]; Cup[0.043]; championship[0.042]; China[0.042]; China[0.042]; China[0.042]; win[0.042]; Uzbekistan[0.042]; Asian[0.041]; title[0.041]; minute[0.040]; advantage[0.040]; began[0.039]; group[0.039]; missed[0.039]; Group[0.039]; Emirates[0.039]; defeat[0.039]; United[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United Arab Emirates ==> ENTITY: \u001b[32mUnited Arab Emirates\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.126:0.126[0]; log p(e|m)= -0.250:-0.250[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAsian[0.046]; match[0.044]; match[0.044]; match[0.044]; Uzbekistan[0.043]; Cup[0.043]; championship[0.043]; China[0.043]; China[0.043]; China[0.043]; Syria[0.042]; Chinese[0.042]; Soccer[0.041]; win[0.041]; win[0.041]; Japan[0.041]; Japan[0.041]; Uzbek[0.040]; second[0.040]; Al[0.040]; took[0.040]; title[0.039]; defeat[0.039]; missed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hiroshige Yanagimoto\u001b[39m ==> ENTITY: \u001b[32mHiroshige Yanagimoto\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplaying[0.052]; Takuya[0.046]; title[0.045]; Japan[0.043]; Japan[0.043]; Cup[0.043]; winning[0.043]; game[0.043]; Takagi[0.042]; finals[0.042]; finals[0.042]; Games[0.041]; goals[0.041]; defensive[0.040]; injury[0.040]; ball[0.040]; win[0.040]; opening[0.040]; scored[0.040]; hitting[0.040]; header[0.039]; head[0.039]; points[0.039]; appeared[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syrian ==> ENTITY: \u001b[32mSyria national football team\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.181:0.181[0]; log p(e|m)= -3.170:-3.170[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.049]; Syria[0.049]; Syrian[0.047]; Syrian[0.047]; Syrian[0.047]; Syrians[0.044]; Kuwait[0.044]; Cup[0.043]; matches[0.041]; Hassan[0.041]; early[0.040]; area[0.040]; area[0.040]; South[0.040]; teams[0.040]; Japan[0.039]; Japan[0.039]; Japan[0.039]; Fifa[0.039]; world[0.039]; Abbas[0.038]; World[0.038]; Korea[0.038]; coach[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan national football team\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.169:0.169[0]; log p(e|m)= -3.730:-3.730[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.045]; Cup[0.043]; goal[0.043]; goal[0.043]; finals[0.043]; finals[0.043]; Asian[0.043]; Asian[0.043]; Games[0.042]; win[0.042]; Uzbekistan[0.042]; playing[0.042]; scored[0.042]; goals[0.042]; winning[0.041]; Chinese[0.041]; missed[0.040]; opening[0.040]; minute[0.040]; minute[0.040]; minute[0.040]; advantage[0.040]; lead[0.040]; title[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Igor Shkvyrin\u001b[39m ==> ENTITY: \u001b[32mIgor Shkvyrin\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.129:0.129[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.046]; Uzbekistan[0.045]; Uzbekistan[0.045]; playing[0.044]; Cup[0.043]; Cup[0.043]; began[0.042]; Games[0.042]; goals[0.042]; championship[0.041]; finals[0.041]; finals[0.041]; Oleg[0.041]; Soviet[0.041]; defensive[0.040]; defensive[0.040]; ball[0.040]; time[0.040]; time[0.040]; Uzbek[0.040]; match[0.040]; match[0.040]; match[0.040]; missed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia national football team\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.193:0.193[0]; log p(e|m)= -2.489:-2.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.047]; Cup[0.045]; Korea[0.045]; teams[0.044]; played[0.043]; play[0.043]; game[0.043]; game[0.043]; Japan[0.042]; Japan[0.042]; Japan[0.042]; world[0.041]; South[0.041]; Fifa[0.041]; title[0.041]; Kuwait[0.040]; World[0.040]; seventh[0.040]; hosts[0.040]; ranked[0.040]; goal[0.040]; minute[0.039]; level[0.039]; coach[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syria ==> ENTITY: \u001b[32mSyria national football team\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.171:0.171[0]; log p(e|m)= -3.058:-3.058[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.049]; Syria[0.049]; Syrian[0.046]; Syrian[0.046]; Syrian[0.046]; Syrian[0.046]; Syrians[0.045]; Kuwait[0.044]; Cup[0.043]; Fifa[0.041]; coach[0.041]; Hassan[0.040]; world[0.040]; World[0.040]; seventh[0.039]; Abbas[0.039]; second[0.039]; game[0.038]; title[0.038]; Japan[0.038]; Japan[0.038]; Japan[0.038]; Japan[0.038]; Salem[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan national football team\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.179:0.179[0]; log p(e|m)= -3.730:-3.730[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.049]; second[0.044]; Soccer[0.044]; match[0.044]; match[0.044]; match[0.044]; Cup[0.043]; championship[0.042]; China[0.042]; China[0.042]; China[0.042]; win[0.042]; win[0.042]; Uzbekistan[0.041]; Asian[0.041]; title[0.040]; Chinese[0.040]; minute[0.040]; advantage[0.040]; advancing[0.039]; time[0.039]; began[0.039]; group[0.039]; missed[0.039]; \t\n\r [======>.......................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 16s466ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 729/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1189testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Golden State ==> ENTITY: \u001b[32mGolden State Warriors\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.178:0.178[0]; log p(e|m)= -0.742:-0.742[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSan[0.048]; San[0.048]; Golden[0.043]; Portland[0.043]; Portland[0.043]; Lakers[0.042]; Lakers[0.042]; Cleveland[0.042]; Clippers[0.042]; Division[0.042]; Division[0.042]; Houston[0.041]; State[0.041]; Orlando[0.041]; Miami[0.041]; Charlotte[0.040]; Boston[0.040]; Phoenix[0.040]; Phoenix[0.040]; Pacific[0.040]; Sacramento[0.040]; Sacramento[0.040]; Western[0.040]; Dallas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sacramento ==> ENTITY: \u001b[32mSacramento Kings\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.187:0.187[0]; log p(e|m)= -3.079:-3.079[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDivision[0.044]; Division[0.044]; Cleveland[0.044]; Western[0.043]; Conference[0.043]; Miami[0.042]; Houston[0.042]; Sacramento[0.042]; Pacific[0.042]; Charlotte[0.042]; Clippers[0.042]; York[0.041]; Boston[0.041]; Toronto[0.041]; Orlando[0.041]; Phoenix[0.041]; Phoenix[0.041]; San[0.041]; San[0.041]; Friday[0.041]; Dallas[0.040]; Denver[0.040]; Portland[0.040]; Portland[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cleveland ==> ENTITY: \u001b[32mCleveland Cavaliers\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.212:0.212[0]; log p(e|m)= -3.730:-3.730[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.045]; Basketball[0.045]; games[0.045]; games[0.045]; Miami[0.043]; Philadelphia[0.043]; Washington[0.043]; Houston[0.042]; Boston[0.042]; teams[0.041]; Dallas[0.041]; Thursday[0.041]; Thursday[0.041]; Toronto[0.041]; Charlotte[0.040]; Atlanta[0.040]; Milwaukee[0.040]; York[0.040]; York[0.040]; played[0.040]; Indiana[0.040]; Orlando[0.040]; Detroit[0.040]; Lakers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Utah ==> ENTITY: \u001b[32mUtah Jazz\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.200:0.200[0]; log p(e|m)= -3.576:-3.576[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhoenix[0.045]; Phoenix[0.045]; Charlotte[0.045]; Cleveland[0.045]; Vancouver[0.045]; Vancouver[0.045]; Clippers[0.044]; Lakers[0.044]; Lakers[0.044]; Dallas[0.044]; Division[0.044]; Portland[0.043]; Portland[0.043]; Miami[0.043]; Friday[0.043]; Orlando[0.043]; Denver[0.043]; Boston[0.042]; Sacramento[0.042]; Sacramento[0.042]; State[0.042]; State[0.042]; Detroit[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Charlotte ==> ENTITY: \u001b[32mNew Orleans Pelicans\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.215:0.215[0]; log p(e|m)= -4.828:-4.828[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMiami[0.046]; Cleveland[0.046]; Orlando[0.045]; Phoenix[0.045]; Phoenix[0.045]; Dallas[0.045]; Division[0.044]; Boston[0.044]; Detroit[0.044]; Clippers[0.044]; Friday[0.043]; Lakers[0.043]; Lakers[0.043]; December[0.043]; Denver[0.043]; Sacramento[0.042]; Sacramento[0.042]; York[0.042]; Indiana[0.042]; Seattle[0.042]; Portland[0.042]; Portland[0.042]; Jersey[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York Knicks\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.213:0.213[0]; log p(e|m)= -4.828:-4.828[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.047]; Basketball[0.047]; games[0.044]; games[0.044]; Philadelphia[0.043]; Thursday[0.043]; Thursday[0.043]; Cleveland[0.042]; Boston[0.042]; teams[0.041]; Miami[0.041]; Conference[0.041]; Conference[0.041]; played[0.041]; Houston[0.041]; Washington[0.040]; Toronto[0.040]; York[0.040]; Charlotte[0.040]; Orlando[0.040]; Detroit[0.039]; National[0.039]; Jersey[0.039]; Central[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Toronto ==> ENTITY: \u001b[32mToronto Raptors\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.195:0.195[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; Vancouver[0.044]; played[0.044]; Cleveland[0.043]; Philadelphia[0.043]; Boston[0.042]; Detroit[0.042]; Houston[0.042]; won[0.042]; Atlanta[0.042]; Orlando[0.041]; Charlotte[0.041]; York[0.041]; Clippers[0.041]; Miami[0.041]; Thursday[0.041]; Conference[0.041]; Conference[0.041]; Phoenix[0.041]; Western[0.040]; Indiana[0.040]; Dallas[0.040]; Washington[0.040]; Chicago[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vancouver ==> ENTITY: \u001b[32mMemphis Grizzlies\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.218:0.218[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVancouver[0.048]; Toronto[0.045]; Cleveland[0.043]; Cleveland[0.043]; Charlotte[0.042]; Charlotte[0.042]; Houston[0.042]; Miami[0.042]; Clippers[0.042]; Phoenix[0.041]; Phoenix[0.041]; Orlando[0.041]; Detroit[0.041]; Detroit[0.041]; Boston[0.041]; Portland[0.041]; Portland[0.041]; Conference[0.041]; Atlanta[0.041]; Chicago[0.040]; Lakers[0.040]; Lakers[0.040]; Dallas[0.040]; Denver[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Jersey ==> ENTITY: \u001b[32mBrooklyn Nets\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.224:0.224[0]; log p(e|m)= -3.912:-3.912[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; games[0.045]; Basketball[0.044]; Basketball[0.044]; Philadelphia[0.042]; played[0.042]; Conference[0.042]; Conference[0.042]; teams[0.042]; York[0.042]; York[0.042]; Thursday[0.041]; Thursday[0.041]; Toronto[0.041]; Boston[0.041]; Cleveland[0.041]; Houston[0.041]; won[0.041]; Atlantic[0.041]; Vancouver[0.040]; Miami[0.040]; Charlotte[0.040]; Central[0.040]; Detroit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Phoenix ==> ENTITY: \u001b[32mPhoenix Suns\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.210:0.210[0]; log p(e|m)= -3.474:-3.474[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.045]; Miami[0.044]; Phoenix[0.043]; Toronto[0.043]; Conference[0.043]; Houston[0.043]; Vancouver[0.043]; Vancouver[0.043]; Orlando[0.042]; Boston[0.042]; Dallas[0.042]; Detroit[0.041]; Lakers[0.041]; Lakers[0.041]; Clippers[0.041]; Charlotte[0.041]; Denver[0.041]; Friday[0.040]; Utah[0.040]; Utah[0.040]; Western[0.040]; Pacific[0.040]; Division[0.040]; Division[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Jersey ==> ENTITY: \u001b[32mBrooklyn Nets\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.196:0.196[0]; log p(e|m)= -3.912:-3.912[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConference[0.044]; York[0.043]; Clippers[0.043]; Boston[0.043]; Cleveland[0.043]; Houston[0.043]; Vancouver[0.042]; Vancouver[0.042]; Miami[0.042]; Charlotte[0.042]; Lakers[0.042]; Lakers[0.042]; Detroit[0.042]; Orlando[0.041]; Friday[0.041]; Division[0.041]; Division[0.041]; Dallas[0.041]; Indiana[0.040]; Phoenix[0.040]; Phoenix[0.040]; Denver[0.040]; Minnesota[0.040]; Minnesota[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portland ==> ENTITY: \u001b[32mPortland Trail Blazers\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.206:0.206[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.047]; Boston[0.046]; Miami[0.045]; Portland[0.045]; Clippers[0.044]; Orlando[0.044]; Pacific[0.044]; Seattle[0.044]; Dallas[0.044]; Phoenix[0.044]; Phoenix[0.044]; Division[0.043]; Charlotte[0.043]; Denver[0.043]; Detroit[0.043]; Lakers[0.043]; Lakers[0.043]; Sacramento[0.042]; Sacramento[0.042]; Indiana[0.042]; Minnesota[0.042]; Friday[0.042]; December[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indiana ==> ENTITY: \u001b[32mIndiana Pacers\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.199:0.199[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.045]; Charlotte[0.044]; Miami[0.044]; Boston[0.043]; Orlando[0.042]; Clippers[0.042]; Jersey[0.042]; Dallas[0.042]; Division[0.042]; Detroit[0.042]; Phoenix[0.041]; Phoenix[0.041]; Denver[0.041]; State[0.041]; State[0.041]; Lakers[0.041]; Lakers[0.041]; Friday[0.041]; Sacramento[0.041]; Sacramento[0.041]; Minnesota[0.040]; York[0.040]; Portland[0.040]; Portland[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Houston ==> ENTITY: \u001b[32mHouston Rockets\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.213:0.213[0]; log p(e|m)= -3.817:-3.817[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConference[0.046]; Dallas[0.046]; Cleveland[0.045]; Cleveland[0.045]; Philadelphia[0.045]; Orlando[0.044]; Miami[0.044]; Miami[0.044]; Clippers[0.044]; Boston[0.043]; Boston[0.043]; Phoenix[0.043]; Phoenix[0.043]; Lakers[0.043]; Washington[0.043]; Atlanta[0.043]; Denver[0.042]; Toronto[0.042]; Friday[0.042]; Charlotte[0.042]; December[0.042]; Western[0.042]; Golden[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington Wizards\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.228:0.228[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; games[0.045]; Basketball[0.044]; Basketball[0.044]; Philadelphia[0.044]; Miami[0.043]; Cleveland[0.042]; teams[0.042]; Toronto[0.042]; Houston[0.042]; Orlando[0.042]; Boston[0.042]; Atlanta[0.041]; played[0.041]; Dallas[0.041]; Charlotte[0.041]; won[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Chicago[0.040]; Conference[0.040]; Conference[0.040]; Detroit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pacific ==> ENTITY: \u001b[32mPacific Division (NBA)\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.201:0.201[0]; log p(e|m)= -3.194:-3.194[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLakers[0.046]; Lakers[0.046]; Vancouver[0.045]; Vancouver[0.045]; Division[0.045]; Division[0.045]; Division[0.045]; Western[0.044]; Phoenix[0.044]; Phoenix[0.044]; Dallas[0.043]; Clippers[0.043]; Conference[0.043]; Friday[0.043]; Toronto[0.043]; Houston[0.042]; Detroit[0.042]; Detroit[0.042]; Denver[0.042]; Boston[0.042]; Cleveland[0.042]; Cleveland[0.042]; Chicago[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Boston ==> ENTITY: \u001b[32mBoston Celtics\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.183:0.183[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYork[0.047]; Cleveland[0.045]; Vancouver[0.043]; Vancouver[0.043]; Detroit[0.043]; Houston[0.042]; Miami[0.042]; Conference[0.042]; Lakers[0.042]; Lakers[0.042]; Charlotte[0.042]; Clippers[0.041]; Orlando[0.041]; Western[0.041]; Dallas[0.041]; Phoenix[0.040]; Phoenix[0.040]; Minnesota[0.040]; Minnesota[0.040]; Portland[0.040]; Portland[0.040]; Division[0.040]; Division[0.040]; Denver[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cleveland ==> ENTITY: \u001b[32mCleveland Cavaliers\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.188:0.188[0]; log p(e|m)= -3.730:-3.730[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMiami[0.045]; Houston[0.044]; Boston[0.043]; Clippers[0.043]; Dallas[0.042]; Charlotte[0.042]; York[0.042]; Indiana[0.042]; Orlando[0.042]; Detroit[0.042]; Lakers[0.042]; Lakers[0.042]; Phoenix[0.042]; Phoenix[0.042]; Minnesota[0.042]; Minnesota[0.042]; Conference[0.041]; Friday[0.041]; Midwest[0.041]; Denver[0.041]; December[0.040]; Utah[0.040]; Utah[0.040]; Seattle[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlanta ==> ENTITY: \u001b[32mAtlanta Hawks\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.230:0.230[0]; log p(e|m)= -3.650:-3.650[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.045]; Basketball[0.045]; games[0.044]; games[0.044]; Philadelphia[0.044]; Cleveland[0.043]; Houston[0.042]; Miami[0.042]; teams[0.042]; Boston[0.042]; Conference[0.041]; Conference[0.041]; Detroit[0.041]; Charlotte[0.041]; Dallas[0.041]; won[0.040]; Orlando[0.040]; Washington[0.040]; played[0.040]; Standings[0.040]; Standings[0.040]; Games[0.040]; Lakers[0.040]; Toronto[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Golden State ==> ENTITY: \u001b[32mGolden State Warriors\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.184:0.184[0]; log p(e|m)= -0.742:-0.742[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSan[0.048]; San[0.048]; Golden[0.044]; Portland[0.043]; Portland[0.043]; Lakers[0.043]; Lakers[0.043]; Cleveland[0.043]; Clippers[0.043]; Division[0.042]; State[0.041]; Orlando[0.041]; Miami[0.041]; Charlotte[0.040]; Boston[0.040]; Phoenix[0.040]; Phoenix[0.040]; Pacific[0.040]; Sacramento[0.040]; Sacramento[0.040]; Dallas[0.040]; Detroit[0.039]; Denver[0.039]; Utah[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Orlando ==> ENTITY: \u001b[32mOrlando Magic\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.185:0.185[0]; log p(e|m)= -3.079:-3.079[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.045]; Miami[0.045]; Boston[0.044]; Detroit[0.043]; Charlotte[0.043]; Division[0.043]; Phoenix[0.043]; Phoenix[0.043]; Clippers[0.042]; Denver[0.042]; Friday[0.042]; Indiana[0.042]; Lakers[0.042]; Lakers[0.042]; Jersey[0.041]; December[0.041]; Utah[0.041]; York[0.040]; Minnesota[0.040]; Sacramento[0.040]; Sacramento[0.040]; Seattle[0.040]; Portland[0.039]; Portland[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Utah ==> ENTITY: \u001b[32mUtah Jazz\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.199:0.199[0]; log p(e|m)= -3.576:-3.576[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConference[0.047]; Phoenix[0.045]; Phoenix[0.045]; Charlotte[0.044]; Cleveland[0.044]; Cleveland[0.044]; Houston[0.044]; Vancouver[0.044]; Vancouver[0.044]; Atlanta[0.044]; Clippers[0.044]; Lakers[0.043]; Dallas[0.043]; Division[0.043]; Division[0.043]; Division[0.043]; Portland[0.043]; Miami[0.043]; Friday[0.043]; Orlando[0.042]; Denver[0.042]; Philadelphia[0.042]; Washington[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Seattle ==> ENTITY: \u001b[32mSeattle SuperSonics\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.232:0.232[0]; log p(e|m)= -3.650:-3.650[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConference[0.047]; Cleveland[0.045]; Cleveland[0.045]; Houston[0.044]; Miami[0.044]; Orlando[0.044]; Western[0.044]; Clippers[0.044]; Boston[0.044]; Portland[0.043]; Portland[0.043]; Dallas[0.043]; Toronto[0.043]; Phoenix[0.043]; Phoenix[0.043]; Division[0.043]; Division[0.043]; Denver[0.043]; Charlotte[0.043]; Charlotte[0.043]; Atlanta[0.042]; Lakers[0.042]; Lakers[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago Bulls\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.202:0.202[0]; log p(e|m)= -4.510:-4.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; games[0.047]; Basketball[0.046]; Basketball[0.046]; Cleveland[0.044]; teams[0.044]; Vancouver[0.044]; Toronto[0.044]; won[0.043]; Philadelphia[0.043]; Conference[0.043]; Conference[0.043]; Boston[0.043]; Houston[0.043]; Miami[0.043]; Division[0.042]; Division[0.042]; Division[0.042]; Division[0.042]; Detroit[0.042]; played[0.042]; Orlando[0.042]; Washington[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Detroit ==> ENTITY: \u001b[32mDetroit Pistons\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.184:0.184[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.045]; Vancouver[0.044]; Vancouver[0.044]; Boston[0.043]; Houston[0.043]; Charlotte[0.043]; Miami[0.042]; Dallas[0.042]; Friday[0.042]; Division[0.042]; Division[0.042]; Phoenix[0.042]; Phoenix[0.042]; Orlando[0.041]; Lakers[0.041]; Lakers[0.041]; Minnesota[0.041]; Minnesota[0.041]; Indiana[0.041]; December[0.041]; Clippers[0.041]; Denver[0.040]; York[0.040]; Jersey[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York Knicks\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.191:0.191[0]; log p(e|m)= -4.828:-4.828[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.047]; Boston[0.046]; Miami[0.046]; Friday[0.045]; Houston[0.045]; Lakers[0.044]; Lakers[0.044]; Charlotte[0.044]; Orlando[0.044]; Clippers[0.044]; Dallas[0.043]; Detroit[0.043]; Jersey[0.043]; Phoenix[0.043]; Phoenix[0.043]; Division[0.043]; Denver[0.042]; Indiana[0.042]; December[0.042]; State[0.042]; State[0.042]; Minnesota[0.042]; Minnesota[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Miami ==> ENTITY: \u001b[32mMiami Heat\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.229:0.229[0]; log p(e|m)= -3.730:-3.730[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; games[0.045]; Basketball[0.045]; Basketball[0.045]; teams[0.044]; Cleveland[0.043]; Conference[0.042]; Conference[0.042]; Orlando[0.042]; Houston[0.042]; Philadelphia[0.042]; Boston[0.041]; Charlotte[0.041]; Thursday[0.041]; Thursday[0.041]; played[0.041]; won[0.040]; Toronto[0.040]; Detroit[0.040]; Atlanta[0.040]; Washington[0.040]; Chicago[0.039]; Indiana[0.039]; National[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = LA LAKERS\u001b[39m ==> ENTITY: \u001b[32mLos Angeles Lakers\u001b[39m\t\nSCORES: global= 0.317:0.317[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLakers[0.047]; Cleveland[0.047]; Boston[0.045]; Miami[0.045]; Clippers[0.045]; Orlando[0.044]; Phoenix[0.044]; Phoenix[0.044]; Charlotte[0.044]; Denver[0.044]; Detroit[0.044]; Division[0.043]; Sacramento[0.043]; Sacramento[0.043]; Friday[0.043]; State[0.042]; State[0.042]; Seattle[0.042]; Portland[0.042]; Portland[0.042]; Minnesota[0.042]; Utah[0.042]; Indiana[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew York\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew York Knicks\u001b[39m <---> \u001b[32mNew York\u001b[39m\t\nSCORES: global= 0.244:0.234[\u001b[31m0.010\u001b[39m]; local()= 0.206:0.086[\u001b[31m0.119\u001b[39m]; log p(e|m)= -4.828:-0.289[\u001b[32m4.539\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.049]; Basketball[0.049]; games[0.046]; games[0.046]; Philadelphia[0.045]; Thursday[0.044]; Thursday[0.044]; Cleveland[0.044]; Boston[0.044]; teams[0.043]; Miami[0.043]; Conference[0.043]; played[0.043]; Washington[0.042]; Detroit[0.042]; York[0.042]; Charlotte[0.042]; Orlando[0.042]; National[0.041]; Jersey[0.041]; won[0.041]; Central[0.041]; Atlanta[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Minnesota ==> ENTITY: \u001b[32mMinnesota Timberwolves\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.193:0.193[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.046]; Cleveland[0.046]; Minnesota[0.046]; Detroit[0.045]; Detroit[0.045]; Philadelphia[0.044]; Houston[0.044]; Charlotte[0.044]; Phoenix[0.044]; Phoenix[0.044]; Conference[0.043]; Miami[0.043]; Boston[0.043]; Boston[0.043]; Dallas[0.043]; Clippers[0.043]; Orlando[0.043]; Central[0.042]; Friday[0.042]; Washington[0.042]; Toronto[0.042]; Denver[0.042]; Atlanta[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Philadelphia ==> ENTITY: \u001b[32mPhiladelphia 76ers\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.211:0.211[0]; log p(e|m)= -3.912:-3.912[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteams[0.050]; games[0.049]; games[0.049]; Basketball[0.048]; Basketball[0.048]; Cleveland[0.047]; Toronto[0.046]; played[0.045]; won[0.045]; Boston[0.045]; Miami[0.045]; Washington[0.045]; Orlando[0.044]; Games[0.044]; National[0.044]; Conference[0.044]; Conference[0.044]; York[0.044]; York[0.044]; Houston[0.044]; Jersey[0.043]; Charlotte[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = LA LAKERS\u001b[39m ==> ENTITY: \u001b[32mLos Angeles Lakers\u001b[39m\t\nSCORES: global= 0.318:0.318[0]; local()= 0.217:0.217[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConference[0.045]; Lakers[0.044]; Cleveland[0.044]; Cleveland[0.044]; Houston[0.043]; Boston[0.042]; Miami[0.042]; Western[0.042]; Dallas[0.042]; Clippers[0.042]; Orlando[0.042]; Toronto[0.041]; Phoenix[0.041]; Phoenix[0.041]; Charlotte[0.041]; Charlotte[0.041]; Denver[0.041]; Detroit[0.041]; Detroit[0.041]; Division[0.040]; Division[0.040]; Sacramento[0.040]; Sacramento[0.040]; Friday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Miami ==> ENTITY: \u001b[32mMiami Heat\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.202:0.202[0]; log p(e|m)= -3.730:-3.730[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.045]; Orlando[0.044]; Houston[0.044]; Boston[0.043]; Charlotte[0.043]; Dallas[0.043]; State[0.042]; State[0.042]; Detroit[0.042]; Lakers[0.042]; Lakers[0.042]; Clippers[0.042]; Friday[0.041]; Denver[0.041]; Indiana[0.041]; Phoenix[0.041]; Phoenix[0.041]; Sacramento[0.041]; Sacramento[0.041]; Division[0.040]; York[0.040]; Minnesota[0.040]; Minnesota[0.040]; December[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Phoenix ==> ENTITY: \u001b[32mPhoenix Suns\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.207:0.207[0]; log p(e|m)= -3.474:-3.474[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.046]; Miami[0.044]; Phoenix[0.044]; Houston[0.043]; Vancouver[0.043]; Vancouver[0.043]; Orlando[0.043]; Boston[0.043]; Dallas[0.042]; Detroit[0.042]; Lakers[0.042]; Lakers[0.042]; Clippers[0.042]; Charlotte[0.042]; Denver[0.041]; Friday[0.041]; Utah[0.041]; Utah[0.041]; Pacific[0.040]; Division[0.040]; December[0.040]; Seattle[0.040]; Sacramento[0.040]; Sacramento[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Charlotte ==> ENTITY: \u001b[32mNew Orleans Pelicans\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.243:0.243[0]; log p(e|m)= -4.828:-4.828[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.045]; Houston[0.044]; games[0.044]; games[0.044]; Miami[0.042]; Conference[0.042]; Conference[0.042]; Cleveland[0.042]; teams[0.042]; Philadelphia[0.042]; Atlanta[0.042]; Orlando[0.041]; Dallas[0.041]; Thursday[0.041]; Thursday[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Central[0.041]; Boston[0.041]; Detroit[0.040]; Clippers[0.040]; Toronto[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Orlando ==> ENTITY: \u001b[32mOrlando Magic\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.221:0.221[0]; log p(e|m)= -3.079:-3.079[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; games[0.045]; Basketball[0.045]; Basketball[0.045]; Cleveland[0.042]; Miami[0.042]; Conference[0.042]; Conference[0.042]; teams[0.042]; won[0.042]; Houston[0.042]; Philadelphia[0.041]; Boston[0.041]; Atlanta[0.041]; Thursday[0.041]; Thursday[0.041]; Detroit[0.040]; Charlotte[0.040]; Toronto[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; played[0.040]; Atlantic[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dallas ==> ENTITY: \u001b[32mDallas Mavericks\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.191:0.191[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.047]; Washington[0.046]; Conference[0.046]; Toronto[0.044]; Miami[0.044]; Vancouver[0.044]; Vancouver[0.044]; Philadelphia[0.044]; Cleveland[0.043]; Cleveland[0.043]; Boston[0.043]; Boston[0.043]; York[0.043]; Clippers[0.043]; Detroit[0.043]; Detroit[0.043]; Phoenix[0.043]; Phoenix[0.043]; Charlotte[0.042]; Charlotte[0.042]; Central[0.042]; Lakers[0.042]; Atlanta[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = National Basketball Association\u001b[39m ==> ENTITY: \u001b[32mNational Basketball Association\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.047]; Cleveland[0.044]; games[0.044]; games[0.044]; teams[0.044]; Miami[0.044]; Philadelphia[0.043]; Boston[0.042]; Conference[0.042]; Thursday[0.042]; Thursday[0.042]; Detroit[0.041]; Orlando[0.041]; played[0.041]; Atlanta[0.041]; won[0.041]; Indiana[0.041]; Charlotte[0.040]; Chicago[0.040]; percentage[0.040]; Division[0.039]; Division[0.039]; Games[0.039]; Washington[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portland ==> ENTITY: \u001b[32mPortland Trail Blazers\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.214:0.214[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.044]; Cleveland[0.044]; Conference[0.043]; Boston[0.043]; Houston[0.043]; Miami[0.043]; Portland[0.043]; Toronto[0.042]; Clippers[0.042]; Orlando[0.042]; Pacific[0.042]; Seattle[0.041]; Dallas[0.041]; Atlanta[0.041]; Phoenix[0.041]; Phoenix[0.041]; Division[0.041]; Division[0.041]; Charlotte[0.041]; Charlotte[0.041]; Denver[0.041]; Detroit[0.041]; Lakers[0.040]; Lakers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Detroit ==> ENTITY: \u001b[32mDetroit Pistons\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.217:0.217[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; games[0.047]; Basketball[0.044]; Basketball[0.044]; Cleveland[0.043]; teams[0.042]; Conference[0.042]; Conference[0.042]; Philadelphia[0.042]; Vancouver[0.042]; Toronto[0.042]; Thursday[0.041]; Thursday[0.041]; Boston[0.041]; Houston[0.041]; played[0.041]; Charlotte[0.041]; Miami[0.040]; Dallas[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Orlando[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sacramento ==> ENTITY: \u001b[32mSacramento Kings\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.186:0.186[0]; log p(e|m)= -3.079:-3.079[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDivision[0.047]; Cleveland[0.046]; Miami[0.045]; Sacramento[0.044]; Pacific[0.044]; Charlotte[0.044]; Clippers[0.044]; York[0.044]; Boston[0.044]; Orlando[0.043]; Phoenix[0.043]; Phoenix[0.043]; San[0.043]; San[0.043]; Friday[0.043]; Dallas[0.043]; Denver[0.042]; Portland[0.042]; Portland[0.042]; Detroit[0.042]; Lakers[0.042]; Lakers[0.042]; Indiana[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vancouver ==> ENTITY: \u001b[32mMemphis Grizzlies\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.207:0.207[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVancouver[0.051]; Cleveland[0.046]; Charlotte[0.045]; Miami[0.045]; Clippers[0.044]; Phoenix[0.044]; Phoenix[0.044]; Orlando[0.044]; Detroit[0.044]; Boston[0.044]; Portland[0.043]; Portland[0.043]; Lakers[0.043]; Lakers[0.043]; Dallas[0.042]; Denver[0.042]; Seattle[0.042]; December[0.042]; Friday[0.042]; Pacific[0.042]; Utah[0.042]; Utah[0.042]; Indiana[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Milwaukee ==> ENTITY: \u001b[32mMilwaukee Bucks\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.219:0.219[0]; log p(e|m)= -3.124:-3.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.044]; Basketball[0.044]; games[0.044]; games[0.044]; Cleveland[0.044]; Boston[0.042]; Houston[0.042]; teams[0.042]; Conference[0.042]; Conference[0.042]; Detroit[0.042]; Miami[0.041]; Atlanta[0.041]; Central[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; played[0.041]; lost[0.041]; Orlando[0.040]; Thursday[0.040]; Clippers[0.040]; Dallas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Denver ==> ENTITY: \u001b[32mDenver Nuggets\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.223:0.223[0]; log p(e|m)= -3.270:-3.270[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.045]; Philadelphia[0.045]; Cleveland[0.045]; Cleveland[0.045]; Miami[0.045]; Orlando[0.045]; Conference[0.044]; Phoenix[0.044]; Phoenix[0.044]; Charlotte[0.044]; Charlotte[0.044]; Dallas[0.044]; Boston[0.043]; Boston[0.043]; Clippers[0.043]; Central[0.043]; Utah[0.043]; Utah[0.043]; Toronto[0.042]; Atlanta[0.042]; Indiana[0.042]; Indiana[0.042]; Chicago[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Boston ==> ENTITY: \u001b[32mBoston Celtics\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.218:0.218[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; games[0.046]; York[0.045]; York[0.045]; Basketball[0.044]; Basketball[0.044]; Cleveland[0.043]; Philadelphia[0.043]; teams[0.042]; Toronto[0.042]; Detroit[0.041]; played[0.041]; Houston[0.041]; Miami[0.040]; Conference[0.040]; Conference[0.040]; won[0.040]; Washington[0.040]; Charlotte[0.040]; Thursday[0.039]; Thursday[0.039]; Orlando[0.039]; Western[0.039]; Chicago[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = LA CLIPPERS\u001b[39m ==> ENTITY: \u001b[32mLos Angeles Clippers\u001b[39m\t\nSCORES: global= 0.318:0.318[0]; local()= 0.229:0.229[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.046]; Boston[0.045]; Miami[0.044]; Houston[0.044]; Orlando[0.043]; Phoenix[0.042]; Phoenix[0.042]; Toronto[0.042]; Denver[0.042]; Dallas[0.042]; Charlotte[0.042]; Charlotte[0.042]; Lakers[0.041]; Lakers[0.041]; Sacramento[0.041]; Sacramento[0.041]; Conference[0.041]; Detroit[0.040]; Portland[0.040]; Portland[0.040]; December[0.040]; Utah[0.040]; Utah[0.040]; Milwaukee[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NBA\u001b[39m ==> ENTITY: \u001b[32mNational Basketball Association\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.046]; Basketball[0.046]; games[0.045]; games[0.045]; Conference[0.044]; teams[0.044]; Division[0.043]; Division[0.043]; Cleveland[0.041]; won[0.041]; Boston[0.041]; Thursday[0.041]; Thursday[0.041]; Miami[0.041]; Philadelphia[0.041]; played[0.041]; Orlando[0.041]; Eastern[0.040]; National[0.040]; Detroit[0.040]; Games[0.040]; Central[0.040]; Atlantic[0.039]; Atlanta[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlantic ==> ENTITY: \u001b[32mAtlantic Division (NBA)\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.242:0.242[0]; log p(e|m)= -3.101:-3.101[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; games[0.045]; teams[0.045]; Basketball[0.044]; Basketball[0.044]; won[0.043]; Division[0.043]; Division[0.043]; Division[0.043]; Boston[0.043]; Philadelphia[0.042]; Conference[0.042]; Conference[0.042]; Detroit[0.041]; Cleveland[0.041]; Toronto[0.041]; lost[0.040]; Atlanta[0.040]; played[0.040]; Eastern[0.039]; Jersey[0.039]; Miami[0.039]; Western[0.039]; Charlotte[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Antonio ==> ENTITY: \u001b[32mSan Antonio Spurs\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.200:0.200[0]; log p(e|m)= -2.865:-2.865[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDallas[0.044]; San[0.044]; Cleveland[0.043]; Orlando[0.043]; Charlotte[0.043]; Miami[0.043]; Phoenix[0.042]; Phoenix[0.042]; Detroit[0.042]; Boston[0.042]; Antonio[0.042]; Lakers[0.042]; Lakers[0.042]; Division[0.042]; Clippers[0.041]; Golden[0.041]; Golden[0.041]; Indiana[0.041]; Denver[0.041]; Friday[0.041]; Utah[0.041]; Sacramento[0.040]; Sacramento[0.040]; Pct[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Minnesota ==> ENTITY: \u001b[32mMinnesota Timberwolves\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.192:0.192[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.045]; Minnesota[0.045]; Detroit[0.044]; Charlotte[0.043]; Phoenix[0.043]; Phoenix[0.043]; Miami[0.042]; Boston[0.042]; Dallas[0.042]; Clippers[0.042]; Orlando[0.042]; Friday[0.041]; Denver[0.041]; Division[0.041]; State[0.041]; State[0.041]; Lakers[0.041]; Lakers[0.041]; Indiana[0.041]; December[0.040]; Sacramento[0.040]; Sacramento[0.040]; Portland[0.040]; Portland[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indiana ==> ENTITY: \u001b[32mIndiana Pacers\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.228:0.228[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.046]; games[0.044]; games[0.044]; Cleveland[0.043]; Philadelphia[0.043]; Charlotte[0.043]; Conference[0.043]; Conference[0.043]; Miami[0.042]; Houston[0.041]; Toronto[0.041]; Boston[0.041]; played[0.041]; Orlando[0.041]; teams[0.041]; Clippers[0.041]; Thursday[0.041]; Dallas[0.040]; Washington[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Detroit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Antonio ==> ENTITY: \u001b[32mSan Antonio Spurs\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.208:0.208[0]; log p(e|m)= -2.865:-2.865[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConference[0.044]; Houston[0.044]; Dallas[0.043]; San[0.043]; Cleveland[0.042]; Cleveland[0.042]; Orlando[0.042]; Charlotte[0.042]; Charlotte[0.042]; Miami[0.042]; Phoenix[0.042]; Phoenix[0.042]; Detroit[0.041]; Detroit[0.041]; Boston[0.041]; Boston[0.041]; Antonio[0.041]; Lakers[0.041]; Lakers[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Atlanta[0.040]; Toronto[0.040]; \t\n\r [=======>......................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 16s136ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 781/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1282testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChicago[0.048]; Chicago[0.048]; Chicago[0.048]; Chicago[0.048]; Chicago[0.048]; Chicago[0.048]; according[0.046]; locations[0.045]; Minneapolis[0.043]; Louis[0.043]; Louis[0.043]; Louis[0.043]; Board[0.042]; Toledo[0.041]; Toledo[0.041]; Toledo[0.041]; day[0.041]; delivery[0.041]; previous[0.040]; receipts[0.040]; receipts[0.040]; trading[0.040]; Trade[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = St. Louis ==> ENTITY: \u001b[32mSt. Louis\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.094:0.094[0]; log p(e|m)= -0.285:-0.285[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \naccording[0.048]; Chicago[0.046]; Chicago[0.046]; Chicago[0.046]; Chicago[0.046]; Chicago[0.046]; Chicago[0.046]; Chicago[0.046]; Louis[0.046]; Louis[0.046]; locations[0.044]; delivery[0.042]; Trade[0.042]; previous[0.042]; soybean[0.042]; Minneapolis[0.041]; Toledo[0.041]; Toledo[0.041]; Toledo[0.041]; trading[0.041]; Board[0.041]; day[0.041]; bushels[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Toledo ==> ENTITY: \u001b[32mToledo, Ohio\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.056:0.056[0]; log p(e|m)= -1.211:-1.211[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nToledo[0.045]; Toledo[0.045]; Louis[0.043]; Louis[0.043]; Louis[0.043]; Chicago[0.042]; Chicago[0.042]; Chicago[0.042]; Chicago[0.042]; Chicago[0.042]; Chicago[0.042]; Chicago[0.042]; locations[0.042]; according[0.042]; Shipments[0.041]; day[0.041]; previous[0.040]; Wheat[0.040]; trading[0.040]; Soybeans[0.040]; Minneapolis[0.040]; Corn[0.040]; Board[0.039]; Grain[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Toledo ==> ENTITY: \u001b[32mToledo, Ohio\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.056:0.056[0]; log p(e|m)= -1.211:-1.211[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nToledo[0.045]; Toledo[0.045]; Louis[0.043]; Louis[0.043]; Louis[0.043]; Chicago[0.042]; Chicago[0.042]; Chicago[0.042]; Chicago[0.042]; Chicago[0.042]; Chicago[0.042]; Chicago[0.042]; locations[0.042]; according[0.042]; Shipments[0.041]; day[0.041]; previous[0.040]; Wheat[0.040]; trading[0.040]; Soybeans[0.040]; Minneapolis[0.040]; Corn[0.040]; Board[0.039]; Grain[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Toledo ==> ENTITY: \u001b[32mToledo, Ohio\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.056:0.056[0]; log p(e|m)= -1.211:-1.211[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nToledo[0.045]; Toledo[0.045]; Louis[0.043]; Louis[0.043]; Louis[0.043]; Chicago[0.042]; Chicago[0.042]; Chicago[0.042]; Chicago[0.042]; Chicago[0.042]; Chicago[0.042]; Chicago[0.042]; locations[0.042]; according[0.042]; Shipments[0.041]; day[0.041]; previous[0.040]; Wheat[0.040]; trading[0.040]; Soybeans[0.040]; Minneapolis[0.040]; Corn[0.040]; Board[0.039]; Grain[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChicago[0.048]; Chicago[0.048]; Chicago[0.048]; Chicago[0.048]; Chicago[0.048]; Chicago[0.048]; according[0.046]; locations[0.045]; Minneapolis[0.043]; Louis[0.043]; Louis[0.043]; Louis[0.043]; Board[0.042]; Toledo[0.041]; Toledo[0.041]; Toledo[0.041]; day[0.041]; delivery[0.041]; previous[0.040]; receipts[0.040]; receipts[0.040]; trading[0.040]; Trade[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = CBOT\u001b[39m ==> ENTITY: \u001b[32mChicago Board of Trade\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrading[0.046]; Chicago[0.044]; Chicago[0.044]; Chicago[0.044]; Chicago[0.044]; Chicago[0.044]; Chicago[0.044]; Chicago[0.044]; Board[0.043]; Trade[0.042]; according[0.042]; receipts[0.041]; receipts[0.041]; Corn[0.041]; Wheat[0.041]; Grain[0.040]; day[0.040]; locations[0.040]; delivery[0.040]; grain[0.039]; soybean[0.039]; previous[0.039]; Receipts[0.039]; oilseed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChicago[0.048]; Chicago[0.048]; Chicago[0.048]; Chicago[0.048]; Chicago[0.048]; Chicago[0.048]; according[0.046]; locations[0.045]; Minneapolis[0.043]; Louis[0.043]; Louis[0.043]; Louis[0.043]; Board[0.042]; Toledo[0.041]; Toledo[0.041]; Toledo[0.041]; day[0.041]; delivery[0.041]; previous[0.040]; receipts[0.040]; receipts[0.040]; trading[0.040]; Trade[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Minneapolis ==> ENTITY: \u001b[32mMinneapolis\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.088:0.088[0]; log p(e|m)= -0.013:-0.013[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChicago[0.046]; Chicago[0.046]; Chicago[0.046]; Chicago[0.046]; Chicago[0.046]; Chicago[0.046]; Chicago[0.046]; locations[0.044]; according[0.044]; delivery[0.043]; day[0.043]; Toledo[0.043]; Toledo[0.043]; Toledo[0.043]; Board[0.043]; soybean[0.042]; Louis[0.042]; Louis[0.042]; Louis[0.042]; shipments[0.041]; shipments[0.041]; previous[0.041]; trading[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChicago[0.048]; Chicago[0.048]; Chicago[0.048]; Chicago[0.048]; Chicago[0.048]; Chicago[0.048]; according[0.046]; locations[0.045]; Minneapolis[0.043]; Louis[0.043]; Louis[0.043]; Louis[0.043]; Board[0.042]; Toledo[0.041]; Toledo[0.041]; Toledo[0.041]; day[0.041]; delivery[0.041]; previous[0.040]; receipts[0.040]; receipts[0.040]; trading[0.040]; Trade[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chicago Board of Trade\u001b[39m ==> ENTITY: \u001b[32mChicago Board of Trade\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrading[0.047]; Chicago[0.045]; Chicago[0.045]; Chicago[0.045]; Chicago[0.045]; Chicago[0.045]; Chicago[0.045]; according[0.042]; receipts[0.042]; receipts[0.042]; Corn[0.041]; Wheat[0.041]; Grain[0.041]; day[0.041]; delivery[0.040]; grain[0.040]; soybean[0.040]; previous[0.040]; Receipts[0.040]; oilseed[0.040]; Oats[0.039]; Toledo[0.039]; Toledo[0.039]; Toledo[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChicago[0.048]; Chicago[0.048]; Chicago[0.048]; Chicago[0.048]; Chicago[0.048]; Chicago[0.048]; according[0.046]; locations[0.045]; Minneapolis[0.043]; Louis[0.043]; Louis[0.043]; Louis[0.043]; Board[0.042]; Toledo[0.041]; Toledo[0.041]; Toledo[0.041]; day[0.041]; delivery[0.041]; previous[0.040]; receipts[0.040]; receipts[0.040]; trading[0.040]; Trade[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = St. Louis ==> ENTITY: \u001b[32mSt. Louis\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.094:0.094[0]; log p(e|m)= -0.285:-0.285[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \naccording[0.048]; Chicago[0.046]; Chicago[0.046]; Chicago[0.046]; Chicago[0.046]; Chicago[0.046]; Chicago[0.046]; Chicago[0.046]; Louis[0.046]; Louis[0.046]; locations[0.044]; delivery[0.042]; Trade[0.042]; previous[0.042]; soybean[0.042]; Minneapolis[0.041]; Toledo[0.041]; Toledo[0.041]; Toledo[0.041]; trading[0.041]; Board[0.041]; day[0.041]; bushels[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = St. Louis ==> ENTITY: \u001b[32mSt. Louis\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.094:0.094[0]; log p(e|m)= -0.285:-0.285[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \naccording[0.048]; Chicago[0.046]; Chicago[0.046]; Chicago[0.046]; Chicago[0.046]; Chicago[0.046]; Chicago[0.046]; Chicago[0.046]; Louis[0.046]; Louis[0.046]; locations[0.044]; delivery[0.042]; Trade[0.042]; previous[0.042]; soybean[0.042]; Minneapolis[0.041]; Toledo[0.041]; Toledo[0.041]; Toledo[0.041]; trading[0.041]; Board[0.041]; day[0.041]; bushels[0.040]; \t\n\r [=======>......................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 16s230ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 795/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1351testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.151:0.151[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.046]; Norway[0.046]; Biathlon[0.043]; Men[0.043]; biathlon[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Nordic[0.042]; Cup[0.041]; Cup[0.041]; Cup[0.041]; Belarus[0.041]; Women[0.041]; Sven[0.041]; Russia[0.041]; Russia[0.041]; Andresen[0.040]; World[0.040]; World[0.040]; World[0.040]; Pavel[0.040]; standings[0.040]; seconds[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paramygina\u001b[39m ==> ENTITY: \u001b[32mSvetlana Paramygina\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.049]; Russia[0.049]; Cup[0.046]; Cup[0.046]; Belorus[0.046]; standings[0.045]; standings[0.045]; Svetlana[0.045]; World[0.044]; World[0.044]; Pavel[0.044]; Muslimov[0.043]; Ole[0.042]; Norway[0.042]; Norway[0.042]; Olga[0.042]; points[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Women[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Svetlana Paramygina\u001b[39m ==> ENTITY: \u001b[32mSvetlana Paramygina\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbiathlon[0.050]; Biathlon[0.047]; Russia[0.046]; Russia[0.046]; Belarus[0.046]; Cup[0.044]; Cup[0.044]; Cup[0.044]; Cup[0.044]; Belorus[0.044]; standings[0.043]; standings[0.043]; Frode[0.043]; races[0.042]; World[0.042]; World[0.042]; World[0.042]; World[0.042]; Pavel[0.042]; Nordic[0.041]; Muslimov[0.041]; Vadim[0.041]; Ole[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Greiner-Petter-Memm\u001b[39m ==> ENTITY: \u001b[32mSimone Greiner-Petter-Memm\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstandings[0.050]; standings[0.050]; Petter[0.050]; Petter[0.050]; Greiner[0.049]; Muslimov[0.048]; points[0.048]; Ole[0.048]; Simone[0.048]; Cup[0.048]; Cup[0.048]; World[0.047]; World[0.047]; Norway[0.047]; Norway[0.047]; Margit[0.047]; kms[0.046]; Pavel[0.046]; Olga[0.045]; Women[0.045]; Einar[0.045]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWorld Cup\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFIS Alpine Ski World Cup\u001b[39m <---> \u001b[32mBiathlon World Cup\u001b[39m\t\nSCORES: global= 0.245:0.229[\u001b[31m0.016\u001b[39m]; local()= 0.202:0.205[\u001b[32m0.003\u001b[39m]; log p(e|m)= -2.765:-4.343[\u001b[31m1.578\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.045]; Norway[0.044]; Norway[0.044]; Norway[0.044]; World[0.043]; World[0.043]; World[0.043]; Cup[0.043]; Cup[0.043]; Cup[0.043]; Skiing[0.041]; biathlon[0.041]; Biathlon[0.041]; races[0.041]; standings[0.040]; standings[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Nordic[0.040]; Women[0.040]; Men[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.164:0.164[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.046]; Norway[0.046]; Sweden[0.046]; Men[0.043]; World[0.042]; World[0.042]; World[0.042]; World[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Cup[0.041]; Cup[0.041]; Cup[0.041]; Cup[0.041]; Biathlon[0.040]; Nordic[0.040]; Sven[0.040]; Women[0.039]; standings[0.039]; standings[0.039]; Andreassen[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Norway[0.043]; Norway[0.043]; Norway[0.043]; standings[0.042]; standings[0.042]; Cup[0.042]; Cup[0.042]; Russia[0.041]; Russia[0.041]; World[0.041]; World[0.041]; Sven[0.040]; Fischer[0.039]; Fischer[0.039]; Belorus[0.039]; points[0.039]; penalty[0.039]; Gross[0.039]; minutes[0.039]; rounds[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Frode Andresen\u001b[39m ==> ENTITY: \u001b[32mFrode Andresen\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBiathlon[0.049]; biathlon[0.047]; standings[0.046]; standings[0.046]; Norway[0.045]; Norway[0.045]; Norway[0.045]; races[0.045]; seconds[0.044]; Sweden[0.044]; Nordic[0.044]; Muslimov[0.043]; Ole[0.043]; Skiing[0.042]; points[0.042]; Cup[0.042]; Cup[0.042]; Cup[0.042]; Cup[0.042]; Pavel[0.042]; Petter[0.040]; Einar[0.040]; Saturday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sashurin\u001b[39m ==> ENTITY: \u001b[32mVadim Sashurin\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVadim[0.048]; Pavel[0.046]; Belarus[0.045]; biathlon[0.045]; Svetlana[0.045]; Belorus[0.045]; Olga[0.043]; Frode[0.043]; Biathlon[0.042]; Russia[0.041]; Russia[0.041]; Melnik[0.041]; Muslimov[0.041]; Nordic[0.040]; Ole[0.040]; Norway[0.040]; Norway[0.040]; Norway[0.040]; Sweden[0.040]; Skiing[0.040]; Saturday[0.039]; Margit[0.038]; Petter[0.038]; Petter[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Simone Greiner-Petter-Memm\u001b[39m ==> ENTITY: \u001b[32mSimone Greiner-Petter-Memm\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseconds[0.046]; biathlon[0.046]; Frode[0.045]; standings[0.045]; standings[0.045]; races[0.045]; Petter[0.045]; Petter[0.045]; Greiner[0.044]; penalty[0.043]; Muslimov[0.043]; points[0.043]; Ole[0.043]; Cup[0.043]; Cup[0.043]; Cup[0.043]; World[0.042]; World[0.042]; World[0.042]; Norway[0.042]; Norway[0.042]; Norway[0.042]; Margit[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.167:0.167[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.048]; Germany[0.048]; Germany[0.048]; Germany[0.048]; Sweden[0.045]; Norway[0.044]; Norway[0.044]; Norway[0.044]; Belarus[0.043]; standings[0.043]; standings[0.043]; Cup[0.043]; Cup[0.043]; Cup[0.043]; Cup[0.043]; Russia[0.042]; Russia[0.042]; World[0.042]; World[0.042]; World[0.042]; World[0.042]; Sven[0.041]; biathlon[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.164:0.164[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; biathlon[0.043]; Sweden[0.043]; Biathlon[0.042]; Norway[0.042]; Norway[0.042]; Norway[0.042]; Belarus[0.041]; standings[0.041]; standings[0.041]; Cup[0.041]; Cup[0.041]; Cup[0.041]; Cup[0.041]; Saturday[0.040]; Russia[0.040]; Russia[0.040]; World[0.040]; World[0.040]; World[0.040]; World[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.169:0.169[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.046]; Norway[0.046]; Sweden[0.045]; Men[0.043]; World[0.042]; World[0.042]; World[0.042]; World[0.042]; Biathlon[0.042]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Nordic[0.041]; Cup[0.041]; Cup[0.041]; Cup[0.041]; Cup[0.041]; biathlon[0.041]; Sven[0.040]; Women[0.039]; Andreassen[0.039]; Skiing[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Belorus\u001b[39m ==> ENTITY: \u001b[32mBelarus\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.047]; Russia[0.046]; Russia[0.046]; Frode[0.044]; Melnik[0.043]; Pavel[0.043]; Svetlana[0.042]; Olga[0.042]; Muslimov[0.041]; Nordic[0.041]; Ole[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Vadim[0.041]; Skiing[0.040]; biathlon[0.040]; Norway[0.040]; Norway[0.040]; Norway[0.040]; Sweden[0.039]; Biathlon[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWorld Cup\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFIS Alpine Ski World Cup\u001b[39m <---> \u001b[32mBiathlon World Cup\u001b[39m\t\nSCORES: global= 0.242:0.228[\u001b[31m0.014\u001b[39m]; local()= 0.175:0.188[\u001b[32m0.013\u001b[39m]; log p(e|m)= -2.765:-4.343[\u001b[31m1.578\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.045]; Norway[0.045]; Norway[0.045]; Norway[0.045]; World[0.044]; World[0.044]; Cup[0.044]; Cup[0.044]; Skiing[0.042]; biathlon[0.042]; Biathlon[0.041]; races[0.041]; standings[0.041]; Nordic[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Women[0.040]; rounds[0.040]; Men[0.039]; Russia[0.038]; Russia[0.038]; penalty[0.038]; Belorus[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWorld Cup\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFIS Alpine Ski World Cup\u001b[39m <---> \u001b[32mBiathlon World Cup\u001b[39m\t\nSCORES: global= 0.242:0.228[\u001b[31m0.014\u001b[39m]; local()= 0.142:0.172[\u001b[32m0.030\u001b[39m]; log p(e|m)= -2.765:-4.343[\u001b[31m1.578\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.047]; Norway[0.047]; Norway[0.047]; World[0.046]; Cup[0.045]; standings[0.043]; standings[0.043]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Women[0.042]; rounds[0.041]; Russia[0.040]; Russia[0.040]; penalty[0.040]; Belorus[0.040]; Ole[0.039]; points[0.039]; Sven[0.038]; Andreassen[0.038]; Einar[0.038]; Pavel[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Olga Melnik\u001b[39m ==> ENTITY: \u001b[32mOlga Melnik\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.092:0.092[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbiathlon[0.049]; Biathlon[0.049]; Frode[0.047]; Svetlana[0.046]; Belorus[0.045]; Muslimov[0.044]; Vadim[0.044]; Nordic[0.044]; Ole[0.044]; Belarus[0.043]; Skiing[0.043]; Andreassen[0.043]; Pavel[0.043]; Russia[0.042]; Russia[0.042]; Andresen[0.042]; Petter[0.041]; Petter[0.041]; Norway[0.041]; Norway[0.041]; Norway[0.041]; Women[0.041]; Einar[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.167:0.167[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.048]; Germany[0.048]; Germany[0.048]; Germany[0.048]; Sweden[0.045]; Norway[0.044]; Norway[0.044]; Norway[0.044]; Belarus[0.043]; standings[0.043]; standings[0.043]; Cup[0.043]; Cup[0.043]; Cup[0.043]; Cup[0.043]; Russia[0.042]; Russia[0.042]; World[0.042]; World[0.042]; World[0.042]; World[0.042]; Sven[0.041]; biathlon[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ole Einar Bjorndalen\u001b[39m ==> ENTITY: \u001b[32mOle Einar Bjørndalen\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbiathlon[0.055]; Biathlon[0.055]; Nordic[0.050]; Skiing[0.049]; Norway[0.049]; Norway[0.049]; Norway[0.049]; Frode[0.048]; races[0.048]; Sweden[0.047]; standings[0.047]; standings[0.047]; seconds[0.047]; points[0.047]; Muslimov[0.046]; rounds[0.045]; Andresen[0.045]; Cup[0.045]; Cup[0.045]; Cup[0.045]; Cup[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Norway[0.042]; Norway[0.042]; Norway[0.042]; Belarus[0.042]; standings[0.042]; standings[0.042]; Cup[0.041]; Cup[0.041]; Russia[0.041]; Russia[0.041]; World[0.041]; World[0.041]; Sven[0.040]; Fischer[0.039]; Fischer[0.039]; Belorus[0.039]; points[0.039]; penalty[0.039]; Gross[0.039]; Men[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pavel Muslimov\u001b[39m ==> ENTITY: \u001b[32mPavel Muslimov\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.238:0.238[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.046]; Vadim[0.046]; Svetlana[0.046]; Belorus[0.045]; Russia[0.043]; Russia[0.043]; biathlon[0.043]; Olga[0.042]; Sweden[0.041]; Frode[0.041]; Melnik[0.041]; Biathlon[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Norway[0.041]; Norway[0.041]; Norway[0.041]; Nordic[0.039]; Fischer[0.039]; Fischer[0.039]; Ole[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gunn Margit Andreassen\u001b[39m ==> ENTITY: \u001b[32mGunn Margit Andreassen\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.047]; Norway[0.047]; Norway[0.047]; Frode[0.046]; biathlon[0.046]; Andresen[0.045]; races[0.045]; Muslimov[0.043]; Ole[0.043]; Sweden[0.043]; World[0.043]; World[0.043]; World[0.043]; Einar[0.042]; Cup[0.042]; Cup[0.042]; Cup[0.042]; standings[0.042]; standings[0.042]; Pavel[0.042]; Petter[0.041]; Petter[0.041]; Sven[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fischer\u001b[39m ==> ENTITY: \u001b[32mSven Fischer\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; biathlon[0.043]; Sven[0.043]; Biathlon[0.042]; Fischer[0.042]; World[0.042]; World[0.042]; World[0.042]; World[0.042]; Frode[0.042]; Sweden[0.041]; Norway[0.041]; Norway[0.041]; Norway[0.041]; Cup[0.040]; Cup[0.040]; Cup[0.040]; Cup[0.040]; Ole[0.040]; races[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sven Fischer\u001b[39m ==> ENTITY: \u001b[32mSven Fischer\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; biathlon[0.043]; Biathlon[0.042]; Fischer[0.042]; World[0.042]; World[0.042]; World[0.042]; World[0.042]; Frode[0.042]; Sweden[0.041]; Norway[0.041]; Norway[0.041]; Norway[0.041]; Cup[0.040]; Cup[0.040]; Cup[0.040]; Cup[0.040]; Ole[0.040]; races[0.040]; Muslimov[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Behle\u001b[39m ==> ENTITY: \u001b[32mPetra Behle\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBehle[0.045]; Frode[0.044]; Petra[0.043]; standings[0.043]; standings[0.043]; World[0.043]; World[0.043]; kms[0.042]; Muslimov[0.042]; Ole[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Pavel[0.041]; Norway[0.040]; Norway[0.040]; Norway[0.040]; Fischer[0.040]; Fischer[0.040]; Petter[0.040]; Petter[0.040]; Andreassen[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Oestersund\u001b[39m ==> ENTITY: \u001b[32mÖstersund\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.113:0.113[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.050]; Norway[0.048]; Norway[0.048]; Frode[0.047]; Muslimov[0.045]; Nordic[0.045]; Ole[0.045]; Skiing[0.044]; Biathlon[0.044]; Pavel[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Einar[0.043]; biathlon[0.042]; Men[0.041]; Andresen[0.041]; Melnik[0.041]; Belorus[0.041]; Sven[0.041]; minutes[0.041]; Women[0.040]; Saturday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Vadim Sashurin\u001b[39m ==> ENTITY: \u001b[32mVadim Sashurin\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPavel[0.047]; Belarus[0.046]; biathlon[0.046]; Svetlana[0.046]; Belorus[0.045]; Olga[0.044]; Frode[0.043]; Biathlon[0.043]; Russia[0.042]; Russia[0.042]; Melnik[0.041]; Muslimov[0.041]; Nordic[0.041]; Ole[0.041]; Norway[0.040]; Norway[0.040]; Norway[0.040]; Sweden[0.040]; Skiing[0.040]; Saturday[0.039]; Margit[0.039]; Petter[0.039]; Fischer[0.038]; Fischer[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.154:0.154[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.047]; Russia[0.046]; Sweden[0.043]; Pavel[0.042]; Belorus[0.042]; Cup[0.042]; Cup[0.042]; Cup[0.042]; Cup[0.042]; Norway[0.042]; Norway[0.042]; Norway[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Biathlon[0.041]; penalty[0.040]; Melnik[0.040]; Svetlana[0.040]; standings[0.040]; standings[0.040]; Nordic[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Petra Behle\u001b[39m ==> ENTITY: \u001b[32mPetra Behle\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBehle[0.045]; Frode[0.045]; standings[0.043]; standings[0.043]; World[0.043]; World[0.043]; kms[0.042]; Muslimov[0.042]; Ole[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Pavel[0.041]; Norway[0.041]; Norway[0.041]; Norway[0.041]; Fischer[0.040]; Fischer[0.040]; Petter[0.040]; Petter[0.040]; Andreassen[0.040]; Belorus[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.154:0.154[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.047]; Russia[0.046]; Sweden[0.043]; Pavel[0.042]; Belorus[0.042]; Cup[0.042]; Cup[0.042]; Cup[0.042]; Cup[0.042]; Norway[0.042]; Norway[0.042]; Norway[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Biathlon[0.041]; penalty[0.040]; Melnik[0.040]; Svetlana[0.040]; standings[0.040]; standings[0.040]; Nordic[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ricco Gross\u001b[39m ==> ENTITY: \u001b[32mRicco Groß\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.125:0.125[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbiathlon[0.047]; Biathlon[0.046]; Pavel[0.045]; Frode[0.044]; Margit[0.042]; Olga[0.042]; Muslimov[0.042]; Nordic[0.042]; Ole[0.042]; Sweden[0.041]; Sven[0.041]; Belarus[0.041]; Skiing[0.041]; Vadim[0.041]; Petter[0.041]; Petter[0.041]; Svetlana[0.041]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; standings[0.040]; standings[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.152:0.152[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.047]; Norway[0.047]; Men[0.044]; World[0.043]; World[0.043]; World[0.043]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Cup[0.041]; Cup[0.041]; Cup[0.041]; Sven[0.040]; Women[0.040]; standings[0.040]; standings[0.040]; Andreassen[0.040]; biathlon[0.040]; Russia[0.039]; Russia[0.039]; Ole[0.039]; Andresen[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belarus ==> ENTITY: \u001b[32mBelarus\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.135:0.135[0]; log p(e|m)= -0.422:-0.422[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWorld[0.048]; World[0.048]; World[0.048]; Men[0.046]; Biathlon[0.043]; Russia[0.042]; Russia[0.042]; Sweden[0.041]; Norway[0.041]; Norway[0.041]; Norway[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; biathlon[0.041]; Women[0.040]; Skiing[0.040]; Belorus[0.040]; Nordic[0.039]; Melnik[0.039]; Cup[0.039]; Cup[0.039]; Cup[0.039]; \t\n\r [=======>......................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 16s21ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 828/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1240testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Harare\u001b[39m ==> ENTITY: \u001b[32mHarare\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.118:0.118[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZimbabwe[0.049]; Zimbabwe[0.049]; Mugabe[0.045]; Ncube[0.044]; number[0.044]; international[0.044]; Friday[0.043]; Sindiso[0.043]; government[0.041]; Robert[0.041]; said[0.041]; past[0.041]; convicted[0.041]; convicted[0.041]; local[0.040]; resisted[0.040]; rights[0.040]; pressure[0.039]; abolish[0.039]; dawn[0.039]; bringing[0.039]; hanged[0.039]; hanged[0.039]; statement[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zimbabwe ==> ENTITY: \u001b[32mZimbabwe\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.367:-0.367[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZimbabwe[0.049]; rights[0.045]; Mugabe[0.045]; government[0.044]; President[0.044]; Harare[0.043]; international[0.043]; groups[0.043]; number[0.042]; Friday[0.042]; said[0.041]; statement[0.041]; Sindiso[0.041]; executions[0.041]; Ncube[0.040]; past[0.040]; abolish[0.040]; convicted[0.040]; convicted[0.040]; local[0.040]; Robert[0.040]; resisted[0.039]; pressure[0.039]; death[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Robert Mugabe\u001b[39m ==> ENTITY: \u001b[32mRobert Mugabe\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZimbabwe[0.050]; Zimbabwe[0.050]; President[0.046]; Harare[0.045]; government[0.044]; international[0.044]; Ncube[0.044]; rights[0.042]; Sindiso[0.041]; said[0.041]; statement[0.041]; groups[0.041]; local[0.040]; number[0.040]; abolish[0.040]; convicted[0.040]; convicted[0.040]; pressure[0.039]; Friday[0.039]; bringing[0.039]; resisted[0.039]; executions[0.038]; death[0.038]; carried[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zimbabwe ==> ENTITY: \u001b[32mZimbabwe\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.367:-0.367[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZimbabwe[0.049]; rights[0.045]; Mugabe[0.045]; government[0.044]; President[0.044]; Harare[0.043]; international[0.043]; groups[0.043]; number[0.042]; Friday[0.042]; said[0.041]; statement[0.041]; Sindiso[0.041]; executions[0.041]; Ncube[0.040]; past[0.040]; abolish[0.040]; convicted[0.040]; convicted[0.040]; local[0.040]; Robert[0.040]; resisted[0.039]; pressure[0.039]; death[0.039]; \t\n\r [=======>......................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 16s467ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 832/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1369testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Glasgow ==> ENTITY: \u001b[32mGlasgow\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.140:0.140[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScottish[0.047]; Scottish[0.047]; Aberdeen[0.046]; Dundee[0.044]; league[0.044]; matches[0.044]; League[0.043]; Dunfermline[0.043]; Kilmarnock[0.042]; won[0.042]; played[0.042]; Celtic[0.041]; Motherwell[0.041]; Rangers[0.041]; Hibernian[0.040]; Saturday[0.040]; Soccer[0.040]; Premier[0.039]; Raith[0.039]; Hearts[0.039]; Standings[0.039]; division[0.039]; lost[0.039]; United[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Berwick ==> ENTITY: \u001b[32mBerwick Rangers F.C.\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.151:0.151[0]; log p(e|m)= -3.689:-3.689[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLivingston[0.045]; Park[0.044]; Fife[0.044]; Stirling[0.043]; Stirling[0.043]; Cowdenbeath[0.043]; Stenhousemuir[0.043]; Inverness[0.042]; Ross[0.042]; Clyde[0.042]; Brechin[0.041]; Hamilton[0.041]; Queen[0.041]; Queen[0.041]; East[0.041]; East[0.041]; County[0.040]; Dumbarton[0.040]; Forfar[0.040]; South[0.040]; Division[0.040]; Division[0.040]; Thistle[0.040]; Alloa[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Montrose ==> ENTITY: \u001b[32mMontrose F.C.\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.169:0.169[0]; log p(e|m)= -2.180:-2.180[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCowdenbeath[0.047]; Stenhousemuir[0.046]; Brechin[0.046]; Inverness[0.045]; Fife[0.045]; Park[0.045]; Alloa[0.044]; Stirling[0.044]; Livingston[0.044]; Forfar[0.044]; Dumbarton[0.043]; Thistle[0.043]; Ross[0.043]; Stranraer[0.043]; Arbroath[0.043]; South[0.043]; Clyde[0.043]; Ayr[0.043]; East[0.042]; East[0.042]; Queen[0.041]; Queen[0.041]; County[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = St Johnstone\u001b[39m ==> ENTITY: \u001b[32mSt. Johnstone F.C.\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.045]; Livingston[0.043]; Dundee[0.043]; Dundee[0.043]; Aberdeen[0.043]; Dunfermline[0.042]; Partick[0.042]; Hearts[0.042]; Hibernian[0.042]; Kilmarnock[0.042]; Stirling[0.042]; Rangers[0.042]; goals[0.042]; goals[0.042]; Raith[0.041]; Falkirk[0.041]; matches[0.041]; Celtic[0.041]; played[0.041]; Motherwell[0.041]; Greenock[0.040]; Fife[0.040]; Division[0.040]; Division[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dundee United\u001b[39m ==> ENTITY: \u001b[32mDundee United F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDundee[0.045]; Airdrieonians[0.045]; Aberdeen[0.044]; League[0.044]; matches[0.042]; Hibernian[0.042]; Glasgow[0.042]; Celtic[0.042]; Dunfermline[0.042]; Kilmarnock[0.042]; Rangers[0.042]; Scottish[0.042]; Scottish[0.042]; Hearts[0.041]; league[0.041]; played[0.041]; Raith[0.041]; won[0.040]; Falkirk[0.040]; Partick[0.040]; Motherwell[0.040]; Soccer[0.040]; standings[0.040]; Premier[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Stranraer ==> ENTITY: \u001b[32mStranraer F.C.\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.214:0.214[0]; log p(e|m)= -1.030:-1.030[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLivingston[0.045]; Stenhousemuir[0.044]; Cowdenbeath[0.043]; Greenock[0.043]; Inverness[0.043]; Brechin[0.043]; Dumbarton[0.043]; Division[0.042]; Division[0.042]; Stirling[0.042]; Fife[0.042]; Thistle[0.042]; Ross[0.042]; Morton[0.041]; Park[0.041]; Alloa[0.041]; Forfar[0.041]; Ayr[0.041]; Clydebank[0.040]; Hamilton[0.040]; Clyde[0.040]; Berwick[0.040]; Albion[0.040]; Montrose[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Aberdeen ==> ENTITY: \u001b[32mAberdeen F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.204:0.204[0]; log p(e|m)= -1.448:-1.448[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScottish[0.044]; Scottish[0.044]; Airdrieonians[0.044]; matches[0.043]; Dundee[0.043]; Dundee[0.043]; league[0.043]; played[0.043]; League[0.043]; Glasgow[0.043]; Dunfermline[0.042]; Hibernian[0.041]; Motherwell[0.041]; Kilmarnock[0.041]; Rangers[0.041]; Celtic[0.041]; Falkirk[0.041]; Hearts[0.040]; Partick[0.040]; Soccer[0.040]; Raith[0.040]; won[0.039]; Johnstone[0.039]; Premier[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = St Mirren\u001b[39m ==> ENTITY: \u001b[32mSt. Mirren F.C.\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.045]; Stenhousemuir[0.044]; Dundee[0.044]; Dundee[0.044]; Livingston[0.043]; Dunfermline[0.043]; Kilmarnock[0.043]; Partick[0.043]; Hearts[0.042]; Hibernian[0.042]; Motherwell[0.042]; Stirling[0.042]; Raith[0.041]; Falkirk[0.041]; Greenock[0.041]; Fife[0.041]; Stranraer[0.041]; Johnstone[0.040]; Clydebank[0.040]; United[0.040]; Division[0.039]; Division[0.039]; Ayr[0.039]; Hamilton[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Arbroath ==> ENTITY: \u001b[32mArbroath F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.158:0.158[0]; log p(e|m)= -0.860:-0.860[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCowdenbeath[0.063]; Inverness[0.062]; Brechin[0.062]; Dumbarton[0.062]; Ross[0.061]; Alloa[0.061]; Forfar[0.061]; Stirling[0.060]; Park[0.059]; Division[0.059]; Thistle[0.058]; Albion[0.057]; Montrose[0.056]; East[0.056]; Berwick[0.055]; Queen[0.055]; County[0.052]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = East Stirling\u001b[39m ==> ENTITY: \u001b[32mEast Stirlingshire F.C.\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCowdenbeath[0.075]; Brechin[0.070]; Inverness[0.070]; Thistle[0.069]; Park[0.069]; Alloa[0.069]; Forfar[0.068]; Division[0.067]; Albion[0.067]; Arbroath[0.066]; Ross[0.065]; Montrose[0.064]; Berwick[0.062]; Queen[0.062]; County[0.056]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Airdrieonians\u001b[39m ==> ENTITY: \u001b[32mAirdrieonians F.C.\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAberdeen[0.045]; Livingston[0.044]; Dundee[0.043]; Dundee[0.043]; Dunfermline[0.043]; Kilmarnock[0.043]; Rangers[0.043]; Partick[0.042]; Stirling[0.042]; Hibernian[0.042]; Motherwell[0.042]; Clydebank[0.041]; Celtic[0.041]; Greenock[0.041]; Falkirk[0.041]; Fife[0.041]; Hearts[0.041]; Johnstone[0.041]; Raith[0.040]; Division[0.040]; Division[0.040]; Clyde[0.040]; Ayr[0.040]; Morton[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dunfermline ==> ENTITY: \u001b[32mDunfermline Athletic F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.236:0.236[0]; log p(e|m)= -1.995:-1.995[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.045]; Aberdeen[0.045]; Dundee[0.044]; Dundee[0.044]; League[0.042]; played[0.042]; Kilmarnock[0.042]; matches[0.042]; Celtic[0.042]; league[0.042]; Rangers[0.042]; Scottish[0.042]; Scottish[0.042]; Glasgow[0.041]; Hibernian[0.041]; Motherwell[0.041]; Falkirk[0.041]; Hearts[0.041]; Partick[0.041]; Stirling[0.040]; Raith[0.040]; Soccer[0.040]; Clydebank[0.040]; Greenock[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Inverness Thistle\u001b[39m ==> ENTITY: \u001b[32mInverness Thistle F.C.\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.132:0.132[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLivingston[0.048]; Stenhousemuir[0.047]; Cowdenbeath[0.047]; Stirling[0.046]; Ross[0.046]; Park[0.045]; Alloa[0.045]; Dumbarton[0.044]; Ayr[0.044]; Stranraer[0.044]; Forfar[0.044]; Brechin[0.044]; Arbroath[0.043]; Albion[0.043]; Clyde[0.043]; Hamilton[0.043]; Berwick[0.042]; Montrose[0.042]; Division[0.042]; Division[0.042]; South[0.040]; East[0.040]; County[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hearts ==> ENTITY: \u001b[32mHeart of Midlothian F.C.\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.222:0.222[0]; log p(e|m)= -0.298:-0.298[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; League[0.043]; Aberdeen[0.043]; Dundee[0.043]; Dundee[0.043]; matches[0.043]; league[0.043]; Glasgow[0.042]; Hibernian[0.042]; Rangers[0.042]; Celtic[0.042]; played[0.042]; Scottish[0.042]; Scottish[0.042]; Dunfermline[0.041]; Motherwell[0.041]; Kilmarnock[0.041]; won[0.041]; Falkirk[0.040]; Soccer[0.040]; Raith[0.040]; Partick[0.039]; Premier[0.039]; Standings[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = East Fife\u001b[39m ==> ENTITY: \u001b[32mEast Fife F.C.\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; Stenhousemuir[0.045]; Dundee[0.044]; Kilmarnock[0.044]; Dumbarton[0.043]; Partick[0.043]; Brechin[0.043]; Clydebank[0.043]; Stirling[0.042]; Livingston[0.042]; Stranraer[0.042]; Greenock[0.042]; Falkirk[0.042]; Raith[0.042]; Division[0.041]; Division[0.041]; Division[0.041]; Johnstone[0.041]; Ayr[0.041]; Clyde[0.040]; Hamilton[0.039]; Berwick[0.038]; Morton[0.038]; Montrose[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ross County\u001b[39m ==> ENTITY: \u001b[32mRoss County F.C.\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLivingston[0.047]; Division[0.047]; Inverness[0.046]; Cowdenbeath[0.046]; Stirling[0.045]; Hamilton[0.045]; Stenhousemuir[0.045]; Park[0.044]; Clyde[0.044]; Alloa[0.044]; South[0.043]; Dumbarton[0.043]; Thistle[0.043]; Stranraer[0.043]; Brechin[0.043]; Albion[0.043]; Arbroath[0.043]; Forfar[0.043]; Montrose[0.043]; East[0.042]; Berwick[0.041]; Queen[0.039]; Queen[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scottish ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.129:0.129[0]; log p(e|m)= -0.480:-0.480[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScottish[0.045]; League[0.043]; Aberdeen[0.043]; Hibernian[0.043]; matches[0.043]; Celtic[0.043]; Dundee[0.043]; league[0.043]; Dunfermline[0.043]; Rangers[0.043]; played[0.043]; Hearts[0.042]; Kilmarnock[0.042]; Glasgow[0.042]; division[0.041]; Motherwell[0.041]; Soccer[0.041]; Raith[0.041]; won[0.041]; Division[0.040]; Premier[0.039]; United[0.039]; goals[0.038]; goals[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dumbarton ==> ENTITY: \u001b[32mDumbarton F.C.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.189:0.189[0]; log p(e|m)= -1.255:-1.255[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStenhousemuir[0.046]; Fife[0.045]; Cowdenbeath[0.045]; Stranraer[0.044]; Arbroath[0.043]; Greenock[0.043]; Division[0.043]; Division[0.043]; Alloa[0.043]; Brechin[0.042]; Inverness[0.042]; Clyde[0.042]; Livingston[0.041]; Stirling[0.041]; Forfar[0.041]; Clydebank[0.041]; Thistle[0.041]; Park[0.040]; Ross[0.040]; Hamilton[0.039]; Ayr[0.039]; County[0.039]; Queen[0.039]; Queen[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Greenock Morton\u001b[39m ==> ENTITY: \u001b[32mGreenock Morton F.C.\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; Stenhousemuir[0.044]; Partick[0.044]; Dunfermline[0.043]; Kilmarnock[0.043]; Livingston[0.043]; Dundee[0.043]; Dundee[0.043]; Fife[0.042]; Motherwell[0.042]; Stranraer[0.042]; Dumbarton[0.042]; Clydebank[0.042]; Raith[0.042]; Stirling[0.042]; Falkirk[0.041]; Johnstone[0.041]; Division[0.040]; Division[0.040]; Hamilton[0.040]; Clyde[0.040]; Ayr[0.039]; United[0.038]; East[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kilmarnock ==> ENTITY: \u001b[32mKilmarnock F.C.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.244:0.244[0]; log p(e|m)= -1.431:-1.431[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.045]; Aberdeen[0.044]; Dundee[0.043]; Dundee[0.043]; played[0.043]; matches[0.042]; Celtic[0.042]; Dunfermline[0.042]; Glasgow[0.042]; Hibernian[0.042]; Scottish[0.042]; Scottish[0.042]; League[0.042]; Motherwell[0.041]; Falkirk[0.041]; league[0.041]; Raith[0.041]; Partick[0.041]; Greenock[0.041]; Stirling[0.041]; Rangers[0.040]; Fife[0.040]; Hearts[0.040]; Clydebank[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hibernian\u001b[39m ==> ENTITY: \u001b[32mHibernian F.C.\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.045]; Aberdeen[0.044]; Dunfermline[0.043]; Partick[0.043]; Dundee[0.043]; Dundee[0.043]; matches[0.043]; Celtic[0.043]; Glasgow[0.042]; Kilmarnock[0.042]; League[0.042]; Scottish[0.042]; Scottish[0.042]; Falkirk[0.041]; Motherwell[0.041]; Rangers[0.041]; played[0.041]; Hearts[0.041]; league[0.041]; won[0.040]; Raith[0.040]; Soccer[0.040]; Johnstone[0.039]; division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clyde ==> ENTITY: \u001b[32mClyde F.C.\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.194:0.194[0]; log p(e|m)= -0.983:-0.983[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.047]; County[0.044]; Livingston[0.044]; Dundee[0.043]; Stenhousemuir[0.043]; Cowdenbeath[0.043]; Stirling[0.043]; Partick[0.042]; Greenock[0.042]; Fife[0.042]; Thistle[0.042]; Brechin[0.042]; Dumbarton[0.042]; Inverness[0.041]; Alloa[0.041]; Ross[0.041]; Clydebank[0.041]; Stranraer[0.041]; Montrose[0.040]; Ayr[0.040]; South[0.039]; Division[0.039]; Division[0.039]; Hamilton[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAlloa\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAlloa\u001b[39m <---> \u001b[32mAlloa Athletic F.C.\u001b[39m\t\nSCORES: global= 0.289:0.284[\u001b[31m0.005\u001b[39m]; local()= 0.156:0.174[\u001b[32m0.018\u001b[39m]; log p(e|m)= 0.000:-0.794[\u001b[31m0.794\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCowdenbeath[0.048]; Stenhousemuir[0.048]; Stirling[0.048]; Brechin[0.046]; Inverness[0.046]; Thistle[0.045]; Arbroath[0.045]; Dumbarton[0.045]; Stranraer[0.045]; Forfar[0.045]; Park[0.044]; Ross[0.044]; Albion[0.044]; Clyde[0.044]; Hamilton[0.043]; Division[0.043]; Montrose[0.042]; East[0.041]; Berwick[0.040]; South[0.039]; Queen[0.039]; Queen[0.039]; County[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cowdenbeath ==> ENTITY: \u001b[32mCowdenbeath F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.176:0.176[0]; log p(e|m)= -0.247:-0.247[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStenhousemuir[0.052]; Brechin[0.048]; Inverness[0.048]; Dumbarton[0.047]; Stranraer[0.047]; Stirling[0.047]; Division[0.047]; Thistle[0.046]; Park[0.046]; Alloa[0.046]; Ross[0.046]; Arbroath[0.046]; Forfar[0.046]; Albion[0.045]; Clyde[0.045]; East[0.044]; Berwick[0.044]; Montrose[0.043]; South[0.043]; County[0.042]; Queen[0.040]; Queen[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAyr\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAyr\u001b[39m <---> \u001b[32mAyr United F.C.\u001b[39m\t\nSCORES: global= 0.266:0.256[\u001b[31m0.010\u001b[39m]; local()= 0.174:0.223[\u001b[32m0.049\u001b[39m]; log p(e|m)= -0.066:-3.352[\u001b[31m3.286\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.044]; Stenhousemuir[0.044]; Thistle[0.044]; Dundee[0.044]; Fife[0.043]; Kilmarnock[0.043]; Partick[0.043]; Livingston[0.043]; Greenock[0.043]; Inverness[0.043]; Clydebank[0.042]; Stirling[0.042]; Dumbarton[0.042]; Stranraer[0.042]; Falkirk[0.042]; Brechin[0.042]; Berwick[0.040]; Johnstone[0.040]; Clyde[0.040]; Division[0.039]; Division[0.039]; Division[0.039]; Montrose[0.039]; Hamilton[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Falkirk ==> ENTITY: \u001b[32mFalkirk F.C.\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.210:0.210[0]; log p(e|m)= -1.204:-1.204[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; Dunfermline[0.045]; Fife[0.044]; Ayr[0.043]; Motherwell[0.043]; Kilmarnock[0.042]; Division[0.042]; Division[0.042]; Premier[0.042]; Dundee[0.042]; Dundee[0.042]; Partick[0.042]; Clydebank[0.042]; Aberdeen[0.041]; Livingston[0.041]; Hibernian[0.041]; Stirling[0.041]; Raith[0.041]; Hearts[0.041]; Celtic[0.041]; Rangers[0.040]; Greenock[0.040]; Johnstone[0.038]; United[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Motherwell ==> ENTITY: \u001b[32mMotherwell F.C.\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.224:0.224[0]; log p(e|m)= -0.335:-0.335[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.045]; Scottish[0.043]; Scottish[0.043]; Aberdeen[0.043]; Dundee[0.043]; Dundee[0.043]; Premier[0.042]; Glasgow[0.042]; Hibernian[0.042]; League[0.042]; Kilmarnock[0.042]; Dunfermline[0.042]; league[0.042]; matches[0.042]; played[0.042]; Celtic[0.041]; Rangers[0.041]; Falkirk[0.041]; Hearts[0.041]; Division[0.040]; Partick[0.040]; Greenock[0.040]; division[0.039]; Raith[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brechin ==> ENTITY: \u001b[32mBrechin City F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.182:0.182[0]; log p(e|m)= -1.917:-1.917[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nForfar[0.046]; Ayr[0.046]; Stenhousemuir[0.045]; Arbroath[0.044]; Cowdenbeath[0.044]; Alloa[0.043]; Dumbarton[0.043]; Stranraer[0.043]; Inverness[0.043]; Thistle[0.042]; Livingston[0.042]; Fife[0.042]; Stirling[0.041]; Stirling[0.041]; Clydebank[0.041]; Park[0.041]; Division[0.041]; Division[0.041]; Albion[0.039]; Berwick[0.039]; Ross[0.039]; Clyde[0.039]; Hamilton[0.038]; Montrose[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scottish ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.126:0.126[0]; log p(e|m)= -0.480:-0.480[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScottish[0.045]; League[0.044]; Aberdeen[0.043]; Hibernian[0.043]; matches[0.043]; Celtic[0.043]; Dundee[0.043]; league[0.043]; Dunfermline[0.043]; Rangers[0.043]; played[0.043]; Hearts[0.043]; Kilmarnock[0.042]; Glasgow[0.042]; division[0.041]; Motherwell[0.041]; Soccer[0.041]; Raith[0.041]; won[0.041]; Premier[0.039]; United[0.039]; goals[0.038]; goals[0.038]; Standings[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rangers ==> ENTITY: \u001b[32mRangers F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.211:0.211[0]; log p(e|m)= -0.302:-0.302[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.047]; league[0.044]; matches[0.044]; League[0.043]; Aberdeen[0.043]; Glasgow[0.043]; Scottish[0.043]; Scottish[0.043]; Dundee[0.043]; Celtic[0.042]; played[0.042]; Dunfermline[0.042]; Hearts[0.042]; Hibernian[0.041]; Soccer[0.041]; Kilmarnock[0.041]; won[0.041]; Motherwell[0.040]; Raith[0.040]; Falkirk[0.040]; Saturday[0.039]; United[0.039]; Premier[0.038]; division[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clydebank ==> ENTITY: \u001b[32mClydebank F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.221:0.221[0]; log p(e|m)= -1.328:-1.328[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.047]; Dundee[0.044]; Stenhousemuir[0.044]; Greenock[0.044]; Dunfermline[0.044]; Fife[0.043]; Stirling[0.043]; Kilmarnock[0.042]; Brechin[0.042]; Partick[0.042]; Dumbarton[0.042]; Motherwell[0.042]; Livingston[0.042]; Stranraer[0.042]; Johnstone[0.041]; Falkirk[0.041]; Clyde[0.041]; Ayr[0.040]; Raith[0.040]; Division[0.040]; Division[0.040]; Hamilton[0.039]; Queen[0.038]; Morton[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Stenhousemuir\u001b[39m ==> ENTITY: \u001b[32mStenhousemuir F.C.\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStirling[0.044]; Cowdenbeath[0.044]; Dumbarton[0.044]; Fife[0.043]; Livingston[0.043]; Inverness[0.043]; Thistle[0.043]; Alloa[0.043]; Division[0.042]; Division[0.042]; Partick[0.042]; Greenock[0.042]; Brechin[0.042]; Stranraer[0.042]; Forfar[0.042]; Ross[0.042]; Clydebank[0.041]; Berwick[0.041]; Albion[0.041]; Ayr[0.041]; Clyde[0.039]; Morton[0.039]; Montrose[0.039]; Hamilton[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Partick ==> ENTITY: \u001b[32mPartick Thistle F.C.\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.228:0.228[0]; log p(e|m)= -1.284:-1.284[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; Stenhousemuir[0.044]; Livingston[0.044]; Dundee[0.043]; Dundee[0.043]; Kilmarnock[0.043]; Aberdeen[0.043]; Motherwell[0.043]; Dunfermline[0.043]; Hearts[0.042]; Falkirk[0.042]; Greenock[0.042]; Hibernian[0.042]; Stirling[0.041]; Clyde[0.041]; Raith[0.041]; Fife[0.041]; Johnstone[0.041]; Clydebank[0.040]; Division[0.039]; Division[0.039]; Hamilton[0.039]; Ayr[0.038]; Morton[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Raith\u001b[39m ==> ENTITY: \u001b[32mRaith Rovers F.C.\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.252:0.252[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.044]; Dundee[0.044]; Dundee[0.044]; league[0.043]; Dunfermline[0.043]; matches[0.043]; Aberdeen[0.042]; Kilmarnock[0.042]; Fife[0.042]; Rangers[0.042]; Glasgow[0.041]; Scottish[0.041]; Scottish[0.041]; Partick[0.041]; Celtic[0.041]; Greenock[0.041]; Motherwell[0.041]; League[0.041]; Clydebank[0.041]; Falkirk[0.041]; Stirling[0.041]; played[0.041]; Hibernian[0.040]; Soccer[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Forfar ==> ENTITY: \u001b[32mForfar Athletic F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.195:0.195[0]; log p(e|m)= -1.766:-1.766[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStenhousemuir[0.056]; Cowdenbeath[0.055]; Inverness[0.053]; Arbroath[0.053]; Brechin[0.052]; Alloa[0.052]; Stranraer[0.052]; Stirling[0.052]; Thistle[0.052]; Dumbarton[0.051]; Park[0.050]; Division[0.050]; Ross[0.049]; Montrose[0.048]; Queen[0.048]; Berwick[0.047]; Albion[0.047]; County[0.046]; South[0.044]; East[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Livingston ==> ENTITY: \u001b[32mLivingston F.C.\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.179:0.179[0]; log p(e|m)= -1.864:-1.864[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.047]; Fife[0.045]; Inverness[0.044]; Dundee[0.043]; Stenhousemuir[0.043]; Stirling[0.043]; Greenock[0.043]; County[0.043]; Dumbarton[0.042]; Falkirk[0.042]; Brechin[0.042]; Thistle[0.042]; Ross[0.041]; Partick[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Stranraer[0.041]; Clydebank[0.040]; Clyde[0.040]; Hamilton[0.040]; East[0.039]; Johnstone[0.039]; South[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Celtic ==> ENTITY: \u001b[32mCeltic F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.191:0.191[0]; log p(e|m)= -0.978:-0.978[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGlasgow[0.044]; Dunfermline[0.043]; Airdrieonians[0.043]; matches[0.043]; Scottish[0.043]; Scottish[0.043]; Dundee[0.043]; Dundee[0.043]; Aberdeen[0.043]; League[0.043]; Kilmarnock[0.042]; Falkirk[0.042]; league[0.042]; Rangers[0.042]; played[0.041]; Hibernian[0.041]; Soccer[0.041]; Motherwell[0.041]; won[0.040]; Hearts[0.040]; Raith[0.039]; Division[0.039]; Johnstone[0.039]; division[0.039]; \t\n\r [=======>......................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 16s11ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 873/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1343testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Asian ==> ENTITY: \u001b[32mAsia\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.113:0.113[0]; log p(e|m)= -1.959:-1.959[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAsian[0.052]; Asia[0.051]; country[0.049]; launches[0.048]; Malaysia[0.047]; launched[0.046]; launched[0.046]; Friday[0.046]; today[0.045]; format[0.045]; format[0.045]; second[0.044]; second[0.044]; opening[0.044]; group[0.044]; service[0.044]; said[0.044]; city[0.044]; recently[0.044]; store[0.043]; store[0.043]; complement[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kuala Lumpur\u001b[39m ==> ENTITY: \u001b[32mKuala Lumpur\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMalaysia[0.052]; Johor[0.046]; service[0.045]; country[0.044]; Bahru[0.044]; Asia[0.044]; Asia[0.044]; Asian[0.044]; Asian[0.044]; today[0.044]; second[0.044]; second[0.044]; opening[0.043]; plans[0.043]; Friday[0.042]; city[0.042]; announced[0.042]; region[0.042]; open[0.042]; launches[0.042]; recently[0.042]; recently[0.042]; supermarket[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ahold\u001b[39m ==> ENTITY: \u001b[32mAhold\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAhold[0.048]; Ahold[0.048]; Ahold[0.048]; Ahold[0.048]; retail[0.045]; stores[0.043]; Europe[0.042]; Bahru[0.042]; company[0.042]; Dutch[0.041]; announced[0.041]; supermarket[0.041]; plans[0.041]; sales[0.040]; Tops[0.040]; Tops[0.040]; Lumpur[0.039]; store[0.039]; worldwide[0.039]; launched[0.039]; employs[0.039]; recently[0.038]; recently[0.038]; service[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dutch ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.104:0.104[0]; log p(e|m)= -0.794:-0.794[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurope[0.046]; country[0.045]; region[0.043]; worldwide[0.043]; Amsterdam[0.043]; Brazil[0.043]; company[0.042]; Tops[0.042]; Tops[0.042]; format[0.041]; recently[0.041]; recently[0.041]; Asian[0.041]; Asia[0.041]; Asia[0.041]; today[0.041]; city[0.041]; second[0.041]; service[0.041]; Malaysia[0.041]; located[0.040]; approximately[0.040]; formats[0.040]; people[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mTOPS\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mTops Supermarket\u001b[39m <---> \u001b[32mTops Friendly Markets\u001b[39m\t\nSCORES: global= 0.262:0.260[\u001b[31m0.002\u001b[39m]; local()= 0.195:0.166[\u001b[31m0.030\u001b[39m]; log p(e|m)= -2.303:-2.198[\u001b[32m0.104\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstores[0.044]; stores[0.044]; stores[0.044]; retail[0.043]; group[0.043]; group[0.043]; Tops[0.042]; announced[0.042]; store[0.042]; store[0.042]; supermarket[0.041]; opening[0.041]; Ahold[0.041]; Ahold[0.041]; Ahold[0.041]; Ahold[0.041]; Ahold[0.041]; Ahold[0.041]; Bahru[0.041]; plans[0.041]; located[0.040]; service[0.040]; today[0.040]; sales[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Johor Bahru\u001b[39m ==> ENTITY: \u001b[32mJohor Bahru\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKuala[0.049]; Malaysia[0.049]; city[0.047]; service[0.045]; Lumpur[0.044]; plans[0.044]; located[0.044]; region[0.044]; Asian[0.043]; Asian[0.043]; open[0.043]; second[0.043]; second[0.043]; launches[0.043]; today[0.042]; announced[0.042]; opening[0.042]; Asia[0.042]; Asia[0.042]; launched[0.042]; launched[0.042]; capital[0.042]; agreement[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Asian ==> ENTITY: \u001b[32mAsia\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.134:0.134[0]; log p(e|m)= -1.959:-1.959[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAsian[0.047]; Asia[0.046]; Asia[0.046]; country[0.044]; launches[0.043]; region[0.043]; Malaysia[0.042]; launched[0.042]; launched[0.042]; Friday[0.042]; today[0.041]; format[0.041]; format[0.041]; household[0.040]; second[0.040]; second[0.040]; opening[0.040]; group[0.040]; service[0.040]; said[0.040]; said[0.040]; city[0.040]; recently[0.039]; coming[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Europe ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.101:0.101[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nregion[0.044]; country[0.044]; discount[0.044]; discount[0.044]; sales[0.043]; stores[0.043]; stores[0.043]; stores[0.043]; retail[0.042]; plans[0.042]; announced[0.042]; Asia[0.042]; Asia[0.042]; company[0.041]; worldwide[0.041]; store[0.041]; recently[0.040]; recently[0.040]; second[0.040]; household[0.040]; price[0.040]; billion[0.040]; Brazil[0.040]; launched[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ahold\u001b[39m ==> ENTITY: \u001b[32mAhold\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAhold[0.049]; Ahold[0.049]; Ahold[0.049]; Ahold[0.049]; retail[0.046]; stores[0.044]; stores[0.044]; stores[0.044]; Europe[0.043]; Bahru[0.043]; company[0.043]; Dutch[0.043]; announced[0.042]; supermarket[0.042]; plans[0.042]; sales[0.041]; today[0.041]; Tops[0.041]; Tops[0.041]; Lumpur[0.040]; store[0.040]; store[0.040]; worldwide[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Amsterdam ==> ENTITY: \u001b[32mAmsterdam\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.103:0.103[0]; log p(e|m)= -0.064:-0.064[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nservice[0.046]; Europe[0.046]; Dutch[0.045]; region[0.045]; Asia[0.042]; Asia[0.042]; worldwide[0.042]; recently[0.042]; recently[0.042]; plans[0.042]; newsroom[0.041]; Brazil[0.041]; people[0.041]; group[0.040]; open[0.040]; launched[0.040]; announced[0.040]; Ahold[0.040]; Ahold[0.040]; Ahold[0.040]; Ahold[0.040]; Ahold[0.040]; said[0.040]; coming[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ahold\u001b[39m ==> ENTITY: \u001b[32mAhold\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.249:0.249[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAhold[0.048]; Ahold[0.048]; Ahold[0.048]; Ahold[0.048]; Ahold[0.048]; Ahold[0.048]; retail[0.045]; stores[0.043]; stores[0.043]; stores[0.043]; stores[0.043]; Europe[0.042]; supermarkets[0.042]; Bahru[0.042]; company[0.042]; Dutch[0.042]; Dutch[0.042]; announced[0.041]; supermarket[0.041]; plans[0.041]; Netherlands[0.041]; sales[0.040]; today[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ahold\u001b[39m ==> ENTITY: \u001b[32mAhold\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAhold[0.049]; Ahold[0.049]; stores[0.044]; stores[0.044]; stores[0.044]; supermarkets[0.043]; Bahru[0.043]; Dutch[0.043]; supermarket[0.042]; Netherlands[0.042]; today[0.041]; Tops[0.041]; Lumpur[0.040]; store[0.040]; store[0.040]; launches[0.040]; launched[0.040]; recently[0.039]; Nv[0.039]; service[0.039]; Kuala[0.039]; opening[0.039]; consumers[0.039]; Zaandam[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Netherlands ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.096:0.096[0]; log p(e|m)= -0.250:-0.250[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDutch[0.048]; country[0.046]; Malaysia[0.044]; located[0.044]; second[0.043]; second[0.043]; today[0.043]; opening[0.042]; group[0.042]; recently[0.042]; format[0.041]; format[0.041]; Asian[0.041]; Asian[0.041]; open[0.041]; Asia[0.041]; launches[0.041]; service[0.040]; Zaandam[0.040]; city[0.040]; Friday[0.040]; plans[0.040]; launched[0.040]; launched[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Asia ==> ENTITY: \u001b[32mAsia\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.142:0.142[0]; log p(e|m)= -0.150:-0.150[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurope[0.047]; Asia[0.047]; Asian[0.046]; country[0.046]; region[0.043]; Malaysia[0.043]; worldwide[0.043]; recently[0.043]; recently[0.043]; service[0.042]; Brazil[0.041]; billion[0.041]; statement[0.041]; presence[0.041]; city[0.040]; group[0.040]; second[0.040]; sales[0.040]; format[0.039]; format[0.039]; said[0.039]; launched[0.039]; people[0.039]; today[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mTOPS\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mTops Supermarket\u001b[39m <---> \u001b[32mTops Friendly Markets\u001b[39m\t\nSCORES: global= 0.264:0.262[\u001b[31m0.002\u001b[39m]; local()= 0.210:0.177[\u001b[31m0.033\u001b[39m]; log p(e|m)= -2.303:-2.198[\u001b[32m0.104\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstores[0.044]; stores[0.044]; stores[0.044]; stores[0.044]; retail[0.043]; group[0.043]; group[0.043]; Tops[0.042]; announced[0.042]; store[0.042]; store[0.042]; supermarkets[0.041]; supermarket[0.041]; opening[0.041]; Ahold[0.041]; Ahold[0.041]; Ahold[0.041]; Ahold[0.041]; Ahold[0.041]; Ahold[0.041]; Ahold[0.041]; Bahru[0.041]; plans[0.040]; located[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Malaysia ==> ENTITY: \u001b[32mMalaysia\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.164:0.164[0]; log p(e|m)= -0.202:-0.202[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJohor[0.048]; country[0.047]; Asian[0.047]; Asian[0.047]; Asia[0.046]; Asia[0.046]; Kuala[0.045]; Netherlands[0.044]; region[0.044]; Dutch[0.043]; Dutch[0.043]; Bahru[0.043]; opening[0.043]; second[0.043]; second[0.043]; Friday[0.042]; service[0.042]; Europe[0.042]; recently[0.041]; recently[0.041]; making[0.041]; Lumpur[0.041]; today[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazil ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.046]; region[0.044]; city[0.044]; Europe[0.043]; agreement[0.043]; Asia[0.042]; Asia[0.042]; presence[0.042]; worldwide[0.042]; names[0.042]; second[0.042]; coming[0.041]; activities[0.041]; household[0.041]; strong[0.041]; billion[0.041]; said[0.040]; joint[0.040]; Dutch[0.040]; statement[0.040]; service[0.040]; recently[0.040]; recently[0.040]; group[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ahold\u001b[39m ==> ENTITY: \u001b[32mAhold\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAhold[0.049]; Ahold[0.049]; Ahold[0.049]; Ahold[0.049]; retail[0.046]; stores[0.045]; stores[0.045]; stores[0.045]; Europe[0.044]; Bahru[0.043]; company[0.043]; Dutch[0.043]; announced[0.043]; supermarket[0.043]; plans[0.042]; sales[0.042]; Tops[0.041]; Tops[0.041]; Lumpur[0.041]; store[0.040]; worldwide[0.040]; launched[0.040]; employs[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zaandam\u001b[39m ==> ENTITY: \u001b[32mZaandam\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.093:0.093[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.045]; Dutch[0.044]; Bahru[0.043]; city[0.043]; open[0.043]; located[0.043]; supermarket[0.042]; service[0.042]; store[0.042]; store[0.042]; today[0.042]; Friday[0.041]; opening[0.041]; supermarkets[0.041]; Lumpur[0.041]; country[0.041]; stores[0.041]; stores[0.041]; stores[0.041]; Ahold[0.040]; Ahold[0.040]; Ahold[0.040]; Ahold[0.040]; Nv[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Asia ==> ENTITY: \u001b[32mAsia\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.157:0.157[0]; log p(e|m)= -0.150:-0.150[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurope[0.046]; Asia[0.046]; Asian[0.045]; Asian[0.045]; country[0.045]; region[0.043]; Malaysia[0.042]; worldwide[0.042]; service[0.042]; recently[0.042]; recently[0.042]; second[0.041]; second[0.041]; Brazil[0.040]; approximately[0.040]; presence[0.040]; launched[0.040]; launched[0.040]; city[0.040]; statement[0.040]; sales[0.039]; group[0.039]; group[0.039]; Zaandam[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = US$\u001b[39m ==> ENTITY: \u001b[32mUnited States dollar\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.066:0.066[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.045]; Bahru[0.044]; price[0.044]; worldwide[0.044]; billion[0.044]; Europe[0.043]; approximately[0.043]; recently[0.042]; recently[0.042]; company[0.041]; retail[0.041]; Ahold[0.041]; Ahold[0.041]; Ahold[0.041]; Ahold[0.041]; Ahold[0.041]; region[0.041]; Asia[0.041]; Asia[0.041]; service[0.041]; said[0.040]; launched[0.040]; sales[0.040]; formats[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dutch ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.794:-0.794[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.052]; country[0.046]; Zaandam[0.043]; recently[0.042]; Asian[0.042]; Asian[0.042]; Asia[0.042]; today[0.042]; Tops[0.042]; city[0.041]; second[0.041]; second[0.041]; service[0.041]; Malaysia[0.041]; located[0.041]; store[0.040]; store[0.040]; launches[0.040]; said[0.040]; Ahold[0.040]; Ahold[0.040]; Ahold[0.040]; Ahold[0.040]; Friday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.055:0.055[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.045]; format[0.045]; recently[0.043]; recently[0.043]; Tops[0.043]; Tops[0.043]; sales[0.043]; worldwide[0.042]; billion[0.042]; formats[0.042]; Malaysia[0.042]; second[0.042]; Europe[0.042]; announced[0.041]; coming[0.041]; statement[0.041]; Brazil[0.040]; group[0.040]; price[0.040]; launched[0.040]; open[0.040]; company[0.040]; names[0.040]; strong[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ahold\u001b[39m ==> ENTITY: \u001b[32mAhold\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.240:0.240[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAhold[0.046]; Ahold[0.046]; Ahold[0.046]; Ahold[0.046]; Ahold[0.046]; retail[0.043]; stores[0.042]; stores[0.042]; stores[0.042]; stores[0.042]; Europe[0.041]; supermarkets[0.041]; Bahru[0.041]; company[0.040]; Dutch[0.040]; Dutch[0.040]; announced[0.040]; supermarket[0.040]; plans[0.040]; Netherlands[0.039]; sales[0.039]; today[0.039]; Tops[0.039]; Tops[0.039]; \t\n\r [========>.....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s907ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 897/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1379testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hobart ==> ENTITY: \u001b[32mHobart\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.390:-0.390[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOval[0.046]; match[0.045]; Australia[0.044]; cricket[0.044]; Cricket[0.044]; play[0.043]; Victoria[0.043]; Victoria[0.043]; Bellerive[0.042]; Tasmania[0.042]; Tasmania[0.042]; second[0.042]; Shield[0.041]; Shield[0.041]; Saturday[0.041]; Dean[0.041]; David[0.041]; declared[0.040]; day[0.040]; day[0.040]; Michael[0.039]; Shaun[0.039]; Young[0.038]; score[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tasmania ==> ENTITY: \u001b[32mTasmania cricket team\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.188:0.188[0]; log p(e|m)= -2.430:-2.430[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.047]; Australia[0.046]; Oval[0.045]; Tasmania[0.044]; Victoria[0.044]; Victoria[0.044]; match[0.044]; Cricket[0.043]; Hobart[0.042]; play[0.042]; day[0.042]; day[0.042]; second[0.041]; score[0.041]; David[0.040]; Bellerive[0.040]; Boon[0.040]; Shaun[0.040]; Shield[0.040]; Shield[0.040]; Saturday[0.039]; Dean[0.039]; Michael[0.039]; declared[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Michael DiVenuto\u001b[39m ==> ENTITY: \u001b[32mMichael Di Venuto\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.047]; Oval[0.046]; Australia[0.046]; Cricket[0.046]; day[0.046]; day[0.046]; score[0.045]; match[0.045]; Shaun[0.044]; second[0.043]; Tasmania[0.043]; Tasmania[0.043]; David[0.043]; Boon[0.043]; Victoria[0.042]; Victoria[0.042]; declared[0.042]; Hobart[0.042]; play[0.042]; Dean[0.042]; Saturday[0.041]; Sheffield[0.041]; Sheffield[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Victoria ==> ENTITY: \u001b[32mVictoria cricket team\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.181:0.181[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVictoria[0.047]; Oval[0.047]; cricket[0.047]; Australia[0.047]; Cricket[0.046]; match[0.046]; Tasmania[0.045]; Tasmania[0.045]; play[0.044]; second[0.044]; Shield[0.043]; Shield[0.043]; Hobart[0.043]; Dean[0.042]; David[0.042]; Michael[0.041]; Young[0.041]; score[0.041]; Shaun[0.041]; Bellerive[0.041]; Sheffield[0.041]; Sheffield[0.041]; Saturday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = David Boon\u001b[39m ==> ENTITY: \u001b[32mDavid Boon\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.048]; Australia[0.047]; Oval[0.045]; Cricket[0.045]; match[0.044]; Hobart[0.043]; Tasmania[0.042]; Tasmania[0.042]; second[0.042]; Michael[0.042]; Dean[0.041]; play[0.041]; Shaun[0.041]; Jones[0.041]; Victoria[0.041]; Victoria[0.041]; Shield[0.040]; Shield[0.040]; day[0.040]; day[0.040]; Sheffield[0.039]; Sheffield[0.039]; Bellerive[0.039]; Young[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sheffield Shield\u001b[39m ==> ENTITY: \u001b[32mSheffield Shield\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; cricket[0.046]; Shield[0.045]; Oval[0.044]; Tasmania[0.044]; Tasmania[0.044]; match[0.044]; Cricket[0.043]; Victoria[0.042]; Victoria[0.042]; Sheffield[0.042]; Hobart[0.041]; play[0.041]; score[0.041]; day[0.041]; day[0.041]; second[0.040]; Saturday[0.040]; Score[0.040]; Bellerive[0.039]; Boon[0.038]; Shaun[0.038]; Dean[0.038]; Michael[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sheffield Shield\u001b[39m ==> ENTITY: \u001b[32mSheffield Shield\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; cricket[0.046]; Shield[0.045]; Oval[0.044]; Tasmania[0.044]; Tasmania[0.044]; match[0.044]; Cricket[0.043]; Victoria[0.042]; Victoria[0.042]; Sheffield[0.042]; Hobart[0.041]; play[0.041]; score[0.041]; day[0.041]; day[0.041]; second[0.040]; Saturday[0.040]; Score[0.040]; Bellerive[0.039]; Boon[0.038]; Shaun[0.038]; Dean[0.038]; Michael[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shaun Young\u001b[39m ==> ENTITY: \u001b[32mShaun Young\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.047]; Oval[0.045]; Cricket[0.044]; Australia[0.044]; Shield[0.043]; Shield[0.043]; Tasmania[0.042]; Tasmania[0.042]; match[0.042]; Dean[0.042]; day[0.041]; day[0.041]; play[0.041]; Michael[0.041]; Victoria[0.041]; Victoria[0.041]; Sheffield[0.041]; Sheffield[0.041]; Hobart[0.041]; David[0.041]; Boon[0.040]; Jones[0.039]; second[0.038]; Saturday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bellerive Oval\u001b[39m ==> ENTITY: \u001b[32mBellerive Oval\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.046]; cricket[0.046]; Australia[0.045]; Cricket[0.045]; second[0.043]; Hobart[0.043]; Victoria[0.043]; Victoria[0.043]; Tasmania[0.043]; Tasmania[0.043]; Saturday[0.043]; play[0.043]; Shield[0.041]; Shield[0.041]; day[0.041]; day[0.041]; David[0.040]; score[0.040]; Michael[0.039]; Shaun[0.039]; Dean[0.038]; Young[0.038]; Sheffield[0.038]; Sheffield[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.047]; match[0.045]; Cricket[0.045]; Oval[0.045]; Victoria[0.044]; Victoria[0.044]; Tasmania[0.043]; Tasmania[0.043]; second[0.042]; day[0.042]; day[0.042]; Shield[0.041]; Shield[0.041]; play[0.041]; Shaun[0.040]; David[0.040]; Hobart[0.040]; Bellerive[0.040]; Boon[0.040]; Sheffield[0.040]; Sheffield[0.040]; Saturday[0.039]; Jones[0.039]; Young[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Victoria ==> ENTITY: \u001b[32mVictoria cricket team\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.181:0.181[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVictoria[0.047]; Oval[0.047]; cricket[0.047]; Australia[0.047]; Cricket[0.046]; match[0.046]; Tasmania[0.045]; Tasmania[0.045]; play[0.044]; second[0.044]; Shield[0.043]; Shield[0.043]; Hobart[0.043]; Dean[0.042]; David[0.042]; Michael[0.041]; Young[0.041]; score[0.041]; Shaun[0.041]; Bellerive[0.041]; Sheffield[0.041]; Sheffield[0.041]; Saturday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tasmania ==> ENTITY: \u001b[32mTasmania cricket team\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.188:0.188[0]; log p(e|m)= -2.430:-2.430[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.047]; Australia[0.046]; Oval[0.045]; Tasmania[0.044]; Victoria[0.044]; Victoria[0.044]; match[0.044]; Cricket[0.043]; Hobart[0.042]; play[0.042]; day[0.042]; day[0.042]; second[0.041]; score[0.041]; David[0.040]; Bellerive[0.040]; Boon[0.040]; Shaun[0.040]; Shield[0.040]; Shield[0.040]; Saturday[0.039]; Dean[0.039]; Michael[0.039]; declared[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dean Jones ==> ENTITY: \u001b[32mDean Jones (cricketer)\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.049]; Cricket[0.045]; Australia[0.045]; David[0.044]; match[0.044]; Oval[0.044]; day[0.043]; day[0.043]; Michael[0.043]; second[0.042]; Shaun[0.042]; play[0.041]; Hobart[0.041]; Boon[0.041]; score[0.040]; Victoria[0.040]; Victoria[0.040]; Saturday[0.040]; Tasmania[0.040]; Tasmania[0.040]; Young[0.039]; Sheffield[0.039]; Sheffield[0.039]; Bellerive[0.038]; \t\n\r [========>.....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s878ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 910/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1375testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national rugby union team\u001b[39m\t\nSCORES: global= 0.241:0.241[0]; local()= 0.182:0.182[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.050]; Australia[0.050]; Australia[0.050]; Barbarians[0.044]; Barbarians[0.044]; Barbarians[0.044]; match[0.044]; Campese[0.043]; Quinnell[0.042]; final[0.041]; Horan[0.041]; David[0.040]; Scott[0.040]; tour[0.040]; European[0.039]; Saturday[0.039]; Alan[0.039]; Scorers[0.039]; Matthew[0.039]; Union[0.039]; Andrew[0.039]; beat[0.038]; Roff[0.038]; halftime[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national rugby union team\u001b[39m\t\nSCORES: global= 0.241:0.241[0]; local()= 0.182:0.182[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.050]; Australia[0.050]; Australia[0.050]; Barbarians[0.044]; Barbarians[0.044]; Barbarians[0.044]; match[0.044]; Campese[0.043]; Quinnell[0.042]; final[0.041]; Horan[0.041]; David[0.040]; Scott[0.040]; tour[0.040]; European[0.039]; Saturday[0.039]; Alan[0.039]; Scorers[0.039]; Matthew[0.039]; Union[0.039]; Andrew[0.039]; beat[0.038]; Roff[0.038]; halftime[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rob Andrew\u001b[39m ==> ENTITY: \u001b[32mRob Andrew\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.049]; match[0.046]; Australia[0.045]; Australia[0.045]; Australia[0.045]; Barbarians[0.045]; Barbarians[0.045]; Barbarians[0.045]; final[0.044]; Quinnell[0.044]; David[0.044]; Matthew[0.044]; European[0.044]; Scott[0.043]; Saturday[0.043]; London[0.042]; Scorers[0.042]; halftime[0.042]; Tim[0.042]; Campese[0.041]; Alan[0.041]; Conversions[0.041]; Union[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Burke\u001b[39m ==> ENTITY: \u001b[32mMatt Burke\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.045]; Horan[0.044]; Burke[0.044]; Burke[0.044]; Australia[0.044]; Australia[0.044]; Australia[0.044]; Matthew[0.042]; Barbarians[0.042]; Barbarians[0.042]; Barbarians[0.042]; Quinnell[0.042]; Campese[0.042]; Andrew[0.041]; match[0.041]; Rob[0.040]; Tim[0.040]; Scott[0.040]; Scorers[0.039]; Conversions[0.039]; final[0.039]; Tries[0.039]; Tries[0.039]; Joe[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Matthew Burke\u001b[39m ==> ENTITY: \u001b[32mMatt Burke\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.045]; Horan[0.044]; Burke[0.044]; Burke[0.044]; Australia[0.044]; Australia[0.044]; Australia[0.044]; Barbarians[0.042]; Barbarians[0.042]; Barbarians[0.042]; Quinnell[0.042]; Campese[0.042]; Andrew[0.041]; match[0.041]; Rob[0.041]; Tim[0.040]; Scott[0.040]; Scorers[0.040]; Conversions[0.040]; final[0.040]; Tries[0.039]; Tries[0.039]; Joe[0.039]; David[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barbarians ==> ENTITY: \u001b[32mBarbarian F.C.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.138:0.138[0]; log p(e|m)= -0.171:-0.171[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.049]; Barbarians[0.046]; Barbarians[0.046]; Australia[0.045]; Australia[0.045]; Australia[0.045]; tour[0.043]; match[0.043]; Saturday[0.042]; Campese[0.042]; Union[0.041]; Quinnell[0.041]; European[0.041]; Horan[0.041]; David[0.040]; London[0.039]; Rob[0.039]; Joe[0.039]; final[0.039]; Roff[0.039]; Andrew[0.039]; Scott[0.039]; Alan[0.039]; Matthew[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barbarians ==> ENTITY: \u001b[32mBarbarian F.C.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.138:0.138[0]; log p(e|m)= -0.171:-0.171[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.049]; Barbarians[0.046]; Barbarians[0.046]; Australia[0.045]; Australia[0.045]; Australia[0.045]; tour[0.043]; match[0.043]; Saturday[0.042]; Campese[0.042]; Union[0.041]; Quinnell[0.041]; European[0.041]; Horan[0.041]; David[0.040]; London[0.039]; Rob[0.039]; Joe[0.039]; final[0.039]; Roff[0.039]; Andrew[0.039]; Scott[0.039]; Alan[0.039]; Matthew[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Joe Roff\u001b[39m ==> ENTITY: \u001b[32mJoe Roff\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.050]; Australia[0.049]; Australia[0.049]; Australia[0.049]; Barbarians[0.045]; Barbarians[0.045]; Barbarians[0.045]; match[0.045]; Campese[0.044]; final[0.043]; Union[0.043]; Rob[0.043]; Andrew[0.042]; halftime[0.042]; David[0.042]; European[0.041]; Scorers[0.041]; Tim[0.041]; Horan[0.041]; Alan[0.041]; Scott[0.040]; Matthew[0.040]; London[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rugby Union\u001b[39m ==> ENTITY: \u001b[32mRugby union\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.047]; Australia[0.047]; Australia[0.047]; Barbarians[0.044]; Barbarians[0.044]; Barbarians[0.044]; match[0.043]; Quinnell[0.042]; Horan[0.042]; Campese[0.042]; final[0.041]; Burke[0.040]; Burke[0.040]; Burke[0.040]; Saturday[0.040]; European[0.040]; Scorers[0.040]; Andrew[0.040]; Rob[0.039]; David[0.039]; Matthew[0.039]; tour[0.039]; London[0.039]; Scott[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.060:0.060[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.046]; final[0.045]; tour[0.043]; Saturday[0.043]; David[0.043]; Alan[0.043]; Rugby[0.043]; Rob[0.042]; Scott[0.042]; beat[0.042]; European[0.042]; Andrew[0.042]; Joe[0.041]; Scorers[0.041]; Tim[0.041]; Beat[0.041]; Burke[0.040]; Burke[0.040]; Burke[0.040]; Australia[0.040]; Australia[0.040]; Australia[0.040]; Matthew[0.040]; Union[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Scott Quinnell\u001b[39m ==> ENTITY: \u001b[32mScott Quinnell\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.049]; Australia[0.044]; Australia[0.044]; Australia[0.044]; final[0.044]; match[0.043]; Barbarians[0.042]; Barbarians[0.042]; Barbarians[0.042]; David[0.041]; Alan[0.041]; Matthew[0.041]; tour[0.041]; beat[0.041]; Saturday[0.041]; Rob[0.041]; European[0.041]; London[0.040]; halftime[0.040]; Scorers[0.040]; Tries[0.040]; Tries[0.040]; Tim[0.039]; Campese[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = David Campese\u001b[39m ==> ENTITY: \u001b[32mDavid Campese\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.045]; Australia[0.045]; Australia[0.045]; Rugby[0.044]; Quinnell[0.043]; Horan[0.043]; Matthew[0.043]; Scott[0.043]; Andrew[0.042]; Barbarians[0.042]; Barbarians[0.042]; Barbarians[0.042]; Tim[0.041]; Rob[0.041]; match[0.041]; Tries[0.041]; Tries[0.041]; Scorers[0.040]; Joe[0.040]; Burke[0.039]; Burke[0.039]; Burke[0.039]; Alan[0.039]; Roff[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = European ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.241:0.241[0]; local()= 0.025:0.025[0]; log p(e|m)= -0.256:-0.256[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.045]; match[0.045]; final[0.044]; tour[0.044]; Australia[0.044]; Australia[0.044]; Australia[0.044]; Union[0.043]; Alan[0.043]; David[0.042]; Joe[0.042]; Saturday[0.042]; Andrew[0.042]; London[0.041]; Matthew[0.041]; beat[0.040]; Scott[0.040]; Burke[0.040]; Burke[0.040]; Burke[0.040]; Bateman[0.039]; Tim[0.039]; Rob[0.039]; Penalties[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBarbarians\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBarbarian F.C.\u001b[39m <---> \u001b[32mNew Zealand Barbarians\u001b[39m\t\nSCORES: global= 0.270:0.246[\u001b[31m0.024\u001b[39m]; local()= 0.138:0.099[\u001b[31m0.039\u001b[39m]; log p(e|m)= -0.171:-1.790[\u001b[31m1.619\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.049]; Barbarians[0.046]; Barbarians[0.046]; Australia[0.045]; Australia[0.045]; Australia[0.045]; tour[0.043]; match[0.043]; Saturday[0.042]; Campese[0.042]; Union[0.041]; Quinnell[0.041]; European[0.041]; Horan[0.041]; David[0.040]; London[0.039]; Rob[0.039]; Joe[0.039]; final[0.039]; Roff[0.039]; Andrew[0.039]; Scott[0.039]; Alan[0.039]; Matthew[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tim Horan\u001b[39m ==> ENTITY: \u001b[32mTim Horan\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.045]; Australia[0.044]; Australia[0.044]; Australia[0.044]; Campese[0.044]; Quinnell[0.043]; Burke[0.043]; Burke[0.043]; Burke[0.043]; match[0.042]; Barbarians[0.041]; Barbarians[0.041]; Barbarians[0.041]; Joe[0.041]; Union[0.040]; Andrew[0.040]; Matthew[0.040]; Scott[0.040]; David[0.040]; halftime[0.040]; Scorers[0.040]; Rob[0.040]; tour[0.039]; final[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national rugby union team\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.179:0.179[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.050]; Australia[0.050]; Australia[0.050]; Barbarians[0.044]; Barbarians[0.044]; Barbarians[0.044]; match[0.043]; Campese[0.043]; Quinnell[0.042]; final[0.041]; Horan[0.041]; David[0.040]; European[0.040]; tour[0.040]; Alan[0.039]; Scorers[0.039]; Saturday[0.039]; Matthew[0.039]; Scott[0.039]; Union[0.039]; Andrew[0.039]; Roff[0.038]; halftime[0.038]; London[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Burke\u001b[39m ==> ENTITY: \u001b[32mMatt Burke\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.045]; Horan[0.044]; Burke[0.044]; Burke[0.044]; Australia[0.044]; Australia[0.044]; Australia[0.044]; Matthew[0.042]; Barbarians[0.042]; Barbarians[0.042]; Barbarians[0.042]; Quinnell[0.042]; Campese[0.042]; Andrew[0.041]; match[0.041]; Rob[0.040]; Tim[0.040]; Scott[0.040]; Scorers[0.039]; Conversions[0.039]; final[0.039]; Tries[0.039]; Tries[0.039]; Joe[0.039]; \t\n\r [========>.....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s796ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 927/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1207testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.050]; team[0.049]; soccer[0.046]; Austrian[0.045]; men[0.044]; fans[0.043]; fans[0.043]; won[0.043]; taken[0.042]; play[0.041]; Vienna[0.041]; Vienna[0.041]; brawl[0.040]; Manchester[0.040]; Manchester[0.040]; Manchester[0.040]; Wednesday[0.040]; leg[0.039]; Friday[0.039]; Britons[0.039]; involved[0.039]; United[0.038]; United[0.038]; United[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Manchester United\u001b[39m ==> ENTITY: \u001b[32mManchester United F.C.\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nManchester[0.048]; Manchester[0.048]; team[0.048]; United[0.044]; United[0.044]; soccer[0.044]; shot[0.043]; fans[0.042]; fans[0.042]; Britons[0.041]; leg[0.041]; play[0.041]; Wednesday[0.041]; hit[0.041]; won[0.040]; brawl[0.040]; Friday[0.040]; Gunmen[0.040]; unidentified[0.039]; earlier[0.039]; Fans[0.039]; watch[0.038]; involved[0.038]; taken[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Manchester United\u001b[39m ==> ENTITY: \u001b[32mManchester United F.C.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nManchester[0.049]; Manchester[0.049]; team[0.048]; United[0.044]; United[0.044]; soccer[0.044]; fans[0.042]; fans[0.042]; Britons[0.042]; Wednesday[0.041]; won[0.041]; play[0.041]; brawl[0.041]; Friday[0.041]; Gunmen[0.040]; unidentified[0.040]; earlier[0.040]; hit[0.039]; leg[0.039]; Fans[0.039]; watch[0.039]; involved[0.039]; taken[0.038]; men[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.099:0.099[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntold[0.046]; said[0.043]; said[0.043]; said[0.043]; Wednesday[0.043]; earlier[0.043]; Britons[0.042]; involved[0.042]; Friday[0.042]; Manchester[0.042]; Manchester[0.042]; Manchester[0.042]; danger[0.042]; unidentified[0.041]; Austrian[0.041]; spokeswoman[0.041]; spokeswoman[0.041]; brawl[0.041]; capital[0.041]; team[0.041]; watch[0.040]; taken[0.040]; Austria[0.040]; Austria[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mercedes ==> ENTITY: \u001b[32mMercedes-Benz\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.104:0.104[0]; log p(e|m)= -0.783:-0.783[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncar[0.051]; told[0.047]; won[0.045]; team[0.045]; said[0.045]; said[0.045]; said[0.045]; hit[0.045]; watch[0.044]; fans[0.044]; fans[0.044]; involved[0.044]; Fans[0.043]; earlier[0.042]; Austrian[0.042]; lives[0.042]; shot[0.042]; Wednesday[0.041]; play[0.041]; Austria[0.041]; Austria[0.041]; Rapid[0.041]; Friday[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.050]; soccer[0.046]; team[0.046]; won[0.046]; Austrian[0.044]; play[0.043]; fans[0.043]; fans[0.043]; men[0.043]; Vienna[0.042]; Vienna[0.042]; Manchester[0.041]; Manchester[0.041]; Manchester[0.041]; Wednesday[0.040]; leg[0.040]; Friday[0.039]; Britons[0.039]; capital[0.039]; United[0.039]; United[0.039]; United[0.039]; Mercedes[0.039]; earlier[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vienna ==> ENTITY: \u001b[32mVienna\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.126:-0.126[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.046]; Austria[0.046]; Vienna[0.046]; Wednesday[0.043]; Austrian[0.042]; Friday[0.042]; won[0.042]; play[0.042]; hit[0.042]; said[0.041]; said[0.041]; said[0.041]; team[0.041]; lives[0.041]; Manchester[0.041]; Manchester[0.041]; Manchester[0.041]; shot[0.040]; earlier[0.040]; soccer[0.040]; capital[0.040]; car[0.040]; involved[0.040]; hospital[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBritons\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUnited Kingdom\u001b[39m <---> \u001b[32mBritish people\u001b[39m\t\nSCORES: global= 0.267:0.249[\u001b[31m0.018\u001b[39m]; local()= 0.093:0.063[\u001b[31m0.030\u001b[39m]; log p(e|m)= -1.505:-1.091[\u001b[32m0.414\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWound[0.048]; leg[0.045]; hospital[0.045]; Manchester[0.044]; Manchester[0.044]; Manchester[0.044]; earlier[0.044]; Wednesday[0.044]; United[0.044]; United[0.044]; United[0.044]; pelvis[0.043]; Friday[0.043]; team[0.043]; Austria[0.043]; Austria[0.043]; hit[0.043]; pub[0.043]; play[0.042]; said[0.042]; said[0.042]; said[0.042]; told[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austrian ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.406:-0.406[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.051]; Austria[0.051]; team[0.043]; soccer[0.042]; Vienna[0.042]; Vienna[0.042]; won[0.042]; play[0.042]; Manchester[0.041]; Manchester[0.041]; Manchester[0.041]; Wednesday[0.041]; Britons[0.041]; leg[0.040]; fans[0.040]; fans[0.040]; said[0.040]; said[0.040]; said[0.040]; Friday[0.040]; United[0.039]; United[0.039]; United[0.039]; earlier[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rapid Vienna\u001b[39m ==> ENTITY: \u001b[32mSK Rapid Wien\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.045]; won[0.045]; Wound[0.044]; fans[0.044]; fans[0.044]; team[0.044]; Austria[0.044]; Austria[0.044]; play[0.043]; leg[0.042]; hospital[0.041]; United[0.041]; United[0.041]; United[0.041]; Manchester[0.041]; Manchester[0.041]; Manchester[0.041]; hit[0.040]; pelvis[0.040]; Austrian[0.040]; watch[0.040]; Wednesday[0.040]; Fans[0.039]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Manchester United\u001b[39m ==> ENTITY: \u001b[32mManchester United F.C.\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nManchester[0.048]; Manchester[0.048]; team[0.048]; United[0.044]; United[0.044]; soccer[0.044]; shot[0.043]; fans[0.042]; fans[0.042]; Britons[0.041]; leg[0.041]; play[0.041]; Wednesday[0.041]; hit[0.041]; won[0.040]; brawl[0.040]; Friday[0.040]; Gunmen[0.040]; unidentified[0.039]; earlier[0.039]; Fans[0.039]; watch[0.038]; involved[0.038]; taken[0.038]; \t\n\r [========>.....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s841ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 938/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1227testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mElectronic Data Systems\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mElectronic Data Systems\u001b[39m <---> \u001b[32mHP Enterprise Services\u001b[39m\t\nSCORES: global= 0.295:0.290[\u001b[31m0.005\u001b[39m]; local()= 0.160:0.132[\u001b[31m0.028\u001b[39m]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nServices[0.045]; subsidiary[0.044]; contract[0.044]; contract[0.044]; contract[0.044]; contract[0.044]; firm[0.043]; technology[0.043]; America[0.043]; North[0.042]; Data[0.042]; Aviation[0.042]; Europe[0.041]; Authority[0.041]; million[0.040]; said[0.040]; said[0.040]; Private[0.040]; upgrading[0.040]; Information[0.040]; data[0.039]; data[0.039]; funded[0.039]; National[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPrestwick\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mGlasgow Prestwick Airport\u001b[39m <---> \u001b[32mPrestwick\u001b[39m\t\nSCORES: global= 0.281:0.278[\u001b[31m0.004\u001b[39m]; local()= 0.204:0.159[\u001b[31m0.046\u001b[39m]; log p(e|m)= -1.291:0.000[\u001b[32m1.291\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.046]; flight[0.045]; flight[0.045]; service[0.044]; Aviation[0.043]; routes[0.043]; Air[0.042]; aircraft[0.042]; transatlantic[0.042]; North[0.041]; south[0.041]; region[0.041]; Traffic[0.041]; Control[0.041]; National[0.041]; London[0.041]; London[0.041]; scheduled[0.040]; Centre[0.040]; west[0.040]; Authority[0.040]; Services[0.040]; flying[0.040]; air[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Europe ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.082:0.082[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsatellite[0.045]; America[0.044]; Aviation[0.043]; aircraft[0.043]; transatlantic[0.043]; service[0.042]; North[0.042]; Scotland[0.042]; region[0.042]; Private[0.042]; contract[0.042]; contract[0.042]; contract[0.042]; Air[0.042]; flight[0.041]; flying[0.041]; scheduled[0.041]; National[0.041]; subsidiary[0.041]; technology[0.040]; data[0.040]; funded[0.040]; Systems[0.040]; routes[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Civil Aviation Authority ==> ENTITY: \u001b[32mCivil Aviation Authority (United Kingdom)\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.168:0.168[0]; log p(e|m)= -0.844:-0.844[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \naircraft[0.044]; service[0.044]; Prestwick[0.043]; Air[0.043]; Data[0.043]; Data[0.043]; Control[0.043]; Services[0.042]; Information[0.042]; data[0.042]; data[0.042]; subsidiary[0.042]; London[0.041]; London[0.041]; flight[0.041]; flight[0.041]; Private[0.041]; Traffic[0.041]; air[0.040]; flying[0.040]; control[0.040]; control[0.040]; routes[0.040]; use[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = North America ==> ENTITY: \u001b[32mNorth America\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.056:0.056[0]; log p(e|m)= -0.012:-0.012[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurope[0.047]; service[0.046]; region[0.045]; use[0.044]; National[0.043]; aircraft[0.042]; million[0.042]; Scotland[0.042]; Friday[0.042]; scheduled[0.041]; satellite[0.041]; contract[0.041]; contract[0.041]; contract[0.041]; processing[0.041]; flight[0.041]; Private[0.040]; Air[0.040]; routes[0.040]; Civil[0.040]; Authority[0.040]; involved[0.040]; Services[0.040]; Oceanic[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEDS\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mElectronic Data Systems\u001b[39m <---> \u001b[32mHP Enterprise Services\u001b[39m\t\nSCORES: global= 0.265:0.258[\u001b[31m0.008\u001b[39m]; local()= 0.155:0.129[\u001b[31m0.026\u001b[39m]; log p(e|m)= -0.308:-1.370[\u001b[31m1.063\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nServices[0.046]; subsidiary[0.045]; Systems[0.045]; contract[0.045]; contract[0.045]; contract[0.045]; contract[0.045]; firm[0.044]; technology[0.044]; technology[0.044]; America[0.044]; service[0.043]; processing[0.043]; North[0.043]; Data[0.043]; Data[0.043]; Aviation[0.043]; Europe[0.042]; data[0.042]; data[0.042]; Authority[0.042]; London[0.042]; London[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NATS\u001b[39m ==> ENTITY: \u001b[32mNATS Holdings\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.224:0.224[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nservice[0.046]; Aviation[0.045]; Prestwick[0.043]; Services[0.043]; Authority[0.043]; National[0.043]; Traffic[0.043]; Systems[0.042]; Air[0.042]; aircraft[0.042]; contract[0.041]; contract[0.041]; contract[0.041]; contract[0.041]; routes[0.041]; subsidiary[0.041]; Control[0.041]; traffic[0.041]; Information[0.041]; stg[0.040]; firm[0.040]; Europe[0.040]; air[0.039]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scotland ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.090:0.090[0]; log p(e|m)= -0.229:-0.229[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPrestwick[0.045]; National[0.043]; scheduled[0.043]; Europe[0.043]; North[0.043]; contract[0.043]; contract[0.043]; contract[0.043]; contract[0.043]; London[0.042]; London[0.042]; south[0.042]; west[0.041]; involved[0.041]; service[0.041]; bagged[0.041]; Friday[0.041]; America[0.041]; use[0.040]; Centre[0.040]; statement[0.040]; firm[0.040]; funded[0.040]; Private[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.051:0.051[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nflight[0.046]; flight[0.046]; Prestwick[0.044]; traffic[0.044]; firm[0.044]; Scotland[0.043]; subsidiary[0.043]; stg[0.043]; aircraft[0.042]; Traffic[0.041]; Air[0.041]; west[0.041]; south[0.041]; Centre[0.041]; Authority[0.041]; Services[0.041]; Friday[0.041]; Finance[0.041]; air[0.040]; Aviation[0.040]; Private[0.040]; National[0.039]; statement[0.039]; Control[0.039]; \t\n\r [========>.....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s876ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 947/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1377testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = N Ireland ==> ENTITY: \u001b[32mNorthern Ireland national football team\u001b[39m\t\nSCORES: global= 0.230:0.230[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.049]; Ireland[0.049]; qualifier[0.043]; Albania[0.043]; Albania[0.043]; Albania[0.043]; Cup[0.043]; Soccer[0.043]; soccer[0.043]; soccer[0.043]; Football[0.042]; players[0.042]; country[0.041]; country[0.041]; Saturday[0.040]; Saturday[0.040]; Squad[0.039]; reinstated[0.039]; World[0.039]; Fifa[0.039]; Fifa[0.039]; Tuesday[0.039]; Hafizi[0.039]; coach[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Altin Haxhi\u001b[39m ==> ENTITY: \u001b[32mAltin Haxhi\u001b[39m\t\nSCORES: global= 0.228:0.228[0]; local()= 0.135:0.135[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndefender[0.044]; Malko[0.044]; win[0.043]; Dervishi[0.043]; Altin[0.043]; squad[0.043]; Afrim[0.043]; match[0.042]; Midfielders[0.042]; Ilir[0.042]; Hafizi[0.042]; Hafizi[0.042]; Northern[0.041]; Arjan[0.041]; Arjan[0.041]; Fatmir[0.041]; Saturday[0.041]; draw[0.041]; Vata[0.041]; Vata[0.041]; Saimir[0.040]; Bogdani[0.040]; Erjon[0.040]; Forwards[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Artur Lekbello\u001b[39m ==> ENTITY: \u001b[32mArtur Lekbello\u001b[39m\t\nSCORES: global= 0.226:0.226[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.047]; Albania[0.045]; Albania[0.045]; Albanian[0.044]; Football[0.043]; Malko[0.043]; Fifa[0.041]; Haxhi[0.041]; sports[0.041]; Midfielders[0.041]; Hafizi[0.041]; Hafizi[0.041]; defender[0.041]; Northern[0.041]; Balkan[0.041]; Arjan[0.041]; Arjan[0.041]; suspensions[0.041]; Erjon[0.041]; Ireland[0.040]; Dervishi[0.040]; Afrim[0.040]; Vata[0.040]; Vata[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = FIFA\u001b[39m ==> ENTITY: \u001b[32mFIFA\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.049]; soccer[0.049]; Soccer[0.048]; Fifa[0.048]; Football[0.047]; players[0.047]; Ireland[0.046]; Ireland[0.046]; Ireland[0.046]; sports[0.046]; country[0.046]; country[0.046]; match[0.045]; Cup[0.045]; banned[0.044]; Play[0.044]; World[0.044]; squad[0.044]; Association[0.043]; reinstated[0.043]; qualifier[0.043]; draw[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Northern Ireland ==> ENTITY: \u001b[32mNorthern Ireland national football team\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.156:0.156[0]; log p(e|m)= -2.207:-2.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.050]; Ireland[0.050]; Belfast[0.048]; qualifier[0.046]; Cup[0.044]; governing[0.043]; Albania[0.043]; Albania[0.043]; Albania[0.043]; Association[0.043]; Soccer[0.043]; soccer[0.043]; soccer[0.043]; Football[0.043]; players[0.042]; banned[0.042]; country[0.042]; country[0.042]; draw[0.041]; Northern[0.041]; win[0.041]; reinstated[0.041]; committee[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ilir Shulku\u001b[39m ==> ENTITY: \u001b[32mIlir Shulku\u001b[39m\t\nSCORES: global= 0.217:0.217[0]; local()= 0.088:0.088[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlbanian[0.048]; named[0.047]; Malko[0.046]; Football[0.045]; Haxhi[0.044]; Midfielders[0.044]; squad[0.044]; Hafizi[0.044]; Hafizi[0.044]; Afrim[0.043]; Northern[0.043]; Fatmir[0.043]; Arjan[0.043]; Arjan[0.043]; Dervishi[0.043]; Vata[0.042]; Vata[0.042]; Altin[0.042]; Altin[0.042]; Bogdani[0.042]; Erjon[0.042]; Forwards[0.042]; Rudi[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fatmir Vata\u001b[39m ==> ENTITY: \u001b[32mFatmir Vata\u001b[39m\t\nSCORES: global= 0.225:0.225[0]; local()= 0.105:0.105[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsquad[0.045]; defender[0.044]; Vata[0.044]; Malko[0.044]; named[0.043]; Midfielders[0.042]; Haxhi[0.042]; Hafizi[0.042]; Hafizi[0.042]; Goalkeepers[0.042]; Northern[0.042]; Arjan[0.041]; Arjan[0.041]; Ireland[0.041]; Afrim[0.041]; Erjon[0.041]; Altin[0.040]; Altin[0.040]; Bogdani[0.040]; Forwards[0.040]; Rudi[0.040]; injured[0.040]; suspension[0.040]; Defenders[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Blendi Nallbani\u001b[39m ==> ENTITY: \u001b[32mBlendi Nallbani\u001b[39m\t\nSCORES: global= 0.228:0.228[0]; local()= 0.092:0.092[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.045]; Albania[0.044]; Malko[0.043]; match[0.043]; happy[0.043]; Albanian[0.042]; Haxhi[0.042]; sports[0.042]; Midfielders[0.042]; Erjon[0.042]; squad[0.041]; Fifa[0.041]; Hafizi[0.041]; Hafizi[0.041]; draw[0.041]; Northern[0.041]; Football[0.041]; Arjan[0.041]; Arjan[0.041]; keen[0.041]; Ilir[0.041]; difficult[0.041]; Dervishi[0.041]; defender[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Albanian Football Association\u001b[39m ==> ENTITY: \u001b[32mAlbanian Football Association\u001b[39m\t\nSCORES: global= 0.236:0.236[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.046]; soccer[0.046]; named[0.044]; reinstated[0.043]; Albania[0.043]; Albania[0.043]; banned[0.043]; Malko[0.042]; sports[0.042]; Albanian[0.041]; lifted[0.041]; squad[0.041]; governing[0.041]; players[0.041]; match[0.041]; qualifier[0.041]; suspensions[0.041]; Hafizi[0.040]; Hafizi[0.040]; Hafizi[0.040]; Play[0.040]; said[0.040]; said[0.040]; officials[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Albania ==> ENTITY: \u001b[32mAlbania\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.411:-0.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlbania[0.045]; Albania[0.045]; country[0.044]; country[0.044]; soccer[0.044]; soccer[0.044]; qualifier[0.043]; players[0.042]; Soccer[0.042]; Balkan[0.041]; Ireland[0.041]; Ireland[0.041]; Ireland[0.041]; sports[0.041]; Football[0.041]; Albanian[0.041]; Albanian[0.041]; Cup[0.041]; draw[0.040]; Association[0.040]; win[0.039]; Play[0.039]; lifted[0.039]; suspensions[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ervin Fakaj\u001b[39m ==> ENTITY: \u001b[32mErvin Fakaj\u001b[39m\t\nSCORES: global= 0.227:0.227[0]; local()= 0.128:0.128[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndefender[0.046]; squad[0.045]; Malko[0.044]; Ilir[0.042]; Fatmir[0.042]; Haxhi[0.042]; Midfielders[0.042]; match[0.042]; Hafizi[0.042]; Hafizi[0.042]; Erjon[0.041]; Northern[0.041]; committee[0.041]; named[0.041]; Arjan[0.041]; Arjan[0.041]; Ireland[0.041]; Saimir[0.041]; Afrim[0.041]; Vata[0.041]; Vata[0.041]; Altin[0.040]; Altin[0.040]; Bogdani[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tirana\u001b[39m ==> ENTITY: \u001b[32mTirana\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayers[0.047]; Albania[0.046]; Albania[0.046]; Albania[0.046]; soccer[0.046]; soccer[0.046]; sports[0.044]; Albanian[0.044]; Albanian[0.044]; qualifier[0.043]; Cup[0.043]; Football[0.043]; Soccer[0.043]; Association[0.043]; government[0.043]; governing[0.042]; country[0.042]; country[0.042]; important[0.042]; various[0.042]; coach[0.042]; officials[0.041]; Tuesday[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nevil Dede\u001b[39m ==> ENTITY: \u001b[32mNevil Dede\u001b[39m\t\nSCORES: global= 0.223:0.223[0]; local()= 0.089:0.089[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndefender[0.044]; Malko[0.044]; Football[0.043]; squad[0.042]; Haxhi[0.042]; Midfielders[0.042]; Hafizi[0.042]; Hafizi[0.042]; match[0.042]; Albanian[0.042]; Northern[0.042]; Arjan[0.042]; Arjan[0.042]; Dervishi[0.041]; Afrim[0.041]; Vata[0.041]; Vata[0.041]; Altin[0.041]; Altin[0.041]; Bogdani[0.041]; Erjon[0.041]; Forwards[0.041]; Rudi[0.040]; Association[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Northern Ireland ==> ENTITY: \u001b[32mNorthern Ireland national football team\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.156:0.156[0]; log p(e|m)= -2.207:-2.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.048]; Belfast[0.047]; Belfast[0.047]; qualifier[0.044]; squad[0.044]; match[0.044]; soccer[0.043]; soccer[0.043]; Albania[0.042]; Albania[0.042]; Football[0.041]; Defenders[0.041]; Squad[0.041]; country[0.041]; sports[0.040]; defender[0.040]; draw[0.040]; Northern[0.040]; win[0.040]; Saturday[0.039]; ordered[0.039]; lifted[0.039]; reinstated[0.038]; World[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Balkan ==> ENTITY: \u001b[32mBalkans\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.045]; country[0.045]; Albania[0.044]; Albania[0.044]; Albania[0.044]; soccer[0.043]; soccer[0.043]; squad[0.043]; Soccer[0.043]; Football[0.042]; Albanian[0.042]; Albanian[0.042]; sports[0.041]; qualifier[0.040]; players[0.040]; Cup[0.040]; win[0.040]; Northern[0.040]; Northern[0.040]; Ireland[0.040]; Ireland[0.040]; Ireland[0.040]; various[0.040]; group[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Erjon Bogdani\u001b[39m ==> ENTITY: \u001b[32mErjon Bogdani\u001b[39m\t\nSCORES: global= 0.229:0.229[0]; local()= 0.087:0.087[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMalko[0.044]; match[0.043]; Afrim[0.043]; Haxhi[0.043]; Midfielders[0.043]; Hafizi[0.042]; win[0.042]; Ilir[0.042]; Northern[0.042]; Arjan[0.042]; Arjan[0.042]; defender[0.042]; injured[0.042]; named[0.041]; Ireland[0.041]; Vata[0.041]; Vata[0.041]; Altin[0.041]; Altin[0.041]; squad[0.041]; Forwards[0.041]; Rudi[0.041]; suspension[0.040]; Fatmir[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rudi Vata\u001b[39m ==> ENTITY: \u001b[32mRudi Vata\u001b[39m\t\nSCORES: global= 0.229:0.229[0]; local()= 0.127:0.127[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlbania[0.045]; Ireland[0.045]; Football[0.043]; Malko[0.043]; match[0.043]; sports[0.042]; Vata[0.042]; Belfast[0.042]; Belfast[0.042]; Erjon[0.042]; Albanian[0.042]; squad[0.042]; defender[0.041]; Haxhi[0.041]; Midfielders[0.041]; Hafizi[0.041]; Hafizi[0.041]; Northern[0.041]; Afrim[0.041]; Fatmir[0.041]; Fifa[0.041]; Arjan[0.041]; Arjan[0.041]; Artur[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Altin Rraklli\u001b[39m ==> ENTITY: \u001b[32mAltin Rraklli\u001b[39m\t\nSCORES: global= 0.230:0.230[0]; local()= 0.103:0.103[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsquad[0.044]; Malko[0.044]; match[0.043]; defender[0.043]; Haxhi[0.042]; Midfielders[0.042]; win[0.042]; Rudi[0.042]; Hafizi[0.042]; Hafizi[0.042]; Northern[0.042]; Altin[0.042]; Arjan[0.042]; Arjan[0.042]; draw[0.041]; Afrim[0.041]; Vata[0.041]; Vata[0.041]; happy[0.041]; Bogdani[0.041]; Erjon[0.040]; Forwards[0.040]; Goalkeepers[0.040]; Ireland[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belfast ==> ENTITY: \u001b[32mBelfast\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.073:0.073[0]; log p(e|m)= -0.131:-0.131[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelfast[0.051]; Ireland[0.049]; soccer[0.046]; match[0.045]; squad[0.044]; sports[0.043]; Football[0.043]; Saturday[0.043]; win[0.043]; Squad[0.041]; defender[0.041]; named[0.041]; Northern[0.040]; draw[0.040]; Fifa[0.040]; Defenders[0.040]; keen[0.040]; said[0.039]; banned[0.039]; Association[0.039]; lifted[0.039]; officials[0.039]; happy[0.038]; Forwards[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Albania ==> ENTITY: \u001b[32mAlbania\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.161:0.161[0]; log p(e|m)= -0.411:-0.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlbania[0.045]; Albania[0.045]; country[0.044]; country[0.044]; match[0.044]; soccer[0.043]; soccer[0.043]; qualifier[0.043]; squad[0.043]; players[0.042]; Soccer[0.041]; Balkan[0.041]; Ireland[0.041]; Ireland[0.041]; Ireland[0.041]; sports[0.041]; Football[0.041]; Albanian[0.041]; Albanian[0.041]; Cup[0.040]; draw[0.040]; Association[0.039]; win[0.039]; Play[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAlbania\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAlbania\u001b[39m <---> \u001b[32mAlbania national football team\u001b[39m\t\nSCORES: global= 0.246:0.242[\u001b[31m0.004\u001b[39m]; local()= 0.142:0.171[\u001b[32m0.030\u001b[39m]; log p(e|m)= -0.411:-2.617[\u001b[31m2.206\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlbania[0.046]; Albania[0.046]; country[0.045]; country[0.045]; soccer[0.044]; soccer[0.044]; qualifier[0.044]; players[0.043]; Ireland[0.043]; Ireland[0.043]; Soccer[0.042]; Balkan[0.042]; sports[0.042]; Albanian[0.041]; Cup[0.041]; group[0.040]; Play[0.039]; lifted[0.039]; suspensions[0.039]; Tirana[0.039]; coach[0.039]; suspension[0.039]; Fifa[0.038]; Fifa[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAlbanian\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAlbania\u001b[39m <---> \u001b[32mAlbania national football team\u001b[39m\t\nSCORES: global= 0.248:0.240[\u001b[31m0.008\u001b[39m]; local()= 0.161:0.178[\u001b[32m0.017\u001b[39m]; log p(e|m)= -1.556:-3.124[\u001b[31m1.568\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlbania[0.047]; Albania[0.047]; Albania[0.047]; Albanian[0.046]; country[0.046]; country[0.046]; soccer[0.045]; soccer[0.045]; qualifier[0.045]; Balkan[0.044]; players[0.044]; Soccer[0.043]; Ireland[0.043]; Ireland[0.043]; sports[0.043]; Football[0.043]; Cup[0.042]; World[0.040]; Play[0.040]; suspensions[0.040]; Tirana[0.040]; coach[0.040]; suspension[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bledar Kola\u001b[39m ==> ENTITY: \u001b[32mBledar Kola\u001b[39m\t\nSCORES: global= 0.217:0.217[0]; local()= 0.062:0.062[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMalko[0.044]; Haxhi[0.043]; Midfielders[0.043]; Hafizi[0.042]; Hafizi[0.042]; Northern[0.042]; Arjan[0.042]; Arjan[0.042]; Afrim[0.042]; squad[0.042]; secretary[0.042]; named[0.042]; Dervishi[0.042]; Goalkeepers[0.041]; Vata[0.041]; Vata[0.041]; Altin[0.041]; Altin[0.041]; Bogdani[0.041]; takes[0.041]; Erjon[0.041]; Forwards[0.041]; Rudi[0.041]; suspension[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arjan Xhumba\u001b[39m ==> ENTITY: \u001b[32mArjan Xhumba\u001b[39m\t\nSCORES: global= 0.226:0.226[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArjan[0.047]; Albanian[0.046]; squad[0.045]; Fatmir[0.044]; Malko[0.043]; Haxhi[0.041]; Midfielders[0.041]; Dervishi[0.041]; Hafizi[0.041]; Hafizi[0.041]; Saturday[0.041]; Northern[0.041]; defender[0.041]; Ireland[0.041]; Saimir[0.041]; sports[0.041]; Vata[0.040]; Vata[0.040]; match[0.040]; draw[0.040]; Afrim[0.040]; Erjon[0.040]; win[0.040]; Ilir[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belfast ==> ENTITY: \u001b[32mBelfast\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.120:0.120[0]; log p(e|m)= -0.131:-0.131[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelfast[0.048]; Ireland[0.047]; Ireland[0.047]; soccer[0.044]; soccer[0.044]; match[0.043]; Cup[0.043]; squad[0.042]; players[0.042]; sports[0.042]; Football[0.042]; Saturday[0.041]; Saturday[0.041]; win[0.041]; qualifier[0.041]; country[0.040]; country[0.040]; Squad[0.039]; defender[0.039]; named[0.039]; group[0.039]; Northern[0.039]; Northern[0.039]; draw[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Viktor Paco\u001b[39m ==> ENTITY: \u001b[32mViktor Paço\u001b[39m\t\nSCORES: global= 0.219:0.219[0]; local()= 0.088:0.088[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArtur[0.045]; Malko[0.044]; squad[0.043]; Erjon[0.042]; Haxhi[0.042]; Midfielders[0.042]; Goalkeepers[0.042]; Hafizi[0.042]; Hafizi[0.042]; Northern[0.042]; Arjan[0.042]; Arjan[0.042]; Afrim[0.041]; Vata[0.041]; Vata[0.041]; defender[0.041]; Altin[0.041]; Altin[0.041]; Ilir[0.041]; Bogdani[0.041]; Forwards[0.041]; Rudi[0.040]; suspension[0.040]; Fatmir[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = FIFA\u001b[39m ==> ENTITY: \u001b[32mFIFA\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.045]; soccer[0.045]; Soccer[0.044]; Fifa[0.044]; Football[0.044]; players[0.043]; Ireland[0.043]; Ireland[0.043]; sports[0.042]; country[0.042]; country[0.042]; qualifier[0.042]; Cup[0.041]; banned[0.041]; Play[0.041]; World[0.040]; Association[0.040]; reinstated[0.040]; Albania[0.039]; Albania[0.039]; Albania[0.039]; ban[0.039]; suspension[0.039]; committee[0.039]; \t\n\r [========>.....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s731ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 974/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1380testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Abu Dhabi ==> ENTITY: \u001b[32mAbu Dhabi\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.135:0.135[0]; log p(e|m)= -0.066:-0.066[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.052]; match[0.050]; match[0.050]; Cup[0.049]; Cup[0.049]; final[0.048]; Korea[0.048]; Korea[0.048]; Korea[0.048]; Korea[0.048]; fifth[0.048]; Saturday[0.048]; minutes[0.046]; spot[0.046]; spot[0.046]; Asian[0.046]; Asian[0.046]; Indonesia[0.046]; Indonesia[0.046]; allowed[0.046]; win[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hendro Kartiko\u001b[39m ==> ENTITY: \u001b[32mHendro Kartiko\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.045]; Indonesia[0.044]; Indonesia[0.044]; Indonesia[0.044]; Kuwait[0.043]; Hendro[0.043]; Ki[0.042]; Koreans[0.042]; Koreans[0.042]; South[0.042]; goalkeeper[0.041]; Suwandi[0.041]; Putra[0.041]; Indonesians[0.041]; Korea[0.041]; Seok[0.041]; Widodo[0.041]; Widodo[0.041]; kick[0.040]; Chul[0.040]; Indonesian[0.040]; Young[0.040]; Byung[0.040]; Byung[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mKorean\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mKorea\u001b[39m <---> \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.252:0.251[\u001b[31m0.001\u001b[39m]; local()= 0.132:0.144[\u001b[32m0.011\u001b[39m]; log p(e|m)= -1.754:-2.703[\u001b[31m0.949\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorean[0.048]; Koreans[0.047]; Koreans[0.047]; Koreans[0.047]; match[0.047]; second[0.045]; seventh[0.045]; Jeong[0.044]; beat[0.043]; Hwang[0.043]; Kim[0.043]; Byung[0.042]; scored[0.042]; scored[0.042]; Indonesia[0.042]; Indonesia[0.042]; Indonesia[0.042]; Kuwait[0.042]; Hong[0.041]; Wednesday[0.041]; goal[0.041]; Indonesian[0.041]; score[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSouth Korea\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSouth Korea\u001b[39m <---> \u001b[32mSouth Korea national football team\u001b[39m\t\nSCORES: global= 0.264:0.241[\u001b[31m0.024\u001b[39m]; local()= 0.180:0.203[\u001b[32m0.023\u001b[39m]; log p(e|m)= -0.224:-2.847[\u001b[31m2.623\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.046]; Korea[0.046]; Korea[0.046]; fifth[0.044]; win[0.044]; match[0.043]; match[0.043]; seventh[0.043]; Soccer[0.042]; finals[0.042]; final[0.042]; second[0.041]; Koreans[0.041]; Koreans[0.041]; South[0.041]; South[0.041]; South[0.041]; Asian[0.040]; Asian[0.040]; Cup[0.039]; Cup[0.039]; Jeong[0.039]; Hwang[0.039]; Kim[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSouth Korea\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSouth Korea\u001b[39m <---> \u001b[32mSouth Korea national football team\u001b[39m\t\nSCORES: global= 0.264:0.241[\u001b[31m0.023\u001b[39m]; local()= 0.171:0.198[\u001b[32m0.027\u001b[39m]; log p(e|m)= -0.224:-2.847[\u001b[31m2.623\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.046]; Korea[0.046]; Korea[0.046]; fifth[0.044]; win[0.044]; match[0.043]; match[0.043]; seventh[0.043]; Soccer[0.042]; finals[0.042]; final[0.042]; second[0.042]; Koreans[0.041]; South[0.041]; South[0.041]; South[0.041]; Asian[0.040]; Asian[0.040]; Cup[0.040]; Cup[0.040]; Hwang[0.039]; Kim[0.038]; Hong[0.038]; score[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hendro Kartiko\u001b[39m ==> ENTITY: \u001b[32mHendro Kartiko\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.045]; Indonesia[0.044]; Indonesia[0.044]; Indonesia[0.044]; Kuwait[0.043]; Hendro[0.043]; Ki[0.042]; Koreans[0.042]; Koreans[0.042]; South[0.042]; goalkeeper[0.041]; Suwandi[0.041]; Putra[0.041]; Indonesians[0.041]; Korea[0.041]; Seok[0.041]; Widodo[0.041]; Widodo[0.041]; kick[0.040]; kick[0.040]; Indonesian[0.040]; Young[0.040]; Byung[0.040]; Byung[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hwang Sun Hong\u001b[39m ==> ENTITY: \u001b[32mHwang Sun-Hong\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.047]; Koreans[0.044]; Kim[0.044]; Kim[0.044]; Kim[0.044]; Kim[0.044]; Huh[0.043]; Ki[0.042]; Joo[0.042]; Keun[0.041]; Lee[0.041]; Jeong[0.041]; Tae[0.041]; Tae[0.041]; Suwandi[0.041]; Young[0.041]; Ji[0.040]; Byung[0.040]; low[0.040]; South[0.040]; Seok[0.040]; Widodo[0.040]; string[0.040]; Sung[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mKuwait\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mKuwait\u001b[39m <---> \u001b[32mKuwait national football team\u001b[39m\t\nSCORES: global= 0.255:0.249[\u001b[31m0.006\u001b[39m]; local()= 0.084:0.136[\u001b[32m0.052\u001b[39m]; log p(e|m)= -0.337:-2.364[\u001b[31m2.028\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.045]; Korea[0.045]; match[0.045]; score[0.042]; direct[0.042]; half[0.042]; beat[0.042]; failed[0.042]; failed[0.042]; Pan[0.042]; prevent[0.042]; goal[0.041]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Huh[0.041]; Koreans[0.041]; Koreans[0.041]; Teams[0.041]; minute[0.040]; Korean[0.040]; Korean[0.040]; Woon[0.040]; started[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIndonesia\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIndonesia\u001b[39m <---> \u001b[32mIndonesia national football team\u001b[39m\t\nSCORES: global= 0.262:0.248[\u001b[31m0.014\u001b[39m]; local()= 0.148:0.220[\u001b[32m0.072\u001b[39m]; log p(e|m)= -0.232:-2.489[\u001b[31m2.257\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.052]; match[0.046]; match[0.046]; Korea[0.045]; Korea[0.045]; Korea[0.045]; Korea[0.045]; Asian[0.044]; Asian[0.044]; Cup[0.043]; Cup[0.043]; finals[0.043]; Soccer[0.043]; second[0.043]; fifth[0.042]; win[0.042]; South[0.041]; South[0.041]; South[0.041]; South[0.041]; Koreans[0.041]; final[0.041]; seventh[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Agung Setyabudi\u001b[39m ==> ENTITY: \u001b[32mAgung Setyabudi\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.045]; Indonesia[0.045]; Indonesia[0.045]; Indonesian[0.044]; Putra[0.044]; Hendro[0.044]; Hendro[0.044]; Widodo[0.042]; Widodo[0.042]; Marzuki[0.042]; South[0.041]; Ki[0.041]; Bima[0.041]; goalkeeper[0.041]; Indonesians[0.041]; match[0.041]; Suwandi[0.040]; Wednesday[0.040]; Pan[0.040]; Seok[0.040]; keeper[0.040]; keeper[0.040]; kick[0.039]; Chul[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSouth Korea\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSouth Korea\u001b[39m <---> \u001b[32mSouth Korea national football team\u001b[39m\t\nSCORES: global= 0.265:0.242[\u001b[31m0.023\u001b[39m]; local()= 0.163:0.197[\u001b[32m0.034\u001b[39m]; log p(e|m)= -0.224:-2.847[\u001b[31m2.623\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.047]; Korea[0.047]; Korea[0.047]; fifth[0.044]; win[0.044]; match[0.044]; match[0.044]; Soccer[0.043]; finals[0.043]; final[0.042]; South[0.041]; South[0.041]; South[0.041]; Asian[0.041]; Asian[0.041]; Cup[0.040]; Cup[0.040]; Hwang[0.039]; Kim[0.039]; Hong[0.039]; minute[0.038]; quarter[0.038]; scoring[0.038]; Group[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIndonesia\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIndonesia\u001b[39m <---> \u001b[32mIndonesia national football team\u001b[39m\t\nSCORES: global= 0.264:0.244[\u001b[31m0.020\u001b[39m]; local()= 0.185:0.187[\u001b[32m0.002\u001b[39m]; log p(e|m)= -0.232:-2.489[\u001b[31m2.257\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.051]; Indonesia[0.051]; Indonesian[0.046]; match[0.045]; Indonesians[0.044]; second[0.042]; Putra[0.041]; Koreans[0.041]; Koreans[0.041]; Koreans[0.041]; Widodo[0.040]; Widodo[0.040]; heading[0.040]; Teams[0.040]; Korean[0.040]; Korean[0.040]; Kuwait[0.040]; seventh[0.040]; spot[0.040]; Wednesday[0.040]; failed[0.040]; failed[0.040]; Hendro[0.039]; Hendro[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Asian Cup ==> ENTITY: \u001b[32mAFC Asian Cup\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.210:0.210[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.045]; Asian[0.044]; Korea[0.043]; Korea[0.043]; Korea[0.043]; Korea[0.043]; finals[0.043]; Saturday[0.043]; final[0.043]; Soccer[0.043]; match[0.042]; match[0.042]; second[0.042]; Koreans[0.041]; fifth[0.041]; win[0.041]; penalty[0.040]; seventh[0.040]; Indonesia[0.040]; Indonesia[0.040]; spot[0.039]; spot[0.039]; added[0.039]; conceding[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mKorean\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mKorea\u001b[39m <---> \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.252:0.251[\u001b[31m0.000\u001b[39m]; local()= 0.151:0.172[\u001b[32m0.021\u001b[39m]; log p(e|m)= -1.754:-2.703[\u001b[31m0.949\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.048]; Korean[0.046]; Koreans[0.045]; Koreans[0.045]; match[0.045]; second[0.043]; Jeong[0.042]; Jeong[0.042]; beat[0.041]; Kim[0.041]; Kim[0.041]; Kim[0.041]; Kim[0.041]; Byung[0.040]; Byung[0.040]; scored[0.040]; scored[0.040]; Indonesia[0.040]; Indonesia[0.040]; Indonesia[0.040]; Kuwait[0.040]; Ji[0.039]; South[0.039]; Pan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bima Sakti\u001b[39m ==> ENTITY: \u001b[32mBima Sakti\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.048]; Indonesia[0.048]; Indonesian[0.047]; Hendro[0.046]; Hendro[0.046]; Putra[0.045]; Ki[0.044]; Marzuki[0.044]; Suwandi[0.043]; Widodo[0.043]; Widodo[0.043]; Agung[0.043]; Seok[0.042]; kick[0.042]; Chul[0.042]; spectacular[0.042]; Byung[0.042]; Byung[0.042]; Korea[0.042]; Woon[0.041]; low[0.041]; goalkeeper[0.041]; Pan[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIndonesia\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIndonesia\u001b[39m <---> \u001b[32mIndonesia national football team\u001b[39m\t\nSCORES: global= 0.263:0.241[\u001b[31m0.022\u001b[39m]; local()= 0.209:0.193[\u001b[31m0.017\u001b[39m]; log p(e|m)= -0.232:-2.489[\u001b[31m2.257\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.052]; Indonesia[0.052]; Indonesian[0.047]; match[0.047]; Korea[0.045]; Indonesians[0.045]; Putra[0.044]; second[0.043]; Suwandi[0.043]; Hendro[0.042]; Hendro[0.042]; South[0.042]; Koreans[0.042]; Koreans[0.042]; Koreans[0.042]; Widodo[0.042]; Widodo[0.042]; heading[0.041]; Teams[0.041]; Korean[0.041]; Korean[0.041]; Kuwait[0.041]; Marzuki[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hwang Sun Hong\u001b[39m ==> ENTITY: \u001b[32mHwang Sun-Hong\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.045]; Korea[0.045]; Korea[0.045]; Cup[0.043]; Cup[0.043]; Asian[0.042]; Asian[0.042]; Korean[0.042]; Koreans[0.042]; Koreans[0.042]; second[0.042]; Kim[0.042]; Kim[0.042]; match[0.041]; match[0.041]; match[0.041]; scoring[0.041]; finals[0.041]; fifth[0.040]; win[0.040]; goals[0.040]; scored[0.040]; failed[0.040]; started[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Korea ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.047]; Korea[0.047]; Korea[0.047]; fifth[0.044]; win[0.044]; match[0.044]; match[0.044]; Soccer[0.043]; finals[0.042]; final[0.042]; Cup[0.042]; Cup[0.042]; South[0.041]; South[0.041]; South[0.041]; Group[0.041]; Asian[0.040]; Asian[0.040]; Final[0.040]; Kim[0.039]; quarter[0.038]; Saturday[0.038]; minute[0.038]; scoring[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Asian Cup ==> ENTITY: \u001b[32mAFC Asian Cup\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.213:0.213[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.045]; Asian[0.044]; Korea[0.044]; Korea[0.044]; Korea[0.044]; Korea[0.044]; finals[0.043]; Saturday[0.043]; final[0.043]; Soccer[0.043]; match[0.042]; match[0.042]; Koreans[0.041]; fifth[0.041]; win[0.041]; penalty[0.040]; seventh[0.040]; Indonesia[0.040]; Indonesia[0.040]; spot[0.039]; spot[0.039]; added[0.039]; conceding[0.039]; goals[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIndonesia\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIndonesia\u001b[39m <---> \u001b[32mIndonesia national football team\u001b[39m\t\nSCORES: global= 0.263:0.241[\u001b[31m0.022\u001b[39m]; local()= 0.207:0.188[\u001b[31m0.018\u001b[39m]; log p(e|m)= -0.232:-2.489[\u001b[31m2.257\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.050]; Indonesia[0.050]; Indonesian[0.046]; match[0.045]; Korea[0.044]; Indonesians[0.043]; Putra[0.042]; Koreans[0.042]; Koreans[0.042]; Suwandi[0.041]; Hendro[0.041]; Hendro[0.041]; Teams[0.040]; South[0.040]; Widodo[0.040]; Widodo[0.040]; Korean[0.040]; Korean[0.040]; Kuwait[0.039]; Marzuki[0.039]; Sakti[0.039]; goal[0.039]; Wednesday[0.039]; Agung[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesian ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.218:0.218[0]; log p(e|m)= -0.742:-0.742[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.053]; Indonesia[0.053]; Indonesia[0.053]; Indonesians[0.048]; Marzuki[0.043]; Koreans[0.042]; Koreans[0.042]; match[0.042]; Widodo[0.042]; Widodo[0.042]; South[0.042]; Korea[0.042]; Putra[0.042]; appeared[0.042]; Suwandi[0.041]; Hendro[0.041]; Hendro[0.041]; Agung[0.041]; Kuwait[0.041]; Bima[0.041]; Woon[0.041]; Woon[0.041]; Hong[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSouth Korea\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSouth Korea\u001b[39m <---> \u001b[32mSouth Korea national football team\u001b[39m\t\nSCORES: global= 0.267:0.238[\u001b[31m0.029\u001b[39m]; local()= 0.186:0.158[\u001b[31m0.028\u001b[39m]; log p(e|m)= -0.224:-2.847[\u001b[31m2.623\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSeok[0.047]; Hong[0.047]; Joo[0.047]; Korean[0.047]; Chul[0.046]; Jeong[0.046]; Sang[0.046]; Sang[0.046]; Byung[0.046]; Byung[0.046]; Koreans[0.046]; Huh[0.045]; Yoo[0.045]; Yong[0.045]; beat[0.045]; Tae[0.045]; Tae[0.045]; Ji[0.045]; Jin[0.044]; Hwang[0.044]; Keun[0.044]; Lee[0.044]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIndonesia\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIndonesia\u001b[39m <---> \u001b[32mIndonesia national football team\u001b[39m\t\nSCORES: global= 0.260:0.246[\u001b[31m0.014\u001b[39m]; local()= 0.141:0.215[\u001b[32m0.074\u001b[39m]; log p(e|m)= -0.232:-2.489[\u001b[31m2.257\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.049]; Asian[0.047]; Asian[0.047]; match[0.044]; match[0.044]; Korea[0.043]; Korea[0.043]; Korea[0.043]; Korea[0.043]; Cup[0.041]; Cup[0.041]; Koreans[0.041]; finals[0.041]; Soccer[0.041]; fifth[0.040]; win[0.040]; South[0.039]; South[0.039]; South[0.039]; South[0.039]; final[0.039]; seventh[0.039]; spot[0.039]; spot[0.039]; \t\n\r [=========>....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s643ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1000/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1166testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.175:0.175[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.049]; France[0.049]; France[0.049]; Germany[0.043]; Cup[0.043]; Cup[0.043]; Canada[0.042]; Canada[0.042]; Canada[0.042]; Sweden[0.042]; Norway[0.041]; competition[0.041]; points[0.041]; Russia[0.041]; World[0.040]; World[0.040]; Luc[0.040]; Finland[0.039]; Ryan[0.039]; Jeremie[0.039]; Tignes[0.039]; Jean[0.038]; Men[0.038]; Johnson[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Donna Weinbrecht\u001b[39m ==> ENTITY: \u001b[32mDonna Weinbrecht\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.119:0.119[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSatoya[0.046]; Canada[0.043]; Canada[0.043]; Canada[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; Gilg[0.043]; Karhu[0.043]; Johann[0.042]; Tatjana[0.041]; Friday[0.041]; Ann[0.041]; Troy[0.041]; Koroleva[0.041]; Jim[0.041]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Women[0.040]; Norway[0.040]; Men[0.040]; Finland[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.073:0.073[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.053]; Canada[0.053]; Canada[0.053]; Sweden[0.049]; Russia[0.048]; Russia[0.048]; Russia[0.048]; Germany[0.048]; Finland[0.047]; World[0.047]; World[0.047]; Norway[0.047]; Cup[0.047]; Cup[0.047]; Johnson[0.047]; Freestyle[0.046]; Ryan[0.046]; Japan[0.046]; Gauthier[0.045]; Satoya[0.045]; competition[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.051]; Men[0.048]; World[0.048]; World[0.048]; Germany[0.047]; competition[0.046]; Finland[0.046]; Freestyle[0.046]; Cup[0.046]; Cup[0.046]; Canada[0.045]; Canada[0.045]; Canada[0.045]; Satoya[0.044]; Women[0.044]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; skiing[0.043]; freestyle[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.052]; Canada[0.052]; Sweden[0.042]; Finland[0.042]; Cup[0.042]; Cup[0.042]; Women[0.042]; Satoya[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; World[0.041]; World[0.041]; competition[0.041]; Russia[0.040]; Russia[0.040]; Japan[0.040]; Men[0.040]; skiing[0.040]; Norway[0.040]; Germany[0.040]; Weinbrecht[0.040]; Skiing[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tae Satoya\u001b[39m ==> ENTITY: \u001b[32mTae Satoya\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.105:0.105[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmoguls[0.046]; skiing[0.046]; freestyle[0.045]; Women[0.045]; competition[0.045]; Gilg[0.045]; Karhu[0.044]; Finland[0.044]; World[0.044]; Johann[0.044]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Sweden[0.043]; Norway[0.043]; Japan[0.043]; Tatjana[0.043]; Men[0.042]; Troy[0.042]; Cup[0.042]; Weinbrecht[0.042]; Germany[0.041]; Jesper[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.085:0.085[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.048]; Canada[0.046]; Canada[0.046]; Canada[0.046]; Finland[0.044]; Johnson[0.044]; Norway[0.043]; Germany[0.043]; Jim[0.042]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; Satoya[0.041]; Troy[0.040]; Koroleva[0.040]; Japan[0.040]; Brassard[0.039]; Weinbrecht[0.039]; Elena[0.039]; Tatjana[0.039]; Ryan[0.039]; Gauthier[0.039]; Women[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFreestyle[0.045]; Canada[0.044]; Canada[0.044]; Canada[0.044]; Friday[0.044]; Sweden[0.044]; competition[0.044]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; Satoya[0.043]; Cup[0.043]; Cup[0.043]; skiing[0.043]; Finland[0.043]; moguls[0.043]; Norway[0.043]; Weinbrecht[0.042]; Russia[0.042]; Russia[0.042]; Russia[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.073:0.073[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.053]; Canada[0.053]; Canada[0.053]; Sweden[0.049]; Russia[0.048]; Russia[0.048]; Russia[0.048]; Germany[0.048]; Finland[0.047]; World[0.047]; World[0.047]; Norway[0.047]; Cup[0.047]; Cup[0.047]; Johnson[0.047]; Freestyle[0.046]; Ryan[0.046]; Japan[0.046]; Gauthier[0.045]; Satoya[0.045]; competition[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jean-Luc Brassard\u001b[39m ==> ENTITY: \u001b[32mJean-Luc Brassard\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.109:0.109[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSatoya[0.045]; Freestyle[0.044]; skiing[0.044]; Gauthier[0.044]; Tignes[0.044]; freestyle[0.044]; moguls[0.043]; Gilg[0.043]; Karhu[0.043]; Skiing[0.042]; Brassard[0.042]; Johann[0.042]; Canada[0.041]; Canada[0.041]; Canada[0.041]; competition[0.041]; Tatjana[0.041]; Troy[0.041]; Koroleva[0.040]; Norway[0.040]; Men[0.040]; Germany[0.039]; Jesper[0.039]; Women[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.077:0.077[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.047]; Canada[0.047]; Canada[0.047]; Sweden[0.044]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Germany[0.043]; Finland[0.042]; Norway[0.042]; Johnson[0.042]; Ryan[0.041]; Japan[0.041]; Gauthier[0.040]; Satoya[0.040]; France[0.040]; France[0.040]; France[0.040]; France[0.040]; Troy[0.039]; Jim[0.039]; Women[0.039]; Weinbrecht[0.038]; Karhu[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.107:0.107[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSatoya[0.044]; Cup[0.044]; skiing[0.043]; Canada[0.043]; Canada[0.043]; Canada[0.043]; Men[0.043]; Germany[0.043]; Norway[0.043]; Women[0.042]; World[0.042]; competition[0.042]; Sweden[0.041]; Jim[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; Finland[0.040]; freestyle[0.040]; points[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.202:0.202[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.049]; France[0.049]; France[0.049]; France[0.049]; Germany[0.042]; Cup[0.042]; Cup[0.042]; Canada[0.042]; Canada[0.042]; Canada[0.042]; Finland[0.041]; Sweden[0.041]; competition[0.041]; Norway[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; World[0.039]; World[0.039]; Japan[0.039]; Luc[0.039]; Ryan[0.038]; Tignes[0.038]; Jean[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.129:0.129[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.052]; Canada[0.052]; Sweden[0.042]; Finland[0.042]; Cup[0.042]; Cup[0.042]; Women[0.042]; Satoya[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; World[0.041]; World[0.041]; competition[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Japan[0.040]; Men[0.040]; skiing[0.040]; Norway[0.040]; Germany[0.040]; Weinbrecht[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.078:0.078[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.049]; Canada[0.046]; Canada[0.046]; Finland[0.045]; Norway[0.044]; Germany[0.044]; Jim[0.043]; France[0.042]; France[0.042]; France[0.042]; France[0.042]; Satoya[0.041]; Troy[0.041]; Koroleva[0.041]; Japan[0.040]; Brassard[0.040]; Weinbrecht[0.040]; Elena[0.040]; Tatjana[0.040]; Gauthier[0.039]; Women[0.039]; Jeremie[0.039]; Benson[0.039]; Karhu[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.202:0.202[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.049]; France[0.049]; France[0.049]; France[0.049]; Germany[0.042]; Cup[0.042]; Cup[0.042]; Canada[0.042]; Canada[0.042]; Canada[0.042]; Finland[0.041]; Sweden[0.041]; competition[0.041]; Norway[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; World[0.039]; World[0.039]; Japan[0.039]; Luc[0.039]; Ryan[0.038]; Tignes[0.038]; Jean[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Liz McIntyre\u001b[39m ==> ENTITY: \u001b[32mElizabeth McIntyre\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.110:0.110[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.045]; Canada[0.045]; Canada[0.045]; Norway[0.044]; Satoya[0.044]; Weinbrecht[0.043]; Gilg[0.043]; Karhu[0.043]; Johann[0.042]; Women[0.042]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Johnson[0.041]; Tatjana[0.041]; Finland[0.041]; Troy[0.041]; Koroleva[0.040]; Japan[0.040]; Gauthier[0.040]; Ann[0.040]; Germany[0.039]; Brassard[0.039]; Hole[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tatjana Mittermayer\u001b[39m ==> ENTITY: \u001b[32mTatjana Mittermayer\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmoguls[0.049]; skiing[0.048]; freestyle[0.048]; Satoya[0.047]; Karhu[0.047]; Freestyle[0.047]; Koroleva[0.047]; Tignes[0.047]; Gilg[0.046]; Skiing[0.046]; Germany[0.045]; Johann[0.045]; Finland[0.045]; Norway[0.044]; Sweden[0.044]; Weinbrecht[0.044]; Russia[0.044]; Russia[0.044]; Russia[0.044]; Troy[0.044]; competition[0.043]; Mogul[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.202:0.202[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.049]; France[0.049]; France[0.049]; France[0.049]; Germany[0.042]; Cup[0.042]; Cup[0.042]; Canada[0.042]; Canada[0.042]; Canada[0.042]; Finland[0.041]; Sweden[0.041]; competition[0.041]; Norway[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; World[0.039]; World[0.039]; Japan[0.039]; Luc[0.039]; Ryan[0.038]; Tignes[0.038]; Jean[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tignes\u001b[39m ==> ENTITY: \u001b[32mTignes\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nskiing[0.047]; freestyle[0.043]; Skiing[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; Gilg[0.042]; Karhu[0.042]; Cup[0.042]; Cup[0.042]; Freestyle[0.042]; competition[0.042]; Johann[0.042]; Friday[0.041]; Luc[0.041]; Tatjana[0.041]; Troy[0.040]; Jean[0.040]; Gauthier[0.040]; Norway[0.040]; Canada[0.040]; Canada[0.040]; Canada[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.047]; Finland[0.044]; Germany[0.044]; Cup[0.043]; Cup[0.043]; Canada[0.043]; Canada[0.043]; Canada[0.043]; Satoya[0.042]; skiing[0.042]; World[0.041]; World[0.041]; Jesper[0.041]; Men[0.041]; moguls[0.040]; competition[0.040]; Japan[0.040]; France[0.040]; France[0.040]; France[0.040]; France[0.040]; Karhu[0.040]; Andrei[0.040]; Freestyle[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.087:0.087[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.047]; Canada[0.047]; Canada[0.047]; Sweden[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Germany[0.043]; Finland[0.042]; Norway[0.042]; Johnson[0.042]; Ryan[0.041]; Japan[0.041]; Gauthier[0.040]; Satoya[0.040]; competition[0.040]; France[0.040]; France[0.040]; France[0.040]; France[0.040]; freestyle[0.040]; Friday[0.039]; Troy[0.039]; Jim[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.052]; France[0.052]; France[0.052]; Germany[0.045]; Canada[0.045]; Finland[0.044]; Norway[0.043]; Russia[0.043]; Russia[0.043]; Japan[0.042]; Weinbrecht[0.040]; Troy[0.040]; Jim[0.040]; Gauthier[0.040]; Jeremie[0.040]; Women[0.039]; Satoya[0.039]; Koroleva[0.039]; Collomb[0.038]; Candice[0.038]; Ann[0.038]; Donna[0.038]; Tatjana[0.037]; Liz[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Finland ==> ENTITY: \u001b[32mFinland\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.245:-0.245[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.048]; Norway[0.047]; Canada[0.044]; Canada[0.044]; Canada[0.044]; skiing[0.044]; Germany[0.043]; Karhu[0.042]; Cup[0.041]; competition[0.041]; Weinbrecht[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Jesper[0.041]; moguls[0.040]; Men[0.040]; Tignes[0.040]; Satoya[0.040]; World[0.040]; points[0.040]; freestyle[0.040]; Japan[0.039]; Women[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.052]; Canada[0.052]; Sweden[0.042]; Finland[0.042]; Cup[0.042]; Cup[0.042]; Women[0.042]; Satoya[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; World[0.040]; World[0.040]; competition[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Japan[0.040]; Men[0.040]; skiing[0.040]; Norway[0.040]; Germany[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.069:0.069[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.048]; Canada[0.048]; Canada[0.048]; Russia[0.044]; Russia[0.044]; Russia[0.044]; Finland[0.043]; Johnson[0.043]; Norway[0.042]; Ryan[0.042]; Satoya[0.041]; Gauthier[0.041]; Germany[0.041]; Troy[0.040]; Jim[0.040]; Japan[0.039]; Women[0.039]; Karhu[0.039]; Brassard[0.039]; France[0.039]; France[0.039]; France[0.039]; France[0.039]; Weinbrecht[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.101:0.101[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.045]; Canada[0.045]; Canada[0.045]; points[0.043]; Sweden[0.043]; Finland[0.043]; Johnson[0.042]; Andrei[0.042]; Cup[0.042]; Cup[0.042]; competition[0.042]; Norway[0.042]; Germany[0.042]; Jim[0.041]; Freestyle[0.041]; France[0.040]; France[0.040]; France[0.040]; France[0.040]; World[0.040]; World[0.040]; Satoya[0.040]; Troy[0.039]; Friday[0.039]; \t\n\r [=========>....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s504ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1028/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1287testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = USDA\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Agriculture\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.047:0.047[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNebraska[0.048]; Wk[0.046]; limited[0.043]; Omaha[0.043]; fed[0.043]; Steers[0.043]; lbs[0.042]; lbs[0.042]; Thursday[0.042]; Thursday[0.042]; cattle[0.042]; established[0.042]; Yr[0.041]; tested[0.041]; Holiday[0.041]; heifers[0.040]; Select[0.040]; Select[0.040]; Week[0.040]; Week[0.040]; Pd[0.040]; Pd[0.040]; Heifers[0.040]; early[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Omaha ==> ENTITY: \u001b[32mOmaha, Nebraska\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.041:0.041[0]; log p(e|m)= -0.330:-0.330[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNebraska[0.052]; roundup[0.048]; Week[0.047]; Week[0.047]; early[0.047]; Select[0.047]; Select[0.047]; Thursday[0.046]; Thursday[0.046]; established[0.046]; limited[0.046]; cattle[0.045]; Holiday[0.045]; Heifers[0.045]; Steers[0.044]; Slaughter[0.044]; Choice[0.043]; Choice[0.043]; steers[0.043]; heifers[0.043]; Wk[0.042]; Date[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = USDA\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Agriculture\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.047:0.047[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNebraska[0.048]; Wk[0.046]; limited[0.043]; Omaha[0.043]; fed[0.043]; Steers[0.043]; lbs[0.042]; lbs[0.042]; Thursday[0.042]; Thursday[0.042]; cattle[0.042]; established[0.042]; Yr[0.041]; tested[0.041]; Holiday[0.041]; heifers[0.040]; Select[0.040]; Select[0.040]; Week[0.040]; Week[0.040]; Pd[0.040]; Pd[0.040]; Heifers[0.040]; early[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nebraska ==> ENTITY: \u001b[32mNebraska\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.025:0.025[0]; log p(e|m)= -0.423:-0.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChoice[0.046]; Choice[0.046]; Omaha[0.045]; early[0.044]; Thursday[0.043]; Thursday[0.043]; steers[0.042]; Select[0.042]; Select[0.042]; Week[0.042]; Week[0.042]; roundup[0.042]; Slaughter[0.041]; Wk[0.041]; load[0.041]; established[0.041]; Steers[0.040]; cattle[0.040]; limited[0.040]; Holiday[0.040]; Heifers[0.040]; tested[0.039]; Yr[0.039]; Seller[0.038]; \t\n\r [=========>....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s598ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1032/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1319testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Greek ==> ENTITY: \u001b[32mGreece\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.101:0.101[0]; log p(e|m)= -2.323:-2.323[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGreece[0.062]; country[0.055]; government[0.048]; introduce[0.046]; promised[0.046]; port[0.045]; Port[0.045]; scheme[0.045]; Feb[0.045]; strike[0.045]; called[0.044]; closed[0.044]; giving[0.044]; Dec[0.044]; conditions[0.044]; kept[0.044]; workers[0.044]; ports[0.044]; Shipping[0.043]; update[0.042]; Lloyds[0.041]; bonus[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Greece ==> ENTITY: \u001b[32mGreece\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.102:0.102[0]; log p(e|m)= -0.276:-0.276[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGreek[0.058]; country[0.057]; government[0.051]; closed[0.051]; strike[0.049]; promised[0.049]; workers[0.049]; called[0.047]; Feb[0.047]; bonus[0.047]; Port[0.046]; port[0.046]; Dec[0.046]; kept[0.046]; introduce[0.045]; ports[0.045]; giving[0.045]; conditions[0.045]; Shipping[0.045]; scheme[0.043]; update[0.043]; \t\n\r [=========>....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s695ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1034/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1314testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moroccan ==> ENTITY: \u001b[32mMorocco\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.268:-0.268[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoroccan[0.050]; Moroccan[0.050]; Moroccan[0.050]; Paris[0.046]; Paris[0.046]; Paris[0.046]; Mohamed[0.044]; army[0.044]; Thursday[0.043]; night[0.043]; March[0.042]; bomb[0.042]; Friday[0.042]; Canadian[0.042]; colonel[0.042]; bombing[0.041]; married[0.041]; said[0.041]; said[0.041]; days[0.041]; train[0.041]; train[0.041]; woman[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moroccan ==> ENTITY: \u001b[32mMorocco\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.149:0.149[0]; log p(e|m)= -0.268:-0.268[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoroccan[0.048]; Moroccan[0.048]; Moroccan[0.048]; Paris[0.044]; Paris[0.044]; Paris[0.044]; Mohamed[0.042]; army[0.042]; Thursday[0.041]; night[0.041]; March[0.041]; bomb[0.041]; Friday[0.040]; Canadian[0.040]; colonel[0.040]; bombing[0.040]; married[0.040]; said[0.039]; said[0.039]; said[0.039]; days[0.039]; woman[0.039]; embassy[0.039]; embassy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moslem ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.117:0.117[0]; log p(e|m)= -1.214:-1.214[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndoctorate[0.047]; doctorate[0.047]; people[0.046]; fundamentalists[0.046]; Canadian[0.046]; Moroccan[0.045]; Moroccan[0.045]; said[0.044]; said[0.044]; said[0.044]; mathematics[0.044]; Algerian[0.044]; teacher[0.043]; March[0.043]; hoped[0.042]; married[0.042]; army[0.042]; Thursday[0.042]; student[0.042]; bomb[0.042]; newly[0.041]; days[0.041]; Friday[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Caledonia ==> ENTITY: \u001b[32mNew Caledonia\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.076:0.076[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseries[0.047]; Canadian[0.046]; Thursday[0.045]; March[0.045]; student[0.045]; student[0.045]; Friday[0.044]; people[0.044]; newly[0.044]; days[0.044]; married[0.043]; night[0.043]; injuries[0.043]; died[0.043]; died[0.043]; army[0.042]; said[0.042]; said[0.042]; said[0.042]; hoped[0.042]; staged[0.042]; Moslem[0.042]; colonel[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canadian ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.057:0.057[0]; log p(e|m)= -0.412:-0.412[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCaledonia[0.045]; people[0.043]; married[0.043]; series[0.043]; March[0.042]; Thursday[0.042]; Paris[0.042]; Paris[0.042]; Paris[0.042]; army[0.042]; days[0.041]; Friday[0.041]; Moslem[0.041]; hoped[0.041]; old[0.041]; said[0.041]; said[0.041]; said[0.041]; hour[0.041]; explosion[0.041]; colonel[0.041]; bore[0.041]; newly[0.040]; staged[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moroccan ==> ENTITY: \u001b[32mMorocco\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.268:-0.268[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoroccan[0.048]; Moroccan[0.048]; Moroccan[0.048]; Moslem[0.045]; Mohamed[0.043]; people[0.043]; Algerian[0.042]; army[0.041]; Paris[0.041]; Paris[0.041]; Paris[0.041]; Thursday[0.041]; person[0.041]; night[0.041]; series[0.040]; March[0.040]; bomb[0.040]; Friday[0.040]; Canadian[0.040]; colonel[0.039]; said[0.039]; said[0.039]; said[0.039]; married[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Paris ==> ENTITY: \u001b[32mParis\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.115:-0.115[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nParis[0.051]; Paris[0.051]; doctorate[0.047]; doctorate[0.047]; woman[0.045]; mathematics[0.044]; student[0.043]; student[0.043]; son[0.043]; March[0.043]; man[0.043]; Thursday[0.042]; victim[0.042]; night[0.042]; married[0.042]; old[0.042]; died[0.042]; died[0.042]; Friday[0.042]; Caledonia[0.041]; embassy[0.041]; embassy[0.041]; person[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moroccan ==> ENTITY: \u001b[32mMorocco\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.268:-0.268[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoroccan[0.047]; Moroccan[0.047]; Moroccan[0.047]; Paris[0.044]; Paris[0.044]; Paris[0.044]; people[0.042]; Algerian[0.042]; Moslem[0.042]; Mohamed[0.041]; army[0.041]; Thursday[0.041]; person[0.041]; night[0.040]; series[0.040]; March[0.040]; bomb[0.040]; Friday[0.040]; Canadian[0.039]; colonel[0.039]; said[0.039]; said[0.039]; said[0.039]; married[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Paris ==> ENTITY: \u001b[32mParis\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.138:0.138[0]; log p(e|m)= -0.115:-0.115[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nParis[0.049]; Paris[0.049]; woman[0.044]; married[0.043]; Thursday[0.042]; son[0.042]; series[0.042]; March[0.042]; man[0.042]; victim[0.041]; night[0.041]; Friday[0.041]; old[0.041]; died[0.041]; died[0.041]; staged[0.041]; hour[0.040]; Caledonia[0.040]; embassy[0.040]; embassy[0.040]; person[0.040]; said[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Algerian ==> ENTITY: \u001b[32mAlgeria\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.209:-0.209[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoroccan[0.046]; Moroccan[0.046]; Moslem[0.045]; people[0.043]; army[0.042]; bombings[0.042]; embassy[0.042]; embassy[0.042]; killed[0.041]; bombing[0.041]; Paris[0.041]; said[0.041]; said[0.041]; said[0.041]; Thursday[0.041]; Canadian[0.041]; series[0.041]; March[0.041]; Friday[0.041]; colonel[0.040]; newly[0.040]; days[0.040]; explosion[0.040]; married[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Paris ==> ENTITY: \u001b[32mParis\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.146:0.146[0]; log p(e|m)= -0.115:-0.115[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nParis[0.049]; Paris[0.049]; doctorate[0.046]; doctorate[0.046]; woman[0.043]; mathematics[0.043]; student[0.042]; student[0.042]; son[0.042]; man[0.041]; Thursday[0.041]; victim[0.041]; night[0.041]; married[0.041]; old[0.040]; died[0.040]; died[0.040]; Friday[0.040]; Caledonia[0.039]; embassy[0.039]; embassy[0.039]; person[0.039]; said[0.039]; said[0.039]; \t\n\r [=========>....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s970ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1045/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1193testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Amsterdam ==> ENTITY: \u001b[32mAmsterdam\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.064:-0.064[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmsterdam[0.049]; Kerkrade[0.045]; Dutch[0.044]; Dutch[0.044]; Waalwijk[0.043]; Doetinchem[0.043]; Eindhoven[0.043]; Soccer[0.043]; Enschede[0.042]; soccer[0.042]; Arnhem[0.042]; Tilburg[0.041]; won[0.041]; played[0.041]; played[0.041]; Friday[0.040]; tabulate[0.040]; Graafschap[0.040]; Willem[0.040]; Standings[0.039]; Standings[0.039]; Feyenoord[0.039]; goals[0.039]; goals[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Twente Enschede\u001b[39m ==> ENTITY: \u001b[32mFC Twente\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDoetinchem[0.045]; Feyenoord[0.045]; Soccer[0.044]; match[0.044]; Eindhoven[0.043]; soccer[0.043]; Kerkrade[0.043]; Ajax[0.043]; Groningen[0.043]; Heerenveen[0.042]; Waalwijk[0.042]; Dutch[0.041]; Dutch[0.041]; Utrecht[0.041]; played[0.041]; played[0.041]; won[0.040]; goals[0.040]; goals[0.040]; Graafschap[0.040]; division[0.040]; Nac[0.040]; Tilburg[0.039]; Tilburg[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = RKC Waalwijk\u001b[39m ==> ENTITY: \u001b[32mRKC Waalwijk\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.118:0.118[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKerkrade[0.057]; Heerenveen[0.054]; Alkmaar[0.053]; Groningen[0.052]; Utrecht[0.052]; Volendam[0.051]; Nac[0.051]; Breda[0.050]; Tilburg[0.050]; Roda[0.050]; Sittard[0.050]; Nijmegen[0.049]; Rotterdam[0.049]; Willem[0.048]; Sparta[0.048]; Jc[0.048]; Fortuna[0.047]; Ii[0.047]; Az[0.047]; Nec[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Willem II Tilburg\u001b[39m ==> ENTITY: \u001b[32mWillem II (football club)\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDoetinchem[0.048]; Feyenoord[0.046]; Ajax[0.044]; Twente[0.044]; Enschede[0.044]; Kerkrade[0.044]; Heerenveen[0.043]; Waalwijk[0.043]; Alkmaar[0.043]; Groningen[0.043]; Nijmegen[0.042]; Volendam[0.042]; Utrecht[0.041]; Graafschap[0.040]; Sparta[0.040]; Breda[0.040]; Vitesse[0.040]; Nac[0.040]; Sittard[0.040]; Amsterdam[0.039]; Roda[0.039]; Arnhem[0.039]; Rotterdam[0.039]; Jc[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Volendam ==> ENTITY: \u001b[32mFC Volendam\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.896:-0.896[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDoetinchem[0.045]; played[0.044]; Feyenoord[0.044]; Sittard[0.044]; Heerenveen[0.043]; Kerkrade[0.043]; Alkmaar[0.043]; Groningen[0.042]; Waalwijk[0.042]; Waalwijk[0.042]; Amsterdam[0.041]; Utrecht[0.041]; Ajax[0.041]; Twente[0.041]; Enschede[0.041]; Graafschap[0.041]; Nac[0.041]; Standings[0.041]; Eindhoven[0.041]; Arnhem[0.040]; Rotterdam[0.040]; won[0.040]; Nijmegen[0.040]; Vitesse[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Vitesse Arnhem\u001b[39m ==> ENTITY: \u001b[32mVitesse\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFeyenoord[0.045]; soccer[0.044]; Doetinchem[0.044]; Ajax[0.043]; Kerkrade[0.043]; Soccer[0.043]; Heerenveen[0.043]; Twente[0.042]; Groningen[0.042]; Waalwijk[0.042]; Dutch[0.042]; Dutch[0.042]; Alkmaar[0.042]; match[0.042]; Eindhoven[0.041]; Tilburg[0.040]; Tilburg[0.040]; won[0.040]; played[0.040]; played[0.040]; Utrecht[0.040]; Graafschap[0.040]; Nac[0.039]; Enschede[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = PSV Eindhoven\u001b[39m ==> ENTITY: \u001b[32mPSV Eindhoven\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFeyenoord[0.047]; Doetinchem[0.047]; Kerkrade[0.045]; Ajax[0.045]; soccer[0.045]; Twente[0.045]; Heerenveen[0.045]; match[0.044]; Enschede[0.044]; Soccer[0.044]; Waalwijk[0.043]; Standings[0.043]; Standings[0.043]; Dutch[0.043]; Dutch[0.043]; Utrecht[0.043]; won[0.042]; points[0.042]; played[0.042]; played[0.042]; Rotterdam[0.041]; goals[0.041]; goals[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = AZ Alkmaar\u001b[39m ==> ENTITY: \u001b[32mAZ Alkmaar\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDoetinchem[0.048]; Waalwijk[0.046]; Kerkrade[0.046]; Ajax[0.045]; Heerenveen[0.044]; Groningen[0.043]; Tilburg[0.042]; Graafschap[0.042]; Utrecht[0.042]; Sittard[0.042]; Nac[0.042]; Nijmegen[0.041]; Amsterdam[0.041]; Sparta[0.041]; Vitesse[0.040]; Arnhem[0.040]; Breda[0.040]; Volendam[0.040]; Rotterdam[0.040]; Roda[0.039]; Willem[0.039]; Jc[0.039]; Fortuna[0.038]; Ii[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sparta Rotterdam\u001b[39m ==> ENTITY: \u001b[32mSparta Rotterdam\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDoetinchem[0.047]; Feyenoord[0.044]; Waalwijk[0.043]; Ajax[0.043]; Heerenveen[0.043]; Kerkrade[0.043]; Nijmegen[0.043]; Groningen[0.043]; Eindhoven[0.043]; Enschede[0.042]; played[0.042]; Alkmaar[0.042]; Utrecht[0.042]; Twente[0.041]; goals[0.041]; goals[0.041]; Amsterdam[0.041]; Tilburg[0.040]; Volendam[0.040]; Graafschap[0.040]; Arnhem[0.039]; Sittard[0.039]; Nac[0.039]; Breda[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ajax Amsterdam\u001b[39m ==> ENTITY: \u001b[32mAFC Ajax\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDoetinchem[0.045]; Feyenoord[0.045]; Soccer[0.044]; soccer[0.043]; Amsterdam[0.043]; Kerkrade[0.043]; Eindhoven[0.042]; Twente[0.042]; match[0.042]; Utrecht[0.042]; Heerenveen[0.041]; Alkmaar[0.041]; Waalwijk[0.041]; played[0.041]; played[0.041]; Groningen[0.041]; won[0.041]; Enschede[0.041]; Division[0.041]; Dutch[0.040]; Dutch[0.040]; Sittard[0.040]; Nijmegen[0.040]; Graafschap[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dutch ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.794:-0.794[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDutch[0.047]; match[0.044]; Doetinchem[0.044]; Eindhoven[0.043]; soccer[0.043]; Feyenoord[0.043]; won[0.042]; played[0.042]; played[0.042]; Soccer[0.042]; Waalwijk[0.042]; Heerenveen[0.041]; Ajax[0.041]; Enschede[0.041]; Amsterdam[0.041]; Amsterdam[0.041]; goals[0.041]; goals[0.041]; Twente[0.041]; Tilburg[0.040]; lost[0.039]; Willem[0.039]; Arnhem[0.039]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = RKC Waalwijk\u001b[39m ==> ENTITY: \u001b[32mRKC Waalwijk\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDoetinchem[0.045]; soccer[0.044]; Kerkrade[0.044]; Ajax[0.044]; Feyenoord[0.044]; Soccer[0.043]; played[0.043]; played[0.043]; match[0.042]; Heerenveen[0.042]; Eindhoven[0.041]; goals[0.041]; goals[0.041]; Dutch[0.041]; Dutch[0.041]; Amsterdam[0.041]; Amsterdam[0.041]; Utrecht[0.040]; won[0.040]; Twente[0.040]; division[0.040]; Volendam[0.040]; Division[0.040]; Graafschap[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NAC Breda\u001b[39m ==> ENTITY: \u001b[32mNAC Breda\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDoetinchem[0.045]; Kerkrade[0.045]; Feyenoord[0.044]; Ajax[0.043]; Waalwijk[0.043]; Heerenveen[0.043]; Alkmaar[0.043]; Twente[0.043]; Enschede[0.042]; Eindhoven[0.042]; goals[0.042]; Groningen[0.042]; Sittard[0.041]; Graafschap[0.041]; Tilburg[0.041]; Nijmegen[0.041]; Utrecht[0.041]; Vitesse[0.040]; Volendam[0.040]; Arnhem[0.040]; Sparta[0.040]; Willem[0.040]; Amsterdam[0.039]; Rotterdam[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Heerenveen ==> ENTITY: \u001b[32mSC Heerenveen\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.239:0.239[0]; log p(e|m)= -0.635:-0.635[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAjax[0.047]; Doetinchem[0.047]; Feyenoord[0.046]; Waalwijk[0.045]; Waalwijk[0.045]; Kerkrade[0.045]; Alkmaar[0.045]; Groningen[0.045]; soccer[0.044]; Twente[0.044]; Eindhoven[0.044]; Soccer[0.043]; Enschede[0.043]; played[0.042]; played[0.042]; Utrecht[0.042]; Nijmegen[0.042]; Tilburg[0.042]; Tilburg[0.042]; Breda[0.042]; Dutch[0.041]; Dutch[0.041]; Sparta[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Feyenoord\u001b[39m ==> ENTITY: \u001b[32mFeyenoord\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDoetinchem[0.046]; Soccer[0.046]; Ajax[0.046]; Eindhoven[0.046]; Kerkrade[0.046]; soccer[0.045]; Heerenveen[0.045]; match[0.045]; Enschede[0.045]; Twente[0.045]; played[0.044]; played[0.044]; Waalwijk[0.043]; won[0.043]; Utrecht[0.043]; Dutch[0.042]; Dutch[0.042]; Graafschap[0.041]; division[0.041]; Nac[0.041]; goals[0.041]; goals[0.041]; Sparta[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NEC Nijmegen\u001b[39m ==> ENTITY: \u001b[32mN.E.C. (football club)\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.132:0.132[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKerkrade[0.052]; Heerenveen[0.052]; Waalwijk[0.051]; Ajax[0.051]; Groningen[0.051]; Alkmaar[0.050]; Utrecht[0.049]; Tilburg[0.048]; Nac[0.048]; Volendam[0.048]; Amsterdam[0.047]; Breda[0.047]; Rotterdam[0.047]; Sittard[0.047]; Sparta[0.046]; Roda[0.046]; Willem[0.045]; Jc[0.045]; Fortuna[0.044]; Ii[0.044]; Az[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Groningen ==> ENTITY: \u001b[32mFC Groningen\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.181:0.181[0]; log p(e|m)= -1.966:-1.966[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKerkrade[0.046]; Utrecht[0.045]; Enschede[0.045]; Sittard[0.044]; Doetinchem[0.044]; Tilburg[0.044]; Volendam[0.043]; Waalwijk[0.043]; Nijmegen[0.043]; Rotterdam[0.043]; Amsterdam[0.042]; Ajax[0.042]; Graafschap[0.042]; Heerenveen[0.042]; Breda[0.042]; Arnhem[0.041]; Alkmaar[0.041]; Twente[0.041]; Willem[0.040]; Roda[0.039]; Vitesse[0.038]; Sparta[0.038]; Fortuna[0.037]; Nac[0.035]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dutch ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.794:-0.794[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDutch[0.047]; match[0.044]; Kerkrade[0.044]; Doetinchem[0.044]; Eindhoven[0.043]; soccer[0.043]; Feyenoord[0.042]; won[0.042]; played[0.042]; played[0.042]; Soccer[0.042]; Waalwijk[0.042]; Heerenveen[0.041]; Ajax[0.041]; Enschede[0.041]; Amsterdam[0.041]; Amsterdam[0.041]; Twente[0.041]; Tilburg[0.040]; Utrecht[0.040]; lost[0.039]; Willem[0.039]; goals[0.039]; goals[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Roda JC Kerkrade\u001b[39m ==> ENTITY: \u001b[32mRoda JC Kerkrade\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.229:0.229[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDoetinchem[0.046]; Enschede[0.044]; Feyenoord[0.043]; Groningen[0.043]; Waalwijk[0.043]; Waalwijk[0.043]; match[0.043]; Ajax[0.043]; Soccer[0.042]; Eindhoven[0.042]; Twente[0.042]; Alkmaar[0.042]; Heerenveen[0.042]; Utrecht[0.041]; Nijmegen[0.041]; goals[0.041]; goals[0.041]; Sittard[0.041]; Tilburg[0.040]; Tilburg[0.040]; soccer[0.040]; played[0.040]; played[0.040]; won[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Willem II Tilburg\u001b[39m ==> ENTITY: \u001b[32mWillem II (football club)\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.223:0.223[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDoetinchem[0.047]; Feyenoord[0.045]; Ajax[0.043]; Twente[0.043]; Enschede[0.043]; Eindhoven[0.043]; soccer[0.043]; Kerkrade[0.043]; Heerenveen[0.042]; Waalwijk[0.042]; match[0.042]; played[0.041]; played[0.041]; Division[0.041]; Soccer[0.041]; Dutch[0.041]; Dutch[0.041]; Volendam[0.041]; division[0.040]; Utrecht[0.040]; won[0.040]; Graafschap[0.039]; goals[0.039]; goals[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fortuna Sittard\u001b[39m ==> ENTITY: \u001b[32mFortuna Sittard\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKerkrade[0.049]; Waalwijk[0.048]; Groningen[0.047]; Heerenveen[0.046]; Ajax[0.046]; Alkmaar[0.046]; Utrecht[0.045]; Amsterdam[0.044]; Nijmegen[0.044]; Tilburg[0.044]; Volendam[0.043]; Rotterdam[0.043]; Nac[0.043]; Breda[0.043]; Roda[0.043]; Arnhem[0.043]; Sparta[0.042]; Vitesse[0.042]; Willem[0.041]; Jc[0.040]; Ii[0.039]; Az[0.039]; Nec[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Utrecht ==> ENTITY: \u001b[32mFC Utrecht\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.202:0.202[0]; log p(e|m)= -2.244:-2.244[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGroningen[0.046]; Doetinchem[0.044]; Feyenoord[0.043]; Dutch[0.043]; Nijmegen[0.043]; Kerkrade[0.043]; Eindhoven[0.042]; soccer[0.042]; Alkmaar[0.041]; Sittard[0.041]; Ajax[0.041]; Arnhem[0.041]; Waalwijk[0.041]; Waalwijk[0.041]; Heerenveen[0.041]; Enschede[0.041]; Amsterdam[0.041]; Willem[0.041]; Willem[0.041]; match[0.041]; Tilburg[0.040]; Tilburg[0.040]; played[0.040]; played[0.040]; \t\n\r [=========>....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s821ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1067/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1344testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colorado ==> ENTITY: \u001b[32mColorado\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.023:0.023[0]; log p(e|m)= -0.269:-0.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWinner[0.045]; winner[0.045]; Saturday[0.044]; Alpine[0.043]; Vail[0.043]; Age[0.043]; Olympics[0.043]; won[0.043]; Championships[0.043]; Russia[0.042]; victories[0.042]; Skiing[0.042]; Lillehammer[0.041]; medal[0.041]; medal[0.041]; World[0.040]; World[0.040]; World[0.040]; World[0.040]; Cup[0.040]; Cup[0.040]; Cup[0.040]; Nation[0.040]; women[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.075:0.075[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympics[0.046]; medal[0.046]; medal[0.046]; won[0.045]; Championships[0.044]; victories[0.043]; Cup[0.042]; Cup[0.042]; Cup[0.042]; Svetlana[0.041]; Winter[0.041]; World[0.041]; World[0.041]; World[0.041]; World[0.041]; bronze[0.041]; Lillehammer[0.040]; winner[0.040]; Alpine[0.040]; women[0.040]; silver[0.040]; Skiing[0.040]; Colorado[0.040]; Saturday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Cup ==> ENTITY: \u001b[32mFIS Alpine Ski World Cup\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.202:0.202[0]; log p(e|m)= -2.765:-2.765[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.045]; World[0.044]; World[0.044]; World[0.044]; Cup[0.044]; Cup[0.044]; Lillehammer[0.044]; medal[0.043]; medal[0.043]; Olympics[0.043]; downhill[0.042]; Skiing[0.042]; Championships[0.042]; won[0.042]; race[0.041]; Women[0.040]; Super[0.040]; women[0.040]; winner[0.039]; Vail[0.039]; victories[0.039]; Colorado[0.039]; Previous[0.039]; Winter[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lillehammer\u001b[39m ==> ENTITY: \u001b[32mLillehammer\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.137:0.137[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympics[0.051]; medal[0.045]; medal[0.045]; Alpine[0.044]; Championships[0.044]; Skiing[0.043]; bronze[0.042]; won[0.042]; Russia[0.042]; downhill[0.042]; Winter[0.041]; women[0.041]; Previous[0.040]; World[0.040]; World[0.040]; World[0.040]; World[0.040]; Saturday[0.040]; Cup[0.040]; Cup[0.040]; Cup[0.040]; race[0.039]; Women[0.039]; Facts[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Cup ==> ENTITY: \u001b[32mFIS Alpine Ski World Cup\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.202:0.202[0]; log p(e|m)= -2.765:-2.765[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.045]; World[0.044]; World[0.044]; World[0.044]; Cup[0.044]; Cup[0.044]; Lillehammer[0.044]; medal[0.043]; medal[0.043]; Olympics[0.043]; downhill[0.042]; Skiing[0.042]; Championships[0.042]; won[0.042]; race[0.041]; Women[0.040]; Super[0.040]; women[0.040]; winner[0.039]; Vail[0.039]; victories[0.039]; Colorado[0.039]; Previous[0.039]; Winter[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vail ==> ENTITY: \u001b[32mVail, Colorado\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.392:-0.392[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.047]; Skiing[0.044]; Saturday[0.044]; Olympics[0.043]; race[0.043]; downhill[0.043]; Colorado[0.043]; women[0.043]; Championships[0.042]; Winter[0.042]; Lillehammer[0.042]; won[0.042]; Winner[0.041]; Women[0.041]; World[0.040]; World[0.040]; World[0.040]; World[0.040]; Cup[0.040]; Cup[0.040]; Cup[0.040]; winner[0.040]; Profile[0.040]; Profile[0.040]; \t\n\r [==========>...................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s832ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1075/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1185testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = National stadium\u001b[39m ==> ENTITY: \u001b[32mNational stadium\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.088:0.088[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.047]; second[0.045]; final[0.044]; Cup[0.044]; Cup[0.044]; Cup[0.044]; Cup[0.044]; leg[0.043]; Zaire[0.043]; won[0.043]; halftime[0.041]; Aly[0.041]; Scorers[0.041]; Egypt[0.041]; Cairo[0.041]; Friday[0.040]; African[0.040]; African[0.040]; aggregate[0.040]; Mohamed[0.040]; Ashour[0.040]; Win[0.039]; Result[0.039]; penalty[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mArab Contractors\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mArab Contractors SC\u001b[39m <---> \u001b[32mArab Contractors (company)\u001b[39m\t\nSCORES: global= 0.280:0.278[\u001b[31m0.002\u001b[39m]; local()= 0.159:0.101[\u001b[31m0.059\u001b[39m]; log p(e|m)= -0.166:0.000[\u001b[32m0.166\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMohamed[0.047]; Egypt[0.045]; Contractors[0.045]; Contractors[0.045]; Soccer[0.043]; Cairo[0.043]; leg[0.043]; stadium[0.042]; Arab[0.042]; Ashour[0.042]; African[0.041]; African[0.041]; Cup[0.041]; Cup[0.041]; Cup[0.041]; Cup[0.041]; halftime[0.040]; Ouda[0.040]; Aly[0.040]; Scorers[0.040]; National[0.040]; won[0.039]; Zaire[0.039]; Result[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Egypt ==> ENTITY: \u001b[32mEgypt\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.107:0.107[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArab[0.049]; Arab[0.049]; Cairo[0.045]; second[0.045]; final[0.043]; Mohamed[0.043]; Soccer[0.043]; won[0.042]; Ashour[0.042]; African[0.041]; African[0.041]; Cup[0.041]; Cup[0.041]; Cup[0.041]; Cup[0.041]; Win[0.041]; leg[0.041]; stadium[0.040]; Zaire[0.040]; penalty[0.039]; Aly[0.039]; National[0.038]; Friday[0.038]; Ouda[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mZaire\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mDR Congo national football team\u001b[39m <---> \u001b[32mDemocratic Republic of the Congo\u001b[39m\t\nSCORES: global= 0.260:0.245[\u001b[31m0.015\u001b[39m]; local()= 0.156:0.055[\u001b[31m0.101\u001b[39m]; log p(e|m)= -1.645:-2.002[\u001b[31m0.357\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.048]; Cup[0.045]; Cup[0.045]; Cup[0.045]; Cup[0.045]; African[0.045]; African[0.045]; final[0.044]; National[0.043]; won[0.043]; Egypt[0.042]; second[0.041]; penalty[0.041]; Cairo[0.040]; Scorers[0.040]; stadium[0.040]; Win[0.040]; aggregate[0.039]; leg[0.039]; Friday[0.039]; Mohamed[0.039]; halftime[0.038]; Arab[0.038]; Arab[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cairo ==> ENTITY: \u001b[32mCairo\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.152:-0.152[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstadium[0.048]; Egypt[0.048]; Arab[0.046]; Arab[0.046]; Soccer[0.044]; Ashour[0.043]; Mohamed[0.043]; won[0.043]; Cup[0.042]; Cup[0.042]; Cup[0.042]; Cup[0.042]; Friday[0.041]; second[0.041]; final[0.040]; African[0.040]; African[0.040]; National[0.040]; Zaire[0.040]; Ouda[0.039]; Winners[0.038]; Winners[0.038]; aggregate[0.038]; Aly[0.038]; \t\n\r [==========>...................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s864ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1080/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1361testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = British Columbia ==> ENTITY: \u001b[32mBritish Columbia\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.065:0.065[0]; log p(e|m)= -0.028:-0.028[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.045]; Cup[0.045]; Cup[0.045]; Cup[0.045]; weekend[0.043]; race[0.043]; run[0.043]; run[0.043]; run[0.043]; racing[0.042]; Saturday[0.042]; Saturday[0.042]; Friday[0.041]; Whistler[0.041]; Whistler[0.041]; Mountain[0.041]; consecutive[0.041]; scheduled[0.040]; scheduled[0.040]; Officials[0.040]; Alpine[0.040]; events[0.040]; runs[0.039]; Organisers[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWhistler Mountain\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mWhistler Blackcomb\u001b[39m <---> \u001b[32mWhistler Mountain\u001b[39m\t\nSCORES: global= 0.278:0.263[\u001b[31m0.015\u001b[39m]; local()= 0.209:0.120[\u001b[31m0.089\u001b[39m]; log p(e|m)= -0.514:0.000[\u001b[32m0.514\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.047]; Whistler[0.047]; ski[0.046]; downhill[0.043]; downhill[0.043]; downhill[0.043]; Skiing[0.042]; Canadian[0.042]; Bernd[0.041]; events[0.041]; weekend[0.041]; run[0.041]; run[0.041]; run[0.041]; run[0.041]; run[0.041]; slalom[0.040]; snow[0.040]; snow[0.040]; giant[0.040]; completed[0.040]; national[0.040]; originally[0.039]; Organisers[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Cup ==> ENTITY: \u001b[32mFIS Alpine Ski World Cup\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.218:0.218[0]; log p(e|m)= -2.765:-2.765[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.046]; ski[0.044]; downhill[0.043]; downhill[0.043]; Skiing[0.043]; World[0.043]; World[0.043]; World[0.043]; Cup[0.042]; Cup[0.042]; Cup[0.042]; race[0.041]; race[0.041]; day[0.041]; day[0.041]; consecutive[0.041]; events[0.041]; Whistler[0.041]; Whistler[0.041]; weekend[0.040]; racing[0.040]; training[0.040]; training[0.040]; training[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Cup ==> ENTITY: \u001b[32mFIS Alpine Ski World Cup\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.225:0.225[0]; log p(e|m)= -2.765:-2.765[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.045]; ski[0.044]; national[0.043]; downhill[0.043]; downhill[0.043]; downhill[0.043]; Skiing[0.042]; World[0.042]; World[0.042]; World[0.042]; Cup[0.042]; Cup[0.042]; Cup[0.042]; slalom[0.042]; race[0.041]; race[0.041]; day[0.041]; day[0.041]; consecutive[0.041]; events[0.040]; today[0.040]; Whistler[0.040]; Whistler[0.040]; weekend[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Cup ==> ENTITY: \u001b[32mFIS Alpine Ski World Cup\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.225:0.225[0]; log p(e|m)= -2.765:-2.765[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.045]; ski[0.044]; national[0.043]; downhill[0.043]; downhill[0.043]; downhill[0.043]; Skiing[0.042]; World[0.042]; World[0.042]; World[0.042]; Cup[0.042]; Cup[0.042]; Cup[0.042]; slalom[0.042]; race[0.041]; race[0.041]; day[0.041]; day[0.041]; consecutive[0.041]; events[0.040]; today[0.040]; Whistler[0.040]; Whistler[0.040]; weekend[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWhistler\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mWhistler Blackcomb\u001b[39m <---> \u001b[32mWhistler, British Columbia\u001b[39m\t\nSCORES: global= 0.267:0.265[\u001b[31m0.002\u001b[39m]; local()= 0.193:0.160[\u001b[31m0.033\u001b[39m]; log p(e|m)= -2.465:-0.664[\u001b[32m1.802\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.049]; Whistler[0.049]; ski[0.048]; Mountain[0.046]; downhill[0.046]; downhill[0.046]; Skiing[0.044]; events[0.043]; weekend[0.043]; run[0.043]; run[0.043]; run[0.043]; British[0.042]; snow[0.042]; World[0.042]; World[0.042]; World[0.042]; World[0.042]; completed[0.042]; Saturday[0.041]; Saturday[0.041]; Organisers[0.041]; runs[0.041]; \t\n\r [==========>...................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s918ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1086/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1258testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Greenspan\u001b[39m ==> ENTITY: \u001b[32mAlan Greenspan\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.137:0.137[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGreenspan[0.047]; Greenspan[0.047]; Federal[0.045]; financial[0.043]; Alan[0.043]; volatility[0.043]; worldwide[0.042]; Chairman[0.042]; recovery[0.042]; Reserve[0.042]; referring[0.041]; said[0.041]; sharply[0.041]; markets[0.041]; shares[0.041]; industrial[0.040]; Mexico[0.040]; Mexico[0.040]; comments[0.040]; index[0.040]; Dow[0.040]; stocks[0.040]; stocks[0.040]; abated[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alan Greenspan\u001b[39m ==> ENTITY: \u001b[32mAlan Greenspan\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGreenspan[0.047]; Greenspan[0.047]; Greenspan[0.047]; Federal[0.044]; financial[0.043]; rates[0.042]; volatility[0.042]; worldwide[0.042]; Chairman[0.042]; trade[0.041]; Treasury[0.041]; recovery[0.041]; Reserve[0.041]; employment[0.041]; broad[0.041]; referring[0.041]; said[0.041]; market[0.040]; data[0.040]; sharply[0.040]; markets[0.040]; shares[0.040]; stock[0.040]; industrial[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexico ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.183:-0.183[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.052]; Mexican[0.048]; Carlos[0.045]; Televisa[0.043]; earlier[0.042]; Friday[0.042]; week[0.042]; group[0.041]; Ponce[0.041]; Ponce[0.041]; York[0.041]; York[0.041]; significant[0.041]; Santander[0.040]; Telmex[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; takes[0.040]; long[0.040]; ended[0.040]; pesos[0.040]; lack[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = ADRs\u001b[39m ==> ENTITY: \u001b[32mAmerican depositary receipt\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrading[0.051]; stock[0.050]; market[0.048]; market[0.048]; market[0.048]; traded[0.046]; Treasury[0.046]; suffered[0.046]; volume[0.045]; healthy[0.045]; closed[0.045]; Telmex[0.044]; percent[0.044]; Traders[0.044]; Traders[0.044]; Traders[0.044]; long[0.044]; stocks[0.044]; stocks[0.044]; ended[0.044]; significant[0.044]; pesos[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Treasury ==> ENTITY: \u001b[32mUnited States Department of the Treasury\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.107:0.107[0]; log p(e|m)= -1.013:-1.013[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfinancial[0.045]; stock[0.044]; Federal[0.044]; shares[0.043]; trading[0.043]; bonds[0.043]; Alan[0.043]; debt[0.042]; bond[0.041]; market[0.041]; million[0.041]; ADRs[0.041]; assets[0.041]; traded[0.041]; traded[0.041]; stocks[0.041]; stocks[0.041]; stocks[0.041]; rates[0.041]; cent[0.040]; cent[0.040]; employment[0.040]; Reserve[0.040]; markets[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexican ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.090:0.090[0]; log p(e|m)= -0.397:-0.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexican[0.051]; Mexico[0.048]; Mexico[0.048]; Mexico[0.048]; worldwide[0.045]; financial[0.042]; market[0.041]; percent[0.041]; percent[0.041]; volume[0.041]; trade[0.040]; markets[0.040]; trading[0.040]; sharply[0.040]; group[0.040]; employment[0.040]; debt[0.040]; rates[0.039]; Volume[0.039]; lack[0.039]; lower[0.039]; hurt[0.039]; volatility[0.038]; unchanged[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dow ==> ENTITY: \u001b[32mDow Jones Industrial Average\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.223:0.223[0]; log p(e|m)= -1.519:-1.519[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclosed[0.046]; lows[0.045]; traded[0.044]; sharply[0.044]; stocks[0.044]; stocks[0.044]; stocks[0.044]; index[0.044]; lower[0.044]; financial[0.044]; trade[0.044]; volatility[0.044]; ended[0.043]; percent[0.043]; City[0.043]; average[0.043]; began[0.043]; rise[0.043]; million[0.042]; worldwide[0.042]; shares[0.042]; hit[0.042]; markets[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Telmex\u001b[39m ==> ENTITY: \u001b[32mTelmex\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.050]; Mexico[0.050]; Televisa[0.050]; Mexican[0.049]; pesos[0.046]; market[0.046]; market[0.046]; market[0.046]; debt[0.045]; Santander[0.045]; Carlos[0.045]; stock[0.045]; earlier[0.045]; week[0.044]; share[0.044]; firm[0.043]; Thursday[0.043]; percent[0.043]; buying[0.043]; respectively[0.043]; trading[0.043]; Volume[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexico ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.183:-0.183[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.052]; Mexican[0.048]; Mexican[0.048]; Televisa[0.043]; worldwide[0.042]; group[0.042]; trade[0.041]; began[0.041]; financial[0.041]; Thursday[0.041]; Telmex[0.041]; closed[0.040]; ended[0.040]; ended[0.040]; York[0.040]; pesos[0.040]; markets[0.040]; released[0.040]; percent[0.040]; percent[0.040]; unchanged[0.040]; begun[0.040]; market[0.040]; employment[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexico ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.100:0.100[0]; log p(e|m)= -0.183:-0.183[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.053]; Carlos[0.047]; week[0.043]; Ponce[0.043]; Ponce[0.043]; Santander[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; foreign[0.041]; market[0.041]; market[0.041]; Friday[0.041]; respectively[0.041]; certainly[0.040]; earlier[0.040]; share[0.040]; long[0.040]; analysts[0.040]; gradually[0.040]; firm[0.040]; toll[0.040]; relative[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmillion[0.046]; released[0.044]; worldwide[0.043]; volume[0.043]; Mexico[0.043]; Mexico[0.043]; Volume[0.043]; Federal[0.042]; Thursday[0.042]; Mexican[0.042]; stock[0.041]; began[0.041]; employment[0.041]; Alan[0.041]; financial[0.041]; noted[0.041]; Friday[0.041]; changed[0.041]; traded[0.040]; traded[0.040]; tentative[0.040]; closed[0.040]; closed[0.040]; data[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexico ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.073:0.073[0]; log p(e|m)= -0.183:-0.183[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.053]; Mexican[0.049]; City[0.044]; Friday[0.043]; worldwide[0.043]; tentative[0.042]; ended[0.041]; closed[0.041]; financial[0.041]; markets[0.041]; percent[0.041]; saw[0.041]; Alan[0.041]; Reserve[0.040]; upset[0.040]; points[0.040]; hit[0.040]; recovery[0.040]; lower[0.040]; average[0.040]; stepped[0.039]; said[0.039]; Jones[0.039]; initial[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexican ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.081:0.081[0]; log p(e|m)= -0.397:-0.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexican[0.053]; Mexico[0.050]; Mexico[0.050]; financial[0.044]; percent[0.043]; markets[0.042]; sharply[0.042]; worldwide[0.040]; lower[0.040]; hit[0.040]; tentative[0.040]; hurt[0.040]; volatility[0.040]; saw[0.040]; traded[0.040]; recovery[0.040]; said[0.040]; million[0.039]; referring[0.039]; Alan[0.039]; stocks[0.039]; stocks[0.039]; stocks[0.039]; closed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexican ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.087:0.087[0]; log p(e|m)= -0.397:-0.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.049]; Mexico[0.049]; Mexico[0.049]; Carlos[0.044]; market[0.042]; market[0.042]; market[0.042]; volume[0.042]; pesos[0.041]; trading[0.041]; group[0.041]; debt[0.041]; healthy[0.041]; Telmex[0.041]; week[0.040]; analysts[0.040]; Santander[0.040]; Volume[0.040]; plunge[0.040]; Televisa[0.040]; ADRs[0.039]; centavos[0.039]; centavos[0.039]; centavos[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Greenspan\u001b[39m ==> ENTITY: \u001b[32mAlan Greenspan\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGreenspan[0.047]; Greenspan[0.047]; Federal[0.044]; financial[0.043]; Alan[0.043]; rates[0.042]; volatility[0.042]; noted[0.042]; worldwide[0.042]; Chairman[0.042]; trade[0.041]; Treasury[0.041]; recovery[0.041]; Reserve[0.041]; employment[0.041]; broad[0.041]; referring[0.041]; said[0.041]; market[0.040]; data[0.040]; lack[0.040]; markets[0.040]; shares[0.040]; stock[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Federal Reserve\u001b[39m ==> ENTITY: \u001b[32mFederal Reserve System\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfinancial[0.046]; Chairman[0.044]; Treasury[0.044]; Greenspan[0.043]; Greenspan[0.043]; Greenspan[0.043]; Greenspan[0.043]; rates[0.042]; Thursday[0.042]; released[0.042]; said[0.041]; data[0.041]; trading[0.041]; referring[0.041]; Alan[0.041]; Jones[0.041]; assets[0.041]; unchanged[0.041]; trade[0.041]; employment[0.040]; recovery[0.040]; markets[0.040]; Friday[0.040]; million[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.060:0.060[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntraded[0.045]; York[0.045]; firm[0.044]; shares[0.044]; director[0.043]; ADRs[0.042]; week[0.042]; ended[0.042]; winners[0.042]; market[0.041]; market[0.041]; takes[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; noted[0.041]; fell[0.041]; stocks[0.041]; stocks[0.041]; Mexico[0.040]; Mexico[0.040]; lackluster[0.040]; Friday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.056:0.056[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrading[0.048]; stock[0.048]; traded[0.046]; York[0.046]; firm[0.045]; closed[0.044]; director[0.044]; ADRs[0.043]; Thursday[0.043]; week[0.043]; ended[0.043]; winners[0.043]; market[0.043]; market[0.043]; market[0.043]; takes[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; Treasury[0.042]; noted[0.042]; fell[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexico City ==> ENTITY: \u001b[32mMexico City\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.028:-0.028[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.048]; Mexican[0.046]; Mexican[0.046]; industrial[0.044]; said[0.042]; hit[0.042]; ended[0.042]; Friday[0.042]; million[0.042]; Alan[0.041]; stepped[0.041]; crash[0.041]; Federal[0.041]; average[0.041]; saw[0.041]; closed[0.041]; lower[0.040]; referring[0.040]; worldwide[0.040]; upset[0.040]; tentative[0.040]; percent[0.040]; points[0.040]; financial[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexico ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.104:0.104[0]; log p(e|m)= -0.183:-0.183[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.052]; Mexican[0.049]; Mexican[0.049]; worldwide[0.042]; group[0.042]; trade[0.042]; began[0.042]; financial[0.042]; Thursday[0.041]; said[0.041]; closed[0.040]; ended[0.040]; pesos[0.040]; markets[0.040]; released[0.040]; percent[0.040]; percent[0.040]; unchanged[0.040]; begun[0.040]; market[0.040]; employment[0.040]; Reserve[0.040]; regular[0.040]; million[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Santander ==> ENTITY: \u001b[32mSantander Group\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.168:0.168[0]; log p(e|m)= -1.650:-1.650[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexican[0.046]; Telmex[0.045]; Mexico[0.045]; Mexico[0.045]; Carlos[0.044]; Ponce[0.043]; Ponce[0.043]; week[0.042]; pesos[0.042]; market[0.042]; market[0.042]; firm[0.042]; shares[0.041]; group[0.041]; ADRs[0.041]; earlier[0.041]; Televisa[0.040]; debt[0.040]; Friday[0.040]; long[0.039]; share[0.039]; ended[0.039]; buying[0.039]; certainly[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Greenspan\u001b[39m ==> ENTITY: \u001b[32mAlan Greenspan\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGreenspan[0.049]; Greenspan[0.049]; Greenspan[0.049]; Federal[0.046]; financial[0.045]; Alan[0.044]; rates[0.044]; volatility[0.044]; worldwide[0.043]; Chairman[0.043]; trade[0.043]; recovery[0.043]; Reserve[0.043]; employment[0.042]; referring[0.042]; said[0.042]; data[0.042]; sharply[0.042]; markets[0.042]; shares[0.042]; industrial[0.041]; Mexico[0.041]; Mexico[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Televisa\u001b[39m ==> ENTITY: \u001b[32mTelevisa\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.118:0.118[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTelmex[0.046]; Mexico[0.045]; Mexico[0.045]; Mexican[0.045]; Carlos[0.044]; Thursday[0.042]; week[0.042]; Santander[0.042]; unchanged[0.041]; Friday[0.041]; Ponce[0.041]; Ponce[0.041]; broad[0.041]; pesos[0.041]; market[0.041]; market[0.041]; market[0.041]; headed[0.040]; firm[0.040]; ended[0.040]; group[0.040]; traded[0.040]; debt[0.040]; director[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.070:0.070[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmillion[0.046]; released[0.044]; worldwide[0.044]; volume[0.043]; Mexico[0.043]; Mexico[0.043]; Volume[0.043]; Federal[0.042]; Mexican[0.042]; Mexican[0.042]; changed[0.042]; began[0.041]; stock[0.041]; employment[0.041]; Alan[0.041]; financial[0.041]; noted[0.041]; pesos[0.041]; fell[0.041]; closed[0.040]; data[0.040]; Thursday[0.040]; ended[0.040]; market[0.040]; \t\n\r [==========>...................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s765ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1110/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1346testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colorado ==> ENTITY: \u001b[32mColorado\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.037:0.037[0]; log p(e|m)= -0.269:-0.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.046]; Sweden[0.046]; Alpine[0.046]; Vail[0.046]; won[0.045]; Carole[0.045]; Russia[0.044]; Pernilla[0.044]; finished[0.044]; Skiing[0.044]; Wins[0.044]; according[0.044]; came[0.043]; Saturday[0.043]; World[0.043]; World[0.043]; France[0.042]; women[0.042]; Cup[0.042]; Cup[0.042]; provisional[0.042]; race[0.042]; Svetlana[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pernilla Wiberg\u001b[39m ==> ENTITY: \u001b[32mPernilla Wiberg\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwon[0.047]; Alpine[0.046]; race[0.044]; Skiing[0.044]; Sweden[0.044]; second[0.043]; finished[0.043]; Vail[0.043]; women[0.043]; Cup[0.042]; Cup[0.042]; Colorado[0.042]; Wins[0.042]; World[0.041]; World[0.041]; provisional[0.040]; France[0.040]; Super[0.040]; Super[0.040]; results[0.039]; Russia[0.039]; Saturday[0.039]; came[0.038]; Svetlana[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.071:0.071[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.045]; Cup[0.045]; second[0.045]; Sweden[0.044]; won[0.044]; finished[0.043]; women[0.043]; Super[0.043]; Super[0.043]; Russia[0.043]; came[0.043]; World[0.042]; World[0.042]; provisional[0.042]; Saturday[0.042]; race[0.040]; Pernilla[0.040]; according[0.039]; Alpine[0.039]; Wins[0.039]; Colorado[0.038]; results[0.038]; Skiing[0.038]; Carole[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.070:0.070[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwon[0.048]; second[0.046]; Sweden[0.045]; finished[0.045]; provisional[0.044]; Cup[0.044]; Cup[0.044]; Wins[0.043]; Svetlana[0.042]; France[0.042]; World[0.042]; World[0.042]; Saturday[0.041]; Alpine[0.041]; race[0.041]; Skiing[0.040]; Colorado[0.040]; came[0.040]; Super[0.039]; Super[0.039]; according[0.038]; Pernilla[0.038]; women[0.038]; results[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Carole Montillet\u001b[39m ==> ENTITY: \u001b[32mCarole Montillet\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.045]; Skiing[0.045]; race[0.044]; Vail[0.043]; women[0.043]; Cup[0.043]; Cup[0.043]; Colorado[0.042]; Saturday[0.042]; Super[0.042]; Super[0.042]; finished[0.042]; provisional[0.042]; France[0.041]; World[0.041]; World[0.041]; won[0.041]; Pernilla[0.041]; results[0.040]; second[0.040]; Sweden[0.040]; Wins[0.040]; Wiberg[0.039]; Svetlana[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.095:0.095[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.047]; Cup[0.047]; won[0.046]; second[0.044]; World[0.043]; World[0.043]; finished[0.043]; women[0.042]; Pernilla[0.042]; France[0.042]; Colorado[0.041]; Wins[0.041]; Svetlana[0.041]; came[0.041]; Russia[0.041]; Skiing[0.040]; race[0.040]; provisional[0.040]; Alpine[0.040]; Super[0.040]; Super[0.040]; Saturday[0.040]; Wiberg[0.038]; according[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vail ==> ENTITY: \u001b[32mVail, Colorado\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.392:-0.392[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.048]; Skiing[0.045]; Saturday[0.045]; race[0.044]; according[0.044]; Colorado[0.044]; women[0.044]; second[0.043]; won[0.043]; finished[0.042]; World[0.041]; World[0.041]; Cup[0.041]; Cup[0.041]; came[0.041]; Carole[0.040]; Sweden[0.040]; results[0.040]; Wins[0.040]; Russia[0.039]; Super[0.039]; Super[0.039]; France[0.039]; provisional[0.038]; \t\n\r [==========>...................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s819ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1117/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1372testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Glasgow ==> ENTITY: \u001b[32mGlasgow\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.194:0.194[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScottish[0.045]; Scottish[0.045]; Aberdeen[0.044]; Fife[0.043]; Greenock[0.043]; Inverness[0.043]; Dundee[0.043]; Dundee[0.043]; league[0.042]; matches[0.042]; Airdrieonians[0.042]; Stirling[0.041]; League[0.041]; Ross[0.041]; Dunfermline[0.041]; Clydebank[0.041]; Kilmarnock[0.040]; Forfar[0.040]; Stenhousemuir[0.040]; played[0.040]; Stranraer[0.040]; Cup[0.040]; Partick[0.040]; County[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Inverness Thistle\u001b[39m ==> ENTITY: \u001b[32mInverness Thistle F.C.\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; Livingston[0.043]; Stenhousemuir[0.043]; Dundee[0.042]; Dundee[0.042]; Aberdeen[0.042]; Dunfermline[0.042]; matches[0.042]; Kilmarnock[0.042]; Stirling[0.042]; Ross[0.042]; Scottish[0.041]; Scottish[0.041]; Hibernian[0.041]; Glasgow[0.041]; Motherwell[0.041]; league[0.041]; Greenock[0.041]; Park[0.041]; Rangers[0.041]; Alloa[0.041]; Alloa[0.041]; played[0.041]; Falkirk[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scottish Cup ==> ENTITY: \u001b[32mScottish Cup\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.214:0.214[0]; log p(e|m)= -0.059:-0.059[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStenhousemuir[0.043]; Aberdeen[0.043]; Airdrieonians[0.043]; matches[0.042]; Dundee[0.042]; Dundee[0.042]; Alloa[0.042]; Alloa[0.042]; Dunfermline[0.042]; Inverness[0.042]; league[0.042]; Hibernian[0.042]; Hearts[0.042]; Greenock[0.041]; Kilmarnock[0.041]; Rangers[0.041]; Brechin[0.041]; Stirling[0.041]; Stranraer[0.041]; Park[0.041]; Thistle[0.041]; Celtic[0.041]; round[0.041]; cup[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dundee United\u001b[39m ==> ENTITY: \u001b[32mDundee United F.C.\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.206:0.206[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDundee[0.044]; Airdrieonians[0.044]; Stenhousemuir[0.044]; Livingston[0.044]; Aberdeen[0.043]; League[0.043]; Inverness[0.042]; matches[0.042]; league[0.042]; Hibernian[0.041]; Glasgow[0.041]; Rangers[0.041]; Dunfermline[0.041]; Celtic[0.041]; Kilmarnock[0.041]; Cup[0.041]; Cup[0.041]; Hearts[0.041]; Scottish[0.041]; Scottish[0.041]; Scottish[0.041]; played[0.040]; Ross[0.040]; Thistle[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Montrose ==> ENTITY: \u001b[32mMontrose F.C.\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.204:0.204[0]; log p(e|m)= -2.180:-2.180[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.047]; Greenock[0.045]; Stenhousemuir[0.044]; Brechin[0.044]; Inverness[0.044]; Scottish[0.044]; Scottish[0.044]; Scottish[0.044]; Aberdeen[0.044]; Fife[0.043]; Park[0.043]; Alloa[0.043]; Alloa[0.043]; Stirling[0.043]; Livingston[0.043]; matches[0.043]; Forfar[0.043]; Forfar[0.043]; Rangers[0.043]; League[0.043]; Dunfermline[0.043]; Thistle[0.042]; played[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hearts ==> ENTITY: \u001b[32mHeart of Midlothian F.C.\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.228:0.228[0]; log p(e|m)= -0.298:-0.298[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; League[0.043]; Aberdeen[0.043]; Dundee[0.043]; Dundee[0.043]; matches[0.043]; league[0.042]; Livingston[0.042]; Glasgow[0.042]; Hibernian[0.041]; Stenhousemuir[0.041]; Rangers[0.041]; Celtic[0.041]; played[0.041]; Inverness[0.041]; Scottish[0.041]; Scottish[0.041]; Scottish[0.041]; Dunfermline[0.041]; Motherwell[0.041]; Kilmarnock[0.041]; Park[0.040]; Stirling[0.040]; Falkirk[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clydebank ==> ENTITY: \u001b[32mClydebank F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.241:0.241[0]; log p(e|m)= -1.328:-1.328[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.047]; Aberdeen[0.045]; Dundee[0.045]; Dundee[0.045]; Stenhousemuir[0.044]; Greenock[0.044]; Dunfermline[0.044]; Fife[0.044]; Stirling[0.043]; Glasgow[0.043]; played[0.043]; Inverness[0.043]; Kilmarnock[0.043]; matches[0.043]; Alloa[0.043]; Alloa[0.043]; Brechin[0.043]; Partick[0.043]; Motherwell[0.043]; league[0.042]; League[0.042]; Livingston[0.042]; Stranraer[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Huntly ==> ENTITY: \u001b[32mHuntly F.C.\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.149:0.149[0]; log p(e|m)= -1.645:-1.645[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCounty[0.046]; Fife[0.044]; Kilmarnock[0.043]; Rangers[0.043]; Inverness[0.042]; Park[0.042]; Scottish[0.042]; Airdrieonians[0.042]; Clyde[0.042]; Stranraer[0.042]; Alloa[0.042]; Alloa[0.042]; Greenock[0.042]; Stirling[0.042]; Stenhousemuir[0.041]; Thistle[0.041]; Dundee[0.041]; Dundee[0.041]; Hawick[0.041]; Forfar[0.041]; Forfar[0.041]; Falkirk[0.040]; Queen[0.040]; Brechin[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clyde ==> ENTITY: \u001b[32mClyde F.C.\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.212:0.212[0]; log p(e|m)= -0.983:-0.983[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; County[0.044]; Livingston[0.043]; Dundee[0.043]; Dundee[0.043]; Stenhousemuir[0.043]; Park[0.042]; Stirling[0.042]; Rangers[0.042]; Partick[0.042]; Greenock[0.042]; Fife[0.041]; Thistle[0.041]; Brechin[0.041]; Celtic[0.041]; Inverness[0.041]; Motherwell[0.041]; Alloa[0.041]; Alloa[0.041]; Ross[0.041]; Falkirk[0.041]; Clydebank[0.040]; Albion[0.040]; Cup[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Stranraer ==> ENTITY: \u001b[32mStranraer F.C.\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.246:0.246[0]; log p(e|m)= -1.030:-1.030[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.044]; Livingston[0.043]; League[0.043]; Aberdeen[0.043]; Stenhousemuir[0.042]; Greenock[0.042]; Inverness[0.042]; Dundee[0.042]; Dundee[0.042]; Brechin[0.042]; Glasgow[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Johnstone[0.041]; Rangers[0.041]; Stirling[0.041]; Fife[0.041]; Dunfermline[0.041]; Thistle[0.041]; Partick[0.041]; Scottish[0.041]; Scottish[0.041]; Scottish[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Greenock Morton\u001b[39m ==> ENTITY: \u001b[32mGreenock Morton F.C.\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.228:0.228[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.044]; Stenhousemuir[0.043]; Aberdeen[0.043]; League[0.043]; Scottish[0.042]; Scottish[0.042]; Scottish[0.042]; Partick[0.042]; matches[0.042]; league[0.042]; played[0.042]; Dunfermline[0.042]; Kilmarnock[0.042]; Livingston[0.041]; Rangers[0.041]; Dundee[0.041]; Dundee[0.041]; Glasgow[0.041]; Hibernian[0.041]; Brechin[0.041]; Fife[0.041]; Inverness[0.041]; play[0.040]; Motherwell[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Celtic ==> ENTITY: \u001b[32mCeltic F.C.\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.978:-0.978[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInverness[0.044]; Fife[0.043]; Greenock[0.042]; Clydebank[0.042]; Stranraer[0.042]; Glasgow[0.042]; Brechin[0.042]; Stirling[0.042]; Stenhousemuir[0.042]; Dunfermline[0.041]; Airdrieonians[0.041]; matches[0.041]; Forfar[0.041]; Forfar[0.041]; Scottish[0.041]; Scottish[0.041]; Scottish[0.041]; Dundee[0.041]; Dundee[0.041]; Aberdeen[0.041]; Alloa[0.041]; Alloa[0.041]; Thistle[0.041]; Partick[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAlloa\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAlloa\u001b[39m <---> \u001b[32mAlloa Athletic F.C.\u001b[39m\t\nSCORES: global= 0.288:0.285[\u001b[31m0.004\u001b[39m]; local()= 0.176:0.220[\u001b[32m0.044\u001b[39m]; log p(e|m)= 0.000:-0.794[\u001b[31m0.794\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.044]; Stenhousemuir[0.043]; Stirling[0.043]; Alloa[0.043]; Livingston[0.043]; Aberdeen[0.043]; Dundee[0.043]; Dundee[0.043]; Greenock[0.042]; Brechin[0.042]; Dunfermline[0.042]; Raith[0.042]; Inverness[0.042]; played[0.041]; Kilmarnock[0.041]; Thistle[0.041]; Motherwell[0.041]; Partick[0.041]; Falkirk[0.041]; Stranraer[0.041]; Fife[0.040]; Forfar[0.040]; Forfar[0.040]; Scottish[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Raith\u001b[39m ==> ENTITY: \u001b[32mRaith Rovers F.C.\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.265:0.265[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.045]; Dundee[0.045]; Dundee[0.045]; Stenhousemuir[0.044]; Inverness[0.044]; league[0.044]; Dunfermline[0.044]; matches[0.044]; Aberdeen[0.044]; Kilmarnock[0.043]; Livingston[0.043]; Fife[0.043]; Rangers[0.043]; Glasgow[0.043]; Scottish[0.043]; Scottish[0.043]; Scottish[0.043]; Partick[0.043]; Celtic[0.043]; Greenock[0.043]; Motherwell[0.042]; League[0.042]; Stranraer[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = St Mirren\u001b[39m ==> ENTITY: \u001b[32mSt. Mirren F.C.\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.227:0.227[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.043]; matches[0.043]; league[0.043]; League[0.043]; Stenhousemuir[0.043]; played[0.043]; Dundee[0.043]; Dundee[0.043]; Aberdeen[0.042]; Livingston[0.042]; Dunfermline[0.042]; Kilmarnock[0.042]; Rangers[0.041]; Partick[0.041]; Celtic[0.041]; Hearts[0.041]; Hibernian[0.041]; Cup[0.041]; Cup[0.041]; play[0.041]; Park[0.041]; Brechin[0.040]; Albion[0.040]; Motherwell[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mHawick\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHawick\u001b[39m <---> \u001b[32mHawick RFC\u001b[39m\t\nSCORES: global= 0.270:0.260[\u001b[31m0.010\u001b[39m]; local()= 0.159:0.133[\u001b[31m0.026\u001b[39m]; log p(e|m)= 0.000:-1.682[\u001b[31m1.682\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.043]; Fife[0.043]; Park[0.043]; Scottish[0.043]; Stenhousemuir[0.043]; Stirling[0.042]; Greenock[0.042]; Inverness[0.042]; Cup[0.042]; Aberdeen[0.042]; Thistle[0.042]; County[0.042]; play[0.041]; Rangers[0.041]; Dundee[0.041]; Dundee[0.041]; Saturday[0.041]; Ross[0.041]; Forfar[0.041]; Forfar[0.041]; Clydebank[0.041]; December[0.041]; Alloa[0.040]; Alloa[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scottish ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.183:0.183[0]; log p(e|m)= -0.480:-0.480[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.045]; Scottish[0.044]; Stenhousemuir[0.043]; League[0.042]; Aberdeen[0.042]; Queen[0.042]; Hibernian[0.042]; Livingston[0.042]; matches[0.042]; Celtic[0.042]; Dundee[0.042]; Dundee[0.042]; league[0.042]; Dunfermline[0.041]; Rangers[0.041]; played[0.041]; Hearts[0.041]; Fife[0.041]; Kilmarnock[0.041]; Greenock[0.041]; Glasgow[0.041]; Inverness[0.040]; Falkirk[0.040]; Ross[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Forfar ==> ENTITY: \u001b[32mForfar Athletic F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.234:0.234[0]; log p(e|m)= -1.766:-1.766[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStenhousemuir[0.044]; Airdrieonians[0.043]; Inverness[0.042]; Aberdeen[0.042]; Scottish[0.042]; Scottish[0.042]; Fife[0.042]; Dundee[0.042]; Dundee[0.042]; Dunfermline[0.042]; Greenock[0.042]; league[0.042]; Brechin[0.041]; Alloa[0.041]; Alloa[0.041]; Partick[0.041]; Stranraer[0.041]; Kilmarnock[0.041]; League[0.041]; Forfar[0.041]; Stirling[0.041]; Clydebank[0.041]; Thistle[0.041]; Motherwell[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = East Fife\u001b[39m ==> ENTITY: \u001b[32mEast Fife F.C.\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; Aberdeen[0.046]; Stenhousemuir[0.045]; Dunfermline[0.044]; Dundee[0.044]; Dundee[0.044]; Kilmarnock[0.044]; matches[0.044]; Alloa[0.043]; Alloa[0.043]; Cup[0.043]; Cup[0.043]; Glasgow[0.043]; Partick[0.043]; Inverness[0.043]; Motherwell[0.043]; Brechin[0.043]; Clydebank[0.043]; Stirling[0.043]; Livingston[0.043]; Hibernian[0.042]; Stranraer[0.042]; Greenock[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hibernian\u001b[39m ==> ENTITY: \u001b[32mHibernian F.C.\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.223:0.223[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.045]; Aberdeen[0.044]; Inverness[0.043]; Livingston[0.043]; Dundee[0.042]; Dundee[0.042]; matches[0.042]; Scottish[0.042]; Scottish[0.042]; Scottish[0.042]; Celtic[0.042]; Stenhousemuir[0.042]; Glasgow[0.042]; League[0.041]; Dunfermline[0.041]; Kilmarnock[0.041]; league[0.041]; Falkirk[0.041]; Rangers[0.041]; Stirling[0.041]; played[0.040]; Hearts[0.040]; Fife[0.040]; Motherwell[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kilmarnock ==> ENTITY: \u001b[32mKilmarnock F.C.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.248:0.248[0]; log p(e|m)= -1.431:-1.431[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.045]; Aberdeen[0.043]; Dundee[0.043]; Dundee[0.043]; Stenhousemuir[0.043]; played[0.042]; Livingston[0.042]; matches[0.042]; Celtic[0.042]; Dunfermline[0.042]; Glasgow[0.042]; Hibernian[0.041]; Scottish[0.041]; Scottish[0.041]; Scottish[0.041]; League[0.041]; Motherwell[0.041]; Falkirk[0.041]; league[0.041]; Raith[0.041]; Brechin[0.041]; Partick[0.040]; Greenock[0.040]; Park[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brechin ==> ENTITY: \u001b[32mBrechin City F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.205:0.205[0]; log p(e|m)= -1.917:-1.917[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; Forfar[0.044]; Forfar[0.044]; Stenhousemuir[0.043]; Huntly[0.042]; Alloa[0.042]; Alloa[0.042]; League[0.041]; Partick[0.041]; Stranraer[0.041]; Aberdeen[0.041]; Inverness[0.041]; Scottish[0.041]; Scottish[0.041]; Scottish[0.041]; Dundee[0.041]; Dundee[0.041]; Thistle[0.041]; Dunfermline[0.041]; Kilmarnock[0.041]; Greenock[0.040]; Rangers[0.040]; Livingston[0.040]; Motherwell[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = St Johnstone\u001b[39m ==> ENTITY: \u001b[32mSt. Johnstone F.C.\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStenhousemuir[0.047]; Airdrieonians[0.046]; League[0.046]; Livingston[0.044]; Dundee[0.044]; Dundee[0.044]; Aberdeen[0.044]; Dunfermline[0.043]; league[0.043]; Partick[0.043]; Hearts[0.043]; Hibernian[0.043]; Kilmarnock[0.043]; Stirling[0.043]; Rangers[0.043]; Ross[0.043]; Raith[0.043]; Inverness[0.042]; Falkirk[0.042]; Albion[0.042]; matches[0.042]; Park[0.042]; Thistle[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Elgin City\u001b[39m ==> ENTITY: \u001b[32mElgin City F.C.\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.213:0.213[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.044]; Livingston[0.043]; Aberdeen[0.043]; Inverness[0.043]; Dundee[0.042]; Dundee[0.042]; Stenhousemuir[0.042]; Raith[0.042]; Dunfermline[0.042]; Fife[0.042]; Forfar[0.042]; Forfar[0.042]; Kilmarnock[0.042]; Scottish[0.042]; Rangers[0.041]; Greenock[0.041]; Hearts[0.041]; Partick[0.041]; Alloa[0.041]; Alloa[0.041]; Stranraer[0.041]; Brechin[0.041]; Ross[0.041]; Clydebank[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Whitehill ==> ENTITY: \u001b[32mWhitehill Welfare F.C.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.170:0.170[0]; log p(e|m)= -2.865:-2.865[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; Stenhousemuir[0.046]; County[0.045]; Raith[0.044]; Park[0.044]; Division[0.044]; Division[0.044]; Division[0.044]; Aberdeen[0.044]; Cup[0.043]; Celtic[0.043]; play[0.043]; Fife[0.043]; Hearts[0.043]; Clyde[0.043]; East[0.043]; Dunfermline[0.043]; Clydebank[0.043]; Dundee[0.042]; Dundee[0.042]; Inverness[0.042]; Thistle[0.042]; Albion[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Livingston ==> ENTITY: \u001b[32mLivingston F.C.\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.204:0.204[0]; log p(e|m)= -1.864:-1.864[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; Fife[0.043]; Aberdeen[0.043]; League[0.043]; Inverness[0.042]; league[0.042]; Dundee[0.042]; Dundee[0.042]; Stenhousemuir[0.042]; Kilmarnock[0.042]; Stirling[0.041]; Greenock[0.041]; County[0.041]; matches[0.041]; Hibernian[0.041]; Dunfermline[0.041]; Motherwell[0.041]; Falkirk[0.041]; Brechin[0.041]; Albion[0.041]; Scottish[0.041]; Scottish[0.041]; Scottish[0.041]; Glasgow[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Falkirk ==> ENTITY: \u001b[32mFalkirk F.C.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.203:0.203[0]; log p(e|m)= -1.204:-1.204[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.046]; Cup[0.046]; Stenhousemuir[0.044]; Airdrieonians[0.044]; Dunfermline[0.043]; League[0.043]; Fife[0.042]; Scottish[0.042]; Scottish[0.042]; Scottish[0.042]; Alloa[0.041]; Alloa[0.041]; Stranraer[0.041]; Brechin[0.041]; Inverness[0.041]; Motherwell[0.041]; Kilmarnock[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Premier[0.040]; Dundee[0.040]; Dundee[0.040]; Partick[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Motherwell ==> ENTITY: \u001b[32mMotherwell F.C.\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.227:0.227[0]; log p(e|m)= -0.335:-0.335[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.044]; Scottish[0.043]; Scottish[0.043]; Scottish[0.043]; Aberdeen[0.042]; Dundee[0.042]; Dundee[0.042]; Livingston[0.042]; Stenhousemuir[0.042]; Premier[0.042]; Glasgow[0.042]; Hibernian[0.042]; League[0.042]; Kilmarnock[0.041]; Dunfermline[0.041]; league[0.041]; matches[0.041]; played[0.041]; Celtic[0.041]; Rangers[0.041]; Park[0.041]; Brechin[0.041]; Fife[0.041]; Falkirk[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAlloa\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAlloa\u001b[39m <---> \u001b[32mAlloa Athletic F.C.\u001b[39m\t\nSCORES: global= 0.288:0.285[\u001b[31m0.003\u001b[39m]; local()= 0.169:0.221[\u001b[32m0.053\u001b[39m]; log p(e|m)= 0.000:-0.794[\u001b[31m0.794\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.044]; Stenhousemuir[0.043]; Stirling[0.043]; Alloa[0.043]; Livingston[0.043]; Aberdeen[0.042]; Dundee[0.042]; Dundee[0.042]; league[0.042]; Greenock[0.042]; Brechin[0.042]; Dunfermline[0.042]; Raith[0.041]; Inverness[0.041]; matches[0.041]; played[0.041]; Kilmarnock[0.041]; Glasgow[0.041]; Thistle[0.041]; Motherwell[0.041]; Partick[0.041]; Falkirk[0.041]; Stranraer[0.040]; Fife[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scottish ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.173:0.173[0]; log p(e|m)= -0.480:-0.480[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.045]; Scottish[0.044]; Stenhousemuir[0.043]; League[0.042]; Aberdeen[0.042]; Hibernian[0.042]; Livingston[0.042]; matches[0.042]; Celtic[0.042]; Dundee[0.042]; Dundee[0.042]; league[0.042]; Dunfermline[0.042]; Rangers[0.041]; played[0.041]; Hearts[0.041]; Fife[0.041]; Kilmarnock[0.041]; Greenock[0.041]; Glasgow[0.041]; Falkirk[0.040]; Ross[0.040]; division[0.040]; Motherwell[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dunfermline ==> ENTITY: \u001b[32mDunfermline Athletic F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.244:0.244[0]; log p(e|m)= -1.995:-1.995[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.044]; Aberdeen[0.044]; Dundee[0.043]; Dundee[0.043]; Stenhousemuir[0.043]; League[0.042]; played[0.042]; Kilmarnock[0.042]; Livingston[0.042]; matches[0.042]; Celtic[0.042]; league[0.041]; Fife[0.041]; Rangers[0.041]; Scottish[0.041]; Scottish[0.041]; Scottish[0.041]; Glasgow[0.041]; Hibernian[0.041]; Motherwell[0.041]; Falkirk[0.041]; Inverness[0.041]; Hearts[0.040]; Partick[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Partick ==> ENTITY: \u001b[32mPartick Thistle F.C.\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.246:0.246[0]; log p(e|m)= -1.284:-1.284[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.045]; Stenhousemuir[0.043]; Livingston[0.043]; Dundee[0.042]; Dundee[0.042]; League[0.042]; Kilmarnock[0.042]; Rangers[0.042]; Aberdeen[0.042]; Motherwell[0.042]; matches[0.042]; Dunfermline[0.042]; Glasgow[0.042]; Hearts[0.041]; Falkirk[0.041]; Celtic[0.041]; league[0.041]; Greenock[0.041]; Hibernian[0.041]; Thistle[0.041]; Inverness[0.040]; Stranraer[0.040]; Stirling[0.040]; Brechin[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ross County\u001b[39m ==> ENTITY: \u001b[32mRoss County F.C.\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.043]; League[0.043]; Aberdeen[0.043]; Fife[0.043]; Livingston[0.042]; Division[0.042]; Division[0.042]; Division[0.042]; Cup[0.042]; Cup[0.042]; league[0.042]; matches[0.042]; Scottish[0.042]; Scottish[0.042]; Scottish[0.042]; Inverness[0.041]; Dundee[0.041]; Dundee[0.041]; Dunfermline[0.041]; Stirling[0.041]; Rangers[0.041]; Greenock[0.040]; Glasgow[0.040]; Raith[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Airdrieonians\u001b[39m ==> ENTITY: \u001b[32mAirdrieonians F.C.\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAberdeen[0.046]; Stenhousemuir[0.045]; Livingston[0.044]; Dundee[0.044]; Dundee[0.044]; Dunfermline[0.044]; Brechin[0.044]; Kilmarnock[0.044]; Thistle[0.043]; Inverness[0.043]; matches[0.043]; Scottish[0.043]; Scottish[0.043]; Scottish[0.043]; Rangers[0.043]; Glasgow[0.043]; Partick[0.043]; League[0.043]; Stirling[0.043]; Stranraer[0.043]; Hibernian[0.043]; Motherwell[0.043]; league[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rangers ==> ENTITY: \u001b[32mRangers F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.211:0.211[0]; log p(e|m)= -0.302:-0.302[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; league[0.044]; matches[0.043]; League[0.043]; Aberdeen[0.042]; Glasgow[0.042]; Scottish[0.042]; Scottish[0.042]; Scottish[0.042]; Dundee[0.042]; Dundee[0.042]; Celtic[0.041]; played[0.041]; Dunfermline[0.041]; Livingston[0.041]; Hearts[0.041]; Hibernian[0.041]; play[0.041]; Soccer[0.041]; Cup[0.041]; Cup[0.041]; Thistle[0.040]; Kilmarnock[0.040]; Stenhousemuir[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Forfar ==> ENTITY: \u001b[32mForfar Athletic F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.231:0.231[0]; log p(e|m)= -1.766:-1.766[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStenhousemuir[0.044]; Airdrieonians[0.043]; Inverness[0.043]; Scottish[0.042]; Fife[0.042]; Dundee[0.042]; Dundee[0.042]; Greenock[0.042]; Brechin[0.042]; Alloa[0.042]; Alloa[0.042]; Partick[0.042]; Stranraer[0.042]; Kilmarnock[0.042]; Forfar[0.041]; Stirling[0.041]; Clydebank[0.041]; Thistle[0.041]; Motherwell[0.041]; Falkirk[0.041]; Johnstone[0.041]; Rangers[0.040]; Cup[0.040]; Raith[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Stenhousemuir\u001b[39m ==> ENTITY: \u001b[32mStenhousemuir F.C.\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; League[0.045]; Park[0.043]; league[0.043]; Cup[0.042]; Cup[0.042]; matches[0.042]; Stirling[0.042]; Dunfermline[0.042]; Fife[0.041]; Livingston[0.041]; played[0.041]; Aberdeen[0.041]; Inverness[0.041]; Rangers[0.041]; Thistle[0.041]; Alloa[0.041]; Alloa[0.041]; Scottish[0.041]; Scottish[0.041]; Scottish[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Aberdeen ==> ENTITY: \u001b[32mAberdeen F.C.\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.203:0.203[0]; log p(e|m)= -1.448:-1.448[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFife[0.044]; Scottish[0.043]; Scottish[0.043]; Scottish[0.043]; Airdrieonians[0.043]; matches[0.042]; Dundee[0.042]; Dundee[0.042]; league[0.042]; played[0.042]; Inverness[0.042]; League[0.042]; Glasgow[0.042]; Dunfermline[0.041]; Livingston[0.041]; Stenhousemuir[0.041]; Cup[0.041]; Cup[0.041]; Stirling[0.041]; Greenock[0.041]; Hibernian[0.040]; Motherwell[0.040]; Kilmarnock[0.040]; Rangers[0.040]; \t\n\r [==========>...................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s479ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1158/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1226testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.070:0.070[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.047]; Conservative[0.047]; parliamentary[0.045]; seat[0.045]; office[0.044]; office[0.044]; office[0.044]; office[0.044]; Conservatives[0.044]; Street[0.043]; vote[0.043]; John[0.043]; John[0.043]; Sir[0.043]; Mp[0.043]; Mp[0.043]; majority[0.043]; majority[0.043]; majority[0.043]; parliament[0.042]; Friday[0.042]; Minister[0.041]; party[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Major\u001b[39m ==> ENTITY: \u001b[32mJohn Major\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCommons[0.048]; Conservative[0.047]; Conservatives[0.045]; British[0.044]; parliamentary[0.042]; seat[0.041]; seat[0.041]; seat[0.041]; Major[0.041]; Major[0.041]; Major[0.041]; John[0.041]; John[0.041]; vote[0.041]; parliament[0.041]; parliament[0.041]; London[0.040]; formal[0.040]; majority[0.040]; majority[0.040]; majority[0.040]; majority[0.040]; resigned[0.040]; resigned[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mGorst\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mJohn Eldon Gorst\u001b[39m <---> \u001b[32mJohn Michael Gorst\u001b[39m\t\nSCORES: global= 0.297:0.280[\u001b[31m0.018\u001b[39m]; local()= 0.127:0.109[\u001b[31m0.018\u001b[39m]; log p(e|m)= 0.000:-0.012[\u001b[31m0.012\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCommons[0.045]; Commons[0.045]; hospital[0.044]; resigned[0.044]; constituency[0.043]; parliament[0.042]; seat[0.042]; seat[0.042]; said[0.042]; said[0.042]; government[0.042]; government[0.042]; withdrawing[0.041]; Gorst[0.041]; Gorst[0.041]; conference[0.041]; closure[0.041]; advised[0.040]; press[0.040]; today[0.040]; attend[0.040]; party[0.040]; Major[0.040]; week[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mDowning Street\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mDowning Street\u001b[39m <---> \u001b[32m10 Downing Street\u001b[39m\t\nSCORES: global= 0.289:0.282[\u001b[31m0.008\u001b[39m]; local()= 0.210:0.202[\u001b[31m0.008\u001b[39m]; log p(e|m)= 0.000:-1.168[\u001b[31m1.168\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCommons[0.047]; British[0.046]; office[0.044]; office[0.044]; office[0.044]; office[0.044]; London[0.043]; John[0.042]; John[0.042]; Conservative[0.042]; week[0.041]; statement[0.041]; statement[0.041]; Minister[0.040]; today[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; press[0.040]; seat[0.040]; seat[0.040]; seat[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mJohn Gorst\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mJohn Eldon Gorst\u001b[39m <---> \u001b[32mJohn Michael Gorst\u001b[39m\t\nSCORES: global= 0.298:0.281[\u001b[31m0.017\u001b[39m]; local()= 0.154:0.131[\u001b[31m0.023\u001b[39m]; log p(e|m)= 0.000:-0.012[\u001b[31m0.012\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJohn[0.044]; Conservative[0.044]; Conservatives[0.044]; parliamentary[0.043]; resigned[0.043]; resigned[0.043]; British[0.043]; Minister[0.042]; quit[0.042]; Sir[0.042]; parliament[0.041]; seat[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; office[0.041]; office[0.041]; office[0.041]; office[0.041]; rebel[0.041]; government[0.041]; government[0.041]; government[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mGorst\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mJohn Eldon Gorst\u001b[39m <---> \u001b[32mJohn Michael Gorst\u001b[39m\t\nSCORES: global= 0.299:0.282[\u001b[31m0.016\u001b[39m]; local()= 0.162:0.149[\u001b[31m0.013\u001b[39m]; log p(e|m)= 0.000:-0.012[\u001b[31m0.012\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJohn[0.046]; John[0.046]; Conservative[0.045]; Commons[0.045]; Commons[0.045]; Conservatives[0.045]; parliamentary[0.044]; resigned[0.044]; resigned[0.044]; British[0.044]; Minister[0.043]; quit[0.043]; Sir[0.043]; parliament[0.043]; parliament[0.043]; seat[0.042]; seat[0.042]; seat[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Major\u001b[39m ==> ENTITY: \u001b[32mJohn Major\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConservative[0.051]; Conservatives[0.049]; British[0.049]; parliamentary[0.047]; seat[0.046]; Major[0.045]; Major[0.045]; John[0.045]; John[0.045]; vote[0.045]; parliament[0.045]; London[0.045]; majority[0.044]; majority[0.044]; majority[0.044]; resigned[0.044]; resigned[0.044]; Prime[0.044]; office[0.044]; office[0.044]; office[0.044]; office[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = John Major\u001b[39m ==> ENTITY: \u001b[32mJohn Major\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConservative[0.052]; Conservatives[0.049]; British[0.049]; parliamentary[0.047]; seat[0.046]; Major[0.046]; Major[0.046]; John[0.045]; vote[0.045]; parliament[0.045]; London[0.045]; majority[0.045]; majority[0.045]; majority[0.045]; resigned[0.044]; resigned[0.044]; Prime[0.044]; office[0.044]; office[0.044]; office[0.044]; office[0.044]; conference[0.044]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mGorst\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mJohn Eldon Gorst\u001b[39m <---> \u001b[32mJohn Michael Gorst\u001b[39m\t\nSCORES: global= 0.299:0.281[\u001b[31m0.018\u001b[39m]; local()= 0.164:0.132[\u001b[31m0.032\u001b[39m]; log p(e|m)= 0.000:-0.012[\u001b[31m0.012\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCommons[0.044]; Commons[0.044]; parliamentary[0.043]; hospital[0.043]; resigned[0.043]; resigned[0.043]; resigned[0.043]; constituency[0.042]; quit[0.042]; parliament[0.041]; parliament[0.041]; seat[0.041]; seat[0.041]; seat[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; office[0.041]; office[0.041]; government[0.041]; government[0.041]; government[0.041]; government[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mGorst\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mJohn Eldon Gorst\u001b[39m <---> \u001b[32mJohn Michael Gorst\u001b[39m\t\nSCORES: global= 0.299:0.282[\u001b[31m0.016\u001b[39m]; local()= 0.161:0.149[\u001b[31m0.012\u001b[39m]; log p(e|m)= 0.000:-0.012[\u001b[31m0.012\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJohn[0.050]; John[0.050]; Conservative[0.050]; Commons[0.050]; Conservatives[0.049]; parliamentary[0.048]; resigned[0.048]; resigned[0.048]; British[0.048]; Minister[0.047]; quit[0.047]; Sir[0.047]; parliament[0.047]; parliament[0.047]; seat[0.047]; seat[0.047]; seat[0.047]; said[0.046]; said[0.046]; said[0.046]; said[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = British ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.100:0.100[0]; log p(e|m)= -0.489:-0.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJohn[0.049]; John[0.049]; London[0.049]; Sir[0.047]; parliament[0.047]; government[0.047]; government[0.047]; government[0.047]; Conservative[0.046]; parliamentary[0.045]; office[0.045]; office[0.045]; office[0.045]; office[0.045]; Major[0.044]; Major[0.044]; Major[0.044]; seat[0.044]; Conservatives[0.044]; Downing[0.043]; seen[0.043]; Minister[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Major\u001b[39m ==> ENTITY: \u001b[32mJohn Major\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCommons[0.055]; Commons[0.055]; parliamentary[0.049]; constituency[0.049]; seat[0.048]; seat[0.048]; seat[0.048]; Major[0.047]; vote[0.047]; vote[0.047]; parliament[0.047]; parliament[0.047]; formal[0.046]; majority[0.046]; majority[0.046]; majority[0.046]; resigned[0.046]; resigned[0.046]; office[0.046]; office[0.046]; conference[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Conservative ==> ENTITY: \u001b[32mConservative Party (UK)\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.197:0.197[0]; log p(e|m)= -0.559:-0.559[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConservatives[0.045]; British[0.044]; majority[0.043]; majority[0.043]; majority[0.043]; party[0.043]; party[0.043]; seat[0.043]; John[0.042]; John[0.042]; vote[0.042]; parliamentary[0.042]; government[0.041]; government[0.041]; government[0.041]; office[0.041]; office[0.041]; office[0.041]; office[0.041]; Sir[0.040]; London[0.040]; parliament[0.040]; resigned[0.040]; resigned[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mGorst\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mJohn Eldon Gorst\u001b[39m <---> \u001b[32mJohn Michael Gorst\u001b[39m\t\nSCORES: global= 0.298:0.280[\u001b[31m0.018\u001b[39m]; local()= 0.146:0.125[\u001b[31m0.021\u001b[39m]; log p(e|m)= 0.000:-0.012[\u001b[31m0.012\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCommons[0.045]; Commons[0.045]; hospital[0.043]; resigned[0.043]; resigned[0.043]; constituency[0.042]; parliament[0.042]; seat[0.042]; seat[0.042]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; office[0.041]; office[0.041]; government[0.041]; government[0.041]; government[0.041]; withdrawing[0.041]; Gorst[0.041]; Gorst[0.041]; Gorst[0.041]; conference[0.041]; Downing[0.040]; \t\n\r [===========>..................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s491ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1172/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1295testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = New Delhi\u001b[39m ==> ENTITY: \u001b[32mNew Delhi\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.081:0.081[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.050]; Indian[0.047]; Sept[0.044]; Sept[0.044]; Sept[0.044]; Sept[0.044]; Sept[0.044]; million[0.042]; million[0.042]; government[0.042]; September[0.042]; month[0.041]; Friday[0.041]; fell[0.040]; tonnes[0.040]; tonnes[0.040]; thousands[0.040]; preliminary[0.040]; said[0.039]; falls[0.039]; Apr[0.039]; Apr[0.039]; Petroleum[0.038]; mln[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = India ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.062:0.062[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndian[0.052]; Delhi[0.048]; government[0.046]; million[0.045]; million[0.045]; month[0.045]; September[0.044]; products[0.044]; Sept[0.044]; Sept[0.044]; Sept[0.044]; Sept[0.044]; Sept[0.044]; Friday[0.043]; preliminary[0.042]; said[0.042]; thousands[0.041]; Figures[0.041]; Output[0.041]; fell[0.041]; Apr[0.041]; Apr[0.041]; Petroleum[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indian ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.070:0.070[0]; log p(e|m)= -0.618:-0.618[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.055]; Delhi[0.048]; month[0.044]; million[0.043]; million[0.043]; September[0.042]; government[0.042]; said[0.042]; Sept[0.041]; Sept[0.041]; Sept[0.041]; Sept[0.041]; Sept[0.041]; thousands[0.041]; Apr[0.041]; Apr[0.041]; Friday[0.040]; fell[0.040]; products[0.040]; oil[0.039]; mln[0.039]; tonnes[0.039]; tonnes[0.039]; falls[0.039]; \t\n\r [===========>..................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s557ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1175/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1350testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colorado ==> ENTITY: \u001b[32mColorado\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.032:0.032[0]; log p(e|m)= -0.269:-0.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.044]; Switzerland[0.044]; Vail[0.044]; Norway[0.044]; minute[0.043]; Heidi[0.043]; Russia[0.042]; Stefanie[0.042]; Skiing[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Helen[0.041]; seconds[0.041]; Italy[0.041]; Saturday[0.041]; World[0.041]; World[0.041]; France[0.040]; women[0.040]; Megan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Slovenia ==> ENTITY: \u001b[32mSlovenia\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.230:0.230[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Norway[0.043]; Norway[0.043]; Switzerland[0.043]; Switzerland[0.043]; France[0.042]; France[0.042]; France[0.042]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Canada[0.041]; Russia[0.039]; Russia[0.039]; States[0.039]; Martina[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.254:0.254[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.052]; Italy[0.052]; Cup[0.042]; Switzerland[0.042]; Switzerland[0.042]; Slovenia[0.042]; France[0.042]; France[0.042]; France[0.042]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Canada[0.040]; Sweden[0.039]; Martina[0.039]; Norway[0.039]; World[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Regine Cavagnoud\u001b[39m ==> ENTITY: \u001b[32mRégine Cavagnoud\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGrete[0.049]; Gerg[0.049]; Hilde[0.049]; Austria[0.049]; Austria[0.049]; Austria[0.049]; Ingeborg[0.048]; Zurbriggen[0.048]; Switzerland[0.048]; Switzerland[0.048]; Martina[0.048]; Regina[0.048]; Wachter[0.046]; Stefanie[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; France[0.046]; France[0.046]; France[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.257:0.257[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.046]; Austria[0.046]; Austria[0.046]; Switzerland[0.045]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; women[0.042]; Slovenia[0.042]; Norway[0.041]; Italy[0.041]; Italy[0.041]; Sweden[0.040]; Women[0.039]; Skiing[0.039]; Canada[0.039]; France[0.039]; France[0.039]; Alpine[0.039]; Pernilla[0.038]; World[0.038]; World[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.178:0.178[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.055]; Italy[0.055]; Switzerland[0.045]; Slovenia[0.044]; France[0.044]; France[0.044]; Austria[0.043]; Austria[0.043]; Canada[0.042]; Borghi[0.042]; Sweden[0.042]; Norway[0.041]; official[0.040]; States[0.040]; declared[0.040]; United[0.040]; Patrizia[0.040]; Alessandra[0.039]; Russia[0.039]; Perez[0.037]; Pace[0.037]; Bibiana[0.037]; results[0.036]; Cavagnoud[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alexandra Meissnitzer\u001b[39m ==> ENTITY: \u001b[32mAlexandra Meissnitzer\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.048]; Cup[0.047]; Cup[0.047]; Downhill[0.046]; race[0.045]; Switzerland[0.044]; Norway[0.044]; Gerg[0.044]; downhill[0.043]; Skiing[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Seizinger[0.043]; Sweden[0.042]; Hilde[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Ingeborg[0.042]; Kostner[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.168:0.168[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.050]; Canada[0.048]; Slovenia[0.047]; Switzerland[0.046]; Switzerland[0.046]; Austria[0.046]; Austria[0.046]; Austria[0.046]; Austria[0.046]; Sweden[0.046]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Norway[0.045]; Norway[0.045]; Italy[0.044]; Italy[0.044]; Italy[0.044]; Italy[0.044]; Italy[0.044]; Svetlana[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Switzerland[0.044]; Sweden[0.042]; Italy[0.042]; Italy[0.042]; France[0.041]; Cup[0.040]; Cup[0.040]; Saturday[0.040]; Norway[0.040]; women[0.040]; Russia[0.040]; World[0.040]; World[0.040]; Downhill[0.039]; Kostner[0.039]; Women[0.039]; Skiing[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Heidi Zurbriggen\u001b[39m ==> ENTITY: \u001b[32mHeidi Zurbriggen\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.044]; Cup[0.044]; race[0.043]; Switzerland[0.043]; Switzerland[0.043]; Pernilla[0.042]; Seizinger[0.042]; Martina[0.042]; Downhill[0.042]; Alpine[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Hilde[0.041]; Gerg[0.041]; Ingeborg[0.041]; Helen[0.041]; Vail[0.041]; Sweden[0.041]; Regina[0.041]; World[0.040]; World[0.040]; Skiing[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Switzerland ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.222:0.222[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.047]; Canada[0.044]; France[0.044]; France[0.044]; Italy[0.044]; Italy[0.044]; Italy[0.044]; Italy[0.044]; Slovenia[0.043]; Sweden[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Austria[0.042]; Austria[0.042]; Russia[0.041]; Russia[0.041]; Norway[0.040]; official[0.039]; Martina[0.038]; States[0.037]; Gerg[0.037]; United[0.037]; results[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Carole Montillet\u001b[39m ==> ENTITY: \u001b[32mCarole Montillet\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.074:0.074[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGrete[0.045]; Gerg[0.045]; Hilde[0.044]; Cavagnoud[0.044]; Martina[0.044]; Zurbriggen[0.042]; Wachter[0.042]; France[0.042]; France[0.042]; Sybille[0.042]; Norway[0.041]; Bibiana[0.041]; Pernilla[0.041]; Germany[0.041]; Germany[0.041]; Austria[0.041]; Austria[0.041]; results[0.041]; Slovenia[0.041]; Sweden[0.040]; Pace[0.040]; Switzerland[0.040]; Switzerland[0.040]; Brauner[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.249:0.249[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Switzerland[0.043]; Switzerland[0.043]; Sweden[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Slovenia[0.040]; France[0.040]; France[0.040]; France[0.040]; Canada[0.040]; Norway[0.039]; Norway[0.039]; Martina[0.038]; minute[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Florence Masnada\u001b[39m ==> ENTITY: \u001b[32mFlorence Masnada\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.046]; downhill[0.044]; Kostner[0.043]; Seizinger[0.042]; Cavagnoud[0.042]; Switzerland[0.042]; Zurbriggen[0.042]; Hilde[0.042]; France[0.042]; France[0.042]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Gerg[0.041]; Ingeborg[0.041]; Downhill[0.041]; Martina[0.041]; Norway[0.041]; Vail[0.041]; Regina[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.087:0.087[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.044]; Sweden[0.043]; Russia[0.042]; Russia[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Slovenia[0.042]; Italy[0.042]; Italy[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Colorado[0.041]; World[0.041]; World[0.041]; Norway[0.041]; Cup[0.041]; Cup[0.041]; Kostner[0.039]; women[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.217:0.217[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.047]; France[0.047]; Switzerland[0.044]; Switzerland[0.044]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Martina[0.042]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Canada[0.041]; Sweden[0.040]; Slovenia[0.040]; Norway[0.039]; Norway[0.039]; Russia[0.039]; Russia[0.039]; Stefanie[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Isolde Kostner\u001b[39m ==> ENTITY: \u001b[32mIsolde Kostner\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.111:0.111[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.045]; Seizinger[0.043]; Cup[0.043]; Cup[0.043]; Downhill[0.042]; Hilde[0.042]; race[0.042]; Gerg[0.042]; Skiing[0.042]; Ingeborg[0.042]; Italy[0.042]; Italy[0.042]; Martina[0.042]; Vail[0.042]; Regina[0.041]; World[0.041]; World[0.041]; Colorado[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; downhill[0.041]; Renate[0.040]; Stefanie[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hilde Gerg\u001b[39m ==> ENTITY: \u001b[32mHilde Gerg\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCavagnoud[0.043]; Seizinger[0.043]; Cup[0.043]; Grete[0.043]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; downhill[0.042]; Ingeborg[0.042]; Martina[0.042]; World[0.042]; Regina[0.042]; Canada[0.041]; Kostner[0.041]; Sweden[0.041]; Zurbriggen[0.041]; race[0.041]; Renate[0.041]; Pernilla[0.041]; Norway[0.040]; Ertl[0.040]; Wachter[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kate Pace Lindsay\u001b[39m ==> ENTITY: \u001b[32mKate Pace\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.048]; Grete[0.045]; Hilde[0.045]; Gerg[0.045]; Ingeborg[0.045]; Martina[0.044]; Hilary[0.044]; Regina[0.044]; Catherine[0.044]; Carole[0.043]; Wachter[0.043]; Stefanie[0.043]; Norway[0.043]; Norway[0.043]; Cavagnoud[0.043]; Zurbriggen[0.043]; Heidi[0.042]; Sybille[0.042]; Barbara[0.042]; Helen[0.042]; Bibiana[0.042]; Pernilla[0.042]; Marken[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Renate Goetschl\u001b[39m ==> ENTITY: \u001b[32mRenate Götschl\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndownhill[0.046]; Alpine[0.045]; Downhill[0.043]; Cup[0.043]; Cup[0.043]; Kostner[0.043]; Saturday[0.042]; Seizinger[0.042]; Switzerland[0.042]; Gerg[0.042]; Women[0.042]; Skiing[0.041]; race[0.041]; women[0.041]; Hilde[0.041]; results[0.041]; Norway[0.041]; Ingeborg[0.041]; Vail[0.040]; Regina[0.040]; Zurbriggen[0.040]; Italy[0.040]; Italy[0.040]; seconds[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Anita Wachter\u001b[39m ==> ENTITY: \u001b[32mAnita Wachter\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Switzerland[0.043]; Switzerland[0.043]; Norway[0.042]; Norway[0.042]; Pernilla[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Sweden[0.042]; Gerg[0.041]; Grete[0.041]; Hilde[0.041]; Sybille[0.041]; Ingeborg[0.041]; Martina[0.041]; Regina[0.040]; Canada[0.040]; Cavagnoud[0.040]; Stefanie[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Switzerland ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.259:0.259[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.048]; Canada[0.045]; France[0.045]; France[0.045]; France[0.045]; Italy[0.044]; Italy[0.044]; Italy[0.044]; Slovenia[0.044]; Sweden[0.044]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Russia[0.042]; Russia[0.042]; Norway[0.040]; Downhill[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.207:0.207[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Switzerland[0.043]; Sweden[0.042]; Italy[0.041]; Italy[0.041]; Slovenia[0.041]; France[0.040]; France[0.040]; Saturday[0.040]; Norway[0.040]; Cup[0.040]; Cup[0.040]; women[0.039]; Downhill[0.039]; Martina[0.039]; Women[0.039]; Skiing[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.184:0.184[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.048]; Switzerland[0.045]; Switzerland[0.045]; Italy[0.044]; Italy[0.044]; Italy[0.044]; Italy[0.044]; Martina[0.043]; Austria[0.043]; Austria[0.043]; Germany[0.042]; Germany[0.042]; Canada[0.042]; Sweden[0.041]; Slovenia[0.041]; Norway[0.040]; Russia[0.040]; Russia[0.040]; Catherine[0.039]; States[0.039]; Brigitte[0.038]; Svetlana[0.038]; Sybille[0.038]; official[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.192:0.192[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.053]; Cup[0.044]; Cup[0.044]; Switzerland[0.044]; France[0.043]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Sweden[0.041]; Alpine[0.040]; Norway[0.040]; Martina[0.040]; World[0.040]; World[0.040]; Stefanie[0.039]; Florence[0.039]; minute[0.039]; Skiing[0.039]; Saturday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pernilla Wiberg\u001b[39m ==> ENTITY: \u001b[32mPernilla Wiberg\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.044]; Norway[0.044]; Zurbriggen[0.044]; Gerg[0.044]; Seizinger[0.043]; Sweden[0.043]; Grete[0.043]; Hilde[0.043]; Ingeborg[0.042]; Martina[0.042]; Regina[0.042]; Kostner[0.041]; Switzerland[0.041]; Switzerland[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Wachter[0.040]; Stefanie[0.040]; Canada[0.040]; Sybille[0.040]; Bibiana[0.040]; Katja[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.219:0.219[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.048]; Switzerland[0.047]; Germany[0.046]; Germany[0.046]; Slovenia[0.044]; Norway[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Sweden[0.042]; Canada[0.041]; France[0.041]; France[0.041]; Pernilla[0.040]; Russia[0.040]; Russia[0.040]; declared[0.039]; Sybille[0.039]; Martina[0.039]; Cavagnoud[0.038]; Gerg[0.038]; States[0.038]; official[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.048]; Switzerland[0.045]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Martina[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Cup[0.041]; Cup[0.041]; Canada[0.041]; Sweden[0.040]; Slovenia[0.040]; Women[0.040]; Norway[0.039]; Russia[0.039]; Russia[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Switzerland[0.042]; Switzerland[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; France[0.042]; France[0.042]; France[0.042]; Russia[0.041]; Russia[0.041]; Norway[0.041]; Norway[0.041]; Regina[0.041]; States[0.041]; Slovenia[0.039]; results[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.247:0.247[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.052]; Italy[0.052]; Italy[0.052]; Switzerland[0.042]; Switzerland[0.042]; Slovenia[0.042]; France[0.042]; France[0.042]; France[0.042]; Austria[0.041]; Austria[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Canada[0.040]; Sweden[0.039]; Martina[0.039]; Norway[0.039]; Norway[0.039]; Alessandra[0.039]; Florence[0.038]; Catherine[0.038]; Borghi[0.038]; States[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.253:0.253[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Switzerland[0.043]; Switzerland[0.043]; Sweden[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Slovenia[0.040]; France[0.040]; France[0.040]; France[0.040]; Canada[0.040]; Norway[0.039]; Norway[0.039]; Martina[0.038]; minute[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.048]; Switzerland[0.045]; World[0.045]; World[0.045]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Slovenia[0.043]; Cup[0.043]; Cup[0.043]; Canada[0.043]; Italy[0.042]; Italy[0.042]; Italy[0.042]; race[0.042]; Women[0.042]; Downhill[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.245:0.245[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.047]; Austria[0.047]; Austria[0.047]; Switzerland[0.045]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; women[0.043]; Norway[0.041]; Italy[0.041]; Italy[0.041]; Sweden[0.041]; Women[0.040]; Skiing[0.039]; France[0.039]; Alpine[0.039]; Pernilla[0.038]; World[0.038]; World[0.038]; Russia[0.038]; Russia[0.038]; Downhill[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.288:0.288[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.046]; Austria[0.046]; Austria[0.046]; Switzerland[0.045]; Switzerland[0.045]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Slovenia[0.042]; Norway[0.041]; Norway[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Sweden[0.040]; Canada[0.039]; France[0.039]; France[0.039]; France[0.039]; Pernilla[0.038]; Russia[0.038]; Russia[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vail ==> ENTITY: \u001b[32mVail, Colorado\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.392:-0.392[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.049]; Downhill[0.047]; Skiing[0.047]; Saturday[0.046]; race[0.045]; downhill[0.045]; Colorado[0.045]; women[0.045]; Regina[0.043]; Switzerland[0.043]; Women[0.043]; World[0.042]; World[0.042]; Cup[0.042]; Cup[0.042]; Helen[0.042]; Stefanie[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; seconds[0.041]; minute[0.041]; results[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.204:0.204[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.047]; Austria[0.047]; Switzerland[0.045]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; women[0.043]; Cup[0.042]; Cup[0.042]; Women[0.042]; World[0.042]; World[0.042]; Norway[0.041]; Italy[0.041]; Italy[0.041]; France[0.040]; Skiing[0.039]; Russia[0.039]; Alpine[0.039]; Downhill[0.038]; Katja[0.038]; Stefanie[0.038]; Schuster[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.163:0.163[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.048]; Canada[0.045]; Slovenia[0.045]; Switzerland[0.044]; Switzerland[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Sweden[0.044]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Cup[0.043]; Norway[0.043]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Svetlana[0.041]; Provisional[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.048]; Switzerland[0.045]; Germany[0.044]; Germany[0.044]; Austria[0.044]; Austria[0.044]; Slovenia[0.044]; Canada[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Pernilla[0.042]; Russia[0.041]; France[0.041]; France[0.041]; Stroem[0.040]; Martina[0.040]; States[0.039]; United[0.039]; Patrizia[0.038]; declared[0.038]; Ertl[0.038]; Grete[0.038]; official[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.194:0.194[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.045]; Norway[0.045]; Switzerland[0.044]; Switzerland[0.044]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Slovenia[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Canada[0.041]; Russia[0.039]; Russia[0.039]; Gerg[0.039]; Pernilla[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.182:0.182[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.054]; Italy[0.054]; Switzerland[0.044]; Slovenia[0.044]; France[0.044]; France[0.044]; Austria[0.043]; Austria[0.043]; Germany[0.043]; Canada[0.042]; Sweden[0.041]; Norway[0.041]; Borghi[0.040]; States[0.040]; official[0.040]; United[0.040]; Alessandra[0.039]; Russia[0.039]; Patrizia[0.039]; declared[0.039]; Lindsay[0.037]; Perez[0.037]; Pace[0.037]; Bibiana[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.228:0.228[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Switzerland[0.043]; Switzerland[0.043]; Sweden[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Slovenia[0.040]; France[0.040]; France[0.040]; Saturday[0.040]; Canada[0.040]; Norway[0.040]; Cup[0.039]; Cup[0.039]; women[0.039]; Downhill[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Miriam Vogt\u001b[39m ==> ENTITY: \u001b[32mMiriam Vogt\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Switzerland[0.043]; Alpine[0.043]; Seizinger[0.042]; Downhill[0.042]; Renate[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Hilde[0.041]; Vail[0.041]; Skiing[0.041]; Gerg[0.041]; Ingeborg[0.041]; Martina[0.041]; downhill[0.041]; women[0.040]; Regina[0.040]; Kostner[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hilary Lindh\u001b[39m ==> ENTITY: \u001b[32mHilary Lindh\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.049]; Norway[0.046]; Switzerland[0.046]; Switzerland[0.046]; Sweden[0.045]; Grete[0.044]; Hilde[0.044]; Austria[0.044]; Austria[0.044]; Gerg[0.044]; Martina[0.043]; Regina[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Pernilla[0.042]; States[0.042]; Wachter[0.042]; France[0.042]; France[0.042]; France[0.042]; Zurbriggen[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.088:0.088[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.046]; Switzerland[0.044]; Switzerland[0.044]; Russia[0.043]; Russia[0.043]; Sweden[0.043]; Norway[0.042]; Austria[0.042]; Austria[0.042]; Slovenia[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; France[0.040]; France[0.040]; France[0.040]; Borghi[0.039]; official[0.038]; Lindsay[0.038]; Martina[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Martina Ertl\u001b[39m ==> ENTITY: \u001b[32mMartina Ertl-Renz\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Kostner[0.042]; Seizinger[0.042]; seconds[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Grete[0.041]; Norway[0.041]; Norway[0.041]; Hilde[0.041]; Sweden[0.041]; Gerg[0.041]; Ingeborg[0.041]; Canada[0.040]; Regina[0.040]; Switzerland[0.040]; Switzerland[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Katja Seizinger\u001b[39m ==> ENTITY: \u001b[32mKatja Seizinger\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrace[0.045]; Cup[0.043]; Cup[0.043]; Kostner[0.043]; Alpine[0.042]; Zurbriggen[0.042]; Gerg[0.042]; Downhill[0.042]; women[0.042]; results[0.042]; Hilde[0.042]; downhill[0.042]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Ingeborg[0.041]; Martina[0.041]; Vail[0.041]; Regina[0.041]; Norway[0.041]; Skiing[0.041]; Colorado[0.040]; \t\n\r [===========>..................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s245ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1221/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1298testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.092:0.092[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.050]; Conservative[0.047]; Conservative[0.047]; Labour[0.044]; held[0.043]; William[0.043]; William[0.043]; election[0.042]; election[0.042]; months[0.042]; Friday[0.042]; Hill[0.041]; Hill[0.041]; said[0.040]; said[0.040]; said[0.040]; longest[0.040]; bookmakers[0.039]; bookmakers[0.039]; victory[0.039]; victory[0.039]; opposition[0.039]; lengthened[0.039]; general[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Conservative ==> ENTITY: \u001b[32mConservative Party (UK)\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.559:-0.559[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConservative[0.050]; Labour[0.045]; election[0.044]; election[0.044]; William[0.044]; William[0.044]; opposition[0.043]; London[0.042]; London[0.042]; held[0.042]; Hill[0.042]; Hill[0.042]; victory[0.042]; victory[0.042]; general[0.041]; months[0.041]; Uk[0.040]; Uk[0.040]; Friday[0.040]; said[0.039]; said[0.039]; said[0.039]; longest[0.038]; lengthened[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = William Hill ==> ENTITY: \u001b[32mWilliam Hill (bookmaker)\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.124:0.124[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWilliam[0.048]; Hill[0.048]; bookmakers[0.048]; bookmakers[0.048]; London[0.046]; London[0.046]; Conservative[0.046]; Conservative[0.046]; election[0.044]; election[0.044]; months[0.043]; Friday[0.043]; longest[0.042]; Uk[0.041]; Uk[0.041]; held[0.041]; opposition[0.041]; Labour[0.041]; odds[0.041]; odds[0.041]; odds[0.041]; victory[0.040]; victory[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Labour ==> ENTITY: \u001b[32mLabour Party (UK)\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.341:-0.341[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConservative[0.050]; Conservative[0.050]; election[0.044]; election[0.044]; William[0.043]; William[0.043]; London[0.043]; London[0.043]; opposition[0.042]; general[0.042]; months[0.041]; said[0.041]; said[0.041]; said[0.041]; held[0.041]; Friday[0.040]; Hill[0.040]; Hill[0.040]; victory[0.039]; victory[0.039]; lengthened[0.039]; longest[0.038]; Uk[0.038]; Uk[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = William Hill ==> ENTITY: \u001b[32mWilliam Hill (bookmaker)\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.124:0.124[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWilliam[0.048]; Hill[0.048]; bookmakers[0.048]; bookmakers[0.048]; London[0.046]; London[0.046]; Conservative[0.046]; Conservative[0.046]; election[0.044]; election[0.044]; months[0.043]; Friday[0.043]; longest[0.042]; Uk[0.041]; Uk[0.041]; held[0.041]; opposition[0.041]; Labour[0.041]; odds[0.041]; odds[0.041]; odds[0.041]; victory[0.040]; victory[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Conservative ==> ENTITY: \u001b[32mConservative Party (UK)\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.559:-0.559[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConservative[0.050]; Labour[0.045]; election[0.044]; election[0.044]; William[0.044]; William[0.044]; opposition[0.043]; London[0.042]; London[0.042]; held[0.042]; Hill[0.042]; Hill[0.042]; victory[0.042]; victory[0.042]; general[0.041]; months[0.041]; Uk[0.040]; Uk[0.040]; Friday[0.040]; said[0.039]; said[0.039]; said[0.039]; longest[0.038]; lengthened[0.038]; \t\n\r [===========>..................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s303ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1227/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1384testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew York\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew York Knicks\u001b[39m <---> \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.244:0.237[\u001b[31m0.008\u001b[39m]; local()= 0.194:0.078[\u001b[31m0.117\u001b[39m]; log p(e|m)= -4.828:-1.790[\u001b[32m3.039\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.048]; Basketball[0.048]; team[0.047]; games[0.046]; Cleveland[0.044]; Boston[0.043]; Miami[0.043]; Friday[0.042]; Friday[0.042]; York[0.041]; Lakers[0.041]; Charlotte[0.041]; Orlando[0.041]; Detroit[0.041]; National[0.040]; Jersey[0.040]; Phoenix[0.040]; Indiana[0.039]; State[0.039]; Minnesota[0.039]; home[0.039]; Sacramento[0.039]; Association[0.038]; Utah[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York Knicks\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.194:0.194[0]; log p(e|m)= -4.828:-4.828[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.048]; Basketball[0.048]; team[0.047]; games[0.046]; Cleveland[0.044]; Boston[0.043]; Miami[0.043]; Friday[0.042]; Friday[0.042]; York[0.041]; Lakers[0.041]; Charlotte[0.041]; Orlando[0.041]; Detroit[0.041]; National[0.040]; Jersey[0.040]; Phoenix[0.040]; Indiana[0.039]; State[0.039]; Minnesota[0.039]; home[0.039]; Sacramento[0.039]; Association[0.038]; Utah[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indiana ==> ENTITY: \u001b[32mIndiana Pacers\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.205:0.205[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.047]; Basketball[0.047]; games[0.045]; team[0.044]; Cleveland[0.044]; Charlotte[0.044]; Miami[0.043]; Boston[0.042]; Orlando[0.042]; Jersey[0.041]; Detroit[0.041]; Phoenix[0.041]; State[0.041]; Lakers[0.041]; National[0.040]; Friday[0.040]; Friday[0.040]; Sacramento[0.040]; home[0.040]; Minnesota[0.040]; York[0.040]; York[0.040]; Portland[0.039]; Utah[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Minnesota ==> ENTITY: \u001b[32mMinnesota Timberwolves\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.181:0.181[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.047]; Basketball[0.047]; Basketball[0.047]; games[0.046]; Cleveland[0.043]; Detroit[0.042]; Vancouver[0.042]; Charlotte[0.042]; Phoenix[0.042]; Miami[0.042]; Orlando[0.041]; Boston[0.041]; Caps[0.041]; Friday[0.041]; Friday[0.041]; State[0.040]; Lakers[0.040]; Indiana[0.040]; National[0.040]; Sacramento[0.039]; Portland[0.039]; home[0.039]; York[0.039]; York[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Miami ==> ENTITY: \u001b[32mMiami Heat\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.205:0.205[0]; log p(e|m)= -3.730:-3.730[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.047]; games[0.046]; Basketball[0.046]; Basketball[0.046]; Cleveland[0.044]; Orlando[0.043]; Boston[0.043]; Charlotte[0.043]; State[0.041]; Phoenix[0.041]; Detroit[0.041]; Lakers[0.041]; Friday[0.041]; Friday[0.041]; Indiana[0.041]; National[0.040]; York[0.040]; York[0.040]; Minnesota[0.040]; Caps[0.039]; Jersey[0.039]; Utah[0.039]; Portland[0.039]; Antonio[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Detroit ==> ENTITY: \u001b[32mDetroit Pistons\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.193:0.193[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.048]; Basketball[0.046]; Basketball[0.046]; team[0.046]; Caps[0.044]; Cleveland[0.044]; Vancouver[0.043]; Boston[0.042]; Charlotte[0.042]; Miami[0.042]; Friday[0.041]; Friday[0.041]; Phoenix[0.041]; Orlando[0.041]; Lakers[0.040]; Indiana[0.040]; National[0.040]; York[0.039]; York[0.039]; Jersey[0.039]; home[0.039]; Minnesota[0.039]; State[0.039]; Golden[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = LA LAKERS\u001b[39m ==> ENTITY: \u001b[32mLos Angeles Lakers\u001b[39m\t\nSCORES: global= 0.317:0.317[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.049]; Basketball[0.048]; Basketball[0.048]; team[0.047]; Cleveland[0.046]; Boston[0.045]; Miami[0.045]; Orlando[0.044]; Phoenix[0.044]; Charlotte[0.043]; Detroit[0.043]; Sacramento[0.042]; National[0.042]; Friday[0.042]; Friday[0.042]; State[0.041]; Portland[0.041]; Minnesota[0.041]; Utah[0.041]; Indiana[0.041]; Golden[0.041]; Jersey[0.041]; Caps[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portland ==> ENTITY: \u001b[32mPortland Trail Blazers\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.192:0.192[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.048]; Basketball[0.045]; Basketball[0.045]; Cleveland[0.045]; team[0.045]; Boston[0.044]; Miami[0.043]; Orlando[0.042]; Phoenix[0.042]; Charlotte[0.042]; Detroit[0.041]; Lakers[0.041]; National[0.040]; Sacramento[0.040]; Indiana[0.040]; Minnesota[0.040]; Vancouver[0.040]; Utah[0.040]; Friday[0.040]; Friday[0.040]; York[0.040]; York[0.040]; home[0.039]; Golden[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Phoenix ==> ENTITY: \u001b[32mPhoenix Suns\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.212:0.212[0]; log p(e|m)= -3.474:-3.474[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; team[0.046]; Basketball[0.046]; Basketball[0.046]; Cleveland[0.045]; Miami[0.043]; Vancouver[0.043]; Caps[0.043]; Orlando[0.042]; Boston[0.042]; Detroit[0.041]; Lakers[0.041]; Charlotte[0.041]; Friday[0.040]; Friday[0.040]; Utah[0.040]; home[0.039]; National[0.039]; Sacramento[0.039]; Portland[0.039]; Indiana[0.039]; Minnesota[0.039]; Antonio[0.039]; Golden[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NBA\u001b[39m ==> ENTITY: \u001b[32mNational Basketball Association\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.048]; Basketball[0.048]; games[0.046]; team[0.046]; Lakers[0.043]; Cleveland[0.043]; Charlotte[0.042]; Phoenix[0.042]; Miami[0.042]; Orlando[0.041]; Boston[0.041]; National[0.041]; Portland[0.040]; Utah[0.040]; Friday[0.040]; Friday[0.040]; Detroit[0.040]; Caps[0.040]; Sacramento[0.040]; Minnesota[0.040]; Indiana[0.040]; Jersey[0.039]; York[0.038]; York[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Jersey ==> ENTITY: \u001b[32mBrooklyn Nets\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.200:0.200[0]; log p(e|m)= -3.912:-3.912[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; Basketball[0.046]; Basketball[0.046]; team[0.045]; York[0.043]; York[0.043]; Caps[0.043]; Boston[0.042]; Cleveland[0.042]; Vancouver[0.042]; Miami[0.042]; Charlotte[0.041]; Lakers[0.041]; Detroit[0.041]; Orlando[0.041]; Friday[0.041]; Friday[0.041]; Indiana[0.040]; Phoenix[0.040]; National[0.040]; Minnesota[0.039]; Utah[0.039]; Sacramento[0.038]; home[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCharlotte\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mCharlotte, North Carolina\u001b[39m <---> \u001b[32mCharlotte Bobcats\u001b[39m\t\nSCORES: global= 0.244:0.239[\u001b[31m0.005\u001b[39m]; local()= 0.102:0.163[\u001b[32m0.062\u001b[39m]; log p(e|m)= -0.736:-4.269[\u001b[31m3.533\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.046]; Basketball[0.046]; games[0.045]; team[0.045]; Miami[0.044]; Cleveland[0.043]; Orlando[0.043]; Phoenix[0.043]; Boston[0.042]; Detroit[0.042]; Friday[0.041]; Friday[0.041]; National[0.041]; Lakers[0.041]; Sacramento[0.040]; York[0.040]; York[0.040]; Indiana[0.040]; Portland[0.040]; Jersey[0.040]; home[0.039]; Golden[0.039]; Utah[0.039]; Minnesota[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sacramento ==> ENTITY: \u001b[32mSacramento Kings\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.198:0.198[0]; log p(e|m)= -3.079:-3.079[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.046]; Basketball[0.046]; games[0.046]; team[0.046]; Cleveland[0.044]; Miami[0.042]; Charlotte[0.042]; York[0.041]; York[0.041]; Boston[0.041]; Orlando[0.041]; Phoenix[0.041]; San[0.041]; Friday[0.041]; Friday[0.041]; National[0.040]; Portland[0.040]; Detroit[0.040]; Lakers[0.040]; Indiana[0.040]; Golden[0.040]; State[0.039]; Minnesota[0.039]; Association[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Golden State ==> ENTITY: \u001b[32mGolden State Warriors\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.192:0.192[0]; log p(e|m)= -0.742:-0.742[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.046]; Basketball[0.046]; team[0.045]; games[0.045]; Portland[0.044]; Lakers[0.044]; Cleveland[0.044]; San[0.042]; Orlando[0.042]; Miami[0.042]; Association[0.041]; Charlotte[0.041]; Boston[0.041]; Phoenix[0.041]; National[0.041]; Sacramento[0.041]; Detroit[0.040]; Utah[0.040]; Friday[0.040]; Friday[0.040]; home[0.039]; Indiana[0.039]; Vancouver[0.038]; Caps[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = National Basketball Association\u001b[39m ==> ENTITY: \u001b[32mNational Basketball Association\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.049]; Basketball[0.048]; Cleveland[0.045]; games[0.045]; Miami[0.044]; Boston[0.043]; Detroit[0.042]; Orlando[0.042]; Friday[0.042]; Friday[0.042]; Indiana[0.042]; Lakers[0.042]; Charlotte[0.041]; Phoenix[0.041]; Utah[0.041]; Minnesota[0.040]; Sacramento[0.040]; Portland[0.040]; Antonio[0.039]; State[0.039]; Jersey[0.039]; York[0.039]; York[0.039]; Golden[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Antonio ==> ENTITY: \u001b[32mSan Antonio Spurs\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.203:0.203[0]; log p(e|m)= -2.865:-2.865[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.048]; Basketball[0.048]; games[0.047]; team[0.046]; Cleveland[0.043]; Orlando[0.043]; Charlotte[0.042]; Miami[0.042]; Phoenix[0.042]; Detroit[0.042]; Boston[0.041]; Lakers[0.041]; Golden[0.040]; Indiana[0.040]; Friday[0.040]; Friday[0.040]; Utah[0.040]; Sacramento[0.040]; home[0.040]; Jersey[0.039]; La[0.039]; National[0.039]; State[0.039]; Minnesota[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Boston ==> ENTITY: \u001b[32mBoston Celtics\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.190:0.190[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; York[0.046]; York[0.046]; team[0.046]; Basketball[0.045]; Basketball[0.045]; Cleveland[0.044]; Caps[0.044]; Vancouver[0.042]; Detroit[0.042]; Miami[0.041]; Lakers[0.041]; Charlotte[0.041]; Orlando[0.040]; Phoenix[0.040]; Minnesota[0.040]; Portland[0.040]; Jersey[0.039]; National[0.039]; Friday[0.039]; Friday[0.039]; home[0.039]; Indiana[0.038]; Sacramento[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Orlando ==> ENTITY: \u001b[32mOrlando Magic\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.189:0.189[0]; log p(e|m)= -3.079:-3.079[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; Basketball[0.047]; Basketball[0.047]; team[0.046]; Cleveland[0.044]; Miami[0.044]; Boston[0.043]; Detroit[0.042]; Charlotte[0.042]; Phoenix[0.042]; Friday[0.041]; Friday[0.041]; Indiana[0.041]; Lakers[0.041]; Jersey[0.041]; Utah[0.040]; York[0.040]; York[0.040]; National[0.039]; Minnesota[0.039]; Sacramento[0.039]; home[0.039]; Portland[0.038]; Caps[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Utah ==> ENTITY: \u001b[32mUtah Jazz\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.203:0.203[0]; log p(e|m)= -3.576:-3.576[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.049]; team[0.047]; Basketball[0.047]; Basketball[0.047]; Phoenix[0.045]; Charlotte[0.044]; Cleveland[0.044]; Vancouver[0.044]; Lakers[0.044]; Caps[0.043]; Portland[0.043]; Miami[0.043]; National[0.043]; Friday[0.043]; Friday[0.043]; Orlando[0.042]; home[0.042]; Boston[0.042]; Sacramento[0.042]; State[0.042]; Detroit[0.041]; Association[0.041]; Minnesota[0.040]; \t\n\r [===========>..................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s388ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1247/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1220testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.063:0.063[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.063]; London[0.063]; recorded[0.057]; shares[0.055]; Depository[0.055]; day[0.053]; day[0.053]; Thursday[0.053]; Global[0.053]; Friday[0.052]; represents[0.052]; closed[0.051]; closed[0.051]; high[0.049]; Receipt[0.049]; low[0.049]; Jakarta[0.047]; common[0.046]; newsroom[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.063:0.063[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.063]; London[0.063]; recorded[0.057]; shares[0.055]; Depository[0.055]; day[0.053]; day[0.053]; Thursday[0.053]; Global[0.053]; Friday[0.052]; represents[0.052]; closed[0.051]; closed[0.051]; high[0.049]; Receipt[0.049]; low[0.049]; Jakarta[0.047]; common[0.046]; newsroom[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jakarta\u001b[39m ==> ENTITY: \u001b[32mJakarta\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.073:0.073[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThursday[0.049]; Timah[0.048]; Timah[0.048]; Friday[0.047]; low[0.047]; day[0.047]; day[0.047]; closed[0.047]; closed[0.047]; high[0.047]; Global[0.046]; recorded[0.046]; represents[0.045]; Receipt[0.044]; London[0.044]; London[0.044]; London[0.044]; Depository[0.044]; newsroom[0.043]; common[0.043]; Pt[0.043]; shares[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.063:0.063[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.063]; London[0.063]; recorded[0.057]; shares[0.055]; Depository[0.055]; day[0.053]; day[0.053]; Thursday[0.053]; Global[0.053]; Friday[0.052]; represents[0.052]; closed[0.051]; closed[0.051]; high[0.049]; Receipt[0.049]; low[0.049]; Jakarta[0.047]; common[0.046]; newsroom[0.046]; \t\n\r [===========>..................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s454ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1251/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1389testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Madrid ==> ENTITY: \u001b[32mMadrid\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.074:0.074[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMadrid[0.066]; Soccer[0.063]; Spanish[0.061]; Spanish[0.061]; match[0.060]; Barcelona[0.059]; Real[0.056]; Predrag[0.054]; division[0.054]; Saturday[0.053]; Division[0.053]; Davor[0.053]; Summary[0.052]; Summary[0.052]; Attendance[0.051]; Halftime[0.051]; Mijatovic[0.051]; Suker[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Madrid\u001b[39m ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.068]; Madrid[0.064]; Barcelona[0.062]; match[0.062]; Division[0.061]; Spanish[0.059]; Spanish[0.059]; division[0.059]; Summary[0.059]; Summary[0.059]; Mijatovic[0.057]; Saturday[0.056]; Attendance[0.056]; Predrag[0.055]; Halftime[0.055]; Davor[0.055]; Suker[0.053]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barcelona ==> ENTITY: \u001b[32mFC Barcelona\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.125:0.125[0]; log p(e|m)= -1.709:-1.709[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.064]; Madrid[0.061]; Madrid[0.061]; match[0.059]; Spanish[0.059]; Spanish[0.059]; Division[0.057]; division[0.057]; Real[0.056]; Predrag[0.055]; Davor[0.054]; Mijatovic[0.053]; Saturday[0.053]; Attendance[0.051]; Halftime[0.051]; Suker[0.050]; Summary[0.050]; Summary[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spanish ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.068:0.068[0]; log p(e|m)= -1.241:-1.241[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpanish[0.071]; Soccer[0.062]; match[0.061]; Madrid[0.059]; Madrid[0.059]; Division[0.058]; Barcelona[0.057]; division[0.057]; Real[0.055]; Saturday[0.054]; Davor[0.053]; Suker[0.052]; Mijatovic[0.052]; Halftime[0.051]; Attendance[0.051]; Predrag[0.051]; Summary[0.049]; Summary[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Predrag Mijatovic\u001b[39m ==> ENTITY: \u001b[32mPredrag Mijatović\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.101:0.101[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHalftime[0.062]; match[0.062]; Davor[0.062]; Soccer[0.061]; Madrid[0.060]; Madrid[0.060]; Real[0.060]; Attendance[0.059]; Barcelona[0.058]; Spanish[0.058]; Spanish[0.058]; Division[0.058]; division[0.057]; Suker[0.056]; Summary[0.056]; Summary[0.056]; Saturday[0.055]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Davor Suker\u001b[39m ==> ENTITY: \u001b[32mDavor Šuker\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.066]; match[0.066]; Halftime[0.062]; Predrag[0.060]; Madrid[0.060]; Madrid[0.060]; Barcelona[0.059]; Spanish[0.059]; Spanish[0.059]; Attendance[0.058]; Division[0.058]; Mijatovic[0.057]; Real[0.056]; division[0.056]; Saturday[0.055]; Summary[0.055]; Summary[0.055]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spanish ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.068:0.068[0]; log p(e|m)= -1.241:-1.241[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpanish[0.071]; Soccer[0.062]; match[0.061]; Madrid[0.059]; Madrid[0.059]; Division[0.058]; Barcelona[0.057]; division[0.057]; Real[0.055]; Saturday[0.054]; Davor[0.053]; Suker[0.052]; Mijatovic[0.052]; Halftime[0.051]; Attendance[0.051]; Predrag[0.051]; Summary[0.049]; Summary[0.049]; \t\n\r [===========>..................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s489ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1258/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1313testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zaire ==> ENTITY: \u001b[32mDemocratic Republic of the Congo\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.153:0.153[0]; log p(e|m)= -2.002:-2.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaire[0.054]; Zaire[0.054]; northern[0.052]; Ugandan[0.049]; Garamba[0.048]; Africans[0.048]; troops[0.048]; rebels[0.048]; government[0.048]; national[0.047]; Italy[0.046]; Italy[0.046]; strip[0.046]; make[0.046]; took[0.046]; fighting[0.046]; group[0.046]; ethnic[0.046]; capital[0.046]; Europeans[0.045]; official[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Europeans ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.110:0.110[0]; log p(e|m)= -0.234:-0.234[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.047]; Africans[0.045]; Italians[0.044]; national[0.043]; ethnic[0.043]; northern[0.043]; Italy[0.043]; Italy[0.043]; group[0.042]; group[0.042]; Zaire[0.041]; Zaire[0.041]; Zaire[0.041]; Belgium[0.041]; Zambia[0.041]; Italian[0.041]; Ugandan[0.040]; government[0.040]; took[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; make[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.101:0.101[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnorthern[0.046]; national[0.046]; Belgium[0.043]; seven[0.043]; seven[0.043]; fighting[0.043]; Europeans[0.043]; Italians[0.043]; troops[0.043]; government[0.042]; Foreign[0.041]; rebels[0.041]; mission[0.041]; Italian[0.041]; group[0.041]; group[0.041]; Zaire[0.040]; Zaire[0.040]; ethnic[0.040]; Africans[0.040]; priests[0.040]; took[0.040]; capital[0.040]; official[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italian ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.180:0.180[0]; log p(e|m)= -0.759:-0.759[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.053]; Italy[0.053]; Italians[0.047]; Rome[0.044]; Spain[0.044]; Roman[0.042]; Zaire[0.041]; Zaire[0.041]; Zaire[0.041]; Catholic[0.041]; northern[0.041]; Belgium[0.041]; national[0.041]; government[0.040]; troops[0.040]; ethnic[0.040]; Europeans[0.040]; seven[0.039]; seven[0.039]; priests[0.039]; priests[0.039]; priests[0.039]; rebels[0.039]; Tutsi[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Africans ==> ENTITY: \u001b[32mAfrica\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.132:0.132[0]; log p(e|m)= -1.187:-1.187[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaire[0.050]; Zaire[0.050]; Zaire[0.050]; Belgium[0.049]; ethnic[0.047]; northern[0.047]; Ugandan[0.046]; Europeans[0.046]; Friday[0.046]; Spain[0.045]; national[0.045]; Zambia[0.045]; fighting[0.045]; Italy[0.044]; Italy[0.044]; said[0.044]; said[0.044]; said[0.044]; said[0.044]; government[0.043]; Italians[0.043]; Italian[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zaire ==> ENTITY: \u001b[32mDemocratic Republic of the Congo\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.148:0.148[0]; log p(e|m)= -2.002:-2.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaire[0.048]; Zaire[0.048]; northern[0.045]; Ugandan[0.043]; Belgium[0.043]; Garamba[0.042]; Africans[0.042]; troops[0.042]; rebels[0.042]; government[0.042]; national[0.041]; Italy[0.041]; Italy[0.041]; strip[0.041]; make[0.040]; took[0.040]; fighting[0.040]; group[0.040]; group[0.040]; ethnic[0.040]; capital[0.040]; consisted[0.040]; Europeans[0.039]; Spain[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italians ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.128:0.128[0]; log p(e|m)= -1.556:-1.556[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItalian[0.051]; Spain[0.045]; northern[0.044]; ethnic[0.044]; Roman[0.043]; Zaire[0.042]; Zaire[0.042]; Catholic[0.042]; Belgium[0.042]; national[0.042]; Europeans[0.042]; government[0.041]; troops[0.041]; fighting[0.040]; seven[0.040]; seven[0.040]; priests[0.040]; priests[0.040]; rebels[0.040]; Tutsi[0.040]; capital[0.040]; Africans[0.040]; group[0.039]; group[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Garamba\u001b[39m ==> ENTITY: \u001b[32mGaramba National Park\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.132:0.132[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npark[0.046]; park[0.046]; seven[0.045]; seven[0.045]; group[0.045]; group[0.045]; rebels[0.044]; northern[0.044]; national[0.044]; Zaire[0.044]; Zaire[0.044]; Zaire[0.044]; Ugandan[0.043]; official[0.043]; evacuated[0.043]; fighting[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; government[0.042]; Tutsi[0.042]; troops[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.055]; Italian[0.048]; northern[0.047]; Rome[0.045]; Roman[0.044]; troops[0.043]; national[0.043]; government[0.041]; fighting[0.041]; seven[0.041]; Europeans[0.040]; ethnic[0.040]; Zaire[0.040]; Zaire[0.040]; Zaire[0.040]; Catholic[0.040]; priests[0.039]; priests[0.039]; Friday[0.039]; Africans[0.039]; capital[0.039]; took[0.039]; landing[0.039]; special[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.055]; Italian[0.048]; northern[0.047]; Rome[0.045]; Roman[0.044]; troops[0.043]; national[0.043]; government[0.041]; fighting[0.041]; seven[0.041]; Europeans[0.040]; ethnic[0.040]; Zaire[0.040]; Zaire[0.040]; Zaire[0.040]; Catholic[0.040]; priests[0.039]; priests[0.039]; Friday[0.039]; Africans[0.039]; capital[0.039]; took[0.039]; landing[0.039]; special[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgium ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.242:-0.242[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.047]; troops[0.045]; fighting[0.044]; seven[0.044]; seven[0.044]; national[0.043]; said[0.043]; said[0.043]; said[0.043]; northern[0.042]; Europeans[0.041]; Zaire[0.041]; Zaire[0.041]; Italian[0.041]; Italians[0.041]; official[0.041]; looked[0.040]; flight[0.040]; took[0.040]; capital[0.040]; government[0.040]; make[0.040]; group[0.040]; group[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zaire ==> ENTITY: \u001b[32mDemocratic Republic of the Congo\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.153:0.153[0]; log p(e|m)= -2.002:-2.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaire[0.047]; Zaire[0.047]; northern[0.045]; Ugandan[0.043]; Belgium[0.043]; Garamba[0.042]; Africans[0.042]; troops[0.042]; rebels[0.042]; Zambia[0.042]; government[0.042]; national[0.041]; Italy[0.041]; Italy[0.041]; strip[0.040]; make[0.040]; took[0.040]; fighting[0.040]; group[0.040]; group[0.040]; ethnic[0.040]; capital[0.040]; consisted[0.040]; Europeans[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tutsi\u001b[39m ==> ENTITY: \u001b[32mTutsi\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nethnic[0.052]; Ugandan[0.050]; Africans[0.049]; Zaire[0.049]; Zaire[0.049]; Zaire[0.049]; rebels[0.049]; government[0.048]; group[0.048]; group[0.048]; national[0.048]; Europeans[0.047]; Garamba[0.047]; troops[0.047]; Belgium[0.047]; fighting[0.046]; northern[0.046]; priests[0.045]; priests[0.045]; priests[0.045]; Zambia[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zambia ==> ENTITY: \u001b[32mZambia\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.320:-0.320[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnational[0.047]; northern[0.044]; Ugandan[0.044]; Zaire[0.044]; priests[0.042]; Africans[0.042]; Foreign[0.042]; seven[0.041]; seven[0.041]; Garamba[0.041]; ministry[0.041]; ministry[0.041]; official[0.041]; Europeans[0.041]; government[0.041]; said[0.041]; said[0.041]; said[0.041]; Ministry[0.041]; Spain[0.040]; group[0.040]; group[0.040]; embassy[0.040]; Kampala[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rome ==> ENTITY: \u001b[32mRome\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.168:0.168[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRoman[0.049]; Catholic[0.048]; Italy[0.048]; Italy[0.048]; priests[0.043]; priests[0.043]; Italian[0.043]; nuns[0.043]; nuns[0.043]; government[0.042]; northern[0.042]; fighting[0.040]; Europeans[0.040]; official[0.040]; ministry[0.039]; ministry[0.039]; said[0.039]; said[0.039]; said[0.039]; said[0.039]; national[0.039]; Ministry[0.039]; special[0.039]; Foreign[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ugandan\u001b[39m ==> ENTITY: \u001b[32mUganda\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngovernment[0.046]; northern[0.046]; troops[0.046]; rebels[0.045]; Africans[0.045]; Zaire[0.045]; Zaire[0.045]; Zaire[0.045]; Garamba[0.044]; national[0.044]; Zambia[0.044]; fighting[0.043]; Kampala[0.043]; make[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; Europeans[0.042]; mission[0.042]; official[0.042]; Foreign[0.042]; ethnic[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Roman Catholic\u001b[39m ==> ENTITY: \u001b[32mCatholic Church\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npriests[0.046]; priests[0.046]; priests[0.046]; nuns[0.044]; nuns[0.044]; nuns[0.044]; Rome[0.044]; Italy[0.043]; Italy[0.043]; national[0.043]; mission[0.041]; Italian[0.041]; Italians[0.040]; government[0.040]; ethnic[0.040]; seven[0.040]; seven[0.040]; group[0.040]; group[0.040]; special[0.039]; said[0.039]; said[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kampala\u001b[39m ==> ENTITY: \u001b[32mKampala\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUgandan[0.045]; Catholic[0.044]; northern[0.043]; priests[0.043]; priests[0.043]; priests[0.043]; troops[0.042]; Zaire[0.042]; Zaire[0.042]; Zaire[0.042]; nuns[0.042]; nuns[0.042]; nuns[0.042]; national[0.041]; government[0.041]; Zambia[0.041]; embassy[0.041]; Friday[0.041]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; official[0.040]; capital[0.040]; \t\n\r [============>.................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s391ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1276/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1365testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.100:0.100[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.048]; matches[0.046]; British[0.044]; British[0.044]; Irish[0.044]; Bristol[0.044]; Cardiff[0.043]; Swansea[0.043]; Cup[0.042]; Bath[0.041]; Reading[0.041]; Rugby[0.041]; fourth[0.041]; Gloucester[0.041]; Newbridge[0.041]; Bridgend[0.041]; Saturday[0.040]; West[0.040]; division[0.040]; division[0.040]; division[0.040]; English[0.040]; Orrell[0.039]; Caerphilly[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dunvant ==> ENTITY: \u001b[32mDunvant RFC\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.267:0.267[0]; log p(e|m)= -1.082:-1.082[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.049]; Rugby[0.048]; Pontypridd[0.045]; Harlequins[0.043]; Welsh[0.043]; Bridgend[0.042]; Wasps[0.042]; union[0.042]; Neath[0.042]; Bath[0.042]; Llanelli[0.042]; Widnes[0.042]; Cardiff[0.042]; Caerphilly[0.041]; Swansea[0.041]; Union[0.041]; Orrell[0.040]; Ebbw[0.040]; West[0.040]; Treorchy[0.039]; Bristol[0.039]; Newbridge[0.039]; Gloucester[0.039]; Irish[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wasps ==> ENTITY: \u001b[32mLondon Wasps\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.212:0.212[0]; log p(e|m)= -0.070:-0.070[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.047]; rugby[0.046]; Harlequins[0.044]; Cardiff[0.044]; Gloucester[0.044]; Bristol[0.044]; Bath[0.043]; division[0.042]; division[0.042]; division[0.042]; matches[0.042]; Welsh[0.041]; Widnes[0.041]; Pontypridd[0.041]; Swansea[0.040]; Llanelli[0.040]; London[0.040]; London[0.040]; Cup[0.040]; Bridgend[0.040]; Caerphilly[0.040]; Saturday[0.039]; Boroughmuir[0.039]; Irish[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Harlequins\u001b[39m ==> ENTITY: \u001b[32mHarlequin F.C.\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.222:0.222[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.048]; rugby[0.047]; matches[0.045]; Widnes[0.044]; Bath[0.044]; Cardiff[0.044]; Wasps[0.043]; Bristol[0.043]; Gloucester[0.042]; Swansea[0.041]; Pontypridd[0.041]; Welsh[0.041]; Boroughmuir[0.040]; Cup[0.040]; London[0.040]; London[0.040]; division[0.040]; division[0.040]; division[0.040]; Orrell[0.040]; Bridgend[0.040]; Llanelli[0.040]; Irish[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scottish ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.480:-0.480[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.046]; British[0.046]; Cardiff[0.046]; Welsh[0.045]; matches[0.044]; Irish[0.044]; division[0.043]; division[0.043]; division[0.043]; Cup[0.042]; Rugby[0.042]; Swansea[0.042]; Bristol[0.042]; fourth[0.041]; London[0.040]; London[0.040]; rugby[0.039]; English[0.039]; Bath[0.039]; Caerphilly[0.039]; Hartlepool[0.039]; Newport[0.039]; Gloucester[0.039]; Saturday[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWidnes\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mWidnes\u001b[39m <---> \u001b[32mWidnes Vikings\u001b[39m\t\nSCORES: global= 0.267:0.265[\u001b[31m0.002\u001b[39m]; local()= 0.145:0.204[\u001b[32m0.059\u001b[39m]; log p(e|m)= -0.004:-1.461[\u001b[31m1.457\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.047]; Harlequins[0.046]; Cardiff[0.046]; rugby[0.046]; Swansea[0.045]; Bristol[0.045]; matches[0.044]; Pontypridd[0.044]; Bath[0.044]; Cup[0.043]; Ebbw[0.043]; Llanelli[0.043]; Newbridge[0.043]; Saturday[0.043]; Neath[0.042]; Hartlepool[0.042]; fourth[0.042]; Gloucester[0.042]; Bridgend[0.042]; Boroughmuir[0.042]; Orrell[0.042]; Welsh[0.041]; Wasps[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mLlanelli\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mLlanelli\u001b[39m <---> \u001b[32mLlanelli RFC\u001b[39m\t\nSCORES: global= 0.256:0.255[\u001b[31m0.001\u001b[39m]; local()= 0.162:0.198[\u001b[32m0.036\u001b[39m]; log p(e|m)= -0.056:-1.911[\u001b[31m1.855\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.045]; Rugby[0.045]; Welsh[0.044]; Harlequins[0.044]; Bath[0.043]; Pontypridd[0.043]; Swansea[0.043]; Cardiff[0.043]; Bridgend[0.042]; Caerphilly[0.042]; Neath[0.041]; Bristol[0.041]; Wasps[0.041]; matches[0.041]; Widnes[0.041]; fourth[0.041]; West[0.040]; Ebbw[0.040]; Gloucester[0.040]; Newport[0.040]; Cup[0.040]; Treorchy[0.040]; Hartlepool[0.040]; Dunvant[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Neath ==> ENTITY: \u001b[32mNeath RFC\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.231:0.231[0]; log p(e|m)= -1.884:-1.884[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.045]; Welsh[0.044]; Caerphilly[0.044]; Rugby[0.044]; Pontypridd[0.044]; Bridgend[0.044]; Cardiff[0.043]; Llanelli[0.043]; Harlequins[0.042]; matches[0.042]; Gloucester[0.042]; Bath[0.042]; Bristol[0.041]; Swansea[0.041]; Widnes[0.041]; Wasps[0.041]; Newport[0.041]; Vale[0.040]; Orrell[0.040]; Ebbw[0.040]; Boroughmuir[0.039]; Scottish[0.039]; Treorchy[0.039]; Dunvant[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Welsh ==> ENTITY: \u001b[32mWales\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.140:0.140[0]; log p(e|m)= -1.079:-1.079[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.047]; rugby[0.047]; Cardiff[0.045]; Swansea[0.043]; matches[0.043]; English[0.043]; Scottish[0.042]; Bridgend[0.042]; Pontypridd[0.041]; Widnes[0.041]; Bath[0.041]; Llanelli[0.041]; Harlequins[0.041]; Bristol[0.041]; British[0.041]; British[0.041]; union[0.041]; Neath[0.040]; Treorchy[0.040]; Caerphilly[0.040]; Wasps[0.040]; Hartlepool[0.040]; Cup[0.040]; Irish[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = London Irish\u001b[39m ==> ENTITY: \u001b[32mLondon Irish\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.225:0.225[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.048]; rugby[0.047]; Harlequins[0.047]; Wasps[0.044]; Cardiff[0.044]; Bath[0.042]; Swansea[0.042]; Pontypridd[0.042]; Widnes[0.042]; Welsh[0.042]; matches[0.042]; British[0.041]; British[0.041]; Gloucester[0.041]; Bristol[0.041]; Llanelli[0.041]; Cup[0.040]; Bridgend[0.039]; Newbridge[0.039]; Neath[0.039]; Boroughmuir[0.039]; Caerphilly[0.039]; Ebbw[0.039]; London[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = English ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.106:0.106[0]; log p(e|m)= -1.355:-1.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.048]; rugby[0.048]; Cardiff[0.046]; British[0.045]; British[0.045]; Welsh[0.045]; matches[0.045]; Swansea[0.044]; Bath[0.044]; Bristol[0.044]; Harlequins[0.044]; fourth[0.042]; Gloucester[0.042]; division[0.042]; division[0.042]; division[0.042]; Reading[0.042]; Cup[0.042]; Irish[0.042]; Pontypridd[0.042]; Wasps[0.041]; Scottish[0.041]; Llanelli[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPontypridd\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPontypridd\u001b[39m <---> \u001b[32mPontypridd RFC\u001b[39m\t\nSCORES: global= 0.269:0.266[\u001b[31m0.003\u001b[39m]; local()= 0.161:0.185[\u001b[32m0.024\u001b[39m]; log p(e|m)= 0.000:-1.820[\u001b[31m1.820\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.044]; rugby[0.044]; Welsh[0.043]; Cardiff[0.043]; Swansea[0.042]; Bristol[0.042]; Bath[0.042]; Dunvant[0.042]; Ebbw[0.042]; Caerphilly[0.042]; Vale[0.042]; Neath[0.042]; Harlequins[0.042]; Bridgend[0.042]; matches[0.041]; Treorchy[0.041]; Orrell[0.041]; division[0.041]; division[0.041]; division[0.041]; Widnes[0.040]; Llanelli[0.040]; West[0.040]; Wasps[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = British ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.111:0.111[0]; log p(e|m)= -0.489:-0.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.049]; Welsh[0.047]; English[0.045]; Bristol[0.045]; Cardiff[0.043]; London[0.043]; London[0.043]; matches[0.042]; Bath[0.042]; Irish[0.042]; Scottish[0.041]; West[0.041]; Swansea[0.041]; rugby[0.040]; Pontypridd[0.040]; Neath[0.040]; Union[0.040]; Gloucester[0.040]; Hartlepool[0.040]; Newport[0.040]; division[0.040]; division[0.040]; division[0.040]; Bridgend[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Boroughmuir\u001b[39m ==> ENTITY: \u001b[32mBoroughmuir RFC\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.047]; Rugby[0.047]; matches[0.044]; Harlequins[0.043]; Welsh[0.043]; Cardiff[0.043]; Cup[0.042]; Watsonians[0.042]; Wasps[0.042]; Scottish[0.042]; Orrell[0.041]; Bath[0.041]; fourth[0.041]; British[0.041]; British[0.041]; Widnes[0.041]; Pontypridd[0.041]; Llanelli[0.041]; Bridgend[0.040]; Caerphilly[0.040]; Saturday[0.040]; Treorchy[0.040]; Irish[0.040]; Swansea[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNewbridge\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNewbridge, Caerphilly\u001b[39m <---> \u001b[32mNewbridge RFC\u001b[39m\t\nSCORES: global= 0.256:0.243[\u001b[31m0.013\u001b[39m]; local()= 0.195:0.140[\u001b[31m0.054\u001b[39m]; log p(e|m)= -1.833:-2.386[\u001b[31m0.553\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPontypridd[0.044]; rugby[0.044]; Caerphilly[0.044]; Bridgend[0.043]; Welsh[0.043]; Widnes[0.042]; Rugby[0.042]; Irish[0.042]; Bristol[0.042]; Llanelli[0.042]; Swansea[0.042]; Neath[0.041]; Treorchy[0.041]; Dunvant[0.041]; Cardiff[0.041]; Scottish[0.041]; Hartlepool[0.041]; West[0.041]; Ebbw[0.041]; Bath[0.040]; Newport[0.040]; Vale[0.040]; Harlequins[0.040]; Orrell[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Hartlepool ==> ENTITY: \u001b[32mWest Hartlepool R.F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.166:0.166[0]; log p(e|m)= -0.865:-0.865[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.045]; rugby[0.044]; matches[0.044]; Bristol[0.043]; Cardiff[0.043]; Widnes[0.042]; Swansea[0.042]; division[0.042]; division[0.042]; division[0.042]; Orrell[0.042]; Newport[0.042]; Bridgend[0.041]; Harlequins[0.041]; Caerphilly[0.041]; round[0.041]; Vale[0.041]; Cup[0.041]; fourth[0.040]; Boroughmuir[0.040]; Pontypridd[0.040]; Ebbw[0.040]; Saturday[0.040]; Scottish[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bath ==> ENTITY: \u001b[32mBath Rugby\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.189:0.189[0]; log p(e|m)= -2.604:-2.604[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBristol[0.049]; Gloucester[0.048]; Cardiff[0.046]; Swansea[0.045]; Pontypridd[0.044]; Llanelli[0.043]; Neath[0.042]; Widnes[0.042]; Rugby[0.042]; rugby[0.042]; Harlequins[0.041]; Dunvant[0.041]; Wasps[0.041]; Hartlepool[0.041]; Welsh[0.041]; Bridgend[0.040]; Scottish[0.040]; Caerphilly[0.039]; British[0.039]; British[0.039]; London[0.039]; London[0.039]; Orrell[0.039]; Boroughmuir[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mTreorchy\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mTreorchy\u001b[39m <---> \u001b[32mTreorchy RFC\u001b[39m\t\nSCORES: global= 0.289:0.283[\u001b[31m0.006\u001b[39m]; local()= 0.131:0.190[\u001b[32m0.059\u001b[39m]; log p(e|m)= 0.000:-0.942[\u001b[31m0.942\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.046]; Rugby[0.046]; Welsh[0.045]; Pontypridd[0.044]; Bridgend[0.044]; Caerphilly[0.043]; Cardiff[0.043]; Harlequins[0.043]; Swansea[0.042]; Wasps[0.042]; Bristol[0.041]; Llanelli[0.041]; Neath[0.041]; Ebbw[0.041]; Widnes[0.041]; Vale[0.041]; West[0.040]; Boroughmuir[0.040]; Gloucester[0.040]; matches[0.040]; Irish[0.040]; Newport[0.040]; Scottish[0.040]; Newbridge[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Caerphilly ==> ENTITY: \u001b[32mCaerphilly RFC\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.192:0.192[0]; log p(e|m)= -1.645:-1.645[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.045]; rugby[0.045]; Welsh[0.044]; Pontypridd[0.044]; Swansea[0.043]; Bridgend[0.043]; Cardiff[0.043]; Widnes[0.042]; Harlequins[0.042]; Llanelli[0.041]; Neath[0.041]; matches[0.041]; Wasps[0.041]; Treorchy[0.041]; Bristol[0.041]; Newport[0.041]; Ebbw[0.041]; Orrell[0.040]; Hartlepool[0.040]; division[0.040]; division[0.040]; division[0.040]; Cup[0.040]; Bath[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bridgend ==> ENTITY: \u001b[32mBridgend Ravens\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.217:0.217[0]; log p(e|m)= -2.040:-2.040[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.046]; Harlequins[0.045]; Rugby[0.045]; Cardiff[0.044]; Pontypridd[0.043]; Swansea[0.043]; Welsh[0.043]; Wasps[0.042]; Caerphilly[0.042]; Bristol[0.042]; Bath[0.042]; Llanelli[0.042]; Widnes[0.042]; matches[0.042]; Neath[0.041]; Gloucester[0.041]; Treorchy[0.041]; Ebbw[0.041]; Irish[0.040]; Dunvant[0.039]; Newport[0.039]; Newbridge[0.039]; Orrell[0.039]; Boroughmuir[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBristol\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBristol\u001b[39m <---> \u001b[32mBristol Rugby\u001b[39m\t\nSCORES: global= 0.258:0.255[\u001b[31m0.003\u001b[39m]; local()= 0.130:0.214[\u001b[32m0.084\u001b[39m]; log p(e|m)= -0.435:-3.772[\u001b[31m3.337\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.046]; rugby[0.046]; Cardiff[0.045]; Harlequins[0.045]; Swansea[0.044]; Bath[0.044]; Wasps[0.043]; Gloucester[0.043]; Welsh[0.042]; Pontypridd[0.042]; Newport[0.041]; Widnes[0.041]; matches[0.041]; Llanelli[0.041]; London[0.040]; London[0.040]; Saturday[0.040]; British[0.040]; British[0.040]; Irish[0.040]; Neath[0.039]; Caerphilly[0.039]; Bridgend[0.039]; Boroughmuir[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gloucester ==> ENTITY: \u001b[32mGloucester Rugby\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.226:0.226[0]; log p(e|m)= -3.058:-3.058[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.048]; rugby[0.047]; Cardiff[0.045]; Bristol[0.044]; Harlequins[0.044]; Bath[0.044]; Welsh[0.043]; Wasps[0.043]; matches[0.042]; Pontypridd[0.042]; Newport[0.042]; Swansea[0.042]; Widnes[0.041]; Llanelli[0.041]; Bridgend[0.041]; Caerphilly[0.040]; West[0.040]; Neath[0.040]; Treorchy[0.039]; Scottish[0.039]; Cup[0.039]; English[0.038]; London[0.038]; London[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Watsonians\u001b[39m ==> ENTITY: \u001b[32mWatsonians RFC\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.048]; rugby[0.048]; matches[0.043]; Scottish[0.043]; Boroughmuir[0.043]; Welsh[0.043]; Cardiff[0.043]; Cup[0.042]; Gloucester[0.042]; Harlequins[0.041]; Swansea[0.041]; Bath[0.041]; Wasps[0.041]; Pontypridd[0.041]; Llanelli[0.041]; Widnes[0.041]; Caerphilly[0.040]; Bristol[0.040]; Dunvant[0.040]; Bridgend[0.040]; Neath[0.040]; Ebbw[0.040]; Saturday[0.039]; Irish[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rugby Union\u001b[39m ==> ENTITY: \u001b[32mRugby union\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.054]; Cardiff[0.045]; Bath[0.044]; matches[0.043]; Pontypridd[0.043]; Harlequins[0.043]; Bristol[0.043]; Welsh[0.043]; Irish[0.042]; Gloucester[0.042]; Widnes[0.041]; Bridgend[0.041]; Llanelli[0.041]; Caerphilly[0.041]; Cup[0.041]; Wasps[0.040]; Swansea[0.040]; Neath[0.040]; Scottish[0.039]; Boroughmuir[0.039]; Newport[0.039]; fourth[0.039]; Saturday[0.039]; Newbridge[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pilkington Cup\u001b[39m ==> ENTITY: \u001b[32mAnglo-Welsh Cup\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.209:0.209[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.051]; rugby[0.050]; Harlequins[0.049]; Wasps[0.048]; matches[0.048]; Cardiff[0.048]; Bristol[0.047]; Gloucester[0.047]; Bath[0.047]; Widnes[0.045]; Welsh[0.045]; Llanelli[0.044]; Irish[0.044]; fourth[0.044]; Swansea[0.044]; Pontypridd[0.043]; Scottish[0.043]; Neath[0.043]; Saturday[0.042]; Boroughmuir[0.042]; Bridgend[0.042]; Ebbw[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEbbw Vale\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEbbw Vale\u001b[39m <---> \u001b[32mEbbw Vale RFC\u001b[39m\t\nSCORES: global= 0.270:0.261[\u001b[31m0.009\u001b[39m]; local()= 0.149:0.200[\u001b[32m0.051\u001b[39m]; log p(e|m)= 0.000:-1.864[\u001b[31m1.864\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.045]; rugby[0.045]; Cardiff[0.044]; Swansea[0.043]; Llanelli[0.043]; matches[0.042]; Caerphilly[0.042]; Bridgend[0.042]; Gloucester[0.042]; Pontypridd[0.042]; Harlequins[0.042]; Bristol[0.042]; Welsh[0.042]; Newport[0.042]; Bath[0.041]; Widnes[0.041]; Neath[0.041]; Wasps[0.040]; Newbridge[0.040]; Orrell[0.040]; Cup[0.040]; Boroughmuir[0.040]; Hartlepool[0.039]; Dunvant[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCardiff\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mCardiff\u001b[39m <---> \u001b[32mCardiff RFC\u001b[39m\t\nSCORES: global= 0.248:0.240[\u001b[31m0.008\u001b[39m]; local()= 0.166:0.186[\u001b[32m0.020\u001b[39m]; log p(e|m)= -0.205:-2.781[\u001b[31m2.576\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.046]; rugby[0.046]; Swansea[0.044]; Harlequins[0.044]; Welsh[0.044]; matches[0.043]; Bristol[0.043]; Pontypridd[0.043]; Wasps[0.042]; Bath[0.042]; Llanelli[0.041]; Newport[0.041]; Gloucester[0.041]; Bridgend[0.041]; London[0.041]; London[0.041]; Caerphilly[0.040]; Widnes[0.040]; fourth[0.040]; Reading[0.040]; Hartlepool[0.040]; Neath[0.040]; Saturday[0.039]; Scottish[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Orrell ==> ENTITY: \u001b[32mOrrell R.U.F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.177:0.177[0]; log p(e|m)= -0.931:-0.931[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.044]; Cardiff[0.044]; Bristol[0.044]; Widnes[0.044]; rugby[0.044]; Gloucester[0.043]; Swansea[0.043]; Bath[0.043]; Irish[0.043]; Harlequins[0.043]; Bridgend[0.042]; Welsh[0.042]; Pontypridd[0.041]; Ebbw[0.041]; matches[0.041]; Boroughmuir[0.041]; Caerphilly[0.040]; English[0.040]; Wasps[0.040]; Dunvant[0.040]; Reading[0.040]; Newport[0.039]; Cup[0.039]; Llanelli[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swansea ==> ENTITY: \u001b[32mSwansea RFC\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.216:0.216[0]; log p(e|m)= -2.375:-2.375[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.046]; Rugby[0.046]; Cardiff[0.045]; Pontypridd[0.044]; Welsh[0.044]; Llanelli[0.043]; Caerphilly[0.043]; Bristol[0.043]; matches[0.043]; Harlequins[0.042]; Bridgend[0.042]; Neath[0.041]; Widnes[0.041]; Ebbw[0.041]; fourth[0.041]; Bath[0.040]; Reading[0.040]; Gloucester[0.040]; Saturday[0.040]; Wasps[0.040]; Treorchy[0.040]; Dunvant[0.039]; Cup[0.039]; Newport[0.039]; \t\n\r [============>.................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s104ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1307/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1178testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Indies ==> ENTITY: \u001b[32mWest Indies cricket team\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.167:0.167[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWest[0.045]; West[0.045]; Lara[0.044]; Australia[0.044]; Australia[0.044]; Chanderpaul[0.043]; Waugh[0.042]; Waugh[0.042]; Cricket[0.042]; match[0.042]; Indies[0.042]; Indies[0.042]; overs[0.042]; Samuels[0.041]; Gillespie[0.041]; Gillespie[0.041]; Warne[0.040]; Warne[0.040]; Warne[0.040]; lbw[0.040]; World[0.040]; World[0.040]; run[0.039]; run[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moody ==> ENTITY: \u001b[32mTom Moody\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.125:0.125[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReiffel[0.043]; Reiffel[0.043]; Moody[0.043]; Moody[0.043]; Moody[0.043]; Australia[0.043]; overs[0.042]; overs[0.042]; Taylor[0.042]; bat[0.042]; Gillespie[0.041]; wickets[0.041]; wickets[0.041]; wickets[0.041]; Healy[0.041]; Healy[0.041]; Bevan[0.041]; Waugh[0.041]; Waugh[0.041]; Waugh[0.041]; Warne[0.041]; Warne[0.041]; Warne[0.041]; Warne[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Healy ==> ENTITY: \u001b[32mIan Healy\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.156:0.156[0]; log p(e|m)= -3.650:-3.650[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.046]; Australia[0.046]; match[0.044]; Healy[0.044]; Reiffel[0.043]; Waugh[0.043]; Waugh[0.043]; overs[0.042]; overs[0.042]; Campbell[0.041]; Melbourne[0.041]; Gillespie[0.041]; Gillespie[0.041]; Gillespie[0.041]; Warne[0.041]; Warne[0.041]; Warne[0.041]; Cricket[0.041]; run[0.040]; run[0.040]; Walsh[0.040]; wickets[0.040]; Bowling[0.039]; Lara[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Healy ==> ENTITY: \u001b[32mIan Healy\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.153:0.153[0]; log p(e|m)= -3.650:-3.650[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.047]; Australia[0.047]; match[0.045]; Healy[0.044]; Waugh[0.043]; Waugh[0.043]; overs[0.042]; overs[0.042]; Campbell[0.042]; Melbourne[0.041]; Gillespie[0.041]; Gillespie[0.041]; Warne[0.041]; Warne[0.041]; Warne[0.041]; Cricket[0.041]; run[0.040]; run[0.040]; Walsh[0.040]; wickets[0.040]; Lara[0.039]; Murray[0.039]; World[0.039]; World[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.171:0.171[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.050]; match[0.045]; Cricket[0.045]; Waugh[0.044]; Waugh[0.044]; Melbourne[0.043]; Warne[0.041]; Warne[0.041]; Warne[0.041]; overs[0.041]; overs[0.041]; Gillespie[0.041]; Gillespie[0.041]; wickets[0.041]; Murray[0.040]; Lara[0.040]; Scoreboard[0.040]; Scoreboard[0.040]; World[0.040]; World[0.040]; West[0.040]; West[0.040]; West[0.040]; Samuels[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Warne ==> ENTITY: \u001b[32mShane Warne\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.208:0.208[0]; log p(e|m)= -0.493:-0.493[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarne[0.044]; Warne[0.044]; Warne[0.044]; Reiffel[0.043]; Australia[0.043]; Australia[0.043]; Australia[0.043]; Waugh[0.042]; Waugh[0.042]; Waugh[0.042]; Waugh[0.042]; Cricket[0.042]; Ponting[0.041]; match[0.041]; overs[0.041]; overs[0.041]; Lara[0.041]; wickets[0.041]; Gillespie[0.040]; Gillespie[0.040]; Gillespie[0.040]; Chanderpaul[0.040]; lbw[0.040]; lbw[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hooper ==> ENTITY: \u001b[32mCarl Hooper\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.181:0.181[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReiffel[0.045]; Reiffel[0.045]; bat[0.043]; Waugh[0.043]; Waugh[0.043]; Bevan[0.043]; Gillespie[0.043]; Gillespie[0.043]; Australia[0.041]; Australia[0.041]; Ponting[0.041]; Taylor[0.041]; wickets[0.041]; wickets[0.041]; wickets[0.041]; wickets[0.041]; Warne[0.041]; Warne[0.041]; Hooper[0.041]; Hooper[0.041]; lbw[0.040]; Bowling[0.040]; Bowling[0.040]; Healy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moody ==> ENTITY: \u001b[32mTom Moody\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.119:0.119[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReiffel[0.047]; Moody[0.047]; Moody[0.047]; Cricket[0.047]; Australia[0.046]; Australia[0.046]; overs[0.046]; overs[0.046]; West[0.046]; West[0.046]; West[0.046]; match[0.045]; Gillespie[0.045]; Gillespie[0.045]; Gillespie[0.045]; wickets[0.044]; Healy[0.044]; Healy[0.044]; Waugh[0.044]; Waugh[0.044]; Waugh[0.044]; Friday[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gillespie ==> ENTITY: \u001b[32mJason Gillespie\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.169:0.169[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.046]; Australia[0.046]; Reiffel[0.045]; Gillespie[0.044]; Gillespie[0.044]; Chanderpaul[0.044]; Waugh[0.044]; Waugh[0.044]; lbw[0.044]; wickets[0.044]; overs[0.044]; overs[0.044]; Warne[0.043]; Warne[0.043]; Warne[0.043]; match[0.043]; Adams[0.043]; Lara[0.043]; Scoreboard[0.042]; Scoreboard[0.042]; Cricket[0.042]; World[0.042]; World[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Indies ==> ENTITY: \u001b[32mWest Indies cricket team\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.176:0.176[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLara[0.046]; Australia[0.046]; Australia[0.046]; Chanderpaul[0.046]; match[0.045]; Cricket[0.045]; Waugh[0.044]; Waugh[0.044]; overs[0.044]; overs[0.044]; Gillespie[0.044]; Gillespie[0.044]; Indies[0.043]; Indies[0.043]; Samuels[0.043]; wickets[0.042]; lbw[0.042]; Warne[0.042]; Warne[0.042]; Warne[0.042]; Adams[0.042]; World[0.042]; World[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Blewett ==> ENTITY: \u001b[32mGreg Blewett\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.178:0.178[0]; log p(e|m)= -2.079:-2.079[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.044]; Blewett[0.044]; Blewett[0.044]; Blewett[0.044]; bat[0.043]; Reiffel[0.042]; Reiffel[0.042]; Warne[0.042]; Warne[0.042]; Warne[0.042]; Warne[0.042]; Waugh[0.041]; Waugh[0.041]; Waugh[0.041]; overs[0.041]; overs[0.041]; Ponting[0.041]; Gillespie[0.040]; Gillespie[0.040]; wickets[0.040]; wickets[0.040]; wickets[0.040]; Bevan[0.040]; Chanderpaul[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWorld Series\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustralian Tri-Series\u001b[39m <---> \u001b[32mWorld Series Cricket\u001b[39m\t\nSCORES: global= 0.257:0.255[\u001b[31m0.002\u001b[39m]; local()= 0.177:0.163[\u001b[31m0.014\u001b[39m]; log p(e|m)= -3.324:-3.324[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; Australia[0.048]; Cricket[0.045]; match[0.044]; Waugh[0.043]; Waugh[0.043]; Warne[0.042]; Warne[0.042]; Warne[0.042]; Series[0.042]; Melbourne[0.041]; World[0.041]; Gillespie[0.041]; Gillespie[0.041]; Scoreboard[0.040]; Scoreboard[0.040]; Friday[0.040]; Adams[0.040]; Samuels[0.040]; run[0.039]; run[0.039]; overs[0.039]; Healy[0.039]; Healy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hooper ==> ENTITY: \u001b[32mCarl Hooper\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.169:0.169[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReiffel[0.048]; bat[0.046]; Waugh[0.046]; Bevan[0.046]; Gillespie[0.046]; Australia[0.044]; Ponting[0.044]; Taylor[0.044]; wickets[0.044]; wickets[0.044]; wickets[0.044]; Warne[0.044]; Hooper[0.043]; Hooper[0.043]; lbw[0.043]; Bowling[0.042]; Healy[0.042]; overs[0.042]; Blewett[0.042]; won[0.041]; Ambrose[0.041]; Benjamin[0.040]; Benjamin[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reiffel\u001b[39m ==> ENTITY: \u001b[32mPaul Reiffel\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.081:0.081[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.046]; Lara[0.045]; Waugh[0.045]; Waugh[0.045]; Waugh[0.045]; Waugh[0.045]; Warne[0.043]; Warne[0.043]; Warne[0.043]; Warne[0.043]; lb[0.043]; lb[0.043]; Bevan[0.043]; lbw[0.043]; lbw[0.043]; Ponting[0.043]; Fall[0.043]; Fall[0.043]; Total[0.043]; Total[0.043]; Bowling[0.043]; overs[0.042]; overs[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moody ==> ENTITY: \u001b[32mTom Moody\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.119:0.119[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReiffel[0.043]; Moody[0.043]; Moody[0.043]; Cricket[0.043]; Australia[0.043]; Australia[0.043]; overs[0.042]; overs[0.042]; West[0.042]; West[0.042]; West[0.042]; match[0.042]; Gillespie[0.041]; Gillespie[0.041]; Gillespie[0.041]; wickets[0.041]; Healy[0.041]; Healy[0.041]; Waugh[0.041]; Waugh[0.041]; Friday[0.041]; Warne[0.041]; Warne[0.041]; Warne[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Indies ==> ENTITY: \u001b[32mWest Indies cricket team\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.176:0.176[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLara[0.046]; Australia[0.046]; Australia[0.046]; Chanderpaul[0.046]; match[0.045]; Cricket[0.045]; Waugh[0.044]; Waugh[0.044]; overs[0.044]; overs[0.044]; Gillespie[0.044]; Gillespie[0.044]; Indies[0.043]; Indies[0.043]; Samuels[0.043]; wickets[0.042]; lbw[0.042]; Warne[0.042]; Warne[0.042]; Warne[0.042]; Adams[0.042]; World[0.042]; World[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hooper ==> ENTITY: \u001b[32mCarl Hooper\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.181:0.181[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReiffel[0.045]; Reiffel[0.045]; bat[0.043]; Waugh[0.043]; Waugh[0.043]; Bevan[0.043]; Gillespie[0.043]; Gillespie[0.043]; Australia[0.041]; Australia[0.041]; Ponting[0.041]; Taylor[0.041]; wickets[0.041]; wickets[0.041]; wickets[0.041]; wickets[0.041]; Warne[0.041]; Warne[0.041]; Hooper[0.041]; Hooper[0.041]; lbw[0.040]; Bowling[0.040]; Bowling[0.040]; Healy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Blewett ==> ENTITY: \u001b[32mGreg Blewett\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.187:0.187[0]; log p(e|m)= -2.079:-2.079[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.044]; Australia[0.044]; Australia[0.044]; Blewett[0.044]; Blewett[0.044]; match[0.043]; Reiffel[0.042]; Warne[0.042]; Warne[0.042]; Warne[0.042]; Warne[0.042]; Cricket[0.042]; Waugh[0.041]; Waugh[0.041]; Waugh[0.041]; Lara[0.041]; overs[0.040]; overs[0.040]; Gillespie[0.040]; Gillespie[0.040]; Gillespie[0.040]; Friday[0.040]; wickets[0.040]; Chanderpaul[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Warne ==> ENTITY: \u001b[32mShane Warne\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.193:0.193[0]; log p(e|m)= -0.493:-0.493[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarne[0.044]; Warne[0.044]; Warne[0.044]; Reiffel[0.044]; Reiffel[0.044]; Australia[0.043]; Waugh[0.043]; Waugh[0.043]; Waugh[0.043]; Ponting[0.042]; overs[0.042]; overs[0.042]; wickets[0.041]; wickets[0.041]; wickets[0.041]; bat[0.041]; Bevan[0.041]; Gillespie[0.040]; Gillespie[0.040]; lbw[0.040]; lbw[0.040]; Bowling[0.039]; Bowling[0.039]; Taylor[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Blewett ==> ENTITY: \u001b[32mGreg Blewett\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.176:0.176[0]; log p(e|m)= -2.079:-2.079[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.045]; Australia[0.045]; Blewett[0.045]; Blewett[0.045]; match[0.044]; Reiffel[0.043]; Warne[0.042]; Warne[0.042]; Warne[0.042]; Cricket[0.042]; Waugh[0.042]; Waugh[0.042]; Lara[0.041]; overs[0.041]; overs[0.041]; Gillespie[0.041]; Gillespie[0.041]; Gillespie[0.041]; Friday[0.040]; wickets[0.040]; Chanderpaul[0.040]; lbw[0.040]; Samuels[0.039]; Benjamin[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gillespie ==> ENTITY: \u001b[32mJason Gillespie\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.172:0.172[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.044]; Australia[0.044]; Reiffel[0.044]; Gillespie[0.042]; Chanderpaul[0.042]; Waugh[0.042]; Waugh[0.042]; lbw[0.042]; wickets[0.042]; overs[0.042]; overs[0.042]; Warne[0.042]; Warne[0.042]; Warne[0.042]; match[0.042]; Adams[0.041]; Lara[0.041]; Scoreboard[0.041]; Scoreboard[0.041]; Cricket[0.040]; World[0.040]; World[0.040]; run[0.039]; run[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.092:0.092[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReiffel[0.045]; bat[0.045]; Warne[0.044]; overs[0.044]; Gillespie[0.043]; wickets[0.043]; wickets[0.043]; wickets[0.043]; Bevan[0.043]; won[0.043]; Bowling[0.042]; Healy[0.041]; Ambrose[0.040]; Murray[0.040]; Blewett[0.040]; McLean[0.040]; McLean[0.040]; Walsh[0.040]; lbw[0.040]; Hooper[0.040]; Hooper[0.040]; Hooper[0.040]; Benjamin[0.039]; Moody[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.175:0.175[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.050]; match[0.046]; Cricket[0.045]; Waugh[0.044]; Waugh[0.044]; Melbourne[0.043]; Warne[0.042]; Warne[0.042]; Warne[0.042]; overs[0.042]; Gillespie[0.041]; Gillespie[0.041]; Lara[0.041]; Scoreboard[0.040]; Scoreboard[0.040]; World[0.040]; World[0.040]; West[0.040]; West[0.040]; West[0.040]; Samuels[0.040]; Indies[0.040]; Indies[0.040]; Indies[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Melbourne ==> ENTITY: \u001b[32mMelbourne\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.081:0.081[0]; log p(e|m)= -0.168:-0.168[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.049]; Australia[0.049]; match[0.048]; Cricket[0.045]; Waugh[0.043]; Waugh[0.043]; Scoreboard[0.042]; Scoreboard[0.042]; Murray[0.042]; Friday[0.041]; Warne[0.040]; Warne[0.040]; Warne[0.040]; Campbell[0.040]; McLean[0.040]; overs[0.040]; West[0.040]; West[0.040]; West[0.040]; Walsh[0.040]; Healy[0.040]; Healy[0.040]; run[0.039]; run[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gillespie ==> ENTITY: \u001b[32mJason Gillespie\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.187:0.187[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.045]; Reiffel[0.043]; bat[0.043]; Ponting[0.043]; Gillespie[0.042]; Chanderpaul[0.042]; Waugh[0.042]; Waugh[0.042]; Waugh[0.042]; lbw[0.042]; lbw[0.042]; wickets[0.042]; wickets[0.042]; wickets[0.042]; overs[0.041]; overs[0.041]; Bevan[0.041]; Warne[0.041]; Warne[0.041]; Warne[0.041]; Warne[0.041]; Adams[0.041]; Lara[0.041]; Taylor[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ambrose ==> ENTITY: \u001b[32mCurtly Ambrose\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.190:0.190[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGillespie[0.045]; Waugh[0.045]; Waugh[0.045]; overs[0.043]; bat[0.043]; Taylor[0.043]; Reiffel[0.042]; wickets[0.042]; wickets[0.042]; wickets[0.042]; Bevan[0.042]; Warne[0.042]; Warne[0.042]; Walsh[0.042]; Australia[0.041]; Australia[0.041]; Ponting[0.041]; Healy[0.041]; lbw[0.040]; Murray[0.039]; Murray[0.039]; Blewett[0.039]; Blewett[0.039]; Bowling[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Warne ==> ENTITY: \u001b[32mShane Warne\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.493:-0.493[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarne[0.044]; Warne[0.044]; Warne[0.044]; Reiffel[0.044]; Australia[0.043]; Australia[0.043]; Australia[0.043]; Waugh[0.043]; Waugh[0.043]; Waugh[0.043]; Cricket[0.042]; match[0.042]; overs[0.042]; overs[0.042]; Lara[0.041]; wickets[0.041]; Gillespie[0.040]; Gillespie[0.040]; Gillespie[0.040]; Chanderpaul[0.040]; lbw[0.040]; Bowling[0.039]; Taylor[0.039]; Adams[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Warne ==> ENTITY: \u001b[32mShane Warne\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.191:0.191[0]; log p(e|m)= -0.493:-0.493[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarne[0.047]; Warne[0.047]; Reiffel[0.046]; Australia[0.046]; Australia[0.046]; Waugh[0.045]; Waugh[0.045]; Cricket[0.045]; match[0.044]; overs[0.044]; overs[0.044]; Lara[0.044]; wickets[0.044]; Gillespie[0.042]; Gillespie[0.042]; Gillespie[0.042]; Chanderpaul[0.042]; lbw[0.042]; Bowling[0.042]; Melbourne[0.041]; Samuels[0.041]; run[0.040]; run[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.148:0.148[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWaugh[0.047]; Waugh[0.047]; Waugh[0.047]; Reiffel[0.046]; Reiffel[0.046]; Walsh[0.046]; Walsh[0.046]; bat[0.046]; Taylor[0.046]; Warne[0.045]; Warne[0.045]; Warne[0.045]; Warne[0.045]; overs[0.045]; overs[0.045]; Gillespie[0.045]; Gillespie[0.045]; wickets[0.044]; wickets[0.044]; wickets[0.044]; Bevan[0.044]; Ponting[0.044]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWorld Series\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustralian Tri-Series\u001b[39m <---> \u001b[32mWorld Series Cricket\u001b[39m\t\nSCORES: global= 0.258:0.256[\u001b[31m0.002\u001b[39m]; local()= 0.186:0.171[\u001b[31m0.015\u001b[39m]; log p(e|m)= -3.324:-3.324[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; Australia[0.048]; Cricket[0.045]; match[0.044]; Waugh[0.043]; Waugh[0.043]; Warne[0.042]; Warne[0.042]; Warne[0.042]; Series[0.042]; Melbourne[0.041]; World[0.041]; Gillespie[0.041]; Gillespie[0.041]; Scoreboard[0.040]; Scoreboard[0.040]; Friday[0.040]; Adams[0.040]; Samuels[0.040]; run[0.039]; run[0.039]; overs[0.039]; overs[0.039]; wickets[0.039]; \t\n\r [============>.................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s904ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1338/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1196testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Amsterdam ==> ENTITY: \u001b[32mAmsterdam\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.050:0.050[0]; log p(e|m)= -0.064:-0.064[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDutch[0.053]; Dutch[0.053]; Waalwijk[0.052]; Van[0.051]; Soccer[0.051]; soccer[0.051]; Tilburg[0.049]; played[0.049]; Friday[0.048]; Attendance[0.048]; Willem[0.047]; Summary[0.047]; der[0.046]; match[0.046]; division[0.046]; Halftime[0.045]; Starbuck[0.044]; Konterman[0.044]; Division[0.044]; Summaries[0.043]; Ii[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = RKC Waalwijk\u001b[39m ==> ENTITY: \u001b[32mRKC Waalwijk\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.128:0.128[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.053]; Soccer[0.052]; played[0.052]; match[0.051]; Dutch[0.049]; Dutch[0.049]; Amsterdam[0.049]; Halftime[0.049]; Konterman[0.048]; division[0.048]; Division[0.048]; Friday[0.047]; Tilburg[0.047]; Summaries[0.047]; Van[0.046]; Attendance[0.046]; der[0.046]; Willem[0.045]; Ii[0.044]; Summary[0.044]; Starbuck[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dutch ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.072:0.072[0]; log p(e|m)= -0.794:-0.794[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDutch[0.059]; match[0.056]; played[0.055]; soccer[0.055]; Van[0.053]; Soccer[0.053]; Waalwijk[0.053]; Amsterdam[0.052]; Tilburg[0.051]; Willem[0.049]; Konterman[0.048]; Friday[0.048]; Division[0.048]; Halftime[0.047]; der[0.047]; Attendance[0.047]; division[0.047]; Summary[0.046]; Starbuck[0.042]; Summaries[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Konterman\u001b[39m ==> ENTITY: \u001b[32mBert Konterman\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.066:0.066[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.053]; Halftime[0.051]; Soccer[0.050]; played[0.050]; Dutch[0.049]; Dutch[0.049]; match[0.049]; Waalwijk[0.049]; Summaries[0.049]; Attendance[0.048]; der[0.048]; Division[0.047]; Van[0.047]; Willem[0.047]; division[0.046]; Ii[0.046]; Summary[0.045]; Tilburg[0.045]; Amsterdam[0.045]; Friday[0.044]; Starbuck[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Willem II Tilburg\u001b[39m ==> ENTITY: \u001b[32mWillem II (football club)\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.137:0.137[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.057]; Waalwijk[0.056]; match[0.056]; played[0.055]; Division[0.055]; Soccer[0.055]; Dutch[0.054]; Dutch[0.054]; Konterman[0.054]; division[0.054]; Halftime[0.053]; Summary[0.051]; Amsterdam[0.051]; Summaries[0.051]; Friday[0.050]; Attendance[0.050]; der[0.050]; Van[0.050]; Starbuck[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dutch ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.072:0.072[0]; log p(e|m)= -0.794:-0.794[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDutch[0.059]; match[0.056]; played[0.055]; soccer[0.055]; Van[0.053]; Soccer[0.053]; Waalwijk[0.053]; Amsterdam[0.052]; Tilburg[0.051]; Willem[0.049]; Konterman[0.048]; Friday[0.048]; Division[0.048]; Halftime[0.047]; der[0.047]; Attendance[0.047]; division[0.047]; Summary[0.046]; Starbuck[0.042]; Summaries[0.042]; \t\n\r [============>.................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s893ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1344/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1311testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = N.M.\u001b[39m ==> ENTITY: \u001b[32mNew Mexico\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.047:0.047[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.048]; Mexico[0.048]; scholastic[0.043]; Albuquerque[0.042]; Albuquerque[0.042]; forearm[0.042]; school[0.042]; school[0.042]; football[0.042]; football[0.042]; School[0.042]; School[0.042]; inter[0.041]; Activities[0.041]; helmet[0.040]; October[0.040]; razor[0.040]; buckles[0.040]; buckles[0.040]; expelled[0.040]; expelled[0.040]; game[0.040]; stitches[0.040]; strap[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Mexico ==> ENTITY: \u001b[32mNew Mexico\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.047:0.047[0]; log p(e|m)= -0.200:-0.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfootball[0.047]; football[0.047]; Mexico[0.045]; game[0.045]; Albuquerque[0.043]; Albuquerque[0.043]; players[0.043]; players[0.043]; player[0.043]; player[0.043]; Mike[0.043]; Thursday[0.041]; opponents[0.040]; opponents[0.040]; competition[0.040]; competition[0.040]; school[0.040]; school[0.040]; high[0.040]; high[0.040]; October[0.040]; helmet[0.039]; Activities[0.039]; scholastic[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Albuquerque ==> ENTITY: \u001b[32mAlbuquerque, New Mexico\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.080:0.080[0]; log p(e|m)= -0.109:-0.109[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayer[0.046]; player[0.046]; players[0.044]; players[0.044]; players[0.044]; Mexico[0.044]; Mexico[0.044]; Albuquerque[0.043]; football[0.043]; football[0.043]; game[0.043]; game[0.043]; October[0.041]; October[0.041]; saying[0.041]; said[0.040]; Thursday[0.040]; competition[0.039]; competition[0.039]; filing[0.039]; Association[0.038]; Officials[0.038]; bar[0.038]; fine[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = New Mexico Activities Association\u001b[39m ==> ENTITY: \u001b[32mNew Mexico Activities Association\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.048]; football[0.047]; football[0.047]; High[0.044]; school[0.043]; school[0.043]; school[0.043]; scholastic[0.043]; School[0.043]; School[0.043]; Officials[0.043]; player[0.043]; player[0.043]; forearm[0.043]; Albuquerque[0.043]; Albuquerque[0.043]; competition[0.042]; competition[0.042]; high[0.042]; high[0.042]; attends[0.042]; game[0.042]; game[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Albuquerque ==> ENTITY: \u001b[32mAlbuquerque, New Mexico\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.087:0.087[0]; log p(e|m)= -0.109:-0.109[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayer[0.047]; player[0.047]; players[0.045]; players[0.045]; Mexico[0.045]; Mexico[0.045]; Albuquerque[0.044]; football[0.044]; football[0.044]; game[0.044]; October[0.042]; said[0.041]; Thursday[0.041]; competition[0.040]; Association[0.040]; bar[0.039]; Officials[0.039]; decided[0.039]; school[0.039]; school[0.039]; Mike[0.038]; high[0.038]; high[0.038]; High[0.038]; \t\n\r [============>.................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1349/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1206testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Weah\u001b[39m ==> ENTITY: \u001b[32mGeorge Weah\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWeah[0.047]; Weah[0.047]; soccer[0.044]; Soccer[0.044]; Fifa[0.043]; Fifa[0.043]; Fifa[0.043]; Milan[0.043]; Liberian[0.042]; award[0.042]; award[0.042]; striker[0.042]; Havelange[0.041]; Havelange[0.041]; Havelange[0.041]; named[0.040]; Sport[0.040]; Boss[0.040]; George[0.040]; Italian[0.039]; nose[0.039]; despite[0.039]; Porto[0.039]; captain[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Costa\u001b[39m ==> ENTITY: \u001b[32mJorge Costa\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.045]; Champions[0.043]; players[0.043]; Costa[0.043]; Costa[0.043]; Costa[0.043]; League[0.043]; Havelange[0.043]; Lisbon[0.042]; match[0.042]; match[0.042]; Cup[0.042]; Uefa[0.042]; investigation[0.041]; suing[0.041]; surgery[0.041]; January[0.040]; Weah[0.040]; Weah[0.040]; Weah[0.040]; Weah[0.040]; decision[0.040]; admitted[0.040]; Italians[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = FIFA\u001b[39m ==> ENTITY: \u001b[32mFIFA\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.045]; soccer[0.045]; Soccer[0.044]; Fifa[0.044]; Fifa[0.044]; players[0.043]; Sport[0.042]; Champions[0.042]; League[0.042]; play[0.042]; match[0.042]; match[0.042]; Uefa[0.041]; award[0.041]; award[0.041]; award[0.041]; Play[0.041]; Havelange[0.040]; Havelange[0.040]; Havelange[0.040]; European[0.040]; European[0.040]; competition[0.040]; place[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jorge Costa\u001b[39m ==> ENTITY: \u001b[32mJorge Costa\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.045]; soccer[0.045]; Porto[0.043]; Rome[0.043]; Costa[0.043]; Costa[0.043]; Havelange[0.043]; Havelange[0.043]; Havelange[0.043]; Lisbon[0.042]; Joao[0.042]; dello[0.041]; despite[0.041]; January[0.040]; Sport[0.040]; Weah[0.040]; Weah[0.040]; Weah[0.040]; Weah[0.040]; decision[0.040]; Milan[0.040]; Fifa[0.040]; Fifa[0.040]; Fifa[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNorway\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNorway national football team\u001b[39m <---> \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.239:0.235[\u001b[31m0.004\u001b[39m]; local()= 0.156:0.069[\u001b[31m0.087\u001b[39m]; log p(e|m)= -3.147:-0.227[\u001b[32m2.920\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.046]; match[0.044]; match[0.044]; players[0.044]; Uefa[0.043]; Cup[0.043]; European[0.043]; European[0.043]; Champions[0.043]; League[0.043]; admitted[0.042]; accepted[0.042]; head[0.041]; place[0.041]; November[0.040]; considering[0.040]; Rosenborg[0.040]; defeat[0.040]; defeat[0.040]; Wednesday[0.040]; served[0.040]; suspended[0.039]; governing[0.039]; needed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = AC Milan\u001b[39m ==> ENTITY: \u001b[32mA.C. Milan\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.047]; Soccer[0.047]; Italian[0.044]; Sport[0.043]; despite[0.043]; January[0.043]; Porto[0.042]; Havelange[0.042]; Havelange[0.042]; Havelange[0.042]; happy[0.042]; play[0.041]; striker[0.041]; Gazzetta[0.041]; dello[0.040]; Fifa[0.040]; Fifa[0.040]; Fifa[0.040]; competition[0.040]; Boss[0.040]; captain[0.040]; Play[0.040]; nose[0.040]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = UEFA\u001b[39m ==> ENTITY: \u001b[32mUEFA\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.044]; Sport[0.044]; match[0.044]; match[0.044]; League[0.044]; players[0.043]; Fifa[0.043]; Champions[0.043]; European[0.043]; European[0.043]; place[0.042]; competition[0.042]; Milan[0.042]; Norway[0.041]; Havelange[0.040]; admitted[0.040]; Lisbon[0.040]; suspended[0.040]; suspension[0.040]; January[0.040]; Rosenborg[0.040]; considering[0.039]; home[0.039]; November[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Weah\u001b[39m ==> ENTITY: \u001b[32mGeorge Weah\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWeah[0.047]; Weah[0.047]; Weah[0.047]; soccer[0.044]; League[0.044]; Milan[0.043]; players[0.042]; Cup[0.042]; admitted[0.042]; Champions[0.042]; Uefa[0.041]; European[0.041]; European[0.041]; surgery[0.041]; Italians[0.040]; match[0.040]; match[0.040]; Rosenborg[0.040]; acknowledges[0.040]; offered[0.039]; nose[0.039]; considering[0.039]; place[0.039]; November[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lisbon ==> ENTITY: \u001b[32mLisbon\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.124:0.124[0]; log p(e|m)= -0.098:-0.098[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngoverning[0.045]; soccer[0.045]; soccer[0.045]; match[0.044]; match[0.044]; European[0.043]; European[0.043]; George[0.043]; accepted[0.042]; Porto[0.042]; Uefa[0.041]; place[0.041]; Costa[0.041]; Costa[0.041]; Costa[0.041]; Costa[0.041]; Costa[0.041]; Costa[0.041]; Jorge[0.040]; League[0.040]; Champions[0.040]; decision[0.040]; January[0.040]; Sport[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gazzetta dello Sport\u001b[39m ==> ENTITY: \u001b[32mLa Gazzetta dello Sport\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItalian[0.046]; Milan[0.046]; newspaper[0.044]; saying[0.044]; January[0.043]; quoted[0.043]; soccer[0.043]; soccer[0.043]; Soccer[0.042]; named[0.041]; interview[0.041]; European[0.041]; Havelange[0.041]; Havelange[0.041]; Havelange[0.041]; personally[0.041]; personally[0.041]; Uefa[0.040]; said[0.040]; said[0.040]; Friday[0.040]; award[0.040]; award[0.040]; award[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = George Weah\u001b[39m ==> ENTITY: \u001b[32mGeorge Weah\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWeah[0.047]; Weah[0.047]; soccer[0.044]; Soccer[0.043]; Fifa[0.043]; Fifa[0.043]; Fifa[0.043]; Milan[0.043]; Liberian[0.042]; award[0.042]; award[0.042]; award[0.042]; striker[0.041]; Havelange[0.041]; Havelange[0.041]; Havelange[0.041]; named[0.040]; Sport[0.040]; Boss[0.040]; January[0.040]; competition[0.039]; Italian[0.039]; nose[0.039]; despite[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Costa\u001b[39m ==> ENTITY: \u001b[32mJorge Costa\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.045]; Champions[0.044]; players[0.043]; Costa[0.043]; Costa[0.043]; Costa[0.043]; League[0.043]; Lisbon[0.042]; match[0.042]; match[0.042]; Cup[0.042]; Uefa[0.042]; investigation[0.042]; suing[0.041]; surgery[0.041]; January[0.040]; Weah[0.040]; Weah[0.040]; Weah[0.040]; Weah[0.040]; admitted[0.040]; Italians[0.040]; Milan[0.040]; head[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rome ==> ENTITY: \u001b[32mRome\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItalian[0.047]; Jorge[0.047]; Milan[0.047]; despite[0.046]; play[0.045]; George[0.045]; soccer[0.044]; Friday[0.044]; Costa[0.043]; Costa[0.043]; world[0.043]; said[0.043]; said[0.043]; dello[0.043]; quoted[0.043]; saying[0.042]; decision[0.042]; Sport[0.042]; Gazzetta[0.042]; Soccer[0.042]; fair[0.041]; interview[0.041]; Joao[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mItalians\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mItaly\u001b[39m <---> \u001b[32mItaly national football team\u001b[39m\t\nSCORES: global= 0.250:0.247[\u001b[31m0.002\u001b[39m]; local()= 0.115:0.154[\u001b[32m0.039\u001b[39m]; log p(e|m)= -1.556:-3.244[\u001b[31m1.688\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.046]; match[0.045]; Milan[0.045]; League[0.044]; Champions[0.044]; European[0.044]; European[0.044]; Cup[0.043]; players[0.043]; place[0.042]; Norway[0.041]; admitted[0.041]; home[0.040]; Wednesday[0.040]; defeat[0.040]; defeat[0.040]; Costa[0.040]; Costa[0.040]; Costa[0.040]; November[0.040]; governing[0.039]; said[0.039]; served[0.039]; needed[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEuropean\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUEFA\u001b[39m <---> \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.246:0.238[\u001b[31m0.008\u001b[39m]; local()= 0.189:0.050[\u001b[31m0.138\u001b[39m]; log p(e|m)= -3.576:-0.256[\u001b[32m3.319\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUefa[0.048]; League[0.047]; European[0.047]; soccer[0.046]; Champions[0.046]; Sport[0.046]; Fifa[0.045]; match[0.044]; match[0.044]; competition[0.044]; Milan[0.044]; players[0.043]; place[0.042]; Havelange[0.042]; admitted[0.042]; suspended[0.042]; Norway[0.042]; suspension[0.042]; January[0.041]; Rosenborg[0.041]; considering[0.041]; award[0.040]; award[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Costa\u001b[39m ==> ENTITY: \u001b[32mJorge Costa\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.050]; soccer[0.050]; soccer[0.050]; Jorge[0.049]; Porto[0.048]; players[0.048]; Rome[0.048]; Costa[0.048]; Costa[0.048]; Havelange[0.048]; Havelange[0.048]; Havelange[0.048]; Lisbon[0.047]; Joao[0.047]; match[0.047]; Uefa[0.047]; investigation[0.046]; dello[0.046]; despite[0.045]; January[0.045]; Sport[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Porto ==> ENTITY: \u001b[32mF.C. Porto\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.161:0.161[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJoao[0.051]; Lisbon[0.047]; Soccer[0.046]; soccer[0.045]; Rome[0.045]; Milan[0.045]; Italian[0.044]; Costa[0.044]; Costa[0.044]; Costa[0.044]; Gazzetta[0.044]; Sport[0.043]; dello[0.043]; Havelange[0.042]; Havelange[0.042]; Havelange[0.042]; Jorge[0.042]; world[0.042]; despite[0.041]; competition[0.041]; January[0.040]; striker[0.040]; play[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Milan ==> ENTITY: \u001b[32mA.C. Milan\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.163:0.163[0]; log p(e|m)= -1.959:-1.959[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.047]; soccer[0.046]; Champions[0.044]; Italians[0.043]; match[0.043]; match[0.043]; Cup[0.043]; Uefa[0.043]; served[0.043]; players[0.043]; home[0.042]; European[0.042]; European[0.042]; Wednesday[0.041]; admitted[0.040]; November[0.040]; accepted[0.040]; governing[0.040]; defeat[0.040]; defeat[0.040]; offered[0.039]; took[0.039]; considering[0.039]; Rosenborg[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Costa\u001b[39m ==> ENTITY: \u001b[32mJorge Costa\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.044]; soccer[0.044]; Jorge[0.043]; Champions[0.043]; Porto[0.043]; players[0.043]; Costa[0.042]; Costa[0.042]; Costa[0.042]; Costa[0.042]; Costa[0.042]; League[0.042]; Havelange[0.042]; Lisbon[0.042]; match[0.041]; match[0.041]; Uefa[0.041]; investigation[0.041]; dello[0.041]; suing[0.040]; surgery[0.040]; despite[0.040]; January[0.040]; Sport[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Costa\u001b[39m ==> ENTITY: \u001b[32mJorge Costa\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.045]; Champions[0.044]; players[0.043]; Costa[0.043]; Costa[0.043]; Costa[0.043]; League[0.043]; Lisbon[0.042]; match[0.042]; match[0.042]; Cup[0.042]; Uefa[0.042]; investigation[0.042]; suing[0.041]; surgery[0.041]; January[0.040]; Weah[0.040]; Weah[0.040]; Weah[0.040]; Weah[0.040]; admitted[0.040]; Italians[0.040]; Milan[0.040]; head[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mLiberian\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mLiberia national football team\u001b[39m <---> \u001b[32mLiberia\u001b[39m\t\nSCORES: global= 0.271:0.259[\u001b[31m0.012\u001b[39m]; local()= 0.168:0.064[\u001b[31m0.104\u001b[39m]; log p(e|m)= -1.076:0.000[\u001b[32m1.076\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.045]; soccer[0.045]; governing[0.045]; players[0.044]; Weah[0.043]; Weah[0.043]; Weah[0.043]; Weah[0.043]; Weah[0.043]; Soccer[0.043]; captain[0.041]; Fifa[0.041]; Fifa[0.041]; Fifa[0.041]; match[0.040]; match[0.040]; League[0.040]; body[0.040]; striker[0.040]; November[0.040]; Champions[0.039]; Havelange[0.039]; Havelange[0.039]; Havelange[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italian ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.094:0.094[0]; log p(e|m)= -0.759:-0.759[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.047]; Soccer[0.046]; Rome[0.046]; Milan[0.046]; match[0.046]; Jorge[0.045]; Sport[0.045]; January[0.044]; Fifa[0.043]; Fifa[0.043]; Fifa[0.043]; dello[0.043]; despite[0.043]; Havelange[0.042]; Havelange[0.042]; Havelange[0.042]; play[0.042]; striker[0.042]; George[0.042]; world[0.042]; Gazzetta[0.042]; competition[0.042]; Play[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = FIFA\u001b[39m ==> ENTITY: \u001b[32mFIFA\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.047]; Soccer[0.047]; Fifa[0.046]; Fifa[0.046]; Sport[0.044]; award[0.044]; award[0.044]; Havelange[0.043]; Havelange[0.043]; play[0.041]; world[0.041]; Play[0.041]; named[0.041]; Milan[0.040]; captain[0.040]; Porto[0.040]; Weah[0.040]; Weah[0.040]; Weah[0.040]; despite[0.039]; Italian[0.039]; striker[0.039]; George[0.039]; Gazzetta[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = FIFA\u001b[39m ==> ENTITY: \u001b[32mFIFA\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.047]; Soccer[0.047]; Fifa[0.046]; Fifa[0.046]; Sport[0.044]; play[0.044]; award[0.043]; award[0.043]; Play[0.042]; Havelange[0.042]; Havelange[0.042]; Havelange[0.042]; world[0.041]; named[0.040]; Milan[0.040]; Porto[0.040]; Weah[0.039]; Weah[0.039]; Weah[0.039]; decision[0.039]; despite[0.039]; Italian[0.039]; George[0.039]; striker[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Weah\u001b[39m ==> ENTITY: \u001b[32mGeorge Weah\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWeah[0.047]; Weah[0.047]; Weah[0.047]; soccer[0.043]; League[0.043]; Milan[0.042]; players[0.042]; Cup[0.042]; Liberian[0.042]; admitted[0.042]; Champions[0.042]; award[0.041]; award[0.041]; Havelange[0.041]; Uefa[0.041]; European[0.041]; European[0.041]; surgery[0.041]; Italians[0.040]; match[0.040]; match[0.040]; Rosenborg[0.039]; January[0.039]; acknowledges[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rosenborg\u001b[39m ==> ENTITY: \u001b[32mRosenborg BK\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.050]; League[0.046]; soccer[0.044]; Uefa[0.044]; players[0.043]; November[0.043]; Champions[0.043]; match[0.043]; match[0.043]; considering[0.041]; Cup[0.041]; admitted[0.041]; European[0.041]; European[0.041]; offered[0.041]; place[0.040]; home[0.040]; acknowledges[0.040]; accepted[0.039]; reported[0.039]; took[0.039]; defeat[0.039]; defeat[0.039]; Milan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Weah\u001b[39m ==> ENTITY: \u001b[32mGeorge Weah\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWeah[0.046]; Weah[0.046]; Weah[0.046]; Weah[0.046]; soccer[0.043]; League[0.043]; Fifa[0.042]; Milan[0.042]; players[0.042]; Liberian[0.041]; admitted[0.041]; Champions[0.041]; award[0.041]; award[0.041]; Havelange[0.041]; Uefa[0.041]; European[0.040]; European[0.040]; surgery[0.040]; named[0.039]; Sport[0.039]; match[0.039]; match[0.039]; Rosenborg[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Weah\u001b[39m ==> ENTITY: \u001b[32mGeorge Weah\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWeah[0.046]; Weah[0.046]; Weah[0.046]; soccer[0.043]; soccer[0.043]; Soccer[0.043]; Fifa[0.042]; Fifa[0.042]; Fifa[0.042]; Milan[0.042]; Liberian[0.041]; award[0.041]; award[0.041]; award[0.041]; striker[0.041]; Havelange[0.041]; Havelange[0.041]; Havelange[0.041]; Uefa[0.041]; European[0.040]; named[0.039]; Sport[0.039]; match[0.039]; Boss[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Weah\u001b[39m ==> ENTITY: \u001b[32mGeorge Weah\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWeah[0.047]; Weah[0.047]; Weah[0.047]; soccer[0.044]; League[0.044]; Milan[0.043]; players[0.042]; Cup[0.042]; admitted[0.042]; Champions[0.042]; Uefa[0.041]; European[0.041]; European[0.041]; surgery[0.041]; Italians[0.040]; match[0.040]; match[0.040]; Rosenborg[0.040]; January[0.040]; acknowledges[0.040]; offered[0.039]; nose[0.039]; considering[0.039]; place[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Joao Havelange\u001b[39m ==> ENTITY: \u001b[32mJoão Havelange\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHavelange[0.049]; Havelange[0.049]; Fifa[0.046]; Fifa[0.046]; Fifa[0.046]; soccer[0.044]; Soccer[0.044]; Sport[0.043]; chairman[0.042]; said[0.040]; said[0.040]; named[0.040]; interview[0.040]; nose[0.040]; decision[0.040]; award[0.039]; award[0.039]; world[0.039]; saying[0.039]; Weah[0.039]; Weah[0.039]; Weah[0.039]; George[0.039]; quoted[0.039]; \t\n\r [=============>................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s802ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1379/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1378testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hobart ==> ENTITY: \u001b[32mHobart\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.171:0.171[0]; log p(e|m)= -0.390:-0.390[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOval[0.044]; Hobart[0.044]; match[0.043]; match[0.043]; match[0.043]; match[0.043]; Australia[0.043]; Australia[0.043]; Australia[0.043]; Cricket[0.042]; play[0.041]; Victoria[0.041]; Victoria[0.041]; Victoria[0.041]; Bellerive[0.041]; Tasmania[0.041]; second[0.041]; Tasmanian[0.040]; fourth[0.040]; batsman[0.040]; Shield[0.040]; Saturday[0.040]; Dean[0.040]; David[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tasmania ==> ENTITY: \u001b[32mTasmania cricket team\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.235:0.235[0]; log p(e|m)= -2.430:-2.430[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.044]; Australia[0.044]; Australia[0.044]; Australia[0.044]; Oval[0.043]; match[0.042]; match[0.042]; match[0.042]; match[0.042]; Victoria[0.042]; Victoria[0.042]; Victoria[0.042]; Cricket[0.041]; Tasmanian[0.041]; wicket[0.041]; spinner[0.041]; fightback[0.040]; Hobart[0.040]; Hobart[0.040]; century[0.040]; day[0.040]; day[0.040]; Shield[0.040]; Warne[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shaun Young\u001b[39m ==> ENTITY: \u001b[32mShaun Young\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOval[0.045]; Young[0.044]; Australia[0.044]; Australia[0.044]; Shaun[0.044]; batsman[0.043]; bowler[0.042]; match[0.042]; match[0.042]; match[0.042]; Tasmanian[0.042]; leg[0.041]; Ian[0.041]; Michael[0.041]; Victoria[0.041]; Hobart[0.041]; David[0.041]; David[0.041]; rounder[0.041]; wicket[0.040]; Harvey[0.040]; took[0.040]; Boon[0.040]; Boon[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shane Warne\u001b[39m ==> ENTITY: \u001b[32mShane Warne\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.047]; batsman[0.047]; bowler[0.046]; Australia[0.045]; Australia[0.045]; match[0.044]; match[0.044]; match[0.044]; match[0.044]; wicket[0.044]; spinner[0.044]; Oval[0.044]; rounder[0.043]; Ian[0.042]; Michael[0.042]; David[0.042]; David[0.042]; Harvey[0.042]; innings[0.042]; captain[0.042]; Shaun[0.041]; Shaun[0.041]; dismissed[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shaun Young\u001b[39m ==> ENTITY: \u001b[32mShaun Young\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOval[0.044]; Young[0.043]; Cricket[0.043]; Australia[0.043]; Australia[0.043]; Australia[0.043]; Australia[0.043]; Shaun[0.043]; batsman[0.042]; batsman[0.042]; Shield[0.042]; bowler[0.041]; Tasmania[0.041]; match[0.041]; match[0.041]; match[0.041]; match[0.041]; Dean[0.041]; Tasmanian[0.041]; day[0.040]; day[0.040]; leg[0.040]; play[0.040]; Ian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Victoria ==> ENTITY: \u001b[32mVictoria cricket team\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.229:0.229[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.044]; Victoria[0.044]; Victoria[0.044]; Oval[0.044]; Australia[0.043]; Australia[0.043]; Australia[0.043]; Australia[0.043]; Cricket[0.042]; match[0.042]; match[0.042]; match[0.042]; match[0.042]; Tasmania[0.042]; play[0.041]; innings[0.040]; fourth[0.040]; second[0.040]; hander[0.040]; wicket[0.040]; Shield[0.040]; Hobart[0.040]; Hobart[0.040]; Tasmanian[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Victoria ==> ENTITY: \u001b[32mVictoria cricket team\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.225:0.225[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.044]; Victoria[0.044]; Victoria[0.044]; Oval[0.044]; Australia[0.044]; Australia[0.044]; Australia[0.044]; Cricket[0.042]; match[0.042]; match[0.042]; match[0.042]; match[0.042]; Tasmania[0.042]; play[0.041]; innings[0.041]; fourth[0.040]; second[0.040]; hander[0.040]; wicket[0.040]; Shield[0.040]; Hobart[0.040]; Hobart[0.040]; Tasmanian[0.039]; spinner[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Harper\u001b[39m ==> ENTITY: \u001b[32mLaurie Harper\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDavid[0.044]; David[0.044]; Australia[0.044]; Australia[0.044]; Victoria[0.043]; Victoria[0.043]; Tasmanian[0.043]; Harper[0.043]; batsman[0.042]; Michael[0.042]; Laurie[0.042]; rounder[0.042]; Tasmania[0.042]; Ian[0.042]; Bellerive[0.041]; Harvey[0.041]; bowler[0.041]; leg[0.041]; Hobart[0.040]; hander[0.040]; play[0.040]; innings[0.039]; Shane[0.039]; left[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Michael DiVenuto\u001b[39m ==> ENTITY: \u001b[32mMichael Di Venuto\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.206:0.206[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.046]; batsman[0.046]; bowler[0.044]; Oval[0.042]; Australia[0.042]; Australia[0.042]; Australia[0.042]; Australia[0.042]; Cricket[0.042]; wicket[0.042]; day[0.042]; day[0.042]; match[0.041]; match[0.041]; match[0.041]; match[0.041]; Warne[0.041]; rounder[0.041]; innings[0.041]; dismissed[0.040]; spinner[0.040]; fourth[0.040]; Shaun[0.040]; Shaun[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Victoria ==> ENTITY: \u001b[32mVictoria cricket team\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.175:0.175[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.047]; Australia[0.046]; Australia[0.046]; bowler[0.045]; match[0.044]; Harvey[0.042]; hander[0.042]; wicket[0.042]; rounder[0.042]; Michael[0.042]; Ian[0.041]; spinner[0.041]; captain[0.041]; runs[0.041]; hit[0.040]; pair[0.040]; pair[0.040]; Warne[0.040]; David[0.040]; David[0.040]; Shane[0.040]; Young[0.040]; Young[0.040]; joined[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hobart ==> ENTITY: \u001b[32mHobart\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.175:0.175[0]; log p(e|m)= -0.390:-0.390[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOval[0.044]; Hobart[0.043]; match[0.043]; match[0.043]; match[0.043]; match[0.043]; Australia[0.043]; Australia[0.043]; Australia[0.043]; Australia[0.043]; Cricket[0.042]; play[0.041]; Victoria[0.041]; Victoria[0.041]; Victoria[0.041]; Bellerive[0.041]; Tasmania[0.041]; second[0.041]; Tasmanian[0.040]; fourth[0.040]; batsman[0.040]; batsman[0.040]; Shield[0.040]; Saturday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Victoria ==> ENTITY: \u001b[32mVictoria cricket team\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.214:0.214[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.045]; Victoria[0.044]; Victoria[0.044]; Oval[0.044]; Australia[0.044]; Australia[0.044]; Cricket[0.043]; match[0.042]; match[0.042]; match[0.042]; match[0.042]; Tasmania[0.042]; play[0.041]; innings[0.041]; fourth[0.041]; second[0.041]; Shield[0.040]; Hobart[0.040]; Hobart[0.040]; Tasmanian[0.040]; captain[0.039]; unbeaten[0.039]; Dean[0.039]; hit[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAustralia\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustralia national cricket team\u001b[39m <---> \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.245:0.243[\u001b[31m0.003\u001b[39m]; local()= 0.193:0.159[\u001b[31m0.034\u001b[39m]; log p(e|m)= -3.507:-0.172[\u001b[32m3.335\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; Australia[0.048]; match[0.043]; match[0.043]; match[0.043]; match[0.043]; batsman[0.043]; Cricket[0.043]; Oval[0.043]; Victoria[0.042]; Victoria[0.042]; Victoria[0.042]; Tasmania[0.041]; captain[0.040]; second[0.040]; day[0.040]; day[0.040]; Shield[0.039]; fourth[0.039]; play[0.039]; Shaun[0.039]; David[0.039]; Hobart[0.039]; Hobart[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.210:0.210[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.047]; Australia[0.047]; Australia[0.047]; match[0.043]; match[0.043]; match[0.043]; match[0.043]; batsman[0.043]; batsman[0.043]; Oval[0.042]; Victoria[0.041]; Victoria[0.041]; Victoria[0.041]; bowler[0.041]; Tasmania[0.041]; captain[0.040]; rounder[0.040]; second[0.040]; day[0.040]; day[0.040]; Warne[0.039]; spinner[0.039]; Shield[0.039]; fourth[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bellerive Oval\u001b[39m ==> ENTITY: \u001b[32mBellerive Oval\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.209:0.209[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.044]; match[0.044]; match[0.044]; match[0.044]; Australia[0.043]; Australia[0.043]; Australia[0.043]; Australia[0.043]; Cricket[0.043]; second[0.041]; Hobart[0.041]; Hobart[0.041]; Victoria[0.041]; Victoria[0.041]; Victoria[0.041]; fourth[0.041]; batsman[0.041]; batsman[0.041]; Tasmania[0.041]; Saturday[0.040]; play[0.040]; Tasmanian[0.040]; leg[0.040]; captain[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sheffield Shield\u001b[39m ==> ENTITY: \u001b[32mSheffield Shield\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.230:0.230[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.046]; Australia[0.046]; Australia[0.046]; batsman[0.044]; Oval[0.043]; Tasmania[0.042]; match[0.042]; match[0.042]; match[0.042]; match[0.042]; Tasmanian[0.042]; Cricket[0.042]; Victoria[0.041]; Victoria[0.041]; Victoria[0.041]; innings[0.041]; Hobart[0.040]; Hobart[0.040]; play[0.040]; wicket[0.040]; day[0.039]; day[0.039]; Warne[0.039]; captain[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = David Boon\u001b[39m ==> ENTITY: \u001b[32mDavid Boon\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.227:0.227[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.046]; Australia[0.046]; batsman[0.044]; Oval[0.043]; match[0.042]; match[0.042]; match[0.042]; Boon[0.042]; Ian[0.042]; Hobart[0.042]; captain[0.041]; bowler[0.041]; Harvey[0.041]; Warne[0.041]; rounder[0.041]; fourth[0.041]; spinner[0.041]; David[0.041]; second[0.040]; Michael[0.040]; Tasmanian[0.040]; play[0.040]; Shaun[0.040]; Shaun[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Laurie Harper\u001b[39m ==> ENTITY: \u001b[32mLaurie Harper\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDavid[0.045]; David[0.045]; Australia[0.045]; Australia[0.045]; Victoria[0.045]; Victoria[0.045]; Tasmanian[0.045]; Harper[0.044]; batsman[0.044]; Michael[0.044]; rounder[0.044]; Tasmania[0.044]; Ian[0.043]; Bellerive[0.043]; Harvey[0.043]; bowler[0.043]; leg[0.042]; Hobart[0.042]; hander[0.042]; play[0.041]; innings[0.041]; Shane[0.041]; left[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAustralia\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustralia national cricket team\u001b[39m <---> \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.245:0.243[\u001b[31m0.003\u001b[39m]; local()= 0.193:0.159[\u001b[31m0.034\u001b[39m]; log p(e|m)= -3.507:-0.172[\u001b[32m3.335\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; Australia[0.048]; match[0.043]; match[0.043]; match[0.043]; match[0.043]; batsman[0.043]; Cricket[0.043]; Oval[0.043]; Victoria[0.042]; Victoria[0.042]; Victoria[0.042]; Tasmania[0.041]; captain[0.040]; second[0.040]; day[0.040]; day[0.040]; Shield[0.039]; fourth[0.039]; play[0.039]; Shaun[0.039]; David[0.039]; Hobart[0.039]; Hobart[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tasmanian ==> ENTITY: \u001b[32mTasmania cricket team\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.239:0.239[0]; log p(e|m)= -1.715:-1.715[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.044]; batsman[0.044]; Australia[0.044]; Australia[0.044]; Australia[0.044]; Australia[0.044]; Oval[0.043]; bowler[0.042]; Tasmania[0.042]; Victoria[0.042]; Victoria[0.042]; Victoria[0.042]; match[0.042]; match[0.042]; match[0.042]; match[0.042]; Cricket[0.041]; rounder[0.041]; wicket[0.041]; spinner[0.040]; fightback[0.040]; Hobart[0.040]; Hobart[0.040]; play[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ian Harvey\u001b[39m ==> ENTITY: \u001b[32mIan Harvey\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.045]; wicket[0.044]; bowler[0.043]; Australia[0.043]; Australia[0.043]; Shaun[0.043]; Shaun[0.043]; David[0.043]; David[0.043]; Michael[0.042]; match[0.042]; match[0.042]; spinner[0.042]; rounder[0.042]; Warne[0.042]; runs[0.041]; leg[0.041]; hit[0.040]; captain[0.040]; boundaries[0.039]; Young[0.039]; Young[0.039]; scored[0.039]; Shane[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dean Jones ==> ENTITY: \u001b[32mDean Jones (cricketer)\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.044]; Cricket[0.043]; Australia[0.043]; Australia[0.043]; Australia[0.043]; David[0.042]; match[0.042]; match[0.042]; match[0.042]; match[0.042]; Oval[0.042]; day[0.042]; day[0.042]; Michael[0.041]; captain[0.041]; Shane[0.041]; Warne[0.041]; century[0.041]; second[0.040]; leg[0.040]; Jones[0.040]; Jones[0.040]; Jones[0.040]; Shaun[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.186:0.186[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.049]; match[0.044]; match[0.044]; match[0.044]; batsman[0.044]; Oval[0.044]; Victoria[0.043]; Victoria[0.043]; bowler[0.042]; captain[0.041]; rounder[0.041]; second[0.041]; day[0.041]; day[0.041]; Warne[0.040]; spinner[0.040]; fourth[0.040]; play[0.040]; Shaun[0.040]; Shaun[0.040]; wicket[0.040]; David[0.039]; David[0.039]; Hobart[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = David Boon\u001b[39m ==> ENTITY: \u001b[32mDavid Boon\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.249:0.249[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.045]; Australia[0.045]; Australia[0.045]; Australia[0.045]; batsman[0.043]; batsman[0.043]; Oval[0.042]; Cricket[0.042]; match[0.041]; match[0.041]; match[0.041]; match[0.041]; Boon[0.041]; Ian[0.041]; Hobart[0.041]; Hobart[0.041]; captain[0.041]; bowler[0.040]; Harvey[0.040]; Warne[0.040]; rounder[0.040]; fourth[0.040]; Tasmania[0.040]; spinner[0.040]; \t\n\r [=============>................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s681ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1403/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1340testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Honda ==> ENTITY: \u001b[32mHonda\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.199:0.199[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHonda[0.077]; Honda[0.077]; vehicle[0.046]; minivan[0.046]; Motor[0.043]; styling[0.043]; rear[0.042]; version[0.042]; launched[0.040]; launch[0.040]; featuring[0.039]; Mx[0.039]; Mx[0.039]; Tokyo[0.039]; domestic[0.039]; received[0.039]; weeks[0.039]; lower[0.038]; sales[0.038]; sales[0.038]; sales[0.038]; November[0.038]; road[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tokyo ==> ENTITY: \u001b[32mTokyo\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.050:0.050[0]; log p(e|m)= -0.093:-0.093[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFriday[0.051]; featuring[0.049]; launched[0.049]; November[0.049]; launch[0.047]; Honda[0.046]; Honda[0.046]; Honda[0.046]; monthly[0.045]; weeks[0.045]; said[0.045]; received[0.045]; domestic[0.044]; exceeds[0.044]; Motor[0.044]; Rv[0.044]; lower[0.044]; accounted[0.044]; light[0.044]; recreational[0.044]; version[0.044]; minivan[0.044]; \t\n\r [=============>................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s775ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1405/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1233testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= -0.010:-0.010[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYork[0.076]; York[0.076]; Maritime[0.076]; Queen[0.075]; Desk[0.071]; China[0.066]; Dec[0.066]; Commodities[0.064]; scrap[0.063]; Steel[0.063]; Kaohsiung[0.062]; Ore[0.062]; tonnes[0.060]; Dampier[0.060]; fixtures[0.059]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = China Steel\u001b[39m ==> ENTITY: \u001b[32mChina Steel\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.041:0.041[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKaohsiung[0.069]; Dec[0.069]; Maritime[0.068]; tonnes[0.065]; Commodities[0.064]; Desk[0.064]; fixtures[0.063]; ore[0.062]; Ore[0.061]; coal[0.061]; Dampier[0.061]; scrap[0.061]; Queen[0.059]; York[0.058]; York[0.058]; York[0.058]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= -0.010:-0.010[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYork[0.076]; York[0.076]; Maritime[0.076]; Queen[0.075]; Desk[0.071]; China[0.066]; Dec[0.066]; Commodities[0.064]; scrap[0.063]; Steel[0.063]; Kaohsiung[0.062]; Ore[0.062]; tonnes[0.060]; Dampier[0.060]; fixtures[0.059]; \t\n\r [=============>................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s805ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1408/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1180testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.184:0.184[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.082]; match[0.074]; Cricket[0.074]; Melbourne[0.069]; overs[0.067]; overs[0.067]; World[0.065]; West[0.065]; West[0.065]; Indies[0.064]; Indies[0.064]; Friday[0.063]; Overs[0.062]; limited[0.060]; Series[0.058]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Indies ==> ENTITY: \u001b[32mWest Indies cricket team\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.191:0.191[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.078]; Australia[0.078]; match[0.075]; Cricket[0.075]; overs[0.074]; overs[0.074]; Indies[0.073]; World[0.070]; West[0.069]; Melbourne[0.068]; Friday[0.068]; Overs[0.066]; Series[0.065]; limited[0.065]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Melbourne ==> ENTITY: \u001b[32mMelbourne\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.168:-0.168[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.079]; Australia[0.079]; match[0.078]; Cricket[0.073]; Friday[0.066]; overs[0.064]; overs[0.064]; West[0.064]; West[0.064]; limited[0.063]; World[0.063]; Indies[0.062]; Indies[0.062]; Overs[0.061]; Series[0.060]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.184:0.184[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.082]; match[0.074]; Cricket[0.074]; Melbourne[0.069]; overs[0.067]; overs[0.067]; World[0.065]; West[0.065]; West[0.065]; Indies[0.064]; Indies[0.064]; Friday[0.063]; Overs[0.062]; limited[0.060]; Series[0.058]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWorld Series\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustralian Tri-Series\u001b[39m <---> \u001b[32mWorld Series Cricket\u001b[39m\t\nSCORES: global= 0.256:0.251[\u001b[31m0.005\u001b[39m]; local()= 0.182:0.155[\u001b[31m0.027\u001b[39m]; log p(e|m)= -3.324:-3.324[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.083]; Australia[0.083]; Cricket[0.078]; match[0.076]; Melbourne[0.072]; Friday[0.071]; overs[0.068]; overs[0.068]; Overs[0.068]; West[0.067]; West[0.067]; Indies[0.067]; Indies[0.067]; limited[0.066]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Indies ==> ENTITY: \u001b[32mWest Indies cricket team\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.191:0.191[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.078]; Australia[0.078]; match[0.075]; Cricket[0.075]; overs[0.074]; overs[0.074]; Indies[0.073]; World[0.070]; West[0.069]; Melbourne[0.068]; Friday[0.068]; Overs[0.066]; Series[0.065]; limited[0.065]; \t\n\r [=============>................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s865ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1414/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1391testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Volendam ==> ENTITY: \u001b[32mFC Volendam\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.169:0.169[0]; log p(e|m)= -0.896:-0.896[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; Volendam[0.045]; Volendam[0.045]; Dutch[0.045]; games[0.044]; matches[0.043]; played[0.043]; Soccer[0.043]; Feyenoord[0.043]; season[0.043]; game[0.042]; Belgian[0.041]; champions[0.040]; Amsterdam[0.040]; Amsterdam[0.040]; Ajax[0.040]; division[0.040]; department[0.040]; Eindhoven[0.040]; twice[0.039]; Zeljko[0.039]; striker[0.039]; striker[0.039]; midfielder[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Feyenoord\u001b[39m ==> ENTITY: \u001b[32mFeyenoord\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.218:0.218[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; Feyenoord[0.045]; Doetinchem[0.044]; Doetinchem[0.044]; matches[0.043]; Ajax[0.043]; Ajax[0.043]; games[0.043]; Eindhoven[0.043]; season[0.042]; season[0.042]; twice[0.041]; Cup[0.041]; Uefa[0.041]; played[0.041]; Alkmaar[0.041]; game[0.041]; striker[0.041]; win[0.040]; champions[0.040]; midfielder[0.040]; defender[0.040]; Dutch[0.039]; Graafschap[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Luc Nilis\u001b[39m ==> ENTITY: \u001b[32mLuc Nilis\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; twice[0.045]; striker[0.043]; striker[0.043]; season[0.043]; Alkmaar[0.042]; Feyenoord[0.042]; Feyenoord[0.042]; matches[0.042]; midfielder[0.041]; defender[0.041]; Ajax[0.041]; Ajax[0.041]; Doetinchem[0.041]; Doetinchem[0.041]; scored[0.041]; Dutch[0.041]; games[0.041]; Graafschap[0.041]; Eindhoven[0.041]; department[0.041]; game[0.040]; played[0.040]; Zeljko[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = PSV Eindhoven\u001b[39m ==> ENTITY: \u001b[32mPSV Eindhoven\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFeyenoord[0.047]; Feyenoord[0.047]; matches[0.046]; league[0.046]; Ajax[0.045]; Ajax[0.045]; games[0.044]; season[0.044]; Alkmaar[0.044]; twice[0.044]; Soccer[0.043]; champions[0.043]; Dutch[0.043]; Cup[0.042]; Uefa[0.042]; win[0.042]; striker[0.042]; striker[0.042]; points[0.042]; played[0.042]; game[0.042]; defender[0.041]; midfielder[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mYugoslav\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mYugoslavia national football team\u001b[39m <---> \u001b[32mYugoslavia\u001b[39m\t\nSCORES: global= 0.244:0.235[\u001b[31m0.008\u001b[39m]; local()= 0.139:0.073[\u001b[31m0.066\u001b[39m]; log p(e|m)= -2.283:-0.759[\u001b[32m1.523\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.048]; matches[0.047]; season[0.045]; played[0.045]; games[0.044]; Zeljko[0.044]; Soccer[0.043]; champions[0.042]; twice[0.042]; win[0.042]; game[0.042]; Petrovic[0.041]; division[0.041]; title[0.041]; leaders[0.040]; Eindhoven[0.040]; points[0.040]; scored[0.040]; boast[0.040]; Feyenoord[0.039]; Feyenoord[0.039]; Saturday[0.039]; defender[0.039]; midfielder[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = PSV\u001b[39m ==> ENTITY: \u001b[32mPSV Eindhoven\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFeyenoord[0.045]; Feyenoord[0.045]; Doetinchem[0.045]; Doetinchem[0.045]; matches[0.044]; league[0.043]; Ajax[0.043]; Ajax[0.043]; Eindhoven[0.043]; games[0.042]; season[0.042]; season[0.042]; Alkmaar[0.041]; twice[0.041]; champions[0.041]; Dutch[0.041]; Uefa[0.040]; win[0.040]; striker[0.040]; striker[0.040]; Cup[0.040]; played[0.040]; game[0.040]; defender[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Volendam ==> ENTITY: \u001b[32mFC Volendam\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.191:0.191[0]; log p(e|m)= -0.896:-0.896[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; Volendam[0.044]; Volendam[0.044]; Dutch[0.044]; games[0.043]; Doetinchem[0.043]; Doetinchem[0.043]; matches[0.042]; played[0.042]; Soccer[0.042]; Feyenoord[0.042]; Feyenoord[0.042]; season[0.042]; season[0.042]; game[0.041]; Cup[0.041]; Alkmaar[0.041]; packages[0.040]; Belgian[0.040]; champions[0.040]; Amsterdam[0.040]; Amsterdam[0.040]; Ajax[0.039]; Ajax[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazilian ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.273:-0.273[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.048]; Brazilian[0.048]; Soccer[0.047]; matches[0.047]; Marcelo[0.046]; games[0.045]; season[0.045]; Cup[0.045]; game[0.044]; division[0.043]; played[0.043]; twice[0.043]; win[0.042]; Uefa[0.042]; Doetinchem[0.042]; Doetinchem[0.042]; title[0.042]; Super[0.041]; champions[0.041]; Ajax[0.041]; Ajax[0.041]; Dutch[0.041]; scored[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zeljko Petrovic\u001b[39m ==> ENTITY: \u001b[32mŽeljko Petrović\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.046]; Alkmaar[0.046]; Feyenoord[0.045]; Feyenoord[0.045]; division[0.045]; matches[0.044]; Brazilian[0.044]; Brazilian[0.044]; department[0.044]; league[0.044]; Eindhoven[0.043]; Dutch[0.043]; played[0.043]; midfielder[0.043]; striker[0.043]; striker[0.043]; defender[0.043]; champions[0.043]; win[0.042]; Volendam[0.042]; Volendam[0.042]; Volendam[0.042]; scored[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Amsterdam ==> ENTITY: \u001b[32mAmsterdam\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.094:0.094[0]; log p(e|m)= -0.064:-0.064[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmsterdam[0.049]; Dutch[0.045]; season[0.044]; twice[0.044]; Eindhoven[0.043]; league[0.043]; Soccer[0.043]; matches[0.043]; Saturday[0.043]; games[0.042]; played[0.042]; boast[0.042]; way[0.041]; win[0.041]; game[0.041]; Belgian[0.040]; title[0.040]; Yugoslav[0.040]; champions[0.040]; Brazilian[0.039]; Brazilian[0.039]; Petrovic[0.039]; clear[0.039]; defender[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ajax ==> ENTITY: \u001b[32mAFC Ajax\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.195:0.195[0]; log p(e|m)= -1.435:-1.435[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; Feyenoord[0.044]; Feyenoord[0.044]; matches[0.044]; Doetinchem[0.043]; Doetinchem[0.043]; Ajax[0.043]; games[0.042]; game[0.042]; Cup[0.042]; season[0.042]; season[0.042]; played[0.042]; Eindhoven[0.042]; twice[0.041]; Uefa[0.041]; win[0.040]; title[0.040]; Alkmaar[0.040]; striker[0.040]; striker[0.040]; champions[0.040]; midfielder[0.040]; Dutch[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgian ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.101:0.101[0]; log p(e|m)= -0.124:-0.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; matches[0.045]; Doetinchem[0.043]; Doetinchem[0.043]; Soccer[0.043]; played[0.043]; games[0.042]; Uefa[0.042]; season[0.042]; Feyenoord[0.042]; Feyenoord[0.042]; Cup[0.041]; Dutch[0.041]; game[0.041]; midfielder[0.041]; twice[0.041]; champions[0.040]; win[0.040]; defender[0.040]; Eindhoven[0.040]; striker[0.040]; striker[0.040]; Ajax[0.040]; Ajax[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dutch ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.107:0.107[0]; log p(e|m)= -0.794:-0.794[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; matches[0.045]; season[0.043]; games[0.043]; twice[0.043]; Eindhoven[0.043]; played[0.042]; Belgian[0.042]; Feyenoord[0.042]; Feyenoord[0.042]; game[0.042]; Soccer[0.042]; win[0.041]; Uefa[0.041]; Cup[0.041]; Ajax[0.041]; Ajax[0.041]; Amsterdam[0.041]; Amsterdam[0.041]; striker[0.040]; striker[0.040]; midfielder[0.040]; Alkmaar[0.039]; champions[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Feyenoord\u001b[39m ==> ENTITY: \u001b[32mFeyenoord\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.217:0.217[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; Feyenoord[0.045]; Doetinchem[0.043]; Doetinchem[0.043]; matches[0.043]; Soccer[0.043]; Ajax[0.043]; Ajax[0.043]; games[0.043]; Eindhoven[0.042]; season[0.042]; season[0.042]; twice[0.041]; Cup[0.041]; Uefa[0.041]; played[0.041]; Alkmaar[0.041]; game[0.040]; striker[0.040]; striker[0.040]; win[0.040]; champions[0.040]; midfielder[0.040]; defender[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tenerife ==> ENTITY: \u001b[32mCD Tenerife\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.179:0.179[0]; log p(e|m)= -1.565:-1.565[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.049]; season[0.047]; season[0.047]; matches[0.045]; games[0.044]; department[0.044]; champions[0.044]; played[0.043]; Cup[0.043]; Feyenoord[0.043]; Feyenoord[0.043]; fourth[0.043]; game[0.043]; eliminated[0.042]; title[0.042]; travel[0.042]; win[0.042]; Alkmaar[0.042]; Super[0.042]; Doetinchem[0.042]; Doetinchem[0.042]; Brazilian[0.042]; Saturday[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = PSV\u001b[39m ==> ENTITY: \u001b[32mPSV Eindhoven\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.218:0.218[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFeyenoord[0.045]; Feyenoord[0.045]; Doetinchem[0.045]; Doetinchem[0.045]; matches[0.043]; league[0.043]; Ajax[0.043]; Ajax[0.043]; Eindhoven[0.042]; games[0.042]; season[0.041]; season[0.041]; Alkmaar[0.041]; twice[0.041]; Soccer[0.041]; champions[0.041]; Dutch[0.040]; Uefa[0.040]; win[0.040]; striker[0.040]; striker[0.040]; Cup[0.040]; played[0.040]; game[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Volendam ==> ENTITY: \u001b[32mFC Volendam\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.187:0.187[0]; log p(e|m)= -0.896:-0.896[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; Volendam[0.044]; Volendam[0.044]; Dutch[0.044]; games[0.044]; Doetinchem[0.043]; matches[0.043]; played[0.042]; Soccer[0.042]; Feyenoord[0.042]; Feyenoord[0.042]; season[0.042]; game[0.042]; Cup[0.042]; Alkmaar[0.041]; Belgian[0.040]; champions[0.040]; Amsterdam[0.040]; Amsterdam[0.040]; Ajax[0.040]; Ajax[0.040]; Uefa[0.039]; Graafschap[0.039]; division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = PSV\u001b[39m ==> ENTITY: \u001b[32mPSV Eindhoven\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFeyenoord[0.046]; matches[0.045]; league[0.045]; Ajax[0.044]; Eindhoven[0.044]; games[0.043]; season[0.043]; twice[0.043]; Soccer[0.042]; champions[0.042]; Dutch[0.042]; win[0.041]; striker[0.041]; striker[0.041]; played[0.041]; game[0.041]; defender[0.040]; midfielder[0.040]; Amsterdam[0.040]; Amsterdam[0.040]; department[0.039]; Volendam[0.039]; Volendam[0.039]; Volendam[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazilian ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.273:-0.273[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; Brazilian[0.046]; Soccer[0.046]; matches[0.046]; Marcelo[0.044]; games[0.044]; season[0.043]; game[0.043]; division[0.042]; played[0.042]; twice[0.042]; win[0.041]; title[0.041]; champions[0.040]; Ajax[0.040]; Ajax[0.040]; Dutch[0.040]; scored[0.039]; Feyenoord[0.039]; Belgian[0.039]; defender[0.039]; striker[0.039]; striker[0.039]; points[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = AZ Alkmaar\u001b[39m ==> ENTITY: \u001b[32mAZ Alkmaar\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; Doetinchem[0.044]; Doetinchem[0.044]; league[0.044]; Feyenoord[0.043]; Feyenoord[0.043]; season[0.043]; season[0.043]; games[0.042]; twice[0.042]; Eindhoven[0.042]; Ajax[0.042]; Ajax[0.042]; champions[0.041]; Cup[0.041]; win[0.041]; played[0.041]; striker[0.040]; striker[0.040]; leaders[0.040]; game[0.040]; midfielder[0.040]; division[0.040]; defender[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ajax Amsterdam\u001b[39m ==> ENTITY: \u001b[32mAFC Ajax\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; Doetinchem[0.044]; Doetinchem[0.044]; Feyenoord[0.044]; Feyenoord[0.044]; Ajax[0.043]; matches[0.043]; Soccer[0.043]; Amsterdam[0.042]; season[0.042]; season[0.042]; games[0.042]; Eindhoven[0.041]; twice[0.041]; Cup[0.041]; Uefa[0.041]; Alkmaar[0.040]; win[0.040]; played[0.040]; striker[0.040]; striker[0.040]; champions[0.040]; midfielder[0.040]; Dutch[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Vampeta\u001b[39m ==> ENTITY: \u001b[32mVampeta\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazilian[0.045]; Brazilian[0.045]; league[0.044]; Marcelo[0.043]; Alkmaar[0.043]; Doetinchem[0.043]; Doetinchem[0.043]; season[0.042]; Feyenoord[0.042]; Feyenoord[0.042]; games[0.042]; Graafschap[0.041]; Soccer[0.041]; department[0.041]; Zeljko[0.041]; game[0.040]; striker[0.040]; striker[0.040]; Ajax[0.040]; Ajax[0.040]; Volendam[0.040]; Volendam[0.040]; Volendam[0.040]; Cup[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = De Graafschap Doetinchem\u001b[39m ==> ENTITY: \u001b[32mDe Graafschap\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.213:0.213[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDoetinchem[0.047]; league[0.045]; Feyenoord[0.045]; Feyenoord[0.045]; matches[0.045]; season[0.044]; season[0.044]; Ajax[0.043]; Ajax[0.043]; games[0.042]; striker[0.042]; Alkmaar[0.042]; game[0.041]; defender[0.041]; played[0.040]; win[0.040]; Uefa[0.040]; Cup[0.040]; champions[0.039]; Volendam[0.039]; Volendam[0.039]; eliminated[0.039]; department[0.039]; Saturday[0.039]; \t\n\r [=============>................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s873ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1439/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1301testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Major\u001b[39m ==> ENTITY: \u001b[32mJohn Major\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.142:0.142[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChancellor[0.044]; Major[0.043]; Major[0.043]; Major[0.043]; Major[0.043]; John[0.043]; Clarke[0.043]; Clarke[0.043]; Clarke[0.043]; Exchequer[0.042]; London[0.042]; European[0.042]; Prime[0.042]; office[0.042]; despite[0.041]; policy[0.041]; job[0.041]; political[0.040]; Minister[0.040]; aware[0.040]; meetings[0.040]; good[0.040]; good[0.040]; good[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Major\u001b[39m ==> ENTITY: \u001b[32mJohn Major\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nparliamentary[0.044]; Chancellor[0.043]; Major[0.042]; Major[0.042]; Major[0.042]; Major[0.042]; Major[0.042]; John[0.042]; Clarke[0.042]; Clarke[0.042]; Clarke[0.042]; Clarke[0.042]; Exchequer[0.042]; London[0.042]; prime[0.041]; Prime[0.041]; office[0.041]; despite[0.041]; policy[0.041]; minister[0.041]; speculation[0.040]; political[0.040]; job[0.040]; Minister[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Clarke\u001b[39m ==> ENTITY: \u001b[32mKenneth Clarke\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClarke[0.052]; Clarke[0.052]; Clarke[0.052]; John[0.049]; Chancellor[0.048]; parliamentary[0.046]; London[0.046]; policy[0.046]; Exchequer[0.045]; Kenneth[0.045]; minister[0.044]; office[0.044]; job[0.043]; spokesman[0.043]; spokesman[0.043]; spokesman[0.043]; spokesman[0.043]; spokesman[0.043]; Minister[0.043]; Prime[0.043]; despite[0.043]; constituencies[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = European ==> ENTITY: \u001b[32mEuropean Union\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.083:0.083[0]; log p(e|m)= -2.937:-2.937[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nconstituencies[0.047]; said[0.045]; said[0.045]; said[0.045]; said[0.045]; parliamentary[0.045]; policy[0.045]; despite[0.043]; office[0.043]; Major[0.043]; Major[0.043]; Major[0.043]; Major[0.043]; Major[0.043]; Major[0.043]; business[0.043]; aware[0.043]; Relations[0.042]; Relations[0.042]; Relations[0.042]; John[0.042]; London[0.042]; day[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Major\u001b[39m ==> ENTITY: \u001b[32mJohn Major\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nparliamentary[0.044]; Chancellor[0.044]; Major[0.042]; Major[0.042]; Major[0.042]; Major[0.042]; Major[0.042]; John[0.042]; Clarke[0.042]; Clarke[0.042]; Clarke[0.042]; Exchequer[0.042]; London[0.042]; prime[0.041]; Prime[0.041]; office[0.041]; despite[0.041]; policy[0.041]; minister[0.041]; speculation[0.041]; political[0.041]; job[0.040]; Minister[0.040]; aware[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Clarke\u001b[39m ==> ENTITY: \u001b[32mKenneth Clarke\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClarke[0.048]; Clarke[0.048]; John[0.046]; Chancellor[0.045]; London[0.043]; policy[0.043]; Exchequer[0.042]; Kenneth[0.042]; office[0.041]; job[0.041]; spokesman[0.041]; spokesman[0.041]; spokesman[0.041]; spokesman[0.041]; spokesman[0.041]; Minister[0.040]; Prime[0.040]; despite[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; political[0.040]; European[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kenneth Clarke\u001b[39m ==> ENTITY: \u001b[32mKenneth Clarke\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClarke[0.048]; Clarke[0.048]; Clarke[0.048]; John[0.045]; Chancellor[0.044]; parliamentary[0.042]; London[0.042]; policy[0.042]; Exchequer[0.041]; minister[0.040]; office[0.040]; job[0.040]; spokesman[0.040]; spokesman[0.040]; spokesman[0.040]; spokesman[0.040]; spokesman[0.040]; Minister[0.040]; Prime[0.040]; despite[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Major\u001b[39m ==> ENTITY: \u001b[32mJohn Major\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nparliamentary[0.044]; Major[0.043]; Major[0.043]; Major[0.043]; Major[0.043]; John[0.043]; Clarke[0.043]; Clarke[0.043]; Clarke[0.043]; prime[0.042]; Prime[0.042]; office[0.042]; despite[0.041]; policy[0.041]; minister[0.041]; speculation[0.041]; political[0.041]; job[0.041]; Minister[0.040]; aware[0.040]; constituencies[0.040]; meetings[0.040]; good[0.040]; good[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Major\u001b[39m ==> ENTITY: \u001b[32mJohn Major\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nparliamentary[0.044]; Chancellor[0.043]; Major[0.042]; Major[0.042]; Major[0.042]; Major[0.042]; Major[0.042]; John[0.042]; Clarke[0.042]; Clarke[0.042]; Clarke[0.042]; Clarke[0.042]; Exchequer[0.042]; London[0.042]; European[0.041]; prime[0.041]; Prime[0.041]; office[0.041]; despite[0.041]; policy[0.041]; minister[0.041]; job[0.040]; political[0.040]; Minister[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.083:0.083[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \noffice[0.046]; said[0.045]; said[0.045]; said[0.045]; said[0.045]; John[0.044]; European[0.044]; day[0.044]; Chancellor[0.044]; despite[0.043]; Friday[0.043]; Minister[0.043]; Clarke[0.043]; Clarke[0.043]; Clarke[0.043]; reports[0.043]; reports[0.043]; media[0.043]; media[0.043]; prime[0.042]; job[0.042]; meetings[0.042]; policy[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Clarke\u001b[39m ==> ENTITY: \u001b[32mKenneth Clarke\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClarke[0.051]; Clarke[0.051]; John[0.048]; parliamentary[0.045]; policy[0.045]; minister[0.043]; office[0.043]; job[0.043]; spokesman[0.043]; spokesman[0.043]; spokesman[0.043]; spokesman[0.043]; Minister[0.042]; Prime[0.042]; despite[0.042]; constituencies[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; business[0.042]; political[0.042]; European[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = John Major\u001b[39m ==> ENTITY: \u001b[32mJohn Major\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nparliamentary[0.044]; Chancellor[0.044]; Major[0.042]; Major[0.042]; Major[0.042]; Major[0.042]; Major[0.042]; Clarke[0.042]; Clarke[0.042]; Clarke[0.042]; Clarke[0.042]; Exchequer[0.042]; London[0.042]; prime[0.041]; Prime[0.041]; office[0.041]; despite[0.041]; policy[0.041]; minister[0.041]; speculation[0.041]; political[0.041]; job[0.040]; Minister[0.040]; aware[0.040]; \t\n\r [=============>................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s837ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1451/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1252testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tallinn\u001b[39m ==> ENTITY: \u001b[32mTallinn\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.098:0.098[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEstonia[0.047]; Estonian[0.046]; largest[0.044]; Tallinna[0.044]; Tallinna[0.044]; Tallinna[0.044]; banks[0.044]; month[0.044]; Friday[0.043]; bank[0.043]; bank[0.043]; assets[0.043]; kroons[0.043]; kroons[0.043]; kroons[0.043]; kroons[0.043]; kroons[0.043]; kroons[0.043]; kroons[0.043]; kroons[0.043]; kroons[0.043]; November[0.042]; time[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Estonia ==> ENTITY: \u001b[32mEstonia\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.090:0.090[0]; log p(e|m)= -0.411:-0.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEstonian[0.048]; Riga[0.045]; Tallinn[0.043]; month[0.043]; November[0.042]; half[0.042]; kroons[0.041]; kroons[0.041]; kroons[0.041]; kroons[0.041]; kroons[0.041]; kroons[0.041]; kroons[0.041]; kroons[0.041]; kroons[0.041]; kroons[0.041]; time[0.041]; increased[0.041]; Tallinna[0.041]; Tallinna[0.041]; Tallinna[0.041]; Friday[0.040]; rose[0.040]; rose[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Estonian ==> ENTITY: \u001b[32mEstonia\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.101:0.101[0]; log p(e|m)= -0.968:-0.968[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEstonia[0.051]; month[0.045]; Tallinn[0.045]; said[0.045]; said[0.045]; said[0.045]; November[0.044]; kroons[0.043]; kroons[0.043]; kroons[0.043]; kroons[0.043]; kroons[0.043]; kroons[0.043]; kroons[0.043]; kroons[0.043]; statement[0.043]; Tallinna[0.042]; Tallinna[0.042]; Tallinna[0.042]; Friday[0.042]; rose[0.042]; rose[0.042]; largest[0.042]; \t\n\r [=============>................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s895ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1454/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1238testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWall Street\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mWall Street\u001b[39m <---> \u001b[32mNew York Stock Exchange\u001b[39m\t\nSCORES: global= 0.263:0.250[\u001b[31m0.013\u001b[39m]; local()= 0.119:0.133[\u001b[32m0.014\u001b[39m]; log p(e|m)= -0.040:-2.293[\u001b[31m2.253\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstock[0.048]; Street[0.045]; Corp[0.044]; Corp[0.044]; York[0.044]; Wall[0.044]; Wall[0.044]; companies[0.043]; announced[0.042]; Thursday[0.042]; white[0.042]; tops[0.041]; swap[0.041]; speculates[0.041]; said[0.040]; bid[0.040]; deal[0.040]; proposal[0.040]; speculated[0.040]; list[0.040]; necessarily[0.039]; takeover[0.039]; billion[0.039]; likely[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = PaineWebber\u001b[39m ==> ENTITY: \u001b[32mPaine Webber\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.121:0.121[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmerged[0.044]; Marc[0.044]; operations[0.043]; investors[0.043]; today[0.043]; Montana[0.042]; Cohen[0.042]; million[0.042]; million[0.042]; million[0.042]; million[0.042]; headquartered[0.042]; share[0.041]; share[0.041]; operates[0.041]; Newmont[0.041]; Newmont[0.041]; Newmont[0.041]; Newmont[0.041]; deal[0.041]; deal[0.041]; second[0.040]; number[0.040]; available[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nevada ==> ENTITY: \u001b[32mNevada\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.135:0.135[0]; log p(e|m)= -0.267:-0.267[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNevada[0.049]; second[0.045]; United[0.044]; States[0.044]; California[0.043]; half[0.043]; Mexico[0.042]; Montana[0.042]; Central[0.042]; South[0.042]; America[0.041]; percent[0.041]; number[0.041]; tops[0.041]; Canada[0.040]; Canada[0.040]; million[0.040]; million[0.040]; million[0.040]; million[0.040]; based[0.040]; operations[0.040]; Albuquerque[0.040]; white[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Newmont\u001b[39m ==> ENTITY: \u001b[32mNewmont Mining Corporation\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmining[0.049]; Newmont[0.046]; Newmont[0.046]; Newmont[0.046]; operations[0.044]; exploration[0.044]; Canada[0.044]; Canada[0.044]; mines[0.044]; gold[0.043]; million[0.043]; million[0.043]; million[0.043]; million[0.043]; Mexico[0.043]; Chile[0.042]; Chile[0.042]; Nevada[0.042]; Australia[0.042]; Australia[0.042]; Montana[0.042]; Homestake[0.041]; Homestake[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Central Asia\u001b[39m ==> ENTITY: \u001b[32mCentral Asia\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.095:0.095[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKazakstan[0.047]; considered[0.046]; geographic[0.045]; America[0.045]; million[0.044]; million[0.044]; million[0.044]; million[0.044]; Canada[0.044]; Canada[0.044]; vast[0.044]; areas[0.044]; Corp[0.043]; California[0.043]; operations[0.043]; stock[0.043]; expansion[0.042]; necessary[0.042]; accounting[0.042]; South[0.042]; States[0.042]; Indonesia[0.042]; headquartered[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Newmont\u001b[39m ==> ENTITY: \u001b[32mNewmont Mining Corporation\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmining[0.048]; Minerals[0.045]; Newmont[0.045]; Newmont[0.045]; Newmont[0.045]; Newmont[0.045]; Newmont[0.045]; Barrick[0.044]; Gold[0.043]; gold[0.042]; deposit[0.042]; Nevada[0.041]; Busang[0.041]; Corp[0.040]; Indonesia[0.040]; Dome[0.040]; acquisition[0.039]; bid[0.039]; Thursday[0.039]; companies[0.039]; Placer[0.039]; benefit[0.039]; announced[0.039]; considered[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kazakstan\u001b[39m ==> ENTITY: \u001b[32mKazakhstan\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.080:0.080[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.049]; Canada[0.049]; Asia[0.048]; Montana[0.047]; second[0.046]; Australia[0.046]; Australia[0.046]; number[0.045]; PaineWebber[0.045]; Newmont[0.045]; Newmont[0.045]; Newmont[0.045]; Newmont[0.045]; States[0.045]; Mexico[0.045]; gold[0.045]; available[0.044]; million[0.044]; million[0.044]; million[0.044]; million[0.044]; Albuquerque[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Newmont\u001b[39m ==> ENTITY: \u001b[32mNewmont Mining Corporation\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmining[0.048]; Mining[0.048]; Newmont[0.045]; Newmont[0.045]; Newmont[0.045]; Newmont[0.045]; Newmont[0.045]; Newmont[0.045]; Gold[0.043]; Nevada[0.041]; Corp[0.041]; Corp[0.041]; acquisition[0.040]; bid[0.040]; Thursday[0.040]; companies[0.039]; Pacific[0.039]; benefit[0.039]; billion[0.039]; announced[0.039]; includes[0.039]; said[0.039]; said[0.039]; payoff[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.106:0.106[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.055]; Australia[0.048]; Australia[0.048]; match[0.047]; Mexico[0.045]; gold[0.045]; gold[0.045]; reported[0.045]; America[0.044]; second[0.044]; South[0.042]; States[0.042]; United[0.042]; based[0.041]; vast[0.041]; today[0.041]; presence[0.041]; operations[0.041]; Central[0.041]; Indonesia[0.041]; operates[0.041]; Chile[0.041]; Chile[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Newmont\u001b[39m ==> ENTITY: \u001b[32mNewmont Mining Corporation\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNewmont[0.047]; Newmont[0.047]; Newmont[0.047]; Newmont[0.047]; arb[0.043]; proposed[0.043]; money[0.042]; bid[0.041]; bid[0.041]; shares[0.041]; shares[0.041]; pocket[0.041]; wants[0.041]; cent[0.040]; said[0.040]; said[0.040]; near[0.040]; proposal[0.040]; Desk[0.040]; ratio[0.040]; board[0.040]; term[0.040]; merger[0.040]; stock[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Newmont\u001b[39m ==> ENTITY: \u001b[32mNewmont Mining Corporation\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNewmont[0.047]; Newmont[0.047]; Newmont[0.047]; Newmont[0.047]; arb[0.043]; proposed[0.043]; money[0.042]; bid[0.041]; bid[0.041]; shares[0.041]; shares[0.041]; pocket[0.041]; wants[0.041]; cent[0.040]; said[0.040]; said[0.040]; near[0.040]; proposal[0.040]; Desk[0.040]; ratio[0.040]; board[0.040]; term[0.040]; merger[0.040]; stock[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazil ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.046]; Australia[0.046]; Canada[0.043]; Canada[0.043]; States[0.043]; Mexico[0.043]; Central[0.042]; America[0.042]; second[0.042]; Chile[0.042]; Chile[0.042]; Francisco[0.042]; tops[0.042]; white[0.041]; South[0.041]; Asia[0.041]; Ghana[0.041]; operations[0.041]; presence[0.040]; term[0.040]; million[0.039]; million[0.039]; million[0.039]; million[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.048]; Canada[0.048]; match[0.048]; Australia[0.048]; Australia[0.048]; Mexico[0.046]; Brazil[0.046]; America[0.044]; Chile[0.043]; Chile[0.043]; Ghana[0.043]; California[0.043]; South[0.042]; Central[0.042]; gold[0.041]; gold[0.041]; based[0.041]; today[0.041]; considered[0.041]; Asia[0.041]; splitting[0.041]; list[0.041]; tops[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.103:0.103[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.047]; Asia[0.046]; Australia[0.045]; includes[0.044]; operates[0.043]; million[0.042]; million[0.042]; territory[0.042]; Canada[0.042]; Chile[0.041]; South[0.041]; America[0.041]; stock[0.041]; negotiations[0.041]; expansion[0.041]; Central[0.041]; Corp[0.041]; revenues[0.041]; States[0.040]; friendly[0.040]; list[0.040]; areas[0.040]; gold[0.040]; gold[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Newmont\u001b[39m ==> ENTITY: \u001b[32mNewmont Mining Corporation\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.216:0.216[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmining[0.049]; Newmont[0.046]; Newmont[0.046]; Newmont[0.046]; Newmont[0.046]; operations[0.044]; exploration[0.044]; Canada[0.044]; Canada[0.044]; mines[0.044]; gold[0.043]; million[0.043]; million[0.043]; million[0.043]; million[0.043]; Mexico[0.043]; Chile[0.042]; Chile[0.042]; Nevada[0.042]; Australia[0.041]; Australia[0.041]; Montana[0.041]; Homestake[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Newmont\u001b[39m ==> ENTITY: \u001b[32mNewmont Mining Corporation\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.229:0.229[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmining[0.051]; Mining[0.051]; Minerals[0.049]; Newmont[0.048]; Newmont[0.048]; Newmont[0.048]; Newmont[0.048]; Newmont[0.048]; Newmont[0.048]; Barrick[0.047]; Gold[0.046]; Nevada[0.044]; Corp[0.043]; Corp[0.043]; acquisition[0.042]; bid[0.042]; Thursday[0.042]; companies[0.042]; benefit[0.042]; billion[0.042]; announced[0.042]; includes[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = California ==> ENTITY: \u001b[32mCalifornia\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.119:0.119[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.047]; Chile[0.047]; Chile[0.047]; America[0.044]; Nevada[0.043]; Nevada[0.043]; second[0.042]; Brazil[0.041]; South[0.041]; Montana[0.041]; today[0.041]; based[0.041]; Canada[0.041]; Canada[0.041]; Australia[0.041]; Australia[0.041]; tops[0.040]; Francisco[0.040]; Central[0.040]; States[0.040]; Marc[0.039]; mines[0.039]; half[0.039]; mining[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Newmont\u001b[39m ==> ENTITY: \u001b[32mNewmont Mining Corporation\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmining[0.047]; Newmont[0.044]; Newmont[0.044]; Newmont[0.044]; Newmont[0.044]; Newmont[0.044]; operations[0.043]; exploration[0.043]; Canada[0.043]; million[0.042]; million[0.042]; Mexico[0.042]; Chile[0.041]; Nevada[0.041]; Australia[0.040]; Montana[0.040]; proposed[0.040]; Homestake[0.040]; revenues[0.040]; second[0.039]; PaineWebber[0.039]; Brazil[0.039]; shares[0.038]; shares[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Newmont\u001b[39m ==> ENTITY: \u001b[32mNewmont Mining Corporation\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.208:0.208[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmining[0.048]; Mining[0.048]; Mining[0.048]; Newmont[0.045]; Newmont[0.045]; Newmont[0.045]; Newmont[0.045]; Newmont[0.045]; Gold[0.043]; Nevada[0.041]; Corp[0.040]; Corp[0.040]; Homestake[0.040]; bid[0.039]; Thursday[0.039]; companies[0.039]; Pacific[0.039]; billion[0.039]; announced[0.039]; includes[0.039]; Brendan[0.039]; said[0.039]; said[0.039]; payoff[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Albuquerque ==> ENTITY: \u001b[32mAlbuquerque, New Mexico\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.075:0.075[0]; log p(e|m)= -0.109:-0.109[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.052]; Mexico[0.049]; second[0.047]; South[0.047]; operations[0.046]; Chile[0.046]; Chile[0.046]; based[0.046]; expansion[0.046]; said[0.045]; said[0.045]; Canada[0.045]; Canada[0.045]; match[0.045]; California[0.044]; United[0.044]; today[0.044]; list[0.044]; operates[0.044]; white[0.044]; think[0.043]; Central[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = South America\u001b[39m ==> ENTITY: \u001b[32mSouth America\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.119:0.119[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.048]; Canada[0.047]; Canada[0.047]; Francisco[0.047]; considered[0.047]; free[0.047]; areas[0.046]; Chile[0.046]; San[0.046]; vast[0.046]; geographic[0.045]; Australia[0.045]; Australia[0.045]; Central[0.045]; match[0.045]; States[0.044]; operations[0.044]; Asia[0.044]; locked[0.044]; revenues[0.044]; revenues[0.044]; necessary[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nevada ==> ENTITY: \u001b[32mNevada\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.056:0.056[0]; log p(e|m)= -0.267:-0.267[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhighest[0.049]; territory[0.049]; Street[0.048]; rival[0.047]; bid[0.046]; includes[0.046]; previous[0.046]; Thursday[0.046]; match[0.045]; Gold[0.045]; limits[0.045]; necessary[0.045]; vast[0.045]; free[0.045]; preserve[0.044]; gold[0.044]; tax[0.044]; necessarily[0.044]; fact[0.044]; acquisition[0.044]; buyer[0.044]; buyer[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.051]; Canada[0.045]; Canada[0.045]; Brazil[0.044]; second[0.043]; Ghana[0.043]; Mexico[0.041]; gold[0.041]; number[0.041]; America[0.041]; United[0.041]; million[0.041]; million[0.041]; million[0.041]; million[0.041]; today[0.041]; States[0.041]; Asia[0.040]; half[0.040]; rating[0.040]; Central[0.040]; list[0.040]; Chile[0.039]; Chile[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.051]; match[0.046]; Canada[0.044]; Canada[0.044]; Brazil[0.043]; second[0.043]; South[0.043]; Asia[0.042]; Ghana[0.042]; Indonesia[0.042]; geographic[0.042]; Mexico[0.041]; vast[0.041]; America[0.041]; United[0.040]; today[0.040]; States[0.040]; areas[0.040]; Central[0.039]; million[0.039]; million[0.039]; million[0.039]; million[0.039]; list[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Newmont\u001b[39m ==> ENTITY: \u001b[32mNewmont Mining Corporation\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNewmont[0.046]; Newmont[0.046]; Newmont[0.046]; Newmont[0.046]; Newmont[0.046]; Newmont[0.046]; Newmont[0.046]; arb[0.042]; proposed[0.042]; Homestake[0.041]; money[0.041]; second[0.041]; bid[0.040]; bid[0.040]; shares[0.040]; shares[0.040]; pocket[0.040]; wants[0.039]; cent[0.039]; said[0.039]; said[0.039]; said[0.039]; near[0.039]; earnings[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Newmont\u001b[39m ==> ENTITY: \u001b[32mNewmont Mining Corporation\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNewmont[0.046]; Newmont[0.046]; Newmont[0.046]; Newmont[0.046]; Newmont[0.046]; Newmont[0.046]; arb[0.042]; proposed[0.042]; Homestake[0.041]; money[0.041]; bid[0.041]; bid[0.041]; shares[0.040]; shares[0.040]; pocket[0.040]; wants[0.040]; cent[0.040]; said[0.040]; said[0.040]; said[0.040]; near[0.039]; earnings[0.039]; proposal[0.039]; does[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Montana ==> ENTITY: \u001b[32mMontana\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.113:0.113[0]; log p(e|m)= -0.308:-0.308[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStates[0.051]; United[0.046]; America[0.046]; Canada[0.046]; Canada[0.046]; available[0.046]; Brazil[0.045]; Mexico[0.043]; Asia[0.043]; Central[0.043]; Nevada[0.043]; Nevada[0.043]; second[0.043]; South[0.042]; Chile[0.042]; Chile[0.042]; tops[0.042]; California[0.042]; white[0.042]; number[0.042]; list[0.041]; Australia[0.041]; Australia[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Francisco ==> ENTITY: \u001b[32mSan Francisco\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.073:0.073[0]; log p(e|m)= -0.090:-0.090[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCalifornia[0.045]; operations[0.044]; Central[0.043]; Mexico[0.043]; Nevada[0.043]; Nevada[0.043]; million[0.042]; million[0.042]; million[0.042]; million[0.042]; South[0.042]; based[0.042]; America[0.042]; gold[0.041]; gold[0.041]; today[0.041]; Gold[0.041]; considered[0.041]; match[0.040]; deposit[0.040]; revenues[0.040]; revenues[0.040]; lowered[0.040]; Albuquerque[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Newmont\u001b[39m ==> ENTITY: \u001b[32mNewmont Mining Corporation\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmining[0.048]; Minerals[0.046]; Newmont[0.045]; Newmont[0.045]; Newmont[0.045]; Newmont[0.045]; Barrick[0.044]; Gold[0.043]; gold[0.043]; deposit[0.042]; Nevada[0.041]; Busang[0.041]; Corp[0.041]; Indonesia[0.040]; Dome[0.040]; acquisition[0.040]; bid[0.040]; Thursday[0.039]; companies[0.039]; Placer[0.039]; benefit[0.039]; geographic[0.039]; considered[0.039]; includes[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ghana ==> ENTITY: \u001b[32mGhana\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.392:-0.392[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.046]; Mexico[0.044]; Canada[0.044]; Canada[0.044]; second[0.043]; Australia[0.042]; Australia[0.042]; Chile[0.042]; Chile[0.042]; States[0.042]; tops[0.041]; number[0.041]; United[0.041]; term[0.041]; gold[0.041]; Earnings[0.041]; based[0.040]; headquartered[0.040]; California[0.040]; positive[0.040]; half[0.040]; lowered[0.040]; earnings[0.040]; earnings[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.028:0.028[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstock[0.051]; Corp[0.047]; Corp[0.047]; announced[0.046]; Thursday[0.045]; list[0.044]; speculates[0.044]; Street[0.043]; Street[0.043]; tops[0.043]; bid[0.043]; swap[0.043]; rejects[0.043]; speculated[0.043]; deal[0.043]; buyer[0.042]; buyer[0.042]; likely[0.042]; reluctantly[0.042]; takeover[0.042]; billion[0.041]; unsolicited[0.041]; Gold[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWall Street\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mWall Street\u001b[39m <---> \u001b[32mNew York Stock Exchange\u001b[39m\t\nSCORES: global= 0.263:0.250[\u001b[31m0.013\u001b[39m]; local()= 0.126:0.143[\u001b[32m0.018\u001b[39m]; log p(e|m)= -0.040:-2.293[\u001b[31m2.253\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstock[0.047]; Corp[0.044]; Corp[0.044]; Street[0.043]; term[0.043]; companies[0.043]; declined[0.043]; York[0.042]; announced[0.042]; Wall[0.042]; Wall[0.042]; Thursday[0.042]; white[0.041]; includes[0.041]; payoff[0.041]; tops[0.041]; swap[0.040]; offer[0.040]; speculates[0.040]; said[0.040]; said[0.040]; bid[0.040]; deal[0.040]; deal[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = N.M.\u001b[39m ==> ENTITY: \u001b[32mNew Mexico\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.049]; Nevada[0.046]; Nevada[0.046]; California[0.046]; America[0.045]; States[0.045]; South[0.044]; Canada[0.044]; Canada[0.044]; Australia[0.044]; Australia[0.044]; Albuquerque[0.043]; geographic[0.043]; Montana[0.043]; Chile[0.042]; Chile[0.042]; Dome[0.042]; Central[0.042]; Brazil[0.042]; considered[0.041]; Asia[0.041]; PaineWebber[0.041]; vast[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chile ==> ENTITY: \u001b[32mChile\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChile[0.048]; Mexico[0.046]; Brazil[0.044]; States[0.044]; list[0.044]; Canada[0.043]; Canada[0.043]; Australia[0.042]; Australia[0.042]; today[0.041]; second[0.041]; rating[0.041]; United[0.041]; million[0.040]; million[0.040]; million[0.040]; million[0.040]; said[0.040]; said[0.040]; said[0.040]; California[0.040]; going[0.040]; Ghana[0.040]; Asia[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chile ==> ENTITY: \u001b[32mChile\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.162:0.162[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChile[0.047]; Mexico[0.045]; San[0.043]; match[0.043]; Brazil[0.043]; States[0.043]; list[0.043]; Francisco[0.042]; Canada[0.042]; Canada[0.042]; Australia[0.041]; Australia[0.041]; America[0.041]; today[0.041]; said[0.041]; said[0.041]; second[0.041]; rating[0.040]; United[0.040]; areas[0.040]; negotiations[0.040]; South[0.040]; considered[0.040]; geographic[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nevada ==> ENTITY: \u001b[32mNevada\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.267:-0.267[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNevada[0.049]; United[0.044]; States[0.044]; California[0.043]; Dome[0.043]; Placer[0.043]; areas[0.043]; Montana[0.042]; geographic[0.042]; Central[0.042]; South[0.042]; America[0.041]; considered[0.041]; tops[0.041]; Canada[0.040]; Canada[0.040]; million[0.040]; million[0.040]; million[0.040]; million[0.040]; based[0.040]; Albuquerque[0.040]; operations[0.040]; San[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = New York Stock Exchange\u001b[39m ==> ENTITY: \u001b[32mNew York Stock Exchange\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstock[0.058]; shares[0.054]; shares[0.054]; Friday[0.052]; trade[0.050]; arb[0.050]; said[0.050]; said[0.050]; said[0.050]; discuss[0.049]; Wall[0.049]; bid[0.049]; bid[0.049]; referring[0.049]; board[0.049]; number[0.049]; merger[0.048]; added[0.048]; added[0.048]; proposed[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Newmont\u001b[39m ==> ENTITY: \u001b[32mNewmont Mining Corporation\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNewmont[0.047]; Newmont[0.047]; Newmont[0.047]; Newmont[0.047]; Newmont[0.047]; arb[0.042]; proposed[0.042]; money[0.042]; bid[0.041]; bid[0.041]; shares[0.040]; shares[0.040]; pocket[0.040]; wants[0.040]; cent[0.040]; said[0.040]; said[0.040]; said[0.040]; near[0.040]; proposal[0.040]; does[0.040]; Desk[0.040]; ratio[0.040]; board[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.061]; Australia[0.053]; Australia[0.053]; Mexico[0.049]; reported[0.049]; today[0.049]; number[0.048]; America[0.048]; second[0.048]; South[0.046]; States[0.046]; United[0.046]; deal[0.046]; going[0.045]; based[0.045]; presence[0.045]; operations[0.045]; Central[0.045]; operates[0.045]; Chile[0.045]; Chile[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Newmont\u001b[39m ==> ENTITY: \u001b[32mNewmont Mining Corporation\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.215:0.215[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmining[0.047]; Mining[0.047]; Mining[0.047]; Newmont[0.044]; Newmont[0.044]; Newmont[0.044]; Newmont[0.044]; Newmont[0.044]; Newmont[0.044]; Gold[0.042]; Nevada[0.041]; Corp[0.040]; Corp[0.040]; Homestake[0.040]; acquisition[0.039]; bid[0.039]; Thursday[0.039]; companies[0.039]; Pacific[0.039]; benefit[0.039]; billion[0.039]; announced[0.038]; includes[0.038]; Brendan[0.038]; \t\n\r [==============>...............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s540ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1495/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1336testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Guilin\u001b[39m ==> ENTITY: \u001b[32mGuilin\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGuilin[0.046]; southern[0.045]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; southwest[0.044]; Guangxi[0.044]; Guangxi[0.044]; Beijing[0.043]; Xinhua[0.042]; Xinhua[0.042]; region[0.042]; airport[0.041]; airport[0.041]; aircraft[0.041]; aircraft[0.041]; Aviation[0.040]; planes[0.040]; tourist[0.039]; State[0.039]; Friday[0.038]; popular[0.038]; authorities[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = State Council ==> ENTITY: \u001b[32mState Council of the People's Republic of China\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.204:0.204[0]; log p(e|m)= -0.335:-0.335[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.047]; China[0.047]; China[0.047]; China[0.047]; China[0.047]; Beijing[0.046]; Administration[0.045]; Administration[0.045]; Xinhua[0.044]; Xinhua[0.044]; granted[0.043]; Office[0.042]; economic[0.042]; Guilin[0.042]; Guilin[0.042]; Guilin[0.042]; Civil[0.042]; Foreign[0.042]; Guangxi[0.041]; Guangxi[0.041]; General[0.041]; gave[0.041]; authorities[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Guilin\u001b[39m ==> ENTITY: \u001b[32mGuilin\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGuilin[0.046]; southern[0.045]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; southwest[0.044]; Guangxi[0.044]; Guangxi[0.044]; Xinhua[0.042]; Xinhua[0.042]; region[0.042]; airport[0.041]; airport[0.041]; aircraft[0.041]; aircraft[0.041]; Aviation[0.041]; river[0.039]; mountain[0.039]; tourist[0.039]; tourist[0.039]; State[0.039]; Friday[0.039]; scenery[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Civil Aviation Administration of China\u001b[39m ==> ENTITY: \u001b[32mCivil Aviation Administration of China\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.045]; China[0.045]; China[0.045]; China[0.045]; aircraft[0.044]; aircraft[0.044]; Beijing[0.044]; planes[0.043]; Xinhua[0.042]; Xinhua[0.042]; Administration[0.042]; expansion[0.041]; State[0.041]; agency[0.041]; agency[0.041]; Friday[0.040]; expected[0.040]; airport[0.040]; airport[0.040]; Guangxi[0.039]; Guangxi[0.039]; handle[0.039]; region[0.039]; Customs[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Guangxi\u001b[39m ==> ENTITY: \u001b[32mGuangxi\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.223:0.223[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.045]; China[0.045]; China[0.045]; China[0.045]; China[0.045]; river[0.044]; southern[0.044]; Guangxi[0.043]; region[0.043]; Beijing[0.042]; Guilin[0.042]; Guilin[0.042]; Guilin[0.042]; southwest[0.041]; Xinhua[0.041]; Xinhua[0.041]; mountain[0.040]; land[0.039]; popular[0.039]; popular[0.039]; foreign[0.039]; foreign[0.039]; State[0.038]; Council[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.162:0.162[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.055]; China[0.055]; southern[0.049]; region[0.048]; Xinhua[0.046]; Xinhua[0.046]; southwest[0.045]; Guilin[0.045]; foreign[0.045]; foreign[0.045]; Guangxi[0.045]; Guangxi[0.045]; authorities[0.044]; group[0.044]; destinations[0.043]; expected[0.043]; Administration[0.043]; Administration[0.043]; economic[0.043]; Office[0.043]; aircraft[0.043]; aircraft[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = General Administration of Customs\u001b[39m ==> ENTITY: \u001b[32mGeneral Administration of Customs\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAdministration[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; China[0.046]; foreign[0.045]; foreign[0.045]; agency[0.044]; agency[0.044]; Foreign[0.044]; Beijing[0.043]; Xinhua[0.043]; Xinhua[0.043]; State[0.043]; airport[0.042]; airport[0.042]; economic[0.042]; authorities[0.041]; Guangxi[0.041]; Guangxi[0.041]; handle[0.041]; arm[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.173:0.173[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.052]; China[0.052]; China[0.052]; Beijing[0.049]; southern[0.046]; region[0.045]; Xinhua[0.043]; Xinhua[0.043]; southwest[0.042]; foreign[0.042]; foreign[0.042]; Foreign[0.041]; group[0.041]; spot[0.041]; handle[0.041]; expected[0.041]; authorities[0.041]; Administration[0.041]; Administration[0.041]; economic[0.041]; Guangxi[0.041]; Guangxi[0.041]; Office[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.173:0.173[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.052]; China[0.052]; China[0.052]; Beijing[0.049]; southern[0.046]; region[0.045]; Xinhua[0.043]; Xinhua[0.043]; southwest[0.042]; foreign[0.042]; foreign[0.042]; Foreign[0.041]; group[0.041]; spot[0.041]; handle[0.041]; expected[0.041]; authorities[0.041]; Administration[0.041]; Administration[0.041]; economic[0.041]; Guangxi[0.041]; Guangxi[0.041]; Office[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Xinhua\u001b[39m ==> ENTITY: \u001b[32mXinhua News Agency\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nXinhua[0.047]; Beijing[0.044]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; Administration[0.044]; Administration[0.044]; southwest[0.042]; Guangxi[0.042]; Guangxi[0.042]; region[0.042]; southern[0.041]; news[0.041]; Guilin[0.039]; Guilin[0.039]; Guilin[0.039]; tourist[0.039]; tourist[0.039]; river[0.039]; agency[0.039]; agency[0.039]; land[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Xinhua\u001b[39m ==> ENTITY: \u001b[32mXinhua News Agency\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nXinhua[0.047]; Beijing[0.044]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; China[0.044]; Administration[0.044]; Administration[0.044]; southwest[0.042]; Guangxi[0.042]; Guangxi[0.042]; region[0.042]; southern[0.041]; news[0.041]; Guilin[0.039]; Guilin[0.039]; Guilin[0.039]; tourist[0.039]; tourist[0.039]; river[0.039]; agency[0.039]; agency[0.039]; land[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Beijing ==> ENTITY: \u001b[32mBeijing\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.042:-0.042[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.048]; China[0.048]; China[0.048]; China[0.048]; airport[0.045]; airport[0.045]; southern[0.042]; Friday[0.041]; region[0.041]; southwest[0.041]; Aviation[0.040]; Xinhua[0.040]; Xinhua[0.040]; planes[0.040]; aircraft[0.040]; aircraft[0.040]; expansion[0.040]; Council[0.039]; Administration[0.039]; Administration[0.039]; spot[0.039]; Guilin[0.039]; Guilin[0.039]; gave[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Guangxi\u001b[39m ==> ENTITY: \u001b[32mGuangxi\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.223:0.223[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.045]; China[0.045]; China[0.045]; China[0.045]; China[0.045]; river[0.044]; southern[0.044]; Guangxi[0.043]; region[0.043]; Beijing[0.042]; Guilin[0.042]; Guilin[0.042]; Guilin[0.042]; southwest[0.041]; Xinhua[0.041]; Xinhua[0.041]; mountain[0.040]; land[0.039]; popular[0.039]; popular[0.039]; foreign[0.039]; foreign[0.039]; State[0.038]; Council[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Guilin\u001b[39m ==> ENTITY: \u001b[32mGuilin\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGuilin[0.045]; Guilin[0.045]; southern[0.045]; China[0.043]; China[0.043]; China[0.043]; China[0.043]; southwest[0.043]; Guangxi[0.043]; Guangxi[0.043]; Beijing[0.043]; Xinhua[0.042]; Xinhua[0.042]; region[0.041]; airport[0.040]; airport[0.040]; aircraft[0.040]; aircraft[0.040]; Aviation[0.040]; planes[0.040]; river[0.039]; mountain[0.038]; tourist[0.038]; State[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.191:0.191[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.051]; China[0.051]; China[0.051]; China[0.051]; Beijing[0.048]; southern[0.044]; Office[0.044]; region[0.044]; Administration[0.043]; Administration[0.043]; Council[0.042]; Xinhua[0.042]; Xinhua[0.042]; southwest[0.041]; economic[0.041]; foreign[0.041]; foreign[0.041]; General[0.041]; State[0.041]; said[0.040]; said[0.040]; said[0.040]; group[0.040]; \t\n\r [==============>...............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s456ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1510/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1225testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.106:0.106[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.048]; Britain[0.048]; Britain[0.048]; Manchester[0.044]; said[0.043]; day[0.042]; second[0.042]; hit[0.042]; linked[0.042]; city[0.042]; School[0.042]; Friday[0.042]; hits[0.041]; weeks[0.041]; northern[0.040]; attended[0.040]; authorities[0.040]; officials[0.040]; contracted[0.040]; scare[0.039]; scare[0.039]; school[0.039]; Grammar[0.039]; evidence[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sale Grammar School\u001b[39m ==> ENTITY: \u001b[32mSale Grammar School\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nschool[0.047]; school[0.047]; university[0.046]; cluster[0.046]; hospital[0.045]; campus[0.045]; old[0.044]; old[0.044]; antibiotics[0.044]; England[0.044]; students[0.044]; London[0.044]; day[0.043]; meningitis[0.043]; meningitis[0.043]; meningitis[0.043]; Cardiff[0.042]; second[0.042]; time[0.042]; health[0.042]; Manchester[0.041]; Wales[0.041]; weeks[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cardiff ==> ENTITY: \u001b[32mCardiff\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.205:-0.205[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.049]; Wales[0.047]; Manchester[0.045]; Scotland[0.045]; area[0.043]; given[0.042]; time[0.041]; cases[0.041]; cases[0.041]; cases[0.041]; people[0.041]; people[0.041]; people[0.041]; people[0.041]; people[0.041]; having[0.041]; suspected[0.041]; linked[0.040]; Lanarkshire[0.040]; evidence[0.040]; day[0.040]; northern[0.040]; School[0.039]; came[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wales ==> ENTITY: \u001b[32mWales\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.164:0.164[0]; log p(e|m)= -0.373:-0.373[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.051]; Cardiff[0.050]; Scotland[0.047]; Manchester[0.045]; Britain[0.044]; attended[0.042]; time[0.042]; weeks[0.041]; northern[0.041]; given[0.041]; area[0.041]; having[0.040]; School[0.040]; authority[0.040]; Lanarkshire[0.040]; county[0.040]; school[0.040]; university[0.039]; people[0.039]; people[0.039]; people[0.039]; people[0.039]; people[0.039]; day[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scotland ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.229:-0.229[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWales[0.048]; Cardiff[0.048]; area[0.045]; county[0.044]; Lanarkshire[0.043]; having[0.043]; time[0.042]; officials[0.042]; people[0.041]; people[0.041]; people[0.041]; people[0.041]; people[0.041]; given[0.041]; linked[0.040]; widespread[0.040]; came[0.040]; moment[0.040]; listed[0.040]; described[0.040]; old[0.039]; university[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.055]; England[0.052]; Wales[0.045]; London[0.045]; northern[0.043]; Manchester[0.042]; second[0.041]; day[0.041]; hit[0.041]; given[0.041]; time[0.041]; Grammar[0.041]; cases[0.040]; city[0.040]; contracted[0.040]; School[0.040]; weeks[0.040]; officials[0.039]; authorities[0.039]; linked[0.039]; died[0.039]; died[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Manchester ==> ENTITY: \u001b[32mManchester\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.322:-0.322[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.049]; London[0.046]; Britain[0.046]; Britain[0.046]; Cardiff[0.045]; Wales[0.043]; city[0.042]; School[0.042]; school[0.041]; school[0.041]; health[0.041]; people[0.041]; old[0.041]; old[0.041]; second[0.040]; northern[0.040]; time[0.040]; said[0.040]; said[0.040]; campus[0.040]; day[0.040]; Friday[0.039]; hospital[0.039]; Grammar[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.190:-0.190[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCardiff[0.048]; Wales[0.048]; Britain[0.048]; Britain[0.048]; Manchester[0.047]; second[0.043]; London[0.042]; day[0.042]; weeks[0.041]; time[0.041]; School[0.041]; Grammar[0.040]; northern[0.040]; people[0.039]; said[0.039]; said[0.039]; contracted[0.039]; given[0.039]; school[0.039]; school[0.039]; Friday[0.039]; linked[0.039]; hit[0.039]; Sale[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lanarkshire\u001b[39m ==> ENTITY: \u001b[32mLanarkshire\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.121:0.121[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.049]; health[0.044]; university[0.043]; Wales[0.043]; having[0.042]; area[0.042]; people[0.042]; people[0.042]; people[0.042]; people[0.042]; people[0.042]; Cardiff[0.042]; antibiotics[0.041]; students[0.041]; school[0.041]; campus[0.041]; came[0.040]; county[0.040]; listed[0.040]; cases[0.040]; cases[0.040]; pensioners[0.040]; meningitis[0.040]; given[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.056]; England[0.052]; London[0.045]; northern[0.044]; Manchester[0.043]; evidence[0.042]; second[0.042]; hit[0.041]; city[0.041]; Grammar[0.041]; authorities[0.041]; day[0.040]; contracted[0.040]; School[0.040]; weeks[0.040]; died[0.039]; died[0.039]; linked[0.039]; said[0.039]; described[0.039]; old[0.039]; old[0.039]; cases[0.039]; attended[0.039]; \t\n\r [==============>...............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s463ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1520/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1242testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Port Louis ==> ENTITY: \u001b[32mPort Louis\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.157:0.157[0]; log p(e|m)= -0.035:-0.035[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMauritius[0.045]; Mauritius[0.045]; Mauritius[0.045]; north[0.045]; island[0.045]; island[0.045]; island[0.045]; south[0.045]; capital[0.044]; southwest[0.044]; northwest[0.043]; office[0.043]; Friday[0.043]; Friday[0.043]; Friday[0.043]; km[0.042]; km[0.042]; km[0.042]; Mauritian[0.042]; miles[0.042]; Winds[0.042]; Saturday[0.042]; authorities[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mauritius ==> ENTITY: \u001b[32mMauritius\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.331:-0.331[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMauritius[0.049]; Mauritius[0.049]; Madagascar[0.045]; island[0.045]; island[0.045]; island[0.045]; island[0.045]; Mauritian[0.044]; night[0.043]; Indian[0.043]; Louis[0.043]; Louis[0.043]; coast[0.043]; km[0.042]; km[0.042]; km[0.042]; afternoon[0.042]; afternoon[0.042]; hour[0.042]; hour[0.042]; northeastern[0.041]; Saturday[0.041]; capital[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Madagascar ==> ENTITY: \u001b[32mMadagascar\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.152:0.152[0]; log p(e|m)= -0.307:-0.307[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntropical[0.045]; island[0.045]; island[0.045]; island[0.045]; coast[0.044]; Mauritius[0.043]; Mauritius[0.043]; northeastern[0.043]; south[0.043]; southwest[0.042]; north[0.042]; Saturday[0.042]; northwest[0.041]; Friday[0.041]; Friday[0.041]; Friday[0.041]; said[0.040]; Louis[0.040]; capital[0.040]; closer[0.039]; change[0.039]; gone[0.039]; alert[0.039]; miles[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mauritian ==> ENTITY: \u001b[32mMauritius\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.170:0.170[0]; log p(e|m)= -0.546:-0.546[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMauritius[0.048]; Mauritius[0.048]; Mauritius[0.048]; island[0.044]; island[0.044]; island[0.044]; Indian[0.042]; Louis[0.041]; Port[0.041]; Port[0.041]; Saturday[0.041]; said[0.040]; said[0.040]; office[0.040]; Friday[0.040]; Friday[0.040]; Friday[0.040]; tropical[0.040]; south[0.040]; change[0.040]; capital[0.039]; added[0.039]; services[0.039]; north[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Port Louis ==> ENTITY: \u001b[32mPort Louis\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.159:0.159[0]; log p(e|m)= -0.035:-0.035[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMauritius[0.043]; Mauritius[0.043]; north[0.043]; island[0.043]; island[0.043]; island[0.043]; island[0.043]; south[0.043]; coast[0.043]; capital[0.042]; southwest[0.042]; northwest[0.041]; office[0.041]; Friday[0.041]; Friday[0.041]; Friday[0.041]; Friday[0.041]; km[0.041]; km[0.041]; km[0.041]; Mauritian[0.040]; northeastern[0.040]; miles[0.040]; Winds[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mauritius ==> ENTITY: \u001b[32mMauritius\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.331:-0.331[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMauritius[0.049]; Mauritius[0.049]; Madagascar[0.045]; island[0.045]; island[0.045]; island[0.045]; island[0.045]; Mauritian[0.044]; night[0.043]; Indian[0.043]; Louis[0.043]; Louis[0.043]; coast[0.043]; km[0.042]; km[0.042]; km[0.042]; afternoon[0.042]; afternoon[0.042]; hour[0.042]; hour[0.042]; northeastern[0.041]; Saturday[0.041]; capital[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mauritius ==> ENTITY: \u001b[32mMauritius\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.331:-0.331[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMauritius[0.053]; Mauritius[0.053]; island[0.048]; island[0.048]; island[0.048]; Mauritian[0.048]; Indian[0.046]; km[0.045]; km[0.045]; km[0.045]; Port[0.045]; tropical[0.044]; coming[0.044]; south[0.043]; said[0.043]; Ocean[0.043]; southwest[0.043]; office[0.043]; northwest[0.043]; moving[0.043]; closer[0.043]; added[0.043]; \t\n\r [==============>...............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s468ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1527/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1218testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Prague ==> ENTITY: \u001b[32mPrague\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.053:-0.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nweek[0.048]; Czech[0.047]; Havel[0.047]; Havel[0.047]; Havel[0.047]; Thursday[0.047]; place[0.046]; Friday[0.046]; statement[0.046]; took[0.046]; Vaclav[0.046]; yesterday[0.046]; earlier[0.045]; performed[0.045]; performed[0.045]; president[0.044]; president[0.044]; help[0.044]; help[0.044]; President[0.044]; presidential[0.043]; procedure[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Vaclav Havel\u001b[39m ==> ENTITY: \u001b[32mVáclav Havel\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHavel[0.049]; Havel[0.049]; Havel[0.049]; Prague[0.047]; Czech[0.044]; President[0.044]; president[0.042]; president[0.042]; president[0.042]; lung[0.041]; throat[0.041]; surgery[0.041]; emergency[0.040]; tracheotomy[0.040]; tracheotomy[0.040]; lungs[0.039]; presidential[0.039]; Doctors[0.039]; statement[0.039]; week[0.038]; said[0.038]; said[0.038]; said[0.038]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Havel\u001b[39m ==> ENTITY: \u001b[32mVáclav Havel\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.117:0.117[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHavel[0.053]; Havel[0.053]; president[0.045]; president[0.045]; lung[0.045]; lung[0.045]; lung[0.045]; surgery[0.044]; surgery[0.044]; tracheotomy[0.044]; pneumonia[0.043]; presidential[0.042]; Doctors[0.042]; statement[0.042]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; performed[0.041]; improved[0.041]; improved[0.041]; smoker[0.041]; Thursday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Havel\u001b[39m ==> ENTITY: \u001b[32mVáclav Havel\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHavel[0.051]; Havel[0.051]; president[0.043]; president[0.043]; president[0.043]; lung[0.043]; lung[0.043]; lung[0.043]; throat[0.042]; surgery[0.042]; surgery[0.042]; tracheotomy[0.042]; pneumonia[0.041]; presidential[0.040]; Doctors[0.040]; statement[0.040]; said[0.039]; said[0.039]; said[0.039]; said[0.039]; performed[0.039]; improved[0.039]; improved[0.039]; smoker[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.120:0.120[0]; log p(e|m)= -1.207:-1.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPrague[0.048]; Havel[0.047]; Havel[0.047]; Havel[0.047]; Vaclav[0.045]; week[0.043]; Thursday[0.043]; presidential[0.043]; Friday[0.043]; President[0.043]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; president[0.042]; president[0.042]; president[0.042]; performed[0.042]; performed[0.042]; place[0.042]; statement[0.042]; yesterday[0.042]; supportive[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Havel\u001b[39m ==> ENTITY: \u001b[32mVáclav Havel\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.135:0.135[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHavel[0.053]; Havel[0.053]; Havel[0.053]; Havel[0.053]; Vaclav[0.045]; president[0.045]; president[0.045]; president[0.045]; lung[0.045]; lung[0.045]; lung[0.045]; throat[0.044]; surgery[0.044]; surgery[0.044]; surgery[0.044]; tracheotomy[0.044]; pneumonia[0.043]; lungs[0.042]; presidential[0.042]; Doctors[0.042]; statement[0.042]; week[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Havel\u001b[39m ==> ENTITY: \u001b[32mVáclav Havel\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHavel[0.049]; Havel[0.049]; Havel[0.049]; Prague[0.046]; Czech[0.044]; President[0.043]; Vaclav[0.042]; president[0.041]; president[0.041]; president[0.041]; lung[0.041]; lung[0.041]; throat[0.041]; surgery[0.041]; surgery[0.041]; emergency[0.040]; tracheotomy[0.040]; tracheotomy[0.040]; lungs[0.039]; presidential[0.039]; Doctors[0.039]; Doctors[0.039]; statement[0.039]; week[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Havel\u001b[39m ==> ENTITY: \u001b[32mVáclav Havel\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.127:0.127[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHavel[0.050]; Havel[0.050]; Prague[0.047]; Czech[0.045]; President[0.044]; Vaclav[0.043]; president[0.042]; president[0.042]; lung[0.042]; throat[0.042]; surgery[0.041]; emergency[0.041]; tracheotomy[0.041]; tracheotomy[0.041]; lungs[0.040]; presidential[0.040]; Doctors[0.040]; statement[0.039]; week[0.039]; said[0.039]; said[0.039]; said[0.039]; procedure[0.038]; breathe[0.038]; \t\n\r [==============>...............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s486ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1535/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1317testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Algerians\u001b[39m ==> ENTITY: \u001b[32mAlgeria\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoslem[0.047]; Moslem[0.047]; Algerian[0.047]; Algerian[0.047]; south[0.045]; Blida[0.045]; province[0.044]; Algiers[0.044]; Paris[0.044]; security[0.043]; security[0.043]; civilians[0.043]; official[0.043]; group[0.042]; forces[0.042]; forces[0.042]; terrorists[0.042]; killed[0.042]; killed[0.042]; Friday[0.041]; overnight[0.041]; said[0.041]; said[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moslem ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.112:0.112[0]; log p(e|m)= -1.214:-1.214[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoslem[0.051]; fundamentalists[0.045]; fundamentalists[0.045]; said[0.043]; said[0.043]; Algerian[0.042]; Algerian[0.042]; civilians[0.042]; group[0.042]; province[0.041]; Algerians[0.041]; south[0.041]; official[0.041]; news[0.041]; security[0.041]; security[0.041]; forces[0.041]; forces[0.041]; terrorists[0.040]; overnight[0.040]; Friday[0.040]; carried[0.040]; killed[0.040]; killed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Algerian ==> ENTITY: \u001b[32mAlgeria\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.209:-0.209[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlgerian[0.046]; Algiers[0.046]; Algerians[0.045]; security[0.045]; security[0.045]; forces[0.045]; forces[0.045]; Blida[0.044]; terrorists[0.044]; civilians[0.044]; south[0.043]; group[0.043]; province[0.043]; Paris[0.043]; overnight[0.043]; killed[0.042]; killed[0.042]; official[0.042]; Moslem[0.042]; Moslem[0.042]; said[0.042]; said[0.042]; Friday[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Algerian ==> ENTITY: \u001b[32mAlgeria\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.209:-0.209[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlgerian[0.046]; Algiers[0.046]; Algerians[0.045]; security[0.045]; security[0.045]; forces[0.045]; forces[0.045]; Blida[0.044]; terrorists[0.044]; civilians[0.044]; south[0.043]; group[0.043]; province[0.043]; Paris[0.043]; overnight[0.043]; killed[0.042]; killed[0.042]; official[0.042]; Moslem[0.042]; Moslem[0.042]; said[0.042]; said[0.042]; Friday[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Paris ==> ENTITY: \u001b[32mParis\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.053:0.053[0]; log p(e|m)= -0.115:-0.115[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nofficial[0.043]; south[0.043]; news[0.043]; Algiers[0.043]; Friday[0.042]; kill[0.042]; forces[0.042]; forces[0.042]; security[0.042]; security[0.042]; agency[0.041]; agency[0.041]; statement[0.041]; killed[0.041]; killed[0.041]; Algerian[0.041]; Algerian[0.041]; said[0.041]; said[0.041]; terrorists[0.041]; overnight[0.041]; Algerians[0.040]; group[0.040]; province[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAlgiers\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAlgeria\u001b[39m <---> \u001b[32mAlgiers\u001b[39m\t\nSCORES: global= 0.269:0.269[\u001b[31m0.001\u001b[39m]; local()= 0.133:0.119[\u001b[31m0.013\u001b[39m]; log p(e|m)= -2.749:-0.104[\u001b[32m2.645\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlgerian[0.047]; Algerian[0.047]; Blida[0.047]; Algerians[0.045]; south[0.043]; civilians[0.043]; official[0.043]; province[0.043]; Paris[0.043]; forces[0.042]; forces[0.042]; Friday[0.041]; security[0.041]; security[0.041]; killed[0.040]; killed[0.040]; Moslem[0.040]; Moslem[0.040]; overnight[0.040]; group[0.039]; terrorists[0.039]; agency[0.038]; agency[0.038]; kill[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBlida\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBlida\u001b[39m <---> \u001b[32mBlida Province\u001b[39m\t\nSCORES: global= 0.289:0.249[\u001b[31m0.041\u001b[39m]; local()= 0.138:0.086[\u001b[31m0.052\u001b[39m]; log p(e|m)= 0.000:-1.291[\u001b[31m1.291\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nprovince[0.047]; said[0.044]; said[0.044]; south[0.044]; Algiers[0.043]; official[0.043]; Algerians[0.042]; Algerian[0.042]; Algerian[0.042]; security[0.042]; security[0.042]; civilians[0.042]; group[0.041]; news[0.041]; Friday[0.041]; Aps[0.041]; killed[0.040]; killed[0.040]; agency[0.040]; agency[0.040]; statement[0.040]; Paris[0.040]; Moslem[0.039]; Moslem[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moslem ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.112:0.112[0]; log p(e|m)= -1.214:-1.214[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoslem[0.051]; fundamentalists[0.045]; fundamentalists[0.045]; said[0.043]; said[0.043]; Algerian[0.042]; Algerian[0.042]; civilians[0.042]; group[0.042]; province[0.041]; Algerians[0.041]; south[0.041]; official[0.041]; news[0.041]; security[0.041]; security[0.041]; forces[0.041]; forces[0.041]; terrorists[0.040]; overnight[0.040]; Friday[0.040]; carried[0.040]; killed[0.040]; killed[0.040]; \t\n\r [==============>...............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s508ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1543/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1360testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jan Bos\u001b[39m ==> ENTITY: \u001b[32mJan Bos\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.044]; Canada[0.044]; Anke[0.044]; Netherlands[0.043]; Netherlands[0.043]; Netherlands[0.043]; Netherlands[0.043]; Velde[0.043]; Norway[0.042]; Van[0.042]; minute[0.041]; Baler[0.041]; Timmer[0.041]; Timmer[0.041]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; South[0.040]; South[0.040]; South[0.040]; Wotherspoon[0.040]; Witty[0.040]; hyuk[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Netherlands ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.174:0.174[0]; log p(e|m)= -0.250:-0.250[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.053]; Netherlands[0.053]; Netherlands[0.053]; Norway[0.049]; Germany[0.049]; Germany[0.049]; Germany[0.049]; Germany[0.049]; Canada[0.049]; Canada[0.049]; Russia[0.046]; Russia[0.046]; Russia[0.046]; round[0.046]; round[0.046]; round[0.046]; Van[0.045]; Korea[0.045]; Korea[0.045]; Wotherspoon[0.044]; Men[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Korea ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.221:0.221[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.047]; Korea[0.047]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Men[0.042]; Men[0.042]; round[0.042]; round[0.042]; round[0.042]; Yoon[0.041]; South[0.041]; South[0.041]; China[0.041]; China[0.041]; Chonju[0.041]; Germany[0.040]; Germany[0.040]; Jin[0.040]; Cup[0.040]; Cup[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.183:0.183[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.048]; Japan[0.048]; Japan[0.048]; Japan[0.048]; Korea[0.045]; Korea[0.045]; China[0.042]; Germany[0.041]; Germany[0.041]; Canada[0.041]; Canada[0.041]; Women[0.040]; Netherlands[0.040]; Netherlands[0.040]; Netherlands[0.040]; Chris[0.040]; Norway[0.039]; round[0.039]; round[0.039]; Jeremy[0.039]; minute[0.039]; Eriko[0.039]; seconds[0.039]; Wotherspoon[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marianne Timmer\u001b[39m ==> ENTITY: \u001b[32mMarianne Timmer\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWotherspoon[0.044]; Van[0.043]; Germany[0.043]; Germany[0.043]; Netherlands[0.043]; Netherlands[0.043]; Netherlands[0.043]; Jan[0.043]; Norway[0.042]; Women[0.042]; Canada[0.042]; Canada[0.042]; Franziska[0.041]; Witty[0.041]; Yoon[0.041]; Schenk[0.041]; Korea[0.041]; Korea[0.041]; Kim[0.041]; China[0.040]; hyuk[0.040]; Xue[0.040]; Kyou[0.040]; Inoue[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.216:0.216[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.047]; Japan[0.047]; Japan[0.047]; Japan[0.047]; Japan[0.047]; Korea[0.044]; Korea[0.044]; Korea[0.044]; Cup[0.041]; Cup[0.041]; round[0.041]; round[0.041]; China[0.040]; China[0.040]; Germany[0.040]; Chonju[0.039]; Canada[0.039]; Skating[0.039]; Skating[0.039]; Men[0.039]; Women[0.039]; Netherlands[0.039]; World[0.039]; World[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.119:0.119[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.046]; Canada[0.043]; Canada[0.043]; Netherlands[0.043]; Netherlands[0.043]; Men[0.042]; Men[0.042]; Cup[0.042]; World[0.042]; Korea[0.042]; Korea[0.042]; Korea[0.042]; Norway[0.041]; Norway[0.041]; Women[0.041]; seconds[0.041]; seconds[0.041]; Japan[0.040]; Japan[0.040]; Japan[0.040]; Japan[0.040]; Japan[0.040]; Japan[0.040]; Japan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chris Witty\u001b[39m ==> ENTITY: \u001b[32mChris Witty\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.089:0.089[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnke[0.047]; Baler[0.044]; Jan[0.043]; China[0.042]; China[0.042]; Hiroyasu[0.042]; hyuk[0.042]; Timmer[0.042]; Canada[0.041]; Canada[0.041]; Canada[0.041]; Franziska[0.041]; Women[0.041]; Women[0.041]; Norway[0.041]; Korea[0.041]; Korea[0.041]; Korea[0.041]; Hiroaki[0.041]; Germany[0.041]; Germany[0.041]; Men[0.041]; Bos[0.041]; seconds[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sylvain Bouchard\u001b[39m ==> ENTITY: \u001b[32mSylvain Bouchard\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWorld[0.050]; World[0.050]; Anke[0.048]; Canada[0.047]; Canada[0.047]; Races[0.047]; Sylvain[0.047]; Bouchard[0.047]; skating[0.046]; Baler[0.045]; Skating[0.045]; Skating[0.045]; speed[0.045]; races[0.044]; Hiroyasu[0.044]; Witty[0.044]; Speed[0.044]; Speed[0.044]; Timmer[0.043]; Chonju[0.043]; Franziska[0.043]; Norway[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMen[0.048]; Netherlands[0.048]; Netherlands[0.048]; Netherlands[0.048]; Netherlands[0.048]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Canada[0.046]; Canada[0.046]; Women[0.044]; Women[0.044]; China[0.044]; China[0.044]; Korea[0.044]; Korea[0.044]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Wotherspoon[0.043]; seconds[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Korea ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.209:0.209[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.047]; Women[0.043]; Men[0.043]; Japan[0.043]; Japan[0.043]; Japan[0.043]; Japan[0.043]; Japan[0.043]; Japan[0.043]; round[0.042]; round[0.042]; Yoon[0.042]; South[0.041]; China[0.041]; China[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Jin[0.040]; Canada[0.040]; Canada[0.040]; Kim[0.040]; Lee[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Netherlands ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.142:0.142[0]; log p(e|m)= -0.250:-0.250[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.047]; Netherlands[0.047]; Norway[0.044]; Germany[0.044]; Germany[0.044]; Canada[0.044]; Canada[0.044]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Russia[0.041]; Russia[0.041]; round[0.041]; Van[0.041]; Korea[0.040]; Korea[0.040]; China[0.039]; South[0.038]; South[0.038]; Lee[0.038]; Women[0.038]; minute[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.048]; Canada[0.045]; Netherlands[0.044]; Netherlands[0.044]; Netherlands[0.044]; Korea[0.043]; Korea[0.043]; Norway[0.043]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; China[0.042]; Schenk[0.041]; Women[0.041]; round[0.040]; Russia[0.040]; Russia[0.040]; Van[0.039]; South[0.039]; South[0.039]; Kim[0.038]; Franziska[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.240:0.240[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.051]; Japan[0.051]; Japan[0.051]; Japan[0.051]; Japan[0.051]; Japan[0.051]; Korea[0.047]; Korea[0.047]; Korea[0.047]; China[0.044]; China[0.044]; Germany[0.043]; Germany[0.043]; Canada[0.043]; Canada[0.043]; Russia[0.042]; Russia[0.042]; Russia[0.042]; Russia[0.042]; Men[0.042]; Men[0.042]; Women[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.058]; Russia[0.058]; Cup[0.056]; Cup[0.056]; Canada[0.055]; World[0.053]; World[0.053]; China[0.052]; China[0.052]; Skating[0.052]; Skating[0.052]; Norway[0.052]; Germany[0.051]; Germany[0.051]; Netherlands[0.051]; skating[0.050]; Korea[0.050]; Korea[0.050]; Korea[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.048]; Germany[0.048]; Canada[0.045]; Canada[0.045]; Canada[0.045]; Netherlands[0.045]; Netherlands[0.045]; Netherlands[0.045]; Korea[0.043]; Korea[0.043]; Korea[0.043]; Norway[0.043]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; China[0.042]; China[0.042]; Men[0.042]; Schenk[0.041]; Schenk[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.228:0.228[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.047]; Japan[0.047]; Japan[0.047]; Japan[0.047]; Japan[0.047]; Japan[0.047]; Korea[0.044]; Korea[0.044]; China[0.041]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Canada[0.040]; Canada[0.040]; Men[0.039]; Women[0.039]; Netherlands[0.039]; Netherlands[0.039]; Netherlands[0.039]; Netherlands[0.039]; Chris[0.039]; Chris[0.039]; Norway[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gerard Van Velde\u001b[39m ==> ENTITY: \u001b[32mGerard van Velde\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnke[0.047]; Canada[0.045]; Canada[0.045]; Witty[0.045]; Witty[0.045]; Baler[0.045]; Netherlands[0.044]; Netherlands[0.044]; Netherlands[0.044]; Netherlands[0.044]; Timmer[0.044]; Timmer[0.044]; minute[0.043]; Wotherspoon[0.043]; hyuk[0.043]; seconds[0.043]; Bos[0.043]; Norway[0.042]; Franziska[0.042]; Jan[0.042]; Korea[0.042]; Korea[0.042]; Kim[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.228:0.228[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.049]; Japan[0.049]; Japan[0.049]; Japan[0.049]; Japan[0.049]; Korea[0.045]; Korea[0.045]; Korea[0.045]; Cup[0.043]; Cup[0.043]; China[0.042]; China[0.042]; Germany[0.041]; Germany[0.041]; Chonju[0.041]; Canada[0.041]; Canada[0.041]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Skating[0.040]; Skating[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.093:0.093[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.047]; Russia[0.044]; Korea[0.043]; Korea[0.043]; Germany[0.043]; Germany[0.043]; Norway[0.042]; Netherlands[0.042]; Netherlands[0.042]; Netherlands[0.042]; South[0.041]; South[0.041]; Japan[0.041]; Japan[0.041]; Japan[0.041]; Japan[0.041]; Japan[0.041]; China[0.041]; round[0.041]; Witty[0.040]; Chris[0.039]; Lee[0.039]; Kim[0.039]; Women[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.175:0.175[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.049]; Japan[0.049]; Japan[0.049]; Japan[0.049]; Korea[0.045]; Korea[0.045]; China[0.042]; Germany[0.041]; Germany[0.041]; Canada[0.041]; Women[0.040]; Netherlands[0.040]; Netherlands[0.040]; Chris[0.040]; Norway[0.040]; round[0.040]; Jeremy[0.040]; Eriko[0.039]; Wotherspoon[0.039]; Kim[0.038]; Inoue[0.038]; Lee[0.038]; Witty[0.038]; Yoon[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.252:0.252[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.051]; Japan[0.051]; Japan[0.051]; Japan[0.051]; Japan[0.051]; Japan[0.051]; Korea[0.047]; Korea[0.047]; Korea[0.047]; Korea[0.047]; China[0.043]; China[0.043]; Germany[0.043]; Germany[0.043]; Canada[0.042]; Canada[0.042]; Russia[0.042]; Russia[0.042]; Russia[0.042]; Russia[0.042]; Men[0.042]; Women[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sylvain Bouchard\u001b[39m ==> ENTITY: \u001b[32mSylvain Bouchard\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.097:0.097[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnke[0.046]; Canada[0.045]; Canada[0.045]; Canada[0.045]; Sylvain[0.044]; Bouchard[0.044]; Baler[0.043]; Wotherspoon[0.042]; Van[0.041]; Witty[0.041]; hyuk[0.041]; Timmer[0.041]; Timmer[0.041]; Franziska[0.041]; Franziska[0.041]; Norway[0.040]; Jan[0.040]; Korea[0.040]; Korea[0.040]; Korea[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Men[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.052]; Bouchard[0.043]; Bouchard[0.043]; skating[0.043]; Cup[0.042]; Cup[0.042]; Women[0.042]; races[0.042]; Skating[0.042]; Skating[0.042]; World[0.041]; World[0.041]; Netherlands[0.041]; day[0.041]; Germany[0.041]; Germany[0.041]; Korea[0.040]; Korea[0.040]; Korea[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; China[0.040]; China[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Netherlands ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.159:0.159[0]; log p(e|m)= -0.250:-0.250[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.053]; Netherlands[0.053]; Netherlands[0.053]; Norway[0.049]; Germany[0.049]; Germany[0.049]; Germany[0.049]; Canada[0.049]; Canada[0.049]; Men[0.047]; round[0.046]; round[0.046]; Russia[0.046]; Russia[0.046]; Russia[0.046]; Van[0.046]; Women[0.045]; Korea[0.045]; Korea[0.045]; Schenk[0.044]; Wotherspoon[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kim Yoon-man\u001b[39m ==> ENTITY: \u001b[32mKim Yoon-man\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnke[0.045]; Korea[0.044]; Korea[0.044]; man[0.043]; Wotherspoon[0.043]; South[0.043]; South[0.043]; Lee[0.043]; Jin[0.043]; Baler[0.042]; hyuk[0.041]; Canada[0.041]; Canada[0.041]; Inoue[0.041]; Timmer[0.040]; Timmer[0.040]; Witty[0.040]; Witty[0.040]; Xue[0.040]; round[0.040]; round[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.085:0.085[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMen[0.051]; Netherlands[0.051]; World[0.050]; World[0.050]; Germany[0.049]; Cup[0.048]; Cup[0.048]; Canada[0.048]; Women[0.047]; China[0.047]; China[0.047]; Korea[0.047]; Korea[0.047]; Korea[0.047]; Skating[0.047]; Skating[0.047]; skating[0.046]; Russia[0.046]; Russia[0.046]; Russia[0.046]; seconds[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.072:0.072[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.047]; China[0.045]; Germany[0.044]; Germany[0.044]; Netherlands[0.043]; Netherlands[0.043]; Korea[0.042]; Korea[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; round[0.042]; Women[0.041]; South[0.041]; South[0.041]; Oksana[0.040]; Wotherspoon[0.040]; man[0.039]; Witty[0.038]; Jeremy[0.038]; Lee[0.038]; Sundstrom[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.167:0.167[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.049]; Japan[0.049]; Japan[0.049]; Japan[0.049]; Korea[0.046]; China[0.042]; Germany[0.042]; Germany[0.042]; Canada[0.041]; Women[0.041]; Netherlands[0.041]; Netherlands[0.041]; Chris[0.040]; round[0.040]; Jeremy[0.040]; Eriko[0.040]; Wotherspoon[0.039]; Kim[0.039]; Inoue[0.039]; Witty[0.039]; Yoon[0.038]; Xue[0.038]; Russia[0.038]; Kyoko[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWorld Cup\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFIFA World Cup\u001b[39m <---> \u001b[32mISU Speed Skating World Cup\u001b[39m\t\nSCORES: global= 0.247:0.222[\u001b[31m0.026\u001b[39m]; local()= 0.103:0.158[\u001b[32m0.055\u001b[39m]; log p(e|m)= -1.002:-3.963[\u001b[31m2.961\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.050]; World[0.049]; Cup[0.049]; Canada[0.047]; Races[0.046]; races[0.046]; Germany[0.046]; Women[0.045]; skating[0.045]; Skating[0.045]; Skating[0.045]; China[0.045]; South[0.044]; South[0.044]; South[0.044]; round[0.044]; round[0.044]; Korea[0.044]; Korea[0.044]; Korea[0.044]; Speed[0.044]; Speed[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marianne Timmer\u001b[39m ==> ENTITY: \u001b[32mMarianne Timmer\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnke[0.045]; Wotherspoon[0.045]; Van[0.045]; Germany[0.044]; Germany[0.044]; Netherlands[0.044]; Netherlands[0.044]; Netherlands[0.044]; Jan[0.044]; Norway[0.044]; Norway[0.044]; Women[0.043]; Canada[0.043]; Canada[0.043]; Koroleva[0.043]; Jin[0.043]; Baler[0.043]; Sung[0.043]; Franziska[0.043]; Witty[0.042]; Yoon[0.042]; Yoon[0.042]; Schenk[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.210:0.210[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.047]; Japan[0.047]; Japan[0.047]; Japan[0.047]; Japan[0.047]; Korea[0.044]; Korea[0.044]; Korea[0.044]; Cup[0.041]; Cup[0.041]; China[0.040]; China[0.040]; round[0.040]; round[0.040]; Germany[0.040]; Germany[0.040]; Chonju[0.039]; Canada[0.039]; Skating[0.039]; Skating[0.039]; Men[0.039]; Men[0.039]; Women[0.039]; Netherlands[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.059]; Canada[0.059]; Bouchard[0.049]; Bouchard[0.049]; Women[0.047]; Women[0.047]; Netherlands[0.047]; Netherlands[0.047]; Netherlands[0.047]; Netherlands[0.047]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Korea[0.046]; Korea[0.046]; Korea[0.046]; Russia[0.046]; Russia[0.046]; Russia[0.046]; China[0.046]; China[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Korea ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.210:0.210[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.047]; Women[0.043]; Men[0.043]; Japan[0.043]; Japan[0.043]; Japan[0.043]; Japan[0.043]; Japan[0.043]; Japan[0.043]; round[0.042]; round[0.042]; Yoon[0.042]; South[0.041]; China[0.041]; China[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Jin[0.041]; Canada[0.040]; Canada[0.040]; Kim[0.040]; Lee[0.039]; hyuk[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jeremy Wotherspoon\u001b[39m ==> ENTITY: \u001b[32mJeremy Wotherspoon\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.116:0.116[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.046]; Canada[0.046]; Anke[0.045]; Korea[0.043]; Korea[0.043]; Baler[0.043]; seconds[0.042]; Kim[0.041]; Japan[0.041]; Japan[0.041]; Japan[0.041]; Japan[0.041]; Japan[0.041]; minute[0.041]; hyuk[0.041]; China[0.041]; China[0.041]; Norway[0.041]; Timmer[0.041]; Timmer[0.041]; Inoue[0.040]; Franziska[0.040]; Jin[0.040]; Lee[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.049]; Korea[0.045]; Korea[0.045]; Korea[0.045]; Korea[0.045]; Russia[0.044]; Russia[0.044]; Russia[0.044]; Russia[0.044]; Japan[0.044]; Japan[0.044]; Japan[0.044]; Japan[0.044]; Japan[0.044]; Japan[0.044]; Cup[0.042]; Germany[0.042]; Germany[0.042]; Canada[0.041]; Canada[0.041]; Norway[0.041]; Norway[0.041]; World[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chris Witty\u001b[39m ==> ENTITY: \u001b[32mChris Witty\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.063:0.063[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.044]; Timmer[0.043]; Canada[0.043]; Franziska[0.043]; Women[0.043]; Korea[0.042]; Germany[0.042]; Germany[0.042]; Kim[0.042]; round[0.042]; South[0.042]; Netherlands[0.041]; Netherlands[0.041]; Marianne[0.041]; Schenk[0.041]; Becky[0.041]; Velde[0.041]; Eriko[0.041]; Oksana[0.041]; Shiho[0.041]; Wotherspoon[0.041]; Miyabe[0.041]; Jeremy[0.041]; Anka[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.066:0.066[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.049]; Russia[0.045]; round[0.044]; Germany[0.043]; Germany[0.043]; South[0.043]; Women[0.043]; China[0.042]; Korea[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Netherlands[0.041]; Netherlands[0.041]; Witty[0.041]; Chris[0.041]; Kim[0.041]; Jeremy[0.040]; Baier[0.039]; Wotherspoon[0.039]; Anka[0.039]; Gerard[0.039]; Schenk[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sergey Klevchenya\u001b[39m ==> ENTITY: \u001b[32mSergey Klevchenya\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.045]; Russia[0.045]; Russia[0.045]; Anke[0.043]; Skating[0.043]; Skating[0.043]; skating[0.043]; seconds[0.043]; races[0.043]; speed[0.042]; Speed[0.042]; Speed[0.042]; Koroleva[0.041]; round[0.041]; round[0.041]; Baler[0.041]; China[0.040]; China[0.040]; Chonju[0.040]; Cup[0.040]; Cup[0.040]; Hiroyasu[0.039]; Svetlana[0.039]; Timmer[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Netherlands ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.149:0.149[0]; log p(e|m)= -0.250:-0.250[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.051]; Netherlands[0.051]; Norway[0.047]; Norway[0.047]; Germany[0.047]; Germany[0.047]; Canada[0.047]; Canada[0.047]; Canada[0.047]; Russia[0.044]; Russia[0.044]; Russia[0.044]; Russia[0.044]; round[0.044]; round[0.044]; Van[0.044]; Korea[0.043]; Korea[0.043]; Korea[0.043]; Korea[0.043]; Wotherspoon[0.042]; Men[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.213:0.213[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.047]; Japan[0.047]; Japan[0.047]; Japan[0.047]; Japan[0.047]; Korea[0.044]; Korea[0.044]; Korea[0.044]; Cup[0.041]; Cup[0.041]; China[0.041]; round[0.040]; round[0.040]; Germany[0.040]; Chonju[0.040]; Canada[0.040]; Skating[0.039]; Skating[0.039]; Men[0.039]; Women[0.039]; World[0.039]; World[0.039]; Norway[0.039]; Okazaki[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sergey Klevchenya\u001b[39m ==> ENTITY: \u001b[32mSergey Klevchenya\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.102:0.102[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.046]; Russia[0.046]; Russia[0.046]; Anke[0.045]; seconds[0.044]; Koroleva[0.043]; round[0.042]; round[0.042]; round[0.042]; Baler[0.042]; China[0.041]; China[0.041]; Wotherspoon[0.041]; hyuk[0.040]; Svetlana[0.040]; Timmer[0.040]; Timmer[0.040]; Franziska[0.040]; Franziska[0.040]; minute[0.040]; Witty[0.040]; Norway[0.040]; Jan[0.039]; Men[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kim Yoon-man\u001b[39m ==> ENTITY: \u001b[32mKim Yoon-man\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnke[0.044]; Korea[0.044]; Korea[0.044]; Korea[0.044]; skating[0.043]; man[0.043]; South[0.043]; South[0.043]; South[0.043]; Skating[0.042]; Skating[0.042]; speed[0.042]; Jin[0.042]; Baler[0.042]; Saturday[0.042]; Chonju[0.040]; Sung[0.040]; Hiroyasu[0.040]; Canada[0.040]; Canada[0.040]; Inoue[0.040]; Timmer[0.040]; Witty[0.040]; Yeol[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.095:0.095[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.055]; Bouchard[0.045]; Women[0.044]; Netherlands[0.043]; Netherlands[0.043]; Netherlands[0.043]; Netherlands[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Korea[0.043]; Korea[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; China[0.043]; China[0.043]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Men[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.190:0.190[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.048]; Japan[0.048]; Japan[0.048]; Japan[0.048]; Korea[0.045]; Korea[0.045]; China[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Canada[0.040]; Canada[0.040]; Men[0.040]; Women[0.040]; Netherlands[0.040]; Netherlands[0.040]; Netherlands[0.040]; Chris[0.039]; Chris[0.039]; Norway[0.039]; round[0.039]; round[0.039]; Jeremy[0.039]; minute[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.169:0.169[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.058]; Russia[0.058]; Russia[0.058]; Cup[0.055]; Canada[0.055]; Canada[0.055]; World[0.052]; China[0.052]; China[0.052]; Norway[0.052]; Norway[0.052]; Germany[0.051]; Germany[0.051]; Netherlands[0.050]; Netherlands[0.050]; skating[0.050]; Korea[0.050]; Korea[0.050]; Korea[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.074:0.074[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.048]; Korea[0.047]; Japan[0.046]; Japan[0.046]; Japan[0.046]; Japan[0.046]; Germany[0.044]; Germany[0.044]; Russia[0.043]; Women[0.042]; Netherlands[0.041]; South[0.041]; Kim[0.041]; Chris[0.040]; round[0.040]; Xue[0.040]; Witty[0.040]; Yoon[0.039]; Schenk[0.038]; Wotherspoon[0.038]; man[0.038]; Timmer[0.038]; Jeremy[0.038]; Eriko[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.050]; Canada[0.048]; Canada[0.048]; Canada[0.048]; Korea[0.047]; Korea[0.047]; Korea[0.047]; Korea[0.047]; Japan[0.046]; Japan[0.046]; Japan[0.046]; Japan[0.046]; Japan[0.046]; Japan[0.046]; Men[0.044]; Germany[0.043]; Germany[0.043]; Women[0.043]; Norway[0.043]; Norway[0.043]; Witty[0.043]; seconds[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Casey Fitzrandolph\u001b[39m ==> ENTITY: \u001b[32mCasey FitzRandolph\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.119:0.119[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnke[0.050]; World[0.049]; World[0.049]; Baler[0.047]; Hiroyasu[0.046]; Saturday[0.045]; Timmer[0.045]; Chonju[0.045]; Witty[0.045]; Franziska[0.045]; Bouchard[0.045]; Bouchard[0.045]; Norway[0.045]; Kim[0.045]; Canada[0.044]; Canada[0.044]; Sylvain[0.044]; Sylvain[0.044]; Men[0.044]; Men[0.044]; skating[0.044]; races[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.050]; Russia[0.050]; Canada[0.048]; Canada[0.048]; round[0.046]; round[0.046]; round[0.046]; Chris[0.046]; Jeremy[0.045]; Norway[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Netherlands[0.044]; Netherlands[0.044]; Netherlands[0.044]; Netherlands[0.044]; Sergey[0.044]; China[0.043]; China[0.043]; Svetlana[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Korea ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.190:0.190[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.047]; Korea[0.047]; Women[0.043]; Men[0.043]; Japan[0.043]; Japan[0.043]; Japan[0.043]; Japan[0.043]; round[0.042]; round[0.042]; South[0.042]; South[0.042]; China[0.041]; Chonju[0.041]; Skating[0.041]; Skating[0.041]; Cup[0.040]; Cup[0.040]; Canada[0.040]; World[0.039]; World[0.039]; Kim[0.039]; Russia[0.039]; skating[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Korea ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.217:0.217[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.047]; Korea[0.047]; Women[0.043]; Men[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; round[0.042]; round[0.042]; Canada[0.041]; South[0.041]; South[0.041]; China[0.041]; China[0.041]; Chonju[0.041]; Germany[0.040]; Cup[0.040]; Cup[0.040]; Norway[0.040]; World[0.039]; World[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Franziska Schenk\u001b[39m ==> ENTITY: \u001b[32mFranziska Schenk\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.080:0.080[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnke[0.046]; Timmer[0.045]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Baler[0.043]; Becky[0.042]; Wotherspoon[0.042]; seconds[0.042]; Jan[0.041]; hyuk[0.041]; Marianne[0.041]; Netherlands[0.041]; Netherlands[0.041]; Netherlands[0.041]; Canada[0.041]; Canada[0.041]; Norway[0.041]; Bos[0.040]; Korea[0.040]; Korea[0.040]; Men[0.040]; Van[0.040]; Inoue[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.049]; Germany[0.049]; Canada[0.045]; Canada[0.045]; Netherlands[0.045]; Netherlands[0.045]; Netherlands[0.045]; Men[0.045]; Korea[0.044]; Korea[0.044]; Norway[0.044]; Women[0.043]; seconds[0.043]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; China[0.042]; Schenk[0.041]; minute[0.041]; Russia[0.040]; Russia[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Franziska Schenk\u001b[39m ==> ENTITY: \u001b[32mFranziska Schenk\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.103:0.103[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnke[0.047]; Timmer[0.046]; skating[0.046]; races[0.045]; World[0.045]; Germany[0.045]; Germany[0.045]; Baler[0.044]; speed[0.043]; seconds[0.043]; seconds[0.043]; Hiroyasu[0.043]; Jan[0.043]; hyuk[0.043]; Marianne[0.042]; Netherlands[0.042]; Netherlands[0.042]; Canada[0.042]; Canada[0.042]; Cup[0.042]; Norway[0.042]; Norway[0.042]; Bos[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.165:0.165[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.050]; Russia[0.050]; Russia[0.050]; Canada[0.048]; Canada[0.048]; Canada[0.048]; round[0.046]; round[0.046]; Chris[0.046]; Germany[0.045]; Germany[0.045]; Jeremy[0.045]; Norway[0.045]; Norway[0.045]; Netherlands[0.044]; Netherlands[0.044]; Netherlands[0.044]; Sergey[0.043]; Sergey[0.043]; China[0.043]; China[0.043]; Svetlana[0.043]; \t\n\r [===============>..............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s93ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1599/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1358testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Finland ==> ENTITY: \u001b[32mFinland\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.245:-0.245[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFinland[0.049]; ski[0.045]; Ski[0.044]; Austria[0.044]; Austria[0.044]; Pekka[0.043]; Germany[0.043]; second[0.043]; Kuusamo[0.042]; Standings[0.042]; Cup[0.041]; Cup[0.041]; Jumping[0.041]; jump[0.041]; jump[0.041]; jumping[0.040]; event[0.040]; points[0.040]; Japan[0.039]; Japan[0.039]; Japan[0.039]; Japan[0.039]; World[0.039]; World[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Soininen\u001b[39m ==> ENTITY: \u001b[32mJani Soininen\u001b[39m\t\nSCORES: global= 0.225:0.225[0]; local()= 0.142:0.142[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSchwarzenberger[0.045]; Primoz[0.045]; standings[0.044]; Finland[0.043]; Finland[0.043]; Brenden[0.042]; Brenden[0.042]; Pekka[0.042]; Norway[0.042]; Norway[0.042]; Soininen[0.042]; Jani[0.042]; Funaki[0.041]; Peterka[0.041]; Reinhard[0.041]; Hiroya[0.041]; Saitoh[0.040]; Saitoh[0.040]; Espen[0.040]; points[0.040]; Austria[0.040]; events[0.040]; Kristian[0.040]; Goldberger[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.135:0.135[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.050]; Germany[0.044]; Austria[0.044]; Austria[0.044]; Finland[0.044]; Finland[0.044]; Slovakia[0.044]; Cup[0.043]; second[0.042]; standings[0.042]; World[0.042]; Japan[0.040]; Japan[0.040]; Japan[0.040]; Japan[0.040]; Saturday[0.040]; Primoz[0.040]; Espen[0.040]; ski[0.039]; points[0.039]; points[0.039]; event[0.039]; Andreas[0.039]; events[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Primoz Peterka\u001b[39m ==> ENTITY: \u001b[32mPrimož Peterka\u001b[39m\t\nSCORES: global= 0.227:0.227[0]; local()= 0.119:0.119[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSchwarzenberger[0.045]; Slovakia[0.044]; standings[0.043]; Austria[0.043]; events[0.043]; Brenden[0.043]; Brenden[0.043]; points[0.042]; Pekka[0.042]; Finland[0.041]; Finland[0.041]; Funaki[0.041]; Reinhard[0.041]; World[0.041]; Hiroya[0.041]; Norway[0.041]; Norway[0.041]; Cup[0.041]; Soininen[0.041]; Soininen[0.041]; Saitoh[0.040]; Saitoh[0.040]; Espen[0.040]; Kristian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reinhard Schwarzenberger\u001b[39m ==> ENTITY: \u001b[32mReinhard Schwarzenberger\u001b[39m\t\nSCORES: global= 0.227:0.227[0]; local()= 0.118:0.118[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSki[0.046]; Primoz[0.045]; ski[0.044]; hill[0.043]; Kuusamo[0.043]; Brenden[0.042]; Brenden[0.042]; Pekka[0.042]; standings[0.042]; Andreas[0.041]; Funaki[0.041]; Funaki[0.041]; Peterka[0.041]; Hiroya[0.041]; Dieter[0.041]; World[0.041]; World[0.041]; World[0.041]; Soininen[0.040]; Soininen[0.040]; Saitoh[0.040]; Saitoh[0.040]; Austria[0.040]; Austria[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.167:0.167[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.048]; Germany[0.046]; Cup[0.043]; Cup[0.043]; Cup[0.043]; World[0.043]; World[0.043]; World[0.043]; Norway[0.042]; Andreas[0.042]; points[0.041]; points[0.041]; Finland[0.041]; Finland[0.041]; Finland[0.041]; Dieter[0.041]; Kuusamo[0.040]; event[0.040]; ski[0.040]; Japan[0.040]; Japan[0.040]; Japan[0.040]; Japan[0.040]; standings[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Espen Bredesen\u001b[39m ==> ENTITY: \u001b[32mEspen Bredesen\u001b[39m\t\nSCORES: global= 0.220:0.220[0]; local()= 0.092:0.092[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSchwarzenberger[0.045]; Primoz[0.045]; Finland[0.043]; Finland[0.043]; standings[0.043]; Brenden[0.042]; Brenden[0.042]; Pekka[0.042]; Norway[0.042]; Norway[0.042]; events[0.042]; Kristian[0.041]; Funaki[0.041]; Peterka[0.041]; Reinhard[0.041]; Hiroya[0.041]; Soininen[0.041]; Soininen[0.041]; Saitoh[0.040]; Saitoh[0.040]; World[0.040]; Dieter[0.040]; Austria[0.040]; Austria[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kazuyoshi Funaki\u001b[39m ==> ENTITY: \u001b[32mKazuyoshi Funaki\u001b[39m\t\nSCORES: global= 0.230:0.230[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \njump[0.045]; jump[0.045]; Schwarzenberger[0.044]; second[0.043]; Ski[0.043]; event[0.043]; jumping[0.042]; Cup[0.042]; Cup[0.042]; Cup[0.042]; Kuusamo[0.042]; standings[0.041]; ski[0.041]; World[0.041]; World[0.041]; World[0.041]; Brenden[0.041]; Brenden[0.041]; Finland[0.040]; Finland[0.040]; Finland[0.040]; Pekka[0.040]; points[0.040]; points[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jani Soininen\u001b[39m ==> ENTITY: \u001b[32mJani Soininen\u001b[39m\t\nSCORES: global= 0.225:0.225[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSchwarzenberger[0.045]; ski[0.044]; Primoz[0.044]; standings[0.043]; hill[0.043]; Finland[0.043]; Finland[0.043]; jump[0.043]; jump[0.043]; Brenden[0.042]; Brenden[0.042]; Pekka[0.041]; Norway[0.041]; Norway[0.041]; Soininen[0.041]; Funaki[0.041]; Funaki[0.041]; Peterka[0.041]; Reinhard[0.041]; Hiroya[0.040]; jumping[0.040]; Andreas[0.040]; Saitoh[0.040]; Saitoh[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Goldberger\u001b[39m ==> ENTITY: \u001b[32mAndreas Goldberger\u001b[39m\t\nSCORES: global= 0.220:0.220[0]; local()= 0.034:0.034[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSchwarzenberger[0.046]; Primoz[0.045]; Brenden[0.043]; Brenden[0.043]; Austria[0.043]; Austria[0.043]; Pekka[0.042]; Funaki[0.042]; Funaki[0.042]; Peterka[0.042]; Reinhard[0.042]; Hiroya[0.041]; Andreas[0.041]; Germany[0.041]; Soininen[0.041]; Soininen[0.041]; Saitoh[0.041]; Saitoh[0.041]; Espen[0.041]; World[0.041]; Dieter[0.040]; Finland[0.040]; Finland[0.040]; Kristian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Noriaki Kasai\u001b[39m ==> ENTITY: \u001b[32mNoriaki Kasai\u001b[39m\t\nSCORES: global= 0.234:0.234[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKuusamo[0.043]; Schwarzenberger[0.043]; event[0.043]; Ski[0.043]; Primoz[0.042]; Cup[0.042]; Cup[0.042]; Cup[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; events[0.042]; ski[0.042]; World[0.041]; World[0.041]; World[0.041]; Norway[0.041]; Norway[0.041]; Finland[0.040]; Finland[0.040]; Finland[0.040]; Brenden[0.040]; Brenden[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.138:0.138[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.045]; Austria[0.045]; Cup[0.043]; Cup[0.043]; Cup[0.043]; Norway[0.042]; Norway[0.042]; standings[0.042]; Finland[0.041]; Finland[0.041]; Finland[0.041]; Japan[0.041]; Japan[0.041]; Japan[0.041]; Japan[0.041]; World[0.041]; World[0.041]; World[0.041]; second[0.041]; Standings[0.041]; Andreas[0.040]; jump[0.040]; jump[0.040]; Dieter[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Slovakia ==> ENTITY: \u001b[32mSlovakia\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.080:0.080[0]; log p(e|m)= -0.367:-0.367[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.048]; Cup[0.046]; Finland[0.045]; Finland[0.045]; Norway[0.044]; Norway[0.044]; Japan[0.044]; Japan[0.044]; World[0.043]; points[0.043]; Primoz[0.043]; events[0.042]; standings[0.042]; Peterka[0.041]; Kristian[0.040]; Leading[0.040]; Pekka[0.039]; Brenden[0.039]; Brenden[0.039]; Goldberger[0.038]; Reinhard[0.037]; Espen[0.037]; Soininen[0.037]; Soininen[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.166:0.166[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.049]; Japan[0.049]; Japan[0.049]; second[0.044]; Cup[0.043]; Cup[0.043]; Cup[0.043]; Germany[0.042]; Austria[0.041]; Austria[0.041]; Ski[0.041]; World[0.040]; World[0.040]; World[0.040]; Norway[0.040]; Noriaki[0.040]; Hiroya[0.040]; event[0.039]; Finland[0.039]; Finland[0.039]; Finland[0.039]; points[0.039]; points[0.039]; standings[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWorld Cup\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFIFA World Cup\u001b[39m <---> \u001b[32mFIS Ski Jumping World Cup\u001b[39m\t\nSCORES: global= 0.233:0.231[\u001b[31m0.002\u001b[39m]; local()= 0.095:0.188[\u001b[32m0.093\u001b[39m]; log p(e|m)= -1.002:-3.772[\u001b[31m2.770\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFinland[0.047]; Finland[0.047]; Norway[0.047]; Norway[0.047]; Austria[0.044]; Austria[0.044]; events[0.044]; Slovakia[0.043]; standings[0.042]; Germany[0.042]; jump[0.042]; jump[0.042]; second[0.042]; Japan[0.040]; Japan[0.040]; Japan[0.040]; Japan[0.040]; points[0.039]; points[0.039]; Pekka[0.038]; Kristian[0.038]; Andreas[0.038]; Espen[0.038]; Primoz[0.037]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWorld Cup\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFIS Alpine Ski World Cup\u001b[39m <---> \u001b[32mFIS Ski Jumping World Cup\u001b[39m\t\nSCORES: global= 0.234:0.233[\u001b[31m0.001\u001b[39m]; local()= 0.176:0.223[\u001b[32m0.047\u001b[39m]; log p(e|m)= -2.765:-3.772[\u001b[31m1.008\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFinland[0.046]; Finland[0.046]; Finland[0.046]; Ski[0.045]; World[0.044]; Cup[0.044]; ski[0.044]; Austria[0.043]; Austria[0.043]; event[0.043]; Standings[0.042]; Kuusamo[0.041]; Germany[0.041]; jump[0.041]; jump[0.041]; second[0.041]; Jumping[0.040]; jumping[0.039]; Japan[0.039]; Japan[0.039]; Japan[0.039]; Japan[0.039]; Saturday[0.038]; points[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.119:0.119[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.051]; World[0.047]; Germany[0.046]; Austria[0.046]; Finland[0.046]; Finland[0.046]; Slovakia[0.045]; Cup[0.045]; standings[0.043]; Japan[0.042]; Japan[0.042]; Primoz[0.041]; Kristian[0.041]; Espen[0.041]; events[0.040]; points[0.040]; Dieter[0.039]; Pekka[0.038]; Leading[0.038]; Thoma[0.037]; Thoma[0.037]; Peterka[0.037]; Jani[0.037]; Reinhard[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Finland ==> ENTITY: \u001b[32mFinland\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.177:0.177[0]; log p(e|m)= -0.245:-0.245[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFinland[0.049]; Norway[0.047]; Norway[0.047]; Slovakia[0.044]; Austria[0.044]; Austria[0.044]; standings[0.043]; Pekka[0.043]; Germany[0.043]; second[0.042]; Cup[0.041]; Cup[0.041]; Primoz[0.041]; Andreas[0.040]; World[0.040]; World[0.040]; points[0.039]; points[0.039]; Jani[0.039]; Espen[0.039]; Japan[0.039]; Japan[0.039]; Japan[0.039]; Japan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Finland ==> ENTITY: \u001b[32mFinland\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.177:0.177[0]; log p(e|m)= -0.245:-0.245[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFinland[0.048]; Finland[0.048]; Norway[0.046]; Norway[0.046]; Slovakia[0.043]; Austria[0.043]; Austria[0.043]; standings[0.042]; Pekka[0.042]; Germany[0.042]; second[0.041]; Kuusamo[0.041]; Standings[0.041]; Cup[0.040]; Cup[0.040]; Cup[0.040]; Primoz[0.040]; Jumping[0.040]; Andreas[0.039]; events[0.039]; event[0.039]; World[0.039]; World[0.039]; World[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andreas Goldberger\u001b[39m ==> ENTITY: \u001b[32mAndreas Goldberger\u001b[39m\t\nSCORES: global= 0.226:0.226[0]; local()= 0.087:0.087[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSchwarzenberger[0.045]; jump[0.044]; jump[0.044]; Brenden[0.042]; Brenden[0.042]; Austria[0.042]; Austria[0.042]; Pekka[0.042]; Ski[0.042]; jumping[0.042]; ski[0.042]; Funaki[0.041]; Funaki[0.041]; Reinhard[0.041]; Hiroya[0.041]; Germany[0.041]; Kuusamo[0.041]; Soininen[0.041]; Saitoh[0.040]; Saitoh[0.040]; event[0.040]; World[0.040]; World[0.040]; World[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Okabe\u001b[39m ==> ENTITY: \u001b[32mTakanobu Okabe\u001b[39m\t\nSCORES: global= 0.221:0.221[0]; local()= 0.116:0.116[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSchwarzenberger[0.047]; Primoz[0.046]; Cup[0.045]; Japan[0.045]; Japan[0.045]; Norway[0.044]; Norway[0.044]; World[0.044]; Brenden[0.044]; Brenden[0.044]; Finland[0.044]; Finland[0.044]; events[0.043]; Pekka[0.043]; Funaki[0.043]; Peterka[0.042]; Reinhard[0.042]; Austria[0.042]; Austria[0.042]; Hiroya[0.042]; Kasai[0.042]; Noriaki[0.042]; Andreas[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.175:0.175[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.049]; Japan[0.049]; Japan[0.049]; Cup[0.043]; Cup[0.043]; Cup[0.043]; second[0.042]; Austria[0.041]; Austria[0.041]; Germany[0.041]; Ski[0.041]; Slovakia[0.041]; World[0.040]; World[0.040]; World[0.040]; Norway[0.040]; Norway[0.040]; Noriaki[0.040]; Hiroya[0.040]; Finland[0.039]; Finland[0.039]; Finland[0.039]; event[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Thoma\u001b[39m ==> ENTITY: \u001b[32mDieter Thoma\u001b[39m\t\nSCORES: global= 0.222:0.222[0]; local()= 0.112:0.112[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThoma[0.048]; Schwarzenberger[0.047]; Primoz[0.046]; jump[0.045]; jump[0.045]; standings[0.045]; Austria[0.044]; Austria[0.044]; Brenden[0.043]; Brenden[0.043]; Pekka[0.043]; Dieter[0.043]; Norway[0.043]; Norway[0.043]; Cup[0.043]; Funaki[0.042]; Funaki[0.042]; Peterka[0.042]; Reinhard[0.042]; Hiroya[0.042]; Andreas[0.042]; second[0.042]; World[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.191:0.191[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.053]; Germany[0.050]; Slovakia[0.050]; Andreas[0.047]; Norway[0.047]; Norway[0.047]; points[0.045]; points[0.045]; Finland[0.045]; Finland[0.045]; Finland[0.045]; event[0.045]; Dieter[0.045]; Kuusamo[0.044]; ski[0.044]; standings[0.044]; Primoz[0.044]; Ski[0.044]; second[0.044]; World[0.043]; World[0.043]; World[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dieter Thoma\u001b[39m ==> ENTITY: \u001b[32mDieter Thoma\u001b[39m\t\nSCORES: global= 0.225:0.225[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThoma[0.045]; Kuusamo[0.045]; Schwarzenberger[0.044]; Ski[0.044]; ski[0.043]; hill[0.043]; jump[0.042]; jump[0.042]; standings[0.042]; jumping[0.042]; Austria[0.041]; Austria[0.041]; Jumping[0.041]; Brenden[0.041]; Brenden[0.041]; event[0.041]; Pekka[0.041]; Norway[0.040]; Norway[0.040]; Cup[0.040]; Cup[0.040]; Cup[0.040]; Funaki[0.040]; Funaki[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nikkola\u001b[39m ==> ENTITY: \u001b[32mAri-Pekka Nikkola\u001b[39m\t\nSCORES: global= 0.223:0.223[0]; local()= 0.107:0.107[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSchwarzenberger[0.045]; Primoz[0.045]; Norway[0.044]; Norway[0.044]; Finland[0.044]; Finland[0.044]; standings[0.043]; Pekka[0.042]; Brenden[0.042]; Brenden[0.042]; Austria[0.041]; Funaki[0.041]; Peterka[0.041]; Reinhard[0.041]; Hiroya[0.041]; Soininen[0.040]; Soininen[0.040]; Saitoh[0.040]; Saitoh[0.040]; Espen[0.040]; Dieter[0.040]; Kristian[0.039]; World[0.039]; Germany[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kuusamo\u001b[39m ==> ENTITY: \u001b[32mKuusamo\u001b[39m\t\nSCORES: global= 0.231:0.231[0]; local()= 0.100:0.100[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFinland[0.052]; Finland[0.052]; Schwarzenberger[0.044]; Austria[0.044]; Austria[0.044]; Pekka[0.043]; ski[0.042]; World[0.042]; World[0.042]; Germany[0.041]; Ski[0.041]; Saturday[0.041]; Funaki[0.040]; Reinhard[0.040]; Hiroya[0.040]; Andreas[0.040]; event[0.040]; Saitoh[0.039]; Dieter[0.039]; hill[0.039]; Japan[0.038]; Japan[0.038]; Japan[0.038]; Japan[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Takanobu Okabe\u001b[39m ==> ENTITY: \u001b[32mTakanobu Okabe\u001b[39m\t\nSCORES: global= 0.227:0.227[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKuusamo[0.045]; ski[0.044]; Schwarzenberger[0.044]; Ski[0.042]; Cup[0.042]; Cup[0.042]; Cup[0.042]; hill[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; Norway[0.041]; jump[0.041]; jump[0.041]; World[0.041]; World[0.041]; World[0.041]; Brenden[0.041]; event[0.041]; Finland[0.041]; Finland[0.041]; Finland[0.041]; jumping[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kristian Brenden\u001b[39m ==> ENTITY: \u001b[32mKristian Brenden\u001b[39m\t\nSCORES: global= 0.229:0.229[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrenden[0.045]; Cup[0.044]; Schwarzenberger[0.044]; hill[0.044]; second[0.043]; Primoz[0.043]; ski[0.043]; standings[0.043]; Norway[0.042]; Norway[0.042]; event[0.042]; jump[0.041]; jump[0.041]; events[0.041]; Pekka[0.041]; Funaki[0.040]; Funaki[0.040]; Peterka[0.040]; Reinhard[0.040]; Finland[0.040]; Finland[0.040]; World[0.040]; Hiroya[0.040]; Andreas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Funaki\u001b[39m ==> ENTITY: \u001b[32mKazuyoshi Funaki\u001b[39m\t\nSCORES: global= 0.224:0.224[0]; local()= 0.118:0.118[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSchwarzenberger[0.045]; Primoz[0.045]; Cup[0.043]; standings[0.043]; World[0.043]; Brenden[0.042]; Brenden[0.042]; Finland[0.042]; Finland[0.042]; Pekka[0.042]; points[0.041]; Peterka[0.041]; Reinhard[0.041]; Hiroya[0.041]; events[0.041]; Andreas[0.041]; Japan[0.041]; Japan[0.041]; Norway[0.041]; Norway[0.041]; Soininen[0.040]; Soininen[0.040]; Saitoh[0.040]; Saitoh[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.175:0.175[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.049]; Japan[0.049]; Japan[0.049]; second[0.044]; Cup[0.043]; Cup[0.043]; Cup[0.043]; Germany[0.042]; Austria[0.041]; Austria[0.041]; Slovakia[0.041]; World[0.040]; World[0.040]; World[0.040]; Norway[0.040]; Norway[0.040]; Noriaki[0.040]; Hiroya[0.040]; event[0.039]; Finland[0.039]; Finland[0.039]; Finland[0.039]; points[0.039]; points[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ari-Pekka Nikkola\u001b[39m ==> ENTITY: \u001b[32mAri-Pekka Nikkola\u001b[39m\t\nSCORES: global= 0.225:0.225[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKuusamo[0.046]; Schwarzenberger[0.044]; Primoz[0.044]; Norway[0.043]; Norway[0.043]; Finland[0.043]; Finland[0.043]; Finland[0.043]; ski[0.042]; standings[0.042]; Brenden[0.041]; Brenden[0.041]; hill[0.041]; Ski[0.041]; Austria[0.040]; Austria[0.040]; Funaki[0.040]; Funaki[0.040]; Peterka[0.040]; Reinhard[0.040]; jump[0.040]; jump[0.040]; Hiroya[0.040]; Andreas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Brenden\u001b[39m ==> ENTITY: \u001b[32mKristian Brenden\u001b[39m\t\nSCORES: global= 0.227:0.227[0]; local()= 0.112:0.112[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrenden[0.046]; Cup[0.045]; Schwarzenberger[0.045]; Primoz[0.044]; standings[0.044]; Kristian[0.043]; Norway[0.043]; Norway[0.043]; jump[0.042]; events[0.041]; Pekka[0.041]; Funaki[0.041]; Funaki[0.041]; Peterka[0.041]; Reinhard[0.041]; Finland[0.041]; Finland[0.041]; World[0.040]; Hiroya[0.040]; Andreas[0.040]; Soininen[0.040]; Soininen[0.040]; Saitoh[0.040]; Saitoh[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.163:0.163[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.049]; Japan[0.049]; Japan[0.049]; second[0.044]; Cup[0.043]; Cup[0.043]; Cup[0.043]; Germany[0.042]; Austria[0.041]; Austria[0.041]; Ski[0.041]; World[0.040]; World[0.040]; World[0.040]; Norway[0.040]; Noriaki[0.040]; Hiroya[0.040]; event[0.039]; Finland[0.039]; Finland[0.039]; Finland[0.039]; points[0.039]; standings[0.039]; Standings[0.039]; \t\n\r [===============>..............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s863ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1633/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1296testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brussels ==> ENTITY: \u001b[32mBrussels\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.091:0.091[0]; log p(e|m)= -0.025:-0.025[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuxembourg[0.047]; Luxembourg[0.047]; Brussels[0.046]; Saturday[0.044]; traditional[0.043]; Market[0.043]; World[0.043]; December[0.042]; market[0.042]; market[0.042]; concert[0.042]; site[0.041]; lu[0.041]; starts[0.041]; activities[0.040]; various[0.040]; taken[0.040]; world[0.040]; wide[0.040]; gives[0.039]; runs[0.039]; retailers[0.039]; Christmas[0.038]; Christmas[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Luxembourg ==> ENTITY: \u001b[32mLuxembourg\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.077:0.077[0]; log p(e|m)= -0.499:-0.499[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuxembourg[0.050]; Brussels[0.046]; Brussels[0.046]; traditional[0.044]; wide[0.044]; taken[0.043]; December[0.043]; world[0.043]; Saturday[0.042]; World[0.042]; various[0.041]; starts[0.041]; market[0.041]; market[0.041]; Wide[0.041]; lu[0.040]; gives[0.040]; details[0.040]; Web[0.039]; site[0.039]; runs[0.039]; web[0.038]; web[0.038]; concert[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Luxembourg ==> ENTITY: \u001b[32mLuxembourg\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.078:0.078[0]; log p(e|m)= -0.499:-0.499[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuxembourg[0.049]; Brussels[0.046]; Brussels[0.046]; World[0.045]; traditional[0.044]; wide[0.043]; taken[0.043]; world[0.043]; December[0.043]; Saturday[0.042]; various[0.041]; starts[0.041]; market[0.041]; market[0.041]; Wide[0.041]; lu[0.040]; gives[0.040]; details[0.040]; Web[0.039]; site[0.039]; runs[0.038]; web[0.038]; web[0.038]; concert[0.038]; \t\n\r [===============>..............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s46ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1636/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1236testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.041:0.041[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCity[0.045]; adding[0.044]; January[0.043]; book[0.043]; York[0.043]; floating[0.042]; floating[0.042]; floating[0.042]; floating[0.042]; refunding[0.042]; refunding[0.042]; hopes[0.042]; said[0.042]; said[0.042]; Sachs[0.041]; million[0.041]; explained[0.041]; European[0.041]; selling[0.041]; Friday[0.041]; summer[0.041]; promoted[0.041]; declined[0.040]; Goldman[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = European ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.075:0.075[0]; log p(e|m)= -0.256:-0.256[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuro[0.044]; benchmark[0.043]; summer[0.042]; said[0.042]; said[0.042]; rate[0.042]; rate[0.042]; rate[0.042]; rate[0.042]; official[0.042]; official[0.042]; January[0.042]; deal[0.041]; million[0.041]; large[0.041]; offering[0.041]; tier[0.041]; June[0.041]; proposed[0.041]; sold[0.041]; second[0.041]; says[0.041]; needed[0.040]; selling[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = London Stock Exchange\u001b[39m ==> ENTITY: \u001b[32mLondon Stock Exchange\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlisted[0.046]; European[0.046]; markets[0.044]; offering[0.043]; investors[0.043]; City[0.042]; offer[0.042]; sold[0.042]; York[0.042]; benchmark[0.042]; June[0.041]; sale[0.041]; sale[0.041]; January[0.041]; January[0.041]; selling[0.041]; exempt[0.041]; bidding[0.041]; deal[0.041]; November[0.040]; establish[0.040]; said[0.040]; said[0.040]; added[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = European ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.256:-0.256[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbenchmark[0.044]; date[0.043]; specific[0.043]; rate[0.043]; official[0.043]; official[0.043]; January[0.043]; January[0.043]; potential[0.042]; deal[0.042]; bidding[0.042]; million[0.042]; offering[0.041]; offering[0.041]; June[0.041]; sold[0.041]; late[0.041]; November[0.041]; needed[0.041]; selling[0.041]; Stock[0.041]; said[0.040]; said[0.040]; promoted[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = New York City\u001b[39m ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.041:0.041[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYork[0.045]; proposed[0.045]; city[0.044]; city[0.044]; city[0.044]; refunding[0.043]; refunding[0.043]; million[0.042]; needed[0.041]; planned[0.041]; said[0.041]; said[0.041]; rate[0.041]; rate[0.041]; rate[0.041]; rate[0.041]; debt[0.041]; debt[0.041]; January[0.040]; tier[0.040]; second[0.040]; issue[0.040]; issue[0.040]; adding[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.073:0.073[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmillion[0.046]; selling[0.044]; sold[0.044]; date[0.043]; late[0.042]; European[0.042]; competitive[0.042]; deal[0.042]; City[0.042]; listed[0.042]; January[0.041]; January[0.041]; rate[0.041]; rate[0.041]; includes[0.041]; specific[0.041]; costs[0.041]; build[0.041]; June[0.041]; exempt[0.041]; pool[0.041]; book[0.041]; November[0.040]; York[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = New York City\u001b[39m ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.054:0.054[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nproposed[0.045]; build[0.044]; city[0.044]; city[0.044]; city[0.044]; refunding[0.043]; costs[0.043]; includes[0.042]; tax[0.042]; London[0.042]; late[0.041]; potential[0.041]; million[0.041]; needed[0.041]; planned[0.041]; said[0.040]; said[0.040]; rate[0.040]; rate[0.040]; Exchange[0.040]; debt[0.040]; debt[0.040]; January[0.040]; January[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NYC ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.081:0.081[0]; log p(e|m)= -0.386:-0.386[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCity[0.045]; York[0.044]; York[0.044]; named[0.044]; offering[0.043]; planned[0.043]; Friday[0.043]; proposed[0.043]; city[0.042]; city[0.042]; second[0.042]; include[0.042]; million[0.042]; selling[0.041]; said[0.041]; declined[0.041]; tier[0.041]; January[0.040]; debt[0.040]; debt[0.040]; summer[0.040]; promoted[0.040]; official[0.040]; Jan[0.040]; \t\n\r [===============>..............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s42ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1644/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1203testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spanish ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.137:0.137[0]; log p(e|m)= -1.241:-1.241[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpanish[0.053]; country[0.046]; Soccer[0.045]; Santiago[0.044]; Madrid[0.043]; Madrid[0.043]; Madrid[0.043]; game[0.043]; game[0.043]; sports[0.043]; Barcelona[0.042]; Mundo[0.042]; stadium[0.041]; Real[0.041]; Bernabeu[0.040]; fans[0.040]; governing[0.039]; said[0.039]; El[0.038]; Saturday[0.038]; events[0.038]; legislation[0.038]; Friday[0.038]; precedents[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Madrid ==> ENTITY: \u001b[32mMadrid\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMadrid[0.047]; Madrid[0.047]; stadium[0.046]; Santiago[0.045]; events[0.044]; Spanish[0.044]; Spanish[0.044]; Mundo[0.043]; country[0.042]; sports[0.042]; Barcelona[0.042]; Soccer[0.042]; supporters[0.040]; supporters[0.040]; Friday[0.040]; governing[0.040]; daily[0.040]; El[0.040]; said[0.040]; gates[0.039]; action[0.039]; game[0.038]; game[0.038]; Saturday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Santiago Bernabeu stadium\u001b[39m ==> ENTITY: \u001b[32mSantiago Bernabéu Stadium\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMadrid[0.046]; Madrid[0.046]; Madrid[0.046]; Soccer[0.044]; game[0.044]; game[0.044]; Barcelona[0.043]; Spanish[0.043]; Spanish[0.043]; fans[0.043]; Real[0.042]; Showcase[0.041]; sports[0.041]; supporters[0.041]; supporters[0.041]; Saturday[0.041]; Friday[0.040]; big[0.040]; El[0.039]; Fans[0.039]; Says[0.039]; Tickets[0.039]; events[0.039]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spanish ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.133:0.133[0]; log p(e|m)= -1.241:-1.241[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpanish[0.053]; country[0.046]; Soccer[0.046]; Santiago[0.044]; Madrid[0.043]; Madrid[0.043]; Madrid[0.043]; game[0.043]; game[0.043]; Barcelona[0.042]; Mundo[0.042]; sports[0.042]; stadium[0.041]; Real[0.041]; Bernabeu[0.040]; fans[0.040]; governing[0.039]; El[0.039]; said[0.039]; Saturday[0.038]; events[0.038]; legislation[0.038]; precedents[0.038]; police[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = El Mundo\u001b[39m ==> ENTITY: \u001b[32mEl Mundo (Spain)\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.107:0.107[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMadrid[0.045]; Madrid[0.045]; Madrid[0.045]; Spanish[0.045]; Spanish[0.045]; Santiago[0.044]; Real[0.044]; Friday[0.043]; country[0.042]; events[0.042]; daily[0.042]; sports[0.041]; barred[0.041]; Barcelona[0.041]; big[0.040]; Face[0.040]; action[0.040]; said[0.040]; Showcase[0.040]; Paper[0.039]; supporters[0.039]; supporters[0.039]; stadium[0.039]; governing[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Madrid ==> ENTITY: \u001b[32mMadrid\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.135:0.135[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMadrid[0.049]; Madrid[0.049]; Santiago[0.047]; Spanish[0.046]; Spanish[0.046]; events[0.046]; stadium[0.045]; Mundo[0.045]; country[0.044]; Barcelona[0.044]; sports[0.043]; Soccer[0.043]; supporters[0.042]; supporters[0.042]; Friday[0.042]; governing[0.042]; daily[0.042]; El[0.042]; said[0.042]; gates[0.041]; action[0.040]; Saturday[0.040]; Real[0.040]; \t\n\r [===============>..............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s58ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1650/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1326testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Euro-sceptic\u001b[39m ==> ENTITY: \u001b[32mEuroscepticism\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuropean[0.048]; Conservative[0.047]; Conservatives[0.046]; British[0.046]; British[0.046]; Parliament[0.043]; Member[0.042]; member[0.042]; Euro[0.042]; party[0.042]; party[0.042]; finance[0.041]; chancellor[0.040]; election[0.040]; issue[0.040]; Tony[0.040]; general[0.040]; Cabinet[0.040]; minister[0.039]; currency[0.039]; says[0.039]; said[0.039]; said[0.039]; ruling[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Michael Heseltine\u001b[39m ==> ENTITY: \u001b[32mMichael Heseltine\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConservative[0.045]; John[0.044]; Clarke[0.043]; Clarke[0.043]; Clarke[0.043]; Clarke[0.043]; Parliament[0.043]; Minister[0.042]; said[0.041]; Prime[0.041]; denied[0.041]; resign[0.041]; resign[0.041]; ministers[0.041]; policy[0.041]; minister[0.041]; prime[0.041]; statement[0.041]; resigned[0.041]; Thursday[0.041]; Thursday[0.041]; government[0.041]; told[0.041]; told[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Newsnight\u001b[39m ==> ENTITY: \u001b[32mNewsnight\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.046]; British[0.046]; television[0.044]; Bbc[0.044]; Conservative[0.044]; London[0.043]; John[0.042]; Conservatives[0.042]; told[0.042]; told[0.042]; Clarke[0.041]; Clarke[0.041]; Clarke[0.041]; Clarke[0.041]; Clarke[0.041]; Thursday[0.041]; Thursday[0.041]; Thursday[0.041]; Thursday[0.041]; prime[0.040]; said[0.040]; said[0.040]; said[0.040]; says[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = European ==> ENTITY: \u001b[32mEuropean Union\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.109:0.109[0]; log p(e|m)= -2.937:-2.937[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuro[0.044]; Euro[0.044]; British[0.044]; British[0.044]; election[0.044]; said[0.042]; said[0.042]; party[0.041]; party[0.041]; place[0.041]; government[0.041]; make[0.041]; Member[0.041]; changed[0.041]; resignation[0.041]; Parliament[0.041]; says[0.041]; ruling[0.041]; resign[0.040]; resign[0.040]; resign[0.040]; resign[0.040]; currency[0.040]; member[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Conservative ==> ENTITY: \u001b[32mConservative Party (UK)\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.176:0.176[0]; log p(e|m)= -0.559:-0.559[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConservatives[0.046]; British[0.044]; British[0.044]; Tony[0.044]; party[0.044]; party[0.044]; Member[0.043]; Parliament[0.042]; election[0.042]; member[0.042]; Cabinet[0.041]; Clarke[0.041]; Clarke[0.041]; Clarke[0.041]; London[0.041]; split[0.041]; chancellor[0.041]; minister[0.040]; minister[0.040]; general[0.040]; Marlow[0.040]; Marlow[0.040]; Thursday[0.039]; Thursday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = John Major\u001b[39m ==> ENTITY: \u001b[32mJohn Major\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConservative[0.048]; Major[0.042]; Clarke[0.042]; Clarke[0.042]; Clarke[0.042]; Clarke[0.042]; Clarke[0.042]; parliament[0.042]; Parliament[0.042]; resigned[0.041]; ministers[0.041]; prime[0.041]; prime[0.041]; Prime[0.041]; position[0.041]; television[0.041]; Newsnight[0.041]; Heseltine[0.041]; policy[0.041]; Michael[0.041]; minister[0.041]; minister[0.041]; government[0.041]; speculation[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Clarke\u001b[39m ==> ENTITY: \u001b[32mKenneth Clarke\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConservative[0.047]; Clarke[0.047]; Clarke[0.047]; John[0.045]; Michael[0.044]; resigned[0.042]; position[0.042]; Parliament[0.042]; deputy[0.042]; ministers[0.041]; resign[0.041]; Heseltine[0.041]; fellow[0.041]; Thursday[0.040]; Thursday[0.040]; parliament[0.040]; minister[0.040]; Currie[0.040]; Minister[0.040]; Prime[0.040]; statement[0.040]; denied[0.039]; said[0.039]; declared[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marlow\u001b[39m ==> ENTITY: \u001b[32mAntony Marlow\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nParliament[0.044]; Parliament[0.044]; election[0.044]; Conservatives[0.043]; Marlow[0.042]; Marlow[0.042]; parliament[0.042]; Prime[0.042]; Cabinet[0.042]; British[0.042]; Newsnight[0.041]; resign[0.041]; resign[0.041]; resign[0.041]; resign[0.041]; minister[0.041]; minister[0.041]; John[0.041]; chancellor[0.041]; Minister[0.041]; party[0.040]; electable[0.040]; Major[0.040]; Major[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCabinet\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mCabinet of the United Kingdom\u001b[39m <---> \u001b[32mCabinet (government)\u001b[39m\t\nSCORES: global= 0.241:0.239[\u001b[31m0.002\u001b[39m]; local()= 0.178:0.159[\u001b[31m0.020\u001b[39m]; log p(e|m)= -2.513:-2.104[\u001b[32m0.410\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConservative[0.044]; Parliament[0.044]; Parliament[0.044]; British[0.043]; British[0.043]; minister[0.042]; minister[0.042]; Conservatives[0.042]; government[0.042]; resign[0.041]; resign[0.041]; resign[0.041]; resign[0.041]; resign[0.041]; Member[0.041]; position[0.041]; general[0.041]; member[0.041]; Clarke[0.041]; Clarke[0.041]; Clarke[0.041]; Clarke[0.041]; Clarke[0.041]; policy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = British ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.106:0.106[0]; log p(e|m)= -0.489:-0.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.053]; London[0.047]; Parliament[0.046]; Tony[0.045]; television[0.045]; European[0.045]; Conservative[0.044]; Clarke[0.044]; Clarke[0.044]; Clarke[0.044]; Newsnight[0.043]; Kenneth[0.043]; Kenneth[0.043]; member[0.043]; Bbc[0.042]; prime[0.042]; says[0.042]; minister[0.042]; minister[0.042]; general[0.041]; single[0.041]; Conservatives[0.041]; Cabinet[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Conservative ==> ENTITY: \u001b[32mConservative Party (UK)\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.142:0.142[0]; log p(e|m)= -0.559:-0.559[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJohn[0.046]; Currie[0.045]; Parliament[0.045]; Heseltine[0.045]; government[0.045]; Clarke[0.044]; Clarke[0.044]; Clarke[0.044]; Michael[0.043]; parliament[0.043]; position[0.043]; resigned[0.043]; minister[0.043]; victory[0.043]; membership[0.043]; Minister[0.043]; ministers[0.043]; deputy[0.043]; Mp[0.043]; pro[0.042]; Thursday[0.042]; Thursday[0.042]; policy[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Clarke\u001b[39m ==> ENTITY: \u001b[32mKenneth Clarke\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.217:0.217[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConservative[0.048]; Clarke[0.048]; Clarke[0.048]; British[0.047]; British[0.047]; Conservatives[0.046]; Cabinet[0.045]; Newsnight[0.044]; chancellor[0.043]; London[0.043]; Parliament[0.042]; member[0.042]; Member[0.042]; finance[0.042]; election[0.042]; Tony[0.042]; resign[0.041]; resign[0.041]; Kenneth[0.041]; Kenneth[0.041]; resignation[0.041]; party[0.041]; party[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Major\u001b[39m ==> ENTITY: \u001b[32mJohn Major\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConservative[0.048]; Major[0.042]; John[0.042]; Clarke[0.042]; Clarke[0.042]; Clarke[0.042]; Clarke[0.042]; Clarke[0.042]; parliament[0.042]; Parliament[0.042]; resigned[0.041]; ministers[0.041]; prime[0.041]; prime[0.041]; Prime[0.041]; position[0.041]; television[0.041]; Newsnight[0.041]; Heseltine[0.041]; policy[0.041]; Michael[0.041]; minister[0.041]; minister[0.041]; government[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = BBC ==> ENTITY: \u001b[32mBBC\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.095:0.095[0]; log p(e|m)= -0.014:-0.014[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntelevision[0.053]; British[0.051]; British[0.051]; Newsnight[0.049]; London[0.049]; told[0.048]; told[0.048]; Conservative[0.048]; member[0.048]; chancellor[0.047]; said[0.047]; said[0.047]; said[0.047]; resignation[0.046]; Conservatives[0.046]; Member[0.046]; resign[0.046]; resign[0.046]; resign[0.046]; resign[0.046]; resign[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kenneth Clarke\u001b[39m ==> ENTITY: \u001b[32mKenneth Clarke\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConservative[0.048]; Clarke[0.048]; Clarke[0.048]; British[0.047]; British[0.047]; Conservatives[0.046]; Cabinet[0.045]; Newsnight[0.044]; chancellor[0.043]; London[0.043]; Parliament[0.042]; member[0.042]; Member[0.042]; finance[0.042]; election[0.042]; Tony[0.042]; resign[0.041]; resign[0.041]; resign[0.041]; Kenneth[0.041]; resignation[0.041]; party[0.041]; party[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Clarke\u001b[39m ==> ENTITY: \u001b[32mKenneth Clarke\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.213:0.213[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClarke[0.046]; Clarke[0.046]; Clarke[0.046]; Conservatives[0.044]; John[0.044]; Cabinet[0.043]; Michael[0.043]; Newsnight[0.043]; chancellor[0.042]; position[0.041]; policy[0.041]; Parliament[0.041]; Parliament[0.041]; deputy[0.041]; Member[0.040]; election[0.040]; Tony[0.040]; resign[0.040]; resign[0.040]; resign[0.040]; Kenneth[0.040]; Heseltine[0.040]; fellow[0.040]; resignation[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Edwina Currie\u001b[39m ==> ENTITY: \u001b[32mEdwina Currie\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConservative[0.047]; Bbc[0.044]; Clarke[0.043]; Clarke[0.043]; Clarke[0.043]; Michael[0.043]; John[0.043]; resigned[0.042]; Thursday[0.042]; Thursday[0.042]; statement[0.041]; Parliament[0.041]; denied[0.041]; prime[0.041]; told[0.041]; told[0.041]; Heseltine[0.041]; position[0.041]; resign[0.041]; declared[0.040]; Pro[0.040]; minister[0.040]; said[0.040]; Minister[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kenneth Clarke\u001b[39m ==> ENTITY: \u001b[32mKenneth Clarke\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.224:0.224[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConservative[0.052]; Clarke[0.051]; Clarke[0.051]; Clarke[0.051]; Clarke[0.051]; British[0.050]; Conservatives[0.050]; John[0.049]; Cabinet[0.048]; Newsnight[0.048]; chancellor[0.046]; position[0.046]; London[0.046]; policy[0.046]; Parliament[0.045]; Parliament[0.045]; member[0.045]; Member[0.045]; finance[0.045]; election[0.045]; Tony[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tony Marlow\u001b[39m ==> ENTITY: \u001b[32mAntony Marlow\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConservative[0.046]; Parliament[0.044]; election[0.044]; Conservatives[0.043]; Marlow[0.042]; Marlow[0.042]; London[0.042]; Cabinet[0.042]; British[0.042]; British[0.042]; Newsnight[0.041]; resign[0.041]; resign[0.041]; resign[0.041]; resign[0.041]; minister[0.041]; minister[0.041]; chancellor[0.041]; party[0.040]; party[0.040]; electable[0.040]; government[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = British ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.489:-0.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.051]; London[0.045]; Parliament[0.045]; Tony[0.044]; television[0.043]; European[0.043]; Conservative[0.043]; Clarke[0.042]; Clarke[0.042]; Newsnight[0.041]; Kenneth[0.041]; member[0.041]; Bbc[0.041]; says[0.040]; minister[0.040]; general[0.040]; single[0.040]; Conservatives[0.040]; Cabinet[0.040]; Marlow[0.039]; Marlow[0.039]; place[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.080:0.080[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMember[0.047]; British[0.047]; British[0.047]; Conservative[0.047]; member[0.045]; Parliament[0.044]; single[0.044]; Conservatives[0.044]; television[0.043]; said[0.043]; said[0.043]; European[0.043]; chancellor[0.043]; Thursday[0.042]; Thursday[0.042]; Cabinet[0.042]; election[0.042]; place[0.042]; told[0.042]; Tony[0.042]; Clarke[0.041]; Clarke[0.041]; Clarke[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Clarke\u001b[39m ==> ENTITY: \u001b[32mKenneth Clarke\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConservative[0.047]; Clarke[0.047]; Clarke[0.047]; Clarke[0.047]; John[0.045]; Michael[0.043]; resigned[0.042]; position[0.042]; policy[0.041]; Parliament[0.041]; deputy[0.041]; ministers[0.040]; resign[0.040]; resign[0.040]; Heseltine[0.040]; fellow[0.040]; Thursday[0.040]; Thursday[0.040]; parliament[0.040]; minister[0.040]; minister[0.040]; Currie[0.039]; Minister[0.039]; Prime[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Clarke\u001b[39m ==> ENTITY: \u001b[32mKenneth Clarke\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConservative[0.048]; Clarke[0.047]; Clarke[0.047]; Clarke[0.047]; Clarke[0.047]; Conservatives[0.046]; John[0.045]; Cabinet[0.045]; Michael[0.044]; Newsnight[0.044]; resigned[0.042]; position[0.042]; policy[0.042]; Parliament[0.042]; deputy[0.042]; election[0.041]; ministers[0.041]; resign[0.041]; resign[0.041]; resign[0.041]; Kenneth[0.041]; Heseltine[0.041]; fellow[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marlow\u001b[39m ==> ENTITY: \u001b[32mAntony Marlow\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConservative[0.047]; Parliament[0.046]; Parliament[0.046]; election[0.045]; Conservatives[0.044]; Marlow[0.044]; Marlow[0.044]; London[0.044]; Prime[0.043]; Cabinet[0.043]; British[0.043]; British[0.043]; Newsnight[0.043]; resign[0.043]; resign[0.043]; resign[0.043]; resign[0.043]; resign[0.043]; minister[0.043]; minister[0.043]; John[0.042]; chancellor[0.042]; Minister[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = BBC ==> ENTITY: \u001b[32mBBC\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.081:0.081[0]; log p(e|m)= -0.014:-0.014[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nresigned[0.044]; told[0.044]; told[0.044]; Conservative[0.043]; statement[0.043]; said[0.043]; Michael[0.042]; resign[0.042]; Clarke[0.042]; Clarke[0.042]; Clarke[0.042]; Thursday[0.042]; Thursday[0.042]; position[0.041]; denied[0.041]; prime[0.041]; tune[0.041]; single[0.040]; single[0.040]; single[0.040]; policy[0.040]; John[0.040]; European[0.040]; European[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Euro-sceptic\u001b[39m ==> ENTITY: \u001b[32mEuroscepticism\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuropean[0.048]; Conservative[0.047]; Conservatives[0.046]; British[0.046]; British[0.046]; Parliament[0.043]; Member[0.042]; member[0.042]; Euro[0.042]; party[0.042]; party[0.042]; finance[0.041]; chancellor[0.040]; election[0.040]; issue[0.040]; Tony[0.040]; general[0.040]; Cabinet[0.040]; does[0.039]; minister[0.039]; currency[0.039]; says[0.039]; said[0.039]; said[0.039]; \t\n\r [================>.............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s884ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1676/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1371testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Arsenal ==> ENTITY: \u001b[32mArsenal F.C.\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.182:0.182[0]; log p(e|m)= -0.158:-0.158[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; league[0.047]; club[0.046]; Arsenal[0.045]; Arsenal[0.045]; Soccer[0.044]; Liverpool[0.043]; season[0.043]; goal[0.042]; goal[0.042]; draw[0.042]; minute[0.041]; Tony[0.041]; minutes[0.040]; suffered[0.040]; Derby[0.040]; defeat[0.039]; Saturday[0.039]; Vieira[0.039]; Vieira[0.039]; English[0.039]; home[0.039]; home[0.039]; captain[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Derby ==> ENTITY: \u001b[32mDerby County F.C.\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.189:0.189[0]; log p(e|m)= -3.689:-3.689[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.045]; league[0.045]; league[0.045]; club[0.045]; Sheffield[0.045]; Sunderland[0.044]; season[0.044]; games[0.043]; Soccer[0.042]; second[0.042]; place[0.041]; Liverpool[0.041]; Liverpool[0.041]; draw[0.041]; unbeaten[0.040]; London[0.040]; London[0.040]; run[0.040]; English[0.040]; winning[0.040]; Saturday[0.039]; home[0.039]; home[0.039]; defeat[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wimbledon ==> ENTITY: \u001b[32mWimbledon F.C.\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.190:0.190[0]; log p(e|m)= -1.890:-1.890[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWimbledon[0.048]; league[0.045]; league[0.045]; Sheffield[0.045]; Sunderland[0.044]; season[0.043]; Ekoku[0.043]; upset[0.042]; second[0.042]; Liverpool[0.042]; Liverpool[0.042]; victory[0.042]; games[0.041]; secured[0.040]; Tony[0.040]; winning[0.040]; late[0.040]; home[0.040]; defeat[0.039]; Adams[0.039]; Arsenal[0.039]; Arsenal[0.039]; cup[0.039]; suffered[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wimbledon ==> ENTITY: \u001b[32mWimbledon F.C.\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.201:0.201[0]; log p(e|m)= -1.890:-1.890[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWimbledon[0.049]; club[0.047]; league[0.046]; league[0.046]; league[0.046]; Sheffield[0.046]; Sunderland[0.045]; London[0.045]; season[0.044]; Ekoku[0.044]; upset[0.043]; second[0.043]; draw[0.043]; Liverpool[0.043]; Liverpool[0.043]; games[0.042]; Derby[0.042]; victory[0.041]; secured[0.041]; Tony[0.041]; winning[0.041]; English[0.041]; late[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Liverpool ==> ENTITY: \u001b[32mLiverpool F.C.\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.190:0.190[0]; log p(e|m)= -1.171:-1.171[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.046]; league[0.046]; league[0.046]; league[0.046]; Ekoku[0.045]; Liverpool[0.045]; Sunderland[0.044]; Sheffield[0.044]; Soccer[0.044]; London[0.044]; London[0.044]; second[0.043]; season[0.043]; games[0.043]; Arsenal[0.042]; Arsenal[0.042]; Arsenal[0.042]; Arsenal[0.042]; secured[0.042]; draw[0.042]; defeat[0.042]; half[0.042]; cup[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.095:0.095[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.046]; club[0.044]; league[0.044]; league[0.044]; league[0.044]; games[0.044]; Ekoku[0.043]; Sheffield[0.043]; Sunderland[0.042]; Soccer[0.042]; season[0.041]; Liverpool[0.041]; Liverpool[0.041]; draw[0.041]; Arsenal[0.041]; Arsenal[0.041]; Arsenal[0.041]; second[0.040]; Adams[0.040]; home[0.040]; home[0.040]; Guy[0.039]; Wednesday[0.039]; Derby[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = English ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.105:0.105[0]; log p(e|m)= -1.355:-1.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.049]; league[0.049]; club[0.046]; Sheffield[0.046]; second[0.045]; Sunderland[0.044]; captain[0.044]; Adams[0.044]; season[0.044]; London[0.044]; London[0.044]; Soccer[0.043]; Tony[0.043]; Dean[0.043]; Guy[0.042]; place[0.042]; Saturday[0.042]; forwards[0.042]; winning[0.041]; Liverpool[0.041]; Liverpool[0.041]; Wednesday[0.041]; Patrick[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sunderland ==> ENTITY: \u001b[32mSunderland A.F.C.\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.396:-0.396[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.045]; league[0.045]; league[0.045]; Sheffield[0.044]; second[0.043]; club[0.043]; Ekoku[0.043]; games[0.042]; season[0.042]; goal[0.042]; goal[0.042]; goal[0.042]; Liverpool[0.041]; Liverpool[0.041]; half[0.041]; draw[0.041]; minutes[0.040]; cup[0.040]; Arsenal[0.040]; Arsenal[0.040]; Arsenal[0.040]; late[0.040]; secured[0.040]; forward[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Guy Whittingham\u001b[39m ==> ENTITY: \u001b[32mGuy Whittingham\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSunderland[0.045]; second[0.044]; season[0.044]; club[0.043]; league[0.043]; league[0.043]; league[0.043]; games[0.043]; Sheffield[0.043]; Wednesday[0.042]; suffered[0.041]; Saturday[0.041]; Ekoku[0.041]; late[0.041]; English[0.041]; captain[0.041]; secured[0.041]; Liverpool[0.041]; Liverpool[0.041]; Derby[0.040]; draw[0.040]; half[0.040]; defeat[0.040]; premier[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dean Sturridge\u001b[39m ==> ENTITY: \u001b[32mDean Sturridge\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; league[0.044]; league[0.044]; season[0.044]; games[0.043]; club[0.043]; Sheffield[0.043]; second[0.042]; Sunderland[0.042]; home[0.041]; home[0.041]; Liverpool[0.041]; Liverpool[0.041]; Ekoku[0.041]; Saturday[0.041]; Minute[0.041]; late[0.041]; Wednesday[0.041]; Derby[0.041]; Dean[0.040]; goal[0.040]; goal[0.040]; goal[0.040]; goal[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Vieira\u001b[39m ==> ENTITY: \u001b[32mPatrick Vieira\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArsenal[0.045]; Arsenal[0.045]; Arsenal[0.045]; club[0.044]; league[0.044]; league[0.044]; Vieira[0.044]; Soccer[0.044]; Frenchman[0.043]; Liverpool[0.041]; season[0.041]; goal[0.041]; goal[0.041]; draw[0.041]; captain[0.040]; English[0.040]; Equaliser[0.040]; minute[0.040]; upset[0.040]; minutes[0.040]; halftime[0.039]; Sturridge[0.039]; premier[0.039]; Tony[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Frenchman ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.057:0.057[0]; log p(e|m)= -0.719:-0.719[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; league[0.044]; Arsenal[0.043]; Arsenal[0.043]; Arsenal[0.043]; Soccer[0.042]; minute[0.042]; minutes[0.042]; club[0.042]; season[0.042]; goal[0.042]; goal[0.042]; goal[0.042]; Minute[0.042]; Patrick[0.041]; captain[0.041]; forwards[0.041]; English[0.041]; Derby[0.041]; blasted[0.040]; Guy[0.040]; Vieira[0.040]; Vieira[0.040]; Saturday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sheffield Wednesday\u001b[39m ==> ENTITY: \u001b[32mSheffield Wednesday F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; league[0.047]; league[0.047]; club[0.043]; Sunderland[0.043]; second[0.043]; Ekoku[0.043]; games[0.043]; season[0.042]; Saturday[0.042]; half[0.041]; cup[0.041]; English[0.040]; goal[0.040]; goal[0.040]; goal[0.040]; goal[0.040]; draw[0.040]; secured[0.040]; Liverpool[0.040]; Liverpool[0.040]; place[0.040]; run[0.039]; forward[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Darryl Powell\u001b[39m ==> ENTITY: \u001b[32mDarryl Powell\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.045]; league[0.045]; league[0.045]; Sunderland[0.045]; club[0.044]; Soccer[0.043]; games[0.043]; Ekoku[0.042]; Sheffield[0.042]; Saturday[0.041]; draw[0.041]; Derby[0.041]; London[0.041]; London[0.041]; English[0.041]; Guy[0.040]; season[0.040]; Wednesday[0.040]; home[0.040]; home[0.040]; premier[0.040]; Tony[0.040]; Efan[0.040]; Equaliser[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Arsenal ==> ENTITY: \u001b[32mArsenal F.C.\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.188:0.188[0]; log p(e|m)= -0.158:-0.158[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; league[0.046]; league[0.046]; club[0.045]; Arsenal[0.044]; Arsenal[0.044]; Sheffield[0.042]; Soccer[0.042]; Sunderland[0.042]; Liverpool[0.042]; Liverpool[0.042]; second[0.041]; season[0.041]; goal[0.040]; goal[0.040]; goal[0.040]; Tony[0.040]; draw[0.040]; minute[0.040]; winning[0.040]; captain[0.039]; minutes[0.039]; cup[0.039]; suffered[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dean Holdsworth\u001b[39m ==> ENTITY: \u001b[32mDean Holdsworth\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEkoku[0.046]; league[0.044]; league[0.044]; Adams[0.044]; second[0.043]; Sheffield[0.043]; season[0.042]; Sunderland[0.042]; games[0.042]; Wednesday[0.041]; Dean[0.041]; Tony[0.041]; forward[0.041]; Darryl[0.041]; Powell[0.041]; goal[0.040]; goal[0.040]; Wimbledon[0.040]; Wimbledon[0.040]; place[0.040]; Guy[0.040]; secured[0.040]; late[0.040]; captain[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Liverpool ==> ENTITY: \u001b[32mLiverpool F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.180:0.180[0]; log p(e|m)= -1.171:-1.171[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.045]; league[0.044]; league[0.044]; league[0.044]; Ekoku[0.044]; Liverpool[0.043]; Sunderland[0.043]; Sheffield[0.043]; London[0.042]; second[0.042]; season[0.042]; games[0.042]; Arsenal[0.041]; Arsenal[0.041]; Arsenal[0.041]; secured[0.041]; draw[0.041]; defeat[0.040]; half[0.040]; cup[0.040]; Dean[0.040]; Dean[0.040]; English[0.040]; unbeaten[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.050]; club[0.047]; league[0.047]; league[0.047]; Sheffield[0.046]; Soccer[0.045]; season[0.044]; Liverpool[0.044]; draw[0.044]; Arsenal[0.043]; Arsenal[0.043]; Arsenal[0.043]; Adams[0.042]; home[0.042]; home[0.042]; Guy[0.042]; Wednesday[0.042]; Derby[0.042]; Patrick[0.041]; defeat[0.041]; English[0.041]; beaten[0.041]; Tony[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Arsenal ==> ENTITY: \u001b[32mArsenal F.C.\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.196:0.196[0]; log p(e|m)= -0.158:-0.158[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.045]; league[0.045]; league[0.045]; club[0.044]; Arsenal[0.043]; Arsenal[0.043]; games[0.043]; Ekoku[0.042]; Sheffield[0.042]; Soccer[0.042]; Liverpool[0.041]; Liverpool[0.041]; Sunderland[0.041]; second[0.041]; season[0.041]; goal[0.040]; goal[0.040]; goal[0.040]; goal[0.040]; draw[0.040]; forward[0.040]; minute[0.040]; secured[0.040]; winning[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Arsenal ==> ENTITY: \u001b[32mArsenal F.C.\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.179:0.179[0]; log p(e|m)= -0.158:-0.158[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; league[0.047]; games[0.044]; Ekoku[0.044]; Sheffield[0.043]; Liverpool[0.043]; Liverpool[0.043]; Sunderland[0.042]; second[0.042]; season[0.042]; goal[0.041]; goal[0.041]; forward[0.041]; secured[0.041]; winning[0.041]; Tony[0.041]; half[0.040]; cup[0.040]; suffered[0.040]; unbeaten[0.039]; place[0.039]; defeat[0.039]; Wednesday[0.039]; late[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Patrick Vieira\u001b[39m ==> ENTITY: \u001b[32mPatrick Vieira\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArsenal[0.045]; Arsenal[0.045]; Arsenal[0.045]; club[0.044]; league[0.044]; league[0.044]; Vieira[0.043]; Soccer[0.043]; Frenchman[0.043]; Liverpool[0.041]; season[0.041]; goal[0.041]; goal[0.041]; goal[0.041]; draw[0.041]; captain[0.040]; English[0.040]; Equaliser[0.040]; minute[0.040]; Sheffield[0.040]; upset[0.040]; minutes[0.040]; Wednesday[0.039]; halftime[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Efan Ekoku\u001b[39m ==> ENTITY: \u001b[32mEfan Ekoku\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.048]; league[0.048]; Sheffield[0.045]; games[0.045]; season[0.044]; Sunderland[0.044]; club[0.043]; Dean[0.042]; Dean[0.042]; secured[0.042]; Liverpool[0.041]; Liverpool[0.041]; second[0.041]; Tony[0.040]; London[0.040]; Wimbledon[0.040]; Wimbledon[0.040]; Holdsworth[0.040]; points[0.040]; half[0.039]; cup[0.039]; Adams[0.039]; captain[0.039]; winning[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tony Adams\u001b[39m ==> ENTITY: \u001b[32mTony Adams (footballer)\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGuy[0.048]; Darryl[0.047]; Patrick[0.046]; Dean[0.045]; Dean[0.045]; league[0.044]; league[0.044]; league[0.044]; Holdsworth[0.042]; club[0.042]; Ekoku[0.041]; Sunderland[0.041]; Powell[0.041]; season[0.040]; Sheffield[0.040]; games[0.040]; Whittingham[0.040]; winning[0.039]; Saturday[0.039]; Liverpool[0.039]; Liverpool[0.039]; captain[0.038]; fellow[0.038]; Soccer[0.038]; \t\n\r [================>.............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s725ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1699/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1188testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Eagles ==> ENTITY: \u001b[32mPhiladelphia Eagles\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.180:0.180[0]; log p(e|m)= -2.477:-2.477[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.043]; Eagles[0.043]; Eagles[0.043]; interception[0.043]; Jim[0.043]; games[0.043]; yards[0.042]; yards[0.042]; yards[0.042]; playoff[0.042]; quarterback[0.042]; Colts[0.042]; Colts[0.042]; Colts[0.042]; Football[0.042]; fourth[0.041]; Justin[0.041]; Faulk[0.041]; touchdowns[0.041]; rushed[0.040]; Jason[0.040]; mark[0.040]; returned[0.040]; starting[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kerwin Bell\u001b[39m ==> ENTITY: \u001b[32mKerwin Bell\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDetmer[0.045]; Td[0.044]; game[0.044]; game[0.044]; quarterback[0.044]; Ricky[0.042]; Ty[0.042]; linebacker[0.042]; Packers[0.042]; defensive[0.042]; second[0.041]; playoff[0.041]; Tony[0.041]; Miami[0.041]; cornerback[0.041]; Eagles[0.041]; draft[0.041]; Faulk[0.041]; Faulk[0.041]; return[0.041]; Dolphins[0.041]; tackle[0.040]; seven[0.040]; Troy[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPhiladelphia\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPhiladelphia Eagles\u001b[39m <---> \u001b[32mPhiladelphia\u001b[39m\t\nSCORES: global= 0.252:0.237[\u001b[31m0.015\u001b[39m]; local()= 0.223:0.066[\u001b[31m0.157\u001b[39m]; log p(e|m)= -4.510:-0.211[\u001b[32m4.299\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.045]; games[0.045]; games[0.045]; playoff[0.043]; season[0.043]; Eagles[0.042]; Eagles[0.042]; game[0.042]; Cowboys[0.042]; Jets[0.041]; yard[0.041]; yard[0.041]; touchdown[0.041]; touchdown[0.041]; Redskins[0.041]; Cincinnati[0.041]; Washington[0.041]; linebacker[0.041]; Tony[0.041]; second[0.040]; fourth[0.040]; defensive[0.040]; played[0.040]; Dallas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Miami Dolphins ==> ENTITY: \u001b[32mMiami Dolphins\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngame[0.045]; game[0.045]; playoff[0.045]; second[0.042]; Colts[0.042]; quarterback[0.042]; left[0.042]; left[0.042]; final[0.042]; clinch[0.042]; straight[0.042]; tackle[0.041]; seven[0.041]; seven[0.041]; Ricky[0.041]; defensive[0.041]; cornerback[0.041]; Harrison[0.041]; Faulk[0.041]; Faulk[0.041]; starting[0.041]; linebacker[0.040]; leads[0.040]; return[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mKansas City\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mKansas City Chiefs\u001b[39m <---> \u001b[32mKansas City, Missouri\u001b[39m\t\nSCORES: global= 0.256:0.235[\u001b[31m0.021\u001b[39m]; local()= 0.222:0.087[\u001b[31m0.135\u001b[39m]; log p(e|m)= -3.170:-0.805[\u001b[32m2.365\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.046]; game[0.044]; games[0.044]; games[0.044]; Faulk[0.044]; Faulk[0.044]; starting[0.044]; starting[0.044]; Ray[0.044]; Jets[0.043]; tackle[0.043]; fourth[0.043]; Cowboys[0.043]; Philadelphia[0.043]; Philadelphia[0.043]; Philadelphia[0.043]; yard[0.043]; defensive[0.043]; linebacker[0.043]; Cincinnati[0.043]; Dallas[0.043]; cornerback[0.043]; touchdowns[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Eagles ==> ENTITY: \u001b[32mPhiladelphia Eagles\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.214:0.214[0]; log p(e|m)= -2.477:-2.477[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntouchdown[0.049]; touchdown[0.049]; Philadelphia[0.048]; Philadelphia[0.048]; Philadelphia[0.048]; Jets[0.048]; Cowboys[0.048]; Eagles[0.048]; Eagles[0.048]; Eagles[0.048]; interception[0.048]; season[0.048]; game[0.047]; Jim[0.047]; games[0.047]; games[0.047]; yards[0.047]; yards[0.047]; yards[0.047]; playoff[0.047]; quarterback[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Justin\u001b[39m ==> ENTITY: \u001b[32mPaul Justin\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngame[0.044]; quarterback[0.044]; Td[0.043]; Justin[0.043]; Redskins[0.043]; starting[0.042]; Jets[0.042]; Cowboys[0.042]; Harrison[0.042]; Faulk[0.042]; Faulk[0.042]; second[0.041]; Packers[0.041]; defensive[0.041]; Troy[0.041]; Philadelphia[0.041]; Philadelphia[0.041]; cornerback[0.041]; Arizona[0.041]; sprained[0.041]; Eagles[0.041]; draft[0.041]; linebacker[0.041]; Dolphins[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marshall Faulk\u001b[39m ==> ENTITY: \u001b[32mMarshall Faulk\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFootball[0.045]; fourth[0.045]; games[0.045]; games[0.045]; season[0.044]; touchdowns[0.044]; game[0.044]; quarterback[0.044]; Cincinnati[0.044]; Colts[0.043]; Colts[0.043]; Colts[0.043]; Colts[0.043]; Afc[0.043]; Jason[0.043]; injury[0.043]; Eagles[0.043]; Eagles[0.043]; Eagles[0.043]; Eagles[0.043]; Indianapolis[0.043]; Indianapolis[0.043]; Indianapolis[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington Redskins ==> ENTITY: \u001b[32mWashington Redskins\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.208:0.208[0]; log p(e|m)= -0.078:-0.078[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJets[0.043]; season[0.043]; fourth[0.043]; Philadelphia[0.042]; Philadelphia[0.042]; linebacker[0.042]; Cowboys[0.042]; games[0.042]; games[0.042]; game[0.042]; second[0.042]; yard[0.042]; yard[0.042]; quarter[0.041]; touchdown[0.041]; touchdown[0.041]; Colts[0.041]; Colts[0.041]; Colts[0.041]; Dallas[0.041]; clinch[0.041]; Cincinnati[0.041]; Jason[0.041]; cornerback[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ty Detmer\u001b[39m ==> ENTITY: \u001b[32mTy Detmer\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.045]; quarterback[0.044]; game[0.044]; game[0.044]; benched[0.043]; Faulk[0.042]; Ricky[0.042]; Troy[0.042]; start[0.042]; Eagles[0.042]; rushing[0.042]; Harrison[0.042]; seven[0.041]; sprained[0.041]; draft[0.041]; Miami[0.041]; capped[0.041]; yards[0.041]; yards[0.041]; yards[0.041]; straight[0.041]; knee[0.040]; Philadelphia[0.040]; Justin[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ricky Watters\u001b[39m ==> ENTITY: \u001b[32mRicky Watters\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDetmer[0.046]; game[0.044]; game[0.044]; Ty[0.043]; rushing[0.043]; quarterback[0.042]; capped[0.042]; Harrison[0.042]; Eagles[0.042]; scoring[0.041]; Miami[0.041]; Troy[0.041]; seven[0.041]; benched[0.041]; return[0.041]; Philadelphia[0.041]; yards[0.041]; yards[0.041]; yards[0.041]; debut[0.041]; Dolphins[0.040]; left[0.040]; left[0.040]; score[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ray Buchanan\u001b[39m ==> ENTITY: \u001b[32mRay Buchanan\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTd[0.048]; Redskins[0.047]; run[0.047]; game[0.046]; Cowboys[0.046]; Harrison[0.046]; Colts[0.046]; Colts[0.046]; Washington[0.046]; Indianapolis[0.046]; Packers[0.046]; Philadelphia[0.045]; Philadelphia[0.045]; Philadelphia[0.045]; touchdown[0.045]; played[0.044]; Eagles[0.044]; linebacker[0.044]; Faulk[0.044]; defensive[0.044]; tie[0.044]; tackle[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Eagles ==> ENTITY: \u001b[32mPhiladelphia Eagles\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.204:0.204[0]; log p(e|m)= -2.477:-2.477[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntouchdown[0.047]; touchdown[0.047]; Philadelphia[0.046]; Philadelphia[0.046]; Philadelphia[0.046]; yard[0.046]; Ray[0.046]; Jets[0.046]; Cowboys[0.046]; Eagles[0.045]; Eagles[0.045]; interception[0.045]; season[0.045]; game[0.045]; Jim[0.045]; games[0.045]; games[0.045]; yards[0.045]; yards[0.045]; yards[0.045]; playoff[0.045]; playoff[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jim Harbaugh\u001b[39m ==> ENTITY: \u001b[32mJim Harbaugh\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nquarterback[0.045]; games[0.043]; games[0.043]; game[0.043]; season[0.042]; Cincinnati[0.042]; Indianapolis[0.042]; Indianapolis[0.042]; Indianapolis[0.042]; starting[0.042]; starting[0.042]; Cowboys[0.041]; playoff[0.041]; playoff[0.041]; touchdowns[0.041]; touchdown[0.041]; touchdown[0.041]; Packers[0.041]; won[0.041]; Jets[0.041]; Redskins[0.041]; fourth[0.041]; Football[0.040]; Arizona[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NFL ==> ENTITY: \u001b[32mNational Football League\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.173:0.173[0]; log p(e|m)= -0.019:-0.019[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwon[0.043]; game[0.043]; game[0.043]; Dolphins[0.043]; Faulk[0.043]; Faulk[0.043]; final[0.042]; benched[0.042]; Philadelphia[0.042]; linebacker[0.042]; second[0.041]; Colts[0.041]; seven[0.041]; seven[0.041]; Harrison[0.041]; defensive[0.041]; Ricky[0.041]; quarterback[0.041]; left[0.041]; left[0.041]; starting[0.041]; return[0.041]; Miami[0.041]; injury[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jason Belser\u001b[39m ==> ENTITY: \u001b[32mJason Belser\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayed[0.048]; game[0.045]; Football[0.044]; games[0.043]; games[0.043]; Cowboys[0.042]; Faulk[0.042]; quarterback[0.042]; Jim[0.041]; Redskins[0.041]; touchdowns[0.041]; Colts[0.041]; Colts[0.041]; Colts[0.041]; Colts[0.041]; Dallas[0.041]; Cincinnati[0.041]; Kansas[0.040]; Eagles[0.040]; Eagles[0.040]; Eagles[0.040]; Eagles[0.040]; touchdown[0.040]; touchdown[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Green Bay Packers ==> ENTITY: \u001b[32mGreen Bay Packers\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.175:0.175[0]; log p(e|m)= -0.100:-0.100[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngame[0.044]; season[0.043]; games[0.043]; games[0.043]; fourth[0.042]; second[0.042]; Jets[0.042]; playoff[0.042]; Cowboys[0.042]; won[0.042]; touchdown[0.042]; touchdown[0.042]; seven[0.042]; yard[0.041]; yard[0.041]; final[0.041]; final[0.041]; Eagles[0.041]; Eagles[0.041]; clinch[0.041]; Faulk[0.041]; Justin[0.041]; Justin[0.041]; Justin[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marvin Harrison\u001b[39m ==> ENTITY: \u001b[32mMarvin Harrison\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.248:0.248[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFaulk[0.045]; Faulk[0.045]; Detmer[0.043]; game[0.042]; game[0.042]; Ricky[0.042]; rushing[0.042]; Td[0.042]; second[0.042]; cornerback[0.042]; quarterback[0.042]; Troy[0.042]; yards[0.041]; yards[0.041]; yards[0.041]; straight[0.041]; tackle[0.041]; linebacker[0.041]; yard[0.041]; yard[0.041]; yard[0.041]; yard[0.041]; yard[0.041]; Philadelphia[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIndianapolis\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIndianapolis Colts\u001b[39m <---> \u001b[32mIndianapolis\u001b[39m\t\nSCORES: global= 0.258:0.235[\u001b[31m0.022\u001b[39m]; local()= 0.181:0.041[\u001b[31m0.140\u001b[39m]; log p(e|m)= -3.411:-0.302[\u001b[32m3.109\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfourth[0.045]; Colts[0.044]; Colts[0.044]; Colts[0.044]; Colts[0.044]; Indianapolis[0.043]; touchdown[0.043]; interception[0.043]; games[0.042]; playoff[0.042]; score[0.042]; quarterback[0.042]; Philadelphia[0.041]; touchdowns[0.041]; Faulk[0.041]; Afc[0.041]; Paul[0.040]; Football[0.040]; Justin[0.040]; rallied[0.040]; returned[0.040]; yards[0.040]; yards[0.040]; yards[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Harbaugh\u001b[39m ==> ENTITY: \u001b[32mJim Harbaugh\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nquarterback[0.046]; Detmer[0.044]; game[0.043]; Td[0.043]; starting[0.042]; Cowboys[0.042]; playoff[0.042]; Miami[0.042]; Ty[0.042]; Dolphins[0.041]; Tony[0.041]; linebacker[0.041]; Harrison[0.041]; Packers[0.041]; second[0.041]; Troy[0.041]; won[0.041]; Jets[0.041]; Redskins[0.041]; Arizona[0.041]; cornerback[0.041]; tackle[0.041]; Philadelphia[0.041]; Philadelphia[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York Jets ==> ENTITY: \u001b[32mNew York Jets\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.211:0.211[0]; log p(e|m)= -0.090:-0.090[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.043]; touchdown[0.043]; touchdown[0.043]; fourth[0.043]; Colts[0.042]; Colts[0.042]; Colts[0.042]; yard[0.042]; yard[0.042]; Philadelphia[0.042]; Philadelphia[0.042]; game[0.042]; won[0.042]; playoff[0.042]; Cowboys[0.041]; seven[0.041]; season[0.041]; final[0.041]; final[0.041]; cornerback[0.041]; quarter[0.041]; Dallas[0.041]; games[0.041]; games[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indianapolis ==> ENTITY: \u001b[32mIndianapolis Colts\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.214:0.214[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfourth[0.044]; game[0.044]; Colts[0.043]; Colts[0.043]; Colts[0.043]; season[0.043]; touchdown[0.042]; touchdown[0.042]; interception[0.042]; games[0.041]; games[0.041]; playoff[0.041]; playoff[0.041]; Cincinnati[0.041]; score[0.041]; quarterback[0.041]; tie[0.041]; clinch[0.041]; Jets[0.041]; final[0.041]; Philadelphia[0.040]; Philadelphia[0.040]; Philadelphia[0.040]; linebacker[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colts ==> ENTITY: \u001b[32mIndianapolis Colts\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.219:0.219[0]; log p(e|m)= -0.555:-0.555[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfourth[0.044]; game[0.044]; Colts[0.043]; Colts[0.043]; Colts[0.043]; season[0.042]; touchdown[0.042]; touchdown[0.042]; interception[0.042]; playoff[0.041]; games[0.041]; games[0.041]; Indianapolis[0.041]; Indianapolis[0.041]; Indianapolis[0.041]; Cincinnati[0.041]; score[0.041]; quarterback[0.041]; tie[0.041]; Football[0.041]; Philadelphia[0.041]; Philadelphia[0.041]; Jets[0.041]; final[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Quentin Coryatt\u001b[39m ==> ENTITY: \u001b[32mQuentin Coryatt\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlinebacker[0.044]; Td[0.044]; tackle[0.043]; Ray[0.043]; Dallas[0.043]; Cowboys[0.042]; Indianapolis[0.042]; second[0.042]; cornerback[0.042]; Justin[0.042]; Justin[0.042]; defensive[0.042]; Harrison[0.042]; Packers[0.042]; Faulk[0.042]; games[0.041]; Colts[0.041]; game[0.041]; Eagles[0.041]; starting[0.041]; season[0.040]; touchdown[0.040]; Cincinnati[0.040]; Troy[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPhiladelphia\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPhiladelphia Eagles\u001b[39m <---> \u001b[32mPhiladelphia\u001b[39m\t\nSCORES: global= 0.248:0.234[\u001b[31m0.015\u001b[39m]; local()= 0.220:0.059[\u001b[31m0.161\u001b[39m]; log p(e|m)= -4.510:-0.211[\u001b[32m4.299\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.045]; games[0.044]; games[0.044]; playoff[0.043]; playoff[0.043]; season[0.042]; game[0.042]; Eagles[0.042]; Eagles[0.042]; Cowboys[0.042]; fourth[0.041]; Jets[0.041]; yard[0.041]; yard[0.041]; played[0.041]; touchdown[0.041]; touchdown[0.041]; won[0.041]; Redskins[0.041]; Cincinnati[0.041]; Washington[0.041]; linebacker[0.041]; Tony[0.040]; defensive[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Philadelphia ==> ENTITY: \u001b[32mPhiladelphia Eagles\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.187:0.187[0]; log p(e|m)= -4.510:-4.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngame[0.045]; game[0.045]; Eagles[0.045]; Dolphins[0.045]; second[0.044]; yard[0.044]; yard[0.044]; yard[0.044]; yard[0.044]; yard[0.044]; debut[0.044]; start[0.044]; Miami[0.044]; linebacker[0.043]; Harrison[0.043]; quarterback[0.043]; seven[0.043]; seven[0.043]; capped[0.042]; rout[0.042]; left[0.042]; left[0.042]; straight[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Troy Vincent\u001b[39m ==> ENTITY: \u001b[32mTroy Vincent\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDetmer[0.045]; linebacker[0.044]; Td[0.044]; game[0.043]; game[0.043]; return[0.042]; quarterback[0.042]; Ty[0.042]; benched[0.042]; Dolphins[0.042]; left[0.041]; left[0.041]; Ricky[0.041]; Harrison[0.041]; injury[0.041]; Justin[0.041]; Justin[0.041]; draft[0.041]; start[0.041]; Eagles[0.041]; shoulder[0.041]; second[0.040]; knee[0.040]; helmet[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colts ==> ENTITY: \u001b[32mIndianapolis Colts\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.201:0.201[0]; log p(e|m)= -0.555:-0.555[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayoff[0.044]; game[0.044]; Colts[0.043]; second[0.043]; season[0.042]; touchdown[0.042]; touchdown[0.042]; yard[0.042]; yard[0.042]; quarter[0.042]; quarter[0.042]; games[0.041]; clinch[0.041]; Indianapolis[0.041]; final[0.041]; final[0.041]; Cincinnati[0.041]; score[0.041]; tie[0.041]; run[0.041]; straight[0.041]; Philadelphia[0.041]; Philadelphia[0.041]; Jets[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCincinnati\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mCincinnati Bengals\u001b[39m <---> \u001b[32mCincinnati\u001b[39m\t\nSCORES: global= 0.250:0.231[\u001b[31m0.019\u001b[39m]; local()= 0.221:0.078[\u001b[31m0.143\u001b[39m]; log p(e|m)= -3.963:-0.460[\u001b[32m3.503\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngame[0.044]; season[0.043]; games[0.043]; games[0.043]; linebacker[0.042]; fourth[0.042]; final[0.042]; quarter[0.042]; Indianapolis[0.042]; starting[0.042]; starting[0.042]; Cowboys[0.042]; Jets[0.042]; Philadelphia[0.041]; Philadelphia[0.041]; Philadelphia[0.041]; tackle[0.041]; yard[0.041]; cornerback[0.041]; Marshall[0.041]; Faulk[0.041]; Faulk[0.041]; Arizona[0.041]; Redskins[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paul Justin\u001b[39m ==> ENTITY: \u001b[32mPaul Justin\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.208:0.208[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngame[0.044]; quarterback[0.044]; games[0.043]; games[0.043]; Redskins[0.042]; starting[0.042]; Football[0.042]; Jets[0.042]; Cowboys[0.042]; Faulk[0.041]; touchdowns[0.041]; Cincinnati[0.041]; Packers[0.041]; Indianapolis[0.041]; Indianapolis[0.041]; Indianapolis[0.041]; Marshall[0.041]; season[0.041]; played[0.041]; Jason[0.041]; Philadelphia[0.041]; Philadelphia[0.041]; Philadelphia[0.041]; Arizona[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Justin\u001b[39m ==> ENTITY: \u001b[32mPaul Justin\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDetmer[0.044]; game[0.044]; quarterback[0.044]; Td[0.043]; Justin[0.043]; starting[0.042]; Jets[0.042]; Harrison[0.042]; Faulk[0.041]; Faulk[0.041]; Ty[0.041]; second[0.041]; benched[0.041]; Packers[0.041]; defensive[0.041]; Troy[0.041]; Philadelphia[0.041]; Philadelphia[0.041]; cornerback[0.041]; Arizona[0.041]; sprained[0.041]; Eagles[0.041]; draft[0.041]; linebacker[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NFL ==> ENTITY: \u001b[32mNational Football League\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.164:0.164[0]; log p(e|m)= -0.019:-0.019[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; playoff[0.044]; Faulk[0.044]; Football[0.043]; fourth[0.043]; Philadelphia[0.043]; Indianapolis[0.042]; Indianapolis[0.042]; quarterback[0.042]; Colts[0.042]; Colts[0.042]; Colts[0.042]; starting[0.041]; Eagles[0.041]; Eagles[0.041]; Eagles[0.041]; Afc[0.041]; mark[0.041]; interception[0.041]; Thursday[0.040]; Playoff[0.040]; American[0.040]; Justin[0.040]; score[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indianapolis Colts ==> ENTITY: \u001b[32mIndianapolis Colts\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.062:-0.062[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfourth[0.044]; Colts[0.044]; Colts[0.044]; Colts[0.044]; game[0.044]; season[0.043]; touchdown[0.043]; interception[0.043]; playoff[0.042]; games[0.042]; Indianapolis[0.042]; score[0.042]; quarterback[0.041]; Afc[0.041]; Philadelphia[0.041]; touchdowns[0.040]; Faulk[0.040]; rallied[0.040]; starting[0.040]; Football[0.040]; Playoff[0.040]; Eagles[0.040]; Eagles[0.040]; Eagles[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mArizona\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mArizona Cardinals\u001b[39m <---> \u001b[32mArizona\u001b[39m\t\nSCORES: global= 0.248:0.219[\u001b[31m0.029\u001b[39m]; local()= 0.207:0.010[\u001b[31m0.197\u001b[39m]; log p(e|m)= -4.343:-0.226[\u001b[32m4.117\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; games[0.044]; season[0.043]; game[0.043]; Jets[0.042]; Cowboys[0.042]; Philadelphia[0.042]; Philadelphia[0.042]; quarter[0.042]; Dallas[0.042]; Cincinnati[0.041]; second[0.041]; fourth[0.041]; touchdown[0.041]; touchdown[0.041]; yard[0.041]; yard[0.041]; linebacker[0.041]; clinch[0.041]; playoff[0.041]; Redskins[0.041]; seven[0.041]; tie[0.041]; played[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Faulk\u001b[39m ==> ENTITY: \u001b[32mMarshall Faulk\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFaulk[0.045]; Td[0.044]; Harrison[0.044]; games[0.043]; Ray[0.043]; Cowboys[0.042]; Tony[0.042]; Packers[0.042]; season[0.042]; Troy[0.041]; second[0.041]; Cincinnati[0.041]; tackle[0.041]; Dolphins[0.041]; defensive[0.041]; Dallas[0.041]; Colts[0.041]; Miami[0.041]; linebacker[0.041]; Arizona[0.041]; injury[0.041]; Eagles[0.041]; Jets[0.040]; Indianapolis[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dallas Cowboys ==> ENTITY: \u001b[32mDallas Cowboys\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.129:-0.129[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayoff[0.043]; season[0.043]; games[0.043]; games[0.043]; game[0.042]; Jason[0.042]; Jets[0.042]; fourth[0.042]; returned[0.042]; Eagles[0.042]; Eagles[0.042]; second[0.042]; defensive[0.041]; loss[0.041]; touchdown[0.041]; touchdown[0.041]; cornerback[0.041]; Justin[0.041]; Justin[0.041]; linebacker[0.041]; Colts[0.041]; Colts[0.041]; Colts[0.041]; quarter[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colts ==> ENTITY: \u001b[32mIndianapolis Colts\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.218:0.218[0]; log p(e|m)= -0.555:-0.555[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfourth[0.044]; game[0.044]; Colts[0.043]; Colts[0.043]; Colts[0.043]; season[0.042]; touchdown[0.042]; touchdown[0.042]; interception[0.042]; playoff[0.041]; playoff[0.041]; games[0.041]; games[0.041]; Indianapolis[0.041]; Indianapolis[0.041]; Indianapolis[0.041]; Cincinnati[0.041]; score[0.041]; quarterback[0.041]; tie[0.041]; clinch[0.041]; Philadelphia[0.041]; Philadelphia[0.041]; Philadelphia[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Faulk\u001b[39m ==> ENTITY: \u001b[32mMarshall Faulk\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFaulk[0.049]; Detmer[0.049]; Ricky[0.048]; Td[0.048]; Harrison[0.048]; Ty[0.046]; rushing[0.045]; game[0.045]; game[0.045]; Troy[0.045]; second[0.045]; quarterback[0.045]; Dolphins[0.045]; carries[0.045]; Miami[0.044]; benched[0.044]; injury[0.044]; Eagles[0.044]; Bell[0.044]; seven[0.044]; seven[0.044]; return[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Eagles ==> ENTITY: \u001b[32mPhiladelphia Eagles\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.168:0.168[0]; log p(e|m)= -2.477:-2.477[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleft[0.044]; left[0.044]; Philadelphia[0.043]; yard[0.043]; yard[0.043]; yard[0.043]; Dolphins[0.042]; game[0.042]; game[0.042]; yards[0.042]; yards[0.042]; yards[0.042]; quarterback[0.042]; capped[0.041]; second[0.041]; Harrison[0.041]; Justin[0.041]; Justin[0.041]; right[0.041]; scoring[0.041]; Ricky[0.040]; sprained[0.040]; Miami[0.040]; Faulk[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tony Siragusa\u001b[39m ==> ENTITY: \u001b[32mTony Siragusa\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.045]; game[0.045]; Td[0.044]; games[0.044]; linebacker[0.043]; defensive[0.043]; Cowboys[0.042]; played[0.042]; Packers[0.042]; Jets[0.042]; Harrison[0.041]; Redskins[0.041]; Dolphins[0.041]; playoff[0.041]; Eagles[0.041]; tackle[0.041]; Colts[0.041]; Colts[0.041]; touchdown[0.041]; touchdown[0.041]; yard[0.040]; yard[0.040]; yard[0.040]; yard[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Philadelphia Eagles ==> ENTITY: \u001b[32mPhiladelphia Eagles\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.199:0.199[0]; log p(e|m)= -0.114:-0.114[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntouchdown[0.045]; touchdown[0.045]; Colts[0.044]; Colts[0.044]; Colts[0.044]; Colts[0.044]; game[0.044]; yards[0.044]; yards[0.044]; yards[0.044]; Eagles[0.044]; Eagles[0.044]; Eagles[0.044]; interception[0.043]; season[0.043]; games[0.043]; games[0.043]; fourth[0.043]; quarterback[0.043]; playoff[0.042]; Football[0.042]; Jason[0.042]; Cincinnati[0.042]; \t\n\r [================>.............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s443ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1740/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1381testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jerusalem ==> ENTITY: \u001b[32mJerusalem\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.217:0.217[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJerusalem[0.048]; Jerusalem[0.048]; Bnei[0.046]; Haifa[0.045]; Haifa[0.045]; Tikva[0.045]; Tikva[0.045]; Israeli[0.045]; Beit[0.044]; Yehuda[0.044]; Beersheba[0.043]; Aviv[0.043]; Aviv[0.043]; Rishon[0.043]; Herzliya[0.043]; Kfar[0.043]; Ironi[0.042]; Petah[0.042]; Petah[0.042]; Betar[0.041]; Holon[0.040]; Tel[0.039]; Tel[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hapoel Haifa\u001b[39m ==> ENTITY: \u001b[32mHapoel Haifa F.C.\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.272:0.272[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMaccabi[0.046]; Maccabi[0.046]; Maccabi[0.046]; Maccabi[0.046]; Maccabi[0.046]; Maccabi[0.046]; Maccabi[0.046]; Betar[0.043]; Betar[0.043]; Aviv[0.043]; Aviv[0.043]; Aviv[0.043]; Hapoel[0.042]; Hapoel[0.042]; Hapoel[0.042]; Hapoel[0.042]; Hapoel[0.042]; Hapoel[0.042]; Hapoel[0.042]; Hapoel[0.042]; Hapoel[0.042]; Hapoel[0.042]; Hapoel[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Maccabi Petah Tikva\u001b[39m ==> ENTITY: \u001b[32mMaccabi Petah Tikva F.C.\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.276:0.276[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBetar[0.079]; Betar[0.079]; Maccabi[0.077]; Maccabi[0.077]; Maccabi[0.077]; Maccabi[0.077]; Maccabi[0.077]; Maccabi[0.077]; Tikva[0.076]; Tikva[0.076]; Tikva[0.076]; Bnei[0.075]; Bnei[0.075]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Betar Jerusalem\u001b[39m ==> ENTITY: \u001b[32mBeitar Jerusalem F.C.\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.279:0.279[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBetar[0.068]; Maccabi[0.068]; Maccabi[0.068]; Maccabi[0.068]; Maccabi[0.068]; Maccabi[0.068]; Maccabi[0.068]; Bnei[0.067]; Bnei[0.067]; Aviv[0.066]; Aviv[0.066]; Aviv[0.066]; Soccer[0.065]; soccer[0.065]; Ironi[0.065]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hapoel Taibe\u001b[39m ==> ENTITY: \u001b[32mHapoel Tayibe F.C.\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.263:0.263[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBetar[0.044]; Betar[0.044]; Bnei[0.042]; Rishon[0.042]; Hapoel[0.042]; Hapoel[0.042]; Hapoel[0.042]; Hapoel[0.042]; Hapoel[0.042]; Hapoel[0.042]; Hapoel[0.042]; Hapoel[0.042]; Hapoel[0.042]; Tikva[0.042]; Tikva[0.042]; Tikva[0.042]; Tikva[0.042]; matches[0.041]; Beit[0.041]; Aviv[0.041]; Aviv[0.041]; Aviv[0.041]; division[0.041]; Kfar[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Maccabi Tel Aviv ==> ENTITY: \u001b[32mMaccabi Tel Aviv F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.348:0.348[0]; log p(e|m)= -0.229:-0.229[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMaccabi[0.045]; Maccabi[0.045]; Maccabi[0.045]; Maccabi[0.045]; Maccabi[0.045]; Maccabi[0.045]; Maccabi[0.045]; Betar[0.043]; Betar[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Maccabi Herzliya\u001b[39m ==> ENTITY: \u001b[32mMaccabi Herzliya F.C.\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.275:0.275[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIroni[0.044]; Bnei[0.044]; Maccabi[0.044]; Maccabi[0.044]; Maccabi[0.044]; Aviv[0.043]; Aviv[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Rishon[0.041]; Haifa[0.041]; Haifa[0.041]; Tikva[0.041]; Holon[0.040]; Tel[0.040]; Tel[0.040]; Beersheba[0.039]; Yehuda[0.038]; Kfar[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Betar Jerusalem\u001b[39m ==> ENTITY: \u001b[32mBeitar Jerusalem F.C.\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.285:0.285[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBetar[0.064]; Maccabi[0.063]; Maccabi[0.063]; Maccabi[0.063]; Maccabi[0.063]; Maccabi[0.063]; Maccabi[0.063]; Maccabi[0.063]; Bnei[0.063]; Bnei[0.063]; Aviv[0.061]; Aviv[0.061]; Aviv[0.061]; soccer[0.061]; Ironi[0.061]; Ironi[0.061]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Maccabi Tel Aviv ==> ENTITY: \u001b[32mMaccabi Tel Aviv F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.322:0.322[0]; log p(e|m)= -0.229:-0.229[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMaccabi[0.045]; Maccabi[0.045]; Maccabi[0.045]; Maccabi[0.045]; Maccabi[0.045]; Betar[0.044]; Betar[0.044]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Hapoel[0.043]; Soccer[0.043]; Herzliya[0.042]; Aviv[0.042]; Aviv[0.042]; soccer[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bnei Yehuda\u001b[39m ==> ENTITY: \u001b[32mBnei Yehuda Tel Aviv F.C.\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.319:0.319[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBetar[0.043]; Betar[0.043]; Maccabi[0.043]; Maccabi[0.043]; Maccabi[0.043]; Maccabi[0.043]; Maccabi[0.043]; Maccabi[0.043]; Bnei[0.042]; Aviv[0.041]; Aviv[0.041]; Aviv[0.041]; matches[0.041]; Herzliya[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Maccabi Haifa\u001b[39m ==> ENTITY: \u001b[32mMaccabi Haifa F.C.\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.304:0.304[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMaccabi[0.044]; Maccabi[0.044]; Maccabi[0.044]; Maccabi[0.044]; Maccabi[0.044]; Betar[0.042]; Haifa[0.042]; Haifa[0.042]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Herzliya[0.040]; Herzliya[0.040]; Ironi[0.040]; Ironi[0.040]; Bnei[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zafririm Holon\u001b[39m ==> ENTITY: \u001b[32mHapoel Tzafririm Holon F.C.\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.297:0.297[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMaccabi[0.044]; Maccabi[0.044]; Maccabi[0.044]; Maccabi[0.044]; Bnei[0.042]; Hapoel[0.042]; Hapoel[0.042]; Hapoel[0.042]; Hapoel[0.042]; Hapoel[0.042]; Hapoel[0.042]; Hapoel[0.042]; Hapoel[0.042]; Hapoel[0.042]; Hapoel[0.042]; Betar[0.041]; Betar[0.041]; Herzliya[0.041]; Tikva[0.041]; Tikva[0.041]; Tikva[0.041]; Rishon[0.041]; Beersheba[0.040]; Beersheba[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Maccabi Petah Tikva\u001b[39m ==> ENTITY: \u001b[32mMaccabi Petah Tikva F.C.\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.282:0.282[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBetar[0.079]; Betar[0.079]; Maccabi[0.077]; Maccabi[0.077]; Maccabi[0.077]; Maccabi[0.077]; Maccabi[0.077]; Maccabi[0.077]; Maccabi[0.077]; Tikva[0.076]; Tikva[0.076]; Bnei[0.075]; Bnei[0.075]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hapoel Beersheba\u001b[39m ==> ENTITY: \u001b[32mHapoel Be'er Sheva F.C.\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.336:0.336[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBetar[0.047]; Betar[0.047]; Maccabi[0.047]; Maccabi[0.047]; Maccabi[0.047]; Maccabi[0.047]; Maccabi[0.047]; Maccabi[0.047]; Maccabi[0.047]; Bnei[0.045]; Bnei[0.045]; Hapoel[0.045]; Hapoel[0.045]; Hapoel[0.045]; Hapoel[0.045]; Hapoel[0.045]; Hapoel[0.045]; Hapoel[0.045]; Hapoel[0.045]; Hapoel[0.045]; Hapoel[0.045]; Hapoel[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hapoel Tel Aviv\u001b[39m ==> ENTITY: \u001b[32mHapoel Tel Aviv F.C.\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.286:0.286[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMaccabi[0.044]; Maccabi[0.044]; Maccabi[0.044]; Maccabi[0.044]; Maccabi[0.044]; Maccabi[0.044]; Betar[0.042]; Betar[0.042]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Haifa[0.041]; Haifa[0.041]; matches[0.040]; Soccer[0.040]; Aviv[0.040]; Aviv[0.040]; Herzliya[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hapoel Tel Aviv\u001b[39m ==> ENTITY: \u001b[32mHapoel Tel Aviv F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.222:0.222[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMaccabi[0.057]; Maccabi[0.057]; Hapoel[0.052]; Hapoel[0.052]; Hapoel[0.052]; Hapoel[0.052]; Hapoel[0.052]; Haifa[0.052]; Haifa[0.052]; Herzliya[0.051]; Bnei[0.050]; Rishon[0.050]; Ironi[0.050]; Kfar[0.047]; Yehuda[0.047]; Holon[0.047]; Jerusalem[0.046]; Beit[0.045]; Taiba[0.044]; Sava[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hapoel Petah Tikva\u001b[39m ==> ENTITY: \u001b[32mHapoel Petah Tikva F.C.\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.314:0.314[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMaccabi[0.048]; Maccabi[0.048]; Maccabi[0.048]; Maccabi[0.048]; Maccabi[0.048]; Maccabi[0.048]; Maccabi[0.048]; Haifa[0.048]; Haifa[0.048]; Haifa[0.048]; Haifa[0.048]; Bnei[0.048]; Bnei[0.048]; Betar[0.048]; Betar[0.048]; Tikva[0.047]; Tikva[0.047]; Tikva[0.047]; Herzliya[0.047]; Rishon[0.046]; Rishon[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hapoel Jerusalem ==> ENTITY: \u001b[32mHapoel Jerusalem F.C.\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.318:0.318[0]; log p(e|m)= -0.311:-0.311[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBetar[0.046]; Maccabi[0.044]; Maccabi[0.044]; Maccabi[0.044]; Maccabi[0.044]; Bnei[0.042]; Tikva[0.042]; Tikva[0.042]; Herzliya[0.042]; Ironi[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Rishon[0.040]; Haifa[0.039]; Haifa[0.039]; Haifa[0.039]; Aviv[0.039]; Aviv[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hapoel Haifa\u001b[39m ==> ENTITY: \u001b[32mHapoel Haifa F.C.\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.273:0.273[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMaccabi[0.079]; Maccabi[0.079]; Maccabi[0.079]; Maccabi[0.079]; Maccabi[0.079]; Maccabi[0.079]; Maccabi[0.079]; Betar[0.075]; Betar[0.075]; Soccer[0.074]; Aviv[0.074]; Aviv[0.074]; Aviv[0.074]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hapoel Jerusalem ==> ENTITY: \u001b[32mHapoel Jerusalem F.C.\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.327:0.327[0]; log p(e|m)= -0.311:-0.311[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBetar[0.056]; Betar[0.056]; Maccabi[0.054]; Maccabi[0.054]; Maccabi[0.054]; Maccabi[0.054]; Maccabi[0.054]; Maccabi[0.054]; Soccer[0.052]; soccer[0.052]; Bnei[0.051]; Bnei[0.051]; matches[0.051]; Tikva[0.051]; Tikva[0.051]; Tikva[0.051]; Tikva[0.051]; Herzliya[0.051]; Ironi[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.218:0.218[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAviv[0.044]; Aviv[0.044]; Maccabi[0.043]; Maccabi[0.043]; Maccabi[0.043]; Maccabi[0.043]; Herzliya[0.043]; Tikva[0.042]; Tikva[0.042]; Yehuda[0.042]; Ironi[0.042]; Jerusalem[0.042]; Jerusalem[0.042]; Jerusalem[0.042]; Rishon[0.041]; matches[0.041]; Bnei[0.041]; Soccer[0.041]; Haifa[0.040]; Haifa[0.040]; soccer[0.040]; Beersheba[0.040]; Petah[0.040]; Petah[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hapoel Petah Tikva\u001b[39m ==> ENTITY: \u001b[32mHapoel Petah Tikva F.C.\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.306:0.306[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMaccabi[0.043]; Maccabi[0.043]; Maccabi[0.043]; Maccabi[0.043]; Maccabi[0.043]; Haifa[0.042]; Haifa[0.042]; Bnei[0.042]; Betar[0.042]; Betar[0.042]; Tikva[0.042]; Tikva[0.042]; Herzliya[0.041]; Rishon[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; division[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bnei Yehuda\u001b[39m ==> ENTITY: \u001b[32mBnei Yehuda Tel Aviv F.C.\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.327:0.327[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBetar[0.064]; Betar[0.064]; Maccabi[0.063]; Maccabi[0.063]; Maccabi[0.063]; Maccabi[0.063]; Maccabi[0.063]; Maccabi[0.063]; Maccabi[0.063]; Bnei[0.063]; Aviv[0.061]; Aviv[0.061]; Aviv[0.061]; Aviv[0.061]; Herzliya[0.060]; Herzliya[0.060]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hapoel Beersheba\u001b[39m ==> ENTITY: \u001b[32mHapoel Be'er Sheva F.C.\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.337:0.337[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBetar[0.043]; Betar[0.043]; Maccabi[0.043]; Maccabi[0.043]; Maccabi[0.043]; Maccabi[0.043]; Maccabi[0.043]; Maccabi[0.043]; Maccabi[0.043]; Maccabi[0.043]; Bnei[0.041]; Bnei[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Maccabi Herzliya\u001b[39m ==> ENTITY: \u001b[32mMaccabi Herzliya F.C.\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.308:0.308[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIroni[0.072]; Bnei[0.072]; Bnei[0.072]; Maccabi[0.072]; Maccabi[0.072]; Maccabi[0.072]; Maccabi[0.072]; Maccabi[0.072]; Maccabi[0.072]; Betar[0.071]; Betar[0.071]; Aviv[0.071]; Aviv[0.071]; Aviv[0.071]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zafririm Holon\u001b[39m ==> ENTITY: \u001b[32mHapoel Tzafririm Holon F.C.\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.294:0.294[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMaccabi[0.046]; Maccabi[0.046]; Maccabi[0.046]; Maccabi[0.046]; Bnei[0.044]; Hapoel[0.044]; Hapoel[0.044]; Hapoel[0.044]; Hapoel[0.044]; Hapoel[0.044]; Hapoel[0.044]; Hapoel[0.044]; Hapoel[0.044]; Herzliya[0.043]; Tikva[0.043]; Tikva[0.043]; Rishon[0.043]; Beersheba[0.042]; Kfar[0.042]; Aviv[0.042]; Aviv[0.042]; Ironi[0.042]; Beit[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Maccabi Haifa\u001b[39m ==> ENTITY: \u001b[32mMaccabi Haifa F.C.\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.293:0.293[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.044]; Soccer[0.044]; Maccabi[0.043]; Maccabi[0.043]; Maccabi[0.043]; Maccabi[0.043]; Maccabi[0.043]; matches[0.042]; Betar[0.041]; Betar[0.041]; Haifa[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Hapoel[0.041]; Herzliya[0.040]; Ironi[0.040]; Bnei[0.040]; \t\n\r [=================>............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s291ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1767/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1195testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Paris ==> ENTITY: \u001b[32mParis\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.085:0.085[0]; log p(e|m)= -0.115:-0.115[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nParis[0.049]; matches[0.047]; Soccer[0.046]; Christophe[0.046]; Claude[0.045]; division[0.044]; League[0.044]; Friday[0.044]; Paul[0.043]; French[0.042]; French[0.042]; Germain[0.042]; Phil[0.042]; Nancy[0.041]; Nantes[0.041]; Gray[0.040]; Lens[0.040]; Bruno[0.040]; Attendance[0.039]; Jocelyn[0.039]; Doram[0.036]; Fischer[0.036]; Halftime[0.036]; Makelele[0.035]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lens ==> ENTITY: \u001b[32mRC Lens\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.140:0.140[0]; log p(e|m)= -0.860:-0.860[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.048]; matches[0.047]; Nantes[0.046]; Soccer[0.046]; Claude[0.044]; Christophe[0.044]; Paul[0.044]; French[0.043]; French[0.043]; Makelele[0.042]; division[0.041]; Paris[0.041]; Paris[0.041]; Bruno[0.041]; Nancy[0.040]; Gray[0.040]; Friday[0.039]; Attendance[0.039]; Germain[0.038]; Jocelyn[0.038]; Fischer[0.038]; Phil[0.038]; Japhet[0.038]; Halftime[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paris St Germain\u001b[39m ==> ENTITY: \u001b[32mParis Saint-Germain F.C.\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.048]; matches[0.046]; Soccer[0.045]; Nantes[0.044]; Christophe[0.043]; Paul[0.043]; division[0.043]; Lens[0.043]; French[0.042]; French[0.042]; Halftime[0.042]; Paris[0.042]; Claude[0.042]; Summaries[0.041]; Summaries[0.041]; Friday[0.040]; Phil[0.040]; Bruno[0.040]; Attendance[0.040]; Makelele[0.039]; Nancy[0.039]; Gotty[0.039]; Gray[0.037]; Japhet[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jocelyn Gourvennec\u001b[39m ==> ENTITY: \u001b[32mJocelyn Gourvennec\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.050]; French[0.045]; French[0.045]; Nantes[0.045]; Christophe[0.045]; matches[0.044]; Claude[0.043]; League[0.043]; Lens[0.042]; Makelele[0.042]; Nancy[0.042]; Halftime[0.042]; Germain[0.041]; Summaries[0.040]; Summaries[0.040]; Paris[0.040]; Paris[0.040]; Attendance[0.039]; Japhet[0.039]; division[0.039]; Gray[0.039]; Bruno[0.039]; Paul[0.038]; Friday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nancy ==> ENTITY: \u001b[32mAS Nancy\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.167:0.167[0]; log p(e|m)= -1.687:-1.687[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.048]; League[0.047]; Christophe[0.047]; Nantes[0.045]; Claude[0.044]; Soccer[0.044]; French[0.044]; French[0.044]; Paris[0.044]; Paris[0.044]; Lens[0.043]; Jocelyn[0.041]; division[0.041]; Gray[0.040]; Paul[0.039]; Germain[0.039]; Bruno[0.039]; Phil[0.039]; Makelele[0.039]; Friday[0.039]; Doram[0.038]; Fischer[0.037]; Halftime[0.037]; Gotty[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.088:0.088[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrench[0.048]; matches[0.047]; Soccer[0.046]; League[0.045]; Paris[0.045]; Paris[0.045]; Christophe[0.043]; Paul[0.043]; Claude[0.043]; Bruno[0.043]; Nantes[0.043]; Nancy[0.042]; Lens[0.042]; Germain[0.042]; Jocelyn[0.040]; Makelele[0.040]; Friday[0.040]; division[0.039]; Halftime[0.039]; Phil[0.038]; Summaries[0.037]; Summaries[0.037]; Attendance[0.037]; Gray[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Claude Makelele\u001b[39m ==> ENTITY: \u001b[32mClaude Makélélé\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.097:0.097[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.046]; Soccer[0.046]; Halftime[0.044]; Christophe[0.044]; Paul[0.044]; League[0.043]; Nantes[0.043]; Lens[0.042]; French[0.042]; French[0.042]; Summaries[0.042]; Summaries[0.042]; Phil[0.042]; Germain[0.042]; Attendance[0.041]; Nancy[0.041]; Bruno[0.041]; Gotty[0.040]; division[0.040]; Gray[0.040]; Japhet[0.039]; Jocelyn[0.039]; Paris[0.039]; Paris[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nantes ==> ENTITY: \u001b[32mFC Nantes\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.149:0.149[0]; log p(e|m)= -1.457:-1.457[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.048]; League[0.046]; Soccer[0.046]; Paris[0.044]; Paris[0.044]; Christophe[0.044]; French[0.044]; French[0.044]; division[0.044]; Lens[0.044]; Halftime[0.044]; Phil[0.041]; Claude[0.041]; Paul[0.041]; Jocelyn[0.040]; Attendance[0.040]; Friday[0.040]; Germain[0.040]; Bruno[0.039]; Makelele[0.039]; Nancy[0.038]; Summaries[0.037]; Summaries[0.037]; Gray[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Phil Gray\u001b[39m ==> ENTITY: \u001b[32mPhil Gray\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.053:0.053[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.049]; Halftime[0.047]; matches[0.046]; Paul[0.046]; Soccer[0.045]; Summaries[0.045]; Summaries[0.045]; Nancy[0.045]; Attendance[0.044]; Nantes[0.044]; Lens[0.043]; Jocelyn[0.043]; Gotty[0.043]; division[0.043]; Friday[0.042]; Bruno[0.042]; Makelele[0.042]; Claude[0.042]; Japhet[0.042]; Doram[0.040]; Christophe[0.040]; Fischer[0.040]; Germain[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrench[0.048]; matches[0.047]; Soccer[0.046]; League[0.045]; Paris[0.045]; Paris[0.045]; Christophe[0.044]; Claude[0.043]; Paul[0.043]; Bruno[0.043]; Nantes[0.043]; Nancy[0.042]; Lens[0.042]; Germain[0.042]; Jocelyn[0.041]; Makelele[0.040]; Friday[0.040]; division[0.039]; Halftime[0.039]; Phil[0.039]; Attendance[0.037]; Gray[0.036]; Fischer[0.036]; Gotty[0.035]; \t\n\r [=================>............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s257ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1777/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1283testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.085:0.085[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouse[0.046]; House[0.046]; election[0.044]; President[0.044]; Mike[0.042]; said[0.042]; said[0.042]; president[0.042]; second[0.042]; second[0.042]; reporters[0.042]; Clinton[0.041]; Clinton[0.041]; Clinton[0.041]; won[0.041]; November[0.041]; term[0.041]; term[0.041]; term[0.041]; January[0.041]; conferences[0.040]; conferences[0.040]; conferences[0.040]; conferences[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = White House\u001b[39m ==> ENTITY: \u001b[32mWhite House\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.140:0.140[0]; log p(e|m)= -0.006:-0.006[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWashington[0.048]; President[0.047]; Clinton[0.046]; Clinton[0.046]; Clinton[0.046]; president[0.045]; January[0.044]; reporters[0.044]; election[0.043]; November[0.043]; said[0.043]; said[0.043]; Mike[0.043]; press[0.042]; House[0.042]; asked[0.042]; second[0.042]; second[0.042]; hold[0.042]; Friday[0.042]; White[0.042]; saying[0.042]; formal[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clinton ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.144:0.144[0]; log p(e|m)= -1.917:-1.917[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinton[0.048]; Clinton[0.048]; President[0.046]; Washington[0.043]; president[0.042]; saying[0.042]; January[0.041]; November[0.041]; various[0.041]; reporters[0.041]; press[0.041]; added[0.041]; election[0.041]; said[0.041]; said[0.041]; White[0.041]; White[0.041]; House[0.041]; House[0.041]; Friday[0.040]; news[0.040]; news[0.040]; news[0.040]; news[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clinton ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.144:0.144[0]; log p(e|m)= -1.917:-1.917[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinton[0.048]; Clinton[0.048]; President[0.046]; Washington[0.043]; president[0.042]; saying[0.042]; January[0.041]; November[0.041]; various[0.041]; reporters[0.041]; press[0.041]; added[0.041]; election[0.041]; said[0.041]; said[0.041]; White[0.041]; White[0.041]; House[0.041]; House[0.041]; Friday[0.040]; news[0.040]; news[0.040]; news[0.040]; news[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mike McCurry\u001b[39m ==> ENTITY: \u001b[32mMike McCurry (press secretary)\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinton[0.044]; Clinton[0.044]; Clinton[0.044]; reporters[0.043]; press[0.043]; Washington[0.043]; news[0.042]; news[0.042]; news[0.042]; news[0.042]; said[0.042]; said[0.042]; January[0.041]; saying[0.041]; President[0.041]; conferences[0.041]; conferences[0.041]; conferences[0.041]; conferences[0.041]; president[0.041]; plans[0.040]; regular[0.040]; asked[0.040]; Friday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = White House\u001b[39m ==> ENTITY: \u001b[32mWhite House\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.141:0.141[0]; log p(e|m)= -0.006:-0.006[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWashington[0.048]; President[0.047]; Clinton[0.045]; Clinton[0.045]; Clinton[0.045]; saying[0.045]; president[0.044]; January[0.044]; reporters[0.044]; election[0.043]; November[0.043]; said[0.043]; said[0.043]; Mike[0.043]; press[0.042]; House[0.042]; second[0.042]; second[0.042]; asked[0.042]; hold[0.042]; Friday[0.042]; plans[0.042]; White[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clinton ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.143:0.143[0]; log p(e|m)= -1.917:-1.917[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinton[0.052]; Clinton[0.052]; President[0.050]; Washington[0.047]; president[0.046]; saying[0.045]; January[0.045]; November[0.045]; various[0.045]; press[0.044]; election[0.044]; said[0.044]; said[0.044]; White[0.044]; White[0.044]; House[0.044]; House[0.044]; Friday[0.044]; news[0.044]; news[0.044]; news[0.044]; news[0.044]; \t\n\r [=================>............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s290ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1784/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1249testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tirana\u001b[39m ==> ENTITY: \u001b[32mTirana\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.120:0.120[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlbanian[0.045]; Albanian[0.045]; Berisha[0.045]; Berisha[0.045]; office[0.043]; outside[0.043]; outside[0.043]; conditions[0.042]; President[0.042]; President[0.042]; Friday[0.042]; palace[0.042]; April[0.041]; Italian[0.041]; hour[0.041]; said[0.041]; began[0.041]; Evaluating[0.040]; presidential[0.040]; unless[0.040]; arms[0.040]; jail[0.040]; meeting[0.040]; time[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sali Berisha\u001b[39m ==> ENTITY: \u001b[32mSali Berisha\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBerisha[0.051]; Albanian[0.047]; Albanian[0.047]; President[0.046]; President[0.046]; Tirana[0.046]; Tirana[0.046]; office[0.043]; denied[0.043]; presidential[0.043]; said[0.042]; jailed[0.042]; April[0.042]; charges[0.042]; sentenced[0.042]; meeting[0.042]; added[0.042]; Friday[0.042]; saying[0.041]; Italian[0.040]; Qazim[0.040]; Buza[0.040]; Buza[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tirana\u001b[39m ==> ENTITY: \u001b[32mTirana\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.135:0.135[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBerisha[0.044]; Berisha[0.044]; Medical[0.043]; court[0.042]; court[0.042]; office[0.042]; outside[0.042]; outside[0.042]; conditions[0.042]; added[0.042]; authorities[0.041]; President[0.041]; President[0.041]; Qazim[0.041]; Friday[0.041]; palace[0.041]; experts[0.041]; April[0.041]; committed[0.041]; Italian[0.041]; hour[0.040]; said[0.040]; said[0.040]; began[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Oscar Luigi Scalfaro\u001b[39m ==> ENTITY: \u001b[32mOscar Luigi Scalfaro\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItalian[0.046]; President[0.045]; President[0.045]; Medical[0.044]; office[0.044]; judge[0.042]; April[0.042]; presidential[0.042]; allowed[0.042]; mentally[0.041]; said[0.041]; said[0.041]; case[0.041]; denied[0.041]; sentence[0.041]; sentenced[0.041]; saying[0.041]; court[0.040]; court[0.040]; Buza[0.040]; Buza[0.040]; Buza[0.040]; responsible[0.040]; added[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Albanian ==> ENTITY: \u001b[32mAlbania\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.098:0.098[0]; log p(e|m)= -1.556:-1.556[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlbanian[0.050]; Italian[0.046]; guerrilla[0.044]; speak[0.043]; Berisha[0.043]; Berisha[0.043]; Tirana[0.042]; President[0.042]; President[0.042]; presidential[0.042]; riot[0.041]; said[0.041]; conditions[0.041]; April[0.041]; began[0.041]; threatened[0.041]; action[0.040]; action[0.040]; illegal[0.040]; thinks[0.040]; Friday[0.040]; lower[0.040]; jailed[0.040]; case[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italian ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.759:-0.759[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlbanian[0.044]; Albanian[0.044]; Luigi[0.044]; speak[0.044]; saying[0.043]; guerrilla[0.043]; Tirana[0.042]; Tirana[0.042]; April[0.042]; palace[0.042]; outside[0.042]; outside[0.042]; Berisha[0.041]; Berisha[0.041]; denied[0.041]; allowed[0.041]; committed[0.041]; said[0.041]; said[0.041]; authorities[0.040]; added[0.040]; President[0.040]; President[0.040]; Friday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Berisha\u001b[39m ==> ENTITY: \u001b[32mSali Berisha\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBerisha[0.048]; Albanian[0.045]; Albanian[0.045]; President[0.044]; President[0.044]; Tirana[0.044]; Tirana[0.044]; Medical[0.043]; office[0.041]; denied[0.041]; presidential[0.041]; said[0.040]; said[0.040]; jailed[0.040]; April[0.040]; charges[0.040]; sentenced[0.040]; committed[0.040]; meeting[0.040]; mentally[0.040]; authorities[0.040]; concluded[0.040]; added[0.040]; Friday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Albanian ==> ENTITY: \u001b[32mAlbania\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.097:0.097[0]; log p(e|m)= -1.556:-1.556[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlbanian[0.050]; Italian[0.046]; guerrilla[0.044]; speak[0.043]; Berisha[0.043]; Berisha[0.043]; Tirana[0.042]; President[0.042]; President[0.042]; presidential[0.042]; riot[0.041]; said[0.041]; conditions[0.041]; April[0.041]; began[0.041]; threatened[0.041]; action[0.040]; action[0.040]; illegal[0.040]; Friday[0.040]; jailed[0.040]; outside[0.040]; outside[0.040]; Luigi[0.039]; \t\n\r [=================>............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s267ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1792/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1353testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Brian Shimer\u001b[39m ==> ENTITY: \u001b[32mBrian Shimer\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.046]; Shimer[0.045]; Shimer[0.045]; man[0.044]; bobsleigh[0.044]; Lueders[0.043]; Cup[0.043]; event[0.043]; Americans[0.042]; race[0.042]; Bobsleigh[0.042]; World[0.041]; World[0.041]; Igls[0.041]; fifth[0.040]; Dave[0.040]; time[0.040]; trip[0.040]; second[0.040]; second[0.040]; second[0.040]; seconds[0.039]; finished[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.165:0.165[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItalians[0.046]; fifth[0.045]; France[0.045]; second[0.045]; second[0.045]; second[0.045]; second[0.045]; finished[0.044]; Austria[0.044]; Germany[0.044]; Cup[0.044]; Cup[0.044]; Olympic[0.043]; event[0.043]; victory[0.043]; Canada[0.043]; points[0.043]; World[0.042]; World[0.042]; World[0.042]; place[0.042]; time[0.042]; represented[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.113:0.113[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.045]; Germany[0.044]; Olympic[0.043]; second[0.043]; second[0.043]; second[0.043]; second[0.043]; Cup[0.043]; Canada[0.043]; standings[0.042]; ahead[0.042]; event[0.041]; finished[0.041]; points[0.041]; place[0.041]; Italians[0.041]; represented[0.040]; cup[0.040]; World[0.040]; World[0.040]; Americans[0.040]; Canadians[0.040]; time[0.040]; lead[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Altenberg ==> ENTITY: \u001b[32mAltenberg, Saxony\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.172:0.172[0]; log p(e|m)= -0.167:-0.167[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.053]; Lueders[0.047]; place[0.047]; Olympic[0.047]; Usa[0.046]; event[0.046]; Cup[0.046]; time[0.045]; ahead[0.045]; events[0.045]; World[0.045]; World[0.045]; Plagne[0.045]; opening[0.045]; near[0.045]; Canada[0.045]; France[0.044]; Saturday[0.044]; Saturday[0.044]; winter[0.044]; Italy[0.044]; race[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Americans ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.050:0.050[0]; log p(e|m)= -0.559:-0.559[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.049]; bobsleigh[0.046]; Lueders[0.044]; Canada[0.044]; Win[0.044]; finished[0.043]; second[0.043]; second[0.043]; second[0.043]; second[0.043]; Igls[0.043]; World[0.043]; World[0.043]; World[0.043]; event[0.043]; Brian[0.043]; Jim[0.043]; Bobsleigh[0.043]; Jones[0.043]; Italy[0.042]; fifth[0.042]; seconds[0.042]; points[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Guenther Huber\u001b[39m ==> ENTITY: \u001b[32mGünther Huber\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbobsleigh[0.045]; Lueders[0.044]; Olympic[0.043]; event[0.043]; Bobsleigh[0.043]; World[0.042]; World[0.042]; World[0.042]; man[0.042]; Igls[0.042]; Americans[0.042]; Altenberg[0.042]; events[0.042]; Shimer[0.041]; Shimer[0.041]; Shimer[0.041]; race[0.041]; Italians[0.041]; Cup[0.041]; Cup[0.041]; Canada[0.040]; Canadians[0.040]; standings[0.040]; place[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.100:0.100[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.044]; Olympic[0.044]; France[0.044]; fifth[0.043]; second[0.043]; second[0.043]; second[0.043]; second[0.043]; Canada[0.043]; hundredth[0.042]; standings[0.042]; Cup[0.042]; World[0.042]; World[0.042]; time[0.041]; place[0.040]; event[0.040]; opening[0.040]; Canadians[0.040]; aggregate[0.040]; minute[0.040]; run[0.040]; run[0.040]; Americans[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pierre Lueders\u001b[39m ==> ENTITY: \u001b[32mPierre Lueders\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.054]; bobsleigh[0.052]; Canada[0.052]; Canadians[0.049]; Igls[0.048]; place[0.048]; race[0.048]; winter[0.047]; finished[0.047]; Dave[0.047]; Cup[0.046]; Cup[0.046]; World[0.046]; World[0.046]; World[0.046]; fifth[0.046]; Shimer[0.046]; Shimer[0.046]; time[0.046]; event[0.046]; Plagne[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.113:0.113[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.045]; World[0.045]; World[0.045]; Italy[0.044]; Olympic[0.044]; represented[0.043]; victory[0.043]; finished[0.042]; Canada[0.042]; Americans[0.042]; second[0.041]; second[0.041]; second[0.041]; Igls[0.041]; fifth[0.041]; bobsleigh[0.040]; Bobsleigh[0.040]; Saturday[0.040]; Surprise[0.040]; Randy[0.040]; race[0.039]; minute[0.039]; trip[0.039]; Huber[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Randy Jones ==> ENTITY: \u001b[32mRandy Jones (bobsleigh)\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.182:0.182[0]; log p(e|m)= -1.224:-1.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.046]; bobsleigh[0.043]; Lueders[0.043]; Saturday[0.043]; Canada[0.042]; second[0.042]; second[0.042]; second[0.042]; second[0.042]; Igls[0.042]; Dave[0.042]; event[0.042]; fifth[0.042]; Surprise[0.041]; Brian[0.041]; time[0.041]; World[0.041]; World[0.041]; World[0.041]; race[0.041]; events[0.040]; Shimer[0.040]; Shimer[0.040]; Shimer[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = La Plagne\u001b[39m ==> ENTITY: \u001b[32mLa Plagne\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.121:0.121[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.047]; winter[0.044]; Lueders[0.044]; run[0.043]; run[0.043]; near[0.043]; France[0.042]; Canada[0.042]; Pierre[0.042]; place[0.042]; events[0.042]; Saturday[0.041]; event[0.041]; Guenther[0.041]; Italy[0.041]; course[0.041]; World[0.041]; World[0.041]; sleigh[0.041]; ahead[0.040]; Germany[0.040]; opening[0.040]; Canadians[0.040]; Altenberg[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dave MacEachern\u001b[39m ==> ENTITY: \u001b[32mDavid MacEachern\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.209:0.209[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.047]; bobsleigh[0.047]; Canadians[0.046]; Lueders[0.045]; Igls[0.044]; Olympic[0.044]; man[0.043]; Altenberg[0.042]; Shimer[0.041]; Shimer[0.041]; event[0.041]; World[0.041]; World[0.041]; World[0.041]; race[0.040]; Pierre[0.040]; Austria[0.040]; Italy[0.040]; place[0.039]; sleigh[0.039]; Lying[0.039]; Brian[0.039]; Jim[0.039]; Italians[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Antonio Tartaglia\u001b[39m ==> ENTITY: \u001b[32mAntonio Tartaglia\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbobsleigh[0.048]; event[0.047]; Lueders[0.046]; man[0.045]; Igls[0.045]; Olympic[0.045]; Huber[0.044]; Italy[0.044]; Canada[0.044]; race[0.044]; Bobsleigh[0.043]; Altenberg[0.043]; World[0.042]; World[0.042]; World[0.042]; Austria[0.042]; events[0.042]; Shimer[0.042]; Shimer[0.042]; Shimer[0.042]; standings[0.042]; Italians[0.042]; Plagne[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Garrett Hines\u001b[39m ==> ENTITY: \u001b[32mGarrett Hines\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.093:0.093[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLueders[0.045]; event[0.044]; winter[0.044]; Jim[0.043]; World[0.042]; World[0.042]; place[0.042]; Cup[0.042]; Plagne[0.042]; Canada[0.042]; Guenther[0.042]; events[0.042]; Saturday[0.042]; opening[0.041]; Altenberg[0.041]; second[0.041]; second[0.041]; second[0.041]; finished[0.041]; completed[0.040]; Germany[0.040]; Dave[0.040]; Americans[0.040]; Italy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shimer\u001b[39m ==> ENTITY: \u001b[32mBrian Shimer\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.046]; Shimer[0.045]; Shimer[0.045]; man[0.044]; bobsleigh[0.044]; Lueders[0.043]; Cup[0.043]; event[0.043]; Americans[0.042]; race[0.042]; Bobsleigh[0.042]; events[0.042]; Brian[0.041]; World[0.041]; World[0.041]; Igls[0.041]; fifth[0.040]; Dave[0.040]; time[0.040]; trip[0.040]; second[0.039]; second[0.039]; second[0.039]; second[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italians ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.160:0.160[0]; log p(e|m)= -1.556:-1.556[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.053]; Olympic[0.044]; fifth[0.043]; time[0.043]; France[0.043]; finished[0.042]; second[0.042]; second[0.042]; second[0.042]; second[0.042]; Germany[0.041]; event[0.041]; Canada[0.041]; bobsleigh[0.041]; points[0.041]; Antonio[0.041]; hundredth[0.041]; place[0.041]; Canadians[0.040]; represented[0.040]; Americans[0.040]; events[0.040]; standings[0.040]; winter[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canadians ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.096:0.096[0]; log p(e|m)= -0.819:-0.819[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.053]; Americans[0.046]; Italians[0.045]; second[0.043]; second[0.043]; second[0.043]; second[0.043]; Cup[0.042]; Cup[0.042]; Olympic[0.042]; bobsleigh[0.041]; Dave[0.041]; fifth[0.041]; Altenberg[0.040]; Italy[0.040]; France[0.040]; World[0.039]; World[0.039]; World[0.039]; Lueders[0.039]; finished[0.039]; Jim[0.039]; Garrett[0.039]; time[0.039]; \t\n\r [=================>............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s196ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1810/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1179testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Indies ==> ENTITY: \u001b[32mWest Indies cricket team\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.182:0.182[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.045]; Australia[0.045]; Australia[0.045]; Chanderpaul[0.044]; Shivnarine[0.044]; match[0.043]; Cricket[0.043]; Cricket[0.043]; overs[0.043]; overs[0.043]; overs[0.043]; Indies[0.042]; Indies[0.042]; wickets[0.041]; World[0.041]; beat[0.040]; West[0.040]; West[0.040]; Melbourne[0.039]; Melbourne[0.039]; Friday[0.039]; Greg[0.039]; Ground[0.039]; Blewett[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.187:0.187[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.050]; Australia[0.050]; match[0.045]; Cricket[0.045]; Cricket[0.045]; Melbourne[0.043]; Melbourne[0.043]; Shivnarine[0.042]; overs[0.042]; overs[0.042]; overs[0.042]; wickets[0.041]; World[0.040]; West[0.040]; West[0.040]; West[0.040]; Indies[0.040]; Indies[0.040]; Indies[0.040]; Ground[0.039]; Friday[0.039]; Chanderpaul[0.038]; Wickets[0.038]; Scores[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Greg Blewett\u001b[39m ==> ENTITY: \u001b[32mGreg Blewett\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.047]; Australia[0.047]; Australia[0.047]; match[0.046]; Cricket[0.044]; Cricket[0.044]; Shivnarine[0.043]; overs[0.043]; overs[0.043]; overs[0.043]; wickets[0.042]; Chanderpaul[0.042]; Melbourne[0.040]; Melbourne[0.040]; limited[0.040]; Series[0.040]; Scores[0.039]; Ground[0.039]; Wickets[0.039]; Indies[0.038]; Indies[0.038]; Indies[0.038]; beat[0.038]; World[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Indies ==> ENTITY: \u001b[32mWest Indies cricket team\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.182:0.182[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.045]; Australia[0.045]; Australia[0.045]; Chanderpaul[0.044]; Shivnarine[0.044]; match[0.043]; Cricket[0.043]; Cricket[0.043]; overs[0.043]; overs[0.043]; overs[0.043]; Indies[0.042]; Indies[0.042]; wickets[0.041]; World[0.041]; beat[0.040]; West[0.040]; West[0.040]; Melbourne[0.039]; Melbourne[0.039]; Friday[0.039]; Greg[0.039]; Ground[0.039]; Blewett[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Melbourne Cricket Ground\u001b[39m ==> ENTITY: \u001b[32mMelbourne Cricket Ground\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.049]; Australia[0.048]; Australia[0.048]; Australia[0.048]; Cricket[0.046]; Melbourne[0.045]; Friday[0.042]; overs[0.041]; overs[0.041]; overs[0.041]; beat[0.040]; Series[0.040]; limited[0.040]; West[0.040]; West[0.040]; West[0.040]; World[0.040]; Shivnarine[0.039]; wickets[0.039]; Indies[0.039]; Indies[0.039]; Indies[0.039]; Scores[0.039]; Wickets[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shivnarine Chanderpaul\u001b[39m ==> ENTITY: \u001b[32mShivnarine Chanderpaul\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwickets[0.046]; overs[0.046]; overs[0.046]; overs[0.046]; match[0.046]; Cricket[0.045]; Cricket[0.045]; Australia[0.044]; Australia[0.044]; Australia[0.044]; Indies[0.044]; Indies[0.044]; Indies[0.044]; Wickets[0.043]; West[0.043]; West[0.043]; West[0.043]; Blewett[0.042]; beat[0.042]; limited[0.041]; Series[0.041]; Greg[0.041]; Friday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Melbourne ==> ENTITY: \u001b[32mMelbourne\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.113:0.113[0]; log p(e|m)= -0.168:-0.168[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.051]; Australia[0.051]; Australia[0.051]; match[0.050]; Melbourne[0.050]; Cricket[0.046]; Cricket[0.046]; Greg[0.042]; Friday[0.042]; beat[0.042]; Ground[0.042]; overs[0.041]; overs[0.041]; overs[0.041]; West[0.041]; West[0.041]; West[0.041]; Series[0.041]; limited[0.040]; Scores[0.040]; World[0.040]; Shivnarine[0.040]; wickets[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWorld Series\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustralian Tri-Series\u001b[39m <---> \u001b[32mWorld Series Cricket\u001b[39m\t\nSCORES: global= 0.259:0.256[\u001b[31m0.003\u001b[39m]; local()= 0.170:0.150[\u001b[31m0.020\u001b[39m]; log p(e|m)= -3.324:-3.324[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.050]; Australia[0.050]; Australia[0.050]; Cricket[0.048]; Cricket[0.048]; match[0.046]; Melbourne[0.044]; Melbourne[0.044]; Friday[0.043]; Shivnarine[0.043]; overs[0.042]; overs[0.042]; overs[0.042]; beat[0.042]; wickets[0.042]; limited[0.042]; Blewett[0.041]; Ground[0.041]; Indies[0.040]; Indies[0.040]; Indies[0.040]; Greg[0.040]; Wickets[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Indies ==> ENTITY: \u001b[32mWest Indies cricket team\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.182:0.182[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.045]; Australia[0.045]; Australia[0.045]; Chanderpaul[0.044]; Shivnarine[0.044]; match[0.043]; Cricket[0.043]; Cricket[0.043]; overs[0.043]; overs[0.043]; overs[0.043]; Indies[0.042]; Indies[0.042]; wickets[0.041]; World[0.041]; beat[0.040]; West[0.040]; West[0.040]; Melbourne[0.039]; Melbourne[0.039]; Friday[0.039]; Greg[0.039]; Ground[0.039]; Blewett[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.187:0.187[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.050]; Australia[0.050]; match[0.045]; Cricket[0.045]; Cricket[0.045]; Melbourne[0.043]; Melbourne[0.043]; Shivnarine[0.042]; overs[0.042]; overs[0.042]; overs[0.042]; wickets[0.041]; World[0.040]; West[0.040]; West[0.040]; West[0.040]; Indies[0.040]; Indies[0.040]; Indies[0.040]; Ground[0.039]; Friday[0.039]; Chanderpaul[0.038]; Wickets[0.038]; Scores[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.187:0.187[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.050]; Australia[0.050]; match[0.045]; Cricket[0.045]; Cricket[0.045]; Melbourne[0.043]; Melbourne[0.043]; Shivnarine[0.042]; overs[0.042]; overs[0.042]; overs[0.042]; wickets[0.041]; World[0.040]; West[0.040]; West[0.040]; West[0.040]; Indies[0.040]; Indies[0.040]; Indies[0.040]; Ground[0.039]; Friday[0.039]; Chanderpaul[0.038]; Wickets[0.038]; Scores[0.038]; \t\n\r [=================>............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s173ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1821/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1197testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Paris ==> ENTITY: \u001b[32mParis\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.119:0.119[0]; log p(e|m)= -0.115:-0.115[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nParis[0.048]; matches[0.046]; Soccer[0.045]; won[0.044]; division[0.043]; League[0.043]; Bastia[0.043]; Guingamp[0.043]; Friday[0.043]; played[0.042]; French[0.041]; French[0.041]; Germain[0.041]; Monaco[0.041]; lost[0.041]; Saint[0.041]; points[0.040]; Strasbourg[0.040]; Bordeaux[0.040]; Nantes[0.040]; Metz[0.039]; Auxerre[0.039]; goals[0.038]; Lille[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bastia ==> ENTITY: \u001b[32mSC Bastia\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.229:0.229[0]; log p(e|m)= -0.703:-0.703[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.047]; League[0.046]; Guingamp[0.046]; Rennes[0.045]; Montpellier[0.045]; Marseille[0.044]; Nantes[0.044]; Auxerre[0.044]; division[0.044]; Soccer[0.044]; French[0.044]; French[0.044]; played[0.043]; won[0.043]; Lille[0.043]; Lens[0.043]; Monaco[0.043]; Bordeaux[0.043]; Paris[0.042]; Paris[0.042]; Strasbourg[0.041]; lost[0.041]; Friday[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nancy ==> ENTITY: \u001b[32mAS Nancy\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.245:0.245[0]; log p(e|m)= -1.687:-1.687[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGuingamp[0.080]; Rennes[0.075]; Nantes[0.075]; Montpellier[0.074]; Marseille[0.073]; Lille[0.073]; Caen[0.073]; Lyon[0.073]; Le[0.072]; Lens[0.071]; Nice[0.070]; Metz[0.069]; Cannes[0.062]; Havre[0.059]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lens ==> ENTITY: \u001b[32mRC Lens\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.238:0.238[0]; log p(e|m)= -0.860:-0.860[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGuingamp[0.057]; Rennes[0.057]; Bastia[0.056]; Marseille[0.056]; Nantes[0.055]; Montpellier[0.054]; Lille[0.054]; Caen[0.054]; Auxerre[0.054]; Lyon[0.053]; Le[0.053]; Bordeaux[0.053]; Monaco[0.052]; Strasbourg[0.051]; Nice[0.051]; Metz[0.050]; Nancy[0.047]; Cannes[0.046]; Havre[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Marseille ==> ENTITY: \u001b[32mOlympique de Marseille\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.194:0.194[0]; log p(e|m)= -1.435:-1.435[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.048]; Lyon[0.045]; Guingamp[0.045]; Paris[0.043]; won[0.043]; played[0.043]; Lille[0.043]; Nantes[0.042]; Montpellier[0.042]; Rennes[0.042]; Bastia[0.042]; Bordeaux[0.041]; French[0.041]; Monaco[0.041]; Auxerre[0.041]; Nice[0.041]; Lens[0.040]; Caen[0.040]; points[0.040]; Nancy[0.040]; Friday[0.040]; goals[0.040]; Saint[0.039]; Strasbourg[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Auxerre ==> ENTITY: \u001b[32mAJ Auxerre\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.203:0.203[0]; log p(e|m)= -0.305:-0.305[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.044]; Nantes[0.044]; Rennes[0.044]; Bastia[0.043]; Soccer[0.043]; Montpellier[0.043]; Bordeaux[0.043]; Guingamp[0.043]; Lille[0.042]; League[0.042]; Caen[0.042]; Marseille[0.042]; Saint[0.041]; French[0.041]; French[0.041]; Paris[0.041]; Paris[0.041]; Havre[0.041]; Lens[0.040]; Le[0.040]; Lyon[0.040]; won[0.040]; played[0.039]; Strasbourg[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bordeaux ==> ENTITY: \u001b[32mFC Girondins de Bordeaux\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.205:0.205[0]; log p(e|m)= -1.609:-1.609[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGuingamp[0.045]; matches[0.044]; Soccer[0.044]; Lille[0.043]; League[0.043]; Marseille[0.043]; French[0.043]; French[0.043]; Lyon[0.042]; Bastia[0.042]; Rennes[0.042]; Nantes[0.042]; Paris[0.042]; Paris[0.042]; Auxerre[0.042]; won[0.041]; Monaco[0.041]; goals[0.040]; Lens[0.040]; played[0.040]; Le[0.040]; Saint[0.039]; Metz[0.039]; division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cannes ==> ENTITY: \u001b[32mAS Cannes\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.232:0.232[0]; log p(e|m)= -2.198:-2.198[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBastia[0.066]; Guingamp[0.065]; Rennes[0.063]; Montpellier[0.061]; Lyon[0.061]; Auxerre[0.061]; Marseille[0.061]; Nantes[0.061]; Lille[0.060]; Lens[0.059]; Caen[0.059]; Metz[0.057]; Nice[0.056]; Strasbourg[0.056]; Le[0.053]; Nancy[0.053]; Havre[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lyon ==> ENTITY: \u001b[32mOlympique Lyonnais\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.220:0.220[0]; log p(e|m)= -1.917:-1.917[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nParis[0.046]; Guingamp[0.045]; Marseille[0.044]; Bastia[0.044]; Rennes[0.044]; Auxerre[0.043]; Lens[0.043]; Monaco[0.042]; Lille[0.042]; Bordeaux[0.042]; Le[0.042]; Nantes[0.042]; Montpellier[0.041]; goals[0.041]; played[0.041]; Strasbourg[0.041]; won[0.040]; Saint[0.040]; Caen[0.040]; points[0.040]; Metz[0.040]; Nice[0.039]; Germain[0.039]; Nancy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Metz ==> ENTITY: \u001b[32mFC Metz\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.218:0.218[0]; log p(e|m)= -1.924:-1.924[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; Guingamp[0.045]; Rennes[0.043]; Soccer[0.043]; Bastia[0.043]; matches[0.043]; Cannes[0.042]; Marseille[0.042]; won[0.042]; Lille[0.041]; Caen[0.041]; Friday[0.041]; Auxerre[0.041]; Bordeaux[0.041]; division[0.041]; Nantes[0.041]; Montpellier[0.041]; played[0.041]; Paris[0.041]; Paris[0.041]; Lyon[0.040]; Lens[0.040]; Nancy[0.040]; Le[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.141:0.141[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrench[0.050]; matches[0.048]; Soccer[0.047]; League[0.047]; Paris[0.046]; Paris[0.046]; Guingamp[0.046]; Monaco[0.045]; played[0.044]; Nantes[0.044]; Saint[0.044]; won[0.043]; Strasbourg[0.042]; Bordeaux[0.042]; goals[0.042]; Auxerre[0.042]; Bastia[0.042]; Friday[0.041]; division[0.040]; Germain[0.040]; lost[0.040]; Metz[0.039]; points[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paris Saint-Germain\u001b[39m ==> ENTITY: \u001b[32mParis Saint-Germain F.C.\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.225:0.225[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermain[0.046]; Paris[0.045]; Soccer[0.044]; League[0.044]; Guingamp[0.044]; Marseille[0.043]; matches[0.043]; Rennes[0.043]; Lyon[0.042]; Bastia[0.042]; Nantes[0.042]; Strasbourg[0.041]; Monaco[0.041]; division[0.041]; Auxerre[0.041]; Bordeaux[0.041]; French[0.041]; French[0.041]; Lille[0.040]; played[0.039]; won[0.039]; goals[0.039]; Metz[0.038]; Friday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rennes ==> ENTITY: \u001b[32mStade Rennais F.C.\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.205:0.205[0]; log p(e|m)= -1.013:-1.013[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGuingamp[0.048]; Bastia[0.048]; Nantes[0.047]; Montpellier[0.046]; Marseille[0.045]; Paris[0.045]; Caen[0.044]; Lille[0.044]; Auxerre[0.044]; Le[0.044]; Lens[0.044]; Saint[0.044]; Lyon[0.044]; Bordeaux[0.044]; Nice[0.043]; Monaco[0.043]; Metz[0.042]; Strasbourg[0.042]; points[0.040]; Germain[0.040]; Havre[0.040]; Nancy[0.040]; Cannes[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nantes ==> ENTITY: \u001b[32mFC Nantes\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.263:0.263[0]; log p(e|m)= -1.457:-1.457[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGuingamp[0.046]; matches[0.045]; Marseille[0.044]; Rennes[0.043]; Bastia[0.043]; Montpellier[0.043]; League[0.043]; Auxerre[0.042]; Soccer[0.042]; played[0.042]; Lille[0.042]; Lyon[0.041]; Paris[0.041]; Paris[0.041]; Bordeaux[0.041]; French[0.041]; French[0.041]; Lens[0.040]; Strasbourg[0.040]; Monaco[0.040]; Caen[0.040]; Metz[0.040]; Nice[0.039]; division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Strasbourg ==> ENTITY: \u001b[32mRC Strasbourg\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.233:0.233[0]; log p(e|m)= -2.216:-2.216[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.045]; Guingamp[0.044]; matches[0.044]; Bastia[0.044]; Rennes[0.043]; Lyon[0.043]; Marseille[0.043]; Soccer[0.043]; Lens[0.043]; played[0.042]; Auxerre[0.042]; Monaco[0.041]; Nantes[0.041]; won[0.041]; Lille[0.041]; French[0.041]; French[0.041]; Paris[0.041]; Paris[0.041]; division[0.040]; Bordeaux[0.040]; Metz[0.039]; Friday[0.038]; lost[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Caen ==> ENTITY: \u001b[32mStade Malherbe Caen\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.210:0.210[0]; log p(e|m)= -1.614:-1.614[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRennes[0.075]; Nantes[0.073]; Guingamp[0.072]; Montpellier[0.070]; Marseille[0.069]; Lille[0.068]; Auxerre[0.068]; Nice[0.067]; Lens[0.066]; Lyon[0.065]; Le[0.064]; Metz[0.063]; Havre[0.063]; Nancy[0.059]; Cannes[0.058]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Le Havre ==> ENTITY: \u001b[32mLe Havre AC\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.245:0.245[0]; log p(e|m)= -1.347:-1.347[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGuingamp[0.063]; Nantes[0.063]; Rennes[0.061]; Bastia[0.061]; Auxerre[0.061]; Marseille[0.061]; Montpellier[0.061]; Caen[0.061]; Bordeaux[0.060]; Lille[0.059]; Cannes[0.059]; Lyon[0.058]; Strasbourg[0.057]; Lens[0.057]; Nice[0.056]; Metz[0.053]; Nancy[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrench[0.047]; matches[0.046]; Soccer[0.045]; League[0.044]; Paris[0.044]; Paris[0.044]; Marseille[0.044]; Guingamp[0.044]; Monaco[0.043]; played[0.042]; Nantes[0.042]; Saint[0.042]; won[0.041]; Lille[0.041]; Strasbourg[0.040]; Bordeaux[0.040]; goals[0.040]; Auxerre[0.040]; Bastia[0.040]; Friday[0.039]; division[0.038]; Germain[0.038]; lost[0.038]; Metz[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Guingamp\u001b[39m ==> ENTITY: \u001b[32mEn Avant de Guingamp\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.221:0.221[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRennes[0.046]; Bastia[0.044]; Montpellier[0.043]; Auxerre[0.043]; played[0.043]; Nantes[0.043]; matches[0.043]; Caen[0.042]; League[0.042]; Marseille[0.042]; Soccer[0.041]; Bordeaux[0.041]; division[0.041]; Lille[0.041]; won[0.041]; Havre[0.041]; French[0.041]; French[0.041]; Lens[0.040]; Nice[0.040]; Paris[0.040]; Paris[0.040]; Le[0.040]; Nancy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nice ==> ENTITY: \u001b[32mOGC Nice\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.230:0.230[0]; log p(e|m)= -2.137:-2.137[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGuingamp[0.084]; Rennes[0.082]; Marseille[0.082]; Nantes[0.081]; Lille[0.079]; Lens[0.079]; Montpellier[0.078]; Lyon[0.078]; Caen[0.077]; Le[0.074]; Cannes[0.071]; Nancy[0.070]; Havre[0.066]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Monaco ==> ENTITY: \u001b[32mAS Monaco FC\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.206:0.206[0]; log p(e|m)= -2.453:-2.453[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.047]; played[0.045]; Guingamp[0.044]; League[0.044]; Soccer[0.043]; Marseille[0.043]; Rennes[0.042]; Bastia[0.042]; Lyon[0.042]; Auxerre[0.042]; Nantes[0.042]; points[0.041]; Paris[0.041]; Paris[0.041]; won[0.041]; Lille[0.041]; Lens[0.041]; goals[0.040]; Bordeaux[0.040]; French[0.040]; French[0.040]; Standings[0.039]; Standings[0.039]; Germain[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Montpellier ==> ENTITY: \u001b[32mMontpellier HSC\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.234:0.234[0]; log p(e|m)= -1.478:-1.478[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGuingamp[0.069]; Bastia[0.068]; Marseille[0.067]; Rennes[0.067]; Nantes[0.067]; Lille[0.064]; Lyon[0.064]; Caen[0.063]; Lens[0.063]; Auxerre[0.063]; Le[0.061]; Nice[0.060]; Metz[0.059]; Cannes[0.056]; Nancy[0.056]; Havre[0.052]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lille ==> ENTITY: \u001b[32mLille OSC\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.233:0.233[0]; log p(e|m)= -1.351:-1.351[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; Guingamp[0.045]; League[0.044]; Marseille[0.043]; Rennes[0.043]; Bastia[0.042]; Paris[0.042]; Paris[0.042]; Auxerre[0.042]; Bordeaux[0.042]; Lyon[0.042]; Nantes[0.042]; Montpellier[0.041]; Caen[0.041]; played[0.041]; French[0.041]; Le[0.041]; Monaco[0.041]; Lens[0.041]; Strasbourg[0.040]; goals[0.040]; won[0.040]; Saint[0.039]; division[0.039]; \t\n\r [=================>............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s138ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1844/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1230testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canadian ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.056:0.056[0]; log p(e|m)= -0.412:-0.412[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nManitoba[0.048]; Manitoba[0.048]; dollars[0.048]; Winnipeg[0.046]; contract[0.043]; contract[0.043]; Contract[0.042]; minimum[0.042]; price[0.041]; Range[0.041]; Pork[0.041]; Pork[0.041]; closing[0.041]; Previous[0.040]; Feb[0.040]; guaranteed[0.040]; Closing[0.040]; Minimum[0.040]; prices[0.039]; prices[0.039]; Close[0.039]; Close[0.039]; Close[0.039]; Apr[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = CAN ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.053:0.053[0]; log p(e|m)= -1.766:-1.766[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nManitoba[0.050]; Manitoba[0.050]; Manitoba[0.050]; Winnipeg[0.048]; Aug[0.044]; Source[0.044]; Sep[0.044]; Oct[0.044]; Jul[0.044]; Nov[0.044]; Dec[0.043]; Jun[0.043]; Note[0.043]; Pork[0.042]; Rate[0.042]; Exchange[0.042]; Price[0.042]; Price[0.042]; Dollar[0.041]; Government[0.041]; Index[0.040]; bureau[0.039]; Hog[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= -0.003:-0.003[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSource[0.044]; Price[0.044]; Price[0.044]; Exchange[0.044]; cwt[0.044]; cwt[0.044]; Index[0.043]; bureau[0.043]; Rate[0.042]; Note[0.042]; Dollar[0.042]; Government[0.042]; Dec[0.042]; Aug[0.041]; Oct[0.041]; Nov[0.041]; Pork[0.041]; Hog[0.041]; Sep[0.040]; Jul[0.040]; Range[0.040]; Manitoba[0.039]; Manitoba[0.039]; Manitoba[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Winnipeg ==> ENTITY: \u001b[32mWinnipeg\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.032:0.032[0]; log p(e|m)= -0.123:-0.123[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nExchange[0.050]; Manitoba[0.048]; Manitoba[0.048]; Manitoba[0.048]; Dec[0.042]; Source[0.042]; Nov[0.042]; Oct[0.042]; Dollar[0.042]; Sep[0.042]; Aug[0.041]; Range[0.041]; Government[0.041]; bureau[0.040]; Price[0.040]; Price[0.040]; Index[0.040]; Jul[0.040]; Pork[0.039]; Rate[0.039]; cwt[0.039]; cwt[0.039]; Hog[0.038]; Note[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Winnipeg ==> ENTITY: \u001b[32mWinnipeg\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.064:0.064[0]; log p(e|m)= -0.123:-0.123[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanadian[0.047]; Manitoba[0.046]; Manitoba[0.046]; contract[0.045]; contract[0.045]; guaranteed[0.043]; prices[0.042]; prices[0.042]; Contract[0.042]; forward[0.042]; forward[0.042]; Closing[0.041]; closing[0.041]; price[0.040]; Close[0.040]; Close[0.040]; Close[0.040]; Range[0.040]; lbs[0.040]; dollars[0.040]; Previous[0.039]; minimum[0.039]; Cst[0.039]; Feb[0.038]; \t\n\r [=================>............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s160ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1849/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1231testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Winnipeg ==> ENTITY: \u001b[32mWinnipeg\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.123:-0.123[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanadian[0.047]; Canadian[0.047]; Canadian[0.047]; Vancouver[0.046]; ships[0.043]; ships[0.043]; Friday[0.042]; seven[0.042]; East[0.042]; ship[0.042]; Thunder[0.041]; West[0.041]; West[0.041]; West[0.041]; Coast[0.040]; Coast[0.040]; Coast[0.040]; Coast[0.040]; Waiting[0.040]; waiting[0.040]; days[0.039]; Board[0.039]; longest[0.039]; Prince[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canadian ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.082:0.082[0]; log p(e|m)= -0.412:-0.412[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanadian[0.050]; Canadian[0.050]; Vancouver[0.045]; Winnipeg[0.044]; ships[0.044]; ships[0.044]; ship[0.044]; Bay[0.042]; West[0.042]; West[0.042]; West[0.042]; East[0.040]; seven[0.040]; Wheat[0.040]; Coast[0.040]; Coast[0.040]; Coast[0.040]; Coast[0.040]; days[0.039]; Vessel[0.039]; Prince[0.039]; Port[0.039]; load[0.038]; load[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Canadian Wheat Board\u001b[39m ==> ENTITY: \u001b[32mCanadian Wheat Board\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.059:0.059[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanadian[0.049]; Canadian[0.049]; Thunder[0.043]; load[0.042]; load[0.042]; Winnipeg[0.042]; Gras[0.042]; Bay[0.042]; Vancouver[0.042]; Loading[0.041]; Coast[0.041]; Coast[0.041]; Coast[0.041]; Coast[0.041]; wait[0.041]; loading[0.040]; West[0.040]; West[0.040]; West[0.040]; ship[0.040]; Port[0.040]; Friday[0.040]; waiting[0.040]; reported[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vancouver ==> ENTITY: \u001b[32mVancouver\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.081:0.081[0]; log p(e|m)= -0.121:-0.121[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanadian[0.046]; Canadian[0.046]; Canadian[0.046]; ship[0.045]; Port[0.044]; Bay[0.043]; ships[0.043]; ships[0.043]; Winnipeg[0.042]; Coast[0.041]; Coast[0.041]; Coast[0.041]; Coast[0.041]; West[0.040]; West[0.040]; West[0.040]; East[0.040]; loaded[0.039]; loaded[0.039]; Friday[0.039]; Prince[0.039]; days[0.039]; load[0.039]; load[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Thunder Bay ==> ENTITY: \u001b[32mThunder Bay\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.106:0.106[0]; log p(e|m)= -0.264:-0.264[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVancouver[0.044]; Canadian[0.044]; Canadian[0.044]; Canadian[0.044]; Port[0.044]; Winnipeg[0.043]; ships[0.042]; ships[0.042]; Prince[0.042]; Coast[0.042]; Coast[0.042]; Coast[0.042]; Coast[0.042]; Wheat[0.041]; East[0.041]; load[0.041]; load[0.041]; seven[0.041]; Friday[0.041]; ship[0.041]; West[0.040]; West[0.040]; West[0.040]; loading[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canadian ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.082:0.082[0]; log p(e|m)= -0.412:-0.412[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanadian[0.050]; Canadian[0.050]; Vancouver[0.045]; Winnipeg[0.044]; ships[0.044]; ships[0.044]; ship[0.044]; Bay[0.042]; West[0.042]; West[0.042]; West[0.042]; East[0.040]; seven[0.040]; Wheat[0.040]; Coast[0.040]; Coast[0.040]; Coast[0.040]; Coast[0.040]; days[0.039]; Vessel[0.039]; Prince[0.039]; Port[0.039]; load[0.038]; load[0.038]; \t\n\r [==================>...........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s166ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1855/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1247testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.046:0.046[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.065]; Russian[0.061]; December[0.056]; weather[0.056]; Friday[0.055]; said[0.052]; Weather[0.052]; Weather[0.052]; airports[0.052]; airports[0.052]; Service[0.052]; closures[0.052]; related[0.051]; Newsroom[0.050]; Conditions[0.050]; expected[0.049]; Dec[0.049]; Cis[0.047]; Cis[0.047]; \t\n\r [==================>...........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s211ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1856/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1345testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colorado ==> ENTITY: \u001b[32mColorado\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.025:0.025[0]; log p(e|m)= -0.269:-0.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.044]; Alpine[0.044]; Switzerland[0.044]; Vail[0.044]; minute[0.043]; Carole[0.042]; Russia[0.042]; Russia[0.042]; Stefanie[0.042]; Skiing[0.042]; Austria[0.042]; Martina[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; seconds[0.041]; Italy[0.041]; Italy[0.041]; Saturday[0.041]; World[0.041]; World[0.041]; France[0.040]; France[0.040]; women[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.044]; Switzerland[0.044]; Switzerland[0.044]; Switzerland[0.044]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Slovenia[0.042]; Italy[0.042]; Italy[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Norway[0.041]; Shannon[0.039]; France[0.039]; France[0.039]; France[0.039]; France[0.039]; Borghi[0.039]; official[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.260:0.260[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Switzerland[0.044]; Switzerland[0.044]; Switzerland[0.044]; Italy[0.043]; Italy[0.043]; Italy[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; Sweden[0.042]; Slovenia[0.042]; Martina[0.039]; minute[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Regine Cavagnoud\u001b[39m ==> ENTITY: \u001b[32mRégine Cavagnoud\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSeizinger[0.049]; Dorfmeister[0.049]; Austria[0.048]; Austria[0.048]; Austria[0.048]; Austria[0.048]; Austria[0.048]; Ingeborg[0.048]; Zurbriggen[0.048]; Switzerland[0.048]; Switzerland[0.048]; Switzerland[0.048]; Martina[0.048]; Monika[0.047]; Renate[0.046]; Wachter[0.046]; Stefanie[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.239:0.239[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.048]; France[0.048]; France[0.048]; France[0.048]; Switzerland[0.045]; Switzerland[0.045]; Italy[0.044]; Italy[0.044]; Italy[0.044]; Austria[0.043]; Austria[0.043]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Cup[0.042]; Cup[0.042]; Sweden[0.041]; Slovenia[0.041]; Women[0.040]; women[0.040]; Super[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Switzerland ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.171:0.171[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.051]; Switzerland[0.051]; Austria[0.045]; Austria[0.045]; Austria[0.045]; Austria[0.045]; Slovenia[0.044]; France[0.043]; France[0.043]; Norway[0.043]; official[0.042]; Monika[0.040]; results[0.039]; Marianne[0.039]; Hilary[0.039]; Cavagnoud[0.039]; Megan[0.039]; Catherine[0.039]; Renate[0.039]; Borghi[0.038]; Zurbriggen[0.038]; Michaela[0.038]; Alexandra[0.038]; Helen[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Isolde Kostner\u001b[39m ==> ENTITY: \u001b[32mIsolde Kostner\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.115:0.115[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.045]; Seizinger[0.043]; Cup[0.043]; Cup[0.043]; Hilde[0.042]; race[0.042]; Gerg[0.042]; Skiing[0.042]; Super[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Martina[0.042]; Vail[0.042]; World[0.041]; World[0.041]; Colorado[0.041]; Austria[0.041]; Austria[0.041]; Renate[0.040]; Stefanie[0.040]; Madlen[0.040]; Zurbriggen[0.040]; Florence[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Heidi Zurbriggen\u001b[39m ==> ENTITY: \u001b[32mHeidi Zurbriggen\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.047]; Switzerland[0.047]; Switzerland[0.047]; Switzerland[0.047]; Seizinger[0.047]; Martina[0.046]; Austria[0.046]; Austria[0.046]; Austria[0.046]; Austria[0.046]; Austria[0.046]; Ingeborg[0.045]; Helen[0.045]; Sybille[0.045]; Marianne[0.045]; Cavagnoud[0.044]; Anita[0.044]; Renate[0.044]; Hilary[0.044]; Lindh[0.044]; Monika[0.044]; Michaela[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.197:0.197[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.048]; France[0.048]; Switzerland[0.045]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Austria[0.042]; Austria[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Cup[0.042]; Cup[0.042]; Sweden[0.041]; Women[0.040]; women[0.040]; Super[0.039]; Russia[0.039]; Russia[0.039]; World[0.039]; World[0.039]; Saturday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.294:0.294[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.049]; Austria[0.049]; Switzerland[0.047]; Switzerland[0.047]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Slovenia[0.044]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Sweden[0.042]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; World[0.040]; World[0.040]; Russia[0.040]; Russia[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Switzerland ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.214:0.214[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.050]; Switzerland[0.050]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Italy[0.044]; Germany[0.043]; Germany[0.043]; Slovenia[0.043]; France[0.042]; France[0.042]; France[0.042]; Norway[0.042]; official[0.041]; Monika[0.039]; results[0.038]; Marianne[0.038]; Hilary[0.038]; Cavagnoud[0.038]; Megan[0.038]; Catherine[0.038]; Renate[0.038]; Borghi[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Michaela Dorfmeister\u001b[39m ==> ENTITY: \u001b[32mMichaela Dorfmeister\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Switzerland[0.043]; Switzerland[0.043]; Switzerland[0.043]; Zurbriggen[0.042]; Norway[0.042]; France[0.042]; France[0.042]; France[0.042]; Ingeborg[0.042]; Monika[0.041]; Renate[0.041]; Slovenia[0.041]; Germany[0.040]; Germany[0.040]; Alexandra[0.040]; Wachter[0.040]; Nobis[0.040]; Helen[0.040]; Italy[0.039]; Catherine[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Renate Goetschl\u001b[39m ==> ENTITY: \u001b[32mRenate Götschl\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKostner[0.047]; Seizinger[0.047]; Dorfmeister[0.047]; Switzerland[0.047]; Switzerland[0.047]; Switzerland[0.047]; Switzerland[0.047]; Gerg[0.046]; Hilde[0.046]; Norway[0.045]; Ingeborg[0.045]; Martina[0.045]; Ertl[0.045]; Zurbriggen[0.045]; Italy[0.044]; Italy[0.044]; Italy[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.240:0.240[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Austria[0.045]; Austria[0.045]; Austria[0.045]; Switzerland[0.044]; Switzerland[0.044]; Italy[0.043]; Italy[0.043]; Italy[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; Sweden[0.042]; Slovenia[0.042]; Cup[0.040]; Cup[0.040]; women[0.040]; Martina[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.049]; Slovenia[0.045]; Switzerland[0.045]; Switzerland[0.045]; Austria[0.045]; Austria[0.045]; Sweden[0.045]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Cup[0.044]; Cup[0.044]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Svetlana[0.042]; Provisional[0.042]; France[0.042]; France[0.042]; France[0.042]; France[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.223:0.223[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.047]; France[0.047]; France[0.047]; Switzerland[0.044]; Switzerland[0.044]; Switzerland[0.044]; Italy[0.043]; Italy[0.043]; Martina[0.042]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Slovenia[0.040]; Norway[0.039]; Stefanie[0.038]; Catherine[0.038]; Monika[0.038]; Michaela[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hilde Gerg\u001b[39m ==> ENTITY: \u001b[32mHilde Gerg\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSkiing[0.046]; Super[0.043]; Seizinger[0.043]; Alpine[0.043]; Cup[0.043]; Cup[0.043]; Austria[0.042]; Austria[0.042]; Martina[0.042]; Vail[0.042]; World[0.042]; World[0.042]; Kostner[0.041]; Colorado[0.041]; Sweden[0.041]; race[0.041]; Renate[0.041]; Provisional[0.041]; Ertl[0.040]; Stefanie[0.040]; Madlen[0.040]; Wiberg[0.040]; Switzerland[0.040]; Results[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Katharina Gutensohn\u001b[39m ==> ENTITY: \u001b[32mKatharina Gutensohn\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Switzerland[0.042]; Switzerland[0.042]; Switzerland[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Seizinger[0.041]; Cavagnoud[0.041]; Hilde[0.041]; Gerg[0.041]; Kostner[0.040]; Ingeborg[0.040]; Sweden[0.040]; Martina[0.040]; Katja[0.039]; Renate[0.039]; Dorfmeister[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Anita Wachter\u001b[39m ==> ENTITY: \u001b[32mAnita Wachter\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Switzerland[0.043]; Switzerland[0.043]; Switzerland[0.043]; Seizinger[0.042]; Norway[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Sybille[0.041]; Ingeborg[0.041]; Renate[0.041]; Dorfmeister[0.041]; Martina[0.041]; Marianne[0.040]; Monika[0.040]; Cavagnoud[0.040]; Stefanie[0.040]; Zurbriggen[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.129:0.129[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.047]; Switzerland[0.044]; Austria[0.044]; Sweden[0.044]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Cup[0.043]; Cup[0.043]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Svetlana[0.041]; Provisional[0.041]; France[0.041]; France[0.041]; World[0.040]; World[0.040]; Saturday[0.040]; Women[0.040]; Alpine[0.040]; race[0.039]; women[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Switzerland ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.286:0.286[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.047]; Switzerland[0.047]; Switzerland[0.047]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Russia[0.041]; Slovenia[0.041]; France[0.040]; France[0.040]; France[0.040]; France[0.040]; Norway[0.039]; official[0.039]; Martina[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.254:0.254[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.052]; Italy[0.052]; Cup[0.042]; Switzerland[0.042]; Switzerland[0.042]; Slovenia[0.042]; France[0.042]; France[0.042]; France[0.042]; France[0.042]; France[0.042]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Sweden[0.039]; Martina[0.039]; World[0.038]; Stefanie[0.038]; Florence[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Florence Masnada\u001b[39m ==> ENTITY: \u001b[32mFlorence Masnada\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.046]; Kostner[0.043]; Seizinger[0.042]; Cavagnoud[0.042]; Switzerland[0.042]; Switzerland[0.042]; Zurbriggen[0.042]; Hilde[0.042]; France[0.042]; France[0.042]; France[0.042]; France[0.042]; France[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Gerg[0.041]; Martina[0.041]; Vail[0.041]; Austria[0.041]; Austria[0.041]; Slovenia[0.041]; Skiing[0.041]; Colorado[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.233:0.233[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.048]; Austria[0.048]; Austria[0.048]; Switzerland[0.046]; Switzerland[0.046]; Switzerland[0.046]; Germany[0.046]; Germany[0.046]; Slovenia[0.044]; Norway[0.042]; France[0.040]; France[0.040]; France[0.040]; Monika[0.039]; declared[0.039]; Sybille[0.039]; Renate[0.039]; Katharina[0.039]; Cavagnoud[0.038]; official[0.037]; Borghi[0.037]; Heidi[0.037]; Dorfmeister[0.037]; Zurbriggen[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.210:0.210[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Austria[0.043]; Austria[0.043]; Switzerland[0.043]; Italy[0.042]; Italy[0.042]; Italy[0.042]; France[0.042]; France[0.042]; France[0.042]; France[0.042]; Sweden[0.041]; Cup[0.040]; Cup[0.040]; women[0.039]; Russia[0.039]; Russia[0.039]; World[0.039]; World[0.039]; Kostner[0.038]; Women[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.326:0.326[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.045]; Austria[0.045]; Austria[0.045]; Austria[0.045]; Switzerland[0.044]; Switzerland[0.044]; Switzerland[0.044]; Switzerland[0.044]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Slovenia[0.041]; Norway[0.040]; Italy[0.040]; Italy[0.040]; Italy[0.040]; France[0.038]; France[0.038]; France[0.038]; France[0.038]; France[0.038]; Russia[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alexandra Meissnitzer\u001b[39m ==> ENTITY: \u001b[32mAlexandra Meissnitzer\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.045]; Switzerland[0.045]; Switzerland[0.045]; Norway[0.044]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Germany[0.042]; Germany[0.042]; Ingeborg[0.042]; Cavagnoud[0.041]; Dorfmeister[0.041]; Monika[0.041]; Renate[0.040]; France[0.040]; France[0.040]; France[0.040]; Wachter[0.040]; Zurbriggen[0.040]; Nobis[0.040]; Katharina[0.040]; Michaela[0.040]; Helen[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Martina Ertl\u001b[39m ==> ENTITY: \u001b[32mMartina Ertl-Renz\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.047]; Austria[0.046]; Austria[0.046]; Austria[0.046]; Skiing[0.045]; World[0.044]; World[0.044]; Kostner[0.043]; Seizinger[0.043]; Cup[0.043]; Cup[0.043]; seconds[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Hilde[0.042]; race[0.042]; Sweden[0.042]; Gerg[0.042]; Vail[0.042]; Super[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.280:0.280[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.046]; Austria[0.046]; Austria[0.046]; Austria[0.046]; Switzerland[0.045]; Switzerland[0.045]; Switzerland[0.045]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Slovenia[0.042]; Norway[0.041]; Italy[0.041]; Italy[0.041]; France[0.039]; France[0.039]; France[0.039]; Ertl[0.038]; Stefanie[0.037]; Schuster[0.037]; Monika[0.037]; Sybille[0.037]; Renate[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.081:0.081[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.045]; Switzerland[0.045]; Switzerland[0.045]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Slovenia[0.043]; Italy[0.043]; Italy[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Norway[0.042]; Shannon[0.040]; France[0.040]; France[0.040]; France[0.040]; Borghi[0.040]; official[0.039]; results[0.039]; Perez[0.039]; Stefanie[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.261:0.261[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Austria[0.042]; Switzerland[0.042]; Switzerland[0.042]; Switzerland[0.042]; Italy[0.041]; Italy[0.041]; Italy[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; Sweden[0.040]; Slovenia[0.040]; Martina[0.038]; minute[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Katja Seizinger\u001b[39m ==> ENTITY: \u001b[32mKatja Seizinger\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrace[0.049]; Cup[0.047]; Cup[0.047]; Kostner[0.047]; Alpine[0.046]; Zurbriggen[0.046]; Gerg[0.046]; women[0.046]; results[0.045]; Hilde[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Martina[0.045]; Vail[0.045]; Skiing[0.044]; Colorado[0.044]; Renate[0.044]; Switzerland[0.044]; Switzerland[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.240:0.240[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.052]; Italy[0.052]; Cup[0.043]; Cup[0.043]; Switzerland[0.042]; Switzerland[0.042]; France[0.042]; France[0.042]; France[0.042]; France[0.042]; Austria[0.041]; Austria[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Super[0.041]; Sweden[0.039]; Martina[0.039]; World[0.038]; World[0.038]; Stefanie[0.038]; Florence[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Slovenia ==> ENTITY: \u001b[32mSlovenia\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.236:0.236[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.044]; Italy[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Austria[0.044]; Norway[0.043]; Switzerland[0.043]; Switzerland[0.043]; Switzerland[0.043]; Switzerland[0.043]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Russia[0.040]; France[0.039]; France[0.039]; France[0.039]; France[0.039]; Martina[0.038]; official[0.038]; Ertl[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vail ==> ENTITY: \u001b[32mVail, Colorado\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.119:0.119[0]; log p(e|m)= -0.392:-0.392[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.050]; Skiing[0.047]; Saturday[0.047]; race[0.046]; Colorado[0.046]; women[0.046]; Switzerland[0.044]; Women[0.043]; World[0.043]; World[0.043]; Cup[0.043]; Cup[0.043]; Stefanie[0.043]; Austria[0.042]; Carole[0.042]; seconds[0.042]; Sweden[0.042]; minute[0.042]; results[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Florence[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.146:0.146[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.049]; Austria[0.048]; women[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Italy[0.046]; Italy[0.046]; Italy[0.046]; Cup[0.045]; Cup[0.045]; Women[0.045]; Russia[0.044]; Russia[0.044]; World[0.044]; World[0.044]; Gerg[0.044]; Alpine[0.043]; seconds[0.043]; minute[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.258:0.258[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.052]; Italy[0.052]; Switzerland[0.042]; Switzerland[0.042]; Switzerland[0.042]; Slovenia[0.042]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Cup[0.041]; Sweden[0.039]; Martina[0.039]; Stefanie[0.038]; World[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.218:0.218[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.049]; Austria[0.049]; Austria[0.049]; Switzerland[0.047]; Switzerland[0.047]; Switzerland[0.047]; Germany[0.046]; Slovenia[0.044]; Norway[0.043]; France[0.041]; France[0.041]; France[0.041]; Monika[0.040]; declared[0.039]; Renate[0.039]; Katharina[0.039]; Cavagnoud[0.039]; official[0.038]; Borghi[0.038]; Heidi[0.038]; Dorfmeister[0.037]; Zurbriggen[0.037]; Megan[0.036]; Ingeborg[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.248:0.248[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.046]; France[0.046]; France[0.046]; Switzerland[0.043]; Switzerland[0.043]; Switzerland[0.043]; Switzerland[0.043]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Martina[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Slovenia[0.039]; Norway[0.038]; Stefanie[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.119:0.119[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.047]; Switzerland[0.047]; Switzerland[0.047]; Austria[0.046]; Austria[0.046]; Austria[0.046]; Austria[0.046]; Slovenia[0.045]; France[0.043]; France[0.043]; France[0.043]; Ingeborg[0.041]; Marianne[0.039]; declared[0.039]; Michaela[0.039]; official[0.039]; Monika[0.039]; Renate[0.039]; Helen[0.039]; Heidi[0.038]; Shannon[0.038]; Catherine[0.038]; results[0.038]; Dorfmeister[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.240:0.240[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Austria[0.045]; Austria[0.045]; Austria[0.045]; Switzerland[0.044]; Switzerland[0.044]; Italy[0.043]; Italy[0.043]; Italy[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; Sweden[0.042]; Slovenia[0.042]; Cup[0.040]; Cup[0.040]; women[0.040]; Martina[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Switzerland ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.243:0.243[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.047]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Slovenia[0.042]; Sweden[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Austria[0.041]; Austria[0.041]; Alpine[0.041]; Russia[0.040]; Russia[0.040]; Cup[0.039]; Cup[0.039]; Skiing[0.038]; minute[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hilary Lindh\u001b[39m ==> ENTITY: \u001b[32mHilary Lindh\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.045]; Switzerland[0.044]; Switzerland[0.044]; Switzerland[0.044]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Ingeborg[0.042]; Italy[0.041]; Italy[0.041]; Monika[0.041]; Renate[0.041]; Wachter[0.040]; Stefanie[0.040]; France[0.040]; France[0.040]; France[0.040]; Shannon[0.040]; Nobis[0.040]; Zurbriggen[0.040]; Katharina[0.040]; Sybille[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.073:0.073[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.045]; Switzerland[0.045]; Switzerland[0.045]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Austria[0.043]; Slovenia[0.043]; Italy[0.043]; Italy[0.043]; Germany[0.043]; Germany[0.043]; Norway[0.042]; Shannon[0.040]; France[0.040]; France[0.040]; France[0.040]; Borghi[0.040]; official[0.040]; results[0.039]; Perez[0.039]; declared[0.038]; Gerety[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.264:0.264[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.046]; France[0.046]; France[0.046]; France[0.046]; Switzerland[0.043]; Switzerland[0.043]; Switzerland[0.043]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Martina[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Sweden[0.039]; Slovenia[0.039]; Norway[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Carole Montillet\u001b[39m ==> ENTITY: \u001b[32mCarole Montillet\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.129:0.129[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWomen[0.044]; Alpine[0.044]; Skiing[0.043]; Seizinger[0.043]; race[0.043]; Gerg[0.043]; Hilde[0.043]; Martina[0.042]; Vail[0.042]; women[0.042]; Cup[0.042]; Cup[0.042]; Colorado[0.041]; Saturday[0.041]; Super[0.041]; Renate[0.041]; Stefanie[0.040]; France[0.040]; France[0.040]; France[0.040]; World[0.040]; World[0.040]; Kostner[0.040]; Madlen[0.040]; \t\n\r [==================>...........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s892ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1902/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1175testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBucharest\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFC Steaua București\u001b[39m <---> \u001b[32mBucharest\u001b[39m\t\nSCORES: global= 0.267:0.267[\u001b[31m0.000\u001b[39m]; local()= 0.173:0.128[\u001b[31m0.045\u001b[39m]; log p(e|m)= -2.564:-0.018[\u001b[32m2.546\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOtelul[0.046]; Otelul[0.046]; Cup[0.045]; qualifier[0.044]; Romania[0.044]; League[0.043]; Bucharest[0.043]; Bucharest[0.043]; World[0.042]; Soccer[0.042]; squad[0.041]; players[0.041]; Macedonia[0.041]; Macedonia[0.041]; Macedonia[0.041]; European[0.040]; week[0.040]; National[0.040]; National[0.040]; championship[0.040]; Friday[0.040]; defender[0.039]; December[0.039]; title[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ovidiu Stinga\u001b[39m ==> ENTITY: \u001b[32mOvidiu Stîngă\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.047]; Munteanu[0.045]; club[0.044]; Champions[0.044]; players[0.044]; squad[0.044]; European[0.043]; Cup[0.043]; Steaua[0.042]; Steaua[0.042]; Bucharest[0.042]; Galatasaray[0.041]; Iulian[0.041]; Stefan[0.041]; Filipescu[0.041]; Popescu[0.040]; Tibor[0.040]; recently[0.040]; Craioveanu[0.040]; Ilie[0.040]; Midfielders[0.040]; Dorinel[0.040]; Defenders[0.040]; Hagi[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dan Petrescu\u001b[39m ==> ENTITY: \u001b[32mDan Petrescu\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOtelul[0.047]; club[0.046]; League[0.044]; League[0.044]; players[0.044]; Munteanu[0.043]; Champions[0.042]; squad[0.042]; Bogdan[0.042]; Ilie[0.041]; Steaua[0.041]; Steaua[0.041]; Dorinel[0.041]; Daniel[0.041]; European[0.040]; Cup[0.040]; Anton[0.040]; Adrian[0.040]; Filipescu[0.040]; Viorel[0.040]; Stefan[0.040]; Hagi[0.040]; Bucharest[0.040]; Bucharest[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gheorghe Craioveanu\u001b[39m ==> ENTITY: \u001b[32mGheorghe Craioveanu\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.107:0.107[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.047]; League[0.046]; Stefan[0.043]; Cup[0.043]; players[0.043]; Steaua[0.042]; Tibor[0.042]; Midfielders[0.042]; Galatasaray[0.042]; Munteanu[0.041]; midfielder[0.041]; recently[0.041]; Hagi[0.041]; European[0.041]; Iulian[0.041]; Anton[0.041]; squad[0.041]; Adrian[0.040]; Gheorghe[0.040]; Gheorghe[0.040]; Forwards[0.040]; Ilie[0.040]; Filipescu[0.040]; group[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Otelul Galati\u001b[39m ==> ENTITY: \u001b[32mFC Oțelul Galați\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.048]; League[0.048]; Otelul[0.047]; club[0.046]; Romania[0.045]; Soccer[0.044]; players[0.044]; players[0.044]; Steaua[0.043]; Steaua[0.043]; Bucharest[0.043]; Bucharest[0.043]; Bucharest[0.043]; Bucharest[0.043]; qualifier[0.043]; performances[0.042]; Champions[0.042]; squad[0.042]; squad[0.042]; Iulian[0.042]; Viorel[0.042]; Stefan[0.042]; Cornel[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Liviu Ciobotariu\u001b[39m ==> ENTITY: \u001b[32mLiviu Ciobotariu\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.045]; Romania[0.044]; players[0.043]; players[0.043]; Otelul[0.042]; Otelul[0.042]; December[0.042]; defender[0.042]; League[0.042]; League[0.042]; Cup[0.042]; Cup[0.042]; Stefan[0.042]; qualifier[0.041]; Macedonia[0.041]; Macedonia[0.041]; Macedonia[0.041]; club[0.041]; Champions[0.040]; striker[0.040]; midfielder[0.040]; Galatasaray[0.040]; championship[0.040]; Players[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Constantin Galca\u001b[39m ==> ENTITY: \u001b[32mConstantin Gâlcă\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.116:0.116[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.045]; Champions[0.044]; players[0.044]; Munteanu[0.043]; Stefan[0.043]; European[0.042]; Tibor[0.042]; squad[0.042]; Midfielders[0.042]; Galatasaray[0.041]; Cup[0.041]; League[0.041]; League[0.041]; title[0.041]; midfielder[0.041]; Steaua[0.041]; Steaua[0.041]; Bucharest[0.041]; Craioveanu[0.041]; Ilie[0.041]; Popescu[0.041]; recently[0.041]; Bogdan[0.041]; Iulian[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Cup ==> ENTITY: \u001b[32mFIFA World Cup\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.153:0.153[0]; log p(e|m)= -1.002:-1.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.049]; squad[0.049]; squad[0.049]; Champions[0.047]; Romania[0.047]; qualifier[0.047]; League[0.046]; League[0.046]; Cup[0.046]; players[0.046]; players[0.046]; European[0.046]; European[0.046]; Otelul[0.045]; Otelul[0.045]; week[0.045]; National[0.043]; National[0.043]; December[0.043]; finished[0.043]; midfielder[0.043]; striker[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Galatasaray ==> ENTITY: \u001b[32mGalatasaray S.K. (football)\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.178:0.178[0]; log p(e|m)= -1.897:-1.897[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.045]; League[0.045]; League[0.045]; Turkish[0.044]; Champions[0.043]; players[0.043]; Otelul[0.043]; Otelul[0.043]; Cup[0.042]; championship[0.042]; European[0.042]; European[0.042]; squad[0.041]; Hagi[0.041]; midfielder[0.040]; group[0.040]; group[0.040]; title[0.040]; recently[0.040]; defender[0.040]; finished[0.039]; Macedonia[0.039]; moved[0.039]; Petrescu[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iulian Filipescu\u001b[39m ==> ENTITY: \u001b[32mIulian Filipescu\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOtelul[0.046]; championship[0.043]; Munteanu[0.043]; players[0.043]; club[0.043]; midfielder[0.042]; Steaua[0.042]; Steaua[0.042]; League[0.042]; League[0.042]; title[0.042]; Florin[0.042]; Viorel[0.041]; Petrescu[0.041]; Stefan[0.041]; Ilie[0.041]; Champions[0.041]; Hagi[0.041]; Bucharest[0.041]; Bucharest[0.041]; Craioveanu[0.040]; Tibor[0.040]; Bogdan[0.040]; Galatasaray[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = European Cup ==> ENTITY: \u001b[32mUEFA Champions League\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.173:0.173[0]; log p(e|m)= -0.534:-0.534[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOtelul[0.044]; Otelul[0.044]; League[0.043]; League[0.043]; club[0.043]; championship[0.043]; Champions[0.043]; qualifier[0.043]; European[0.043]; Cup[0.042]; players[0.042]; title[0.041]; squad[0.041]; squad[0.041]; holders[0.040]; Bucharest[0.040]; Bucharest[0.040]; Bucharest[0.040]; Munteanu[0.040]; Steaua[0.040]; Steaua[0.040]; clash[0.040]; defender[0.040]; Galatasaray[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEuropean\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUEFA\u001b[39m <---> \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.242:0.237[\u001b[31m0.005\u001b[39m]; local()= 0.170:0.056[\u001b[31m0.115\u001b[39m]; log p(e|m)= -3.576:-0.256[\u001b[32m3.319\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; League[0.044]; European[0.044]; Otelul[0.044]; Otelul[0.044]; club[0.043]; Champions[0.043]; Soccer[0.043]; Macedonia[0.041]; Macedonia[0.041]; Macedonia[0.041]; Cup[0.041]; Cup[0.041]; qualifier[0.041]; week[0.041]; players[0.040]; players[0.040]; squad[0.040]; squad[0.040]; Steaua[0.040]; Steaua[0.040]; Romania[0.040]; Galatasaray[0.040]; domestic[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ionel Danciulescu\u001b[39m ==> ENTITY: \u001b[32mIonel Dănciulescu\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.046]; Iulian[0.043]; squad[0.043]; players[0.042]; Munteanu[0.042]; Florin[0.042]; Stefan[0.042]; Ilie[0.042]; Gheorghe[0.042]; Gheorghe[0.042]; Gheorghe[0.042]; Petrescu[0.042]; Popescu[0.042]; Viorel[0.041]; recently[0.041]; Bogdan[0.041]; Tibor[0.041]; Hagi[0.041]; Steaua[0.041]; Midfielders[0.041]; Galatasaray[0.041]; Valentin[0.040]; Filipescu[0.040]; Anton[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Viorel Ion\u001b[39m ==> ENTITY: \u001b[32mViorel Ion\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.046]; players[0.045]; players[0.045]; Otelul[0.045]; Otelul[0.045]; club[0.045]; League[0.044]; League[0.044]; Champions[0.044]; title[0.043]; Stefan[0.043]; Ilie[0.043]; Liviu[0.043]; Romania[0.043]; championship[0.043]; Adrian[0.043]; chance[0.043]; qualifier[0.042]; Steaua[0.042]; Steaua[0.042]; Galatasaray[0.042]; defender[0.042]; Midfielder[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Daniel Prodan\u001b[39m ==> ENTITY: \u001b[32mDaniel Prodan\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.045]; players[0.044]; Otelul[0.044]; League[0.043]; League[0.043]; Cup[0.043]; Stefan[0.043]; Champions[0.042]; championship[0.042]; Tibor[0.042]; title[0.042]; Midfielders[0.042]; Munteanu[0.041]; Galatasaray[0.041]; squad[0.041]; Iulian[0.041]; Filipescu[0.040]; Petrescu[0.040]; Anton[0.040]; Steaua[0.040]; Steaua[0.040]; Adrian[0.040]; holders[0.040]; midfielder[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Viorel Ion\u001b[39m ==> ENTITY: \u001b[32mViorel Ion\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIulian[0.045]; players[0.044]; club[0.043]; Munteanu[0.043]; Stefan[0.042]; Ilie[0.042]; Florin[0.042]; Dorinel[0.042]; Liviu[0.042]; Constantin[0.041]; Adrian[0.041]; Petrescu[0.041]; Gheorghe[0.041]; Gheorghe[0.041]; Gheorghe[0.041]; Tibor[0.041]; Bogdan[0.041]; Steaua[0.041]; Midfielders[0.041]; Galatasaray[0.041]; Dan[0.040]; Ioan[0.040]; Hagi[0.040]; Ovidiu[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mRomania\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRomania\u001b[39m <---> \u001b[32mRomania national football team\u001b[39m\t\nSCORES: global= 0.249:0.245[\u001b[31m0.004\u001b[39m]; local()= 0.111:0.161[\u001b[32m0.050\u001b[39m]; log p(e|m)= -0.212:-3.058[\u001b[31m2.846\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.045]; players[0.044]; squad[0.044]; Otelul[0.043]; Otelul[0.043]; qualifier[0.043]; League[0.043]; Macedonia[0.043]; Macedonia[0.043]; Macedonia[0.043]; Cup[0.042]; European[0.042]; Bucharest[0.042]; Bucharest[0.042]; Bucharest[0.042]; week[0.041]; uncapped[0.041]; Players[0.041]; Steaua[0.040]; championship[0.040]; World[0.039]; clash[0.039]; group[0.039]; Uncapped[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Basarab Panduru\u001b[39m ==> ENTITY: \u001b[32mBasarab Panduru\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.046]; players[0.045]; League[0.045]; League[0.045]; Cornel[0.043]; squad[0.042]; midfielder[0.042]; Stefan[0.042]; Tibor[0.042]; Steaua[0.041]; Steaua[0.041]; Hagi[0.041]; Midfielders[0.041]; Petrescu[0.041]; Galatasaray[0.041]; Munteanu[0.041]; Cup[0.041]; Viorel[0.041]; Iulian[0.040]; Goalkeepers[0.040]; Anton[0.040]; Filipescu[0.040]; Bogdan[0.040]; Champions[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bogdan Stelea\u001b[39m ==> ENTITY: \u001b[32mBogdan Stelea\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.046]; Champions[0.045]; League[0.044]; League[0.044]; Otelul[0.044]; players[0.044]; squad[0.043]; European[0.042]; European[0.042]; Stefan[0.042]; Cup[0.041]; Tibor[0.041]; recently[0.041]; championship[0.041]; Midfielders[0.040]; Galatasaray[0.040]; Steaua[0.040]; Steaua[0.040]; midfielder[0.040]; domestic[0.040]; National[0.040]; moved[0.040]; Filipescu[0.040]; Munteanu[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = National Bucharest\u001b[39m ==> ENTITY: \u001b[32mFC Progresul București\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.044]; club[0.044]; League[0.043]; League[0.043]; Champions[0.043]; Otelul[0.043]; Otelul[0.043]; Bucharest[0.042]; Bucharest[0.042]; Bucharest[0.042]; qualifier[0.042]; squad[0.042]; squad[0.042]; Cup[0.042]; Cup[0.042]; Romania[0.041]; players[0.041]; players[0.041]; Galatasaray[0.041]; Stefan[0.040]; recently[0.040]; Steaua[0.040]; Steaua[0.040]; Tibor[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Otelul Galati\u001b[39m ==> ENTITY: \u001b[32mFC Oțelul Galați\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; League[0.046]; Otelul[0.045]; club[0.044]; Romania[0.044]; Soccer[0.042]; players[0.042]; players[0.042]; Steaua[0.041]; Steaua[0.041]; Bucharest[0.041]; Bucharest[0.041]; Bucharest[0.041]; Bucharest[0.041]; qualifier[0.041]; performances[0.041]; Champions[0.041]; squad[0.040]; squad[0.040]; Viorel[0.040]; Stefan[0.040]; group[0.040]; group[0.040]; championship[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gheorghe Hagi\u001b[39m ==> ENTITY: \u001b[32mGheorghe Hagi\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.045]; players[0.044]; Munteanu[0.044]; League[0.043]; League[0.043]; Otelul[0.043]; Galatasaray[0.043]; Petrescu[0.043]; Cup[0.042]; squad[0.042]; Steaua[0.041]; Steaua[0.041]; Stefan[0.041]; midfielder[0.041]; Iulian[0.041]; Craioveanu[0.041]; Filipescu[0.041]; Tibor[0.040]; Ilie[0.040]; European[0.040]; Midfielders[0.040]; Viorel[0.040]; holders[0.040]; recently[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Adrian Ilie\u001b[39m ==> ENTITY: \u001b[32mAdrian Ilie\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayers[0.045]; midfielder[0.044]; defender[0.043]; squad[0.043]; Viorel[0.043]; Viorel[0.043]; European[0.042]; European[0.042]; striker[0.042]; Otelul[0.042]; Otelul[0.042]; club[0.041]; Iulian[0.041]; Munteanu[0.041]; League[0.041]; League[0.041]; Craioveanu[0.041]; Ioan[0.041]; Stefan[0.041]; Stefan[0.041]; Popescu[0.041]; Filipescu[0.040]; Valentin[0.040]; Valentin[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Anton Dobos\u001b[39m ==> ENTITY: \u001b[32mAnton Doboș\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.043]; Stefan[0.043]; Munteanu[0.043]; Otelul[0.042]; championship[0.042]; League[0.042]; League[0.042]; Galatasaray[0.042]; Champions[0.042]; Tibor[0.042]; Steaua[0.042]; Steaua[0.042]; title[0.042]; squad[0.042]; think[0.042]; Midfielders[0.042]; Hagi[0.041]; Cup[0.041]; Cornel[0.041]; Bucharest[0.041]; Bucharest[0.041]; Iulian[0.041]; Liviu[0.040]; players[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Steaua Bucharest ==> ENTITY: \u001b[32mFC Steaua București\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.202:0.202[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.047]; Otelul[0.047]; Otelul[0.047]; Steaua[0.045]; players[0.043]; players[0.043]; squad[0.042]; squad[0.042]; Romania[0.042]; League[0.042]; League[0.042]; Champions[0.041]; Bucharest[0.041]; Bucharest[0.041]; qualifier[0.041]; championship[0.040]; striker[0.040]; Ilie[0.040]; Cup[0.040]; Cup[0.040]; midfielder[0.040]; Munteanu[0.039]; defender[0.039]; Petrescu[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = National Bucharest\u001b[39m ==> ENTITY: \u001b[32mFC Progresul București\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.216:0.216[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.044]; club[0.044]; League[0.043]; League[0.043]; Champions[0.043]; Otelul[0.043]; Otelul[0.043]; Bucharest[0.042]; Bucharest[0.042]; Bucharest[0.042]; qualifier[0.042]; squad[0.042]; squad[0.042]; Cup[0.042]; Cup[0.042]; Romania[0.041]; players[0.041]; players[0.041]; Galatasaray[0.041]; Stefan[0.040]; recently[0.040]; Steaua[0.040]; Steaua[0.040]; moved[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dorinel Munteanu\u001b[39m ==> ENTITY: \u001b[32mDorinel Munteanu\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.045]; Bogdan[0.043]; Bucharest[0.043]; League[0.043]; squad[0.042]; Champions[0.042]; midfielder[0.042]; Stefan[0.042]; Petrescu[0.042]; Filipescu[0.042]; players[0.042]; Iulian[0.042]; Tibor[0.041]; Gheorghe[0.041]; Gheorghe[0.041]; Gheorghe[0.041]; Hagi[0.041]; European[0.041]; Midfielders[0.041]; Craioveanu[0.041]; Steaua[0.041]; Steaua[0.041]; Galatasaray[0.041]; Valentin[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Macedonia ==> ENTITY: \u001b[32mMacedonia national football team\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.174:0.174[0]; log p(e|m)= -2.303:-2.303[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMacedonia[0.048]; Macedonia[0.048]; Soccer[0.045]; qualifier[0.045]; League[0.044]; Romania[0.043]; squad[0.043]; players[0.042]; Players[0.042]; European[0.042]; Cup[0.041]; Otelul[0.041]; Otelul[0.041]; World[0.040]; week[0.040]; National[0.040]; National[0.040]; December[0.040]; striker[0.040]; championship[0.040]; defender[0.039]; face[0.039]; clash[0.039]; performances[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Anghel Iordanescu\u001b[39m ==> ENTITY: \u001b[32mAnghel Iordănescu\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOtelul[0.045]; Otelul[0.045]; Romania[0.045]; players[0.043]; League[0.043]; Viorel[0.042]; Cup[0.042]; squad[0.042]; Stefan[0.042]; Bucharest[0.042]; Bucharest[0.042]; Bucharest[0.042]; Bucharest[0.042]; qualifier[0.041]; European[0.041]; Soccer[0.041]; Steaua[0.041]; championship[0.040]; Midfielder[0.040]; Macedonia[0.040]; Macedonia[0.040]; Macedonia[0.040]; National[0.040]; National[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Florin Prunea\u001b[39m ==> ENTITY: \u001b[32mFlorin Prunea\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOtelul[0.046]; championship[0.043]; Bucharest[0.042]; Bucharest[0.042]; players[0.042]; Cup[0.042]; Petrescu[0.042]; League[0.042]; League[0.042]; chance[0.042]; finished[0.042]; Munteanu[0.042]; Stefan[0.042]; European[0.041]; European[0.041]; title[0.041]; Hagi[0.041]; Champions[0.041]; good[0.041]; club[0.041]; Tibor[0.041]; Midfielders[0.040]; Bogdan[0.040]; Galatasaray[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gheorghe Popescu\u001b[39m ==> ENTITY: \u001b[32mGheorghe Popescu\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.044]; Munteanu[0.043]; Iulian[0.043]; League[0.043]; League[0.043]; Gheorghe[0.043]; Gheorghe[0.043]; Galatasaray[0.043]; Steaua[0.042]; Steaua[0.042]; European[0.042]; players[0.042]; Ilie[0.042]; Filipescu[0.041]; Petrescu[0.041]; Bucharest[0.041]; Craioveanu[0.041]; Cup[0.041]; Champions[0.040]; Adrian[0.040]; midfielder[0.040]; Bogdan[0.040]; recently[0.040]; Viorel[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Steaua\u001b[39m ==> ENTITY: \u001b[32mFC Steaua București\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.050]; Otelul[0.049]; Otelul[0.049]; Steaua[0.047]; players[0.045]; squad[0.044]; League[0.044]; League[0.044]; Champions[0.044]; Bucharest[0.043]; Bucharest[0.043]; Bucharest[0.043]; championship[0.042]; striker[0.042]; Ilie[0.042]; Cup[0.042]; midfielder[0.042]; Munteanu[0.042]; defender[0.041]; Petrescu[0.041]; Craioveanu[0.041]; Galatasaray[0.041]; Bogdan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Macedonia ==> ENTITY: \u001b[32mRepublic of Macedonia\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.119:0.119[0]; log p(e|m)= -1.115:-1.115[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntitle[0.049]; Macedonia[0.049]; Macedonia[0.049]; Soccer[0.047]; qualifier[0.045]; championship[0.045]; Romania[0.044]; squad[0.044]; squad[0.044]; Champions[0.043]; players[0.043]; players[0.043]; Players[0.042]; finished[0.042]; Turkish[0.042]; European[0.042]; European[0.042]; Cup[0.042]; Cup[0.042]; Bucharest[0.041]; Bucharest[0.041]; Bucharest[0.041]; Bucharest[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Liviu Ciobotariu\u001b[39m ==> ENTITY: \u001b[32mLiviu Ciobotariu\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayers[0.044]; Otelul[0.044]; League[0.043]; League[0.043]; Cup[0.043]; Stefan[0.043]; club[0.042]; Tibor[0.042]; Midfielders[0.042]; Champions[0.042]; midfielder[0.042]; Galatasaray[0.041]; championship[0.041]; Munteanu[0.041]; Bogdan[0.041]; Galati[0.041]; Viorel[0.041]; Cornel[0.041]; Defenders[0.041]; Iulian[0.041]; title[0.041]; Dan[0.040]; European[0.040]; Anton[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Macedonia ==> ENTITY: \u001b[32mMacedonia national football team\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.183:0.183[0]; log p(e|m)= -2.303:-2.303[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMacedonia[0.047]; Macedonia[0.047]; Soccer[0.044]; qualifier[0.044]; League[0.043]; League[0.043]; Romania[0.042]; squad[0.042]; squad[0.042]; players[0.042]; players[0.042]; Players[0.041]; European[0.041]; European[0.041]; Champions[0.041]; finished[0.040]; Cup[0.040]; Cup[0.040]; Otelul[0.040]; Otelul[0.040]; World[0.039]; week[0.039]; National[0.039]; National[0.039]; \t\n\r [==================>...........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s631ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1937/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1292testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rome ==> ENTITY: \u001b[32mRome\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.052]; Italy[0.052]; Italian[0.046]; Milan[0.045]; officially[0.043]; Tuesday[0.042]; start[0.042]; financial[0.041]; Saturday[0.041]; upper[0.041]; commission[0.041]; commission[0.041]; concludes[0.041]; house[0.040]; approve[0.040]; assembly[0.040]; concluded[0.040]; budget[0.039]; budget[0.039]; budget[0.039]; measure[0.039]; examination[0.039]; examination[0.039]; examination[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italian ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.124:0.124[0]; log p(e|m)= -0.759:-0.759[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.061]; Italy[0.061]; Rome[0.050]; Milan[0.049]; Tuesday[0.044]; officially[0.043]; concluded[0.042]; approve[0.042]; financial[0.042]; commission[0.042]; commission[0.042]; house[0.041]; assembly[0.041]; budget[0.041]; budget[0.041]; budget[0.041]; Saturday[0.041]; start[0.040]; upper[0.040]; measure[0.040]; package[0.040]; Senate[0.039]; Senate[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.057]; Italian[0.049]; Rome[0.047]; Milan[0.046]; officially[0.044]; Tuesday[0.043]; financial[0.043]; commission[0.042]; commission[0.042]; start[0.042]; package[0.042]; concluded[0.041]; Saturday[0.041]; budget[0.040]; budget[0.040]; budget[0.040]; approve[0.039]; assembly[0.039]; measure[0.038]; concludes[0.037]; examination[0.037]; examination[0.037]; examination[0.037]; upper[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Milan ==> ENTITY: \u001b[32mMilan\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.301:-0.301[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.055]; Italy[0.055]; Rome[0.053]; Italian[0.051]; house[0.048]; financial[0.046]; commission[0.045]; commission[0.045]; start[0.045]; Tuesday[0.045]; Saturday[0.045]; officially[0.045]; concluded[0.044]; upper[0.043]; assembly[0.043]; budget[0.042]; budget[0.042]; budget[0.042]; Senate[0.042]; Senate[0.042]; approve[0.042]; measure[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.057]; Italian[0.049]; Rome[0.047]; Milan[0.046]; officially[0.044]; Tuesday[0.043]; financial[0.043]; commission[0.042]; commission[0.042]; start[0.042]; package[0.042]; concluded[0.041]; Saturday[0.041]; budget[0.040]; budget[0.040]; budget[0.040]; approve[0.039]; assembly[0.039]; measure[0.038]; concludes[0.037]; examination[0.037]; examination[0.037]; examination[0.037]; upper[0.036]; \t\n\r [==================>...........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s655ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1942/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1229testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Zealand ==> ENTITY: \u001b[32mNew Zealand national cricket team\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.163:0.163[0]; log p(e|m)= -3.147:-3.147[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCricket[0.059]; match[0.056]; Pakistan[0.056]; Pakistan[0.056]; Pakistan[0.056]; Pakistan[0.056]; series[0.055]; Zealand[0.055]; Zealand[0.055]; day[0.053]; Scores[0.051]; beat[0.050]; beat[0.050]; runs[0.050]; runs[0.050]; Sialkot[0.050]; Friday[0.049]; lead[0.048]; unbeatable[0.046]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPakistan\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPakistan national cricket team\u001b[39m <---> \u001b[32mPakistan\u001b[39m\t\nSCORES: global= 0.259:0.249[\u001b[31m0.010\u001b[39m]; local()= 0.165:0.132[\u001b[31m0.034\u001b[39m]; log p(e|m)= -2.749:-0.219[\u001b[32m2.529\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.067]; Pakistan[0.067]; Pakistan[0.067]; series[0.062]; match[0.057]; Sialkot[0.055]; Cricket[0.055]; day[0.051]; runs[0.049]; runs[0.049]; Friday[0.048]; Zealand[0.048]; Zealand[0.048]; Zealand[0.048]; beat[0.048]; beat[0.048]; Scores[0.046]; lead[0.045]; unbeatable[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Zealand ==> ENTITY: \u001b[32mNew Zealand national cricket team\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.163:0.163[0]; log p(e|m)= -3.147:-3.147[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCricket[0.059]; match[0.056]; Pakistan[0.056]; Pakistan[0.056]; Pakistan[0.056]; Pakistan[0.056]; series[0.055]; Zealand[0.055]; Zealand[0.055]; day[0.053]; Scores[0.051]; beat[0.050]; beat[0.050]; runs[0.050]; runs[0.050]; Sialkot[0.050]; Friday[0.049]; lead[0.048]; unbeatable[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pakistan ==> ENTITY: \u001b[32mPakistan national cricket team\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.165:0.165[0]; log p(e|m)= -2.749:-2.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.067]; Pakistan[0.067]; Pakistan[0.067]; series[0.062]; match[0.057]; Sialkot[0.055]; Cricket[0.055]; day[0.051]; runs[0.049]; runs[0.049]; Friday[0.048]; Zealand[0.048]; Zealand[0.048]; Zealand[0.048]; beat[0.048]; beat[0.048]; Scores[0.046]; lead[0.045]; unbeatable[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sialkot\u001b[39m ==> ENTITY: \u001b[32mSialkot\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.122:0.122[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.064]; Pakistan[0.064]; Pakistan[0.064]; Pakistan[0.064]; Cricket[0.054]; series[0.053]; Scores[0.052]; Friday[0.052]; day[0.052]; lead[0.050]; match[0.050]; runs[0.049]; runs[0.049]; beat[0.048]; beat[0.048]; Zealand[0.047]; Zealand[0.047]; Zealand[0.047]; unbeatable[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Zealand ==> ENTITY: \u001b[32mNew Zealand national cricket team\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.163:0.163[0]; log p(e|m)= -3.147:-3.147[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCricket[0.059]; match[0.056]; Pakistan[0.056]; Pakistan[0.056]; Pakistan[0.056]; Pakistan[0.056]; series[0.055]; Zealand[0.055]; Zealand[0.055]; day[0.053]; Scores[0.051]; beat[0.050]; beat[0.050]; runs[0.050]; runs[0.050]; Sialkot[0.050]; Friday[0.049]; lead[0.048]; unbeatable[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pakistan ==> ENTITY: \u001b[32mPakistan national cricket team\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.165:0.165[0]; log p(e|m)= -2.749:-2.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.067]; Pakistan[0.067]; Pakistan[0.067]; series[0.062]; match[0.057]; Sialkot[0.055]; Cricket[0.055]; day[0.051]; runs[0.049]; runs[0.049]; Friday[0.048]; Zealand[0.048]; Zealand[0.048]; Zealand[0.048]; beat[0.048]; beat[0.048]; Scores[0.046]; lead[0.045]; unbeatable[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pakistan ==> ENTITY: \u001b[32mPakistan national cricket team\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.165:0.165[0]; log p(e|m)= -2.749:-2.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.067]; Pakistan[0.067]; Pakistan[0.067]; series[0.062]; match[0.057]; Sialkot[0.055]; Cricket[0.055]; day[0.051]; runs[0.049]; runs[0.049]; Friday[0.048]; Zealand[0.048]; Zealand[0.048]; Zealand[0.048]; beat[0.048]; beat[0.048]; Scores[0.046]; lead[0.045]; unbeatable[0.043]; \t\n\r [===================>..........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s648ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1950/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1321testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Munich ==> ENTITY: \u001b[32mMunich\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.119:-0.119[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.048]; Munich[0.048]; Munich[0.048]; Munich[0.048]; Munich[0.048]; stock[0.045]; stock[0.045]; tenth[0.045]; planned[0.043]; plans[0.043]; company[0.042]; September[0.042]; September[0.042]; says[0.042]; expected[0.042]; shares[0.042]; shares[0.042]; Friday[0.042]; world[0.041]; largest[0.041]; marks[0.041]; reinsurer[0.041]; today[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Munich Re\u001b[39m ==> ENTITY: \u001b[32mMunich Re\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAg[0.048]; Germany[0.048]; company[0.046]; Frankfurt[0.046]; Munich[0.045]; Munich[0.045]; Munich[0.045]; Munich[0.045]; stock[0.044]; stock[0.044]; trading[0.043]; largest[0.043]; Newsroom[0.042]; shares[0.042]; shares[0.042]; shares[0.042]; reinsurer[0.042]; shareholders[0.042]; world[0.042]; September[0.042]; September[0.042]; group[0.041]; today[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = DAX\u001b[39m ==> ENTITY: \u001b[32mDAX\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrankfurt[0.046]; trading[0.045]; stock[0.044]; index[0.044]; company[0.043]; largest[0.042]; September[0.042]; September[0.042]; blue[0.042]; shares[0.042]; shares[0.042]; shares[0.042]; today[0.042]; world[0.041]; said[0.040]; said[0.040]; drop[0.040]; value[0.040]; value[0.040]; value[0.040]; seek[0.040]; Munich[0.040]; Munich[0.040]; Munich[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMunich[0.046]; Munich[0.046]; Munich[0.046]; Munich[0.046]; Munich[0.046]; September[0.043]; September[0.043]; value[0.042]; value[0.042]; value[0.042]; tenth[0.041]; Friday[0.041]; probably[0.041]; world[0.041]; drop[0.040]; lower[0.040]; price[0.040]; expected[0.040]; today[0.039]; company[0.039]; marks[0.039]; group[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Munich Re\u001b[39m ==> ENTITY: \u001b[32mMunich Re\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAg[0.047]; Germany[0.046]; company[0.044]; Frankfurt[0.044]; Munich[0.043]; Munich[0.043]; Munich[0.043]; stock[0.042]; stock[0.042]; trading[0.042]; largest[0.041]; Newsroom[0.041]; shares[0.041]; shares[0.041]; shares[0.041]; reinsurer[0.041]; shareholders[0.040]; world[0.040]; September[0.040]; September[0.040]; group[0.040]; today[0.040]; Friday[0.040]; Friday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Munich Re\u001b[39m ==> ENTITY: \u001b[32mMunich Re\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAg[0.048]; Germany[0.048]; company[0.046]; Frankfurt[0.046]; Munich[0.045]; Munich[0.045]; Munich[0.045]; Munich[0.045]; stock[0.044]; stock[0.044]; trading[0.043]; largest[0.043]; Newsroom[0.042]; shares[0.042]; shares[0.042]; shares[0.042]; reinsurer[0.042]; shareholders[0.042]; world[0.042]; September[0.042]; September[0.042]; group[0.041]; today[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Munich Re\u001b[39m ==> ENTITY: \u001b[32mMunich Re\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAg[0.047]; Germany[0.047]; company[0.045]; Munich[0.043]; Munich[0.043]; Munich[0.043]; stock[0.043]; stock[0.043]; largest[0.042]; shares[0.041]; shares[0.041]; reinsurer[0.041]; shareholders[0.041]; world[0.041]; September[0.041]; September[0.041]; group[0.040]; today[0.040]; Friday[0.040]; tenth[0.040]; said[0.039]; said[0.039]; switch[0.039]; switch[0.039]; \t\n\r [===================>..........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s646ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1957/4485 \r\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kurdistan Workers Party\u001b[39m ==> ENTITY: \u001b[32mKurdistan Workers' Party\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.045]; Turkey[0.045]; Kurdish[0.045]; Kurdish[0.045]; Kurdish[0.045]; Turkish[0.044]; Turkish[0.044]; government[0.044]; guerrilla[0.043]; groups[0.041]; militia[0.041]; militia[0.041]; Diyarbakir[0.041]; guerrillas[0.041]; region[0.040]; forces[0.040]; said[0.040]; said[0.040]; civilian[0.040]; security[0.039]; security[0.039]; autonomy[0.039]; weapons[0.039]; conflict[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = PKK\u001b[39m ==> ENTITY: \u001b[32mKurdistan Workers' Party\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdish[0.046]; Kurdish[0.046]; Kurdish[0.046]; Turkish[0.045]; Kurdistan[0.045]; forces[0.044]; government[0.043]; conflict[0.042]; guerrillas[0.042]; region[0.041]; civilian[0.041]; security[0.041]; security[0.041]; fighting[0.041]; said[0.041]; militia[0.040]; crime[0.040]; weapons[0.039]; fight[0.039]; people[0.039]; killed[0.039]; rebels[0.039]; autonomy[0.039]; involved[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = PKK\u001b[39m ==> ENTITY: \u001b[32mKurdistan Workers' Party\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.046]; Turkey[0.046]; Kurdish[0.045]; Kurdish[0.045]; Kurdish[0.045]; Turkish[0.044]; Turkish[0.044]; Kurdistan[0.043]; forces[0.042]; Diyarbakir[0.042]; government[0.042]; guerrilla[0.041]; conflict[0.041]; guerrillas[0.040]; region[0.040]; civilian[0.040]; security[0.040]; security[0.040]; fighting[0.040]; gun[0.039]; groups[0.039]; said[0.039]; said[0.039]; militia[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mKurd\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mKurdish people\u001b[39m <---> \u001b[32mKurdistan\u001b[39m\t\nSCORES: global= 0.277:0.250[\u001b[31m0.027\u001b[39m]; local()= 0.179:0.182[\u001b[32m0.002\u001b[39m]; log p(e|m)= 0.000:-2.604[\u001b[31m2.604\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdish[0.050]; Kurdish[0.050]; Turkey[0.046]; Turkey[0.046]; southeastern[0.044]; Diyarbakir[0.044]; Turkish[0.044]; people[0.041]; guerrilla[0.041]; militia[0.041]; families[0.040]; mainly[0.040]; state[0.040]; groups[0.040]; militiamen[0.040]; city[0.040]; killed[0.040]; civilian[0.039]; rebels[0.039]; village[0.039]; village[0.039]; said[0.039]; said[0.039]; caused[0.038]; \t\n\r [===================>..........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s644ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1964/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1264testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ramos Horta\u001b[39m ==> ENTITY: \u001b[32mJosé Ramos-Horta\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.234:0.234[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNobel[0.049]; Timor[0.048]; Timor[0.048]; Alatas[0.048]; Alatas[0.048]; Alatas[0.048]; Alatas[0.048]; Alatas[0.048]; Indonesia[0.048]; Indonesia[0.048]; conference[0.048]; ambassador[0.047]; Prize[0.047]; leader[0.047]; ceremony[0.047]; ceremony[0.047]; Indonesian[0.047]; government[0.047]; Peace[0.047]; Horta[0.047]; Horta[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = East Timor ==> ENTITY: \u001b[32mEast Timor\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.182:-0.182[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimor[0.046]; Indonesia[0.046]; government[0.044]; following[0.044]; Indonesian[0.043]; territory[0.043]; Pope[0.042]; East[0.042]; Nations[0.041]; ministerial[0.041]; secure[0.041]; Jakarta[0.041]; Jakarta[0.041]; Jakarta[0.041]; Vatican[0.041]; peaceful[0.041]; conference[0.041]; Portuguese[0.040]; Norway[0.040]; questions[0.040]; Asked[0.040]; Peace[0.040]; week[0.040]; ceremony[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jakarta\u001b[39m ==> ENTITY: \u001b[32mJakarta\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.051]; Indonesia[0.051]; Indonesian[0.047]; Indonesian[0.047]; territory[0.047]; people[0.047]; Tuesday[0.046]; Australia[0.045]; Timor[0.045]; East[0.045]; East[0.045]; East[0.045]; Friday[0.045]; Friday[0.045]; capital[0.044]; capital[0.044]; jointly[0.044]; Dili[0.044]; Dili[0.044]; arrive[0.043]; Alatas[0.043]; officially[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Belo\u001b[39m ==> ENTITY: \u001b[32mCarlos Filipe Ximenes Belo\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimor[0.043]; Timor[0.043]; Indonesia[0.043]; Indonesia[0.043]; Belo[0.043]; Belo[0.043]; Belo[0.043]; bishop[0.042]; Helmut[0.042]; Pope[0.042]; month[0.042]; Bonn[0.041]; Nobel[0.041]; Peace[0.041]; Portuguese[0.041]; government[0.041]; visit[0.041]; visit[0.041]; ceremony[0.041]; peaceful[0.041]; Jakarta[0.040]; Jakarta[0.040]; ambassador[0.040]; position[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vatican ==> ENTITY: \u001b[32mVatican City\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.929:-0.929[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPope[0.047]; bishop[0.045]; ceremony[0.043]; Norway[0.043]; month[0.043]; ambassador[0.042]; official[0.042]; following[0.042]; visit[0.042]; visit[0.042]; government[0.042]; position[0.041]; Tuesday[0.041]; Portuguese[0.041]; said[0.041]; said[0.041]; solely[0.041]; attended[0.041]; issue[0.040]; come[0.040]; Indonesia[0.040]; Indonesia[0.040]; wanted[0.040]; way[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = East Timorese\u001b[39m ==> ENTITY: \u001b[32mEast Timor\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimor[0.048]; East[0.045]; East[0.045]; Indonesia[0.045]; Timorese[0.044]; Australia[0.044]; territory[0.043]; people[0.043]; Indonesian[0.042]; Indonesian[0.042]; government[0.042]; Dili[0.042]; Dili[0.042]; Catholic[0.042]; Horta[0.040]; capital[0.040]; Jakarta[0.040]; said[0.039]; Norway[0.039]; inclusion[0.039]; Peace[0.039]; ceremony[0.039]; ceremony[0.039]; Jose[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesian ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.194:0.194[0]; log p(e|m)= -0.742:-0.742[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.051]; Indonesian[0.046]; people[0.044]; Jakarta[0.043]; Jakarta[0.043]; Australia[0.043]; Timor[0.042]; Timor[0.042]; Catholic[0.042]; Norway[0.041]; Ali[0.041]; government[0.040]; said[0.040]; said[0.040]; said[0.040]; Islamic[0.040]; territory[0.040]; ceremony[0.040]; ceremony[0.040]; Alatas[0.040]; Alatas[0.040]; discuss[0.040]; figure[0.040]; Norwegian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alatas\u001b[39m ==> ENTITY: \u001b[32mAli Alatas\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlatas[0.045]; Alatas[0.045]; Indonesia[0.044]; Indonesia[0.044]; Jakarta[0.044]; Jakarta[0.044]; Jakarta[0.044]; Australia[0.043]; ministerial[0.042]; Foreign[0.041]; East[0.041]; East[0.041]; East[0.041]; Timor[0.041]; Timor[0.041]; Indonesian[0.041]; controversial[0.041]; Minister[0.040]; ceremony[0.040]; government[0.040]; government[0.040]; exile[0.040]; Conference[0.040]; Nations[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nobel Peace Prize\u001b[39m ==> ENTITY: \u001b[32mNobel Peace Prize\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNobel[0.047]; Nobel[0.047]; award[0.044]; award[0.044]; award[0.044]; said[0.042]; said[0.042]; recipient[0.041]; ceremony[0.041]; ceremony[0.041]; ceremony[0.041]; Tuesday[0.041]; activist[0.041]; Norway[0.041]; awards[0.041]; receive[0.041]; Oslo[0.041]; Friday[0.040]; Friday[0.040]; Alatas[0.040]; Norwegian[0.040]; inclusion[0.040]; exile[0.040]; condemned[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portuguese ==> ENTITY: \u001b[32mPortugal\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.103:0.103[0]; log p(e|m)= -1.127:-1.127[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nterritory[0.045]; colony[0.045]; Indonesia[0.045]; government[0.043]; Timor[0.043]; Timor[0.043]; Indonesian[0.042]; Jakarta[0.042]; Jakarta[0.042]; Jakarta[0.042]; Norway[0.041]; rule[0.041]; annexed[0.041]; Horta[0.041]; Horta[0.041]; Horta[0.041]; East[0.040]; East[0.040]; Ramos[0.040]; Ramos[0.040]; Ramos[0.040]; following[0.040]; ambassador[0.040]; different[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.189:0.189[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesian[0.047]; ambassador[0.045]; Jakarta[0.044]; Jakarta[0.044]; Jakarta[0.044]; Timor[0.043]; Timor[0.043]; government[0.042]; Norway[0.042]; territory[0.042]; efforts[0.041]; Alatas[0.041]; Alatas[0.041]; Alatas[0.041]; Alatas[0.041]; visit[0.041]; ceremony[0.041]; Portuguese[0.040]; discuss[0.040]; week[0.040]; colony[0.040]; following[0.040]; Tuesday[0.040]; conference[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.110:0.110[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPortuguese[0.048]; Norway[0.044]; bishop[0.044]; Vatican[0.043]; Helmut[0.042]; ambassador[0.042]; government[0.042]; Nations[0.042]; official[0.041]; attended[0.041]; invaded[0.041]; busy[0.041]; month[0.041]; Pope[0.041]; said[0.041]; said[0.041]; following[0.041]; Bonn[0.041]; East[0.041]; East[0.041]; visit[0.041]; visit[0.041]; ceremony[0.040]; involved[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Helmut Kohl\u001b[39m ==> ENTITY: \u001b[32mHelmut Kohl\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.137:0.137[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKohl[0.046]; chancellor[0.045]; German[0.044]; Chancellor[0.043]; Tuesday[0.043]; Norway[0.043]; Bonn[0.042]; month[0.042]; visit[0.042]; visit[0.042]; ambassador[0.041]; said[0.041]; said[0.041]; government[0.041]; Prize[0.041]; Nations[0.041]; Nobel[0.041]; official[0.040]; ceremony[0.040]; Oslo[0.040]; attended[0.040]; Timor[0.040]; Probably[0.039]; awarded[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ramos Horta\u001b[39m ==> ENTITY: \u001b[32mJosé Ramos-Horta\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.222:0.222[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDili[0.044]; Timorese[0.044]; Nobel[0.043]; Nobel[0.043]; Timor[0.042]; Alatas[0.042]; Alatas[0.042]; Indonesia[0.042]; conference[0.042]; award[0.042]; award[0.042]; Minister[0.041]; Prize[0.041]; leader[0.041]; ceremony[0.041]; ceremony[0.041]; Indonesian[0.041]; Indonesian[0.041]; government[0.041]; Peace[0.041]; Horta[0.041]; Horta[0.041]; Horta[0.041]; exile[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Carlos Belo\u001b[39m ==> ENTITY: \u001b[32mCarlos Filipe Ximenes Belo\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimorese[0.048]; Timorese[0.048]; Dili[0.047]; Dili[0.047]; Timor[0.047]; Bishop[0.047]; Indonesia[0.047]; Belo[0.046]; Belo[0.046]; bishop[0.046]; Horta[0.045]; Horta[0.045]; Catholic[0.045]; Nobel[0.044]; Nobel[0.044]; Nobel[0.044]; Peace[0.044]; Jose[0.044]; government[0.044]; ceremony[0.044]; ceremony[0.044]; Jakarta[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Organisation of the Islamic Conference\u001b[39m ==> ENTITY: \u001b[32mOrganisation of Islamic Cooperation\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.046]; Indonesia[0.046]; conference[0.044]; Australia[0.043]; meeting[0.043]; Foreign[0.043]; inclusion[0.042]; week[0.042]; ambassador[0.042]; ministerial[0.041]; condemned[0.041]; Norway[0.041]; government[0.041]; government[0.041]; ceremony[0.041]; ceremony[0.041]; efforts[0.041]; represented[0.041]; officially[0.040]; following[0.040]; said[0.040]; said[0.040]; said[0.040]; territory[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = United Nations\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.104:0.104[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPeace[0.044]; ambassador[0.044]; Norway[0.043]; Nobel[0.042]; territory[0.042]; issue[0.042]; government[0.042]; Organisation[0.042]; following[0.042]; Prize[0.042]; official[0.042]; annexed[0.041]; called[0.041]; position[0.041]; Tuesday[0.041]; Indonesia[0.041]; meeting[0.041]; week[0.041]; efforts[0.041]; Ramos[0.041]; Ramos[0.041]; Ramos[0.041]; secure[0.041]; Friday[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alatas\u001b[39m ==> ENTITY: \u001b[32mAli Alatas\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlatas[0.045]; Alatas[0.045]; Alatas[0.045]; Indonesia[0.043]; Indonesia[0.043]; ambassador[0.043]; Jakarta[0.043]; Jakarta[0.043]; Jakarta[0.043]; ministerial[0.042]; Helmut[0.042]; visit[0.041]; visit[0.041]; Bonn[0.041]; month[0.041]; East[0.040]; East[0.040]; Timor[0.040]; Timor[0.040]; Indonesian[0.040]; ceremony[0.040]; government[0.039]; Conference[0.039]; Nations[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oslo ==> ENTITY: \u001b[32mOslo\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.036:-0.036[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.050]; Norwegian[0.048]; government[0.045]; Peace[0.045]; territory[0.044]; Nobel[0.043]; Nobel[0.043]; Nobel[0.043]; East[0.043]; East[0.043]; East[0.043]; East[0.043]; capital[0.043]; capital[0.043]; people[0.043]; solution[0.043]; officially[0.042]; condemned[0.042]; lives[0.042]; said[0.042]; said[0.042]; Foreign[0.042]; jointly[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bonn\u001b[39m ==> ENTITY: \u001b[32mBonn\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.129:0.129[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nchancellor[0.045]; East[0.044]; German[0.044]; Tuesday[0.043]; said[0.042]; said[0.042]; Pope[0.042]; Chancellor[0.042]; Vatican[0.042]; month[0.042]; attended[0.041]; position[0.041]; Oslo[0.041]; Nobel[0.041]; government[0.041]; Kohl[0.041]; Kohl[0.041]; bishop[0.041]; visit[0.041]; visit[0.041]; ambassador[0.041]; solely[0.041]; following[0.041]; come[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kohl\u001b[39m ==> ENTITY: \u001b[32mHelmut Kohl\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.142:0.142[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKohl[0.045]; chancellor[0.045]; Helmut[0.044]; German[0.044]; Chancellor[0.043]; Tuesday[0.043]; Norway[0.043]; Bonn[0.042]; month[0.042]; visit[0.041]; visit[0.041]; ambassador[0.041]; said[0.041]; said[0.041]; government[0.041]; Prize[0.041]; Nations[0.041]; Nobel[0.040]; official[0.040]; ceremony[0.040]; Oslo[0.040]; attended[0.040]; Timor[0.040]; Probably[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ali Alatas\u001b[39m ==> ENTITY: \u001b[32mAli Alatas\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlatas[0.046]; Indonesia[0.044]; Jakarta[0.044]; Jakarta[0.044]; Jakarta[0.044]; Australia[0.044]; ministerial[0.043]; Foreign[0.041]; East[0.041]; East[0.041]; Timor[0.041]; Indonesian[0.041]; Indonesian[0.041]; controversial[0.041]; Dili[0.041]; Minister[0.041]; ceremony[0.040]; ceremony[0.040]; government[0.040]; awards[0.040]; exile[0.040]; Conference[0.040]; meeting[0.040]; Oslo[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = OIC\u001b[39m ==> ENTITY: \u001b[32mOrganisation of Islamic Cooperation\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOrganisation[0.045]; Islamic[0.045]; ambassador[0.044]; conference[0.044]; Conference[0.043]; meeting[0.043]; Foreign[0.043]; different[0.042]; inclusion[0.042]; Indonesia[0.041]; Indonesia[0.041]; week[0.041]; Indonesian[0.041]; Indonesian[0.041]; ministerial[0.041]; Peace[0.041]; condemned[0.041]; government[0.040]; government[0.040]; Timor[0.040]; Timor[0.040]; ceremony[0.040]; ceremony[0.040]; responding[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Belo\u001b[39m ==> ENTITY: \u001b[32mCarlos Filipe Ximenes Belo\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimorese[0.044]; Timorese[0.044]; Dili[0.043]; Dili[0.043]; Timor[0.043]; Bishop[0.043]; Indonesia[0.043]; Indonesia[0.043]; Belo[0.042]; Belo[0.042]; bishop[0.042]; Horta[0.041]; Horta[0.041]; Catholic[0.041]; Nobel[0.041]; Nobel[0.041]; Nobel[0.041]; Peace[0.041]; Jose[0.040]; government[0.040]; ceremony[0.040]; ceremony[0.040]; ceremony[0.040]; believe[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nobel Peace Prize\u001b[39m ==> ENTITY: \u001b[32mNobel Peace Prize\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnews[0.044]; prize[0.043]; visit[0.043]; visit[0.043]; ambassador[0.043]; awarded[0.042]; month[0.042]; said[0.042]; said[0.042]; said[0.042]; week[0.042]; ceremony[0.042]; Tuesday[0.042]; Norway[0.041]; Pope[0.041]; conference[0.041]; Oslo[0.041]; opposition[0.041]; discuss[0.041]; leader[0.041]; Friday[0.041]; Alatas[0.040]; Alatas[0.040]; Alatas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Belo\u001b[39m ==> ENTITY: \u001b[32mCarlos Filipe Ximenes Belo\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimor[0.045]; Timor[0.045]; Indonesia[0.045]; Indonesia[0.045]; Belo[0.045]; Belo[0.045]; Belo[0.045]; Belo[0.045]; bishop[0.044]; Helmut[0.044]; Pope[0.044]; Horta[0.043]; Horta[0.043]; month[0.043]; Bonn[0.043]; Nobel[0.043]; Peace[0.043]; Portuguese[0.043]; government[0.042]; visit[0.042]; visit[0.042]; ceremony[0.042]; peaceful[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = East Timor ==> ENTITY: \u001b[32mEast Timor\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.164:0.164[0]; log p(e|m)= -0.182:-0.182[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.045]; Timorese[0.044]; Timorese[0.044]; government[0.044]; Australia[0.043]; Catholic[0.043]; territory[0.043]; Dili[0.042]; Dili[0.042]; East[0.042]; East[0.042]; awards[0.041]; Indonesian[0.041]; Indonesian[0.041]; Prize[0.041]; Belo[0.040]; Belo[0.040]; Belo[0.040]; Norway[0.040]; arrive[0.040]; Peace[0.040]; Horta[0.040]; award[0.040]; award[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Belo\u001b[39m ==> ENTITY: \u001b[32mCarlos Filipe Ximenes Belo\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.142:0.142[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimor[0.044]; Indonesia[0.044]; Belo[0.043]; Belo[0.043]; Belo[0.043]; bishop[0.043]; Helmut[0.043]; Pope[0.042]; month[0.042]; Bonn[0.042]; Nobel[0.042]; Peace[0.042]; government[0.041]; visit[0.041]; visit[0.041]; ceremony[0.041]; Jakarta[0.041]; Jakarta[0.041]; ambassador[0.041]; position[0.041]; Alatas[0.040]; Alatas[0.040]; Alatas[0.040]; Indonesian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.162:0.162[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesian[0.049]; ambassador[0.048]; Jakarta[0.046]; Jakarta[0.046]; Timor[0.045]; Norway[0.044]; month[0.043]; government[0.043]; Alatas[0.043]; Alatas[0.043]; Alatas[0.043]; visit[0.043]; visit[0.043]; ceremony[0.043]; wanted[0.043]; Tuesday[0.043]; East[0.042]; meet[0.042]; meet[0.042]; Oslo[0.042]; awarded[0.041]; official[0.041]; attended[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Belo\u001b[39m ==> ENTITY: \u001b[32mCarlos Filipe Ximenes Belo\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimorese[0.044]; Timorese[0.044]; Dili[0.044]; Dili[0.044]; Timor[0.043]; Bishop[0.043]; Indonesia[0.043]; Belo[0.043]; Belo[0.043]; bishop[0.042]; Catholic[0.041]; Nobel[0.041]; Nobel[0.041]; Nobel[0.041]; Peace[0.041]; Jose[0.041]; ceremony[0.040]; ceremony[0.040]; Jakarta[0.040]; award[0.040]; award[0.040]; Ramos[0.040]; Carlos[0.040]; receive[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.193:0.193[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesian[0.047]; Indonesian[0.047]; Australia[0.044]; Jakarta[0.044]; Jakarta[0.044]; Jakarta[0.044]; people[0.043]; Timor[0.042]; Norway[0.042]; leader[0.042]; government[0.041]; territory[0.040]; territory[0.040]; ceremony[0.040]; ceremony[0.040]; capital[0.040]; capital[0.040]; Alatas[0.040]; Alatas[0.040]; week[0.040]; conference[0.040]; opposition[0.040]; Islamic[0.040]; Norwegian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dili\u001b[39m ==> ENTITY: \u001b[32mDili\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbishop[0.044]; Timor[0.044]; Dili[0.043]; Catholic[0.042]; Belo[0.042]; Belo[0.042]; Belo[0.042]; capital[0.042]; Bishop[0.042]; Carlos[0.042]; Timorese[0.042]; Timorese[0.042]; Horta[0.042]; Horta[0.042]; government[0.042]; Jose[0.041]; Indonesia[0.041]; East[0.041]; East[0.041]; East[0.041]; Jakarta[0.040]; Tuesday[0.040]; Indonesian[0.040]; Indonesian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jakarta\u001b[39m ==> ENTITY: \u001b[32mJakarta\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.050]; Indonesia[0.050]; Jakarta[0.047]; Jakarta[0.047]; Indonesian[0.044]; Indonesian[0.044]; Australia[0.042]; colony[0.042]; Timor[0.041]; Timor[0.041]; government[0.041]; government[0.041]; Norway[0.040]; annexed[0.040]; East[0.040]; East[0.040]; Friday[0.039]; Friday[0.039]; territory[0.039]; capital[0.039]; represented[0.039]; conference[0.038]; efforts[0.038]; following[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesian ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.224:0.224[0]; log p(e|m)= -0.742:-0.742[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.051]; Indonesia[0.051]; Jakarta[0.043]; Jakarta[0.043]; Jakarta[0.043]; Jakarta[0.043]; Timor[0.042]; Timor[0.042]; Norway[0.041]; East[0.041]; East[0.041]; said[0.040]; said[0.040]; Islamic[0.040]; territory[0.040]; ceremony[0.040]; Portuguese[0.040]; different[0.040]; Alatas[0.040]; Alatas[0.040]; Alatas[0.040]; colony[0.040]; attended[0.040]; month[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.087:0.087[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nambassador[0.045]; Nobel[0.044]; Oslo[0.043]; government[0.042]; awarded[0.042]; Nations[0.042]; Conference[0.042]; Indonesia[0.042]; Indonesia[0.042]; month[0.042]; solely[0.042]; following[0.042]; United[0.041]; territory[0.041]; Chancellor[0.041]; chancellor[0.041]; opposition[0.041]; prize[0.041]; position[0.041]; Probably[0.041]; official[0.041]; ceremony[0.041]; visit[0.041]; visit[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nobel ==> ENTITY: \u001b[32mNobel Peace Prize\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.206:0.206[0]; log p(e|m)= -2.674:-2.674[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNobel[0.047]; Nobel[0.047]; award[0.043]; award[0.043]; award[0.043]; Peace[0.043]; ceremony[0.043]; ceremony[0.043]; ceremony[0.043]; Prize[0.042]; awards[0.042]; Norwegian[0.041]; recipient[0.041]; Oslo[0.041]; Norway[0.040]; activist[0.040]; Foreign[0.040]; condemned[0.040]; receive[0.040]; said[0.040]; said[0.040]; said[0.040]; inclusion[0.039]; Tuesday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = East Timor ==> ENTITY: \u001b[32mEast Timor\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.182:-0.182[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.047]; Pope[0.044]; month[0.043]; Asked[0.043]; Indonesian[0.043]; Vatican[0.042]; involved[0.042]; Norway[0.042]; Peace[0.042]; official[0.042]; ceremony[0.041]; Jakarta[0.041]; visit[0.041]; visit[0.041]; Tuesday[0.041]; wanted[0.041]; ambassador[0.041]; position[0.041]; said[0.040]; Alatas[0.040]; Alatas[0.040]; yes[0.040]; Oslo[0.040]; bishop[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jakarta\u001b[39m ==> ENTITY: \u001b[32mJakarta\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.118:0.118[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.049]; Indonesian[0.045]; Tuesday[0.044]; Timor[0.043]; East[0.043]; month[0.042]; visit[0.042]; visit[0.042]; bishop[0.042]; Vatican[0.041]; Alatas[0.041]; Alatas[0.041]; different[0.041]; official[0.041]; ceremony[0.041]; Helmut[0.040]; German[0.040]; attended[0.040]; Bonn[0.040]; awarded[0.040]; involved[0.040]; Norway[0.040]; Pope[0.040]; chancellor[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Roman Catholic\u001b[39m ==> ENTITY: \u001b[32mCatholic Church\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.047:0.047[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.049]; Bishop[0.048]; Jose[0.045]; bishop[0.045]; ceremony[0.045]; ceremony[0.045]; inclusion[0.043]; Carlos[0.043]; territory[0.043]; people[0.043]; Timorese[0.043]; Timorese[0.043]; Indonesia[0.043]; Peace[0.043]; said[0.043]; Belo[0.042]; Belo[0.042]; Belo[0.042]; Dili[0.042]; Dili[0.042]; Ramos[0.042]; Ramos[0.042]; Tuesday[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norwegian ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.088:0.088[0]; log p(e|m)= -0.607:-0.607[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nofficially[0.044]; Oslo[0.044]; award[0.043]; award[0.043]; born[0.043]; ceremony[0.043]; Nobel[0.043]; Nobel[0.043]; exile[0.043]; conference[0.042]; week[0.042]; prize[0.042]; inclusion[0.041]; territory[0.041]; territory[0.041]; people[0.041]; finding[0.041]; represented[0.040]; figure[0.040]; receive[0.040]; Australia[0.040]; Conference[0.040]; Tuesday[0.040]; secure[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nobel ==> ENTITY: \u001b[32mNobel Peace Prize\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.187:0.187[0]; log p(e|m)= -2.674:-2.674[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNobel[0.048]; Nobel[0.048]; award[0.044]; award[0.044]; Peace[0.044]; ceremony[0.044]; ceremony[0.044]; Prize[0.043]; awards[0.043]; recipient[0.042]; Oslo[0.041]; Norway[0.041]; activist[0.041]; receive[0.041]; said[0.041]; Tuesday[0.040]; people[0.040]; Friday[0.039]; Indonesia[0.039]; Carlos[0.039]; Roman[0.039]; Jose[0.039]; accompanied[0.038]; Indonesian[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ramos Horta\u001b[39m ==> ENTITY: \u001b[32mJosé Ramos-Horta\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.224:0.224[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNobel[0.043]; Timor[0.042]; Timor[0.042]; Alatas[0.042]; Alatas[0.042]; Alatas[0.042]; Indonesia[0.042]; Indonesia[0.042]; conference[0.042]; award[0.042]; Minister[0.042]; ambassador[0.042]; Prize[0.041]; leader[0.041]; ceremony[0.041]; ceremony[0.041]; Indonesian[0.041]; Indonesian[0.041]; government[0.041]; government[0.041]; Peace[0.041]; Horta[0.041]; Horta[0.041]; Horta[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesian ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.203:0.203[0]; log p(e|m)= -0.742:-0.742[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.051]; Indonesia[0.051]; Indonesian[0.046]; Jakarta[0.044]; Australia[0.043]; people[0.042]; Timor[0.042]; Catholic[0.041]; Norway[0.041]; ceremony[0.041]; ceremony[0.041]; ceremony[0.041]; East[0.041]; East[0.041]; East[0.041]; said[0.040]; said[0.040]; territory[0.040]; Alatas[0.040]; Tuesday[0.040]; Norwegian[0.039]; born[0.039]; Friday[0.039]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Belo\u001b[39m ==> ENTITY: \u001b[32mCarlos Filipe Ximenes Belo\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimor[0.045]; Timor[0.045]; Indonesia[0.045]; Indonesia[0.045]; Belo[0.045]; Belo[0.045]; Belo[0.045]; Belo[0.045]; bishop[0.044]; Helmut[0.044]; Pope[0.044]; Horta[0.043]; Horta[0.043]; month[0.043]; Bonn[0.043]; Nobel[0.043]; Peace[0.043]; Portuguese[0.043]; government[0.042]; visit[0.042]; visit[0.042]; ceremony[0.042]; peaceful[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alatas\u001b[39m ==> ENTITY: \u001b[32mAli Alatas\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlatas[0.047]; Alatas[0.047]; Indonesia[0.045]; Indonesia[0.045]; ambassador[0.045]; Jakarta[0.045]; Jakarta[0.045]; Jakarta[0.045]; Helmut[0.044]; visit[0.043]; visit[0.043]; Bonn[0.043]; month[0.043]; East[0.042]; East[0.042]; Timor[0.042]; Timor[0.042]; Indonesian[0.042]; ceremony[0.042]; government[0.041]; Nations[0.041]; Chancellor[0.041]; Oslo[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jose Ramos Horta\u001b[39m ==> ENTITY: \u001b[32mJosé Ramos-Horta\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.231:0.231[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDili[0.044]; Dili[0.044]; Timorese[0.043]; Timorese[0.043]; Nobel[0.042]; Nobel[0.042]; Nobel[0.042]; Timor[0.042]; Timor[0.042]; Alatas[0.042]; Alatas[0.042]; Indonesia[0.042]; Indonesia[0.042]; conference[0.041]; award[0.041]; award[0.041]; award[0.041]; Minister[0.041]; Prize[0.041]; ceremony[0.041]; ceremony[0.041]; ceremony[0.041]; Indonesian[0.041]; Indonesian[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ramos Horta\u001b[39m ==> ENTITY: \u001b[32mJosé Ramos-Horta\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimorese[0.044]; Nobel[0.043]; Timor[0.042]; Timor[0.042]; Alatas[0.042]; Alatas[0.042]; Indonesia[0.042]; Indonesia[0.042]; conference[0.042]; award[0.042]; award[0.042]; Minister[0.042]; leader[0.041]; ceremony[0.041]; Indonesian[0.041]; government[0.041]; Horta[0.041]; Horta[0.041]; Horta[0.041]; Horta[0.041]; exile[0.041]; said[0.041]; said[0.041]; ministerial[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.069:0.069[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNobel[0.044]; Nobel[0.044]; Nobel[0.044]; Oslo[0.043]; Foreign[0.042]; Australia[0.042]; government[0.042]; territory[0.042]; Indonesia[0.042]; Indonesia[0.042]; award[0.041]; award[0.041]; award[0.041]; receive[0.041]; East[0.041]; East[0.041]; East[0.041]; inclusion[0.041]; left[0.041]; left[0.041]; way[0.041]; immediately[0.041]; people[0.041]; awards[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Belo\u001b[39m ==> ENTITY: \u001b[32mCarlos Filipe Ximenes Belo\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimor[0.044]; Timor[0.044]; Indonesia[0.043]; Indonesia[0.043]; Belo[0.043]; Belo[0.043]; Horta[0.042]; Horta[0.042]; Horta[0.042]; conference[0.042]; Nobel[0.041]; Peace[0.041]; Portuguese[0.041]; government[0.041]; ceremony[0.041]; ceremony[0.041]; believe[0.041]; peaceful[0.041]; Jakarta[0.041]; Jakarta[0.041]; Jakarta[0.041]; ambassador[0.041]; position[0.041]; week[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesian[0.049]; Jakarta[0.045]; Timor[0.044]; Norway[0.044]; territory[0.042]; East[0.042]; East[0.042]; East[0.042]; ceremony[0.042]; ceremony[0.042]; people[0.041]; Tuesday[0.041]; Peace[0.041]; Timorese[0.040]; Timorese[0.040]; awards[0.040]; Dili[0.040]; Dili[0.040]; jointly[0.040]; said[0.040]; capital[0.040]; Oslo[0.040]; award[0.040]; award[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dili\u001b[39m ==> ENTITY: \u001b[32mDili\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbishop[0.044]; Timor[0.044]; Dili[0.043]; Catholic[0.043]; Belo[0.042]; Belo[0.042]; Belo[0.042]; capital[0.042]; Bishop[0.042]; Carlos[0.042]; Timorese[0.042]; Timorese[0.042]; Horta[0.042]; Jose[0.042]; Indonesia[0.041]; East[0.041]; East[0.041]; East[0.041]; Jakarta[0.041]; Tuesday[0.040]; Indonesian[0.040]; ceremony[0.040]; ceremony[0.040]; territory[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alatas\u001b[39m ==> ENTITY: \u001b[32mAli Alatas\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlatas[0.045]; Alatas[0.045]; Indonesia[0.044]; Indonesia[0.044]; ambassador[0.044]; Jakarta[0.044]; Jakarta[0.044]; Helmut[0.042]; visit[0.042]; visit[0.042]; Bonn[0.041]; month[0.041]; East[0.041]; East[0.041]; Timor[0.041]; Timor[0.041]; Indonesian[0.041]; ceremony[0.040]; government[0.040]; Nations[0.040]; Chancellor[0.040]; Oslo[0.040]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oslo ==> ENTITY: \u001b[32mOslo\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.114:0.114[0]; log p(e|m)= -0.036:-0.036[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.048]; chancellor[0.043]; following[0.043]; busy[0.043]; government[0.042]; Nobel[0.042]; awarded[0.042]; East[0.042]; East[0.042]; ambassador[0.042]; Bonn[0.042]; solution[0.041]; month[0.041]; official[0.041]; visit[0.041]; visit[0.041]; Peace[0.041]; said[0.041]; said[0.041]; secure[0.041]; invaded[0.041]; attended[0.040]; issue[0.040]; Probably[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.087:0.087[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nweek[0.044]; government[0.044]; Indonesia[0.043]; Timor[0.043]; Timor[0.043]; Norway[0.043]; inclusion[0.042]; territory[0.042]; Tuesday[0.042]; Timorese[0.042]; Timorese[0.042]; ministerial[0.042]; Indonesian[0.041]; Indonesian[0.041]; officially[0.041]; jointly[0.041]; represented[0.041]; receive[0.041]; Foreign[0.041]; ceremony[0.041]; ceremony[0.041]; Friday[0.040]; Friday[0.040]; Friday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = East Timor ==> ENTITY: \u001b[32mEast Timor\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.174:0.174[0]; log p(e|m)= -0.182:-0.182[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTimor[0.045]; Indonesia[0.045]; Indonesia[0.045]; government[0.044]; government[0.044]; Timorese[0.043]; following[0.043]; Australia[0.043]; territory[0.042]; East[0.041]; East[0.041]; week[0.041]; Foreign[0.041]; Tuesday[0.041]; Nations[0.041]; ministerial[0.041]; secure[0.041]; conference[0.040]; Indonesian[0.040]; peaceful[0.040]; Friday[0.040]; Friday[0.040]; questions[0.040]; Jose[0.040]; \t\n\r [===================>..........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s505ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2019/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1335testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Seoul ==> ENTITY: \u001b[32mSeoul\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.038:-0.038[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHyundai[0.045]; won[0.045]; won[0.045]; won[0.045]; won[0.045]; Korean[0.044]; South[0.042]; set[0.042]; players[0.042]; opening[0.041]; world[0.041]; week[0.041]; earlier[0.041]; Friday[0.041]; Bank[0.041]; day[0.041]; Koram[0.041]; ended[0.040]; million[0.040]; Monday[0.040]; Monday[0.040]; dollar[0.039]; dollar[0.039]; dollar[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.076:0.076[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayers[0.045]; million[0.045]; won[0.045]; won[0.045]; won[0.045]; won[0.045]; dollar[0.044]; dollar[0.044]; dollar[0.044]; week[0.042]; South[0.041]; day[0.041]; ended[0.041]; earlier[0.041]; Bank[0.040]; opening[0.040]; world[0.040]; set[0.039]; slid[0.039]; Friday[0.039]; Monday[0.039]; Monday[0.039]; helped[0.039]; market[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = South Korean\u001b[39m ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.125:0.125[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHyundai[0.047]; Seoul[0.046]; dollar[0.046]; dollar[0.046]; million[0.045]; players[0.045]; Bank[0.043]; won[0.043]; won[0.043]; won[0.043]; yen[0.042]; Koram[0.041]; day[0.040]; market[0.040]; world[0.040]; earlier[0.040]; Monday[0.040]; Monday[0.040]; opening[0.039]; ended[0.039]; Friday[0.039]; helped[0.038]; set[0.038]; Heavy[0.038]; \t\n\r [===================>..........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s593ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2022/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1376testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Harare\u001b[39m ==> ENTITY: \u001b[32mHarare\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZimbabwe[0.046]; Zimbabwe[0.046]; Zimbabwe[0.046]; Zimbabwe[0.046]; Zimbabwe[0.046]; South[0.044]; Saturday[0.043]; African[0.043]; Scores[0.042]; Hugh[0.041]; Brett[0.040]; Trevor[0.040]; scores[0.040]; Mark[0.040]; Mark[0.040]; Mark[0.040]; Cayeux[0.040]; Golf[0.040]; round[0.039]; Schalk[0.039]; Merwe[0.039]; Greg[0.039]; der[0.039]; Steve[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nick Price\u001b[39m ==> ENTITY: \u001b[32mNick Price\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.045]; Open[0.045]; round[0.045]; Golf[0.044]; Murless[0.043]; Greg[0.043]; Cayeux[0.043]; Cayeux[0.043]; Mark[0.042]; Mark[0.042]; Mark[0.042]; Steve[0.042]; Glen[0.042]; Brett[0.041]; Trevor[0.041]; Justin[0.041]; Nic[0.040]; African[0.040]; South[0.040]; scores[0.040]; Andrew[0.040]; Schalk[0.039]; Merwe[0.039]; Sean[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bradford Vaughan\u001b[39m ==> ENTITY: \u001b[32mBradford Vaughan\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.102:0.102[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCayeux[0.048]; Cayeux[0.048]; Nic[0.044]; Murless[0.043]; Schalk[0.043]; Merwe[0.043]; der[0.042]; Hobday[0.042]; Glen[0.042]; Brett[0.042]; Botes[0.041]; Henning[0.041]; Zimbabwe[0.041]; Zimbabwe[0.041]; Zimbabwe[0.041]; Hennie[0.041]; Sean[0.040]; Vuuren[0.040]; Greg[0.040]; Trevor[0.040]; Namibia[0.040]; Namibia[0.040]; Swart[0.039]; Hugh[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Namibia ==> ENTITY: \u001b[32mNamibia\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.160:0.160[0]; log p(e|m)= -0.368:-0.368[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZimbabwe[0.045]; Zimbabwe[0.045]; Zimbabwe[0.045]; Zimbabwe[0.045]; Zimbabwe[0.045]; Zimbabwe[0.045]; Zimbabwe[0.045]; Namibia[0.045]; South[0.043]; African[0.043]; round[0.041]; scores[0.041]; Terblanche[0.040]; Harare[0.040]; Round[0.040]; Fourie[0.040]; Mark[0.039]; Mark[0.039]; Mark[0.039]; Swart[0.039]; Glen[0.039]; Merwe[0.039]; Steve[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Brett Liddle\u001b[39m ==> ENTITY: \u001b[32mBrett Liddle\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.049]; Cayeux[0.046]; Cayeux[0.046]; African[0.046]; Murless[0.044]; Hennie[0.044]; Golf[0.043]; Terblanche[0.043]; Nick[0.043]; Nic[0.043]; Trevor[0.043]; Mark[0.043]; Mark[0.043]; Mark[0.043]; Andrew[0.043]; Andrew[0.043]; round[0.043]; Fourie[0.042]; Saturday[0.042]; Botes[0.042]; Schalk[0.042]; Steven[0.042]; Swart[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zimbabwe ==> ENTITY: \u001b[32mZimbabwe\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.224:0.224[0]; log p(e|m)= -0.367:-0.367[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZimbabwe[0.049]; Zimbabwe[0.049]; Zimbabwe[0.049]; Zimbabwe[0.049]; Zimbabwe[0.049]; Zimbabwe[0.049]; South[0.043]; Harare[0.042]; African[0.042]; Namibia[0.041]; Namibia[0.041]; Swart[0.040]; Cayeux[0.039]; Cayeux[0.039]; Saturday[0.039]; Golf[0.038]; scores[0.038]; Brett[0.038]; Fourie[0.038]; Glen[0.038]; Steve[0.038]; Trevor[0.038]; Andrew[0.038]; round[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zimbabwe ==> ENTITY: \u001b[32mZimbabwe\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.223:0.223[0]; log p(e|m)= -0.367:-0.367[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZimbabwe[0.049]; Zimbabwe[0.049]; Zimbabwe[0.049]; Zimbabwe[0.049]; Zimbabwe[0.049]; Zimbabwe[0.049]; South[0.043]; Harare[0.042]; African[0.042]; Namibia[0.041]; Namibia[0.041]; Swart[0.040]; Cayeux[0.039]; Cayeux[0.039]; Saturday[0.039]; Golf[0.038]; scores[0.038]; Brett[0.038]; Glen[0.038]; Steve[0.038]; Trevor[0.038]; Andrew[0.038]; round[0.038]; Terblanche[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.054:0.054[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.043]; Steve[0.043]; Mark[0.043]; Mark[0.043]; Mark[0.043]; round[0.043]; Steven[0.043]; Park[0.043]; Price[0.042]; Bradford[0.042]; Glen[0.042]; Sean[0.042]; Saturday[0.042]; Greg[0.042]; Open[0.042]; African[0.041]; stated[0.040]; Dion[0.040]; McNulty[0.040]; Andrew[0.040]; Andrew[0.040]; Farrell[0.040]; Nick[0.040]; Pitts[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hugh Baiocchi\u001b[39m ==> ENTITY: \u001b[32mHugh Baiocchi\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nround[0.046]; Golf[0.045]; Cayeux[0.044]; Cayeux[0.044]; South[0.043]; African[0.043]; Nic[0.043]; Murless[0.042]; Open[0.041]; Open[0.041]; Schalk[0.041]; Merwe[0.041]; der[0.041]; Glen[0.041]; Round[0.041]; Des[0.041]; Brett[0.041]; McNulty[0.040]; Justin[0.040]; Hennie[0.040]; scores[0.040]; Park[0.040]; Sean[0.040]; Henning[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Clinton Whitelaw\u001b[39m ==> ENTITY: \u001b[32mClinton Whitelaw\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrican[0.045]; Cayeux[0.045]; Cayeux[0.045]; Open[0.044]; Open[0.044]; round[0.044]; McNulty[0.043]; South[0.042]; Nic[0.041]; Murless[0.041]; Glen[0.041]; Golf[0.041]; Saturday[0.041]; Harare[0.040]; Henning[0.040]; Zimbabwe[0.040]; Zimbabwe[0.040]; Zimbabwe[0.040]; Zimbabwe[0.040]; Zimbabwe[0.040]; Zimbabwe[0.040]; Zimbabwe[0.040]; Schalk[0.040]; Merwe[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zimbabwe Open\u001b[39m ==> ENTITY: \u001b[32mZimbabwe Open\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.046]; Golf[0.045]; Cayeux[0.045]; Murless[0.043]; round[0.042]; Steve[0.042]; Zimbabwe[0.042]; Zimbabwe[0.042]; Zimbabwe[0.042]; Zimbabwe[0.042]; Mark[0.042]; Mark[0.042]; Mark[0.042]; Nick[0.041]; African[0.041]; Saturday[0.041]; Greg[0.041]; Justin[0.041]; Schalk[0.041]; Merwe[0.041]; South[0.040]; McNulty[0.040]; der[0.040]; Price[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South African ==> ENTITY: \u001b[32mSouth Africa\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.044]; Open[0.044]; Zimbabwe[0.044]; Zimbabwe[0.044]; Zimbabwe[0.044]; Zimbabwe[0.044]; Zimbabwe[0.044]; Zimbabwe[0.044]; round[0.043]; Saturday[0.042]; Namibia[0.042]; Namibia[0.042]; Golf[0.042]; Murless[0.042]; Cayeux[0.041]; Brett[0.041]; Harare[0.040]; Schalk[0.040]; scores[0.039]; Swart[0.039]; Round[0.039]; Scores[0.039]; Vuuren[0.039]; Andrew[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Trevor Dodds\u001b[39m ==> ENTITY: \u001b[32mTrevor Dodds\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCayeux[0.046]; Cayeux[0.046]; Open[0.043]; Open[0.043]; Golf[0.042]; Zimbabwe[0.042]; Zimbabwe[0.042]; Zimbabwe[0.042]; Zimbabwe[0.042]; Zimbabwe[0.042]; Zimbabwe[0.042]; Zimbabwe[0.042]; Namibia[0.042]; Namibia[0.042]; Nic[0.041]; Glen[0.041]; Harare[0.041]; Botes[0.041]; Murless[0.041]; African[0.040]; round[0.040]; Schalk[0.040]; Merwe[0.040]; der[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zimbabwe ==> ENTITY: \u001b[32mZimbabwe\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.224:0.224[0]; log p(e|m)= -0.367:-0.367[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZimbabwe[0.050]; Zimbabwe[0.050]; Zimbabwe[0.050]; Zimbabwe[0.050]; Zimbabwe[0.050]; Zimbabwe[0.050]; South[0.044]; Harare[0.044]; African[0.044]; Namibia[0.042]; Namibia[0.042]; Swart[0.041]; Vaughan[0.041]; Cayeux[0.041]; Cayeux[0.041]; Saturday[0.041]; Golf[0.040]; scores[0.040]; Brett[0.040]; Fourie[0.039]; Glen[0.039]; Steve[0.039]; Trevor[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Des Terblanche\u001b[39m ==> ENTITY: \u001b[32mDes Terblanche\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCayeux[0.052]; Cayeux[0.052]; Murless[0.051]; South[0.050]; Golf[0.050]; African[0.050]; Open[0.050]; Open[0.050]; Hennie[0.047]; Nic[0.047]; Namibia[0.046]; Namibia[0.046]; van[0.046]; van[0.046]; Schalk[0.046]; Merwe[0.046]; Trevor[0.045]; der[0.045]; round[0.045]; McNulty[0.045]; Brett[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zimbabwe ==> ENTITY: \u001b[32mZimbabwe\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.166:0.166[0]; log p(e|m)= -0.367:-0.367[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZimbabwe[0.051]; Zimbabwe[0.051]; Zimbabwe[0.051]; Zimbabwe[0.051]; Namibia[0.043]; Namibia[0.043]; Swart[0.042]; Vaughan[0.041]; Cayeux[0.041]; Cayeux[0.041]; Brett[0.040]; Fourie[0.040]; Glen[0.040]; Steve[0.040]; Botes[0.039]; Trevor[0.039]; Andrew[0.039]; Andrew[0.039]; Park[0.039]; Terblanche[0.039]; Hugh[0.039]; Greg[0.039]; Mark[0.039]; Mark[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Namibia ==> ENTITY: \u001b[32mNamibia\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.160:0.160[0]; log p(e|m)= -0.368:-0.368[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZimbabwe[0.045]; Zimbabwe[0.045]; Zimbabwe[0.045]; Zimbabwe[0.045]; Zimbabwe[0.045]; Zimbabwe[0.045]; Zimbabwe[0.045]; Namibia[0.045]; South[0.043]; African[0.043]; round[0.041]; scores[0.041]; Terblanche[0.040]; Harare[0.040]; Round[0.040]; Fourie[0.040]; Mark[0.039]; Mark[0.039]; Mark[0.039]; Swart[0.039]; Glen[0.039]; Merwe[0.039]; Steve[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zimbabwe Open\u001b[39m ==> ENTITY: \u001b[32mZimbabwe Open\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.046]; Golf[0.045]; Cayeux[0.045]; Murless[0.043]; round[0.042]; Steve[0.042]; Zimbabwe[0.041]; Zimbabwe[0.041]; Zimbabwe[0.041]; Zimbabwe[0.041]; Mark[0.041]; Mark[0.041]; Mark[0.041]; Hennie[0.041]; Nick[0.041]; African[0.041]; Andrew[0.041]; Saturday[0.041]; Greg[0.040]; Justin[0.040]; Schalk[0.040]; Merwe[0.040]; South[0.040]; McNulty[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Justin Hobday\u001b[39m ==> ENTITY: \u001b[32mJustin Hobday\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCayeux[0.048]; Cayeux[0.048]; round[0.046]; Open[0.046]; Open[0.046]; African[0.045]; Golf[0.045]; South[0.045]; Nick[0.044]; Trevor[0.043]; Murless[0.043]; Nic[0.043]; Steve[0.042]; Mark[0.042]; Mark[0.042]; Mark[0.042]; McNulty[0.042]; scores[0.042]; Greg[0.042]; Glen[0.041]; Schalk[0.041]; Merwe[0.041]; Botes[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Desvonde Botes\u001b[39m ==> ENTITY: \u001b[32mDesvonde Botes\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.121:0.121[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMurless[0.046]; Cayeux[0.046]; Nic[0.044]; Schalk[0.043]; Merwe[0.043]; Hennie[0.042]; der[0.042]; Fourie[0.042]; Henning[0.042]; Brett[0.042]; Namibia[0.042]; Namibia[0.042]; Swart[0.041]; Mark[0.041]; van[0.041]; van[0.041]; Sean[0.041]; Vuuren[0.040]; Andrew[0.040]; Andrew[0.040]; Glen[0.040]; Greg[0.040]; Zimbabwe[0.040]; Zimbabwe[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mark McNulty\u001b[39m ==> ENTITY: \u001b[32mMark McNulty\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCayeux[0.044]; Cayeux[0.044]; Open[0.043]; Open[0.043]; Zimbabwe[0.043]; Zimbabwe[0.043]; Zimbabwe[0.043]; Zimbabwe[0.043]; Zimbabwe[0.043]; Zimbabwe[0.043]; Zimbabwe[0.043]; round[0.043]; Golf[0.043]; African[0.041]; Round[0.040]; Harare[0.040]; Murless[0.040]; Schalk[0.040]; Merwe[0.040]; der[0.039]; Glen[0.039]; Scores[0.039]; Namibia[0.039]; Namibia[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zimbabwe ==> ENTITY: \u001b[32mZimbabwe\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.185:0.185[0]; log p(e|m)= -0.367:-0.367[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZimbabwe[0.050]; Zimbabwe[0.050]; Zimbabwe[0.050]; Zimbabwe[0.050]; South[0.044]; African[0.043]; Namibia[0.042]; Namibia[0.042]; Swart[0.041]; Vaughan[0.041]; Cayeux[0.040]; Cayeux[0.040]; Brett[0.039]; Fourie[0.039]; Glen[0.039]; Steve[0.039]; Botes[0.039]; Trevor[0.039]; Andrew[0.039]; Andrew[0.039]; Park[0.039]; Terblanche[0.039]; Hugh[0.038]; Greg[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Steve van Vuuren\u001b[39m ==> ENTITY: \u001b[32mSteve van Vuuren\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCayeux[0.051]; Cayeux[0.051]; Murless[0.050]; Hennie[0.050]; Golf[0.049]; van[0.049]; Swart[0.048]; Mark[0.048]; Mark[0.048]; Mark[0.048]; South[0.047]; Fourie[0.047]; Schalk[0.047]; Namibia[0.046]; Namibia[0.046]; Terblanche[0.046]; Andrew[0.046]; Andrew[0.046]; African[0.046]; Nic[0.046]; Nick[0.046]; \t\n\r [===================>..........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s525ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2045/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1276testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= -0.021:-0.021[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYork[0.120]; York[0.120]; Desk[0.106]; China[0.099]; Dec[0.099]; Commodities[0.097]; tonnes[0.091]; freight[0.090]; soybeans[0.089]; fixtures[0.089]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= -0.000:-0.000[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYork[0.102]; York[0.102]; York[0.102]; Dec[0.089]; Desk[0.085]; fixtures[0.080]; Commodities[0.079]; freight[0.075]; grain[0.073]; soybeans[0.073]; tonnes[0.071]; Mana[0.070]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= -0.021:-0.021[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYork[0.120]; York[0.120]; Desk[0.106]; China[0.099]; Dec[0.099]; Commodities[0.097]; tonnes[0.091]; freight[0.090]; soybeans[0.089]; fixtures[0.089]; \t\n\r [====================>.........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s522ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2048/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1172testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.023:0.023[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.049]; won[0.046]; played[0.045]; championship[0.043]; Thursday[0.043]; Basketball[0.042]; men[0.042]; basketball[0.041]; lost[0.041]; Spain[0.041]; Maccabi[0.041]; Zagreb[0.041]; points[0.041]; Estudiantes[0.041]; Madrid[0.041]; Berlin[0.041]; Moscow[0.040]; EuroLeague[0.040]; Standings[0.040]; Standings[0.040]; Group[0.040]; Group[0.040]; Milan[0.040]; France[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Croatia ==> ENTITY: \u001b[32mCroatia\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.196:0.196[0]; log p(e|m)= -0.350:-0.350[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCroatia[0.048]; Slovenia[0.047]; Yugoslavia[0.045]; Italy[0.043]; Italy[0.043]; Zagreb[0.043]; Belgium[0.042]; Group[0.041]; Group[0.041]; Group[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Belgrade[0.041]; France[0.041]; France[0.041]; France[0.041]; Germany[0.041]; Germany[0.041]; Turkey[0.040]; Turkey[0.040]; Cibona[0.040]; Maccabi[0.040]; Russia[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCharleroi\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mR. Charleroi S.C.\u001b[39m <---> \u001b[32mSpirou Charleroi\u001b[39m\t\nSCORES: global= 0.257:0.251[\u001b[31m0.005\u001b[39m]; local()= 0.136:0.206[\u001b[32m0.070\u001b[39m]; log p(e|m)= -1.064:-1.938[\u001b[31m0.874\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; Belgium[0.045]; EuroLeague[0.044]; France[0.043]; France[0.043]; France[0.043]; Spain[0.042]; Spain[0.042]; Spain[0.042]; played[0.042]; basketball[0.042]; Partizan[0.042]; Villeurbanne[0.041]; Thursday[0.041]; won[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Panathinaikos[0.041]; Turkey[0.040]; Turkey[0.040]; Berlin[0.040]; Olympiakos[0.040]; Madrid[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBayer Leverkusen\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBayer 04 Leverkusen\u001b[39m <---> \u001b[32mBayer Giants Leverkusen\u001b[39m\t\nSCORES: global= 0.271:0.253[\u001b[31m0.018\u001b[39m]; local()= 0.135:0.147[\u001b[32m0.013\u001b[39m]; log p(e|m)= 0.000:-1.378[\u001b[31m1.378\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.044]; Germany[0.044]; Estudiantes[0.043]; Italy[0.043]; Italy[0.043]; Dynamo[0.043]; Madrid[0.043]; Olympiakos[0.042]; Partizan[0.042]; Barcelona[0.042]; Sevilla[0.042]; Panionios[0.042]; Panathinaikos[0.041]; Spor[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Charleroi[0.041]; Tel[0.041]; Orthez[0.040]; Berlin[0.040]; France[0.040]; France[0.040]; France[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.221:0.221[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.051]; EuroLeague[0.044]; matches[0.043]; Milan[0.042]; Spain[0.042]; Spain[0.042]; won[0.042]; Yugoslavia[0.042]; Slovenia[0.042]; France[0.041]; France[0.041]; France[0.041]; Basketball[0.041]; played[0.041]; Euroleague[0.041]; Bologna[0.041]; Germany[0.041]; Germany[0.041]; Greece[0.040]; Greece[0.040]; Greece[0.040]; Croatia[0.040]; Croatia[0.040]; Maccabi[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Limoges ==> ENTITY: \u001b[32mLimoges CSP\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.190:0.190[0]; log p(e|m)= -2.154:-2.154[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuroLeague[0.045]; Euroleague[0.045]; Basketball[0.044]; championship[0.043]; Cibona[0.043]; played[0.043]; France[0.043]; France[0.043]; points[0.042]; Leverkusen[0.042]; basketball[0.042]; matches[0.042]; won[0.042]; Spor[0.041]; Efes[0.040]; Villeurbanne[0.040]; Group[0.040]; Group[0.040]; Group[0.040]; Group[0.040]; Belgium[0.040]; Standings[0.040]; Standings[0.040]; lost[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBarcelona\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBarcelona\u001b[39m <---> \u001b[32mFC Barcelona Bàsquet\u001b[39m\t\nSCORES: global= 0.253:0.233[\u001b[31m0.020\u001b[39m]; local()= 0.185:0.173[\u001b[31m0.012\u001b[39m]; log p(e|m)= -0.371:-4.017[\u001b[31m3.646\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.049]; Spain[0.049]; Spain[0.049]; Estudiantes[0.047]; Madrid[0.046]; Italy[0.044]; Italy[0.044]; Italy[0.044]; Maccabi[0.043]; Cibona[0.042]; Villeurbanne[0.042]; Sevilla[0.042]; Dynamo[0.042]; France[0.042]; France[0.042]; France[0.042]; Milan[0.042]; Bologna[0.042]; Bologna[0.042]; Berlin[0.041]; Charleroi[0.041]; Ulker[0.041]; Belgium[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.201:0.201[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.051]; EuroLeague[0.046]; Euroleague[0.046]; Basketball[0.046]; matches[0.043]; Milan[0.042]; Spain[0.042]; Spain[0.042]; won[0.041]; Slovenia[0.041]; basketball[0.041]; Bologna[0.041]; France[0.041]; France[0.041]; played[0.041]; Germany[0.040]; Maccabi[0.040]; Greece[0.040]; Greece[0.040]; Greece[0.040]; Croatia[0.040]; championship[0.039]; points[0.039]; Cibona[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; Germany[0.045]; EuroLeague[0.045]; Euroleague[0.044]; Basketball[0.044]; Maccabi[0.043]; Spain[0.042]; Spain[0.042]; Spain[0.042]; Yugoslavia[0.041]; Slovenia[0.041]; Berlin[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; played[0.041]; Croatia[0.041]; Croatia[0.041]; Partizan[0.040]; championship[0.040]; basketball[0.040]; France[0.040]; France[0.040]; France[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kinder Bologna\u001b[39m ==> ENTITY: \u001b[32mVirtus Pallacanestro Bologna\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.050]; Italy[0.050]; Cibona[0.046]; Partizan[0.046]; Panathinaikos[0.045]; Leverkusen[0.045]; Villeurbanne[0.044]; Efes[0.043]; Olympiakos[0.043]; Greece[0.043]; Greece[0.043]; Croatia[0.042]; Croatia[0.042]; Zagreb[0.042]; Belgrade[0.042]; Slovenia[0.042]; Spain[0.042]; Spain[0.042]; Spain[0.042]; Yugoslavia[0.042]; Bayer[0.042]; Dynamo[0.041]; Barcelona[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.170:0.170[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.044]; Spain[0.044]; Spain[0.044]; Yugoslavia[0.044]; Slovenia[0.043]; Bologna[0.043]; France[0.043]; France[0.043]; Germany[0.042]; Germany[0.042]; Greece[0.042]; Greece[0.042]; Croatia[0.042]; Croatia[0.042]; Cibona[0.041]; Belgium[0.041]; Panathinaikos[0.041]; Turkey[0.040]; Estudiantes[0.040]; Dynamo[0.040]; Group[0.040]; Group[0.040]; Olympiakos[0.040]; Partizan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Slovenia ==> ENTITY: \u001b[32mSlovenia\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.218:0.218[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCroatia[0.049]; Croatia[0.049]; Italy[0.047]; Italy[0.047]; Italy[0.047]; Yugoslavia[0.047]; Cibona[0.046]; Zagreb[0.046]; Germany[0.045]; Germany[0.045]; Turkey[0.045]; Turkey[0.045]; Belgrade[0.045]; Spain[0.045]; Spain[0.045]; Spain[0.045]; won[0.045]; played[0.044]; points[0.044]; Belgium[0.044]; Russia[0.043]; Russia[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cibona Zagreb\u001b[39m ==> ENTITY: \u001b[32mKK Cibona\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.229:0.229[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuroLeague[0.049]; Euroleague[0.048]; Basketball[0.046]; Maccabi[0.045]; Belgrade[0.044]; won[0.044]; played[0.044]; basketball[0.044]; Croatia[0.044]; Croatia[0.044]; matches[0.044]; championship[0.044]; Leverkusen[0.043]; Partizan[0.043]; Ulker[0.043]; Yugoslavia[0.043]; Efes[0.043]; points[0.042]; Slovenia[0.042]; Spor[0.042]; Greece[0.040]; Greece[0.040]; Greece[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPanionios\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPanionios G.S.S.\u001b[39m <---> \u001b[32mPanionios B.C.\u001b[39m\t\nSCORES: global= 0.285:0.279[\u001b[31m0.006\u001b[39m]; local()= 0.184:0.247[\u001b[32m0.063\u001b[39m]; log p(e|m)= 0.000:-0.587[\u001b[31m0.587\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuroleague[0.045]; Basketball[0.044]; EuroLeague[0.044]; won[0.043]; Greece[0.043]; Greece[0.043]; Greece[0.043]; Maccabi[0.043]; basketball[0.042]; Spor[0.042]; Efes[0.042]; played[0.042]; Cibona[0.042]; Panathinaikos[0.042]; matches[0.041]; Ulker[0.041]; Leverkusen[0.041]; Olympiakos[0.040]; Turkey[0.040]; Turkey[0.040]; points[0.040]; championship[0.040]; Slovenia[0.039]; Thursday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.048]; Spain[0.045]; Spain[0.045]; Spain[0.045]; Yugoslavia[0.044]; Slovenia[0.044]; Berlin[0.044]; Italy[0.044]; Italy[0.044]; Croatia[0.044]; Croatia[0.044]; Partizan[0.043]; France[0.043]; France[0.043]; France[0.043]; Olympiakos[0.043]; Bayer[0.043]; Cibona[0.042]; Belgium[0.042]; Leverkusen[0.042]; Panathinaikos[0.042]; Israel[0.042]; Spor[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Maccabi Tel Aviv ==> ENTITY: \u001b[32mMaccabi Tel Aviv B.C.\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.229:0.229[0]; log p(e|m)= -0.330:-0.330[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuroleague[0.049]; EuroLeague[0.048]; Israel[0.045]; Basketball[0.043]; Cibona[0.043]; basketball[0.042]; matches[0.042]; played[0.042]; won[0.042]; Leverkusen[0.042]; Italy[0.041]; Italy[0.041]; championship[0.041]; Panathinaikos[0.041]; Olympiakos[0.040]; Turkey[0.040]; points[0.040]; Spor[0.040]; Estudiantes[0.040]; Thursday[0.040]; Ulker[0.040]; Spain[0.040]; Spain[0.040]; Germany[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgium ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.210:0.210[0]; log p(e|m)= -0.242:-0.242[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.045]; France[0.045]; France[0.045]; matches[0.044]; won[0.043]; Germany[0.043]; Germany[0.043]; Spain[0.042]; Spain[0.042]; Spain[0.042]; Slovenia[0.041]; Croatia[0.041]; Croatia[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; played[0.041]; championship[0.041]; Yugoslavia[0.040]; Standings[0.040]; EuroLeague[0.040]; Turkey[0.039]; Turkey[0.039]; Belgrade[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Efes Pilsen\u001b[39m ==> ENTITY: \u001b[32mAnadolu Efes S.K.\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCibona[0.046]; Turkey[0.044]; Turkey[0.044]; Partizan[0.044]; Leverkusen[0.042]; Zagreb[0.042]; Belgrade[0.042]; Panathinaikos[0.042]; Belgium[0.042]; Slovenia[0.041]; Group[0.041]; Group[0.041]; Group[0.041]; Panionios[0.041]; Bayer[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Italy[0.041]; Italy[0.041]; Croatia[0.041]; Croatia[0.041]; Dynamo[0.040]; Olympiakos[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = EuroLeague\u001b[39m ==> ENTITY: \u001b[32mEuroleague Basketball\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.215:0.215[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuroleague[0.049]; Basketball[0.046]; Maccabi[0.046]; matches[0.044]; basketball[0.044]; played[0.043]; Cibona[0.042]; Italy[0.042]; Italy[0.042]; points[0.042]; Panathinaikos[0.041]; won[0.041]; Ulker[0.041]; Spain[0.040]; Croatia[0.040]; Germany[0.040]; Olympiakos[0.040]; Panionios[0.040]; Spor[0.040]; championship[0.040]; men[0.039]; Greece[0.039]; Greece[0.039]; Milan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.201:0.201[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.049]; Spain[0.049]; Italy[0.044]; Italy[0.044]; Italy[0.044]; France[0.043]; France[0.043]; France[0.043]; Russia[0.043]; Madrid[0.042]; Germany[0.041]; Germany[0.041]; Slovenia[0.041]; Moscow[0.040]; Croatia[0.040]; Croatia[0.040]; Turkey[0.040]; Turkey[0.040]; Belgium[0.039]; Maccabi[0.039]; Barcelona[0.039]; Greece[0.039]; Greece[0.039]; Greece[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Estudiantes Madrid\u001b[39m ==> ENTITY: \u001b[32mCB Estudiantes\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.045]; Spain[0.045]; Spain[0.045]; EuroLeague[0.045]; Cibona[0.044]; Euroleague[0.044]; Leverkusen[0.043]; basketball[0.043]; Basketball[0.042]; Spor[0.041]; Efes[0.041]; Maccabi[0.041]; won[0.041]; played[0.040]; Sevilla[0.040]; championship[0.040]; Standings[0.040]; Standings[0.040]; Partizan[0.040]; points[0.040]; Barcelona[0.040]; Zagreb[0.040]; matches[0.040]; Bayer[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Teamsystem Bologna\u001b[39m ==> ENTITY: \u001b[32mFortitudo Pallacanestro Bologna\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.235:0.235[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuroleague[0.047]; Milan[0.045]; Italy[0.045]; Italy[0.045]; Maccabi[0.043]; EuroLeague[0.043]; Basketball[0.042]; Cibona[0.042]; Ulker[0.042]; Olympiakos[0.041]; Partizan[0.041]; Leverkusen[0.041]; Panathinaikos[0.041]; matches[0.040]; Spain[0.040]; Spain[0.040]; Panionios[0.040]; Thursday[0.040]; won[0.040]; Greece[0.040]; Greece[0.040]; Greece[0.040]; Stefanel[0.040]; Spor[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSevilla\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSeville\u001b[39m <---> \u001b[32mCB Sevilla\u001b[39m\t\nSCORES: global= 0.259:0.245[\u001b[31m0.014\u001b[39m]; local()= 0.133:0.153[\u001b[32m0.020\u001b[39m]; log p(e|m)= -0.837:-2.865[\u001b[31m2.028\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.047]; Spain[0.047]; Spain[0.047]; France[0.045]; France[0.045]; Madrid[0.045]; Italy[0.043]; Barcelona[0.043]; Belgium[0.042]; Dynamo[0.042]; Estudiantes[0.042]; Germany[0.041]; Germany[0.041]; Panathinaikos[0.040]; Villeurbanne[0.040]; Bologna[0.040]; Partizan[0.039]; Moscow[0.039]; Leverkusen[0.039]; Turkey[0.039]; Russia[0.039]; Alba[0.039]; Berlin[0.039]; Pau[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.140:0.140[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMaccabi[0.046]; matches[0.044]; EuroLeague[0.043]; Slovenia[0.043]; played[0.043]; Euroleague[0.043]; Aviv[0.043]; Croatia[0.042]; Croatia[0.042]; Italy[0.041]; Italy[0.041]; Basketball[0.041]; Germany[0.041]; Germany[0.041]; Turkey[0.041]; won[0.041]; Spain[0.041]; Spain[0.041]; France[0.041]; France[0.041]; Belgium[0.040]; Greece[0.040]; Greece[0.040]; Greece[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mVilleurbanne\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mVilleurbanne\u001b[39m <---> \u001b[32mASVEL Basket\u001b[39m\t\nSCORES: global= 0.280:0.280[\u001b[31m0.000\u001b[39m]; local()= 0.061:0.146[\u001b[32m0.086\u001b[39m]; log p(e|m)= 0.000:-0.662[\u001b[31m0.662\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCibona[0.046]; Leverkusen[0.045]; Spor[0.043]; Efes[0.043]; France[0.043]; France[0.043]; France[0.043]; Limoges[0.042]; Partizan[0.042]; Zagreb[0.041]; Maccabi[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Panathinaikos[0.041]; Bayer[0.041]; Ulker[0.041]; Yugoslavia[0.041]; Dynamo[0.041]; Tel[0.041]; Panionios[0.040]; Belgium[0.040]; Olympiakos[0.040]; Charleroi[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.169:0.169[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.050]; Spain[0.050]; Italy[0.045]; France[0.044]; France[0.044]; Russia[0.043]; Madrid[0.043]; Germany[0.042]; Germany[0.042]; Slovenia[0.042]; Moscow[0.041]; Croatia[0.041]; Croatia[0.041]; Turkey[0.040]; Belgium[0.040]; Barcelona[0.040]; Greece[0.040]; Yugoslavia[0.040]; Estudiantes[0.040]; Group[0.039]; Group[0.039]; Sevilla[0.038]; Dynamo[0.038]; Cibona[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Croatia ==> ENTITY: \u001b[32mCroatia\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.177:0.177[0]; log p(e|m)= -0.350:-0.350[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCroatia[0.049]; Slovenia[0.048]; EuroLeague[0.046]; Yugoslavia[0.045]; matches[0.044]; basketball[0.044]; Italy[0.044]; Italy[0.044]; Italy[0.044]; Zagreb[0.044]; championship[0.044]; Basketball[0.043]; Belgium[0.043]; won[0.042]; Group[0.042]; Group[0.042]; Group[0.042]; Group[0.042]; Spain[0.042]; Spain[0.042]; Spain[0.042]; Euroleague[0.042]; Belgrade[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pau-Orthez\u001b[39m ==> ENTITY: \u001b[32mÉlan Béarnais Pau-Orthez\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCibona[0.046]; Villeurbanne[0.045]; Leverkusen[0.044]; France[0.044]; France[0.044]; Partizan[0.043]; Belgrade[0.043]; Orthez[0.043]; Efes[0.043]; Italy[0.042]; Italy[0.042]; Belgium[0.041]; Zagreb[0.041]; Bayer[0.041]; Charleroi[0.041]; Yugoslavia[0.040]; Panathinaikos[0.040]; Turkey[0.040]; Panionios[0.040]; Dynamo[0.040]; Spain[0.040]; Spain[0.040]; Spain[0.040]; Madrid[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.049]; Belgium[0.044]; Italy[0.044]; Italy[0.044]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Germany[0.042]; Germany[0.042]; Slovenia[0.042]; Villeurbanne[0.042]; Croatia[0.041]; Croatia[0.041]; Charleroi[0.041]; Panionios[0.041]; Partizan[0.041]; Yugoslavia[0.040]; Turkey[0.040]; Pau[0.040]; Bologna[0.040]; Bologna[0.040]; Russia[0.039]; Cibona[0.039]; Olympiakos[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Panathinaikos ==> ENTITY: \u001b[32mPanathinaikos B.C.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.208:0.208[0]; log p(e|m)= -0.805:-0.805[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuroLeague[0.048]; Maccabi[0.045]; matches[0.043]; basketball[0.043]; Greece[0.042]; Greece[0.042]; Greece[0.042]; Ulker[0.042]; points[0.042]; Panionios[0.042]; won[0.042]; championship[0.042]; Cibona[0.042]; Olympiakos[0.042]; played[0.042]; Efes[0.041]; Belgrade[0.040]; Milan[0.040]; Spor[0.040]; Zagreb[0.040]; Italy[0.040]; Italy[0.040]; Italy[0.040]; Slovenia[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Greece ==> ENTITY: \u001b[32mGreece\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.252:0.252[0]; log p(e|m)= -0.276:-0.276[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGreece[0.046]; Greece[0.046]; Euroleague[0.045]; Basketball[0.045]; EuroLeague[0.044]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Turkey[0.041]; Turkey[0.041]; Maccabi[0.041]; Yugoslavia[0.041]; Croatia[0.041]; Croatia[0.041]; matches[0.041]; Slovenia[0.040]; won[0.040]; played[0.040]; Germany[0.040]; Germany[0.040]; Panathinaikos[0.040]; Spain[0.040]; Spain[0.040]; Olympiakos[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.047]; EuroLeague[0.045]; Euroleague[0.044]; Belgium[0.043]; matches[0.043]; Italy[0.043]; Italy[0.043]; Basketball[0.043]; basketball[0.042]; Spain[0.042]; Spain[0.042]; Germany[0.041]; Germany[0.041]; Slovenia[0.041]; championship[0.041]; Villeurbanne[0.041]; played[0.040]; Croatia[0.040]; Croatia[0.040]; Maccabi[0.040]; Charleroi[0.040]; Panionios[0.040]; won[0.039]; points[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSplit\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSplit, Croatia\u001b[39m <---> \u001b[32mKK Split\u001b[39m\t\nSCORES: global= 0.273:0.259[\u001b[31m0.014\u001b[39m]; local()= 0.149:0.175[\u001b[32m0.026\u001b[39m]; log p(e|m)= -0.654:-3.147[\u001b[31m2.493\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYugoslavia[0.045]; Zagreb[0.044]; Croatia[0.044]; Croatia[0.044]; Belgrade[0.044]; Partizan[0.043]; Maccabi[0.043]; Cibona[0.043]; Slovenia[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Berlin[0.041]; Panathinaikos[0.041]; Group[0.040]; Group[0.040]; Group[0.040]; Charleroi[0.040]; Ljubljana[0.040]; Bayer[0.040]; Panionios[0.040]; Ulker[0.040]; Olympiakos[0.040]; Efes[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Euroleague\u001b[39m ==> ENTITY: \u001b[32mEuroleague Basketball\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuroLeague[0.050]; Basketball[0.047]; Maccabi[0.046]; matches[0.044]; basketball[0.044]; played[0.043]; Cibona[0.043]; Italy[0.042]; Italy[0.042]; points[0.042]; won[0.041]; Ulker[0.041]; Spain[0.040]; Croatia[0.040]; Germany[0.040]; Olympiakos[0.040]; Panionios[0.040]; Spor[0.040]; championship[0.040]; men[0.040]; Greece[0.039]; Greece[0.039]; Milan[0.039]; Zagreb[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = CSKA Moscow ==> ENTITY: \u001b[32mPBC CSKA Moscow\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.182:0.182[0]; log p(e|m)= -0.853:-0.853[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuroLeague[0.046]; Basketball[0.046]; Euroleague[0.046]; matches[0.045]; basketball[0.045]; Russia[0.044]; played[0.044]; Maccabi[0.043]; won[0.043]; Estudiantes[0.041]; championship[0.041]; Panathinaikos[0.041]; Olympiakos[0.041]; points[0.040]; Milan[0.040]; Cibona[0.040]; Panionios[0.040]; lost[0.039]; Zagreb[0.039]; Standings[0.039]; Standings[0.039]; Spor[0.039]; Villeurbanne[0.039]; London[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.177:0.177[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.050]; Spain[0.050]; Italy[0.045]; Italy[0.045]; Italy[0.045]; matches[0.045]; France[0.044]; France[0.044]; France[0.044]; Russia[0.044]; Russia[0.044]; Madrid[0.043]; EuroLeague[0.043]; Germany[0.042]; Germany[0.042]; Slovenia[0.042]; Moscow[0.042]; Moscow[0.042]; championship[0.041]; Croatia[0.041]; Croatia[0.041]; points[0.041]; London[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mDynamo Moscow\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFC Dynamo Moscow\u001b[39m <---> \u001b[32mBC Dynamo Moscow\u001b[39m\t\nSCORES: global= 0.257:0.254[\u001b[31m0.003\u001b[39m]; local()= 0.122:0.099[\u001b[31m0.023\u001b[39m]; log p(e|m)= -0.453:-1.280[\u001b[31m0.828\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.046]; Partizan[0.045]; Belgrade[0.044]; Yugoslavia[0.043]; Estudiantes[0.043]; Panathinaikos[0.043]; Greece[0.043]; Croatia[0.042]; Croatia[0.042]; Efes[0.042]; Leverkusen[0.041]; Slovenia[0.041]; Madrid[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Germany[0.041]; Germany[0.041]; Belgium[0.040]; Turkey[0.040]; Sevilla[0.040]; Villeurbanne[0.040]; Group[0.040]; Group[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.191:0.191[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.050]; France[0.050]; Belgium[0.045]; Italy[0.045]; Italy[0.045]; Italy[0.045]; Spain[0.044]; Spain[0.044]; Spain[0.044]; Germany[0.043]; Germany[0.043]; Slovenia[0.043]; Villeurbanne[0.043]; Croatia[0.042]; Croatia[0.042]; Maccabi[0.042]; Charleroi[0.042]; Panionios[0.042]; Partizan[0.041]; Yugoslavia[0.041]; Turkey[0.041]; Turkey[0.041]; Pau[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.107:0.107[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCroatia[0.046]; Moscow[0.045]; Slovenia[0.045]; Dynamo[0.043]; Yugoslavia[0.043]; Belgrade[0.043]; Germany[0.043]; Germany[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Turkey[0.042]; Belgium[0.042]; Italy[0.042]; France[0.041]; France[0.041]; Greece[0.041]; Panathinaikos[0.039]; Partizan[0.039]; Efes[0.039]; Ljubljana[0.039]; Group[0.039]; Group[0.039]; Berlin[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alba Berlin\u001b[39m ==> ENTITY: \u001b[32mAlba Berlin\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.046]; EuroLeague[0.045]; Euroleague[0.044]; basketball[0.043]; Cibona[0.043]; won[0.043]; championship[0.043]; matches[0.043]; Leverkusen[0.042]; Standings[0.042]; Standings[0.042]; Maccabi[0.041]; played[0.041]; Slovenia[0.041]; points[0.041]; Spor[0.040]; Efes[0.040]; Belgrade[0.040]; Dynamo[0.040]; Panathinaikos[0.040]; Yugoslavia[0.040]; Partizan[0.040]; Croatia[0.039]; Croatia[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Greece ==> ENTITY: \u001b[32mGreece\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.241:0.241[0]; log p(e|m)= -0.276:-0.276[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGreece[0.047]; Greece[0.047]; Euroleague[0.045]; Basketball[0.045]; EuroLeague[0.044]; Italy[0.042]; Italy[0.042]; Turkey[0.042]; Turkey[0.042]; Maccabi[0.042]; Croatia[0.041]; Croatia[0.041]; matches[0.041]; Slovenia[0.040]; won[0.040]; played[0.040]; Germany[0.040]; Germany[0.040]; Panathinaikos[0.040]; Spain[0.040]; Spain[0.040]; Olympiakos[0.040]; basketball[0.040]; Panionios[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Partizan Belgrade ==> ENTITY: \u001b[32mKK Partizan\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.909:-0.909[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCibona[0.045]; Zagreb[0.045]; Croatia[0.044]; Croatia[0.044]; Yugoslavia[0.043]; Dynamo[0.043]; Italy[0.043]; Italy[0.043]; Slovenia[0.042]; Panathinaikos[0.042]; Efes[0.041]; Greece[0.041]; Greece[0.041]; Turkey[0.041]; Villeurbanne[0.041]; Russia[0.040]; Group[0.040]; Group[0.040]; Group[0.040]; Olympiakos[0.040]; Moscow[0.040]; Ljubljana[0.040]; Germany[0.040]; Germany[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Yugoslavia ==> ENTITY: \u001b[32mYugoslavia\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.654:-0.654[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCroatia[0.045]; Croatia[0.045]; Slovenia[0.044]; Italy[0.044]; Italy[0.044]; Turkey[0.044]; Belgrade[0.044]; Zagreb[0.043]; Germany[0.042]; Germany[0.042]; Spain[0.042]; Spain[0.042]; Spain[0.042]; Greece[0.041]; Greece[0.041]; Russia[0.041]; Belgium[0.041]; France[0.040]; France[0.040]; Ljubljana[0.040]; Berlin[0.039]; Cibona[0.039]; Madrid[0.039]; Moscow[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Greece ==> ENTITY: \u001b[32mGreece\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.249:0.249[0]; log p(e|m)= -0.276:-0.276[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGreece[0.047]; Greece[0.047]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Turkey[0.042]; Turkey[0.042]; Maccabi[0.042]; Yugoslavia[0.042]; Croatia[0.042]; Croatia[0.042]; matches[0.042]; Slovenia[0.041]; won[0.041]; played[0.041]; Germany[0.041]; Germany[0.041]; Panathinaikos[0.041]; Spain[0.040]; Spain[0.040]; Spain[0.040]; Olympiakos[0.040]; basketball[0.040]; Panionios[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.096:0.096[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuroLeague[0.047]; matches[0.044]; Croatia[0.044]; Moscow[0.043]; won[0.043]; Euroleague[0.043]; Slovenia[0.043]; championship[0.042]; Basketball[0.042]; played[0.042]; points[0.042]; Zagreb[0.042]; Cibona[0.041]; Germany[0.041]; Maccabi[0.041]; Spain[0.041]; basketball[0.041]; Turkey[0.040]; Belgium[0.040]; Italy[0.040]; Italy[0.040]; Standings[0.040]; Standings[0.040]; lost[0.039]; \t\n\r [====================>.........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s105ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2093/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1339testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Barrick\u001b[39m ==> ENTITY: \u001b[32mBarrick Gold\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBarrick[0.048]; Barrick[0.048]; Minerals[0.045]; gold[0.044]; Gold[0.044]; Canada[0.044]; Corp[0.043]; Arasu[0.042]; continue[0.042]; continue[0.042]; held[0.042]; returned[0.041]; Friday[0.041]; Busang[0.040]; Busang[0.040]; said[0.040]; said[0.040]; said[0.040]; Toronto[0.040]; Toronto[0.040]; progress[0.039]; declined[0.039]; close[0.039]; develop[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Barrick\u001b[39m ==> ENTITY: \u001b[32mBarrick Gold\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmining[0.050]; Barrick[0.048]; company[0.044]; potential[0.044]; development[0.042]; development[0.042]; Duta[0.042]; President[0.042]; million[0.041]; companies[0.041]; companies[0.041]; failed[0.041]; Group[0.041]; Group[0.041]; Busang[0.041]; Busang[0.041]; months[0.040]; said[0.040]; said[0.040]; Pt[0.040]; stake[0.040]; government[0.040]; explore[0.040]; earlier[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bre-X\u001b[39m ==> ENTITY: \u001b[32mBre-X\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngold[0.050]; Busang[0.048]; Busang[0.048]; Barrick[0.048]; Barrick[0.048]; Barrick[0.048]; Barrick[0.048]; Mines[0.048]; shareholders[0.046]; companies[0.045]; property[0.045]; property[0.045]; property[0.045]; deal[0.044]; deal[0.044]; advised[0.044]; procedures[0.043]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bre-X\u001b[39m ==> ENTITY: \u001b[32mBre-X\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngold[0.050]; gold[0.050]; Busang[0.048]; Busang[0.048]; Barrick[0.048]; Barrick[0.048]; Mines[0.048]; Mines[0.048]; property[0.045]; property[0.045]; Said[0.045]; ounces[0.044]; ounces[0.044]; Kalimantan[0.044]; returned[0.044]; million[0.043]; million[0.043]; estimated[0.043]; estimated[0.043]; report[0.043]; Toronto[0.043]; Toronto[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Barrick\u001b[39m ==> ENTITY: \u001b[32mBarrick Gold\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBarrick[0.046]; Barrick[0.046]; Barrick[0.046]; Mines[0.044]; Mines[0.044]; Minerals[0.043]; gold[0.043]; gold[0.043]; Gold[0.043]; Canada[0.042]; region[0.041]; Corp[0.041]; Arasu[0.041]; General[0.040]; continue[0.040]; continue[0.040]; held[0.040]; property[0.040]; property[0.040]; hold[0.040]; returned[0.039]; north[0.039]; million[0.039]; ounces[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Suharto\u001b[39m ==> ENTITY: \u001b[32mSuharto\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.045]; government[0.044]; government[0.044]; government[0.044]; government[0.044]; Suharto[0.042]; son[0.042]; richest[0.042]; Sigit[0.041]; Duta[0.041]; months[0.041]; daughter[0.041]; property[0.041]; property[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; Thursday[0.041]; million[0.040]; eldest[0.040]; eldest[0.040]; earlier[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jakarta\u001b[39m ==> ENTITY: \u001b[32mJakarta\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJakarta[0.050]; Indonesia[0.048]; Indonesian[0.044]; Thursday[0.043]; East[0.042]; source[0.042]; source[0.042]; source[0.042]; told[0.042]; Friday[0.042]; submit[0.042]; held[0.041]; central[0.041]; December[0.041]; agreement[0.041]; coordinating[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; hold[0.040]; General[0.040]; partnership[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bre-X\u001b[39m ==> ENTITY: \u001b[32mBre-X\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngold[0.053]; gold[0.053]; Busang[0.051]; Barrick[0.051]; Barrick[0.051]; Mines[0.050]; shareholders[0.049]; companies[0.047]; property[0.047]; ounces[0.047]; deal[0.046]; deal[0.046]; advised[0.046]; million[0.046]; estimated[0.046]; report[0.045]; procedures[0.045]; said[0.045]; said[0.045]; said[0.045]; said[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bre-X\u001b[39m ==> ENTITY: \u001b[32mBre-X\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngold[0.052]; Busang[0.050]; Busang[0.050]; Barrick[0.050]; Barrick[0.050]; Barrick[0.050]; Mines[0.050]; Duta[0.048]; companies[0.047]; companies[0.047]; companies[0.047]; property[0.047]; property[0.047]; property[0.047]; Pt[0.046]; deal[0.046]; deal[0.046]; deal[0.046]; advised[0.045]; Sigit[0.045]; procedures[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bre-X\u001b[39m ==> ENTITY: \u001b[32mBre-X\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngold[0.052]; gold[0.052]; Busang[0.050]; Busang[0.050]; Busang[0.050]; Barrick[0.050]; Barrick[0.050]; Mines[0.049]; Mines[0.049]; shareholders[0.048]; property[0.046]; property[0.046]; Said[0.046]; ounces[0.046]; ounces[0.046]; Kalimantan[0.046]; million[0.045]; million[0.045]; estimated[0.045]; estimated[0.045]; report[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bre-X\u001b[39m ==> ENTITY: \u001b[32mBre-X\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngold[0.045]; gold[0.045]; gold[0.045]; Busang[0.043]; Busang[0.043]; Busang[0.043]; Busang[0.043]; Barrick[0.043]; Barrick[0.043]; Barrick[0.043]; Mines[0.043]; Mines[0.043]; Arasu[0.042]; property[0.040]; property[0.040]; Said[0.040]; Canada[0.040]; ounces[0.040]; ounces[0.040]; Gold[0.040]; Kalimantan[0.040]; Corp[0.040]; returned[0.039]; Minerals[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.137:0.137[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntold[0.044]; region[0.043]; Canada[0.043]; Corp[0.042]; source[0.042]; source[0.042]; source[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; Thursday[0.042]; Toronto[0.041]; Toronto[0.041]; estimated[0.041]; key[0.041]; negotiators[0.041]; say[0.041]; central[0.041]; sources[0.041]; sources[0.041]; Friday[0.040]; partnership[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Barrick\u001b[39m ==> ENTITY: \u001b[32mBarrick Gold\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBarrick[0.048]; Barrick[0.048]; Barrick[0.048]; Mines[0.046]; gold[0.044]; region[0.043]; property[0.041]; property[0.041]; property[0.041]; shareholders[0.041]; companies[0.041]; given[0.040]; Busang[0.040]; Busang[0.040]; Energy[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; Thursday[0.040]; December[0.040]; deadline[0.040]; deadline[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.064:0.064[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nToronto[0.046]; Toronto[0.046]; gold[0.045]; Thursday[0.044]; Friday[0.043]; held[0.043]; partnership[0.042]; returned[0.042]; Gold[0.042]; key[0.042]; hold[0.042]; agreement[0.041]; Indonesia[0.041]; talks[0.041]; talks[0.041]; talks[0.041]; declined[0.041]; close[0.040]; progress[0.040]; Corp[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Barrick\u001b[39m ==> ENTITY: \u001b[32mBarrick Gold\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmining[0.050]; Barrick[0.048]; gold[0.044]; company[0.044]; potential[0.043]; development[0.042]; development[0.042]; Duta[0.041]; property[0.041]; property[0.041]; President[0.041]; million[0.041]; companies[0.041]; companies[0.041]; failed[0.040]; Group[0.040]; Group[0.040]; Busang[0.040]; Busang[0.040]; Busang[0.040]; months[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Barrick\u001b[39m ==> ENTITY: \u001b[32mBarrick Gold\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBarrick[0.048]; Mines[0.046]; gold[0.045]; region[0.043]; region[0.043]; property[0.042]; shareholders[0.042]; north[0.041]; million[0.041]; companies[0.041]; ounces[0.041]; given[0.041]; Busang[0.041]; feasibility[0.041]; report[0.041]; Energy[0.041]; south[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; December[0.040]; December[0.040]; December[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = East Kalimantan\u001b[39m ==> ENTITY: \u001b[32mEast Kalimantan\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nregion[0.046]; region[0.046]; south[0.044]; Indonesia[0.044]; north[0.043]; Jakarta[0.043]; Mines[0.042]; Mines[0.042]; hold[0.042]; million[0.041]; million[0.041]; Indonesian[0.041]; parts[0.041]; Busang[0.041]; Busang[0.041]; Busang[0.041]; property[0.041]; property[0.041]; estimated[0.040]; estimated[0.040]; central[0.040]; central[0.040]; Said[0.040]; Umar[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bre-X\u001b[39m ==> ENTITY: \u001b[32mBre-X\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngold[0.045]; gold[0.045]; Busang[0.043]; Busang[0.043]; Busang[0.043]; Busang[0.043]; Barrick[0.043]; Barrick[0.043]; Barrick[0.043]; Barrick[0.043]; Mines[0.043]; Mines[0.043]; Arasu[0.042]; property[0.040]; property[0.040]; Said[0.040]; Canada[0.040]; ounces[0.040]; Gold[0.040]; Kalimantan[0.040]; Corp[0.040]; returned[0.039]; Minerals[0.039]; million[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Barrick\u001b[39m ==> ENTITY: \u001b[32mBarrick Gold\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBarrick[0.057]; Barrick[0.057]; Mines[0.055]; gold[0.053]; development[0.050]; Duta[0.049]; property[0.049]; property[0.049]; property[0.049]; President[0.049]; million[0.049]; companies[0.048]; companies[0.048]; companies[0.048]; given[0.048]; failed[0.048]; Group[0.048]; Busang[0.048]; Busang[0.048]; Energy[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bre-X\u001b[39m ==> ENTITY: \u001b[32mBre-X\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngold[0.048]; gold[0.048]; Busang[0.047]; Busang[0.047]; Barrick[0.047]; Mines[0.046]; shareholders[0.045]; property[0.043]; property[0.043]; ounces[0.043]; ounces[0.043]; Kalimantan[0.043]; million[0.042]; million[0.042]; estimated[0.042]; estimated[0.042]; report[0.042]; procedures[0.042]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bre-X\u001b[39m ==> ENTITY: \u001b[32mBre-X\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngold[0.046]; Busang[0.045]; Barrick[0.044]; Barrick[0.044]; Barrick[0.044]; Barrick[0.044]; Mines[0.044]; shareholders[0.043]; companies[0.041]; companies[0.041]; companies[0.041]; property[0.041]; property[0.041]; property[0.041]; deal[0.040]; deal[0.040]; advised[0.040]; procedures[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bre-X\u001b[39m ==> ENTITY: \u001b[32mBre-X\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngold[0.052]; gold[0.052]; Busang[0.050]; Busang[0.050]; Busang[0.050]; Barrick[0.050]; Barrick[0.050]; Mines[0.049]; Mines[0.049]; shareholders[0.048]; property[0.046]; property[0.046]; Said[0.046]; ounces[0.046]; ounces[0.046]; Kalimantan[0.046]; million[0.045]; million[0.045]; estimated[0.045]; estimated[0.045]; report[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Toronto ==> ENTITY: \u001b[32mToronto\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.123:-0.123[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nToronto[0.056]; Canada[0.055]; region[0.049]; north[0.049]; south[0.049]; Thursday[0.049]; December[0.048]; Friday[0.047]; Mines[0.047]; Mines[0.047]; million[0.046]; close[0.046]; East[0.046]; key[0.046]; central[0.046]; held[0.046]; hold[0.045]; Barrick[0.045]; Barrick[0.045]; Barrick[0.045]; Barrick[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Suharto\u001b[39m ==> ENTITY: \u001b[32mSuharto\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.106:0.106[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.045]; government[0.045]; government[0.045]; Suharto[0.043]; son[0.042]; richest[0.042]; Sigit[0.042]; Duta[0.042]; months[0.042]; daughter[0.042]; property[0.041]; said[0.041]; said[0.041]; said[0.041]; million[0.041]; eldest[0.041]; eldest[0.041]; earlier[0.041]; Pt[0.040]; Siti[0.040]; stake[0.040]; deal[0.040]; receive[0.040]; receive[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bre-X\u001b[39m ==> ENTITY: \u001b[32mBre-X\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngold[0.047]; Busang[0.045]; Busang[0.045]; Barrick[0.045]; Barrick[0.045]; Barrick[0.045]; Arasu[0.044]; Canada[0.042]; Gold[0.041]; Corp[0.041]; returned[0.041]; Minerals[0.041]; Toronto[0.040]; Toronto[0.040]; said[0.040]; said[0.040]; said[0.040]; Bre[0.040]; Bre[0.040]; Reuters[0.040]; Indonesia[0.039]; hammer[0.039]; Jakarta[0.039]; Jakarta[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.100:0.100[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJakarta[0.054]; Jakarta[0.054]; Kalimantan[0.052]; speaking[0.052]; Canada[0.050]; continue[0.050]; continue[0.050]; talks[0.050]; talks[0.050]; talks[0.050]; negotiations[0.049]; negotiations[0.049]; negotiations[0.049]; negotiations[0.049]; held[0.049]; East[0.049]; progress[0.049]; agreement[0.049]; hold[0.049]; key[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bre-X\u001b[39m ==> ENTITY: \u001b[32mBre-X\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngold[0.053]; Busang[0.051]; Barrick[0.051]; Barrick[0.051]; Barrick[0.051]; Mines[0.050]; shareholders[0.049]; companies[0.047]; property[0.047]; ounces[0.047]; deal[0.046]; deal[0.046]; advised[0.046]; million[0.046]; estimated[0.046]; report[0.046]; procedures[0.045]; said[0.045]; said[0.045]; said[0.045]; said[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bre-X\u001b[39m ==> ENTITY: \u001b[32mBre-X\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngold[0.051]; mining[0.050]; Busang[0.050]; Busang[0.050]; Busang[0.050]; Barrick[0.050]; Barrick[0.050]; Barrick[0.050]; Duta[0.048]; months[0.047]; company[0.047]; potential[0.046]; companies[0.046]; companies[0.046]; property[0.046]; property[0.046]; property[0.046]; Pt[0.046]; deal[0.045]; million[0.045]; Sigit[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesian ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.742:-0.742[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesian[0.048]; region[0.044]; region[0.044]; region[0.044]; agreement[0.043]; million[0.042]; million[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; formally[0.041]; seek[0.041]; say[0.041]; report[0.041]; estimated[0.041]; estimated[0.041]; south[0.041]; government[0.040]; government[0.040]; expected[0.040]; set[0.040]; added[0.040]; declined[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Barrick\u001b[39m ==> ENTITY: \u001b[32mBarrick Gold\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBarrick[0.046]; Barrick[0.046]; Mines[0.044]; Mines[0.044]; Minerals[0.043]; gold[0.043]; gold[0.043]; gold[0.043]; Gold[0.043]; Canada[0.042]; region[0.041]; region[0.041]; Corp[0.041]; Arasu[0.041]; General[0.040]; continue[0.040]; held[0.040]; property[0.040]; property[0.040]; hold[0.040]; returned[0.040]; north[0.039]; million[0.039]; million[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Barrick\u001b[39m ==> ENTITY: \u001b[32mBarrick Gold\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBarrick[0.048]; Barrick[0.048]; Mines[0.046]; gold[0.044]; development[0.042]; property[0.042]; property[0.042]; property[0.042]; shareholders[0.042]; companies[0.041]; companies[0.041]; companies[0.041]; given[0.041]; failed[0.041]; Busang[0.040]; Energy[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; Thursday[0.040]; December[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Toronto ==> ENTITY: \u001b[32mToronto\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.123:-0.123[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nToronto[0.050]; Canada[0.049]; Thursday[0.043]; December[0.043]; Friday[0.042]; Mines[0.042]; Mines[0.042]; close[0.041]; East[0.041]; key[0.041]; held[0.041]; hold[0.040]; Barrick[0.040]; Barrick[0.040]; Barrick[0.040]; Barrick[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; agreement[0.040]; gold[0.040]; returned[0.040]; \t\n\r [====================>.........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s941ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2125/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1241testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nairobi\u001b[39m ==> ENTITY: \u001b[32mNairobi\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNairobi[0.051]; land[0.047]; west[0.042]; living[0.042]; countryside[0.042]; miles[0.042]; east[0.042]; valley[0.042]; charge[0.042]; main[0.041]; large[0.041]; city[0.041]; Friday[0.041]; going[0.041]; going[0.041]; Rwandan[0.041]; conference[0.041]; General[0.040]; decided[0.040]; general[0.040]; eastern[0.040]; eastern[0.040]; group[0.040]; plight[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mKinshasa\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mDemocratic Republic of the Congo\u001b[39m <---> \u001b[32mKinshasa\u001b[39m\t\nSCORES: global= 0.288:0.288[\u001b[31m0.000\u001b[39m]; local()= 0.152:0.140[\u001b[31m0.013\u001b[39m]; log p(e|m)= -1.382:0.000[\u001b[32m1.382\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \neastern[0.046]; Zaire[0.046]; Zairean[0.046]; Zairean[0.046]; north[0.044]; refugees[0.043]; Kindu[0.043]; west[0.042]; Bukavu[0.042]; government[0.042]; preserve[0.041]; town[0.041]; Masisi[0.041]; groups[0.041]; rebels[0.040]; lives[0.040]; denied[0.040]; said[0.040]; said[0.040]; general[0.039]; group[0.039]; Baril[0.039]; peace[0.039]; thing[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Doug Young ==> ENTITY: \u001b[32mDoug Young (politician)\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.066:0.066[0]; log p(e|m)= -1.130:-1.130[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYoung[0.045]; Canadian[0.044]; leader[0.043]; Thursday[0.043]; home[0.043]; time[0.042]; probably[0.042]; fall[0.042]; Defence[0.042]; second[0.041]; Minister[0.041]; supporters[0.041]; group[0.041]; hands[0.041]; said[0.041]; said[0.041]; said[0.041]; militarily[0.041]; rebel[0.041]; apparently[0.041]; airdrops[0.041]; airdrops[0.041]; airdrops[0.041]; forward[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zairean\u001b[39m ==> ENTITY: \u001b[32mDemocratic Republic of the Congo\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaire[0.049]; eastern[0.047]; Zairean[0.046]; north[0.044]; Kinshasa[0.044]; refugees[0.043]; Kindu[0.043]; west[0.042]; government[0.042]; security[0.041]; Masisi[0.041]; groups[0.041]; rebels[0.041]; town[0.040]; tens[0.040]; Kisangani[0.040]; said[0.040]; said[0.040]; general[0.040]; reconnaissance[0.039]; group[0.039]; peace[0.039]; thousands[0.039]; ground[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zaire ==> ENTITY: \u001b[32mDemocratic Republic of the Congo\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.190:0.190[0]; log p(e|m)= -2.002:-2.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaire[0.046]; Rwanda[0.044]; eastern[0.044]; eastern[0.044]; eastern[0.044]; Zairean[0.043]; Zairean[0.043]; army[0.042]; Rwandan[0.042]; Kinshasa[0.041]; Kinshasa[0.041]; force[0.041]; land[0.041]; deploy[0.041]; refugees[0.040]; refugees[0.040]; refugees[0.040]; rebels[0.040]; rebels[0.040]; rebels[0.040]; rebels[0.040]; government[0.040]; west[0.040]; rebel[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Goma\u001b[39m ==> ENTITY: \u001b[32mGoma\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGoma[0.068]; Kinshasa[0.059]; Nairobi[0.056]; Nairobi[0.056]; Kabila[0.052]; Rwanda[0.050]; Rwandan[0.049]; rebels[0.048]; rebels[0.048]; Zairean[0.048]; refugees[0.048]; refugees[0.048]; refugees[0.048]; rebel[0.047]; east[0.047]; eastern[0.046]; eastern[0.046]; eastern[0.046]; eastern[0.046]; west[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.105:0.105[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecurity[0.046]; Council[0.045]; government[0.045]; government[0.045]; Security[0.045]; mandate[0.044]; amounted[0.044]; added[0.044]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; position[0.043]; position[0.043]; peace[0.043]; absolutely[0.043]; denied[0.043]; Seven[0.042]; Seven[0.042]; Baril[0.042]; general[0.042]; coordinate[0.042]; powers[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rwandan\u001b[39m ==> ENTITY: \u001b[32mRwanda\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRwanda[0.048]; Zaire[0.044]; Zaire[0.044]; Zaire[0.044]; Zaire[0.044]; refugees[0.043]; refugees[0.043]; refugees[0.043]; Zairean[0.042]; Kinshasa[0.042]; Nairobi[0.041]; Nairobi[0.041]; force[0.041]; government[0.041]; eastern[0.041]; eastern[0.041]; eastern[0.041]; miles[0.040]; main[0.040]; commander[0.040]; countryside[0.040]; rebels[0.040]; group[0.039]; city[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zairean\u001b[39m ==> ENTITY: \u001b[32mDemocratic Republic of the Congo\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaire[0.047]; Zaire[0.047]; Zaire[0.047]; Rwanda[0.045]; Rwanda[0.045]; eastern[0.044]; eastern[0.044]; Zairean[0.044]; Rwandan[0.042]; Kinshasa[0.042]; Kinshasa[0.042]; border[0.041]; refugees[0.041]; government[0.040]; rebel[0.040]; area[0.039]; area[0.039]; major[0.039]; multinational[0.039]; multinational[0.039]; rebels[0.039]; rebels[0.039]; rebels[0.039]; rebels[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Medecins sans Frontieres\u001b[39m ==> ENTITY: \u001b[32mMédecins Sans Frontières\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.104:0.104[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncharity[0.048]; refugees[0.045]; refugees[0.045]; refugees[0.045]; group[0.044]; medical[0.044]; Masisi[0.043]; groups[0.042]; said[0.042]; said[0.042]; said[0.042]; British[0.041]; Goma[0.041]; Kindu[0.041]; ground[0.041]; coordinate[0.040]; area[0.040]; Walikale[0.040]; Bukavu[0.039]; Bukavu[0.039]; plans[0.039]; Kisangani[0.039]; risk[0.039]; office[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Walikale\u001b[39m ==> ENTITY: \u001b[32mWalikale\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.115:0.115[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKindu[0.044]; Masisi[0.044]; area[0.044]; groups[0.043]; Bukavu[0.043]; Bukavu[0.043]; Goma[0.043]; town[0.042]; town[0.042]; north[0.042]; Kisangani[0.042]; west[0.041]; northwest[0.041]; medical[0.041]; road[0.041]; said[0.041]; said[0.041]; refugees[0.041]; refugees[0.041]; refugees[0.041]; general[0.040]; group[0.040]; office[0.040]; Frontieres[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baril\u001b[39m ==> ENTITY: \u001b[32mMaurice Baril\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.069:0.069[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDefence[0.045]; General[0.045]; Canadian[0.045]; Rwanda[0.043]; Thursday[0.042]; militarily[0.042]; airdrops[0.042]; airdrops[0.042]; airdrops[0.042]; airdrops[0.042]; rebels[0.042]; force[0.042]; army[0.041]; multinational[0.041]; air[0.041]; air[0.041]; fall[0.041]; refugees[0.041]; Minister[0.041]; reconnaissance[0.041]; Zairean[0.040]; said[0.040]; said[0.040]; humanitarian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Laurent Kabila\u001b[39m ==> ENTITY: \u001b[32mLaurent-Désiré Kabila\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZairean[0.048]; Zairean[0.048]; government[0.046]; Rwanda[0.046]; Zaire[0.044]; Zaire[0.044]; rebels[0.044]; rebels[0.044]; rebels[0.044]; rebels[0.044]; rebel[0.044]; Rwandan[0.044]; Kinshasa[0.044]; Kinshasa[0.044]; leader[0.043]; army[0.043]; supporters[0.042]; eastern[0.041]; eastern[0.041]; eastern[0.041]; force[0.041]; force[0.041]; militarily[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zairean\u001b[39m ==> ENTITY: \u001b[32mDemocratic Republic of the Congo\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaire[0.048]; Zaire[0.048]; Zaire[0.048]; Rwanda[0.046]; eastern[0.046]; eastern[0.046]; eastern[0.046]; Zairean[0.045]; Rwandan[0.044]; Rwandan[0.044]; Kinshasa[0.043]; Kinshasa[0.043]; refugees[0.042]; refugees[0.042]; refugees[0.042]; refugees[0.042]; west[0.041]; government[0.041]; rebel[0.041]; area[0.041]; living[0.040]; Nairobi[0.040]; multinational[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baril\u001b[39m ==> ENTITY: \u001b[32mMaurice Baril\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.027:0.027[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmedical[0.044]; ground[0.044]; Masisi[0.043]; security[0.043]; Security[0.043]; Frontieres[0.042]; Council[0.042]; rebels[0.042]; Medecins[0.042]; Friday[0.042]; Goma[0.041]; Kindu[0.041]; refugees[0.041]; refugees[0.041]; refugees[0.041]; reconnaissance[0.041]; Zairean[0.041]; Zairean[0.041]; office[0.041]; amounted[0.041]; British[0.041]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Maurice Baril\u001b[39m ==> ENTITY: \u001b[32mMaurice Baril\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.079:0.079[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGeneral[0.047]; Canadian[0.047]; commander[0.045]; Rwanda[0.045]; aircrew[0.045]; Multinational[0.044]; airdrops[0.044]; rebels[0.044]; force[0.044]; charge[0.043]; Lieutenant[0.043]; multinational[0.043]; conference[0.043]; Friday[0.043]; Goma[0.043]; refugees[0.042]; refugees[0.042]; refugees[0.042]; Zairean[0.042]; Rwandan[0.042]; said[0.042]; said[0.042]; government[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zaire ==> ENTITY: \u001b[32mDemocratic Republic of the Congo\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.140:0.140[0]; log p(e|m)= -2.002:-2.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRwanda[0.046]; Zairean[0.046]; army[0.044]; Kinshasa[0.043]; force[0.043]; force[0.043]; border[0.043]; deploy[0.043]; refugees[0.042]; rebels[0.042]; make[0.041]; area[0.041]; area[0.041]; major[0.041]; group[0.041]; humanitarian[0.041]; multinational[0.040]; multinational[0.040]; hands[0.040]; probably[0.040]; probably[0.040]; ground[0.039]; Canadian[0.039]; Thursday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.057:0.057[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.046]; groups[0.046]; amounted[0.045]; added[0.045]; position[0.045]; position[0.045]; Friday[0.044]; tracked[0.044]; ground[0.044]; mention[0.043]; group[0.043]; office[0.043]; strip[0.043]; plans[0.043]; area[0.043]; northwest[0.043]; know[0.043]; know[0.043]; know[0.043]; know[0.043]; smaller[0.042]; general[0.042]; general[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canadian ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.073:0.073[0]; log p(e|m)= -0.412:-0.412[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncity[0.045]; Maurice[0.043]; force[0.043]; eastern[0.043]; eastern[0.043]; miles[0.043]; large[0.042]; land[0.042]; west[0.042]; multinational[0.042]; east[0.042]; living[0.042]; group[0.041]; countryside[0.041]; refugees[0.041]; refugees[0.041]; refugees[0.041]; Zaire[0.041]; Zaire[0.041]; Zaire[0.041]; news[0.041]; charge[0.040]; going[0.040]; going[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Security Council\u001b[39m ==> ENTITY: \u001b[32mUnited Nations Security Council\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecurity[0.046]; government[0.044]; government[0.044]; general[0.043]; coordinate[0.043]; mandate[0.042]; contacts[0.042]; negotiations[0.042]; powers[0.041]; added[0.041]; conscience[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; necessary[0.041]; necessary[0.041]; Asked[0.041]; wide[0.041]; amounted[0.041]; negotiate[0.041]; disregard[0.041]; position[0.041]; position[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mKinshasa\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mDemocratic Republic of the Congo\u001b[39m <---> \u001b[32mKinshasa\u001b[39m\t\nSCORES: global= 0.288:0.288[\u001b[31m0.000\u001b[39m]; local()= 0.196:0.183[\u001b[31m0.013\u001b[39m]; log p(e|m)= -1.382:0.000[\u001b[32m1.382\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRwanda[0.046]; Rwanda[0.046]; eastern[0.046]; eastern[0.046]; Zaire[0.046]; Zaire[0.046]; Zaire[0.046]; Zairean[0.045]; Zairean[0.045]; Rwandan[0.044]; border[0.043]; refugees[0.043]; General[0.042]; General[0.042]; area[0.042]; area[0.042]; Laurent[0.042]; major[0.042]; rebel[0.041]; Kabila[0.041]; Goma[0.040]; multinational[0.040]; multinational[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rwanda ==> ENTITY: \u001b[32mRwanda\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.429:-0.429[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRwandan[0.046]; Rwandan[0.046]; Zaire[0.046]; Zaire[0.046]; refugees[0.044]; refugees[0.044]; refugees[0.044]; refugees[0.044]; army[0.044]; Zairean[0.044]; Zairean[0.044]; government[0.043]; force[0.043]; Kinshasa[0.043]; Kinshasa[0.043]; Nairobi[0.042]; eastern[0.042]; eastern[0.042]; eastern[0.042]; ground[0.042]; ground[0.042]; rebel[0.042]; Kabila[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nairobi\u001b[39m ==> ENTITY: \u001b[32mNairobi\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNairobi[0.050]; land[0.046]; Rwanda[0.043]; west[0.042]; living[0.042]; countryside[0.042]; miles[0.042]; east[0.042]; valley[0.042]; charge[0.042]; Kinshasa[0.041]; main[0.041]; large[0.041]; city[0.041]; Friday[0.041]; going[0.041]; going[0.041]; going[0.041]; try[0.041]; Rwandan[0.041]; conference[0.040]; second[0.040]; General[0.040]; decided[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zairean\u001b[39m ==> ENTITY: \u001b[32mDemocratic Republic of the Congo\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.129:0.129[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaire[0.050]; eastern[0.048]; Zairean[0.047]; refugees[0.044]; government[0.043]; government[0.043]; humanitarian[0.042]; security[0.042]; tens[0.041]; said[0.041]; said[0.041]; fly[0.041]; reconnaissance[0.040]; group[0.040]; airdrops[0.040]; peace[0.040]; agencies[0.040]; thousands[0.040]; probably[0.040]; air[0.039]; air[0.039]; lives[0.039]; action[0.039]; access[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Goma\u001b[39m ==> ENTITY: \u001b[32mGoma\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGoma[0.057]; Kinshasa[0.050]; Kinshasa[0.050]; Kabila[0.044]; Rwanda[0.042]; Rwandan[0.042]; area[0.041]; rebels[0.041]; rebels[0.041]; rebels[0.041]; rebels[0.041]; Zairean[0.040]; Zairean[0.040]; refugees[0.040]; refugees[0.040]; rebel[0.040]; eastern[0.039]; eastern[0.039]; eastern[0.039]; west[0.039]; valley[0.039]; km[0.039]; Zaire[0.038]; Zaire[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = British ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.062:0.062[0]; log p(e|m)= -0.489:-0.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecurity[0.045]; refugees[0.044]; refugees[0.044]; refugees[0.044]; government[0.044]; medical[0.042]; group[0.042]; reconnaissance[0.042]; planes[0.041]; groups[0.041]; general[0.041]; general[0.041]; contacts[0.041]; north[0.041]; area[0.041]; added[0.041]; peace[0.041]; office[0.041]; sensitive[0.040]; plans[0.040]; ground[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bukavu\u001b[39m ==> ENTITY: \u001b[32mBukavu\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrefugees[0.045]; refugees[0.045]; refugees[0.045]; north[0.044]; Bukavu[0.044]; Goma[0.043]; Kindu[0.043]; west[0.043]; Kinshasa[0.043]; town[0.042]; town[0.042]; Masisi[0.042]; Medecins[0.041]; Zairean[0.041]; area[0.041]; northwest[0.041]; rebels[0.040]; medical[0.040]; Walikale[0.040]; Kisangani[0.040]; groups[0.039]; road[0.039]; Frontieres[0.039]; coordinate[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rwanda ==> ENTITY: \u001b[32mRwanda\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.429:-0.429[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaire[0.046]; refugees[0.044]; Kinshasa[0.044]; army[0.044]; Zairean[0.044]; force[0.043]; force[0.043]; border[0.043]; humanitarian[0.042]; ground[0.042]; major[0.041]; rebels[0.041]; Canadian[0.041]; governmental[0.041]; consider[0.041]; multinational[0.041]; agencies[0.041]; group[0.040]; NGOs[0.040]; General[0.040]; told[0.040]; area[0.040]; area[0.040]; deploy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baril\u001b[39m ==> ENTITY: \u001b[32mMaurice Baril\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.078:0.078[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDefence[0.044]; deploy[0.044]; General[0.044]; Canadian[0.044]; ground[0.043]; ground[0.043]; Rwanda[0.042]; aircrew[0.042]; Thursday[0.041]; rebel[0.041]; airdrops[0.041]; airdrops[0.041]; airdrops[0.041]; rebels[0.041]; rebels[0.041]; rebels[0.041]; rebels[0.041]; force[0.041]; force[0.041]; army[0.041]; multinational[0.041]; multinational[0.041]; Goma[0.040]; Goma[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mKinshasa\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mDemocratic Republic of the Congo\u001b[39m <---> \u001b[32mKinshasa\u001b[39m\t\nSCORES: global= 0.289:0.288[\u001b[31m0.001\u001b[39m]; local()= 0.209:0.178[\u001b[31m0.031\u001b[39m]; log p(e|m)= -1.382:0.000[\u001b[32m1.382\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRwanda[0.044]; eastern[0.044]; eastern[0.044]; eastern[0.044]; eastern[0.044]; Zaire[0.043]; Zaire[0.043]; Zaire[0.043]; Zaire[0.043]; Zairean[0.043]; Zairean[0.043]; Rwandan[0.042]; Rwandan[0.042]; refugees[0.040]; refugees[0.040]; refugees[0.040]; refugees[0.040]; General[0.040]; General[0.040]; west[0.040]; Laurent[0.040]; government[0.039]; Nairobi[0.039]; rebel[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zairean\u001b[39m ==> ENTITY: \u001b[32mDemocratic Republic of the Congo\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaire[0.050]; eastern[0.047]; Zairean[0.047]; Zairean[0.047]; Kinshasa[0.045]; government[0.043]; government[0.043]; security[0.042]; rebels[0.041]; tens[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; general[0.041]; reconnaissance[0.040]; group[0.040]; peace[0.040]; thousands[0.039]; ground[0.039]; denied[0.039]; lives[0.039]; British[0.038]; added[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zaire ==> ENTITY: \u001b[32mDemocratic Republic of the Congo\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.162:0.162[0]; log p(e|m)= -2.002:-2.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaire[0.048]; Zaire[0.048]; eastern[0.046]; eastern[0.046]; Rwandan[0.043]; force[0.043]; land[0.043]; east[0.043]; west[0.043]; refugees[0.042]; refugees[0.042]; living[0.040]; Nairobi[0.040]; Nairobi[0.040]; commander[0.040]; multinational[0.040]; charge[0.039]; miles[0.039]; Canadian[0.039]; said[0.039]; general[0.039]; valley[0.039]; main[0.039]; group[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zaire ==> ENTITY: \u001b[32mDemocratic Republic of the Congo\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.166:0.166[0]; log p(e|m)= -2.002:-2.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaire[0.048]; Zaire[0.048]; eastern[0.045]; eastern[0.045]; Rwandan[0.043]; Kinshasa[0.043]; force[0.042]; land[0.042]; east[0.042]; west[0.042]; refugees[0.042]; refugees[0.042]; refugees[0.042]; government[0.042]; living[0.040]; Nairobi[0.040]; Nairobi[0.040]; commander[0.040]; multinational[0.039]; order[0.039]; charge[0.039]; miles[0.039]; Canadian[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canadian ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.076:0.076[0]; log p(e|m)= -0.412:-0.412[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \narea[0.045]; area[0.045]; eastern[0.044]; major[0.043]; Laurent[0.043]; force[0.043]; force[0.043]; probably[0.042]; fall[0.042]; multinational[0.041]; multinational[0.041]; Doug[0.041]; rule[0.041]; non[0.041]; refugees[0.041]; time[0.041]; border[0.041]; home[0.041]; Zaire[0.041]; Zaire[0.041]; Thursday[0.040]; second[0.040]; local[0.040]; significant[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zaire ==> ENTITY: \u001b[32mDemocratic Republic of the Congo\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.145:0.145[0]; log p(e|m)= -2.002:-2.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \neastern[0.047]; Zairean[0.046]; Zairean[0.046]; Zairean[0.046]; Kinshasa[0.044]; rebels[0.043]; government[0.043]; government[0.043]; security[0.041]; resort[0.040]; tens[0.040]; added[0.040]; peace[0.040]; mandate[0.040]; ground[0.040]; denied[0.040]; said[0.040]; said[0.040]; said[0.040]; general[0.040]; coordinate[0.040]; fly[0.040]; operate[0.039]; wide[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rwandan\u001b[39m ==> ENTITY: \u001b[32mRwanda\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRwanda[0.055]; Zaire[0.050]; Zaire[0.050]; refugees[0.049]; refugees[0.049]; army[0.048]; Zairean[0.048]; Zairean[0.048]; Kinshasa[0.047]; Kinshasa[0.047]; force[0.047]; force[0.047]; area[0.047]; government[0.047]; eastern[0.046]; eastern[0.046]; local[0.046]; Kabila[0.046]; leader[0.046]; rebel[0.046]; Defence[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zaire ==> ENTITY: \u001b[32mDemocratic Republic of the Congo\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.201:0.201[0]; log p(e|m)= -2.002:-2.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaire[0.048]; Zaire[0.048]; Rwanda[0.046]; eastern[0.045]; eastern[0.045]; eastern[0.045]; Zairean[0.045]; Zairean[0.045]; army[0.044]; Rwandan[0.043]; Rwandan[0.043]; Kinshasa[0.043]; Kinshasa[0.043]; force[0.042]; land[0.042]; west[0.042]; deploy[0.042]; refugees[0.042]; refugees[0.042]; refugees[0.042]; refugees[0.042]; government[0.042]; rebel[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kindu\u001b[39m ==> ENTITY: \u001b[32mKindu\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \narea[0.049]; Bukavu[0.048]; Bukavu[0.048]; Kinshasa[0.047]; Kisangani[0.047]; northwest[0.047]; north[0.046]; Zairean[0.046]; Goma[0.046]; rebels[0.046]; medical[0.045]; Walikale[0.045]; west[0.045]; town[0.045]; town[0.045]; Masisi[0.044]; road[0.044]; planes[0.044]; Frontieres[0.044]; groups[0.043]; holding[0.043]; Medecins[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kisangani\u001b[39m ==> ENTITY: \u001b[32mKisangani\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBukavu[0.047]; Bukavu[0.047]; Kindu[0.046]; Goma[0.045]; Walikale[0.045]; Masisi[0.043]; north[0.042]; rebels[0.042]; Zairean[0.042]; west[0.041]; area[0.041]; medical[0.041]; northwest[0.040]; planes[0.040]; refugees[0.040]; refugees[0.040]; refugees[0.040]; general[0.040]; office[0.040]; town[0.040]; town[0.040]; groups[0.040]; thing[0.039]; road[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Young ==> ENTITY: \u001b[32mDoug Young (politician)\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.061:0.061[0]; log p(e|m)= -1.130:-1.130[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYoung[0.045]; Canadian[0.044]; Doug[0.044]; Thursday[0.043]; probably[0.042]; probably[0.042]; fall[0.042]; Defence[0.042]; Minister[0.042]; supporters[0.042]; group[0.042]; hands[0.041]; said[0.041]; said[0.041]; militarily[0.041]; apparently[0.041]; airdrops[0.041]; airdrops[0.041]; airdrops[0.041]; airdrops[0.041]; forward[0.041]; rebels[0.041]; action[0.041]; consider[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zaire ==> ENTITY: \u001b[32mDemocratic Republic of the Congo\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.191:0.191[0]; log p(e|m)= -2.002:-2.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaire[0.046]; Zaire[0.046]; Zaire[0.046]; eastern[0.044]; eastern[0.044]; eastern[0.044]; Zairean[0.044]; Rwandan[0.042]; Kinshasa[0.042]; force[0.041]; land[0.041]; east[0.041]; west[0.041]; refugees[0.041]; refugees[0.041]; refugees[0.041]; government[0.041]; rebels[0.041]; living[0.039]; Nairobi[0.039]; Nairobi[0.039]; commander[0.039]; multinational[0.038]; order[0.038]; \t\n\r [=====================>........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s605ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2166/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1280testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.073:0.073[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNational[0.044]; National[0.044]; January[0.044]; late[0.043]; ended[0.043]; said[0.042]; said[0.042]; early[0.042]; day[0.042]; million[0.042]; British[0.042]; half[0.041]; Friday[0.041]; ends[0.041]; taking[0.041]; lower[0.041]; lower[0.041]; lower[0.041]; lower[0.041]; Service[0.041]; Service[0.041]; Feb[0.041]; profit[0.041]; pressured[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.076:0.076[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmillion[0.045]; week[0.043]; cash[0.043]; units[0.043]; January[0.043]; January[0.043]; late[0.042]; anticipated[0.042]; British[0.042]; early[0.042]; early[0.042]; November[0.041]; normal[0.041]; normal[0.041]; day[0.041]; calling[0.041]; December[0.041]; December[0.041]; contract[0.041]; set[0.041]; states[0.041]; thermal[0.041]; settled[0.041]; settled[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NYMEX\u001b[39m ==> ENTITY: \u001b[32mNew York Mercantile Exchange\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfutures[0.048]; Midcontinent[0.046]; traded[0.045]; Nymex[0.045]; Nymex[0.045]; Nymex[0.045]; contracts[0.045]; Permian[0.045]; Permian[0.045]; cash[0.043]; close[0.043]; cents[0.042]; cents[0.042]; cents[0.042]; cents[0.042]; cents[0.042]; cents[0.042]; cents[0.042]; cents[0.042]; cents[0.042]; cents[0.042]; cents[0.042]; ended[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NWS\u001b[39m ==> ENTITY: \u001b[32mNational Weather Service\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWeather[0.045]; Weather[0.045]; Weather[0.045]; late[0.044]; Henry[0.044]; early[0.042]; early[0.042]; November[0.042]; Friday[0.042]; Coast[0.041]; forecasts[0.041]; forecast[0.041]; forecast[0.041]; January[0.041]; January[0.041]; cash[0.041]; National[0.040]; National[0.040]; December[0.040]; December[0.040]; million[0.040]; week[0.040]; Service[0.039]; Service[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = British ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.093:0.093[0]; log p(e|m)= -0.489:-0.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nearly[0.044]; early[0.044]; ended[0.044]; late[0.043]; Friday[0.042]; states[0.042]; week[0.042]; January[0.042]; January[0.042]; taking[0.042]; December[0.041]; December[0.041]; Henry[0.041]; York[0.041]; half[0.041]; Service[0.041]; Service[0.041]; units[0.041]; National[0.041]; National[0.041]; million[0.041]; day[0.041]; support[0.040]; low[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NYMEX\u001b[39m ==> ENTITY: \u001b[32mNew York Mercantile Exchange\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfutures[0.047]; Nymex[0.045]; Permian[0.044]; Permian[0.044]; cash[0.042]; close[0.042]; cents[0.041]; cents[0.041]; cents[0.041]; cents[0.041]; cents[0.041]; cents[0.041]; cents[0.041]; cents[0.041]; ended[0.041]; Southwest[0.041]; prices[0.041]; prices[0.041]; pressured[0.041]; Waha[0.040]; lost[0.040]; narrowed[0.040]; session[0.040]; premium[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.101:0.101[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntraded[0.045]; January[0.043]; January[0.043]; nomination[0.042]; Thursday[0.042]; Thursday[0.042]; week[0.042]; ended[0.042]; December[0.042]; December[0.042]; market[0.042]; November[0.042]; anticipated[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; shortly[0.041]; early[0.041]; Canada[0.041]; support[0.041]; expects[0.041]; set[0.040]; cash[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NYMEX\u001b[39m ==> ENTITY: \u001b[32mNew York Mercantile Exchange\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfutures[0.049]; Nymex[0.046]; million[0.044]; cents[0.043]; cents[0.043]; day[0.042]; ended[0.042]; pressured[0.042]; said[0.041]; settled[0.041]; settled[0.041]; sharply[0.041]; profit[0.041]; bearish[0.041]; calling[0.041]; early[0.040]; low[0.040]; units[0.040]; lower[0.040]; lower[0.040]; lower[0.040]; lower[0.040]; Hub[0.040]; January[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Henry Hub\u001b[39m ==> ENTITY: \u001b[32mHenry Hub\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNymex[0.045]; Nymex[0.045]; futures[0.044]; futures[0.044]; thermal[0.044]; million[0.043]; units[0.043]; bearish[0.042]; York[0.041]; settled[0.041]; settled[0.041]; weather[0.041]; said[0.041]; said[0.041]; said[0.041]; British[0.041]; pressured[0.041]; Weather[0.041]; Weather[0.041]; Weather[0.041]; low[0.040]; cents[0.040]; cents[0.040]; driven[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NYMEX\u001b[39m ==> ENTITY: \u001b[32mNew York Mercantile Exchange\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfutures[0.046]; Midcontinent[0.044]; traded[0.044]; Nymex[0.044]; Nymex[0.044]; contracts[0.043]; Permian[0.043]; Permian[0.043]; market[0.043]; traders[0.042]; cash[0.041]; cash[0.041]; cents[0.040]; cents[0.040]; cents[0.040]; cents[0.040]; contract[0.040]; ended[0.040]; prices[0.040]; prices[0.040]; prices[0.040]; prices[0.040]; pressured[0.040]; storage[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = National Weather Service\u001b[39m ==> ENTITY: \u001b[32mNational Weather Service\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWeather[0.049]; Weather[0.049]; weather[0.046]; forecasts[0.045]; Service[0.045]; forecast[0.045]; forecast[0.045]; million[0.044]; National[0.044]; Gulf[0.043]; Nymex[0.042]; Nymex[0.042]; late[0.042]; Coast[0.042]; expects[0.042]; states[0.042]; early[0.042]; early[0.042]; week[0.042]; anticipated[0.042]; high[0.042]; low[0.042]; low[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NYMEX\u001b[39m ==> ENTITY: \u001b[32mNew York Mercantile Exchange\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfutures[0.055]; futures[0.055]; Nymex[0.052]; million[0.049]; cents[0.048]; cents[0.048]; trader[0.048]; day[0.048]; ended[0.047]; pressured[0.047]; said[0.046]; said[0.046]; settled[0.046]; settled[0.046]; sharply[0.046]; profit[0.046]; bearish[0.046]; calling[0.046]; early[0.045]; low[0.045]; units[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NYMEX\u001b[39m ==> ENTITY: \u001b[32mNew York Mercantile Exchange\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfutures[0.046]; Midcontinent[0.044]; traded[0.044]; Nymex[0.044]; Nymex[0.044]; contracts[0.043]; Permian[0.043]; Permian[0.043]; market[0.043]; traders[0.042]; cash[0.041]; cash[0.041]; cents[0.040]; cents[0.040]; cents[0.040]; cents[0.040]; cents[0.040]; cents[0.040]; contract[0.040]; ended[0.040]; prices[0.040]; prices[0.040]; prices[0.040]; prices[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = National Weather Service\u001b[39m ==> ENTITY: \u001b[32mNational Weather Service\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWeather[0.047]; Weather[0.047]; weather[0.045]; forecasts[0.043]; Service[0.043]; forecast[0.043]; forecast[0.043]; million[0.043]; National[0.042]; Nymex[0.041]; Nymex[0.041]; late[0.041]; expects[0.041]; states[0.041]; early[0.041]; low[0.040]; January[0.040]; December[0.040]; December[0.040]; said[0.040]; said[0.040]; said[0.040]; Friday[0.039]; bearish[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.082:0.082[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nweekend[0.044]; weekend[0.044]; weekend[0.044]; Thursday[0.044]; Thursday[0.044]; Alberta[0.044]; western[0.043]; February[0.042]; weather[0.042]; January[0.042]; January[0.042]; West[0.041]; finished[0.041]; lost[0.041]; settling[0.041]; tally[0.040]; earlier[0.040]; East[0.040]; unseasonably[0.040]; slipped[0.040]; similarly[0.040]; close[0.040]; ended[0.040]; traded[0.040]; \t\n\r [=====================>........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s551ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2181/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1299testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.094:0.094[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.048]; league[0.048]; week[0.043]; week[0.043]; week[0.043]; ended[0.043]; Friday[0.043]; Thursday[0.043]; topped[0.042]; performing[0.042]; place[0.042]; tops[0.041]; weekly[0.041]; High[0.040]; close[0.040]; rose[0.040]; said[0.040]; returns[0.040]; returns[0.040]; returns[0.040]; terms[0.040]; terms[0.040]; managed[0.040]; world[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.056]; Canadian[0.051]; Australia[0.048]; territory[0.048]; league[0.047]; league[0.047]; country[0.046]; Friday[0.045]; Sweden[0.045]; Thursday[0.045]; week[0.044]; week[0.044]; nation[0.044]; returns[0.044]; returns[0.044]; returns[0.044]; returns[0.044]; return[0.043]; France[0.043]; paltry[0.043]; British[0.043]; Italy[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.051]; league[0.050]; country[0.049]; Italy[0.046]; France[0.046]; returns[0.046]; returns[0.046]; return[0.046]; British[0.046]; managed[0.045]; Returns[0.045]; government[0.045]; government[0.045]; lost[0.044]; International[0.044]; Stephen[0.044]; Japan[0.044]; followed[0.044]; gain[0.044]; Spanish[0.044]; terms[0.043]; terms[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.106:0.106[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.046]; country[0.046]; dollar[0.046]; dollar[0.046]; dollar[0.046]; Canada[0.045]; Canada[0.045]; British[0.044]; British[0.044]; Australia[0.043]; Australia[0.043]; weekly[0.043]; biggest[0.043]; included[0.043]; Returns[0.043]; followed[0.042]; paltry[0.042]; November[0.042]; performers[0.042]; Italy[0.042]; certain[0.041]; Japan[0.041]; available[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japanese ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.173:0.173[0]; log p(e|m)= -0.992:-0.992[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.069]; country[0.063]; country[0.063]; nation[0.061]; Spanish[0.059]; territory[0.058]; Australia[0.058]; Australia[0.058]; British[0.058]; British[0.058]; Canada[0.057]; Canada[0.057]; Canada[0.057]; world[0.057]; Italy[0.057]; government[0.055]; government[0.055]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canadian ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.412:-0.412[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.054]; Canada[0.054]; country[0.045]; dollar[0.045]; British[0.042]; terms[0.042]; terms[0.042]; territory[0.042]; Australia[0.041]; Sweden[0.040]; league[0.040]; performers[0.040]; world[0.040]; paltry[0.040]; close[0.040]; Spanish[0.040]; week[0.040]; week[0.040]; place[0.039]; performing[0.039]; nation[0.039]; Thursday[0.039]; currency[0.039]; currency[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; Canada[0.045]; Canada[0.045]; country[0.044]; Italy[0.044]; Italy[0.044]; nation[0.042]; Australia[0.042]; place[0.042]; France[0.041]; British[0.041]; week[0.041]; week[0.041]; week[0.041]; world[0.041]; bloc[0.040]; minus[0.040]; territory[0.040]; Canadian[0.040]; terms[0.040]; terms[0.040]; government[0.040]; government[0.040]; government[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.126:0.126[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.046]; country[0.044]; country[0.044]; Australia[0.044]; Australia[0.044]; league[0.044]; Canada[0.043]; government[0.042]; government[0.042]; British[0.042]; lost[0.042]; lost[0.042]; Spanish[0.042]; performers[0.041]; International[0.041]; Japan[0.041]; performance[0.040]; biggest[0.040]; followed[0.040]; losers[0.040]; return[0.040]; return[0.040]; November[0.040]; included[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spanish ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.150:0.150[0]; log p(e|m)= -1.241:-1.241[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.047]; country[0.047]; Italy[0.045]; territory[0.044]; nation[0.043]; France[0.042]; Canada[0.042]; Canada[0.042]; Canada[0.042]; world[0.042]; British[0.041]; British[0.041]; return[0.041]; return[0.041]; gain[0.041]; government[0.040]; government[0.040]; government[0.040]; league[0.040]; poorest[0.040]; lost[0.040]; lost[0.040]; lost[0.040]; November[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.056]; Canadian[0.051]; Australia[0.049]; territory[0.048]; league[0.047]; country[0.047]; country[0.047]; Sweden[0.045]; Thursday[0.045]; week[0.045]; nation[0.044]; returns[0.044]; returns[0.044]; returns[0.044]; return[0.044]; return[0.044]; world[0.043]; France[0.043]; paltry[0.043]; British[0.043]; managed[0.043]; managed[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.055]; league[0.047]; Italian[0.047]; France[0.044]; week[0.042]; week[0.042]; week[0.042]; Sweden[0.042]; Thursday[0.041]; government[0.041]; government[0.041]; world[0.041]; Friday[0.041]; place[0.041]; currency[0.040]; currency[0.040]; managed[0.039]; BTPs[0.039]; topped[0.039]; offered[0.039]; ended[0.039]; said[0.039]; close[0.039]; tops[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Salomon Brothers\u001b[39m ==> ENTITY: \u001b[32mSalomon Brothers\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \noffered[0.044]; Salomon[0.044]; lost[0.043]; performing[0.043]; bonds[0.043]; bonds[0.043]; London[0.042]; bond[0.042]; bond[0.042]; bond[0.042]; week[0.042]; week[0.042]; week[0.042]; said[0.041]; close[0.041]; Treasuries[0.041]; weekly[0.041]; managed[0.041]; ended[0.040]; rose[0.040]; Thursday[0.040]; returns[0.040]; returns[0.040]; returns[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Salomon Brothers\u001b[39m ==> ENTITY: \u001b[32mSalomon Brothers\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSalomon[0.046]; lost[0.045]; lost[0.045]; lost[0.045]; market[0.044]; performing[0.044]; bonds[0.044]; bonds[0.044]; bonds[0.044]; bonds[0.044]; offering[0.044]; bond[0.044]; bond[0.044]; gain[0.043]; dollar[0.043]; dollar[0.043]; dollar[0.043]; performers[0.043]; performance[0.043]; Treasuries[0.042]; weekly[0.042]; managed[0.042]; followed[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = British ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.489:-0.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.048]; country[0.048]; Canada[0.046]; Canada[0.046]; nation[0.045]; Australia[0.045]; France[0.044]; territory[0.044]; Sweden[0.043]; Canadian[0.043]; government[0.043]; government[0.043]; government[0.043]; world[0.043]; poorest[0.043]; Japan[0.042]; local[0.042]; performers[0.042]; performing[0.041]; week[0.041]; terms[0.041]; terms[0.041]; November[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.157:0.157[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.047]; Italy[0.044]; Italy[0.044]; country[0.043]; world[0.042]; league[0.042]; Canada[0.042]; Canada[0.042]; week[0.042]; week[0.042]; week[0.042]; Sweden[0.042]; government[0.041]; government[0.041]; government[0.041]; government[0.041]; British[0.041]; place[0.040]; territory[0.040]; Friday[0.040]; nation[0.040]; Thursday[0.040]; Canadian[0.039]; topped[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = British ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.140:0.140[0]; log p(e|m)= -0.489:-0.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.059]; Canada[0.057]; Stephen[0.057]; Australia[0.056]; France[0.054]; government[0.053]; government[0.053]; Italy[0.053]; Japan[0.052]; included[0.051]; performers[0.051]; biggest[0.051]; account[0.051]; International[0.050]; terms[0.050]; terms[0.050]; followed[0.050]; certain[0.050]; November[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapanese[0.045]; Australia[0.045]; Australia[0.045]; country[0.043]; country[0.043]; league[0.043]; Italy[0.043]; territory[0.042]; Canada[0.042]; Canada[0.042]; minus[0.041]; minus[0.041]; minus[0.041]; minus[0.041]; British[0.041]; British[0.041]; France[0.041]; government[0.040]; government[0.040]; included[0.040]; Stephen[0.040]; performers[0.040]; Brothers[0.040]; November[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.152:0.152[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.054]; league[0.047]; Italian[0.047]; France[0.044]; topped[0.042]; week[0.042]; week[0.042]; week[0.042]; Canada[0.042]; ended[0.041]; Sweden[0.041]; Thursday[0.041]; government[0.040]; government[0.040]; government[0.040]; world[0.040]; Friday[0.040]; place[0.040]; rose[0.040]; lost[0.039]; managed[0.039]; offered[0.039]; said[0.039]; close[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.045]; country[0.045]; league[0.045]; Canada[0.044]; Canada[0.044]; Japan[0.042]; week[0.042]; government[0.042]; government[0.042]; government[0.042]; British[0.042]; France[0.041]; France[0.041]; territory[0.041]; terms[0.041]; terms[0.041]; return[0.040]; return[0.040]; Sweden[0.040]; world[0.040]; Canadian[0.040]; paltry[0.040]; better[0.040]; Japanese[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.046]; league[0.046]; Canada[0.045]; Japan[0.043]; Italy[0.043]; performers[0.042]; performance[0.042]; France[0.042]; terms[0.041]; terms[0.041]; return[0.041]; returns[0.041]; returns[0.041]; Stephen[0.041]; British[0.041]; certain[0.041]; November[0.041]; minus[0.041]; minus[0.041]; minus[0.041]; Spanish[0.040]; followed[0.040]; International[0.040]; included[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italian ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.174:0.174[0]; log p(e|m)= -0.759:-0.759[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.061]; Italy[0.061]; league[0.053]; France[0.049]; local[0.048]; local[0.048]; Canada[0.047]; Canada[0.047]; British[0.047]; week[0.047]; week[0.047]; week[0.047]; world[0.046]; Sweden[0.045]; said[0.045]; Canadian[0.045]; managed[0.044]; nation[0.044]; Thursday[0.044]; terms[0.044]; terms[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.149:0.149[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.071]; country[0.071]; France[0.070]; league[0.069]; Australia[0.069]; lost[0.067]; Canada[0.067]; November[0.066]; government[0.065]; government[0.065]; British[0.065]; managed[0.065]; Japan[0.065]; followed[0.064]; gain[0.064]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.145:0.145[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanadian[0.047]; Canada[0.047]; Canada[0.047]; Canada[0.047]; country[0.045]; country[0.045]; Italy[0.044]; British[0.043]; British[0.043]; nation[0.042]; Spanish[0.041]; territory[0.041]; France[0.041]; France[0.041]; Australia[0.040]; government[0.040]; government[0.040]; world[0.040]; gain[0.039]; Japan[0.038]; league[0.038]; terms[0.038]; terms[0.038]; terms[0.038]; \t\n\r [=====================>........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s418ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2204/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1256testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sofia ==> ENTITY: \u001b[32mSofia\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntown[0.047]; Kazanluk[0.045]; Kazanluk[0.045]; Bulgaria[0.044]; Bulgaria[0.044]; Bulgaria[0.044]; central[0.044]; central[0.044]; Bulgarian[0.044]; area[0.042]; state[0.042]; Friday[0.042]; Turkish[0.040]; soared[0.040]; economy[0.040]; collapse[0.040]; residents[0.040]; Bombings[0.039]; Violent[0.039]; case[0.039]; said[0.039]; Ottoman[0.039]; scores[0.038]; communism[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kazanluk\u001b[39m ==> ENTITY: \u001b[32mKazanlak\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKazanluk[0.049]; Bulgarian[0.046]; area[0.044]; Bulgaria[0.044]; Bulgaria[0.044]; Bulgaria[0.044]; town[0.044]; Turkish[0.044]; Sofia[0.042]; exploded[0.042]; Ottoman[0.041]; central[0.041]; central[0.041]; soared[0.041]; injuries[0.040]; residents[0.040]; communism[0.040]; state[0.039]; economy[0.039]; collapse[0.039]; mosque[0.039]; mosque[0.039]; mosque[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Islam ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.140:0.140[0]; log p(e|m)= -0.121:-0.121[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoslems[0.053]; Christians[0.053]; mosque[0.049]; mosque[0.049]; mosque[0.049]; Ottoman[0.047]; Turkish[0.043]; communism[0.042]; saying[0.040]; said[0.039]; majority[0.039]; central[0.039]; central[0.039]; rule[0.039]; area[0.039]; state[0.039]; criminals[0.038]; converted[0.038]; Violent[0.038]; carried[0.038]; crime[0.038]; clear[0.038]; scores[0.037]; Bombings[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Moslems\u001b[39m ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIslam[0.050]; Christians[0.047]; mosque[0.044]; mosque[0.044]; mosque[0.044]; Ottoman[0.042]; Turkish[0.042]; majority[0.041]; area[0.041]; saying[0.041]; said[0.041]; town[0.041]; state[0.041]; rule[0.041]; central[0.041]; central[0.041]; Friday[0.040]; residents[0.040]; converted[0.040]; injuries[0.040]; economy[0.039]; carried[0.039]; quoted[0.039]; scores[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mTurkish\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mTurkey\u001b[39m <---> \u001b[32mOttoman Empire\u001b[39m\t\nSCORES: global= 0.252:0.230[\u001b[31m0.022\u001b[39m]; local()= 0.155:0.174[\u001b[32m0.019\u001b[39m]; log p(e|m)= -1.546:-2.937[\u001b[31m1.391\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOttoman[0.049]; Bulgaria[0.044]; Bulgaria[0.044]; Bulgaria[0.044]; Christians[0.043]; Moslems[0.043]; Bulgarian[0.043]; town[0.041]; mosque[0.041]; mosque[0.041]; mosque[0.041]; Islam[0.041]; Kazanluk[0.041]; Kazanluk[0.041]; central[0.041]; central[0.041]; area[0.040]; state[0.040]; quoted[0.040]; said[0.040]; rule[0.040]; majority[0.040]; soared[0.040]; Sofia[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bulgarian ==> ENTITY: \u001b[32mBulgaria\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.196:0.196[0]; log p(e|m)= -1.155:-1.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBulgaria[0.047]; Bulgaria[0.047]; Bulgaria[0.047]; Ottoman[0.046]; Kazanluk[0.043]; Kazanluk[0.043]; Turkish[0.043]; Christians[0.042]; Sofia[0.042]; soared[0.041]; central[0.041]; central[0.041]; Moslems[0.041]; communism[0.041]; town[0.041]; rule[0.040]; said[0.040]; state[0.040]; area[0.039]; carried[0.039]; residents[0.039]; saying[0.039]; quoted[0.038]; scores[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kazanluk\u001b[39m ==> ENTITY: \u001b[32mKazanlak\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKazanluk[0.049]; Bulgarian[0.046]; area[0.044]; Bulgaria[0.044]; Bulgaria[0.044]; Bulgaria[0.044]; town[0.044]; Turkish[0.044]; Sofia[0.042]; exploded[0.042]; Ottoman[0.041]; central[0.041]; central[0.041]; soared[0.041]; injuries[0.040]; residents[0.040]; communism[0.040]; state[0.039]; economy[0.039]; collapse[0.039]; mosque[0.039]; mosque[0.039]; mosque[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bulgaria ==> ENTITY: \u001b[32mBulgaria\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.190:0.190[0]; log p(e|m)= -0.317:-0.317[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBulgaria[0.047]; Bulgaria[0.047]; Ottoman[0.046]; Bulgarian[0.045]; communism[0.044]; Kazanluk[0.043]; Kazanluk[0.043]; Christians[0.043]; Sofia[0.042]; Turkish[0.042]; soared[0.041]; central[0.041]; central[0.041]; rule[0.040]; scores[0.040]; state[0.040]; town[0.040]; Moslems[0.040]; area[0.039]; carried[0.039]; said[0.039]; saying[0.039]; economy[0.038]; broke[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bulgaria ==> ENTITY: \u001b[32mBulgaria\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.190:0.190[0]; log p(e|m)= -0.317:-0.317[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBulgaria[0.047]; Bulgaria[0.047]; Ottoman[0.046]; Bulgarian[0.045]; communism[0.044]; Kazanluk[0.043]; Kazanluk[0.043]; Christians[0.043]; Sofia[0.042]; Turkish[0.042]; soared[0.041]; central[0.041]; central[0.041]; rule[0.040]; scores[0.040]; state[0.040]; town[0.040]; Moslems[0.040]; area[0.039]; carried[0.039]; said[0.039]; saying[0.039]; economy[0.038]; broke[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bulgaria ==> ENTITY: \u001b[32mBulgaria\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.191:0.191[0]; log p(e|m)= -0.317:-0.317[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBulgaria[0.047]; Bulgaria[0.047]; Ottoman[0.046]; Bulgarian[0.045]; communism[0.044]; Kazanluk[0.043]; Kazanluk[0.043]; Christians[0.043]; Sofia[0.042]; Turkish[0.042]; soared[0.041]; central[0.041]; central[0.041]; rule[0.040]; scores[0.040]; state[0.040]; town[0.040]; Moslems[0.040]; area[0.039]; carried[0.039]; said[0.039]; saying[0.039]; economy[0.038]; collapse[0.038]; \t\n\r [=====================>........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s335ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2214/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1348testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colorado ==> ENTITY: \u001b[32mColorado\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.033:0.033[0]; log p(e|m)= -0.269:-0.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWinner[0.045]; winner[0.044]; history[0.044]; Saturday[0.043]; Alpine[0.043]; Vail[0.043]; Norway[0.043]; Age[0.043]; won[0.042]; finals[0.042]; victories[0.042]; Skiing[0.041]; Austria[0.041]; Austria[0.041]; Lillehammer[0.040]; victor[0.040]; World[0.040]; World[0.040]; World[0.040]; Cup[0.040]; Cup[0.040]; Cup[0.040]; Cup[0.040]; Nation[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.048]; women[0.044]; Norway[0.044]; Cup[0.043]; Cup[0.043]; Cup[0.043]; Cup[0.043]; Women[0.043]; qualifier[0.043]; World[0.043]; World[0.043]; World[0.043]; Lillehammer[0.043]; finals[0.041]; victories[0.041]; won[0.041]; Flachau[0.040]; Alpine[0.040]; history[0.040]; slalom[0.039]; slalom[0.039]; Skiing[0.039]; Renate[0.039]; Saturday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.048]; women[0.044]; Norway[0.044]; Cup[0.043]; Cup[0.043]; Cup[0.043]; Cup[0.043]; Women[0.043]; qualifier[0.043]; World[0.043]; World[0.043]; World[0.043]; Lillehammer[0.043]; finals[0.041]; victories[0.041]; won[0.041]; Flachau[0.040]; Alpine[0.040]; history[0.040]; slalom[0.039]; slalom[0.039]; Skiing[0.039]; Renate[0.039]; Saturday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Flachau\u001b[39m ==> ENTITY: \u001b[32mFlachau\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.135:0.135[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSkiing[0.051]; history[0.050]; Alpine[0.050]; downhill[0.050]; Austria[0.050]; Austria[0.050]; Vail[0.049]; slalom[0.049]; slalom[0.049]; qualifier[0.048]; finals[0.048]; Lillehammer[0.047]; Colorado[0.047]; Downhill[0.047]; Renate[0.046]; Europa[0.046]; Norway[0.045]; results[0.044]; victories[0.044]; Winner[0.044]; super[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Cup ==> ENTITY: \u001b[32mFIS Alpine Ski World Cup\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.226:0.226[0]; log p(e|m)= -2.765:-2.765[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.044]; Alpine[0.044]; Downhill[0.044]; World[0.043]; World[0.043]; Cup[0.043]; Cup[0.043]; Cup[0.043]; Flachau[0.043]; Lillehammer[0.043]; Austria[0.042]; Austria[0.042]; downhill[0.042]; Skiing[0.041]; qualifier[0.041]; finals[0.041]; slalom[0.041]; slalom[0.041]; race[0.040]; Women[0.040]; won[0.039]; Vail[0.038]; victories[0.038]; Europa[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Renate Goetschl\u001b[39m ==> ENTITY: \u001b[32mRenate Götschl\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndownhill[0.045]; victories[0.044]; Alpine[0.044]; slalom[0.044]; slalom[0.044]; Flachau[0.043]; won[0.043]; Lillehammer[0.042]; Downhill[0.042]; Cup[0.042]; Cup[0.042]; Cup[0.042]; Cup[0.042]; Saturday[0.041]; Women[0.041]; Skiing[0.040]; race[0.040]; women[0.040]; finals[0.040]; history[0.040]; youngest[0.040]; results[0.040]; Norway[0.040]; Vail[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWorld[0.044]; World[0.044]; World[0.044]; qualifier[0.044]; Austria[0.043]; Austria[0.043]; finals[0.043]; Cup[0.043]; Cup[0.043]; Cup[0.043]; Cup[0.043]; won[0.041]; Lillehammer[0.041]; Europa[0.041]; race[0.041]; Women[0.041]; Downhill[0.041]; women[0.041]; victories[0.040]; Colorado[0.040]; history[0.039]; Skiing[0.039]; Alpine[0.039]; winner[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lillehammer\u001b[39m ==> ENTITY: \u001b[32mLillehammer\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.045]; Alpine[0.044]; Skiing[0.043]; history[0.043]; Flachau[0.043]; Downhill[0.043]; Austria[0.043]; Austria[0.043]; won[0.042]; downhill[0.042]; qualifier[0.042]; finals[0.042]; women[0.041]; slalom[0.041]; slalom[0.041]; Previous[0.041]; World[0.040]; World[0.040]; World[0.040]; Saturday[0.040]; Cup[0.040]; Cup[0.040]; Cup[0.040]; Cup[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vail ==> ENTITY: \u001b[32mVail, Colorado\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.392:-0.392[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.047]; Downhill[0.045]; Skiing[0.044]; Saturday[0.044]; race[0.043]; downhill[0.043]; Colorado[0.043]; women[0.043]; Lillehammer[0.042]; won[0.042]; history[0.041]; Winner[0.041]; youngest[0.040]; Women[0.040]; qualifier[0.040]; Flachau[0.040]; World[0.040]; World[0.040]; World[0.040]; finals[0.040]; Cup[0.040]; Cup[0.040]; Cup[0.040]; Cup[0.040]; \t\n\r [=====================>........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s380ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2223/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1356testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kanpur ==> ENTITY: \u001b[32mKanpur\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.160:0.160[0]; log p(e|m)= -0.036:-0.036[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.045]; India[0.045]; India[0.045]; India[0.045]; Kanpur[0.045]; Kanpur[0.045]; Kanpur[0.045]; northern[0.044]; South[0.042]; South[0.042]; cricketer[0.042]; cricketer[0.042]; Africa[0.041]; city[0.040]; played[0.040]; Cricket[0.040]; cricketing[0.040]; England[0.040]; England[0.040]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Uttar Pradesh\u001b[39m ==> ENTITY: \u001b[32mUttar Pradesh\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.128:0.128[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricketer[0.046]; cricketer[0.046]; India[0.045]; India[0.045]; India[0.045]; Ranji[0.044]; northern[0.042]; Kanpur[0.042]; Kanpur[0.042]; cricketing[0.042]; birthplace[0.042]; Trophy[0.041]; Calcutta[0.041]; England[0.041]; England[0.041]; state[0.041]; championship[0.040]; national[0.040]; captained[0.039]; city[0.039]; return[0.039]; place[0.039]; visit[0.039]; came[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South African ==> ENTITY: \u001b[32mSouth Africa national cricket team\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.161:0.161[0]; log p(e|m)= -2.617:-2.617[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricketer[0.050]; cricketer[0.050]; England[0.047]; England[0.047]; England[0.047]; South[0.047]; Cricket[0.046]; captained[0.045]; cricketing[0.045]; India[0.045]; India[0.045]; India[0.045]; India[0.045]; India[0.045]; Africa[0.045]; Greig[0.044]; played[0.044]; Saturday[0.044]; national[0.043]; Ranji[0.043]; Cape[0.043]; Trophy[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bob Woolmer\u001b[39m ==> ENTITY: \u001b[32mBob Woolmer\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.234:0.234[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWoolmer[0.050]; Woolmer[0.050]; Woolmer[0.050]; Woolmer[0.050]; Woolmer[0.050]; cricketer[0.048]; cricketer[0.048]; England[0.046]; England[0.046]; England[0.046]; Cricket[0.046]; India[0.044]; India[0.044]; India[0.044]; India[0.044]; India[0.044]; cricketing[0.044]; coach[0.042]; South[0.041]; South[0.041]; African[0.041]; played[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Calcutta ==> ENTITY: \u001b[32mKolkata\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.101:0.101[0]; log p(e|m)= -0.022:-0.022[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.045]; India[0.045]; India[0.045]; England[0.044]; England[0.044]; stadium[0.044]; Kanpur[0.043]; cricketing[0.043]; South[0.043]; born[0.042]; Ranji[0.041]; place[0.041]; championship[0.041]; Trophy[0.041]; played[0.040]; Woolmer[0.040]; Woolmer[0.040]; Woolmer[0.040]; Woolmer[0.040]; house[0.040]; captained[0.040]; left[0.040]; tests[0.039]; member[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kanpur ==> ENTITY: \u001b[32mKanpur\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.177:0.177[0]; log p(e|m)= -0.036:-0.036[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.053]; India[0.053]; India[0.053]; India[0.053]; Kanpur[0.053]; Kanpur[0.053]; Kanpur[0.053]; northern[0.052]; South[0.049]; South[0.049]; cricketer[0.049]; cricketer[0.049]; Africa[0.048]; city[0.048]; played[0.048]; Cricket[0.048]; cricketing[0.047]; England[0.047]; England[0.047]; England[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland cricket team\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.180:0.180[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; England[0.047]; cricketer[0.046]; cricketer[0.046]; captained[0.044]; represented[0.042]; national[0.042]; India[0.042]; India[0.042]; India[0.042]; India[0.042]; cricketing[0.042]; played[0.041]; South[0.040]; Greig[0.040]; Trophy[0.040]; Saturday[0.039]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cape Town\u001b[39m ==> ENTITY: \u001b[32mCape Town\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstadium[0.047]; South[0.046]; South[0.046]; England[0.045]; England[0.045]; played[0.044]; Saturday[0.042]; cricketer[0.042]; place[0.042]; African[0.042]; captained[0.041]; Trophy[0.041]; represented[0.041]; Province[0.040]; national[0.040]; Africa[0.040]; championship[0.040]; left[0.039]; India[0.039]; India[0.039]; India[0.039]; India[0.039]; came[0.039]; cricketing[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland cricket team\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.192:0.192[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.056]; England[0.056]; cricketer[0.054]; cricketer[0.054]; captained[0.052]; Cricket[0.050]; represented[0.050]; national[0.050]; India[0.050]; India[0.050]; India[0.050]; India[0.050]; India[0.050]; cricketing[0.049]; played[0.049]; South[0.047]; South[0.047]; Greig[0.047]; Trophy[0.046]; Saturday[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kanpur ==> ENTITY: \u001b[32mKanpur\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.172:0.172[0]; log p(e|m)= -0.036:-0.036[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.044]; India[0.044]; India[0.044]; India[0.044]; India[0.044]; Kanpur[0.044]; Kanpur[0.044]; Kanpur[0.044]; Pradesh[0.044]; northern[0.043]; South[0.041]; South[0.041]; cricketer[0.041]; cricketer[0.041]; state[0.041]; Uttar[0.041]; Africa[0.040]; city[0.039]; played[0.039]; Cricket[0.039]; cricketing[0.039]; England[0.039]; England[0.039]; England[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIndia\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIndia\u001b[39m <---> \u001b[32mIndia national cricket team\u001b[39m\t\nSCORES: global= 0.254:0.246[\u001b[31m0.007\u001b[39m]; local()= 0.144:0.188[\u001b[32m0.044\u001b[39m]; log p(e|m)= -0.136:-3.507[\u001b[31m3.371\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.048]; India[0.048]; India[0.048]; cricketer[0.046]; cricketer[0.046]; England[0.043]; England[0.043]; England[0.043]; Cricket[0.043]; played[0.041]; cricketing[0.041]; African[0.040]; represented[0.040]; Kanpur[0.040]; Kanpur[0.040]; Kanpur[0.040]; Kanpur[0.040]; northern[0.039]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = India ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.058]; India[0.058]; cricketer[0.055]; England[0.052]; England[0.052]; captained[0.051]; played[0.050]; cricketing[0.050]; Ranji[0.049]; Calcutta[0.049]; African[0.048]; championship[0.048]; represented[0.048]; came[0.048]; Trophy[0.048]; place[0.048]; Kanpur[0.048]; Kanpur[0.048]; national[0.047]; northern[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kanpur ==> ENTITY: \u001b[32mKanpur\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.160:0.160[0]; log p(e|m)= -0.036:-0.036[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.045]; India[0.045]; India[0.045]; India[0.045]; Kanpur[0.045]; Kanpur[0.045]; Kanpur[0.045]; northern[0.044]; South[0.042]; South[0.042]; cricketer[0.042]; cricketer[0.042]; Africa[0.041]; city[0.040]; played[0.040]; Cricket[0.040]; cricketing[0.040]; England[0.040]; England[0.040]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWoolmer\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBob Woolmer\u001b[39m <---> \u001b[32mClarence Woolmer\u001b[39m\t\nSCORES: global= 0.285:0.275[\u001b[31m0.009\u001b[39m]; local()= 0.109:0.141[\u001b[32m0.032\u001b[39m]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWoolmer[0.050]; Woolmer[0.050]; India[0.044]; born[0.044]; captained[0.043]; Calcutta[0.043]; South[0.043]; father[0.042]; father[0.042]; Kanpur[0.042]; hospital[0.041]; son[0.041]; Africa[0.041]; senior[0.040]; cobra[0.040]; hose[0.040]; played[0.040]; test[0.040]; national[0.040]; championship[0.039]; came[0.039]; stadium[0.039]; pipe[0.039]; Cape[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Clarence Woolmer\u001b[39m ==> ENTITY: \u001b[32mClarence Woolmer\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricketer[0.046]; cricketer[0.046]; Woolmer[0.043]; Woolmer[0.043]; Woolmer[0.043]; Woolmer[0.043]; Woolmer[0.043]; Woolmer[0.043]; Ranji[0.042]; coach[0.042]; India[0.041]; India[0.041]; India[0.041]; India[0.041]; born[0.041]; England[0.041]; England[0.041]; England[0.041]; captained[0.041]; Calcutta[0.040]; South[0.040]; Trophy[0.040]; cricketing[0.040]; African[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Woolmer ==> ENTITY: \u001b[32mBob Woolmer\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.241:0.241[0]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWoolmer[0.047]; Woolmer[0.047]; Woolmer[0.047]; Woolmer[0.047]; Woolmer[0.047]; Woolmer[0.047]; Woolmer[0.047]; cricketer[0.046]; cricketer[0.046]; England[0.043]; England[0.043]; England[0.043]; Cricket[0.043]; Ranji[0.042]; coach[0.041]; India[0.041]; India[0.041]; India[0.041]; India[0.041]; India[0.041]; cricketing[0.041]; born[0.041]; captained[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland cricket team\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.177:0.177[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.048]; England[0.048]; cricketer[0.047]; cricketer[0.047]; Cricket[0.043]; represented[0.043]; India[0.042]; India[0.042]; India[0.042]; India[0.042]; cricketing[0.042]; played[0.042]; South[0.040]; South[0.040]; Greig[0.040]; Saturday[0.039]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; United[0.039]; coach[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWoolmer\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBob Woolmer\u001b[39m <---> \u001b[32mClarence Woolmer\u001b[39m\t\nSCORES: global= 0.286:0.278[\u001b[31m0.008\u001b[39m]; local()= 0.195:0.197[\u001b[32m0.002\u001b[39m]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWoolmer[0.046]; Woolmer[0.046]; Woolmer[0.046]; Woolmer[0.046]; Woolmer[0.046]; cricketer[0.045]; England[0.043]; England[0.043]; Ranji[0.041]; coach[0.041]; India[0.041]; India[0.041]; India[0.041]; India[0.041]; cricketing[0.040]; born[0.040]; captained[0.040]; Calcutta[0.040]; South[0.040]; South[0.040]; Trophy[0.039]; African[0.039]; Uttar[0.038]; United[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = India ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.173:0.173[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.056]; India[0.056]; India[0.056]; India[0.056]; cricketer[0.054]; cricketer[0.054]; England[0.050]; England[0.050]; England[0.050]; captained[0.049]; played[0.048]; cricketing[0.048]; Ranji[0.048]; Calcutta[0.047]; African[0.047]; championship[0.047]; represented[0.047]; came[0.046]; Trophy[0.046]; place[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ranji Trophy\u001b[39m ==> ENTITY: \u001b[32mRanji Trophy\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.142:0.142[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.046]; India[0.046]; India[0.046]; cricketer[0.046]; England[0.044]; England[0.044]; captained[0.043]; Kanpur[0.043]; Kanpur[0.043]; cricketing[0.042]; played[0.042]; championship[0.042]; Calcutta[0.042]; national[0.040]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; South[0.039]; coach[0.039]; senior[0.039]; hose[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Woolmer ==> ENTITY: \u001b[32mBob Woolmer\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.241:0.241[0]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWoolmer[0.047]; Woolmer[0.047]; Woolmer[0.047]; Woolmer[0.047]; Woolmer[0.047]; Woolmer[0.047]; Woolmer[0.047]; cricketer[0.046]; cricketer[0.046]; England[0.043]; England[0.043]; England[0.043]; Cricket[0.043]; Ranji[0.042]; coach[0.041]; India[0.041]; India[0.041]; India[0.041]; India[0.041]; India[0.041]; cricketing[0.041]; born[0.041]; captained[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Woolmer ==> ENTITY: \u001b[32mBob Woolmer\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.222:0.222[0]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWoolmer[0.045]; Woolmer[0.045]; Woolmer[0.045]; Woolmer[0.045]; Woolmer[0.045]; cricketer[0.044]; cricketer[0.044]; England[0.042]; England[0.042]; England[0.042]; Cricket[0.042]; Ranji[0.041]; coach[0.040]; India[0.040]; India[0.040]; India[0.040]; India[0.040]; India[0.040]; cricketing[0.040]; born[0.040]; captained[0.039]; South[0.039]; South[0.039]; Trophy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Woolmer ==> ENTITY: \u001b[32mBob Woolmer\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.198:0.198[0]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWoolmer[0.047]; Woolmer[0.047]; Woolmer[0.047]; Woolmer[0.047]; cricketer[0.046]; cricketer[0.046]; England[0.043]; England[0.043]; Cricket[0.043]; coach[0.041]; India[0.041]; India[0.041]; India[0.041]; born[0.041]; South[0.040]; South[0.040]; African[0.039]; father[0.039]; Kanpur[0.039]; Kanpur[0.039]; Kanpur[0.039]; Kanpur[0.039]; Africa[0.038]; Bob[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = India ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.048]; India[0.048]; India[0.048]; cricketer[0.046]; cricketer[0.046]; England[0.043]; England[0.043]; Cricket[0.043]; played[0.042]; cricketing[0.041]; African[0.040]; Kanpur[0.040]; Kanpur[0.040]; Kanpur[0.040]; Kanpur[0.040]; northern[0.039]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; Woolmer[0.039]; South[0.039]; South[0.039]; birthplace[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United Province ==> ENTITY: \u001b[32mUttar Pradesh\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.143:0.143[0]; log p(e|m)= -1.088:-1.088[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.045]; India[0.045]; India[0.045]; state[0.045]; Pradesh[0.043]; northern[0.042]; born[0.042]; came[0.042]; cricketer[0.042]; cricketer[0.042]; birthplace[0.041]; Calcutta[0.041]; Kanpur[0.041]; Kanpur[0.041]; Uttar[0.041]; national[0.041]; given[0.040]; England[0.040]; England[0.040]; England[0.040]; renamed[0.040]; senior[0.040]; Clarence[0.040]; Ranji[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kanpur ==> ENTITY: \u001b[32mKanpur\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.036:-0.036[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPradesh[0.046]; India[0.046]; India[0.046]; India[0.046]; India[0.046]; Calcutta[0.044]; cricketer[0.043]; state[0.043]; Uttar[0.043]; played[0.041]; played[0.041]; England[0.041]; England[0.041]; stadium[0.040]; cricketing[0.040]; renamed[0.040]; Ranji[0.040]; Woolmer[0.040]; Woolmer[0.040]; Woolmer[0.040]; Woolmer[0.040]; Woolmer[0.040]; Woolmer[0.040]; national[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tony Greig\u001b[39m ==> ENTITY: \u001b[32mTony Greig\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricketer[0.046]; cricketer[0.046]; England[0.045]; England[0.045]; England[0.045]; cricketing[0.044]; Woolmer[0.041]; Woolmer[0.041]; Woolmer[0.041]; Woolmer[0.041]; Woolmer[0.041]; Woolmer[0.041]; Woolmer[0.041]; India[0.040]; India[0.040]; India[0.040]; India[0.040]; played[0.040]; born[0.040]; championship[0.040]; South[0.040]; South[0.040]; Saturday[0.039]; captained[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = India ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.166:0.166[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.047]; India[0.047]; India[0.047]; India[0.047]; cricketer[0.045]; cricketer[0.045]; England[0.042]; England[0.042]; England[0.042]; Cricket[0.042]; captained[0.041]; played[0.041]; cricketing[0.040]; Ranji[0.040]; African[0.039]; championship[0.039]; represented[0.039]; Trophy[0.039]; place[0.039]; Kanpur[0.039]; Kanpur[0.039]; Kanpur[0.039]; Kanpur[0.039]; Kanpur[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWoolmer\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBob Woolmer\u001b[39m <---> \u001b[32mClarence Woolmer\u001b[39m\t\nSCORES: global= 0.286:0.278[\u001b[31m0.008\u001b[39m]; local()= 0.195:0.194[\u001b[31m0.001\u001b[39m]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWoolmer[0.050]; Woolmer[0.050]; Woolmer[0.050]; Woolmer[0.050]; Woolmer[0.050]; cricketer[0.049]; England[0.046]; England[0.046]; Ranji[0.045]; coach[0.044]; India[0.044]; India[0.044]; India[0.044]; cricketing[0.044]; born[0.044]; captained[0.043]; Calcutta[0.043]; South[0.043]; Trophy[0.043]; African[0.042]; Uttar[0.042]; United[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Africa ==> ENTITY: \u001b[32mSouth Africa national cricket team\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.162:0.162[0]; log p(e|m)= -3.016:-3.016[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricketer[0.053]; cricketer[0.053]; England[0.051]; England[0.051]; South[0.049]; Cricket[0.049]; African[0.048]; cricketing[0.048]; India[0.048]; India[0.048]; India[0.048]; India[0.048]; Greig[0.047]; played[0.046]; Saturday[0.046]; Woolmer[0.045]; Woolmer[0.045]; Woolmer[0.045]; Woolmer[0.045]; Woolmer[0.045]; coach[0.045]; \t\n\r [======================>.......................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s178ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2253/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1306testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Okla ==> ENTITY: \u001b[32mOklahoma\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.072:0.072[0]; log p(e|m)= -1.100:-1.100[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWk[0.046]; Texas[0.046]; Panhandle[0.043]; Steers[0.043]; lbs[0.042]; Thursday[0.042]; Thursday[0.042]; cattle[0.042]; cattle[0.042]; area[0.042]; Amarillo[0.042]; confirmed[0.042]; heifers[0.041]; reporting[0.041]; includes[0.041]; Sales[0.041]; contracted[0.041]; Holiday[0.041]; Select[0.040]; Select[0.040]; Week[0.040]; Week[0.040]; head[0.040]; head[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = USDA\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Agriculture\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.088:0.088[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWk[0.045]; includes[0.044]; fed[0.043]; Steers[0.042]; reporting[0.042]; lbs[0.042]; lbs[0.042]; Thursday[0.042]; Thursday[0.042]; cattle[0.042]; cattle[0.042]; week[0.042]; week[0.042]; Amarillo[0.041]; contracted[0.041]; Texas[0.041]; shipped[0.041]; Panhandle[0.041]; reported[0.041]; tested[0.041]; Holiday[0.041]; choice[0.041]; Okla[0.041]; Sales[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Amarillo ==> ENTITY: \u001b[32mAmarillo, Texas\u001b[39m\t\nSCORES: global= 0.241:0.241[0]; local()= 0.085:0.085[0]; log p(e|m)= -0.177:-0.177[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTexas[0.047]; Wk[0.046]; area[0.043]; Panhandle[0.043]; Steers[0.043]; lbs[0.042]; lbs[0.042]; contracted[0.042]; cattle[0.042]; cattle[0.042]; Thursday[0.042]; Thursday[0.042]; Okla[0.041]; Sales[0.041]; heifers[0.040]; week[0.040]; week[0.040]; date[0.040]; Trade[0.040]; Select[0.040]; Select[0.040]; Week[0.040]; Week[0.040]; reporting[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = USDA\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Agriculture\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.088:0.088[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWk[0.045]; includes[0.044]; fed[0.043]; Steers[0.042]; reporting[0.042]; lbs[0.042]; lbs[0.042]; Thursday[0.042]; Thursday[0.042]; cattle[0.042]; cattle[0.042]; week[0.042]; week[0.042]; Amarillo[0.041]; contracted[0.041]; Texas[0.041]; shipped[0.041]; Panhandle[0.041]; reported[0.041]; tested[0.041]; Holiday[0.041]; choice[0.041]; Okla[0.041]; Sales[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Texas ==> ENTITY: \u001b[32mTexas\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.076:0.076[0]; log p(e|m)= -0.161:-0.161[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThursday[0.046]; Thursday[0.046]; week[0.046]; week[0.046]; area[0.045]; Panhandle[0.045]; Amarillo[0.044]; Week[0.044]; Week[0.044]; roundup[0.044]; Okla[0.044]; includes[0.043]; slow[0.043]; Slaughter[0.043]; Slaughter[0.043]; Wk[0.043]; head[0.042]; head[0.042]; Choice[0.042]; date[0.042]; confirmed[0.042]; Holiday[0.042]; reported[0.041]; \t\n\r [======================>.......................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s485ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2258/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1248testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bratislava\u001b[39m ==> ENTITY: \u001b[32mBratislava\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.094:0.094[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBratislava[0.048]; Bratislava[0.048]; city[0.044]; second[0.044]; Friday[0.043]; added[0.042]; police[0.041]; police[0.041]; police[0.041]; group[0.041]; attack[0.041]; attack[0.041]; unable[0.041]; rabbi[0.041]; rabbi[0.041]; local[0.040]; caused[0.040]; American[0.040]; escaped[0.040]; attacked[0.040]; attacked[0.040]; procedure[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bratislava\u001b[39m ==> ENTITY: \u001b[32mBratislava\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.086:0.086[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBratislava[0.050]; Bratislava[0.050]; city[0.046]; second[0.045]; Friday[0.045]; available[0.043]; added[0.043]; police[0.043]; police[0.043]; police[0.043]; group[0.043]; attack[0.043]; attack[0.043]; unable[0.042]; rabbi[0.042]; rabbi[0.042]; local[0.042]; caused[0.042]; American[0.042]; escaped[0.042]; attacked[0.042]; attacked[0.042]; procedure[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.110:0.110[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntold[0.049]; American[0.048]; local[0.047]; said[0.046]; said[0.046]; added[0.046]; available[0.046]; group[0.046]; insulted[0.045]; spokesman[0.045]; spokesman[0.045]; spokesman[0.045]; spokesman[0.045]; attack[0.045]; attack[0.045]; Friday[0.045]; second[0.045]; attacked[0.044]; attacked[0.044]; city[0.044]; attackers[0.044]; caused[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = American ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.006:0.006[0]; log p(e|m)= -0.130:-0.130[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.046]; escaped[0.045]; said[0.044]; said[0.044]; Meyers[0.044]; Meyers[0.044]; Meyers[0.044]; group[0.044]; attack[0.044]; apprehended[0.044]; Friday[0.044]; available[0.043]; told[0.043]; local[0.043]; added[0.043]; kicked[0.043]; attackers[0.042]; details[0.042]; skinheads[0.042]; skinheads[0.042]; skinheads[0.042]; detained[0.042]; caused[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bratislava\u001b[39m ==> ENTITY: \u001b[32mBratislava\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.094:0.094[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBratislava[0.048]; Bratislava[0.048]; city[0.044]; second[0.044]; Friday[0.043]; added[0.042]; police[0.041]; police[0.041]; police[0.041]; group[0.041]; attack[0.041]; attack[0.041]; unable[0.041]; rabbi[0.041]; rabbi[0.041]; local[0.040]; caused[0.040]; escaped[0.040]; attacked[0.040]; attacked[0.040]; procedure[0.040]; said[0.040]; said[0.040]; details[0.040]; \t\n\r [======================>.......................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s425ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2263/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1352testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colorado ==> ENTITY: \u001b[32mColorado\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.045:0.045[0]; log p(e|m)= -0.269:-0.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.045]; Saturday[0.045]; Alpine[0.045]; Vail[0.044]; won[0.043]; finished[0.043]; Skiing[0.043]; Austria[0.042]; according[0.042]; took[0.042]; Germany[0.042]; Italy[0.041]; World[0.041]; World[0.041]; Cup[0.041]; Cup[0.041]; provisional[0.040]; race[0.040]; Wins[0.040]; Renate[0.040]; women[0.040]; downhill[0.040]; Downhill[0.039]; Kostner[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; women[0.046]; Cup[0.045]; Cup[0.045]; World[0.044]; World[0.044]; won[0.044]; Italy[0.044]; took[0.043]; finished[0.042]; Alpine[0.041]; second[0.041]; Skiing[0.040]; Renate[0.040]; Katja[0.040]; Saturday[0.040]; Downhill[0.040]; provisional[0.040]; Seizinger[0.039]; downhill[0.039]; race[0.039]; Kostner[0.039]; Wins[0.039]; Colorado[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwon[0.048]; Cup[0.046]; Cup[0.046]; Austria[0.045]; Germany[0.044]; second[0.044]; finished[0.044]; took[0.042]; Alpine[0.042]; Wins[0.042]; World[0.042]; World[0.042]; Skiing[0.041]; provisional[0.041]; Saturday[0.040]; Kostner[0.040]; race[0.040]; Seizinger[0.040]; Downhill[0.039]; women[0.039]; Colorado[0.039]; according[0.038]; Vail[0.037]; downhill[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.090:0.090[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.047]; Downhill[0.045]; Italy[0.044]; won[0.044]; Skiing[0.043]; Cup[0.043]; Cup[0.043]; Saturday[0.043]; World[0.042]; World[0.042]; second[0.042]; provisional[0.042]; Kostner[0.041]; Wins[0.041]; Alpine[0.041]; women[0.041]; race[0.041]; finished[0.040]; downhill[0.040]; took[0.040]; Renate[0.039]; according[0.038]; Vail[0.038]; Seizinger[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Renate Goetschl\u001b[39m ==> ENTITY: \u001b[32mRenate Götschl\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.206:0.206[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndownhill[0.046]; Alpine[0.045]; second[0.045]; won[0.044]; Downhill[0.043]; Cup[0.043]; Cup[0.043]; Kostner[0.043]; Saturday[0.042]; Seizinger[0.042]; finished[0.042]; Skiing[0.041]; race[0.041]; women[0.041]; results[0.041]; Vail[0.041]; Italy[0.040]; Colorado[0.040]; Germany[0.040]; provisional[0.040]; Austria[0.040]; World[0.039]; World[0.039]; took[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Katja Seizinger\u001b[39m ==> ENTITY: \u001b[32mKatja Seizinger\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrace[0.046]; won[0.044]; Cup[0.043]; Cup[0.043]; Kostner[0.043]; Alpine[0.043]; Downhill[0.043]; women[0.043]; results[0.042]; downhill[0.042]; Germany[0.042]; Vail[0.042]; finished[0.042]; second[0.042]; Skiing[0.041]; Wins[0.041]; provisional[0.041]; Colorado[0.041]; Renate[0.041]; Austria[0.041]; World[0.038]; World[0.038]; Italy[0.038]; Saturday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Cup ==> ENTITY: \u001b[32mFIS Alpine Ski World Cup\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.206:0.206[0]; log p(e|m)= -2.765:-2.765[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.047]; Downhill[0.046]; World[0.046]; Cup[0.045]; Austria[0.045]; downhill[0.044]; Skiing[0.044]; Italy[0.043]; Germany[0.042]; race[0.042]; second[0.042]; won[0.041]; Vail[0.041]; finished[0.040]; Seizinger[0.040]; Wins[0.040]; Colorado[0.040]; women[0.040]; provisional[0.040]; Saturday[0.039]; Kostner[0.039]; results[0.038]; Renate[0.038]; took[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vail ==> ENTITY: \u001b[32mVail, Colorado\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.392:-0.392[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlpine[0.048]; Downhill[0.046]; Skiing[0.045]; Saturday[0.045]; race[0.044]; downhill[0.044]; according[0.043]; Colorado[0.043]; women[0.043]; second[0.043]; won[0.042]; finished[0.041]; World[0.041]; World[0.041]; Cup[0.041]; Cup[0.041]; Austria[0.040]; took[0.040]; results[0.039]; Wins[0.039]; Germany[0.039]; Italy[0.038]; Kostner[0.038]; Renate[0.038]; \t\n\r [======================>.......................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s524ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2271/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1316testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Copenhagen ==> ENTITY: \u001b[32mCopenhagen\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.091:0.091[0]; log p(e|m)= -0.119:-0.119[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.051]; Copenhagen[0.047]; Wednesday[0.042]; failure[0.042]; Jakobsen[0.042]; group[0.042]; group[0.042]; half[0.042]; months[0.042]; largely[0.041]; introduce[0.041]; share[0.041]; crowns[0.041]; percent[0.041]; result[0.041]; pre[0.041]; seven[0.041]; products[0.041]; million[0.040]; million[0.040]; said[0.040]; profit[0.040]; medical[0.040]; consensus[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWednesday\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSheffield Wednesday F.C.\u001b[39m <---> \u001b[32mWednesday\u001b[39m\t\nSCORES: global= 0.256:0.244[\u001b[31m0.012\u001b[39m]; local()= 0.078:0.067[\u001b[31m0.011\u001b[39m]; log p(e|m)= -1.720:-0.288[\u001b[32m1.433\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nH1[0.046]; months[0.046]; seven[0.044]; million[0.043]; million[0.043]; percent[0.043]; reporting[0.043]; half[0.042]; profit[0.042]; pre[0.042]; equipment[0.041]; analysts[0.041]; failure[0.041]; performance[0.041]; said[0.041]; result[0.041]; predicted[0.040]; group[0.040]; group[0.040]; sees[0.040]; forecasts[0.040]; tax[0.040]; weak[0.040]; unchanged[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Copenhagen ==> ENTITY: \u001b[32mCopenhagen\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.096:0.096[0]; log p(e|m)= -0.119:-0.119[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.051]; Copenhagen[0.043]; Wednesday[0.042]; failure[0.042]; introduce[0.042]; Jakobsen[0.042]; months[0.042]; largely[0.042]; seven[0.042]; share[0.041]; crowns[0.041]; sees[0.041]; percent[0.041]; group[0.041]; group[0.041]; products[0.041]; million[0.041]; million[0.041]; said[0.041]; profit[0.041]; medical[0.041]; far[0.040]; result[0.040]; Reuter[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Denmark ==> ENTITY: \u001b[32mDenmark\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.075:0.075[0]; log p(e|m)= -0.268:-0.268[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJakobsen[0.045]; lost[0.044]; seven[0.044]; Copenhagen[0.042]; Copenhagen[0.042]; Wednesday[0.042]; crowns[0.042]; forecasts[0.042]; unchanged[0.042]; months[0.042]; average[0.042]; pre[0.042]; compared[0.042]; half[0.041]; percent[0.041]; largely[0.041]; million[0.041]; million[0.041]; products[0.041]; group[0.041]; group[0.041]; far[0.040]; Soeren[0.040]; reporting[0.040]; \t\n\r [======================>.......................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s570ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2275/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1177testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Melbourne ==> ENTITY: \u001b[32mMelbourne\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.168:-0.168[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; Australia[0.048]; match[0.047]; match[0.047]; Cricket[0.044]; opening[0.042]; balls[0.042]; day[0.042]; series[0.041]; Wednesday[0.041]; victory[0.040]; rounder[0.040]; Greg[0.040]; Friday[0.040]; Brian[0.040]; beat[0.040]; overhauled[0.040]; tour[0.040]; disappointing[0.040]; misconduct[0.039]; Day[0.039]; total[0.039]; overs[0.039]; defeats[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Melbourne ==> ENTITY: \u001b[32mMelbourne\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.124:0.124[0]; log p(e|m)= -0.168:-0.168[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.052]; Australia[0.052]; coach[0.049]; team[0.049]; tournament[0.047]; captain[0.046]; crowd[0.045]; balls[0.045]; fifth[0.045]; Geoff[0.044]; month[0.044]; Stuart[0.044]; Pakistan[0.044]; groin[0.044]; Shane[0.044]; Tuesday[0.044]; rounder[0.044]; despite[0.044]; mid[0.043]; tour[0.043]; Ian[0.043]; Warne[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shane Warne\u001b[39m ==> ENTITY: \u001b[32mShane Warne\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.208:0.208[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStuart[0.044]; Shivnarine[0.044]; Australia[0.043]; match[0.043]; wicket[0.042]; wicket[0.042]; wicket[0.042]; wicketkeeper[0.042]; balls[0.042]; balls[0.042]; Lara[0.042]; Lara[0.042]; Lara[0.042]; rounder[0.042]; rounder[0.042]; Ian[0.041]; Chanderpaul[0.041]; Chanderpaul[0.041]; Clive[0.040]; Marsh[0.040]; innings[0.040]; innings[0.040]; captain[0.040]; day[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Indies ==> ENTITY: \u001b[32mWest Indies cricket team\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.215:0.215[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLara[0.045]; Lara[0.045]; Lara[0.045]; Lara[0.045]; Australia[0.045]; Australia[0.045]; rounder[0.044]; rounder[0.044]; series[0.044]; day[0.044]; match[0.043]; match[0.043]; match[0.043]; Cricket[0.043]; Stuart[0.043]; wicket[0.043]; wicket[0.043]; fifth[0.043]; overs[0.042]; innings[0.042]; Indies[0.042]; opening[0.041]; Wednesday[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWorld Series\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustralian Tri-Series\u001b[39m <---> \u001b[32mWorld Series Cricket\u001b[39m\t\nSCORES: global= 0.255:0.253[\u001b[31m0.002\u001b[39m]; local()= 0.182:0.168[\u001b[31m0.015\u001b[39m]; log p(e|m)= -3.324:-3.324[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.047]; Australia[0.047]; Cricket[0.045]; series[0.045]; match[0.044]; match[0.044]; match[0.044]; beat[0.042]; rounder[0.041]; tour[0.041]; Melbourne[0.041]; day[0.040]; Wednesday[0.040]; opening[0.040]; Friday[0.040]; Robert[0.040]; total[0.040]; total[0.040]; came[0.040]; balls[0.040]; balls[0.040]; victory[0.040]; successive[0.040]; dismissed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.178:0.178[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.050]; match[0.045]; Pakistan[0.043]; Stuart[0.043]; Melbourne[0.042]; captain[0.042]; rounder[0.042]; Shivnarine[0.042]; despite[0.042]; wicketkeeper[0.042]; Marsh[0.041]; Warne[0.041]; coach[0.041]; fifth[0.041]; wicket[0.040]; wicket[0.040]; wicket[0.040]; month[0.040]; Lara[0.040]; Lara[0.040]; Lara[0.040]; balls[0.040]; balls[0.040]; Tuesday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pakistan ==> ENTITY: \u001b[32mPakistan national cricket team\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.149:0.149[0]; log p(e|m)= -2.749:-2.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.047]; tournament[0.046]; Australia[0.045]; Australia[0.045]; Shivnarine[0.044]; wicketkeeper[0.043]; month[0.042]; innings[0.042]; innings[0.042]; Tuesday[0.042]; Marsh[0.042]; Chanderpaul[0.041]; Chanderpaul[0.041]; hander[0.041]; despite[0.041]; crease[0.040]; captain[0.040]; West[0.040]; Lara[0.040]; arrive[0.040]; competing[0.040]; run[0.040]; gritty[0.040]; Melbourne[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Clive Lloyd\u001b[39m ==> ENTITY: \u001b[32mClive Lloyd\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStuart[0.048]; wicketkeeper[0.047]; Marsh[0.047]; Shivnarine[0.046]; Lara[0.046]; Lara[0.046]; Lara[0.046]; match[0.046]; Chanderpaul[0.046]; Chanderpaul[0.046]; balls[0.045]; balls[0.045]; Ian[0.045]; Australia[0.045]; rounder[0.045]; rounder[0.045]; Warne[0.045]; captain[0.044]; wicket[0.044]; wicket[0.044]; wicket[0.044]; Geoff[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Blewett\u001b[39m ==> ENTITY: \u001b[32mGreg Blewett\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.044]; Australia[0.044]; match[0.044]; match[0.044]; match[0.044]; Stuart[0.043]; wicketkeeper[0.042]; rounder[0.042]; rounder[0.042]; Warne[0.042]; wicket[0.041]; wicket[0.041]; wicket[0.041]; Lara[0.041]; Lara[0.041]; Lara[0.041]; Lara[0.041]; crease[0.041]; overs[0.041]; balls[0.040]; balls[0.040]; balls[0.040]; Shane[0.040]; series[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lara\u001b[39m ==> ENTITY: \u001b[32mBrian Lara\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLara[0.045]; Lara[0.045]; match[0.044]; match[0.044]; match[0.044]; Cricket[0.043]; balls[0.043]; balls[0.043]; rounder[0.043]; wicket[0.042]; series[0.042]; Australia[0.041]; Australia[0.041]; wickets[0.041]; overs[0.041]; day[0.040]; fours[0.040]; run[0.040]; run[0.040]; Indies[0.040]; Indies[0.040]; total[0.040]; total[0.040]; dismissed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Indies ==> ENTITY: \u001b[32mWest Indies cricket team\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.182:0.182[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLara[0.044]; Lara[0.044]; Lara[0.044]; Australia[0.043]; Australia[0.043]; Chanderpaul[0.043]; Chanderpaul[0.043]; Shivnarine[0.043]; Stuart[0.042]; wicket[0.042]; wicket[0.042]; rounder[0.042]; team[0.041]; innings[0.041]; innings[0.041]; hander[0.041]; catch[0.041]; Marsh[0.041]; wicketkeeper[0.041]; Indies[0.041]; tournament[0.040]; Pakistan[0.040]; Tuesday[0.040]; Ian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.189:0.189[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.050]; match[0.045]; match[0.045]; Cricket[0.045]; series[0.043]; Melbourne[0.043]; rounder[0.042]; day[0.042]; overs[0.041]; wickets[0.041]; Lara[0.040]; Lara[0.040]; Lara[0.040]; balls[0.040]; balls[0.040]; Wednesday[0.040]; Robert[0.040]; successive[0.040]; World[0.040]; victory[0.040]; tour[0.040]; Brian[0.040]; West[0.040]; West[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Geoff Marsh\u001b[39m ==> ENTITY: \u001b[32mGeoff Marsh\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.047]; Australia[0.047]; Stuart[0.044]; team[0.044]; captain[0.044]; coach[0.043]; match[0.043]; despite[0.042]; Melbourne[0.042]; Ian[0.042]; Healy[0.041]; fifth[0.041]; rounder[0.041]; Shivnarine[0.041]; Pakistan[0.041]; wicketkeeper[0.040]; month[0.040]; Clive[0.040]; Tuesday[0.040]; vice[0.040]; Shane[0.040]; Warne[0.039]; groin[0.039]; hander[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lara\u001b[39m ==> ENTITY: \u001b[32mBrian Lara\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.224:0.224[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLara[0.044]; Lara[0.044]; match[0.044]; match[0.044]; Stuart[0.043]; balls[0.042]; balls[0.042]; balls[0.042]; rounder[0.042]; rounder[0.042]; Warne[0.042]; wicket[0.042]; wicket[0.042]; wicket[0.042]; Marsh[0.042]; series[0.041]; wicketkeeper[0.041]; Australia[0.041]; day[0.040]; fours[0.040]; Ian[0.040]; hander[0.040]; catch[0.040]; run[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.204:0.204[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.049]; match[0.045]; match[0.045]; match[0.045]; Cricket[0.045]; series[0.042]; Stuart[0.042]; Melbourne[0.042]; rounder[0.042]; rounder[0.042]; day[0.041]; overs[0.041]; wickets[0.040]; fifth[0.040]; wicket[0.040]; wicket[0.040]; Lara[0.040]; Lara[0.040]; Lara[0.040]; balls[0.040]; balls[0.040]; Wednesday[0.040]; crucial[0.040]; successive[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAustralia\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustralia national cricket team\u001b[39m <---> \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.250:0.238[\u001b[31m0.012\u001b[39m]; local()= 0.141:0.084[\u001b[31m0.057\u001b[39m]; log p(e|m)= -3.507:-0.172[\u001b[32m3.335\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.051]; team[0.046]; Pakistan[0.044]; tournament[0.043]; Melbourne[0.043]; captain[0.043]; Shivnarine[0.043]; despite[0.042]; wicketkeeper[0.042]; Marsh[0.042]; coach[0.042]; competing[0.042]; month[0.041]; World[0.040]; vice[0.040]; West[0.040]; Indies[0.040]; innings[0.040]; innings[0.040]; hander[0.039]; Healy[0.039]; middle[0.039]; Geoff[0.039]; Ian[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lara\u001b[39m ==> ENTITY: \u001b[32mBrian Lara\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.217:0.217[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nShivnarine[0.044]; Lara[0.044]; Lara[0.044]; match[0.044]; Stuart[0.043]; balls[0.042]; balls[0.042]; rounder[0.042]; Warne[0.042]; wicket[0.042]; wicket[0.042]; wicket[0.042]; Marsh[0.042]; Chanderpaul[0.042]; Chanderpaul[0.042]; wicketkeeper[0.041]; Australia[0.041]; fours[0.040]; Ian[0.040]; hander[0.040]; catch[0.040]; run[0.040]; added[0.040]; Clive[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shivnarine Chanderpaul\u001b[39m ==> ENTITY: \u001b[32mShivnarine Chanderpaul\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChanderpaul[0.045]; wicketkeeper[0.044]; Lara[0.043]; Lara[0.043]; hander[0.043]; wicket[0.043]; innings[0.043]; innings[0.043]; despite[0.042]; catch[0.042]; crease[0.041]; crease[0.041]; Marsh[0.041]; Australia[0.041]; Australia[0.041]; Warne[0.041]; captain[0.041]; Pakistan[0.041]; Indies[0.041]; Indies[0.041]; Ian[0.040]; team[0.040]; West[0.040]; West[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Indies ==> ENTITY: \u001b[32mWest Indies cricket team\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.188:0.188[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLara[0.043]; Lara[0.043]; Lara[0.043]; Australia[0.043]; Chanderpaul[0.042]; Chanderpaul[0.042]; rounder[0.042]; rounder[0.042]; Shivnarine[0.042]; day[0.042]; match[0.042]; Stuart[0.042]; wicket[0.042]; wicket[0.042]; wicket[0.042]; fifth[0.041]; wicketkeeper[0.041]; innings[0.041]; innings[0.041]; innings[0.041]; hander[0.041]; catch[0.041]; Marsh[0.041]; Indies[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Greg Blewett\u001b[39m ==> ENTITY: \u001b[32mGreg Blewett\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.045]; Australia[0.045]; match[0.044]; match[0.044]; match[0.044]; Stuart[0.044]; rounder[0.042]; rounder[0.042]; Warne[0.042]; wicket[0.041]; wicket[0.041]; wicket[0.041]; Lara[0.041]; Lara[0.041]; Lara[0.041]; crease[0.041]; overs[0.041]; balls[0.040]; balls[0.040]; balls[0.040]; Shane[0.040]; series[0.040]; wickets[0.040]; Blewett[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lara\u001b[39m ==> ENTITY: \u001b[32mBrian Lara\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLara[0.045]; Lara[0.045]; match[0.045]; match[0.045]; Cricket[0.044]; balls[0.043]; rounder[0.043]; series[0.042]; Australia[0.042]; Australia[0.042]; wickets[0.042]; overs[0.042]; day[0.041]; run[0.041]; run[0.041]; Indies[0.040]; Indies[0.040]; total[0.040]; dismissed[0.040]; successive[0.040]; limited[0.039]; disappointing[0.039]; beat[0.039]; gave[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Brian Lara\u001b[39m ==> ENTITY: \u001b[32mBrian Lara\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLara[0.045]; Lara[0.045]; match[0.045]; match[0.045]; Cricket[0.044]; balls[0.043]; balls[0.043]; rounder[0.043]; series[0.042]; Australia[0.042]; Australia[0.042]; wickets[0.041]; overs[0.041]; day[0.041]; run[0.040]; run[0.040]; Indies[0.040]; Indies[0.040]; total[0.040]; dismissed[0.040]; successive[0.040]; victory[0.039]; Greg[0.039]; limited[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ian Healy\u001b[39m ==> ENTITY: \u001b[32mIan Healy\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.047]; Australia[0.047]; Stuart[0.046]; team[0.043]; wicketkeeper[0.042]; match[0.042]; captain[0.042]; Marsh[0.042]; Shivnarine[0.042]; Warne[0.042]; Pakistan[0.041]; Geoff[0.041]; month[0.041]; despite[0.041]; Tuesday[0.041]; rounder[0.040]; Melbourne[0.040]; Clive[0.040]; vice[0.040]; Lara[0.040]; Lara[0.040]; Lara[0.040]; Shane[0.040]; tournament[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWorld Series\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustralian Tri-Series\u001b[39m <---> \u001b[32mWorld Series Cricket\u001b[39m\t\nSCORES: global= 0.253:0.252[\u001b[31m0.001\u001b[39m]; local()= 0.138:0.143[\u001b[32m0.005\u001b[39m]; log p(e|m)= -3.324:-3.324[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.049]; team[0.046]; tournament[0.044]; Pakistan[0.043]; Melbourne[0.043]; captain[0.043]; despite[0.042]; Ian[0.042]; Shivnarine[0.042]; wicketkeeper[0.042]; competing[0.041]; month[0.041]; hander[0.041]; run[0.041]; middle[0.041]; Healy[0.041]; left[0.040]; left[0.040]; innings[0.040]; innings[0.040]; angered[0.040]; runner[0.040]; Indies[0.040]; crowd[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lara\u001b[39m ==> ENTITY: \u001b[32mBrian Lara\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.218:0.218[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nShivnarine[0.044]; Lara[0.044]; Lara[0.044]; match[0.044]; Stuart[0.043]; balls[0.042]; balls[0.042]; rounder[0.042]; Warne[0.042]; wicket[0.042]; wicket[0.042]; wicket[0.042]; Marsh[0.042]; Chanderpaul[0.041]; Chanderpaul[0.041]; wicketkeeper[0.041]; Australia[0.041]; Pakistan[0.040]; fours[0.040]; Ian[0.040]; hander[0.040]; catch[0.040]; run[0.040]; added[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Stuart Law\u001b[39m ==> ENTITY: \u001b[32mStuart Law\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwicket[0.043]; wicket[0.043]; wicket[0.043]; wicketkeeper[0.043]; Ian[0.043]; balls[0.043]; balls[0.043]; balls[0.043]; Geoff[0.043]; Australia[0.042]; Australia[0.042]; Lara[0.041]; Lara[0.041]; Lara[0.041]; Greg[0.041]; fours[0.041]; captain[0.041]; rounder[0.041]; rounder[0.041]; Shane[0.041]; crease[0.040]; innings[0.040]; successive[0.040]; Clive[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chanderpaul\u001b[39m ==> ENTITY: \u001b[32mShivnarine Chanderpaul\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nShivnarine[0.047]; Chanderpaul[0.045]; wicketkeeper[0.044]; Lara[0.043]; Lara[0.043]; hander[0.043]; wicket[0.043]; innings[0.042]; innings[0.042]; despite[0.042]; crease[0.041]; Marsh[0.041]; Australia[0.041]; Australia[0.041]; Warne[0.041]; captain[0.041]; Pakistan[0.041]; Indies[0.040]; Indies[0.040]; Ian[0.040]; team[0.040]; West[0.039]; West[0.039]; run[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Indies ==> ENTITY: \u001b[32mWest Indies cricket team\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.207:0.207[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLara[0.044]; Lara[0.044]; Lara[0.044]; Australia[0.044]; Australia[0.044]; rounder[0.043]; series[0.043]; day[0.043]; match[0.042]; match[0.042]; match[0.042]; Cricket[0.042]; overs[0.042]; Indies[0.041]; opening[0.040]; Wednesday[0.040]; balls[0.040]; balls[0.040]; victory[0.040]; wickets[0.040]; successive[0.040]; tour[0.040]; World[0.040]; dismissed[0.039]; \t\n\r [======================>.......................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s500ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2303/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1199testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bonn\u001b[39m ==> ENTITY: \u001b[32mBonn\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.059:0.059[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWosz[0.046]; matches[0.046]; Munich[0.045]; Leverkusen[0.044]; Soccer[0.044]; Freiburg[0.044]; German[0.044]; German[0.044]; played[0.044]; division[0.044]; Werder[0.044]; Bremen[0.043]; Friday[0.043]; Karlsruhe[0.043]; Halftime[0.043]; Halftime[0.043]; Halftime[0.043]; Halftime[0.043]; Carl[0.043]; Schalke[0.042]; Bayer[0.042]; Rostock[0.042]; Bochum[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kirsten ==> ENTITY: \u001b[32mUlf Kirsten\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.144:0.144[0]; log p(e|m)= -1.073:-1.073[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.047]; Soccer[0.046]; Wosz[0.045]; played[0.045]; Bayer[0.044]; Leverkusen[0.044]; Bremen[0.043]; Rostock[0.042]; Bochum[0.042]; Freiburg[0.042]; Ramelow[0.041]; German[0.041]; German[0.041]; Hansa[0.041]; Munich[0.041]; Carl[0.041]; Werder[0.040]; Bonn[0.040]; Karlsruhe[0.040]; Friday[0.040]; Schalke[0.039]; Bode[0.039]; 8th[0.038]; 2nd[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Freiburg ==> ENTITY: \u001b[32mSC Freiburg\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.152:0.152[0]; log p(e|m)= -1.743:-1.743[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; Bonn[0.044]; Soccer[0.044]; Munich[0.044]; Wosz[0.044]; German[0.042]; German[0.042]; Leverkusen[0.042]; Ramelow[0.042]; Bremen[0.042]; Werder[0.042]; Bochum[0.041]; Karlsruhe[0.041]; played[0.041]; Halftime[0.041]; Halftime[0.041]; Halftime[0.041]; Halftime[0.041]; Bayer[0.040]; Rostock[0.040]; Schalke[0.040]; division[0.040]; Summaries[0.039]; Summaries[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ramelow\u001b[39m ==> ENTITY: \u001b[32mCarsten Ramelow\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBayer[0.046]; Wosz[0.046]; matches[0.045]; German[0.045]; German[0.045]; played[0.045]; Soccer[0.045]; Leverkusen[0.044]; Werder[0.044]; Freiburg[0.044]; Bochum[0.043]; Karlsruhe[0.043]; Halftime[0.043]; Halftime[0.043]; Halftime[0.043]; Halftime[0.043]; Bremen[0.043]; Schalke[0.042]; Friday[0.042]; Bonn[0.042]; Reich[0.042]; Rostock[0.041]; division[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wosz\u001b[39m ==> ENTITY: \u001b[32mDariusz Wosz\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.047]; matches[0.047]; played[0.046]; Bremen[0.045]; Leverkusen[0.045]; Werder[0.044]; Karlsruhe[0.044]; Munich[0.044]; Bochum[0.044]; Freiburg[0.043]; Halftime[0.043]; Halftime[0.043]; Halftime[0.043]; Halftime[0.043]; Schalke[0.043]; Bonn[0.042]; Rostock[0.042]; Ramelow[0.042]; Summaries[0.042]; Summaries[0.042]; Bayer[0.042]; German[0.041]; German[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMulder\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mErwin Mulder\u001b[39m <---> \u001b[32mYouri Mulder\u001b[39m\t\nSCORES: global= 0.253:0.249[\u001b[31m0.004\u001b[39m]; local()= 0.088:0.102[\u001b[32m0.014\u001b[39m]; log p(e|m)= -2.865:-2.564[\u001b[32m0.301\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.048]; Soccer[0.047]; played[0.045]; Wosz[0.043]; Freiburg[0.043]; Hansa[0.043]; Leverkusen[0.042]; Ramelow[0.042]; Bremen[0.042]; Bayer[0.042]; Karlsruhe[0.042]; Werder[0.042]; Schalke[0.042]; Bochum[0.041]; Munich[0.041]; Friday[0.040]; German[0.040]; German[0.040]; Rostock[0.040]; Carl[0.040]; Bonn[0.039]; pen[0.039]; division[0.039]; Kirsten[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Werder Bremen\u001b[39m ==> ENTITY: \u001b[32mSV Werder Bremen\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.047]; Wosz[0.045]; Soccer[0.045]; Leverkusen[0.044]; Ramelow[0.043]; Schalke[0.043]; Bochum[0.042]; German[0.042]; German[0.042]; played[0.042]; Freiburg[0.042]; Munich[0.042]; Karlsruhe[0.041]; Halftime[0.041]; Halftime[0.041]; Halftime[0.041]; Halftime[0.041]; Bayer[0.041]; Bonn[0.040]; Rostock[0.040]; 2nd[0.040]; division[0.039]; Summaries[0.039]; Summaries[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.080:0.080[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGerman[0.049]; matches[0.048]; Munich[0.046]; played[0.044]; Soccer[0.044]; Wosz[0.044]; Reich[0.043]; Ramelow[0.043]; Bochum[0.042]; Leverkusen[0.042]; Bremen[0.041]; Carl[0.040]; Bayer[0.040]; Bonn[0.040]; Schalke[0.040]; Werder[0.040]; Halftime[0.039]; Halftime[0.039]; Halftime[0.039]; Halftime[0.039]; Kirsten[0.039]; Freiburg[0.039]; Dundee[0.038]; Friday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bochum ==> ENTITY: \u001b[32mVfL Bochum\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.197:0.197[0]; log p(e|m)= -1.510:-1.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.048]; matches[0.048]; Wosz[0.048]; Freiburg[0.045]; played[0.045]; Leverkusen[0.044]; Munich[0.043]; Ramelow[0.043]; German[0.043]; German[0.043]; Bremen[0.043]; Werder[0.043]; Friday[0.043]; Karlsruhe[0.043]; division[0.042]; Halftime[0.042]; Halftime[0.042]; Halftime[0.042]; Halftime[0.042]; Division[0.042]; Schalke[0.042]; Bonn[0.041]; Rostock[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bode ==> ENTITY: \u001b[32mMarco Bode\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.209:0.209[0]; log p(e|m)= -1.398:-1.398[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWosz[0.048]; German[0.046]; German[0.046]; Bochum[0.046]; Ramelow[0.045]; matches[0.045]; Munich[0.045]; Bremen[0.045]; Leverkusen[0.045]; Werder[0.044]; Karlsruhe[0.044]; Soccer[0.044]; Reich[0.044]; Freiburg[0.043]; Bayer[0.042]; Schalke[0.042]; Rostock[0.042]; Carl[0.042]; Bonn[0.042]; played[0.041]; Friday[0.040]; Hansa[0.039]; Kirsten[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bayer Leverkusen\u001b[39m ==> ENTITY: \u001b[32mBayer 04 Leverkusen\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.048]; Soccer[0.047]; division[0.046]; Wosz[0.046]; Ramelow[0.045]; Karlsruhe[0.045]; Werder[0.044]; Schalke[0.044]; Freiburg[0.044]; Bochum[0.044]; Bremen[0.043]; Halftime[0.043]; Halftime[0.043]; Halftime[0.043]; Halftime[0.043]; played[0.043]; Rostock[0.043]; Munich[0.043]; Division[0.042]; German[0.041]; German[0.041]; 8th[0.041]; Bonn[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Schalke\u001b[39m ==> ENTITY: \u001b[32mFC Schalke 04\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.049]; Ramelow[0.046]; Wosz[0.046]; Soccer[0.046]; German[0.045]; German[0.045]; Bremen[0.044]; played[0.044]; Leverkusen[0.044]; Munich[0.043]; Bochum[0.043]; Werder[0.043]; Rostock[0.043]; Karlsruhe[0.043]; Bayer[0.043]; Freiburg[0.042]; Halftime[0.042]; Halftime[0.042]; Halftime[0.042]; Halftime[0.042]; Friday[0.041]; division[0.041]; Bonn[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.080:0.080[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGerman[0.049]; matches[0.048]; Munich[0.046]; played[0.044]; Soccer[0.044]; Wosz[0.044]; Reich[0.043]; Ramelow[0.043]; Bochum[0.042]; Leverkusen[0.042]; Bremen[0.041]; Carl[0.040]; Bayer[0.040]; Bonn[0.040]; Schalke[0.040]; Werder[0.040]; Halftime[0.039]; Halftime[0.039]; Halftime[0.039]; Halftime[0.039]; Kirsten[0.039]; Freiburg[0.039]; Dundee[0.038]; Friday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = 1860 Munich\u001b[39m ==> ENTITY: \u001b[32mTSV 1860 München\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWosz[0.045]; Soccer[0.045]; matches[0.044]; Bochum[0.044]; Freiburg[0.043]; Leverkusen[0.043]; played[0.043]; Werder[0.042]; Bremen[0.042]; division[0.041]; Karlsruhe[0.041]; Halftime[0.041]; Halftime[0.041]; Halftime[0.041]; Halftime[0.041]; 2nd[0.041]; 8th[0.041]; Schalke[0.041]; German[0.041]; German[0.041]; Ramelow[0.040]; Bonn[0.040]; Bayer[0.040]; Rostock[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Stickroth\u001b[39m ==> ENTITY: \u001b[32mThomas Stickroth\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWosz[0.046]; Soccer[0.044]; Freiburg[0.044]; Werder[0.043]; Leverkusen[0.043]; German[0.043]; German[0.043]; Bayer[0.042]; played[0.042]; Bochum[0.042]; Karlsruhe[0.042]; Halftime[0.041]; Halftime[0.041]; Halftime[0.041]; Halftime[0.041]; Schalke[0.041]; Bremen[0.041]; Ramelow[0.041]; 8th[0.041]; Bonn[0.040]; Rostock[0.040]; matches[0.040]; Summaries[0.040]; Summaries[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hansa Rostock\u001b[39m ==> ENTITY: \u001b[32mF.C. Hansa Rostock\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; Soccer[0.045]; Bremen[0.044]; Wosz[0.044]; played[0.043]; Bochum[0.043]; Freiburg[0.043]; Leverkusen[0.042]; German[0.042]; German[0.042]; Werder[0.042]; Ramelow[0.041]; Karlsruhe[0.041]; Halftime[0.041]; Halftime[0.041]; Halftime[0.041]; Halftime[0.041]; Munich[0.041]; Schalke[0.041]; Bonn[0.040]; division[0.040]; Friday[0.040]; Summaries[0.039]; Summaries[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Karlsruhe ==> ENTITY: \u001b[32mKarlsruher SC\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.211:0.211[0]; log p(e|m)= -2.226:-2.226[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.049]; matches[0.047]; Wosz[0.047]; Bremen[0.045]; played[0.045]; Leverkusen[0.044]; Freiburg[0.044]; Ramelow[0.043]; Bayer[0.043]; Munich[0.043]; Werder[0.043]; German[0.043]; German[0.043]; Bochum[0.043]; Hansa[0.043]; Schalke[0.042]; Halftime[0.042]; Halftime[0.042]; Halftime[0.042]; Halftime[0.042]; Bonn[0.042]; division[0.041]; Rostock[0.041]; \t\n\r [======================>.......................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s333ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2323/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1297testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.004:0.004[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.061]; Royal[0.050]; respectively[0.050]; Clipper[0.049]; Peak[0.049]; Point[0.048]; Coe[0.048]; Newcastle[0.048]; Coal[0.048]; China[0.048]; China[0.048]; Lantau[0.048]; Steel[0.047]; Steel[0.047]; tones[0.046]; fixtures[0.046]; Hay[0.045]; tonnes[0.043]; tonnes[0.043]; Kaohsiung[0.043]; Kaohsiung[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Royal Clipper\u001b[39m ==> ENTITY: \u001b[32mRoyal Clipper\u001b[39m\t\nSCORES: global= 0.205:0.205[0]; local()= 0.070:0.070[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrespectively[0.045]; fixtures[0.045]; Lantau[0.045]; Steel[0.044]; Steel[0.044]; tonnes[0.044]; tonnes[0.044]; coal[0.044]; coal[0.044]; coal[0.044]; Peak[0.044]; Fos[0.044]; Dampier[0.044]; Coal[0.044]; Newcastle[0.043]; China[0.043]; China[0.043]; ore[0.042]; Ore[0.042]; Point[0.042]; Coe[0.042]; tones[0.041]; Hay[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Maracaibo\u001b[39m ==> ENTITY: \u001b[32mMaracaibo\u001b[39m\t\nSCORES: global= 0.229:0.229[0]; local()= -0.048:-0.048[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKaohsiung[0.044]; Kaohsiung[0.044]; ore[0.043]; Peak[0.043]; tonnes[0.043]; tonnes[0.043]; Clipper[0.043]; respectively[0.043]; Dampier[0.043]; fixtures[0.042]; Point[0.042]; Lantau[0.042]; Ore[0.042]; Fos[0.041]; Royal[0.041]; China[0.041]; China[0.041]; Newcastle[0.041]; coal[0.040]; coal[0.040]; coal[0.040]; Clerici[0.040]; Hay[0.039]; tones[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kaohsiung\u001b[39m ==> ENTITY: \u001b[32mKaohsiung\u001b[39m\t\nSCORES: global= 0.237:0.237[0]; local()= 0.031:0.031[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKaohsiung[0.049]; China[0.047]; China[0.047]; Lantau[0.045]; Clipper[0.044]; respectively[0.042]; Point[0.042]; tonnes[0.042]; tonnes[0.042]; Peak[0.041]; Fos[0.041]; Steel[0.041]; Steel[0.041]; Newcastle[0.041]; Maracaibo[0.041]; Dampier[0.040]; fixtures[0.040]; Royal[0.040]; ore[0.040]; Ore[0.039]; Coe[0.039]; coal[0.039]; coal[0.039]; coal[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mFos\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFos, Hérault\u001b[39m <---> \u001b[32mFos-sur-Mer\u001b[39m\t\nSCORES: global= 0.223:0.210[\u001b[31m0.013\u001b[39m]; local()= 0.031:0.093[\u001b[32m0.062\u001b[39m]; log p(e|m)= -0.155:-2.263[\u001b[31m2.108\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMaracaibo[0.045]; tonnes[0.043]; tonnes[0.043]; Kaohsiung[0.043]; Kaohsiung[0.043]; ore[0.043]; China[0.042]; China[0.042]; London[0.042]; London[0.042]; coal[0.042]; coal[0.042]; coal[0.042]; Ore[0.042]; respectively[0.042]; Newcastle[0.041]; Lantau[0.041]; Coal[0.041]; tones[0.040]; Peak[0.040]; Point[0.040]; Steel[0.040]; Steel[0.040]; Clipper[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.004:0.004[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.061]; Royal[0.050]; respectively[0.050]; Clipper[0.049]; Peak[0.049]; Point[0.048]; Coe[0.048]; Newcastle[0.048]; Coal[0.048]; China[0.048]; China[0.048]; Lantau[0.048]; Steel[0.047]; Steel[0.047]; tones[0.046]; fixtures[0.046]; Hay[0.045]; tonnes[0.043]; tonnes[0.043]; Kaohsiung[0.043]; Kaohsiung[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hay Point\u001b[39m ==> ENTITY: \u001b[32mHay Point, Queensland\u001b[39m\t\nSCORES: global= 0.210:0.210[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntonnes[0.047]; tonnes[0.047]; coal[0.046]; coal[0.046]; coal[0.046]; Coal[0.045]; Steel[0.044]; Steel[0.044]; Clipper[0.042]; Dampier[0.041]; respectively[0.041]; Peak[0.041]; Lantau[0.041]; Newcastle[0.040]; Fos[0.040]; ore[0.040]; fixtures[0.039]; China[0.039]; China[0.039]; Ore[0.039]; Maracaibo[0.039]; Royal[0.038]; tones[0.038]; Coe[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dampier ==> ENTITY: \u001b[32mDampier, Western Australia\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.092:0.092[0]; log p(e|m)= -0.837:-0.837[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHay[0.048]; Clipper[0.047]; Point[0.047]; ore[0.046]; coal[0.045]; coal[0.045]; coal[0.045]; tonnes[0.045]; tonnes[0.045]; Peak[0.045]; Lantau[0.044]; Royal[0.044]; Coal[0.043]; China[0.043]; China[0.043]; Ore[0.042]; Maracaibo[0.041]; London[0.041]; London[0.041]; Coe[0.041]; Steel[0.041]; Steel[0.041]; Newcastle[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = China Steel\u001b[39m ==> ENTITY: \u001b[32mChina Steel\u001b[39m\t\nSCORES: global= 0.218:0.218[0]; local()= 0.046:0.046[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.049]; Steel[0.046]; Kaohsiung[0.044]; Kaohsiung[0.044]; Newcastle[0.044]; respectively[0.042]; London[0.042]; London[0.042]; Lantau[0.042]; tonnes[0.042]; tonnes[0.042]; Peak[0.041]; Fos[0.041]; fixtures[0.041]; Coal[0.041]; Clipper[0.040]; Coe[0.040]; Maracaibo[0.040]; Royal[0.040]; ore[0.040]; Ore[0.039]; coal[0.039]; coal[0.039]; coal[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lantau Peak\u001b[39m ==> ENTITY: \u001b[32mLantau Peak\u001b[39m\t\nSCORES: global= 0.202:0.202[0]; local()= 0.021:0.021[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrespectively[0.048]; Kaohsiung[0.047]; Kaohsiung[0.047]; tonnes[0.047]; tonnes[0.047]; Steel[0.047]; Steel[0.047]; Point[0.047]; China[0.046]; China[0.046]; Fos[0.046]; Newcastle[0.046]; fixtures[0.045]; Royal[0.044]; ore[0.044]; Ore[0.044]; Coal[0.044]; Clipper[0.044]; Dampier[0.044]; Coe[0.044]; London[0.044]; London[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kaohsiung\u001b[39m ==> ENTITY: \u001b[32mKaohsiung\u001b[39m\t\nSCORES: global= 0.237:0.237[0]; local()= 0.031:0.031[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKaohsiung[0.049]; China[0.047]; China[0.047]; Lantau[0.045]; Clipper[0.044]; respectively[0.042]; Point[0.042]; tonnes[0.042]; tonnes[0.042]; Peak[0.041]; Fos[0.041]; Steel[0.041]; Steel[0.041]; Newcastle[0.041]; Maracaibo[0.041]; Dampier[0.040]; fixtures[0.040]; Royal[0.040]; ore[0.040]; Ore[0.039]; Coe[0.039]; coal[0.039]; coal[0.039]; coal[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNewcastle\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNewcastle, New South Wales\u001b[39m <---> \u001b[32mNewcastle upon Tyne\u001b[39m\t\nSCORES: global= 0.246:0.241[\u001b[31m0.005\u001b[39m]; local()= 0.062:0.027[\u001b[31m0.035\u001b[39m]; log p(e|m)= -1.406:-1.309[\u001b[32m0.097\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.047]; London[0.047]; fixtures[0.046]; Coal[0.044]; Royal[0.043]; Point[0.043]; Clipper[0.043]; Coe[0.042]; Hay[0.042]; Peak[0.042]; Steel[0.041]; Steel[0.041]; coal[0.041]; coal[0.041]; coal[0.041]; respectively[0.041]; Dampier[0.040]; ore[0.040]; China[0.039]; China[0.039]; Ore[0.039]; Lantau[0.039]; Kaohsiung[0.039]; Kaohsiung[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = China Steel\u001b[39m ==> ENTITY: \u001b[32mChina Steel\u001b[39m\t\nSCORES: global= 0.218:0.218[0]; local()= 0.046:0.046[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.049]; Steel[0.046]; Kaohsiung[0.044]; Kaohsiung[0.044]; Newcastle[0.044]; respectively[0.042]; London[0.042]; London[0.042]; Lantau[0.042]; tonnes[0.042]; tonnes[0.042]; Peak[0.041]; Fos[0.041]; fixtures[0.041]; Coal[0.041]; Clipper[0.040]; Coe[0.040]; Maracaibo[0.040]; Royal[0.040]; ore[0.040]; Ore[0.039]; coal[0.039]; coal[0.039]; coal[0.039]; \t\n\r [======================>.......................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s349ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2336/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1267testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rangoon\u001b[39m ==> ENTITY: \u001b[32mYangon\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRangoon[0.044]; University[0.044]; University[0.044]; Yangon[0.043]; Yangon[0.043]; Yangon[0.043]; Burmese[0.043]; Burmese[0.043]; northern[0.042]; institute[0.041]; institute[0.041]; institute[0.041]; march[0.041]; marched[0.041]; closed[0.041]; engines[0.041]; students[0.041]; students[0.041]; standby[0.040]; Institute[0.040]; campus[0.040]; campus[0.040]; campus[0.040]; roads[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = University of Yangon\u001b[39m ==> ENTITY: \u001b[32mUniversity of Yangon\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.241:0.241[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nuniversity[0.048]; university[0.048]; University[0.048]; Rangoon[0.047]; Rangoon[0.047]; gathered[0.046]; Yangon[0.046]; Yangon[0.046]; institute[0.045]; institute[0.045]; institute[0.045]; institute[0.045]; march[0.045]; marched[0.045]; students[0.044]; students[0.044]; students[0.044]; students[0.044]; protests[0.044]; protests[0.044]; military[0.044]; Institute[0.044]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAung San\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAung San Suu Kyi\u001b[39m <---> \u001b[32mAung San\u001b[39m\t\nSCORES: global= 0.289:0.236[\u001b[31m0.053\u001b[39m]; local()= 0.161:0.157[\u001b[31m0.003\u001b[39m]; log p(e|m)= 0.000:-6.908[\u001b[31m6.908\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlaureate[0.053]; military[0.051]; Nobel[0.051]; government[0.051]; government[0.051]; government[0.051]; government[0.051]; Democracy[0.051]; political[0.051]; independence[0.050]; National[0.050]; Kyi[0.050]; Kyi[0.050]; Kyi[0.050]; protests[0.049]; uprising[0.049]; Suu[0.048]; Suu[0.048]; Suu[0.048]; State[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NLD ==> ENTITY: \u001b[32mNational League for Democracy\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.170:0.170[0]; log p(e|m)= -1.217:-1.217[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDemocracy[0.049]; political[0.047]; week[0.047]; National[0.047]; protests[0.046]; protests[0.046]; Kyi[0.046]; Kyi[0.046]; nearly[0.046]; held[0.045]; military[0.045]; government[0.045]; government[0.045]; government[0.045]; government[0.045]; dissatisfied[0.045]; parties[0.044]; October[0.044]; published[0.044]; Aung[0.044]; Suu[0.044]; Suu[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Aung San Suu Kyi\u001b[39m ==> ENTITY: \u001b[32mAung San Suu Kyi\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.112:0.112[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.044]; government[0.043]; university[0.043]; university[0.043]; imprisoned[0.043]; junta[0.042]; leader[0.042]; Rangoon[0.042]; opposition[0.042]; University[0.042]; protests[0.042]; protests[0.042]; protests[0.042]; protests[0.042]; uprising[0.041]; held[0.041]; students[0.041]; students[0.041]; students[0.041]; students[0.041]; Yangon[0.041]; campus[0.040]; campus[0.040]; protest[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = University of Yangon\u001b[39m ==> ENTITY: \u001b[32mUniversity of Yangon\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.233:0.233[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nuniversity[0.045]; University[0.044]; Rangoon[0.044]; Rangoon[0.044]; gathered[0.042]; Yangon[0.042]; Yangon[0.042]; institute[0.042]; institute[0.042]; institute[0.042]; march[0.042]; marched[0.041]; students[0.041]; students[0.041]; students[0.041]; students[0.041]; Institute[0.041]; campus[0.041]; campus[0.041]; campus[0.041]; campus[0.041]; Monday[0.040]; Burmese[0.040]; Burmese[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yangon\u001b[39m ==> ENTITY: \u001b[32mYangon\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.126:0.126[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRangoon[0.048]; government[0.046]; government[0.046]; central[0.045]; nearly[0.045]; closed[0.045]; October[0.044]; road[0.044]; remained[0.043]; key[0.043]; demonstration[0.043]; lakeside[0.043]; officials[0.043]; institute[0.042]; institute[0.042]; institute[0.042]; near[0.042]; involved[0.042]; Aung[0.042]; shut[0.042]; National[0.042]; residence[0.042]; uprising[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Suu Kyi\u001b[39m ==> ENTITY: \u001b[32mAung San Suu Kyi\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlaureate[0.044]; military[0.043]; Nobel[0.043]; government[0.043]; government[0.043]; government[0.043]; government[0.043]; Democracy[0.043]; political[0.043]; independence[0.042]; National[0.042]; Kyi[0.042]; protests[0.041]; protests[0.041]; Aung[0.041]; Suu[0.041]; held[0.041]; State[0.040]; students[0.040]; students[0.040]; students[0.040]; students[0.040]; students[0.040]; students[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rangoon\u001b[39m ==> ENTITY: \u001b[32mYangon\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.125:0.125[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nuniversity[0.045]; military[0.044]; military[0.044]; institute[0.042]; demonstrations[0.042]; student[0.042]; students[0.041]; students[0.041]; students[0.041]; capital[0.041]; protests[0.041]; protests[0.041]; protests[0.041]; protests[0.041]; junta[0.041]; central[0.041]; government[0.041]; government[0.041]; government[0.041]; Monday[0.041]; protest[0.041]; detained[0.041]; campus[0.041]; campus[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burmese ==> ENTITY: \u001b[32mBurma\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.102:0.102[0]; log p(e|m)= -1.228:-1.228[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurmese[0.048]; northern[0.048]; Rangoon[0.043]; Rangoon[0.043]; returned[0.043]; Yangon[0.042]; Yangon[0.042]; students[0.042]; students[0.042]; said[0.041]; declined[0.041]; troubled[0.041]; official[0.041]; institute[0.041]; institute[0.041]; march[0.041]; leading[0.040]; clashes[0.040]; marched[0.040]; near[0.040]; campus[0.040]; campus[0.040]; campus[0.040]; armed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yangon Institute of Technology\u001b[39m ==> ENTITY: \u001b[32mYangon Technological University\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.209:0.209[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nuniversity[0.046]; University[0.045]; University[0.045]; Rangoon[0.044]; Rangoon[0.044]; campus[0.043]; campus[0.043]; campus[0.043]; campus[0.043]; Yangon[0.042]; Yangon[0.042]; students[0.042]; students[0.042]; students[0.042]; institute[0.041]; institute[0.041]; institute[0.041]; closed[0.040]; outside[0.039]; leading[0.039]; Seven[0.039]; Yit[0.039]; gathered[0.038]; northern[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nobel ==> ENTITY: \u001b[32mNobel Peace Prize\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.126:0.126[0]; log p(e|m)= -2.674:-2.674[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlaureate[0.047]; political[0.043]; National[0.043]; published[0.043]; Council[0.042]; told[0.042]; Democracy[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; Reuters[0.041]; institute[0.041]; institute[0.041]; institute[0.041]; protests[0.041]; protests[0.041]; demanded[0.041]; government[0.041]; government[0.041]; government[0.041]; government[0.041]; key[0.041]; newspapers[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npublished[0.045]; newspapers[0.044]; told[0.044]; political[0.044]; week[0.042]; government[0.042]; government[0.042]; government[0.042]; denied[0.042]; said[0.042]; said[0.042]; said[0.042]; protest[0.041]; Thursday[0.041]; key[0.041]; independent[0.041]; National[0.041]; Tuesday[0.041]; military[0.041]; military[0.041]; killed[0.041]; September[0.041]; Democracy[0.040]; central[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.086:0.086[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nroads[0.044]; told[0.044]; leading[0.043]; authorities[0.043]; northern[0.042]; official[0.042]; truckloads[0.042]; launched[0.042]; government[0.042]; said[0.042]; said[0.042]; said[0.042]; Technology[0.041]; Wednesday[0.041]; Monday[0.041]; institute[0.041]; institute[0.041]; institute[0.041]; institute[0.041]; Tuesday[0.041]; Tuesday[0.041]; military[0.041]; standby[0.041]; miles[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rangoon\u001b[39m ==> ENTITY: \u001b[32mYangon\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.126:0.126[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYangon[0.045]; government[0.044]; government[0.044]; nearly[0.043]; institute[0.043]; institute[0.043]; institute[0.043]; central[0.043]; closed[0.042]; students[0.042]; students[0.042]; students[0.042]; students[0.042]; road[0.042]; remained[0.041]; demonstration[0.041]; uprising[0.041]; Friday[0.041]; near[0.040]; October[0.039]; leading[0.039]; residence[0.039]; shut[0.039]; latest[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NLD ==> ENTITY: \u001b[32mNational League for Democracy\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.169:0.169[0]; log p(e|m)= -1.217:-1.217[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDemocracy[0.049]; political[0.047]; National[0.047]; protests[0.047]; Kyi[0.047]; Kyi[0.047]; Kyi[0.047]; nearly[0.046]; military[0.045]; government[0.045]; government[0.045]; government[0.045]; government[0.045]; uprising[0.045]; dissatisfied[0.045]; parties[0.044]; October[0.044]; published[0.044]; Aung[0.044]; Suu[0.044]; Suu[0.044]; Suu[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burmese ==> ENTITY: \u001b[32mBurma\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.101:0.101[0]; log p(e|m)= -1.228:-1.228[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurmese[0.048]; northern[0.047]; Rangoon[0.043]; Rangoon[0.043]; returned[0.043]; Yangon[0.042]; Yangon[0.042]; Yangon[0.042]; students[0.041]; students[0.041]; students[0.041]; closed[0.041]; university[0.041]; said[0.041]; declined[0.041]; troubled[0.041]; official[0.041]; institute[0.040]; institute[0.040]; institute[0.040]; march[0.040]; authorities[0.040]; leading[0.040]; clashes[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Suu Kyi\u001b[39m ==> ENTITY: \u001b[32mAung San Suu Kyi\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlaureate[0.053]; military[0.052]; Nobel[0.051]; government[0.051]; government[0.051]; government[0.051]; government[0.051]; Democracy[0.051]; political[0.051]; independence[0.050]; National[0.050]; Kyi[0.050]; imprisoned[0.050]; protests[0.049]; protests[0.049]; uprising[0.049]; Aung[0.049]; Suu[0.049]; held[0.048]; State[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rangoon\u001b[39m ==> ENTITY: \u001b[32mYangon\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nuniversity[0.046]; Rangoon[0.046]; University[0.045]; University[0.045]; Yangon[0.045]; Yangon[0.045]; Yangon[0.045]; Burmese[0.044]; Burmese[0.044]; northern[0.044]; gathered[0.043]; institute[0.043]; institute[0.043]; institute[0.043]; march[0.043]; marched[0.042]; closed[0.042]; engines[0.042]; students[0.042]; students[0.042]; students[0.042]; standby[0.042]; Institute[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Suu Kyi\u001b[39m ==> ENTITY: \u001b[32mAung San Suu Kyi\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.082:0.082[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngovernment[0.045]; government[0.045]; Rangoon[0.043]; uprising[0.043]; Aung[0.043]; students[0.042]; students[0.042]; students[0.042]; students[0.042]; Yangon[0.042]; denied[0.041]; said[0.041]; said[0.041]; parties[0.041]; institute[0.041]; institute[0.041]; institute[0.041]; central[0.041]; nearly[0.041]; close[0.041]; October[0.041]; officials[0.040]; lakeside[0.040]; residence[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = National League for Democracy\u001b[39m ==> ENTITY: \u001b[32mNational League for Democracy\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nheld[0.050]; political[0.047]; protests[0.047]; protests[0.047]; Kyi[0.047]; Kyi[0.047]; parties[0.046]; October[0.046]; Friday[0.045]; military[0.045]; government[0.045]; government[0.045]; government[0.045]; government[0.045]; imprisoned[0.045]; Aung[0.044]; week[0.044]; Suu[0.044]; Suu[0.044]; laureate[0.044]; denied[0.044]; State[0.044]; \t\n\r [=======================>......................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s162ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2357/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1210testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bombay ==> ENTITY: \u001b[32mMumbai\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.076:0.076[0]; log p(e|m)= -0.242:-0.242[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.052]; tournament[0.045]; England[0.044]; Squash[0.043]; International[0.043]; number[0.043]; fifth[0.042]; praise[0.042]; title[0.041]; World[0.041]; World[0.041]; final[0.041]; moved[0.041]; Mahindra[0.041]; outside[0.041]; Scotland[0.041]; circuit[0.040]; squash[0.040]; battled[0.040]; return[0.040]; overcame[0.040]; richest[0.040]; Parke[0.039]; brushed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mark Cairns ==> ENTITY: \u001b[32mMark Cairns (squash player)\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.183:-0.183[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayer[0.051]; Squash[0.048]; World[0.048]; World[0.048]; tournament[0.047]; title[0.046]; Englishman[0.046]; Scotland[0.046]; England[0.046]; Simon[0.046]; squash[0.045]; unseeded[0.045]; meets[0.045]; Peter[0.045]; Australian[0.044]; outside[0.044]; International[0.044]; India[0.044]; circuit[0.044]; Bombay[0.043]; fifth[0.043]; Parke[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nicol\u001b[39m ==> ENTITY: \u001b[32mPeter Nicol\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.046]; Squash[0.045]; squash[0.044]; player[0.044]; title[0.044]; Open[0.043]; Scotland[0.043]; World[0.042]; World[0.042]; Nicol[0.042]; Nicol[0.042]; unseeded[0.041]; England[0.041]; number[0.041]; fifth[0.041]; final[0.041]; titleholder[0.041]; Englishman[0.040]; Simon[0.040]; Mark[0.040]; praise[0.040]; overcame[0.040]; Peter[0.040]; Sight[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Eyles\u001b[39m ==> ENTITY: \u001b[32mRodney Eyles\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralian[0.045]; World[0.044]; World[0.044]; Squash[0.044]; player[0.044]; Open[0.043]; tournament[0.043]; final[0.043]; title[0.042]; Rodney[0.041]; India[0.041]; unseeded[0.041]; Nicol[0.041]; Nicol[0.041]; Nicol[0.041]; fifth[0.041]; squash[0.040]; Eyles[0.040]; Eyles[0.040]; Peter[0.040]; International[0.040]; England[0.040]; Title[0.040]; Mark[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australian ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.491:-0.491[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.051]; tournament[0.050]; Open[0.050]; India[0.049]; player[0.047]; Scotland[0.046]; Peter[0.046]; Mark[0.046]; final[0.045]; fifth[0.045]; number[0.045]; title[0.045]; Squash[0.045]; Simon[0.044]; squash[0.044]; unseeded[0.044]; Cairns[0.043]; titleholder[0.043]; circuit[0.043]; Englishman[0.043]; World[0.043]; World[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nicol\u001b[39m ==> ENTITY: \u001b[32mPeter Nicol\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.046]; squash[0.045]; player[0.044]; title[0.044]; Open[0.043]; Scotland[0.043]; World[0.043]; Nicol[0.042]; Nicol[0.042]; unseeded[0.042]; England[0.042]; fifth[0.041]; final[0.041]; titleholder[0.041]; Englishman[0.041]; Simon[0.041]; Mark[0.040]; praise[0.040]; overcame[0.040]; Peter[0.040]; International[0.040]; richest[0.039]; return[0.039]; seeded[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Peter Nicol\u001b[39m ==> ENTITY: \u001b[32mPeter Nicol\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.046]; Squash[0.045]; squash[0.044]; player[0.044]; title[0.044]; Open[0.043]; Scotland[0.043]; World[0.042]; World[0.042]; Nicol[0.042]; Nicol[0.042]; unseeded[0.041]; England[0.041]; number[0.041]; fifth[0.041]; final[0.041]; titleholder[0.041]; Englishman[0.040]; Simon[0.040]; Mark[0.040]; praise[0.040]; overcame[0.040]; Sight[0.039]; International[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Eyles\u001b[39m ==> ENTITY: \u001b[32mRodney Eyles\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralian[0.046]; World[0.045]; World[0.045]; Squash[0.044]; Open[0.043]; tournament[0.043]; final[0.043]; title[0.042]; Rodney[0.042]; India[0.042]; unseeded[0.041]; Nicol[0.041]; Nicol[0.041]; fifth[0.041]; squash[0.041]; Eyles[0.040]; Eyles[0.040]; Peter[0.040]; International[0.040]; England[0.040]; Title[0.040]; Mark[0.040]; Cairns[0.040]; titleholder[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rodney Eyles\u001b[39m ==> ENTITY: \u001b[32mRodney Eyles\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralian[0.046]; World[0.045]; World[0.045]; Squash[0.044]; player[0.044]; Open[0.043]; tournament[0.043]; final[0.043]; title[0.042]; India[0.041]; unseeded[0.041]; Nicol[0.041]; Nicol[0.041]; Nicol[0.041]; fifth[0.041]; squash[0.041]; Eyles[0.040]; Eyles[0.040]; Peter[0.040]; International[0.040]; England[0.040]; Title[0.040]; Mark[0.040]; Cairns[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scotland ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.075:0.075[0]; log p(e|m)= -0.229:-0.229[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.049]; player[0.045]; tournament[0.044]; India[0.044]; number[0.044]; final[0.043]; fifth[0.043]; return[0.042]; World[0.041]; World[0.041]; Englishman[0.041]; Squash[0.041]; outside[0.041]; International[0.041]; Nicol[0.041]; Nicol[0.041]; Nicol[0.041]; Australian[0.040]; overcame[0.040]; Mark[0.040]; Cairns[0.040]; title[0.040]; Peter[0.040]; minutes[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEnglishman\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUnited Kingdom\u001b[39m <---> \u001b[32mEnglish people\u001b[39m\t\nSCORES: global= 0.248:0.242[\u001b[31m0.006\u001b[39m]; local()= 0.070:0.103[\u001b[32m0.034\u001b[39m]; log p(e|m)= -1.858:-0.241[\u001b[32m1.617\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.049]; India[0.045]; Scotland[0.044]; Simon[0.043]; Peter[0.043]; Mark[0.043]; outside[0.042]; Australian[0.042]; fifth[0.042]; tournament[0.041]; player[0.041]; Cairns[0.041]; title[0.041]; number[0.041]; Eyles[0.040]; Eyles[0.040]; Eyles[0.040]; final[0.040]; Nicol[0.040]; Nicol[0.040]; Nicol[0.040]; Parke[0.040]; return[0.040]; Squash[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = India ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.042:0.042[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.047]; England[0.047]; final[0.044]; fifth[0.044]; Bombay[0.044]; Mahindra[0.042]; moved[0.042]; World[0.042]; World[0.042]; Cairns[0.041]; overcame[0.041]; Open[0.041]; title[0.041]; return[0.041]; unseeded[0.041]; International[0.041]; Scotland[0.041]; Australian[0.041]; number[0.040]; seeded[0.040]; battled[0.040]; Eyles[0.040]; Eyles[0.040]; Eyles[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Simon Parke\u001b[39m ==> ENTITY: \u001b[32mSimon Parke\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.046]; Squash[0.044]; Open[0.043]; fifth[0.043]; circuit[0.043]; title[0.043]; World[0.043]; World[0.043]; tournament[0.042]; Englishman[0.042]; unseeded[0.042]; return[0.042]; cancer[0.042]; testicular[0.041]; squash[0.041]; final[0.041]; player[0.040]; titleholder[0.040]; Scotland[0.040]; overcame[0.040]; Sight[0.040]; seeded[0.039]; Cairns[0.039]; International[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Open ==> ENTITY: \u001b[32mWorld Open (squash)\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.175:0.175[0]; log p(e|m)= -1.625:-1.625[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfinal[0.045]; tournament[0.045]; World[0.044]; fifth[0.044]; Australian[0.043]; Simon[0.043]; Squash[0.043]; Mark[0.043]; player[0.042]; title[0.042]; squash[0.042]; number[0.042]; Friday[0.041]; England[0.041]; Scotland[0.041]; Peter[0.040]; titleholder[0.040]; Nicol[0.040]; Nicol[0.040]; Nicol[0.040]; Englishman[0.040]; India[0.040]; richest[0.040]; circuit[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.239:0.239[0]; local()= 0.062:0.062[0]; log p(e|m)= -0.190:-0.190[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.046]; India[0.045]; player[0.044]; tournament[0.043]; fifth[0.043]; Australian[0.043]; Mark[0.042]; final[0.042]; Cairns[0.041]; Englishman[0.041]; return[0.041]; number[0.041]; Squash[0.041]; World[0.041]; World[0.041]; Eyles[0.041]; Eyles[0.041]; Eyles[0.041]; Simon[0.041]; Peter[0.041]; overcame[0.040]; Parke[0.040]; Friday[0.040]; Rodney[0.040]; \t\n\r [=======================>......................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s148ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2372/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1293testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brussels ==> ENTITY: \u001b[32mBrussels\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.076:0.076[0]; log p(e|m)= -0.025:-0.025[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuropean[0.048]; European[0.048]; Europe[0.046]; Commission[0.044]; Poland[0.043]; Poland[0.043]; Poland[0.043]; Poland[0.043]; regarding[0.043]; related[0.042]; contravened[0.041]; market[0.041]; Friday[0.041]; January[0.040]; import[0.040]; import[0.040]; local[0.040]; called[0.040]; said[0.039]; reached[0.039]; resolved[0.039]; imports[0.039]; Agreement[0.039]; products[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Europe ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.095:0.095[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsector[0.047]; European[0.047]; European[0.047]; Commission[0.042]; Commission[0.042]; market[0.042]; Poland[0.042]; Poland[0.042]; Poland[0.042]; Poland[0.042]; Poland[0.042]; industry[0.041]; Brussels[0.041]; Agreement[0.041]; January[0.040]; January[0.040]; Union[0.040]; import[0.040]; import[0.040]; import[0.040]; import[0.040]; envisaged[0.040]; imports[0.040]; imports[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.157:0.157[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.049]; Poland[0.049]; Poland[0.049]; Polish[0.045]; Polish[0.045]; Polish[0.045]; Polish[0.045]; Europe[0.042]; European[0.041]; Friday[0.041]; final[0.040]; regime[0.040]; regarding[0.040]; said[0.040]; said[0.040]; objected[0.040]; January[0.040]; January[0.040]; restrictions[0.039]; saying[0.039]; protecting[0.039]; obliged[0.038]; statement[0.038]; related[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.184:0.184[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.048]; Poland[0.048]; Poland[0.048]; Poland[0.048]; Polish[0.044]; Polish[0.044]; Polish[0.044]; Polish[0.044]; Polish[0.044]; Europe[0.041]; European[0.040]; European[0.040]; Friday[0.040]; final[0.040]; regime[0.039]; regime[0.039]; regarding[0.039]; Union[0.039]; said[0.039]; said[0.039]; objected[0.039]; January[0.039]; January[0.039]; restrictions[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = European Commission\u001b[39m ==> ENTITY: \u001b[32mEuropean Commission\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuropean[0.048]; Commission[0.045]; Europe[0.043]; related[0.043]; Brussels[0.043]; Agreement[0.042]; Poland[0.042]; Poland[0.042]; Poland[0.042]; Poland[0.042]; Poland[0.042]; statement[0.042]; regarding[0.041]; agreement[0.041]; January[0.041]; Polish[0.040]; Polish[0.040]; Polish[0.040]; Polish[0.040]; envisaged[0.040]; said[0.040]; said[0.040]; objected[0.040]; issues[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Polish ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.140:0.140[0]; log p(e|m)= -0.814:-0.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.049]; Poland[0.049]; Poland[0.049]; Poland[0.049]; Polish[0.048]; Polish[0.048]; Europe[0.043]; local[0.042]; regime[0.041]; early[0.040]; January[0.040]; related[0.039]; issues[0.039]; meetings[0.039]; restrictions[0.039]; market[0.038]; said[0.038]; settlement[0.038]; settlement[0.038]; abolish[0.038]; agreement[0.038]; agreement[0.038]; equal[0.038]; Commission[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Polish ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.204:0.204[0]; log p(e|m)= -0.814:-0.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.047]; Poland[0.047]; Poland[0.047]; Poland[0.047]; Poland[0.047]; Poland[0.047]; Polish[0.045]; Polish[0.045]; Polish[0.045]; Polish[0.045]; Europe[0.041]; European[0.039]; European[0.039]; regime[0.039]; regime[0.039]; Union[0.038]; early[0.038]; January[0.038]; January[0.038]; restrictions[0.038]; objected[0.038]; said[0.038]; said[0.038]; Friday[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.141:0.141[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.049]; Poland[0.049]; Poland[0.049]; Polish[0.046]; Polish[0.046]; Polish[0.046]; Europe[0.043]; European[0.042]; European[0.042]; Friday[0.041]; regime[0.041]; Union[0.040]; said[0.040]; objected[0.040]; January[0.040]; saying[0.040]; protecting[0.039]; import[0.039]; import[0.039]; market[0.038]; reached[0.038]; introduced[0.038]; disagreements[0.038]; local[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.157:0.157[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.048]; Poland[0.048]; Poland[0.048]; Polish[0.045]; Polish[0.045]; Polish[0.045]; Polish[0.045]; Europe[0.042]; European[0.041]; European[0.041]; Friday[0.041]; final[0.040]; regime[0.040]; regime[0.040]; regarding[0.040]; Union[0.040]; said[0.040]; objected[0.040]; January[0.039]; saying[0.039]; protecting[0.038]; related[0.038]; settlement[0.038]; issues[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.050]; Poland[0.050]; Polish[0.046]; Polish[0.046]; Polish[0.046]; final[0.042]; regime[0.041]; regarding[0.041]; said[0.041]; January[0.041]; restrictions[0.041]; protecting[0.040]; obliged[0.040]; statement[0.040]; related[0.040]; settlement[0.040]; settlement[0.040]; issues[0.039]; abolish[0.039]; import[0.039]; import[0.039]; includes[0.039]; market[0.039]; reached[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.198:0.198[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.047]; Poland[0.047]; Poland[0.047]; Poland[0.047]; Poland[0.047]; Polish[0.044]; Polish[0.044]; Polish[0.044]; Polish[0.044]; Polish[0.044]; Europe[0.041]; European[0.040]; European[0.040]; Friday[0.040]; final[0.039]; regime[0.039]; regime[0.039]; regarding[0.039]; Union[0.039]; said[0.039]; said[0.039]; objected[0.038]; January[0.038]; January[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Polish ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.187:0.187[0]; log p(e|m)= -0.814:-0.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.048]; Poland[0.048]; Poland[0.048]; Poland[0.048]; Poland[0.048]; Polish[0.046]; Polish[0.046]; Polish[0.046]; Europe[0.042]; European[0.040]; European[0.040]; regime[0.040]; regime[0.040]; Union[0.039]; January[0.039]; objected[0.039]; said[0.039]; said[0.039]; Friday[0.038]; related[0.038]; Commission[0.038]; Commission[0.038]; helping[0.038]; settlement[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = European Union\u001b[39m ==> ENTITY: \u001b[32mEuropean Union\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.173:0.173[0]; log p(e|m)= -0.006:-0.006[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuropean[0.053]; Europe[0.052]; Commission[0.050]; Poland[0.048]; Poland[0.048]; Poland[0.048]; Poland[0.048]; tariffs[0.048]; tariffs[0.048]; saying[0.047]; regarding[0.047]; Agreement[0.047]; said[0.047]; Brussels[0.047]; helping[0.046]; envisaged[0.046]; issues[0.046]; related[0.046]; regime[0.046]; agree[0.046]; disagreements[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Polish ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.192:0.192[0]; log p(e|m)= -0.814:-0.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.047]; Poland[0.047]; Poland[0.047]; Poland[0.047]; Poland[0.047]; Polish[0.046]; Polish[0.046]; Polish[0.046]; Polish[0.046]; Europe[0.042]; European[0.040]; European[0.040]; regime[0.039]; regime[0.039]; Union[0.039]; early[0.039]; January[0.038]; January[0.038]; restrictions[0.038]; objected[0.038]; said[0.038]; said[0.038]; Friday[0.037]; related[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Polish ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.172:0.172[0]; log p(e|m)= -0.814:-0.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.048]; Poland[0.048]; Poland[0.048]; Poland[0.048]; Polish[0.047]; Polish[0.047]; Polish[0.047]; Europe[0.043]; European[0.041]; European[0.041]; regime[0.040]; regime[0.040]; Union[0.040]; January[0.039]; objected[0.039]; said[0.039]; Friday[0.038]; related[0.038]; Commission[0.038]; Commission[0.038]; helping[0.038]; settlement[0.038]; market[0.038]; saying[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Commission ==> ENTITY: \u001b[32mEuropean Commission\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.161:0.161[0]; log p(e|m)= -0.761:-0.761[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuropean[0.048]; Commission[0.046]; Europe[0.044]; price[0.043]; local[0.043]; offer[0.043]; statement[0.042]; regarding[0.042]; January[0.041]; January[0.041]; related[0.041]; sector[0.041]; restrictions[0.041]; monthly[0.041]; products[0.041]; market[0.041]; includes[0.041]; meetings[0.041]; industry[0.040]; early[0.040]; agreement[0.040]; agreement[0.040]; said[0.040]; said[0.040]; \t\n\r [=======================>......................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 10s992ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2388/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1208testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Perugia ==> ENTITY: \u001b[32mA.C. Perugia Calcio\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.251:0.251[0]; log p(e|m)= -0.753:-0.753[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReggiana[0.046]; Reggiana[0.046]; league[0.043]; season[0.043]; Rome[0.042]; Napoli[0.042]; Milan[0.042]; Milan[0.042]; Milan[0.042]; Cups[0.041]; Italian[0.041]; Perugia[0.041]; Francesco[0.041]; team[0.041]; Cagliari[0.041]; Cagliari[0.041]; Amoruso[0.041]; Fiorentina[0.041]; Fiorentina[0.041]; Lazio[0.041]; Lazio[0.041]; Paolo[0.040]; players[0.040]; Casiraghi[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ivan Zamorano\u001b[39m ==> ENTITY: \u001b[32mIván Zamorano\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.046]; Inter[0.045]; Inter[0.045]; match[0.045]; Angloma[0.044]; Roberto[0.043]; Uruguayan[0.042]; Internazionale[0.042]; successive[0.042]; captain[0.041]; Chilean[0.041]; Sampdoria[0.041]; Del[0.041]; Mancini[0.041]; champions[0.041]; impressive[0.041]; striker[0.041]; defender[0.040]; midfielder[0.040]; draws[0.040]; world[0.040]; Marcelo[0.040]; nonetheless[0.039]; clash[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Napoli ==> ENTITY: \u001b[32mS.S.C. Napoli\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.264:0.264[0]; log p(e|m)= -0.373:-0.373[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNapoli[0.044]; League[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Fabio[0.043]; season[0.043]; Parma[0.042]; Parma[0.042]; Parma[0.042]; Roberto[0.042]; Roberto[0.042]; Marco[0.042]; Champions[0.041]; Cagliari[0.041]; Udinese[0.041]; Udinese[0.041]; Mario[0.041]; second[0.040]; Baggio[0.040]; win[0.040]; win[0.040]; midweek[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Igor Kolyvanov\u001b[39m ==> ENTITY: \u001b[32mIgor Kolyvanov\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.217:0.217[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.048]; league[0.048]; team[0.046]; games[0.046]; Reggiana[0.045]; Reggiana[0.045]; players[0.045]; Bologna[0.043]; Bologna[0.043]; Bologna[0.043]; Soccer[0.043]; played[0.042]; Dietmar[0.042]; Piacenza[0.042]; Piacenza[0.042]; Milan[0.042]; Vicenza[0.042]; Russian[0.042]; Italian[0.042]; Italian[0.042]; Italian[0.042]; Inter[0.042]; win[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Roma ==> ENTITY: \u001b[32mA.S. Roma\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.249:0.249[0]; log p(e|m)= -1.221:-1.221[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRome[0.047]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Lazio[0.043]; Lazio[0.043]; League[0.042]; Roma[0.041]; Udinese[0.041]; Udinese[0.041]; season[0.041]; Casiraghi[0.041]; midweek[0.041]; Italian[0.041]; Amoruso[0.041]; Roberto[0.040]; players[0.040]; Roman[0.040]; Totti[0.040]; teams[0.040]; win[0.040]; match[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.061:0.061[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrench[0.047]; French[0.047]; Angloma[0.045]; match[0.044]; club[0.043]; Pierre[0.042]; champions[0.042]; Fabio[0.042]; ago[0.041]; Juventus[0.041]; Juventus[0.041]; Chilean[0.041]; captain[0.041]; Inter[0.040]; Inter[0.040]; goals[0.040]; Argentine[0.040]; Portuguese[0.040]; weeks[0.040]; Del[0.040]; defender[0.040]; defender[0.040]; defender[0.040]; Cannavaro[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reggiana\u001b[39m ==> ENTITY: \u001b[32mA.C. Reggiana 1919\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.257:0.257[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; league[0.046]; Reggiana[0.046]; games[0.044]; Soccer[0.044]; Serie[0.043]; Lazio[0.042]; team[0.042]; players[0.041]; Milan[0.041]; Milan[0.041]; played[0.041]; Napoli[0.041]; match[0.041]; match[0.041]; win[0.040]; Fiorentina[0.040]; Fiorentina[0.040]; Piacenza[0.040]; Piacenza[0.040]; Italian[0.040]; Italian[0.040]; Italian[0.040]; Inter[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMontenegrin\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMontenegro national football team\u001b[39m <---> \u001b[32mMontenegro\u001b[39m\t\nSCORES: global= 0.267:0.256[\u001b[31m0.011\u001b[39m]; local()= 0.155:0.063[\u001b[31m0.092\u001b[39m]; log p(e|m)= -2.501:-1.088[\u001b[32m1.413\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.045]; Champions[0.045]; League[0.044]; season[0.043]; win[0.043]; win[0.043]; Norwegian[0.043]; Croat[0.043]; Dejan[0.042]; signing[0.041]; despite[0.041]; welcome[0.041]; led[0.041]; midweek[0.040]; Milan[0.040]; Milan[0.040]; Milan[0.040]; Milan[0.040]; Roma[0.040]; Czech[0.040]; midfielder[0.040]; midfielder[0.040]; midfielder[0.040]; chances[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.053:0.053[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.046]; match[0.046]; players[0.044]; Uefa[0.043]; Italian[0.043]; Italian[0.043]; team[0.042]; Reggiana[0.042]; Reggiana[0.042]; Roma[0.042]; season[0.041]; Casiraghi[0.041]; league[0.041]; win[0.041]; injury[0.041]; Czech[0.041]; Lorenzo[0.041]; midfielder[0.040]; midfielder[0.040]; Fausto[0.040]; defender[0.040]; defender[0.040]; teams[0.040]; Emiliano[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Filippo Inzaghi\u001b[39m ==> ENTITY: \u001b[32mFilippo Inzaghi\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.227:0.227[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.050]; season[0.049]; Parma[0.049]; Parma[0.049]; Parma[0.049]; Fabio[0.049]; club[0.048]; injury[0.048]; injury[0.048]; Juventus[0.048]; Juventus[0.048]; win[0.047]; Alessandro[0.047]; Cannavaro[0.047]; Roberto[0.047]; Roberto[0.047]; Internazionale[0.047]; Sampdoria[0.047]; Sampdoria[0.047]; goals[0.047]; Mario[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Croat ==> ENTITY: \u001b[32mCroatia\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.097:0.097[0]; log p(e|m)= -1.440:-1.440[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRoma[0.044]; Roma[0.044]; Czech[0.044]; Italian[0.043]; league[0.042]; team[0.042]; home[0.042]; German[0.042]; Rome[0.042]; Roman[0.042]; Marco[0.042]; Ramon[0.041]; players[0.041]; difficult[0.041]; Argentine[0.041]; teams[0.041]; Sunday[0.041]; following[0.041]; start[0.041]; Uefa[0.040]; Swiss[0.040]; match[0.040]; match[0.040]; Emiliano[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Parma ==> ENTITY: \u001b[32mParma F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.242:0.242[0]; log p(e|m)= -1.085:-1.085[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nParma[0.043]; Parma[0.043]; Juventus[0.043]; Juventus[0.043]; Fabio[0.043]; Milan[0.043]; Milan[0.043]; Filippo[0.042]; Alessandro[0.042]; Napoli[0.042]; Napoli[0.042]; Roberto[0.041]; Roberto[0.041]; season[0.041]; signing[0.041]; signing[0.041]; Sampdoria[0.041]; Piero[0.041]; Verona[0.041]; Verona[0.041]; Verona[0.041]; Udinese[0.041]; Cagliari[0.040]; Eugenio[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mArgentine\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mArgentina national football team\u001b[39m <---> \u001b[32mArgentina\u001b[39m\t\nSCORES: global= 0.256:0.253[\u001b[31m0.002\u001b[39m]; local()= 0.181:0.104[\u001b[31m0.077\u001b[39m]; log p(e|m)= -3.442:-0.189[\u001b[32m3.253\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRoberto[0.055]; League[0.054]; match[0.054]; Marco[0.053]; Champions[0.052]; season[0.051]; teams[0.051]; Emiliano[0.050]; Uefa[0.050]; Italian[0.049]; Fausto[0.049]; midweek[0.049]; win[0.048]; derby[0.048]; Paolo[0.048]; Casiraghi[0.048]; Cups[0.048]; Baggio[0.048]; injury[0.048]; injury[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marcelo Otero\u001b[39m ==> ENTITY: \u001b[32mMarcelo Otero\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUruguayan[0.046]; Chilean[0.046]; Roberto[0.045]; match[0.044]; Angloma[0.044]; Inter[0.043]; Inter[0.043]; Del[0.042]; Vicenza[0.042]; Vicenza[0.042]; Alessandro[0.041]; club[0.041]; Zamorano[0.041]; Internazionale[0.041]; Juventus[0.040]; Juventus[0.040]; defender[0.040]; defender[0.040]; midfielder[0.040]; field[0.040]; Sampdoria[0.040]; Sampdoria[0.040]; scorer[0.040]; Croat[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dejan Savicevic\u001b[39m ==> ENTITY: \u001b[32mDejan Savićević\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.043]; win[0.043]; win[0.043]; Montenegrin[0.043]; knee[0.042]; midfield[0.042]; midweek[0.042]; Roberto[0.042]; Roberto[0.042]; Balbo[0.042]; despite[0.042]; Marco[0.042]; Rosenborg[0.042]; Casiraghi[0.041]; signing[0.041]; Bierhoff[0.041]; Pavel[0.041]; Roma[0.041]; Croat[0.041]; Milan[0.041]; Milan[0.041]; Milan[0.041]; Milan[0.041]; Champions[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vicenza ==> ENTITY: \u001b[32mVicenza Calcio\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.284:0.284[0]; log p(e|m)= -0.703:-0.703[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReggiana[0.047]; Reggiana[0.047]; Serie[0.044]; Piacenza[0.043]; Piacenza[0.043]; league[0.043]; league[0.043]; Milan[0.042]; Rome[0.041]; Bologna[0.041]; Bologna[0.041]; Bologna[0.041]; fourth[0.041]; Cagliari[0.041]; Cagliari[0.041]; Napoli[0.041]; Division[0.041]; relegation[0.040]; Italian[0.040]; Italian[0.040]; Italian[0.040]; team[0.040]; Inter[0.040]; match[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Juventus\u001b[39m ==> ENTITY: \u001b[32mJuventus F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.226:0.226[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.044]; Fabio[0.043]; Juventus[0.043]; Roberto[0.043]; Roberto[0.043]; match[0.042]; Sampdoria[0.042]; Sampdoria[0.042]; Mario[0.042]; Inter[0.042]; Alessandro[0.042]; season[0.041]; Parma[0.041]; Parma[0.041]; Parma[0.041]; recent[0.041]; Internazionale[0.041]; Marcelo[0.041]; Cagliari[0.041]; Piero[0.041]; Cannavaro[0.041]; Atalanta[0.040]; champions[0.040]; Mancini[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = AS Roma\u001b[39m ==> ENTITY: \u001b[32mA.S. Roma\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.264:0.264[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReggiana[0.045]; Reggiana[0.045]; season[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Roma[0.043]; Lazio[0.043]; Lazio[0.043]; Napoli[0.042]; League[0.042]; Fiorentina[0.041]; Fiorentina[0.041]; Inter[0.041]; Udinese[0.041]; Casiraghi[0.040]; Italian[0.040]; midweek[0.040]; Amoruso[0.040]; Uefa[0.040]; players[0.040]; Cagliari[0.040]; Cagliari[0.040]; team[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Croat ==> ENTITY: \u001b[32mCroatia\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.086:0.086[0]; log p(e|m)= -1.440:-1.440[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCroat[0.045]; George[0.044]; Dejan[0.044]; Montenegrin[0.043]; Daniel[0.042]; Mario[0.042]; Roberto[0.042]; Roberto[0.042]; knee[0.042]; midfield[0.042]; ago[0.041]; world[0.041]; Bierhoff[0.041]; conquering[0.041]; German[0.041]; Nestor[0.041]; Eugenio[0.041]; strong[0.041]; Alen[0.041]; despite[0.041]; seen[0.041]; Argentine[0.040]; Argentine[0.040]; season[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bologna ==> ENTITY: \u001b[32mBologna F.C. 1909\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.264:0.264[0]; log p(e|m)= -1.661:-1.661[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReggiana[0.046]; Reggiana[0.046]; Rome[0.043]; Bologna[0.043]; Bologna[0.043]; Serie[0.043]; league[0.043]; league[0.043]; Italian[0.042]; Italian[0.042]; Italian[0.042]; Napoli[0.042]; Milan[0.042]; Fiorentina[0.041]; Fiorentina[0.041]; games[0.041]; team[0.041]; Inter[0.040]; Cagliari[0.040]; Cagliari[0.040]; win[0.040]; relegation[0.040]; Amoruso[0.039]; Soccer[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Francesco Totti\u001b[39m ==> ENTITY: \u001b[32mFrancesco Totti\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.239:0.239[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.046]; Baggio[0.045]; Italian[0.044]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Paolo[0.044]; Lazio[0.043]; Lazio[0.043]; Casiraghi[0.043]; Marco[0.043]; Roberto[0.043]; win[0.043]; Napoli[0.043]; Napoli[0.043]; Roma[0.043]; Roma[0.043]; second[0.042]; injury[0.042]; injury[0.042]; League[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Inter ==> ENTITY: \u001b[32mInter Milan\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.177:0.177[0]; log p(e|m)= -0.015:-0.015[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.046]; Juventus[0.045]; match[0.045]; Inter[0.044]; Internazionale[0.043]; Sampdoria[0.043]; champions[0.042]; signing[0.042]; Roberto[0.042]; recent[0.042]; Mancini[0.042]; strong[0.042]; nonetheless[0.041]; successive[0.041]; captain[0.041]; midfielder[0.040]; Vicenza[0.040]; Vicenza[0.040]; striker[0.040]; Angloma[0.040]; Alessandro[0.040]; defender[0.040]; defender[0.040]; Uruguayan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Napoli ==> ENTITY: \u001b[32mS.S.C. Napoli\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.265:0.265[0]; log p(e|m)= -0.373:-0.373[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReggiana[0.045]; Reggiana[0.045]; Serie[0.043]; games[0.042]; league[0.042]; league[0.042]; Lazio[0.042]; Milan[0.042]; Milan[0.042]; Rome[0.042]; Piacenza[0.042]; Piacenza[0.042]; Inter[0.041]; Cagliari[0.041]; Cagliari[0.041]; Bologna[0.041]; Bologna[0.041]; Bologna[0.041]; Italian[0.041]; Italian[0.041]; season[0.041]; team[0.040]; Fiorentina[0.040]; Fiorentina[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Oliver Bierhoff\u001b[39m ==> ENTITY: \u001b[32mOliver Bierhoff\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwin[0.044]; win[0.044]; season[0.043]; chances[0.043]; German[0.042]; defender[0.042]; defender[0.042]; Roberto[0.042]; Roberto[0.042]; second[0.042]; League[0.041]; needed[0.041]; Milan[0.041]; Milan[0.041]; Milan[0.041]; Milan[0.041]; return[0.041]; knee[0.041]; lift[0.041]; Baggio[0.041]; midfield[0.041]; midfielder[0.041]; midfielder[0.041]; midfielder[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fiorentina\u001b[39m ==> ENTITY: \u001b[32mACF Fiorentina\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.270:0.270[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReggiana[0.047]; Reggiana[0.047]; league[0.045]; season[0.043]; Milan[0.042]; Milan[0.042]; Lazio[0.042]; Lazio[0.042]; Napoli[0.042]; Inter[0.042]; Fiorentina[0.042]; Italian[0.041]; Italian[0.041]; team[0.041]; Roma[0.041]; Bologna[0.040]; Bologna[0.040]; Cagliari[0.040]; Cagliari[0.040]; match[0.040]; match[0.040]; teams[0.040]; derby[0.040]; Casiraghi[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alen Boksic\u001b[39m ==> ENTITY: \u001b[32mAlen Bokšić\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFabio[0.045]; match[0.044]; club[0.043]; season[0.043]; Inter[0.043]; Inter[0.043]; Sensini[0.043]; goals[0.042]; win[0.042]; Angloma[0.042]; champions[0.041]; Juventus[0.041]; Juventus[0.041]; Sampdoria[0.041]; Sampdoria[0.041]; Mancini[0.041]; scorer[0.041]; Alessandro[0.040]; Cagliari[0.040]; injury[0.040]; Internazionale[0.040]; Parma[0.040]; Parma[0.040]; Mario[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Roman ==> ENTITY: \u001b[32mRome\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.153:0.153[0]; log p(e|m)= -3.194:-3.194[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRome[0.052]; Italian[0.045]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Francesco[0.044]; Roma[0.042]; Roma[0.042]; Paolo[0.042]; home[0.041]; Marco[0.041]; Pierluigi[0.040]; Fausto[0.040]; form[0.040]; draws[0.040]; present[0.040]; Lorenzo[0.040]; Reggiana[0.040]; Norwegian[0.040]; scrap[0.040]; Perugia[0.039]; Perugia[0.039]; Argentine[0.039]; Totti[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = GMT\u001b[39m ==> ENTITY: \u001b[32mGreenwich Mean Time\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.079:0.079[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nautumn[0.047]; Sunday[0.047]; Sunday[0.047]; following[0.046]; Kennet[0.044]; parentheses[0.044]; Inter[0.044]; usual[0.043]; start[0.043]; Division[0.043]; times[0.043]; Weekend[0.043]; scrap[0.043]; represent[0.043]; biggest[0.043]; kick[0.042]; positions[0.042]; fourth[0.042]; placed[0.042]; Matches[0.042]; Rome[0.042]; Soccer[0.042]; expect[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fabio Cannavaro\u001b[39m ==> ENTITY: \u001b[32mFabio Cannavaro\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.231:0.231[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJuventus[0.043]; Juventus[0.043]; Roberto[0.043]; Roberto[0.043]; club[0.043]; Alessandro[0.043]; Mario[0.042]; absence[0.042]; Parma[0.042]; Parma[0.042]; Parma[0.042]; Napoli[0.041]; Napoli[0.041]; Piero[0.041]; injury[0.041]; injury[0.041]; injury[0.041]; match[0.041]; Mancini[0.041]; season[0.041]; Argentine[0.041]; Argentine[0.041]; Filippo[0.041]; Internazionale[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = AC Milan\u001b[39m ==> ENTITY: \u001b[32mA.C. Milan\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.250:0.250[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Lazio[0.043]; Lazio[0.043]; Napoli[0.042]; Napoli[0.042]; Champions[0.042]; Roma[0.041]; Roma[0.041]; win[0.041]; Italian[0.041]; Udinese[0.041]; Udinese[0.041]; Uefa[0.040]; Casiraghi[0.040]; Roberto[0.040]; season[0.040]; despite[0.040]; midweek[0.040]; derby[0.040]; teams[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rome ==> ENTITY: \u001b[32mRome\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.046]; fourth[0.045]; played[0.045]; Italian[0.045]; Italian[0.045]; Italian[0.045]; Milan[0.045]; Swede[0.044]; Bologna[0.044]; Bologna[0.044]; Bologna[0.044]; win[0.043]; Division[0.043]; Matches[0.043]; Vicenza[0.043]; biggest[0.042]; league[0.042]; league[0.042]; following[0.042]; Sunday[0.042]; Sunday[0.042]; placed[0.042]; games[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Inter ==> ENTITY: \u001b[32mInter Milan\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.242:0.242[0]; log p(e|m)= -0.015:-0.015[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; league[0.046]; Reggiana[0.046]; Reggiana[0.046]; team[0.045]; Milan[0.044]; Milan[0.044]; Lazio[0.044]; match[0.044]; match[0.044]; games[0.044]; Serie[0.043]; season[0.043]; Napoli[0.043]; played[0.043]; teams[0.043]; players[0.042]; Fiorentina[0.042]; Fiorentina[0.042]; derby[0.042]; Amoruso[0.041]; Roma[0.041]; win[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Verona ==> ENTITY: \u001b[32mHellas Verona F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.263:0.263[0]; log p(e|m)= -2.040:-2.040[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nParma[0.045]; Parma[0.045]; Parma[0.045]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; League[0.042]; Napoli[0.042]; Napoli[0.042]; Cagliari[0.042]; Udinese[0.041]; Udinese[0.041]; season[0.041]; Francesco[0.041]; Atalanta[0.041]; Fabio[0.041]; Verona[0.041]; Verona[0.041]; midweek[0.040]; second[0.040]; Corini[0.039]; Roberto[0.039]; Roberto[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Perugia ==> ENTITY: \u001b[32mA.C. Perugia Calcio\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.256:0.256[0]; log p(e|m)= -0.753:-0.753[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; Reggiana[0.046]; Reggiana[0.046]; season[0.042]; Rome[0.042]; Napoli[0.042]; Bologna[0.042]; Bologna[0.042]; Milan[0.041]; Milan[0.041]; Cups[0.041]; Italian[0.041]; Italian[0.041]; Perugia[0.041]; team[0.041]; Cagliari[0.041]; Cagliari[0.041]; Amoruso[0.041]; Fiorentina[0.041]; Fiorentina[0.041]; Lazio[0.040]; Lazio[0.040]; Paolo[0.040]; players[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pierre Laigle\u001b[39m ==> ENTITY: \u001b[32mPierre Laigle\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncaptain[0.049]; club[0.047]; champions[0.046]; Angloma[0.045]; French[0.045]; French[0.045]; French[0.045]; home[0.044]; scorer[0.044]; striker[0.043]; injury[0.043]; defender[0.043]; defender[0.043]; defender[0.043]; Juventus[0.042]; Juventus[0.042]; match[0.042]; Portuguese[0.042]; midfielder[0.042]; midfielder[0.042]; Uruguayan[0.042]; field[0.042]; Cannavaro[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italian ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.169:0.169[0]; log p(e|m)= -0.759:-0.759[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReggiana[0.046]; League[0.043]; Rome[0.042]; Milan[0.042]; Milan[0.042]; Milan[0.042]; Lazio[0.042]; Lazio[0.042]; match[0.042]; Emiliano[0.042]; Paolo[0.042]; Marco[0.041]; Francesco[0.041]; Casiraghi[0.041]; season[0.041]; Inter[0.041]; Argentine[0.041]; Roma[0.041]; Roma[0.041]; Fiorentina[0.041]; Fiorentina[0.041]; team[0.041]; Amoruso[0.041]; Champions[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sampdoria\u001b[39m ==> ENTITY: \u001b[32mU.C. Sampdoria\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.233:0.233[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.045]; Juventus[0.043]; Juventus[0.043]; Sampdoria[0.043]; Mancini[0.043]; Fabio[0.043]; Parma[0.043]; Parma[0.043]; Parma[0.043]; season[0.043]; match[0.042]; Cagliari[0.042]; Internazionale[0.041]; Inter[0.041]; win[0.041]; signing[0.041]; signing[0.041]; goals[0.040]; Roberto[0.040]; Roberto[0.040]; Atalanta[0.040]; Alessandro[0.040]; champions[0.040]; Sensini[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Milan ==> ENTITY: \u001b[32mA.C. Milan\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.254:0.254[0]; log p(e|m)= -1.959:-1.959[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; Rome[0.044]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Lazio[0.043]; Lazio[0.043]; Italian[0.042]; Napoli[0.041]; Napoli[0.041]; Champions[0.041]; Roma[0.041]; Roma[0.041]; Paolo[0.041]; Udinese[0.040]; Udinese[0.040]; Francesco[0.040]; Marco[0.040]; Uefa[0.040]; Casiraghi[0.040]; Roberto[0.040]; Roberto[0.040]; season[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Piacenza ==> ENTITY: \u001b[32mS.S.D. Piacenza Calcio 1919\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.245:0.245[0]; log p(e|m)= -1.112:-1.112[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReggiana[0.045]; Reggiana[0.045]; league[0.045]; league[0.045]; Serie[0.042]; Italian[0.042]; Italian[0.042]; Italian[0.042]; Piacenza[0.042]; team[0.042]; Napoli[0.041]; Bologna[0.041]; Bologna[0.041]; Bologna[0.041]; Rome[0.041]; win[0.041]; games[0.041]; Soccer[0.040]; played[0.040]; Milan[0.040]; fourth[0.040]; relegation[0.040]; Cagliari[0.040]; Cagliari[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Roberto Ayala\u001b[39m ==> ENTITY: \u001b[32mRoberto Ayala\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRoberto[0.045]; Fabio[0.044]; season[0.043]; Mario[0.042]; Napoli[0.042]; Napoli[0.042]; League[0.042]; Argentine[0.042]; Argentine[0.042]; Eugenio[0.042]; Sensini[0.041]; injury[0.041]; injury[0.041]; injury[0.041]; Parma[0.041]; Parma[0.041]; Parma[0.041]; goals[0.041]; second[0.041]; Juventus[0.041]; signing[0.041]; defender[0.041]; defender[0.041]; chances[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Internazionale\u001b[39m ==> ENTITY: \u001b[32mInter Milan\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRoberto[0.044]; club[0.044]; Juventus[0.044]; Juventus[0.044]; Fabio[0.044]; Mancini[0.043]; Inter[0.043]; Inter[0.043]; Sampdoria[0.042]; Sampdoria[0.042]; match[0.042]; Alessandro[0.042]; impressive[0.041]; signing[0.041]; Marcelo[0.041]; champions[0.041]; midfielder[0.040]; midfielder[0.040]; defender[0.040]; defender[0.040]; defender[0.040]; Vicenza[0.040]; Vicenza[0.040]; Angloma[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.058:0.058[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; team[0.045]; match[0.045]; match[0.045]; players[0.045]; league[0.044]; league[0.044]; Soccer[0.043]; played[0.042]; win[0.041]; fourth[0.041]; represent[0.040]; Igor[0.040]; times[0.040]; midfielder[0.040]; Italian[0.040]; Italian[0.040]; Italian[0.040]; German[0.040]; Matches[0.040]; Swede[0.039]; defender[0.039]; defender[0.039]; placed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pierluigi Casiraghi\u001b[39m ==> ENTITY: \u001b[32mPierluigi Casiraghi\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReggiana[0.047]; season[0.047]; team[0.047]; League[0.046]; match[0.044]; Champions[0.044]; Uefa[0.044]; win[0.043]; players[0.043]; Italian[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; injury[0.043]; derby[0.043]; midfielder[0.043]; midfielder[0.043]; Totti[0.042]; teams[0.042]; Dietmar[0.042]; eliminated[0.042]; midweek[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Juventus\u001b[39m ==> ENTITY: \u001b[32mJuventus F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.224:0.224[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.044]; Fabio[0.043]; Juventus[0.043]; Roberto[0.043]; match[0.043]; Sampdoria[0.042]; Sampdoria[0.042]; Mario[0.042]; Inter[0.042]; Inter[0.042]; Alessandro[0.042]; season[0.042]; Parma[0.041]; Parma[0.041]; Parma[0.041]; recent[0.041]; Internazionale[0.041]; Marcelo[0.041]; Cagliari[0.041]; Piero[0.041]; Cannavaro[0.041]; Atalanta[0.041]; champions[0.040]; Mancini[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = AC Milan\u001b[39m ==> ENTITY: \u001b[32mA.C. Milan\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.262:0.262[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReggiana[0.045]; Reggiana[0.045]; Milan[0.044]; league[0.043]; league[0.043]; Serie[0.043]; Soccer[0.043]; Lazio[0.043]; Inter[0.042]; Napoli[0.041]; games[0.041]; Fiorentina[0.041]; Fiorentina[0.041]; Bologna[0.041]; Bologna[0.041]; Bologna[0.041]; win[0.041]; Italian[0.041]; Italian[0.041]; Italian[0.041]; team[0.040]; match[0.040]; match[0.040]; Amoruso[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mLiberian\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mLiberia national football team\u001b[39m <---> \u001b[32mLiberia\u001b[39m\t\nSCORES: global= 0.281:0.261[\u001b[31m0.020\u001b[39m]; local()= 0.154:0.018[\u001b[31m0.136\u001b[39m]; log p(e|m)= -1.076:0.000[\u001b[32m1.076\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.048]; win[0.047]; League[0.047]; teams[0.047]; striker[0.046]; striker[0.046]; striker[0.046]; Champions[0.046]; second[0.045]; Weah[0.045]; midfielder[0.045]; midfielder[0.045]; return[0.045]; chances[0.045]; defender[0.045]; defender[0.045]; knee[0.045]; field[0.045]; eliminated[0.045]; Uefa[0.045]; midfield[0.045]; Balbo[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ramon Vega\u001b[39m ==> ENTITY: \u001b[32mRamon Vega\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReggiana[0.048]; Reggiana[0.048]; players[0.047]; league[0.045]; season[0.045]; Roma[0.044]; striker[0.043]; striker[0.043]; derby[0.043]; relegation[0.043]; home[0.043]; home[0.043]; Uefa[0.043]; match[0.043]; match[0.043]; Inter[0.042]; Amoruso[0.042]; Dietmar[0.042]; Napoli[0.042]; start[0.042]; Milan[0.042]; Milan[0.042]; midfielder[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mario Stanic\u001b[39m ==> ENTITY: \u001b[32mMario Stanić\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.048]; win[0.046]; Fabio[0.045]; goals[0.044]; Baggio[0.044]; club[0.044]; knee[0.044]; Dejan[0.044]; midfield[0.044]; chances[0.043]; weeks[0.043]; second[0.043]; Bierhoff[0.043]; signing[0.043]; signing[0.043]; striker[0.043]; scorer[0.043]; injury[0.042]; injury[0.042]; injury[0.042]; midfielder[0.042]; midfielder[0.042]; midfielder[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Daniele Carnasciali\u001b[39m ==> ENTITY: \u001b[32mDaniele Carnasciali\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.268:0.268[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReggiana[0.045]; Reggiana[0.045]; season[0.045]; Bologna[0.044]; league[0.044]; Italian[0.042]; Italian[0.042]; Roma[0.042]; Roma[0.042]; team[0.042]; Paolo[0.041]; derby[0.041]; Milan[0.041]; Milan[0.041]; Fiorentina[0.041]; Fiorentina[0.041]; Lazio[0.041]; Lazio[0.041]; Piacenza[0.041]; Cagliari[0.040]; Cagliari[0.040]; teams[0.039]; Dietmar[0.039]; players[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.053:0.053[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrench[0.048]; Angloma[0.045]; match[0.045]; club[0.044]; Pierre[0.043]; champions[0.043]; Chilean[0.042]; captain[0.041]; Inter[0.041]; Inter[0.041]; Portuguese[0.041]; Del[0.041]; defender[0.041]; defender[0.041]; striker[0.041]; home[0.041]; midfielder[0.041]; world[0.041]; Uruguayan[0.040]; Roberto[0.040]; Marcelo[0.040]; Sampdoria[0.040]; Internazionale[0.040]; recent[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Serie A ==> ENTITY: \u001b[32mSerie A\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.246:0.246[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.048]; league[0.048]; Reggiana[0.047]; Italian[0.044]; Italian[0.044]; Italian[0.044]; Bologna[0.044]; Bologna[0.044]; Bologna[0.044]; Milan[0.044]; games[0.044]; Inter[0.043]; Soccer[0.043]; Cagliari[0.043]; Cagliari[0.043]; Napoli[0.043]; played[0.042]; placed[0.042]; win[0.042]; fourth[0.041]; following[0.041]; match[0.041]; Division[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Napoli ==> ENTITY: \u001b[32mS.S.C. Napoli\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.264:0.264[0]; log p(e|m)= -0.373:-0.373[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNapoli[0.044]; League[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Fabio[0.043]; season[0.043]; Parma[0.042]; Parma[0.042]; Parma[0.042]; Roberto[0.042]; Roberto[0.042]; Marco[0.042]; Champions[0.041]; Cagliari[0.041]; Udinese[0.041]; Udinese[0.041]; Mario[0.041]; second[0.040]; Baggio[0.040]; win[0.040]; win[0.040]; midweek[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCzech\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mCzech Republic national football team\u001b[39m <---> \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.248:0.245[\u001b[31m0.003\u001b[39m]; local()= 0.173:0.079[\u001b[31m0.094\u001b[39m]; log p(e|m)= -3.324:-1.207[\u001b[32m2.117\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.045]; League[0.045]; teams[0.044]; players[0.044]; Champions[0.044]; season[0.042]; Uefa[0.042]; win[0.042]; midweek[0.042]; Pavel[0.042]; German[0.041]; Milan[0.041]; Milan[0.041]; Milan[0.041]; Milan[0.041]; Cups[0.041]; return[0.041]; injury[0.041]; Dejan[0.040]; Casiraghi[0.040]; eliminated[0.040]; needed[0.040]; Italian[0.040]; Nedved[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sampdoria\u001b[39m ==> ENTITY: \u001b[32mU.C. Sampdoria\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.045]; Juventus[0.044]; Juventus[0.044]; Sampdoria[0.043]; Mancini[0.043]; Fabio[0.043]; Parma[0.043]; season[0.043]; match[0.043]; Cagliari[0.042]; Internazionale[0.041]; Inter[0.041]; Inter[0.041]; win[0.041]; signing[0.041]; goals[0.041]; Angloma[0.041]; Roberto[0.041]; Atalanta[0.040]; Alessandro[0.040]; champions[0.040]; Sensini[0.040]; midfielder[0.040]; midfielder[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = George Weah\u001b[39m ==> ENTITY: \u001b[32mGeorge Weah\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.045]; Casiraghi[0.045]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Liberian[0.044]; Champions[0.044]; signing[0.043]; Lazio[0.043]; striker[0.043]; striker[0.043]; striker[0.043]; midfielder[0.043]; midfielder[0.043]; midfielder[0.043]; Uefa[0.043]; Roberto[0.043]; Roberto[0.043]; Roma[0.042]; Baggio[0.042]; Cups[0.042]; chances[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Parma ==> ENTITY: \u001b[32mParma F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.243:0.243[0]; log p(e|m)= -1.085:-1.085[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.044]; Parma[0.043]; Parma[0.043]; Juventus[0.043]; Juventus[0.043]; Fabio[0.043]; Milan[0.043]; Filippo[0.042]; Alessandro[0.042]; Napoli[0.042]; Napoli[0.042]; Roberto[0.041]; Roberto[0.041]; season[0.041]; signing[0.041]; signing[0.041]; Sampdoria[0.041]; Sampdoria[0.041]; Piero[0.041]; Verona[0.041]; Verona[0.041]; Verona[0.041]; Udinese[0.041]; Cagliari[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lorenzo Amoruso\u001b[39m ==> ENTITY: \u001b[32mLorenzo Amoruso\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.043]; match[0.043]; Reggiana[0.043]; Reggiana[0.043]; Italian[0.042]; Lazio[0.042]; Lazio[0.042]; Milan[0.042]; Milan[0.042]; Napoli[0.042]; derby[0.042]; Fiorentina[0.042]; Fiorentina[0.042]; players[0.042]; league[0.041]; Uefa[0.041]; defender[0.041]; defender[0.041]; defender[0.041]; season[0.041]; striker[0.040]; striker[0.040]; Casiraghi[0.040]; team[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mArgentine\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mArgentina national football team\u001b[39m <---> \u001b[32mArgentina\u001b[39m\t\nSCORES: global= 0.254:0.253[\u001b[31m0.001\u001b[39m]; local()= 0.192:0.128[\u001b[31m0.063\u001b[39m]; log p(e|m)= -3.442:-0.189[\u001b[32m3.253\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRoberto[0.052]; Roberto[0.052]; League[0.051]; Argentine[0.051]; Champions[0.049]; Mario[0.049]; season[0.049]; Fabio[0.048]; French[0.048]; Eugenio[0.047]; Nestor[0.047]; midweek[0.047]; second[0.046]; win[0.046]; win[0.046]; Filippo[0.045]; Baggio[0.045]; goals[0.045]; injury[0.045]; injury[0.045]; injury[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paolo Negro\u001b[39m ==> ENTITY: \u001b[32mPaolo Negro\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.223:0.223[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLazio[0.044]; Lazio[0.044]; players[0.042]; Milan[0.042]; Milan[0.042]; Milan[0.042]; Milan[0.042]; Milan[0.042]; League[0.042]; Udinese[0.042]; Udinese[0.042]; Perugia[0.042]; Perugia[0.042]; Champions[0.042]; Casiraghi[0.041]; Italian[0.041]; match[0.041]; Fiorentina[0.041]; season[0.041]; derby[0.041]; Roberto[0.040]; Uefa[0.040]; Baggio[0.040]; Amoruso[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bologna ==> ENTITY: \u001b[32mBologna F.C. 1909\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.260:0.260[0]; log p(e|m)= -1.661:-1.661[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReggiana[0.049]; Reggiana[0.049]; Rome[0.045]; Bologna[0.045]; Bologna[0.045]; Serie[0.045]; league[0.045]; league[0.045]; Italian[0.044]; Italian[0.044]; Italian[0.044]; Napoli[0.044]; Milan[0.044]; games[0.042]; Inter[0.042]; Cagliari[0.042]; Cagliari[0.042]; win[0.042]; team[0.041]; relegation[0.041]; Soccer[0.041]; played[0.041]; Vicenza[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vicenza ==> ENTITY: \u001b[32mVicenza Calcio\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.224:0.224[0]; log p(e|m)= -0.703:-0.703[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.045]; Juventus[0.045]; Juventus[0.045]; Sampdoria[0.043]; Sampdoria[0.043]; Vicenza[0.042]; Atalanta[0.042]; captain[0.042]; Inter[0.042]; Inter[0.042]; Alessandro[0.042]; match[0.042]; Internazionale[0.041]; Roberto[0.041]; champions[0.041]; Fabio[0.041]; Sensini[0.041]; Cannavaro[0.041]; Angloma[0.040]; scorer[0.040]; midfielder[0.040]; midfielder[0.040]; Mancini[0.040]; striker[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Roberto Baggio\u001b[39m ==> ENTITY: \u001b[32mRoberto Baggio\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.256:0.256[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMilan[0.044]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Marco[0.043]; Totti[0.043]; Paolo[0.042]; Roberto[0.042]; Mario[0.042]; Casiraghi[0.042]; Napoli[0.041]; Napoli[0.041]; Francesco[0.041]; Udinese[0.041]; Udinese[0.041]; Parma[0.041]; Parma[0.041]; win[0.041]; Corini[0.041]; despite[0.040]; injury[0.040]; injury[0.040]; injury[0.040]; Roma[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mArgentine\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mArgentina national football team\u001b[39m <---> \u001b[32mArgentina\u001b[39m\t\nSCORES: global= 0.259:0.259[\u001b[31m0.000\u001b[39m]; local()= 0.188:0.139[\u001b[31m0.049\u001b[39m]; log p(e|m)= -3.442:-0.189[\u001b[32m3.253\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.045]; Argentine[0.045]; Roberto[0.045]; Roberto[0.045]; season[0.043]; club[0.043]; Fabio[0.042]; French[0.042]; French[0.042]; Eugenio[0.042]; Nestor[0.042]; Mario[0.041]; Portuguese[0.041]; Del[0.041]; second[0.041]; win[0.041]; Juventus[0.040]; Juventus[0.040]; champions[0.040]; goals[0.040]; injury[0.040]; injury[0.040]; field[0.040]; field[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSwede\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSweden national football team\u001b[39m <---> \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.264:0.259[\u001b[31m0.005\u001b[39m]; local()= 0.163:0.076[\u001b[31m0.087\u001b[39m]; log p(e|m)= -3.170:-0.646[\u001b[32m2.524\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.045]; Andersson[0.044]; games[0.044]; players[0.044]; league[0.044]; league[0.044]; Soccer[0.044]; team[0.043]; played[0.043]; Reggiana[0.042]; Reggiana[0.042]; win[0.042]; midfielder[0.041]; autumn[0.041]; Matches[0.041]; kick[0.040]; fourth[0.040]; defender[0.040]; defender[0.040]; German[0.040]; Amoruso[0.040]; Sunday[0.039]; Sunday[0.039]; Russian[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fausto Pizzi\u001b[39m ==> ENTITY: \u001b[32mFausto Pizzi\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInter[0.044]; Reggiana[0.044]; Reggiana[0.044]; Cagliari[0.042]; Cagliari[0.042]; season[0.042]; Milan[0.042]; Milan[0.042]; Milan[0.042]; Milan[0.042]; League[0.042]; Perugia[0.042]; Perugia[0.042]; Italian[0.041]; match[0.041]; Paolo[0.041]; team[0.041]; Dietmar[0.041]; Napoli[0.041]; Emiliano[0.041]; outings[0.041]; Fiorentina[0.040]; Fiorentina[0.040]; Francesco[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Verona ==> ENTITY: \u001b[32mHellas Verona F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.258:0.258[0]; log p(e|m)= -2.040:-2.040[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nParma[0.045]; Parma[0.045]; Parma[0.045]; Milan[0.043]; Milan[0.043]; Juventus[0.043]; Juventus[0.043]; League[0.042]; Sampdoria[0.042]; Napoli[0.042]; Napoli[0.042]; Cagliari[0.042]; Udinese[0.041]; season[0.041]; Filippo[0.041]; Atalanta[0.041]; Fabio[0.041]; Verona[0.041]; Verona[0.041]; midweek[0.040]; second[0.040]; Corini[0.039]; Roberto[0.039]; Roberto[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Abel Balbo\u001b[39m ==> ENTITY: \u001b[32mAbel Balbo\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRoberto[0.046]; Marco[0.044]; Argentine[0.043]; Paolo[0.043]; Baggio[0.042]; Roma[0.042]; Roma[0.042]; Champions[0.042]; Emiliano[0.042]; Udinese[0.042]; Udinese[0.042]; Lazio[0.041]; Lazio[0.041]; Napoli[0.041]; Napoli[0.041]; Milan[0.041]; Milan[0.041]; Milan[0.041]; Milan[0.041]; Milan[0.041]; Italian[0.041]; Fausto[0.041]; match[0.040]; Casiraghi[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vicenza ==> ENTITY: \u001b[32mVicenza Calcio\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.222:0.222[0]; log p(e|m)= -0.703:-0.703[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.047]; Juventus[0.047]; Juventus[0.047]; Sampdoria[0.045]; Sampdoria[0.045]; Vicenza[0.044]; Atalanta[0.044]; captain[0.044]; Inter[0.044]; Inter[0.044]; Alessandro[0.043]; match[0.043]; Internazionale[0.043]; Roberto[0.043]; champions[0.043]; Sensini[0.043]; Angloma[0.042]; scorer[0.042]; midfielder[0.042]; Mancini[0.042]; striker[0.042]; Filippo[0.042]; Piero[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cagliari ==> ENTITY: \u001b[32mCagliari Calcio\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.254:0.254[0]; log p(e|m)= -0.870:-0.870[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReggiana[0.051]; Reggiana[0.051]; league[0.047]; league[0.047]; Serie[0.046]; team[0.046]; Lazio[0.045]; Bologna[0.045]; Bologna[0.045]; Bologna[0.045]; Milan[0.045]; Milan[0.045]; Inter[0.045]; Piacenza[0.045]; Piacenza[0.045]; Cagliari[0.044]; Amoruso[0.044]; games[0.044]; Napoli[0.044]; win[0.044]; Soccer[0.044]; Rome[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Milan ==> ENTITY: \u001b[32mA.C. Milan\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.253:0.253[0]; log p(e|m)= -1.959:-1.959[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.049]; Milan[0.048]; Milan[0.048]; Milan[0.048]; Parma[0.046]; Parma[0.046]; Parma[0.046]; Napoli[0.046]; Napoli[0.046]; Champions[0.046]; Roma[0.045]; Udinese[0.045]; Udinese[0.045]; Francesco[0.044]; Roberto[0.044]; Roberto[0.044]; season[0.044]; Cagliari[0.044]; midweek[0.044]; Paolo[0.044]; win[0.043]; win[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dietmar Beiersdorfer\u001b[39m ==> ENTITY: \u001b[32mDietmar Beiersdorfer\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.121:0.121[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; match[0.046]; match[0.046]; season[0.046]; Reggiana[0.045]; Reggiana[0.045]; players[0.045]; team[0.044]; outings[0.044]; Pavel[0.043]; teams[0.043]; start[0.043]; win[0.043]; German[0.043]; oriented[0.042]; Roma[0.042]; midfielder[0.042]; midfielder[0.042]; Napoli[0.042]; derby[0.042]; man[0.041]; miserable[0.041]; Nedved[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portuguese ==> ENTITY: \u001b[32mPortugal\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.124:0.124[0]; log p(e|m)= -1.127:-1.127[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.046]; season[0.044]; Fabio[0.042]; Inter[0.042]; Inter[0.042]; Argentine[0.042]; match[0.042]; Juventus[0.042]; Juventus[0.042]; signing[0.042]; signing[0.042]; champions[0.042]; Roberto[0.042]; Marcelo[0.041]; Uruguayan[0.041]; Eugenio[0.041]; Mario[0.041]; draws[0.041]; goals[0.041]; scorer[0.041]; weeks[0.041]; Internazionale[0.040]; world[0.040]; ago[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lazio ==> ENTITY: \u001b[32mS.S. Lazio\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.273:0.273[0]; log p(e|m)= -0.633:-0.633[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteams[0.045]; Reggiana[0.045]; Reggiana[0.045]; Lazio[0.044]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; team[0.042]; League[0.042]; Italian[0.042]; Inter[0.042]; Napoli[0.041]; match[0.040]; Fiorentina[0.040]; Fiorentina[0.040]; Casiraghi[0.040]; season[0.040]; Udinese[0.040]; Paolo[0.040]; Marco[0.040]; Roma[0.040]; Roma[0.040]; Emiliano[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cagliari ==> ENTITY: \u001b[32mCagliari Calcio\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.236:0.236[0]; log p(e|m)= -0.870:-0.870[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.045]; club[0.045]; Juventus[0.045]; Juventus[0.045]; Parma[0.044]; Parma[0.044]; Parma[0.044]; Milan[0.044]; Alessandro[0.044]; Roberto[0.044]; Roberto[0.044]; second[0.043]; Eugenio[0.043]; Corini[0.043]; Sampdoria[0.043]; Sampdoria[0.043]; Udinese[0.043]; Napoli[0.043]; Napoli[0.043]; win[0.043]; Fabio[0.042]; Mario[0.042]; Atalanta[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marco Delvecchio\u001b[39m ==> ENTITY: \u001b[32mMarco Delvecchio\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMilan[0.044]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Italian[0.042]; Napoli[0.042]; Napoli[0.042]; Lazio[0.042]; Lazio[0.042]; Rome[0.042]; injury[0.041]; injury[0.041]; Baggio[0.041]; Roma[0.041]; Roma[0.041]; Udinese[0.041]; Udinese[0.041]; midfield[0.041]; Perugia[0.040]; match[0.040]; Balbo[0.040]; Totti[0.040]; Rosenborg[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mChilean\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mChile national football team\u001b[39m <---> \u001b[32mChile\u001b[39m\t\nSCORES: global= 0.255:0.250[\u001b[31m0.005\u001b[39m]; local()= 0.127:0.088[\u001b[31m0.039\u001b[39m]; log p(e|m)= -3.352:-0.257[\u001b[32m3.095\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUruguayan[0.045]; match[0.045]; Inter[0.045]; Inter[0.045]; Roberto[0.045]; Angloma[0.043]; club[0.043]; world[0.042]; French[0.042]; French[0.042]; Marcelo[0.042]; champions[0.041]; suspended[0.040]; nonetheless[0.040]; Internazionale[0.040]; strong[0.040]; Otero[0.040]; home[0.040]; successive[0.040]; captain[0.040]; lead[0.040]; Croat[0.040]; Del[0.039]; defender[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Udinese\u001b[39m ==> ENTITY: \u001b[32mUdinese Calcio\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.258:0.258[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.044]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Parma[0.043]; Parma[0.043]; Parma[0.043]; Napoli[0.043]; Napoli[0.043]; Udinese[0.042]; League[0.042]; Cagliari[0.042]; Roma[0.042]; midweek[0.041]; Paolo[0.040]; win[0.040]; win[0.040]; Roberto[0.040]; Roberto[0.040]; Champions[0.040]; Francesco[0.040]; second[0.039]; Baggio[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alessandro Del Piero\u001b[39m ==> ENTITY: \u001b[32mAlessandro Del Piero\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.226:0.226[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFabio[0.044]; Roberto[0.044]; match[0.043]; Angloma[0.043]; Mario[0.043]; Marcelo[0.042]; Juventus[0.042]; Juventus[0.042]; Cannavaro[0.042]; Parma[0.041]; Parma[0.041]; Parma[0.041]; goals[0.041]; midfielder[0.041]; midfielder[0.041]; midfielder[0.041]; midfielder[0.041]; striker[0.041]; Sampdoria[0.041]; Sampdoria[0.041]; defender[0.041]; defender[0.041]; defender[0.041]; Cagliari[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kennet Andersson\u001b[39m ==> ENTITY: \u001b[32mKennet Andersson\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; league[0.047]; players[0.045]; match[0.045]; team[0.045]; Swede[0.045]; Soccer[0.044]; games[0.044]; played[0.044]; defender[0.043]; defender[0.043]; Dietmar[0.043]; fourth[0.043]; midfielder[0.042]; Reggiana[0.042]; Reggiana[0.042]; win[0.042]; parentheses[0.042]; kick[0.042]; Bologna[0.042]; Bologna[0.042]; Bologna[0.042]; defenders[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rome ==> ENTITY: \u001b[32mRome\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRoman[0.049]; match[0.043]; Pavel[0.043]; Champions[0.043]; Italian[0.043]; Milan[0.042]; Milan[0.042]; Milan[0.042]; Francesco[0.042]; Marco[0.042]; Czech[0.041]; Lorenzo[0.041]; players[0.041]; win[0.041]; Paolo[0.041]; teams[0.041]; Croat[0.041]; eliminated[0.041]; Argentine[0.041]; Swiss[0.041]; Emiliano[0.041]; George[0.040]; Ramon[0.040]; League[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.062:0.062[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrench[0.048]; club[0.044]; Pierre[0.043]; champions[0.042]; Fabio[0.042]; ago[0.042]; second[0.042]; Juventus[0.042]; Juventus[0.042]; German[0.042]; season[0.041]; win[0.041]; captain[0.041]; goals[0.041]; Argentine[0.041]; Argentine[0.041]; Portuguese[0.041]; weeks[0.041]; Del[0.041]; defender[0.041]; defender[0.041]; defender[0.041]; Cannavaro[0.041]; Bierhoff[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Udinese\u001b[39m ==> ENTITY: \u001b[32mUdinese Calcio\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.253:0.253[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.044]; Italian[0.044]; Lazio[0.043]; Lazio[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Napoli[0.042]; Napoli[0.042]; Udinese[0.042]; League[0.042]; Rome[0.042]; Roma[0.041]; Roma[0.041]; teams[0.041]; midweek[0.040]; derby[0.040]; Paolo[0.040]; win[0.040]; Roberto[0.040]; Champions[0.040]; Francesco[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Verona ==> ENTITY: \u001b[32mHellas Verona F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.264:0.264[0]; log p(e|m)= -2.040:-2.040[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nParma[0.047]; Parma[0.047]; Parma[0.047]; Milan[0.045]; Milan[0.045]; Milan[0.045]; Milan[0.045]; League[0.044]; Napoli[0.043]; Napoli[0.043]; Cagliari[0.043]; Udinese[0.043]; Udinese[0.043]; season[0.043]; Francesco[0.043]; Fabio[0.042]; Verona[0.042]; Verona[0.042]; midweek[0.042]; second[0.042]; Corini[0.041]; Roberto[0.041]; Roberto[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bologna ==> ENTITY: \u001b[32mBologna F.C. 1909\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.263:0.263[0]; log p(e|m)= -1.661:-1.661[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReggiana[0.046]; Reggiana[0.046]; Rome[0.043]; Bologna[0.043]; Bologna[0.043]; Serie[0.043]; league[0.043]; league[0.043]; Italian[0.042]; Italian[0.042]; Italian[0.042]; Napoli[0.042]; Milan[0.042]; Fiorentina[0.041]; Fiorentina[0.041]; games[0.041]; Inter[0.040]; Cagliari[0.040]; Cagliari[0.040]; win[0.040]; team[0.040]; relegation[0.040]; Soccer[0.039]; played[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Roberto Mancini\u001b[39m ==> ENTITY: \u001b[32mRoberto Mancini\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInternazionale[0.045]; club[0.044]; Inter[0.044]; Inter[0.044]; Fabio[0.043]; Sampdoria[0.043]; Sampdoria[0.043]; Juventus[0.043]; Juventus[0.043]; Parma[0.043]; match[0.042]; recent[0.042]; Cagliari[0.042]; signing[0.041]; Del[0.041]; goals[0.041]; weeks[0.040]; Alessandro[0.040]; successive[0.040]; captain[0.039]; Atalanta[0.039]; Piero[0.039]; injury[0.039]; striker[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = UEFA Cups\u001b[39m ==> ENTITY: \u001b[32mUEFA Europa League\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.045]; match[0.044]; Champions[0.044]; team[0.043]; Reggiana[0.043]; season[0.042]; Milan[0.042]; Milan[0.042]; Milan[0.042]; Milan[0.042]; win[0.042]; teams[0.042]; Lazio[0.042]; Lazio[0.042]; players[0.041]; Roma[0.041]; Roma[0.041]; Argentine[0.041]; midweek[0.041]; Casiraghi[0.041]; derby[0.040]; eliminated[0.040]; Fiorentina[0.040]; Fiorentina[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atalanta ==> ENTITY: \u001b[32mAtalanta B.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.224:0.224[0]; log p(e|m)= -0.347:-0.347[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJuventus[0.046]; Juventus[0.046]; club[0.046]; Parma[0.045]; Parma[0.045]; Parma[0.045]; Cagliari[0.045]; season[0.044]; Marcelo[0.044]; Sampdoria[0.044]; Sampdoria[0.044]; match[0.043]; Internazionale[0.043]; Alessandro[0.043]; Fabio[0.043]; Mario[0.042]; Roberto[0.042]; Roberto[0.042]; Portuguese[0.042]; champions[0.042]; second[0.042]; win[0.041]; nonetheless[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Piacenza ==> ENTITY: \u001b[32mS.S.D. Piacenza Calcio 1919\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.245:0.245[0]; log p(e|m)= -1.112:-1.112[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReggiana[0.045]; Reggiana[0.045]; league[0.045]; league[0.045]; Serie[0.042]; Italian[0.042]; Italian[0.042]; Italian[0.042]; Piacenza[0.042]; team[0.041]; Napoli[0.041]; Bologna[0.041]; Bologna[0.041]; Bologna[0.041]; Rome[0.041]; win[0.041]; games[0.041]; Soccer[0.040]; played[0.040]; Milan[0.040]; Fiorentina[0.040]; Fiorentina[0.040]; fourth[0.040]; relegation[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNorwegian\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNorway national football team\u001b[39m <---> \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.242:0.239[\u001b[31m0.002\u001b[39m]; local()= 0.141:0.045[\u001b[31m0.096\u001b[39m]; log p(e|m)= -4.135:-0.607[\u001b[32m3.528\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; Uefa[0.045]; teams[0.044]; eliminated[0.044]; midweek[0.044]; Champions[0.043]; season[0.043]; reduced[0.042]; win[0.042]; Rosenborg[0.041]; defender[0.041]; defender[0.041]; elimination[0.041]; German[0.041]; injury[0.041]; injury[0.041]; injury[0.041]; Casiraghi[0.041]; derby[0.040]; striker[0.040]; striker[0.040]; striker[0.040]; despite[0.040]; midfielder[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swiss ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.118:0.118[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.047]; match[0.047]; teams[0.047]; win[0.045]; players[0.045]; team[0.045]; German[0.044]; league[0.044]; league[0.044]; season[0.044]; Uefa[0.043]; Italian[0.043]; Italian[0.043]; fourth[0.043]; eliminated[0.042]; start[0.042]; Rome[0.042]; Reggiana[0.042]; Reggiana[0.042]; Swede[0.042]; form[0.042]; Roma[0.041]; Roman[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Eugenio Corini\u001b[39m ==> ENTITY: \u001b[32mEugenio Corini\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.206:0.206[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.044]; Milan[0.044]; Milan[0.044]; Parma[0.043]; Parma[0.043]; Parma[0.043]; Alessandro[0.043]; Fabio[0.043]; weeks[0.042]; Roberto[0.042]; Roberto[0.042]; Cagliari[0.041]; injury[0.041]; injury[0.041]; injury[0.041]; despite[0.041]; recent[0.041]; Sampdoria[0.041]; Juventus[0.041]; Juventus[0.041]; Filippo[0.041]; Napoli[0.040]; Napoli[0.040]; Udinese[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italian ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.193:0.193[0]; log p(e|m)= -0.759:-0.759[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItalian[0.047]; Italian[0.047]; league[0.046]; league[0.046]; Reggiana[0.045]; Serie[0.044]; games[0.042]; Soccer[0.042]; Rome[0.042]; Milan[0.041]; match[0.041]; Bologna[0.041]; Bologna[0.041]; Bologna[0.041]; Napoli[0.041]; played[0.040]; Cagliari[0.040]; Cagliari[0.040]; Division[0.039]; win[0.039]; Piacenza[0.039]; Piacenza[0.039]; fourth[0.039]; Vicenza[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fiorentina\u001b[39m ==> ENTITY: \u001b[32mACF Fiorentina\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.270:0.270[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReggiana[0.047]; Reggiana[0.047]; league[0.045]; season[0.043]; Milan[0.042]; Milan[0.042]; Lazio[0.042]; Lazio[0.042]; Napoli[0.042]; Inter[0.042]; Fiorentina[0.042]; Italian[0.041]; Italian[0.041]; team[0.041]; Roma[0.041]; Bologna[0.040]; Bologna[0.040]; Cagliari[0.040]; Cagliari[0.040]; match[0.040]; match[0.040]; teams[0.040]; derby[0.040]; Casiraghi[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Uruguayan ==> ENTITY: \u001b[32mUruguay\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.066:-0.066[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.046]; Roberto[0.044]; Angloma[0.044]; club[0.043]; Inter[0.043]; Inter[0.043]; Chilean[0.042]; champions[0.042]; Marcelo[0.042]; recent[0.042]; world[0.042]; strong[0.042]; Portuguese[0.041]; clash[0.041]; successive[0.041]; field[0.041]; home[0.041]; Del[0.040]; goals[0.040]; captain[0.040]; Juventus[0.040]; Juventus[0.040]; injury[0.040]; Zamorano[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lazio ==> ENTITY: \u001b[32mS.S. Lazio\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.260:0.260[0]; log p(e|m)= -0.633:-0.633[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteams[0.046]; Reggiana[0.046]; Lazio[0.044]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; team[0.042]; League[0.042]; Italian[0.042]; match[0.041]; Fiorentina[0.041]; Fiorentina[0.041]; Casiraghi[0.041]; season[0.041]; Udinese[0.041]; Paolo[0.040]; Marco[0.040]; Roma[0.040]; Roma[0.040]; Emiliano[0.040]; Amoruso[0.040]; win[0.040]; Champions[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dimas ==> ENTITY: \u001b[32mDimas Teixeira\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.453:-0.453[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.044]; Alessandro[0.043]; Roberto[0.043]; Angloma[0.042]; midfielder[0.042]; midfielder[0.042]; midfielder[0.042]; midfielder[0.042]; season[0.042]; champions[0.042]; Fabio[0.042]; Uruguayan[0.042]; Marcelo[0.042]; Mario[0.041]; Parma[0.041]; Parma[0.041]; Parma[0.041]; Cagliari[0.041]; goals[0.041]; Eugenio[0.041]; Sensini[0.041]; Inter[0.041]; Inter[0.041]; Portuguese[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rosenborg\u001b[39m ==> ENTITY: \u001b[32mRosenborg BK\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.048]; Norwegian[0.048]; season[0.046]; midweek[0.045]; Uefa[0.045]; teams[0.044]; Champions[0.044]; Cups[0.044]; welcome[0.043]; Nedved[0.043]; Casiraghi[0.043]; derby[0.043]; win[0.043]; despite[0.042]; striker[0.042]; striker[0.042]; striker[0.042]; chances[0.042]; second[0.042]; return[0.042]; injury[0.042]; injury[0.042]; injury[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Parma ==> ENTITY: \u001b[32mParma F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.242:0.242[0]; log p(e|m)= -1.085:-1.085[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nParma[0.043]; Parma[0.043]; Juventus[0.043]; Juventus[0.043]; Fabio[0.043]; Milan[0.043]; Milan[0.043]; Filippo[0.042]; Alessandro[0.042]; Napoli[0.042]; Napoli[0.042]; Roberto[0.041]; Roberto[0.041]; season[0.041]; signing[0.041]; signing[0.041]; Sampdoria[0.041]; Piero[0.041]; Verona[0.041]; Verona[0.041]; Verona[0.041]; Udinese[0.041]; Cagliari[0.040]; Eugenio[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italian ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.188:0.188[0]; log p(e|m)= -0.759:-0.759[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItalian[0.047]; Italian[0.047]; league[0.046]; league[0.046]; Reggiana[0.045]; Reggiana[0.045]; Serie[0.044]; games[0.041]; Soccer[0.041]; Rome[0.041]; Milan[0.041]; match[0.040]; team[0.040]; Bologna[0.040]; Bologna[0.040]; Bologna[0.040]; Napoli[0.040]; Inter[0.040]; Fiorentina[0.040]; Fiorentina[0.040]; players[0.040]; played[0.039]; Cagliari[0.039]; Cagliari[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Inter ==> ENTITY: \u001b[32mInter Milan\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.190:0.190[0]; log p(e|m)= -0.015:-0.015[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.046]; Juventus[0.045]; Juventus[0.045]; Inter[0.043]; Internazionale[0.043]; match[0.043]; Sampdoria[0.043]; Sampdoria[0.043]; signing[0.042]; champions[0.042]; Roberto[0.042]; recent[0.041]; Mancini[0.041]; strong[0.041]; nonetheless[0.041]; captain[0.041]; goals[0.040]; scorer[0.040]; midfielder[0.040]; Vicenza[0.040]; Vicenza[0.040]; striker[0.040]; Angloma[0.039]; Alessandro[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italian ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.187:0.187[0]; log p(e|m)= -0.759:-0.759[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItalian[0.047]; Italian[0.047]; league[0.047]; league[0.047]; Reggiana[0.045]; Serie[0.044]; games[0.042]; Soccer[0.042]; Rome[0.042]; Milan[0.042]; match[0.041]; Bologna[0.041]; Bologna[0.041]; Bologna[0.041]; played[0.040]; Cagliari[0.040]; Cagliari[0.040]; Division[0.039]; win[0.039]; Piacenza[0.039]; Piacenza[0.039]; fourth[0.039]; Vicenza[0.039]; Matches[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.047:0.047[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRoberto[0.043]; Roberto[0.043]; Bierhoff[0.043]; Argentine[0.042]; Argentine[0.042]; Roma[0.042]; Marco[0.042]; League[0.042]; Eugenio[0.042]; midweek[0.042]; season[0.042]; Mario[0.042]; win[0.042]; win[0.042]; injury[0.041]; injury[0.041]; French[0.041]; Croat[0.041]; weeks[0.041]; midfielder[0.041]; midfielder[0.041]; midfielder[0.041]; defender[0.040]; defender[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pavel Nedved\u001b[39m ==> ENTITY: \u001b[32mPavel Nedvěd\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayers[0.047]; Roberto[0.045]; Casiraghi[0.044]; injury[0.044]; Paolo[0.044]; Italian[0.044]; match[0.044]; League[0.044]; win[0.043]; Marco[0.043]; Czech[0.043]; Argentine[0.043]; defender[0.043]; Emiliano[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Lazio[0.043]; Lazio[0.043]; Uefa[0.043]; Dietmar[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Croat ==> ENTITY: \u001b[32mCroatia\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.077:0.077[0]; log p(e|m)= -1.440:-1.440[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCroat[0.045]; world[0.044]; Ivan[0.044]; Portuguese[0.044]; Chilean[0.043]; Argentine[0.042]; Daniel[0.042]; Mario[0.042]; Roberto[0.042]; continue[0.042]; Uruguayan[0.042]; nonetheless[0.041]; Nestor[0.041]; Fabio[0.041]; home[0.041]; club[0.041]; ago[0.041]; conquering[0.041]; Del[0.040]; Marcelo[0.040]; strong[0.040]; Alen[0.040]; successive[0.040]; seen[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cagliari ==> ENTITY: \u001b[32mCagliari Calcio\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.254:0.254[0]; log p(e|m)= -0.870:-0.870[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReggiana[0.051]; Reggiana[0.051]; league[0.047]; league[0.047]; Serie[0.046]; team[0.046]; Lazio[0.045]; Bologna[0.045]; Bologna[0.045]; Bologna[0.045]; Milan[0.045]; Milan[0.045]; Inter[0.045]; Piacenza[0.045]; Piacenza[0.045]; Cagliari[0.044]; Amoruso[0.044]; games[0.044]; Napoli[0.044]; win[0.044]; Soccer[0.044]; Rome[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Milan ==> ENTITY: \u001b[32mA.C. Milan\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.238:0.238[0]; log p(e|m)= -1.959:-1.959[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.045]; Milan[0.045]; Milan[0.045]; Milan[0.045]; Lazio[0.044]; Parma[0.042]; Parma[0.042]; Napoli[0.042]; Napoli[0.042]; Champions[0.042]; Roma[0.042]; Udinese[0.041]; Udinese[0.041]; Casiraghi[0.041]; Francesco[0.041]; Roberto[0.041]; Roberto[0.041]; midweek[0.040]; Paolo[0.040]; win[0.040]; Verona[0.040]; Verona[0.040]; Verona[0.040]; second[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reggiana\u001b[39m ==> ENTITY: \u001b[32mA.C. Reggiana 1919\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.253:0.253[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; Reggiana[0.046]; season[0.045]; derby[0.043]; Lazio[0.043]; Lazio[0.043]; team[0.042]; Cups[0.042]; players[0.042]; Milan[0.041]; Milan[0.041]; Napoli[0.041]; teams[0.041]; match[0.041]; match[0.041]; win[0.040]; Fiorentina[0.040]; Fiorentina[0.040]; Piacenza[0.040]; Piacenza[0.040]; Italian[0.040]; Italian[0.040]; Casiraghi[0.040]; Inter[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Daniel Bravo\u001b[39m ==> ENTITY: \u001b[32mDaniel Bravo\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.047]; club[0.047]; win[0.045]; Mario[0.044]; Napoli[0.044]; Napoli[0.044]; goals[0.044]; knee[0.044]; champions[0.043]; Pierre[0.043]; Bierhoff[0.043]; French[0.043]; French[0.043]; Juventus[0.043]; Juventus[0.043]; Roberto[0.043]; Roberto[0.043]; chances[0.043]; slim[0.043]; Atalanta[0.043]; experienced[0.042]; Fabio[0.042]; scorer[0.042]; \t\n\r [========================>.....................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 10s285ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2496/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1322testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = European Union\u001b[39m ==> ENTITY: \u001b[32mEuropean Union\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.006:-0.006[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuropean[0.053]; official[0.047]; agenda[0.047]; said[0.047]; said[0.047]; Brussels[0.047]; Brussels[0.047]; group[0.046]; talks[0.046]; documents[0.045]; documents[0.045]; postpone[0.044]; paid[0.044]; prices[0.044]; week[0.044]; similar[0.044]; payments[0.044]; payments[0.044]; experts[0.044]; experts[0.044]; experts[0.044]; Thursday[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = European ==> ENTITY: \u001b[32mEuropean Union\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.086:0.086[0]; log p(e|m)= -2.937:-2.937[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlanguages[0.049]; languages[0.049]; European[0.048]; week[0.043]; official[0.042]; group[0.042]; area[0.042]; area[0.042]; area[0.042]; Brussels[0.041]; Brussels[0.041]; translated[0.041]; rice[0.041]; rice[0.041]; rice[0.041]; rice[0.041]; cereal[0.040]; said[0.040]; said[0.040]; similar[0.040]; postponed[0.040]; postponed[0.040]; documents[0.039]; documents[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBrussels\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEuropean Union\u001b[39m <---> \u001b[32mBrussels\u001b[39m\t\nSCORES: global= 0.263:0.250[\u001b[31m0.013\u001b[39m]; local()= 0.123:0.100[\u001b[31m0.023\u001b[39m]; log p(e|m)= -3.324:-0.025[\u001b[32m3.299\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuropean[0.048]; European[0.048]; Brussels[0.045]; area[0.045]; area[0.045]; area[0.045]; documents[0.042]; documents[0.042]; official[0.042]; Thursday[0.042]; week[0.041]; Friday[0.041]; said[0.040]; said[0.040]; paid[0.040]; available[0.040]; postponed[0.040]; postponed[0.040]; similar[0.040]; postpone[0.039]; aid[0.039]; aid[0.039]; aid[0.039]; needed[0.039]; \t\n\r [========================>.....................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 10s291ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2499/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1194testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bonn\u001b[39m ==> ENTITY: \u001b[32mBonn\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.096:0.096[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMunich[0.044]; Munich[0.044]; soccer[0.044]; matches[0.044]; Soccer[0.042]; Stuttgart[0.042]; Freiburg[0.042]; German[0.042]; German[0.042]; played[0.042]; played[0.042]; division[0.042]; Bochum[0.042]; Bochum[0.042]; Dortmund[0.041]; Bremen[0.041]; Friday[0.041]; Division[0.041]; lost[0.040]; Bayer[0.040]; Bayer[0.040]; won[0.040]; Leverkusen[0.040]; Leverkusen[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = FC St. Pauli\u001b[39m ==> ENTITY: \u001b[32mFC St. Pauli\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVfL[0.044]; Borussia[0.043]; Borussia[0.043]; Bielefeld[0.043]; Monchengladbach[0.043]; VfB[0.042]; Duisburg[0.042]; Leverkusen[0.042]; Dortmund[0.042]; Arminia[0.042]; Cologne[0.042]; Bochum[0.042]; Stuttgart[0.042]; Werder[0.041]; Sc[0.041]; Bremen[0.041]; Karlsruhe[0.041]; Freiburg[0.041]; Dusseldorf[0.041]; Munich[0.041]; Munich[0.041]; Schalke[0.040]; Bayern[0.040]; Rostock[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = VfB Stuttgart\u001b[39m ==> ENTITY: \u001b[32mVfB Stuttgart\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.227:0.227[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVfL[0.045]; matches[0.045]; Soccer[0.044]; soccer[0.043]; Bochum[0.042]; Bochum[0.042]; Bremen[0.042]; Bremen[0.042]; Dortmund[0.042]; Borussia[0.042]; Cologne[0.042]; Leverkusen[0.042]; Leverkusen[0.042]; Schalke[0.041]; Schalke[0.041]; Freiburg[0.041]; Werder[0.041]; Werder[0.041]; Duisburg[0.041]; Bayern[0.040]; played[0.040]; played[0.040]; Karlsruhe[0.040]; Karlsruhe[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Freiburg ==> ENTITY: \u001b[32mSC Freiburg\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.195:0.195[0]; log p(e|m)= -1.743:-1.743[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStuttgart[0.046]; matches[0.046]; Cologne[0.045]; Bonn[0.045]; VfL[0.044]; Soccer[0.044]; Munich[0.044]; Munich[0.044]; Karlsruhe[0.044]; Karlsruhe[0.044]; played[0.044]; played[0.044]; division[0.043]; Dortmund[0.043]; German[0.043]; German[0.043]; soccer[0.043]; Bochum[0.042]; Bochum[0.042]; VfB[0.042]; Borussia[0.042]; Schalke[0.042]; Schalke[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Werder Bremen\u001b[39m ==> ENTITY: \u001b[32mSV Werder Bremen\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.229:0.229[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; VfL[0.045]; soccer[0.043]; Soccer[0.043]; Stuttgart[0.043]; Leverkusen[0.042]; Leverkusen[0.042]; Dortmund[0.042]; Borussia[0.042]; Bremen[0.041]; Werder[0.041]; VfB[0.041]; Schalke[0.041]; Schalke[0.041]; Bochum[0.041]; Bochum[0.041]; German[0.041]; German[0.041]; Bayern[0.041]; played[0.041]; played[0.041]; goals[0.040]; goals[0.040]; Freiburg[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Schalke 04\u001b[39m ==> ENTITY: \u001b[32mFC Schalke 04\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.232:0.232[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; VfL[0.043]; Bayer[0.043]; Bayer[0.043]; Borussia[0.043]; Borussia[0.043]; Dortmund[0.043]; Monchengladbach[0.042]; Stuttgart[0.042]; soccer[0.042]; Schalke[0.042]; Duisburg[0.041]; Bochum[0.041]; Bochum[0.041]; Bayern[0.041]; Leverkusen[0.041]; Leverkusen[0.041]; Bremen[0.041]; Bremen[0.041]; Karlsruhe[0.041]; Karlsruhe[0.041]; played[0.041]; played[0.041]; Arminia[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = FC Hansa Rostock\u001b[39m ==> ENTITY: \u001b[32mF.C. Hansa Rostock\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBremen[0.049]; Cologne[0.049]; Borussia[0.048]; VfL[0.048]; Duisburg[0.048]; Monchengladbach[0.048]; Bielefeld[0.047]; Bochum[0.047]; Freiburg[0.047]; Dusseldorf[0.046]; Werder[0.046]; Arminia[0.046]; Sc[0.046]; Munich[0.045]; Schalke[0.045]; Sv[0.044]; Sv[0.044]; Fc[0.042]; Fc[0.042]; Fortuna[0.041]; Hamburger[0.041]; Pauli[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Karlsruhe ==> ENTITY: \u001b[32mKarlsruher SC\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.248:0.248[0]; log p(e|m)= -2.226:-2.226[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.050]; soccer[0.048]; matches[0.048]; VfL[0.047]; Borussia[0.046]; Bielefeld[0.046]; Bremen[0.046]; Bremen[0.046]; Dortmund[0.045]; Stuttgart[0.045]; Arminia[0.045]; played[0.045]; played[0.045]; lost[0.045]; Leverkusen[0.045]; Leverkusen[0.045]; division[0.044]; Freiburg[0.044]; Bayer[0.044]; Bayer[0.044]; won[0.044]; VfB[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hansa Rostock\u001b[39m ==> ENTITY: \u001b[32mF.C. Hansa Rostock\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.236:0.236[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.046]; matches[0.045]; Soccer[0.045]; Bremen[0.045]; Bremen[0.045]; Cologne[0.044]; Borussia[0.044]; VfL[0.044]; Stuttgart[0.044]; VfB[0.043]; Duisburg[0.043]; Dortmund[0.043]; played[0.043]; played[0.043]; Bochum[0.043]; Bochum[0.043]; Freiburg[0.043]; Leverkusen[0.043]; Leverkusen[0.043]; German[0.042]; German[0.042]; Werder[0.042]; Werder[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = 1860 Munich\u001b[39m ==> ENTITY: \u001b[32mTSV 1860 München\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.239:0.239[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.044]; Soccer[0.043]; matches[0.043]; Stuttgart[0.043]; VfL[0.043]; Borussia[0.042]; Bochum[0.042]; Bochum[0.042]; won[0.042]; Cologne[0.042]; Freiburg[0.041]; VfB[0.041]; Leverkusen[0.041]; Leverkusen[0.041]; played[0.041]; played[0.041]; Dortmund[0.041]; Munich[0.041]; Werder[0.041]; Werder[0.041]; goals[0.040]; goals[0.040]; Bremen[0.040]; Bremen[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fortuna Dusseldorf\u001b[39m ==> ENTITY: \u001b[32mFortuna Düsseldorf\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVfL[0.045]; Monchengladbach[0.044]; Borussia[0.043]; Borussia[0.043]; Stuttgart[0.043]; Arminia[0.043]; VfB[0.043]; Bielefeld[0.043]; Dortmund[0.043]; Duisburg[0.042]; Bochum[0.042]; Bremen[0.042]; Cologne[0.042]; Werder[0.042]; Sc[0.041]; Freiburg[0.041]; Karlsruhe[0.041]; Schalke[0.040]; Munich[0.040]; Munich[0.040]; Rostock[0.040]; Sv[0.039]; Sv[0.039]; Hamburger[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.135:0.135[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.049]; German[0.048]; soccer[0.046]; played[0.045]; played[0.045]; Soccer[0.045]; VfL[0.045]; Munich[0.044]; Munich[0.044]; Stuttgart[0.043]; Cologne[0.043]; Dortmund[0.042]; won[0.042]; Bochum[0.042]; Bochum[0.042]; Leverkusen[0.042]; Leverkusen[0.042]; goals[0.042]; goals[0.042]; VfB[0.042]; Bremen[0.042]; Borussia[0.041]; lost[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Werder Bremen\u001b[39m ==> ENTITY: \u001b[32mSV Werder Bremen\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.227:0.227[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVfL[0.049]; Stuttgart[0.047]; Monchengladbach[0.047]; Leverkusen[0.046]; Dortmund[0.046]; Borussia[0.046]; Borussia[0.046]; Arminia[0.045]; Bremen[0.045]; Werder[0.045]; VfB[0.045]; Schalke[0.045]; Schalke[0.045]; Bochum[0.045]; Bayern[0.045]; played[0.045]; Bielefeld[0.044]; goals[0.044]; goals[0.044]; Freiburg[0.044]; Freiburg[0.044]; Cologne[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bayern Munich\u001b[39m ==> ENTITY: \u001b[32mFC Bayern Munich\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.231:0.231[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.047]; soccer[0.046]; Soccer[0.045]; VfL[0.045]; VfB[0.044]; played[0.044]; played[0.044]; Stuttgart[0.044]; Werder[0.044]; Werder[0.044]; Leverkusen[0.043]; Leverkusen[0.043]; Schalke[0.043]; Schalke[0.043]; Cologne[0.043]; Dortmund[0.043]; won[0.043]; Bayer[0.042]; Bayer[0.042]; Borussia[0.042]; Division[0.042]; Bochum[0.042]; Bochum[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Karlsruhe ==> ENTITY: \u001b[32mKarlsruher SC\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.242:0.242[0]; log p(e|m)= -2.226:-2.226[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.046]; soccer[0.044]; matches[0.044]; VfL[0.044]; Borussia[0.042]; Bremen[0.042]; Bremen[0.042]; Dortmund[0.042]; Stuttgart[0.042]; played[0.042]; played[0.042]; lost[0.041]; Leverkusen[0.041]; Leverkusen[0.041]; division[0.041]; Freiburg[0.041]; Bayer[0.041]; Bayer[0.041]; won[0.041]; VfB[0.040]; Munich[0.040]; Munich[0.040]; Werder[0.040]; Werder[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Borussia Monchengladbach\u001b[39m ==> ENTITY: \u001b[32mBorussia Mönchengladbach\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.132:0.132[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBremen[0.051]; Cologne[0.050]; Werder[0.050]; Bielefeld[0.050]; Sc[0.050]; Arminia[0.049]; Freiburg[0.049]; Schalke[0.049]; Duisburg[0.049]; Dusseldorf[0.048]; Munich[0.048]; Rostock[0.048]; Sv[0.047]; Sv[0.047]; Hansa[0.046]; Fc[0.045]; Fc[0.045]; Fc[0.045]; Fortuna[0.045]; Hamburger[0.044]; Pauli[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = VfL Bochum\u001b[39m ==> ENTITY: \u001b[32mVfL Bochum\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.237:0.237[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.045]; matches[0.045]; soccer[0.044]; Dortmund[0.043]; Bielefeld[0.043]; Bochum[0.043]; Freiburg[0.042]; Borussia[0.042]; played[0.042]; played[0.042]; Stuttgart[0.042]; Cologne[0.042]; VfB[0.042]; Leverkusen[0.041]; Leverkusen[0.041]; Arminia[0.041]; Bremen[0.040]; Bremen[0.040]; Werder[0.040]; Werder[0.040]; Friday[0.040]; Dusseldorf[0.040]; Karlsruhe[0.040]; Karlsruhe[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hamburger SV\u001b[39m ==> ENTITY: \u001b[32mHamburger SV\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.232:0.232[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVfL[0.049]; Stuttgart[0.049]; Dortmund[0.048]; Monchengladbach[0.048]; Borussia[0.047]; Borussia[0.047]; Cologne[0.047]; Bochum[0.047]; Bielefeld[0.046]; Bremen[0.046]; Schalke[0.045]; Munich[0.045]; Arminia[0.045]; Werder[0.045]; Karlsruhe[0.045]; Dusseldorf[0.045]; Freiburg[0.044]; Sc[0.044]; Duisburg[0.044]; Rostock[0.043]; Sv[0.042]; Hansa[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Borussia Dortmund\u001b[39m ==> ENTITY: \u001b[32mBorussia Dortmund\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStuttgart[0.047]; matches[0.045]; German[0.045]; German[0.045]; Soccer[0.044]; Cologne[0.044]; VfL[0.044]; soccer[0.043]; Bochum[0.043]; Bochum[0.043]; Leverkusen[0.043]; Leverkusen[0.043]; Munich[0.043]; Munich[0.043]; Munich[0.043]; Bremen[0.043]; Bremen[0.043]; Karlsruhe[0.043]; Karlsruhe[0.043]; Duisburg[0.042]; Dusseldorf[0.042]; Freiburg[0.042]; VfB[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Schalke\u001b[39m ==> ENTITY: \u001b[32mFC Schalke 04\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.234:0.234[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.046]; VfL[0.044]; Stuttgart[0.043]; soccer[0.043]; Soccer[0.043]; German[0.042]; German[0.042]; Borussia[0.042]; Bremen[0.041]; Bremen[0.041]; played[0.041]; played[0.041]; Dortmund[0.041]; Leverkusen[0.041]; Leverkusen[0.041]; VfB[0.041]; Munich[0.041]; Munich[0.041]; Bochum[0.041]; Bochum[0.041]; Werder[0.040]; Werder[0.040]; Cologne[0.040]; Schalke[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.125:0.125[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.047]; German[0.046]; soccer[0.045]; played[0.043]; played[0.043]; Soccer[0.043]; VfL[0.043]; Munich[0.043]; Munich[0.043]; Stuttgart[0.041]; Dortmund[0.041]; won[0.041]; Bochum[0.041]; Bochum[0.041]; Leverkusen[0.041]; Leverkusen[0.041]; goals[0.040]; goals[0.040]; VfB[0.040]; Bremen[0.040]; Borussia[0.040]; lost[0.040]; Bayer[0.039]; Bayer[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bochum ==> ENTITY: \u001b[32mVfL Bochum\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.237:0.237[0]; log p(e|m)= -1.510:-1.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.045]; matches[0.045]; soccer[0.044]; VfL[0.044]; Dortmund[0.043]; Bochum[0.043]; Freiburg[0.042]; Borussia[0.042]; played[0.042]; played[0.042]; Stuttgart[0.042]; Cologne[0.042]; VfB[0.042]; Leverkusen[0.041]; Leverkusen[0.041]; Munich[0.041]; Munich[0.041]; German[0.040]; German[0.040]; Bremen[0.040]; Friday[0.040]; Karlsruhe[0.040]; Karlsruhe[0.040]; Werder[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = MSV Duisburg\u001b[39m ==> ENTITY: \u001b[32mMSV Duisburg\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.240:0.240[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVfL[0.044]; Stuttgart[0.044]; Bochum[0.044]; Monchengladbach[0.043]; Dortmund[0.043]; Bielefeld[0.043]; Leverkusen[0.042]; Cologne[0.042]; Borussia[0.042]; Borussia[0.042]; Bremen[0.042]; Freiburg[0.042]; Freiburg[0.042]; played[0.042]; VfB[0.041]; Bayer[0.041]; Dusseldorf[0.040]; Arminia[0.040]; Munich[0.040]; Munich[0.040]; Karlsruhe[0.040]; Werder[0.040]; Sc[0.040]; won[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = SC Freiburg\u001b[39m ==> ENTITY: \u001b[32mSC Freiburg\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMonchengladbach[0.059]; Borussia[0.058]; Bielefeld[0.056]; Werder[0.055]; Arminia[0.055]; Duisburg[0.055]; Schalke[0.054]; Bremen[0.054]; Rostock[0.052]; Dusseldorf[0.052]; Munich[0.052]; Fc[0.051]; Fc[0.051]; Sv[0.051]; Sv[0.051]; Pauli[0.050]; Hansa[0.049]; Fortuna[0.048]; Hamburger[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = 1. FC Cologne\u001b[39m ==> ENTITY: \u001b[32m1. FC Köln\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.236:0.236[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVfL[0.049]; matches[0.047]; soccer[0.047]; Dortmund[0.046]; Stuttgart[0.046]; Monchengladbach[0.046]; Borussia[0.046]; Borussia[0.046]; Bielefeld[0.046]; Leverkusen[0.046]; Leverkusen[0.046]; VfB[0.045]; played[0.045]; played[0.045]; Bochum[0.045]; Bochum[0.045]; Duisburg[0.045]; Bremen[0.044]; Bremen[0.044]; Freiburg[0.044]; Werder[0.044]; Werder[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bayer Leverkusen\u001b[39m ==> ENTITY: \u001b[32mBayer 04 Leverkusen\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.227:0.227[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBayer[0.052]; matches[0.046]; Soccer[0.046]; soccer[0.045]; Leverkusen[0.044]; VfL[0.044]; Stuttgart[0.044]; division[0.044]; Dortmund[0.043]; Karlsruhe[0.043]; Karlsruhe[0.043]; Borussia[0.043]; Standings[0.043]; Standings[0.043]; Werder[0.043]; Werder[0.043]; Schalke[0.042]; Schalke[0.042]; Freiburg[0.042]; Bochum[0.042]; Bochum[0.042]; VfB[0.042]; Bayern[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arminia Bielefeld\u001b[39m ==> ENTITY: \u001b[32mArminia Bielefeld\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVfL[0.047]; Monchengladbach[0.046]; Bochum[0.046]; Borussia[0.046]; Dortmund[0.045]; Cologne[0.045]; Freiburg[0.044]; Bremen[0.043]; Karlsruhe[0.043]; Werder[0.043]; Schalke[0.043]; Duisburg[0.043]; Dusseldorf[0.041]; Munich[0.041]; Sc[0.041]; Rostock[0.040]; Sv[0.039]; Sv[0.039]; Hansa[0.039]; Fc[0.038]; Fc[0.038]; Fc[0.038]; Fortuna[0.037]; Hamburger[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bayer Leverkusen\u001b[39m ==> ENTITY: \u001b[32mBayer 04 Leverkusen\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.224:0.224[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBayer[0.050]; matches[0.044]; Soccer[0.044]; soccer[0.043]; Leverkusen[0.043]; VfL[0.042]; Stuttgart[0.042]; division[0.042]; Dortmund[0.042]; Karlsruhe[0.041]; Karlsruhe[0.041]; Borussia[0.041]; Standings[0.041]; Standings[0.041]; Werder[0.041]; Schalke[0.041]; Schalke[0.041]; Freiburg[0.040]; Bochum[0.040]; Bochum[0.040]; VfB[0.040]; Bayern[0.040]; won[0.040]; Bremen[0.040]; \t\n\r [========================>.....................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 10s127ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2527/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1334testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Shanghai ==> ENTITY: \u001b[32mShanghai\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.101:-0.101[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChinese[0.049]; China[0.047]; Tianjin[0.045]; Hui[0.045]; Wen[0.043]; city[0.043]; announced[0.042]; Bao[0.042]; Friday[0.042]; nearly[0.042]; small[0.042]; having[0.041]; immediately[0.041]; hours[0.040]; people[0.040]; said[0.040]; said[0.040]; said[0.040]; newspaper[0.040]; newspaper[0.040]; east[0.039]; difficulty[0.039]; relatives[0.039]; party[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.120:0.120[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChinese[0.050]; nearly[0.045]; Tianjin[0.045]; city[0.044]; east[0.044]; Bao[0.044]; Shanghai[0.044]; Wen[0.043]; old[0.043]; Hui[0.043]; small[0.042]; having[0.041]; Friday[0.040]; said[0.040]; said[0.040]; said[0.040]; people[0.040]; hours[0.039]; announced[0.039]; relatives[0.039]; father[0.039]; grandfather[0.039]; difficulty[0.039]; birthday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tianjin\u001b[39m ==> ENTITY: \u001b[32mTianjin\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.048]; Shanghai[0.047]; Chinese[0.046]; nearly[0.044]; city[0.043]; Wen[0.043]; hours[0.043]; small[0.042]; people[0.042]; Bao[0.042]; room[0.042]; Friday[0.041]; east[0.041]; having[0.041]; choked[0.040]; difficulty[0.040]; Hui[0.040]; birthday[0.040]; relatives[0.039]; said[0.039]; said[0.039]; said[0.039]; immediately[0.039]; party[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wen Hui Bao\u001b[39m ==> ENTITY: \u001b[32mWen Hui Bao\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.109:0.109[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnewspaper[0.051]; newspaper[0.051]; Chinese[0.049]; China[0.048]; Shanghai[0.048]; hospital[0.046]; Tianjin[0.046]; announced[0.045]; Friday[0.043]; rushed[0.043]; hours[0.042]; immediately[0.042]; said[0.042]; said[0.042]; said[0.042]; party[0.041]; nearly[0.041]; breathing[0.041]; having[0.041]; difficulty[0.040]; choked[0.040]; smoke[0.040]; smoke[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chinese ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.121:0.121[0]; log p(e|m)= -1.419:-1.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.051]; Hui[0.046]; Shanghai[0.045]; relatives[0.044]; Tianjin[0.044]; people[0.043]; Wen[0.043]; Bao[0.043]; nearly[0.042]; small[0.042]; grandfather[0.041]; father[0.041]; said[0.041]; said[0.041]; said[0.041]; city[0.040]; having[0.040]; east[0.039]; difficulty[0.039]; hours[0.039]; Friday[0.039]; newspaper[0.039]; newspaper[0.039]; party[0.039]; \t\n\r [========================>.....................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 10s128ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2532/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1251testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.057]; Russia[0.056]; Russian[0.054]; Konstantin[0.049]; Friday[0.048]; Chernyshev[0.048]; Alexander[0.047]; regional[0.047]; Minister[0.047]; Livshits[0.047]; Livshits[0.047]; Livshits[0.047]; official[0.046]; Tass[0.046]; said[0.046]; said[0.046]; agency[0.045]; financially[0.045]; overdue[0.045]; frozen[0.044]; federal[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.126:0.126[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.049]; Moscow[0.047]; Moscow[0.047]; authorities[0.043]; group[0.043]; prominent[0.042]; North[0.042]; Livshits[0.042]; Livshits[0.042]; words[0.041]; Tass[0.041]; Bank[0.041]; cabinet[0.040]; said[0.040]; said[0.040]; Chernyshev[0.040]; firms[0.040]; official[0.040]; given[0.040]; Christopher[0.040]; money[0.040]; chief[0.040]; quoted[0.040]; agency[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.124:0.124[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.047]; officials[0.044]; authorities[0.043]; month[0.042]; Livshits[0.042]; worked[0.042]; Duma[0.042]; regional[0.042]; regional[0.042]; regional[0.042]; Tass[0.041]; September[0.041]; threatened[0.041]; Krasnoyarsk[0.041]; led[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; seen[0.041]; total[0.041]; Chernyshev[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = International Monetary Fund\u001b[39m ==> ENTITY: \u001b[32mInternational Monetary Fund\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.112:0.112[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBank[0.046]; billion[0.044]; payments[0.043]; budget[0.043]; agency[0.043]; revenues[0.042]; essentially[0.042]; long[0.042]; bolt[0.042]; economist[0.042]; term[0.041]; firms[0.041]; official[0.041]; United[0.041]; tax[0.041]; tax[0.041]; tax[0.041]; based[0.041]; pay[0.041]; pay[0.041]; loan[0.041]; statement[0.041]; boost[0.041]; money[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Itar-Tass\u001b[39m ==> ENTITY: \u001b[32mInformation Telegraph Agency of Russia\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nagency[0.047]; news[0.047]; Tass[0.046]; Russian[0.046]; Russian[0.046]; Russia[0.045]; Moscow[0.045]; Moscow[0.045]; Moscow[0.045]; official[0.044]; statement[0.043]; bolt[0.042]; International[0.042]; quoted[0.042]; response[0.042]; regional[0.042]; company[0.042]; authorities[0.042]; parliamentary[0.042]; equities[0.042]; Bank[0.042]; Chernyshev[0.041]; cabinet[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.104:0.104[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.048]; Moscow[0.047]; Moscow[0.047]; Konstantin[0.044]; Alexander[0.043]; said[0.042]; said[0.042]; Minister[0.042]; regional[0.042]; Chernyshev[0.042]; official[0.041]; Livshits[0.041]; Livshits[0.041]; pay[0.041]; federal[0.041]; warned[0.040]; company[0.040]; overdue[0.040]; assets[0.040]; expected[0.040]; frozen[0.039]; big[0.039]; financially[0.039]; liquidated[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Norilsk Nickel\u001b[39m ==> ENTITY: \u001b[32mNorilsk Nickel\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorilsk[0.049]; Norilsk[0.049]; Norilsk[0.049]; company[0.046]; Nikel[0.043]; assets[0.041]; Moscow[0.041]; Moscow[0.041]; Chernyshev[0.041]; Russian[0.041]; agency[0.041]; Russia[0.040]; pay[0.040]; pay[0.040]; firm[0.040]; financially[0.040]; warned[0.040]; regional[0.040]; bankruptcy[0.040]; equities[0.040]; troubled[0.039]; liquidate[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.124:0.124[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.047]; Moscow[0.046]; Moscow[0.046]; Moscow[0.046]; government[0.045]; officials[0.042]; authorities[0.042]; North[0.041]; cabinet[0.041]; Bank[0.041]; United[0.041]; statement[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; group[0.040]; International[0.040]; declined[0.040]; Chernyshev[0.040]; hundreds[0.040]; pay[0.040]; given[0.040]; chief[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.049]; Moscow[0.049]; Russian[0.046]; Russia[0.046]; given[0.042]; Konstantin[0.042]; Tass[0.042]; Granville[0.042]; City[0.041]; Chernyshev[0.041]; agency[0.041]; procedures[0.040]; Livshits[0.040]; Livshits[0.040]; pay[0.040]; pay[0.040]; official[0.040]; cabinet[0.040]; authorities[0.040]; essentially[0.040]; Bank[0.040]; industrial[0.040]; response[0.040]; chief[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.122:0.122[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.050]; Moscow[0.047]; Moscow[0.047]; Konstantin[0.044]; Alexander[0.043]; Tass[0.043]; Livshits[0.042]; Livshits[0.042]; Livshits[0.042]; regional[0.041]; words[0.041]; Chernyshev[0.041]; said[0.040]; said[0.040]; parliamentary[0.040]; official[0.040]; troubled[0.040]; quoted[0.040]; agency[0.040]; big[0.039]; pay[0.039]; news[0.039]; firm[0.039]; Minister[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.049]; Moscow[0.049]; Russia[0.049]; Russian[0.047]; Konstantin[0.042]; Friday[0.042]; Tass[0.042]; Granville[0.042]; City[0.042]; Chernyshev[0.041]; Alexander[0.041]; regional[0.041]; Livshits[0.040]; Livshits[0.040]; Livshits[0.040]; Bank[0.040]; response[0.040]; official[0.039]; agency[0.039]; financially[0.039]; Norilsk[0.039]; Norilsk[0.039]; Norilsk[0.039]; Norilsk[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mKrasnoyarsk\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mKrasnoyarsk\u001b[39m <---> \u001b[32mKrasnoyarsk Krai\u001b[39m\t\nSCORES: global= 0.288:0.269[\u001b[31m0.018\u001b[39m]; local()= 0.136:0.144[\u001b[32m0.009\u001b[39m]; log p(e|m)= 0.000:-1.231[\u001b[31m1.231\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.045]; money[0.043]; regional[0.043]; regional[0.043]; regional[0.043]; Chernyshev[0.043]; Tass[0.042]; near[0.042]; near[0.042]; total[0.042]; Norilsk[0.041]; Norilsk[0.041]; Norilsk[0.041]; Norilsk[0.041]; Norilsk[0.041]; Norilsk[0.041]; Norilsk[0.041]; Norilsk[0.041]; future[0.041]; authorities[0.041]; Livshits[0.040]; month[0.040]; worked[0.040]; estimated[0.040]; \t\n\r [=========================>....................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 10s138ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2546/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1217testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Polish ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.072:0.072[0]; log p(e|m)= -0.814:-0.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarsaw[0.048]; despite[0.043]; Steven[0.043]; President[0.043]; months[0.043]; recent[0.043]; Zywiec[0.042]; Zywiec[0.042]; Zywiec[0.042]; said[0.042]; Friday[0.041]; rise[0.041]; high[0.041]; having[0.041]; market[0.041]; interview[0.041]; post[0.040]; firm[0.040]; firm[0.040]; firm[0.040]; percent[0.040]; percent[0.040]; percent[0.040]; significant[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.155:0.155[0]; log p(e|m)= -1.207:-1.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.046]; beer[0.044]; beer[0.044]; Polish[0.044]; Czech[0.043]; Czech[0.043]; Czech[0.043]; Okocim[0.043]; Zywiec[0.042]; Zywiec[0.042]; Zywiec[0.042]; beers[0.042]; beers[0.042]; beers[0.042]; beers[0.042]; make[0.041]; Heineken[0.040]; countries[0.040]; brewers[0.040]; start[0.040]; speaking[0.039]; Republic[0.039]; say[0.039]; Van[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = SAB ==> ENTITY: \u001b[32mSouth African Breweries\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.238:0.238[0]; log p(e|m)= -0.312:-0.312[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncompany[0.045]; Heineken[0.044]; Breweries[0.044]; brewery[0.044]; beer[0.043]; beer[0.043]; brewers[0.042]; brewers[0.042]; brewers[0.042]; brewers[0.042]; country[0.042]; African[0.042]; market[0.041]; market[0.041]; market[0.041]; market[0.041]; Carlsberg[0.041]; owned[0.041]; firm[0.041]; South[0.040]; Okocim[0.040]; currently[0.040]; hold[0.040]; Europe[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.100:0.100[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.048]; country[0.048]; Europe[0.046]; Australia[0.045]; Poland[0.043]; Poland[0.043]; Poland[0.043]; Germany[0.042]; foreign[0.042]; South[0.042]; small[0.041]; population[0.041]; African[0.041]; large[0.041]; currently[0.040]; million[0.040]; certainly[0.040]; largest[0.040]; strategic[0.040]; brewery[0.039]; beer[0.039]; beer[0.039]; beer[0.039]; beer[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPolish[0.050]; Warsaw[0.050]; start[0.050]; country[0.049]; win[0.047]; international[0.047]; war[0.046]; Okocim[0.045]; recent[0.045]; Heineken[0.045]; Heineken[0.045]; boosted[0.045]; zlotys[0.044]; Zywiec[0.044]; Zywiec[0.044]; Zywiec[0.044]; Zywiec[0.044]; accounts[0.044]; overall[0.044]; short[0.043]; beer[0.043]; beer[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.140:0.140[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.049]; Polish[0.046]; Polish[0.046]; country[0.045]; country[0.045]; Germany[0.044]; Europe[0.043]; Lech[0.042]; Tychy[0.041]; competition[0.041]; Okocim[0.041]; Britain[0.041]; Heineken[0.040]; zlotys[0.040]; level[0.040]; Zywiec[0.040]; certainly[0.040]; switch[0.039]; small[0.039]; currently[0.039]; say[0.039]; fall[0.039]; young[0.039]; think[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.168:0.168[0]; log p(e|m)= -1.207:-1.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.046]; Poland[0.046]; beer[0.044]; beer[0.044]; Polish[0.044]; Okocim[0.043]; Czech[0.042]; Czech[0.042]; Czech[0.042]; Zywiec[0.042]; Zywiec[0.042]; Zywiec[0.042]; beers[0.041]; beers[0.041]; beers[0.041]; beers[0.041]; make[0.041]; brewery[0.040]; countries[0.040]; speaking[0.039]; larges[0.039]; Republic[0.039]; say[0.039]; brewers[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Heineken\u001b[39m ==> ENTITY: \u001b[32mHeineken\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.266:0.266[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHeineken[0.048]; brand[0.044]; brand[0.044]; beer[0.044]; beer[0.044]; beer[0.044]; brands[0.042]; brands[0.042]; Zywiec[0.041]; Zywiec[0.041]; Zywiec[0.041]; Zywiec[0.041]; beers[0.041]; beers[0.041]; beers[0.041]; beers[0.041]; Carlsberg[0.041]; country[0.041]; Okocim[0.040]; drinks[0.040]; recent[0.040]; Poland[0.040]; Poland[0.040]; sales[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.100:0.100[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.047]; Britain[0.044]; South[0.044]; Germany[0.043]; Europe[0.043]; Poland[0.043]; Poland[0.043]; million[0.042]; million[0.042]; beer[0.042]; beer[0.042]; African[0.042]; position[0.041]; Heineken[0.041]; competition[0.041]; competition[0.041]; Republic[0.041]; line[0.040]; decade[0.040]; say[0.040]; compete[0.040]; Breweries[0.040]; added[0.040]; costs[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mHeineken\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHeineken\u001b[39m <---> \u001b[32mHeineken International\u001b[39m\t\nSCORES: global= 0.274:0.259[\u001b[31m0.016\u001b[39m]; local()= 0.224:0.207[\u001b[31m0.017\u001b[39m]; log p(e|m)= 0.000:-1.336[\u001b[31m1.336\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbeer[0.046]; beer[0.046]; beer[0.046]; brewery[0.045]; Europe[0.045]; brewers[0.044]; brewers[0.044]; brewers[0.044]; brewers[0.044]; Zywiec[0.044]; Zywiec[0.044]; Carlsberg[0.043]; country[0.043]; company[0.043]; Breweries[0.043]; Okocim[0.043]; Britain[0.043]; Poland[0.042]; Poland[0.042]; Poland[0.042]; population[0.042]; advertising[0.041]; million[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Europe ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.046]; country[0.044]; country[0.044]; Poland[0.044]; Poland[0.044]; competition[0.044]; level[0.043]; market[0.042]; market[0.042]; market[0.042]; market[0.042]; Germany[0.041]; Australia[0.041]; Heineken[0.040]; population[0.040]; growth[0.040]; Polish[0.040]; Polish[0.040]; percent[0.040]; percent[0.040]; percent[0.040]; profits[0.040]; sales[0.040]; foreign[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Polish ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.814:-0.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.050]; Republic[0.044]; Czech[0.043]; Czech[0.043]; Czech[0.043]; Czech[0.043]; decade[0.042]; countries[0.042]; factor[0.041]; Okocim[0.041]; accounts[0.041]; speaking[0.041]; Zywiec[0.041]; Zywiec[0.041]; Zywiec[0.041]; despite[0.041]; compete[0.041]; largely[0.040]; million[0.040]; beer[0.040]; beer[0.040]; added[0.040]; general[0.040]; says[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.119:0.119[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninterview[0.045]; told[0.043]; market[0.042]; market[0.042]; million[0.042]; million[0.042]; million[0.042]; firm[0.042]; firm[0.042]; firm[0.042]; leading[0.042]; percent[0.042]; percent[0.042]; percent[0.042]; share[0.042]; share[0.042]; investment[0.041]; sales[0.041]; sales[0.041]; sales[0.041]; said[0.041]; said[0.041]; recent[0.040]; months[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.050]; Polish[0.046]; Republic[0.044]; Czech[0.043]; Czech[0.043]; Czech[0.043]; Czech[0.043]; despite[0.042]; competition[0.041]; Okocim[0.041]; added[0.041]; start[0.041]; Heineken[0.041]; limit[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; line[0.040]; Zywiec[0.040]; Zywiec[0.040]; Zywiec[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Okocim\u001b[39m ==> ENTITY: \u001b[32mOkocim\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.231:0.231[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbrewery[0.048]; Poland[0.047]; Poland[0.047]; Poland[0.047]; Zywiec[0.047]; Zywiec[0.047]; Carlsberg[0.046]; beer[0.046]; beer[0.046]; beer[0.046]; brewers[0.046]; brewers[0.046]; brewers[0.046]; brewers[0.046]; Polish[0.045]; Heineken[0.045]; Breweries[0.044]; country[0.044]; Lech[0.043]; Tychy[0.043]; Europe[0.043]; company[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.046]; Poland[0.046]; Poland[0.046]; Europe[0.045]; country[0.044]; country[0.044]; million[0.043]; Britain[0.043]; Polish[0.042]; Polish[0.042]; competition[0.041]; bought[0.041]; Australia[0.041]; reached[0.040]; Heineken[0.040]; level[0.040]; foreign[0.040]; beer[0.040]; beer[0.040]; beer[0.040]; beer[0.040]; say[0.039]; South[0.039]; rise[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Polish ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.814:-0.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.053]; Warsaw[0.048]; country[0.048]; Czech[0.045]; war[0.045]; international[0.044]; factor[0.043]; Okocim[0.043]; accounts[0.043]; overall[0.043]; recent[0.043]; Zywiec[0.043]; Zywiec[0.043]; Zywiec[0.043]; Zywiec[0.043]; million[0.042]; beer[0.042]; beer[0.042]; think[0.041]; largest[0.041]; beers[0.041]; beers[0.041]; beers[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.141:0.141[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.049]; Poland[0.049]; country[0.045]; Germany[0.044]; Republic[0.043]; Czech[0.043]; Czech[0.043]; Lech[0.042]; Tychy[0.041]; competition[0.041]; competition[0.041]; Okocim[0.041]; Britain[0.041]; added[0.040]; beer[0.040]; beer[0.040]; Heineken[0.040]; decade[0.040]; Zywiec[0.040]; Zywiec[0.040]; Australia[0.039]; compete[0.039]; tough[0.039]; small[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.173:0.173[0]; log p(e|m)= -1.207:-1.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.045]; country[0.044]; beer[0.043]; beer[0.043]; beer[0.043]; Polish[0.043]; Polish[0.043]; Czech[0.043]; Czech[0.043]; Czech[0.043]; Okocim[0.042]; Zywiec[0.042]; Zywiec[0.042]; Zywiec[0.042]; beers[0.041]; beers[0.041]; beers[0.041]; beers[0.041]; win[0.040]; Heineken[0.040]; Heineken[0.040]; countries[0.039]; brewers[0.039]; start[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tychy\u001b[39m ==> ENTITY: \u001b[32mTychy\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.047]; Poland[0.047]; Poland[0.047]; Europe[0.043]; Okocim[0.042]; Lech[0.042]; Germany[0.042]; brewery[0.041]; company[0.041]; compete[0.041]; country[0.041]; Boxmeer[0.041]; Boxmeer[0.041]; brewers[0.041]; brewers[0.041]; brewers[0.041]; brewers[0.041]; brewers[0.041]; Sab[0.041]; strategic[0.040]; Bv[0.040]; Carlsberg[0.040]; Zywiec[0.040]; Zywiec[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.049]; Poland[0.049]; Polish[0.045]; Polish[0.045]; country[0.044]; country[0.044]; Germany[0.044]; Europe[0.042]; Lech[0.042]; Tychy[0.041]; competition[0.041]; Okocim[0.041]; Britain[0.041]; Heineken[0.040]; Zywiec[0.040]; currently[0.040]; certainly[0.039]; numerous[0.039]; small[0.039]; fall[0.039]; level[0.039]; think[0.039]; Australia[0.039]; large[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Warsaw ==> ENTITY: \u001b[32mWarsaw\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.073:0.073[0]; log p(e|m)= -0.143:-0.143[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPolish[0.046]; recent[0.044]; Friday[0.044]; high[0.043]; percent[0.043]; percent[0.043]; percent[0.043]; months[0.043]; firm[0.042]; firm[0.042]; firm[0.042]; far[0.041]; rise[0.041]; fell[0.041]; slightly[0.041]; Rise[0.041]; chief[0.041]; defend[0.040]; significant[0.040]; despite[0.040]; executive[0.040]; make[0.040]; help[0.040]; Reuters[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Polish ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.814:-0.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.049]; Poland[0.049]; Britain[0.048]; Polish[0.048]; country[0.045]; Europe[0.043]; Germany[0.043]; population[0.042]; despite[0.041]; young[0.040]; Zywiec[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; million[0.039]; million[0.039]; million[0.039]; million[0.039]; fell[0.039]; beer[0.039]; beer[0.039]; beer[0.039]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = CEFTA\u001b[39m ==> ENTITY: \u001b[32mCentral European Free Trade Agreement\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.092:0.092[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountries[0.047]; Poland[0.047]; Republic[0.046]; Okocim[0.043]; Boxmeer[0.042]; Boxmeer[0.042]; Boxmeer[0.042]; tariffs[0.042]; Czech[0.041]; Czech[0.041]; Czech[0.041]; Czech[0.041]; imports[0.041]; Carlsberg[0.041]; Polish[0.041]; Heineken[0.041]; position[0.041]; Company[0.040]; largely[0.040]; Zywiec[0.040]; Zywiec[0.040]; Zywiec[0.040]; general[0.040]; million[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Heineken\u001b[39m ==> ENTITY: \u001b[32mHeineken\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.241:0.241[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHeineken[0.053]; brand[0.049]; brand[0.049]; beer[0.048]; beer[0.048]; brands[0.046]; Zywiec[0.046]; Zywiec[0.046]; Zywiec[0.046]; Zywiec[0.046]; beers[0.045]; Carlsberg[0.045]; country[0.045]; Okocim[0.044]; drinks[0.044]; recent[0.044]; Poland[0.044]; sales[0.044]; sales[0.044]; selling[0.042]; boosted[0.042]; million[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech Republic ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.046]; Poland[0.046]; Czech[0.044]; Czech[0.044]; Czech[0.044]; Australia[0.043]; competition[0.042]; brands[0.042]; despite[0.042]; compete[0.042]; brewery[0.041]; countries[0.041]; Okocim[0.041]; beer[0.041]; Polish[0.041]; added[0.040]; start[0.040]; Zywiec[0.040]; Zywiec[0.040]; quality[0.040]; competitors[0.040]; hold[0.039]; brewers[0.039]; quite[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Carlsberg\u001b[39m ==> ENTITY: \u001b[32mCarlsberg Group\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.216:0.216[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbrewery[0.046]; Heineken[0.045]; beer[0.044]; beer[0.044]; beer[0.044]; Zywiec[0.043]; Zywiec[0.043]; Okocim[0.042]; brewers[0.042]; brewers[0.042]; brewers[0.042]; brewers[0.042]; competition[0.041]; competition[0.041]; Breweries[0.041]; company[0.041]; Europe[0.040]; Poland[0.040]; Poland[0.040]; Poland[0.040]; Britain[0.039]; Boxmeer[0.039]; Boxmeer[0.039]; Boxmeer[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Carlsberg\u001b[39m ==> ENTITY: \u001b[32mCarlsberg Group\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.208:0.208[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHeineken[0.044]; Heineken[0.044]; beer[0.044]; beer[0.044]; beer[0.044]; Zywiec[0.043]; Zywiec[0.043]; Zywiec[0.043]; Zywiec[0.043]; Okocim[0.042]; beers[0.041]; beers[0.041]; beers[0.041]; beers[0.041]; brand[0.041]; brand[0.041]; brands[0.041]; brands[0.041]; win[0.041]; Poland[0.039]; Poland[0.039]; boosted[0.039]; recent[0.039]; says[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Polish ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.814:-0.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.049]; Poland[0.049]; Britain[0.048]; Polish[0.047]; country[0.044]; country[0.044]; Europe[0.043]; Germany[0.043]; population[0.041]; Okocim[0.040]; large[0.040]; young[0.040]; Zywiec[0.040]; Zywiec[0.040]; small[0.039]; said[0.039]; said[0.039]; said[0.039]; said[0.039]; million[0.039]; million[0.039]; million[0.039]; million[0.039]; foreign[0.039]; \t\n\r [=========================>....................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 10s11ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2575/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1216testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swiss ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.053]; Swiss[0.047]; Swiss[0.047]; Polish[0.045]; Friday[0.045]; Poland[0.044]; Poland[0.044]; Poland[0.044]; Poland[0.044]; Warsaw[0.043]; countries[0.043]; unveiling[0.043]; died[0.042]; said[0.042]; said[0.042]; news[0.042]; received[0.042]; received[0.042]; accounts[0.042]; accounts[0.042]; accounts[0.042]; War[0.042]; War[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rosati\u001b[39m ==> ENTITY: \u001b[32mDariusz Rosati\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.051]; Poland[0.051]; communist[0.048]; Warsaw[0.047]; Soviet[0.047]; Polish[0.047]; communists[0.046]; international[0.046]; information[0.045]; Rosati[0.045]; Rosati[0.045]; Rosati[0.045]; data[0.045]; financial[0.044]; German[0.044]; francs[0.043]; francs[0.043]; francs[0.043]; Nazi[0.043]; government[0.043]; payment[0.043]; exports[0.043]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNazi\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNazism\u001b[39m <---> \u001b[32mNazi Party\u001b[39m\t\nSCORES: global= 0.261:0.235[\u001b[31m0.026\u001b[39m]; local()= 0.159:0.110[\u001b[31m0.049\u001b[39m]; log p(e|m)= -0.098:-2.333[\u001b[31m2.235\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJews[0.050]; Jews[0.050]; Poland[0.049]; Poland[0.049]; Poland[0.049]; Poland[0.049]; Poland[0.049]; German[0.046]; Warsaw[0.045]; Polish[0.045]; war[0.045]; Soviet[0.044]; War[0.044]; Switzerland[0.044]; Switzerland[0.044]; Switzerland[0.044]; Switzerland[0.044]; communists[0.044]; communist[0.042]; World[0.042]; authorities[0.042]; imposed[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Switzerland ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.180:0.180[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.050]; Switzerland[0.050]; international[0.044]; Poland[0.043]; Poland[0.043]; Swiss[0.043]; Swiss[0.043]; weeks[0.042]; commission[0.041]; Polish[0.041]; small[0.041]; citizens[0.041]; citizens[0.041]; citizens[0.041]; government[0.041]; German[0.040]; authorities[0.040]; World[0.040]; million[0.040]; million[0.040]; data[0.040]; said[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Switzerland ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.199:0.199[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.050]; Poland[0.043]; Poland[0.043]; Poland[0.043]; Poland[0.043]; Poland[0.043]; Poland[0.043]; Poland[0.043]; Swiss[0.042]; Swiss[0.042]; Swiss[0.042]; Swiss[0.042]; stands[0.042]; nationals[0.041]; countries[0.041]; Polish[0.040]; Polish[0.040]; government[0.040]; citizens[0.039]; citizens[0.039]; German[0.039]; received[0.039]; received[0.039]; World[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = World War Two\u001b[39m ==> ENTITY: \u001b[32mWorld War II\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.077:0.077[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.044]; Poland[0.044]; Poland[0.044]; Poland[0.044]; Poland[0.044]; countries[0.043]; communist[0.043]; Jews[0.043]; War[0.043]; Holocaust[0.042]; Warsaw[0.042]; Polish[0.041]; Polish[0.041]; Switzerland[0.041]; Switzerland[0.041]; government[0.041]; citizens[0.041]; Foreign[0.040]; money[0.040]; money[0.040]; accounts[0.039]; accounts[0.039]; accounts[0.039]; accounts[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swiss ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.052]; Switzerland[0.052]; Switzerland[0.052]; Switzerland[0.052]; Swiss[0.045]; German[0.045]; Poland[0.042]; Poland[0.042]; Poland[0.042]; Warsaw[0.042]; weeks[0.042]; Polish[0.042]; Polish[0.042]; finish[0.041]; nationals[0.041]; powerful[0.041]; million[0.041]; million[0.041]; said[0.041]; said[0.041]; news[0.041]; called[0.040]; War[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Polish ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.814:-0.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.051]; Poland[0.051]; Poland[0.051]; Poland[0.051]; Warsaw[0.046]; War[0.044]; War[0.044]; Jews[0.044]; communist[0.044]; Switzerland[0.042]; Switzerland[0.042]; government[0.042]; Foreign[0.042]; countries[0.042]; accounts[0.041]; accounts[0.041]; accounts[0.041]; Dariusz[0.041]; accused[0.041]; Marcin[0.041]; Accounts[0.040]; Friday[0.040]; news[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swiss ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.189:0.189[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.049]; Switzerland[0.049]; Switzerland[0.049]; Swiss[0.043]; Swiss[0.043]; German[0.042]; francs[0.042]; francs[0.042]; money[0.041]; money[0.041]; million[0.041]; sum[0.041]; Poland[0.040]; Poland[0.040]; Poland[0.040]; Poland[0.040]; Poland[0.040]; bank[0.040]; weeks[0.040]; Polish[0.039]; accounts[0.039]; accounts[0.039]; accounts[0.039]; finish[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mFDP\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFree Democratic Party (Germany)\u001b[39m <---> \u001b[32mFree Democratic Party of Switzerland\u001b[39m\t\nSCORES: global= 0.257:0.253[\u001b[31m0.003\u001b[39m]; local()= 0.077:0.097[\u001b[32m0.021\u001b[39m]; log p(e|m)= -0.420:-0.545[\u001b[31m0.125\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nconservative[0.045]; government[0.043]; Democrats[0.043]; parliament[0.043]; country[0.043]; small[0.043]; Switzerland[0.042]; Swiss[0.042]; Swiss[0.042]; Swiss[0.042]; Swiss[0.042]; Radical[0.041]; million[0.041]; million[0.041]; millions[0.041]; issue[0.041]; ask[0.041]; doubts[0.041]; Jewish[0.040]; Jewish[0.040]; Jewish[0.040]; groups[0.040]; work[0.040]; Polish[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swiss ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.050]; dollars[0.046]; Swiss[0.044]; Swiss[0.044]; Swiss[0.044]; country[0.044]; francs[0.042]; francs[0.042]; francs[0.042]; million[0.042]; million[0.042]; groups[0.040]; bank[0.040]; week[0.040]; Polish[0.040]; accounts[0.040]; accounts[0.040]; accounts[0.040]; international[0.040]; fund[0.040]; Jewish[0.040]; Jewish[0.040]; Jewish[0.040]; wealth[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rosati\u001b[39m ==> ENTITY: \u001b[32mDariusz Rosati\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.047]; Poland[0.047]; Poland[0.047]; communist[0.044]; Warsaw[0.043]; Soviet[0.043]; Polish[0.043]; Committee[0.042]; chairman[0.042]; communists[0.042]; information[0.041]; Rosati[0.041]; Rosati[0.041]; Rosati[0.041]; news[0.041]; German[0.040]; conference[0.040]; Senator[0.040]; francs[0.040]; francs[0.040]; francs[0.040]; Nazi[0.039]; payment[0.039]; exports[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Switzerland ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.209:0.209[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.049]; Switzerland[0.049]; Poland[0.043]; Poland[0.043]; Poland[0.043]; Poland[0.043]; Poland[0.043]; Poland[0.043]; Swiss[0.042]; Swiss[0.042]; Swiss[0.042]; Swiss[0.042]; stands[0.042]; nationals[0.041]; countries[0.041]; Polish[0.040]; Polish[0.040]; government[0.040]; citizens[0.039]; citizens[0.039]; citizens[0.039]; German[0.039]; received[0.039]; received[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.211:0.211[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.047]; Poland[0.047]; Poland[0.047]; Poland[0.047]; Poland[0.047]; Polish[0.043]; Polish[0.043]; Warsaw[0.043]; Switzerland[0.042]; Switzerland[0.042]; communist[0.041]; communist[0.041]; Marcin[0.040]; World[0.040]; War[0.039]; War[0.039]; Jews[0.039]; Jews[0.039]; war[0.039]; Dariusz[0.039]; government[0.039]; received[0.038]; received[0.038]; countries[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Switzerland ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.179:0.179[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.051]; Switzerland[0.051]; international[0.044]; Poland[0.044]; Swiss[0.043]; Swiss[0.043]; Swiss[0.043]; week[0.043]; weeks[0.042]; Polish[0.041]; small[0.041]; government[0.041]; citizens[0.040]; citizens[0.040]; World[0.040]; data[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; legal[0.039]; confirmed[0.039]; finish[0.039]; Warsaw[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jews ==> ENTITY: \u001b[32mJews\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.125:0.125[0]; log p(e|m)= -0.126:-0.126[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHolocaust[0.048]; Poland[0.046]; Poland[0.046]; Poland[0.046]; Poland[0.046]; Warsaw[0.043]; Polish[0.042]; countries[0.042]; Switzerland[0.042]; Switzerland[0.042]; accounts[0.041]; accounts[0.041]; accounts[0.041]; know[0.040]; settlements[0.040]; World[0.040]; government[0.040]; said[0.040]; said[0.040]; War[0.039]; War[0.039]; bank[0.039]; died[0.039]; sum[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mGerman\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mGermany\u001b[39m <---> \u001b[32mNazi Germany\u001b[39m\t\nSCORES: global= 0.261:0.228[\u001b[31m0.033\u001b[39m]; local()= 0.172:0.164[\u001b[31m0.008\u001b[39m]; log p(e|m)= -1.047:-2.551[\u001b[31m1.504\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.046]; Poland[0.046]; Poland[0.046]; Poland[0.046]; Polish[0.044]; Switzerland[0.043]; Switzerland[0.043]; Switzerland[0.043]; Switzerland[0.043]; Nazi[0.043]; Warsaw[0.042]; Jews[0.042]; Jews[0.042]; war[0.041]; Soviet[0.040]; War[0.040]; communist[0.039]; nationals[0.039]; communists[0.039]; invaders[0.039]; Swiss[0.039]; Swiss[0.039]; Swiss[0.039]; authorities[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Polish ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.092:0.092[0]; log p(e|m)= -0.814:-0.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.045]; Jewish[0.045]; Jewish[0.045]; War[0.044]; communists[0.043]; Switzerland[0.042]; Switzerland[0.042]; government[0.042]; international[0.042]; groups[0.041]; citizens[0.041]; citizens[0.041]; accounts[0.041]; accounts[0.041]; accounts[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; small[0.041]; authorities[0.041]; million[0.040]; million[0.040]; million[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Warsaw ==> ENTITY: \u001b[32mWarsaw\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.143:-0.143[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.048]; Poland[0.048]; Poland[0.048]; Poland[0.048]; Jews[0.045]; Polish[0.044]; Holocaust[0.043]; communist[0.042]; accused[0.041]; government[0.041]; settlements[0.040]; accounts[0.040]; accounts[0.040]; accounts[0.040]; Dariusz[0.040]; War[0.039]; War[0.039]; commission[0.039]; died[0.039]; Marcin[0.039]; Friday[0.039]; Post[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = World War Two\u001b[39m ==> ENTITY: \u001b[32mWorld War II\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.106:0.106[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNazi[0.046]; Soviet[0.044]; Poland[0.044]; Poland[0.044]; Poland[0.044]; war[0.043]; communist[0.042]; Jews[0.042]; Jews[0.042]; German[0.042]; Warsaw[0.041]; communists[0.041]; Polish[0.041]; Polish[0.041]; Switzerland[0.041]; Switzerland[0.041]; Switzerland[0.041]; citizens[0.040]; citizens[0.040]; citizens[0.040]; citizens[0.040]; million[0.040]; million[0.040]; called[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.047]; Poland[0.047]; Poland[0.047]; Poland[0.047]; Polish[0.043]; Warsaw[0.043]; Switzerland[0.042]; Switzerland[0.042]; Switzerland[0.042]; Switzerland[0.042]; Nazi[0.042]; Soviet[0.042]; communist[0.041]; communist[0.041]; communists[0.040]; war[0.040]; War[0.039]; Jews[0.039]; Jews[0.039]; government[0.039]; received[0.038]; received[0.038]; German[0.038]; weeks[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.221:0.221[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.047]; Poland[0.047]; Poland[0.047]; Poland[0.047]; Poland[0.047]; Poland[0.047]; Polish[0.043]; Polish[0.043]; Warsaw[0.043]; Switzerland[0.042]; Switzerland[0.042]; communist[0.040]; communist[0.040]; Marcin[0.040]; World[0.039]; War[0.039]; War[0.039]; Jews[0.039]; Jews[0.039]; war[0.039]; Dariusz[0.039]; government[0.039]; received[0.038]; received[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swiss ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndollars[0.046]; Swiss[0.044]; Swiss[0.044]; Swiss[0.044]; country[0.044]; francs[0.043]; francs[0.043]; francs[0.043]; million[0.042]; million[0.042]; groups[0.041]; bank[0.041]; week[0.041]; Polish[0.040]; accounts[0.040]; accounts[0.040]; accounts[0.040]; international[0.040]; fund[0.040]; Jewish[0.040]; Jewish[0.040]; Jewish[0.040]; wealth[0.040]; pay[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Holocaust\u001b[39m ==> ENTITY: \u001b[32mThe Holocaust\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJews[0.047]; Poland[0.046]; Poland[0.046]; Poland[0.046]; Poland[0.046]; Poland[0.046]; Polish[0.043]; Warsaw[0.043]; settlements[0.041]; communist[0.040]; War[0.040]; War[0.040]; government[0.040]; received[0.040]; received[0.040]; World[0.040]; powerful[0.040]; countries[0.040]; accounts[0.040]; accounts[0.040]; accounts[0.040]; Switzerland[0.039]; Switzerland[0.039]; died[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Warsaw ==> ENTITY: \u001b[32mWarsaw\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.111:0.111[0]; log p(e|m)= -0.143:-0.143[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.048]; Poland[0.048]; Poland[0.048]; Jews[0.044]; Jews[0.044]; Polish[0.044]; communists[0.043]; Nazi[0.042]; communist[0.042]; Soviet[0.042]; war[0.041]; accused[0.040]; authorities[0.040]; accounts[0.040]; accounts[0.040]; accounts[0.040]; citizens[0.040]; citizens[0.040]; citizens[0.040]; German[0.039]; murdered[0.039]; War[0.039]; commission[0.039]; imposed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jewish ==> ENTITY: \u001b[32mJews\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.146:0.146[0]; log p(e|m)= -0.805:-0.805[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJewish[0.053]; Jewish[0.053]; country[0.044]; community[0.043]; Polish[0.043]; Holocaust[0.043]; international[0.042]; work[0.042]; groups[0.041]; week[0.041]; conservative[0.041]; million[0.040]; million[0.040]; said[0.040]; said[0.040]; government[0.040]; millions[0.039]; dollars[0.039]; citizens[0.039]; accounts[0.039]; accounts[0.039]; moral[0.039]; small[0.039]; charitable[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.196:0.196[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.047]; Poland[0.047]; Poland[0.047]; Poland[0.047]; Polish[0.043]; Warsaw[0.043]; Switzerland[0.042]; Switzerland[0.042]; Switzerland[0.042]; Nazi[0.042]; Soviet[0.042]; communist[0.041]; communist[0.041]; communists[0.040]; war[0.040]; World[0.040]; World[0.040]; War[0.039]; War[0.039]; Jews[0.039]; Jews[0.039]; Dariusz[0.039]; government[0.039]; Holocaust[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rosati\u001b[39m ==> ENTITY: \u001b[32mDariusz Rosati\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.105:0.105[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPolish[0.045]; international[0.044]; parliament[0.043]; information[0.043]; Rosati[0.043]; Rosati[0.043]; data[0.042]; financial[0.042]; francs[0.041]; francs[0.041]; francs[0.041]; francs[0.041]; government[0.041]; exports[0.041]; Switzerland[0.041]; Switzerland[0.041]; Radical[0.041]; work[0.041]; pressure[0.041]; issue[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jewish ==> ENTITY: \u001b[32mJews\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.805:-0.805[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJewish[0.052]; Jewish[0.052]; country[0.043]; community[0.043]; Polish[0.043]; Holocaust[0.043]; international[0.042]; groups[0.041]; work[0.041]; called[0.041]; conservative[0.041]; Radical[0.040]; million[0.040]; million[0.040]; million[0.040]; week[0.040]; order[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; Switzerland[0.040]; Switzerland[0.040]; government[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.188:0.188[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.051]; Poland[0.051]; Poland[0.051]; Polish[0.047]; Warsaw[0.047]; Switzerland[0.046]; Switzerland[0.046]; Switzerland[0.046]; Switzerland[0.046]; Switzerland[0.046]; Nazi[0.046]; Soviet[0.046]; communist[0.044]; communists[0.044]; war[0.044]; World[0.043]; War[0.043]; Jews[0.043]; Jews[0.043]; received[0.042]; German[0.042]; weeks[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rosati\u001b[39m ==> ENTITY: \u001b[32mDariusz Rosati\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.045]; Poland[0.045]; Poland[0.045]; Poland[0.045]; Poland[0.045]; Poland[0.045]; Dariusz[0.042]; communist[0.042]; communist[0.042]; Warsaw[0.042]; Warsaw[0.042]; Polish[0.041]; Polish[0.041]; Committee[0.041]; Minister[0.040]; chairman[0.040]; countries[0.040]; Marcin[0.040]; Rosati[0.040]; Rosati[0.040]; news[0.039]; received[0.039]; received[0.039]; Accounts[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dariusz Rosati\u001b[39m ==> ENTITY: \u001b[32mDariusz Rosati\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.046]; Poland[0.046]; Poland[0.046]; Poland[0.046]; Poland[0.046]; communist[0.043]; Warsaw[0.043]; Polish[0.042]; Polish[0.042]; Committee[0.042]; Minister[0.041]; chairman[0.041]; countries[0.040]; Marcin[0.040]; Rosati[0.040]; news[0.040]; received[0.040]; received[0.040]; Accounts[0.040]; conference[0.039]; Senator[0.039]; francs[0.039]; francs[0.039]; government[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rosati\u001b[39m ==> ENTITY: \u001b[32mDariusz Rosati\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.051]; Poland[0.051]; communist[0.048]; Warsaw[0.047]; Soviet[0.047]; Polish[0.047]; communists[0.046]; international[0.046]; information[0.045]; Rosati[0.045]; Rosati[0.045]; Rosati[0.045]; data[0.045]; financial[0.044]; German[0.044]; francs[0.043]; francs[0.043]; francs[0.043]; Nazi[0.043]; government[0.043]; payment[0.043]; exports[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.164:0.164[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.051]; Poland[0.051]; Poland[0.051]; Polish[0.047]; Warsaw[0.046]; Switzerland[0.046]; communist[0.044]; Marcin[0.043]; World[0.043]; War[0.043]; War[0.043]; Jews[0.043]; Dariusz[0.042]; government[0.042]; received[0.041]; received[0.041]; countries[0.041]; Friday[0.041]; accounts[0.041]; accounts[0.041]; accounts[0.041]; Holocaust[0.040]; settlements[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.161:0.161[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.048]; Poland[0.048]; Poland[0.048]; Polish[0.045]; Warsaw[0.044]; Switzerland[0.044]; Switzerland[0.044]; communist[0.042]; Marcin[0.041]; World[0.041]; War[0.041]; War[0.041]; Jews[0.041]; Dariusz[0.040]; government[0.040]; received[0.039]; received[0.039]; countries[0.039]; Friday[0.039]; accounts[0.039]; accounts[0.039]; accounts[0.039]; accused[0.039]; Holocaust[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jews ==> ENTITY: \u001b[32mJews\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.126:-0.126[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJews[0.048]; Poland[0.045]; Poland[0.045]; Poland[0.045]; Poland[0.045]; Poland[0.045]; Nazi[0.042]; Warsaw[0.042]; Polish[0.041]; Switzerland[0.041]; Switzerland[0.041]; Switzerland[0.041]; Switzerland[0.041]; accounts[0.041]; accounts[0.041]; accounts[0.041]; citizens[0.040]; citizens[0.040]; citizens[0.040]; say[0.040]; Soviet[0.040]; know[0.040]; war[0.040]; German[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swiss ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndollars[0.049]; Swiss[0.047]; Swiss[0.047]; Swiss[0.047]; country[0.047]; francs[0.045]; million[0.044]; million[0.044]; groups[0.043]; bank[0.043]; week[0.043]; accounts[0.042]; international[0.042]; fund[0.042]; Jewish[0.042]; Jewish[0.042]; Jewish[0.042]; wealth[0.042]; pay[0.042]; millions[0.042]; financial[0.042]; precisely[0.042]; devised[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Switzerland ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.193:0.193[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.050]; Switzerland[0.050]; Switzerland[0.050]; Poland[0.043]; Poland[0.043]; Poland[0.043]; Swiss[0.043]; Swiss[0.043]; stands[0.042]; weeks[0.042]; nationals[0.041]; Polish[0.041]; citizens[0.040]; citizens[0.040]; citizens[0.040]; German[0.040]; World[0.039]; said[0.039]; said[0.039]; chairman[0.039]; confirmed[0.039]; Committee[0.039]; Warsaw[0.038]; authorities[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jewish ==> ENTITY: \u001b[32mJews\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.805:-0.805[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJewish[0.053]; Jewish[0.053]; country[0.044]; community[0.043]; Holocaust[0.043]; international[0.042]; groups[0.042]; work[0.042]; week[0.041]; conservative[0.041]; Radical[0.041]; million[0.040]; million[0.040]; order[0.040]; said[0.040]; said[0.040]; government[0.040]; parliament[0.040]; millions[0.040]; dollars[0.039]; accounts[0.039]; moral[0.039]; small[0.039]; charitable[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Polish ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.814:-0.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.047]; Poland[0.047]; Poland[0.047]; Poland[0.047]; Poland[0.047]; Warsaw[0.043]; Soviet[0.042]; War[0.041]; War[0.041]; Jews[0.041]; Jews[0.041]; communist[0.041]; communist[0.041]; war[0.040]; German[0.040]; communists[0.040]; Switzerland[0.039]; Switzerland[0.039]; Switzerland[0.039]; nationals[0.039]; government[0.039]; Foreign[0.039]; countries[0.039]; Nazi[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swiss ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.207:0.207[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.053]; Switzerland[0.053]; Swiss[0.047]; Swiss[0.047]; Swiss[0.047]; German[0.046]; francs[0.045]; francs[0.045]; Polish[0.045]; Polish[0.045]; money[0.045]; money[0.045]; sum[0.044]; Poland[0.044]; Poland[0.044]; Poland[0.044]; Poland[0.044]; Poland[0.044]; Poland[0.044]; countries[0.043]; bank[0.043]; bank[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swiss ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.172:0.172[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.049]; Switzerland[0.049]; dollars[0.045]; Swiss[0.043]; Swiss[0.043]; Swiss[0.043]; country[0.043]; francs[0.042]; francs[0.042]; francs[0.042]; francs[0.042]; million[0.041]; million[0.041]; million[0.041]; sum[0.041]; international[0.040]; groups[0.040]; accounts[0.040]; accounts[0.040]; accounts[0.040]; bank[0.040]; week[0.039]; weeks[0.039]; Polish[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Holocaust\u001b[39m ==> ENTITY: \u001b[32mThe Holocaust\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJewish[0.048]; Jewish[0.048]; Jewish[0.048]; moral[0.043]; accounts[0.043]; accounts[0.043]; accounts[0.043]; Polish[0.042]; said[0.041]; said[0.041]; said[0.041]; international[0.041]; victims[0.041]; country[0.041]; citizens[0.040]; nature[0.040]; government[0.040]; community[0.040]; order[0.040]; significance[0.040]; groups[0.039]; called[0.039]; ask[0.039]; work[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jews ==> ENTITY: \u001b[32mJews\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.126:-0.126[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJews[0.048]; Poland[0.045]; Poland[0.045]; Poland[0.045]; Nazi[0.042]; Warsaw[0.042]; Polish[0.042]; Switzerland[0.041]; Switzerland[0.041]; Switzerland[0.041]; Switzerland[0.041]; Switzerland[0.041]; accounts[0.041]; accounts[0.041]; citizens[0.041]; citizens[0.041]; citizens[0.041]; say[0.040]; Soviet[0.040]; know[0.040]; war[0.040]; German[0.040]; expect[0.040]; nationals[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swiss ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.162:0.162[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.051]; Switzerland[0.051]; Swiss[0.044]; Swiss[0.044]; Polish[0.043]; Friday[0.042]; Poland[0.041]; Poland[0.041]; Poland[0.041]; Poland[0.041]; Warsaw[0.041]; countries[0.041]; unveiling[0.040]; died[0.040]; said[0.040]; said[0.040]; news[0.040]; received[0.040]; received[0.040]; accounts[0.040]; accounts[0.040]; accounts[0.040]; War[0.039]; War[0.039]; \t\n\r [=========================>....................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s682ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2620/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1278testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Appleton ==> ENTITY: \u001b[32mAppleton, Wisconsin\u001b[39m\t\nSCORES: global= 0.240:0.240[0]; local()= 0.081:0.081[0]; log p(e|m)= -0.919:-0.919[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCounty[0.048]; hired[0.044]; said[0.043]; said[0.043]; said[0.043]; Dennis[0.043]; gets[0.043]; family[0.042]; restitution[0.042]; Thursday[0.042]; term[0.042]; Wis[0.041]; prison[0.041]; Friday[0.041]; authorities[0.041]; offered[0.041]; Darrel[0.041]; stole[0.040]; proceeds[0.040]; prior[0.040]; habit[0.039]; Court[0.039]; adding[0.039]; farming[0.039]; \t\n\r [=========================>....................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s754ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2621/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1318testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brussels ==> ENTITY: \u001b[32mBrussels\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.025:-0.025[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgian[0.045]; Belgian[0.045]; Belgian[0.045]; Belgian[0.045]; Spain[0.044]; Antwerp[0.044]; major[0.042]; people[0.042]; office[0.042]; port[0.042]; Friday[0.041]; city[0.041]; woman[0.041]; public[0.041]; millions[0.041]; said[0.040]; said[0.040]; said[0.040]; street[0.040]; police[0.040]; police[0.040]; police[0.040]; taxi[0.040]; Barcelona[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barcelona ==> ENTITY: \u001b[32mBarcelona\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.371:-0.371[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.052]; Spain[0.052]; city[0.046]; France[0.046]; European[0.044]; Belgium[0.044]; Belgium[0.044]; Germany[0.044]; nearly[0.043]; public[0.043]; street[0.042]; taxi[0.042]; came[0.042]; Friday[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; port[0.041]; Antwerp[0.041]; Antwerp[0.041]; lost[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgium ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.193:0.193[0]; log p(e|m)= -0.242:-0.242[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.056]; France[0.051]; European[0.049]; Germany[0.048]; Spain[0.047]; Belgian[0.046]; Belgian[0.046]; Antwerp[0.044]; Antwerp[0.044]; Turkey[0.044]; Turkish[0.042]; Turkish[0.042]; Barcelona[0.041]; lost[0.041]; public[0.040]; nearly[0.040]; office[0.040]; office[0.040]; came[0.040]; people[0.040]; taken[0.040]; nations[0.040]; containing[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgium ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.198:0.198[0]; log p(e|m)= -0.242:-0.242[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.053]; France[0.048]; European[0.046]; Germany[0.045]; Spain[0.044]; Belgian[0.044]; Belgian[0.044]; Belgian[0.044]; Belgian[0.044]; Antwerp[0.042]; Antwerp[0.042]; Turkey[0.042]; Turkish[0.040]; Turkish[0.040]; Barcelona[0.039]; Friday[0.039]; kg[0.038]; public[0.038]; office[0.038]; office[0.038]; came[0.038]; lost[0.038]; people[0.038]; taken[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Antwerp ==> ENTITY: \u001b[32mAntwerp\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.174:0.174[0]; log p(e|m)= -0.119:-0.119[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.048]; Belgium[0.048]; Antwerp[0.047]; port[0.046]; Spain[0.045]; Spain[0.045]; Belgian[0.044]; Belgian[0.044]; Belgian[0.044]; Belgian[0.044]; France[0.044]; European[0.043]; major[0.043]; city[0.042]; Germany[0.042]; people[0.042]; people[0.042]; woman[0.042]; woman[0.042]; Turkey[0.041]; seized[0.041]; lost[0.041]; came[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEuropean\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEurope\u001b[39m <---> \u001b[32mEuropean Union\u001b[39m\t\nSCORES: global= 0.270:0.242[\u001b[31m0.028\u001b[39m]; local()= 0.129:0.125[\u001b[31m0.004\u001b[39m]; log p(e|m)= -0.256:-2.937[\u001b[31m2.681\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.045]; Spain[0.045]; Germany[0.043]; nearly[0.043]; Turkish[0.042]; Turkish[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; Belgium[0.042]; Belgium[0.042]; nations[0.042]; France[0.042]; Belgian[0.041]; Belgian[0.041]; people[0.041]; containing[0.040]; spokesman[0.040]; Barcelona[0.040]; police[0.040]; came[0.039]; left[0.039]; taken[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Antwerp ==> ENTITY: \u001b[32mAntwerp\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.167:0.167[0]; log p(e|m)= -0.119:-0.119[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAntwerp[0.051]; Belgium[0.050]; Belgium[0.050]; Spain[0.047]; Belgian[0.047]; Belgian[0.047]; Belgian[0.047]; France[0.047]; nearly[0.046]; European[0.045]; city[0.044]; seized[0.044]; Germany[0.044]; port[0.044]; people[0.044]; woman[0.044]; woman[0.044]; old[0.043]; old[0.043]; Turkey[0.043]; lost[0.043]; came[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgian ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.135:0.135[0]; log p(e|m)= -0.124:-0.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgian[0.048]; Belgian[0.048]; Belgian[0.048]; Spain[0.047]; Antwerp[0.045]; said[0.044]; said[0.044]; said[0.044]; hundreds[0.044]; people[0.044]; dozens[0.043]; Brussels[0.043]; containing[0.043]; major[0.043]; woman[0.043]; francs[0.042]; Friday[0.042]; public[0.041]; spokesman[0.041]; city[0.041]; office[0.041]; left[0.040]; left[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgian ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.124:-0.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.053]; Belgian[0.045]; Belgian[0.045]; Belgian[0.045]; Spain[0.044]; francs[0.042]; Antwerp[0.042]; said[0.041]; said[0.041]; said[0.041]; hundreds[0.041]; people[0.041]; office[0.041]; office[0.041]; lb[0.041]; dozens[0.040]; Brussels[0.040]; major[0.040]; woman[0.040]; woman[0.040]; seized[0.040]; public[0.039]; lost[0.039]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgian ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.187:0.187[0]; log p(e|m)= -0.124:-0.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.053]; Belgium[0.053]; France[0.048]; Germany[0.045]; Belgian[0.044]; Spain[0.044]; European[0.043]; Antwerp[0.042]; nearly[0.041]; said[0.041]; said[0.041]; Turkey[0.040]; people[0.040]; office[0.040]; lb[0.040]; woman[0.039]; woman[0.039]; nations[0.039]; lost[0.039]; Turkish[0.038]; Turkish[0.038]; came[0.038]; man[0.038]; man[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.129:0.129[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuropean[0.046]; nearly[0.044]; Germany[0.044]; France[0.044]; Belgium[0.044]; Belgium[0.044]; Barcelona[0.043]; Turkey[0.043]; nations[0.042]; lost[0.042]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; millions[0.041]; Turkish[0.040]; Turkish[0.040]; people[0.040]; came[0.040]; port[0.040]; city[0.040]; Antwerp[0.040]; Antwerp[0.040]; public[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgian ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.229:0.229[0]; log p(e|m)= -0.124:-0.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.051]; Belgium[0.051]; France[0.046]; Germany[0.043]; Belgian[0.043]; Belgian[0.043]; Belgian[0.043]; Belgian[0.043]; Spain[0.042]; Spain[0.042]; European[0.041]; francs[0.040]; Antwerp[0.040]; Antwerp[0.040]; nearly[0.040]; said[0.039]; said[0.039]; said[0.039]; said[0.039]; said[0.039]; Turkey[0.039]; hundreds[0.039]; people[0.039]; people[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.069:0.069[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBarcelona[0.047]; lost[0.046]; major[0.045]; said[0.045]; said[0.045]; said[0.045]; people[0.045]; millions[0.044]; Turkish[0.043]; came[0.043]; port[0.043]; city[0.043]; seized[0.043]; public[0.043]; containing[0.043]; Friday[0.042]; office[0.042]; Brussels[0.042]; drugs[0.042]; drugs[0.042]; smuggling[0.042]; rings[0.042]; rings[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.141:0.141[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.048]; Belgium[0.048]; Spain[0.046]; Germany[0.046]; European[0.045]; Belgian[0.042]; Belgian[0.042]; Belgian[0.042]; nearly[0.042]; nations[0.041]; Antwerp[0.041]; Antwerp[0.041]; Turkey[0.041]; came[0.041]; public[0.040]; Barcelona[0.040]; office[0.040]; office[0.040]; city[0.040]; lost[0.040]; lasting[0.039]; millions[0.039]; hundreds[0.039]; people[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgian ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.124:-0.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.052]; France[0.047]; Germany[0.044]; Belgian[0.043]; Belgian[0.043]; Belgian[0.043]; Spain[0.043]; Spain[0.043]; francs[0.041]; Antwerp[0.041]; Antwerp[0.041]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; Turkey[0.040]; hundreds[0.040]; people[0.040]; people[0.040]; lb[0.040]; dozens[0.039]; Brussels[0.039]; major[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.052]; European[0.050]; France[0.049]; Belgium[0.049]; Belgium[0.049]; Turkey[0.047]; lost[0.046]; nations[0.046]; came[0.045]; city[0.045]; Antwerp[0.044]; Antwerp[0.044]; Turkish[0.044]; Turkish[0.044]; Barcelona[0.044]; nearly[0.044]; containing[0.043]; left[0.043]; people[0.043]; public[0.043]; old[0.043]; old[0.043]; \t\n\r [==========================>...................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s591ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2637/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1263testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexico ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.183:-0.183[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.052]; Chile[0.046]; Chile[0.046]; Chile[0.046]; Eduardo[0.046]; January[0.043]; Santiago[0.042]; seek[0.042]; start[0.042]; February[0.041]; provisions[0.040]; close[0.040]; services[0.040]; services[0.040]; said[0.040]; said[0.040]; free[0.040]; free[0.040]; covering[0.039]; Finance[0.039]; hopes[0.039]; treaty[0.039]; add[0.039]; areas[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexico ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.142:0.142[0]; log p(e|m)= -0.183:-0.183[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.050]; Mexico[0.050]; Chile[0.045]; Chile[0.045]; Chile[0.045]; Eduardo[0.044]; Canada[0.043]; Chilean[0.043]; unions[0.042]; services[0.041]; services[0.041]; services[0.041]; Santiago[0.041]; unanimously[0.040]; start[0.040]; January[0.040]; Congress[0.040]; Congress[0.040]; signing[0.039]; deal[0.039]; deal[0.039]; provisions[0.039]; ratify[0.039]; February[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.071:0.071[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.046]; Mexico[0.046]; Chile[0.043]; Chile[0.043]; February[0.043]; January[0.043]; Roger[0.041]; trade[0.041]; trade[0.041]; trade[0.041]; advance[0.041]; want[0.041]; covering[0.041]; close[0.041]; add[0.041]; add[0.041]; signing[0.041]; hopes[0.041]; greater[0.041]; Eduardo[0.041]; ratify[0.040]; treaty[0.040]; treaty[0.040]; treaty[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mFinance\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFinance\u001b[39m <---> \u001b[32mMinistry of Finance (Chile)\u001b[39m\t\nSCORES: global= 0.262:0.230[\u001b[31m0.032\u001b[39m]; local()= 0.158:0.183[\u001b[32m0.025\u001b[39m]; log p(e|m)= -0.494:-3.297[\u001b[31m2.803\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nservices[0.046]; services[0.046]; services[0.046]; investments[0.045]; investments[0.045]; investment[0.045]; January[0.044]; trade[0.044]; trade[0.044]; trade[0.044]; trade[0.044]; provisions[0.043]; Minister[0.043]; Chile[0.043]; Chile[0.043]; Chile[0.043]; February[0.043]; imports[0.042]; reduction[0.042]; reduction[0.042]; exports[0.042]; include[0.042]; tariff[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexico ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.118:0.118[0]; log p(e|m)= -0.183:-0.183[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.053]; Chile[0.047]; Chile[0.047]; Eduardo[0.047]; Canada[0.046]; Chilean[0.045]; unions[0.045]; services[0.043]; services[0.043]; Santiago[0.043]; unanimously[0.043]; January[0.042]; Congress[0.042]; Congress[0.042]; signing[0.042]; deal[0.041]; provisions[0.041]; ratify[0.041]; February[0.041]; close[0.041]; free[0.041]; free[0.041]; confident[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chile ==> ENTITY: \u001b[32mChile\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChile[0.048]; Chile[0.048]; Mexico[0.046]; Mexico[0.046]; said[0.042]; said[0.042]; agreement[0.041]; trade[0.041]; trade[0.041]; trade[0.041]; Santiago[0.041]; Eduardo[0.041]; free[0.041]; free[0.041]; areas[0.041]; negotiations[0.041]; negotiations[0.041]; start[0.040]; seek[0.040]; deal[0.040]; signed[0.040]; services[0.040]; services[0.040]; Minister[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chile ==> ENTITY: \u001b[32mChile\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.161:0.161[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChile[0.047]; Chile[0.047]; Mexico[0.045]; Mexico[0.045]; Mexico[0.045]; provisions[0.042]; said[0.041]; said[0.041]; agreement[0.041]; trade[0.041]; trade[0.041]; trade[0.041]; contacts[0.040]; treaty[0.040]; Santiago[0.040]; Eduardo[0.040]; free[0.040]; free[0.040]; areas[0.040]; negotiations[0.040]; negotiations[0.040]; start[0.040]; seek[0.040]; deal[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chile ==> ENTITY: \u001b[32mChile\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChile[0.048]; Mexico[0.046]; Mexico[0.046]; Chilean[0.045]; Santiago[0.043]; Canada[0.043]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; trade[0.041]; trade[0.041]; trade[0.041]; contacts[0.041]; Eduardo[0.040]; free[0.040]; free[0.040]; areas[0.040]; negotiations[0.040]; deal[0.040]; signed[0.040]; signing[0.039]; told[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chile ==> ENTITY: \u001b[32mChile\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.186:0.186[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChile[0.047]; Chile[0.047]; Chile[0.047]; Mexico[0.045]; Mexico[0.045]; Mexico[0.045]; Mexico[0.045]; Canada[0.042]; provisions[0.041]; said[0.040]; said[0.040]; agreement[0.040]; trade[0.040]; trade[0.040]; trade[0.040]; trade[0.040]; contacts[0.040]; treaty[0.040]; treaty[0.040]; Santiago[0.040]; Eduardo[0.040]; free[0.040]; free[0.040]; free[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Santiago ==> ENTITY: \u001b[32mSantiago\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.465:-0.465[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChile[0.047]; Chile[0.047]; Chile[0.047]; Mexico[0.046]; Mexico[0.046]; Mexico[0.046]; areas[0.046]; services[0.045]; services[0.045]; services[0.045]; Eduardo[0.043]; January[0.043]; February[0.042]; said[0.042]; said[0.042]; provisions[0.042]; start[0.042]; covering[0.041]; issue[0.041]; Finance[0.041]; trade[0.041]; trade[0.041]; trade[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chilean ==> ENTITY: \u001b[32mChile\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.257:-0.257[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChile[0.048]; Mexico[0.046]; Mexico[0.046]; Santiago[0.044]; Canada[0.043]; tend[0.042]; business[0.042]; free[0.042]; free[0.042]; trade[0.041]; trade[0.041]; trade[0.041]; greater[0.041]; positive[0.041]; seen[0.041]; want[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; good[0.039]; reduction[0.039]; treaty[0.039]; treaty[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexico ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.183:-0.183[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.051]; Mexico[0.051]; Chile[0.045]; Chile[0.045]; Chile[0.045]; Eduardo[0.045]; services[0.042]; services[0.042]; services[0.042]; Santiago[0.041]; start[0.041]; January[0.040]; deal[0.040]; provisions[0.040]; February[0.040]; close[0.040]; free[0.039]; free[0.039]; issue[0.039]; issue[0.039]; said[0.039]; said[0.039]; covering[0.039]; laden[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Santiago ==> ENTITY: \u001b[32mSantiago\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.465:-0.465[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChile[0.047]; Mexico[0.046]; Chilean[0.045]; advance[0.043]; reporters[0.042]; Roger[0.042]; greater[0.042]; said[0.041]; said[0.041]; said[0.041]; services[0.041]; quickly[0.041]; issue[0.041]; issue[0.041]; problem[0.041]; good[0.041]; trade[0.041]; trade[0.041]; reactions[0.041]; business[0.040]; treaty[0.040]; treaty[0.040]; goods[0.040]; want[0.040]; \t\n\r [==========================>...................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s635ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2650/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1222testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belfast ==> ENTITY: \u001b[32mBelfast\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.126:0.126[0]; log p(e|m)= -0.131:-0.131[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelfast[0.052]; Ireland[0.050]; Irish[0.047]; Army[0.043]; incendiaries[0.041]; forces[0.041]; bombs[0.041]; bombs[0.041]; bombs[0.041]; house[0.041]; house[0.041]; charged[0.040]; charged[0.040]; charged[0.040]; Friday[0.040]; Saturday[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; released[0.040]; terrorist[0.040]; terrorism[0.039]; ammunition[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belfast ==> ENTITY: \u001b[32mBelfast\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.131:-0.131[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelfast[0.055]; Ireland[0.053]; Ireland[0.053]; Irish[0.050]; Britain[0.047]; Army[0.046]; Northern[0.045]; incendiaries[0.044]; forces[0.044]; bombs[0.044]; bombs[0.044]; bombs[0.044]; house[0.043]; house[0.043]; guerrilla[0.043]; bombing[0.043]; charged[0.043]; charged[0.043]; charged[0.043]; Friday[0.043]; group[0.043]; Saturday[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Irish Republican Army\u001b[39m ==> ENTITY: \u001b[32mIrish Republican Army\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.128:0.128[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelfast[0.048]; Belfast[0.048]; Ireland[0.047]; guerrilla[0.043]; ammunition[0.043]; Britain[0.042]; forces[0.042]; terrorist[0.042]; bombs[0.042]; bombs[0.042]; bombs[0.042]; bombing[0.041]; group[0.041]; explosives[0.040]; campaign[0.040]; charged[0.040]; charged[0.040]; charged[0.040]; Northern[0.040]; released[0.040]; arms[0.040]; terrorism[0.039]; oust[0.039]; mortar[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Northern Ireland ==> ENTITY: \u001b[32mNorthern Ireland\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.251:-0.251[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelfast[0.049]; Army[0.047]; forces[0.045]; Britain[0.044]; police[0.043]; police[0.043]; Police[0.043]; use[0.042]; Security[0.041]; campaign[0.041]; Saturday[0.041]; guerrilla[0.040]; bombing[0.040]; charged[0.040]; released[0.040]; group[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; explosives[0.040]; pre[0.040]; ammunition[0.040]; possession[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.067:0.067[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArmy[0.049]; Ireland[0.049]; Belfast[0.045]; released[0.043]; forces[0.043]; Northern[0.043]; group[0.041]; ammunition[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; use[0.041]; Security[0.041]; guerrilla[0.040]; bombs[0.040]; bombs[0.040]; bombs[0.040]; campaign[0.040]; possession[0.040]; explosives[0.040]; terrorism[0.040]; police[0.040]; police[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = N. Ireland\u001b[39m ==> ENTITY: \u001b[32mNorthern Ireland\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelfast[0.052]; Belfast[0.052]; Irish[0.049]; Army[0.045]; Saturday[0.043]; Friday[0.041]; possession[0.041]; charged[0.041]; charged[0.041]; charged[0.041]; old[0.040]; released[0.040]; Police[0.040]; police[0.040]; police[0.040]; available[0.040]; Republican[0.040]; use[0.040]; forces[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; making[0.039]; \t\n\r [==========================>...................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s651ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2656/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1328testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sydney ==> ENTITY: \u001b[32mSydney\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.144:-0.144[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSydney[0.053]; Australian[0.047]; Australian[0.047]; home[0.046]; Wales[0.046]; South[0.045]; house[0.044]; Tony[0.044]; Paul[0.043]; shoot[0.043]; February[0.043]; leg[0.043]; female[0.042]; misconduct[0.042]; contracted[0.042]; Friday[0.042]; come[0.042]; knock[0.042]; went[0.041]; punish[0.041]; friend[0.040]; told[0.040]; Court[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australian ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.491:-0.491[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralian[0.050]; Sydney[0.049]; Sydney[0.049]; Wales[0.047]; Tony[0.046]; Michael[0.046]; told[0.046]; South[0.046]; Paul[0.045]; leg[0.043]; February[0.043]; Friday[0.043]; home[0.041]; material[0.041]; contracted[0.041]; man[0.041]; man[0.041]; went[0.041]; female[0.041]; Court[0.040]; knock[0.040]; house[0.040]; house[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = New South Wales Supreme Court\u001b[39m ==> ENTITY: \u001b[32mSupreme Court of New South Wales\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.137:0.137[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJudge[0.045]; Australian[0.044]; Australian[0.044]; leg[0.044]; February[0.043]; Michael[0.042]; Sydney[0.042]; Sydney[0.042]; guilty[0.042]; said[0.042]; said[0.042]; placed[0.042]; sentenced[0.041]; misconduct[0.041]; sentencing[0.041]; Friday[0.041]; brought[0.041]; house[0.040]; house[0.040]; Betcher[0.040]; Betcher[0.040]; Paul[0.040]; pleaded[0.040]; account[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sydney ==> ENTITY: \u001b[32mSydney\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.144:-0.144[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSydney[0.050]; Australian[0.045]; Australian[0.045]; home[0.044]; Wales[0.043]; South[0.043]; Michael[0.042]; house[0.042]; house[0.042]; Tony[0.042]; Paul[0.041]; shoot[0.041]; February[0.041]; unnamed[0.041]; took[0.040]; leg[0.040]; horrendous[0.040]; contracted[0.040]; state[0.040]; female[0.040]; misconduct[0.040]; ghastly[0.040]; Friday[0.039]; come[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australian ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.491:-0.491[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralian[0.048]; Sydney[0.047]; Sydney[0.047]; Wales[0.046]; Tony[0.045]; South[0.045]; Paul[0.044]; leg[0.042]; told[0.042]; February[0.042]; Friday[0.041]; home[0.040]; contracted[0.040]; man[0.040]; man[0.040]; went[0.040]; female[0.040]; knock[0.039]; identified[0.039]; house[0.039]; material[0.039]; hitman[0.039]; hitman[0.039]; shot[0.038]; \t\n\r [==========================>...................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s656ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2661/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1323testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mFrankfurt\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFrankfurt Stock Exchange\u001b[39m <---> \u001b[32mFrankfurt\u001b[39m\t\nSCORES: global= 0.243:0.238[\u001b[31m0.004\u001b[39m]; local()= 0.147:0.108[\u001b[31m0.038\u001b[39m]; log p(e|m)= -3.352:-0.070[\u001b[32m3.282\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrankfurt[0.096]; Frankfurt[0.096]; Bundesbank[0.085]; Thursday[0.083]; Friday[0.082]; dollar[0.082]; dollar[0.082]; fixed[0.081]; marks[0.080]; marks[0.080]; marks[0.080]; fix[0.074]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Frankfurt ==> ENTITY: \u001b[32mFrankfurt Stock Exchange\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.147:0.147[0]; log p(e|m)= -3.352:-3.352[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrankfurt[0.096]; Frankfurt[0.096]; Bundesbank[0.085]; Thursday[0.083]; Friday[0.082]; dollar[0.082]; dollar[0.082]; fixed[0.081]; marks[0.080]; marks[0.080]; marks[0.080]; fix[0.074]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bundesbank\u001b[39m ==> ENTITY: \u001b[32mDeutsche Bundesbank\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndollar[0.082]; dollar[0.082]; Frankfurt[0.081]; Frankfurt[0.081]; Frankfurt[0.081]; fixed[0.078]; marks[0.075]; marks[0.075]; marks[0.075]; Thursday[0.074]; Friday[0.073]; intervention[0.072]; fix[0.071]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Frankfurt ==> ENTITY: \u001b[32mFrankfurt Stock Exchange\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.147:0.147[0]; log p(e|m)= -3.352:-3.352[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrankfurt[0.096]; Frankfurt[0.096]; Bundesbank[0.085]; Thursday[0.083]; Friday[0.082]; dollar[0.082]; dollar[0.082]; fixed[0.081]; marks[0.080]; marks[0.080]; marks[0.080]; fix[0.074]; \t\n\r [==========================>...................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s762ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2665/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1271testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moslem ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.148:0.148[0]; log p(e|m)= -1.214:-1.214[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChristian[0.044]; living[0.044]; Jewish[0.042]; territory[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanon[0.042]; Lebanon[0.042]; citizens[0.041]; said[0.041]; said[0.041]; said[0.041]; soldiers[0.041]; Israel[0.041]; Israel[0.041]; Israel[0.041]; regards[0.041]; held[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jewish ==> ENTITY: \u001b[32mJews\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.157:0.157[0]; log p(e|m)= -0.805:-0.805[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.045]; Israel[0.045]; Israel[0.045]; Lebanon[0.043]; Lebanon[0.043]; Lebanese[0.043]; Lebanese[0.043]; Lebanese[0.043]; Lebanese[0.043]; Lebanese[0.043]; Lebanese[0.043]; Lebanese[0.043]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; Christian[0.042]; Moslem[0.039]; living[0.039]; Peres[0.039]; Peres[0.039]; Shimon[0.039]; Beirut[0.039]; states[0.038]; rightwing[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanon ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.299:0.299[0]; log p(e|m)= -0.409:-0.409[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.047]; Lebanon[0.047]; Lebanon[0.047]; Lebanon[0.047]; Beirut[0.043]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Israel[0.042]; Israel[0.042]; Israel[0.042]; Israel[0.042]; Israel[0.042]; Israel[0.042]; Israel[0.042]; Syrian[0.042]; Israeli[0.039]; government[0.038]; military[0.038]; northern[0.037]; peace[0.037]; South[0.036]; war[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Benjamin Netanyahu\u001b[39m ==> ENTITY: \u001b[32mBenjamin Netanyahu\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.226:0.226[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.049]; Israel[0.049]; Peres[0.045]; Peres[0.045]; Israeli[0.045]; Israeli[0.045]; Israeli[0.045]; Israeli[0.045]; Lebanon[0.042]; Lebanon[0.042]; Lebanon[0.042]; Shimon[0.040]; Prime[0.040]; leader[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; said[0.038]; said[0.038]; said[0.038]; said[0.038]; negotiations[0.038]; Minister[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanon ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.231:0.231[0]; log p(e|m)= -0.409:-0.409[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.050]; Lebanon[0.050]; Beirut[0.046]; Lebanese[0.045]; Lebanese[0.045]; Lebanese[0.045]; Lebanese[0.045]; Israel[0.045]; Israeli[0.042]; Israeli[0.042]; Israeli[0.042]; Israeli[0.042]; military[0.040]; political[0.040]; peace[0.039]; Netanyahu[0.039]; war[0.039]; Peres[0.038]; Peres[0.038]; militia[0.038]; civil[0.038]; army[0.038]; army[0.038]; negotiations[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Antoine Lahd\u001b[39m ==> ENTITY: \u001b[32mAntoine Lahad\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.044]; Lebanon[0.044]; Lebanon[0.044]; Lebanon[0.044]; Lebanon[0.044]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israeli[0.043]; Lebanese[0.043]; Lebanese[0.043]; Lebanese[0.043]; army[0.040]; Beirut[0.040]; military[0.040]; militia[0.039]; militia[0.039]; said[0.039]; said[0.039]; Jewish[0.039]; Army[0.039]; Haddadin[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanon ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.300:0.300[0]; log p(e|m)= -0.409:-0.409[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.047]; Lebanon[0.047]; Lebanon[0.047]; Beirut[0.043]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Israel[0.042]; Israel[0.042]; Israel[0.042]; Israel[0.042]; Israel[0.042]; Israel[0.042]; Israel[0.042]; Syrian[0.042]; government[0.038]; military[0.038]; northern[0.037]; peace[0.037]; South[0.036]; war[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanese ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.291:0.291[0]; log p(e|m)= -0.559:-0.559[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.047]; Lebanon[0.047]; Lebanese[0.044]; Lebanese[0.044]; Lebanese[0.044]; Lebanese[0.044]; Lebanese[0.044]; Lebanese[0.044]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Syrian[0.042]; Beirut[0.041]; Jewish[0.040]; Jewish[0.040]; Israeli[0.039]; Moslem[0.038]; Christian[0.037]; military[0.037]; territory[0.037]; government[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanese ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.300:0.300[0]; log p(e|m)= -0.559:-0.559[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.047]; Lebanon[0.047]; Lebanon[0.047]; Lebanon[0.047]; Lebanon[0.047]; Lebanon[0.047]; Lebanese[0.044]; Lebanese[0.044]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Syrian[0.042]; Beirut[0.042]; Jewish[0.040]; Israeli[0.039]; northern[0.038]; military[0.037]; border[0.036]; South[0.036]; war[0.036]; major[0.036]; Haitham[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.231:0.231[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.049]; Israel[0.049]; Israeli[0.045]; Israeli[0.045]; Israeli[0.045]; Lebanon[0.044]; Lebanon[0.044]; Jewish[0.042]; Peres[0.041]; Peres[0.041]; military[0.041]; Netanyahu[0.041]; army[0.040]; army[0.040]; army[0.040]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Beirut[0.038]; soldiers[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanese ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.275:0.275[0]; log p(e|m)= -0.559:-0.559[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.048]; Lebanon[0.048]; Lebanon[0.048]; Lebanese[0.044]; Lebanese[0.044]; Lebanese[0.044]; Lebanese[0.044]; Lebanese[0.044]; Lebanese[0.044]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Beirut[0.042]; Jewish[0.040]; Israeli[0.040]; Israeli[0.040]; Israeli[0.040]; Moslem[0.038]; Christian[0.038]; military[0.038]; force[0.037]; political[0.036]; Benjamin[0.036]; civilians[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = SLA ==> ENTITY: \u001b[32mSouth Lebanon Army\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.194:0.194[0]; log p(e|m)= -1.666:-1.666[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.049]; Lebanon[0.049]; Lebanon[0.049]; Israel[0.048]; Lebanese[0.045]; Lebanese[0.045]; Lebanese[0.045]; military[0.045]; Israeli[0.045]; Israeli[0.045]; Israeli[0.045]; Israeli[0.045]; army[0.043]; militia[0.042]; Beirut[0.041]; peace[0.041]; commanders[0.040]; leader[0.040]; commander[0.040]; Christian[0.039]; war[0.039]; civil[0.039]; brigade[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Beirut ==> ENTITY: \u001b[32mBeirut\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.185:0.185[0]; log p(e|m)= -0.047:-0.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.048]; Lebanon[0.048]; Lebanon[0.048]; Israeli[0.045]; Israeli[0.045]; Israeli[0.045]; Israeli[0.045]; Israel[0.045]; Lebanese[0.045]; Lebanese[0.045]; Lebanese[0.045]; military[0.043]; army[0.040]; political[0.039]; peace[0.038]; Netanyahu[0.038]; war[0.038]; civil[0.038]; militia[0.038]; said[0.038]; said[0.038]; said[0.038]; Saqr[0.037]; Saqr[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.262:0.262[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.048]; Israel[0.048]; Israel[0.048]; Israel[0.048]; Israel[0.048]; Israel[0.048]; Israeli[0.044]; Lebanon[0.043]; Lebanon[0.043]; Lebanon[0.043]; Jewish[0.041]; military[0.040]; army[0.039]; army[0.039]; Syrian[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Beirut[0.037]; soldiers[0.037]; government[0.037]; civilians[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = SLA ==> ENTITY: \u001b[32mSouth Lebanon Army\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.225:0.225[0]; log p(e|m)= -1.666:-1.666[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.045]; Lebanon[0.045]; Lebanon[0.045]; Lebanon[0.045]; Lebanon[0.045]; Lebanon[0.045]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Lebanese[0.041]; Lebanese[0.041]; Lebanese[0.041]; military[0.041]; Israeli[0.041]; army[0.039]; Syrian[0.039]; Army[0.039]; militia[0.038]; militia[0.038]; Beirut[0.038]; guerrilla[0.038]; government[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.245:0.245[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.049]; Israel[0.049]; Israel[0.049]; Israel[0.049]; Israel[0.049]; Israeli[0.044]; Lebanon[0.043]; Lebanon[0.043]; Jewish[0.042]; military[0.040]; army[0.039]; army[0.039]; army[0.039]; army[0.039]; Syrian[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Beirut[0.038]; soldiers[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = SLA ==> ENTITY: \u001b[32mSouth Lebanon Army\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.209:0.209[0]; log p(e|m)= -1.666:-1.666[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.046]; Lebanon[0.046]; Lebanon[0.046]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; military[0.042]; Israeli[0.042]; Israeli[0.042]; Israeli[0.042]; army[0.040]; army[0.040]; army[0.040]; Beirut[0.039]; peace[0.038]; Jewish[0.038]; kidnapping[0.037]; soldiers[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanese ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.264:0.264[0]; log p(e|m)= -0.559:-0.559[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.048]; Lebanon[0.048]; Lebanon[0.048]; Lebanese[0.045]; Lebanese[0.045]; Lebanese[0.045]; Lebanese[0.045]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Beirut[0.043]; Beirut[0.043]; Jewish[0.041]; Israeli[0.040]; Israeli[0.040]; Israeli[0.040]; Moslem[0.039]; Christian[0.038]; military[0.038]; military[0.038]; force[0.037]; political[0.037]; Benjamin[0.036]; patriot[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.280:0.280[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.048]; Israel[0.048]; Israel[0.048]; Israel[0.048]; Israel[0.048]; Israel[0.048]; Lebanon[0.043]; Lebanon[0.043]; Lebanon[0.043]; Lebanon[0.043]; Lebanon[0.043]; Jewish[0.041]; Syrian[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Beirut[0.038]; territory[0.037]; military[0.037]; government[0.037]; war[0.036]; legal[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Guardians of the Cedars\u001b[39m ==> ENTITY: \u001b[32mGuardians of the Cedars\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanese[0.050]; Lebanon[0.050]; Lebanon[0.050]; Lebanon[0.050]; Beirut[0.046]; Israel[0.046]; militia[0.044]; Israeli[0.044]; Israeli[0.044]; Israeli[0.044]; Israeli[0.044]; rightwing[0.042]; rightwing[0.042]; Saqr[0.042]; Saqr[0.042]; military[0.042]; Christian[0.041]; empted[0.040]; leader[0.040]; pro[0.040]; action[0.040]; civil[0.039]; calling[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.264:0.264[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.048]; Israel[0.048]; Israel[0.048]; Israel[0.048]; Israel[0.048]; Israel[0.048]; Israeli[0.044]; Lebanon[0.043]; Lebanon[0.043]; Lebanon[0.043]; Jewish[0.041]; military[0.040]; army[0.039]; army[0.039]; Syrian[0.039]; Lebanese[0.038]; Lebanese[0.038]; Lebanese[0.038]; Lebanese[0.038]; Lebanese[0.038]; Beirut[0.037]; soldiers[0.037]; Christian[0.037]; government[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = SLA ==> ENTITY: \u001b[32mSouth Lebanon Army\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.210:0.210[0]; log p(e|m)= -1.666:-1.666[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.045]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; military[0.042]; Israeli[0.042]; Israeli[0.042]; army[0.040]; army[0.040]; army[0.040]; army[0.040]; civilians[0.039]; Beirut[0.038]; government[0.038]; peace[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = SLA ==> ENTITY: \u001b[32mSouth Lebanon Army\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.237:0.237[0]; log p(e|m)= -1.666:-1.666[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.044]; Lebanon[0.044]; Lebanon[0.044]; Lebanon[0.044]; Lebanon[0.044]; Lebanon[0.044]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Lebanese[0.041]; Lebanese[0.041]; Lebanese[0.041]; military[0.041]; Israeli[0.040]; army[0.039]; army[0.039]; Syrian[0.039]; Army[0.039]; militia[0.038]; militia[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanese ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.259:0.259[0]; log p(e|m)= -0.559:-0.559[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.048]; Lebanon[0.048]; Lebanon[0.048]; Lebanese[0.045]; Lebanese[0.045]; Lebanese[0.045]; Lebanese[0.045]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Beirut[0.043]; Jewish[0.041]; Israeli[0.040]; Israeli[0.040]; Israeli[0.040]; Moslem[0.039]; Christian[0.039]; military[0.038]; political[0.038]; force[0.038]; Saqr[0.037]; Benjamin[0.037]; patriot[0.036]; south[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.236:0.236[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.049]; Israel[0.049]; Israeli[0.045]; Israeli[0.045]; Israeli[0.045]; Jewish[0.045]; Lebanon[0.044]; Lebanon[0.044]; Lebanon[0.044]; Shimon[0.042]; army[0.041]; army[0.041]; Peres[0.041]; Peres[0.041]; military[0.040]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Beirut[0.039]; Netanyahu[0.038]; force[0.037]; calling[0.037]; peace[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanon ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.262:0.262[0]; log p(e|m)= -0.409:-0.409[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.048]; Lebanon[0.048]; Lebanon[0.048]; Lebanon[0.048]; Beirut[0.044]; Lebanese[0.044]; Lebanese[0.044]; Lebanese[0.044]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Israeli[0.041]; military[0.039]; northern[0.039]; peace[0.038]; South[0.038]; war[0.038]; Haitham[0.037]; border[0.037]; militia[0.037]; militia[0.037]; major[0.037]; army[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Saqr ==> ENTITY: \u001b[32mEtienne Saqr\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.168:0.168[0]; log p(e|m)= -0.605:-0.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.048]; Lebanon[0.048]; Lebanon[0.048]; Lebanese[0.046]; Beirut[0.044]; Israel[0.044]; Cedars[0.043]; Israeli[0.042]; Israeli[0.042]; Israeli[0.042]; Saqr[0.041]; charges[0.041]; Christian[0.040]; absentia[0.040]; militia[0.040]; war[0.040]; trial[0.039]; military[0.039]; earlier[0.039]; treason[0.039]; Peres[0.039]; Peres[0.039]; convicted[0.039]; court[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.176:0.176[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.053]; Israeli[0.048]; Israeli[0.048]; Lebanon[0.047]; Lebanon[0.047]; Lebanon[0.047]; Peres[0.044]; military[0.043]; Lebanese[0.042]; Beirut[0.042]; sentenced[0.041]; Netanyahu[0.040]; militia[0.039]; charges[0.039]; said[0.039]; said[0.039]; Benjamin[0.039]; convicted[0.039]; Christian[0.039]; absentia[0.038]; war[0.038]; life[0.038]; commanders[0.038]; leader[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Beirut ==> ENTITY: \u001b[32mBeirut\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.202:0.202[0]; log p(e|m)= -0.047:-0.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.047]; Lebanon[0.047]; Lebanon[0.047]; Lebanon[0.047]; Lebanon[0.047]; Israeli[0.044]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Lebanese[0.044]; Lebanese[0.044]; military[0.042]; northern[0.039]; army[0.039]; Army[0.039]; war[0.038]; militia[0.037]; militia[0.037]; border[0.037]; major[0.037]; frontier[0.037]; Haitham[0.037]; Friday[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanese ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.294:0.294[0]; log p(e|m)= -0.559:-0.559[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.047]; Lebanon[0.047]; Lebanon[0.047]; Lebanese[0.044]; Lebanese[0.044]; Lebanese[0.044]; Lebanese[0.044]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Syrian[0.042]; Beirut[0.042]; Jewish[0.040]; Israeli[0.039]; Moslem[0.038]; northern[0.038]; Christian[0.037]; military[0.037]; government[0.037]; territory[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.257:0.257[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.049]; Israel[0.049]; Israel[0.049]; Israel[0.049]; Israeli[0.044]; Lebanon[0.043]; Lebanon[0.043]; Lebanon[0.043]; Lebanon[0.043]; Lebanon[0.043]; Lebanon[0.043]; Jewish[0.042]; military[0.040]; Army[0.040]; army[0.040]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Syrian[0.039]; Beirut[0.038]; government[0.037]; war[0.037]; commander[0.037]; legal[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanon ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.265:0.265[0]; log p(e|m)= -0.409:-0.409[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBeirut[0.044]; Lebanese[0.044]; Lebanese[0.044]; Lebanese[0.044]; Lebanese[0.044]; Lebanese[0.044]; Lebanese[0.044]; Lebanese[0.044]; Lebanese[0.044]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Israeli[0.041]; Israeli[0.041]; Jewish[0.039]; military[0.039]; political[0.039]; peace[0.038]; territory[0.038]; allies[0.038]; Peres[0.037]; force[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shimon Peres\u001b[39m ==> ENTITY: \u001b[32mShimon Peres\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.240:0.240[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.048]; Israel[0.048]; Peres[0.048]; Israeli[0.046]; Israeli[0.046]; Israeli[0.046]; Israeli[0.046]; Netanyahu[0.044]; Jewish[0.043]; Lebanon[0.042]; Lebanon[0.042]; Lebanon[0.042]; Lebanese[0.040]; Lebanese[0.040]; Lebanese[0.040]; Lebanese[0.040]; leader[0.038]; Beirut[0.038]; Prime[0.038]; peace[0.038]; said[0.037]; said[0.037]; said[0.037]; said[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanese ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.252:0.252[0]; log p(e|m)= -0.559:-0.559[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.049]; Lebanon[0.049]; Lebanon[0.049]; Lebanese[0.045]; Lebanese[0.045]; Lebanese[0.045]; Israel[0.044]; Israel[0.044]; Beirut[0.043]; Jewish[0.041]; Israeli[0.041]; Israeli[0.041]; Israeli[0.041]; Israeli[0.041]; Cedars[0.040]; Christian[0.039]; Christian[0.039]; military[0.039]; political[0.038]; Guardians[0.038]; Moslem[0.038]; Saqr[0.037]; Saqr[0.037]; war[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanese ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.304:0.304[0]; log p(e|m)= -0.559:-0.559[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.047]; Lebanon[0.047]; Lebanon[0.047]; Lebanon[0.047]; Lebanon[0.047]; Lebanese[0.043]; Lebanese[0.043]; Lebanese[0.043]; Lebanese[0.043]; Israel[0.042]; Israel[0.042]; Israel[0.042]; Israel[0.042]; Israel[0.042]; Israel[0.042]; Israel[0.042]; Syrian[0.042]; Jewish[0.040]; northern[0.037]; territory[0.037]; government[0.037]; border[0.036]; war[0.036]; Antoine[0.035]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.171:0.171[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsraeli[0.049]; Israeli[0.049]; Israeli[0.049]; Lebanon[0.048]; Lebanon[0.048]; Lebanon[0.048]; Netanyahu[0.044]; Peres[0.044]; negotiations[0.042]; Christian[0.041]; Cedars[0.041]; Beirut[0.040]; military[0.040]; war[0.040]; said[0.040]; ousted[0.039]; civil[0.039]; hostile[0.038]; Benjamin[0.037]; Guardians[0.037]; insulted[0.037]; Saqr[0.037]; Saqr[0.037]; pro[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanon ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.172:0.172[0]; log p(e|m)= -0.409:-0.409[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.053]; Beirut[0.049]; Israel[0.048]; Israeli[0.045]; Israeli[0.045]; Israeli[0.045]; military[0.043]; Netanyahu[0.042]; war[0.041]; Peres[0.041]; militia[0.041]; civil[0.041]; negotiations[0.040]; Saqr[0.040]; Saqr[0.040]; Christian[0.039]; Cedars[0.039]; small[0.039]; pro[0.038]; leader[0.038]; absentia[0.038]; said[0.038]; hostile[0.038]; Benjamin[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = South Lebanon Army\u001b[39m ==> ENTITY: \u001b[32mSouth Lebanon Army\u001b[39m\t\nSCORES: global= 0.311:0.311[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.045]; Lebanon[0.045]; Lebanon[0.045]; Lebanon[0.045]; Lebanon[0.045]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; military[0.042]; Israeli[0.042]; Syrian[0.040]; army[0.039]; Beirut[0.038]; government[0.038]; border[0.038]; militia[0.037]; militia[0.037]; Jewish[0.037]; guerrilla[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanese ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.291:0.291[0]; log p(e|m)= -0.559:-0.559[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.047]; Lebanon[0.047]; Lebanese[0.044]; Lebanese[0.044]; Lebanese[0.044]; Lebanese[0.044]; Lebanese[0.044]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Syrian[0.042]; Beirut[0.042]; Jewish[0.040]; Jewish[0.040]; Israeli[0.039]; Moslem[0.038]; northern[0.038]; Christian[0.037]; military[0.037]; government[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanese ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.260:0.260[0]; log p(e|m)= -0.559:-0.559[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.049]; Lebanon[0.049]; Lebanon[0.049]; Lebanon[0.049]; Lebanon[0.049]; Lebanese[0.046]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Beirut[0.043]; Israeli[0.041]; northern[0.039]; military[0.039]; border[0.038]; South[0.037]; war[0.037]; major[0.037]; Haitham[0.037]; officially[0.037]; Antoine[0.037]; south[0.037]; south[0.037]; convicted[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.237:0.237[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.049]; Israel[0.049]; Israel[0.049]; Israel[0.049]; Israeli[0.045]; Israeli[0.045]; Lebanon[0.043]; Jewish[0.042]; Peres[0.041]; military[0.041]; army[0.040]; army[0.040]; army[0.040]; army[0.040]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Beirut[0.038]; soldiers[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.283:0.283[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.047]; Israel[0.047]; Israel[0.047]; Israel[0.047]; Israel[0.047]; Israel[0.047]; Israeli[0.043]; Lebanon[0.042]; Lebanon[0.042]; Lebanon[0.042]; Lebanon[0.042]; Lebanon[0.042]; Lebanon[0.042]; Jewish[0.040]; military[0.039]; Army[0.039]; army[0.038]; army[0.038]; Syrian[0.038]; Lebanese[0.038]; Lebanese[0.038]; Lebanese[0.038]; Lebanese[0.038]; Beirut[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanon ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.232:0.232[0]; log p(e|m)= -0.409:-0.409[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.050]; Beirut[0.046]; Lebanese[0.045]; Lebanese[0.045]; Lebanese[0.045]; Lebanese[0.045]; Israel[0.045]; Israel[0.045]; Israeli[0.042]; Israeli[0.042]; Israeli[0.042]; Israeli[0.042]; military[0.040]; political[0.040]; peace[0.040]; Netanyahu[0.039]; allies[0.039]; war[0.039]; Peres[0.038]; Peres[0.038]; militia[0.038]; force[0.038]; civil[0.038]; Jewish[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = SLA ==> ENTITY: \u001b[32mSouth Lebanon Army\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.196:0.196[0]; log p(e|m)= -1.666:-1.666[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.049]; Lebanon[0.049]; Israel[0.047]; Israel[0.047]; Lebanese[0.045]; Lebanese[0.045]; Lebanese[0.045]; Lebanese[0.045]; military[0.045]; Israeli[0.044]; Israeli[0.044]; Israeli[0.044]; Israeli[0.044]; army[0.043]; army[0.043]; militia[0.041]; Beirut[0.041]; peace[0.040]; Jewish[0.040]; commanders[0.040]; leader[0.040]; commander[0.039]; fighters[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jewish ==> ENTITY: \u001b[32mJews\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.157:0.157[0]; log p(e|m)= -0.805:-0.805[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.045]; Israel[0.045]; Israel[0.045]; Israel[0.045]; Israel[0.045]; Israel[0.045]; Syrian[0.043]; Lebanon[0.043]; Lebanon[0.043]; Lebanon[0.043]; Lebanon[0.043]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Beirut[0.038]; border[0.038]; northern[0.038]; said[0.038]; said[0.038]; war[0.038]; territory[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanese ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.284:0.284[0]; log p(e|m)= -0.559:-0.559[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.047]; Lebanon[0.047]; Lebanese[0.044]; Lebanese[0.044]; Lebanese[0.044]; Lebanese[0.044]; Lebanese[0.044]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Syrian[0.043]; Beirut[0.042]; Jewish[0.040]; Jewish[0.040]; Israeli[0.040]; Christian[0.038]; military[0.038]; Moslem[0.037]; territory[0.037]; south[0.036]; fighters[0.035]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanon ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.295:0.295[0]; log p(e|m)= -0.409:-0.409[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.047]; Lebanon[0.047]; Lebanon[0.047]; Lebanon[0.047]; Lebanon[0.047]; Beirut[0.043]; Lebanese[0.043]; Lebanese[0.043]; Lebanese[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Syrian[0.042]; Israeli[0.040]; government[0.038]; military[0.038]; northern[0.038]; peace[0.037]; South[0.037]; war[0.037]; Haitham[0.036]; border[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = SLA ==> ENTITY: \u001b[32mSouth Lebanon Army\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.215:0.215[0]; log p(e|m)= -1.666:-1.666[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.045]; Lebanon[0.045]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; military[0.041]; Israeli[0.041]; Israeli[0.041]; army[0.040]; army[0.040]; army[0.040]; civilians[0.038]; Beirut[0.038]; peace[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanon ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.229:0.229[0]; log p(e|m)= -0.409:-0.409[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.050]; Lebanon[0.050]; Lebanon[0.050]; Beirut[0.046]; Lebanese[0.045]; Lebanese[0.045]; Israel[0.045]; Israel[0.045]; Israel[0.045]; Israeli[0.042]; military[0.040]; South[0.039]; war[0.039]; Haitham[0.039]; border[0.039]; militia[0.038]; militia[0.038]; major[0.038]; army[0.038]; Friday[0.038]; south[0.037]; south[0.037]; Army[0.037]; officially[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.227:0.227[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israel[0.050]; Israel[0.050]; Israeli[0.045]; Lebanon[0.045]; Lebanon[0.045]; Lebanon[0.045]; Lebanon[0.045]; Lebanon[0.045]; military[0.042]; Army[0.041]; army[0.041]; Lebanese[0.040]; Lebanese[0.040]; Lebanese[0.040]; Beirut[0.039]; war[0.038]; commander[0.038]; northern[0.038]; border[0.037]; frontier[0.037]; February[0.037]; Friday[0.036]; general[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Peres\u001b[39m ==> ENTITY: \u001b[32mShimon Peres\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.240:0.240[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.054]; Israel[0.054]; Peres[0.053]; Israeli[0.052]; Israeli[0.052]; Israeli[0.052]; Israeli[0.052]; Netanyahu[0.049]; Jewish[0.048]; Lebanon[0.047]; Lebanon[0.047]; Lebanon[0.047]; Shimon[0.045]; Lebanese[0.045]; Lebanese[0.045]; Lebanese[0.045]; Lebanese[0.045]; leader[0.043]; Beirut[0.042]; Prime[0.042]; peace[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.242:0.242[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.052]; Israel[0.052]; Israeli[0.047]; Israeli[0.047]; Israeli[0.047]; Jewish[0.047]; Lebanon[0.046]; Lebanon[0.046]; Lebanon[0.046]; Shimon[0.044]; Peres[0.043]; Peres[0.043]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Lebanese[0.042]; Beirut[0.041]; Netanyahu[0.040]; army[0.039]; army[0.039]; military[0.039]; calling[0.038]; peace[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.275:0.275[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.049]; Israel[0.049]; Israel[0.049]; Israel[0.049]; Israel[0.049]; Israel[0.049]; Lebanon[0.043]; Lebanon[0.043]; Lebanon[0.043]; Lebanon[0.043]; Jewish[0.042]; Syrian[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Lebanese[0.039]; Beirut[0.038]; territory[0.037]; military[0.037]; government[0.037]; war[0.037]; legal[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Beirut ==> ENTITY: \u001b[32mBeirut\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.205:0.205[0]; log p(e|m)= -0.047:-0.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.046]; Lebanon[0.046]; Israeli[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Lebanese[0.043]; Lebanese[0.043]; Lebanese[0.043]; Lebanese[0.043]; Lebanese[0.043]; Lebanese[0.043]; Syrian[0.042]; military[0.041]; government[0.039]; civilians[0.038]; army[0.038]; army[0.038]; army[0.038]; army[0.038]; soldiers[0.038]; \t\n\r [==========================>...................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s346ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2720/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1325testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = GMT\u001b[39m ==> ENTITY: \u001b[32mGreenwich Mean Time\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.098:0.098[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThursday[0.059]; Friday[0.059]; London[0.058]; London[0.058]; London[0.058]; Global[0.055]; represents[0.054]; close[0.052]; Jakarta[0.052]; previous[0.052]; low[0.051]; high[0.051]; common[0.051]; recorded[0.051]; traded[0.049]; newsroom[0.049]; Receipt[0.048]; Depository[0.048]; shares[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.065:0.065[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.067]; London[0.067]; recorded[0.060]; shares[0.059]; traded[0.058]; Depository[0.058]; Thursday[0.056]; Global[0.056]; Friday[0.055]; previous[0.055]; represents[0.055]; close[0.053]; high[0.052]; Receipt[0.052]; low[0.052]; Jakarta[0.049]; common[0.049]; newsroom[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.065:0.065[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.067]; London[0.067]; recorded[0.060]; shares[0.059]; traded[0.058]; Depository[0.058]; Thursday[0.056]; Global[0.056]; Friday[0.055]; previous[0.055]; represents[0.055]; close[0.053]; high[0.052]; Receipt[0.052]; low[0.052]; Jakarta[0.049]; common[0.049]; newsroom[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jakarta\u001b[39m ==> ENTITY: \u001b[32mJakarta\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.062:0.062[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThursday[0.052]; London[0.051]; London[0.051]; London[0.051]; Friday[0.050]; low[0.050]; high[0.050]; Global[0.049]; recorded[0.048]; represents[0.048]; Receipt[0.047]; previous[0.046]; Timah[0.046]; Timah[0.046]; close[0.046]; newsroom[0.046]; Pt[0.046]; Depository[0.045]; common[0.045]; shares[0.044]; traded[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = GMT\u001b[39m ==> ENTITY: \u001b[32mGreenwich Mean Time\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.098:0.098[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThursday[0.059]; Friday[0.059]; London[0.058]; London[0.058]; London[0.058]; Global[0.055]; represents[0.054]; close[0.052]; Jakarta[0.052]; previous[0.052]; low[0.051]; high[0.051]; common[0.051]; recorded[0.051]; traded[0.049]; newsroom[0.049]; Receipt[0.048]; Depository[0.048]; shares[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.065:0.065[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.067]; London[0.067]; recorded[0.060]; shares[0.059]; traded[0.058]; Depository[0.058]; Thursday[0.056]; Global[0.056]; Friday[0.055]; previous[0.055]; represents[0.055]; close[0.053]; high[0.052]; Receipt[0.052]; low[0.052]; Jakarta[0.049]; common[0.049]; newsroom[0.048]; \t\n\r [==========================>...................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s319ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2726/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1235testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew York\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew York\u001b[39m <---> \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.239:0.237[\u001b[31m0.002\u001b[39m]; local()= -0.014:-0.017[\u001b[31m0.003\u001b[39m]; log p(e|m)= -0.289:-1.790[\u001b[31m1.501\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYork[0.055]; Sea[0.052]; Sea[0.052]; Mediterranean[0.051]; Dirty[0.050]; Desk[0.049]; Mobil[0.048]; Red[0.047]; Red[0.047]; Augusta[0.047]; Augusta[0.047]; Mideast[0.047]; Thai[0.047]; tanker[0.047]; Spetses[0.046]; Exxon[0.046]; Exxon[0.046]; Exxon[0.046]; Resource[0.044]; Ras[0.044]; Lula[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Exxon\u001b[39m ==> ENTITY: \u001b[32mExxon\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.013:0.013[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nExxon[0.052]; Exxon[0.052]; Mobil[0.048]; Sea[0.046]; Sea[0.046]; tanker[0.045]; York[0.044]; York[0.044]; Red[0.044]; Red[0.044]; Mediterranean[0.043]; Ras[0.042]; Thai[0.042]; Commodities[0.042]; Fos[0.042]; Desk[0.042]; Augusta[0.041]; Augusta[0.041]; fixtures[0.041]; Dirty[0.040]; Spetses[0.040]; Mideast[0.040]; Resource[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Spetses\u001b[39m ==> ENTITY: \u001b[32mSpetses\u001b[39m\t\nSCORES: global= 0.229:0.229[0]; local()= 0.024:0.024[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMediterranean[0.052]; Sea[0.047]; Sea[0.047]; tanker[0.047]; Ras[0.045]; Commodities[0.044]; Fos[0.044]; Desk[0.044]; Mobil[0.044]; fixtures[0.043]; Thai[0.043]; Mideast[0.043]; Exxon[0.043]; Exxon[0.043]; Exxon[0.043]; Resource[0.042]; Red[0.042]; Red[0.042]; Sidi[0.041]; Sidi[0.041]; Augusta[0.041]; Augusta[0.041]; Dirty[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mobil\u001b[39m ==> ENTITY: \u001b[32mMobil\u001b[39m\t\nSCORES: global= 0.231:0.231[0]; local()= 0.040:0.040[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nExxon[0.051]; Exxon[0.051]; Exxon[0.051]; tanker[0.044]; York[0.042]; York[0.042]; Ras[0.042]; Fos[0.041]; Commodities[0.041]; Desk[0.041]; Red[0.041]; Red[0.041]; Sea[0.041]; Sea[0.041]; Resource[0.040]; Augusta[0.040]; Augusta[0.040]; fixtures[0.040]; Mideast[0.039]; Thai[0.039]; Mediterranean[0.038]; Dirty[0.038]; Sidi[0.038]; Sidi[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Exxon\u001b[39m ==> ENTITY: \u001b[32mExxon\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.013:0.013[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nExxon[0.052]; Exxon[0.052]; Mobil[0.048]; Sea[0.046]; Sea[0.046]; tanker[0.045]; York[0.044]; York[0.044]; Red[0.044]; Red[0.044]; Mediterranean[0.043]; Ras[0.042]; Thai[0.042]; Commodities[0.042]; Fos[0.042]; Desk[0.042]; Augusta[0.041]; Augusta[0.041]; fixtures[0.041]; Dirty[0.040]; Spetses[0.040]; Mideast[0.040]; Resource[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = GMT\u001b[39m ==> ENTITY: \u001b[32mGreenwich Mean Time\u001b[39m\t\nSCORES: global= 0.234:0.234[0]; local()= 0.020:0.020[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMediterranean[0.047]; Mideast[0.045]; Sea[0.044]; Sea[0.044]; Thai[0.044]; York[0.043]; York[0.043]; Ras[0.042]; Desk[0.042]; Commodities[0.042]; Fos[0.042]; Mobil[0.042]; fixtures[0.042]; tanker[0.041]; Dirty[0.041]; Spetses[0.040]; Exxon[0.040]; Exxon[0.040]; Exxon[0.040]; Resource[0.040]; Red[0.039]; Red[0.039]; Sidi[0.039]; Sidi[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ras Tanura\u001b[39m ==> ENTITY: \u001b[32mRas Tanura\u001b[39m\t\nSCORES: global= 0.218:0.218[0]; local()= 0.056:0.056[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntanker[0.055]; Mediterranean[0.055]; Sea[0.053]; Sea[0.053]; Mideast[0.051]; Exxon[0.051]; Exxon[0.051]; Exxon[0.051]; Commodities[0.051]; Fos[0.051]; Thai[0.050]; Desk[0.050]; Mobil[0.050]; fixtures[0.049]; Resource[0.048]; Sidi[0.047]; Sidi[0.047]; Spetses[0.046]; York[0.046]; York[0.046]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mFos\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFos, Hérault\u001b[39m <---> \u001b[32mFos-sur-Mer\u001b[39m\t\nSCORES: global= 0.227:0.214[\u001b[31m0.013\u001b[39m]; local()= 0.015:0.066[\u001b[32m0.052\u001b[39m]; log p(e|m)= -0.155:-2.263[\u001b[31m2.108\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMediterranean[0.048]; Sea[0.046]; Sea[0.046]; tanker[0.045]; Exxon[0.044]; Exxon[0.044]; Exxon[0.044]; Spetses[0.044]; Red[0.044]; Red[0.044]; Ras[0.044]; Commodities[0.043]; Augusta[0.043]; Augusta[0.043]; Desk[0.043]; Mobil[0.043]; Lula[0.043]; York[0.042]; York[0.042]; Dirty[0.042]; fixtures[0.042]; Mideast[0.041]; Resource[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mediterranean ==> ENTITY: \u001b[32mMediterranean Sea\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.040:0.040[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSea[0.047]; Sea[0.047]; Mideast[0.045]; York[0.043]; York[0.043]; Spetses[0.043]; fixtures[0.043]; Sidi[0.043]; Sidi[0.043]; Exxon[0.042]; Exxon[0.042]; Exxon[0.042]; tanker[0.042]; Augusta[0.041]; Augusta[0.041]; Red[0.040]; Red[0.040]; Thai[0.040]; Fos[0.040]; Commodities[0.039]; Resource[0.039]; Ras[0.038]; Desk[0.038]; Lula[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Red Sea\u001b[39m ==> ENTITY: \u001b[32mRed Sea\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.049:0.049[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMediterranean[0.053]; Sidi[0.046]; Sidi[0.046]; Sea[0.045]; Mideast[0.044]; Red[0.044]; Ras[0.044]; Thai[0.042]; Exxon[0.041]; Exxon[0.041]; Exxon[0.041]; Dirty[0.041]; York[0.041]; York[0.041]; tanker[0.040]; Spetses[0.040]; Fos[0.040]; Desk[0.040]; Mobil[0.039]; Resource[0.039]; Commodities[0.039]; fixtures[0.038]; Augusta[0.038]; Augusta[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Exxon\u001b[39m ==> ENTITY: \u001b[32mExxon\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.013:0.013[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nExxon[0.052]; Exxon[0.052]; Mobil[0.048]; Sea[0.046]; Sea[0.046]; tanker[0.045]; York[0.044]; York[0.044]; Red[0.044]; Red[0.044]; Mediterranean[0.043]; Ras[0.042]; Thai[0.042]; Commodities[0.042]; Fos[0.042]; Desk[0.042]; Augusta[0.041]; Augusta[0.041]; fixtures[0.041]; Dirty[0.040]; Spetses[0.040]; Mideast[0.040]; Resource[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Red Sea\u001b[39m ==> ENTITY: \u001b[32mRed Sea\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.049:0.049[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMediterranean[0.053]; Sidi[0.046]; Sidi[0.046]; Sea[0.045]; Mideast[0.044]; Red[0.044]; Ras[0.044]; Thai[0.042]; Exxon[0.041]; Exxon[0.041]; Exxon[0.041]; Dirty[0.041]; York[0.041]; York[0.041]; tanker[0.040]; Spetses[0.040]; Fos[0.040]; Desk[0.040]; Mobil[0.039]; Resource[0.039]; Commodities[0.039]; fixtures[0.038]; Augusta[0.038]; Augusta[0.038]; \t\n\r [===========================>..................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s283ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2738/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1212testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Durban ==> ENTITY: \u001b[32mDurban\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.170:0.170[0]; log p(e|m)= -0.075:-0.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKwaZulu[0.047]; Anc[0.047]; Mandela[0.047]; Natal[0.046]; South[0.044]; South[0.044]; Africa[0.044]; Africa[0.044]; Africa[0.044]; African[0.044]; Zulu[0.043]; Zulu[0.043]; south[0.041]; Thursday[0.039]; night[0.039]; Friday[0.039]; coast[0.039]; province[0.039]; province[0.039]; youth[0.038]; Province[0.038]; youths[0.038]; National[0.038]; supporters[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = ANC ==> ENTITY: \u001b[32mAfrican National Congress\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.059:-0.059[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrican[0.045]; South[0.044]; South[0.044]; Mandela[0.044]; National[0.043]; KwaZulu[0.043]; Durban[0.042]; attacks[0.042]; Africa[0.041]; Africa[0.041]; Africa[0.041]; attack[0.041]; political[0.041]; province[0.041]; province[0.041]; Congress[0.041]; opened[0.041]; linked[0.041]; Zulu[0.041]; Zulu[0.041]; Province[0.040]; supporters[0.040]; police[0.040]; police[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mZulu\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mZulu language\u001b[39m <---> \u001b[32mZulu Kingdom\u001b[39m\t\nSCORES: global= 0.259:0.245[\u001b[31m0.014\u001b[39m]; local()= 0.103:0.123[\u001b[32m0.020\u001b[39m]; log p(e|m)= -0.209:-2.564[\u001b[31m2.354\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKwaZulu[0.046]; hours[0.045]; old[0.044]; came[0.044]; south[0.043]; home[0.042]; Natal[0.042]; people[0.042]; people[0.042]; region[0.042]; massacred[0.042]; Inkatha[0.042]; season[0.041]; province[0.041]; lives[0.041]; wars[0.041]; Mangosuthu[0.041]; coast[0.041]; Buthelezi[0.040]; lost[0.040]; brewed[0.040]; Freedom[0.040]; recent[0.039]; villages[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = KwaZulu-Natal\u001b[39m ==> ENTITY: \u001b[32mKwaZulu-Natal\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.050]; South[0.050]; KwaZulu[0.048]; season[0.046]; African[0.046]; Durban[0.045]; Natal[0.045]; Natal[0.045]; south[0.045]; south[0.045]; south[0.045]; Mandela[0.045]; Thursday[0.045]; Thursday[0.045]; Friday[0.045]; rural[0.045]; region[0.044]; Province[0.044]; Umkomaas[0.044]; National[0.044]; Zulu[0.044]; Zulu[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nelson Mandela\u001b[39m ==> ENTITY: \u001b[32mNelson Mandela\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrican[0.045]; South[0.045]; South[0.045]; Africa[0.042]; Africa[0.042]; Africa[0.042]; National[0.042]; President[0.042]; KwaZulu[0.042]; people[0.042]; Zulu[0.042]; Zulu[0.042]; youth[0.041]; political[0.041]; Durban[0.041]; rural[0.041]; supporters[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; hours[0.040]; Umkomaas[0.040]; Congress[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mangosuthu Buthelezi\u001b[39m ==> ENTITY: \u001b[32mMangosuthu Buthelezi\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.117:0.117[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKwaZulu[0.047]; Inkatha[0.045]; Zulu[0.045]; political[0.044]; political[0.044]; Anc[0.044]; said[0.043]; said[0.043]; Natal[0.043]; Party[0.043]; Freedom[0.042]; season[0.041]; keen[0.041]; province[0.040]; region[0.040]; peaceful[0.040]; pointed[0.040]; decade[0.040]; violence[0.039]; lives[0.039]; Chief[0.039]; villages[0.039]; festive[0.039]; recent[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Umkomaas\u001b[39m ==> ENTITY: \u001b[32mUmkomaas\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncoast[0.046]; coast[0.046]; coast[0.046]; near[0.046]; South[0.045]; south[0.043]; south[0.043]; south[0.043]; region[0.042]; rural[0.042]; KwaZulu[0.042]; KwaZulu[0.042]; African[0.041]; Natal[0.040]; Natal[0.040]; recent[0.040]; cities[0.040]; season[0.040]; attack[0.040]; Zulu[0.039]; wounded[0.039]; thousands[0.039]; migrant[0.039]; believed[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mZulu\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mZulu language\u001b[39m <---> \u001b[32mZulu Kingdom\u001b[39m\t\nSCORES: global= 0.263:0.244[\u001b[31m0.019\u001b[39m]; local()= 0.189:0.152[\u001b[31m0.038\u001b[39m]; log p(e|m)= -0.209:-2.564[\u001b[31m2.354\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZulu[0.049]; African[0.047]; South[0.044]; South[0.044]; KwaZulu[0.044]; Thursday[0.043]; Friday[0.042]; rural[0.042]; Africa[0.042]; Africa[0.042]; Africa[0.042]; south[0.041]; Mandela[0.041]; Natal[0.041]; people[0.041]; Province[0.040]; Durban[0.040]; province[0.040]; said[0.039]; said[0.039]; said[0.039]; believed[0.039]; police[0.039]; police[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Africa ==> ENTITY: \u001b[32mSouth Africa\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.263:-0.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.046]; African[0.046]; Africa[0.045]; Africa[0.045]; Durban[0.045]; KwaZulu[0.043]; Province[0.042]; Mandela[0.042]; Thursday[0.042]; Thursday[0.042]; National[0.041]; Friday[0.041]; Anc[0.041]; President[0.041]; Zulu[0.040]; Zulu[0.040]; settlement[0.040]; Congress[0.040]; province[0.040]; province[0.040]; night[0.040]; night[0.040]; Natal[0.040]; youth[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Inkatha Freedom Party\u001b[39m ==> ENTITY: \u001b[32mInkatha Freedom Party\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.127:0.127[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMangosuthu[0.049]; KwaZulu[0.046]; Buthelezi[0.045]; political[0.044]; political[0.044]; recent[0.043]; Zulu[0.042]; province[0.042]; region[0.042]; Natal[0.042]; said[0.041]; said[0.041]; feuding[0.041]; season[0.041]; upset[0.041]; Anc[0.041]; pointed[0.040]; keen[0.040]; villages[0.040]; festive[0.040]; decade[0.039]; Chief[0.039]; Christmas[0.039]; tense[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.078:0.078[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntold[0.044]; cities[0.043]; region[0.043]; Congress[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; migrant[0.042]; rural[0.042]; Thursday[0.042]; Thursday[0.042]; south[0.041]; south[0.041]; south[0.041]; workers[0.041]; recent[0.041]; National[0.041]; massacred[0.041]; decade[0.041]; thousands[0.041]; linked[0.041]; killed[0.041]; opened[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = African National Congress\u001b[39m ==> ENTITY: \u001b[32mAfrican National Congress\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.045]; South[0.045]; Mandela[0.044]; President[0.044]; Zulu[0.043]; Zulu[0.043]; KwaZulu[0.043]; supporters[0.043]; Africa[0.042]; Africa[0.042]; Africa[0.042]; linked[0.042]; political[0.042]; rural[0.042]; Durban[0.041]; politics[0.041]; youth[0.040]; Anc[0.040]; violence[0.039]; Naidoo[0.039]; said[0.039]; said[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Africa ==> ENTITY: \u001b[32mSouth Africa\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.263:-0.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.048]; African[0.048]; Africa[0.047]; Africa[0.047]; Durban[0.047]; KwaZulu[0.045]; Province[0.044]; Mandela[0.044]; Thursday[0.043]; National[0.043]; Friday[0.043]; Anc[0.043]; President[0.042]; Zulu[0.042]; Zulu[0.042]; settlement[0.042]; Congress[0.042]; province[0.042]; province[0.042]; night[0.042]; Natal[0.041]; youth[0.041]; opened[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = KwaZulu-Natal\u001b[39m ==> ENTITY: \u001b[32mKwaZulu-Natal\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKwaZulu[0.045]; season[0.043]; Mangosuthu[0.043]; Natal[0.042]; Natal[0.042]; south[0.042]; south[0.042]; south[0.042]; Thursday[0.042]; Thursday[0.042]; rural[0.042]; region[0.041]; Buthelezi[0.041]; Umkomaas[0.041]; Zulu[0.041]; coast[0.041]; coast[0.041]; coast[0.041]; province[0.041]; province[0.041]; province[0.041]; home[0.041]; Inkatha[0.040]; scores[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mZulu\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mZulu language\u001b[39m <---> \u001b[32mZulu Kingdom\u001b[39m\t\nSCORES: global= 0.261:0.241[\u001b[31m0.020\u001b[39m]; local()= 0.204:0.161[\u001b[31m0.043\u001b[39m]; log p(e|m)= -0.209:-2.564[\u001b[31m2.354\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZulu[0.050]; African[0.049]; South[0.046]; South[0.046]; KwaZulu[0.045]; Thursday[0.045]; Thursday[0.045]; Friday[0.044]; rural[0.043]; Africa[0.043]; Africa[0.043]; Africa[0.043]; south[0.043]; south[0.043]; Mandela[0.043]; Natal[0.042]; people[0.042]; Province[0.042]; Durban[0.041]; province[0.041]; province[0.041]; believed[0.041]; Umkomaas[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = ANC ==> ENTITY: \u001b[32mAfrican National Congress\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.124:0.124[0]; log p(e|m)= -0.059:-0.059[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInkatha[0.046]; Mangosuthu[0.045]; KwaZulu[0.044]; Buthelezi[0.044]; region[0.043]; decade[0.043]; political[0.042]; political[0.042]; potential[0.042]; Zulu[0.042]; Freedom[0.041]; recent[0.041]; Party[0.041]; keen[0.041]; people[0.041]; people[0.041]; cities[0.041]; hours[0.041]; south[0.040]; wars[0.040]; spots[0.040]; scores[0.040]; said[0.040]; said[0.040]; \t\n\r [===========================>..................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s209ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2754/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1310testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.063:0.063[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUnited[0.048]; States[0.048]; State[0.046]; Department[0.045]; Thursday[0.043]; said[0.042]; Burns[0.041]; release[0.041]; release[0.041]; world[0.041]; stand[0.041]; need[0.041]; convicted[0.041]; convicted[0.041]; Nicholas[0.040]; airliner[0.040]; terrorist[0.040]; does[0.040]; danger[0.040]; saying[0.040]; blasted[0.040]; prison[0.040]; teenager[0.040]; deserves[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Athens ==> ENTITY: \u001b[32mAthens\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.288:-0.288[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairport[0.048]; Greece[0.048]; Greece[0.048]; Olympic[0.047]; Greek[0.045]; Greek[0.045]; flight[0.044]; transit[0.044]; outside[0.043]; government[0.043]; Korydallos[0.043]; Thursday[0.043]; air[0.043]; headquarters[0.042]; early[0.042]; Cairo[0.042]; mid[0.042]; ruled[0.042]; Clinton[0.042]; recently[0.041]; American[0.041]; Palestine[0.041]; security[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burns ==> ENTITY: \u001b[32mR. Nicholas Burns\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.017:-0.017[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDepartment[0.045]; Burns[0.045]; Burns[0.045]; Thursday[0.043]; Palestine[0.043]; Nicholas[0.043]; briefing[0.043]; said[0.042]; said[0.042]; State[0.042]; teenager[0.042]; administration[0.042]; Clinton[0.041]; security[0.041]; told[0.041]; spokesman[0.041]; Organisation[0.041]; serving[0.040]; serving[0.040]; news[0.040]; deserves[0.040]; Mahammad[0.040]; world[0.040]; conveyed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.045:0.045[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.050]; States[0.047]; mid[0.045]; plane[0.044]; pre[0.044]; term[0.044]; Brazil[0.044]; air[0.043]; group[0.043]; Athens[0.043]; airliner[0.043]; bomb[0.043]; bomb[0.043]; testified[0.043]; guerrilla[0.043]; said[0.043]; said[0.043]; planted[0.043]; United[0.042]; accomplished[0.042]; evidence[0.042]; approaching[0.042]; counted[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = State Department\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of State\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecurity[0.049]; Washington[0.048]; Clinton[0.048]; administration[0.046]; States[0.044]; terrorist[0.043]; world[0.043]; release[0.043]; release[0.043]; news[0.043]; views[0.043]; briefing[0.043]; maximum[0.042]; government[0.042]; Burns[0.042]; Burns[0.042]; Burns[0.042]; Thursday[0.042]; Thursday[0.042]; said[0.042]; said[0.042]; serving[0.042]; serving[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pan American ==> ENTITY: \u001b[32mPan American World Airways\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.067:0.067[0]; log p(e|m)= -0.790:-0.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.050]; flight[0.045]; held[0.043]; Airways[0.043]; student[0.043]; airport[0.042]; recently[0.042]; Thursday[0.042]; mid[0.042]; United[0.042]; airliner[0.042]; early[0.041]; time[0.041]; master[0.041]; Cairo[0.041]; regular[0.041]; Organisation[0.040]; Athens[0.040]; Palestinian[0.040]; Palestinian[0.040]; States[0.040]; Palestine[0.040]; security[0.040]; air[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazil ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.064:0.064[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStates[0.046]; American[0.044]; term[0.043]; Pan[0.043]; time[0.043]; United[0.042]; citizens[0.042]; said[0.042]; said[0.042]; guerrilla[0.042]; freed[0.042]; recently[0.042]; group[0.042]; pre[0.041]; court[0.041]; serving[0.041]; ruled[0.041]; mid[0.041]; Greece[0.041]; held[0.040]; accomplished[0.040]; expelled[0.040]; accuses[0.040]; immediately[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cairo ==> ENTITY: \u001b[32mCairo\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.124:0.124[0]; log p(e|m)= -0.152:-0.152[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRashid[0.045]; Rashid[0.045]; Rashid[0.045]; Rashid[0.045]; Rashid[0.045]; airport[0.045]; Palestine[0.043]; Tunis[0.043]; said[0.042]; said[0.042]; Mahammad[0.042]; flight[0.041]; news[0.041]; Airways[0.041]; Thursday[0.041]; outside[0.041]; headquarters[0.040]; Athens[0.040]; security[0.039]; serving[0.039]; serving[0.039]; serving[0.039]; told[0.039]; boarded[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clinton ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.093:0.093[0]; log p(e|m)= -1.917:-1.917[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nadministration[0.046]; American[0.045]; State[0.044]; Department[0.044]; outside[0.043]; serving[0.042]; serving[0.042]; Thursday[0.042]; mid[0.042]; boarded[0.042]; Athens[0.042]; security[0.041]; early[0.041]; airport[0.041]; Palestine[0.041]; headquarters[0.041]; Burns[0.041]; Burns[0.041]; Burns[0.041]; Cairo[0.040]; does[0.040]; said[0.040]; said[0.040]; government[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nicholas Burns ==> ENTITY: \u001b[32mR. Nicholas Burns\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.017:-0.017[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDepartment[0.045]; Burns[0.044]; Burns[0.044]; Washington[0.044]; Thursday[0.043]; Thursday[0.043]; States[0.043]; briefing[0.042]; said[0.042]; said[0.042]; State[0.042]; teenager[0.042]; administration[0.042]; Clinton[0.041]; Palestinian[0.041]; security[0.041]; told[0.041]; spokesman[0.040]; serving[0.040]; serving[0.040]; news[0.040]; deserves[0.040]; Mahammad[0.040]; world[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.064:0.064[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWashington[0.046]; world[0.044]; Thursday[0.043]; serving[0.043]; does[0.043]; State[0.043]; Department[0.042]; Greece[0.042]; allowed[0.042]; said[0.042]; stand[0.041]; deserves[0.041]; release[0.041]; release[0.041]; Burns[0.041]; teenager[0.041]; leave[0.041]; blasted[0.041]; Nicholas[0.040]; saying[0.040]; inexplicable[0.040]; let[0.040]; make[0.040]; need[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Greece ==> ENTITY: \u001b[32mGreece\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.124:0.124[0]; log p(e|m)= -0.276:-0.276[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGreek[0.048]; Athens[0.046]; Olympic[0.044]; group[0.043]; Brazil[0.043]; Palestine[0.042]; Korydallos[0.042]; States[0.042]; guerrilla[0.042]; said[0.041]; said[0.041]; United[0.041]; Organisation[0.041]; Palestinian[0.041]; Palestinian[0.041]; ruled[0.041]; time[0.040]; mid[0.040]; mid[0.040]; Tunis[0.040]; recently[0.040]; American[0.040]; American[0.040]; held[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burns ==> ENTITY: \u001b[32mR. Nicholas Burns\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.017:-0.017[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDepartment[0.045]; Burns[0.044]; Burns[0.044]; American[0.044]; Thursday[0.043]; Palestine[0.043]; Nicholas[0.042]; briefing[0.042]; said[0.042]; said[0.042]; State[0.042]; administration[0.042]; time[0.042]; Clinton[0.041]; security[0.041]; told[0.041]; spokesman[0.040]; Organisation[0.040]; serving[0.040]; serving[0.040]; serving[0.040]; news[0.040]; deserves[0.040]; Mahammad[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Greece ==> ENTITY: \u001b[32mGreece\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.276:-0.276[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGreece[0.051]; Greek[0.046]; Greek[0.046]; Athens[0.045]; government[0.043]; Olympic[0.043]; Palestine[0.041]; Korydallos[0.041]; world[0.041]; States[0.041]; saying[0.041]; guerrilla[0.041]; Liberation[0.040]; said[0.040]; said[0.040]; administration[0.040]; security[0.040]; terrorist[0.040]; outside[0.040]; United[0.040]; news[0.040]; serving[0.040]; serving[0.040]; Palestinian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = TWA\u001b[39m ==> ENTITY: \u001b[32mTrans World Airlines\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.087:0.087[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.046]; mid[0.045]; plane[0.044]; recently[0.044]; term[0.043]; United[0.043]; air[0.042]; States[0.042]; airliner[0.042]; group[0.041]; serving[0.041]; Pan[0.041]; ample[0.041]; counted[0.041]; held[0.041]; belonging[0.041]; accuses[0.041]; agents[0.040]; time[0.040]; said[0.040]; said[0.040]; expelled[0.040]; Brazil[0.040]; planted[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tunis ==> ENTITY: \u001b[32mTunis\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.033:-0.033[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCairo[0.051]; Palestine[0.048]; government[0.047]; security[0.047]; Greece[0.046]; Greece[0.046]; Greece[0.046]; strong[0.045]; serving[0.045]; serving[0.045]; serving[0.045]; reduced[0.045]; American[0.045]; left[0.045]; Thursday[0.044]; outside[0.044]; Palestinian[0.044]; expelled[0.044]; administration[0.044]; flight[0.044]; group[0.044]; allowed[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = FBI\u001b[39m ==> ENTITY: \u001b[32mFederal Bureau of Investigation\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.078:0.078[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbomb[0.047]; bomb[0.047]; bombing[0.046]; accuses[0.046]; agents[0.045]; bombmaker[0.045]; recently[0.044]; American[0.044]; American[0.044]; Abu[0.044]; trial[0.043]; trial[0.043]; convicted[0.043]; Ibrahim[0.043]; group[0.043]; States[0.043]; held[0.042]; testified[0.042]; plane[0.042]; Rashid[0.041]; Rashid[0.041]; Rashid[0.041]; Rashid[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Olympic Airways\u001b[39m ==> ENTITY: \u001b[32mOlympic Airlines\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.129:0.129[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAthens[0.046]; Greece[0.045]; Greece[0.045]; Greece[0.045]; Greek[0.044]; Greek[0.044]; flight[0.044]; airport[0.043]; airliner[0.042]; air[0.041]; Thursday[0.041]; told[0.041]; United[0.040]; government[0.040]; recently[0.040]; outside[0.040]; American[0.040]; headquarters[0.040]; Mahammad[0.040]; mid[0.040]; transit[0.040]; pre[0.040]; administration[0.039]; regular[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Greece ==> ENTITY: \u001b[32mGreece\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.276:-0.276[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGreece[0.052]; Greek[0.047]; Greek[0.047]; Athens[0.046]; Olympic[0.043]; government[0.042]; Korydallos[0.042]; world[0.042]; early[0.041]; Liberation[0.041]; Palestine[0.041]; outside[0.041]; serving[0.040]; serving[0.040]; Organisation[0.040]; Thursday[0.040]; terrorist[0.040]; mid[0.040]; Tunis[0.040]; security[0.039]; Department[0.039]; strong[0.039]; saying[0.039]; news[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.062:0.062[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrelease[0.046]; release[0.046]; Washington[0.046]; States[0.044]; Department[0.044]; does[0.043]; world[0.042]; saying[0.042]; bars[0.042]; Thursday[0.042]; inexplicable[0.041]; State[0.041]; bomber[0.041]; need[0.041]; sense[0.041]; airliner[0.040]; guerrilla[0.040]; said[0.040]; United[0.040]; teenager[0.040]; Burns[0.040]; stand[0.040]; bombing[0.040]; make[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Greek ==> ENTITY: \u001b[32mGreece\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.103:0.103[0]; log p(e|m)= -2.323:-2.323[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGreece[0.057]; guerrilla[0.045]; Thursday[0.044]; saying[0.044]; States[0.044]; civilians[0.044]; world[0.044]; Department[0.044]; said[0.043]; ca[0.043]; serving[0.043]; Washington[0.043]; does[0.042]; Nicholas[0.042]; teenager[0.042]; spokesman[0.042]; sense[0.042]; bombing[0.042]; bars[0.042]; allowed[0.042]; State[0.042]; leave[0.042]; terrorist[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.063:0.063[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.050]; Brazil[0.045]; American[0.045]; American[0.045]; Athens[0.044]; group[0.043]; time[0.043]; said[0.042]; said[0.042]; serving[0.041]; pre[0.041]; recently[0.041]; mid[0.041]; mid[0.041]; counted[0.040]; Greece[0.040]; agents[0.040]; regular[0.040]; held[0.040]; Palestine[0.040]; Pan[0.039]; Pan[0.039]; belonging[0.039]; accomplished[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Athens ==> ENTITY: \u001b[32mAthens\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.088:0.088[0]; log p(e|m)= -0.288:-0.288[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGreece[0.048]; citizens[0.045]; States[0.044]; Korydallos[0.043]; air[0.043]; mid[0.043]; American[0.042]; said[0.042]; said[0.042]; student[0.042]; plane[0.042]; group[0.041]; United[0.041]; accuses[0.041]; term[0.040]; time[0.040]; serving[0.040]; airliner[0.040]; held[0.040]; prison[0.040]; counted[0.040]; killed[0.039]; Pan[0.039]; master[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Greek ==> ENTITY: \u001b[32mGreece\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.166:0.166[0]; log p(e|m)= -2.323:-2.323[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGreece[0.055]; Greece[0.055]; Greek[0.050]; Athens[0.048]; Olympic[0.045]; Palestine[0.043]; government[0.043]; guerrilla[0.042]; American[0.042]; Thursday[0.042]; security[0.042]; States[0.042]; spoke[0.041]; term[0.041]; said[0.041]; said[0.041]; recently[0.041]; Korydallos[0.041]; belonging[0.041]; group[0.041]; outside[0.041]; ruled[0.041]; early[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Greek ==> ENTITY: \u001b[32mGreece\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.169:0.169[0]; log p(e|m)= -2.323:-2.323[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGreece[0.053]; Greece[0.053]; Greek[0.049]; Athens[0.046]; Olympic[0.043]; Palestine[0.042]; government[0.041]; American[0.041]; world[0.040]; spoke[0.040]; Department[0.040]; strong[0.040]; said[0.040]; said[0.040]; Cairo[0.040]; Korydallos[0.040]; early[0.040]; outside[0.040]; ruled[0.039]; mid[0.039]; recently[0.039]; news[0.039]; does[0.039]; Nicholas[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pan American ==> ENTITY: \u001b[32mPan American World Airways\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.078:0.078[0]; log p(e|m)= -0.790:-0.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplane[0.045]; held[0.044]; student[0.044]; Brazil[0.044]; recently[0.043]; mid[0.043]; United[0.043]; airliner[0.043]; airliner[0.043]; time[0.043]; master[0.042]; Athens[0.041]; Palestinian[0.041]; Palestinian[0.041]; States[0.041]; air[0.041]; group[0.041]; said[0.040]; said[0.040]; serving[0.040]; immediately[0.040]; expelled[0.039]; reduced[0.039]; term[0.039]; \t\n\r [===========================>..................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s21ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2783/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1300testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.064:0.064[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.052]; Dubai[0.044]; news[0.044]; said[0.044]; Thursday[0.043]; Thursday[0.043]; Wednesday[0.043]; seven[0.042]; Saudi[0.042]; Light[0.041]; Light[0.041]; price[0.041]; price[0.041]; agency[0.041]; Mexico[0.041]; Arabia[0.040]; quoting[0.040]; Newsroom[0.040]; Algeria[0.039]; Indonesia[0.039]; Arabian[0.039]; stood[0.039]; Nigeria[0.039]; Juana[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Saudi Arabia ==> ENTITY: \u001b[32mSaudi Arabia\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.255:-0.255[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlgeria[0.050]; Dubai[0.049]; Arabian[0.048]; Venezuela[0.046]; Mexico[0.046]; Nigeria[0.046]; seven[0.044]; London[0.044]; London[0.044]; Indonesia[0.044]; Wednesday[0.043]; said[0.043]; news[0.043]; Thursday[0.042]; Thursday[0.042]; Saharan[0.042]; barrel[0.042]; stood[0.042]; crudes[0.041]; quoting[0.040]; agency[0.040]; Light[0.040]; Light[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dubai ==> ENTITY: \u001b[32mDubai\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.114:0.114[0]; log p(e|m)= -0.082:-0.082[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncrudes[0.046]; Saudi[0.044]; barrel[0.044]; London[0.044]; London[0.044]; Arabian[0.043]; seven[0.043]; Wednesday[0.042]; Thursday[0.042]; Thursday[0.042]; Arabia[0.042]; Indonesia[0.041]; Mexico[0.041]; Saharan[0.041]; Venezuela[0.041]; price[0.041]; price[0.041]; said[0.040]; Blend[0.040]; Algeria[0.040]; Newsroom[0.039]; news[0.039]; agency[0.039]; Minas[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.082:0.082[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDubai[0.047]; seven[0.045]; Algeria[0.043]; comprises[0.043]; Wednesday[0.043]; Venezuela[0.043]; Mexico[0.043]; Thursday[0.042]; Thursday[0.042]; Nigeria[0.042]; said[0.041]; barrel[0.041]; Arabia[0.041]; news[0.041]; London[0.041]; London[0.041]; agency[0.041]; Saudi[0.041]; secretariat[0.040]; Arabian[0.040]; Newsroom[0.040]; quoting[0.039]; stood[0.039]; crudes[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = UAE ==> ENTITY: \u001b[32mUnited Arab Emirates\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.149:0.149[0]; log p(e|m)= -0.195:-0.195[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDubai[0.050]; Algeria[0.046]; Saudi[0.046]; seven[0.044]; London[0.044]; London[0.044]; Venezuela[0.044]; Indonesia[0.044]; Arabia[0.043]; Nigeria[0.043]; Mexico[0.042]; Arabian[0.042]; Wednesday[0.042]; Thursday[0.041]; Thursday[0.041]; secretariat[0.040]; said[0.039]; news[0.038]; agency[0.038]; comprises[0.038]; price[0.038]; price[0.038]; Saharan[0.038]; stood[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = OPEC\u001b[39m ==> ENTITY: \u001b[32mOPEC\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.132:0.132[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVenezuela[0.048]; Algeria[0.045]; Saudi[0.045]; Nigeria[0.044]; price[0.043]; price[0.043]; Mexico[0.043]; Newsroom[0.042]; barrel[0.041]; Dubai[0.041]; said[0.041]; Wednesday[0.041]; agency[0.041]; Thursday[0.041]; Thursday[0.041]; Indonesia[0.041]; secretariat[0.040]; London[0.040]; London[0.040]; Arabia[0.040]; Saharan[0.040]; seven[0.040]; news[0.039]; crudes[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = OPEC\u001b[39m ==> ENTITY: \u001b[32mOPEC\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.132:0.132[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVenezuela[0.048]; Algeria[0.045]; Saudi[0.045]; Nigeria[0.044]; price[0.043]; price[0.043]; Mexico[0.043]; Newsroom[0.042]; barrel[0.041]; Dubai[0.041]; said[0.041]; Wednesday[0.041]; agency[0.041]; Thursday[0.041]; Thursday[0.041]; Indonesia[0.041]; secretariat[0.040]; London[0.040]; London[0.040]; Arabia[0.040]; Saharan[0.040]; seven[0.040]; news[0.039]; crudes[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Venezuela ==> ENTITY: \u001b[32mVenezuela\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.068:0.068[0]; log p(e|m)= -0.212:-0.212[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.050]; Light[0.046]; Light[0.046]; Wednesday[0.045]; seven[0.045]; stood[0.045]; Thursday[0.044]; Thursday[0.044]; Juana[0.044]; Algeria[0.044]; secretariat[0.043]; Nigeria[0.043]; news[0.043]; Bonny[0.043]; Minas[0.042]; crudes[0.042]; agency[0.042]; Indonesia[0.042]; Blend[0.042]; said[0.042]; Saharan[0.041]; barrel[0.041]; Newsroom[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nigeria ==> ENTITY: \u001b[32mNigeria\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.315:-0.315[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlgeria[0.047]; Mexico[0.044]; Venezuela[0.044]; seven[0.043]; Indonesia[0.043]; London[0.043]; London[0.043]; Wednesday[0.043]; Thursday[0.042]; Thursday[0.042]; secretariat[0.042]; said[0.042]; Saudi[0.041]; agency[0.041]; Dubai[0.041]; price[0.041]; price[0.041]; barrel[0.040]; Saharan[0.040]; Bonny[0.039]; Arabia[0.039]; comprises[0.039]; quoting[0.039]; crudes[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = OPEC\u001b[39m ==> ENTITY: \u001b[32mOPEC\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.132:0.132[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVenezuela[0.048]; Algeria[0.045]; Saudi[0.045]; Nigeria[0.044]; price[0.043]; price[0.043]; Mexico[0.043]; Newsroom[0.042]; barrel[0.041]; Dubai[0.041]; said[0.041]; Wednesday[0.041]; agency[0.041]; Thursday[0.041]; Thursday[0.041]; Indonesia[0.041]; secretariat[0.040]; London[0.040]; London[0.040]; Arabia[0.040]; Saharan[0.040]; seven[0.040]; news[0.039]; crudes[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexico ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.052:0.052[0]; log p(e|m)= -0.183:-0.183[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVenezuela[0.049]; Algeria[0.046]; Light[0.044]; Light[0.044]; Thursday[0.043]; Thursday[0.043]; Wednesday[0.043]; Nigeria[0.043]; seven[0.043]; Indonesia[0.042]; Dubai[0.042]; Juana[0.041]; comprises[0.041]; Saudi[0.041]; Arabian[0.040]; Saharan[0.040]; stood[0.040]; agency[0.040]; said[0.039]; London[0.039]; London[0.039]; Newsroom[0.039]; Minas[0.039]; Blend[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Algeria ==> ENTITY: \u001b[32mAlgeria\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.086:0.086[0]; log p(e|m)= -0.337:-0.337[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaharan[0.045]; Venezuela[0.045]; Arabia[0.045]; Saudi[0.044]; Nigeria[0.044]; Mexico[0.043]; Thursday[0.043]; Thursday[0.043]; Wednesday[0.043]; seven[0.042]; said[0.042]; Dubai[0.042]; news[0.041]; Indonesia[0.041]; agency[0.040]; London[0.040]; London[0.040]; secretariat[0.040]; Juana[0.040]; Blend[0.040]; Arabian[0.040]; Newsroom[0.040]; Minas[0.039]; comprises[0.039]; \t\n\r [===========================>..................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s956ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2796/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1268testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = ILO\u001b[39m ==> ENTITY: \u001b[32mInternational Labour Organization\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.121:0.121[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIlo[0.047]; Ilo[0.047]; Organisation[0.044]; International[0.044]; International[0.044]; International[0.044]; Unions[0.043]; speaking[0.042]; Confederation[0.042]; Trade[0.041]; Union[0.041]; Labour[0.041]; Singapore[0.040]; Singapore[0.040]; world[0.040]; group[0.040]; command[0.040]; general[0.040]; general[0.040]; small[0.040]; conference[0.040]; nations[0.040]; director[0.039]; respect[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Thailand ==> ENTITY: \u001b[32mThailand\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.163:0.163[0]; log p(e|m)= -0.241:-0.241[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMalaysia[0.047]; Asian[0.046]; Indonesia[0.045]; Vietnam[0.044]; Philippines[0.044]; Singapore[0.044]; Asean[0.043]; Asean[0.043]; rights[0.043]; rights[0.043]; groups[0.041]; Association[0.041]; included[0.040]; chairman[0.040]; final[0.040]; Nations[0.040]; Brunei[0.040]; said[0.040]; said[0.040]; said[0.040]; disappointed[0.040]; news[0.039]; statement[0.039]; making[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.127:0.127[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnews[0.049]; told[0.044]; trade[0.044]; trade[0.044]; trade[0.044]; trade[0.044]; trade[0.044]; groups[0.044]; international[0.044]; credibility[0.044]; day[0.043]; Jordan[0.043]; Jordan[0.043]; issues[0.043]; statement[0.043]; Group[0.042]; conference[0.042]; conference[0.042]; conference[0.042]; conference[0.042]; governments[0.042]; agenda[0.042]; agenda[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vietnam ==> ENTITY: \u001b[32mVietnam\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.390:-0.390[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThailand[0.047]; Philippines[0.045]; Malaysia[0.044]; Asian[0.044]; Indonesia[0.043]; Singapore[0.043]; rights[0.043]; rights[0.043]; social[0.042]; social[0.042]; social[0.042]; governments[0.041]; groups[0.041]; Asean[0.041]; Asean[0.041]; Association[0.040]; governmental[0.040]; news[0.040]; discrimination[0.040]; organization[0.040]; guarantee[0.039]; conference[0.039]; statement[0.039]; Brunei[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = World Trade Organisation\u001b[39m ==> ENTITY: \u001b[32mWorld Trade Organization\u001b[39m\t\nSCORES: global= 0.317:0.317[0]; local()= 0.240:0.240[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrade[0.046]; trade[0.046]; trade[0.046]; trade[0.046]; international[0.045]; International[0.043]; International[0.043]; nations[0.043]; Organisation[0.043]; Trade[0.042]; agenda[0.041]; Confederation[0.040]; talks[0.040]; meeting[0.040]; meeting[0.040]; issues[0.040]; developing[0.040]; Singapore[0.039]; Singapore[0.039]; credibility[0.039]; conference[0.039]; conference[0.039]; conference[0.039]; conference[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = International Confederation of Free Trade Unions\u001b[39m ==> ENTITY: \u001b[32mInternational Confederation of Free Trade Unions\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUnion[0.045]; Organisation[0.044]; Organisation[0.044]; International[0.044]; International[0.044]; international[0.043]; trade[0.043]; trade[0.043]; Trade[0.043]; nations[0.042]; union[0.042]; links[0.041]; conference[0.041]; conference[0.041]; Singapore[0.041]; Singapore[0.041]; group[0.040]; Ilo[0.040]; Ilo[0.040]; Ilo[0.040]; secretary[0.040]; meeting[0.040]; meeting[0.040]; Labour[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Malaysia ==> ENTITY: \u001b[32mMalaysia\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.170:0.170[0]; log p(e|m)= -0.202:-0.202[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThailand[0.048]; Indonesia[0.048]; Singapore[0.046]; Vietnam[0.046]; Asian[0.045]; Brunei[0.044]; Philippines[0.043]; final[0.042]; Asean[0.041]; Asean[0.041]; level[0.040]; Association[0.040]; speaking[0.040]; secretariat[0.040]; Group[0.040]; leaders[0.040]; wants[0.040]; making[0.040]; particularly[0.039]; withdrawn[0.039]; groups[0.039]; Nations[0.039]; conference[0.039]; conference[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = WTO\u001b[39m ==> ENTITY: \u001b[32mWorld Trade Organization\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.218:0.218[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrade[0.045]; trade[0.045]; International[0.044]; International[0.044]; International[0.044]; World[0.042]; meeting[0.042]; meeting[0.042]; nations[0.042]; Organisation[0.042]; Organisation[0.042]; developing[0.042]; conference[0.041]; Trade[0.041]; Trade[0.041]; Singapore[0.041]; Singapore[0.041]; Confederation[0.040]; respect[0.040]; world[0.040]; group[0.040]; week[0.040]; talks[0.040]; Union[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = WTO\u001b[39m ==> ENTITY: \u001b[32mWorld Trade Organization\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.242:0.242[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninternational[0.045]; trade[0.044]; trade[0.044]; trade[0.044]; trade[0.044]; trade[0.044]; World[0.041]; meeting[0.041]; nations[0.041]; standards[0.041]; Organisation[0.041]; agenda[0.041]; issues[0.041]; Asean[0.041]; governments[0.041]; rights[0.041]; developing[0.041]; conference[0.041]; conference[0.041]; conference[0.041]; conference[0.041]; credibility[0.041]; Trade[0.040]; Asian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = ILO\u001b[39m ==> ENTITY: \u001b[32mInternational Labour Organization\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.092:0.092[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIlo[0.049]; clause[0.044]; clause[0.044]; included[0.042]; Thailand[0.042]; delegates[0.042]; Malaysia[0.042]; gathering[0.042]; Singapore[0.041]; significant[0.041]; efforts[0.041]; Philippines[0.041]; making[0.041]; social[0.041]; social[0.041]; social[0.041]; level[0.041]; issued[0.041]; capable[0.041]; secretariat[0.041]; organization[0.041]; groups[0.040]; director[0.040]; governmental[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = ASEAN\u001b[39m ==> ENTITY: \u001b[32mAssociation of Southeast Asian Nations\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.225:0.225[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAsean[0.050]; Asian[0.046]; Thailand[0.046]; Malaysia[0.046]; trade[0.045]; trade[0.045]; trade[0.045]; Philippines[0.045]; Singapore[0.044]; Indonesia[0.044]; governments[0.044]; stake[0.043]; organization[0.042]; Vietnam[0.042]; discussed[0.042]; members[0.042]; efforts[0.042]; Association[0.042]; Brunei[0.041]; leaders[0.041]; governmental[0.041]; delegates[0.041]; groups[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = ICFTU\u001b[39m ==> ENTITY: \u001b[32mInternational Confederation of Free Trade Unions\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOrganisation[0.046]; international[0.045]; trade[0.045]; trade[0.045]; trade[0.045]; trade[0.045]; Trade[0.045]; Trade[0.045]; Unions[0.044]; nations[0.044]; union[0.044]; union[0.044]; governments[0.043]; links[0.043]; conference[0.042]; conference[0.042]; conference[0.042]; conference[0.042]; conference[0.042]; Singapore[0.042]; Group[0.042]; group[0.042]; agenda[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = WTO\u001b[39m ==> ENTITY: \u001b[32mWorld Trade Organization\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.246:0.246[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninternational[0.045]; trade[0.044]; trade[0.044]; trade[0.044]; trade[0.044]; International[0.043]; International[0.043]; International[0.043]; World[0.041]; meeting[0.041]; meeting[0.041]; nations[0.041]; standards[0.041]; Organisation[0.041]; Organisation[0.041]; developing[0.041]; conference[0.041]; conference[0.041]; conference[0.041]; Trade[0.040]; Trade[0.040]; Singapore[0.040]; Singapore[0.040]; Confederation[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = ILO\u001b[39m ==> ENTITY: \u001b[32mInternational Labour Organization\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.082:0.082[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIlo[0.047]; Ilo[0.047]; Ilo[0.047]; Workers[0.044]; clause[0.043]; clause[0.043]; governments[0.042]; Southeast[0.042]; speaking[0.042]; included[0.041]; Thailand[0.041]; right[0.041]; discrimination[0.040]; delegates[0.040]; Malaysia[0.040]; gathering[0.040]; Singapore[0.040]; significant[0.040]; efforts[0.040]; Philippines[0.040]; making[0.040]; social[0.040]; social[0.040]; social[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.104:0.104[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.053]; respect[0.042]; head[0.042]; head[0.042]; World[0.042]; Unions[0.042]; world[0.042]; said[0.042]; International[0.042]; International[0.042]; International[0.042]; week[0.041]; Jordan[0.041]; Jordan[0.041]; outrage[0.041]; conference[0.041]; command[0.041]; Confederation[0.041]; trade[0.041]; trade[0.041]; group[0.041]; union[0.041]; general[0.040]; general[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Association of Southeast Asian Nations\u001b[39m ==> ENTITY: \u001b[32mAssociation of Southeast Asian Nations\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.242:0.242[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAsean[0.049]; Asean[0.049]; Thailand[0.045]; Malaysia[0.045]; Singapore[0.043]; trade[0.043]; trade[0.043]; trade[0.043]; Indonesia[0.043]; Philippines[0.042]; Vietnam[0.041]; Brunei[0.040]; leaders[0.040]; governments[0.040]; issues[0.040]; meeting[0.040]; delegates[0.040]; efforts[0.040]; agenda[0.039]; agenda[0.039]; secretariat[0.039]; ministerial[0.039]; ministerial[0.039]; groups[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = WTO\u001b[39m ==> ENTITY: \u001b[32mWorld Trade Organization\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrade[0.045]; trade[0.045]; organization[0.043]; meeting[0.042]; agenda[0.042]; clause[0.042]; clause[0.042]; Asean[0.042]; Asean[0.042]; governmental[0.042]; governments[0.042]; rights[0.042]; rights[0.042]; conference[0.042]; Philippines[0.041]; Thailand[0.041]; secretariat[0.041]; Asian[0.041]; groups[0.041]; Singapore[0.041]; gathering[0.040]; Malaysia[0.040]; delegates[0.040]; included[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.106:0.106[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.053]; respect[0.042]; head[0.042]; head[0.042]; Unions[0.042]; world[0.042]; said[0.042]; International[0.042]; International[0.042]; International[0.042]; week[0.041]; Jordan[0.041]; Jordan[0.041]; outrage[0.041]; conference[0.041]; command[0.041]; Confederation[0.041]; trade[0.041]; trade[0.041]; group[0.041]; union[0.041]; general[0.040]; general[0.040]; nations[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Philippines ==> ENTITY: \u001b[32mPhilippines\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.180:-0.180[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThailand[0.046]; Malaysia[0.046]; Asian[0.045]; Singapore[0.044]; Indonesia[0.044]; Vietnam[0.043]; Asean[0.042]; Asean[0.042]; included[0.042]; adding[0.042]; Brunei[0.042]; speaking[0.042]; Monday[0.041]; statement[0.041]; final[0.040]; particularly[0.040]; groups[0.040]; governments[0.040]; said[0.040]; said[0.040]; said[0.040]; members[0.040]; conference[0.040]; conference[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = ICFTU\u001b[39m ==> ENTITY: \u001b[32mInternational Confederation of Free Trade Unions\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOrganisation[0.044]; Asean[0.043]; Asean[0.043]; international[0.043]; trade[0.043]; trade[0.043]; trade[0.043]; trade[0.043]; trade[0.043]; Trade[0.043]; nations[0.042]; union[0.042]; union[0.042]; Southeast[0.041]; governments[0.041]; links[0.041]; groups[0.041]; conference[0.040]; conference[0.040]; conference[0.040]; conference[0.040]; Singapore[0.040]; Group[0.040]; group[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Michel Hansenne\u001b[39m ==> ENTITY: \u001b[32mMichel Hansenne\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.116:0.116[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nconditions[0.044]; Ilo[0.042]; Ilo[0.042]; Ilo[0.042]; Organisation[0.042]; Organisation[0.042]; International[0.042]; International[0.042]; International[0.042]; Labour[0.042]; conference[0.042]; conference[0.042]; conference[0.042]; opening[0.042]; Unions[0.041]; international[0.041]; head[0.041]; head[0.041]; secretary[0.041]; director[0.041]; standards[0.041]; day[0.041]; said[0.041]; said[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBill Jordan\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mWilliam Joseph Jordan\u001b[39m <---> \u001b[32mBill Jordan, Baron Jordan\u001b[39m\t\nSCORES: global= 0.283:0.262[\u001b[31m0.021\u001b[39m]; local()= 0.116:0.132[\u001b[32m0.017\u001b[39m]; log p(e|m)= -1.324:-0.777[\u001b[32m0.548\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJordan[0.045]; Jordan[0.045]; Confederation[0.045]; Union[0.043]; Labour[0.043]; secretary[0.043]; Unions[0.042]; union[0.042]; Organisation[0.042]; Organisation[0.042]; leaders[0.041]; leaders[0.041]; World[0.041]; trade[0.041]; trade[0.041]; International[0.041]; International[0.041]; International[0.041]; Ilo[0.041]; Ilo[0.041]; Ilo[0.041]; general[0.040]; general[0.040]; taking[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brunei ==> ENTITY: \u001b[32mBrunei\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.214:0.214[0]; log p(e|m)= -0.387:-0.387[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMalaysia[0.049]; Thailand[0.048]; Indonesia[0.047]; Asian[0.046]; Asean[0.046]; Asean[0.046]; Philippines[0.045]; Singapore[0.044]; Association[0.042]; Vietnam[0.041]; organization[0.040]; Monday[0.040]; withdrawn[0.039]; contribution[0.039]; particularly[0.039]; delegates[0.039]; ministerial[0.039]; ministerial[0.039]; meeting[0.039]; Nations[0.039]; making[0.038]; Group[0.038]; conference[0.038]; conference[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nday[0.047]; international[0.045]; respect[0.044]; guarantee[0.044]; Monday[0.044]; World[0.044]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; issue[0.043]; issue[0.043]; Jordan[0.043]; Jordan[0.043]; Jordan[0.043]; conference[0.043]; conference[0.043]; conference[0.043]; conference[0.043]; hiring[0.043]; command[0.043]; issues[0.043]; chairman[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = ICFTU\u001b[39m ==> ENTITY: \u001b[32mInternational Confederation of Free Trade Unions\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecretariat[0.046]; Asean[0.045]; Asean[0.045]; trade[0.045]; trade[0.045]; organization[0.045]; Philippines[0.044]; members[0.044]; clause[0.044]; clause[0.044]; Malaysia[0.044]; Southeast[0.043]; governments[0.043]; issued[0.043]; statement[0.043]; groups[0.043]; Indonesia[0.043]; delegates[0.042]; Singapore[0.042]; Asian[0.042]; Association[0.042]; agenda[0.042]; Thailand[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = WTO\u001b[39m ==> ENTITY: \u001b[32mWorld Trade Organization\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrade[0.045]; trade[0.045]; organization[0.043]; agenda[0.042]; clause[0.042]; clause[0.042]; Asean[0.042]; Asean[0.042]; statement[0.042]; governmental[0.042]; governments[0.042]; rights[0.042]; rights[0.042]; Philippines[0.041]; Thailand[0.041]; secretariat[0.041]; groups[0.041]; Richard[0.041]; gathering[0.041]; Malaysia[0.040]; director[0.040]; delegates[0.040]; impose[0.040]; efforts[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = ASEAN\u001b[39m ==> ENTITY: \u001b[32mAssociation of Southeast Asian Nations\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.221:0.221[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAsean[0.050]; Asian[0.046]; Thailand[0.046]; Malaysia[0.046]; trade[0.045]; trade[0.045]; trade[0.045]; Philippines[0.045]; Singapore[0.044]; Indonesia[0.044]; governments[0.044]; stake[0.043]; organization[0.042]; Vietnam[0.042]; discussed[0.042]; efforts[0.042]; Association[0.042]; Brunei[0.041]; leaders[0.041]; governmental[0.041]; delegates[0.041]; Jordan[0.041]; groups[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = WTO\u001b[39m ==> ENTITY: \u001b[32mWorld Trade Organization\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.244:0.244[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninternational[0.045]; trade[0.044]; trade[0.044]; trade[0.044]; trade[0.044]; International[0.043]; International[0.043]; World[0.041]; meeting[0.041]; meeting[0.041]; nations[0.041]; standards[0.041]; Organisation[0.041]; Organisation[0.041]; agenda[0.041]; issues[0.041]; developing[0.041]; conference[0.041]; conference[0.041]; conference[0.041]; conference[0.041]; credibility[0.040]; Trade[0.040]; Trade[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = ICFTU\u001b[39m ==> ENTITY: \u001b[32mInternational Confederation of Free Trade Unions\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConfederation[0.045]; Union[0.044]; Organisation[0.043]; Organisation[0.043]; International[0.043]; International[0.043]; International[0.043]; international[0.042]; trade[0.042]; trade[0.042]; Trade[0.042]; Trade[0.042]; Unions[0.042]; nations[0.042]; union[0.041]; links[0.040]; conference[0.040]; conference[0.040]; Singapore[0.040]; Singapore[0.040]; group[0.040]; Ilo[0.039]; Ilo[0.039]; Ilo[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.151:0.151[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMalaysia[0.049]; Asian[0.048]; Thailand[0.047]; Indonesia[0.044]; Philippines[0.044]; Brunei[0.044]; Asean[0.043]; Asean[0.043]; Vietnam[0.043]; organization[0.041]; final[0.041]; level[0.040]; guarantee[0.040]; Monday[0.040]; delegates[0.040]; Association[0.040]; issue[0.040]; conference[0.039]; hiring[0.039]; chairman[0.039]; Speaking[0.039]; particularly[0.039]; trade[0.039]; trade[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = WTO\u001b[39m ==> ENTITY: \u001b[32mWorld Trade Organization\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrade[0.046]; International[0.045]; International[0.045]; International[0.045]; meeting[0.043]; nations[0.043]; Organisation[0.043]; developing[0.042]; conference[0.042]; Trade[0.042]; Singapore[0.041]; Singapore[0.041]; Confederation[0.041]; respect[0.041]; world[0.041]; group[0.040]; week[0.040]; Union[0.040]; said[0.040]; general[0.040]; general[0.040]; director[0.040]; leaders[0.040]; leaders[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = International Labour Organisation\u001b[39m ==> ENTITY: \u001b[32mInternational Labour Organization\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOrganisation[0.046]; International[0.046]; International[0.046]; Unions[0.045]; Confederation[0.043]; Trade[0.043]; Trade[0.043]; Union[0.042]; World[0.042]; Ilo[0.041]; Ilo[0.041]; Ilo[0.041]; director[0.041]; developing[0.041]; meeting[0.040]; conditions[0.040]; trade[0.040]; trade[0.040]; conference[0.040]; union[0.040]; head[0.040]; head[0.040]; nations[0.040]; secretary[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = WTO\u001b[39m ==> ENTITY: \u001b[32mWorld Trade Organization\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.241:0.241[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninternational[0.045]; trade[0.044]; trade[0.044]; trade[0.044]; trade[0.044]; trade[0.044]; World[0.041]; meeting[0.041]; standards[0.041]; Organisation[0.041]; agenda[0.041]; agenda[0.041]; issues[0.041]; Asean[0.041]; Asean[0.041]; governments[0.041]; rights[0.041]; conference[0.041]; conference[0.041]; conference[0.041]; conference[0.041]; credibility[0.041]; Trade[0.040]; Asian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMalaysia[0.048]; Thailand[0.048]; Asian[0.046]; Philippines[0.045]; Singapore[0.043]; Asean[0.043]; Asean[0.043]; Vietnam[0.043]; leaders[0.043]; governments[0.042]; speaking[0.041]; members[0.041]; trade[0.041]; trade[0.041]; groups[0.040]; final[0.040]; Brunei[0.040]; disappointed[0.039]; particularly[0.039]; conference[0.039]; conference[0.039]; organization[0.039]; ministerial[0.039]; ministerial[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = WTO\u001b[39m ==> ENTITY: \u001b[32mWorld Trade Organization\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.239:0.239[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninternational[0.045]; trade[0.044]; trade[0.044]; trade[0.044]; trade[0.044]; trade[0.044]; World[0.041]; meeting[0.041]; nations[0.041]; standards[0.041]; Organisation[0.041]; agenda[0.041]; issues[0.041]; governments[0.041]; rights[0.041]; developing[0.041]; conference[0.041]; conference[0.041]; conference[0.041]; conference[0.041]; credibility[0.041]; Trade[0.040]; Asian[0.040]; Singapore[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = ILO\u001b[39m ==> ENTITY: \u001b[32mInternational Labour Organization\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIlo[0.047]; Ilo[0.047]; Organisation[0.044]; Organisation[0.044]; International[0.044]; International[0.044]; International[0.044]; Unions[0.043]; international[0.042]; speaking[0.041]; conditions[0.041]; Confederation[0.041]; Trade[0.041]; Trade[0.041]; standards[0.041]; Union[0.040]; Labour[0.040]; linking[0.040]; platform[0.040]; World[0.040]; world[0.040]; command[0.040]; general[0.039]; general[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = WTO\u001b[39m ==> ENTITY: \u001b[32mWorld Trade Organization\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrade[0.047]; trade[0.047]; organization[0.045]; agenda[0.044]; clause[0.044]; clause[0.044]; Asean[0.044]; Asean[0.044]; statement[0.044]; governmental[0.044]; rights[0.044]; rights[0.044]; Philippines[0.043]; Thailand[0.043]; secretariat[0.043]; groups[0.043]; Richard[0.043]; gathering[0.042]; Malaysia[0.042]; director[0.042]; delegates[0.042]; impose[0.042]; efforts[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = WTO\u001b[39m ==> ENTITY: \u001b[32mWorld Trade Organization\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrade[0.046]; organization[0.044]; agenda[0.043]; clause[0.042]; clause[0.042]; Asean[0.042]; Asean[0.042]; statement[0.042]; governmental[0.042]; rights[0.042]; Philippines[0.042]; Thailand[0.042]; secretariat[0.042]; groups[0.041]; Richard[0.041]; gathering[0.041]; Malaysia[0.041]; director[0.041]; delegates[0.041]; impose[0.041]; efforts[0.040]; said[0.040]; said[0.040]; Asian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = ILO\u001b[39m ==> ENTITY: \u001b[32mInternational Labour Organization\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.090:0.090[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIlo[0.047]; Ilo[0.047]; Workers[0.044]; clause[0.043]; clause[0.043]; governments[0.043]; Southeast[0.042]; speaking[0.042]; included[0.042]; Thailand[0.041]; right[0.041]; discrimination[0.041]; delegates[0.041]; Malaysia[0.041]; gathering[0.041]; Singapore[0.041]; significant[0.040]; efforts[0.040]; Philippines[0.040]; beginning[0.040]; making[0.040]; social[0.040]; social[0.040]; social[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = ILO\u001b[39m ==> ENTITY: \u001b[32mInternational Labour Organization\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.128:0.128[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIlo[0.047]; Ilo[0.047]; Organisation[0.044]; Organisation[0.044]; International[0.044]; International[0.044]; International[0.044]; Unions[0.043]; speaking[0.042]; conditions[0.041]; Confederation[0.041]; Trade[0.041]; Trade[0.041]; Union[0.041]; Labour[0.040]; linking[0.040]; platform[0.040]; World[0.040]; world[0.040]; command[0.040]; general[0.039]; general[0.039]; small[0.039]; director[0.039]; \t\n\r [============================>.................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s645ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2836/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1290testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rome ==> ENTITY: \u001b[32mRome\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.051]; Giuseppe[0.044]; Mussolini[0.043]; Mussolini[0.043]; Mussolini[0.043]; Mussolini[0.043]; month[0.043]; state[0.042]; National[0.042]; Fascist[0.041]; Gianfranco[0.041]; Alessandra[0.041]; said[0.041]; said[0.041]; policy[0.041]; Friday[0.041]; met[0.041]; Benito[0.040]; resolved[0.040]; offices[0.040]; told[0.040]; told[0.040]; lose[0.040]; lose[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fini\u001b[39m ==> ENTITY: \u001b[32mGianfranco Fini\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFini[0.046]; Italy[0.045]; parliamentary[0.043]; parliamentary[0.043]; Alliance[0.042]; party[0.042]; party[0.042]; party[0.042]; party[0.042]; party[0.042]; Mussolini[0.041]; Mussolini[0.041]; Mussolini[0.041]; Mussolini[0.041]; Mussolini[0.041]; Mussolini[0.041]; political[0.041]; Giuseppe[0.041]; Gianfranco[0.041]; leaders[0.041]; group[0.040]; month[0.040]; month[0.040]; resigned[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = La Stampa\u001b[39m ==> ENTITY: \u001b[32mLa Stampa\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntold[0.047]; told[0.047]; newspaper[0.047]; Giuseppe[0.044]; political[0.043]; month[0.043]; Fini[0.043]; Mussolini[0.042]; Mussolini[0.042]; Mussolini[0.042]; resigned[0.042]; said[0.042]; Fiamma[0.041]; World[0.041]; quitting[0.040]; accused[0.040]; radio[0.040]; resolved[0.040]; considering[0.040]; Movement[0.040]; joining[0.039]; internal[0.039]; fascist[0.039]; formed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = RAI ==> ENTITY: \u001b[32mRAI\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.140:0.140[0]; log p(e|m)= -0.014:-0.014[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.059]; party[0.054]; party[0.054]; party[0.054]; party[0.054]; party[0.054]; political[0.054]; Rome[0.054]; Giuseppe[0.053]; National[0.052]; Gianfranco[0.052]; month[0.052]; month[0.052]; leaders[0.051]; radio[0.051]; radio[0.051]; offices[0.050]; parliamentary[0.050]; parliamentary[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = National Alliance ==> ENTITY: \u001b[32mNational Alliance (Italy)\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.892:-0.892[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.045]; Fini[0.045]; Fini[0.045]; party[0.043]; party[0.043]; party[0.043]; parliamentary[0.043]; Gianfranco[0.042]; Giuseppe[0.042]; Fascist[0.041]; political[0.041]; group[0.041]; Alessandra[0.041]; leader[0.041]; leader[0.041]; quit[0.041]; rejoined[0.040]; Mussolini[0.040]; Mussolini[0.040]; Mussolini[0.040]; Mussolini[0.040]; Mussolini[0.040]; radio[0.040]; radio[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = World War Two\u001b[39m ==> ENTITY: \u001b[32mWorld War II\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.052:0.052[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmonth[0.046]; fascist[0.046]; Mussolini[0.044]; Mussolini[0.044]; political[0.043]; lower[0.042]; leaders[0.042]; said[0.042]; overcome[0.041]; leader[0.041]; Social[0.041]; joining[0.041]; reasons[0.041]; debate[0.041]; Movement[0.041]; considering[0.041]; accused[0.041]; group[0.041]; quitting[0.040]; newspaper[0.040]; house[0.040]; followers[0.040]; stifling[0.040]; resigned[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMussolini\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBenito Mussolini\u001b[39m <---> \u001b[32mAlessandra Mussolini\u001b[39m\t\nSCORES: global= 0.255:0.244[\u001b[31m0.010\u001b[39m]; local()= 0.180:0.178[\u001b[31m0.003\u001b[39m]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMussolini[0.054]; Mussolini[0.054]; Mussolini[0.054]; Mussolini[0.054]; Mussolini[0.054]; fascist[0.053]; Italy[0.052]; Fascist[0.052]; Fini[0.049]; Fini[0.049]; Rome[0.049]; Giuseppe[0.049]; political[0.048]; Gianfranco[0.048]; policy[0.047]; dictator[0.047]; National[0.047]; group[0.047]; Social[0.047]; Alessandra[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mussolini ==> ENTITY: \u001b[32mBenito Mussolini\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.159:0.159[0]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMussolini[0.049]; Mussolini[0.049]; Mussolini[0.049]; Italy[0.047]; Fascist[0.047]; Fini[0.045]; Fini[0.045]; Rome[0.045]; Gianfranco[0.043]; policy[0.043]; dictator[0.043]; National[0.043]; Alessandra[0.042]; party[0.042]; party[0.042]; Alliance[0.042]; right[0.042]; right[0.042]; Benito[0.041]; leader[0.041]; month[0.041]; quit[0.041]; state[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.129:0.129[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRome[0.046]; Mussolini[0.045]; Mussolini[0.045]; Mussolini[0.045]; Mussolini[0.045]; Gianfranco[0.044]; Giuseppe[0.043]; month[0.043]; Fascist[0.042]; quit[0.042]; row[0.041]; National[0.041]; Friday[0.041]; Fini[0.041]; Fini[0.041]; Alliance[0.040]; said[0.040]; said[0.040]; Alessandra[0.039]; shortly[0.039]; told[0.039]; told[0.039]; rejoined[0.039]; resolved[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMussolini\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBenito Mussolini\u001b[39m <---> \u001b[32mAlessandra Mussolini\u001b[39m\t\nSCORES: global= 0.253:0.245[\u001b[31m0.008\u001b[39m]; local()= 0.150:0.178[\u001b[32m0.028\u001b[39m]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMussolini[0.046]; Mussolini[0.046]; Mussolini[0.046]; fascist[0.046]; Fascist[0.045]; Fini[0.042]; Fini[0.042]; Giuseppe[0.042]; political[0.041]; Gianfranco[0.041]; policy[0.041]; dictator[0.041]; National[0.041]; group[0.040]; Social[0.040]; party[0.040]; party[0.040]; party[0.040]; party[0.040]; stifling[0.040]; Alliance[0.040]; parliamentary[0.040]; parliamentary[0.040]; right[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gianfranco Fini\u001b[39m ==> ENTITY: \u001b[32mGianfranco Fini\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFini[0.046]; Italy[0.045]; parliamentary[0.043]; Alliance[0.042]; party[0.042]; party[0.042]; party[0.042]; party[0.042]; Mussolini[0.042]; Mussolini[0.042]; Mussolini[0.042]; Mussolini[0.042]; Mussolini[0.042]; Mussolini[0.042]; political[0.041]; Giuseppe[0.041]; leaders[0.041]; group[0.041]; month[0.041]; resigned[0.040]; Fascist[0.040]; leader[0.040]; leader[0.040]; shortly[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alessandra Mussolini\u001b[39m ==> ENTITY: \u001b[32mAlessandra Mussolini\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.047]; Fini[0.046]; Fini[0.046]; Mussolini[0.046]; Mussolini[0.046]; Mussolini[0.046]; Giuseppe[0.045]; Fascist[0.045]; Rome[0.044]; party[0.043]; party[0.043]; Gianfranco[0.043]; Alliance[0.043]; leader[0.042]; leader[0.042]; month[0.042]; National[0.042]; Benito[0.042]; quit[0.042]; told[0.041]; told[0.041]; said[0.041]; said[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sophia Loren\u001b[39m ==> ENTITY: \u001b[32mSophia Loren\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.072:0.072[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstar[0.048]; Giuseppe[0.044]; Mussolini[0.044]; Mussolini[0.044]; Mussolini[0.044]; Fiamma[0.043]; Gianfranco[0.042]; month[0.042]; screen[0.042]; World[0.041]; Ms[0.041]; neo[0.041]; met[0.041]; Stampa[0.041]; said[0.041]; said[0.041]; interviewed[0.041]; internal[0.040]; La[0.040]; Chamber[0.040]; Chamber[0.040]; Social[0.040]; quitting[0.040]; differences[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Benito Mussolini\u001b[39m ==> ENTITY: \u001b[32mBenito Mussolini\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMussolini[0.049]; Mussolini[0.049]; Mussolini[0.049]; Mussolini[0.049]; Italy[0.047]; Fascist[0.047]; Rome[0.045]; Giuseppe[0.044]; Gianfranco[0.043]; policy[0.043]; dictator[0.043]; National[0.043]; Fini[0.042]; Fini[0.042]; Alliance[0.042]; right[0.042]; right[0.042]; month[0.040]; state[0.040]; leader[0.040]; leader[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMussolini\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBenito Mussolini\u001b[39m <---> \u001b[32mAlessandra Mussolini\u001b[39m\t\nSCORES: global= 0.252:0.245[\u001b[31m0.007\u001b[39m]; local()= 0.131:0.171[\u001b[32m0.040\u001b[39m]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMussolini[0.047]; Mussolini[0.047]; fascist[0.047]; Fini[0.043]; Fini[0.043]; Giuseppe[0.043]; political[0.042]; group[0.041]; Social[0.041]; party[0.041]; party[0.041]; party[0.041]; stifling[0.041]; parliamentary[0.041]; parliamentary[0.041]; strictly[0.040]; leader[0.040]; reasons[0.040]; Movement[0.040]; leaders[0.040]; month[0.040]; politician[0.040]; World[0.040]; neo[0.040]; \t\n\r [============================>.................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s494ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2851/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1181testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hobart ==> ENTITY: \u001b[32mHobart\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.390:-0.390[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOval[0.049]; match[0.048]; Australia[0.047]; Cricket[0.047]; Victoria[0.046]; Victoria[0.046]; Bellerive[0.045]; Tasmania[0.045]; Tasmania[0.045]; Shield[0.044]; Shield[0.044]; Friday[0.044]; David[0.044]; day[0.042]; day[0.042]; Michael[0.042]; Shaun[0.042]; Young[0.041]; Sheffield[0.040]; Sheffield[0.040]; Boon[0.040]; Score[0.038]; Score[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tasmania ==> ENTITY: \u001b[32mTasmania cricket team\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.159:0.159[0]; log p(e|m)= -2.430:-2.430[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.049]; match[0.049]; Oval[0.048]; Tasmania[0.047]; Victoria[0.047]; Victoria[0.047]; Cricket[0.046]; Shield[0.045]; Shield[0.045]; Hobart[0.044]; day[0.044]; day[0.044]; Bellerive[0.043]; Boon[0.043]; Friday[0.042]; Michael[0.041]; Shaun[0.041]; Sheffield[0.040]; Sheffield[0.040]; David[0.040]; Young[0.039]; Score[0.039]; Score[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = David Boon\u001b[39m ==> ENTITY: \u001b[32mDavid Boon\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.053]; Oval[0.050]; Cricket[0.050]; match[0.049]; Hobart[0.048]; Tasmania[0.047]; Tasmania[0.047]; Michael[0.046]; Shaun[0.046]; Victoria[0.045]; Victoria[0.045]; Shield[0.045]; Shield[0.045]; day[0.044]; day[0.044]; Sheffield[0.044]; Sheffield[0.044]; Bellerive[0.044]; Young[0.044]; Friday[0.042]; Score[0.039]; Score[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Victoria ==> ENTITY: \u001b[32mVictoria cricket team\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.167:0.167[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVictoria[0.049]; Oval[0.049]; Australia[0.048]; Cricket[0.047]; match[0.047]; Tasmania[0.046]; Tasmania[0.046]; Shield[0.045]; Shield[0.045]; Hobart[0.044]; David[0.043]; Michael[0.043]; Young[0.042]; Shaun[0.042]; Bellerive[0.042]; Sheffield[0.042]; Sheffield[0.042]; Friday[0.042]; day[0.041]; day[0.041]; Boon[0.040]; Score[0.037]; Score[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shaun Young\u001b[39m ==> ENTITY: \u001b[32mShaun Young\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOval[0.050]; Cricket[0.049]; Australia[0.049]; Shield[0.048]; Shield[0.048]; Tasmania[0.047]; Tasmania[0.047]; match[0.047]; day[0.046]; day[0.046]; Michael[0.046]; Victoria[0.046]; Victoria[0.046]; Sheffield[0.046]; Sheffield[0.046]; Hobart[0.045]; David[0.045]; Boon[0.044]; Friday[0.042]; Bellerive[0.040]; Score[0.040]; Score[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sheffield Shield\u001b[39m ==> ENTITY: \u001b[32mSheffield Shield\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.053]; Shield[0.050]; Oval[0.049]; Tasmania[0.048]; Tasmania[0.048]; match[0.048]; Cricket[0.047]; Victoria[0.047]; Victoria[0.047]; Sheffield[0.046]; Hobart[0.046]; day[0.045]; day[0.045]; Score[0.044]; Score[0.044]; Bellerive[0.043]; Friday[0.043]; Boon[0.042]; Shaun[0.042]; Michael[0.042]; David[0.041]; Young[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sheffield Shield\u001b[39m ==> ENTITY: \u001b[32mSheffield Shield\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.053]; Shield[0.050]; Oval[0.049]; Tasmania[0.048]; Tasmania[0.048]; match[0.048]; Cricket[0.047]; Victoria[0.047]; Victoria[0.047]; Sheffield[0.046]; Hobart[0.046]; day[0.045]; day[0.045]; Score[0.044]; Score[0.044]; Bellerive[0.043]; Friday[0.043]; Boon[0.042]; Shaun[0.042]; Michael[0.042]; David[0.041]; Young[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Michael DiVenuto\u001b[39m ==> ENTITY: \u001b[32mMichael Di Venuto\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.142:0.142[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOval[0.047]; Australia[0.047]; Cricket[0.047]; day[0.047]; day[0.047]; match[0.046]; Shaun[0.045]; Tasmania[0.044]; Tasmania[0.044]; David[0.044]; Boon[0.044]; Victoria[0.043]; Victoria[0.043]; Friday[0.043]; Hobart[0.043]; Sheffield[0.042]; Sheffield[0.042]; Shield[0.042]; Shield[0.042]; Bellerive[0.042]; Young[0.041]; Score[0.038]; Score[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bellerive Oval\u001b[39m ==> ENTITY: \u001b[32mBellerive Oval\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.051]; Australia[0.050]; Cricket[0.050]; Hobart[0.048]; Victoria[0.048]; Victoria[0.048]; Tasmania[0.048]; Tasmania[0.048]; Friday[0.047]; Shield[0.046]; Shield[0.046]; day[0.045]; day[0.045]; David[0.044]; Michael[0.044]; Shaun[0.043]; Young[0.042]; Sheffield[0.042]; Sheffield[0.042]; Boon[0.042]; Score[0.040]; Score[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.107:0.107[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.048]; Cricket[0.048]; Oval[0.048]; Victoria[0.046]; Victoria[0.046]; Tasmania[0.046]; Tasmania[0.046]; day[0.045]; day[0.045]; Shield[0.044]; Shield[0.044]; Shaun[0.043]; Hobart[0.043]; Bellerive[0.043]; Boon[0.042]; Sheffield[0.042]; Sheffield[0.042]; David[0.041]; Young[0.041]; Friday[0.041]; Michael[0.041]; Score[0.037]; Score[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Victoria ==> ENTITY: \u001b[32mVictoria cricket team\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.167:0.167[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVictoria[0.049]; Oval[0.049]; Australia[0.048]; Cricket[0.047]; match[0.047]; Tasmania[0.046]; Tasmania[0.046]; Shield[0.045]; Shield[0.045]; Hobart[0.044]; David[0.043]; Michael[0.043]; Young[0.042]; Shaun[0.042]; Bellerive[0.042]; Sheffield[0.042]; Sheffield[0.042]; Friday[0.042]; day[0.041]; day[0.041]; Boon[0.040]; Score[0.037]; Score[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tasmania ==> ENTITY: \u001b[32mTasmania cricket team\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.159:0.159[0]; log p(e|m)= -2.430:-2.430[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.049]; match[0.049]; Oval[0.048]; Tasmania[0.047]; Victoria[0.047]; Victoria[0.047]; Cricket[0.046]; Shield[0.045]; Shield[0.045]; Hobart[0.044]; day[0.044]; day[0.044]; Bellerive[0.043]; Boon[0.043]; Friday[0.042]; Michael[0.041]; Shaun[0.041]; Sheffield[0.040]; Sheffield[0.040]; David[0.040]; Young[0.039]; Score[0.039]; Score[0.039]; \t\n\r [============================>.................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s474ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2863/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1302testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNewfoundland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNewfoundland and Labrador\u001b[39m <---> \u001b[32mNewfoundland\u001b[39m\t\nSCORES: global= 0.267:0.252[\u001b[31m0.015\u001b[39m]; local()= 0.143:0.160[\u001b[32m0.017\u001b[39m]; log p(e|m)= -1.328:-0.730[\u001b[32m0.598\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNewfoundland[0.052]; Newfoundland[0.052]; coast[0.050]; Canadian[0.045]; east[0.042]; miles[0.042]; Ireland[0.042]; aboard[0.042]; local[0.041]; David[0.041]; began[0.040]; people[0.040]; Michigan[0.040]; route[0.040]; time[0.040]; said[0.039]; said[0.039]; said[0.039]; reported[0.039]; Friday[0.039]; Stephenville[0.039]; Stephenville[0.039]; jet[0.039]; jet[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canadian ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.126:0.126[0]; log p(e|m)= -0.412:-0.412[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNewfoundland[0.048]; Newfoundland[0.048]; province[0.045]; route[0.043]; David[0.043]; considered[0.042]; east[0.042]; miles[0.042]; stop[0.042]; Ireland[0.041]; Ireland[0.041]; Snow[0.041]; Snow[0.041]; Snow[0.041]; time[0.041]; Michigan[0.040]; Michigan[0.040]; coast[0.040]; traffic[0.040]; local[0.040]; Grand[0.040]; km[0.040]; originated[0.040]; Shannon[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ireland ==> ENTITY: \u001b[32mRepublic of Ireland\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.097:0.097[0]; log p(e|m)= -1.505:-1.505[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfinal[0.046]; Shannon[0.044]; time[0.044]; originated[0.044]; Canadian[0.043]; going[0.042]; Friday[0.042]; local[0.042]; Grand[0.042]; added[0.042]; began[0.041]; east[0.041]; flight[0.041]; control[0.041]; missing[0.041]; province[0.041]; considered[0.041]; communication[0.041]; fly[0.040]; said[0.040]; said[0.040]; coast[0.040]; Stephenville[0.040]; Stephenville[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = GMT\u001b[39m ==> ENTITY: \u001b[32mGreenwich Mean Time\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.111:0.111[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.046]; Friday[0.046]; time[0.044]; interview[0.044]; flight[0.043]; traffic[0.043]; David[0.042]; reported[0.042]; manager[0.042]; refueling[0.041]; began[0.041]; miles[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; added[0.041]; local[0.040]; coast[0.040]; east[0.040]; air[0.040]; aircraft[0.040]; communication[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNewfoundland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNewfoundland and Labrador\u001b[39m <---> \u001b[32mNewfoundland\u001b[39m\t\nSCORES: global= 0.267:0.252[\u001b[31m0.015\u001b[39m]; local()= 0.164:0.175[\u001b[32m0.011\u001b[39m]; log p(e|m)= -1.328:-0.730[\u001b[32m0.598\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNewfoundland[0.052]; Newfoundland[0.052]; coast[0.050]; province[0.045]; Canadian[0.045]; east[0.042]; miles[0.042]; Ireland[0.042]; aboard[0.041]; local[0.041]; David[0.041]; began[0.040]; considered[0.039]; people[0.039]; Michigan[0.039]; added[0.039]; route[0.039]; time[0.039]; said[0.039]; said[0.039]; said[0.039]; reported[0.039]; Friday[0.039]; missing[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mStephenville\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mStephenville, Newfoundland and Labrador\u001b[39m <---> \u001b[32mStephenville International Airport\u001b[39m\t\nSCORES: global= 0.267:0.245[\u001b[31m0.022\u001b[39m]; local()= 0.150:0.144[\u001b[31m0.006\u001b[39m]; log p(e|m)= -0.734:-1.760[\u001b[31m1.026\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \neast[0.045]; Stephenville[0.044]; airport[0.044]; airport[0.044]; province[0.044]; coast[0.044]; miles[0.044]; flight[0.043]; Canadian[0.042]; refueling[0.042]; fly[0.042]; aircraft[0.042]; Shannon[0.041]; km[0.041]; cargo[0.041]; Friday[0.040]; Snow[0.040]; began[0.040]; going[0.040]; crash[0.040]; said[0.040]; said[0.040]; added[0.039]; considered[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Grand Rapids ==> ENTITY: \u001b[32mGrand Rapids, Michigan\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.060:0.060[0]; log p(e|m)= -0.285:-0.285[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairport[0.048]; airport[0.048]; airport[0.048]; Michigan[0.045]; cargo[0.045]; refueling[0.045]; aircraft[0.045]; Friday[0.044]; flight[0.044]; east[0.043]; traffic[0.043]; miles[0.043]; air[0.043]; reported[0.043]; stop[0.042]; considered[0.041]; province[0.041]; Canadian[0.041]; local[0.041]; plane[0.041]; Snow[0.041]; Snow[0.041]; Snow[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNewfoundland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNewfoundland and Labrador\u001b[39m <---> \u001b[32mNewfoundland\u001b[39m\t\nSCORES: global= 0.267:0.251[\u001b[31m0.015\u001b[39m]; local()= 0.128:0.139[\u001b[32m0.011\u001b[39m]; log p(e|m)= -1.328:-0.730[\u001b[32m0.598\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNewfoundland[0.053]; Newfoundland[0.053]; east[0.043]; miles[0.043]; Ireland[0.043]; aboard[0.042]; local[0.042]; David[0.042]; began[0.041]; people[0.040]; Michigan[0.040]; route[0.040]; time[0.040]; said[0.040]; said[0.040]; said[0.040]; reported[0.040]; Friday[0.040]; Stephenville[0.039]; Stephenville[0.039]; jet[0.039]; jet[0.039]; pilot[0.039]; pilot[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Michigan ==> ENTITY: \u001b[32mMichigan\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.087:0.087[0]; log p(e|m)= -0.248:-0.248[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfinal[0.047]; coast[0.047]; miles[0.045]; east[0.045]; province[0.045]; David[0.044]; route[0.044]; Newfoundland[0.044]; Newfoundland[0.044]; Newfoundland[0.044]; Canadian[0.043]; began[0.043]; Friday[0.043]; executive[0.043]; executive[0.043]; considered[0.043]; time[0.043]; air[0.042]; Ireland[0.042]; traffic[0.042]; said[0.041]; said[0.041]; said[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Michigan ==> ENTITY: \u001b[32mMichigan\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.091:0.091[0]; log p(e|m)= -0.248:-0.248[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfinal[0.045]; coast[0.045]; miles[0.044]; east[0.044]; province[0.044]; Grand[0.042]; Canadian[0.042]; began[0.042]; Friday[0.042]; David[0.042]; local[0.042]; considered[0.042]; Rapids[0.041]; time[0.041]; air[0.041]; Ireland[0.041]; traffic[0.040]; originated[0.040]; stop[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; Shannon[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = GMT\u001b[39m ==> ENTITY: \u001b[32mGreenwich Mean Time\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.101:0.101[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.045]; Ireland[0.045]; Friday[0.045]; time[0.044]; interview[0.043]; flight[0.043]; route[0.042]; traffic[0.042]; David[0.042]; reported[0.042]; manager[0.042]; refueling[0.041]; executive[0.040]; executive[0.040]; began[0.040]; miles[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; added[0.040]; local[0.040]; coast[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ireland ==> ENTITY: \u001b[32mRepublic of Ireland\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.102:0.102[0]; log p(e|m)= -1.505:-1.505[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfinal[0.046]; David[0.045]; Newfoundland[0.043]; Newfoundland[0.043]; Newfoundland[0.043]; time[0.043]; manager[0.042]; Canadian[0.042]; Friday[0.042]; local[0.041]; coast[0.041]; people[0.041]; killed[0.041]; killed[0.041]; added[0.041]; began[0.041]; east[0.041]; Michigan[0.041]; miles[0.041]; route[0.041]; authorities[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mStephenville\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mStephenville, Newfoundland and Labrador\u001b[39m <---> \u001b[32mStephenville International Airport\u001b[39m\t\nSCORES: global= 0.267:0.246[\u001b[31m0.022\u001b[39m]; local()= 0.148:0.142[\u001b[31m0.006\u001b[39m]; log p(e|m)= -0.734:-1.760[\u001b[31m1.026\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \neast[0.045]; Stephenville[0.044]; airport[0.044]; airport[0.044]; province[0.044]; coast[0.043]; miles[0.043]; flight[0.043]; air[0.042]; Canadian[0.042]; refueling[0.041]; fly[0.041]; aircraft[0.041]; local[0.041]; reported[0.041]; Shannon[0.041]; km[0.040]; cargo[0.040]; Friday[0.040]; Snow[0.040]; Snow[0.040]; plane[0.040]; began[0.040]; traffic[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mStephenville\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mStephenville, Newfoundland and Labrador\u001b[39m <---> \u001b[32mStephenville International Airport\u001b[39m\t\nSCORES: global= 0.269:0.245[\u001b[31m0.023\u001b[39m]; local()= 0.174:0.148[\u001b[31m0.026\u001b[39m]; log p(e|m)= -0.734:-1.760[\u001b[31m1.026\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNewfoundland[0.044]; Newfoundland[0.044]; Newfoundland[0.044]; east[0.043]; pilot[0.043]; pilot[0.043]; Stephenville[0.043]; airport[0.043]; airport[0.043]; airport[0.043]; airport[0.043]; province[0.043]; coast[0.042]; miles[0.042]; route[0.041]; air[0.041]; Canadian[0.041]; aircraft[0.040]; local[0.040]; reported[0.040]; km[0.039]; Friday[0.039]; Snow[0.039]; Snow[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Shannon ==> ENTITY: \u001b[32mShannon Airport\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.186:0.186[0]; log p(e|m)= -2.847:-2.847[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.048]; airport[0.047]; airport[0.047]; refueling[0.044]; flight[0.043]; coast[0.043]; aircraft[0.042]; east[0.042]; cargo[0.041]; miles[0.041]; communication[0.041]; missing[0.041]; fly[0.041]; considered[0.040]; traffic[0.040]; discovered[0.040]; time[0.040]; local[0.040]; Friday[0.040]; stop[0.040]; reported[0.040]; province[0.040]; final[0.039]; crash[0.039]; \t\n\r [============================>.................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s390ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2878/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1390testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Madrid ==> ENTITY: \u001b[32mMadrid\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMadrid[0.047]; Spain[0.047]; league[0.046]; Soccer[0.044]; games[0.044]; match[0.043]; game[0.042]; force[0.042]; Barcelona[0.042]; Barcelona[0.042]; Barcelona[0.042]; international[0.042]; spending[0.040]; strike[0.040]; Croatian[0.040]; ahead[0.040]; ahead[0.040]; Real[0.040]; Real[0.040]; Real[0.040]; Real[0.040]; win[0.039]; scheduled[0.039]; striker[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.048]; games[0.046]; Soccer[0.046]; match[0.046]; Madrid[0.046]; Madrid[0.046]; international[0.045]; twice[0.044]; win[0.044]; game[0.044]; points[0.043]; Barcelona[0.043]; Barcelona[0.043]; Barcelona[0.043]; Barcelona[0.043]; Brazilian[0.042]; ahead[0.042]; ahead[0.042]; Real[0.041]; Real[0.041]; Real[0.041]; Real[0.041]; denied[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Clarence Seedorf\u001b[39m ==> ENTITY: \u001b[32mClarence Seedorf\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; games[0.044]; Ronaldo[0.043]; Real[0.043]; Real[0.043]; Real[0.043]; match[0.043]; Brazilian[0.043]; game[0.043]; striker[0.041]; striker[0.041]; international[0.041]; Mijatovic[0.041]; Mijatovic[0.041]; twice[0.041]; Madrid[0.041]; Madrid[0.041]; win[0.041]; Spain[0.041]; Davor[0.041]; Barcelona[0.040]; Barcelona[0.040]; Barcelona[0.040]; Barcelona[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Suker\u001b[39m ==> ENTITY: \u001b[32mDavor Šuker\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.048]; games[0.047]; Soccer[0.046]; match[0.046]; win[0.046]; game[0.046]; Spain[0.045]; twice[0.045]; Croatian[0.044]; Davor[0.044]; international[0.044]; points[0.042]; ahead[0.042]; ahead[0.042]; Predrag[0.042]; Montenegrin[0.042]; Madrid[0.042]; Madrid[0.042]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; half[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Real ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.196:0.196[0]; log p(e|m)= -3.352:-3.352[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazilian[0.047]; Soccer[0.046]; league[0.046]; Madrid[0.045]; Madrid[0.045]; Spain[0.045]; Barcelona[0.044]; Barcelona[0.044]; Barcelona[0.044]; Barcelona[0.044]; Real[0.044]; Real[0.044]; Real[0.044]; match[0.043]; twice[0.043]; games[0.042]; game[0.042]; striker[0.042]; striker[0.042]; Croatian[0.041]; international[0.041]; win[0.041]; spending[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Predrag Mijatovic\u001b[39m ==> ENTITY: \u001b[32mPredrag Mijatović\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; league[0.045]; twice[0.045]; match[0.045]; Davor[0.044]; Soccer[0.044]; Croatian[0.043]; Montenegrin[0.043]; Mijatovic[0.043]; Madrid[0.043]; Madrid[0.043]; Real[0.043]; Real[0.043]; Real[0.043]; Real[0.043]; Balkan[0.043]; Balkan[0.043]; striker[0.043]; game[0.043]; win[0.042]; points[0.042]; rivals[0.042]; Spain[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazilian ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.273:-0.273[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.048]; match[0.045]; games[0.045]; Spain[0.045]; game[0.044]; international[0.043]; twice[0.043]; Real[0.043]; Real[0.043]; win[0.042]; Ronaldo[0.040]; Barcelona[0.040]; Barcelona[0.040]; Barcelona[0.040]; Barcelona[0.040]; mix[0.040]; half[0.040]; striker[0.040]; striker[0.040]; points[0.040]; strongly[0.040]; work[0.040]; Saturday[0.039]; unusually[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMontenegrin\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMontenegro national football team\u001b[39m <---> \u001b[32mMontenegro\u001b[39m\t\nSCORES: global= 0.261:0.259[\u001b[31m0.002\u001b[39m]; local()= 0.155:0.106[\u001b[31m0.049\u001b[39m]; log p(e|m)= -2.501:-1.088[\u001b[32m1.413\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; Balkan[0.045]; match[0.044]; games[0.044]; twice[0.044]; game[0.044]; Croatian[0.043]; win[0.043]; Predrag[0.043]; points[0.041]; Win[0.041]; Davor[0.041]; break[0.041]; break[0.041]; international[0.040]; scheduled[0.040]; half[0.040]; ahead[0.040]; ahead[0.040]; striker[0.040]; striker[0.040]; Spain[0.040]; denied[0.039]; force[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barcelona ==> ENTITY: \u001b[32mFC Barcelona\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.184:0.184[0]; log p(e|m)= -1.709:-1.709[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.047]; league[0.047]; Soccer[0.046]; Madrid[0.044]; Madrid[0.044]; games[0.043]; match[0.042]; ahead[0.042]; ahead[0.042]; Barcelona[0.042]; Barcelona[0.042]; Barcelona[0.042]; Brazilian[0.041]; game[0.041]; twice[0.041]; points[0.040]; win[0.040]; Real[0.040]; Real[0.040]; Real[0.040]; Real[0.040]; international[0.039]; work[0.039]; scheduled[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Real ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.175:0.175[0]; log p(e|m)= -3.352:-3.352[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.046]; Soccer[0.045]; league[0.045]; Real[0.044]; Real[0.044]; Real[0.044]; Madrid[0.044]; Madrid[0.044]; Barcelona[0.042]; Barcelona[0.042]; Barcelona[0.042]; match[0.042]; games[0.041]; game[0.041]; Croatian[0.040]; international[0.040]; win[0.040]; striker[0.040]; spending[0.039]; break[0.039]; Mijatovic[0.039]; Mijatovic[0.039]; work[0.039]; scheduled[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Balkan ==> ENTITY: \u001b[32mBalkans\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.071:0.071[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.052]; Balkan[0.051]; games[0.050]; Croatian[0.048]; Soccer[0.048]; Montenegrin[0.047]; game[0.046]; international[0.045]; Davor[0.045]; win[0.045]; force[0.044]; match[0.044]; points[0.044]; Spain[0.044]; Mijatovic[0.043]; Mijatovic[0.043]; Predrag[0.043]; break[0.043]; break[0.043]; Madrid[0.043]; Madrid[0.043]; Old[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mijatovic\u001b[39m ==> ENTITY: \u001b[32mPredrag Mijatović\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; league[0.044]; twice[0.043]; match[0.043]; Davor[0.043]; Predrag[0.042]; Croatian[0.042]; Montenegrin[0.042]; Mijatovic[0.042]; Madrid[0.042]; Madrid[0.042]; Real[0.041]; Real[0.041]; Real[0.041]; Real[0.041]; Balkan[0.041]; striker[0.041]; striker[0.041]; game[0.041]; Brazilian[0.041]; win[0.041]; points[0.041]; rivals[0.041]; Spain[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Real ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.196:0.196[0]; log p(e|m)= -3.352:-3.352[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazilian[0.047]; Soccer[0.046]; league[0.046]; Madrid[0.045]; Madrid[0.045]; Spain[0.045]; Barcelona[0.044]; Barcelona[0.044]; Barcelona[0.044]; Barcelona[0.044]; Real[0.044]; Real[0.044]; Real[0.044]; match[0.043]; twice[0.043]; games[0.042]; game[0.042]; striker[0.042]; striker[0.042]; Croatian[0.041]; international[0.041]; win[0.041]; spending[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barcelona ==> ENTITY: \u001b[32mFC Barcelona\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.184:0.184[0]; log p(e|m)= -1.709:-1.709[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.047]; league[0.047]; Soccer[0.046]; Madrid[0.044]; Madrid[0.044]; games[0.043]; match[0.042]; ahead[0.042]; ahead[0.042]; Barcelona[0.042]; Barcelona[0.042]; Barcelona[0.042]; Brazilian[0.041]; game[0.041]; twice[0.041]; points[0.040]; win[0.040]; Real[0.040]; Real[0.040]; Real[0.040]; Real[0.040]; international[0.039]; work[0.039]; scheduled[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barcelona ==> ENTITY: \u001b[32mFC Barcelona\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.162:0.162[0]; log p(e|m)= -1.709:-1.709[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.048]; league[0.047]; twice[0.045]; games[0.044]; match[0.043]; Barcelona[0.042]; Barcelona[0.042]; Barcelona[0.042]; Brazilian[0.042]; ahead[0.042]; ahead[0.042]; game[0.042]; win[0.041]; Real[0.041]; Real[0.041]; work[0.040]; scheduled[0.040]; international[0.040]; striker[0.040]; striker[0.040]; mix[0.039]; rivals[0.039]; spending[0.039]; strongly[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Madrid\u001b[39m ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; Soccer[0.047]; Madrid[0.044]; Spain[0.044]; games[0.044]; game[0.043]; Barcelona[0.043]; Barcelona[0.043]; Barcelona[0.043]; match[0.042]; Real[0.042]; Real[0.042]; Real[0.042]; points[0.041]; win[0.040]; international[0.040]; striker[0.040]; ahead[0.039]; ahead[0.039]; Mijatovic[0.039]; Mijatovic[0.039]; Game[0.039]; rivals[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barcelona ==> ENTITY: \u001b[32mFC Barcelona\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.185:0.185[0]; log p(e|m)= -1.709:-1.709[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.049]; league[0.048]; Soccer[0.047]; twice[0.046]; Madrid[0.045]; Madrid[0.045]; games[0.045]; match[0.044]; ahead[0.043]; ahead[0.043]; Barcelona[0.043]; Barcelona[0.043]; Barcelona[0.043]; Brazilian[0.043]; game[0.043]; win[0.042]; Real[0.041]; Real[0.041]; Real[0.041]; Real[0.041]; work[0.041]; scheduled[0.041]; international[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Davor Suker\u001b[39m ==> ENTITY: \u001b[32mDavor Šuker\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; games[0.045]; Soccer[0.045]; match[0.044]; win[0.044]; game[0.044]; Spain[0.043]; twice[0.043]; Croatian[0.043]; international[0.042]; points[0.041]; ahead[0.041]; ahead[0.041]; Predrag[0.040]; Montenegrin[0.040]; Madrid[0.040]; Madrid[0.040]; Barcelona[0.040]; Barcelona[0.040]; Barcelona[0.040]; Barcelona[0.040]; half[0.040]; striker[0.040]; rivals[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ronaldo ==> ENTITY: \u001b[32mRonaldo\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.172:0.172[0]; log p(e|m)= -0.109:-0.109[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; Brazilian[0.046]; striker[0.044]; striker[0.044]; games[0.043]; game[0.043]; Barcelona[0.043]; Barcelona[0.043]; Barcelona[0.043]; twice[0.042]; Real[0.042]; Real[0.042]; Spain[0.042]; match[0.041]; international[0.041]; old[0.041]; ahead[0.041]; ahead[0.041]; Mijatovic[0.041]; Seedorf[0.039]; points[0.039]; denied[0.039]; let[0.039]; rivals[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Croatian ==> ENTITY: \u001b[32mCroatia\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.140:0.140[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.049]; games[0.047]; win[0.047]; match[0.047]; Soccer[0.047]; international[0.046]; Balkan[0.045]; Balkan[0.045]; twice[0.045]; game[0.044]; Davor[0.043]; Predrag[0.043]; Spain[0.043]; Montenegrin[0.042]; Win[0.042]; ahead[0.041]; ahead[0.041]; striker[0.041]; striker[0.041]; rivals[0.041]; points[0.041]; half[0.040]; Saturday[0.040]; \t\n\r [============================>.................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s312ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2898/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1320testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hamburg ==> ENTITY: \u001b[32mHamburg\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.129:0.129[0]; log p(e|m)= -0.272:-0.272[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGerman[0.043]; German[0.043]; Salvador[0.043]; shipped[0.043]; compared[0.043]; seventh[0.043]; million[0.042]; million[0.042]; million[0.042]; million[0.042]; position[0.042]; July[0.041]; July[0.041]; January[0.041]; August[0.041]; August[0.041]; August[0.041]; August[0.041]; August[0.041]; ago[0.041]; outside[0.041]; Brazil[0.041]; Jan[0.040]; period[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colombia ==> ENTITY: \u001b[32mColombia\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.103:0.103[0]; log p(e|m)= -0.223:-0.223[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.048]; Indonesia[0.047]; imports[0.044]; imports[0.044]; Kenya[0.043]; Salvador[0.042]; Ethiopia[0.042]; seventh[0.041]; million[0.041]; million[0.041]; million[0.041]; million[0.041]; German[0.041]; German[0.041]; July[0.041]; July[0.041]; coffee[0.040]; coffee[0.040]; coffee[0.040]; period[0.040]; January[0.040]; said[0.040]; El[0.040]; Imports[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kenya ==> ENTITY: \u001b[32mKenya\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.095:0.095[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEthiopia[0.058]; Indonesia[0.052]; Brazil[0.046]; Colombia[0.045]; August[0.043]; August[0.043]; August[0.043]; August[0.043]; August[0.043]; January[0.042]; coffee[0.042]; coffee[0.042]; coffee[0.042]; Hamburg[0.042]; Hamburg[0.042]; seventh[0.042]; July[0.042]; July[0.042]; detailed[0.041]; compared[0.041]; German[0.041]; German[0.041]; Imports[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ethiopia ==> ENTITY: \u001b[32mEthiopia\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.101:0.101[0]; log p(e|m)= -0.236:-0.236[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKenya[0.048]; January[0.042]; Colombia[0.042]; August[0.042]; August[0.042]; August[0.042]; August[0.042]; August[0.042]; July[0.042]; July[0.042]; million[0.042]; million[0.042]; million[0.042]; million[0.042]; Brazil[0.041]; period[0.041]; shipped[0.041]; Indonesia[0.041]; seventh[0.041]; position[0.040]; imports[0.040]; imports[0.040]; said[0.040]; outside[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazil ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nColombia[0.046]; Salvador[0.046]; Indonesia[0.044]; Kenya[0.043]; seventh[0.042]; million[0.042]; million[0.042]; million[0.042]; million[0.042]; outside[0.042]; ago[0.042]; January[0.041]; July[0.041]; July[0.041]; Hamburg[0.041]; Hamburg[0.041]; August[0.041]; August[0.041]; August[0.041]; August[0.041]; August[0.041]; period[0.040]; compared[0.040]; German[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.142:0.142[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGerman[0.048]; Hamburg[0.043]; seventh[0.043]; August[0.043]; August[0.043]; August[0.043]; August[0.043]; August[0.043]; January[0.042]; July[0.042]; July[0.042]; million[0.042]; million[0.042]; million[0.042]; million[0.042]; Brazil[0.041]; compared[0.041]; Colombia[0.041]; ago[0.040]; Jan[0.040]; shipped[0.040]; period[0.040]; association[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAugust[0.044]; August[0.044]; August[0.044]; August[0.044]; August[0.044]; January[0.044]; July[0.044]; July[0.044]; Kenya[0.043]; Colombia[0.043]; Ethiopia[0.042]; Brazil[0.042]; seventh[0.041]; ago[0.040]; said[0.040]; million[0.040]; million[0.040]; million[0.040]; million[0.040]; period[0.040]; detailed[0.040]; compared[0.040]; outside[0.039]; position[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = El Salvador ==> ENTITY: \u001b[32mEl Salvador\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.126:0.126[0]; log p(e|m)= -0.358:-0.358[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJuly[0.050]; July[0.050]; Brazil[0.046]; Colombia[0.046]; August[0.046]; August[0.046]; August[0.046]; August[0.046]; August[0.046]; January[0.045]; seventh[0.041]; position[0.038]; million[0.038]; million[0.038]; million[0.038]; million[0.038]; said[0.038]; Indonesia[0.038]; shipped[0.038]; outside[0.038]; Ethiopia[0.038]; Kenya[0.038]; Hamburg[0.037]; Hamburg[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hamburg ==> ENTITY: \u001b[32mHamburg\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.272:-0.272[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSalvador[0.048]; compared[0.047]; seventh[0.047]; million[0.046]; million[0.046]; million[0.046]; million[0.046]; position[0.046]; July[0.046]; July[0.046]; January[0.045]; shipped[0.045]; August[0.045]; August[0.045]; August[0.045]; August[0.045]; ago[0.045]; outside[0.045]; Brazil[0.045]; Colombia[0.045]; period[0.044]; said[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.131:0.131[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGerman[0.049]; Hamburg[0.043]; August[0.043]; August[0.043]; August[0.043]; August[0.043]; August[0.043]; January[0.042]; July[0.042]; July[0.042]; million[0.042]; million[0.042]; million[0.042]; million[0.042]; Kenya[0.042]; shipped[0.041]; Indonesia[0.041]; Jan[0.040]; Ethiopia[0.040]; Colombia[0.040]; outside[0.040]; period[0.039]; ago[0.039]; association[0.039]; \t\n\r [============================>.................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s260ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2908/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1382testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Abu Dhabi ==> ENTITY: \u001b[32mAbu Dhabi\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.197:0.197[0]; log p(e|m)= -0.066:-0.066[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKuwait[0.050]; Kuwait[0.050]; Emirates[0.047]; Al[0.046]; Al[0.046]; matches[0.046]; Soccer[0.046]; Saad[0.044]; Arab[0.044]; Ahmed[0.043]; played[0.043]; Cup[0.042]; Cup[0.042]; Korea[0.042]; Korea[0.042]; Adnan[0.042]; Hassan[0.042]; standings[0.041]; Jassem[0.041]; Asian[0.041]; Asian[0.041]; won[0.040]; Saturday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kuwait ==> ENTITY: \u001b[32mKuwait national football team\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.210:0.210[0]; log p(e|m)= -2.364:-2.364[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKuwait[0.051]; Kuwait[0.051]; Arab[0.042]; Al[0.042]; Al[0.042]; matches[0.042]; Asian[0.042]; Asian[0.042]; won[0.042]; Korea[0.041]; Korea[0.041]; Korea[0.041]; Soccer[0.041]; Emirates[0.041]; Hassan[0.040]; Ahmed[0.040]; played[0.040]; Cup[0.040]; Cup[0.040]; Jassem[0.040]; Saad[0.040]; Adnan[0.039]; lost[0.039]; United[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia national football team\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.188:0.188[0]; log p(e|m)= -2.489:-2.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.052]; Indonesia[0.052]; won[0.045]; Korea[0.045]; Korea[0.045]; Korea[0.045]; Putra[0.041]; lost[0.041]; South[0.041]; South[0.041]; South[0.041]; played[0.041]; Widodo[0.041]; Kuwait[0.040]; Kuwait[0.040]; Kuwait[0.040]; standings[0.040]; goals[0.040]; Hong[0.039]; Emirates[0.038]; Woon[0.038]; Group[0.038]; Hwang[0.038]; drawn[0.037]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSouth Korea\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSouth Korea\u001b[39m <---> \u001b[32mSouth Korea national football team\u001b[39m\t\nSCORES: global= 0.255:0.254[\u001b[31m0.000\u001b[39m]; local()= 0.162:0.175[\u001b[32m0.013\u001b[39m]; log p(e|m)= -0.224:-2.847[\u001b[31m2.623\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.049]; Korea[0.049]; won[0.046]; matches[0.045]; South[0.043]; South[0.043]; Asian[0.042]; Cup[0.042]; played[0.042]; Kuwait[0.041]; Kuwait[0.041]; Kuwait[0.041]; Hwang[0.041]; standings[0.041]; Kim[0.040]; Hong[0.040]; Jeon[0.039]; lost[0.039]; group[0.039]; Group[0.039]; Indonesia[0.039]; Indonesia[0.039]; Indonesia[0.039]; Hoon[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Korea ==> ENTITY: \u001b[32mSouth Korea national football team\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.194:0.194[0]; log p(e|m)= -2.847:-2.847[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.052]; Korea[0.052]; Soccer[0.048]; matches[0.048]; Asian[0.047]; Asian[0.047]; won[0.046]; South[0.046]; South[0.046]; Cup[0.045]; Cup[0.045]; Kuwait[0.044]; Kuwait[0.044]; Kuwait[0.044]; Hwang[0.044]; standings[0.044]; played[0.044]; Kim[0.043]; Hong[0.043]; Jeon[0.042]; group[0.042]; Group[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mUAE\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUnited Arab Emirates\u001b[39m <---> \u001b[32mUnited Arab Emirates national football team\u001b[39m\t\nSCORES: global= 0.263:0.251[\u001b[31m0.012\u001b[39m]; local()= 0.222:0.225[\u001b[32m0.003\u001b[39m]; log p(e|m)= -0.195:-2.283[\u001b[31m2.088\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKuwait[0.047]; Kuwait[0.047]; Kuwait[0.047]; Emirates[0.046]; Asian[0.044]; Asian[0.044]; matches[0.043]; Arab[0.043]; United[0.042]; Cup[0.041]; Cup[0.041]; Korea[0.041]; Korea[0.041]; Korea[0.041]; Soccer[0.040]; played[0.040]; Indonesia[0.039]; Indonesia[0.039]; Indonesia[0.039]; Saad[0.039]; Ahmed[0.039]; won[0.039]; Al[0.039]; Al[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mUAE\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUnited Arab Emirates\u001b[39m <---> \u001b[32mUnited Arab Emirates national football team\u001b[39m\t\nSCORES: global= 0.260:0.249[\u001b[31m0.011\u001b[39m]; local()= 0.211:0.234[\u001b[32m0.022\u001b[39m]; log p(e|m)= -0.195:-2.283[\u001b[31m2.088\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKuwait[0.048]; Kuwait[0.048]; Kuwait[0.048]; matches[0.045]; Emirates[0.044]; Arab[0.042]; won[0.042]; Korea[0.041]; Korea[0.041]; Korea[0.041]; played[0.041]; Indonesia[0.040]; Indonesia[0.040]; Indonesia[0.040]; Saad[0.040]; standings[0.040]; Ahmed[0.040]; United[0.040]; Al[0.040]; Al[0.040]; Hassan[0.040]; Saturday[0.039]; lost[0.039]; Jassem[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia national football team\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.214:0.214[0]; log p(e|m)= -2.489:-2.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.049]; Indonesia[0.049]; Asian[0.047]; Asian[0.047]; matches[0.044]; won[0.043]; Korea[0.043]; Korea[0.043]; Korea[0.043]; Putra[0.041]; Cup[0.041]; Cup[0.041]; Soccer[0.041]; group[0.040]; lost[0.039]; South[0.039]; South[0.039]; South[0.039]; played[0.039]; Widodo[0.039]; Kuwait[0.039]; Kuwait[0.039]; Kuwait[0.039]; standings[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Asian Cup ==> ENTITY: \u001b[32mAFC Asian Cup\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.211:0.211[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKuwait[0.045]; Kuwait[0.045]; Cup[0.045]; Asian[0.044]; Korea[0.044]; Korea[0.044]; Saturday[0.044]; Soccer[0.043]; matches[0.043]; Emirates[0.043]; won[0.041]; played[0.040]; Indonesia[0.040]; Indonesia[0.040]; Results[0.040]; Results[0.040]; Arab[0.040]; standings[0.040]; Saad[0.040]; Hassan[0.040]; Ahmed[0.040]; South[0.039]; South[0.039]; Group[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mUnited Arab Emirates\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUnited Arab Emirates\u001b[39m <---> \u001b[32mUnited Arab Emirates national football team\u001b[39m\t\nSCORES: global= 0.253:0.244[\u001b[31m0.009\u001b[39m]; local()= 0.203:0.220[\u001b[32m0.017\u001b[39m]; log p(e|m)= -0.250:-2.216[\u001b[31m1.967\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKuwait[0.047]; Kuwait[0.047]; Kuwait[0.047]; Asian[0.045]; Asian[0.045]; matches[0.044]; Cup[0.042]; Cup[0.042]; won[0.042]; Korea[0.041]; Korea[0.041]; Korea[0.041]; Soccer[0.041]; played[0.040]; Indonesia[0.040]; Indonesia[0.040]; Indonesia[0.040]; Saad[0.040]; standings[0.040]; Ahmed[0.040]; Al[0.039]; Al[0.039]; Hassan[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kuwait ==> ENTITY: \u001b[32mKuwait national football team\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.197:0.197[0]; log p(e|m)= -2.364:-2.364[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKuwait[0.052]; Kuwait[0.052]; Korea[0.043]; Korea[0.043]; Korea[0.043]; Arab[0.043]; matches[0.043]; won[0.043]; Emirates[0.042]; played[0.041]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Saad[0.040]; lost[0.040]; Hassan[0.039]; United[0.039]; standings[0.039]; Al[0.039]; Al[0.039]; Hong[0.039]; Ahmed[0.039]; group[0.039]; goals[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Korea ==> ENTITY: \u001b[32mSouth Korea national football team\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.193:0.193[0]; log p(e|m)= -2.847:-2.847[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.050]; Korea[0.050]; won[0.047]; matches[0.047]; Soccer[0.046]; South[0.044]; South[0.044]; Asian[0.044]; Asian[0.044]; Cup[0.043]; Cup[0.043]; played[0.043]; Kuwait[0.042]; Kuwait[0.042]; Kuwait[0.042]; Hwang[0.042]; standings[0.042]; Kim[0.042]; Hong[0.041]; Jeon[0.041]; lost[0.040]; group[0.040]; Group[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Adnan Al Talyani\u001b[39m ==> ENTITY: \u001b[32mAdnan Al Talyani\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKuwait[0.045]; Kuwait[0.045]; Kuwait[0.045]; Soccer[0.044]; Emirates[0.044]; Al[0.043]; Saad[0.042]; group[0.042]; matches[0.042]; played[0.042]; Saturday[0.042]; United[0.041]; Arab[0.041]; Putra[0.041]; won[0.041]; halftime[0.041]; Scorers[0.040]; Scorers[0.040]; Attendance[0.040]; Attendance[0.040]; Jassem[0.040]; Widodo[0.040]; Ahmed[0.040]; goals[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Asian Cup ==> ENTITY: \u001b[32mAFC Asian Cup\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.220:0.220[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngoals[0.046]; Kuwait[0.045]; Kuwait[0.045]; Cup[0.045]; Asian[0.044]; Korea[0.044]; Korea[0.044]; Korea[0.044]; Saturday[0.043]; Soccer[0.043]; matches[0.043]; won[0.041]; Emirates[0.040]; points[0.040]; played[0.040]; Indonesia[0.040]; Indonesia[0.040]; Results[0.040]; Results[0.040]; Arab[0.040]; standings[0.039]; Saad[0.039]; Hassan[0.039]; Ahmed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hwang Sun Hong\u001b[39m ==> ENTITY: \u001b[32mHwang Sun-Hong\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.050]; Korea[0.050]; Korea[0.050]; Cup[0.048]; Cup[0.048]; matches[0.047]; Asian[0.047]; Asian[0.047]; Soccer[0.047]; Kim[0.046]; won[0.046]; goals[0.045]; played[0.044]; Jeon[0.044]; standings[0.043]; Koo[0.043]; Putra[0.043]; halftime[0.043]; Scorers[0.043]; Scorers[0.043]; Attendance[0.043]; Attendance[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kuwait ==> ENTITY: \u001b[32mKuwait national football team\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.210:0.210[0]; log p(e|m)= -2.364:-2.364[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKuwait[0.051]; Kuwait[0.051]; Arab[0.042]; Al[0.042]; Al[0.042]; matches[0.042]; Asian[0.042]; Asian[0.042]; won[0.042]; Korea[0.041]; Korea[0.041]; Korea[0.041]; Soccer[0.041]; Emirates[0.041]; Hassan[0.040]; Ahmed[0.040]; played[0.040]; Cup[0.040]; Cup[0.040]; Jassem[0.040]; Saad[0.040]; Adnan[0.039]; lost[0.039]; United[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia national football team\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.213:0.213[0]; log p(e|m)= -2.489:-2.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.049]; Indonesia[0.049]; Asian[0.047]; Asian[0.047]; matches[0.044]; won[0.043]; Korea[0.043]; Korea[0.043]; Korea[0.043]; Cup[0.041]; Cup[0.041]; Soccer[0.041]; Emirates[0.040]; Putra[0.040]; lost[0.039]; standings[0.039]; South[0.039]; South[0.039]; South[0.039]; played[0.039]; Widodo[0.039]; Kuwait[0.039]; Kuwait[0.039]; Kuwait[0.039]; \t\n\r [=============================>................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s184ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2925/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1337testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.092:0.092[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTokyo[0.047]; Shimpei[0.046]; released[0.043]; despite[0.043]; Agency[0.043]; Vice[0.042]; September[0.042]; early[0.042]; July[0.042]; April[0.042]; Tuesday[0.042]; domestic[0.042]; Minister[0.041]; recent[0.041]; planned[0.041]; reporters[0.041]; Friday[0.041]; fiscal[0.040]; fiscal[0.040]; says[0.040]; economy[0.040]; economy[0.040]; economy[0.040]; economic[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tokyo ==> ENTITY: \u001b[32mTokyo\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.101:0.101[0]; log p(e|m)= -0.093:-0.093[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.048]; Shimpei[0.044]; September[0.044]; Friday[0.043]; Tuesday[0.042]; economy[0.042]; economy[0.042]; economy[0.042]; July[0.042]; April[0.042]; Economic[0.042]; fiscal[0.042]; fiscal[0.042]; recent[0.041]; economic[0.041]; told[0.041]; domestic[0.041]; despite[0.041]; beginning[0.040]; said[0.040]; said[0.040]; reflected[0.040]; reporters[0.040]; Agency[0.040]; \t\n\r [=============================>................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s173ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2927/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1176testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rio De Janeiro\u001b[39m ==> ENTITY: \u001b[32mRio de Janeiro\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazilian[0.065]; Brazilian[0.065]; matches[0.063]; Championship[0.061]; leg[0.061]; soccer[0.061]; Portuguesa[0.059]; Goias[0.059]; Gremio[0.059]; semifinal[0.058]; Atletico[0.058]; Soccer[0.058]; championship[0.056]; Mineiro[0.056]; Thursday[0.056]; Results[0.052]; Results[0.052]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Goias ==> ENTITY: \u001b[32mGoiás Esporte Clube\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.203:0.203[0]; log p(e|m)= -0.957:-0.957[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.062]; Brazilian[0.060]; Brazilian[0.060]; soccer[0.059]; matches[0.059]; Janeiro[0.059]; Mineiro[0.058]; Rio[0.057]; leg[0.056]; Portuguesa[0.056]; Championship[0.055]; Atletico[0.055]; Gremio[0.054]; semifinal[0.053]; championship[0.051]; Thursday[0.050]; Results[0.048]; Results[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Atletico Mineiro\u001b[39m ==> ENTITY: \u001b[32mClube Atlético Mineiro\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGremio[0.063]; Soccer[0.062]; championship[0.062]; Championship[0.062]; soccer[0.062]; matches[0.061]; Goias[0.061]; Portuguesa[0.061]; Brazilian[0.060]; Brazilian[0.060]; leg[0.060]; Janeiro[0.058]; semifinal[0.057]; Rio[0.057]; Thursday[0.053]; Results[0.051]; Results[0.051]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazilian ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.273:-0.273[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazilian[0.062]; soccer[0.062]; matches[0.061]; Soccer[0.061]; Goias[0.059]; Mineiro[0.058]; Portuguesa[0.056]; Championship[0.056]; Janeiro[0.056]; Gremio[0.056]; Rio[0.056]; championship[0.055]; Atletico[0.055]; semifinal[0.052]; leg[0.050]; Thursday[0.049]; Results[0.048]; Results[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portuguesa ==> ENTITY: \u001b[32mAssociação Portuguesa de Desportos\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.143:-0.143[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRio[0.066]; Janeiro[0.064]; Soccer[0.061]; Goias[0.060]; Brazilian[0.059]; Brazilian[0.059]; Mineiro[0.059]; soccer[0.059]; Gremio[0.057]; Atletico[0.056]; matches[0.055]; Championship[0.053]; leg[0.052]; championship[0.051]; Thursday[0.050]; semifinal[0.047]; Results[0.047]; Results[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gremio\u001b[39m ==> ENTITY: \u001b[32mGrêmio Foot-Ball Porto Alegrense\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMineiro[0.061]; Soccer[0.060]; soccer[0.059]; matches[0.059]; Brazilian[0.058]; Brazilian[0.058]; Atletico[0.058]; Championship[0.057]; Portuguesa[0.057]; Goias[0.057]; Janeiro[0.055]; leg[0.055]; championship[0.055]; Rio[0.054]; semifinal[0.052]; Thursday[0.050]; Results[0.047]; Results[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazilian ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.273:-0.273[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazilian[0.062]; soccer[0.062]; matches[0.061]; Soccer[0.061]; Goias[0.059]; Mineiro[0.058]; Portuguesa[0.056]; Championship[0.056]; Janeiro[0.056]; Gremio[0.056]; Rio[0.056]; championship[0.055]; Atletico[0.055]; semifinal[0.052]; leg[0.050]; Thursday[0.049]; Results[0.048]; Results[0.048]; \t\n\r [=============================>................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s218ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2934/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1198testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Paris ==> ENTITY: \u001b[32mParis\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.115:-0.115[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nParis[0.094]; matches[0.090]; Soccer[0.087]; division[0.085]; League[0.085]; Friday[0.084]; French[0.081]; French[0.081]; Germain[0.080]; Nancy[0.079]; Nantes[0.078]; Lens[0.077]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lens ==> ENTITY: \u001b[32mRC Lens\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.169:0.169[0]; log p(e|m)= -0.860:-0.860[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.081]; matches[0.081]; Nantes[0.078]; Soccer[0.077]; French[0.076]; French[0.076]; division[0.070]; Paris[0.069]; Paris[0.069]; Nancy[0.067]; Friday[0.066]; Germain[0.066]; Results[0.062]; Results[0.062]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nancy ==> ENTITY: \u001b[32mAS Nancy\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.216:0.216[0]; log p(e|m)= -1.687:-1.687[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.080]; League[0.079]; Nantes[0.074]; Soccer[0.074]; French[0.074]; French[0.074]; Paris[0.073]; Paris[0.073]; Lens[0.071]; Friday[0.070]; division[0.068]; Germain[0.066]; Results[0.062]; Results[0.062]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.119:0.119[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrench[0.081]; matches[0.079]; Soccer[0.077]; League[0.076]; Paris[0.075]; Paris[0.075]; Nantes[0.072]; Nancy[0.071]; Lens[0.071]; Germain[0.071]; Friday[0.068]; division[0.066]; Results[0.058]; Results[0.058]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paris St Germain\u001b[39m ==> ENTITY: \u001b[32mParis Saint-Germain F.C.\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.087]; matches[0.085]; Soccer[0.082]; Nantes[0.081]; division[0.079]; Lens[0.078]; French[0.077]; French[0.077]; Paris[0.077]; Friday[0.074]; Nancy[0.071]; Results[0.066]; Results[0.066]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nantes ==> ENTITY: \u001b[32mFC Nantes\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.197:0.197[0]; log p(e|m)= -1.457:-1.457[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.080]; League[0.077]; Soccer[0.077]; Paris[0.075]; Paris[0.075]; French[0.074]; French[0.074]; division[0.074]; Lens[0.073]; Friday[0.068]; Germain[0.068]; Nancy[0.064]; Results[0.061]; Results[0.061]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.119:0.119[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrench[0.081]; matches[0.079]; Soccer[0.077]; League[0.076]; Paris[0.075]; Paris[0.075]; Nantes[0.072]; Nancy[0.071]; Lens[0.071]; Germain[0.071]; Friday[0.068]; division[0.066]; Results[0.058]; Results[0.058]; \t\n\r [=============================>................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s183ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2941/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1354testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.149:0.149[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.049]; France[0.049]; French[0.043]; season[0.043]; Cup[0.043]; final[0.043]; Canada[0.042]; place[0.041]; event[0.041]; seventh[0.041]; seventh[0.041]; Sebastien[0.041]; Blanc[0.041]; Blanc[0.041]; women[0.041]; competition[0.041]; competition[0.041]; overall[0.040]; China[0.040]; ninth[0.040]; finish[0.040]; cup[0.040]; world[0.040]; finishing[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.164:0.164[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.049]; France[0.049]; French[0.042]; season[0.042]; Cup[0.042]; Cup[0.042]; Cup[0.042]; final[0.042]; second[0.042]; Canada[0.042]; Canada[0.042]; Canada[0.042]; win[0.041]; place[0.040]; place[0.040]; event[0.040]; event[0.040]; seventh[0.040]; seventh[0.040]; Sebastien[0.040]; finished[0.040]; Blanc[0.040]; Blanc[0.040]; women[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Evelyne Leu\u001b[39m ==> ENTITY: \u001b[32mEvelyne Leu\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.044]; second[0.044]; Tignes[0.044]; skiers[0.044]; Swiss[0.044]; ninth[0.044]; Cup[0.043]; podium[0.042]; podium[0.042]; finished[0.042]; narrowly[0.041]; ahead[0.041]; wins[0.041]; champion[0.041]; Karin[0.041]; mark[0.041]; trip[0.041]; victory[0.040]; career[0.040]; event[0.040]; Canada[0.040]; Canada[0.040]; World[0.040]; women[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Veronica Brenner\u001b[39m ==> ENTITY: \u001b[32mVeronica Brenner\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.046]; Cup[0.046]; Cup[0.046]; jump[0.045]; Canada[0.045]; Canada[0.045]; Canada[0.045]; aerials[0.045]; aerials[0.045]; Tignes[0.044]; finish[0.044]; finished[0.043]; podium[0.043]; podium[0.043]; skiers[0.043]; World[0.043]; World[0.043]; World[0.043]; Nannan[0.042]; event[0.042]; season[0.041]; career[0.040]; career[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.110:0.110[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.051]; Canada[0.051]; season[0.042]; overall[0.042]; overall[0.042]; wins[0.041]; Cup[0.041]; Cup[0.041]; Cup[0.041]; failed[0.041]; win[0.041]; ninth[0.041]; ninth[0.041]; podium[0.041]; podium[0.041]; World[0.041]; World[0.041]; World[0.041]; second[0.040]; second[0.040]; finished[0.040]; day[0.040]; seventh[0.040]; seventh[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.052]; Canada[0.052]; season[0.042]; Cup[0.041]; Cup[0.041]; Cup[0.041]; failed[0.041]; win[0.041]; finishing[0.041]; podium[0.041]; World[0.041]; World[0.041]; World[0.041]; event[0.041]; event[0.041]; Freestyle[0.041]; finished[0.040]; second[0.040]; day[0.040]; final[0.040]; finish[0.040]; place[0.040]; place[0.040]; world[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chinese ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.132:0.132[0]; log p(e|m)= -1.419:-1.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChinese[0.054]; China[0.053]; Canada[0.049]; Canada[0.049]; world[0.046]; Xu[0.045]; women[0.045]; men[0.042]; respectable[0.042]; gain[0.042]; dominated[0.042]; French[0.042]; second[0.041]; Nannan[0.041]; World[0.041]; World[0.041]; day[0.041]; place[0.041]; France[0.041]; France[0.041]; France[0.041]; event[0.041]; distant[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tignes\u001b[39m ==> ENTITY: \u001b[32mTignes\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nevent[0.046]; skiers[0.044]; place[0.043]; resort[0.043]; France[0.043]; Cup[0.042]; Cup[0.042]; Cup[0.042]; trip[0.042]; overall[0.042]; combined[0.042]; competition[0.042]; competition[0.042]; Sebastien[0.042]; day[0.041]; season[0.041]; ahead[0.041]; places[0.041]; aerials[0.040]; aerials[0.040]; jump[0.040]; finish[0.040]; second[0.040]; second[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jeff Bean\u001b[39m ==> ENTITY: \u001b[32mJeff Bean\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.276:0.276[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.044]; second[0.044]; finished[0.043]; finishing[0.043]; event[0.043]; World[0.042]; World[0.042]; World[0.042]; finish[0.042]; ninth[0.042]; ninth[0.042]; seventh[0.041]; seventh[0.041]; place[0.041]; place[0.041]; jump[0.041]; Cup[0.041]; Cup[0.041]; Cup[0.041]; overall[0.041]; overall[0.041]; aerials[0.041]; aerials[0.041]; season[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sebastien Foucras\u001b[39m ==> ENTITY: \u001b[32mSébastien Foucras\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nskiing[0.047]; Cup[0.046]; Cup[0.046]; Cup[0.046]; Tignes[0.044]; Tignes[0.044]; Tignes[0.044]; season[0.044]; freestyle[0.043]; Freestyle[0.043]; win[0.043]; wins[0.043]; Canada[0.043]; Canada[0.043]; Canada[0.043]; aerials[0.043]; aerials[0.043]; aerials[0.043]; finished[0.043]; finishing[0.042]; event[0.042]; event[0.042]; Saturday[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tignes\u001b[39m ==> ENTITY: \u001b[32mTignes\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTignes[0.047]; skiing[0.046]; event[0.045]; Saturday[0.043]; place[0.043]; freestyle[0.042]; Skiing[0.042]; Skiing[0.042]; resort[0.042]; France[0.042]; France[0.042]; France[0.042]; overall[0.041]; world[0.041]; Freestyle[0.041]; competition[0.041]; competition[0.041]; Sebastien[0.041]; day[0.040]; season[0.040]; finishing[0.040]; aerials[0.039]; aerials[0.039]; finish[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.050]; France[0.050]; France[0.050]; Canada[0.045]; Cup[0.043]; Cup[0.043]; Sebastien[0.042]; Blanc[0.041]; Blanc[0.041]; women[0.041]; debut[0.040]; cup[0.040]; season[0.040]; Cuo[0.040]; Chinese[0.040]; Chinese[0.040]; final[0.040]; win[0.040]; World[0.039]; World[0.039]; career[0.039]; Tignes[0.039]; Tignes[0.039]; men[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.048]; France[0.048]; French[0.042]; Cup[0.042]; Cup[0.042]; Cup[0.042]; final[0.042]; second[0.042]; Canada[0.042]; Canada[0.042]; Canada[0.042]; champion[0.042]; win[0.041]; season[0.041]; victory[0.041]; wins[0.040]; event[0.040]; event[0.040]; seventh[0.040]; seventh[0.040]; Sebastien[0.040]; finished[0.040]; overall[0.040]; overall[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Xu Nannan\u001b[39m ==> ENTITY: \u001b[32mXu Nannan\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \naerials[0.048]; aerials[0.048]; aerials[0.048]; Chinese[0.046]; Chinese[0.046]; freestyle[0.045]; China[0.045]; skiing[0.045]; Freestyle[0.045]; finishing[0.043]; finished[0.043]; Tignes[0.043]; Tignes[0.043]; Skiing[0.042]; Skiing[0.042]; finish[0.042]; jump[0.041]; seventh[0.041]; seventh[0.041]; place[0.041]; place[0.041]; competition[0.041]; competition[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swiss ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.124:0.124[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.044]; second[0.044]; win[0.043]; wins[0.043]; place[0.043]; event[0.043]; ninth[0.042]; overall[0.042]; women[0.042]; French[0.042]; Cup[0.042]; Cup[0.042]; Cup[0.042]; Canada[0.041]; Canada[0.041]; champion[0.041]; competition[0.041]; career[0.041]; ahead[0.040]; score[0.040]; score[0.040]; combined[0.040]; podium[0.040]; podium[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.119:0.119[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.054]; Canada[0.054]; season[0.044]; wins[0.043]; Cup[0.043]; Cup[0.043]; Cup[0.043]; win[0.043]; podium[0.043]; podium[0.043]; World[0.042]; World[0.042]; World[0.042]; event[0.042]; finished[0.042]; second[0.042]; second[0.042]; day[0.042]; finish[0.042]; place[0.042]; France[0.042]; France[0.042]; victory[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.073:0.073[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChinese[0.046]; Chinese[0.046]; Freestyle[0.044]; Xu[0.043]; competition[0.043]; competition[0.043]; event[0.043]; Cup[0.042]; Canada[0.042]; freestyle[0.041]; win[0.041]; finishing[0.041]; World[0.041]; aerials[0.041]; aerials[0.041]; skiing[0.041]; finish[0.040]; debut[0.040]; Cuo[0.040]; world[0.040]; France[0.040]; France[0.040]; France[0.040]; season[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tignes\u001b[39m ==> ENTITY: \u001b[32mTignes\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.217:0.217[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTignes[0.046]; skiing[0.045]; event[0.045]; Saturday[0.042]; place[0.042]; freestyle[0.042]; Skiing[0.042]; Skiing[0.042]; resort[0.042]; France[0.042]; France[0.042]; France[0.042]; Cup[0.041]; Cup[0.041]; overall[0.041]; overall[0.041]; combined[0.041]; world[0.041]; Freestyle[0.041]; competition[0.040]; competition[0.040]; Sebastien[0.040]; day[0.040]; season[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.046]; Canada[0.046]; Cup[0.044]; Cup[0.044]; Cup[0.044]; Blanc[0.042]; women[0.042]; second[0.042]; second[0.042]; season[0.041]; win[0.041]; Swiss[0.041]; seventh[0.041]; Evelyne[0.041]; World[0.040]; World[0.040]; World[0.040]; career[0.040]; career[0.040]; Tignes[0.040]; ninth[0.040]; victory[0.040]; wins[0.040]; beating[0.039]; \t\n\r [=============================>................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s17ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2963/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1266testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burmese ==> ENTITY: \u001b[32mBurma\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.094:0.094[0]; log p(e|m)= -1.228:-1.228[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnorthern[0.047]; troops[0.045]; troops[0.045]; Rangoon[0.043]; Rangoon[0.043]; Rangoon[0.043]; students[0.042]; students[0.042]; students[0.042]; Government[0.041]; said[0.041]; said[0.041]; said[0.041]; student[0.041]; staged[0.041]; protest[0.040]; protest[0.040]; protest[0.040]; sitting[0.040]; street[0.040]; university[0.040]; police[0.040]; police[0.040]; hit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Insein\u001b[39m ==> ENTITY: \u001b[32mInsein Township\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRangoon[0.049]; Rangoon[0.049]; Rangoon[0.049]; Rangoon[0.049]; northern[0.044]; junction[0.042]; junction[0.042]; junction[0.042]; protest[0.041]; protest[0.041]; university[0.040]; jail[0.040]; days[0.040]; near[0.040]; prison[0.040]; campuses[0.040]; University[0.039]; University[0.039]; suburban[0.039]; suburban[0.039]; protesting[0.039]; intersection[0.039]; capital[0.039]; major[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rangoon\u001b[39m ==> ENTITY: \u001b[32mYangon\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nuniversity[0.044]; Rangoon[0.044]; Rangoon[0.044]; Rangoon[0.044]; Rangoon[0.044]; University[0.044]; University[0.044]; Burmese[0.043]; troops[0.041]; troops[0.041]; student[0.041]; student[0.041]; engines[0.041]; students[0.041]; students[0.041]; students[0.041]; students[0.041]; students[0.041]; capital[0.041]; northern[0.041]; protest[0.040]; protest[0.040]; protest[0.040]; protest[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rangoon University\u001b[39m ==> ENTITY: \u001b[32mUniversity of Yangon\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nuniversity[0.046]; Rangoon[0.045]; Rangoon[0.045]; student[0.043]; students[0.043]; students[0.043]; students[0.043]; students[0.043]; students[0.043]; protest[0.042]; protest[0.042]; campuses[0.042]; major[0.041]; capital[0.040]; moved[0.040]; Saturday[0.040]; Insein[0.040]; Friday[0.040]; leaders[0.040]; protesting[0.040]; staged[0.040]; days[0.040]; second[0.039]; started[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rangoon\u001b[39m ==> ENTITY: \u001b[32mYangon\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nuniversity[0.045]; Rangoon[0.045]; Rangoon[0.045]; Rangoon[0.045]; University[0.044]; University[0.044]; northern[0.042]; major[0.042]; troops[0.042]; student[0.041]; engines[0.041]; students[0.041]; students[0.041]; students[0.041]; students[0.041]; students[0.041]; capital[0.041]; protest[0.041]; protest[0.041]; campuses[0.040]; Friday[0.039]; near[0.039]; moved[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rangoon\u001b[39m ==> ENTITY: \u001b[32mYangon\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nuniversity[0.045]; Rangoon[0.045]; Rangoon[0.045]; University[0.045]; Burmese[0.044]; troops[0.042]; troops[0.042]; student[0.042]; engines[0.042]; students[0.042]; students[0.042]; students[0.042]; northern[0.042]; protest[0.041]; protest[0.041]; protest[0.041]; Government[0.041]; near[0.040]; water[0.040]; water[0.040]; moved[0.040]; moved[0.040]; Saturday[0.040]; staged[0.039]; \t\n\r [=============================>................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s62ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2969/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1329testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wellington ==> ENTITY: \u001b[32mWellington\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.926:-0.926[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZealand[0.046]; Zealand[0.046]; Thursday[0.046]; Friday[0.046]; National[0.046]; Sunday[0.045]; Sunday[0.045]; afternoon[0.045]; said[0.044]; said[0.044]; meet[0.044]; meet[0.044]; expected[0.044]; says[0.044]; government[0.042]; Jim[0.041]; formed[0.041]; emerging[0.041]; coalition[0.041]; party[0.041]; Nz[0.041]; Nz[0.041]; Nz[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jim Bolger\u001b[39m ==> ENTITY: \u001b[32mJim Bolger\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.135:0.135[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNational[0.046]; Bolger[0.045]; Bolger[0.045]; Prime[0.045]; Minister[0.044]; government[0.043]; Wellington[0.042]; coalition[0.042]; Zealand[0.042]; Zealand[0.042]; Nz[0.042]; Nz[0.042]; Nz[0.042]; party[0.041]; said[0.040]; said[0.040]; Sunday[0.040]; Sunday[0.040]; Thursday[0.040]; Nats[0.040]; expected[0.039]; Friday[0.039]; emerging[0.039]; formed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bolger\u001b[39m ==> ENTITY: \u001b[32mJim Bolger\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.135:0.135[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNational[0.045]; Bolger[0.045]; Bolger[0.045]; Prime[0.044]; Jim[0.044]; Minister[0.044]; government[0.042]; Wellington[0.042]; coalition[0.042]; Zealand[0.042]; Zealand[0.042]; Nz[0.042]; Nz[0.042]; Nz[0.042]; party[0.041]; said[0.040]; said[0.040]; Sunday[0.040]; Sunday[0.040]; Thursday[0.040]; Nats[0.039]; expected[0.039]; Friday[0.039]; emerging[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NZ First\u001b[39m ==> ENTITY: \u001b[32mNew Zealand First\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNational[0.045]; party[0.045]; coalition[0.044]; Wellington[0.044]; Nz[0.043]; Nz[0.043]; Zealand[0.043]; Zealand[0.043]; Minister[0.042]; government[0.042]; Jim[0.042]; Prime[0.041]; said[0.041]; said[0.041]; emerging[0.041]; formed[0.041]; Thursday[0.041]; Nats[0.040]; talks[0.040]; nationalist[0.040]; Bolger[0.040]; Bolger[0.040]; Bolger[0.040]; says[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Zealand ==> ENTITY: \u001b[32mNew Zealand\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.088:0.088[0]; log p(e|m)= -0.238:-0.238[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWellington[0.047]; National[0.046]; Zealand[0.045]; Sunday[0.043]; Sunday[0.043]; Thursday[0.043]; government[0.042]; Minister[0.042]; says[0.042]; Friday[0.042]; said[0.041]; said[0.041]; expected[0.041]; formed[0.041]; Nz[0.041]; Nz[0.041]; Nz[0.041]; talks[0.041]; Jim[0.040]; meet[0.040]; meet[0.040]; party[0.040]; Prime[0.039]; afternoon[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = National ==> ENTITY: \u001b[32mNew Zealand National Party\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.137:0.137[0]; log p(e|m)= -1.877:-1.877[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWellington[0.049]; coalition[0.048]; Sunday[0.047]; Sunday[0.047]; party[0.047]; Jim[0.046]; Zealand[0.046]; Zealand[0.046]; Minister[0.046]; Friday[0.046]; government[0.046]; Prime[0.045]; expected[0.045]; Thursday[0.045]; formed[0.044]; said[0.044]; said[0.044]; says[0.044]; emerging[0.044]; meet[0.043]; meet[0.043]; nationalist[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = New Zealand First\u001b[39m ==> ENTITY: \u001b[32mNew Zealand First\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.115:0.115[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNational[0.045]; party[0.044]; coalition[0.044]; Wellington[0.044]; Nz[0.043]; Nz[0.043]; Nz[0.043]; Zealand[0.043]; Minister[0.042]; government[0.042]; Jim[0.042]; Prime[0.041]; said[0.041]; said[0.041]; emerging[0.041]; formed[0.041]; Thursday[0.041]; Nats[0.040]; talks[0.040]; nationalist[0.040]; Bolger[0.040]; Bolger[0.040]; Bolger[0.040]; says[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bolger\u001b[39m ==> ENTITY: \u001b[32mJim Bolger\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.135:0.135[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNational[0.045]; Bolger[0.045]; Bolger[0.045]; Prime[0.044]; Jim[0.044]; Minister[0.044]; government[0.042]; Wellington[0.042]; coalition[0.042]; Zealand[0.042]; Zealand[0.042]; Nz[0.042]; Nz[0.042]; Nz[0.042]; party[0.041]; said[0.040]; said[0.040]; Sunday[0.040]; Sunday[0.040]; Thursday[0.040]; Nats[0.039]; expected[0.039]; Friday[0.039]; emerging[0.039]; \t\n\r [=============================>................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s19ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2977/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1315testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.140:0.140[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.055]; Italian[0.047]; Italian[0.047]; Italian[0.047]; League[0.044]; League[0.044]; state[0.043]; Northern[0.042]; Friday[0.041]; Luigi[0.040]; appealed[0.040]; north[0.040]; campaign[0.040]; Mantua[0.039]; Mantua[0.039]; parliament[0.039]; said[0.039]; supporters[0.039]; march[0.039]; heartland[0.039]; invitation[0.039]; direct[0.039]; President[0.039]; President[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italia ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.129:0.129[0]; log p(e|m)= -0.642:-0.642[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItalians[0.046]; Italians[0.046]; Padania[0.045]; Venice[0.044]; Umberto[0.042]; national[0.042]; national[0.042]; September[0.042]; independence[0.042]; day[0.042]; April[0.041]; Republic[0.041]; autonomy[0.041]; heavy[0.041]; federalist[0.041]; march[0.041]; Bossi[0.041]; Po[0.040]; small[0.040]; Mantua[0.040]; percent[0.040]; federalism[0.040]; rulers[0.039]; badly[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italians ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.123:0.123[0]; log p(e|m)= -1.556:-1.556[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.055]; Italians[0.048]; small[0.043]; Italia[0.043]; national[0.041]; national[0.041]; Umberto[0.041]; said[0.041]; said[0.041]; said[0.041]; strong[0.041]; work[0.041]; outnumbered[0.041]; local[0.041]; jeered[0.041]; Bossi[0.040]; autonomy[0.040]; white[0.040]; campaigned[0.040]; September[0.040]; April[0.040]; independence[0.040]; security[0.040]; heavy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Scalfaro\u001b[39m ==> ENTITY: \u001b[32mOscar Luigi Scalfaro\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.046]; Italy[0.046]; Italian[0.044]; Italian[0.044]; Italian[0.044]; President[0.043]; President[0.043]; Luigi[0.042]; Italians[0.042]; Scalfaro[0.042]; Scalfaro[0.042]; Scalfaro[0.042]; parliament[0.042]; Northern[0.040]; appealed[0.040]; federal[0.040]; ceremony[0.040]; supporters[0.040]; supporters[0.040]; said[0.040]; said[0.040]; said[0.040]; campaigned[0.039]; symbolic[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italian ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.154:0.154[0]; log p(e|m)= -0.759:-0.759[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.055]; Italy[0.055]; Italian[0.051]; Italian[0.051]; Luigi[0.042]; campaign[0.041]; march[0.041]; autonomy[0.040]; symbolic[0.039]; local[0.039]; parliament[0.039]; Mantua[0.039]; Mantua[0.039]; Mantua[0.039]; Northern[0.039]; federal[0.039]; called[0.039]; breakaway[0.039]; said[0.039]; said[0.039]; separatist[0.039]; commemorating[0.039]; work[0.039]; strong[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Umberto Bossi\u001b[39m ==> ENTITY: \u001b[32mUmberto Bossi\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfederalist[0.045]; Padania[0.044]; national[0.044]; national[0.044]; Italia[0.044]; Italians[0.043]; Italians[0.043]; Scalfaro[0.042]; leader[0.042]; Venice[0.041]; League[0.041]; League[0.041]; September[0.041]; independence[0.041]; election[0.041]; said[0.041]; won[0.041]; group[0.040]; group[0.040]; Republic[0.040]; supporters[0.040]; general[0.040]; unity[0.040]; campaigned[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.170:0.170[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.054]; Italian[0.046]; Italian[0.046]; Italians[0.044]; Italians[0.044]; League[0.043]; League[0.043]; League[0.043]; state[0.042]; national[0.041]; Northern[0.041]; Mantua[0.041]; Mantua[0.041]; Mantua[0.041]; autonomy[0.040]; Luigi[0.040]; jeered[0.040]; local[0.039]; north[0.039]; campaign[0.039]; small[0.039]; parliament[0.039]; march[0.038]; heartland[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italian ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.759:-0.759[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.056]; Italian[0.052]; Italian[0.052]; Luigi[0.043]; campaign[0.042]; march[0.042]; symbolic[0.040]; parliament[0.040]; Mantua[0.040]; Mantua[0.040]; Northern[0.040]; federal[0.040]; called[0.040]; breakaway[0.040]; separatist[0.040]; work[0.040]; state[0.039]; President[0.039]; President[0.039]; Friday[0.039]; north[0.039]; League[0.039]; League[0.039]; supporters[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Oscar Luigi Scalfaro\u001b[39m ==> ENTITY: \u001b[32mOscar Luigi Scalfaro\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.047]; Italy[0.047]; Italian[0.045]; Italian[0.045]; Italian[0.045]; President[0.044]; President[0.044]; Scalfaro[0.043]; Scalfaro[0.043]; parliament[0.042]; Northern[0.041]; appealed[0.041]; federal[0.041]; supporters[0.040]; said[0.040]; symbolic[0.040]; campaign[0.040]; party[0.040]; federalism[0.039]; state[0.039]; strong[0.039]; heartland[0.039]; unity[0.039]; appeal[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italians ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.139:0.139[0]; log p(e|m)= -1.556:-1.556[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.054]; Italians[0.048]; Padania[0.044]; Venice[0.042]; Italia[0.042]; small[0.042]; Republic[0.042]; national[0.041]; national[0.041]; Umberto[0.041]; said[0.041]; said[0.041]; said[0.041]; strong[0.041]; outnumbered[0.041]; local[0.040]; jeered[0.040]; Bossi[0.040]; autonomy[0.040]; white[0.040]; campaigned[0.040]; September[0.040]; April[0.040]; independence[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Scalfaro\u001b[39m ==> ENTITY: \u001b[32mOscar Luigi Scalfaro\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.047]; Italian[0.045]; Italians[0.043]; Italians[0.043]; Scalfaro[0.043]; Scalfaro[0.043]; Umberto[0.043]; parliament[0.042]; won[0.042]; general[0.041]; April[0.041]; Italia[0.041]; election[0.041]; federal[0.041]; ceremony[0.041]; federalist[0.041]; supporters[0.041]; said[0.041]; said[0.041]; said[0.041]; campaigned[0.040]; campaign[0.040]; party[0.040]; federalism[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Scalfaro\u001b[39m ==> ENTITY: \u001b[32mOscar Luigi Scalfaro\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.046]; Italy[0.046]; Italian[0.044]; Italian[0.044]; Italian[0.044]; President[0.043]; President[0.043]; Luigi[0.042]; Italians[0.042]; Scalfaro[0.042]; Scalfaro[0.042]; Scalfaro[0.042]; parliament[0.042]; Northern[0.040]; appealed[0.040]; federal[0.040]; ceremony[0.040]; supporters[0.040]; said[0.040]; said[0.040]; campaigned[0.039]; symbolic[0.039]; campaign[0.039]; party[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italian ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.151:0.151[0]; log p(e|m)= -0.759:-0.759[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.055]; Italy[0.055]; Italian[0.051]; Italian[0.051]; Luigi[0.043]; campaign[0.041]; march[0.041]; symbolic[0.040]; parliament[0.039]; Mantua[0.039]; Mantua[0.039]; Northern[0.039]; federal[0.039]; called[0.039]; breakaway[0.039]; said[0.039]; separatist[0.039]; work[0.039]; strong[0.039]; state[0.039]; President[0.039]; President[0.039]; Friday[0.039]; north[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Northern League ==> ENTITY: \u001b[32mLega Nord\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.157:0.157[0]; log p(e|m)= -1.115:-1.115[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.049]; Italy[0.046]; Italy[0.046]; Italian[0.045]; Italian[0.045]; Italian[0.045]; Luigi[0.042]; party[0.042]; campaign[0.041]; supporters[0.041]; parliament[0.041]; Friday[0.041]; Scalfaro[0.041]; Scalfaro[0.041]; Scalfaro[0.041]; President[0.040]; President[0.040]; breakaway[0.040]; drop[0.039]; autonomy[0.039]; federal[0.039]; separatist[0.039]; state[0.039]; visited[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Venice ==> ENTITY: \u001b[32mVenice\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.083:0.083[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItalians[0.046]; Padania[0.044]; Umberto[0.044]; national[0.043]; national[0.043]; Italia[0.043]; Republic[0.042]; September[0.042]; won[0.042]; visit[0.041]; Po[0.041]; waved[0.041]; April[0.041]; River[0.041]; supporters[0.041]; votes[0.041]; culminating[0.041]; independence[0.041]; day[0.041]; said[0.041]; arrived[0.040]; outnumbered[0.040]; flags[0.040]; group[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAustrian\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustria\u001b[39m <---> \u001b[32mAustrian Empire\u001b[39m\t\nSCORES: global= 0.254:0.235[\u001b[31m0.019\u001b[39m]; local()= 0.099:0.107[\u001b[32m0.008\u001b[39m]; log p(e|m)= -0.406:-2.590[\u001b[31m2.184\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.046]; Italians[0.044]; Italians[0.044]; national[0.043]; national[0.043]; parliament[0.043]; federalist[0.042]; independence[0.042]; propose[0.042]; party[0.042]; autonomy[0.042]; federalism[0.042]; April[0.041]; said[0.041]; said[0.041]; said[0.041]; Mantua[0.041]; Mantua[0.041]; work[0.040]; agenda[0.040]; campaigned[0.040]; reform[0.040]; small[0.040]; leader[0.040]; \t\n\r [=============================>................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 7s856ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2997/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1262testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rio De Janeiro\u001b[39m ==> ENTITY: \u001b[32mRio de Janeiro\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.128:0.128[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.052]; Brazilian[0.049]; Rio[0.049]; state[0.048]; university[0.047]; university[0.047]; Globo[0.046]; Globo[0.046]; Janeiro[0.046]; students[0.045]; students[0.045]; students[0.045]; reais[0.045]; special[0.044]; disqualified[0.044]; operated[0.044]; exams[0.043]; like[0.043]; officials[0.043]; caught[0.043]; caught[0.043]; gave[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rio de Janeiro\u001b[39m ==> ENTITY: \u001b[32mRio de Janeiro\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.128:0.128[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.052]; Brazilian[0.049]; Rio[0.049]; state[0.048]; university[0.047]; university[0.047]; Globo[0.046]; Globo[0.046]; Janeiro[0.046]; students[0.045]; students[0.045]; students[0.045]; reais[0.045]; special[0.044]; disqualified[0.044]; operated[0.044]; exams[0.043]; like[0.043]; officials[0.043]; caught[0.043]; caught[0.043]; gave[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazilian ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.140:0.140[0]; log p(e|m)= -0.273:-0.273[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.052]; seven[0.045]; Rio[0.045]; Rio[0.045]; said[0.044]; said[0.044]; said[0.044]; officials[0.044]; Globo[0.044]; Globo[0.044]; discovered[0.044]; state[0.044]; gave[0.044]; like[0.043]; Janeiro[0.042]; Janeiro[0.042]; Seventy[0.042]; reais[0.042]; newspaper[0.042]; university[0.041]; university[0.041]; test[0.041]; Friday[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = O Globo\u001b[39m ==> ENTITY: \u001b[32mO Globo\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGlobo[0.046]; newspaper[0.045]; Brazilian[0.045]; university[0.044]; university[0.044]; Brazil[0.044]; students[0.042]; students[0.042]; students[0.042]; Rio[0.042]; Rio[0.042]; Janeiro[0.042]; Janeiro[0.042]; Friday[0.041]; exams[0.041]; officials[0.041]; said[0.040]; said[0.040]; said[0.040]; state[0.040]; telephone[0.040]; seven[0.040]; exam[0.039]; reais[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = O Globo\u001b[39m ==> ENTITY: \u001b[32mO Globo\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGlobo[0.046]; newspaper[0.045]; Brazilian[0.045]; university[0.044]; university[0.044]; Brazil[0.044]; students[0.042]; students[0.042]; students[0.042]; Rio[0.042]; Rio[0.042]; Janeiro[0.042]; Janeiro[0.042]; Friday[0.041]; exams[0.041]; officials[0.041]; said[0.040]; said[0.040]; said[0.040]; state[0.040]; telephone[0.040]; seven[0.040]; exam[0.039]; reais[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazil ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazilian[0.048]; Rio[0.046]; Rio[0.046]; Janeiro[0.044]; Janeiro[0.044]; Globo[0.044]; Globo[0.044]; state[0.043]; like[0.043]; seven[0.042]; reais[0.041]; gave[0.041]; special[0.040]; said[0.040]; said[0.040]; officials[0.040]; Seventy[0.040]; disqualified[0.040]; Friday[0.040]; watches[0.040]; watches[0.040]; watches[0.040]; watches[0.040]; test[0.039]; \t\n\r [=============================>................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 7s815ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3003/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1307testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = USDA\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Agriculture\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThursday[0.043]; Thursday[0.043]; slaughter[0.042]; slaughter[0.042]; slaughter[0.042]; week[0.042]; week[0.042]; week[0.042]; week[0.042]; market[0.042]; City[0.042]; contracted[0.042]; adequate[0.042]; cattle[0.042]; cattle[0.042]; Dodge[0.041]; Select[0.041]; Select[0.041]; head[0.041]; head[0.041]; confirmed[0.041]; confirmed[0.041]; confirmed[0.041]; demand[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dodge City ==> ENTITY: \u001b[32mDodge City, Kansas\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.149:0.149[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncattle[0.045]; cattle[0.045]; week[0.045]; week[0.045]; week[0.045]; week[0.045]; Thursday[0.045]; Thursday[0.045]; Kansas[0.045]; ago[0.044]; ago[0.044]; ago[0.044]; ago[0.044]; good[0.043]; roundup[0.043]; date[0.042]; date[0.042]; contracted[0.042]; lbs[0.042]; lbs[0.042]; steers[0.042]; adequate[0.041]; Trade[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = USDA\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Agriculture\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.095:0.095[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndate[0.044]; date[0.044]; Select[0.042]; Select[0.042]; Thursday[0.042]; Thursday[0.042]; slaughter[0.042]; slaughter[0.042]; slaughter[0.042]; week[0.042]; week[0.042]; week[0.042]; week[0.042]; City[0.042]; contracted[0.041]; adequate[0.041]; cattle[0.041]; cattle[0.041]; market[0.041]; light[0.041]; Dodge[0.041]; lbs[0.041]; lbs[0.041]; sales[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kansas ==> ENTITY: \u001b[32mKansas\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.093:0.093[0]; log p(e|m)= -0.384:-0.384[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDodge[0.049]; week[0.043]; week[0.043]; week[0.043]; week[0.043]; Thursday[0.043]; Thursday[0.043]; Select[0.043]; Select[0.043]; City[0.043]; sales[0.043]; contracted[0.041]; head[0.041]; head[0.041]; Choice[0.040]; date[0.040]; date[0.040]; Trade[0.040]; roundup[0.040]; Sales[0.040]; ago[0.040]; ago[0.040]; ago[0.040]; ago[0.040]; \t\n\r [=============================>................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 7s817ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3007/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1202testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barcelona ==> ENTITY: \u001b[32mFC Barcelona\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.182:0.182[0]; log p(e|m)= -1.709:-1.709[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.046]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; match[0.043]; Spanish[0.042]; Barcelona[0.042]; Barcelona[0.042]; Barcelona[0.042]; division[0.041]; win[0.041]; win[0.041]; win[0.041]; Real[0.041]; Real[0.041]; Real[0.041]; Real[0.041]; William[0.040]; Hill[0.040]; score[0.038]; Saturday[0.038]; betting[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Victor Sanchez\u001b[39m ==> ENTITY: \u001b[32mVíctor Sánchez\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGarcia[0.052]; Juan[0.052]; Enrique[0.051]; Rafael[0.051]; Roberto[0.051]; Miguel[0.051]; match[0.050]; goalscorer[0.050]; Guillermo[0.050]; Raul[0.050]; Luis[0.050]; Luis[0.050]; Luis[0.050]; Jose[0.050]; Fernando[0.049]; Fernando[0.049]; Fernando[0.049]; Carlos[0.049]; Carlos[0.049]; Davor[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barcelona ==> ENTITY: \u001b[32mFC Barcelona\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.226:0.226[0]; log p(e|m)= -1.709:-1.709[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.049]; Luis[0.049]; Luis[0.049]; Madrid[0.049]; Madrid[0.049]; Madrid[0.049]; Madrid[0.049]; Madrid[0.049]; Madrid[0.049]; Madrid[0.049]; Roberto[0.048]; Enrique[0.048]; Jose[0.047]; Carlos[0.047]; Carlos[0.047]; Juan[0.046]; Fernando[0.045]; Fernando[0.045]; Fernando[0.045]; Guillermo[0.045]; match[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Guillermo Amor\u001b[39m ==> ENTITY: \u001b[32mGuillermo Amor\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngoalscorer[0.042]; Miguel[0.042]; Davor[0.042]; Barcelona[0.042]; Barcelona[0.042]; Barcelona[0.042]; Barcelona[0.042]; Barcelona[0.042]; Barcelona[0.042]; Barcelona[0.042]; Carlos[0.042]; Carlos[0.042]; Nadal[0.042]; match[0.041]; Jose[0.041]; Fernando[0.041]; Fernando[0.041]; Fernando[0.041]; Luis[0.041]; Luis[0.041]; Luis[0.041]; Roberto[0.041]; Rafael[0.041]; Juan[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Madrid\u001b[39m ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.253:0.253[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.047]; Jose[0.047]; Carlos[0.047]; Madrid[0.046]; Madrid[0.046]; Madrid[0.046]; Madrid[0.046]; Madrid[0.046]; Madrid[0.046]; Raul[0.045]; Juan[0.045]; Barcelona[0.045]; Barcelona[0.045]; Barcelona[0.045]; Barcelona[0.045]; Barcelona[0.045]; Barcelona[0.045]; Barcelona[0.045]; Roberto[0.045]; match[0.045]; Fernando[0.044]; Guillermo[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Miguel Nadal\u001b[39m ==> ENTITY: \u001b[32mMiguel Ángel Nadal\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.044]; Luis[0.044]; Luis[0.044]; Fernando[0.043]; Fernando[0.043]; Fernando[0.043]; Jose[0.043]; Juan[0.043]; Rafael[0.043]; Raul[0.043]; Enrique[0.043]; Carlos[0.042]; Carlos[0.042]; Hierro[0.042]; Roberto[0.042]; Guillermo[0.042]; Gonzalez[0.041]; Garcia[0.040]; Amavisca[0.040]; Sanchis[0.039]; Ferrer[0.039]; Sanz[0.039]; Roger[0.039]; Redondo[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barcelona ==> ENTITY: \u001b[32mFC Barcelona\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.182:0.182[0]; log p(e|m)= -1.709:-1.709[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.046]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; match[0.043]; Spanish[0.042]; Barcelona[0.042]; Barcelona[0.042]; Barcelona[0.042]; division[0.041]; win[0.041]; win[0.041]; win[0.041]; Real[0.041]; Real[0.041]; Real[0.041]; Real[0.041]; William[0.040]; Hill[0.040]; score[0.038]; Saturday[0.038]; betting[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Raul Gonzalez ==> ENTITY: \u001b[32mRaúl (footballer)\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.280:0.280[0]; log p(e|m)= -0.501:-0.501[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEnrique[0.057]; Roberto[0.056]; Madrid[0.056]; Madrid[0.056]; Madrid[0.056]; Madrid[0.056]; Madrid[0.056]; Madrid[0.056]; Madrid[0.056]; Rafael[0.056]; Jose[0.056]; Fernando[0.055]; Fernando[0.055]; Fernando[0.055]; Miguel[0.055]; Luis[0.054]; Luis[0.054]; Luis[0.054]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Madrid\u001b[39m ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.049]; Madrid[0.046]; Madrid[0.046]; Madrid[0.046]; Madrid[0.046]; Barcelona[0.045]; Barcelona[0.045]; Barcelona[0.045]; Barcelona[0.045]; match[0.045]; Real[0.044]; Real[0.044]; Real[0.044]; Spanish[0.043]; division[0.043]; win[0.043]; win[0.043]; win[0.043]; score[0.040]; Betting[0.040]; Saturday[0.039]; betting[0.039]; Showcase[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barcelona ==> ENTITY: \u001b[32mFC Barcelona\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.224:0.224[0]; log p(e|m)= -1.709:-1.709[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.053]; Luis[0.053]; Luis[0.053]; Roberto[0.051]; Enrique[0.051]; Jose[0.051]; Carlos[0.051]; Madrid[0.050]; Madrid[0.050]; Madrid[0.050]; Madrid[0.050]; Madrid[0.050]; Madrid[0.050]; Madrid[0.050]; Juan[0.049]; Fernando[0.049]; Fernando[0.049]; Fernando[0.049]; Guillermo[0.048]; match[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Laurent Blanc\u001b[39m ==> ENTITY: \u001b[32mLaurent Blanc\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCarlos[0.043]; Carlos[0.043]; Roberto[0.043]; Fernando[0.043]; Fernando[0.043]; Fernando[0.043]; Enrique[0.043]; Rafael[0.042]; Nadal[0.042]; Jose[0.042]; Luis[0.042]; Luis[0.042]; Luis[0.042]; Juan[0.042]; Miguel[0.042]; Figo[0.041]; Raul[0.041]; Victor[0.041]; Redondo[0.041]; Hierro[0.040]; Roger[0.040]; Secretario[0.040]; Guillermo[0.040]; Ivan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Luis Enrique\u001b[39m ==> ENTITY: \u001b[32mLuis Enrique Martínez García\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.044]; Luis[0.044]; Roberto[0.044]; Sanz[0.043]; Jose[0.043]; Miguel[0.042]; Juan[0.042]; Carlos[0.042]; Carlos[0.042]; match[0.042]; Fernando[0.042]; Fernando[0.042]; Fernando[0.042]; Rafael[0.042]; Gonzalez[0.041]; Guillermo[0.041]; Garcia[0.041]; Raul[0.041]; Manolo[0.041]; Sanchez[0.040]; Victor[0.039]; Madrid[0.039]; Madrid[0.039]; Davor[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fernando Sanz\u001b[39m ==> ENTITY: \u001b[32mFernando Sanz\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEnrique[0.044]; Fernando[0.044]; Fernando[0.044]; Miguel[0.043]; Luis[0.043]; Luis[0.043]; Luis[0.043]; Roberto[0.042]; Jose[0.042]; Madrid[0.041]; Madrid[0.041]; Raul[0.041]; Carlos[0.041]; Carlos[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Davor[0.041]; Nadal[0.040]; Amavisca[0.040]; Manolo[0.040]; Guillermo[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Davor Suker\u001b[39m ==> ENTITY: \u001b[32mDavor Šuker\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.046]; goalscorer[0.042]; Rafael[0.042]; Carlos[0.042]; Carlos[0.042]; Raul[0.042]; Madrid[0.041]; Madrid[0.041]; Madrid[0.041]; Madrid[0.041]; Madrid[0.041]; Madrid[0.041]; Madrid[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; half[0.041]; Fernando[0.041]; Fernando[0.041]; Fernando[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barcelona ==> ENTITY: \u001b[32mFC Barcelona\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.222:0.222[0]; log p(e|m)= -1.709:-1.709[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.045]; Roberto[0.044]; Jose[0.043]; Carlos[0.043]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Juan[0.042]; Fernando[0.041]; Fernando[0.041]; Guillermo[0.041]; match[0.041]; Barcelona[0.040]; Barcelona[0.040]; Barcelona[0.040]; Barcelona[0.040]; Barcelona[0.040]; Barcelona[0.040]; Raul[0.040]; Gonzalez[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barcelona ==> ENTITY: \u001b[32mFC Barcelona\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.181:0.181[0]; log p(e|m)= -1.709:-1.709[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.046]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; match[0.043]; Spanish[0.042]; Barcelona[0.042]; Barcelona[0.042]; Barcelona[0.042]; division[0.041]; win[0.040]; win[0.040]; win[0.040]; Real[0.040]; Real[0.040]; Real[0.040]; Real[0.040]; Real[0.040]; time[0.040]; time[0.040]; William[0.040]; Hill[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jose Amavisca\u001b[39m ==> ENTITY: \u001b[32mJosé Emilio Amavisca\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.243:0.243[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEnrique[0.045]; Fernando[0.044]; Fernando[0.044]; Fernando[0.044]; Barcelona[0.044]; Barcelona[0.044]; Barcelona[0.044]; Barcelona[0.044]; Barcelona[0.044]; Barcelona[0.044]; Barcelona[0.044]; Madrid[0.043]; Madrid[0.043]; Madrid[0.043]; Madrid[0.043]; Madrid[0.043]; Madrid[0.043]; Madrid[0.043]; Hierro[0.043]; Luis[0.043]; Luis[0.043]; Luis[0.043]; Roberto[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barcelona ==> ENTITY: \u001b[32mFC Barcelona\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.182:0.182[0]; log p(e|m)= -1.709:-1.709[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.046]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; match[0.043]; Spanish[0.042]; Barcelona[0.042]; Barcelona[0.042]; Barcelona[0.042]; division[0.041]; win[0.041]; win[0.041]; win[0.041]; Real[0.041]; Real[0.041]; Real[0.041]; Real[0.041]; William[0.040]; Hill[0.040]; score[0.038]; Saturday[0.038]; betting[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barcelona ==> ENTITY: \u001b[32mFC Barcelona\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.225:0.225[0]; log p(e|m)= -1.709:-1.709[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.050]; Luis[0.050]; Luis[0.050]; Roberto[0.049]; Enrique[0.049]; Jose[0.048]; Carlos[0.048]; Carlos[0.048]; Madrid[0.047]; Madrid[0.047]; Madrid[0.047]; Madrid[0.047]; Madrid[0.047]; Madrid[0.047]; Madrid[0.047]; Juan[0.047]; Fernando[0.046]; Fernando[0.046]; Fernando[0.046]; Guillermo[0.046]; match[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Roger Garcia\u001b[39m ==> ENTITY: \u001b[32mRoger García Junyent\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFernando[0.043]; Fernando[0.043]; Fernando[0.043]; Luis[0.043]; Luis[0.043]; Luis[0.043]; Miguel[0.043]; Raul[0.042]; Juan[0.042]; Roberto[0.042]; Rafael[0.041]; Jose[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Carlos[0.041]; Carlos[0.041]; Gonzalez[0.041]; Davor[0.041]; match[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Madrid ==> ENTITY: \u001b[32mMadrid\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.166:0.166[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMadrid[0.048]; Madrid[0.048]; Madrid[0.048]; Madrid[0.048]; Soccer[0.045]; Spanish[0.044]; match[0.043]; Barcelona[0.042]; Barcelona[0.042]; Barcelona[0.042]; Barcelona[0.042]; Real[0.040]; Real[0.040]; Real[0.040]; Real[0.040]; win[0.040]; win[0.040]; win[0.040]; Draw[0.039]; division[0.039]; Saturday[0.039]; score[0.038]; William[0.037]; betting[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Juan Pizzi\u001b[39m ==> ENTITY: \u001b[32mJuan Antonio Pizzi\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.228:0.228[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFernando[0.047]; Fernando[0.047]; Fernando[0.047]; Jose[0.047]; Luis[0.046]; Luis[0.046]; Luis[0.046]; Enrique[0.046]; Roberto[0.045]; Barcelona[0.045]; Barcelona[0.045]; Barcelona[0.045]; Barcelona[0.045]; Barcelona[0.045]; Barcelona[0.045]; Barcelona[0.045]; Raul[0.045]; Figo[0.045]; Miguel[0.045]; Hierro[0.044]; Carlos[0.044]; Carlos[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Madrid\u001b[39m ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.256:0.256[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.044]; Luis[0.044]; Luis[0.044]; Jose[0.044]; Carlos[0.044]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; Enrique[0.043]; Raul[0.043]; Juan[0.043]; Barcelona[0.043]; Barcelona[0.043]; Barcelona[0.043]; Barcelona[0.043]; Barcelona[0.043]; Barcelona[0.043]; Barcelona[0.043]; Roberto[0.043]; match[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Madrid\u001b[39m ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.251:0.251[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.051]; Jose[0.051]; Madrid[0.051]; Madrid[0.051]; Madrid[0.051]; Madrid[0.051]; Madrid[0.051]; Madrid[0.051]; Raul[0.050]; Juan[0.050]; Barcelona[0.049]; Barcelona[0.049]; Barcelona[0.049]; Barcelona[0.049]; Barcelona[0.049]; Barcelona[0.049]; Barcelona[0.049]; match[0.049]; Fernando[0.049]; Guillermo[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Carlos Secretario\u001b[39m ==> ENTITY: \u001b[32mCarlos Secretário\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCarlos[0.044]; Enrique[0.043]; Fernando[0.042]; Fernando[0.042]; Fernando[0.042]; Miguel[0.042]; goalscorer[0.042]; Luis[0.042]; Luis[0.042]; Luis[0.042]; Juan[0.042]; Jose[0.042]; Madrid[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Roberto[0.041]; Davor[0.041]; Rafael[0.041]; match[0.041]; Figo[0.041]; Nadal[0.041]; Gonzalez[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Madrid\u001b[39m ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.049]; Madrid[0.046]; Madrid[0.046]; Madrid[0.046]; Madrid[0.046]; Barcelona[0.045]; Barcelona[0.045]; Barcelona[0.045]; Barcelona[0.045]; match[0.045]; Real[0.044]; Real[0.044]; Real[0.044]; Spanish[0.043]; division[0.043]; win[0.043]; win[0.043]; win[0.043]; score[0.040]; Betting[0.040]; Saturday[0.039]; betting[0.039]; Showcase[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ronaldo ==> ENTITY: \u001b[32mRonaldo\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.233:0.233[0]; log p(e|m)= -0.109:-0.109[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRoberto[0.047]; Rafael[0.047]; Enrique[0.046]; Raul[0.046]; goalscorer[0.046]; Fernando[0.046]; Fernando[0.046]; Fernando[0.046]; Carlos[0.046]; Carlos[0.046]; Jose[0.045]; Luis[0.045]; Luis[0.045]; Luis[0.045]; Figo[0.045]; Real[0.045]; Real[0.045]; Real[0.045]; Real[0.045]; Real[0.045]; Real[0.045]; Real[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sergi Barjuan\u001b[39m ==> ENTITY: \u001b[32mSergi Barjuán\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.206:0.206[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.044]; Luis[0.044]; Luis[0.044]; Gonzalez[0.042]; Roberto[0.042]; Miguel[0.042]; Barcelona[0.042]; Barcelona[0.042]; Barcelona[0.042]; Fernando[0.042]; Fernando[0.042]; Fernando[0.042]; Madrid[0.042]; Jose[0.042]; Enrique[0.041]; Carlos[0.041]; Carlos[0.041]; Rafael[0.041]; Sanz[0.041]; Laurent[0.041]; Real[0.041]; Garcia[0.041]; match[0.041]; Sanchis[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Madrid\u001b[39m ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.252:0.252[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.045]; Jose[0.045]; Carlos[0.045]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; Raul[0.043]; Juan[0.043]; Barcelona[0.043]; Barcelona[0.043]; Barcelona[0.043]; Barcelona[0.043]; Barcelona[0.043]; Barcelona[0.043]; Barcelona[0.043]; Roberto[0.043]; match[0.043]; Fernando[0.042]; Fernando[0.042]; Guillermo[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gheorghe Popescu\u001b[39m ==> ENTITY: \u001b[32mGheorghe Popescu\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.241:0.241[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRaul[0.047]; Roberto[0.047]; Enrique[0.046]; Hierro[0.046]; Carlos[0.046]; Carlos[0.046]; Rafael[0.045]; Luis[0.045]; Luis[0.045]; Luis[0.045]; Fernando[0.045]; Fernando[0.045]; Fernando[0.045]; Jose[0.045]; Barcelona[0.045]; Barcelona[0.045]; Barcelona[0.045]; Barcelona[0.045]; Barcelona[0.045]; Barcelona[0.045]; Barcelona[0.045]; Miguel[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Roberto Carlos ==> ENTITY: \u001b[32mRoberto Carlos (footballer)\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.238:0.238[0]; log p(e|m)= -0.057:-0.057[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCarlos[0.053]; Enrique[0.051]; Luis[0.050]; Luis[0.050]; Luis[0.050]; Jose[0.050]; Miguel[0.050]; Juan[0.049]; Fernando[0.049]; Fernando[0.049]; Fernando[0.049]; Rafael[0.048]; Guillermo[0.047]; Raul[0.046]; Sanz[0.045]; Manolo[0.045]; Victor[0.044]; Amor[0.043]; Ivan[0.043]; match[0.043]; Giovanni[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Madrid\u001b[39m ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.256:0.256[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.043]; Luis[0.043]; Luis[0.043]; Jose[0.042]; Carlos[0.042]; Carlos[0.042]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Enrique[0.041]; Rafael[0.041]; Raul[0.041]; Juan[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Roberto[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Madrid\u001b[39m ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.252:0.252[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.045]; Jose[0.045]; Carlos[0.045]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; Raul[0.043]; Juan[0.043]; Barcelona[0.043]; Barcelona[0.043]; Barcelona[0.043]; Barcelona[0.043]; Barcelona[0.043]; Barcelona[0.043]; Barcelona[0.043]; Roberto[0.043]; match[0.043]; Fernando[0.042]; Fernando[0.042]; Guillermo[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Madrid\u001b[39m ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.047]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; Madrid[0.044]; Barcelona[0.043]; Barcelona[0.043]; Barcelona[0.043]; Barcelona[0.043]; match[0.043]; Real[0.042]; Real[0.042]; Real[0.042]; Spanish[0.041]; division[0.041]; win[0.041]; win[0.041]; win[0.041]; time[0.039]; time[0.039]; Double[0.039]; score[0.038]; Betting[0.038]; Saturday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fernando Redondo\u001b[39m ==> ENTITY: \u001b[32mFernando Redondo\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.221:0.221[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFernando[0.044]; Fernando[0.044]; Roberto[0.043]; Raul[0.042]; Jose[0.042]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Sanz[0.042]; Miguel[0.042]; Enrique[0.041]; Carlos[0.041]; Carlos[0.041]; Rafael[0.041]; Juan[0.041]; Victor[0.041]; Luis[0.041]; Luis[0.041]; Luis[0.041]; match[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Madrid\u001b[39m ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.047]; Madrid[0.045]; Madrid[0.045]; Madrid[0.045]; Madrid[0.045]; Barcelona[0.043]; Barcelona[0.043]; Barcelona[0.043]; Barcelona[0.043]; match[0.043]; Real[0.043]; Real[0.043]; Real[0.043]; Spanish[0.041]; division[0.041]; win[0.041]; win[0.041]; win[0.041]; Showcase[0.040]; Saturday[0.039]; score[0.038]; Betting[0.038]; betting[0.038]; Draw[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = William Hill ==> ENTITY: \u001b[32mWilliam Hill (bookmaker)\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.076:0.076[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbetting[0.053]; Betting[0.052]; match[0.049]; Saturday[0.049]; Showcase[0.048]; score[0.048]; win[0.048]; win[0.048]; win[0.048]; Soccer[0.045]; division[0.045]; Spanish[0.044]; Draw[0.043]; Real[0.042]; Real[0.042]; Real[0.042]; Real[0.042]; Correct[0.042]; Barcelona[0.042]; Barcelona[0.042]; Barcelona[0.042]; Barcelona[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barcelona ==> ENTITY: \u001b[32mFC Barcelona\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.225:0.225[0]; log p(e|m)= -1.709:-1.709[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.050]; Luis[0.050]; Luis[0.050]; Roberto[0.049]; Enrique[0.049]; Jose[0.048]; Carlos[0.048]; Carlos[0.048]; Madrid[0.047]; Madrid[0.047]; Madrid[0.047]; Madrid[0.047]; Madrid[0.047]; Madrid[0.047]; Madrid[0.047]; Juan[0.047]; Fernando[0.046]; Fernando[0.046]; Fernando[0.046]; Guillermo[0.046]; match[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rafael Alkorta\u001b[39m ==> ENTITY: \u001b[32mRafael Alkorta\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMiguel[0.043]; Carlos[0.043]; Carlos[0.043]; Luis[0.043]; Luis[0.043]; Luis[0.043]; Enrique[0.043]; Raul[0.043]; Fernando[0.043]; Fernando[0.043]; Fernando[0.043]; Jose[0.042]; Juan[0.042]; Madrid[0.042]; Hierro[0.041]; Guillermo[0.041]; match[0.041]; Barcelona[0.041]; Roberto[0.040]; Gonzalez[0.040]; Davor[0.040]; Nadal[0.040]; Garcia[0.039]; Real[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Luis Figo\u001b[39m ==> ENTITY: \u001b[32mLuís Figo\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEnrique[0.043]; Roberto[0.043]; Rafael[0.043]; Jose[0.043]; Carlos[0.042]; Carlos[0.042]; Raul[0.042]; Luis[0.042]; Luis[0.042]; match[0.042]; Fernando[0.042]; Fernando[0.042]; Fernando[0.042]; Juan[0.041]; Miguel[0.041]; Ronaldo[0.041]; Guillermo[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Albert Ferrer\u001b[39m ==> ENTITY: \u001b[32mAlbert Ferrer\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFernando[0.043]; Fernando[0.043]; Fernando[0.043]; Barcelona[0.043]; Luis[0.043]; Luis[0.043]; Luis[0.043]; Hierro[0.042]; Miguel[0.042]; Rafael[0.042]; Davor[0.042]; Jose[0.042]; Raul[0.042]; Enrique[0.042]; Carlos[0.042]; Carlos[0.042]; Nadal[0.042]; Roberto[0.041]; Juan[0.041]; Redondo[0.040]; Garcia[0.040]; Gonzalez[0.040]; Blanc[0.040]; Guillermo[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barcelona ==> ENTITY: \u001b[32mFC Barcelona\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.224:0.224[0]; log p(e|m)= -1.709:-1.709[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.055]; Luis[0.055]; Luis[0.055]; Roberto[0.054]; Enrique[0.054]; Jose[0.053]; Carlos[0.053]; Madrid[0.052]; Madrid[0.052]; Madrid[0.052]; Madrid[0.052]; Madrid[0.052]; Madrid[0.052]; Madrid[0.052]; Juan[0.051]; Fernando[0.051]; Fernando[0.051]; Guillermo[0.051]; match[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Madrid\u001b[39m ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.252:0.252[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.049]; Jose[0.049]; Madrid[0.049]; Madrid[0.049]; Madrid[0.049]; Madrid[0.049]; Madrid[0.049]; Madrid[0.049]; Raul[0.047]; Juan[0.047]; Barcelona[0.047]; Barcelona[0.047]; Barcelona[0.047]; Barcelona[0.047]; Barcelona[0.047]; Barcelona[0.047]; Barcelona[0.047]; Barcelona[0.047]; match[0.047]; Fernando[0.046]; Guillermo[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fernando Hierro\u001b[39m ==> ENTITY: \u001b[32mFernando Hierro\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEnrique[0.046]; Rafael[0.044]; Miguel[0.044]; Fernando[0.044]; Fernando[0.044]; Roberto[0.043]; Luis[0.043]; Luis[0.043]; Luis[0.043]; Jose[0.043]; Carlos[0.043]; Carlos[0.043]; Juan[0.043]; Raul[0.042]; Guillermo[0.041]; Victor[0.039]; Davor[0.039]; Madrid[0.039]; Madrid[0.039]; Madrid[0.039]; Madrid[0.039]; goalscorer[0.039]; Nadal[0.039]; Roger[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Luis Milla\u001b[39m ==> ENTITY: \u001b[32mLuis Milla\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.230:0.230[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEnrique[0.044]; match[0.043]; Raul[0.043]; Miguel[0.042]; Jose[0.042]; Luis[0.042]; Luis[0.042]; Hierro[0.042]; Rafael[0.042]; Roberto[0.042]; Figo[0.042]; Juan[0.042]; Madrid[0.042]; Madrid[0.042]; Fernando[0.041]; Fernando[0.041]; Fernando[0.041]; Guillermo[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barcelona ==> ENTITY: \u001b[32mFC Barcelona\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.225:0.225[0]; log p(e|m)= -1.709:-1.709[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.048]; Luis[0.048]; Luis[0.048]; Roberto[0.047]; Enrique[0.047]; Jose[0.046]; Carlos[0.046]; Carlos[0.046]; Madrid[0.045]; Madrid[0.045]; Madrid[0.045]; Madrid[0.045]; Madrid[0.045]; Madrid[0.045]; Madrid[0.045]; Rafael[0.045]; Juan[0.044]; Fernando[0.044]; Fernando[0.044]; Fernando[0.044]; Guillermo[0.044]; match[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spanish ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.137:0.137[0]; log p(e|m)= -1.241:-1.241[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.045]; match[0.045]; Madrid[0.043]; Madrid[0.043]; Madrid[0.043]; Madrid[0.043]; Madrid[0.043]; win[0.042]; win[0.042]; win[0.042]; Barcelona[0.042]; Barcelona[0.042]; Barcelona[0.042]; Barcelona[0.042]; division[0.042]; Real[0.041]; Real[0.041]; Real[0.041]; Real[0.041]; Saturday[0.039]; Hill[0.039]; score[0.039]; William[0.038]; betting[0.038]; \t\n\r [==============================>...............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 7s502ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3054/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1368testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.057:0.057[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; League[0.045]; Sheffield[0.044]; Southampton[0.043]; Middlesbrough[0.043]; Coventry[0.043]; Blackburn[0.043]; Sunderland[0.043]; Soccer[0.043]; Leeds[0.043]; Kanchelskis[0.042]; Liverpool[0.042]; Tottenham[0.042]; Arsenal[0.041]; Chelsea[0.041]; Everton[0.041]; Adams[0.040]; Leicester[0.040]; Wednesday[0.040]; Aston[0.040]; Derby[0.040]; Saturday[0.039]; Premier[0.039]; Villa[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sunderland ==> ENTITY: \u001b[32mSunderland A.F.C.\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.183:0.183[0]; log p(e|m)= -0.396:-0.396[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthampton[0.045]; Coventry[0.045]; Middlesbrough[0.045]; Sheffield[0.045]; Blackburn[0.044]; Ekoku[0.044]; Leeds[0.044]; matches[0.043]; Everton[0.043]; Leicester[0.042]; Kanchelskis[0.042]; Liverpool[0.042]; Tottenham[0.041]; Whelan[0.041]; Arsenal[0.041]; Chelsea[0.040]; Sutton[0.040]; Derby[0.039]; Sheringham[0.039]; London[0.039]; Adams[0.039]; Saturday[0.039]; Sturridge[0.039]; Vialli[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Blackburn ==> ENTITY: \u001b[32mBlackburn Rovers F.C.\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.183:0.183[0]; log p(e|m)= -2.733:-2.733[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.045]; Sheffield[0.044]; Sunderland[0.044]; Southampton[0.044]; matches[0.043]; Middlesbrough[0.043]; Leeds[0.043]; Ekoku[0.043]; Coventry[0.043]; Soccer[0.042]; Derby[0.042]; Leicester[0.041]; Kanchelskis[0.041]; Liverpool[0.041]; English[0.041]; English[0.041]; Everton[0.040]; Aston[0.040]; Tottenham[0.040]; Villa[0.040]; Sutton[0.040]; Arsenal[0.039]; Premier[0.039]; Vialli[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Melville ==> ENTITY: \u001b[32mAndy Melville\u001b[39m\t\nSCORES: global= 0.239:0.239[0]; local()= 0.142:0.142[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWednesday[0.048]; Saturday[0.047]; Leeds[0.045]; Coventry[0.045]; Southampton[0.043]; Middlesbrough[0.043]; Ekoku[0.042]; Blackburn[0.042]; Sunderland[0.042]; Leicester[0.042]; Sheffield[0.042]; Kanchelskis[0.041]; Everton[0.041]; Adams[0.041]; Derby[0.040]; Tottenham[0.040]; matches[0.040]; Sutton[0.040]; premier[0.040]; Sheringham[0.039]; Marshall[0.039]; Aston[0.039]; Townsend[0.039]; Arsenal[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Branch ==> ENTITY: \u001b[32mMichael Branch\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.149:0.149[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.049]; Southampton[0.047]; Coventry[0.047]; Sheffield[0.046]; Blackburn[0.046]; Soccer[0.045]; Middlesbrough[0.045]; Sunderland[0.044]; Leeds[0.044]; Adams[0.044]; Liverpool[0.044]; Marshall[0.042]; Townsend[0.042]; Everton[0.042]; Premier[0.042]; Chelsea[0.042]; Saturday[0.042]; Tottenham[0.041]; matches[0.041]; Leicester[0.041]; Derby[0.041]; Ekoku[0.041]; Kanchelskis[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Powell ==> ENTITY: \u001b[32mChris Powell\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.149:0.149[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEkoku[0.045]; Soccer[0.044]; League[0.044]; Leeds[0.043]; Coventry[0.043]; Leicester[0.043]; Southampton[0.043]; matches[0.043]; Middlesbrough[0.043]; Sunderland[0.043]; Sheffield[0.042]; Saturday[0.042]; Kanchelskis[0.042]; Blackburn[0.041]; Wednesday[0.041]; Derby[0.040]; Sutton[0.040]; Vialli[0.040]; Aston[0.040]; Townsend[0.040]; Liverpool[0.040]; Sheringham[0.039]; Adams[0.039]; Tottenham[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Southampton ==> ENTITY: \u001b[32mSouthampton F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.154:0.154[0]; log p(e|m)= -1.149:-1.149[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCoventry[0.048]; Sunderland[0.045]; Liverpool[0.045]; Middlesbrough[0.044]; Chelsea[0.043]; Blackburn[0.042]; Sheringham[0.042]; League[0.042]; Melville[0.042]; Sheffield[0.042]; Leeds[0.041]; Aston[0.041]; Tottenham[0.041]; Villa[0.041]; Arsenal[0.041]; Everton[0.041]; Ekoku[0.040]; Kanchelskis[0.040]; matches[0.040]; Vialli[0.040]; Soccer[0.040]; Premier[0.039]; Zola[0.039]; Leicester[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vieira ==> ENTITY: \u001b[32mPatrick Vieira\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.518:-0.518[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.047]; Soccer[0.046]; Arsenal[0.045]; Chelsea[0.043]; matches[0.043]; Southampton[0.043]; Ekoku[0.042]; Tottenham[0.042]; Vialli[0.041]; Leeds[0.041]; Kanchelskis[0.041]; Villa[0.041]; Blackburn[0.041]; Coventry[0.041]; Whelan[0.041]; Middlesbrough[0.041]; Liverpool[0.041]; Wednesday[0.040]; Everton[0.040]; Leicester[0.040]; Saturday[0.040]; English[0.040]; English[0.040]; Sunderland[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sheringham ==> ENTITY: \u001b[32mTeddy Sheringham\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.136:0.136[0]; log p(e|m)= -1.839:-1.839[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.047]; Ekoku[0.045]; matches[0.043]; Southampton[0.043]; Chelsea[0.043]; Premier[0.043]; Kanchelskis[0.042]; Sheffield[0.042]; Blackburn[0.042]; Coventry[0.042]; Sunderland[0.042]; Tottenham[0.041]; Middlesbrough[0.041]; Melville[0.041]; Aston[0.041]; Leeds[0.041]; Soccer[0.041]; Liverpool[0.040]; Wimbledon[0.040]; Saturday[0.040]; Arsenal[0.040]; London[0.040]; Vialli[0.040]; Leicester[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWhittingham\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPeter Whittingham\u001b[39m <---> \u001b[32mGuy Whittingham\u001b[39m\t\nSCORES: global= 0.272:0.257[\u001b[31m0.015\u001b[39m]; local()= 0.168:0.153[\u001b[31m0.015\u001b[39m]; log p(e|m)= -1.537:-2.847[\u001b[31m1.310\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEkoku[0.045]; Southampton[0.045]; League[0.044]; Coventry[0.044]; Sunderland[0.043]; Sheffield[0.043]; Leeds[0.043]; Leicester[0.043]; Kanchelskis[0.043]; Middlesbrough[0.042]; Blackburn[0.042]; matches[0.042]; Sheringham[0.041]; Wednesday[0.041]; Aston[0.041]; Soccer[0.040]; Tottenham[0.040]; Saturday[0.040]; Everton[0.040]; English[0.040]; English[0.040]; Chelsea[0.040]; Derby[0.039]; Liverpool[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Coventry ==> ENTITY: \u001b[32mCoventry City F.C.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.184:0.184[0]; log p(e|m)= -3.124:-3.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeeds[0.045]; Blackburn[0.045]; League[0.045]; Southampton[0.045]; Leicester[0.044]; Sheffield[0.044]; matches[0.043]; Middlesbrough[0.043]; Sunderland[0.043]; Ekoku[0.042]; Aston[0.041]; Kanchelskis[0.041]; Soccer[0.041]; Tottenham[0.040]; Sutton[0.040]; London[0.040]; Saturday[0.040]; Liverpool[0.040]; Chelsea[0.040]; Everton[0.040]; Adams[0.039]; Wednesday[0.039]; Whelan[0.039]; Premier[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zola ==> ENTITY: \u001b[32mGianfranco Zola\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.156:0.156[0]; log p(e|m)= -1.630:-1.630[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; Ekoku[0.044]; Soccer[0.044]; Kanchelskis[0.043]; matches[0.043]; Vialli[0.043]; Southampton[0.043]; Coventry[0.043]; Blackburn[0.042]; Sunderland[0.042]; Sheffield[0.042]; Middlesbrough[0.042]; Chelsea[0.042]; Saturday[0.041]; Leeds[0.041]; Leicester[0.041]; Wednesday[0.041]; Everton[0.040]; Tottenham[0.040]; Liverpool[0.040]; Villa[0.040]; Whelan[0.040]; English[0.040]; English[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wimbledon ==> ENTITY: \u001b[32mWimbledon F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.200:0.200[0]; log p(e|m)= -1.890:-1.890[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.046]; Southampton[0.045]; Sheffield[0.045]; Coventry[0.045]; Sunderland[0.044]; Middlesbrough[0.043]; Blackburn[0.043]; Leeds[0.043]; Ekoku[0.043]; Kanchelskis[0.042]; Liverpool[0.042]; Leicester[0.042]; Tottenham[0.041]; Aston[0.041]; Derby[0.041]; Chelsea[0.040]; Everton[0.040]; Sutton[0.040]; English[0.040]; Vialli[0.039]; Adams[0.039]; Arsenal[0.039]; Saturday[0.039]; Wednesday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Derby ==> ENTITY: \u001b[32mDerby County F.C.\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.179:0.179[0]; log p(e|m)= -3.689:-3.689[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; Southampton[0.045]; Sheffield[0.044]; Coventry[0.044]; Blackburn[0.044]; matches[0.043]; Middlesbrough[0.043]; Sunderland[0.043]; Ekoku[0.043]; Leeds[0.043]; Soccer[0.042]; Sutton[0.041]; Kanchelskis[0.041]; Leicester[0.041]; Liverpool[0.040]; Aston[0.040]; Tottenham[0.040]; Sheringham[0.040]; Chelsea[0.039]; Everton[0.039]; London[0.039]; Premier[0.039]; English[0.039]; English[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sutton ==> ENTITY: \u001b[32mChris Sutton\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.145:0.145[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; League[0.044]; Sheffield[0.044]; Southampton[0.043]; Middlesbrough[0.043]; Coventry[0.043]; London[0.043]; Ekoku[0.043]; Blackburn[0.043]; Wimbledon[0.042]; Kanchelskis[0.042]; Soccer[0.042]; Leeds[0.041]; Sunderland[0.041]; Chelsea[0.041]; Townsend[0.040]; Liverpool[0.040]; Adams[0.040]; Aston[0.040]; Whelan[0.040]; Vialli[0.040]; Premier[0.040]; Leicester[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ekoku\u001b[39m ==> ENTITY: \u001b[32mEfan Ekoku\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.046]; Coventry[0.046]; Leeds[0.045]; Southampton[0.045]; Sheffield[0.045]; Sunderland[0.044]; Middlesbrough[0.044]; Leicester[0.043]; Blackburn[0.042]; Sutton[0.042]; Kanchelskis[0.042]; Liverpool[0.041]; Chelsea[0.041]; Halftime[0.040]; Tottenham[0.040]; Everton[0.040]; Wimbledon[0.040]; Holdsworth[0.040]; Vialli[0.040]; Adams[0.039]; Aston[0.039]; Derby[0.039]; English[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = English ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.097:0.097[0]; log p(e|m)= -1.355:-1.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEnglish[0.048]; League[0.048]; Leeds[0.045]; matches[0.045]; Coventry[0.044]; Southampton[0.043]; Sheffield[0.043]; Leicester[0.043]; Ekoku[0.042]; Sunderland[0.041]; Premier[0.041]; Adams[0.041]; Middlesbrough[0.041]; London[0.041]; Soccer[0.041]; Aston[0.040]; Saturday[0.040]; Blackburn[0.040]; Liverpool[0.039]; Townsend[0.039]; Sutton[0.039]; Wednesday[0.039]; Marshall[0.038]; Attendance[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Aston Villa\u001b[39m ==> ENTITY: \u001b[32mAston Villa F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; Southampton[0.044]; matches[0.044]; Blackburn[0.044]; Middlesbrough[0.044]; Ekoku[0.043]; Sunderland[0.043]; Coventry[0.043]; Sheffield[0.043]; Soccer[0.042]; Leeds[0.042]; Tottenham[0.041]; Kanchelskis[0.041]; Liverpool[0.041]; Everton[0.041]; Leicester[0.040]; Premier[0.040]; Chelsea[0.040]; Arsenal[0.040]; English[0.040]; English[0.040]; Saturday[0.039]; Sutton[0.039]; Derby[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sinton ==> ENTITY: \u001b[32mAndy Sinton\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.175:0.175[0]; log p(e|m)= -1.041:-1.041[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthampton[0.047]; Coventry[0.046]; Leeds[0.045]; Sheffield[0.045]; League[0.044]; Sunderland[0.044]; Leicester[0.044]; Middlesbrough[0.043]; matches[0.042]; Blackburn[0.042]; Ekoku[0.042]; Soccer[0.041]; Liverpool[0.040]; Wednesday[0.040]; Sutton[0.040]; Aston[0.040]; Halftime[0.040]; Tottenham[0.040]; Adams[0.039]; Chelsea[0.039]; Kanchelskis[0.039]; London[0.039]; Saturday[0.039]; Everton[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Everton ==> ENTITY: \u001b[32mEverton F.C.\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.189:0.189[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.045]; Blackburn[0.044]; Ekoku[0.044]; Coventry[0.043]; Southampton[0.043]; matches[0.043]; Liverpool[0.043]; Sunderland[0.043]; Sheffield[0.043]; Middlesbrough[0.043]; Soccer[0.042]; Leeds[0.042]; Kanchelskis[0.042]; Tottenham[0.042]; Aston[0.041]; Arsenal[0.041]; Holdsworth[0.041]; Chelsea[0.040]; Leicester[0.040]; Villa[0.040]; Sutton[0.039]; English[0.039]; English[0.039]; Derby[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSturridge\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mDaniel Sturridge\u001b[39m <---> \u001b[32mDean Sturridge\u001b[39m\t\nSCORES: global= 0.279:0.269[\u001b[31m0.010\u001b[39m]; local()= 0.136:0.129[\u001b[31m0.007\u001b[39m]; log p(e|m)= 0.000:-1.505[\u001b[31m1.505\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthampton[0.045]; Soccer[0.044]; Coventry[0.043]; matches[0.043]; Middlesbrough[0.043]; League[0.043]; Ekoku[0.043]; Sunderland[0.042]; Blackburn[0.042]; Leicester[0.042]; Kanchelskis[0.042]; Sheffield[0.042]; English[0.042]; English[0.042]; Tottenham[0.042]; Leeds[0.041]; Chelsea[0.041]; Derby[0.041]; Everton[0.040]; London[0.040]; Arsenal[0.040]; Liverpool[0.040]; Halftime[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Middlesbrough ==> ENTITY: \u001b[32mMiddlesbrough F.C.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.178:0.178[0]; log p(e|m)= -0.518:-0.518[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; Ekoku[0.045]; Blackburn[0.044]; Southampton[0.044]; matches[0.044]; Sheffield[0.044]; Sunderland[0.043]; Soccer[0.043]; Leeds[0.043]; Coventry[0.043]; Kanchelskis[0.041]; Sutton[0.041]; Aston[0.041]; Tottenham[0.041]; Chelsea[0.041]; Everton[0.040]; Liverpool[0.040]; Premier[0.040]; Arsenal[0.040]; Villa[0.040]; Leicester[0.040]; Saturday[0.040]; Whelan[0.040]; Adams[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Holdsworth ==> ENTITY: \u001b[32mDean Holdsworth\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.151:0.151[0]; log p(e|m)= -1.168:-1.168[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthampton[0.046]; Leeds[0.046]; Coventry[0.045]; Ekoku[0.044]; Sheffield[0.044]; matches[0.043]; Leicester[0.043]; Aston[0.043]; Sunderland[0.043]; Adams[0.042]; Sutton[0.042]; Middlesbrough[0.042]; Blackburn[0.041]; Townsend[0.041]; Whelan[0.041]; Marshall[0.040]; Wednesday[0.040]; Kanchelskis[0.040]; Powell[0.039]; Wimbledon[0.039]; Chelsea[0.039]; Liverpool[0.039]; Saturday[0.039]; Attendance[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Liverpool ==> ENTITY: \u001b[32mLiverpool F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.173:0.173[0]; log p(e|m)= -1.171:-1.171[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; matches[0.044]; Ekoku[0.043]; Blackburn[0.043]; Middlesbrough[0.043]; Sunderland[0.043]; Leeds[0.043]; Southampton[0.042]; Tottenham[0.042]; Sheffield[0.042]; Soccer[0.042]; London[0.042]; Kanchelskis[0.041]; Coventry[0.041]; Chelsea[0.041]; Leicester[0.041]; Premier[0.041]; Everton[0.041]; Sutton[0.041]; Arsenal[0.040]; Villa[0.040]; English[0.040]; English[0.040]; Branch[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kanchelskis\u001b[39m ==> ENTITY: \u001b[32mAndrei Kanchelskis\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEkoku[0.045]; matches[0.044]; Southampton[0.044]; Chelsea[0.044]; League[0.043]; Blackburn[0.043]; Coventry[0.043]; Sheffield[0.043]; Soccer[0.043]; Whelan[0.042]; Everton[0.041]; Middlesbrough[0.041]; Arsenal[0.041]; Sunderland[0.041]; Halftime[0.041]; Liverpool[0.041]; Tottenham[0.041]; Vialli[0.041]; Sheringham[0.040]; Leeds[0.040]; Sutton[0.040]; Adams[0.040]; Aston[0.040]; Villa[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chelsea ==> ENTITY: \u001b[32mChelsea F.C.\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.174:0.174[0]; log p(e|m)= -0.518:-0.518[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; London[0.043]; matches[0.043]; Sheffield[0.043]; Soccer[0.043]; Southampton[0.043]; Blackburn[0.042]; Coventry[0.042]; Middlesbrough[0.042]; Leeds[0.042]; Liverpool[0.042]; Ekoku[0.042]; Sunderland[0.042]; Tottenham[0.042]; Kanchelskis[0.041]; Arsenal[0.041]; Everton[0.040]; Sutton[0.040]; Leicester[0.040]; Aston[0.040]; Premier[0.040]; Vialli[0.040]; Adams[0.039]; Villa[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Townsend ==> ENTITY: \u001b[32mAndy Townsend\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.177:0.177[0]; log p(e|m)= -3.219:-3.219[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthampton[0.045]; Coventry[0.044]; Sunderland[0.043]; Middlesbrough[0.043]; Ekoku[0.043]; Tottenham[0.043]; Kanchelskis[0.043]; Sheffield[0.042]; League[0.042]; Leeds[0.042]; Aston[0.042]; Chelsea[0.042]; Liverpool[0.041]; matches[0.041]; Blackburn[0.041]; London[0.041]; Arsenal[0.041]; Leicester[0.041]; Everton[0.041]; Sheringham[0.040]; Soccer[0.040]; Wednesday[0.040]; English[0.039]; English[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = English ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.092:0.092[0]; log p(e|m)= -1.355:-1.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEnglish[0.048]; League[0.048]; Leeds[0.045]; matches[0.045]; Coventry[0.045]; Southampton[0.044]; Sheffield[0.043]; Leicester[0.043]; Premier[0.042]; Adams[0.042]; Middlesbrough[0.042]; London[0.042]; Soccer[0.041]; Aston[0.040]; Saturday[0.040]; Blackburn[0.040]; Liverpool[0.039]; Townsend[0.039]; Sutton[0.039]; Wednesday[0.039]; Marshall[0.039]; Attendance[0.038]; Tottenham[0.038]; Powell[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Arsenal ==> ENTITY: \u001b[32mArsenal F.C.\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.170:0.170[0]; log p(e|m)= -0.158:-0.158[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.047]; matches[0.045]; Ekoku[0.044]; Chelsea[0.043]; Sheffield[0.043]; Soccer[0.043]; Middlesbrough[0.042]; Southampton[0.042]; Coventry[0.042]; Sunderland[0.042]; Liverpool[0.042]; Tottenham[0.042]; Blackburn[0.042]; Villa[0.042]; Leeds[0.042]; Kanchelskis[0.041]; Premier[0.041]; Everton[0.040]; Aston[0.040]; Sutton[0.040]; Leicester[0.039]; Derby[0.039]; Adams[0.039]; Saturday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leicester ==> ENTITY: \u001b[32mLeicester City F.C.\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.183:0.183[0]; log p(e|m)= -3.219:-3.219[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; Coventry[0.046]; Southampton[0.044]; Blackburn[0.044]; matches[0.044]; Sunderland[0.043]; Sheffield[0.043]; Leeds[0.043]; Middlesbrough[0.043]; Ekoku[0.043]; London[0.042]; Adams[0.041]; Soccer[0.041]; Saturday[0.041]; Sutton[0.040]; Kanchelskis[0.040]; Wednesday[0.040]; Aston[0.039]; Townsend[0.039]; Tottenham[0.039]; Derby[0.039]; Everton[0.039]; Premier[0.039]; Sheringham[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leeds ==> ENTITY: \u001b[32mLeeds United A.F.C.\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.180:0.180[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.047]; Blackburn[0.046]; Sheffield[0.046]; Coventry[0.046]; Southampton[0.046]; Ekoku[0.045]; Middlesbrough[0.045]; Sunderland[0.045]; matches[0.045]; Leicester[0.044]; London[0.044]; Soccer[0.043]; Liverpool[0.043]; Kanchelskis[0.042]; Derby[0.042]; Tottenham[0.042]; Sheringham[0.041]; Everton[0.041]; Sutton[0.041]; Saturday[0.041]; Aston[0.041]; Townsend[0.041]; Chelsea[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tottenham ==> ENTITY: \u001b[32mTottenham Hotspur F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.201:0.201[0]; log p(e|m)= -0.997:-0.997[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.045]; matches[0.044]; Southampton[0.044]; Ekoku[0.044]; Blackburn[0.043]; Soccer[0.043]; Sheffield[0.043]; London[0.043]; Sunderland[0.043]; Coventry[0.043]; Middlesbrough[0.043]; Kanchelskis[0.042]; Leeds[0.042]; Liverpool[0.042]; Chelsea[0.041]; Everton[0.041]; Leicester[0.040]; Arsenal[0.040]; Derby[0.040]; Saturday[0.040]; Aston[0.039]; Villa[0.039]; Premier[0.039]; Vialli[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Vialli\u001b[39m ==> ENTITY: \u001b[32mGianluca Vialli\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.047]; Kanchelskis[0.046]; matches[0.045]; Soccer[0.045]; Chelsea[0.045]; Southampton[0.045]; Liverpool[0.044]; Leeds[0.044]; Ekoku[0.044]; Middlesbrough[0.044]; Blackburn[0.043]; Arsenal[0.043]; Sunderland[0.043]; Halftime[0.043]; Villa[0.043]; Everton[0.043]; Tottenham[0.043]; Coventry[0.042]; Leicester[0.042]; Sheffield[0.042]; Zola[0.042]; Vieira[0.042]; Whelan[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sheffield Wednesday\u001b[39m ==> ENTITY: \u001b[32mSheffield Wednesday F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.049]; matches[0.044]; Coventry[0.044]; Leeds[0.044]; Soccer[0.044]; Southampton[0.044]; Blackburn[0.043]; Sunderland[0.043]; Ekoku[0.043]; Middlesbrough[0.042]; Leicester[0.042]; Saturday[0.041]; Kanchelskis[0.041]; English[0.040]; English[0.040]; Premier[0.040]; Sutton[0.040]; Everton[0.040]; Tottenham[0.040]; Liverpool[0.040]; Aston[0.040]; Whelan[0.039]; Adams[0.039]; Arsenal[0.038]; \t\n\r [==============================>...............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 7s388ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3090/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1312testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Geneva ==> ENTITY: \u001b[32mGeneva\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.083:0.083[0]; log p(e|m)= -0.247:-0.247[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGeneva[0.046]; laws[0.044]; laws[0.044]; built[0.044]; small[0.043]; procedures[0.042]; officials[0.042]; officials[0.042]; large[0.042]; setting[0.041]; said[0.041]; trying[0.041]; days[0.041]; Friday[0.041]; time[0.041]; gritty[0.041]; got[0.041]; government[0.041]; covering[0.041]; nations[0.040]; level[0.040]; diplomatic[0.040]; groups[0.040]; delegates[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ulan Bator\u001b[39m ==> ENTITY: \u001b[32mUlan Bator\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMongolia[0.049]; Mongolia[0.049]; academic[0.043]; music[0.042]; Moscow[0.042]; visit[0.042]; Net[0.042]; Berne[0.042]; official[0.041]; official[0.041]; officials[0.041]; based[0.041]; conference[0.041]; commerce[0.041]; added[0.041]; Russia[0.040]; literature[0.040]; services[0.040]; existing[0.040]; state[0.040]; foreign[0.040]; ministry[0.040]; arrived[0.040]; ve[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.S.-based\u001b[39m ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.050:0.050[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbased[0.045]; music[0.043]; music[0.043]; business[0.043]; million[0.042]; Tim[0.042]; said[0.042]; said[0.042]; network[0.042]; unknown[0.042]; Network[0.042]; global[0.041]; computer[0.041]; computer[0.041]; Internet[0.041]; Internet[0.041]; Internet[0.041]; cassettes[0.041]; click[0.041]; Casey[0.041]; mouse[0.041]; distributed[0.041]; dollar[0.040]; Communications[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.046:0.046[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerica[0.045]; cash[0.044]; music[0.043]; music[0.043]; European[0.043]; Information[0.043]; foreign[0.043]; based[0.042]; Association[0.042]; Marc[0.041]; lot[0.041]; laws[0.041]; rulebooks[0.041]; business[0.041]; association[0.041]; potential[0.041]; network[0.041]; network[0.041]; trade[0.041]; pirated[0.040]; opposing[0.040]; policy[0.040]; services[0.040]; hold[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = European ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.093:0.093[0]; log p(e|m)= -0.256:-0.256[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerica[0.045]; trade[0.043]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; official[0.043]; policy[0.042]; people[0.042]; unknown[0.042]; companies[0.042]; added[0.042]; foreign[0.041]; treaties[0.041]; legal[0.041]; Association[0.041]; fear[0.041]; Russia[0.041]; understand[0.041]; understand[0.041]; business[0.040]; aware[0.040]; People[0.040]; agenda[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mInternet\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mInternet\u001b[39m <---> \u001b[32mWorld Wide Web\u001b[39m\t\nSCORES: global= 0.272:0.230[\u001b[31m0.042\u001b[39m]; local()= 0.205:0.177[\u001b[31m0.029\u001b[39m]; log p(e|m)= -0.037:-3.381[\u001b[31m3.345\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInternet[0.049]; Internet[0.049]; Internet[0.049]; network[0.046]; computer[0.045]; services[0.043]; copyright[0.043]; copyright[0.043]; copyright[0.043]; copyright[0.043]; digital[0.042]; industry[0.041]; offering[0.041]; groups[0.040]; money[0.039]; based[0.039]; based[0.039]; industries[0.039]; industries[0.039]; survey[0.038]; information[0.038]; entertainment[0.038]; European[0.038]; legal[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mInternet\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mInternet\u001b[39m <---> \u001b[32mWorld Wide Web\u001b[39m\t\nSCORES: global= 0.271:0.230[\u001b[31m0.041\u001b[39m]; local()= 0.179:0.158[\u001b[31m0.021\u001b[39m]; log p(e|m)= -0.037:-3.381[\u001b[31m3.345\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInternet[0.050]; network[0.047]; computer[0.046]; software[0.045]; copyright[0.044]; copyright[0.044]; copyright[0.044]; digital[0.042]; distribution[0.042]; music[0.042]; industry[0.042]; industry[0.042]; groups[0.040]; electronic[0.040]; dollars[0.040]; money[0.039]; based[0.039]; industries[0.039]; survey[0.039]; information[0.039]; entertainment[0.039]; legal[0.039]; consumer[0.039]; extension[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mInternet\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mInternet\u001b[39m <---> \u001b[32mWorld Wide Web\u001b[39m\t\nSCORES: global= 0.275:0.233[\u001b[31m0.042\u001b[39m]; local()= 0.264:0.232[\u001b[31m0.032\u001b[39m]; log p(e|m)= -0.037:-3.381[\u001b[31m3.345\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInternet[0.046]; Internet[0.046]; Internet[0.046]; Internet[0.046]; Internet[0.046]; network[0.043]; network[0.043]; computer[0.043]; computer[0.043]; computer[0.043]; software[0.041]; services[0.041]; pirated[0.041]; copyrights[0.040]; copyright[0.040]; copyright[0.040]; messages[0.040]; companies[0.039]; technology[0.039]; technology[0.039]; business[0.039]; music[0.039]; music[0.039]; distributed[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mInternet\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mInternet\u001b[39m <---> \u001b[32mWorld Wide Web\u001b[39m\t\nSCORES: global= 0.273:0.231[\u001b[31m0.042\u001b[39m]; local()= 0.226:0.194[\u001b[31m0.032\u001b[39m]; log p(e|m)= -0.037:-3.381[\u001b[31m3.345\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInternet[0.047]; Internet[0.047]; Internet[0.047]; Internet[0.047]; Internet[0.047]; network[0.044]; computer[0.044]; services[0.042]; pirated[0.042]; copyrights[0.041]; copyright[0.041]; copyright[0.041]; companies[0.040]; digital[0.040]; Association[0.040]; technology[0.040]; technology[0.040]; business[0.040]; music[0.040]; offering[0.039]; trade[0.039]; association[0.038]; European[0.038]; policy[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mongolia ==> ENTITY: \u001b[32mMongolia\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.361:-0.361[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMongolia[0.052]; rights[0.045]; Russia[0.044]; pacts[0.042]; pacts[0.042]; visit[0.042]; policy[0.042]; groups[0.042]; legal[0.041]; foreign[0.041]; powers[0.041]; Moscow[0.041]; European[0.041]; agenda[0.041]; officials[0.041]; official[0.041]; official[0.041]; conference[0.041]; services[0.040]; opponents[0.040]; president[0.040]; money[0.040]; life[0.040]; potential[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Geneva ==> ENTITY: \u001b[32mGeneva\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.090:0.090[0]; log p(e|m)= -0.247:-0.247[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGeneva[0.046]; laws[0.044]; laws[0.044]; built[0.044]; small[0.043]; procedures[0.042]; officials[0.042]; officials[0.042]; large[0.042]; setting[0.041]; said[0.041]; trying[0.041]; days[0.041]; Friday[0.041]; time[0.041]; gritty[0.041]; got[0.041]; government[0.041]; nations[0.040]; level[0.040]; diplomatic[0.040]; groups[0.040]; delegates[0.040]; conference[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mInternet\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mInternet\u001b[39m <---> \u001b[32mWorld Wide Web\u001b[39m\t\nSCORES: global= 0.270:0.230[\u001b[31m0.040\u001b[39m]; local()= 0.162:0.155[\u001b[31m0.006\u001b[39m]; log p(e|m)= -0.037:-3.381[\u001b[31m3.345\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndatabases[0.045]; online[0.045]; online[0.045]; copyright[0.044]; access[0.044]; use[0.044]; companies[0.043]; digital[0.043]; music[0.043]; licensing[0.041]; dollars[0.041]; subject[0.040]; public[0.040]; articles[0.040]; proposals[0.040]; announced[0.040]; based[0.040]; based[0.040]; industries[0.040]; week[0.040]; kind[0.040]; owned[0.040]; information[0.040]; illegal[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Information Technology Association of America\u001b[39m ==> ENTITY: \u001b[32mInformation Technology Association of America\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nservices[0.048]; companies[0.047]; information[0.046]; business[0.046]; industries[0.045]; technology[0.045]; technology[0.045]; global[0.044]; trade[0.044]; Network[0.044]; president[0.043]; association[0.042]; Internet[0.042]; Internet[0.042]; Internet[0.042]; Internet[0.042]; Internet[0.042]; software[0.042]; policy[0.042]; Net[0.041]; network[0.041]; network[0.041]; distributed[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.104:0.104[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.046]; president[0.043]; lone[0.043]; European[0.043]; Mongolia[0.042]; Mongolia[0.042]; services[0.042]; Alexander[0.042]; state[0.042]; officials[0.042]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; foreign[0.041]; ministry[0.041]; America[0.041]; Association[0.041]; Berne[0.041]; official[0.041]; official[0.041]; policy[0.040]; added[0.040]; chief[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mongolia ==> ENTITY: \u001b[32mMongolia\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.090:0.090[0]; log p(e|m)= -0.361:-0.361[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMongolia[0.052]; rights[0.045]; Russia[0.045]; pacts[0.043]; pacts[0.043]; visit[0.042]; policy[0.042]; groups[0.042]; legal[0.041]; foreign[0.041]; powers[0.041]; Moscow[0.041]; agenda[0.041]; officials[0.041]; official[0.041]; conference[0.041]; opponents[0.040]; money[0.040]; life[0.040]; dollars[0.040]; aware[0.040]; state[0.040]; future[0.039]; Supporters[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mInternet\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mInternet\u001b[39m <---> \u001b[32mWorld Wide Web\u001b[39m\t\nSCORES: global= 0.273:0.232[\u001b[31m0.041\u001b[39m]; local()= 0.201:0.177[\u001b[31m0.024\u001b[39m]; log p(e|m)= -0.037:-3.381[\u001b[31m3.345\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInternet[0.050]; Internet[0.050]; network[0.047]; computer[0.047]; computer[0.047]; software[0.045]; pirated[0.044]; copyrights[0.044]; copyright[0.044]; messages[0.043]; technology[0.043]; business[0.042]; music[0.042]; music[0.042]; distributed[0.042]; million[0.042]; Network[0.042]; Communications[0.042]; based[0.041]; operators[0.041]; Corporation[0.040]; global[0.040]; transmitted[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mInternet\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mInternet\u001b[39m <---> \u001b[32mWorld Wide Web\u001b[39m\t\nSCORES: global= 0.275:0.233[\u001b[31m0.042\u001b[39m]; local()= 0.252:0.221[\u001b[31m0.031\u001b[39m]; log p(e|m)= -0.037:-3.381[\u001b[31m3.345\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInternet[0.046]; Internet[0.046]; Internet[0.046]; Internet[0.046]; network[0.044]; network[0.044]; computer[0.043]; computer[0.043]; computer[0.043]; software[0.042]; services[0.041]; pirated[0.041]; copyrights[0.041]; copyright[0.041]; copyright[0.041]; messages[0.040]; companies[0.040]; technology[0.039]; technology[0.039]; business[0.039]; music[0.039]; music[0.039]; distributed[0.039]; million[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mInternet\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mInternet\u001b[39m <---> \u001b[32mWorld Wide Web\u001b[39m\t\nSCORES: global= 0.272:0.232[\u001b[31m0.040\u001b[39m]; local()= 0.116:0.107[\u001b[31m0.009\u001b[39m]; log p(e|m)= -0.037:-3.381[\u001b[31m3.345\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncopyright[0.046]; copyright[0.046]; copyright[0.046]; digital[0.045]; digital[0.045]; groups[0.043]; laws[0.042]; laws[0.042]; proposals[0.042]; built[0.041]; government[0.041]; agenda[0.041]; delegates[0.040]; said[0.040]; rights[0.040]; Cyberspace[0.040]; Cyberspace[0.040]; works[0.040]; conference[0.040]; meeting[0.040]; debate[0.040]; officials[0.040]; officials[0.040]; time[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = MCI Communications Corporation\u001b[39m ==> ENTITY: \u001b[32mMCI Communications\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnetwork[0.046]; business[0.046]; Internet[0.044]; Internet[0.044]; Internet[0.044]; Network[0.043]; million[0.043]; billion[0.043]; global[0.042]; software[0.042]; messages[0.041]; said[0.041]; said[0.041]; operators[0.041]; based[0.040]; dollar[0.040]; computer[0.040]; computer[0.040]; music[0.040]; music[0.040]; responsible[0.040]; multi[0.039]; Tim[0.039]; hold[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = James Love ==> ENTITY: \u001b[32mJames Love (NGO director)\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.104:0.104[0]; log p(e|m)= -3.772:-3.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWashington[0.043]; works[0.043]; companies[0.043]; rights[0.042]; academic[0.042]; dollars[0.042]; subject[0.042]; announced[0.042]; software[0.042]; licensing[0.042]; read[0.042]; access[0.042]; property[0.042]; public[0.042]; based[0.042]; based[0.042]; Berne[0.041]; old[0.041]; use[0.041]; century[0.040]; music[0.040]; Technology[0.040]; scientific[0.040]; literature[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mInternet\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mInternet\u001b[39m <---> \u001b[32mWorld Wide Web\u001b[39m\t\nSCORES: global= 0.274:0.232[\u001b[31m0.042\u001b[39m]; local()= 0.237:0.205[\u001b[31m0.031\u001b[39m]; log p(e|m)= -0.037:-3.381[\u001b[31m3.345\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInternet[0.047]; Internet[0.047]; Internet[0.047]; Internet[0.047]; network[0.044]; network[0.044]; computer[0.043]; computer[0.043]; software[0.042]; pirated[0.041]; copyrights[0.041]; copyright[0.041]; copyright[0.041]; messages[0.040]; companies[0.040]; technology[0.040]; technology[0.040]; business[0.039]; music[0.039]; music[0.039]; distributed[0.039]; million[0.039]; Network[0.039]; Communications[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mInternet\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mInternet\u001b[39m <---> \u001b[32mWorld Wide Web\u001b[39m\t\nSCORES: global= 0.274:0.233[\u001b[31m0.042\u001b[39m]; local()= 0.212:0.186[\u001b[31m0.026\u001b[39m]; log p(e|m)= -0.037:-3.381[\u001b[31m3.345\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInternet[0.048]; Internet[0.048]; Internet[0.048]; Internet[0.048]; network[0.045]; computer[0.044]; services[0.042]; copyrights[0.042]; copyright[0.042]; copyright[0.042]; copyright[0.042]; companies[0.041]; digital[0.041]; Association[0.040]; technology[0.040]; technology[0.040]; business[0.040]; offering[0.040]; trade[0.039]; association[0.039]; European[0.038]; policy[0.038]; Information[0.038]; money[0.038]; \t\n\r [==============================>...............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 7s281ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3113/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1388testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vienna ==> ENTITY: \u001b[32mVienna\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.026:0.026[0]; log p(e|m)= -0.126:-0.126[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nround[0.050]; International[0.048]; title[0.048]; German[0.048]; Saturday[0.046]; Schulz[0.046]; Schulz[0.046]; Federation[0.046]; Jose[0.046]; Boxing[0.046]; Boxing[0.046]; Axel[0.045]; Heavyweight[0.045]; non[0.045]; fight[0.045]; Defeats[0.045]; heavyweight[0.044]; outpointed[0.044]; Cuba[0.044]; Ribalta[0.043]; Ribalta[0.043]; Fight[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.036:0.036[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVienna[0.049]; Fight[0.048]; Schulz[0.048]; Schulz[0.048]; Axel[0.048]; Federation[0.047]; Cuba[0.047]; round[0.047]; Defeats[0.047]; International[0.046]; title[0.046]; Saturday[0.046]; Jose[0.045]; fight[0.045]; outpointed[0.044]; heavyweight[0.044]; Heavyweight[0.043]; Boxing[0.043]; Boxing[0.043]; Ribalta[0.042]; Ribalta[0.042]; non[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Axel Schulz\u001b[39m ==> ENTITY: \u001b[32mAxel Schulz\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.135:0.135[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nheavyweight[0.054]; fight[0.053]; Heavyweight[0.052]; outpointed[0.052]; Boxing[0.051]; Boxing[0.051]; title[0.049]; German[0.049]; round[0.049]; Fight[0.048]; Ribalta[0.047]; Ribalta[0.047]; Schulz[0.047]; Defeats[0.046]; Vienna[0.046]; Saturday[0.044]; Cuba[0.044]; Federation[0.044]; International[0.043]; Jose[0.043]; non[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cuba ==> ENTITY: \u001b[32mCuba\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.028:0.028[0]; log p(e|m)= -0.240:-0.240[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHeavyweight[0.052]; heavyweight[0.051]; Boxing[0.050]; Boxing[0.050]; Jose[0.049]; round[0.049]; title[0.047]; International[0.046]; Defeats[0.046]; Saturday[0.045]; outpointed[0.045]; Federation[0.045]; Fight[0.044]; Ribalta[0.044]; Ribalta[0.044]; fight[0.044]; Axel[0.043]; non[0.043]; Vienna[0.043]; German[0.042]; Schulz[0.040]; Schulz[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Schulz\u001b[39m ==> ENTITY: \u001b[32mAxel Schulz\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.135:0.135[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nheavyweight[0.052]; fight[0.050]; Heavyweight[0.050]; outpointed[0.049]; Boxing[0.049]; Boxing[0.049]; title[0.047]; German[0.046]; round[0.046]; Fight[0.046]; Ribalta[0.045]; Ribalta[0.045]; Schulz[0.045]; Axel[0.044]; Defeats[0.044]; Vienna[0.044]; Saturday[0.042]; Cuba[0.042]; Federation[0.042]; International[0.041]; Jose[0.041]; non[0.040]; \t\n\r [==============================>...............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 7s267ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3118/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1272testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Okla ==> ENTITY: \u001b[32mOklahoma\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.056:0.056[0]; log p(e|m)= -1.100:-1.100[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTexas[0.046]; Panhandle[0.043]; Steers[0.043]; Heifers[0.042]; lbs[0.042]; lbs[0.042]; cattle[0.042]; cattle[0.042]; lower[0.042]; area[0.042]; Amarillo[0.042]; slaughter[0.042]; heifers[0.041]; heifers[0.041]; cwt[0.041]; reporting[0.041]; Friday[0.041]; market[0.041]; following[0.041]; feedlot[0.041]; Sales[0.041]; based[0.041]; Select[0.040]; Select[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Amarillo ==> ENTITY: \u001b[32mAmarillo, Texas\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.177:-0.177[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTexas[0.049]; area[0.045]; Panhandle[0.045]; Steers[0.045]; lbs[0.044]; lbs[0.044]; cattle[0.044]; cattle[0.044]; based[0.043]; weekly[0.043]; cwt[0.043]; Chicago[0.043]; Friday[0.043]; Okla[0.043]; Week[0.043]; feedlot[0.043]; Sales[0.043]; heifers[0.042]; heifers[0.042]; following[0.042]; Trade[0.042]; Select[0.042]; Select[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.030:0.030[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \narea[0.046]; based[0.044]; Week[0.043]; Amarillo[0.043]; Friday[0.043]; following[0.042]; Slaughter[0.042]; Slaughter[0.042]; Slaughter[0.042]; Steers[0.042]; Select[0.042]; Select[0.042]; weekly[0.042]; Choice[0.041]; Choice[0.041]; Panhandle[0.041]; reported[0.041]; head[0.041]; head[0.041]; reporting[0.040]; Note[0.040]; Confirmed[0.040]; said[0.040]; percent[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = USDA\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Agriculture\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.070:0.070[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfed[0.043]; Steers[0.043]; reporting[0.043]; lbs[0.042]; lbs[0.042]; following[0.042]; slaughter[0.042]; based[0.042]; Chicago[0.042]; cattle[0.042]; cattle[0.042]; Amarillo[0.042]; market[0.042]; Confirmed[0.041]; weights[0.041]; area[0.041]; Texas[0.041]; Friday[0.041]; cwt[0.041]; Panhandle[0.041]; reported[0.041]; Okla[0.041]; feedlot[0.041]; Sales[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = USDA\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Agriculture\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.070:0.070[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfed[0.043]; Steers[0.043]; reporting[0.043]; lbs[0.042]; lbs[0.042]; following[0.042]; slaughter[0.042]; based[0.042]; Chicago[0.042]; cattle[0.042]; cattle[0.042]; Amarillo[0.042]; market[0.042]; Confirmed[0.041]; weights[0.041]; area[0.041]; Texas[0.041]; Friday[0.041]; cwt[0.041]; Panhandle[0.041]; reported[0.041]; Okla[0.041]; feedlot[0.041]; Sales[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPanhandle\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mTexas Panhandle\u001b[39m <---> \u001b[32mOklahoma Panhandle\u001b[39m\t\nSCORES: global= 0.255:0.248[\u001b[31m0.006\u001b[39m]; local()= 0.122:0.108[\u001b[31m0.014\u001b[39m]; log p(e|m)= -0.747:-1.556[\u001b[31m0.809\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTexas[0.048]; area[0.047]; percent[0.045]; Amarillo[0.045]; Okla[0.043]; moderate[0.042]; cattle[0.042]; cattle[0.042]; Friday[0.041]; reporting[0.041]; slow[0.041]; lower[0.041]; said[0.041]; based[0.041]; reported[0.040]; fed[0.040]; feedlot[0.040]; movement[0.040]; market[0.040]; weekly[0.040]; following[0.040]; steers[0.039]; steers[0.039]; prices[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Texas ==> ENTITY: \u001b[32mTexas\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.060:0.060[0]; log p(e|m)= -0.161:-0.161[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPanhandle[0.047]; area[0.045]; Confirmed[0.044]; cattle[0.043]; cattle[0.043]; Amarillo[0.043]; following[0.042]; based[0.041]; feedlot[0.041]; percent[0.041]; Friday[0.041]; Slaughter[0.041]; Slaughter[0.041]; Slaughter[0.041]; slow[0.041]; Okla[0.041]; head[0.041]; head[0.041]; net[0.040]; Choice[0.040]; Choice[0.040]; market[0.040]; prices[0.040]; reported[0.040]; \t\n\r [===============================>..............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 7s194ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3125/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1279testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= -0.016:-0.016[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nweek[0.047]; Canadian[0.046]; Canadian[0.046]; December[0.044]; weekly[0.044]; ending[0.043]; Commission[0.042]; Date[0.042]; Flax[0.042]; Wk[0.042]; Wk[0.042]; Statistics[0.041]; Dec[0.041]; Barley[0.041]; Rye[0.041]; Curr[0.041]; Curr[0.041]; Reuter[0.040]; statistics[0.040]; Visible[0.040]; Oats[0.039]; Wheat[0.039]; Supplies[0.039]; Grain[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Statistics Canada\u001b[39m ==> ENTITY: \u001b[32mStatistics Canada\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.021:0.021[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nindicated[0.049]; addition[0.047]; Total[0.046]; Exports[0.046]; Oats[0.045]; Oats[0.045]; tonnes[0.045]; exports[0.044]; following[0.044]; Canola[0.044]; Canola[0.044]; Flaxseed[0.043]; September[0.043]; Flax[0.043]; Chicago[0.042]; August[0.042]; Corn[0.042]; Corn[0.042]; Ago[0.042]; newsdesk[0.041]; Rye[0.041]; Rye[0.041]; Barley[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Canadian Grain Commission\u001b[39m ==> ENTITY: \u001b[32mCanadian Grain Commission\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.043:0.043[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWk[0.047]; Wk[0.047]; grain[0.046]; Canadian[0.045]; Durum[0.044]; Wheat[0.042]; Yr[0.042]; Yr[0.042]; Yr[0.042]; Statistics[0.041]; Oats[0.041]; Supplies[0.041]; weekly[0.041]; tonnes[0.041]; Farmers[0.040]; Deliveries[0.040]; Canola[0.040]; Flax[0.040]; December[0.040]; Corn[0.040]; Barley[0.040]; Curr[0.039]; Curr[0.039]; Dec[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.026:0.026[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSeptember[0.078]; August[0.077]; following[0.076]; addition[0.075]; Canada[0.075]; indicated[0.072]; Flax[0.071]; Total[0.069]; Rye[0.069]; Corn[0.068]; Corn[0.068]; Statistics[0.068]; newsdesk[0.067]; Oats[0.066]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= -0.000:-0.000[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.050]; following[0.048]; August[0.047]; September[0.047]; addition[0.047]; indicated[0.045]; Total[0.044]; Chicago[0.044]; Statistics[0.044]; Rye[0.043]; Rye[0.043]; tonnes[0.043]; Exports[0.043]; Barley[0.042]; newsdesk[0.042]; Corn[0.042]; Corn[0.042]; exports[0.041]; Flax[0.041]; Ago[0.041]; Oats[0.040]; Canola[0.040]; Canola[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canadian ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.012:0.012[0]; log p(e|m)= -0.412:-0.412[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanadian[0.055]; weekly[0.047]; week[0.047]; December[0.044]; Statistics[0.043]; statistics[0.042]; ending[0.042]; Commission[0.042]; Farmers[0.042]; Chicago[0.042]; Visible[0.041]; Dec[0.041]; Wk[0.040]; Wk[0.040]; Supplies[0.040]; Date[0.040]; Rye[0.040]; Wheat[0.039]; grain[0.039]; Deliveries[0.039]; Grain[0.039]; tonnes[0.039]; Barley[0.039]; Flax[0.039]; \t\n\r [===============================>..............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 7s223ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3131/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1332testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rangoon\u001b[39m ==> ENTITY: \u001b[32mYangon\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nuniversity[0.045]; Rangoon[0.045]; University[0.044]; Yangon[0.044]; Yangon[0.044]; Burmese[0.043]; Burmese[0.043]; northern[0.042]; march[0.041]; march[0.041]; marched[0.041]; students[0.041]; students[0.041]; students[0.041]; protests[0.041]; protests[0.041]; Monday[0.041]; Institute[0.041]; campus[0.040]; Tuesday[0.040]; taking[0.040]; outskirts[0.040]; Friday[0.039]; dawn[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yangon Institute of Technology\u001b[39m ==> ENTITY: \u001b[32mYangon Technological University\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nuniversity[0.049]; University[0.048]; Rangoon[0.047]; Rangoon[0.047]; campus[0.046]; student[0.046]; Yangon[0.046]; students[0.045]; students[0.045]; students[0.045]; central[0.043]; Dagon[0.042]; October[0.042]; Yit[0.041]; Yit[0.041]; Monday[0.041]; moved[0.041]; northern[0.041]; street[0.041]; Burmese[0.041]; Burmese[0.041]; launched[0.041]; briefly[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rangoon\u001b[39m ==> ENTITY: \u001b[32mYangon\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nuniversity[0.045]; university[0.045]; Rangoon[0.045]; University[0.044]; Yangon[0.044]; protesters[0.043]; student[0.042]; march[0.042]; students[0.041]; students[0.041]; students[0.041]; protests[0.041]; protests[0.041]; northern[0.041]; central[0.041]; Monday[0.041]; detained[0.041]; campuses[0.041]; Tuesday[0.040]; Tuesday[0.040]; taking[0.040]; Order[0.040]; outskirts[0.040]; Friday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shwe Dagon\u001b[39m ==> ENTITY: \u001b[32mShwedagon Pagoda\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npagoda[0.047]; university[0.043]; university[0.043]; march[0.043]; University[0.042]; near[0.042]; protesters[0.042]; Monday[0.042]; Yangon[0.042]; said[0.041]; said[0.041]; State[0.041]; students[0.041]; students[0.041]; students[0.041]; Friday[0.041]; Tuesday[0.041]; Tuesday[0.041]; exact[0.041]; Rangoon[0.041]; Rangoon[0.041]; street[0.041]; protests[0.040]; protests[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.121:0.121[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npublished[0.047]; newspapers[0.046]; told[0.046]; launched[0.044]; released[0.044]; taking[0.044]; said[0.044]; said[0.044]; independent[0.043]; Monday[0.043]; Tuesday[0.043]; Tuesday[0.043]; miles[0.043]; October[0.043]; October[0.043]; central[0.043]; university[0.043]; university[0.043]; march[0.042]; Friday[0.042]; punishment[0.042]; handling[0.042]; handling[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burmese ==> ENTITY: \u001b[32mBurma\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.098:0.098[0]; log p(e|m)= -1.228:-1.228[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurmese[0.049]; northern[0.049]; group[0.045]; Rangoon[0.045]; Rangoon[0.045]; moved[0.044]; protests[0.043]; protests[0.043]; Yangon[0.043]; Yangon[0.043]; Institute[0.043]; students[0.043]; students[0.043]; students[0.043]; called[0.042]; said[0.042]; exact[0.042]; handling[0.042]; culminated[0.042]; Monday[0.042]; student[0.042]; owners[0.042]; numbers[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = University of Yangon\u001b[39m ==> ENTITY: \u001b[32mUniversity of Yangon\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.223:0.223[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nuniversity[0.045]; Rangoon[0.044]; Rangoon[0.044]; Rangoon[0.044]; protesters[0.043]; Yangon[0.043]; student[0.042]; march[0.042]; march[0.042]; marched[0.042]; students[0.041]; students[0.041]; students[0.041]; students[0.041]; protests[0.041]; protests[0.041]; Institute[0.041]; campus[0.041]; Monday[0.041]; Burmese[0.041]; Burmese[0.041]; Tuesday[0.040]; Tuesday[0.040]; dawn[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rangoon\u001b[39m ==> ENTITY: \u001b[32mYangon\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nuniversity[0.046]; Rangoon[0.046]; Rangoon[0.046]; University[0.045]; Yangon[0.045]; Yangon[0.045]; Burmese[0.045]; Burmese[0.045]; protesters[0.044]; northern[0.044]; central[0.043]; student[0.043]; march[0.043]; march[0.043]; marched[0.043]; students[0.042]; students[0.042]; students[0.042]; protests[0.042]; protests[0.042]; Monday[0.042]; Institute[0.042]; campus[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burmese ==> ENTITY: \u001b[32mBurma\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.103:0.103[0]; log p(e|m)= -1.228:-1.228[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurmese[0.048]; northern[0.047]; group[0.043]; Rangoon[0.043]; Rangoon[0.043]; moved[0.043]; protests[0.042]; Yangon[0.041]; Yangon[0.041]; Institute[0.041]; students[0.041]; students[0.041]; students[0.041]; called[0.041]; said[0.041]; exact[0.041]; handling[0.041]; Monday[0.041]; owners[0.040]; numbers[0.040]; march[0.040]; march[0.040]; launched[0.040]; Tuesday[0.040]; \t\n\r [===============================>..............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 7s192ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3140/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1309testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.021:0.021[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmonth[0.046]; reports[0.044]; Thursday[0.043]; following[0.043]; add[0.043]; levels[0.043]; Earlier[0.042]; daytime[0.042]; daytime[0.042]; higher[0.041]; Access[0.041]; Access[0.041]; cold[0.041]; session[0.041]; ended[0.041]; added[0.041]; trade[0.041]; trade[0.041]; temperatures[0.040]; futures[0.040]; futures[0.040]; raised[0.040]; cents[0.040]; cents[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NYMEX\u001b[39m ==> ENTITY: \u001b[32mNew York Mercantile Exchange\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfutures[0.046]; futures[0.046]; traded[0.044]; Nymex[0.044]; Nymex[0.044]; Northeastern[0.044]; traders[0.042]; trade[0.042]; trade[0.042]; markets[0.041]; floor[0.041]; roughly[0.041]; cents[0.041]; cents[0.041]; cents[0.041]; cents[0.041]; cents[0.041]; barrel[0.040]; oil[0.040]; trader[0.040]; crude[0.040]; crude[0.040]; ended[0.040]; added[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NYMEX\u001b[39m ==> ENTITY: \u001b[32mNew York Mercantile Exchange\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntraded[0.046]; traded[0.046]; Nymex[0.046]; Northeastern[0.046]; traders[0.045]; trade[0.044]; markets[0.044]; roughly[0.043]; cents[0.043]; cents[0.043]; cents[0.043]; cents[0.043]; cents[0.043]; barrel[0.043]; barrel[0.043]; oil[0.043]; trader[0.043]; crude[0.042]; crude[0.042]; ended[0.042]; months[0.042]; climbed[0.042]; cent[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Los Angeles ==> ENTITY: \u001b[32mLos Angeles\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.082:0.082[0]; log p(e|m)= -0.101:-0.101[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDavid[0.047]; overall[0.044]; overall[0.044]; traded[0.044]; traded[0.044]; roughly[0.044]; month[0.043]; month[0.043]; moved[0.042]; January[0.042]; January[0.042]; January[0.042]; months[0.042]; nearby[0.041]; hours[0.041]; bureau[0.040]; said[0.040]; delivery[0.040]; barely[0.040]; settlement[0.040]; climbed[0.040]; Access[0.040]; edging[0.039]; light[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Los Angeles ==> ENTITY: \u001b[32mLos Angeles\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.046:0.046[0]; log p(e|m)= -0.101:-0.101[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrade[0.045]; trade[0.045]; Thursday[0.044]; key[0.044]; month[0.043]; following[0.042]; helping[0.042]; energy[0.042]; energy[0.042]; added[0.042]; reports[0.041]; weather[0.041]; daytime[0.041]; daytime[0.041]; add[0.041]; said[0.041]; ended[0.041]; Access[0.041]; Access[0.041]; levels[0.040]; raised[0.040]; tighter[0.040]; light[0.040]; markets[0.040]; \t\n\r [===============================>..............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 7s140ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3145/4485 \r\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Astle ==> ENTITY: \u001b[32mNathan Astle\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.175:0.175[0]; log p(e|m)= -1.252:-1.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAstle[0.050]; Vaughan[0.049]; Vaughan[0.049]; Zealand[0.046]; wicket[0.046]; wickets[0.046]; Fleming[0.046]; Mushtaq[0.046]; Mushtaq[0.046]; overs[0.045]; Parore[0.045]; Saqlain[0.045]; Saqlain[0.045]; Saqlain[0.045]; Germon[0.045]; Germon[0.045]; innings[0.044]; Cairns[0.044]; Cairns[0.044]; Waqar[0.044]; Waqar[0.044]; Waqar[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Germon\u001b[39m ==> ENTITY: \u001b[32mLee Germon\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.045]; Germon[0.045]; Vaughan[0.044]; Cricket[0.043]; Doull[0.043]; Zealand[0.043]; Zealand[0.043]; wickets[0.042]; wicket[0.042]; overs[0.042]; Cairns[0.042]; Harris[0.041]; Harris[0.041]; Harris[0.041]; Harris[0.041]; Harris[0.041]; second[0.041]; Kennedy[0.041]; Saqlain[0.040]; Friday[0.040]; Pakistan[0.040]; Pakistan[0.040]; Pakistan[0.040]; Pakistan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wasim\u001b[39m ==> ENTITY: \u001b[32mWasim Akram\u001b[39m\t\nSCORES: global= 0.311:0.311[0]; local()= 0.247:0.247[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAkram[0.045]; Wasim[0.044]; Wasim[0.044]; Wasim[0.044]; Wasim[0.044]; Mushtaq[0.042]; Mushtaq[0.042]; Ijaz[0.042]; Ijaz[0.042]; Ijaz[0.042]; Saqlain[0.042]; Saqlain[0.042]; Saqlain[0.042]; Waqar[0.041]; Waqar[0.041]; Waqar[0.041]; Malik[0.041]; Malik[0.041]; Khan[0.040]; Khan[0.040]; Khan[0.040]; Moin[0.039]; Moin[0.039]; Moin[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Waqar\u001b[39m ==> ENTITY: \u001b[32mWaqar Younis\u001b[39m\t\nSCORES: global= 0.311:0.311[0]; local()= 0.292:0.292[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWasim[0.049]; Wasim[0.049]; Wasim[0.049]; Wasim[0.049]; Wasim[0.049]; Akram[0.047]; Waqar[0.046]; Waqar[0.046]; Saqlain[0.045]; Saqlain[0.045]; Saqlain[0.045]; Pakistan[0.045]; Mushtaq[0.044]; Mushtaq[0.044]; Malik[0.044]; Malik[0.044]; Ijaz[0.044]; Ijaz[0.044]; Ijaz[0.044]; Afridi[0.043]; Afridi[0.043]; Afridi[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pakistan ==> ENTITY: \u001b[32mPakistan national cricket team\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.201:0.201[0]; log p(e|m)= -2.749:-2.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.050]; Pakistan[0.050]; Pakistan[0.050]; cricket[0.045]; Akram[0.043]; Saqlain[0.042]; Wasim[0.042]; Waqar[0.041]; Malik[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Afridi[0.041]; Sialkot[0.041]; Ijaz[0.041]; Cricket[0.040]; Khan[0.040]; Vaughan[0.040]; Saeed[0.039]; Younis[0.039]; Moin[0.039]; overs[0.039]; Zahoor[0.038]; wicket[0.038]; day[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saqlain Mushtaq\u001b[39m ==> ENTITY: \u001b[32mSaqlain Mushtaq\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.046]; Wasim[0.046]; Akram[0.046]; Mushtaq[0.045]; Pakistan[0.045]; Pakistan[0.045]; Pakistan[0.045]; Pakistan[0.045]; Cricket[0.044]; Ijaz[0.044]; wicket[0.043]; wickets[0.043]; Waqar[0.043]; overs[0.043]; Sialkot[0.042]; Malik[0.042]; Afridi[0.042]; Vaughan[0.042]; Vaughan[0.042]; Khan[0.042]; Younis[0.042]; Bowling[0.041]; day[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cairns ==> ENTITY: \u001b[32mChris Cairns\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.150:0.150[0]; log p(e|m)= -3.730:-3.730[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.045]; Doull[0.043]; Vaughan[0.043]; overs[0.043]; wicket[0.043]; Germon[0.042]; Germon[0.042]; Cricket[0.042]; Astle[0.042]; Astle[0.042]; Saqlain[0.042]; international[0.042]; second[0.042]; wickets[0.041]; Akram[0.041]; run[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Waqar[0.041]; Zealand[0.040]; Zealand[0.040]; Wasim[0.040]; Friday[0.040]; day[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Zealand ==> ENTITY: \u001b[32mNew Zealand national cricket team\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.184:0.184[0]; log p(e|m)= -3.147:-3.147[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaqlain[0.043]; Saqlain[0.043]; Doull[0.043]; Doull[0.043]; Parore[0.042]; Wasim[0.042]; Wasim[0.042]; Wasim[0.042]; Wasim[0.042]; Waqar[0.042]; Waqar[0.042]; Vaughan[0.042]; Vaughan[0.042]; Mushtaq[0.041]; Mushtaq[0.041]; Astle[0.041]; Astle[0.041]; Germon[0.041]; Cairns[0.041]; Cairns[0.041]; overs[0.041]; overs[0.041]; Fleming[0.040]; wickets[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Moin Khan\u001b[39m ==> ENTITY: \u001b[32mMoin Khan\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaqlain[0.044]; Saqlain[0.044]; Waqar[0.043]; Waqar[0.043]; Wasim[0.043]; Wasim[0.043]; Wasim[0.043]; Wasim[0.043]; Khan[0.042]; Khan[0.042]; Ijaz[0.042]; Ijaz[0.042]; Moin[0.042]; Moin[0.042]; Malik[0.041]; Malik[0.041]; Afridi[0.041]; Afridi[0.041]; innings[0.040]; Parore[0.040]; wickets[0.040]; wickets[0.040]; Doull[0.039]; Doull[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wasim Akram\u001b[39m ==> ENTITY: \u001b[32mWasim Akram\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.237:0.237[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWasim[0.045]; Wasim[0.045]; Wasim[0.045]; Mushtaq[0.043]; Mushtaq[0.043]; Ijaz[0.043]; Ijaz[0.043]; Ijaz[0.043]; Saqlain[0.042]; Saqlain[0.042]; Saqlain[0.042]; Pakistan[0.042]; Waqar[0.041]; Waqar[0.041]; Malik[0.041]; Malik[0.041]; match[0.040]; Khan[0.040]; Moin[0.040]; Afridi[0.040]; Afridi[0.040]; Afridi[0.040]; Younis[0.040]; Karachi[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Waqar Younis\u001b[39m ==> ENTITY: \u001b[32mWaqar Younis\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.274:0.274[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWasim[0.046]; Wasim[0.046]; Wasim[0.046]; Wasim[0.046]; Akram[0.044]; Waqar[0.043]; Saqlain[0.042]; Saqlain[0.042]; Pakistan[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Malik[0.042]; Malik[0.042]; Ijaz[0.041]; Ijaz[0.041]; match[0.041]; Afridi[0.040]; Afridi[0.040]; Afridi[0.040]; Moin[0.040]; Khan[0.040]; Karachi[0.039]; Shahid[0.038]; Vaughan[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Moin Khan\u001b[39m ==> ENTITY: \u001b[32mMoin Khan\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaqlain[0.046]; Akram[0.046]; Waqar[0.045]; Wasim[0.045]; Pakistan[0.045]; Pakistan[0.045]; Pakistan[0.045]; Pakistan[0.045]; Mushtaq[0.045]; Mushtaq[0.045]; cricket[0.045]; Ijaz[0.044]; Malik[0.043]; Afridi[0.043]; Cricket[0.043]; second[0.042]; Younis[0.042]; wickets[0.042]; Haq[0.042]; Doull[0.041]; overs[0.041]; Sialkot[0.040]; wicket[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pakistan ==> ENTITY: \u001b[32mPakistan national cricket team\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.201:0.201[0]; log p(e|m)= -2.749:-2.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.050]; Pakistan[0.050]; Pakistan[0.050]; cricket[0.045]; Akram[0.043]; Saqlain[0.042]; Wasim[0.042]; Waqar[0.041]; Malik[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Afridi[0.041]; Sialkot[0.041]; Ijaz[0.041]; Cricket[0.040]; Khan[0.040]; Vaughan[0.040]; Saeed[0.039]; Younis[0.039]; Moin[0.039]; overs[0.039]; Zahoor[0.038]; wicket[0.038]; day[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Waqar\u001b[39m ==> ENTITY: \u001b[32mWaqar Younis\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.259:0.259[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWasim[0.048]; Wasim[0.048]; Wasim[0.048]; Wasim[0.048]; Waqar[0.046]; Saqlain[0.045]; Saqlain[0.045]; Malik[0.044]; Malik[0.044]; Ijaz[0.044]; Ijaz[0.044]; Afridi[0.043]; Afridi[0.043]; Moin[0.042]; Moin[0.042]; Moin[0.042]; Khan[0.042]; Khan[0.042]; Khan[0.042]; Parore[0.041]; wicket[0.041]; Vaughan[0.040]; Vaughan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Moin Khan\u001b[39m ==> ENTITY: \u001b[32mMoin Khan\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaqlain[0.046]; Saqlain[0.046]; Waqar[0.045]; Waqar[0.045]; Wasim[0.045]; Wasim[0.045]; Wasim[0.045]; Wasim[0.045]; Khan[0.044]; Khan[0.044]; Ijaz[0.044]; Ijaz[0.044]; Moin[0.044]; Moin[0.044]; Malik[0.043]; Afridi[0.043]; Afridi[0.043]; innings[0.042]; Parore[0.042]; wickets[0.041]; wickets[0.041]; Doull[0.041]; Doull[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Astle ==> ENTITY: \u001b[32mNathan Astle\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.179:0.179[0]; log p(e|m)= -1.252:-1.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAstle[0.046]; cricket[0.045]; Vaughan[0.044]; Cricket[0.044]; Zealand[0.042]; Zealand[0.042]; wicket[0.042]; wickets[0.042]; second[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; overs[0.042]; run[0.041]; Saqlain[0.041]; Germon[0.041]; Germon[0.041]; Pakistan[0.040]; Pakistan[0.040]; Pakistan[0.040]; Pakistan[0.040]; day[0.040]; Cairns[0.040]; Cairns[0.040]; Waqar[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saqlain Mushtaq\u001b[39m ==> ENTITY: \u001b[32mSaqlain Mushtaq\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWasim[0.044]; Wasim[0.044]; Wasim[0.044]; Wasim[0.044]; Akram[0.044]; Mushtaq[0.043]; match[0.043]; Pakistan[0.043]; Ijaz[0.042]; Ijaz[0.042]; wickets[0.041]; Karachi[0.041]; Waqar[0.041]; Waqar[0.041]; overs[0.041]; Malik[0.040]; Malik[0.040]; Afridi[0.040]; Afridi[0.040]; lbw[0.040]; lbw[0.040]; Vaughan[0.040]; Khan[0.040]; Younis[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shahid Afridi\u001b[39m ==> ENTITY: \u001b[32mShahid Afridi\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.225:0.225[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAkram[0.044]; Saqlain[0.044]; Wasim[0.044]; Wasim[0.044]; Wasim[0.044]; Wasim[0.044]; Pakistan[0.043]; Afridi[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Waqar[0.042]; Waqar[0.042]; Ijaz[0.042]; Younis[0.042]; Malik[0.042]; match[0.041]; Karachi[0.040]; Moin[0.040]; Khan[0.040]; Vaughan[0.039]; lbw[0.039]; runs[0.039]; overs[0.038]; wickets[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Astle ==> ENTITY: \u001b[32mNathan Astle\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.182:0.182[0]; log p(e|m)= -1.252:-1.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAstle[0.045]; cricket[0.045]; Vaughan[0.044]; Vaughan[0.044]; Cricket[0.044]; Zealand[0.042]; Zealand[0.042]; wicket[0.042]; wickets[0.042]; second[0.042]; Mushtaq[0.041]; Mushtaq[0.041]; overs[0.041]; run[0.041]; Saqlain[0.041]; Germon[0.040]; Germon[0.040]; Pakistan[0.040]; Pakistan[0.040]; Pakistan[0.040]; Pakistan[0.040]; day[0.040]; Cairns[0.040]; Cairns[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Salim Malik\u001b[39m ==> ENTITY: \u001b[32mSaleem Malik\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.206:0.206[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.044]; Pakistan[0.044]; Pakistan[0.044]; Pakistan[0.044]; cricket[0.044]; Akram[0.043]; Saqlain[0.043]; Ijaz[0.042]; Wasim[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Khan[0.042]; Khan[0.042]; Younis[0.041]; Waqar[0.041]; Waqar[0.041]; Afridi[0.041]; Moin[0.040]; Moin[0.040]; overs[0.040]; Saeed[0.040]; second[0.039]; wickets[0.039]; Shahid[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spearman ==> ENTITY: \u001b[32mCraig Spearman\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.175:0.175[0]; log p(e|m)= -1.945:-1.945[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwicket[0.051]; wickets[0.050]; overs[0.049]; second[0.049]; Astle[0.049]; Astle[0.049]; Saqlain[0.048]; Wasim[0.048]; Mushtaq[0.048]; Mushtaq[0.048]; cricket[0.048]; Akram[0.047]; run[0.047]; Younis[0.047]; Waqar[0.046]; Vaughan[0.046]; Germon[0.046]; Germon[0.046]; Malik[0.046]; Cricket[0.046]; Khan[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wasim\u001b[39m ==> ENTITY: \u001b[32mWasim Akram\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWasim[0.046]; Wasim[0.046]; Wasim[0.046]; Ijaz[0.043]; Ijaz[0.043]; Saqlain[0.043]; Saqlain[0.043]; Waqar[0.042]; Waqar[0.042]; Malik[0.042]; Khan[0.041]; Khan[0.041]; Khan[0.041]; Moin[0.041]; Moin[0.041]; Moin[0.041]; Afridi[0.041]; Afridi[0.041]; Parore[0.040]; overs[0.040]; overs[0.040]; wicket[0.039]; Vaughan[0.039]; Vaughan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saqlain\u001b[39m ==> ENTITY: \u001b[32mSaqlain Mushtaq\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaqlain[0.047]; Wasim[0.046]; Wasim[0.046]; Wasim[0.046]; Wasim[0.046]; Wasim[0.046]; Akram[0.045]; Ijaz[0.043]; Ijaz[0.043]; wicket[0.043]; wickets[0.043]; wickets[0.043]; Waqar[0.043]; Waqar[0.043]; overs[0.043]; overs[0.043]; Malik[0.042]; Afridi[0.042]; Afridi[0.042]; lbw[0.042]; lbw[0.042]; Vaughan[0.042]; Vaughan[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPakistan\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPakistan national cricket team\u001b[39m <---> \u001b[32mPakistan\u001b[39m\t\nSCORES: global= 0.259:0.250[\u001b[31m0.009\u001b[39m]; local()= 0.198:0.213[\u001b[32m0.015\u001b[39m]; log p(e|m)= -2.749:-0.219[\u001b[32m2.529\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.050]; Pakistan[0.050]; Pakistan[0.050]; cricket[0.045]; Akram[0.043]; Saqlain[0.042]; Wasim[0.042]; Waqar[0.041]; Malik[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Afridi[0.041]; Sialkot[0.041]; Ijaz[0.041]; Cricket[0.040]; Khan[0.040]; Vaughan[0.040]; Saeed[0.039]; Younis[0.039]; Moin[0.039]; Zahoor[0.039]; day[0.038]; Shahid[0.038]; Haq[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Salim Malik\u001b[39m ==> ENTITY: \u001b[32mSaleem Malik\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.044]; Akram[0.044]; Saqlain[0.044]; Ijaz[0.043]; Wasim[0.043]; Wasim[0.043]; Wasim[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Khan[0.042]; Younis[0.042]; Waqar[0.042]; Waqar[0.042]; Afridi[0.042]; Afridi[0.042]; match[0.041]; Moin[0.041]; overs[0.041]; wickets[0.040]; Shahid[0.040]; day[0.040]; Karachi[0.039]; runs[0.039]; Doull[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pakistan ==> ENTITY: \u001b[32mPakistan national cricket team\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.192:0.192[0]; log p(e|m)= -2.749:-2.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAkram[0.045]; Saqlain[0.044]; Wasim[0.044]; Wasim[0.044]; match[0.044]; Karachi[0.044]; Waqar[0.043]; Waqar[0.043]; Malik[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; won[0.043]; Afridi[0.043]; Ijaz[0.042]; Younis[0.041]; overs[0.040]; Bowling[0.040]; day[0.040]; Shahid[0.040]; wickets[0.039]; runs[0.038]; Ahmad[0.038]; Ahmad[0.038]; Extras[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Zealand ==> ENTITY: \u001b[32mNew Zealand national cricket team\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.183:0.183[0]; log p(e|m)= -3.147:-3.147[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.048]; Cricket[0.046]; Saqlain[0.045]; Akram[0.044]; Pakistan[0.044]; Pakistan[0.044]; Pakistan[0.044]; Pakistan[0.044]; Wasim[0.044]; Waqar[0.044]; international[0.044]; Vaughan[0.044]; Zealand[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Astle[0.043]; Astle[0.043]; Germon[0.043]; Germon[0.043]; Cairns[0.042]; second[0.042]; day[0.042]; Malik[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saeed Anwar\u001b[39m ==> ENTITY: \u001b[32mSaeed Anwar\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.235:0.235[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.045]; Pakistan[0.045]; Pakistan[0.045]; Pakistan[0.045]; Wasim[0.044]; Saqlain[0.043]; Akram[0.043]; Waqar[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; cricket[0.042]; Ijaz[0.042]; Younis[0.041]; Moin[0.041]; Malik[0.041]; Afridi[0.040]; Cricket[0.040]; wicket[0.040]; Khan[0.040]; wickets[0.039]; Sialkot[0.039]; Vaughan[0.039]; Zahoor[0.039]; day[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Doull ==> ENTITY: \u001b[32mSimon Doull\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.088:0.088[0]; log p(e|m)= -1.720:-1.720[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCairns[0.044]; Cairns[0.044]; Cairns[0.044]; Zealand[0.043]; Parore[0.043]; Saqlain[0.042]; Saqlain[0.042]; Saqlain[0.042]; Fleming[0.042]; Germon[0.042]; Germon[0.042]; Germon[0.042]; Waqar[0.041]; Waqar[0.041]; Young[0.041]; lb[0.041]; Vaughan[0.041]; Vaughan[0.041]; wicket[0.041]; lbw[0.040]; lbw[0.040]; Fall[0.040]; Total[0.040]; Akram[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Afridi\u001b[39m ==> ENTITY: \u001b[32mShahid Afridi\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.240:0.240[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAkram[0.046]; Saqlain[0.045]; Saqlain[0.045]; Wasim[0.045]; Wasim[0.045]; Wasim[0.045]; Wasim[0.045]; Wasim[0.045]; Afridi[0.044]; Waqar[0.044]; Waqar[0.044]; Waqar[0.044]; Ijaz[0.043]; Ijaz[0.043]; Younis[0.043]; Malik[0.043]; Moin[0.042]; Moin[0.042]; Moin[0.042]; Khan[0.041]; Khan[0.041]; Khan[0.041]; Parore[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saqlain\u001b[39m ==> ENTITY: \u001b[32mSaqlain Mushtaq\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaqlain[0.045]; Wasim[0.044]; Wasim[0.044]; Wasim[0.044]; Wasim[0.044]; Wasim[0.044]; Akram[0.044]; Ijaz[0.042]; Ijaz[0.042]; wickets[0.041]; Waqar[0.041]; Waqar[0.041]; Waqar[0.041]; overs[0.041]; Malik[0.040]; Afridi[0.040]; Afridi[0.040]; lbw[0.040]; lbw[0.040]; Vaughan[0.040]; Vaughan[0.040]; Khan[0.040]; Khan[0.040]; Khan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wasim Akram\u001b[39m ==> ENTITY: \u001b[32mWasim Akram\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.208:0.208[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMushtaq[0.044]; Mushtaq[0.044]; Ijaz[0.044]; Saqlain[0.044]; Pakistan[0.043]; Pakistan[0.043]; Pakistan[0.043]; Pakistan[0.043]; cricket[0.043]; Waqar[0.043]; Malik[0.043]; Khan[0.041]; Moin[0.041]; Afridi[0.041]; Younis[0.041]; Cricket[0.041]; Saeed[0.041]; overs[0.040]; Shahid[0.040]; wicket[0.040]; Vaughan[0.040]; wickets[0.040]; Sialkot[0.039]; Zahoor[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Karachi\u001b[39m ==> ENTITY: \u001b[32mKarachi\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.049]; Wasim[0.046]; Malik[0.045]; Ijaz[0.044]; Akram[0.044]; Saqlain[0.043]; match[0.043]; Waqar[0.042]; Afridi[0.042]; Younis[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Shahid[0.041]; Ahmad[0.041]; Ahmad[0.041]; wickets[0.040]; runs[0.040]; day[0.040]; overs[0.040]; December[0.039]; Salim[0.039]; won[0.038]; Bowling[0.038]; Total[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Zealand ==> ENTITY: \u001b[32mNew Zealand national cricket team\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.182:0.182[0]; log p(e|m)= -3.147:-3.147[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.046]; Cricket[0.044]; Saqlain[0.043]; Akram[0.042]; Pakistan[0.042]; Pakistan[0.042]; Pakistan[0.042]; Pakistan[0.042]; Wasim[0.042]; Waqar[0.042]; international[0.042]; Vaughan[0.042]; Zealand[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Astle[0.041]; Astle[0.041]; Germon[0.041]; Germon[0.041]; Cairns[0.041]; overs[0.040]; wickets[0.040]; wicket[0.040]; second[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Waqar Younis\u001b[39m ==> ENTITY: \u001b[32mWaqar Younis\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.239:0.239[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWasim[0.047]; Akram[0.046]; Saqlain[0.043]; Pakistan[0.043]; Pakistan[0.043]; Pakistan[0.043]; Pakistan[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Malik[0.043]; cricket[0.043]; Ijaz[0.042]; Cricket[0.042]; Afridi[0.041]; Moin[0.041]; Khan[0.041]; Sialkot[0.040]; wicket[0.039]; Shahid[0.039]; Vaughan[0.039]; Vaughan[0.039]; Saeed[0.039]; wickets[0.039]; Bowling[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = GMT\u001b[39m ==> ENTITY: \u001b[32mGreenwich Mean Time\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.038:0.038[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFriday[0.046]; day[0.046]; international[0.044]; Scoreboard[0.043]; Scoreboard[0.043]; Pakistan[0.043]; Pakistan[0.043]; Pakistan[0.043]; Pakistan[0.043]; Extras[0.043]; lb[0.042]; Corrected[0.042]; Day[0.041]; run[0.041]; Saqlain[0.041]; Cairns[0.041]; second[0.040]; Germon[0.040]; Germon[0.040]; Ahmad[0.039]; Ahmad[0.039]; Vaughan[0.039]; Zealand[0.039]; Zealand[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cairns ==> ENTITY: \u001b[32mChris Cairns\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.144:0.144[0]; log p(e|m)= -3.730:-3.730[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCairns[0.045]; Parore[0.044]; Doull[0.043]; Doull[0.043]; Fleming[0.043]; Vaughan[0.042]; Vaughan[0.042]; overs[0.042]; wicket[0.042]; Germon[0.041]; Germon[0.041]; Astle[0.041]; Astle[0.041]; Astle[0.041]; Astle[0.041]; Saqlain[0.041]; Saqlain[0.041]; Saqlain[0.041]; wickets[0.041]; Akram[0.041]; Mushtaq[0.040]; Mushtaq[0.040]; lbw[0.040]; lbw[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Salim Malik\u001b[39m ==> ENTITY: \u001b[32mSaleem Malik\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAkram[0.047]; Saqlain[0.047]; Saqlain[0.047]; Saqlain[0.047]; Ijaz[0.046]; Ijaz[0.046]; Wasim[0.046]; Wasim[0.046]; Wasim[0.046]; Wasim[0.046]; Wasim[0.046]; Mushtaq[0.045]; Mushtaq[0.045]; Khan[0.045]; Khan[0.045]; Khan[0.045]; Younis[0.045]; Waqar[0.045]; Waqar[0.045]; Waqar[0.045]; Afridi[0.044]; Afridi[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wasim\u001b[39m ==> ENTITY: \u001b[32mWasim Akram\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.242:0.242[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAkram[0.045]; Wasim[0.044]; Wasim[0.044]; Wasim[0.044]; Wasim[0.044]; Mushtaq[0.043]; Mushtaq[0.043]; Ijaz[0.042]; Ijaz[0.042]; Saqlain[0.042]; Saqlain[0.042]; Saqlain[0.042]; Waqar[0.041]; Waqar[0.041]; Waqar[0.041]; Malik[0.041]; Khan[0.040]; Khan[0.040]; Khan[0.040]; Moin[0.039]; Moin[0.039]; Moin[0.039]; Afridi[0.039]; Afridi[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Germon\u001b[39m ==> ENTITY: \u001b[32mLee Germon\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.046]; Germon[0.046]; Vaughan[0.045]; Vaughan[0.045]; Cricket[0.045]; Doull[0.044]; Zealand[0.044]; Zealand[0.044]; wickets[0.043]; wicket[0.043]; overs[0.043]; Cairns[0.043]; Cairns[0.043]; Harris[0.043]; Harris[0.043]; Harris[0.043]; Harris[0.043]; Harris[0.043]; Harris[0.043]; second[0.043]; Kennedy[0.042]; Saqlain[0.042]; Friday[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zahoor Elahi\u001b[39m ==> ENTITY: \u001b[32mZahoor Elahi\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.250:0.250[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.046]; Pakistan[0.046]; Pakistan[0.046]; Pakistan[0.046]; Sialkot[0.043]; Ijaz[0.043]; Waqar[0.043]; cricket[0.043]; Saqlain[0.042]; Wasim[0.042]; Malik[0.041]; Cricket[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Khan[0.041]; Younis[0.041]; Akram[0.041]; Afridi[0.040]; Saeed[0.039]; Vaughan[0.039]; Moin[0.038]; Shahid[0.038]; wickets[0.038]; Haq[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pakistan ==> ENTITY: \u001b[32mPakistan national cricket team\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.198:0.198[0]; log p(e|m)= -2.749:-2.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.050]; Pakistan[0.050]; Pakistan[0.050]; cricket[0.045]; Akram[0.043]; Saqlain[0.043]; Wasim[0.042]; Waqar[0.042]; Mushtaq[0.041]; Mushtaq[0.041]; Afridi[0.041]; Sialkot[0.041]; Ijaz[0.041]; Cricket[0.041]; Khan[0.040]; Vaughan[0.040]; Saeed[0.039]; Younis[0.039]; Moin[0.039]; Zahoor[0.039]; day[0.038]; Shahid[0.038]; Haq[0.038]; second[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sialkot\u001b[39m ==> ENTITY: \u001b[32mSialkot\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.230:0.230[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.045]; Pakistan[0.045]; Pakistan[0.045]; Pakistan[0.045]; Malik[0.045]; Akram[0.042]; Ijaz[0.042]; Wasim[0.042]; Khan[0.042]; Waqar[0.042]; Saqlain[0.041]; Elahi[0.041]; cricket[0.041]; Afridi[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Saeed[0.041]; Moin[0.040]; Zahoor[0.040]; Younis[0.040]; Shahid[0.040]; Haq[0.039]; Ahmad[0.039]; Ahmad[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Afridi\u001b[39m ==> ENTITY: \u001b[32mShahid Afridi\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.247:0.247[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAkram[0.045]; Saqlain[0.045]; Saqlain[0.045]; Saqlain[0.045]; Wasim[0.045]; Wasim[0.045]; Wasim[0.045]; Wasim[0.045]; Wasim[0.045]; Afridi[0.043]; Afridi[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Waqar[0.043]; Waqar[0.043]; Waqar[0.043]; Ijaz[0.043]; Ijaz[0.043]; Younis[0.043]; Malik[0.042]; Moin[0.041]; Moin[0.041]; Moin[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Moin Khan\u001b[39m ==> ENTITY: \u001b[32mMoin Khan\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaqlain[0.047]; Saqlain[0.047]; Saqlain[0.047]; Akram[0.046]; Waqar[0.046]; Waqar[0.046]; Waqar[0.046]; Wasim[0.046]; Wasim[0.046]; Wasim[0.046]; Wasim[0.046]; Wasim[0.046]; Mushtaq[0.046]; Mushtaq[0.046]; Khan[0.045]; Khan[0.045]; Ijaz[0.045]; Ijaz[0.045]; Moin[0.044]; Moin[0.044]; Malik[0.044]; Malik[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shahid Afridi\u001b[39m ==> ENTITY: \u001b[32mShahid Afridi\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAkram[0.045]; Saqlain[0.044]; Wasim[0.044]; Pakistan[0.043]; Pakistan[0.043]; Pakistan[0.043]; Pakistan[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Waqar[0.043]; Ijaz[0.043]; Younis[0.042]; Malik[0.042]; cricket[0.042]; Moin[0.041]; Cricket[0.040]; Khan[0.040]; Vaughan[0.040]; Vaughan[0.040]; wicket[0.040]; Saeed[0.039]; Sialkot[0.039]; second[0.039]; overs[0.038]; \t\n\r [===============================>..............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 6s769ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3199/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1291testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Santa ==> ENTITY: \u001b[32mSanta Claus\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.329:-0.329[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSanta[0.082]; Claus[0.053]; Nicholas[0.043]; presents[0.043]; presents[0.043]; northern[0.042]; children[0.041]; small[0.041]; city[0.041]; nearly[0.039]; nearly[0.039]; distributing[0.039]; disguise[0.039]; robber[0.038]; traditionally[0.038]; rounds[0.038]; day[0.038]; Friday[0.038]; Saint[0.038]; gets[0.037]; receive[0.037]; Hanover[0.037]; Hanover[0.037]; man[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGerman[0.045]; German[0.045]; German[0.045]; nearly[0.044]; nearly[0.044]; northern[0.043]; Hanover[0.043]; Hanover[0.043]; Friday[0.042]; bank[0.042]; bank[0.042]; traditionally[0.042]; rounds[0.042]; day[0.041]; city[0.041]; ended[0.040]; small[0.040]; receive[0.040]; Nicholas[0.040]; thought[0.040]; eventually[0.039]; gets[0.039]; Claus[0.039]; children[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.138:0.138[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.050]; German[0.049]; German[0.049]; northern[0.045]; small[0.043]; nearly[0.042]; nearly[0.042]; traditionally[0.042]; children[0.041]; Claus[0.041]; Santa[0.041]; Santa[0.041]; receive[0.041]; city[0.040]; Hanover[0.040]; Hanover[0.040]; workers[0.040]; patrol[0.040]; eventually[0.040]; man[0.039]; Friday[0.039]; thought[0.039]; day[0.039]; ended[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hanover ==> ENTITY: \u001b[32mHanover\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.707:-0.707[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHanover[0.047]; Germany[0.047]; nearly[0.044]; nearly[0.044]; northern[0.043]; small[0.043]; city[0.043]; German[0.043]; German[0.043]; German[0.043]; traditionally[0.042]; Friday[0.041]; bank[0.041]; bank[0.041]; receive[0.041]; day[0.040]; police[0.040]; police[0.040]; eventually[0.040]; Saint[0.040]; ended[0.040]; patrol[0.039]; man[0.039]; Nicholas[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saint Nicholas\u001b[39m ==> ENTITY: \u001b[32mSaint Nicholas\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.115:0.115[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSanta[0.048]; Santa[0.048]; Claus[0.044]; traditionally[0.043]; presents[0.043]; presents[0.043]; Germany[0.042]; day[0.042]; children[0.041]; disguise[0.041]; city[0.041]; small[0.041]; receive[0.041]; German[0.041]; German[0.041]; German[0.041]; eventually[0.041]; Friday[0.040]; nearly[0.040]; nearly[0.040]; thought[0.040]; distributing[0.040]; northern[0.039]; gets[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hanover ==> ENTITY: \u001b[32mHanover\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.707:-0.707[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHanover[0.047]; Germany[0.047]; nearly[0.044]; nearly[0.044]; northern[0.043]; small[0.043]; city[0.043]; German[0.043]; German[0.043]; German[0.043]; Friday[0.041]; bank[0.041]; bank[0.041]; traditionally[0.041]; receive[0.041]; day[0.040]; police[0.040]; police[0.040]; eventually[0.040]; Nicholas[0.040]; Saint[0.040]; ended[0.040]; patrol[0.040]; children[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.135:0.135[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.050]; German[0.049]; German[0.049]; northern[0.045]; small[0.043]; nearly[0.042]; nearly[0.042]; traditionally[0.042]; children[0.041]; Claus[0.041]; Santa[0.041]; Santa[0.041]; receive[0.041]; city[0.040]; Nicholas[0.040]; Hanover[0.040]; Hanover[0.040]; workers[0.040]; man[0.040]; eventually[0.040]; thought[0.039]; Friday[0.039]; day[0.039]; ended[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Santa Claus\u001b[39m ==> ENTITY: \u001b[32mSanta Claus\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.106:0.106[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSanta[0.048]; Saint[0.046]; Nicholas[0.044]; traditionally[0.043]; presents[0.043]; presents[0.043]; nearly[0.042]; nearly[0.042]; city[0.042]; Germany[0.042]; day[0.042]; bars[0.041]; small[0.041]; German[0.041]; German[0.041]; German[0.041]; Friday[0.040]; receive[0.040]; distributing[0.040]; children[0.040]; disguise[0.040]; northern[0.039]; gets[0.039]; eventually[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.140:0.140[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.051]; German[0.049]; German[0.049]; northern[0.043]; small[0.042]; nearly[0.042]; nearly[0.042]; patrol[0.041]; traditionally[0.041]; city[0.041]; Nicholas[0.041]; eventually[0.041]; Hanover[0.041]; Hanover[0.041]; man[0.040]; children[0.040]; Friday[0.040]; Claus[0.040]; day[0.040]; police[0.039]; police[0.039]; bank[0.039]; bank[0.039]; receive[0.039]; \t\n\r [===============================>..............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 6s719ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3208/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1364testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.003:0.003[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.047]; Ireland[0.047]; matches[0.046]; Division[0.044]; Coleraine[0.044]; Coleraine[0.044]; Soccer[0.044]; played[0.042]; won[0.041]; Ards[0.041]; Ards[0.041]; Glentoran[0.041]; Glentoran[0.041]; Cliftonville[0.040]; Cliftonville[0.040]; Portadown[0.040]; Portadown[0.040]; goals[0.040]; goals[0.040]; Premier[0.040]; division[0.040]; Saturday[0.040]; Glenavon[0.039]; Glenavon[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Glenavon\u001b[39m ==> ENTITY: \u001b[32mGlenavon F.C.\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGlentoran[0.044]; Glentoran[0.044]; Coleraine[0.043]; Coleraine[0.043]; played[0.043]; Ireland[0.043]; Ireland[0.043]; Cliftonville[0.043]; Cliftonville[0.043]; Portadown[0.042]; Portadown[0.042]; matches[0.042]; Glenavon[0.042]; won[0.042]; Ards[0.041]; Ards[0.041]; division[0.041]; Soccer[0.040]; Linfield[0.040]; Linfield[0.040]; Division[0.040]; Crusaders[0.040]; Crusaders[0.040]; Premier[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portadown ==> ENTITY: \u001b[32mPortadown F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.235:0.235[0]; log p(e|m)= -0.594:-0.594[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nColeraine[0.044]; Coleraine[0.044]; Portadown[0.044]; Glentoran[0.044]; Glentoran[0.044]; Cliftonville[0.043]; Cliftonville[0.043]; Ireland[0.043]; Ireland[0.043]; Ards[0.043]; Ards[0.043]; matches[0.042]; played[0.041]; Glenavon[0.041]; Glenavon[0.041]; Linfield[0.041]; Linfield[0.041]; won[0.040]; Soccer[0.039]; Crusaders[0.039]; Crusaders[0.039]; goals[0.038]; goals[0.038]; Premier[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Crusaders ==> ENTITY: \u001b[32mCrusaders F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.198:0.198[0]; log p(e|m)= -2.025:-2.025[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.043]; Glentoran[0.043]; Glentoran[0.043]; Coleraine[0.043]; Coleraine[0.043]; Crusaders[0.043]; Ireland[0.043]; Ireland[0.043]; Ards[0.042]; Ards[0.042]; played[0.042]; Portadown[0.042]; Portadown[0.042]; Cliftonville[0.041]; Cliftonville[0.041]; Linfield[0.041]; Linfield[0.041]; Glenavon[0.041]; Glenavon[0.041]; Saturday[0.041]; Soccer[0.040]; London[0.040]; division[0.040]; won[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Linfield\u001b[39m ==> ENTITY: \u001b[32mLinfield F.C.\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.221:0.221[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGlentoran[0.044]; Glentoran[0.044]; Portadown[0.044]; Portadown[0.044]; Ireland[0.043]; Ireland[0.043]; matches[0.043]; Cliftonville[0.043]; Cliftonville[0.043]; Coleraine[0.043]; Coleraine[0.043]; played[0.043]; won[0.042]; Linfield[0.042]; Ards[0.041]; Ards[0.041]; Soccer[0.041]; Glenavon[0.040]; Glenavon[0.040]; Crusaders[0.039]; Crusaders[0.039]; division[0.039]; Premier[0.039]; Saturday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cliftonville ==> ENTITY: \u001b[32mCliftonville F.C.\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.188:0.188[0]; log p(e|m)= -0.033:-0.033[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCliftonville[0.044]; matches[0.043]; Ireland[0.043]; Ireland[0.043]; Coleraine[0.043]; Coleraine[0.043]; Glentoran[0.042]; Glentoran[0.042]; Ards[0.042]; Ards[0.042]; Portadown[0.042]; Portadown[0.042]; played[0.042]; won[0.042]; Northern[0.041]; Northern[0.041]; Division[0.040]; Premier[0.040]; Glenavon[0.040]; Glenavon[0.040]; Linfield[0.040]; Linfield[0.040]; Soccer[0.040]; London[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cliftonville ==> ENTITY: \u001b[32mCliftonville F.C.\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.199:0.199[0]; log p(e|m)= -0.033:-0.033[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCliftonville[0.046]; matches[0.045]; Ireland[0.045]; Coleraine[0.045]; Coleraine[0.045]; Glentoran[0.044]; Glentoran[0.044]; Ards[0.044]; Ards[0.044]; Portadown[0.044]; Portadown[0.044]; played[0.044]; won[0.044]; Northern[0.043]; Division[0.042]; Premier[0.042]; Glenavon[0.042]; Glenavon[0.042]; Linfield[0.042]; Linfield[0.042]; London[0.042]; Saturday[0.042]; division[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Glenavon\u001b[39m ==> ENTITY: \u001b[32mGlenavon F.C.\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGlentoran[0.044]; Glentoran[0.044]; Coleraine[0.043]; Coleraine[0.043]; played[0.043]; Ireland[0.043]; Ireland[0.043]; Cliftonville[0.043]; Cliftonville[0.043]; Portadown[0.042]; Portadown[0.042]; matches[0.042]; Glenavon[0.042]; won[0.042]; Ards[0.041]; Ards[0.041]; division[0.041]; Soccer[0.040]; Linfield[0.040]; Linfield[0.040]; Division[0.040]; Crusaders[0.040]; Crusaders[0.040]; Premier[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Glentoran\u001b[39m ==> ENTITY: \u001b[32mGlentoran F.C.\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.245:0.245[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGlentoran[0.046]; Ireland[0.045]; Ireland[0.045]; Portadown[0.044]; Portadown[0.044]; Cliftonville[0.044]; Cliftonville[0.044]; matches[0.043]; Coleraine[0.043]; Coleraine[0.043]; Ards[0.043]; Ards[0.043]; Linfield[0.042]; Linfield[0.042]; Glenavon[0.041]; Glenavon[0.041]; Soccer[0.041]; won[0.039]; played[0.039]; Crusaders[0.038]; Crusaders[0.038]; goals[0.038]; goals[0.038]; Premier[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNorthern Ireland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNIFL Premiership\u001b[39m <---> \u001b[32mNorthern Ireland\u001b[39m\t\nSCORES: global= 0.248:0.244[\u001b[31m0.004\u001b[39m]; local()= 0.214:0.144[\u001b[31m0.071\u001b[39m]; log p(e|m)= -3.411:-0.251[\u001b[32m3.160\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.047]; matches[0.044]; Glentoran[0.043]; Glentoran[0.043]; Coleraine[0.043]; Coleraine[0.043]; Cliftonville[0.043]; Portadown[0.042]; Portadown[0.042]; Linfield[0.042]; Linfield[0.042]; played[0.042]; Ards[0.042]; Ards[0.042]; Soccer[0.041]; won[0.041]; Glenavon[0.040]; Glenavon[0.040]; division[0.040]; Division[0.040]; Saturday[0.039]; goals[0.039]; goals[0.039]; points[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ards ==> ENTITY: \u001b[32mArds F.C.\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.222:0.222[0]; log p(e|m)= -0.270:-0.270[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArds[0.048]; Coleraine[0.046]; Coleraine[0.046]; Glentoran[0.046]; Glentoran[0.046]; Ireland[0.046]; Ireland[0.046]; Portadown[0.045]; Portadown[0.045]; Cliftonville[0.044]; Cliftonville[0.044]; Glenavon[0.043]; Glenavon[0.043]; Northern[0.042]; Northern[0.042]; Soccer[0.042]; matches[0.042]; Linfield[0.041]; Linfield[0.041]; Division[0.041]; played[0.041]; won[0.040]; Premier[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Glentoran\u001b[39m ==> ENTITY: \u001b[32mGlentoran F.C.\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.245:0.245[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGlentoran[0.046]; Ireland[0.045]; Ireland[0.045]; Portadown[0.044]; Portadown[0.044]; Cliftonville[0.044]; Cliftonville[0.044]; matches[0.043]; Coleraine[0.043]; Coleraine[0.043]; Ards[0.043]; Ards[0.043]; Linfield[0.042]; Linfield[0.042]; Glenavon[0.041]; Glenavon[0.041]; Soccer[0.041]; won[0.039]; played[0.039]; Crusaders[0.038]; Crusaders[0.038]; goals[0.038]; goals[0.038]; Premier[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Coleraine ==> ENTITY: \u001b[32mColeraine F.C.\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.195:0.195[0]; log p(e|m)= -1.772:-1.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nColeraine[0.049]; Ireland[0.045]; Ireland[0.045]; Portadown[0.043]; Portadown[0.043]; Glentoran[0.043]; Glentoran[0.043]; Cliftonville[0.043]; Cliftonville[0.043]; matches[0.042]; Ards[0.042]; Ards[0.042]; Glenavon[0.041]; Glenavon[0.041]; won[0.041]; Soccer[0.041]; played[0.040]; Saturday[0.040]; Division[0.040]; Linfield[0.040]; Linfield[0.040]; Northern[0.039]; Northern[0.039]; Premier[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portadown ==> ENTITY: \u001b[32mPortadown F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.235:0.235[0]; log p(e|m)= -0.594:-0.594[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nColeraine[0.044]; Coleraine[0.044]; Portadown[0.044]; Glentoran[0.044]; Glentoran[0.044]; Cliftonville[0.043]; Cliftonville[0.043]; Ireland[0.043]; Ireland[0.043]; Ards[0.043]; Ards[0.043]; matches[0.042]; played[0.041]; Glenavon[0.041]; Glenavon[0.041]; Linfield[0.041]; Linfield[0.041]; won[0.040]; Soccer[0.039]; Crusaders[0.039]; Crusaders[0.039]; goals[0.038]; goals[0.038]; Premier[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ards ==> ENTITY: \u001b[32mArds F.C.\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.222:0.222[0]; log p(e|m)= -0.270:-0.270[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArds[0.048]; Coleraine[0.046]; Coleraine[0.046]; Glentoran[0.046]; Glentoran[0.046]; Ireland[0.046]; Ireland[0.046]; Portadown[0.045]; Portadown[0.045]; Cliftonville[0.044]; Cliftonville[0.044]; Glenavon[0.043]; Glenavon[0.043]; Northern[0.042]; Northern[0.042]; Soccer[0.042]; matches[0.042]; Linfield[0.041]; Linfield[0.041]; Division[0.041]; played[0.041]; won[0.040]; Premier[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNorthern Ireland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNIFL Premiership\u001b[39m <---> \u001b[32mNorthern Ireland\u001b[39m\t\nSCORES: global= 0.248:0.244[\u001b[31m0.004\u001b[39m]; local()= 0.215:0.146[\u001b[31m0.069\u001b[39m]; log p(e|m)= -3.411:-0.251[\u001b[32m3.160\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.047]; matches[0.043]; Glentoran[0.043]; Glentoran[0.043]; Coleraine[0.043]; Coleraine[0.043]; Cliftonville[0.042]; Cliftonville[0.042]; Portadown[0.042]; Portadown[0.042]; Linfield[0.042]; Linfield[0.042]; played[0.042]; Ards[0.042]; Ards[0.042]; Soccer[0.041]; won[0.041]; Glenavon[0.040]; Glenavon[0.040]; division[0.040]; Division[0.040]; Saturday[0.039]; goals[0.039]; goals[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Coleraine ==> ENTITY: \u001b[32mColeraine F.C.\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.195:0.195[0]; log p(e|m)= -1.772:-1.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nColeraine[0.049]; Ireland[0.045]; Ireland[0.045]; Portadown[0.043]; Portadown[0.043]; Glentoran[0.043]; Glentoran[0.043]; Cliftonville[0.043]; Cliftonville[0.043]; matches[0.042]; Ards[0.042]; Ards[0.042]; Glenavon[0.041]; Glenavon[0.041]; won[0.041]; Soccer[0.041]; played[0.040]; Saturday[0.040]; Division[0.040]; Linfield[0.040]; Linfield[0.040]; Northern[0.039]; Northern[0.039]; Premier[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Crusaders ==> ENTITY: \u001b[32mCrusaders F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.198:0.198[0]; log p(e|m)= -2.025:-2.025[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.043]; Glentoran[0.043]; Glentoran[0.043]; Coleraine[0.043]; Coleraine[0.043]; Crusaders[0.043]; Ireland[0.043]; Ireland[0.043]; Ards[0.042]; Ards[0.042]; played[0.042]; Portadown[0.042]; Portadown[0.042]; Cliftonville[0.041]; Cliftonville[0.041]; Linfield[0.041]; Linfield[0.041]; Glenavon[0.041]; Glenavon[0.041]; Saturday[0.041]; Soccer[0.040]; London[0.040]; division[0.040]; won[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Linfield\u001b[39m ==> ENTITY: \u001b[32mLinfield F.C.\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.221:0.221[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGlentoran[0.044]; Glentoran[0.044]; Portadown[0.044]; Portadown[0.044]; Ireland[0.043]; Ireland[0.043]; matches[0.043]; Cliftonville[0.043]; Cliftonville[0.043]; Coleraine[0.043]; Coleraine[0.043]; played[0.043]; won[0.042]; Linfield[0.042]; Ards[0.041]; Ards[0.041]; Soccer[0.041]; Glenavon[0.040]; Glenavon[0.040]; Crusaders[0.039]; Crusaders[0.039]; division[0.039]; Premier[0.039]; Saturday[0.038]; \t\n\r [================================>.............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 6s567ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3227/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1182testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Melbourne ==> ENTITY: \u001b[32mMelbourne\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.168:-0.168[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; Australia[0.048]; match[0.048]; Australian[0.045]; Australian[0.045]; Cricket[0.044]; captain[0.043]; batsman[0.042]; Tour[0.041]; struggled[0.041]; Friday[0.040]; Brian[0.040]; disciplined[0.040]; tour[0.040]; tour[0.040]; Ian[0.040]; bat[0.040]; ball[0.040]; misconduct[0.040]; overs[0.039]; earned[0.039]; West[0.039]; West[0.039]; Healy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Indies ==> ENTITY: \u001b[32mWest Indies cricket team\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.192:0.192[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.043]; cricket[0.043]; Lara[0.043]; Lara[0.043]; Lara[0.043]; Australia[0.043]; Australia[0.043]; Australia[0.043]; Australia[0.043]; Cricket[0.042]; series[0.042]; score[0.042]; score[0.042]; second[0.041]; triangular[0.041]; Pakistan[0.041]; team[0.041]; team[0.041]; innings[0.040]; Ground[0.040]; win[0.040]; Indies[0.040]; tournament[0.040]; played[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lloyd\u001b[39m ==> ENTITY: \u001b[32mClive Lloyd\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.046]; Cricket[0.045]; Cricket[0.045]; Marsh[0.044]; Lara[0.044]; Lara[0.044]; Lara[0.044]; Lara[0.044]; Lara[0.044]; match[0.044]; Clive[0.044]; bat[0.043]; toss[0.043]; Australia[0.043]; Australia[0.043]; Australia[0.043]; Australia[0.043]; Australia[0.043]; Lloyd[0.043]; captain[0.042]; wicket[0.042]; second[0.042]; Geoff[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.205:0.205[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.050]; Australia[0.050]; Australia[0.050]; cricket[0.048]; match[0.045]; Cricket[0.045]; team[0.045]; team[0.045]; Sydney[0.045]; Pakistan[0.043]; series[0.043]; tournament[0.043]; played[0.042]; week[0.042]; win[0.041]; Lara[0.040]; Lara[0.040]; Lara[0.040]; Tuesday[0.040]; highest[0.040]; World[0.040]; failed[0.040]; triangular[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Melbourne Cricket Ground\u001b[39m ==> ENTITY: \u001b[32mMelbourne Cricket Ground\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.047]; Australia[0.046]; Australia[0.046]; Australia[0.046]; second[0.043]; play[0.042]; captain[0.042]; captain[0.042]; crowd[0.042]; crowd[0.042]; Australian[0.042]; series[0.041]; toss[0.041]; Wednesday[0.041]; Tuesday[0.040]; Friday[0.040]; struggled[0.040]; mid[0.040]; tour[0.040]; won[0.040]; Warne[0.040]; innings[0.040]; bat[0.040]; bat[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ian Healy\u001b[39m ==> ENTITY: \u001b[32mIan Healy\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.054]; Australia[0.054]; play[0.049]; batsman[0.049]; wicketkeeper[0.048]; Australian[0.048]; Australian[0.048]; Australian[0.048]; match[0.048]; captain[0.048]; captain[0.048]; Warne[0.047]; Cricket[0.047]; Cricket[0.047]; bat[0.046]; bat[0.046]; Melbourne[0.045]; Melbourne[0.045]; toss[0.045]; vice[0.045]; dismissed[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lara\u001b[39m ==> ENTITY: \u001b[32mBrian Lara\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.215:0.215[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.046]; Lara[0.045]; Lara[0.045]; match[0.044]; Cricket[0.044]; Marsh[0.042]; series[0.042]; Australia[0.041]; Australia[0.041]; Australia[0.041]; Australia[0.041]; Australia[0.041]; century[0.041]; runs[0.041]; Pakistan[0.041]; record[0.041]; tournament[0.041]; score[0.040]; score[0.040]; innings[0.040]; Indies[0.040]; Indies[0.040]; highest[0.040]; Lloyd[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.190:0.190[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.049]; match[0.045]; batsman[0.044]; Cricket[0.044]; Cricket[0.044]; Australian[0.043]; Australian[0.043]; Melbourne[0.042]; Melbourne[0.042]; captain[0.041]; captain[0.041]; bat[0.041]; wicketkeeper[0.041]; Warne[0.041]; overs[0.041]; overs[0.041]; play[0.040]; wicket[0.040]; struggled[0.040]; Lara[0.040]; Lara[0.040]; Lara[0.040]; suffered[0.039]; mid[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lara\u001b[39m ==> ENTITY: \u001b[32mBrian Lara\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.227:0.227[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.045]; Lara[0.044]; Lara[0.044]; Lara[0.044]; match[0.043]; Cricket[0.043]; Cricket[0.043]; toss[0.043]; Warne[0.042]; wicket[0.042]; wicket[0.042]; Marsh[0.041]; bat[0.041]; series[0.041]; series[0.041]; Australia[0.041]; Australia[0.041]; Australia[0.041]; overs[0.040]; runs[0.040]; record[0.040]; score[0.040]; Clive[0.039]; innings[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.205:0.205[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.047]; Australia[0.047]; Australia[0.047]; Australia[0.047]; cricket[0.045]; match[0.043]; Cricket[0.043]; team[0.043]; team[0.043]; Sydney[0.042]; Pakistan[0.041]; series[0.041]; series[0.041]; tournament[0.041]; played[0.040]; second[0.040]; Marsh[0.039]; week[0.039]; coach[0.039]; win[0.039]; Lara[0.038]; Lara[0.038]; Lara[0.038]; Lara[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lara\u001b[39m ==> ENTITY: \u001b[32mBrian Lara\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLara[0.045]; Lara[0.045]; Cricket[0.044]; toss[0.044]; Warne[0.043]; wicket[0.042]; wicket[0.042]; Marsh[0.042]; bat[0.042]; bat[0.042]; series[0.042]; wicketkeeper[0.042]; Australia[0.041]; Australia[0.041]; overs[0.041]; record[0.041]; Ian[0.041]; run[0.040]; Clive[0.040]; innings[0.040]; Indies[0.040]; Indies[0.040]; Lloyd[0.040]; Lloyd[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tom Moody\u001b[39m ==> ENTITY: \u001b[32mTom Moody\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.047]; Cricket[0.045]; Australia[0.045]; Australia[0.045]; overs[0.045]; overs[0.045]; bat[0.044]; bat[0.044]; match[0.044]; wicketkeeper[0.044]; innings[0.043]; toss[0.043]; Warne[0.043]; captain[0.043]; captain[0.043]; Wednesday[0.043]; wicket[0.043]; wicket[0.043]; Melbourne[0.042]; Melbourne[0.042]; Australian[0.042]; Australian[0.042]; mid[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pakistan ==> ENTITY: \u001b[32mPakistan national cricket team\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.184:0.184[0]; log p(e|m)= -2.749:-2.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseries[0.048]; cricket[0.048]; team[0.045]; team[0.045]; match[0.045]; tournament[0.044]; Cricket[0.043]; Australia[0.043]; Australia[0.043]; Australia[0.043]; Australia[0.043]; played[0.042]; test[0.040]; test[0.040]; triangular[0.040]; innings[0.040]; win[0.039]; highest[0.039]; score[0.039]; score[0.039]; runs[0.039]; week[0.038]; West[0.038]; West[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lara\u001b[39m ==> ENTITY: \u001b[32mBrian Lara\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.046]; Lara[0.045]; Lara[0.045]; match[0.044]; Cricket[0.044]; Cricket[0.044]; Warne[0.043]; wicket[0.042]; bat[0.042]; wicketkeeper[0.042]; Australia[0.041]; Australia[0.041]; overs[0.041]; overs[0.041]; Ian[0.041]; run[0.040]; Indies[0.040]; Indies[0.040]; Indies[0.040]; crease[0.040]; play[0.040]; captain[0.040]; captain[0.040]; dismissed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lara\u001b[39m ==> ENTITY: \u001b[32mBrian Lara\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.235:0.235[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.045]; Lara[0.044]; Lara[0.044]; Lara[0.044]; Lara[0.044]; match[0.044]; Cricket[0.043]; toss[0.043]; wicket[0.042]; Marsh[0.042]; bat[0.041]; series[0.041]; series[0.041]; Australia[0.041]; Australia[0.041]; Australia[0.041]; Australia[0.041]; Australia[0.041]; century[0.040]; runs[0.040]; record[0.040]; score[0.040]; score[0.040]; Clive[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWorld Series\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustralian Tri-Series\u001b[39m <---> \u001b[32mWorld Series Cricket\u001b[39m\t\nSCORES: global= 0.257:0.254[\u001b[31m0.002\u001b[39m]; local()= 0.194:0.180[\u001b[31m0.014\u001b[39m]; log p(e|m)= -3.324:-3.324[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.046]; Australia[0.046]; Australia[0.046]; Australia[0.046]; Cricket[0.044]; team[0.044]; team[0.044]; series[0.044]; match[0.043]; played[0.043]; tournament[0.042]; Sydney[0.041]; win[0.041]; Pakistan[0.041]; runs[0.040]; triangular[0.040]; week[0.040]; score[0.039]; went[0.039]; failed[0.039]; Ground[0.038]; ball[0.038]; arrived[0.038]; innings[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Indies ==> ENTITY: \u001b[32mWest Indies cricket team\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.175:0.175[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.045]; Lara[0.045]; Lara[0.045]; Lara[0.045]; Australia[0.044]; Australia[0.044]; match[0.043]; Cricket[0.043]; bat[0.043]; overs[0.042]; wicketkeeper[0.042]; Indies[0.042]; Ian[0.041]; tour[0.040]; tour[0.040]; cheaply[0.040]; captain[0.040]; run[0.040]; earned[0.040]; crease[0.040]; West[0.040]; Healy[0.039]; scored[0.039]; Melbourne[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Indies ==> ENTITY: \u001b[32mWest Indies cricket team\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.192:0.192[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.043]; cricket[0.043]; Lara[0.043]; Lara[0.043]; Lara[0.043]; Australia[0.043]; Australia[0.043]; Australia[0.043]; Australia[0.043]; Cricket[0.042]; series[0.042]; score[0.042]; score[0.042]; second[0.041]; triangular[0.041]; Pakistan[0.041]; team[0.041]; team[0.041]; innings[0.040]; Ground[0.040]; win[0.040]; Indies[0.040]; tournament[0.040]; played[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.199:0.199[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; Australia[0.048]; cricket[0.046]; match[0.044]; Cricket[0.044]; Cricket[0.044]; Sydney[0.043]; Australian[0.042]; series[0.041]; series[0.041]; toss[0.041]; Melbourne[0.041]; captain[0.041]; second[0.041]; bat[0.040]; Marsh[0.040]; Warne[0.040]; coach[0.040]; overs[0.040]; play[0.039]; win[0.039]; won[0.039]; wicket[0.039]; wicket[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Indies ==> ENTITY: \u001b[32mWest Indies cricket team\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.203:0.203[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLara[0.043]; Lara[0.043]; Lara[0.043]; Australia[0.043]; Australia[0.043]; Australia[0.043]; series[0.043]; toss[0.042]; match[0.042]; Cricket[0.042]; wicket[0.042]; wicket[0.042]; wicketkeeper[0.042]; bat[0.041]; bat[0.041]; overs[0.041]; overs[0.041]; innings[0.041]; Marsh[0.041]; Indies[0.041]; Wednesday[0.040]; crease[0.040]; play[0.040]; Warne[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.199:0.199[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.053]; match[0.048]; batsman[0.048]; Cricket[0.048]; Cricket[0.048]; Australian[0.046]; Australian[0.046]; toss[0.045]; Melbourne[0.045]; Melbourne[0.045]; captain[0.045]; captain[0.045]; bat[0.045]; bat[0.045]; wicketkeeper[0.044]; Warne[0.044]; overs[0.044]; overs[0.044]; play[0.043]; won[0.043]; wicket[0.043]; struggled[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sydney Cricket Ground\u001b[39m ==> ENTITY: \u001b[32mSydney Cricket Ground\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.046]; Australia[0.046]; Australia[0.046]; Australia[0.046]; Australia[0.046]; cricket[0.044]; match[0.044]; second[0.041]; series[0.041]; series[0.041]; played[0.041]; Marsh[0.040]; Pakistan[0.040]; tournament[0.040]; highest[0.040]; team[0.040]; team[0.040]; crowd[0.040]; innings[0.039]; innings[0.039]; win[0.039]; Clive[0.039]; failed[0.039]; week[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAustralian\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustralia national cricket team\u001b[39m <---> \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.258:0.243[\u001b[31m0.015\u001b[39m]; local()= 0.176:0.101[\u001b[31m0.074\u001b[39m]; log p(e|m)= -3.612:-0.491[\u001b[32m3.121\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.054]; cricket[0.051]; Cricket[0.049]; series[0.046]; toss[0.046]; Melbourne[0.046]; captain[0.045]; captain[0.045]; bat[0.045]; bat[0.045]; Marsh[0.045]; Clive[0.045]; second[0.044]; Warne[0.044]; coach[0.044]; overs[0.044]; Ian[0.044]; Geoff[0.044]; play[0.044]; won[0.044]; wicket[0.043]; wicket[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Healy\u001b[39m ==> ENTITY: \u001b[32mIan Healy\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.046]; Australia[0.046]; Australia[0.046]; Australia[0.046]; Australia[0.046]; cricket[0.043]; team[0.042]; team[0.042]; Australian[0.041]; match[0.041]; Marsh[0.041]; Cricket[0.041]; played[0.040]; series[0.040]; series[0.040]; second[0.040]; Pakistan[0.040]; Geoff[0.040]; Wednesday[0.040]; week[0.040]; Tuesday[0.040]; Sydney[0.040]; dismissal[0.039]; Clive[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Courtney Walsh\u001b[39m ==> ENTITY: \u001b[32mCourtney Walsh\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.044]; Australia[0.044]; Australia[0.044]; overs[0.043]; toss[0.042]; wicket[0.042]; wicket[0.042]; bat[0.042]; bat[0.042]; Lara[0.042]; Lara[0.042]; Lara[0.042]; Warne[0.042]; Ian[0.041]; Australian[0.041]; wicketkeeper[0.041]; Marsh[0.041]; Cricket[0.041]; Geoff[0.041]; Tom[0.040]; Shane[0.040]; Clive[0.040]; mid[0.040]; second[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Indies ==> ENTITY: \u001b[32mWest Indies cricket team\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.199:0.199[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.044]; Lara[0.043]; Lara[0.043]; Lara[0.043]; Lara[0.043]; Australia[0.043]; Australia[0.043]; toss[0.042]; match[0.042]; Cricket[0.042]; Cricket[0.042]; wicket[0.042]; wicket[0.042]; wicketkeeper[0.041]; bat[0.041]; bat[0.041]; overs[0.041]; overs[0.041]; Indies[0.040]; Indies[0.040]; crease[0.040]; play[0.040]; Warne[0.040]; Ian[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAustralia\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustralia national cricket team\u001b[39m <---> \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.254:0.240[\u001b[31m0.014\u001b[39m]; local()= 0.204:0.147[\u001b[31m0.056\u001b[39m]; log p(e|m)= -3.507:-0.172[\u001b[32m3.335\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; Australia[0.048]; Australia[0.048]; cricket[0.046]; match[0.044]; Cricket[0.044]; team[0.043]; team[0.043]; Sydney[0.043]; Pakistan[0.042]; series[0.041]; tournament[0.041]; played[0.041]; second[0.040]; week[0.040]; win[0.039]; Lara[0.039]; Lara[0.039]; Lara[0.039]; Tuesday[0.039]; highest[0.038]; World[0.038]; failed[0.038]; triangular[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lara\u001b[39m ==> ENTITY: \u001b[32mBrian Lara\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.047]; Lara[0.046]; Lara[0.046]; match[0.045]; Cricket[0.044]; bat[0.043]; wicketkeeper[0.042]; Australia[0.042]; Australia[0.042]; overs[0.042]; Ian[0.041]; run[0.041]; Indies[0.041]; Indies[0.041]; crease[0.041]; captain[0.041]; dismissed[0.041]; scored[0.040]; limited[0.040]; cheaply[0.039]; Healy[0.039]; earned[0.039]; tour[0.039]; tour[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Brian Lara\u001b[39m ==> ENTITY: \u001b[32mBrian Lara\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.047]; Lara[0.045]; Lara[0.045]; match[0.045]; Cricket[0.044]; Warne[0.043]; bat[0.042]; wicketkeeper[0.042]; Australia[0.042]; Australia[0.042]; overs[0.042]; Ian[0.041]; run[0.041]; Indies[0.041]; Indies[0.041]; crease[0.040]; captain[0.040]; dismissed[0.040]; ball[0.040]; pacer[0.040]; scored[0.039]; limited[0.039]; cheaply[0.039]; Healy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Geoff Marsh\u001b[39m ==> ENTITY: \u001b[32mGeoff Marsh\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; Australia[0.048]; Australia[0.048]; cricket[0.047]; Cricket[0.046]; Cricket[0.046]; Australian[0.045]; captain[0.045]; coach[0.044]; match[0.044]; Sydney[0.044]; Melbourne[0.043]; Healy[0.042]; play[0.042]; second[0.042]; Wednesday[0.041]; Clive[0.041]; Tuesday[0.041]; ago[0.041]; toss[0.041]; series[0.041]; series[0.041]; Walsh[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lara\u001b[39m ==> ENTITY: \u001b[32mBrian Lara\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.046]; Lara[0.045]; Lara[0.045]; Lara[0.045]; match[0.044]; Cricket[0.043]; Marsh[0.042]; series[0.042]; Australia[0.041]; Australia[0.041]; Australia[0.041]; Australia[0.041]; Australia[0.041]; century[0.041]; runs[0.041]; Pakistan[0.041]; record[0.041]; tournament[0.040]; score[0.040]; score[0.040]; innings[0.040]; Indies[0.040]; Indies[0.040]; highest[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Clive Lloyd\u001b[39m ==> ENTITY: \u001b[32mClive Lloyd\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.045]; Cricket[0.044]; Cricket[0.044]; Marsh[0.043]; Lara[0.043]; Lara[0.043]; Lara[0.043]; match[0.043]; bat[0.042]; toss[0.042]; Australia[0.041]; Australia[0.041]; Australia[0.041]; Lloyd[0.041]; Warne[0.041]; overs[0.041]; captain[0.041]; wicket[0.041]; wicket[0.041]; second[0.041]; Geoff[0.040]; innings[0.040]; highest[0.040]; Indies[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shane Warne\u001b[39m ==> ENTITY: \u001b[32mShane Warne\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.223:0.223[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.045]; Australia[0.043]; Australia[0.043]; Australia[0.043]; Cricket[0.043]; toss[0.042]; match[0.042]; wicket[0.042]; wicket[0.042]; wicketkeeper[0.042]; Lara[0.042]; Lara[0.042]; Lara[0.042]; Lara[0.042]; bat[0.041]; bat[0.041]; overs[0.041]; overs[0.041]; series[0.041]; Australian[0.041]; Australian[0.041]; Ian[0.040]; cheaply[0.040]; Clive[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.205:0.205[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.047]; Australia[0.047]; Australia[0.047]; Australia[0.047]; cricket[0.045]; match[0.043]; Cricket[0.043]; team[0.043]; team[0.043]; Sydney[0.042]; Pakistan[0.041]; series[0.041]; series[0.041]; tournament[0.041]; played[0.040]; second[0.040]; Marsh[0.039]; week[0.039]; coach[0.039]; win[0.039]; Lara[0.038]; Lara[0.038]; Lara[0.038]; Lara[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAustralian\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustralia national cricket team\u001b[39m <---> \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.258:0.244[\u001b[31m0.014\u001b[39m]; local()= 0.187:0.125[\u001b[31m0.062\u001b[39m]; log p(e|m)= -3.612:-0.491[\u001b[32m3.121\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.054]; Australia[0.054]; batsman[0.049]; Cricket[0.049]; Australian[0.048]; match[0.047]; Melbourne[0.046]; captain[0.045]; bat[0.045]; wicketkeeper[0.045]; Warne[0.044]; overs[0.044]; Ian[0.044]; wicket[0.044]; Lara[0.043]; Lara[0.043]; Lara[0.043]; Brian[0.043]; Shane[0.043]; mid[0.043]; tour[0.043]; tour[0.043]; \t\n\r [================================>.............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 6s354ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3262/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1246testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Africa ==> ENTITY: \u001b[32mSouth Africa\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.085:0.085[0]; log p(e|m)= -0.263:-0.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJohannesburg[0.047]; North[0.047]; West[0.046]; Thursday[0.045]; Friday[0.044]; region[0.044]; state[0.043]; Mafikeng[0.042]; hits[0.042]; radio[0.041]; van[0.041]; preparing[0.041]; people[0.041]; reported[0.041]; delivery[0.041]; freak[0.040]; crashed[0.040]; occurred[0.040]; light[0.040]; plane[0.039]; plane[0.039]; driver[0.039]; pickup[0.039]; pilot[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = North West ==> ENTITY: \u001b[32mNorth West (South African province)\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.109:0.109[0]; log p(e|m)= -1.370:-1.370[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.049]; region[0.048]; Friday[0.046]; Thursday[0.046]; radio[0.044]; Africa[0.043]; Johannesburg[0.043]; state[0.043]; reported[0.043]; Mafikeng[0.042]; people[0.042]; delivery[0.042]; preparing[0.041]; light[0.040]; vehicle[0.040]; occurred[0.040]; van[0.039]; crop[0.039]; crop[0.039]; pickup[0.039]; freak[0.039]; takeoff[0.038]; crash[0.038]; crashed[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mafikeng\u001b[39m ==> ENTITY: \u001b[32mMahikeng\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.101:0.101[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorth[0.048]; South[0.047]; Africa[0.047]; West[0.047]; Thursday[0.045]; Friday[0.044]; region[0.044]; preparing[0.044]; Johannesburg[0.044]; state[0.042]; people[0.041]; occurred[0.040]; pickup[0.040]; killed[0.040]; killed[0.040]; vehicle[0.039]; van[0.039]; radio[0.039]; crashed[0.039]; accident[0.039]; passengers[0.039]; freak[0.038]; survived[0.038]; delivery[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Johannesburg ==> ENTITY: \u001b[32mJohannesburg\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.095:0.095[0]; log p(e|m)= -0.095:-0.095[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.048]; North[0.046]; West[0.045]; Thursday[0.045]; Friday[0.044]; Africa[0.044]; takeoff[0.042]; Mafikeng[0.042]; passengers[0.042]; region[0.042]; delivery[0.041]; plane[0.041]; plane[0.041]; crashed[0.041]; people[0.041]; reported[0.040]; state[0.040]; killed[0.040]; killed[0.040]; occurred[0.040]; freak[0.039]; hits[0.039]; preparing[0.039]; pickup[0.039]; \t\n\r [================================>.............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 6s340ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3266/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1289testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Madrid ==> ENTITY: \u001b[32mMadrid\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.114:0.114[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.047]; Spain[0.047]; Spain[0.047]; Spanish[0.044]; country[0.043]; Basque[0.042]; Basque[0.042]; Pamplona[0.042]; bombs[0.042]; bombs[0.042]; people[0.041]; supporters[0.041]; Friday[0.040]; group[0.040]; political[0.040]; said[0.040]; anniversary[0.040]; protest[0.040]; protest[0.040]; protest[0.040]; cans[0.040]; bottles[0.040]; Herri[0.040]; Batasuna[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBasque\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBasque Country (autonomous community)\u001b[39m <---> \u001b[32mBasque Country (greater region)\u001b[39m\t\nSCORES: global= 0.255:0.235[\u001b[31m0.020\u001b[39m]; local()= 0.196:0.172[\u001b[31m0.024\u001b[39m]; log p(e|m)= -2.025:-2.163[\u001b[31m0.138\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpanish[0.048]; Spain[0.046]; Spain[0.046]; Spain[0.046]; Basque[0.045]; country[0.043]; Eta[0.043]; Batasuna[0.043]; political[0.042]; Madrid[0.041]; separatist[0.041]; Herri[0.041]; group[0.041]; constitution[0.040]; constitution[0.040]; supporters[0.040]; protest[0.040]; protest[0.040]; protest[0.040]; wing[0.040]; people[0.039]; Pamplona[0.039]; said[0.038]; coincided[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBasque\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBasque Country (autonomous community)\u001b[39m <---> \u001b[32mBasque Country (greater region)\u001b[39m\t\nSCORES: global= 0.255:0.235[\u001b[31m0.020\u001b[39m]; local()= 0.196:0.172[\u001b[31m0.024\u001b[39m]; log p(e|m)= -2.025:-2.163[\u001b[31m0.138\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpanish[0.048]; Spain[0.046]; Spain[0.046]; Spain[0.046]; Basque[0.045]; country[0.043]; Eta[0.043]; Batasuna[0.043]; political[0.042]; Madrid[0.041]; separatist[0.041]; Herri[0.041]; group[0.041]; constitution[0.040]; constitution[0.040]; supporters[0.040]; protest[0.040]; protest[0.040]; protest[0.040]; wing[0.040]; people[0.039]; Pamplona[0.039]; said[0.038]; coincided[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Herri Batasuna\u001b[39m ==> ENTITY: \u001b[32mBatasuna\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.128:0.128[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasque[0.045]; Basque[0.045]; Spain[0.043]; Spain[0.043]; Spain[0.043]; supporters[0.043]; group[0.043]; political[0.042]; wing[0.042]; Spanish[0.042]; Eta[0.042]; constitution[0.041]; constitution[0.041]; separatist[0.041]; lb[0.041]; said[0.041]; protest[0.041]; protest[0.041]; protest[0.041]; people[0.040]; Madrid[0.040]; kg[0.040]; disturbances[0.039]; screws[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = ETA ==> ENTITY: \u001b[32mETA\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.151:0.151[0]; log p(e|m)= -0.102:-0.102[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.047]; Spain[0.047]; Spain[0.047]; Basque[0.047]; Basque[0.047]; Batasuna[0.045]; Spanish[0.045]; political[0.045]; country[0.043]; group[0.043]; separatist[0.043]; Madrid[0.043]; protest[0.043]; protest[0.043]; protest[0.043]; wing[0.042]; people[0.042]; Herri[0.041]; said[0.041]; arrested[0.041]; attracted[0.041]; screws[0.041]; objects[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spanish ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.150:0.150[0]; log p(e|m)= -1.241:-1.241[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.053]; Spain[0.053]; Spain[0.053]; country[0.046]; Madrid[0.044]; Basque[0.043]; Basque[0.043]; group[0.042]; thousand[0.041]; people[0.041]; seize[0.040]; political[0.040]; seized[0.039]; Pamplona[0.039]; said[0.039]; Herri[0.039]; police[0.038]; police[0.038]; police[0.038]; attracted[0.038]; separatist[0.038]; arrested[0.038]; coincided[0.038]; Batasuna[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pamplona ==> ENTITY: \u001b[32mPamplona\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.151:0.151[0]; log p(e|m)= -0.038:-0.038[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.045]; Spain[0.045]; Spain[0.045]; people[0.043]; Spanish[0.043]; Madrid[0.043]; Basque[0.043]; Basque[0.043]; seized[0.043]; country[0.042]; attracted[0.041]; Batasuna[0.041]; supporters[0.041]; thousand[0.040]; seize[0.040]; confiscated[0.040]; Friday[0.040]; coincided[0.040]; threw[0.040]; group[0.040]; political[0.040]; said[0.040]; separatist[0.040]; anniversary[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.055]; Spain[0.055]; Spanish[0.048]; country[0.048]; Madrid[0.047]; Basque[0.044]; Basque[0.044]; political[0.044]; said[0.042]; coincided[0.042]; thousand[0.042]; seize[0.041]; group[0.041]; seized[0.041]; separatist[0.041]; people[0.041]; constitution[0.041]; constitution[0.041]; Herri[0.040]; bombs[0.040]; bombs[0.040]; Batasuna[0.040]; Pamplona[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.054]; Spanish[0.047]; country[0.047]; Madrid[0.046]; Basque[0.043]; Basque[0.043]; group[0.042]; political[0.042]; said[0.041]; supporters[0.041]; coincided[0.041]; Friday[0.040]; attracted[0.040]; thousand[0.040]; separatist[0.040]; constitution[0.040]; constitution[0.040]; people[0.039]; seized[0.039]; Herri[0.039]; bombs[0.039]; bombs[0.039]; Batasuna[0.039]; Pamplona[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.054]; Spanish[0.047]; country[0.047]; Madrid[0.046]; Basque[0.043]; Basque[0.043]; political[0.043]; said[0.041]; thousand[0.041]; seize[0.040]; group[0.040]; seized[0.040]; separatist[0.040]; people[0.040]; constitution[0.040]; Herri[0.040]; bombs[0.039]; bombs[0.039]; Batasuna[0.039]; Pamplona[0.039]; protest[0.039]; protest[0.039]; protest[0.039]; Friday[0.039]; \t\n\r [================================>.............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 6s414ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3276/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1374testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Campese\u001b[39m ==> ENTITY: \u001b[32mDavid Campese\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.231:0.231[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWallaby[0.045]; Wallaby[0.045]; Wallaby[0.045]; rugby[0.045]; Wallabies[0.044]; Campese[0.044]; Campese[0.044]; Australia[0.044]; Rugby[0.042]; internationals[0.041]; David[0.041]; jersey[0.041]; tries[0.041]; tries[0.041]; spectacular[0.040]; Barbarians[0.040]; Barbarians[0.040]; game[0.040]; caps[0.040]; match[0.039]; match[0.039]; match[0.039]; international[0.039]; Twickenham[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Twickenham ==> ENTITY: \u001b[32mTwickenham Stadium\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.207:0.207[0]; log p(e|m)= -0.594:-0.594[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.046]; rugby[0.046]; team[0.043]; match[0.043]; match[0.043]; match[0.043]; match[0.043]; caps[0.043]; internationals[0.043]; Barbarians[0.042]; Barbarians[0.042]; Australia[0.042]; game[0.042]; final[0.041]; Wallaby[0.040]; Wallaby[0.040]; Wallaby[0.040]; Wallabies[0.040]; second[0.040]; jersey[0.040]; London[0.040]; Saturday[0.039]; international[0.039]; practice[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Matthew Burke\u001b[39m ==> ENTITY: \u001b[32mMatt Burke\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.048]; Wallabies[0.045]; Wallaby[0.044]; Wallaby[0.044]; Twickenham[0.043]; internationals[0.043]; Barbarians[0.042]; game[0.042]; team[0.042]; Campese[0.042]; jersey[0.042]; practice[0.041]; match[0.041]; match[0.041]; match[0.041]; match[0.041]; caps[0.041]; second[0.040]; touchdown[0.040]; retired[0.040]; tries[0.039]; tries[0.039]; final[0.039]; points[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wallabies\u001b[39m ==> ENTITY: \u001b[32mAustralia national rugby union team\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.244:0.244[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.048]; Rugby[0.047]; Australia[0.045]; internationals[0.044]; Wallaby[0.044]; Wallaby[0.044]; Wallaby[0.044]; Barbarians[0.042]; Barbarians[0.042]; team[0.041]; caps[0.041]; match[0.041]; match[0.041]; match[0.041]; match[0.041]; Twickenham[0.041]; jersey[0.040]; game[0.040]; Campese[0.040]; Campese[0.040]; Campese[0.040]; second[0.039]; final[0.039]; international[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barbarians ==> ENTITY: \u001b[32mBarbarian F.C.\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.203:0.203[0]; log p(e|m)= -0.171:-0.171[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.051]; Rugby[0.050]; team[0.047]; Barbarians[0.047]; Australia[0.047]; Wallabies[0.046]; caps[0.046]; internationals[0.046]; Twickenham[0.046]; jersey[0.045]; Wallaby[0.045]; Wallaby[0.045]; Wallaby[0.045]; tour[0.044]; tour[0.044]; match[0.044]; match[0.044]; match[0.044]; match[0.044]; touring[0.043]; Saturday[0.043]; game[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Campese\u001b[39m ==> ENTITY: \u001b[32mDavid Campese\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.239:0.239[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWallaby[0.049]; Wallaby[0.049]; Wallaby[0.049]; rugby[0.048]; Wallabies[0.048]; Campese[0.048]; Campese[0.048]; Australia[0.047]; Rugby[0.046]; internationals[0.045]; David[0.045]; Matthew[0.044]; jersey[0.044]; tries[0.044]; tries[0.044]; spectacular[0.044]; Barbarians[0.043]; Barbarians[0.043]; game[0.043]; caps[0.043]; team[0.043]; second[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rugby Union\u001b[39m ==> ENTITY: \u001b[32mRugby union\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.055]; Australia[0.046]; internationals[0.046]; Wallaby[0.046]; Wallaby[0.046]; Wallabies[0.046]; Barbarians[0.044]; Barbarians[0.044]; caps[0.044]; Twickenham[0.044]; Campese[0.044]; Campese[0.044]; Campese[0.044]; match[0.043]; match[0.043]; match[0.043]; jersey[0.042]; game[0.041]; international[0.040]; final[0.040]; Saturday[0.040]; European[0.039]; practice[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national rugby union team\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.235:0.235[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.048]; Rugby[0.047]; internationals[0.044]; Wallaby[0.044]; Wallaby[0.044]; Wallaby[0.044]; Wallabies[0.044]; Barbarians[0.042]; Barbarians[0.042]; match[0.041]; match[0.041]; match[0.041]; match[0.041]; caps[0.041]; Twickenham[0.040]; Campese[0.040]; Campese[0.040]; Campese[0.040]; jersey[0.040]; game[0.039]; international[0.039]; final[0.039]; touring[0.039]; David[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wallaby ==> ENTITY: \u001b[32mAustralia national rugby union team\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.245:0.245[0]; log p(e|m)= -0.562:-0.562[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.048]; Rugby[0.047]; Australia[0.045]; internationals[0.044]; Wallaby[0.044]; Wallaby[0.044]; Wallabies[0.044]; Barbarians[0.042]; Barbarians[0.042]; team[0.041]; caps[0.041]; match[0.041]; match[0.041]; match[0.041]; match[0.041]; Twickenham[0.041]; jersey[0.040]; game[0.040]; Campese[0.040]; Campese[0.040]; Campese[0.040]; second[0.039]; final[0.039]; international[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = European ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.106:0.106[0]; log p(e|m)= -0.256:-0.256[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.043]; international[0.043]; Rugby[0.043]; record[0.043]; match[0.043]; match[0.043]; match[0.043]; match[0.043]; touring[0.042]; caps[0.042]; team[0.042]; final[0.042]; world[0.042]; internationals[0.041]; tour[0.041]; tour[0.041]; Australia[0.041]; game[0.041]; time[0.041]; Union[0.041]; practice[0.040]; David[0.040]; half[0.040]; half[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.066:0.066[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrecord[0.046]; match[0.045]; match[0.045]; match[0.045]; match[0.045]; final[0.045]; game[0.043]; tour[0.042]; tour[0.042]; Saturday[0.042]; David[0.042]; Rugby[0.042]; European[0.041]; farewell[0.040]; included[0.040]; internationals[0.040]; rugby[0.040]; international[0.040]; Australia[0.040]; finish[0.039]; Union[0.039]; sessions[0.039]; pre[0.039]; departing[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barbarians ==> ENTITY: \u001b[32mBarbarian F.C.\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.205:0.205[0]; log p(e|m)= -0.171:-0.171[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.049]; Rugby[0.048]; team[0.045]; Barbarians[0.045]; Australia[0.045]; Wallabies[0.044]; caps[0.044]; internationals[0.044]; Twickenham[0.044]; jersey[0.043]; Wallaby[0.043]; Wallaby[0.043]; Wallaby[0.043]; tour[0.042]; tour[0.042]; tour[0.042]; match[0.042]; match[0.042]; match[0.042]; match[0.042]; touring[0.041]; Saturday[0.041]; game[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = David Campese\u001b[39m ==> ENTITY: \u001b[32mDavid Campese\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.231:0.231[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWallaby[0.045]; Wallaby[0.045]; Wallaby[0.045]; rugby[0.045]; Wallabies[0.044]; Campese[0.044]; Campese[0.044]; Australia[0.044]; Rugby[0.042]; internationals[0.041]; jersey[0.041]; tries[0.041]; tries[0.041]; spectacular[0.040]; Barbarians[0.040]; Barbarians[0.040]; game[0.040]; caps[0.040]; team[0.040]; match[0.039]; match[0.039]; match[0.039]; match[0.039]; international[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wallaby ==> ENTITY: \u001b[32mAustralia national rugby union team\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.238:0.238[0]; log p(e|m)= -0.562:-0.562[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.048]; Australia[0.045]; internationals[0.044]; Wallaby[0.044]; Wallaby[0.044]; Wallabies[0.044]; Barbarians[0.042]; Barbarians[0.042]; team[0.042]; caps[0.042]; match[0.041]; match[0.041]; match[0.041]; match[0.041]; Twickenham[0.041]; jersey[0.041]; game[0.040]; Campese[0.040]; Campese[0.040]; Campese[0.040]; second[0.039]; final[0.039]; international[0.039]; test[0.038]; \t\n\r [================================>.............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 6s352ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3290/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1393testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Charlton\u001b[39m ==> ENTITY: \u001b[32mJack Charlton\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.047]; Ireland[0.047]; Ireland[0.047]; Charlton[0.044]; Charlton[0.044]; Charlton[0.044]; Irish[0.043]; manager[0.043]; manager[0.043]; Irishman[0.042]; Republic[0.041]; national[0.041]; soccer[0.041]; Dublin[0.041]; Soccer[0.040]; Jack[0.040]; spent[0.040]; Englishman[0.040]; Englishman[0.040]; received[0.039]; officially[0.039]; Minister[0.038]; Dick[0.038]; said[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIrish\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIreland\u001b[39m <---> \u001b[32mRepublic of Ireland national football team\u001b[39m\t\nSCORES: global= 0.244:0.234[\u001b[31m0.010\u001b[39m]; local()= 0.139:0.204[\u001b[32m0.064\u001b[39m]; log p(e|m)= -1.457:-4.605[\u001b[31m3.148\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.051]; Ireland[0.051]; Ireland[0.051]; England[0.044]; England[0.044]; team[0.044]; team[0.044]; matches[0.043]; Cup[0.041]; Cup[0.041]; caps[0.040]; Leeds[0.040]; Charlton[0.040]; Charlton[0.040]; finals[0.040]; finals[0.040]; qualifier[0.039]; won[0.039]; European[0.039]; tournaments[0.038]; winning[0.038]; successive[0.038]; championship[0.037]; debut[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ireland ==> ENTITY: \u001b[32mRepublic of Ireland\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.209:0.209[0]; log p(e|m)= -1.505:-1.505[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.050]; Ireland[0.050]; Ireland[0.050]; Ireland[0.050]; Ireland[0.050]; Irish[0.044]; Irish[0.044]; England[0.042]; team[0.042]; matches[0.042]; Cup[0.039]; Charlton[0.039]; Charlton[0.039]; Charlton[0.039]; Charlton[0.039]; Charlton[0.039]; soccer[0.038]; Republic[0.038]; finals[0.038]; finals[0.038]; qualifier[0.038]; European[0.037]; tournaments[0.037]; national[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ireland ==> ENTITY: \u001b[32mRepublic of Ireland\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.170:0.170[0]; log p(e|m)= -1.505:-1.505[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.053]; Ireland[0.053]; Ireland[0.053]; Irish[0.047]; Dublin[0.045]; Charlton[0.042]; Charlton[0.042]; Charlton[0.042]; Charlton[0.042]; Charlton[0.042]; soccer[0.041]; Republic[0.041]; Soccer[0.040]; Irishman[0.040]; national[0.039]; Saturday[0.039]; Englishman[0.038]; Englishman[0.038]; Jack[0.038]; manager[0.038]; manager[0.038]; night[0.037]; spent[0.037]; received[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ireland ==> ENTITY: \u001b[32mRepublic of Ireland\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.174:0.174[0]; log p(e|m)= -1.505:-1.505[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.053]; Ireland[0.053]; Ireland[0.053]; Irish[0.047]; Dublin[0.045]; Charlton[0.042]; Charlton[0.042]; Charlton[0.042]; Charlton[0.042]; soccer[0.041]; Republic[0.041]; Soccer[0.040]; Irishman[0.040]; national[0.040]; Saturday[0.039]; Englishman[0.038]; Englishman[0.038]; Jack[0.038]; manager[0.038]; manager[0.038]; night[0.038]; spent[0.038]; received[0.037]; officially[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Cup ==> ENTITY: \u001b[32mFIFA World Cup\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.168:0.168[0]; log p(e|m)= -1.002:-1.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.044]; Ireland[0.044]; Ireland[0.044]; Ireland[0.044]; England[0.044]; England[0.044]; matches[0.044]; Cup[0.043]; tournaments[0.042]; team[0.042]; team[0.042]; finals[0.042]; finals[0.042]; qualifier[0.041]; Germany[0.041]; caps[0.041]; World[0.041]; European[0.040]; championship[0.040]; won[0.040]; successive[0.039]; winning[0.039]; time[0.039]; debut[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = 1966 World Cup\u001b[39m ==> ENTITY: \u001b[32m1966 FIFA World Cup\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; England[0.047]; matches[0.046]; team[0.043]; team[0.043]; qualifier[0.042]; Ireland[0.042]; Ireland[0.042]; Cup[0.042]; Germany[0.042]; won[0.042]; Leeds[0.041]; European[0.041]; tournaments[0.041]; winning[0.041]; Charlton[0.040]; World[0.040]; beat[0.040]; caps[0.040]; finals[0.040]; finals[0.040]; successive[0.040]; Bobby[0.039]; added[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Charlton\u001b[39m ==> ENTITY: \u001b[32mJack Charlton\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.045]; Ireland[0.045]; Ireland[0.045]; Ireland[0.045]; England[0.045]; England[0.045]; Leeds[0.042]; matches[0.042]; Charlton[0.042]; Charlton[0.042]; Irish[0.041]; team[0.041]; team[0.041]; manager[0.041]; manager[0.041]; Bobby[0.041]; European[0.041]; won[0.040]; Republic[0.040]; caps[0.040]; qualifier[0.039]; national[0.039]; Cup[0.039]; Cup[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dick Spring\u001b[39m ==> ENTITY: \u001b[32mDick Spring\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.049]; Ireland[0.049]; Ireland[0.049]; Ireland[0.049]; Ireland[0.049]; Irish[0.048]; Dublin[0.048]; received[0.044]; Republic[0.044]; Minister[0.043]; Irishman[0.043]; spent[0.042]; Jack[0.041]; said[0.041]; said[0.041]; formally[0.041]; Soccer[0.040]; Prime[0.040]; Saturday[0.040]; added[0.040]; night[0.040]; deputy[0.040]; matches[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEnglishman\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEngland\u001b[39m <---> \u001b[32mEnglish people\u001b[39m\t\nSCORES: global= 0.258:0.247[\u001b[31m0.011\u001b[39m]; local()= 0.113:0.134[\u001b[32m0.021\u001b[39m]; log p(e|m)= -0.830:-0.241[\u001b[32m0.589\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.047]; Ireland[0.047]; Ireland[0.047]; Ireland[0.047]; Englishman[0.046]; Irish[0.046]; national[0.042]; Jack[0.042]; Republic[0.042]; Irishman[0.041]; Spring[0.040]; great[0.040]; great[0.040]; Charlton[0.040]; Charlton[0.040]; Charlton[0.040]; Charlton[0.040]; Charlton[0.040]; Dick[0.040]; Dublin[0.040]; wife[0.040]; recognition[0.039]; soccer[0.039]; spent[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland national football team\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.180:0.180[0]; log p(e|m)= -2.781:-2.781[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; matches[0.045]; caps[0.044]; Ireland[0.044]; Ireland[0.044]; Ireland[0.044]; Charlton[0.043]; Charlton[0.043]; team[0.042]; team[0.042]; qualifier[0.042]; Leeds[0.042]; tournaments[0.041]; Cup[0.041]; Cup[0.041]; debut[0.040]; finals[0.040]; finals[0.040]; European[0.039]; won[0.039]; successive[0.039]; beat[0.039]; United[0.039]; time[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Irishman ==> ENTITY: \u001b[32mRepublic of Ireland\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.196:0.196[0]; log p(e|m)= -1.470:-1.470[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.056]; Ireland[0.056]; Ireland[0.056]; Irish[0.050]; Dublin[0.048]; Republic[0.043]; Jack[0.041]; national[0.041]; spent[0.039]; life[0.039]; Englishman[0.039]; Englishman[0.039]; received[0.039]; Minister[0.039]; said[0.038]; deputy[0.038]; formally[0.038]; people[0.038]; Dick[0.038]; achievements[0.038]; soccer[0.038]; wife[0.037]; passports[0.037]; citizens[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = European ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.256:-0.256[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.046]; Cup[0.042]; Cup[0.042]; team[0.042]; team[0.042]; England[0.042]; England[0.042]; tournaments[0.042]; qualifier[0.042]; Ireland[0.042]; Ireland[0.042]; Ireland[0.042]; finals[0.042]; finals[0.042]; caps[0.042]; Germany[0.041]; won[0.041]; Irish[0.041]; group[0.040]; grew[0.040]; United[0.040]; championship[0.040]; shared[0.040]; Leeds[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Republic of Ireland ==> ENTITY: \u001b[32mRepublic of Ireland national football team\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.231:0.231[0]; log p(e|m)= -1.461:-1.461[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.049]; Ireland[0.049]; Ireland[0.049]; Ireland[0.049]; Ireland[0.049]; Irish[0.044]; Dublin[0.043]; matches[0.042]; soccer[0.042]; Soccer[0.041]; Cup[0.040]; Charlton[0.040]; Charlton[0.040]; Charlton[0.040]; Charlton[0.040]; Charlton[0.040]; Charlton[0.040]; tournaments[0.039]; national[0.039]; Irishman[0.038]; finals[0.038]; time[0.038]; manager[0.038]; manager[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Charlton\u001b[39m ==> ENTITY: \u001b[32mJack Charlton\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.047]; Ireland[0.047]; Ireland[0.047]; Ireland[0.047]; Charlton[0.043]; Charlton[0.043]; Charlton[0.043]; Charlton[0.043]; Irish[0.043]; manager[0.042]; manager[0.042]; Irishman[0.042]; Republic[0.041]; national[0.040]; soccer[0.040]; Dublin[0.040]; Soccer[0.040]; Jack[0.039]; spent[0.039]; Englishman[0.039]; Englishman[0.039]; received[0.039]; officially[0.038]; Minister[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBobby\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBobby Robson\u001b[39m <---> \u001b[32mBobby Charlton\u001b[39m\t\nSCORES: global= 0.253:0.233[\u001b[31m0.020\u001b[39m]; local()= 0.151:0.136[\u001b[31m0.014\u001b[39m]; log p(e|m)= -3.772:-4.962[\u001b[31m1.190\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.048]; team[0.048]; England[0.047]; England[0.047]; Charlton[0.045]; Leeds[0.044]; matches[0.044]; United[0.043]; Ireland[0.043]; Ireland[0.043]; Cup[0.043]; Cup[0.043]; championship[0.043]; debut[0.043]; caps[0.042]; European[0.042]; winning[0.042]; won[0.042]; time[0.042]; added[0.042]; finals[0.041]; finals[0.041]; resigned[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIreland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRepublic of Ireland\u001b[39m <---> \u001b[32mRepublic of Ireland national football team\u001b[39m\t\nSCORES: global= 0.244:0.231[\u001b[31m0.013\u001b[39m]; local()= 0.163:0.216[\u001b[32m0.053\u001b[39m]; log p(e|m)= -1.505:-4.423[\u001b[31m2.918\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.050]; Ireland[0.050]; Ireland[0.050]; Irish[0.045]; England[0.043]; England[0.043]; team[0.043]; team[0.043]; matches[0.043]; caps[0.041]; Cup[0.040]; Cup[0.040]; Charlton[0.040]; Charlton[0.040]; Charlton[0.040]; soccer[0.039]; Leeds[0.039]; Republic[0.039]; finals[0.039]; finals[0.039]; qualifier[0.038]; won[0.038]; European[0.038]; tournaments[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Leeds United\u001b[39m ==> ENTITY: \u001b[32mLeeds United A.F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.045]; Charlton[0.045]; England[0.045]; England[0.045]; matches[0.044]; team[0.043]; team[0.043]; debut[0.042]; Cup[0.042]; Cup[0.042]; caps[0.042]; Bobby[0.041]; beat[0.041]; group[0.040]; won[0.040]; defender[0.040]; period[0.040]; finals[0.040]; finals[0.040]; qualifier[0.040]; Ireland[0.040]; Ireland[0.040]; added[0.040]; December[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Charlton\u001b[39m ==> ENTITY: \u001b[32mJack Charlton\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.045]; Ireland[0.045]; Ireland[0.045]; Ireland[0.045]; Ireland[0.045]; England[0.045]; England[0.045]; Leeds[0.042]; matches[0.042]; Charlton[0.041]; Charlton[0.041]; Charlton[0.041]; Irish[0.041]; Irish[0.041]; team[0.041]; manager[0.041]; manager[0.041]; European[0.041]; Republic[0.039]; qualifier[0.039]; national[0.039]; Cup[0.039]; debut[0.039]; soccer[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIrish\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIreland\u001b[39m <---> \u001b[32mRepublic of Ireland\u001b[39m\t\nSCORES: global= 0.247:0.243[\u001b[31m0.004\u001b[39m]; local()= 0.183:0.191[\u001b[32m0.008\u001b[39m]; log p(e|m)= -1.457:-2.254[\u001b[31m0.797\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.052]; Ireland[0.052]; Ireland[0.052]; Ireland[0.052]; Dublin[0.045]; Irishman[0.042]; soccer[0.041]; Republic[0.040]; Charlton[0.040]; Charlton[0.040]; Charlton[0.040]; Charlton[0.040]; Charlton[0.040]; grew[0.040]; Soccer[0.040]; national[0.039]; Jack[0.039]; Saturday[0.038]; manager[0.038]; manager[0.038]; people[0.037]; people[0.037]; spent[0.037]; added[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Englishman ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.107:0.107[0]; log p(e|m)= -0.830:-0.830[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.047]; Ireland[0.047]; Ireland[0.047]; Englishman[0.046]; Irish[0.046]; national[0.043]; Jack[0.043]; Republic[0.042]; Irishman[0.041]; Spring[0.041]; Charlton[0.040]; Charlton[0.040]; Charlton[0.040]; Charlton[0.040]; Dick[0.040]; Dublin[0.040]; wife[0.040]; recognition[0.040]; soccer[0.040]; spent[0.040]; people[0.040]; formally[0.039]; said[0.039]; Peggy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dublin ==> ENTITY: \u001b[32mDublin\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.152:0.152[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.051]; Ireland[0.051]; Ireland[0.051]; Irish[0.047]; Irishman[0.044]; soccer[0.042]; Jack[0.042]; Saturday[0.042]; Republic[0.041]; Soccer[0.041]; spent[0.041]; Peggy[0.041]; national[0.041]; Dick[0.040]; manager[0.039]; manager[0.039]; people[0.039]; received[0.038]; wife[0.038]; citizens[0.038]; lots[0.038]; passports[0.038]; formally[0.038]; declared[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mGermany\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mGermany national football team\u001b[39m <---> \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.244:0.241[\u001b[31m0.003\u001b[39m]; local()= 0.163:0.097[\u001b[31m0.066\u001b[39m]; log p(e|m)= -3.352:-0.155[\u001b[32m3.197\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.046]; England[0.044]; England[0.044]; team[0.044]; team[0.044]; tournaments[0.043]; Ireland[0.042]; Ireland[0.042]; Ireland[0.042]; European[0.042]; qualifier[0.042]; successive[0.041]; key[0.041]; caps[0.041]; finals[0.040]; finals[0.040]; World[0.040]; World[0.040]; won[0.040]; debut[0.040]; championship[0.040]; Cup[0.040]; Cup[0.040]; winning[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIreland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRepublic of Ireland\u001b[39m <---> \u001b[32mRepublic of Ireland national football team\u001b[39m\t\nSCORES: global= 0.244:0.231[\u001b[31m0.013\u001b[39m]; local()= 0.164:0.212[\u001b[32m0.049\u001b[39m]; log p(e|m)= -1.505:-4.423[\u001b[31m2.918\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.051]; Ireland[0.051]; Ireland[0.051]; Irish[0.045]; England[0.043]; England[0.043]; team[0.043]; team[0.043]; matches[0.043]; caps[0.041]; Cup[0.040]; Cup[0.040]; Charlton[0.040]; Charlton[0.040]; Leeds[0.039]; Republic[0.039]; finals[0.039]; finals[0.039]; qualifier[0.039]; won[0.038]; European[0.038]; tournaments[0.038]; winning[0.038]; debut[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Charlton\u001b[39m ==> ENTITY: \u001b[32mJack Charlton\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.048]; Ireland[0.048]; Ireland[0.048]; Ireland[0.048]; Ireland[0.048]; matches[0.044]; Charlton[0.044]; Charlton[0.044]; Charlton[0.044]; Charlton[0.044]; Charlton[0.044]; Irish[0.043]; manager[0.043]; manager[0.043]; Irishman[0.042]; Republic[0.042]; national[0.041]; soccer[0.041]; Dublin[0.041]; Soccer[0.040]; Jack[0.040]; added[0.040]; spent[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland national football team\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.176:0.176[0]; log p(e|m)= -2.781:-2.781[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.048]; matches[0.045]; team[0.044]; team[0.044]; Leeds[0.044]; caps[0.044]; Ireland[0.044]; Ireland[0.044]; qualifier[0.042]; Charlton[0.042]; Charlton[0.042]; tournaments[0.041]; Cup[0.041]; Cup[0.041]; finals[0.041]; finals[0.041]; debut[0.040]; won[0.039]; successive[0.039]; beat[0.039]; time[0.039]; European[0.039]; United[0.039]; championship[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jack Charlton\u001b[39m ==> ENTITY: \u001b[32mJack Charlton\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.047]; Ireland[0.047]; Ireland[0.047]; Ireland[0.047]; Charlton[0.043]; Charlton[0.043]; Charlton[0.043]; Irish[0.043]; manager[0.043]; manager[0.043]; Irishman[0.042]; Republic[0.041]; national[0.040]; soccer[0.040]; Dublin[0.040]; Soccer[0.040]; spent[0.039]; Englishman[0.039]; Englishman[0.039]; received[0.039]; officially[0.039]; Minister[0.038]; Dick[0.038]; said[0.038]; \t\n\r [=================================>............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 6s162ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3317/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1228testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWorld Series\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustralian Tri-Series\u001b[39m <---> \u001b[32mWorld Series Cricket\u001b[39m\t\nSCORES: global= 0.253:0.250[\u001b[31m0.003\u001b[39m]; local()= 0.166:0.142[\u001b[31m0.025\u001b[39m]; log p(e|m)= -3.324:-3.324[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; Australia[0.048]; Cricket[0.046]; match[0.045]; match[0.045]; play[0.044]; Sydney[0.043]; opener[0.042]; Melbourne[0.042]; delay[0.042]; Friday[0.041]; Shivnarine[0.041]; delayed[0.041]; start[0.040]; overs[0.040]; earlier[0.040]; innings[0.039]; Play[0.039]; Play[0.039]; Indies[0.039]; Indies[0.039]; Indies[0.039]; minutes[0.039]; minutes[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.175:0.175[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.051]; match[0.046]; match[0.046]; Cricket[0.046]; Sydney[0.045]; Melbourne[0.043]; Shivnarine[0.043]; overs[0.042]; play[0.042]; opener[0.041]; World[0.041]; West[0.040]; West[0.040]; West[0.040]; Indies[0.040]; Indies[0.040]; Indies[0.040]; start[0.040]; innings[0.040]; resumed[0.039]; Friday[0.039]; Chanderpaul[0.039]; earlier[0.038]; Campbell[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shivnarine Chanderpaul\u001b[39m ==> ENTITY: \u001b[32mShivnarine Chanderpaul\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \novers[0.045]; match[0.045]; match[0.045]; innings[0.045]; Cricket[0.044]; Australia[0.043]; Australia[0.043]; play[0.042]; opener[0.042]; Indies[0.042]; Indies[0.042]; Indies[0.042]; West[0.041]; West[0.041]; West[0.041]; start[0.041]; earlier[0.040]; limited[0.040]; Series[0.040]; Newsroom[0.040]; delay[0.039]; delayed[0.039]; resumed[0.039]; Friday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Indies ==> ENTITY: \u001b[32mWest Indies cricket team\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.181:0.181[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.045]; Australia[0.045]; Chanderpaul[0.045]; Shivnarine[0.044]; match[0.044]; match[0.044]; Cricket[0.044]; overs[0.043]; innings[0.043]; Indies[0.043]; Indies[0.043]; play[0.042]; World[0.041]; Sydney[0.041]; opener[0.040]; West[0.040]; West[0.040]; Melbourne[0.040]; Friday[0.040]; earlier[0.039]; start[0.039]; delayed[0.039]; resumed[0.039]; limited[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sherwin Campbell\u001b[39m ==> ENTITY: \u001b[32mSherwin Campbell\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.044]; match[0.044]; play[0.044]; opener[0.043]; resumed[0.042]; start[0.042]; Australia[0.042]; Australia[0.042]; Shivnarine[0.042]; West[0.042]; West[0.042]; West[0.042]; innings[0.042]; overs[0.041]; Indies[0.041]; Indies[0.041]; Indies[0.041]; Cricket[0.041]; Series[0.040]; minutes[0.040]; minutes[0.040]; Play[0.040]; Play[0.040]; Newsroom[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Indies ==> ENTITY: \u001b[32mWest Indies cricket team\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.181:0.181[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.045]; Australia[0.045]; Chanderpaul[0.045]; Shivnarine[0.044]; match[0.044]; match[0.044]; Cricket[0.044]; overs[0.043]; innings[0.043]; Indies[0.043]; Indies[0.043]; play[0.042]; World[0.041]; Sydney[0.041]; opener[0.040]; West[0.040]; West[0.040]; Melbourne[0.040]; Friday[0.040]; earlier[0.039]; start[0.039]; delayed[0.039]; resumed[0.039]; limited[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Melbourne ==> ENTITY: \u001b[32mMelbourne\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.111:0.111[0]; log p(e|m)= -0.168:-0.168[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSydney[0.049]; Australia[0.048]; Australia[0.048]; match[0.048]; match[0.048]; Cricket[0.045]; play[0.043]; opener[0.041]; start[0.041]; Friday[0.041]; innings[0.040]; Play[0.040]; Play[0.040]; Campbell[0.040]; overs[0.039]; West[0.039]; West[0.039]; West[0.039]; World[0.039]; earlier[0.039]; resumed[0.039]; limited[0.039]; Shivnarine[0.038]; Sherwin[0.038]; \t\n\r [=================================>............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 6s128ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3324/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1250testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Warsaw ==> ENTITY: \u001b[32mWarsaw\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.159:0.159[0]; log p(e|m)= -0.143:-0.143[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.049]; Warsaw[0.046]; Polish[0.045]; Polish[0.045]; communist[0.043]; communist[0.043]; Friday[0.042]; Kwasniewski[0.042]; Kwasniewski[0.042]; Aleksander[0.041]; relations[0.041]; despite[0.041]; early[0.041]; John[0.041]; visit[0.040]; visit[0.040]; visit[0.040]; Dariusz[0.040]; born[0.040]; planned[0.040]; said[0.040]; said[0.040]; conference[0.039]; minister[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.045]; won[0.044]; signed[0.044]; meeting[0.042]; meeting[0.042]; government[0.042]; despite[0.042]; Friday[0.042]; place[0.042]; Scalfaro[0.042]; communist[0.041]; communist[0.041]; communist[0.041]; conference[0.041]; Warsaw[0.041]; Warsaw[0.041]; Warsaw[0.041]; early[0.041]; Alliance[0.041]; Vatican[0.041]; Vatican[0.041]; agreement[0.041]; treaty[0.041]; treaty[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vatican ==> ENTITY: \u001b[32mVatican City\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.929:-0.929[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVatican[0.049]; Catholic[0.048]; Poland[0.045]; Poland[0.045]; law[0.045]; Church[0.044]; Warsaw[0.043]; religious[0.042]; rules[0.041]; month[0.041]; believers[0.041]; rights[0.040]; meeting[0.040]; recent[0.040]; government[0.040]; parliament[0.040]; Kwasniewski[0.040]; relations[0.040]; largely[0.040]; place[0.039]; groups[0.039]; abortion[0.039]; treaty[0.039]; treaty[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Democratic Left Alliance\u001b[39m ==> ENTITY: \u001b[32mDemocratic Left Alliance\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nparty[0.045]; party[0.045]; parliamentary[0.044]; won[0.044]; meeting[0.043]; meeting[0.043]; conference[0.042]; parliament[0.042]; centrist[0.042]; Poland[0.042]; Poland[0.042]; elections[0.041]; ratification[0.041]; ratification[0.041]; President[0.041]; President[0.041]; signed[0.040]; signed[0.040]; minister[0.040]; Kwasniewski[0.040]; Kwasniewski[0.040]; agreement[0.040]; groups[0.040]; Warsaw[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Oscar Scalfaro\u001b[39m ==> ENTITY: \u001b[32mOscar Luigi Scalfaro\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.049]; President[0.046]; President[0.046]; President[0.046]; president[0.044]; parliament[0.044]; won[0.044]; parliamentary[0.044]; government[0.043]; influence[0.043]; elections[0.043]; Pope[0.043]; Pope[0.043]; Pope[0.043]; Catholic[0.043]; Warsaw[0.042]; Warsaw[0.042]; Warsaw[0.042]; Vatican[0.042]; Vatican[0.042]; Kwasniewski[0.042]; Kwasniewski[0.042]; centrist[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Warsaw ==> ENTITY: \u001b[32mWarsaw\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.160:0.160[0]; log p(e|m)= -0.143:-0.143[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.048]; Warsaw[0.046]; Warsaw[0.046]; Polish[0.044]; Polish[0.044]; communist[0.043]; communist[0.043]; communist[0.043]; Friday[0.042]; Kwasniewski[0.041]; Kwasniewski[0.041]; Aleksander[0.041]; relations[0.040]; despite[0.040]; early[0.040]; John[0.040]; visit[0.040]; visit[0.040]; visit[0.040]; Dariusz[0.040]; government[0.040]; born[0.039]; planned[0.039]; parliament[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kwasniewski\u001b[39m ==> ENTITY: \u001b[32mAleksander Kwaśniewski\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.046]; Poland[0.046]; Warsaw[0.045]; parliamentary[0.044]; elections[0.043]; communist[0.042]; won[0.042]; parliament[0.042]; signed[0.042]; signed[0.042]; party[0.041]; party[0.041]; ratification[0.041]; ratification[0.041]; Democratic[0.041]; ex[0.041]; Alliance[0.041]; says[0.040]; month[0.040]; talks[0.040]; agreement[0.040]; relations[0.040]; government[0.040]; recent[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Catholic Church\u001b[39m ==> ENTITY: \u001b[32mCatholic Church\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.126:0.126[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVatican[0.046]; Vatican[0.046]; religious[0.045]; Pope[0.045]; government[0.044]; Poland[0.043]; Poland[0.043]; Italy[0.043]; recent[0.042]; groups[0.042]; believers[0.041]; Warsaw[0.041]; non[0.041]; law[0.040]; said[0.040]; relations[0.040]; largely[0.040]; Alliance[0.040]; rules[0.040]; communist[0.040]; conference[0.040]; rights[0.040]; told[0.039]; abortion[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dariusz Rosati\u001b[39m ==> ENTITY: \u001b[32mDariusz Rosati\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.045]; Poland[0.045]; President[0.042]; President[0.042]; President[0.042]; president[0.042]; Italy[0.042]; communist[0.042]; communist[0.042]; communist[0.042]; Warsaw[0.042]; Warsaw[0.042]; Warsaw[0.042]; Alliance[0.041]; Polish[0.041]; Polish[0.041]; minister[0.041]; Kwasniewski[0.040]; Kwasniewski[0.040]; relations[0.040]; parliamentary[0.040]; ratification[0.040]; ratification[0.040]; parliament[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Polish ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.138:0.138[0]; log p(e|m)= -0.814:-0.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.049]; Polish[0.048]; Warsaw[0.045]; Warsaw[0.045]; John[0.043]; communist[0.042]; communist[0.042]; born[0.042]; Paul[0.041]; Kwasniewski[0.041]; Kwasniewski[0.041]; Pope[0.041]; Pope[0.041]; Pope[0.041]; President[0.040]; President[0.040]; President[0.040]; early[0.040]; Italy[0.040]; Aleksander[0.040]; relations[0.040]; despite[0.040]; president[0.040]; Dariusz[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Aleksander Kwasniewski\u001b[39m ==> ENTITY: \u001b[32mAleksander Kwaśniewski\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKwasniewski[0.047]; Poland[0.046]; President[0.045]; President[0.045]; President[0.045]; Warsaw[0.045]; Warsaw[0.045]; Warsaw[0.045]; Polish[0.044]; Polish[0.044]; president[0.043]; meeting[0.043]; meeting[0.043]; said[0.043]; said[0.043]; communist[0.043]; communist[0.043]; despite[0.042]; Friday[0.042]; conference[0.041]; visit[0.041]; visit[0.041]; visit[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.094:0.094[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.050]; Warsaw[0.045]; won[0.044]; communist[0.043]; parliament[0.043]; parliamentary[0.043]; Kwasniewski[0.042]; Dariusz[0.041]; law[0.041]; government[0.041]; rights[0.041]; month[0.041]; signed[0.041]; signed[0.041]; ratification[0.041]; ratification[0.041]; party[0.040]; party[0.040]; recent[0.040]; elections[0.040]; relations[0.040]; meeting[0.040]; meeting[0.040]; Alliance[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rosati\u001b[39m ==> ENTITY: \u001b[32mDariusz Rosati\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.047]; Poland[0.047]; President[0.044]; President[0.044]; President[0.044]; president[0.044]; Italy[0.044]; Dariusz[0.044]; communist[0.044]; communist[0.044]; communist[0.044]; Warsaw[0.043]; Warsaw[0.043]; Warsaw[0.043]; Alliance[0.043]; Polish[0.043]; Polish[0.043]; minister[0.042]; Kwasniewski[0.042]; Kwasniewski[0.042]; relations[0.042]; relations[0.042]; parliamentary[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kwasniewski\u001b[39m ==> ENTITY: \u001b[32mAleksander Kwaśniewski\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKwasniewski[0.045]; Poland[0.044]; President[0.043]; President[0.043]; President[0.043]; Warsaw[0.043]; Warsaw[0.043]; Warsaw[0.043]; Polish[0.042]; Polish[0.042]; parliamentary[0.042]; president[0.041]; elections[0.041]; meeting[0.041]; meeting[0.041]; said[0.041]; said[0.041]; communist[0.041]; communist[0.041]; communist[0.041]; won[0.040]; parliament[0.040]; despite[0.040]; signed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.050]; Warsaw[0.046]; won[0.044]; communist[0.043]; Kwasniewski[0.042]; law[0.042]; government[0.042]; rights[0.041]; month[0.041]; rules[0.041]; parliament[0.041]; ratification[0.041]; ratification[0.041]; party[0.041]; party[0.041]; recent[0.041]; relations[0.041]; parliamentary[0.041]; Alliance[0.040]; religious[0.040]; treaty[0.040]; treaty[0.040]; influence[0.040]; right[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.146:0.146[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPolish[0.045]; Polish[0.045]; Warsaw[0.045]; Warsaw[0.045]; Italy[0.044]; President[0.044]; President[0.044]; President[0.044]; communist[0.042]; communist[0.042]; president[0.042]; Kwasniewski[0.041]; Kwasniewski[0.041]; despite[0.041]; Dariusz[0.041]; Aleksander[0.040]; visit[0.040]; visit[0.040]; visit[0.040]; relations[0.040]; born[0.039]; Friday[0.039]; place[0.039]; John[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Warsaw ==> ENTITY: \u001b[32mWarsaw\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.143:-0.143[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.048]; Poland[0.048]; Warsaw[0.046]; Polish[0.044]; parliamentary[0.043]; communist[0.042]; communist[0.042]; plans[0.042]; anti[0.041]; parliament[0.041]; government[0.041]; Friday[0.041]; Kwasniewski[0.040]; Kwasniewski[0.040]; Kwasniewski[0.040]; elections[0.040]; month[0.040]; relations[0.040]; relations[0.040]; John[0.040]; despite[0.040]; conference[0.040]; visit[0.040]; visit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vatican ==> ENTITY: \u001b[32mVatican City\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.170:0.170[0]; log p(e|m)= -0.929:-0.929[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVatican[0.048]; Italy[0.047]; Pope[0.044]; Pope[0.044]; Pope[0.044]; Poland[0.044]; John[0.043]; Paul[0.043]; President[0.043]; President[0.043]; President[0.043]; Warsaw[0.040]; Warsaw[0.040]; Warsaw[0.040]; president[0.040]; minister[0.040]; Polish[0.039]; Polish[0.039]; said[0.039]; said[0.039]; signed[0.039]; visit[0.039]; visit[0.039]; visit[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vatican ==> ENTITY: \u001b[32mVatican City\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.929:-0.929[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVatican[0.048]; Vatican[0.048]; Catholic[0.048]; Italy[0.047]; Pope[0.046]; Pope[0.046]; Poland[0.044]; Poland[0.044]; Church[0.044]; John[0.044]; Paul[0.043]; President[0.043]; President[0.043]; President[0.043]; Warsaw[0.043]; Warsaw[0.043]; Polish[0.041]; religious[0.041]; month[0.040]; believers[0.040]; minister[0.040]; said[0.040]; said[0.040]; \t\n\r [=================================>............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 6s | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3343/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1357testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.170:0.170[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.049]; France[0.049]; Switzerland[0.045]; Switzerland[0.045]; Austria[0.043]; Belarus[0.042]; Republic[0.042]; Cup[0.042]; Cup[0.042]; Canada[0.042]; Canada[0.042]; Canada[0.042]; Andy[0.041]; Sebastien[0.041]; Blanc[0.040]; competition[0.040]; points[0.040]; Russia[0.040]; Eric[0.039]; World[0.039]; World[0.039]; Jeff[0.039]; Alexandre[0.039]; Saturday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.048]; Canada[0.046]; Canada[0.046]; Canada[0.046]; Canada[0.046]; Australia[0.044]; Australia[0.044]; Sweden[0.043]; Switzerland[0.042]; Switzerland[0.042]; Belarus[0.042]; Norway[0.041]; Austria[0.041]; Republic[0.041]; Russia[0.041]; Xu[0.040]; Cuo[0.040]; Czech[0.040]; Women[0.040]; Andy[0.038]; Dan[0.038]; Nannan[0.037]; Worthington[0.037]; Jacqui[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Evelyne Leu\u001b[39m ==> ENTITY: \u001b[32mEvelyne Leu\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAerials[0.049]; aerials[0.045]; skiing[0.045]; freestyle[0.044]; Tignes[0.043]; Skiing[0.042]; Hilde[0.042]; Cup[0.042]; Cup[0.042]; Switzerland[0.042]; Switzerland[0.042]; Republic[0.041]; Nannan[0.041]; Belarus[0.041]; China[0.040]; China[0.040]; Austria[0.040]; Valenta[0.040]; Karin[0.040]; Norway[0.040]; Czech[0.039]; Women[0.039]; Saturday[0.039]; Alexandre[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hilde Synnove Lid\u001b[39m ==> ENTITY: \u001b[32mHilde Synnøve Lid\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.091:0.091[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.048]; Sweden[0.045]; Lindgren[0.044]; Nannan[0.043]; Cuo[0.042]; Valenta[0.042]; Women[0.042]; Karin[0.042]; Austria[0.041]; Eric[0.041]; Canada[0.041]; Canada[0.041]; Canada[0.041]; Marie[0.041]; Australia[0.041]; Australia[0.041]; Marshall[0.041]; Kristie[0.041]; Jacqui[0.041]; Switzerland[0.041]; Switzerland[0.041]; Leu[0.040]; Rijavec[0.040]; Dmitri[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech Republic ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.169:0.169[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.044]; Switzerland[0.044]; Cup[0.044]; Cup[0.044]; Sweden[0.043]; Belarus[0.043]; Canada[0.043]; Canada[0.043]; Canada[0.043]; Canada[0.043]; Ales[0.042]; Norway[0.042]; Austria[0.042]; World[0.042]; World[0.042]; competition[0.041]; Skiing[0.040]; Russia[0.040]; Andy[0.040]; France[0.039]; France[0.039]; France[0.039]; Australia[0.039]; Australia[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Eric Bergoust\u001b[39m ==> ENTITY: \u001b[32mEric Bergoust\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \naerials[0.046]; Freestyle[0.046]; Aerials[0.045]; skiing[0.044]; Tignes[0.044]; freestyle[0.044]; Hilde[0.043]; Skiing[0.042]; Canada[0.041]; Canada[0.041]; Canada[0.041]; Canada[0.041]; Jeff[0.041]; Valenta[0.041]; Andy[0.041]; Dan[0.041]; Leu[0.040]; Austria[0.040]; Karin[0.040]; Nannan[0.040]; Men[0.039]; Dmitri[0.039]; Australia[0.039]; Australia[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.048]; Canada[0.045]; Canada[0.045]; Canada[0.045]; Canada[0.045]; Australia[0.044]; Australia[0.044]; Sweden[0.043]; Switzerland[0.042]; Switzerland[0.042]; Women[0.042]; Belarus[0.041]; Norway[0.041]; Austria[0.041]; Republic[0.040]; Russia[0.040]; Cuo[0.040]; France[0.040]; Czech[0.039]; Nannan[0.039]; points[0.039]; Xu[0.039]; Jeff[0.038]; Andy[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.176:0.176[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.048]; France[0.048]; Switzerland[0.045]; Switzerland[0.045]; Austria[0.042]; Australia[0.042]; Australia[0.042]; Belarus[0.042]; Republic[0.042]; Cup[0.042]; Cup[0.042]; Canada[0.041]; Canada[0.041]; Canada[0.041]; Canada[0.041]; Sweden[0.041]; Andy[0.040]; Sebastien[0.040]; Blanc[0.040]; competition[0.040]; points[0.040]; Russia[0.039]; China[0.039]; Eric[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Veronica Brenner\u001b[39m ==> ENTITY: \u001b[32mVeronica Brenner\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFreestyle[0.045]; freestyle[0.044]; Cup[0.044]; Cup[0.044]; Canada[0.043]; Canada[0.043]; Canada[0.043]; Canada[0.043]; aerials[0.043]; skiing[0.042]; Tignes[0.042]; Aerials[0.042]; World[0.041]; World[0.041]; Hilde[0.041]; Skiing[0.041]; Nannan[0.041]; Marie[0.040]; Norway[0.040]; Austria[0.039]; Valenta[0.039]; Jacqui[0.039]; Sweden[0.039]; Caroline[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.124:0.124[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.050]; Canada[0.044]; Canada[0.044]; Canada[0.044]; Canada[0.044]; Cup[0.043]; Switzerland[0.042]; Switzerland[0.042]; China[0.042]; China[0.042]; Austria[0.041]; points[0.041]; Women[0.041]; Norway[0.041]; Belarus[0.041]; Jacqui[0.040]; competition[0.040]; France[0.040]; France[0.040]; Marshall[0.040]; Russia[0.040]; Sweden[0.040]; Saturday[0.039]; Men[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.103:0.103[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.048]; Canada[0.048]; Canada[0.048]; Canada[0.048]; Switzerland[0.046]; Switzerland[0.046]; Belarus[0.045]; Sweden[0.044]; Russia[0.044]; Austria[0.044]; World[0.042]; World[0.042]; Norway[0.042]; Cup[0.042]; Cup[0.042]; Freestyle[0.042]; Australia[0.041]; Australia[0.041]; Republic[0.041]; China[0.041]; China[0.041]; Aerials[0.041]; competition[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.176:0.176[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.046]; Switzerland[0.046]; Canada[0.045]; Canada[0.045]; Canada[0.045]; Canada[0.045]; Norway[0.043]; Sweden[0.043]; Belarus[0.042]; competition[0.042]; Republic[0.041]; Women[0.041]; Czech[0.041]; Skiing[0.040]; France[0.040]; France[0.040]; France[0.040]; Eric[0.040]; skiing[0.040]; Ales[0.039]; World[0.039]; World[0.039]; Russia[0.039]; Tignes[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.111:0.111[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.049]; Switzerland[0.046]; Switzerland[0.046]; Austria[0.045]; Canada[0.044]; Canada[0.044]; Canada[0.044]; Women[0.042]; China[0.042]; China[0.042]; Russia[0.042]; Belarus[0.041]; Australia[0.041]; Australia[0.041]; Hilde[0.040]; Cuo[0.040]; Republic[0.040]; Worthington[0.040]; Kristie[0.039]; Czech[0.039]; Karin[0.039]; Andy[0.039]; Ales[0.038]; Nannan[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jeff Bean\u001b[39m ==> ENTITY: \u001b[32mJeff Bean\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWorld[0.045]; World[0.045]; Cup[0.044]; Cup[0.044]; Freestyle[0.044]; aerials[0.044]; competition[0.043]; Tignes[0.042]; freestyle[0.042]; Canada[0.042]; Canada[0.042]; Canada[0.042]; Canada[0.042]; points[0.041]; Aerials[0.041]; skiing[0.041]; Andy[0.041]; Skiing[0.041]; Sweden[0.040]; Eric[0.040]; Valenta[0.039]; China[0.039]; Austria[0.039]; Dan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Switzerland ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.178:0.178[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.047]; Canada[0.044]; Canada[0.044]; Canada[0.044]; Canada[0.044]; France[0.044]; France[0.044]; France[0.044]; Sweden[0.043]; Austria[0.042]; Belarus[0.041]; Russia[0.041]; points[0.040]; Australia[0.040]; Australia[0.040]; Men[0.040]; Norway[0.040]; Cup[0.040]; Cup[0.040]; skiing[0.039]; Aerials[0.039]; Andy[0.039]; Czech[0.039]; Republic[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.106:0.106[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.051]; Canada[0.044]; Canada[0.044]; Canada[0.044]; Canada[0.044]; Switzerland[0.042]; Switzerland[0.042]; China[0.042]; China[0.042]; Austria[0.042]; Women[0.041]; Norway[0.041]; Belarus[0.041]; Jacqui[0.041]; France[0.041]; Marshall[0.041]; Russia[0.041]; Sweden[0.040]; Kristie[0.039]; Czech[0.039]; Republic[0.039]; Cuo[0.039]; Cooper[0.039]; Worthington[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.157:0.157[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.051]; Canada[0.051]; Canada[0.051]; Australia[0.044]; Australia[0.044]; Sweden[0.041]; Cup[0.041]; Cup[0.041]; Belarus[0.041]; Women[0.041]; Switzerland[0.040]; Switzerland[0.040]; France[0.040]; France[0.040]; France[0.040]; World[0.040]; World[0.040]; competition[0.040]; Russia[0.039]; Men[0.039]; skiing[0.039]; Norway[0.039]; Skiing[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ales Valenta\u001b[39m ==> ENTITY: \u001b[32mAleš Valenta\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \naerials[0.047]; Aerials[0.046]; freestyle[0.044]; Freestyle[0.044]; skiing[0.044]; Tignes[0.043]; Skiing[0.043]; Hilde[0.042]; competition[0.042]; Sweden[0.041]; Canada[0.041]; Canada[0.041]; Canada[0.041]; Canada[0.041]; Czech[0.041]; Belarus[0.040]; Norway[0.040]; Switzerland[0.040]; Switzerland[0.040]; Nannan[0.040]; Karin[0.039]; Austria[0.039]; Men[0.039]; Eric[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.051]; Canada[0.051]; Canada[0.051]; Australia[0.044]; Australia[0.044]; Sweden[0.041]; Cup[0.041]; Cup[0.041]; Belarus[0.041]; Men[0.041]; Aerials[0.040]; Freestyle[0.040]; Switzerland[0.040]; Switzerland[0.040]; France[0.040]; France[0.040]; France[0.040]; World[0.040]; World[0.040]; Women[0.039]; competition[0.039]; China[0.039]; China[0.039]; Russia[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marie Lindgren\u001b[39m ==> ENTITY: \u001b[32mMarie Lindgren\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.102:0.102[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.046]; Norway[0.046]; Hilde[0.044]; Nannan[0.043]; Women[0.042]; Kuster[0.042]; Canada[0.041]; Canada[0.041]; Canada[0.041]; Canada[0.041]; Valenta[0.041]; Austria[0.041]; Belarus[0.041]; Switzerland[0.041]; Switzerland[0.041]; competition[0.041]; Karin[0.041]; Russia[0.041]; Evelyne[0.041]; Xu[0.041]; Men[0.040]; Eric[0.040]; Australia[0.040]; Australia[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dmitri Dashinski\u001b[39m ==> ENTITY: \u001b[32mDmitri Dashinski\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFreestyle[0.046]; Aerials[0.044]; aerials[0.044]; freestyle[0.044]; Belarus[0.044]; Russia[0.043]; Canada[0.042]; Canada[0.042]; Canada[0.042]; Canada[0.042]; skiing[0.042]; Skiing[0.042]; World[0.042]; World[0.042]; Tignes[0.041]; Men[0.041]; Hilde[0.041]; Cup[0.040]; Cup[0.040]; China[0.040]; China[0.040]; competition[0.039]; Norway[0.039]; Women[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.151:0.151[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.052]; Canada[0.052]; Canada[0.052]; Australia[0.045]; Australia[0.045]; Sweden[0.042]; Cup[0.041]; Belarus[0.041]; Women[0.041]; Switzerland[0.040]; Switzerland[0.040]; France[0.040]; France[0.040]; World[0.040]; competition[0.040]; Russia[0.040]; Men[0.040]; skiing[0.039]; Norway[0.039]; Saturday[0.039]; Republic[0.038]; Czech[0.038]; Austria[0.038]; aerials[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.157:0.157[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.051]; Canada[0.051]; Canada[0.051]; Australia[0.044]; Australia[0.044]; Sweden[0.041]; Cup[0.041]; Cup[0.041]; Belarus[0.041]; Women[0.041]; Switzerland[0.040]; Switzerland[0.040]; France[0.040]; France[0.040]; France[0.040]; World[0.040]; World[0.040]; competition[0.040]; Russia[0.039]; Men[0.039]; skiing[0.039]; Norway[0.039]; Skiing[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.S\u001b[39m ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.072:0.072[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.043]; Cup[0.043]; Canada[0.043]; Canada[0.043]; Canada[0.043]; Canada[0.043]; Switzerland[0.042]; Switzerland[0.042]; Kristie[0.042]; World[0.042]; World[0.042]; Women[0.042]; Tignes[0.042]; freestyle[0.042]; competition[0.041]; Australia[0.041]; Australia[0.041]; Hilde[0.041]; Aerials[0.040]; Freestyle[0.040]; Sweden[0.040]; Skiing[0.040]; Men[0.040]; Saturday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Xu Nannan\u001b[39m ==> ENTITY: \u001b[32mXu Nannan\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.075:0.075[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.047]; China[0.047]; Hilde[0.044]; Russia[0.043]; Valenta[0.042]; Belarus[0.042]; Canada[0.042]; Canada[0.042]; Canada[0.042]; Canada[0.042]; Norway[0.041]; Czech[0.041]; Karin[0.041]; Sweden[0.041]; Austria[0.041]; Eric[0.041]; Australia[0.040]; Australia[0.040]; Marshall[0.040]; Dmitri[0.040]; Kristie[0.040]; Jacqui[0.040]; Switzerland[0.040]; Switzerland[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tignes\u001b[39m ==> ENTITY: \u001b[32mTignes\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nskiing[0.047]; Saturday[0.044]; freestyle[0.043]; Skiing[0.043]; France[0.043]; France[0.043]; France[0.043]; Switzerland[0.043]; Switzerland[0.043]; Aerials[0.042]; Cup[0.042]; Cup[0.042]; Freestyle[0.042]; competition[0.042]; Sebastien[0.042]; aerials[0.040]; Valenta[0.040]; Alexandre[0.040]; Canada[0.040]; Canada[0.040]; Canada[0.040]; Karin[0.039]; Austria[0.039]; Men[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.178:0.178[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.048]; France[0.048]; Switzerland[0.045]; Switzerland[0.045]; Austria[0.042]; Australia[0.042]; Belarus[0.042]; Republic[0.042]; Cup[0.042]; Cup[0.042]; Canada[0.041]; Canada[0.041]; Canada[0.041]; Canada[0.041]; Sweden[0.041]; Andy[0.041]; Sebastien[0.040]; Blanc[0.040]; competition[0.040]; points[0.040]; Russia[0.040]; Eric[0.039]; World[0.039]; World[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Switzerland ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.165:0.165[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.048]; Canada[0.045]; Canada[0.045]; Canada[0.045]; Canada[0.045]; Sweden[0.043]; Austria[0.043]; Belarus[0.042]; Russia[0.042]; Men[0.041]; Jeff[0.041]; France[0.040]; France[0.040]; France[0.040]; Andy[0.040]; Cup[0.040]; Cup[0.040]; Freestyle[0.040]; Norway[0.040]; skiing[0.040]; Skiing[0.040]; Aerials[0.040]; Eric[0.040]; Czech[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.142:0.142[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.047]; Switzerland[0.047]; Switzerland[0.047]; Austria[0.045]; Canada[0.044]; Canada[0.044]; Canada[0.044]; Canada[0.044]; Russia[0.042]; Belarus[0.042]; Czech[0.040]; France[0.040]; France[0.040]; Cuo[0.040]; Men[0.040]; Ales[0.040]; Australia[0.040]; Australia[0.040]; points[0.040]; Karin[0.039]; Caroline[0.039]; Women[0.039]; Andy[0.039]; Republic[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belarus ==> ENTITY: \u001b[32mBelarus\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.422:-0.422[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.044]; Canada[0.044]; Canada[0.044]; Canada[0.044]; Russia[0.044]; Austria[0.043]; Sweden[0.043]; Aerials[0.043]; Norway[0.042]; Switzerland[0.042]; Switzerland[0.042]; Czech[0.042]; Republic[0.041]; Men[0.041]; Cup[0.040]; Cup[0.040]; France[0.040]; France[0.040]; France[0.040]; Women[0.040]; skiing[0.040]; points[0.040]; World[0.039]; World[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.111:0.111[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.046]; Canada[0.043]; Canada[0.043]; Canada[0.043]; Canada[0.043]; Eric[0.043]; Andy[0.043]; Switzerland[0.042]; Switzerland[0.042]; Austria[0.042]; Jeff[0.042]; points[0.042]; Sweden[0.042]; Cup[0.041]; Cup[0.041]; competition[0.041]; Norway[0.041]; Czech[0.040]; Dmitri[0.040]; Freestyle[0.040]; Dan[0.040]; Republic[0.040]; China[0.039]; China[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jacqui Cooper\u001b[39m ==> ENTITY: \u001b[32mJacqui Cooper\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.113:0.113[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfreestyle[0.044]; aerials[0.044]; competition[0.044]; Australia[0.044]; Australia[0.044]; skiing[0.043]; Women[0.043]; Hilde[0.043]; Marshall[0.042]; Cup[0.041]; World[0.041]; Canada[0.041]; Canada[0.041]; Canada[0.041]; Canada[0.041]; Valenta[0.041]; Norway[0.040]; Saturday[0.040]; Kristie[0.040]; Karin[0.040]; Russia[0.040]; Nannan[0.040]; Austria[0.040]; Men[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sebastien Foucras\u001b[39m ==> ENTITY: \u001b[32mSébastien Foucras\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nskiing[0.046]; Cup[0.045]; Cup[0.045]; Tignes[0.043]; Evelyne[0.043]; freestyle[0.042]; Freestyle[0.042]; Canada[0.042]; Canada[0.042]; Canada[0.042]; Canada[0.042]; aerials[0.042]; Saturday[0.041]; Aerials[0.041]; France[0.041]; France[0.041]; France[0.041]; Skiing[0.040]; Switzerland[0.040]; Switzerland[0.040]; Marie[0.040]; Austria[0.040]; Alexandre[0.039]; Kuster[0.039]; \t\n\r [=================================>............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 5s767ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3377/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1192testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pace\u001b[39m ==> ENTITY: \u001b[32mOrlando Pace\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlineman[0.045]; lineman[0.045]; defensive[0.044]; Bowl[0.043]; football[0.042]; college[0.042]; award[0.042]; tackle[0.042]; sophomore[0.042]; junior[0.042]; offensive[0.041]; Lombardi[0.041]; Lombardi[0.041]; Pace[0.041]; Pace[0.041]; Pace[0.041]; Pace[0.041]; win[0.041]; Houston[0.040]; Houston[0.040]; played[0.040]; Football[0.040]; Award[0.040]; Award[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rotary Club\u001b[39m ==> ENTITY: \u001b[32mRotary International\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.089:0.089[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNebraska[0.046]; Tech[0.045]; college[0.044]; sophomore[0.043]; award[0.043]; junior[0.043]; football[0.042]; country[0.041]; Cornell[0.041]; Football[0.041]; lineman[0.041]; lineman[0.041]; Houston[0.041]; Houston[0.041]; Award[0.041]; Award[0.041]; helped[0.041]; senior[0.041]; State[0.040]; State[0.040]; State[0.040]; State[0.040]; State[0.040]; American[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lombardi Award\u001b[39m ==> ENTITY: \u001b[32mLombardi Award\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlineman[0.044]; lineman[0.044]; lineman[0.044]; football[0.044]; Nebraska[0.044]; award[0.043]; defensive[0.043]; defensive[0.043]; defensive[0.043]; Lombardi[0.043]; Tech[0.042]; college[0.041]; college[0.041]; sophomore[0.041]; junior[0.041]; finalists[0.041]; Packers[0.040]; Houston[0.040]; State[0.040]; State[0.040]; State[0.040]; field[0.039]; Bowl[0.039]; Cornell[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cornell Brown\u001b[39m ==> ENTITY: \u001b[32mCornell Brown\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.217:0.217[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNebraska[0.046]; football[0.045]; Virginia[0.045]; college[0.045]; college[0.045]; defensive[0.045]; defensive[0.045]; defensive[0.045]; lineman[0.044]; lineman[0.044]; lineman[0.044]; Tech[0.044]; Bowl[0.043]; Houston[0.043]; sophomore[0.043]; coach[0.043]; junior[0.043]; Orlando[0.042]; Packers[0.042]; played[0.042]; tackle[0.041]; tackle[0.041]; field[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pace\u001b[39m ==> ENTITY: \u001b[32mOrlando Pace\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlineman[0.044]; lineman[0.044]; lineman[0.044]; Nebraska[0.044]; defensive[0.043]; defensive[0.043]; defensive[0.043]; Bowl[0.042]; Tech[0.042]; football[0.041]; college[0.041]; college[0.041]; award[0.041]; tackle[0.041]; tackle[0.041]; sophomore[0.041]; junior[0.041]; offensive[0.040]; offensive[0.040]; Packers[0.040]; Lombardi[0.040]; Lombardi[0.040]; Lombardi[0.040]; Lombardi[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mHouston\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHouston\u001b[39m <---> \u001b[32mHouston Cougars football\u001b[39m\t\nSCORES: global= 0.238:0.233[\u001b[31m0.005\u001b[39m]; local()= 0.066:0.155[\u001b[32m0.090\u001b[39m]; log p(e|m)= -0.323:-4.075[\u001b[31m3.752\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.048]; Football[0.044]; football[0.044]; Bowl[0.044]; record[0.043]; college[0.043]; Orlando[0.042]; Thursday[0.042]; lineman[0.041]; lineman[0.041]; played[0.041]; defensive[0.041]; defensive[0.041]; defensive[0.041]; tackle[0.041]; tackle[0.041]; Tech[0.041]; American[0.041]; Lombardi[0.040]; Lombardi[0.040]; Lombardi[0.040]; Arizona[0.040]; Arizona[0.040]; win[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Orlando Pace\u001b[39m ==> ENTITY: \u001b[32mOrlando Pace\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlineman[0.045]; lineman[0.045]; defensive[0.043]; defensive[0.043]; Bowl[0.043]; Tech[0.043]; football[0.042]; college[0.042]; award[0.042]; tackle[0.042]; sophomore[0.042]; junior[0.042]; offensive[0.041]; offensive[0.041]; Lombardi[0.041]; Lombardi[0.041]; Pace[0.041]; Pace[0.041]; Pace[0.041]; Pace[0.041]; win[0.040]; Houston[0.040]; Houston[0.040]; played[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Juan Roque\u001b[39m ==> ENTITY: \u001b[32mJuan Roque\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNebraska[0.049]; Tech[0.048]; Arizona[0.047]; Arizona[0.047]; college[0.047]; college[0.047]; sophomore[0.046]; junior[0.046]; offensive[0.046]; offensive[0.046]; Packers[0.045]; defensive[0.045]; defensive[0.045]; defensive[0.045]; football[0.045]; award[0.044]; Cornell[0.044]; played[0.044]; tackle[0.044]; lineman[0.044]; lineman[0.044]; lineman[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pace\u001b[39m ==> ENTITY: \u001b[32mOrlando Pace\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlineman[0.044]; lineman[0.044]; lineman[0.044]; Nebraska[0.044]; defensive[0.043]; defensive[0.043]; defensive[0.043]; Bowl[0.042]; Tech[0.042]; football[0.041]; college[0.041]; college[0.041]; award[0.041]; tackle[0.041]; tackle[0.041]; sophomore[0.041]; junior[0.041]; offensive[0.040]; offensive[0.040]; Packers[0.040]; Lombardi[0.040]; Lombardi[0.040]; Lombardi[0.040]; Lombardi[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Green Bay Packers ==> ENTITY: \u001b[32mGreen Bay Packers\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.167:0.167[0]; log p(e|m)= -0.100:-0.100[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwin[0.047]; field[0.047]; lineman[0.045]; lineman[0.045]; defensive[0.044]; defensive[0.044]; defensive[0.044]; Lombardi[0.044]; Lombardi[0.044]; goal[0.044]; played[0.044]; line[0.043]; line[0.043]; offensive[0.043]; offensive[0.043]; coach[0.043]; tackle[0.043]; Arizona[0.042]; award[0.042]; Jared[0.042]; Vince[0.041]; Brown[0.041]; outdistanced[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pace\u001b[39m ==> ENTITY: \u001b[32mOrlando Pace\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlineman[0.048]; lineman[0.048]; lineman[0.048]; Nebraska[0.048]; defensive[0.046]; defensive[0.046]; defensive[0.046]; Bowl[0.046]; Tech[0.046]; football[0.045]; college[0.045]; college[0.045]; award[0.045]; tackle[0.045]; tackle[0.045]; sophomore[0.045]; junior[0.044]; offensive[0.044]; offensive[0.044]; Lombardi[0.044]; Lombardi[0.044]; Lombardi[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NCAA ==> ENTITY: \u001b[32mNational Collegiate Athletic Association\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.149:0.149[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfootball[0.045]; played[0.044]; record[0.043]; junior[0.043]; berth[0.043]; college[0.043]; Football[0.043]; State[0.042]; State[0.042]; State[0.042]; State[0.042]; Bowl[0.042]; Thursday[0.041]; night[0.041]; Houston[0.041]; Houston[0.041]; winner[0.041]; Ohio[0.040]; Ohio[0.040]; Ohio[0.040]; goal[0.040]; American[0.040]; Arizona[0.040]; country[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Houston ==> ENTITY: \u001b[32mHouston\u001b[39m\t\nSCORES: global= 0.238:0.238[0]; local()= 0.070:0.070[0]; log p(e|m)= -0.323:-0.323[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.053]; Football[0.048]; football[0.048]; Bowl[0.048]; record[0.047]; college[0.047]; Orlando[0.046]; Thursday[0.045]; lineman[0.045]; lineman[0.045]; played[0.045]; defensive[0.044]; tackle[0.044]; Tech[0.044]; American[0.044]; Lombardi[0.044]; Lombardi[0.044]; Arizona[0.044]; win[0.044]; offensive[0.044]; country[0.043]; junior[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Virginia Tech ==> ENTITY: \u001b[32mVirginia Tech Hokies football\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.195:0.195[0]; log p(e|m)= -1.640:-1.640[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfootball[0.046]; college[0.044]; college[0.044]; Bowl[0.043]; record[0.042]; field[0.042]; award[0.042]; Cornell[0.042]; played[0.042]; State[0.041]; State[0.041]; State[0.041]; State[0.041]; win[0.041]; coach[0.041]; Thursday[0.041]; defensive[0.041]; defensive[0.041]; defensive[0.041]; Award[0.041]; Award[0.041]; lineman[0.041]; lineman[0.041]; lineman[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ohio State ==> ENTITY: \u001b[32mOhio State Buckeyes football\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.226:0.226[0]; log p(e|m)= -1.565:-1.565[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfootball[0.048]; field[0.048]; Ohio[0.046]; Ohio[0.046]; defensive[0.046]; defensive[0.046]; defensive[0.046]; Football[0.046]; Bowl[0.045]; offensive[0.045]; offensive[0.045]; played[0.045]; college[0.045]; college[0.045]; tackle[0.045]; tackle[0.045]; Cornell[0.045]; win[0.045]; State[0.045]; State[0.045]; State[0.045]; State[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lombardi Award\u001b[39m ==> ENTITY: \u001b[32mLombardi Award\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.223:0.223[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlineman[0.045]; lineman[0.045]; football[0.045]; award[0.043]; defensive[0.043]; defensive[0.043]; defensive[0.043]; Lombardi[0.043]; Tech[0.043]; Award[0.042]; Football[0.042]; college[0.041]; sophomore[0.041]; junior[0.041]; finalists[0.041]; Houston[0.040]; Houston[0.040]; State[0.040]; State[0.040]; State[0.040]; State[0.040]; State[0.040]; winner[0.040]; Bowl[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nebraska ==> ENTITY: \u001b[32mNebraska Cornhuskers football\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.195:0.195[0]; log p(e|m)= -3.037:-3.037[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncollege[0.044]; college[0.044]; football[0.044]; record[0.043]; defensive[0.042]; defensive[0.042]; defensive[0.042]; coach[0.042]; field[0.041]; win[0.041]; lineman[0.041]; lineman[0.041]; lineman[0.041]; State[0.041]; State[0.041]; State[0.041]; Bowl[0.041]; award[0.041]; junior[0.041]; played[0.041]; Virginia[0.041]; Brown[0.041]; Arizona[0.041]; Arizona[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Arizona State ==> ENTITY: \u001b[32mArizona State Sun Devils football\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.192:0.192[0]; log p(e|m)= -1.452:-1.452[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfootball[0.049]; Arizona[0.047]; State[0.047]; State[0.047]; Bowl[0.047]; college[0.047]; college[0.047]; field[0.046]; played[0.046]; lineman[0.045]; lineman[0.045]; lineman[0.045]; record[0.045]; tackle[0.044]; tackle[0.044]; Tech[0.044]; offensive[0.044]; offensive[0.044]; coach[0.044]; win[0.044]; Nebraska[0.044]; goal[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lombardi Award\u001b[39m ==> ENTITY: \u001b[32mLombardi Award\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.217:0.217[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlineman[0.045]; lineman[0.045]; football[0.045]; award[0.044]; defensive[0.044]; Lombardi[0.043]; Award[0.043]; Football[0.042]; college[0.042]; sophomore[0.042]; junior[0.042]; finalists[0.041]; Houston[0.041]; Houston[0.041]; State[0.040]; State[0.040]; State[0.040]; State[0.040]; winner[0.040]; Bowl[0.040]; offensive[0.040]; tackle[0.040]; senior[0.040]; win[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rose Bowl ==> ENTITY: \u001b[32mRose Bowl (stadium)\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.118:0.118[0]; log p(e|m)= -0.656:-0.656[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfield[0.052]; football[0.052]; defensive[0.051]; defensive[0.051]; defensive[0.051]; Football[0.051]; played[0.050]; Arizona[0.050]; Arizona[0.050]; win[0.050]; record[0.050]; goal[0.050]; Houston[0.050]; Houston[0.050]; Thursday[0.049]; offensive[0.049]; offensive[0.049]; college[0.049]; college[0.049]; left[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Vince Lombardi\u001b[39m ==> ENTITY: \u001b[32mVince Lombardi\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNebraska[0.046]; coach[0.044]; Tech[0.044]; Lombardi[0.044]; college[0.043]; sophomore[0.043]; played[0.042]; Packers[0.042]; lineman[0.042]; lineman[0.042]; field[0.042]; Bowl[0.042]; defensive[0.041]; defensive[0.041]; defensive[0.041]; Cornell[0.041]; award[0.041]; Bay[0.040]; win[0.040]; senior[0.040]; State[0.040]; State[0.040]; legendary[0.040]; addition[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ohio State ==> ENTITY: \u001b[32mOhio State Buckeyes football\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.195:0.195[0]; log p(e|m)= -1.565:-1.565[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfootball[0.045]; Ohio[0.043]; Ohio[0.043]; defensive[0.043]; defensive[0.043]; Football[0.043]; Bowl[0.042]; Thursday[0.042]; offensive[0.042]; played[0.042]; college[0.042]; Cornell[0.041]; State[0.041]; State[0.041]; State[0.041]; lineman[0.041]; lineman[0.041]; Tech[0.041]; win[0.041]; berth[0.041]; Virginia[0.041]; American[0.041]; record[0.040]; Arizona[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jared Tomich\u001b[39m ==> ENTITY: \u001b[32mJared Tomich\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNebraska[0.045]; football[0.045]; Tech[0.044]; played[0.043]; college[0.043]; college[0.043]; sophomore[0.043]; junior[0.042]; Packers[0.042]; Bay[0.042]; field[0.042]; Ohio[0.041]; defensive[0.041]; defensive[0.041]; defensive[0.041]; Cornell[0.041]; lineman[0.040]; lineman[0.040]; lineman[0.040]; Arizona[0.040]; Arizona[0.040]; State[0.040]; State[0.040]; State[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Arizona State ==> ENTITY: \u001b[32mArizona State Sun Devils football\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.202:0.202[0]; log p(e|m)= -1.452:-1.452[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfootball[0.045]; Arizona[0.043]; State[0.043]; State[0.043]; State[0.043]; State[0.043]; Bowl[0.043]; college[0.042]; college[0.042]; Football[0.042]; field[0.042]; played[0.042]; lineman[0.041]; lineman[0.041]; lineman[0.041]; record[0.041]; tackle[0.041]; tackle[0.041]; Tech[0.040]; offensive[0.040]; offensive[0.040]; coach[0.040]; win[0.040]; Nebraska[0.040]; \t\n\r [=================================>............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 5s688ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3401/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1355testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Brian Shimer\u001b[39m ==> ENTITY: \u001b[32mBrian Shimer\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nman[0.046]; bobsleigh[0.045]; Lueders[0.044]; Cup[0.044]; Cup[0.044]; event[0.044]; Bobsleigh[0.043]; Jiri[0.043]; Reto[0.042]; World[0.042]; World[0.042]; Man[0.042]; Igls[0.042]; Dave[0.041]; Dirk[0.041]; Sepp[0.040]; seconds[0.040]; Saturday[0.040]; Pavel[0.040]; Thomas[0.039]; Guenther[0.039]; Randy[0.039]; Jakobs[0.039]; MacEachern[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sepp Dostthaler\u001b[39m ==> ENTITY: \u001b[32mSepp Dostthaler\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.208:0.208[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbobsleigh[0.047]; event[0.046]; Bobsleigh[0.045]; Igls[0.045]; Cup[0.043]; Cup[0.043]; World[0.043]; World[0.043]; man[0.042]; German[0.042]; Lueders[0.042]; Man[0.042]; Jiri[0.041]; Reto[0.041]; Austria[0.041]; Austria[0.041]; Austria[0.041]; Hannes[0.040]; Dirk[0.040]; Germany[0.039]; Shimer[0.039]; Olsson[0.039]; Czech[0.039]; Pavel[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Garrett Hines\u001b[39m ==> ENTITY: \u001b[32mGarrett Hines\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= -0.012:-0.012[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJiri[0.046]; Reto[0.046]; Hannes[0.045]; Dirk[0.044]; Jim[0.043]; Pavel[0.043]; Olsson[0.043]; Georg[0.042]; Arnold[0.042]; Jakobs[0.042]; States[0.041]; Britain[0.041]; Germany[0.041]; Austria[0.041]; Austria[0.041]; Erwin[0.041]; Sean[0.041]; Switzerland[0.040]; Thomas[0.040]; Sepp[0.040]; Czech[0.039]; Ward[0.039]; Dean[0.039]; Marco[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hubert Schoesser\u001b[39m ==> ENTITY: \u001b[32mHubert Schösser\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.121:0.121[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLueders[0.047]; Austria[0.045]; Austria[0.045]; Jiri[0.044]; Reto[0.044]; Hannes[0.043]; Dirk[0.043]; Switzerland[0.042]; Canada[0.042]; Sepp[0.042]; Pavel[0.041]; Olsson[0.041]; Guenther[0.041]; Georg[0.041]; Germany[0.041]; Arnold[0.041]; Jakobs[0.040]; Shimer[0.040]; Czech[0.040]; German[0.040]; Jim[0.039]; Thomas[0.039]; Dave[0.039]; Erwin[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reto Goetschi\u001b[39m ==> ENTITY: \u001b[32mReto Götschi\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nevent[0.044]; Lueders[0.044]; bobsleigh[0.044]; Jiri[0.043]; Igls[0.043]; Bobsleigh[0.043]; Guido[0.043]; man[0.042]; Hannes[0.042]; Acklin[0.042]; Sepp[0.042]; Dirk[0.042]; Switzerland[0.042]; Marco[0.041]; Man[0.041]; Cup[0.041]; Cup[0.041]; seconds[0.041]; Austria[0.040]; Austria[0.040]; Austria[0.040]; Pavel[0.040]; Guenther[0.040]; Olsson[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pierre Lueders\u001b[39m ==> ENTITY: \u001b[32mPierre Lueders\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.116:0.116[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbobsleigh[0.047]; Canada[0.046]; Jiri[0.044]; Reto[0.043]; Igls[0.043]; Bobsleigh[0.043]; Dirk[0.042]; Dave[0.042]; Cup[0.042]; Cup[0.042]; World[0.042]; World[0.042]; Shimer[0.041]; event[0.041]; Pavel[0.041]; Guenther[0.040]; Olsson[0.040]; man[0.040]; Jim[0.040]; MacEachern[0.040]; Jakobs[0.040]; Brian[0.040]; Randy[0.040]; Britain[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sean Olsson\u001b[39m ==> ENTITY: \u001b[32mSean Olsson\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLueders[0.045]; Switzerland[0.044]; Canada[0.044]; Austria[0.044]; Austria[0.044]; Britain[0.044]; Jiri[0.043]; Hubert[0.043]; Reto[0.043]; Hannes[0.042]; Dirk[0.042]; Germany[0.041]; Czech[0.041]; Thomas[0.041]; Jim[0.041]; Sepp[0.041]; Pavel[0.040]; Dave[0.040]; Guenther[0.040]; Georg[0.040]; Ward[0.040]; Jakobs[0.040]; Garrett[0.040]; Pierre[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Guido Acklin\u001b[39m ==> ENTITY: \u001b[32mGuido Acklin\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbobsleigh[0.046]; event[0.045]; man[0.044]; Lueders[0.044]; Igls[0.043]; Reto[0.043]; Jiri[0.043]; Marco[0.042]; Bobsleigh[0.042]; Man[0.042]; Hannes[0.042]; Switzerland[0.041]; World[0.041]; World[0.041]; Dirk[0.041]; seconds[0.041]; Sepp[0.041]; Shimer[0.040]; Italy[0.040]; Czech[0.040]; Pavel[0.040]; Guenther[0.040]; Olsson[0.040]; Georg[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.048]; Germany[0.047]; Canada[0.047]; German[0.044]; Italy[0.043]; Igls[0.043]; Lueders[0.042]; Sepp[0.042]; event[0.042]; Marco[0.042]; Reto[0.041]; bobsleigh[0.041]; Pierre[0.041]; World[0.041]; World[0.041]; Cup[0.040]; Cup[0.040]; Brian[0.040]; Guido[0.040]; Saturday[0.040]; Huber[0.039]; Bobsleigh[0.039]; Randy[0.039]; minute[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Guenther Huber\u001b[39m ==> ENTITY: \u001b[32mGünther Huber\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbobsleigh[0.046]; Lueders[0.045]; event[0.044]; Bobsleigh[0.043]; World[0.043]; World[0.043]; man[0.043]; Igls[0.043]; Jiri[0.042]; Reto[0.042]; Sepp[0.042]; Shimer[0.041]; Cup[0.041]; Cup[0.041]; Dirk[0.041]; Man[0.041]; Canada[0.041]; Guido[0.040]; German[0.040]; Pavel[0.040]; seconds[0.040]; Austria[0.040]; Austria[0.040]; Germany[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dean Ward\u001b[39m ==> ENTITY: \u001b[32mDean Ward\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.082:0.082[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.046]; Jiri[0.045]; Reto[0.045]; Hannes[0.044]; Lueders[0.044]; Dirk[0.043]; Canada[0.043]; Pavel[0.042]; Olsson[0.042]; Georg[0.042]; Dave[0.042]; Jakobs[0.041]; Austria[0.041]; Austria[0.041]; Hubert[0.041]; MacEachern[0.041]; Germany[0.040]; Switzerland[0.040]; Erwin[0.040]; Garrett[0.040]; Sean[0.040]; Jim[0.039]; Sepp[0.039]; Czech[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dirk Wiese\u001b[39m ==> ENTITY: \u001b[32mDirk Wiese\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.218:0.218[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBobsleigh[0.051]; bobsleigh[0.049]; Igls[0.048]; event[0.048]; Lueders[0.048]; man[0.048]; Man[0.046]; Sepp[0.045]; World[0.045]; World[0.045]; Cup[0.045]; Cup[0.045]; Jiri[0.045]; Reto[0.045]; German[0.044]; Huber[0.044]; Shimer[0.044]; Hannes[0.044]; Olsson[0.043]; Germany[0.043]; Czech[0.043]; Georg[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dave MacEachern\u001b[39m ==> ENTITY: \u001b[32mDavid MacEachern\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.048]; bobsleigh[0.047]; Lueders[0.046]; Igls[0.045]; man[0.043]; Bobsleigh[0.042]; Jiri[0.042]; Reto[0.042]; Shimer[0.042]; event[0.042]; Sepp[0.042]; World[0.041]; World[0.041]; Dirk[0.041]; Pierre[0.040]; Man[0.040]; Austria[0.040]; Austria[0.040]; Italy[0.040]; Britain[0.039]; Marco[0.039]; Brian[0.039]; Switzerland[0.039]; Jim[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Randy Jones ==> ENTITY: \u001b[32mRandy Jones (bobsleigh)\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.117:0.117[0]; log p(e|m)= -1.224:-1.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbobsleigh[0.044]; Lueders[0.044]; Saturday[0.044]; Canada[0.043]; Jiri[0.043]; Reto[0.043]; Igls[0.043]; Dave[0.042]; event[0.042]; Thomas[0.042]; Brian[0.042]; Dirk[0.042]; World[0.042]; World[0.042]; Shimer[0.041]; Pierre[0.041]; seconds[0.040]; Pavel[0.040]; Guenther[0.040]; Bobsleigh[0.040]; man[0.040]; States[0.040]; Jakobs[0.040]; Man[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jiri Dzmura\u001b[39m ==> ENTITY: \u001b[32mJiří Džmura\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbobsleigh[0.045]; Lueders[0.044]; Reto[0.044]; Czech[0.044]; event[0.043]; Hannes[0.043]; Dirk[0.043]; Marco[0.042]; Canada[0.042]; man[0.042]; Republic[0.042]; Italy[0.041]; Pavel[0.041]; Hubert[0.041]; Guenther[0.041]; Olsson[0.041]; Georg[0.041]; Sepp[0.041]; World[0.040]; Austria[0.040]; Austria[0.040]; Jakobs[0.040]; Switzerland[0.040]; Germany[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Antonio Tartaglia\u001b[39m ==> ENTITY: \u001b[32mAntonio Tartaglia\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbobsleigh[0.046]; event[0.045]; Lueders[0.043]; man[0.043]; Igls[0.043]; Guido[0.043]; Jiri[0.043]; Reto[0.043]; Marco[0.042]; Huber[0.042]; Italy[0.042]; Canada[0.042]; Dirk[0.041]; Bobsleigh[0.041]; World[0.040]; World[0.040]; Austria[0.040]; Austria[0.040]; Sepp[0.040]; Shimer[0.040]; Man[0.040]; Pavel[0.040]; Cup[0.040]; Cup[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pavel Polomsky\u001b[39m ==> ENTITY: \u001b[32mPavel Polomský\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCzech[0.045]; bobsleigh[0.045]; Lueders[0.045]; Canada[0.044]; Jiri[0.043]; Republic[0.043]; Reto[0.043]; Austria[0.043]; Austria[0.043]; event[0.042]; Hannes[0.042]; Dirk[0.042]; man[0.041]; Switzerland[0.041]; Brian[0.040]; Sean[0.040]; Guenther[0.040]; Olsson[0.040]; Georg[0.040]; Dave[0.040]; Hubert[0.040]; Jakobs[0.040]; Germany[0.039]; Sepp[0.039]; \t\n\r [==================================>...........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 5s629ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3418/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1363testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.054:0.054[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.046]; league[0.045]; Ekoku[0.044]; Southampton[0.044]; Middlesbrough[0.044]; Blackburn[0.043]; Soccer[0.043]; Yorke[0.042]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; Arsenal[0.042]; Chelsea[0.041]; Alan[0.041]; Everton[0.041]; Ravanelli[0.040]; Aston[0.040]; Ian[0.040]; Saturday[0.040]; Robbie[0.040]; Robbie[0.040]; Goalscorers[0.040]; Chris[0.040]; Gayle[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Steve McManaman\u001b[39m ==> ENTITY: \u001b[32mSteve McManaman\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEkoku[0.045]; Gary[0.044]; Alan[0.043]; Robbie[0.043]; Robbie[0.043]; league[0.042]; Ravanelli[0.042]; Vialli[0.042]; Villa[0.042]; Ian[0.042]; Blackburn[0.042]; Chris[0.042]; Fowler[0.041]; Chelsea[0.041]; matches[0.041]; Liverpool[0.041]; Liverpool[0.041]; Newcastle[0.041]; Newcastle[0.041]; Newcastle[0.041]; Newcastle[0.041]; Middlesbrough[0.040]; Arsenal[0.040]; Shearer[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Les Ferdinand\u001b[39m ==> ENTITY: \u001b[32mLes Ferdinand\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthampton[0.046]; Alan[0.045]; Middlesbrough[0.045]; Ekoku[0.045]; Steve[0.045]; matches[0.045]; Robbie[0.044]; Robbie[0.044]; league[0.044]; Gary[0.044]; Ian[0.043]; Chris[0.043]; Blackburn[0.043]; Vialli[0.043]; Chelsea[0.043]; Aston[0.043]; Ravanelli[0.043]; Everton[0.042]; Soccer[0.042]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Liverpool ==> ENTITY: \u001b[32mLiverpool F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.181:0.181[0]; log p(e|m)= -1.171:-1.171[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; matches[0.044]; Ekoku[0.044]; Blackburn[0.043]; Liverpool[0.043]; Middlesbrough[0.043]; Southampton[0.043]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; Chelsea[0.041]; Matthew[0.041]; Everton[0.041]; Sutton[0.041]; Arsenal[0.041]; Ravanelli[0.041]; Fowler[0.041]; Robbie[0.041]; Robbie[0.041]; Villa[0.040]; Ian[0.040]; Alan[0.040]; Peter[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Les Ferdinand\u001b[39m ==> ENTITY: \u001b[32mLes Ferdinand\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthampton[0.046]; Alan[0.045]; Middlesbrough[0.045]; Ekoku[0.045]; Steve[0.045]; matches[0.045]; Robbie[0.044]; Robbie[0.044]; league[0.044]; Gary[0.044]; Ian[0.043]; Chris[0.043]; Blackburn[0.043]; Vialli[0.043]; Chelsea[0.043]; Aston[0.043]; Ravanelli[0.043]; Everton[0.042]; Soccer[0.042]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Newcastle ==> ENTITY: \u001b[32mNewcastle United F.C.\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.189:0.189[0]; log p(e|m)= -2.489:-2.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.045]; matches[0.044]; Ekoku[0.043]; Blackburn[0.043]; Southampton[0.043]; Middlesbrough[0.043]; Newcastle[0.043]; Newcastle[0.043]; Newcastle[0.043]; Soccer[0.042]; Alan[0.042]; Steve[0.041]; Liverpool[0.041]; Liverpool[0.041]; London[0.041]; Robbie[0.041]; Robbie[0.041]; Ravanelli[0.040]; Gary[0.040]; Chelsea[0.040]; Ian[0.040]; Everton[0.040]; Chris[0.039]; McManaman[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Blackburn ==> ENTITY: \u001b[32mBlackburn Rovers F.C.\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.185:0.185[0]; log p(e|m)= -2.733:-2.733[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; Southampton[0.044]; matches[0.044]; Middlesbrough[0.044]; Ekoku[0.043]; Newcastle[0.043]; Newcastle[0.043]; Newcastle[0.043]; Newcastle[0.043]; Alan[0.042]; Ian[0.042]; Everton[0.041]; Liverpool[0.041]; Liverpool[0.041]; Ravanelli[0.041]; English[0.041]; English[0.041]; Aston[0.041]; Villa[0.040]; Sutton[0.040]; Steve[0.040]; Arsenal[0.040]; McManaman[0.040]; Fowler[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fabrizio Ravanelli\u001b[39m ==> ENTITY: \u001b[32mFabrizio Ravanelli\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.046]; league[0.045]; Chelsea[0.044]; Vialli[0.044]; Ekoku[0.044]; Gianluca[0.044]; Southampton[0.042]; Middlesbrough[0.041]; Blackburn[0.041]; Steve[0.041]; Newcastle[0.041]; Newcastle[0.041]; Newcastle[0.041]; Newcastle[0.041]; Arsenal[0.041]; Villa[0.041]; goalscorers[0.041]; Liverpool[0.040]; Liverpool[0.040]; Saturday[0.040]; Goalscorers[0.040]; Robbie[0.040]; Robbie[0.040]; Alan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chelsea ==> ENTITY: \u001b[32mChelsea F.C.\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.176:0.176[0]; log p(e|m)= -0.518:-0.518[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.045]; London[0.044]; matches[0.044]; Soccer[0.043]; Southampton[0.043]; Blackburn[0.043]; Middlesbrough[0.043]; Liverpool[0.042]; Liverpool[0.042]; Ekoku[0.042]; Newcastle[0.041]; Newcastle[0.041]; Newcastle[0.041]; Newcastle[0.041]; Ravanelli[0.041]; Arsenal[0.041]; Everton[0.041]; Sutton[0.041]; Aston[0.040]; Chris[0.040]; Alan[0.040]; Steve[0.040]; Vialli[0.040]; Fowler[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gary Speed\u001b[39m ==> ENTITY: \u001b[32mGary Speed\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthampton[0.044]; Blackburn[0.044]; Ekoku[0.043]; Alan[0.043]; Steve[0.043]; Middlesbrough[0.042]; Robbie[0.042]; Robbie[0.042]; Ian[0.042]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; league[0.041]; Chris[0.041]; matches[0.041]; Fowler[0.041]; Aston[0.041]; Goalscorers[0.041]; Shearer[0.040]; Sutton[0.040]; Liverpool[0.040]; Liverpool[0.040]; Everton[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Efan Ekoku\u001b[39m ==> ENTITY: \u001b[32mEfan Ekoku\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.208:0.208[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; Steve[0.045]; matches[0.044]; Southampton[0.044]; Gary[0.043]; Middlesbrough[0.043]; Alan[0.042]; Chris[0.042]; Robbie[0.042]; Robbie[0.042]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; Ian[0.042]; Soccer[0.041]; Blackburn[0.041]; Sutton[0.041]; Liverpool[0.040]; Liverpool[0.040]; Chelsea[0.040]; Matthew[0.039]; Everton[0.039]; London[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Southampton ==> ENTITY: \u001b[32mSouthampton F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.150:0.150[0]; log p(e|m)= -1.149:-1.149[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLiverpool[0.045]; Liverpool[0.045]; Middlesbrough[0.044]; Newcastle[0.044]; Newcastle[0.044]; Newcastle[0.044]; Newcastle[0.044]; Chelsea[0.043]; Blackburn[0.043]; matches[0.042]; league[0.042]; Aston[0.042]; Yorke[0.041]; Villa[0.041]; Arsenal[0.041]; Everton[0.041]; Ekoku[0.041]; Vialli[0.040]; Soccer[0.040]; Gayle[0.039]; Steve[0.039]; Shearer[0.039]; London[0.039]; McManaman[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Peter Beardsley\u001b[39m ==> ENTITY: \u001b[32mPeter Beardsley\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.206:0.206[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEkoku[0.044]; league[0.044]; Steve[0.044]; matches[0.044]; Middlesbrough[0.043]; Southampton[0.043]; Alan[0.043]; Blackburn[0.042]; Robbie[0.042]; Robbie[0.042]; Gary[0.042]; Ian[0.041]; Ravanelli[0.041]; Liverpool[0.041]; Liverpool[0.041]; Newcastle[0.041]; Newcastle[0.041]; Newcastle[0.041]; Newcastle[0.041]; Shearer[0.040]; Fowler[0.040]; Chris[0.040]; Chelsea[0.040]; McManaman[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Middlesbrough ==> ENTITY: \u001b[32mMiddlesbrough F.C.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.185:0.185[0]; log p(e|m)= -0.518:-0.518[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEkoku[0.047]; league[0.046]; Blackburn[0.046]; Southampton[0.046]; matches[0.046]; Soccer[0.045]; Alan[0.044]; Newcastle[0.043]; Newcastle[0.043]; Newcastle[0.043]; Newcastle[0.043]; Steve[0.043]; Ravanelli[0.043]; Arsenal[0.043]; Chelsea[0.043]; Chris[0.043]; Sutton[0.042]; Goalscorers[0.042]; Fowler[0.042]; Everton[0.042]; Gary[0.042]; Liverpool[0.042]; Liverpool[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Newcastle ==> ENTITY: \u001b[32mNewcastle United F.C.\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.189:0.189[0]; log p(e|m)= -2.489:-2.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.045]; matches[0.044]; Ekoku[0.043]; Blackburn[0.043]; Southampton[0.043]; Middlesbrough[0.043]; Newcastle[0.043]; Newcastle[0.043]; Newcastle[0.043]; Soccer[0.042]; Alan[0.042]; Steve[0.041]; Liverpool[0.041]; Liverpool[0.041]; London[0.041]; Robbie[0.041]; Robbie[0.041]; Ravanelli[0.040]; Gary[0.040]; Chelsea[0.040]; Ian[0.040]; Everton[0.040]; Chris[0.039]; McManaman[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Newcastle ==> ENTITY: \u001b[32mNewcastle United F.C.\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.187:0.187[0]; log p(e|m)= -2.489:-2.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; Ekoku[0.044]; Blackburn[0.044]; Southampton[0.044]; Middlesbrough[0.044]; Newcastle[0.043]; Newcastle[0.043]; Newcastle[0.043]; Alan[0.042]; Steve[0.042]; Liverpool[0.042]; Liverpool[0.042]; Robbie[0.041]; Robbie[0.041]; Ravanelli[0.041]; Gary[0.041]; Chelsea[0.041]; Ian[0.041]; Everton[0.040]; Chris[0.040]; McManaman[0.040]; Sutton[0.040]; Arsenal[0.040]; Aston[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = English ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.113:0.113[0]; log p(e|m)= -1.355:-1.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEnglish[0.048]; league[0.047]; matches[0.045]; Southampton[0.043]; Ekoku[0.042]; Alan[0.042]; Robbie[0.042]; Robbie[0.042]; Middlesbrough[0.042]; London[0.042]; Matthew[0.041]; Soccer[0.041]; Ian[0.041]; Newcastle[0.041]; Newcastle[0.041]; Newcastle[0.041]; Gary[0.040]; Gayle[0.040]; Steve[0.040]; Aston[0.040]; Saturday[0.040]; Blackburn[0.040]; Chris[0.040]; Liverpool[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Robbie Fowler\u001b[39m ==> ENTITY: \u001b[32mRobbie Fowler\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEkoku[0.045]; Steve[0.044]; Robbie[0.043]; Alan[0.043]; league[0.042]; Ian[0.042]; Gary[0.042]; Middlesbrough[0.042]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; matches[0.042]; Chris[0.042]; Liverpool[0.041]; Liverpool[0.041]; Blackburn[0.041]; Southampton[0.041]; McManaman[0.041]; Matthew[0.040]; Ravanelli[0.040]; Everton[0.039]; Chelsea[0.039]; Sutton[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Robbie Earle\u001b[39m ==> ENTITY: \u001b[32mRobbie Earle\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEkoku[0.045]; Steve[0.044]; Robbie[0.044]; Southampton[0.043]; Ian[0.043]; Gary[0.043]; matches[0.043]; Chris[0.042]; Matthew[0.042]; Alan[0.042]; league[0.042]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; Blackburn[0.041]; Middlesbrough[0.041]; Peter[0.041]; Fowler[0.041]; Soccer[0.040]; Gayle[0.040]; Sutton[0.039]; Marcus[0.039]; Shearer[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Aston Villa\u001b[39m ==> ENTITY: \u001b[32mAston Villa F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; Southampton[0.044]; matches[0.044]; Blackburn[0.044]; Middlesbrough[0.043]; Ekoku[0.043]; Goalscorers[0.043]; Soccer[0.042]; Fowler[0.041]; Chelsea[0.041]; Sutton[0.041]; Alan[0.041]; Newcastle[0.041]; Newcastle[0.041]; Newcastle[0.041]; Newcastle[0.041]; Liverpool[0.041]; Liverpool[0.041]; McManaman[0.041]; Steve[0.041]; Everton[0.040]; Ian[0.040]; goalscorers[0.040]; Shearer[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alan Shearer\u001b[39m ==> ENTITY: \u001b[32mAlan Shearer\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEkoku[0.045]; Steve[0.044]; Robbie[0.043]; Robbie[0.043]; league[0.043]; Southampton[0.042]; Middlesbrough[0.042]; matches[0.042]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; Chris[0.042]; Gary[0.042]; Ian[0.042]; Liverpool[0.041]; Liverpool[0.041]; Ravanelli[0.041]; Chelsea[0.041]; Blackburn[0.041]; Soccer[0.040]; Sutton[0.040]; Matthew[0.040]; Beardsley[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Everton ==> ENTITY: \u001b[32mEverton F.C.\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.187:0.187[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; matches[0.044]; Blackburn[0.044]; Ekoku[0.044]; Southampton[0.043]; Middlesbrough[0.043]; Soccer[0.043]; Steve[0.042]; Aston[0.041]; Ian[0.041]; Gayle[0.041]; Alan[0.041]; Newcastle[0.041]; Newcastle[0.041]; Newcastle[0.041]; Newcastle[0.041]; Liverpool[0.041]; Liverpool[0.041]; Wright[0.041]; Gary[0.040]; Robbie[0.040]; Robbie[0.040]; Ravanelli[0.040]; Chelsea[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wimbledon ==> ENTITY: \u001b[32mWimbledon F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.189:0.189[0]; log p(e|m)= -1.890:-1.890[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWimbledon[0.049]; Wimbledon[0.049]; matches[0.047]; Southampton[0.046]; league[0.045]; London[0.045]; Middlesbrough[0.044]; Blackburn[0.044]; Ekoku[0.043]; Soccer[0.043]; Alan[0.042]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; Liverpool[0.042]; Liverpool[0.042]; Steve[0.042]; Peter[0.042]; Chris[0.042]; Gary[0.041]; Aston[0.041]; Ravanelli[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marcus Gayle\u001b[39m ==> ENTITY: \u001b[32mMarcus Gayle\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; Steve[0.045]; matches[0.044]; Gary[0.044]; Southampton[0.044]; Ekoku[0.044]; Chris[0.043]; Robbie[0.043]; Robbie[0.043]; Middlesbrough[0.042]; Ian[0.042]; Blackburn[0.042]; Alan[0.042]; Matthew[0.041]; Sutton[0.040]; Fowler[0.040]; Goalscorers[0.040]; Chelsea[0.040]; Vialli[0.040]; Newcastle[0.039]; Newcastle[0.039]; Newcastle[0.039]; Newcastle[0.039]; Soccer[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Newcastle ==> ENTITY: \u001b[32mNewcastle United F.C.\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.189:0.189[0]; log p(e|m)= -2.489:-2.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.045]; matches[0.044]; Ekoku[0.043]; Blackburn[0.043]; Southampton[0.043]; Middlesbrough[0.043]; Newcastle[0.043]; Newcastle[0.043]; Newcastle[0.043]; Soccer[0.042]; Alan[0.042]; Steve[0.041]; Liverpool[0.041]; Liverpool[0.041]; London[0.041]; Robbie[0.041]; Robbie[0.041]; Ravanelli[0.040]; Gary[0.040]; Chelsea[0.040]; Ian[0.040]; Everton[0.040]; Chris[0.039]; McManaman[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Matthew Le Tissier\u001b[39m ==> ENTITY: \u001b[32mMatthew Le Tissier\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEkoku[0.045]; Southampton[0.044]; league[0.044]; Steve[0.044]; matches[0.044]; Robbie[0.043]; Robbie[0.043]; Alan[0.042]; Ian[0.042]; Gary[0.042]; Chris[0.042]; Middlesbrough[0.042]; Soccer[0.041]; Newcastle[0.041]; Newcastle[0.041]; Newcastle[0.041]; Newcastle[0.041]; Goalscorers[0.040]; Blackburn[0.040]; Shearer[0.040]; Peter[0.040]; Saturday[0.040]; Yorke[0.039]; Sutton[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Liverpool ==> ENTITY: \u001b[32mLiverpool F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.181:0.181[0]; log p(e|m)= -1.171:-1.171[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; matches[0.044]; Ekoku[0.044]; Blackburn[0.043]; Liverpool[0.043]; Middlesbrough[0.043]; Southampton[0.043]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; Chelsea[0.041]; Matthew[0.041]; Everton[0.041]; Sutton[0.041]; Arsenal[0.041]; Ravanelli[0.041]; Fowler[0.041]; Robbie[0.041]; Robbie[0.041]; Villa[0.040]; Ian[0.040]; Alan[0.040]; Peter[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = English ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.109:0.109[0]; log p(e|m)= -1.355:-1.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEnglish[0.049]; league[0.047]; matches[0.045]; Southampton[0.044]; Ekoku[0.043]; Alan[0.042]; Robbie[0.042]; Middlesbrough[0.042]; London[0.042]; Matthew[0.042]; Soccer[0.042]; Ian[0.041]; Newcastle[0.041]; Newcastle[0.041]; Newcastle[0.041]; Gary[0.041]; Gayle[0.040]; Aston[0.040]; Saturday[0.040]; Blackburn[0.040]; Chris[0.040]; Sutton[0.039]; Goalscorers[0.039]; Earle[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ian Wright ==> ENTITY: \u001b[32mIan Wright\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.040:-0.040[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; Southampton[0.046]; Alan[0.043]; Ekoku[0.043]; Middlesbrough[0.043]; Gary[0.042]; Steve[0.042]; Blackburn[0.042]; Robbie[0.042]; Robbie[0.042]; matches[0.041]; London[0.041]; Chelsea[0.041]; Shearer[0.041]; Newcastle[0.041]; Newcastle[0.041]; Newcastle[0.041]; Newcastle[0.041]; English[0.041]; English[0.041]; Soccer[0.040]; Liverpool[0.040]; Liverpool[0.040]; Everton[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wimbledon ==> ENTITY: \u001b[32mWimbledon F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.189:0.189[0]; log p(e|m)= -1.890:-1.890[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWimbledon[0.049]; Wimbledon[0.049]; matches[0.047]; Southampton[0.046]; league[0.045]; London[0.045]; Middlesbrough[0.044]; Blackburn[0.044]; Ekoku[0.043]; Soccer[0.043]; Alan[0.042]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; Liverpool[0.042]; Liverpool[0.042]; Steve[0.042]; Peter[0.042]; Chris[0.042]; Gary[0.041]; Aston[0.041]; Ravanelli[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chris Sutton\u001b[39m ==> ENTITY: \u001b[32mChris Sutton\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; league[0.044]; Southampton[0.044]; Alan[0.043]; Ekoku[0.043]; Steve[0.043]; Blackburn[0.043]; Ian[0.042]; Robbie[0.042]; Robbie[0.042]; Ravanelli[0.042]; Soccer[0.042]; Gary[0.042]; Middlesbrough[0.042]; Peter[0.041]; Matthew[0.041]; Chelsea[0.041]; Liverpool[0.040]; Liverpool[0.040]; Aston[0.040]; Newcastle[0.040]; Newcastle[0.040]; Newcastle[0.040]; Newcastle[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gianluca Vialli\u001b[39m ==> ENTITY: \u001b[32mGianluca Vialli\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; matches[0.044]; Soccer[0.044]; Chelsea[0.043]; Ravanelli[0.043]; Southampton[0.043]; Liverpool[0.043]; Liverpool[0.043]; Ekoku[0.042]; Middlesbrough[0.042]; Blackburn[0.042]; Arsenal[0.042]; Newcastle[0.041]; Newcastle[0.041]; Newcastle[0.041]; Newcastle[0.041]; Villa[0.041]; Everton[0.041]; goalscorers[0.040]; Goalscorers[0.040]; McManaman[0.040]; Steve[0.040]; Robbie[0.039]; Robbie[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dwight Yorke\u001b[39m ==> ENTITY: \u001b[32mDwight Yorke\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.050]; Ekoku[0.049]; Southampton[0.049]; Chelsea[0.049]; Steve[0.049]; Alan[0.049]; Middlesbrough[0.048]; Liverpool[0.048]; Liverpool[0.048]; Gary[0.048]; league[0.047]; McManaman[0.047]; Soccer[0.047]; Peter[0.047]; Robbie[0.047]; Robbie[0.047]; Villa[0.047]; Aston[0.047]; Everton[0.046]; Ravanelli[0.046]; Chris[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wimbledon ==> ENTITY: \u001b[32mWimbledon F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.189:0.189[0]; log p(e|m)= -1.890:-1.890[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWimbledon[0.049]; Wimbledon[0.049]; matches[0.047]; Southampton[0.046]; league[0.045]; London[0.045]; Middlesbrough[0.044]; Blackburn[0.044]; Ekoku[0.043]; Soccer[0.043]; Alan[0.042]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; Newcastle[0.042]; Liverpool[0.042]; Liverpool[0.042]; Steve[0.042]; Peter[0.042]; Chris[0.042]; Gary[0.041]; Aston[0.041]; Ravanelli[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Arsenal ==> ENTITY: \u001b[32mArsenal F.C.\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.167:0.167[0]; log p(e|m)= -0.158:-0.158[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.045]; matches[0.045]; Ekoku[0.043]; Chelsea[0.043]; Ian[0.043]; Robbie[0.042]; Robbie[0.042]; Soccer[0.042]; Middlesbrough[0.042]; Southampton[0.042]; Liverpool[0.042]; Liverpool[0.042]; Steve[0.042]; Blackburn[0.041]; Villa[0.041]; Alan[0.041]; Gary[0.041]; Chris[0.040]; Newcastle[0.040]; Newcastle[0.040]; Newcastle[0.040]; Newcastle[0.040]; Ravanelli[0.040]; Goalscorers[0.040]; \t\n\r [==================================>...........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 5s354ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3453/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1362testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Glasgow ==> ENTITY: \u001b[32mGlasgow\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.190:0.190[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScottish[0.046]; Scottish[0.046]; Aberdeen[0.045]; Aberdeen[0.045]; Dundee[0.044]; Colin[0.043]; Dean[0.043]; matches[0.042]; Dunfermline[0.042]; Cameron[0.042]; Kilmarnock[0.041]; McCoist[0.041]; Celtic[0.041]; Celtic[0.041]; Robert[0.040]; Billy[0.040]; Peter[0.040]; Brian[0.040]; Darren[0.040]; Rangers[0.040]; Rangers[0.040]; Rangers[0.040]; Rangers[0.040]; Hibernian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rangers ==> ENTITY: \u001b[32mRangers F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.202:0.202[0]; log p(e|m)= -0.302:-0.302[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.044]; Rangers[0.043]; Rangers[0.043]; Rangers[0.043]; Aberdeen[0.043]; Aberdeen[0.043]; Glasgow[0.043]; Scottish[0.043]; Scottish[0.043]; Dundee[0.043]; Celtic[0.042]; Celtic[0.042]; Celtic[0.042]; Dunfermline[0.042]; Hearts[0.042]; Hibernian[0.041]; Soccer[0.041]; Kilmarnock[0.041]; McCoist[0.040]; Colin[0.040]; Saturday[0.039]; United[0.039]; Pierre[0.039]; Brian[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Celtic ==> ENTITY: \u001b[32mCeltic F.C.\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.198:0.198[0]; log p(e|m)= -0.978:-0.978[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGlasgow[0.044]; Dunfermline[0.044]; matches[0.043]; Scottish[0.043]; Scottish[0.043]; Dundee[0.043]; Aberdeen[0.043]; Aberdeen[0.043]; Kilmarnock[0.043]; Celtic[0.042]; Celtic[0.042]; Rangers[0.042]; Rangers[0.042]; Rangers[0.042]; Rangers[0.042]; Hibernian[0.041]; Soccer[0.041]; Hearts[0.040]; McCoist[0.040]; Brian[0.039]; Division[0.039]; division[0.039]; Colin[0.039]; United[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Paul Wright ==> ENTITY: \u001b[32mPaul Wright (footballer)\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.931:-0.931[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDundee[0.045]; Kilmarnock[0.045]; Aberdeen[0.044]; Aberdeen[0.044]; Glasgow[0.044]; Dunfermline[0.044]; Scottish[0.043]; Scottish[0.043]; Soccer[0.042]; Hibernian[0.042]; Brian[0.042]; Thom[0.041]; Celtic[0.041]; Celtic[0.041]; Celtic[0.041]; Paul[0.040]; Robert[0.040]; Dean[0.040]; Rangers[0.040]; Rangers[0.040]; Rangers[0.040]; Rangers[0.040]; Colin[0.040]; Hearts[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paolo Di Canio\u001b[39m ==> ENTITY: \u001b[32mPaolo Di Canio\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.116:0.116[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; Rangers[0.043]; Rangers[0.043]; Rangers[0.043]; Rangers[0.043]; Paul[0.043]; Paul[0.043]; Colin[0.043]; Peter[0.042]; United[0.041]; Darren[0.041]; Gascoigne[0.041]; Celtic[0.041]; Celtic[0.041]; Celtic[0.041]; Brian[0.041]; Windass[0.041]; Andreas[0.041]; Billy[0.040]; Laudrup[0.040]; goalscorers[0.040]; Scorers[0.040]; Aberdeen[0.040]; Aberdeen[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Celtic ==> ENTITY: \u001b[32mCeltic F.C.\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.198:0.198[0]; log p(e|m)= -0.978:-0.978[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGlasgow[0.044]; Dunfermline[0.044]; matches[0.043]; Scottish[0.043]; Scottish[0.043]; Dundee[0.043]; Aberdeen[0.043]; Aberdeen[0.043]; Kilmarnock[0.043]; Celtic[0.042]; Celtic[0.042]; Rangers[0.042]; Rangers[0.042]; Rangers[0.042]; Rangers[0.042]; Hibernian[0.041]; Soccer[0.041]; Hearts[0.040]; McCoist[0.040]; Brian[0.039]; Division[0.039]; division[0.039]; Colin[0.039]; United[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Darren Jackson\u001b[39m ==> ENTITY: \u001b[32mDarren Jackson\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.236:0.236[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDundee[0.045]; Glasgow[0.043]; Hibernian[0.043]; Celtic[0.042]; Celtic[0.042]; Celtic[0.042]; Aberdeen[0.042]; Aberdeen[0.042]; Rangers[0.042]; Rangers[0.042]; Rangers[0.042]; Rangers[0.042]; Kilmarnock[0.042]; McCoist[0.042]; Hearts[0.042]; Scottish[0.042]; Scottish[0.042]; Dunfermline[0.042]; Colin[0.041]; Billy[0.041]; Brian[0.040]; United[0.039]; matches[0.038]; Soccer[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hearts ==> ENTITY: \u001b[32mHeart of Midlothian F.C.\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.229:0.229[0]; log p(e|m)= -0.298:-0.298[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAberdeen[0.046]; Aberdeen[0.046]; Dundee[0.045]; matches[0.045]; Glasgow[0.045]; Hibernian[0.044]; Rangers[0.044]; Rangers[0.044]; Rangers[0.044]; Rangers[0.044]; Celtic[0.044]; Celtic[0.044]; Celtic[0.044]; Scottish[0.044]; Scottish[0.044]; Dunfermline[0.044]; Kilmarnock[0.043]; Soccer[0.043]; McCoist[0.041]; Darren[0.041]; Premier[0.041]; Colin[0.041]; Saturday[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Aberdeen ==> ENTITY: \u001b[32mAberdeen F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.194:0.194[0]; log p(e|m)= -1.448:-1.448[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScottish[0.045]; Scottish[0.045]; Aberdeen[0.044]; matches[0.044]; Dundee[0.044]; Glasgow[0.043]; Dunfermline[0.043]; Hibernian[0.042]; Kilmarnock[0.042]; Rangers[0.041]; Rangers[0.041]; Rangers[0.041]; Rangers[0.041]; Celtic[0.041]; Celtic[0.041]; Celtic[0.041]; Hearts[0.041]; Gerry[0.041]; Soccer[0.040]; Colin[0.040]; Peter[0.040]; Darren[0.040]; Premier[0.039]; McCoist[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ally McCoist\u001b[39m ==> ENTITY: \u001b[32mAlly McCoist\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.044]; Rangers[0.043]; Rangers[0.043]; Rangers[0.043]; Rangers[0.043]; Brian[0.042]; Aberdeen[0.042]; Aberdeen[0.042]; Darren[0.042]; Colin[0.042]; Soccer[0.041]; Billy[0.041]; Dundee[0.041]; Hibernian[0.041]; Dunfermline[0.041]; Celtic[0.041]; Celtic[0.041]; Celtic[0.041]; Scottish[0.041]; Scottish[0.041]; Kilmarnock[0.040]; Gascoigne[0.040]; Glasgow[0.040]; Gerry[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paul Gascoigne\u001b[39m ==> ENTITY: \u001b[32mPaul Gascoigne\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.044]; Darren[0.043]; Scottish[0.043]; Scottish[0.043]; Paul[0.043]; Dundee[0.042]; Colin[0.042]; Aberdeen[0.042]; Aberdeen[0.042]; Celtic[0.042]; Celtic[0.042]; Celtic[0.042]; Hearts[0.042]; Billy[0.041]; Brian[0.041]; Glasgow[0.041]; Peter[0.041]; Robert[0.041]; Hibernian[0.041]; Dunfermline[0.041]; Dean[0.041]; McCoist[0.040]; United[0.040]; Andreas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gerry Britton\u001b[39m ==> ENTITY: \u001b[32mGerry Britton\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAberdeen[0.046]; Aberdeen[0.046]; Rangers[0.043]; Rangers[0.043]; Rangers[0.043]; Rangers[0.043]; matches[0.043]; Dunfermline[0.042]; Dundee[0.042]; Hibernian[0.042]; Kilmarnock[0.042]; Celtic[0.041]; Celtic[0.041]; Celtic[0.041]; Glasgow[0.041]; Hearts[0.041]; Billy[0.041]; McCoist[0.040]; Soccer[0.040]; Colin[0.040]; Scottish[0.040]; Scottish[0.040]; Darren[0.040]; Brian[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Billy Dodds\u001b[39m ==> ENTITY: \u001b[32mBilly Dodds\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAberdeen[0.045]; Aberdeen[0.045]; Dundee[0.043]; Scottish[0.043]; Scottish[0.043]; Rangers[0.043]; Rangers[0.043]; Rangers[0.043]; Rangers[0.043]; matches[0.042]; Glasgow[0.042]; Celtic[0.041]; Celtic[0.041]; Celtic[0.041]; Dunfermline[0.041]; Kilmarnock[0.041]; Darren[0.041]; McCoist[0.040]; Colin[0.040]; Soccer[0.040]; Brian[0.040]; Hibernian[0.040]; United[0.040]; Hearts[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Celtic ==> ENTITY: \u001b[32mCeltic F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.193:0.193[0]; log p(e|m)= -0.978:-0.978[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDunfermline[0.044]; matches[0.044]; Scottish[0.044]; Dundee[0.044]; Aberdeen[0.044]; Aberdeen[0.044]; Kilmarnock[0.043]; Celtic[0.043]; Celtic[0.043]; Rangers[0.043]; Rangers[0.043]; Rangers[0.043]; Rangers[0.043]; Hibernian[0.042]; Hearts[0.040]; McCoist[0.040]; Brian[0.040]; division[0.040]; Colin[0.040]; United[0.039]; Billy[0.039]; Saturday[0.039]; Peter[0.039]; Thom[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hibernian\u001b[39m ==> ENTITY: \u001b[32mHibernian F.C.\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAberdeen[0.044]; Aberdeen[0.044]; Celtic[0.043]; Celtic[0.043]; Celtic[0.043]; Dundee[0.043]; matches[0.043]; Glasgow[0.042]; Scottish[0.042]; Scottish[0.042]; Dunfermline[0.042]; Kilmarnock[0.042]; Rangers[0.042]; Rangers[0.042]; Rangers[0.042]; Rangers[0.042]; Hearts[0.041]; Gerry[0.040]; Soccer[0.040]; Brian[0.040]; Darren[0.039]; Colin[0.039]; Paul[0.039]; Paul[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andreas Thom\u001b[39m ==> ENTITY: \u001b[32mAndreas Thom\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDivision[0.046]; matches[0.046]; Soccer[0.045]; division[0.043]; Celtic[0.042]; Celtic[0.042]; Celtic[0.042]; Saturday[0.042]; Rangers[0.042]; Rangers[0.042]; Rangers[0.042]; Rangers[0.042]; goalscorers[0.041]; Paul[0.041]; Paul[0.041]; Peter[0.041]; Aberdeen[0.041]; Aberdeen[0.041]; Glasgow[0.040]; United[0.040]; Laudrup[0.040]; Scorers[0.040]; Brian[0.040]; Premier[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rangers ==> ENTITY: \u001b[32mRangers F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.211:0.211[0]; log p(e|m)= -0.302:-0.302[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.044]; Rangers[0.043]; Rangers[0.043]; Rangers[0.043]; Aberdeen[0.043]; Aberdeen[0.043]; Glasgow[0.043]; Scottish[0.043]; Scottish[0.043]; Dundee[0.043]; Celtic[0.042]; Celtic[0.042]; Celtic[0.042]; Dunfermline[0.042]; Hearts[0.041]; Hibernian[0.041]; Soccer[0.041]; Kilmarnock[0.041]; McCoist[0.040]; Colin[0.040]; Saturday[0.039]; United[0.039]; Brian[0.039]; division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Robert Winters ==> ENTITY: \u001b[32mRobbie Winters\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAberdeen[0.044]; Aberdeen[0.044]; Rangers[0.043]; Rangers[0.043]; Rangers[0.043]; Rangers[0.043]; Celtic[0.042]; Celtic[0.042]; Celtic[0.042]; Hibernian[0.042]; Dundee[0.042]; Premier[0.041]; Hearts[0.041]; Dunfermline[0.041]; matches[0.041]; Billy[0.041]; premier[0.041]; Kilmarnock[0.040]; McCoist[0.040]; Cameron[0.040]; Paul[0.040]; Paul[0.040]; Gerry[0.040]; Laudrup[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Brian Laudrup\u001b[39m ==> ENTITY: \u001b[32mBrian Laudrup\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.043]; Peter[0.043]; matches[0.043]; Colin[0.043]; Paul[0.042]; Paul[0.042]; Aberdeen[0.042]; Aberdeen[0.042]; Celtic[0.042]; Celtic[0.042]; Celtic[0.042]; Rangers[0.042]; Rangers[0.042]; Rangers[0.042]; Rangers[0.042]; Darren[0.041]; Robert[0.041]; Scottish[0.041]; Scottish[0.041]; Glasgow[0.040]; Hibernian[0.040]; Hearts[0.040]; McCoist[0.040]; goalscorers[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dunfermline ==> ENTITY: \u001b[32mDunfermline Athletic F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.240:0.240[0]; log p(e|m)= -1.995:-1.995[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAberdeen[0.045]; Aberdeen[0.045]; Dundee[0.045]; Kilmarnock[0.043]; matches[0.043]; Celtic[0.043]; Celtic[0.043]; Celtic[0.043]; Rangers[0.042]; Rangers[0.042]; Rangers[0.042]; Rangers[0.042]; Glasgow[0.042]; Hibernian[0.042]; Hearts[0.042]; Scottish[0.041]; Scottish[0.041]; Soccer[0.041]; United[0.039]; Division[0.039]; McCoist[0.039]; Premier[0.038]; Colin[0.038]; goalscorers[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dean Windass\u001b[39m ==> ENTITY: \u001b[32mDean Windass\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDarren[0.045]; Colin[0.044]; matches[0.043]; Brian[0.043]; Soccer[0.043]; Aberdeen[0.042]; Aberdeen[0.042]; Paul[0.042]; Paul[0.042]; Dundee[0.042]; Saturday[0.042]; Peter[0.042]; United[0.042]; division[0.041]; Gascoigne[0.041]; Division[0.041]; Premier[0.040]; Rangers[0.040]; Rangers[0.040]; Rangers[0.040]; Rangers[0.040]; Dodds[0.040]; Andreas[0.040]; Cameron[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kilmarnock ==> ENTITY: \u001b[32mKilmarnock F.C.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.245:0.245[0]; log p(e|m)= -1.431:-1.431[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAberdeen[0.045]; Aberdeen[0.045]; Dundee[0.044]; matches[0.043]; Celtic[0.043]; Celtic[0.043]; Celtic[0.043]; Dunfermline[0.043]; Glasgow[0.043]; Hibernian[0.043]; Scottish[0.042]; Scottish[0.042]; Rangers[0.041]; Rangers[0.041]; Rangers[0.041]; Rangers[0.041]; Hearts[0.041]; McCoist[0.040]; Soccer[0.040]; Colin[0.040]; United[0.039]; Premier[0.039]; Darren[0.039]; Billy[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Scottish Premier Division\u001b[39m ==> ENTITY: \u001b[32mScottish Premier League\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDunfermline[0.044]; Aberdeen[0.043]; Aberdeen[0.043]; Rangers[0.043]; Rangers[0.043]; Rangers[0.043]; Rangers[0.043]; Hibernian[0.043]; Celtic[0.043]; Celtic[0.043]; Dundee[0.043]; Scottish[0.042]; Hearts[0.042]; Kilmarnock[0.042]; matches[0.042]; Glasgow[0.042]; Soccer[0.041]; McCoist[0.040]; division[0.040]; Saturday[0.040]; Colin[0.039]; Darren[0.039]; United[0.038]; Brian[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Aberdeen ==> ENTITY: \u001b[32mAberdeen F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.194:0.194[0]; log p(e|m)= -1.448:-1.448[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScottish[0.045]; Scottish[0.045]; Aberdeen[0.044]; matches[0.044]; Dundee[0.044]; Glasgow[0.043]; Dunfermline[0.043]; Hibernian[0.042]; Kilmarnock[0.042]; Rangers[0.041]; Rangers[0.041]; Rangers[0.041]; Rangers[0.041]; Celtic[0.041]; Celtic[0.041]; Celtic[0.041]; Hearts[0.041]; Gerry[0.041]; Soccer[0.040]; Colin[0.040]; Peter[0.040]; Darren[0.040]; Premier[0.039]; McCoist[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Peter van Vossen\u001b[39m ==> ENTITY: \u001b[32mPeter van Vossen\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRangers[0.044]; Rangers[0.044]; Rangers[0.044]; Rangers[0.044]; Celtic[0.043]; Celtic[0.043]; Celtic[0.043]; Van[0.043]; Hearts[0.042]; Aberdeen[0.042]; Aberdeen[0.042]; matches[0.042]; goalscorers[0.041]; Soccer[0.041]; Division[0.041]; Andreas[0.040]; Hibernian[0.040]; Scottish[0.040]; Scottish[0.040]; Dunfermline[0.040]; premier[0.040]; Dundee[0.040]; Darren[0.040]; Scorers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dundee United\u001b[39m ==> ENTITY: \u001b[32mDundee United F.C.\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAberdeen[0.045]; Aberdeen[0.045]; matches[0.043]; Hibernian[0.043]; Glasgow[0.043]; Celtic[0.043]; Celtic[0.043]; Celtic[0.043]; Rangers[0.042]; Rangers[0.042]; Rangers[0.042]; Rangers[0.042]; Scottish[0.042]; Dunfermline[0.042]; Kilmarnock[0.042]; Hearts[0.041]; Brian[0.040]; Colin[0.040]; Division[0.040]; McCoist[0.040]; Premier[0.040]; Paul[0.040]; Paul[0.040]; Darren[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rangers ==> ENTITY: \u001b[32mRangers F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.202:0.202[0]; log p(e|m)= -0.302:-0.302[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.044]; Rangers[0.043]; Rangers[0.043]; Rangers[0.043]; Aberdeen[0.043]; Aberdeen[0.043]; Glasgow[0.043]; Scottish[0.043]; Scottish[0.043]; Dundee[0.043]; Celtic[0.042]; Celtic[0.042]; Celtic[0.042]; Dunfermline[0.042]; Hearts[0.042]; Hibernian[0.041]; Soccer[0.041]; Kilmarnock[0.041]; McCoist[0.040]; Colin[0.040]; Saturday[0.039]; United[0.039]; Pierre[0.039]; Brian[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rangers ==> ENTITY: \u001b[32mRangers F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.211:0.211[0]; log p(e|m)= -0.302:-0.302[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.044]; Rangers[0.043]; Rangers[0.043]; Rangers[0.043]; Aberdeen[0.043]; Aberdeen[0.043]; Glasgow[0.043]; Scottish[0.043]; Scottish[0.043]; Dundee[0.043]; Celtic[0.042]; Celtic[0.042]; Celtic[0.042]; Dunfermline[0.042]; Hearts[0.041]; Hibernian[0.041]; Soccer[0.041]; Kilmarnock[0.041]; McCoist[0.040]; Colin[0.040]; Saturday[0.039]; United[0.039]; Brian[0.039]; division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colin Cameron ==> ENTITY: \u001b[32mColin Cameron (footballer)\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.170:0.170[0]; log p(e|m)= -0.361:-0.361[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDundee[0.044]; Rangers[0.044]; Rangers[0.044]; Rangers[0.044]; Rangers[0.044]; Aberdeen[0.043]; Aberdeen[0.043]; matches[0.042]; Dunfermline[0.042]; Scottish[0.042]; Scottish[0.042]; Hibernian[0.042]; Kilmarnock[0.041]; Darren[0.041]; Glasgow[0.041]; Celtic[0.041]; Celtic[0.041]; Celtic[0.041]; Soccer[0.040]; Division[0.040]; Hearts[0.040]; McCoist[0.040]; Brian[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pierre Van Hooydonk\u001b[39m ==> ENTITY: \u001b[32mPierre van Hooijdonk\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCeltic[0.044]; Celtic[0.044]; Celtic[0.044]; matches[0.044]; Rangers[0.043]; Rangers[0.043]; Rangers[0.043]; Rangers[0.043]; Laudrup[0.042]; Soccer[0.042]; Hearts[0.042]; Aberdeen[0.041]; Aberdeen[0.041]; Hibernian[0.041]; Paul[0.041]; Paul[0.041]; United[0.041]; Dundee[0.041]; Peter[0.041]; Saturday[0.040]; Andreas[0.040]; goalscorers[0.040]; McCoist[0.039]; Scorers[0.039]; \t\n\r [==================================>...........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 5s165ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3483/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1237testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = USDA\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Agriculture\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.021:0.021[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThursday[0.053]; slaughter[0.053]; compared[0.053]; live[0.052]; Friday[0.052]; estimated[0.052]; gross[0.051]; offal[0.051]; offal[0.051]; hide[0.050]; hide[0.050]; typical[0.050]; value[0.049]; value[0.049]; value[0.049]; cwt[0.049]; dn[0.049]; steer[0.047]; cutout[0.046]; Moines[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = USDA\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Agriculture\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.021:0.021[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThursday[0.053]; slaughter[0.053]; compared[0.053]; live[0.052]; Friday[0.052]; estimated[0.052]; gross[0.051]; offal[0.051]; offal[0.051]; hide[0.050]; hide[0.050]; typical[0.050]; value[0.049]; value[0.049]; value[0.049]; cwt[0.049]; dn[0.049]; steer[0.047]; cutout[0.046]; Moines[0.046]; \t\n\r [==================================>...........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 5s160ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3485/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1213testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.110:0.110[0]; log p(e|m)= -1.207:-1.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.048]; Czech[0.047]; Czech[0.047]; Prague[0.047]; Havel[0.046]; Havel[0.046]; Havel[0.046]; Native[0.045]; origins[0.043]; Vaclav[0.043]; distinguished[0.043]; States[0.042]; States[0.042]; United[0.042]; United[0.042]; Secretary[0.042]; Secretary[0.042]; Praises[0.042]; diplomat[0.041]; Friday[0.041]; welcomed[0.041]; President[0.041]; Klara[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Josef Zieleniec\u001b[39m ==> ENTITY: \u001b[32mJosef Zieleniec\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.110:0.110[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nambassador[0.044]; ambassador[0.044]; studied[0.044]; academic[0.043]; lectured[0.043]; strategic[0.043]; diplomats[0.042]; Karel[0.042]; Minister[0.042]; Foreign[0.042]; problems[0.042]; integration[0.041]; Europe[0.041]; Europe[0.041]; Mlada[0.041]; policy[0.041]; foreign[0.041]; country[0.041]; Nato[0.041]; Nato[0.041]; warned[0.040]; diplomatic[0.040]; Czech[0.040]; Czech[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Czechoslovak\u001b[39m ==> ENTITY: \u001b[32mCzechoslovakia\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.135:0.135[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.046]; Czech[0.045]; Prague[0.044]; Communist[0.043]; democratic[0.043]; Europe[0.043]; Communists[0.042]; government[0.042]; Havel[0.042]; Havel[0.042]; Havel[0.042]; ambassador[0.042]; countries[0.041]; States[0.041]; States[0.041]; bloc[0.041]; foreign[0.040]; foreign[0.040]; ousted[0.040]; policy[0.040]; policy[0.040]; regime[0.040]; strengthening[0.040]; strong[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.040:0.040[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npolicy[0.044]; told[0.044]; told[0.044]; country[0.043]; warned[0.043]; Europe[0.042]; Europe[0.042]; academic[0.042]; strategic[0.042]; countries[0.042]; Dnes[0.041]; key[0.041]; Josef[0.041]; Foreign[0.041]; foreign[0.041]; Karel[0.041]; security[0.041]; lectured[0.041]; heaven[0.041]; Czech[0.041]; Czech[0.041]; diplomats[0.041]; coup[0.040]; Kovanda[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Havel\u001b[39m ==> ENTITY: \u001b[32mVáclav Havel\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHavel[0.048]; Havel[0.048]; Havel[0.048]; Havel[0.048]; Prague[0.045]; Prague[0.045]; Czech[0.043]; Czech[0.043]; President[0.042]; Czechoslovak[0.041]; Vaclav[0.041]; lung[0.040]; hospital[0.040]; surgery[0.040]; academic[0.040]; State[0.039]; State[0.039]; Communist[0.039]; invited[0.039]; Communists[0.039]; democratic[0.038]; appointment[0.038]; Secretary[0.038]; Secretary[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.047:0.047[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.047]; United[0.046]; States[0.046]; Friday[0.045]; position[0.045]; ousted[0.044]; Prague[0.044]; Prague[0.044]; Czech[0.044]; Czech[0.044]; Czech[0.044]; helped[0.044]; hold[0.043]; foreign[0.043]; forward[0.043]; said[0.042]; said[0.042]; Friend[0.042]; diplomat[0.042]; Albright[0.042]; Albright[0.042]; Albright[0.042]; Albright[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NATO\u001b[39m ==> ENTITY: \u001b[32mNATO\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncooperation[0.046]; Nato[0.045]; strategic[0.044]; Europe[0.043]; Europe[0.043]; included[0.042]; security[0.042]; policy[0.042]; policy[0.042]; bloc[0.042]; alliance[0.041]; ambassador[0.041]; ambassador[0.041]; country[0.041]; strengthening[0.041]; enlargement[0.041]; selection[0.041]; countries[0.041]; States[0.041]; seeking[0.041]; government[0.040]; Atlantic[0.040]; key[0.040]; expansion[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Vaclav Havel\u001b[39m ==> ENTITY: \u001b[32mVáclav Havel\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHavel[0.049]; Havel[0.049]; Havel[0.049]; Havel[0.049]; Prague[0.047]; Czech[0.044]; Czech[0.044]; Czech[0.044]; President[0.043]; surgery[0.041]; State[0.040]; State[0.040]; appointment[0.039]; Secretary[0.039]; Secretary[0.039]; statement[0.039]; statement[0.039]; welcomed[0.038]; wish[0.038]; said[0.038]; said[0.038]; distinguished[0.038]; American[0.038]; Praises[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Havel\u001b[39m ==> ENTITY: \u001b[32mVáclav Havel\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHavel[0.050]; Havel[0.050]; Prague[0.047]; Czech[0.045]; Czechoslovak[0.043]; lung[0.042]; hospital[0.042]; surgery[0.041]; academic[0.041]; ambassador[0.041]; State[0.041]; supporter[0.041]; Communist[0.041]; invited[0.040]; Communists[0.040]; country[0.040]; democratic[0.040]; praised[0.040]; integration[0.040]; Secretary[0.039]; statement[0.039]; Nato[0.039]; regime[0.039]; wish[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCzechs\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mCzech Republic\u001b[39m <---> \u001b[32mCzechs\u001b[39m\t\nSCORES: global= 0.257:0.253[\u001b[31m0.005\u001b[39m]; local()= 0.088:0.044[\u001b[31m0.045\u001b[39m]; log p(e|m)= -2.207:-0.024[\u001b[32m2.183\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurope[0.046]; Czech[0.045]; bloc[0.044]; Karel[0.044]; policy[0.043]; countries[0.043]; ambassador[0.043]; Josef[0.042]; Dnes[0.042]; cooperation[0.042]; key[0.042]; influence[0.042]; little[0.041]; United[0.041]; alliance[0.041]; security[0.041]; enlargement[0.040]; strengthening[0.040]; Nato[0.040]; creation[0.040]; expecting[0.040]; foreign[0.040]; strategic[0.040]; daily[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.063:0.063[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.046]; ousted[0.045]; Europe[0.043]; Europe[0.043]; States[0.043]; seeking[0.042]; expected[0.042]; included[0.042]; selection[0.042]; United[0.041]; United[0.041]; United[0.041]; statement[0.041]; Prague[0.041]; Nations[0.041]; supporter[0.041]; Albright[0.041]; Albright[0.041]; Albright[0.041]; Albright[0.041]; Albright[0.041]; ambassador[0.041]; ambassador[0.041]; Atlantic[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Havel\u001b[39m ==> ENTITY: \u001b[32mVáclav Havel\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHavel[0.050]; Havel[0.050]; Havel[0.050]; Havel[0.050]; Prague[0.047]; Prague[0.047]; Czech[0.045]; Czech[0.045]; President[0.044]; Czechoslovak[0.043]; Vaclav[0.042]; lung[0.042]; hospital[0.042]; surgery[0.041]; academic[0.041]; ambassador[0.041]; State[0.041]; State[0.041]; Communist[0.041]; invited[0.040]; Communists[0.040]; democratic[0.040]; appointment[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Madeleine Albright\u001b[39m ==> ENTITY: \u001b[32mMadeleine Albright\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSecretary[0.046]; Secretary[0.046]; President[0.045]; American[0.043]; policy[0.043]; lung[0.042]; Albright[0.042]; Albright[0.042]; Albright[0.042]; surgery[0.041]; invited[0.041]; States[0.041]; States[0.041]; statement[0.041]; statement[0.041]; Vaclav[0.041]; foreign[0.041]; State[0.041]; State[0.041]; distinguished[0.040]; appointment[0.040]; diplomat[0.040]; Prague[0.040]; Prague[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Albright\u001b[39m ==> ENTITY: \u001b[32mMadeleine Albright\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.107:0.107[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npolicy[0.044]; ambassador[0.044]; ambassador[0.044]; Foreign[0.043]; Albright[0.042]; Albright[0.042]; Albright[0.042]; academic[0.042]; strategic[0.042]; lectured[0.042]; foreign[0.042]; Josef[0.042]; country[0.041]; praised[0.041]; diplomats[0.041]; Karel[0.041]; told[0.041]; told[0.041]; integration[0.041]; Mlada[0.040]; diplomatic[0.040]; supporter[0.040]; Nato[0.040]; Nato[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Madeleine Albright\u001b[39m ==> ENTITY: \u001b[32mMadeleine Albright\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSecretary[0.055]; Secretary[0.055]; President[0.054]; American[0.052]; Albright[0.050]; Albright[0.050]; surgery[0.050]; invited[0.050]; States[0.050]; States[0.050]; statement[0.049]; statement[0.049]; Vaclav[0.049]; State[0.049]; State[0.049]; distinguished[0.048]; appointment[0.048]; diplomat[0.048]; Prague[0.048]; Prague[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.082:0.082[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.048]; countries[0.048]; States[0.047]; Europe[0.045]; Europe[0.045]; said[0.044]; bloc[0.043]; key[0.043]; think[0.043]; United[0.043]; United[0.043]; Nations[0.043]; came[0.043]; Monday[0.043]; Atlantic[0.042]; ousted[0.042]; State[0.042]; helped[0.042]; Prague[0.042]; foreign[0.042]; foreign[0.042]; strong[0.042]; included[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NATO\u001b[39m ==> ENTITY: \u001b[32mNATO\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.132:0.132[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncooperation[0.046]; Nato[0.046]; included[0.043]; Europe[0.043]; Europe[0.043]; bloc[0.042]; alliance[0.042]; security[0.042]; strengthening[0.042]; enlargement[0.042]; warned[0.041]; seeking[0.041]; selection[0.041]; countries[0.041]; policy[0.041]; Czech[0.041]; Czech[0.041]; Czechs[0.041]; strategic[0.041]; country[0.040]; praised[0.040]; creation[0.040]; diplomatic[0.040]; foreign[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Albright\u001b[39m ==> ENTITY: \u001b[32mMadeleine Albright\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSecretary[0.046]; Secretary[0.046]; President[0.045]; American[0.044]; Albright[0.042]; Albright[0.042]; surgery[0.042]; States[0.042]; States[0.042]; statement[0.042]; Vaclav[0.042]; State[0.041]; State[0.041]; distinguished[0.041]; appointment[0.041]; diplomat[0.040]; Prague[0.040]; Havel[0.040]; Havel[0.040]; Havel[0.040]; Havel[0.040]; said[0.040]; said[0.040]; welcomed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Albright\u001b[39m ==> ENTITY: \u001b[32mMadeleine Albright\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npolicy[0.045]; policy[0.045]; ambassador[0.045]; ambassador[0.045]; lung[0.044]; hospital[0.044]; Foreign[0.044]; Albright[0.044]; Albright[0.044]; Albright[0.044]; Albright[0.044]; academic[0.043]; States[0.043]; strategic[0.043]; lectured[0.043]; foreign[0.043]; foreign[0.043]; Josef[0.043]; country[0.043]; praised[0.043]; diplomats[0.043]; Czechoslovak[0.042]; Karel[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Albright\u001b[39m ==> ENTITY: \u001b[32mMadeleine Albright\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSecretary[0.048]; American[0.045]; policy[0.045]; policy[0.045]; ambassador[0.044]; lung[0.044]; hospital[0.044]; Albright[0.043]; Albright[0.043]; Albright[0.043]; invited[0.043]; academic[0.043]; States[0.043]; States[0.043]; statement[0.043]; lectured[0.043]; foreign[0.043]; foreign[0.043]; State[0.043]; country[0.042]; praised[0.042]; diplomats[0.042]; Czechoslovak[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.066:0.066[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.045]; United[0.044]; United[0.044]; States[0.043]; States[0.043]; said[0.043]; said[0.043]; think[0.042]; came[0.041]; Monday[0.041]; Friday[0.041]; ousted[0.041]; helped[0.041]; Prague[0.041]; Prague[0.041]; Czech[0.041]; Czech[0.041]; Czech[0.041]; Native[0.041]; policy[0.041]; hold[0.041]; family[0.041]; small[0.041]; position[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.128:0.128[0]; log p(e|m)= -1.207:-1.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPrague[0.046]; Prague[0.046]; Czech[0.045]; Czech[0.045]; American[0.043]; Havel[0.043]; Havel[0.043]; Havel[0.043]; Havel[0.043]; Havel[0.043]; Communist[0.042]; Vaclav[0.041]; origins[0.041]; distinguished[0.040]; Native[0.040]; invited[0.040]; States[0.040]; States[0.040]; United[0.040]; United[0.040]; helped[0.039]; Secretary[0.039]; Secretary[0.039]; Praises[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = United Nations\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStates[0.045]; countries[0.044]; ambassador[0.042]; ambassador[0.042]; security[0.042]; Europe[0.042]; Europe[0.042]; key[0.042]; bloc[0.042]; cooperation[0.042]; country[0.042]; policy[0.042]; policy[0.042]; Albright[0.041]; Albright[0.041]; Albright[0.041]; Albright[0.041]; Albright[0.041]; foreign[0.041]; foreign[0.041]; Foreign[0.041]; diplomats[0.041]; United[0.040]; United[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Prague ==> ENTITY: \u001b[32mPrague\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.053:-0.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nacademic[0.047]; studied[0.045]; Czechoslovak[0.044]; lectured[0.044]; country[0.043]; woman[0.042]; woman[0.042]; Czech[0.042]; Czech[0.042]; Havel[0.042]; Havel[0.042]; Havel[0.042]; Havel[0.042]; family[0.041]; Europe[0.041]; Monday[0.041]; ambassador[0.040]; included[0.040]; Communists[0.040]; States[0.040]; States[0.040]; States[0.040]; American[0.040]; came[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.123:0.123[0]; log p(e|m)= -1.207:-1.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.048]; Czech[0.046]; Europe[0.045]; Czechs[0.044]; ambassador[0.044]; ambassador[0.044]; Josef[0.043]; Karel[0.043]; century[0.043]; countries[0.042]; Dnes[0.041]; foreign[0.041]; Foreign[0.041]; Minister[0.040]; praised[0.040]; key[0.040]; supporter[0.040]; influence[0.040]; told[0.040]; told[0.040]; diplomatic[0.039]; included[0.039]; strong[0.039]; expected[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = United Nations\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.132:0.132[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountries[0.045]; ambassador[0.043]; ambassador[0.043]; security[0.043]; diplomatic[0.043]; key[0.043]; bloc[0.043]; cooperation[0.043]; country[0.043]; policy[0.042]; Europe[0.042]; foreign[0.042]; Foreign[0.041]; United[0.041]; Nations[0.041]; Atlantic[0.041]; diplomats[0.041]; strong[0.040]; strengthening[0.040]; included[0.040]; seeking[0.040]; Albright[0.040]; Albright[0.040]; Albright[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Albright\u001b[39m ==> ENTITY: \u001b[32mMadeleine Albright\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSecretary[0.046]; policy[0.043]; policy[0.043]; ambassador[0.043]; lung[0.042]; hospital[0.042]; Foreign[0.042]; Albright[0.042]; Albright[0.042]; Albright[0.042]; Albright[0.042]; invited[0.041]; academic[0.041]; States[0.041]; strategic[0.041]; statement[0.041]; lectured[0.041]; foreign[0.041]; foreign[0.041]; State[0.041]; Josef[0.041]; country[0.041]; praised[0.041]; diplomats[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.079:0.079[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npolicy[0.045]; American[0.044]; family[0.043]; private[0.043]; statement[0.043]; statement[0.043]; Europe[0.042]; government[0.042]; said[0.042]; said[0.042]; academic[0.042]; Monday[0.041]; working[0.041]; foreign[0.041]; lectured[0.041]; position[0.041]; Czech[0.041]; Czech[0.041]; origins[0.041]; small[0.041]; coup[0.041]; Friday[0.041]; problems[0.040]; think[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Havel\u001b[39m ==> ENTITY: \u001b[32mVáclav Havel\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHavel[0.048]; Havel[0.048]; Havel[0.048]; Havel[0.048]; Prague[0.046]; Prague[0.046]; Czech[0.043]; Czech[0.043]; Czech[0.043]; President[0.043]; Vaclav[0.041]; surgery[0.040]; State[0.039]; State[0.039]; Communist[0.039]; invited[0.039]; democratic[0.039]; appointment[0.039]; Secretary[0.038]; Secretary[0.038]; statement[0.038]; statement[0.038]; regime[0.038]; welcomed[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Europe ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.101:0.101[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurope[0.049]; country[0.045]; bloc[0.045]; policy[0.045]; countries[0.044]; States[0.044]; enlargement[0.043]; creation[0.041]; foreign[0.041]; included[0.041]; century[0.041]; cooperation[0.041]; integration[0.041]; supporter[0.041]; Atlantic[0.041]; Foreign[0.040]; United[0.040]; United[0.040]; United[0.040]; ambassador[0.040]; ambassador[0.040]; alliance[0.039]; Nato[0.039]; Nato[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Havel\u001b[39m ==> ENTITY: \u001b[32mVáclav Havel\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHavel[0.050]; Havel[0.050]; Prague[0.048]; Czech[0.045]; Czech[0.045]; Czech[0.045]; President[0.045]; Vaclav[0.043]; surgery[0.042]; State[0.041]; appointment[0.040]; Secretary[0.040]; Secretary[0.040]; statement[0.040]; welcomed[0.039]; said[0.039]; distinguished[0.039]; American[0.039]; Praises[0.039]; Albright[0.038]; Albright[0.038]; Albright[0.038]; States[0.038]; States[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Prague ==> ENTITY: \u001b[32mPrague\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.093:0.093[0]; log p(e|m)= -0.053:-0.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwoman[0.043]; woman[0.043]; Czech[0.043]; Czech[0.043]; Czech[0.043]; Havel[0.042]; Havel[0.042]; Havel[0.042]; Havel[0.042]; Havel[0.042]; forward[0.042]; Native[0.042]; position[0.041]; Friday[0.041]; Vaclav[0.041]; said[0.041]; said[0.041]; appointment[0.041]; States[0.040]; States[0.040]; American[0.040]; distinguished[0.040]; United[0.040]; United[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marie Korbelova\u001b[39m ==> ENTITY: \u001b[32mMadeleine Albright\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSecretary[0.046]; American[0.043]; policy[0.043]; policy[0.043]; ambassador[0.043]; lung[0.042]; hospital[0.042]; Albright[0.042]; Albright[0.042]; Albright[0.042]; Albright[0.042]; invited[0.041]; academic[0.041]; States[0.041]; States[0.041]; statement[0.041]; lectured[0.041]; foreign[0.041]; foreign[0.041]; State[0.041]; country[0.041]; praised[0.041]; diplomats[0.041]; Czechoslovak[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.170:0.170[0]; log p(e|m)= -1.207:-1.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.047]; Czechoslovak[0.046]; Czech[0.044]; Prague[0.044]; born[0.044]; Europe[0.043]; Europe[0.043]; Havel[0.043]; ambassador[0.042]; ambassador[0.042]; Karel[0.042]; century[0.042]; Communist[0.041]; countries[0.041]; Communists[0.040]; Josef[0.040]; small[0.040]; studied[0.040]; family[0.040]; States[0.039]; strong[0.039]; United[0.039]; United[0.039]; United[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Albright\u001b[39m ==> ENTITY: \u001b[32mMadeleine Albright\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSecretary[0.046]; Secretary[0.046]; American[0.043]; policy[0.043]; ambassador[0.042]; lung[0.042]; hospital[0.042]; Albright[0.041]; Albright[0.041]; Albright[0.041]; surgery[0.041]; invited[0.041]; academic[0.041]; States[0.041]; States[0.041]; States[0.041]; statement[0.041]; statement[0.041]; lectured[0.041]; foreign[0.041]; State[0.041]; State[0.041]; country[0.040]; praised[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = American ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.063:0.063[0]; log p(e|m)= -0.130:-0.130[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNative[0.047]; Friend[0.045]; Czech[0.045]; Czech[0.045]; Czech[0.045]; born[0.045]; small[0.045]; helped[0.044]; States[0.044]; States[0.044]; policy[0.043]; origins[0.043]; diplomat[0.043]; think[0.042]; foreign[0.042]; working[0.042]; said[0.042]; said[0.042]; Monday[0.042]; welcomed[0.042]; woman[0.042]; woman[0.042]; good[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mlada Fronta Dnes\u001b[39m ==> ENTITY: \u001b[32mMladá fronta DNES\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.071:0.071[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCzech[0.045]; daily[0.045]; told[0.044]; told[0.044]; strategic[0.043]; Josef[0.043]; Karel[0.043]; country[0.042]; integration[0.042]; Nato[0.041]; Nato[0.041]; Europe[0.041]; diplomatic[0.041]; Minister[0.041]; bloc[0.041]; trans[0.041]; policy[0.041]; enlargement[0.040]; Kovanda[0.040]; foreign[0.040]; Czechs[0.040]; expansion[0.040]; strengthening[0.040]; nomination[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Albright\u001b[39m ==> ENTITY: \u001b[32mMadeleine Albright\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.086:0.086[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npolicy[0.045]; ambassador[0.044]; Foreign[0.043]; Albright[0.043]; Albright[0.043]; strategic[0.043]; foreign[0.043]; Josef[0.042]; praised[0.042]; Karel[0.042]; told[0.042]; told[0.042]; integration[0.041]; Mlada[0.041]; diplomatic[0.041]; supporter[0.041]; Nato[0.041]; Nato[0.041]; nomination[0.040]; included[0.040]; security[0.040]; countries[0.040]; Minister[0.040]; Dnes[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Europe ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurope[0.051]; country[0.046]; bloc[0.046]; policy[0.046]; policy[0.046]; countries[0.045]; States[0.045]; enlargement[0.044]; Nato[0.042]; Nato[0.042]; creation[0.042]; foreign[0.042]; foreign[0.042]; included[0.042]; Atlantic[0.042]; small[0.042]; century[0.042]; government[0.042]; cooperation[0.042]; integration[0.042]; supporter[0.042]; Foreign[0.042]; key[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.118:0.118[0]; log p(e|m)= -1.207:-1.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCzech[0.045]; Czech[0.045]; Prague[0.045]; Havel[0.044]; Havel[0.044]; Havel[0.044]; Havel[0.044]; Havel[0.044]; American[0.042]; Vaclav[0.041]; distinguished[0.041]; Native[0.041]; States[0.040]; States[0.040]; United[0.040]; United[0.040]; Klara[0.040]; helped[0.040]; Secretary[0.040]; Secretary[0.040]; Praises[0.040]; diplomat[0.040]; President[0.039]; Friday[0.039]; \t\n\r [===================================>..........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 4s974ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3525/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1277testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Des Moines ==> ENTITY: \u001b[32mDes Moines, Iowa\u001b[39m\t\nSCORES: global= 0.241:0.241[0]; local()= 0.025:0.025[0]; log p(e|m)= -0.158:-0.158[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIowa[0.047]; Thursday[0.044]; Minn[0.044]; Steers[0.043]; Steers[0.043]; lbs[0.043]; lbs[0.043]; cattle[0.043]; Thursdays[0.042]; heifers[0.041]; heifers[0.041]; Select[0.041]; Select[0.041]; Select[0.041]; Select[0.041]; Select[0.041]; compared[0.041]; Heifers[0.041]; said[0.040]; Holstein[0.040]; close[0.040]; Carcass[0.040]; market[0.040]; Offerings[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = USDA\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Agriculture\u001b[39m\t\nSCORES: global= 0.238:0.238[0]; local()= 0.037:0.037[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nweight[0.047]; weight[0.047]; weight[0.047]; Minn[0.047]; fed[0.046]; Steers[0.046]; Steers[0.046]; lbs[0.046]; lbs[0.046]; lbs[0.046]; Thursday[0.046]; cattle[0.046]; market[0.045]; Confirmed[0.045]; Moines[0.045]; Thursdays[0.045]; Iowa[0.044]; tested[0.044]; tested[0.044]; Des[0.044]; heifers[0.044]; heifers[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= -0.023:-0.023[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWeek[0.058]; Week[0.058]; Slaughter[0.056]; Choice[0.055]; Choice[0.055]; Choice[0.055]; Choice[0.055]; Choice[0.055]; Choice[0.055]; tested[0.055]; Select[0.055]; Select[0.055]; Select[0.055]; Select[0.055]; Select[0.055]; Select[0.055]; Compared[0.055]; Confirmed[0.054]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Minn\u001b[39m ==> ENTITY: \u001b[32mMinnesota\u001b[39m\t\nSCORES: global= 0.232:0.232[0]; local()= 0.004:0.004[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSteers[0.050]; Steers[0.050]; lbs[0.050]; Iowa[0.049]; cattle[0.049]; Moines[0.049]; Thursdays[0.048]; Des[0.047]; heifers[0.047]; heifers[0.047]; Select[0.047]; Select[0.047]; Select[0.047]; Select[0.047]; Heifers[0.047]; Holstein[0.047]; Thursday[0.046]; Carcass[0.046]; said[0.046]; market[0.046]; Offerings[0.046]; \t\n\r [===================================>..........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 4s989ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3529/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1201testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Helder Cristovao\u001b[39m ==> ENTITY: \u001b[32mHélder Cristóvão\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.245:0.245[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPortugal[0.044]; Benfica[0.044]; League[0.043]; Deportivo[0.042]; Champions[0.042]; Braga[0.042]; Porto[0.042]; Porto[0.042]; Porto[0.042]; Porto[0.042]; Porto[0.042]; Porto[0.042]; games[0.041]; Portuguese[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; November[0.041]; Coruna[0.041]; Costa[0.040]; December[0.040]; twice[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.237:0.237[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.048]; Spain[0.048]; Spain[0.048]; Spain[0.048]; Cadete[0.042]; League[0.042]; European[0.042]; Portugal[0.041]; games[0.041]; Italy[0.041]; Italy[0.041]; Sergio[0.041]; Madrid[0.041]; Carlos[0.040]; Jose[0.040]; Jose[0.040]; Coruna[0.040]; Luis[0.040]; Deportivo[0.040]; Fernando[0.040]; Germany[0.040]; Germany[0.040]; twice[0.040]; qualifying[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ukraine ==> ENTITY: \u001b[32mUkraine national football team\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.176:0.176[0]; log p(e|m)= -3.101:-3.101[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nqualifier[0.044]; games[0.044]; Champions[0.044]; League[0.043]; club[0.043]; Cup[0.042]; European[0.042]; European[0.042]; qualifying[0.042]; twice[0.042]; game[0.042]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Italy[0.041]; played[0.041]; played[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Portugal[0.040]; Portugal[0.040]; host[0.040]; group[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oviedo ==> ENTITY: \u001b[32mReal Oviedo\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.214:0.214[0]; log p(e|m)= -2.254:-2.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Antonio[0.043]; Costa[0.043]; Sergio[0.042]; Cadete[0.042]; Luis[0.042]; Fernando[0.042]; Coruna[0.042]; Cristovao[0.042]; Salamanca[0.041]; Madrid[0.041]; Jose[0.041]; Jose[0.041]; Carlos[0.041]; Deportivo[0.041]; Jorge[0.041]; Manuel[0.040]; Paulo[0.040]; Paulo[0.040]; Cruz[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.255:0.255[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.049]; Spain[0.049]; Spain[0.049]; Spain[0.049]; Cadete[0.042]; Italy[0.042]; Italy[0.042]; Sergio[0.042]; Madrid[0.042]; Carlos[0.041]; Jose[0.041]; Jose[0.041]; Coruna[0.041]; Luis[0.040]; Deportivo[0.040]; Fernando[0.040]; Antonio[0.040]; Jorge[0.040]; Salamanca[0.039]; Manuel[0.039]; Benfica[0.039]; Barcelona[0.039]; Barcelona[0.039]; Barcelona[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany national football team\u001b[39m\t\nSCORES: global= 0.241:0.241[0]; local()= 0.189:0.189[0]; log p(e|m)= -3.352:-3.352[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; Germany[0.047]; played[0.045]; played[0.045]; Soccer[0.044]; club[0.043]; games[0.043]; game[0.042]; twice[0.042]; League[0.042]; qualifier[0.041]; Portugal[0.041]; Portugal[0.041]; Portugal[0.041]; European[0.041]; European[0.041]; host[0.040]; qualifying[0.040]; Champions[0.040]; Cup[0.039]; defender[0.039]; November[0.039]; striker[0.038]; seven[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barcelona ==> ENTITY: \u001b[32mFC Barcelona\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.210:0.210[0]; log p(e|m)= -1.709:-1.709[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.046]; Spain[0.046]; Spain[0.046]; Spain[0.046]; Luis[0.046]; club[0.044]; Jorge[0.044]; Jose[0.044]; Jose[0.044]; Carlos[0.044]; Sergio[0.044]; twice[0.043]; Madrid[0.043]; League[0.043]; games[0.042]; Cup[0.042]; Santos[0.042]; Fernando[0.042]; Champions[0.042]; Italy[0.042]; Coruna[0.041]; Paulo[0.041]; qualifying[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Celtic Glasgow\u001b[39m ==> ENTITY: \u001b[32mCeltic F.C.\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.068:0.068[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.050]; Cadete[0.046]; Midfielders[0.043]; Italy[0.043]; Italy[0.043]; Coruna[0.042]; Deportivo[0.042]; Spain[0.042]; Spain[0.042]; Spain[0.042]; Spain[0.042]; Benfica[0.042]; Forwards[0.041]; Juventus[0.041]; Secretario[0.041]; Figo[0.041]; Madrid[0.041]; Salamanca[0.040]; Sporting[0.040]; Fiorentina[0.040]; Real[0.040]; Oviedo[0.040]; Barcelona[0.040]; Barcelona[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mScotland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mScotland national football team\u001b[39m <---> \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.245:0.235[\u001b[31m0.010\u001b[39m]; local()= 0.093:0.017[\u001b[31m0.076\u001b[39m]; log p(e|m)= -2.703:-0.229[\u001b[32m2.474\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCeltic[0.049]; Glasgow[0.047]; Cadete[0.047]; Italy[0.046]; Italy[0.046]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Forwards[0.042]; Benfica[0.041]; Fiorentina[0.040]; Salamanca[0.040]; Sporting[0.040]; Real[0.039]; Juventus[0.039]; Carlos[0.039]; Rui[0.039]; Rui[0.039]; Coruna[0.039]; Figo[0.039]; Madrid[0.039]; Vieira[0.038]; Cristovao[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Porto ==> ENTITY: \u001b[32mF.C. Porto\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.228:0.228[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPorto[0.045]; Porto[0.045]; Porto[0.045]; Porto[0.045]; Porto[0.045]; Joao[0.044]; Joao[0.044]; Portugal[0.042]; Portuguese[0.042]; Benfica[0.041]; Paulo[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Conceicao[0.040]; games[0.040]; Braga[0.040]; League[0.040]; Cristovao[0.040]; Fernando[0.040]; Deportivo[0.039]; Juventus[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = AC Milan\u001b[39m ==> ENTITY: \u001b[32mA.C. Milan\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.046]; League[0.046]; Soccer[0.045]; games[0.043]; Champions[0.043]; Cup[0.042]; twice[0.042]; played[0.042]; played[0.042]; Spain[0.041]; European[0.041]; European[0.041]; points[0.041]; Portugal[0.041]; Portugal[0.041]; Portugal[0.041]; qualifying[0.041]; Barcelona[0.041]; Barcelona[0.041]; Porto[0.041]; Porto[0.041]; Porto[0.041]; Porto[0.041]; seven[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dimas Teixeira\u001b[39m ==> ENTITY: \u001b[32mDimas Teixeira\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuropean[0.043]; twice[0.043]; qualifying[0.043]; Portugal[0.043]; Spain[0.042]; Spain[0.042]; Spain[0.042]; Spain[0.042]; Spain[0.042]; Cadete[0.042]; Benfica[0.042]; League[0.042]; games[0.041]; Sporting[0.041]; Sporting[0.041]; Portuguese[0.041]; Porto[0.041]; Porto[0.041]; Porto[0.041]; Porto[0.041]; Porto[0.041]; Porto[0.041]; Italy[0.041]; Italy[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Juventus\u001b[39m ==> ENTITY: \u001b[32mJuventus F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.045]; Italy[0.045]; games[0.043]; Cadete[0.043]; Paulo[0.043]; Paulo[0.043]; Sergio[0.042]; played[0.042]; Antonio[0.042]; Fiorentina[0.042]; twice[0.042]; Pinto[0.041]; Pinto[0.041]; Luis[0.041]; Fernando[0.041]; Figo[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; seven[0.040]; Joao[0.040]; Joao[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jose Taira\u001b[39m ==> ENTITY: \u001b[32mJosé Taira\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.241:0.241[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCosta[0.044]; Porto[0.043]; Porto[0.043]; Porto[0.043]; Porto[0.043]; Porto[0.043]; Porto[0.043]; Benfica[0.042]; Paulo[0.042]; Paulo[0.042]; Fernando[0.042]; Braga[0.041]; Alves[0.041]; Bento[0.041]; Santos[0.041]; Spain[0.040]; Spain[0.040]; Spain[0.040]; Spain[0.040]; Spain[0.040]; Joao[0.040]; Joao[0.040]; Cadete[0.040]; Cristovao[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Salamanca ==> ENTITY: \u001b[32mUD Salamanca\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.183:0.183[0]; log p(e|m)= -2.017:-2.017[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.045]; Carlos[0.045]; Sergio[0.044]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Jose[0.043]; Jose[0.043]; Cruz[0.043]; Jorge[0.043]; Manuel[0.042]; Fernando[0.042]; Antonio[0.041]; Madrid[0.041]; Joao[0.040]; Joao[0.040]; Coruna[0.040]; Cadete[0.040]; Santos[0.039]; Teixeira[0.039]; Cristovao[0.038]; Paulo[0.038]; Paulo[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.216:0.216[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.048]; Spain[0.048]; Spain[0.048]; club[0.043]; League[0.042]; European[0.042]; European[0.042]; Portugal[0.042]; games[0.042]; Italy[0.041]; Sergio[0.041]; Madrid[0.041]; Carlos[0.041]; Jose[0.040]; Jose[0.040]; Coruna[0.040]; Luis[0.040]; Deportivo[0.040]; Fernando[0.040]; qualifier[0.040]; Germany[0.040]; Germany[0.040]; twice[0.040]; qualifying[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jorge Costa\u001b[39m ==> ENTITY: \u001b[32mJorge Costa\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.226:0.226[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.044]; Soccer[0.044]; Portuguese[0.043]; Santos[0.043]; Portugal[0.042]; Portugal[0.042]; Portugal[0.042]; Champions[0.042]; Vitor[0.042]; Porto[0.042]; Costa[0.042]; Costa[0.042]; League[0.042]; Spain[0.041]; Barcelona[0.041]; Lisbon[0.041]; Pinto[0.041]; games[0.041]; Joao[0.041]; Cup[0.041]; Baia[0.040]; forward[0.040]; Braga[0.040]; Manuel[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Porto ==> ENTITY: \u001b[32mF.C. Porto\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.232:0.232[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPorto[0.045]; Porto[0.045]; Porto[0.045]; Porto[0.045]; Porto[0.045]; Joao[0.044]; Joao[0.044]; Portugal[0.042]; Portuguese[0.042]; Benfica[0.041]; Paulo[0.041]; Paulo[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Conceicao[0.040]; games[0.040]; Braga[0.040]; League[0.040]; Cristovao[0.040]; Fernando[0.040]; Deportivo[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.196:0.196[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.052]; Scotland[0.043]; Juventus[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Fiorentina[0.042]; games[0.042]; Cadete[0.042]; qualifying[0.042]; Portugal[0.041]; European[0.041]; Germany[0.041]; Germany[0.041]; Figo[0.041]; played[0.041]; twice[0.040]; Coruna[0.039]; seven[0.039]; Glasgow[0.039]; Sergio[0.039]; Benfica[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Carlos Secretario\u001b[39m ==> ENTITY: \u001b[32mCarlos Secretário\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPortuguese[0.045]; Spain[0.045]; Spain[0.045]; Spain[0.045]; Spain[0.045]; Spain[0.045]; Cadete[0.045]; games[0.044]; Fernando[0.043]; Paulo[0.043]; Paulo[0.043]; Benfica[0.043]; Antonio[0.043]; European[0.043]; Luis[0.043]; Sergio[0.043]; Coruna[0.043]; Jose[0.043]; Jose[0.043]; Jorge[0.043]; Deportivo[0.043]; Madrid[0.042]; Braga[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany national football team\u001b[39m\t\nSCORES: global= 0.241:0.241[0]; local()= 0.191:0.191[0]; log p(e|m)= -3.352:-3.352[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; Germany[0.047]; played[0.045]; played[0.045]; club[0.043]; games[0.043]; Spain[0.042]; Spain[0.042]; Spain[0.042]; game[0.042]; twice[0.041]; League[0.041]; Italy[0.041]; qualifier[0.041]; Portugal[0.041]; European[0.040]; European[0.040]; host[0.040]; qualifying[0.040]; Champions[0.040]; Cup[0.039]; defender[0.039]; November[0.039]; striker[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.237:0.237[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.049]; Spain[0.049]; Spain[0.049]; Cadete[0.043]; Italy[0.042]; Italy[0.042]; Sergio[0.042]; Madrid[0.042]; Carlos[0.041]; Jose[0.041]; Jose[0.041]; Coruna[0.041]; Luis[0.041]; Deportivo[0.041]; Fernando[0.041]; Antonio[0.040]; Jorge[0.040]; Salamanca[0.040]; Manuel[0.039]; Benfica[0.039]; Barcelona[0.039]; Barcelona[0.039]; Scotland[0.039]; Real[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Antonio Folha\u001b[39m ==> ENTITY: \u001b[32mAntónio Folha\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.045]; Spain[0.045]; Spain[0.045]; Spain[0.045]; Cadete[0.045]; Benfica[0.044]; Fernando[0.043]; Paulo[0.043]; Paulo[0.043]; Santos[0.043]; Porto[0.043]; Porto[0.043]; Porto[0.043]; Porto[0.043]; Porto[0.043]; Porto[0.043]; Real[0.043]; Deportivo[0.043]; Luis[0.043]; Jose[0.043]; Jose[0.043]; Italy[0.042]; Italy[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rui Correia\u001b[39m ==> ENTITY: \u001b[32mRui Correia\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.239:0.239[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPortugal[0.049]; club[0.048]; Braga[0.046]; games[0.046]; League[0.046]; Cristovao[0.046]; Sporting[0.046]; Paulo[0.046]; Porto[0.045]; Porto[0.045]; Porto[0.045]; Porto[0.045]; Porto[0.045]; Joao[0.045]; Portuguese[0.045]; Deportivo[0.045]; Rui[0.045]; Pinto[0.044]; Champions[0.044]; Helder[0.044]; Bento[0.044]; Goalkeepers[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Oceano Cruz\u001b[39m ==> ENTITY: \u001b[32mOceano da Cruz\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.261:0.261[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPortugal[0.044]; League[0.043]; Spain[0.042]; Spain[0.042]; Spain[0.042]; Spain[0.042]; Spain[0.042]; Fernando[0.042]; Porto[0.042]; Porto[0.042]; Porto[0.042]; Porto[0.042]; Porto[0.042]; Porto[0.042]; Benfica[0.042]; Deportivo[0.041]; Carlos[0.041]; Jose[0.041]; Jose[0.041]; Luis[0.041]; Antonio[0.041]; games[0.041]; Sergio[0.040]; Santos[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portugal ==> ENTITY: \u001b[32mPortugal national football team\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.215:0.215[0]; log p(e|m)= -2.453:-2.453[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.048]; Spain[0.048]; Portugal[0.047]; Portugal[0.047]; Soccer[0.047]; Portuguese[0.045]; League[0.044]; club[0.044]; Italy[0.044]; games[0.044]; game[0.043]; qualifier[0.043]; qualifying[0.043]; Champions[0.043]; Sporting[0.042]; Cup[0.041]; Fernando[0.041]; points[0.041]; Sergio[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; place[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Madrid\u001b[39m ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.227:0.227[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.043]; Jose[0.043]; Jose[0.043]; Carlos[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Cadete[0.042]; games[0.042]; Sergio[0.042]; Deportivo[0.041]; Jorge[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Benfica[0.041]; Fernando[0.041]; Coruna[0.041]; twice[0.041]; Manuel[0.040]; Antonio[0.040]; Juventus[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.243:0.243[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.048]; Spain[0.048]; Spain[0.048]; Spain[0.048]; Cadete[0.042]; games[0.042]; Italy[0.042]; Italy[0.042]; Sergio[0.041]; Madrid[0.041]; Carlos[0.041]; Jose[0.040]; Jose[0.040]; Coruna[0.040]; Luis[0.040]; Deportivo[0.040]; Fernando[0.040]; twice[0.040]; Antonio[0.040]; Jorge[0.039]; Germany[0.039]; Germany[0.039]; points[0.039]; Salamanca[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Porto ==> ENTITY: \u001b[32mF.C. Porto\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.207:0.207[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJoao[0.047]; Portugal[0.045]; Portugal[0.045]; Portugal[0.045]; Lisbon[0.044]; club[0.043]; games[0.043]; League[0.043]; Soccer[0.042]; Milan[0.042]; central[0.042]; Costa[0.041]; Costa[0.041]; Costa[0.041]; Champions[0.041]; Cup[0.040]; Manuel[0.040]; European[0.040]; European[0.040]; Jorge[0.039]; game[0.039]; qualifying[0.039]; played[0.039]; played[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Porto ==> ENTITY: \u001b[32mF.C. Porto\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.204:0.204[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPorto[0.045]; Porto[0.045]; Porto[0.045]; Porto[0.045]; Porto[0.045]; Joao[0.044]; Joao[0.044]; Italy[0.044]; Italy[0.044]; Coruna[0.042]; Salamanca[0.041]; Benfica[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Conceicao[0.040]; Cristovao[0.040]; Fernando[0.039]; Juventus[0.039]; Barros[0.039]; Rui[0.039]; Rui[0.039]; Madrid[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Joao Vieira Pinto\u001b[39m ==> ENTITY: \u001b[32mJoão Vieira Pinto\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.228:0.228[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.045]; Spain[0.045]; Spain[0.045]; Spain[0.045]; Figo[0.044]; Paulo[0.044]; Paulo[0.044]; Madrid[0.044]; Fernando[0.043]; Barcelona[0.043]; Barcelona[0.043]; Cadete[0.043]; Alves[0.043]; Pinto[0.043]; Benfica[0.043]; Carlos[0.043]; Jose[0.043]; Jose[0.043]; Barros[0.043]; Luis[0.043]; Italy[0.042]; Italy[0.042]; Sergio[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portugal ==> ENTITY: \u001b[32mPortugal national football team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.201:0.201[0]; log p(e|m)= -2.453:-2.453[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPortugal[0.047]; Portugal[0.047]; Soccer[0.047]; League[0.044]; club[0.044]; qualifier[0.043]; Champions[0.043]; qualifying[0.043]; Cup[0.042]; game[0.041]; European[0.041]; European[0.041]; Lisbon[0.041]; Germany[0.041]; played[0.041]; World[0.041]; Jorge[0.040]; Porto[0.040]; central[0.040]; group[0.039]; Pinto[0.039]; place[0.039]; November[0.039]; Manuel[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Braga ==> ENTITY: \u001b[32mS.C. Braga\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.196:0.196[0]; log p(e|m)= -1.398:-1.398[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJoao[0.044]; Portugal[0.043]; Porto[0.043]; Porto[0.043]; Porto[0.043]; Porto[0.043]; Porto[0.043]; Costa[0.042]; Costa[0.042]; Paulo[0.042]; Fernando[0.042]; Cristovao[0.041]; club[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Deportivo[0.041]; Portuguese[0.041]; games[0.040]; Santos[0.040]; Jose[0.040]; Jose[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Liberian ==> ENTITY: \u001b[32mLiberia national football team\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.161:0.161[0]; log p(e|m)= -1.076:-1.076[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWeah[0.045]; Weah[0.045]; Soccer[0.045]; qualifying[0.043]; Cup[0.043]; club[0.043]; twice[0.043]; qualifier[0.043]; League[0.042]; games[0.042]; striker[0.041]; game[0.041]; November[0.041]; played[0.041]; played[0.041]; Champions[0.041]; Spain[0.041]; World[0.040]; defender[0.040]; Friday[0.040]; central[0.040]; seven[0.040]; December[0.040]; host[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sporting ==> ENTITY: \u001b[32mSporting Clube de Portugal\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.222:0.222[0]; log p(e|m)= -0.121:-0.121[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSporting[0.050]; Portugal[0.048]; Luis[0.047]; games[0.047]; Portuguese[0.047]; Jose[0.046]; Jose[0.046]; Coruna[0.046]; Benfica[0.046]; League[0.045]; Deportivo[0.045]; Champions[0.045]; Braga[0.045]; Sergio[0.045]; game[0.045]; Carlos[0.044]; Spain[0.044]; Spain[0.044]; Spain[0.044]; Spain[0.044]; Spain[0.044]; Fernando[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rui Barros\u001b[39m ==> ENTITY: \u001b[32mRui Barros\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.251:0.251[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPortuguese[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; games[0.043]; Benfica[0.042]; Porto[0.042]; Porto[0.042]; Porto[0.042]; Porto[0.042]; Porto[0.042]; Porto[0.042]; Fernando[0.042]; Braga[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Deportivo[0.041]; twice[0.041]; Vitor[0.040]; Cadete[0.040]; Madrid[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Porto ==> ENTITY: \u001b[32mF.C. Porto\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.217:0.217[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPorto[0.044]; Porto[0.044]; Porto[0.044]; Porto[0.044]; Porto[0.044]; Joao[0.043]; Joao[0.043]; Italy[0.043]; Italy[0.043]; Portuguese[0.042]; Coruna[0.041]; Salamanca[0.041]; Benfica[0.041]; Paulo[0.040]; Paulo[0.040]; Spain[0.040]; Spain[0.040]; Spain[0.040]; Spain[0.040]; Spain[0.040]; Conceicao[0.040]; Braga[0.040]; Cristovao[0.039]; Fernando[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portuguese ==> ENTITY: \u001b[32mPortugal\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.235:0.235[0]; log p(e|m)= -1.127:-1.127[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPortugal[0.054]; Spain[0.054]; Spain[0.054]; Spain[0.054]; European[0.053]; European[0.053]; club[0.051]; Madrid[0.050]; games[0.050]; Italy[0.050]; Barcelona[0.049]; Barcelona[0.049]; League[0.048]; Milan[0.048]; Champions[0.048]; Fernando[0.048]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Cup[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paulo Alves\u001b[39m ==> ENTITY: \u001b[32mPaulo Alves\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCoruna[0.046]; Fernando[0.045]; Benfica[0.044]; Cadete[0.044]; Spain[0.044]; Spain[0.044]; Spain[0.044]; Spain[0.044]; Luis[0.044]; Deportivo[0.044]; Porto[0.043]; Porto[0.043]; Porto[0.043]; Porto[0.043]; Carlos[0.043]; Paulo[0.043]; Rui[0.043]; Rui[0.043]; Jose[0.043]; Jose[0.043]; Madrid[0.042]; Jorge[0.042]; Antonio[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sporting ==> ENTITY: \u001b[32mSporting Clube de Portugal\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.227:0.227[0]; log p(e|m)= -0.121:-0.121[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSporting[0.047]; Luis[0.044]; Jose[0.043]; Jose[0.043]; Cadete[0.043]; Coruna[0.043]; Benfica[0.043]; Deportivo[0.042]; Jorge[0.042]; Carlos[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Fernando[0.041]; Rui[0.041]; Rui[0.041]; Porto[0.041]; Porto[0.041]; Porto[0.041]; Porto[0.041]; Joao[0.040]; Barcelona[0.040]; Barcelona[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Cup ==> ENTITY: \u001b[32mFIFA World Cup\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.184:0.184[0]; log p(e|m)= -1.002:-1.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.044]; qualifying[0.043]; Champions[0.043]; Portugal[0.043]; Portugal[0.043]; Portugal[0.043]; qualifier[0.042]; host[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; League[0.042]; club[0.042]; games[0.042]; European[0.042]; European[0.042]; twice[0.041]; played[0.041]; played[0.041]; place[0.040]; game[0.040]; Milan[0.040]; November[0.040]; December[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paulinho Santos\u001b[39m ==> ENTITY: \u001b[32mPaulinho Santos\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.223:0.223[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPortugal[0.046]; games[0.045]; Portuguese[0.044]; played[0.044]; played[0.044]; Porto[0.044]; Porto[0.044]; Porto[0.044]; Porto[0.044]; Porto[0.044]; Porto[0.044]; Sporting[0.044]; Champions[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Bento[0.043]; Paulo[0.043]; Helder[0.042]; Fernando[0.042]; Deportivo[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lisbon ==> ENTITY: \u001b[32mLisbon\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.169:0.169[0]; log p(e|m)= -0.098:-0.098[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPortugal[0.046]; Portugal[0.046]; Portugal[0.046]; Soccer[0.044]; central[0.044]; games[0.043]; Joao[0.043]; club[0.043]; Cup[0.042]; Manuel[0.042]; Porto[0.041]; game[0.041]; Pinto[0.041]; played[0.041]; place[0.040]; Costa[0.040]; Costa[0.040]; Costa[0.040]; Jorge[0.040]; League[0.040]; Champions[0.040]; seven[0.039]; qualifier[0.039]; November[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fernando Couto\u001b[39m ==> ENTITY: \u001b[32mFernando Couto\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.226:0.226[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCadete[0.053]; Spain[0.051]; Spain[0.051]; Spain[0.051]; Spain[0.051]; Spain[0.051]; Portugal[0.050]; League[0.050]; Barcelona[0.050]; Barcelona[0.050]; Barcelona[0.050]; Luis[0.049]; games[0.049]; Italy[0.049]; Italy[0.049]; Jose[0.049]; Jose[0.049]; Benfica[0.049]; Paulo[0.049]; Paulo[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = European ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.256:-0.256[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.045]; European[0.045]; club[0.044]; Champions[0.044]; games[0.044]; qualifying[0.042]; Spain[0.042]; Spain[0.042]; Cup[0.042]; qualifier[0.042]; twice[0.041]; Milan[0.041]; Italy[0.041]; game[0.041]; Portugal[0.041]; Portugal[0.041]; Portugal[0.041]; Juventus[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; place[0.040]; played[0.040]; played[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Porto ==> ENTITY: \u001b[32mF.C. Porto\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.217:0.217[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPorto[0.044]; Porto[0.044]; Porto[0.044]; Porto[0.044]; Porto[0.044]; Joao[0.043]; Joao[0.043]; Italy[0.043]; Italy[0.043]; Portuguese[0.042]; Coruna[0.041]; Salamanca[0.041]; Benfica[0.041]; Paulo[0.040]; Paulo[0.040]; Spain[0.040]; Spain[0.040]; Spain[0.040]; Spain[0.040]; Spain[0.040]; Conceicao[0.040]; Braga[0.040]; Cristovao[0.039]; Fernando[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Luis Figo\u001b[39m ==> ENTITY: \u001b[32mLuís Figo\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSergio[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Jose[0.043]; Jose[0.043]; Carlos[0.042]; Cadete[0.042]; Fernando[0.042]; Jorge[0.042]; Rui[0.041]; Rui[0.041]; Rui[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; Paulo[0.041]; Paulo[0.041]; Madrid[0.040]; Real[0.040]; Antonio[0.040]; Juventus[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany national football team\u001b[39m\t\nSCORES: global= 0.241:0.241[0]; local()= 0.193:0.193[0]; log p(e|m)= -3.352:-3.352[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; Germany[0.047]; played[0.045]; played[0.045]; club[0.043]; games[0.043]; Spain[0.042]; Spain[0.042]; Spain[0.042]; game[0.042]; twice[0.041]; League[0.041]; Italy[0.041]; qualifier[0.041]; Portugal[0.041]; European[0.040]; European[0.040]; host[0.040]; qualifying[0.040]; Champions[0.040]; Cup[0.039]; November[0.039]; striker[0.038]; seven[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rui Costa\u001b[39m ==> ENTITY: \u001b[32mRui Costa\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.228:0.228[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBenfica[0.044]; Cadete[0.043]; Juventus[0.042]; Paulo[0.042]; Paulo[0.042]; Porto[0.042]; Porto[0.042]; Porto[0.042]; Porto[0.042]; Rui[0.042]; Barros[0.042]; Real[0.042]; Luis[0.041]; Jorge[0.041]; Fernando[0.041]; Coruna[0.041]; Deportivo[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Jose[0.041]; Jose[0.041]; Fiorentina[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fiorentina\u001b[39m ==> ENTITY: \u001b[32mACF Fiorentina\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.050]; Italy[0.050]; Juventus[0.047]; Cadete[0.046]; Celtic[0.044]; Deportivo[0.044]; Coruna[0.043]; Benfica[0.043]; Bento[0.043]; Costa[0.042]; Couto[0.042]; Porto[0.042]; Porto[0.042]; Porto[0.042]; Spain[0.042]; Spain[0.042]; Spain[0.042]; Spain[0.042]; Barcelona[0.042]; Barcelona[0.042]; Madrid[0.042]; Real[0.042]; Antonio[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Porto ==> ENTITY: \u001b[32mF.C. Porto\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.224:0.224[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPorto[0.045]; Porto[0.045]; Porto[0.045]; Porto[0.045]; Porto[0.045]; Joao[0.044]; Joao[0.044]; Portugal[0.043]; Portuguese[0.042]; Paulo[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Conceicao[0.040]; games[0.040]; Braga[0.040]; League[0.040]; Cristovao[0.040]; Fernando[0.040]; Deportivo[0.039]; Juventus[0.039]; Barros[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCosta\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRui Costa\u001b[39m <---> \u001b[32mJorge Costa\u001b[39m\t\nSCORES: global= 0.270:0.264[\u001b[31m0.007\u001b[39m]; local()= 0.237:0.228[\u001b[31m0.009\u001b[39m]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.044]; Soccer[0.043]; Braga[0.042]; League[0.042]; Portuguese[0.042]; Santos[0.042]; Sergio[0.042]; Jorge[0.042]; Paulinho[0.042]; Portugal[0.042]; Portugal[0.042]; Portugal[0.042]; Champions[0.041]; Vitor[0.041]; Porto[0.041]; Porto[0.041]; Porto[0.041]; Porto[0.041]; games[0.041]; Rui[0.041]; Costa[0.041]; Costa[0.041]; Spain[0.041]; Barcelona[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barcelona ==> ENTITY: \u001b[32mFC Barcelona\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.209:0.209[0]; log p(e|m)= -1.709:-1.709[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.044]; Spain[0.044]; Spain[0.044]; Spain[0.044]; Spain[0.044]; Luis[0.044]; Jorge[0.042]; Jose[0.042]; Jose[0.042]; Carlos[0.042]; Sergio[0.042]; Madrid[0.041]; League[0.041]; Cadete[0.041]; games[0.041]; Santos[0.040]; Fernando[0.040]; Champions[0.040]; Italy[0.040]; Italy[0.040]; Antonio[0.040]; Coruna[0.040]; Paulo[0.040]; Paulo[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Weah\u001b[39m ==> ENTITY: \u001b[32mGeorge Weah\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWeah[0.050]; League[0.044]; Soccer[0.044]; striker[0.043]; club[0.043]; Milan[0.043]; Liberian[0.043]; Cup[0.042]; called[0.042]; Champions[0.042]; European[0.041]; European[0.041]; defender[0.041]; World[0.041]; played[0.041]; qualifier[0.041]; forward[0.040]; Portugal[0.040]; Portugal[0.040]; Portugal[0.040]; game[0.040]; nursing[0.040]; Georg[0.039]; nose[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portugal ==> ENTITY: \u001b[32mPortugal national football team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.202:0.202[0]; log p(e|m)= -2.453:-2.453[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPortugal[0.047]; Portugal[0.047]; Soccer[0.047]; League[0.044]; club[0.044]; games[0.043]; qualifier[0.043]; Champions[0.043]; qualifying[0.042]; Cup[0.041]; game[0.041]; European[0.041]; European[0.041]; Lisbon[0.040]; Germany[0.040]; played[0.040]; World[0.040]; points[0.040]; Jorge[0.040]; Porto[0.039]; central[0.039]; group[0.039]; Pinto[0.039]; place[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barcelona ==> ENTITY: \u001b[32mFC Barcelona\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.211:0.211[0]; log p(e|m)= -1.709:-1.709[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.045]; Spain[0.045]; Spain[0.045]; Spain[0.045]; Spain[0.045]; Luis[0.044]; Jorge[0.042]; Jose[0.042]; Jose[0.042]; Carlos[0.042]; Sergio[0.042]; Madrid[0.041]; Cadete[0.041]; Santos[0.041]; Fernando[0.041]; Italy[0.040]; Italy[0.040]; Antonio[0.040]; Coruna[0.040]; Paulo[0.040]; Paulo[0.040]; Barcelona[0.040]; Barcelona[0.040]; Benfica[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paulo Bento\u001b[39m ==> ENTITY: \u001b[32mPaulo Bento\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCadete[0.055]; Fernando[0.054]; Sergio[0.053]; Spain[0.053]; Spain[0.053]; Spain[0.053]; Spain[0.053]; Spain[0.053]; Jose[0.053]; Jose[0.053]; Luis[0.053]; Paulo[0.053]; Benfica[0.053]; Sporting[0.052]; Sporting[0.052]; Jorge[0.052]; Cristovao[0.052]; Deportivo[0.051]; Braga[0.051]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Vitor Baia\u001b[39m ==> ENTITY: \u001b[32mVítor Baía\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.228:0.228[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPortugal[0.045]; club[0.045]; games[0.045]; Spain[0.045]; Spain[0.045]; Spain[0.045]; Spain[0.045]; Sergio[0.044]; Luis[0.044]; Champions[0.044]; Paulo[0.043]; League[0.043]; Fernando[0.043]; game[0.043]; Jose[0.043]; Carlos[0.043]; twice[0.043]; Madrid[0.043]; Portuguese[0.042]; Sporting[0.042]; Figo[0.042]; Rui[0.042]; Rui[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.154:0.154[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.055]; Scotland[0.047]; Juventus[0.046]; Spain[0.046]; Spain[0.046]; Spain[0.046]; Spain[0.046]; Fiorentina[0.045]; Cadete[0.045]; Figo[0.044]; Coruna[0.042]; Glasgow[0.042]; Benfica[0.042]; Paulo[0.041]; Paulo[0.041]; Sporting[0.041]; Sporting[0.041]; Salamanca[0.041]; Porto[0.041]; Porto[0.041]; Porto[0.041]; Deportivo[0.041]; Fernando[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jorge Cadete\u001b[39m ==> ENTITY: \u001b[32mJorge Cadete\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCeltic[0.045]; Jose[0.043]; Jose[0.043]; Real[0.043]; Luis[0.043]; Scotland[0.042]; Carlos[0.042]; Fernando[0.042]; Coruna[0.042]; Benfica[0.042]; Paulo[0.042]; Paulo[0.042]; Antonio[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Deportivo[0.041]; Porto[0.040]; Porto[0.040]; Porto[0.040]; Glasgow[0.040]; Italy[0.040]; Italy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sergio Conceicao\u001b[39m ==> ENTITY: \u001b[32mSérgio Conceição\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Italy[0.043]; Portugal[0.043]; League[0.043]; game[0.042]; Sporting[0.042]; qualifying[0.042]; Milan[0.041]; points[0.041]; Barcelona[0.041]; Barcelona[0.041]; Barcelona[0.041]; European[0.041]; European[0.041]; Germany[0.040]; Germany[0.040]; twice[0.040]; Juventus[0.040]; Midfielders[0.040]; \t\n\r [===================================>..........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 4s632ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3590/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1243testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Central African Republic ==> ENTITY: \u001b[32mCentral African Republic\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.308:-0.308[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRepublic[0.049]; Bangui[0.049]; African[0.048]; Central[0.048]; violence[0.047]; capital[0.046]; refugee[0.046]; United[0.045]; Nations[0.045]; official[0.045]; official[0.045]; evacuates[0.045]; army[0.044]; chartered[0.044]; Coast[0.044]; agency[0.044]; said[0.044]; said[0.044]; week[0.044]; evacuated[0.044]; Abidjan[0.043]; Abidjan[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = United Nations\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.098:0.098[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nofficial[0.046]; official[0.046]; agency[0.045]; Republic[0.043]; Republic[0.043]; chartered[0.043]; Central[0.042]; Central[0.042]; week[0.042]; said[0.042]; said[0.042]; Friday[0.041]; staff[0.041]; staff[0.041]; staff[0.041]; refugee[0.041]; mounting[0.041]; African[0.040]; African[0.040]; evacuated[0.039]; violence[0.039]; heading[0.039]; evacuates[0.039]; army[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ivory Coast\u001b[39m ==> ENTITY: \u001b[32mIvory Coast\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAbidjan[0.043]; Abidjan[0.043]; African[0.043]; African[0.043]; refugee[0.043]; Friday[0.043]; week[0.043]; Nations[0.043]; official[0.042]; official[0.042]; violence[0.042]; Central[0.042]; Central[0.042]; Republic[0.042]; Republic[0.042]; said[0.041]; said[0.041]; Bangui[0.041]; army[0.040]; capital[0.040]; heading[0.040]; chartered[0.039]; agency[0.039]; mutiny[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = UNHCR\u001b[39m ==> ENTITY: \u001b[32mUnited Nations High Commissioner for Refugees\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrefugee[0.047]; agency[0.045]; Republic[0.043]; Republic[0.043]; official[0.043]; official[0.043]; said[0.042]; said[0.042]; Nations[0.042]; Central[0.041]; Central[0.041]; violence[0.041]; chartered[0.041]; evacuated[0.041]; African[0.041]; African[0.041]; United[0.041]; week[0.041]; evacuates[0.040]; staff[0.040]; staff[0.040]; staff[0.040]; mounting[0.040]; army[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.100:0.100[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nweek[0.044]; agency[0.044]; Republic[0.043]; Republic[0.043]; official[0.042]; official[0.042]; said[0.042]; said[0.042]; staff[0.042]; staff[0.042]; staff[0.042]; Friday[0.042]; United[0.041]; Nations[0.041]; violence[0.041]; Central[0.041]; Central[0.041]; refugee[0.041]; mounting[0.041]; African[0.041]; African[0.041]; chartered[0.041]; Ivory[0.039]; capital[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bangui\u001b[39m ==> ENTITY: \u001b[32mBangui\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCentral[0.046]; Central[0.046]; Abidjan[0.044]; Abidjan[0.044]; Republic[0.043]; Republic[0.043]; Nations[0.043]; capital[0.043]; African[0.042]; African[0.042]; army[0.041]; plane[0.041]; chartered[0.041]; refugee[0.041]; Friday[0.041]; official[0.040]; official[0.040]; mutiny[0.040]; violence[0.040]; agency[0.039]; evacuated[0.039]; heading[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Abidjan\u001b[39m ==> ENTITY: \u001b[32mAbidjan\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAbidjan[0.047]; African[0.045]; African[0.045]; Bangui[0.044]; capital[0.043]; week[0.043]; evacuated[0.042]; staff[0.042]; staff[0.042]; staff[0.042]; Central[0.042]; Central[0.042]; chartered[0.042]; evacuates[0.041]; Friday[0.041]; official[0.040]; official[0.040]; heading[0.040]; said[0.040]; said[0.040]; Ivory[0.040]; plane[0.040]; Nations[0.040]; army[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.100:0.100[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nweek[0.044]; agency[0.044]; Republic[0.043]; Republic[0.043]; official[0.042]; official[0.042]; said[0.042]; said[0.042]; staff[0.042]; staff[0.042]; staff[0.042]; Friday[0.042]; United[0.041]; Nations[0.041]; violence[0.041]; Central[0.041]; Central[0.041]; refugee[0.041]; mounting[0.041]; African[0.041]; African[0.041]; chartered[0.041]; Ivory[0.039]; capital[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Central African Republic ==> ENTITY: \u001b[32mCentral African Republic\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.308:-0.308[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRepublic[0.049]; Bangui[0.049]; African[0.048]; Central[0.048]; violence[0.047]; capital[0.046]; refugee[0.046]; United[0.045]; Nations[0.045]; official[0.045]; official[0.045]; evacuates[0.045]; army[0.044]; chartered[0.044]; Coast[0.044]; agency[0.044]; said[0.044]; said[0.044]; week[0.044]; evacuated[0.044]; Abidjan[0.043]; Abidjan[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Abidjan\u001b[39m ==> ENTITY: \u001b[32mAbidjan\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAbidjan[0.047]; African[0.045]; African[0.045]; Bangui[0.044]; capital[0.043]; week[0.043]; evacuated[0.042]; staff[0.042]; staff[0.042]; staff[0.042]; Central[0.042]; Central[0.042]; chartered[0.042]; evacuates[0.041]; Friday[0.041]; official[0.040]; official[0.040]; heading[0.040]; said[0.040]; said[0.040]; Ivory[0.040]; plane[0.040]; Nations[0.040]; army[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.100:0.100[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nweek[0.044]; agency[0.044]; Republic[0.043]; Republic[0.043]; official[0.042]; official[0.042]; said[0.042]; said[0.042]; staff[0.042]; staff[0.042]; staff[0.042]; Friday[0.042]; United[0.041]; Nations[0.041]; violence[0.041]; Central[0.041]; Central[0.041]; refugee[0.041]; mounting[0.041]; African[0.041]; African[0.041]; chartered[0.041]; Ivory[0.039]; capital[0.039]; \t\n\r [===================================>..........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 4s588ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3601/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1359testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bali ==> ENTITY: \u001b[32mBali\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.160:0.160[0]; log p(e|m)= -0.104:-0.104[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.047]; Indonesia[0.047]; Indonesia[0.047]; Denpasar[0.044]; Kusuma[0.044]; Budi[0.043]; Wijaya[0.042]; semifinals[0.042]; Badminton[0.042]; finals[0.042]; Semifinal[0.042]; Susanti[0.042]; Denmark[0.041]; China[0.040]; China[0.040]; China[0.040]; singles[0.040]; singles[0.040]; Indra[0.040]; Saturday[0.040]; Permadi[0.039]; World[0.039]; World[0.039]; Ye[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.120:0.120[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTaiwan[0.056]; China[0.055]; China[0.055]; finals[0.049]; Denmark[0.048]; Women[0.048]; World[0.047]; World[0.047]; Badminton[0.047]; Indonesia[0.047]; Indonesia[0.047]; Indonesia[0.047]; Gong[0.047]; semifinals[0.046]; Zhichao[0.046]; Men[0.045]; singles[0.045]; singles[0.045]; Jun[0.045]; Grand[0.045]; Grand[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gong Zhichao\u001b[39m ==> ENTITY: \u001b[32mGong Zhichao\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBadminton[0.052]; China[0.050]; China[0.050]; China[0.050]; finals[0.049]; Taiwan[0.048]; World[0.048]; World[0.048]; singles[0.048]; singles[0.048]; Semifinal[0.048]; Ye[0.048]; semifinals[0.047]; Denmark[0.047]; Women[0.047]; Martin[0.047]; Fung[0.046]; Men[0.046]; Denpasar[0.046]; Jun[0.045]; Wijaya[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Denmark ==> ENTITY: \u001b[32mDenmark\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.059:0.059[0]; log p(e|m)= -0.268:-0.268[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBadminton[0.046]; Taiwan[0.043]; finals[0.043]; Indonesia[0.042]; Indonesia[0.042]; Indonesia[0.042]; China[0.042]; China[0.042]; China[0.042]; World[0.042]; World[0.042]; Semifinal[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; singles[0.041]; singles[0.041]; semifinals[0.041]; Gong[0.041]; Grand[0.041]; Grand[0.041]; Martin[0.040]; Camilla[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Allan Budi Kusuma\u001b[39m ==> ENTITY: \u001b[32mAlan Budikusuma\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.044]; China[0.044]; China[0.044]; Badminton[0.043]; finals[0.043]; Semifinal[0.043]; semifinals[0.042]; singles[0.042]; singles[0.042]; Indonesia[0.042]; Indonesia[0.042]; Indonesia[0.042]; Taiwan[0.042]; Denpasar[0.041]; Jun[0.041]; Women[0.041]; World[0.041]; World[0.041]; Zhichao[0.040]; Wijaya[0.040]; Martin[0.040]; Gong[0.040]; Denmark[0.040]; Susanti[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.151:0.151[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.050]; Indonesia[0.050]; China[0.046]; China[0.046]; China[0.046]; Taiwan[0.042]; semifinals[0.042]; Badminton[0.042]; finals[0.042]; Semifinal[0.042]; World[0.042]; World[0.042]; Bali[0.042]; Denmark[0.041]; Denpasar[0.040]; singles[0.040]; singles[0.040]; Kusuma[0.038]; beat[0.038]; beat[0.038]; beat[0.038]; beat[0.038]; Grand[0.038]; Grand[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ye Zhaoying\u001b[39m ==> ENTITY: \u001b[32mYe Zhaoying\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWorld[0.046]; World[0.046]; Denmark[0.045]; Badminton[0.045]; China[0.045]; China[0.045]; China[0.045]; Semifinal[0.044]; Zhichao[0.044]; Taiwan[0.044]; Gong[0.044]; semifinals[0.043]; finals[0.043]; Indonesia[0.042]; Indonesia[0.042]; Indonesia[0.042]; Denpasar[0.042]; Grand[0.042]; Grand[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.163:0.163[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.050]; Indonesia[0.050]; China[0.047]; China[0.047]; China[0.047]; Taiwan[0.043]; World[0.042]; World[0.042]; Bali[0.042]; Denmark[0.042]; finals[0.041]; Denpasar[0.041]; Wijaya[0.040]; Badminton[0.040]; semifinals[0.039]; singles[0.039]; singles[0.039]; Kusuma[0.039]; beat[0.039]; beat[0.039]; beat[0.039]; beat[0.039]; Susanti[0.038]; Men[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWorld Grand Prix\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFIVB World Grand Prix\u001b[39m <---> \u001b[32mWorld Badminton Grand Prix Finals\u001b[39m\t\nSCORES: global= 0.254:0.253[\u001b[31m0.000\u001b[39m]; local()= 0.126:0.148[\u001b[32m0.022\u001b[39m]; log p(e|m)= -0.744:-1.431[\u001b[31m0.687\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWorld[0.053]; Grand[0.046]; Prix[0.044]; finals[0.043]; Badminton[0.043]; semifinals[0.043]; Denmark[0.043]; singles[0.042]; singles[0.042]; Women[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; China[0.041]; China[0.041]; China[0.041]; Semifinal[0.040]; Saturday[0.040]; Martin[0.040]; Men[0.040]; Allan[0.038]; Camilla[0.038]; Sun[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fung Permadi\u001b[39m ==> ENTITY: \u001b[32mFung Permadi\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsingles[0.044]; singles[0.044]; Indonesia[0.043]; Indonesia[0.043]; Indonesia[0.043]; Semifinal[0.043]; Allan[0.043]; semifinals[0.042]; Taiwan[0.042]; finals[0.041]; Jun[0.041]; Gong[0.041]; Denpasar[0.041]; World[0.041]; World[0.041]; China[0.041]; China[0.041]; China[0.041]; Denmark[0.040]; Wijaya[0.040]; Sun[0.040]; Badminton[0.040]; Budi[0.040]; Bali[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.111:0.111[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTaiwan[0.050]; China[0.050]; China[0.050]; finals[0.043]; Denmark[0.043]; Badminton[0.042]; Indonesia[0.042]; Indonesia[0.042]; Indonesia[0.042]; Gong[0.041]; Zhichao[0.041]; World[0.041]; World[0.041]; semifinals[0.041]; singles[0.040]; singles[0.040]; Jun[0.040]; beat[0.039]; beat[0.039]; beat[0.039]; beat[0.039]; Men[0.039]; Martin[0.039]; Fung[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Susi Susanti\u001b[39m ==> ENTITY: \u001b[32mSusi Susanti\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.135:0.135[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBadminton[0.045]; finals[0.043]; Semifinal[0.043]; Women[0.043]; China[0.042]; China[0.042]; China[0.042]; semifinals[0.042]; singles[0.042]; singles[0.042]; World[0.041]; World[0.041]; Denpasar[0.041]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Grand[0.041]; Grand[0.041]; Allan[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; Denmark[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sun Jun ==> ENTITY: \u001b[32mSun Jun (badminton)\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.165:0.165[0]; log p(e|m)= -0.137:-0.137[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.045]; Zhichao[0.044]; singles[0.043]; singles[0.043]; Gong[0.043]; Semifinal[0.042]; finals[0.042]; World[0.042]; World[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; China[0.041]; China[0.041]; China[0.041]; Ye[0.041]; semifinals[0.041]; Badminton[0.041]; Allan[0.041]; Denpasar[0.040]; Men[0.040]; Martin[0.040]; Taiwan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.110:0.110[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTaiwan[0.050]; China[0.049]; China[0.049]; finals[0.043]; Denmark[0.043]; Badminton[0.042]; Indonesia[0.042]; Indonesia[0.042]; Indonesia[0.042]; Gong[0.041]; World[0.041]; World[0.041]; semifinals[0.041]; Zhichao[0.041]; singles[0.040]; singles[0.040]; Jun[0.040]; beat[0.039]; beat[0.039]; beat[0.039]; beat[0.039]; Men[0.039]; Martin[0.039]; Women[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWorld Grand Prix\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFIVB World Grand Prix\u001b[39m <---> \u001b[32mWorld Badminton Grand Prix Finals\u001b[39m\t\nSCORES: global= 0.254:0.253[\u001b[31m0.000\u001b[39m]; local()= 0.126:0.148[\u001b[32m0.022\u001b[39m]; log p(e|m)= -0.744:-1.431[\u001b[31m0.687\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWorld[0.053]; Grand[0.046]; Prix[0.044]; finals[0.043]; Badminton[0.043]; semifinals[0.043]; Denmark[0.043]; singles[0.042]; singles[0.042]; Women[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; China[0.041]; China[0.041]; China[0.041]; Semifinal[0.040]; Saturday[0.040]; Martin[0.040]; Men[0.040]; Allan[0.038]; Camilla[0.038]; Sun[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.163:0.163[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.050]; Indonesia[0.050]; China[0.047]; China[0.047]; China[0.047]; Taiwan[0.043]; World[0.042]; World[0.042]; Bali[0.042]; Denmark[0.042]; finals[0.041]; Denpasar[0.041]; Wijaya[0.040]; Badminton[0.040]; semifinals[0.039]; singles[0.039]; singles[0.039]; Kusuma[0.039]; beat[0.039]; beat[0.039]; beat[0.039]; beat[0.039]; Susanti[0.038]; Men[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Taiwan ==> ENTITY: \u001b[32mTaiwan\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.078:-0.078[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.049]; China[0.049]; China[0.049]; finals[0.044]; Zhichao[0.043]; Indonesia[0.042]; Indonesia[0.042]; Indonesia[0.042]; World[0.042]; World[0.042]; semifinals[0.041]; Denmark[0.041]; Gong[0.040]; Grand[0.040]; Grand[0.040]; Saturday[0.040]; Martin[0.040]; Badminton[0.040]; Jun[0.040]; beat[0.039]; beat[0.039]; beat[0.039]; beat[0.039]; Men[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Camilla Martin\u001b[39m ==> ENTITY: \u001b[32mCamilla Martin\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.046]; Badminton[0.043]; World[0.043]; World[0.043]; China[0.043]; China[0.043]; China[0.043]; Semifinal[0.043]; singles[0.042]; singles[0.042]; Gong[0.042]; Taiwan[0.042]; semifinals[0.042]; Women[0.042]; Men[0.041]; Zhichao[0.041]; finals[0.041]; Jun[0.041]; Denpasar[0.041]; Grand[0.040]; Grand[0.040]; Allan[0.040]; Wijaya[0.039]; Ye[0.039]; \t\n\r [====================================>.........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 4s484ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3619/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1186testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York\u001b[39m\t\nSCORES: global= 0.240:0.240[0]; local()= 0.078:0.078[0]; log p(e|m)= -0.289:-0.289[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; League[0.045]; Philadelphia[0.044]; Thursday[0.044]; Thursday[0.044]; Boston[0.043]; teams[0.043]; Conference[0.043]; won[0.043]; played[0.043]; Hockey[0.041]; Hockey[0.041]; National[0.041]; Buffalo[0.041]; lost[0.041]; Jersey[0.040]; Hartford[0.040]; Pittsburgh[0.040]; Division[0.040]; Division[0.040]; Games[0.040]; Eastern[0.039]; Atlantic[0.039]; points[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Anaheim ==> ENTITY: \u001b[32mAnaheim Ducks\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.246:0.246[0]; log p(e|m)= -2.189:-2.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.050]; Edmonton[0.046]; Edmonton[0.046]; Vancouver[0.045]; Montreal[0.045]; Buffalo[0.045]; Ottawa[0.044]; Calgary[0.044]; Phoenix[0.043]; Toronto[0.043]; Toronto[0.043]; Pittsburgh[0.043]; Philadelphia[0.043]; Rangers[0.043]; Division[0.043]; Colorado[0.042]; Colorado[0.042]; Detroit[0.042]; Dallas[0.042]; Dallas[0.042]; Los[0.042]; Chicago[0.041]; Chicago[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = St Louis ==> ENTITY: \u001b[32mSt. Louis Blues\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.244:0.244[0]; log p(e|m)= -2.564:-2.564[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.047]; Anaheim[0.047]; Edmonton[0.045]; Ottawa[0.045]; Buffalo[0.045]; Montreal[0.045]; Islanders[0.044]; Toronto[0.044]; Toronto[0.044]; Philadelphia[0.044]; Philadelphia[0.044]; Pittsburgh[0.043]; Louis[0.043]; Vancouver[0.043]; Calgary[0.043]; Colorado[0.042]; Detroit[0.042]; Phoenix[0.042]; Washington[0.042]; Washington[0.042]; Rangers[0.042]; Rangers[0.042]; Tampa[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Anaheim ==> ENTITY: \u001b[32mAnaheim Ducks\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.240:0.240[0]; log p(e|m)= -2.189:-2.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.048]; Edmonton[0.044]; Edmonton[0.044]; Vancouver[0.043]; Montreal[0.043]; Buffalo[0.043]; Ottawa[0.043]; Calgary[0.042]; Phoenix[0.042]; Toronto[0.042]; Toronto[0.042]; Pittsburgh[0.042]; Philadelphia[0.041]; Rangers[0.041]; Division[0.041]; Colorado[0.041]; Colorado[0.041]; Dallas[0.040]; Dallas[0.040]; Los[0.040]; Chicago[0.040]; Chicago[0.040]; Louis[0.039]; Louis[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington Capitals\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.215:0.215[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; Hockey[0.045]; Islanders[0.044]; Ottawa[0.044]; Philadelphia[0.044]; Montreal[0.043]; Buffalo[0.043]; teams[0.042]; League[0.042]; Toronto[0.042]; Pittsburgh[0.042]; Boston[0.041]; Tampa[0.041]; Rangers[0.041]; played[0.041]; Vancouver[0.041]; Dallas[0.040]; won[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Chicago[0.040]; Phoenix[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mOttawa\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mOttawa Senators\u001b[39m <---> \u001b[32mOttawa 67's\u001b[39m\t\nSCORES: global= 0.263:0.237[\u001b[31m0.026\u001b[39m]; local()= 0.204:0.180[\u001b[31m0.024\u001b[39m]; log p(e|m)= -3.689:-4.135[\u001b[31m0.446\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.049]; Hockey[0.049]; games[0.049]; teams[0.048]; Buffalo[0.047]; Islanders[0.047]; Montreal[0.046]; League[0.046]; Rangers[0.046]; Pittsburgh[0.046]; won[0.045]; Philadelphia[0.045]; Tampa[0.044]; Boston[0.044]; Central[0.044]; played[0.044]; Detroit[0.044]; Ice[0.043]; Louis[0.043]; Conference[0.043]; Conference[0.043]; Washington[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mLos Angeles\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mLos Angeles Kings\u001b[39m <---> \u001b[32mLos Angeles\u001b[39m\t\nSCORES: global= 0.258:0.237[\u001b[31m0.020\u001b[39m]; local()= 0.246:0.102[\u001b[31m0.144\u001b[39m]; log p(e|m)= -4.605:-0.101[\u001b[32m4.504\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.046]; Anaheim[0.046]; Edmonton[0.043]; Edmonton[0.043]; Islanders[0.043]; Conference[0.043]; Philadelphia[0.043]; Montreal[0.042]; Buffalo[0.042]; Bay[0.042]; Ottawa[0.042]; Vancouver[0.041]; Toronto[0.041]; Toronto[0.041]; Pittsburgh[0.041]; Phoenix[0.041]; Calgary[0.040]; Detroit[0.040]; Tampa[0.040]; Western[0.040]; Dallas[0.040]; Dallas[0.040]; Colorado[0.040]; Colorado[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Boston ==> ENTITY: \u001b[32mBoston Bruins\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.219:0.219[0]; log p(e|m)= -4.343:-4.343[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; Hockey[0.047]; Hockey[0.047]; York[0.046]; Islanders[0.045]; Buffalo[0.045]; Ottawa[0.045]; Montreal[0.044]; Philadelphia[0.044]; Pittsburgh[0.043]; teams[0.043]; League[0.043]; Rangers[0.043]; Detroit[0.042]; played[0.042]; Northeast[0.042]; Bay[0.042]; Conference[0.042]; Conference[0.042]; won[0.041]; Washington[0.041]; Ice[0.041]; Tampa[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Detroit ==> ENTITY: \u001b[32mDetroit Red Wings\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.215:0.215[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.046]; Ottawa[0.045]; Islanders[0.044]; Edmonton[0.044]; Montreal[0.043]; Conference[0.043]; Philadelphia[0.043]; Vancouver[0.043]; Toronto[0.042]; Rangers[0.042]; Pittsburgh[0.042]; Boston[0.042]; Calgary[0.041]; Tampa[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Phoenix[0.040]; Chicago[0.040]; Washington[0.040]; Colorado[0.040]; Louis[0.040]; Dallas[0.039]; Western[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mDallas\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mDallas Stars\u001b[39m <---> \u001b[32mDallas\u001b[39m\t\nSCORES: global= 0.261:0.236[\u001b[31m0.025\u001b[39m]; local()= 0.231:0.097[\u001b[31m0.134\u001b[39m]; log p(e|m)= -4.200:-0.462[\u001b[32m3.738\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.046]; Anaheim[0.046]; Washington[0.044]; Edmonton[0.044]; Edmonton[0.044]; Buffalo[0.044]; Ottawa[0.043]; Toronto[0.042]; Toronto[0.042]; Montreal[0.042]; Vancouver[0.042]; Philadelphia[0.042]; Rangers[0.042]; Pittsburgh[0.042]; Calgary[0.041]; Phoenix[0.041]; Colorado[0.040]; Colorado[0.040]; Chicago[0.040]; Friday[0.039]; December[0.039]; Division[0.039]; Louis[0.038]; Louis[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = St Louis ==> ENTITY: \u001b[32mSt. Louis Blues\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.249:0.249[0]; log p(e|m)= -2.564:-2.564[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.046]; Anaheim[0.046]; Edmonton[0.044]; Edmonton[0.044]; Ottawa[0.043]; Buffalo[0.043]; Montreal[0.043]; Toronto[0.042]; Toronto[0.042]; Philadelphia[0.042]; Pittsburgh[0.042]; Louis[0.041]; Vancouver[0.041]; Calgary[0.041]; Colorado[0.041]; Colorado[0.041]; Phoenix[0.041]; Washington[0.041]; Rangers[0.040]; Chicago[0.040]; Dallas[0.040]; Friday[0.039]; Division[0.039]; December[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colorado ==> ENTITY: \u001b[32mColorado Avalanche\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.256:0.256[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nColorado[0.052]; Anaheim[0.045]; Anaheim[0.045]; Edmonton[0.043]; Edmonton[0.043]; Buffalo[0.043]; Montreal[0.043]; Ottawa[0.042]; Toronto[0.042]; Toronto[0.042]; Vancouver[0.042]; Islanders[0.041]; Philadelphia[0.041]; Calgary[0.041]; Pittsburgh[0.040]; Phoenix[0.040]; Tampa[0.040]; Rangers[0.040]; Rangers[0.040]; Dallas[0.039]; Dallas[0.039]; Louis[0.039]; Louis[0.039]; Detroit[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NY ISLANDERS\u001b[39m ==> ENTITY: \u001b[32mNew York Islanders\u001b[39m\t\nSCORES: global= 0.317:0.317[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMontreal[0.045]; Buffalo[0.045]; Edmonton[0.044]; Ottawa[0.044]; Philadelphia[0.043]; Boston[0.043]; Toronto[0.043]; Vancouver[0.043]; Pittsburgh[0.042]; Rangers[0.042]; Tampa[0.042]; Colorado[0.042]; Phoenix[0.041]; Detroit[0.041]; Florida[0.041]; Jersey[0.040]; Louis[0.040]; Pts[0.040]; Pts[0.040]; Pts[0.040]; Pts[0.040]; Chicago[0.040]; Hartford[0.040]; Dallas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Florida ==> ENTITY: \u001b[32mFlorida Panthers\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.204:0.204[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; Hockey[0.044]; Hockey[0.044]; Buffalo[0.043]; Ottawa[0.043]; Tampa[0.042]; Islanders[0.042]; teams[0.042]; Conference[0.042]; Conference[0.042]; Philadelphia[0.042]; Montreal[0.042]; League[0.041]; Pittsburgh[0.041]; Toronto[0.041]; won[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; played[0.040]; Phoenix[0.040]; Dallas[0.040]; lost[0.040]; Boston[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Buffalo ==> ENTITY: \u001b[32mBuffalo Sabres\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.224:0.224[0]; log p(e|m)= -3.540:-3.540[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.045]; Anaheim[0.045]; Edmonton[0.044]; Edmonton[0.044]; Ottawa[0.044]; Montreal[0.043]; Toronto[0.043]; Toronto[0.043]; Vancouver[0.042]; Pittsburgh[0.042]; Dallas[0.042]; Dallas[0.042]; Philadelphia[0.042]; Calgary[0.041]; Rangers[0.041]; Washington[0.041]; Colorado[0.040]; Colorado[0.040]; Phoenix[0.040]; Chicago[0.040]; Chicago[0.040]; Friday[0.039]; Louis[0.039]; Louis[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMontreal\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMontreal Canadiens\u001b[39m <---> \u001b[32mMontreal\u001b[39m\t\nSCORES: global= 0.261:0.243[\u001b[31m0.018\u001b[39m]; local()= 0.200:0.090[\u001b[31m0.110\u001b[39m]; log p(e|m)= -3.863:-0.110[\u001b[32m3.753\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.045]; Hockey[0.045]; games[0.045]; Ottawa[0.044]; League[0.043]; Islanders[0.043]; Philadelphia[0.042]; Boston[0.042]; teams[0.042]; Buffalo[0.042]; Detroit[0.042]; Pittsburgh[0.041]; Rangers[0.041]; Thursday[0.041]; Thursday[0.041]; Tampa[0.041]; won[0.041]; played[0.041]; Games[0.040]; Ice[0.040]; York[0.040]; Washington[0.040]; Florida[0.040]; lost[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Toronto ==> ENTITY: \u001b[32mToronto Maple Leafs\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.225:0.225[0]; log p(e|m)= -4.075:-4.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nToronto[0.045]; Ottawa[0.045]; Anaheim[0.044]; Anaheim[0.044]; Edmonton[0.044]; Edmonton[0.044]; Montreal[0.044]; Buffalo[0.043]; Islanders[0.042]; Vancouver[0.042]; Calgary[0.042]; Rangers[0.041]; Rangers[0.041]; Philadelphia[0.041]; Philadelphia[0.041]; Pittsburgh[0.041]; Detroit[0.041]; Louis[0.040]; Louis[0.040]; Tampa[0.039]; Conference[0.039]; Phoenix[0.039]; December[0.039]; Western[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mToronto\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mToronto Maple Leafs\u001b[39m <---> \u001b[32mToronto\u001b[39m\t\nSCORES: global= 0.250:0.237[\u001b[31m0.013\u001b[39m]; local()= 0.220:0.133[\u001b[31m0.087\u001b[39m]; log p(e|m)= -4.075:-0.123[\u001b[32m3.951\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nToronto[0.048]; Ottawa[0.047]; Anaheim[0.047]; Anaheim[0.047]; Edmonton[0.046]; Edmonton[0.046]; Montreal[0.046]; Buffalo[0.045]; Vancouver[0.044]; Calgary[0.044]; Rangers[0.043]; Philadelphia[0.043]; Pittsburgh[0.043]; Louis[0.042]; Louis[0.042]; Phoenix[0.041]; December[0.041]; Dallas[0.041]; Washington[0.041]; Friday[0.041]; Colorado[0.041]; Colorado[0.041]; Division[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = National Hockey League\u001b[39m ==> ENTITY: \u001b[32mNational Hockey League\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.049]; Hockey[0.049]; teams[0.047]; Ottawa[0.047]; Montreal[0.046]; won[0.045]; Buffalo[0.045]; played[0.044]; Philadelphia[0.044]; Boston[0.043]; Pittsburgh[0.043]; goals[0.043]; goals[0.043]; Ice[0.042]; Jersey[0.042]; Florida[0.041]; Hartford[0.041]; York[0.041]; tied[0.041]; Games[0.041]; Thursday[0.041]; Thursday[0.041]; points[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Jose ==> ENTITY: \u001b[32mSan Jose Sharks\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.251:0.251[0]; log p(e|m)= -3.124:-3.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.047]; Anaheim[0.047]; Ottawa[0.044]; Edmonton[0.043]; Edmonton[0.043]; Buffalo[0.043]; Montreal[0.042]; Vancouver[0.042]; Toronto[0.042]; Toronto[0.042]; Calgary[0.041]; Colorado[0.041]; Colorado[0.041]; Philadelphia[0.041]; Dallas[0.041]; Dallas[0.041]; Pittsburgh[0.041]; Phoenix[0.040]; Tampa[0.040]; Conference[0.039]; Rangers[0.039]; Detroit[0.039]; Division[0.039]; Division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dallas ==> ENTITY: \u001b[32mDallas Stars\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.223:0.223[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.045]; Anaheim[0.045]; Washington[0.044]; Conference[0.044]; Edmonton[0.043]; Buffalo[0.043]; Ottawa[0.043]; Islanders[0.042]; Toronto[0.042]; Montreal[0.042]; Vancouver[0.042]; Philadelphia[0.042]; Rangers[0.042]; Pittsburgh[0.041]; Calgary[0.041]; Detroit[0.041]; Tampa[0.041]; Phoenix[0.041]; Central[0.040]; Colorado[0.040]; Chicago[0.039]; Florida[0.039]; Friday[0.039]; Western[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Jersey ==> ENTITY: \u001b[32mNew Jersey Devils\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.221:0.221[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; Hockey[0.044]; Islanders[0.044]; League[0.043]; Philadelphia[0.042]; Montreal[0.042]; played[0.042]; Conference[0.042]; Conference[0.042]; Ottawa[0.042]; Buffalo[0.042]; teams[0.042]; York[0.041]; Rangers[0.041]; Thursday[0.041]; Toronto[0.041]; Standings[0.041]; Boston[0.041]; won[0.041]; Atlantic[0.040]; Tampa[0.040]; Pittsburgh[0.040]; Northeast[0.040]; Detroit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colorado ==> ENTITY: \u001b[32mColorado Avalanche\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.245:0.245[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nColorado[0.053]; Anaheim[0.045]; Anaheim[0.045]; Edmonton[0.044]; Edmonton[0.044]; Buffalo[0.043]; Montreal[0.043]; Ottawa[0.043]; Toronto[0.042]; Toronto[0.042]; Vancouver[0.042]; Philadelphia[0.042]; Calgary[0.042]; Pittsburgh[0.041]; Phoenix[0.040]; Rangers[0.040]; Dallas[0.040]; Louis[0.040]; Washington[0.039]; Chicago[0.039]; Division[0.038]; Friday[0.038]; December[0.038]; Jose[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pittsburgh ==> ENTITY: \u001b[32mPittsburgh Penguins\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.199:0.199[0]; log p(e|m)= -3.689:-3.689[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.046]; Hockey[0.046]; games[0.045]; League[0.043]; Ottawa[0.043]; Buffalo[0.043]; Philadelphia[0.043]; Montreal[0.042]; Islanders[0.042]; teams[0.042]; Boston[0.041]; Florida[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Ice[0.041]; played[0.041]; Washington[0.040]; Tampa[0.040]; Detroit[0.040]; won[0.040]; Rangers[0.040]; Dallas[0.040]; National[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mVancouver\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mVancouver Canucks\u001b[39m <---> \u001b[32mVancouver\u001b[39m\t\nSCORES: global= 0.263:0.240[\u001b[31m0.023\u001b[39m]; local()= 0.256:0.150[\u001b[31m0.105\u001b[39m]; log p(e|m)= -3.912:-0.121[\u001b[32m3.791\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.047]; Anaheim[0.047]; Edmonton[0.046]; Edmonton[0.046]; Islanders[0.046]; Ottawa[0.046]; Toronto[0.045]; Toronto[0.045]; Montreal[0.045]; Calgary[0.044]; Buffalo[0.043]; Rangers[0.043]; Rangers[0.043]; Bay[0.042]; Detroit[0.042]; Philadelphia[0.042]; Washington[0.042]; Washington[0.042]; Pittsburgh[0.042]; Friday[0.041]; Colorado[0.041]; Colorado[0.041]; Tampa[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mOttawa\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mOttawa Senators\u001b[39m <---> \u001b[32mOttawa 67's\u001b[39m\t\nSCORES: global= 0.264:0.236[\u001b[31m0.029\u001b[39m]; local()= 0.226:0.175[\u001b[31m0.052\u001b[39m]; log p(e|m)= -3.689:-4.135[\u001b[31m0.446\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.045]; Anaheim[0.045]; Edmonton[0.045]; Edmonton[0.045]; Toronto[0.045]; Toronto[0.045]; Buffalo[0.044]; Montreal[0.044]; Calgary[0.043]; Rangers[0.043]; Vancouver[0.043]; Pittsburgh[0.043]; Philadelphia[0.042]; Louis[0.041]; Phoenix[0.040]; Washington[0.040]; Division[0.040]; Friday[0.039]; Chicago[0.039]; Colorado[0.038]; Colorado[0.038]; Dallas[0.038]; December[0.037]; Pts[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tampa Bay ==> ENTITY: \u001b[32mTampa Bay Lightning\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.229:0.229[0]; log p(e|m)= -2.120:-2.120[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFlorida[0.045]; Edmonton[0.044]; Ottawa[0.044]; Islanders[0.044]; Montreal[0.044]; Philadelphia[0.043]; Buffalo[0.043]; Toronto[0.043]; Vancouver[0.042]; Pittsburgh[0.042]; Dallas[0.042]; Boston[0.042]; Detroit[0.042]; Rangers[0.041]; Phoenix[0.041]; Washington[0.040]; Chicago[0.040]; Conference[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Louis[0.040]; Colorado[0.040]; Hartford[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington Capitals\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.225:0.225[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.045]; Anaheim[0.045]; Ottawa[0.044]; Philadelphia[0.044]; Edmonton[0.044]; Edmonton[0.044]; Montreal[0.044]; Buffalo[0.043]; Toronto[0.043]; Toronto[0.043]; Pittsburgh[0.043]; Rangers[0.041]; Calgary[0.041]; Vancouver[0.041]; Dallas[0.041]; Chicago[0.041]; Chicago[0.041]; Phoenix[0.040]; Colorado[0.040]; Colorado[0.040]; Division[0.039]; Pacific[0.038]; Friday[0.038]; December[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Montreal ==> ENTITY: \u001b[32mMontreal Canadiens\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.224:0.224[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.045]; Anaheim[0.045]; Ottawa[0.045]; Toronto[0.044]; Toronto[0.044]; Edmonton[0.044]; Edmonton[0.044]; Philadelphia[0.042]; Vancouver[0.042]; Louis[0.042]; Louis[0.042]; Buffalo[0.042]; Pittsburgh[0.042]; Rangers[0.042]; Calgary[0.042]; Friday[0.041]; Washington[0.040]; Chicago[0.040]; Phoenix[0.040]; Division[0.039]; Dallas[0.039]; Jose[0.039]; Colorado[0.039]; Colorado[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Phoenix ==> ENTITY: \u001b[32mPhoenix Coyotes\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.253:0.253[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.045]; Anaheim[0.045]; Edmonton[0.044]; Edmonton[0.044]; Buffalo[0.044]; Ottawa[0.043]; Montreal[0.043]; Islanders[0.042]; Toronto[0.041]; Toronto[0.041]; Conference[0.041]; Calgary[0.041]; Vancouver[0.041]; Philadelphia[0.041]; Philadelphia[0.041]; Pittsburgh[0.041]; Tampa[0.040]; Louis[0.040]; Louis[0.040]; Dallas[0.040]; Dallas[0.040]; Rangers[0.040]; Rangers[0.040]; Detroit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NY RANGERS\u001b[39m ==> ENTITY: \u001b[32mNew York Rangers\u001b[39m\t\nSCORES: global= 0.318:0.318[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIslanders[0.047]; Montreal[0.045]; Ottawa[0.044]; Edmonton[0.044]; Buffalo[0.044]; Boston[0.044]; Toronto[0.043]; Philadelphia[0.043]; Pittsburgh[0.042]; Vancouver[0.042]; won[0.042]; Detroit[0.041]; Louis[0.041]; Tampa[0.041]; Phoenix[0.040]; Thursday[0.040]; Colorado[0.040]; Pts[0.040]; Pts[0.040]; Pts[0.040]; Pts[0.040]; Hartford[0.040]; Chicago[0.040]; Jersey[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlantic ==> ENTITY: \u001b[32mAtlantic Division (NHL)\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.240:0.240[0]; log p(e|m)= -3.381:-3.381[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; teams[0.044]; League[0.044]; Hockey[0.043]; Hockey[0.043]; won[0.043]; Division[0.043]; Division[0.043]; Division[0.043]; Boston[0.042]; Pittsburgh[0.042]; Philadelphia[0.042]; Northeast[0.041]; Conference[0.041]; Conference[0.041]; Ottawa[0.041]; Tampa[0.041]; Detroit[0.040]; Buffalo[0.040]; Islanders[0.040]; Florida[0.040]; Toronto[0.040]; Bay[0.040]; lost[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Calgary ==> ENTITY: \u001b[32mCalgary Flames\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.238:0.238[0]; log p(e|m)= -3.147:-3.147[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.046]; Anaheim[0.046]; Edmonton[0.045]; Edmonton[0.045]; Ottawa[0.043]; Montreal[0.043]; Vancouver[0.043]; Buffalo[0.042]; Toronto[0.042]; Toronto[0.042]; Phoenix[0.041]; Philadelphia[0.041]; Pts[0.041]; Pts[0.041]; Colorado[0.041]; Colorado[0.041]; Pittsburgh[0.040]; Dallas[0.040]; Dallas[0.040]; Rangers[0.040]; Detroit[0.040]; Conference[0.040]; Division[0.039]; Division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NHL ==> ENTITY: \u001b[32mNational Hockey League\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.190:0.190[0]; log p(e|m)= -0.036:-0.036[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.048]; Hockey[0.048]; games[0.048]; teams[0.046]; Ottawa[0.046]; won[0.045]; Montreal[0.045]; League[0.044]; Buffalo[0.044]; Division[0.043]; Division[0.043]; Pittsburgh[0.043]; played[0.043]; Boston[0.042]; Conference[0.042]; Standings[0.042]; Standings[0.042]; tied[0.041]; National[0.041]; Hartford[0.041]; Ice[0.041]; Atlantic[0.041]; lost[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hartford ==> ENTITY: \u001b[32mHartford Whalers\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.219:0.219[0]; log p(e|m)= -3.689:-3.689[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.045]; Hockey[0.045]; games[0.045]; Islanders[0.044]; League[0.044]; Buffalo[0.043]; Boston[0.043]; Montreal[0.043]; Ottawa[0.043]; teams[0.043]; Rangers[0.043]; Philadelphia[0.042]; Pittsburgh[0.041]; Tampa[0.041]; played[0.040]; Conference[0.040]; Ice[0.040]; Washington[0.040]; won[0.040]; Division[0.040]; Division[0.040]; Florida[0.039]; Jersey[0.039]; Northeast[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago Blackhawks\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.245:0.245[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.046]; Anaheim[0.046]; Ottawa[0.044]; Islanders[0.043]; Edmonton[0.043]; Buffalo[0.043]; Vancouver[0.042]; Toronto[0.042]; Toronto[0.042]; Philadelphia[0.041]; Conference[0.041]; Calgary[0.041]; Rangers[0.041]; Rangers[0.041]; Pittsburgh[0.041]; Pittsburgh[0.041]; Division[0.040]; Division[0.040]; Division[0.040]; Detroit[0.040]; Phoenix[0.040]; Louis[0.040]; Tampa[0.040]; Colorado[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Edmonton ==> ENTITY: \u001b[32mEdmonton Oilers\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.223:0.223[0]; log p(e|m)= -3.270:-3.270[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOttawa[0.046]; Anaheim[0.046]; Anaheim[0.046]; Calgary[0.045]; Edmonton[0.045]; Buffalo[0.044]; Vancouver[0.044]; Toronto[0.043]; Toronto[0.043]; Montreal[0.043]; Philadelphia[0.042]; Pittsburgh[0.042]; Rangers[0.041]; Louis[0.041]; Phoenix[0.041]; Colorado[0.040]; Colorado[0.040]; Dallas[0.040]; Chicago[0.039]; Friday[0.039]; Division[0.039]; December[0.038]; Washington[0.038]; Los[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Edmonton ==> ENTITY: \u001b[32mEdmonton Oilers\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.233:0.233[0]; log p(e|m)= -3.270:-3.270[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOttawa[0.045]; Anaheim[0.045]; Anaheim[0.045]; Calgary[0.044]; Edmonton[0.044]; Buffalo[0.043]; Vancouver[0.043]; Toronto[0.043]; Toronto[0.043]; Montreal[0.043]; Islanders[0.043]; Philadelphia[0.041]; Pittsburgh[0.041]; Rangers[0.041]; Rangers[0.041]; Tampa[0.040]; Louis[0.040]; Louis[0.040]; Phoenix[0.040]; Detroit[0.040]; Colorado[0.039]; Colorado[0.039]; Dallas[0.039]; Dallas[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Central Division ==> ENTITY: \u001b[32mCentral Division (NHL)\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.200:0.200[0]; log p(e|m)= -1.374:-1.374[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDivision[0.046]; Division[0.046]; Conference[0.044]; Philadelphia[0.044]; Boston[0.043]; Western[0.043]; Pittsburgh[0.042]; Detroit[0.042]; Buffalo[0.042]; Ottawa[0.042]; Dallas[0.041]; Calgary[0.041]; Jersey[0.041]; Vancouver[0.041]; Chicago[0.041]; Montreal[0.041]; Edmonton[0.041]; Toronto[0.040]; Rangers[0.040]; Phoenix[0.040]; Atlantic[0.040]; Islanders[0.040]; Pacific[0.040]; Florida[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Philadelphia ==> ENTITY: \u001b[32mPhiladelphia Flyers\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.219:0.219[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; Hockey[0.045]; Hockey[0.045]; teams[0.044]; Montreal[0.043]; Ottawa[0.043]; Buffalo[0.042]; Islanders[0.042]; Toronto[0.042]; League[0.042]; Boston[0.041]; Pittsburgh[0.041]; played[0.041]; won[0.041]; Washington[0.041]; Conference[0.040]; Conference[0.040]; Rangers[0.040]; Tampa[0.040]; Ice[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago Blackhawks\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.252:0.252[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.047]; Anaheim[0.047]; Ottawa[0.044]; Edmonton[0.044]; Edmonton[0.044]; Montreal[0.043]; Buffalo[0.043]; Vancouver[0.042]; Toronto[0.042]; Toronto[0.042]; Philadelphia[0.042]; Calgary[0.042]; Rangers[0.041]; Pittsburgh[0.041]; Division[0.041]; Phoenix[0.040]; Louis[0.040]; Louis[0.040]; Colorado[0.040]; Colorado[0.040]; Washington[0.040]; Dallas[0.039]; Friday[0.039]; December[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Philadelphia ==> ENTITY: \u001b[32mPhiladelphia Flyers\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.233:0.233[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.046]; Anaheim[0.046]; Montreal[0.044]; Ottawa[0.044]; Edmonton[0.043]; Edmonton[0.043]; Buffalo[0.043]; Toronto[0.043]; Toronto[0.043]; Pittsburgh[0.042]; Washington[0.042]; Vancouver[0.042]; Rangers[0.041]; Calgary[0.041]; Phoenix[0.040]; Dallas[0.040]; Louis[0.040]; Louis[0.040]; Chicago[0.040]; Colorado[0.040]; Colorado[0.040]; Division[0.040]; Friday[0.039]; December[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pacific ==> ENTITY: \u001b[32mPacific Division (NHL)\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.252:0.252[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.045]; Anaheim[0.045]; Edmonton[0.043]; Edmonton[0.043]; Vancouver[0.043]; Division[0.043]; Division[0.043]; Calgary[0.042]; Western[0.042]; Phoenix[0.041]; Dallas[0.041]; Dallas[0.041]; Conference[0.041]; Friday[0.041]; Bay[0.041]; Ottawa[0.041]; Toronto[0.041]; Toronto[0.041]; Islanders[0.041]; Buffalo[0.041]; Detroit[0.040]; Montreal[0.040]; Colorado[0.040]; Colorado[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Buffalo ==> ENTITY: \u001b[32mBuffalo Sabres\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.198:0.198[0]; log p(e|m)= -3.540:-3.540[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; teams[0.045]; Hockey[0.044]; Hockey[0.044]; Islanders[0.043]; Ottawa[0.043]; League[0.042]; National[0.042]; Montreal[0.042]; won[0.042]; Division[0.041]; Division[0.041]; Division[0.041]; played[0.041]; Pittsburgh[0.041]; York[0.041]; Philadelphia[0.041]; Central[0.040]; Tampa[0.040]; Boston[0.040]; Rangers[0.040]; Thursday[0.040]; Thursday[0.040]; Northeast[0.040]; \t\n\r [====================================>.........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 4s212ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3663/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1170testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.060:0.060[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.046]; club[0.046]; games[0.046]; Sheffield[0.045]; Soccer[0.044]; Liverpool[0.043]; players[0.043]; playing[0.042]; Saturday[0.042]; set[0.042]; comeback[0.042]; Wednesday[0.042]; Wednesday[0.042]; Friday[0.041]; appearing[0.041]; missed[0.040]; sign[0.040]; Feyenoord[0.040]; forward[0.040]; committee[0.040]; committee[0.040]; worldwide[0.039]; contracts[0.039]; Dutch[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = FIFA\u001b[39m ==> ENTITY: \u001b[32mFIFA\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.046]; Fifa[0.045]; Fifa[0.045]; players[0.045]; games[0.044]; player[0.043]; club[0.043]; club[0.043]; club[0.043]; fined[0.043]; playing[0.042]; appearing[0.041]; worldwide[0.040]; ban[0.040]; said[0.040]; contract[0.040]; Barcelona[0.040]; suspension[0.040]; committee[0.040]; committee[0.040]; Feyenoord[0.040]; Feyenoord[0.040]; contracts[0.040]; missed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Blinker\u001b[39m ==> ENTITY: \u001b[32mRegi Blinker\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.045]; club[0.045]; club[0.045]; Feyenoord[0.044]; Feyenoord[0.044]; games[0.042]; playing[0.042]; Sheffield[0.042]; player[0.042]; Italian[0.041]; Fifa[0.041]; Fifa[0.041]; Fifa[0.041]; Udinese[0.041]; Udinese[0.041]; Udinese[0.041]; Barcelona[0.041]; Liverpool[0.041]; Wednesday[0.040]; Wednesday[0.040]; worldwide[0.040]; players[0.040]; Swiss[0.039]; regulations[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Udinese\u001b[39m ==> ENTITY: \u001b[32mUdinese Calcio\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItalian[0.046]; club[0.045]; club[0.045]; club[0.045]; games[0.044]; Soccer[0.044]; Udinese[0.044]; Udinese[0.044]; players[0.042]; playing[0.042]; player[0.042]; Feyenoord[0.041]; Feyenoord[0.041]; Liverpool[0.040]; contracts[0.040]; Fifa[0.040]; Fifa[0.040]; Fifa[0.040]; contract[0.040]; missed[0.039]; comeback[0.039]; Barcelona[0.039]; Wednesday[0.039]; Wednesday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wednesday ==> ENTITY: \u001b[32mSheffield Wednesday F.C.\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.165:0.165[0]; log p(e|m)= -1.720:-1.720[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFriday[0.053]; Wednesday[0.053]; Saturday[0.049]; Sheffield[0.045]; club[0.044]; club[0.044]; club[0.044]; games[0.042]; Soccer[0.041]; contract[0.041]; previous[0.040]; worldwide[0.040]; Liverpool[0.040]; players[0.040]; player[0.040]; playing[0.040]; contracts[0.039]; appearing[0.039]; comeback[0.039]; commitment[0.038]; sign[0.038]; document[0.038]; indefinite[0.038]; Italian[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Liverpool ==> ENTITY: \u001b[32mLiverpool F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.166:0.166[0]; log p(e|m)= -1.171:-1.171[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.046]; club[0.046]; club[0.046]; Sheffield[0.044]; Soccer[0.043]; London[0.043]; games[0.043]; player[0.042]; playing[0.042]; players[0.042]; previous[0.041]; lifted[0.041]; Feyenoord[0.040]; Feyenoord[0.040]; missed[0.040]; Wednesday[0.040]; Wednesday[0.040]; set[0.040]; Swiss[0.040]; Saturday[0.040]; comeback[0.040]; contract[0.040]; Italian[0.039]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Udinese\u001b[39m ==> ENTITY: \u001b[32mUdinese Calcio\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItalian[0.046]; club[0.045]; club[0.045]; club[0.045]; games[0.045]; Udinese[0.044]; Udinese[0.044]; players[0.043]; playing[0.042]; player[0.042]; Feyenoord[0.042]; Feyenoord[0.042]; Liverpool[0.041]; contracts[0.040]; Fifa[0.040]; Fifa[0.040]; contract[0.040]; missed[0.040]; Barcelona[0.039]; Wednesday[0.039]; Saturday[0.039]; worldwide[0.039]; sign[0.039]; valid[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italian ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.759:-0.759[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; club[0.045]; club[0.045]; club[0.045]; player[0.045]; players[0.043]; Udinese[0.042]; Udinese[0.042]; Udinese[0.042]; playing[0.042]; Fifa[0.041]; Fifa[0.041]; Fifa[0.041]; Wednesday[0.041]; Wednesday[0.041]; comeback[0.040]; contract[0.040]; Feyenoord[0.040]; Feyenoord[0.040]; worldwide[0.040]; contracts[0.040]; missed[0.039]; Barcelona[0.039]; forward[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Udinese\u001b[39m ==> ENTITY: \u001b[32mUdinese Calcio\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItalian[0.046]; club[0.045]; club[0.045]; club[0.045]; games[0.044]; Soccer[0.044]; Udinese[0.044]; Udinese[0.044]; players[0.042]; playing[0.042]; player[0.042]; Feyenoord[0.041]; Feyenoord[0.041]; Liverpool[0.040]; contracts[0.040]; Fifa[0.040]; Fifa[0.040]; Fifa[0.040]; contract[0.040]; missed[0.039]; comeback[0.039]; Barcelona[0.039]; Wednesday[0.039]; Wednesday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Blinker\u001b[39m ==> ENTITY: \u001b[32mRegi Blinker\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.046]; Feyenoord[0.045]; Dutch[0.044]; games[0.043]; playing[0.043]; Sheffield[0.043]; Soccer[0.043]; London[0.042]; Italian[0.042]; Fifa[0.041]; Fifa[0.041]; Fifa[0.041]; Udinese[0.041]; Udinese[0.041]; Barcelona[0.041]; Liverpool[0.041]; Wednesday[0.041]; Wednesday[0.041]; worldwide[0.041]; players[0.041]; sign[0.040]; Friday[0.040]; Saturday[0.039]; missed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = FIFA\u001b[39m ==> ENTITY: \u001b[32mFIFA\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; Soccer[0.045]; Fifa[0.045]; Fifa[0.045]; playing[0.045]; players[0.044]; player[0.043]; club[0.043]; club[0.043]; fined[0.043]; appearing[0.041]; worldwide[0.041]; ban[0.040]; said[0.040]; missed[0.040]; suspension[0.040]; committee[0.040]; committee[0.040]; Liverpool[0.040]; Feyenoord[0.040]; Feyenoord[0.040]; Sheffield[0.039]; sign[0.039]; Italian[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Feyenoord\u001b[39m ==> ENTITY: \u001b[32mFeyenoord\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.047]; club[0.047]; club[0.047]; Feyenoord[0.047]; games[0.044]; players[0.044]; player[0.042]; playing[0.042]; Liverpool[0.042]; sign[0.040]; Sheffield[0.040]; Fifa[0.040]; Fifa[0.040]; Fifa[0.040]; contract[0.040]; Udinese[0.040]; Udinese[0.040]; Udinese[0.040]; Barcelona[0.040]; missed[0.040]; Wednesday[0.040]; Wednesday[0.040]; Saturday[0.039]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Feyenoord\u001b[39m ==> ENTITY: \u001b[32mFeyenoord\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.049]; club[0.049]; club[0.049]; Feyenoord[0.048]; Soccer[0.046]; games[0.046]; players[0.045]; player[0.044]; playing[0.043]; Liverpool[0.043]; Dutch[0.042]; sign[0.041]; Sheffield[0.041]; Fifa[0.041]; Fifa[0.041]; Fifa[0.041]; contract[0.041]; forward[0.041]; Udinese[0.041]; Udinese[0.041]; Udinese[0.041]; Barcelona[0.041]; missed[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Blinker\u001b[39m ==> ENTITY: \u001b[32mRegi Blinker\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.045]; club[0.045]; club[0.045]; Feyenoord[0.044]; Feyenoord[0.044]; Dutch[0.043]; games[0.042]; playing[0.042]; Sheffield[0.042]; Soccer[0.042]; player[0.041]; London[0.041]; Italian[0.041]; Fifa[0.041]; Fifa[0.041]; Fifa[0.041]; Udinese[0.040]; Udinese[0.040]; Udinese[0.040]; Barcelona[0.040]; Liverpool[0.040]; Wednesday[0.040]; Wednesday[0.040]; worldwide[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swiss ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; player[0.045]; playing[0.045]; players[0.044]; francs[0.043]; Italian[0.043]; Fifa[0.042]; Fifa[0.042]; Fifa[0.042]; club[0.041]; club[0.041]; club[0.041]; Wednesday[0.041]; Wednesday[0.041]; Saturday[0.041]; comeback[0.041]; fined[0.041]; set[0.041]; missed[0.041]; worldwide[0.040]; Friday[0.040]; said[0.040]; failing[0.040]; failing[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mDutch\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNetherlands\u001b[39m <---> \u001b[32mNetherlands national football team\u001b[39m\t\nSCORES: global= 0.255:0.245[\u001b[31m0.010\u001b[39m]; local()= 0.055:0.143[\u001b[32m0.088\u001b[39m]; log p(e|m)= -0.794:-4.711[\u001b[31m3.916\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayers[0.046]; games[0.045]; playing[0.044]; Feyenoord[0.044]; Soccer[0.043]; Fifa[0.042]; Fifa[0.042]; Fifa[0.042]; comeback[0.042]; club[0.042]; club[0.042]; missed[0.041]; Liverpool[0.041]; failing[0.041]; Reggie[0.041]; Saturday[0.041]; Italian[0.041]; make[0.041]; Wednesday[0.040]; Wednesday[0.040]; appearing[0.040]; Friday[0.040]; Sheffield[0.040]; Barcelona[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = FIFA\u001b[39m ==> ENTITY: \u001b[32mFIFA\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.046]; Fifa[0.045]; Fifa[0.045]; players[0.045]; games[0.044]; player[0.043]; club[0.043]; club[0.043]; club[0.043]; fined[0.043]; playing[0.042]; appearing[0.041]; worldwide[0.040]; ban[0.040]; said[0.040]; contract[0.040]; Barcelona[0.040]; suspension[0.040]; committee[0.040]; committee[0.040]; Feyenoord[0.040]; Feyenoord[0.040]; contracts[0.040]; missed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sheffield Wednesday\u001b[39m ==> ENTITY: \u001b[32mSheffield Wednesday F.C.\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayers[0.045]; Soccer[0.045]; club[0.044]; club[0.044]; club[0.044]; games[0.044]; Wednesday[0.043]; player[0.043]; Saturday[0.042]; playing[0.042]; Friday[0.041]; contract[0.041]; Feyenoord[0.041]; Feyenoord[0.041]; missed[0.041]; Liverpool[0.041]; status[0.040]; contracts[0.040]; forward[0.040]; sign[0.040]; comeback[0.039]; Fifa[0.039]; Fifa[0.039]; Fifa[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBarcelona\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFC Barcelona\u001b[39m <---> \u001b[32mBarcelona\u001b[39m\t\nSCORES: global= 0.268:0.240[\u001b[31m0.028\u001b[39m]; local()= 0.156:0.086[\u001b[31m0.071\u001b[39m]; log p(e|m)= -1.709:-0.371[\u001b[32m1.338\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.047]; club[0.046]; club[0.046]; club[0.046]; players[0.044]; player[0.044]; games[0.044]; playing[0.042]; Feyenoord[0.042]; Feyenoord[0.042]; contract[0.041]; Fifa[0.040]; Fifa[0.040]; Fifa[0.040]; previous[0.040]; London[0.040]; worldwide[0.040]; appearing[0.040]; Wednesday[0.040]; Wednesday[0.040]; suspension[0.040]; set[0.040]; Liverpool[0.040]; comeback[0.039]; \t\n\r [====================================>.........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 4s114ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3682/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1294testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = India ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.052]; Hindu[0.047]; Hindu[0.047]; Hindu[0.047]; Hindu[0.047]; Hindu[0.047]; Delhi[0.045]; Janata[0.042]; minute[0.041]; forces[0.040]; Bharatiya[0.040]; Friday[0.040]; mosque[0.040]; mosque[0.040]; December[0.040]; forced[0.039]; Babri[0.039]; resumed[0.039]; reconvened[0.039]; razed[0.038]; parliament[0.038]; parliament[0.038]; observe[0.038]; nationalists[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = BJP\u001b[39m ==> ENTITY: \u001b[32mBharatiya Janata Party\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.046]; India[0.046]; Hindu[0.044]; Hindu[0.044]; Hindu[0.044]; Hindu[0.044]; Hindu[0.044]; Hindu[0.044]; Indian[0.043]; Indian[0.043]; Bharatiya[0.043]; Delhi[0.042]; party[0.042]; Janata[0.042]; Bhan[0.041]; Ayodhya[0.039]; parliament[0.039]; parliament[0.039]; Party[0.039]; December[0.039]; pro[0.038]; Babri[0.038]; nationalist[0.038]; adjourned[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mHindu\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHindu\u001b[39m <---> \u001b[32mHindu nationalism\u001b[39m\t\nSCORES: global= 0.257:0.222[\u001b[31m0.036\u001b[39m]; local()= 0.218:0.155[\u001b[31m0.063\u001b[39m]; log p(e|m)= -0.377:-3.016[\u001b[31m2.639\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHindu[0.050]; temple[0.048]; Hindus[0.048]; India[0.044]; Indian[0.044]; Indian[0.044]; birthplace[0.042]; people[0.042]; people[0.042]; Bhan[0.041]; Ayodhya[0.041]; Moslem[0.041]; Moslem[0.041]; Bombay[0.040]; century[0.040]; mosque[0.040]; mosque[0.040]; mosque[0.040]; Rama[0.040]; city[0.039]; town[0.039]; hardline[0.039]; Lord[0.039]; main[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMoslem\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMuslim\u001b[39m <---> \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.267:0.261[\u001b[31m0.006\u001b[39m]; local()= 0.212:0.179[\u001b[31m0.033\u001b[39m]; log p(e|m)= -0.308:-1.214[\u001b[31m0.906\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoslem[0.047]; Hindu[0.045]; Hindu[0.045]; Hindu[0.045]; Hindus[0.044]; northern[0.043]; mosque[0.042]; mosque[0.042]; mosque[0.042]; Indian[0.042]; Indian[0.042]; people[0.042]; people[0.042]; birthplace[0.042]; India[0.041]; century[0.040]; town[0.040]; months[0.040]; temple[0.039]; hardline[0.039]; political[0.038]; officials[0.038]; violence[0.038]; city[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = BJP\u001b[39m ==> ENTITY: \u001b[32mBharatiya Janata Party\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.045]; India[0.045]; Hindu[0.043]; Hindu[0.043]; Hindu[0.043]; Hindu[0.043]; Hindu[0.043]; Hindu[0.043]; Hindu[0.043]; Indian[0.043]; Indian[0.043]; Bharatiya[0.043]; Delhi[0.042]; party[0.042]; Janata[0.041]; Bhan[0.041]; hardline[0.040]; Hindus[0.039]; Ayodhya[0.039]; months[0.039]; parliament[0.039]; parliament[0.039]; Party[0.039]; December[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mHindu\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHindu\u001b[39m <---> \u001b[32mHindu nationalism\u001b[39m\t\nSCORES: global= 0.257:0.223[\u001b[31m0.034\u001b[39m]; local()= 0.246:0.198[\u001b[31m0.047\u001b[39m]; log p(e|m)= -0.377:-3.016[\u001b[31m2.639\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHindu[0.049]; Hindu[0.049]; Hindu[0.049]; Hindu[0.049]; Hindu[0.049]; India[0.043]; India[0.043]; Indian[0.043]; Indian[0.043]; people[0.041]; Bhan[0.040]; Ayodhya[0.040]; Moslem[0.040]; Delhi[0.039]; century[0.039]; mosque[0.039]; mosque[0.039]; mosque[0.039]; Bharatiya[0.038]; nationalist[0.038]; town[0.038]; observe[0.038]; northern[0.037]; party[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = BJP\u001b[39m ==> ENTITY: \u001b[32mBharatiya Janata Party\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.048]; Hindu[0.046]; Hindu[0.046]; Indian[0.045]; Indian[0.045]; party[0.044]; Bhan[0.043]; hardline[0.043]; Hindus[0.042]; opposition[0.042]; political[0.042]; Ayodhya[0.041]; months[0.041]; Bombay[0.040]; officials[0.040]; Suraj[0.039]; catapulted[0.039]; speaker[0.039]; Monday[0.039]; blame[0.039]; suspended[0.039]; campaign[0.039]; campaign[0.039]; temple[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bombay ==> ENTITY: \u001b[32mMumbai\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.242:-0.242[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.050]; Indian[0.047]; Indian[0.047]; Hindu[0.043]; Hindu[0.043]; Hindu[0.043]; town[0.042]; city[0.042]; Hindus[0.042]; northern[0.041]; temple[0.041]; century[0.041]; build[0.041]; Monday[0.040]; razed[0.040]; Bhan[0.040]; Moslem[0.040]; Moslem[0.040]; Ayodhya[0.040]; officials[0.040]; months[0.040]; Babri[0.039]; string[0.039]; Suraj[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Suraj Bhan\u001b[39m ==> ENTITY: \u001b[32mSuraj Bhan\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.132:0.132[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDecember[0.046]; India[0.043]; India[0.043]; Bharatiya[0.043]; Delhi[0.042]; adjourned[0.042]; speaker[0.042]; Janata[0.042]; Party[0.042]; political[0.041]; months[0.041]; parliament[0.041]; parliament[0.041]; deputy[0.041]; deputy[0.041]; Indian[0.041]; Indian[0.041]; temple[0.041]; Ayodhya[0.041]; Monday[0.041]; Babri[0.041]; communist[0.040]; adjournment[0.040]; adjourn[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mHindu\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHindu\u001b[39m <---> \u001b[32mHindu nationalism\u001b[39m\t\nSCORES: global= 0.257:0.225[\u001b[31m0.032\u001b[39m]; local()= 0.191:0.167[\u001b[31m0.025\u001b[39m]; log p(e|m)= -0.377:-3.016[\u001b[31m2.639\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHindu[0.051]; Hindu[0.051]; Hindu[0.051]; Hindu[0.051]; India[0.045]; India[0.045]; Delhi[0.041]; mosque[0.041]; mosque[0.041]; Bharatiya[0.040]; nationalist[0.040]; observe[0.040]; party[0.039]; house[0.039]; house[0.039]; house[0.039]; Party[0.039]; nationalists[0.039]; remembrance[0.039]; Janata[0.038]; Babri[0.038]; parliament[0.038]; parliament[0.038]; lower[0.037]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mHindu\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHindu\u001b[39m <---> \u001b[32mHindu nationalism\u001b[39m\t\nSCORES: global= 0.257:0.225[\u001b[31m0.032\u001b[39m]; local()= 0.196:0.173[\u001b[31m0.023\u001b[39m]; log p(e|m)= -0.377:-3.016[\u001b[31m2.639\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHindu[0.051]; Hindu[0.051]; Hindu[0.051]; Hindu[0.051]; India[0.045]; India[0.045]; Bhan[0.042]; Delhi[0.041]; mosque[0.041]; mosque[0.041]; Bharatiya[0.040]; nationalist[0.040]; observe[0.040]; party[0.039]; house[0.039]; house[0.039]; house[0.039]; Party[0.039]; nationalists[0.039]; remembrance[0.038]; Janata[0.038]; Babri[0.038]; parliament[0.038]; parliament[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = BJP\u001b[39m ==> ENTITY: \u001b[32mBharatiya Janata Party\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.048]; Hindu[0.046]; Hindu[0.046]; Indian[0.045]; Indian[0.045]; party[0.044]; Bhan[0.043]; hardline[0.043]; Hindus[0.042]; opposition[0.042]; political[0.041]; Ayodhya[0.041]; months[0.041]; December[0.041]; Bombay[0.040]; Babri[0.040]; adjourned[0.039]; officials[0.039]; Suraj[0.039]; catapulted[0.039]; speaker[0.039]; Monday[0.039]; blame[0.039]; suspended[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indian ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.618:-0.618[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.050]; Indian[0.049]; Hindu[0.045]; Hindu[0.045]; Hindu[0.045]; Hindu[0.045]; Hindu[0.045]; Hindus[0.044]; Bombay[0.042]; northern[0.042]; temple[0.041]; century[0.041]; birthplace[0.040]; Bhan[0.040]; people[0.039]; people[0.039]; Moslem[0.039]; Moslem[0.039]; months[0.039]; work[0.039]; city[0.038]; town[0.038]; believe[0.038]; Ayodhya[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bharatiya Janata Party\u001b[39m ==> ENTITY: \u001b[32mBharatiya Janata Party\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.046]; India[0.046]; Hindu[0.045]; Hindu[0.045]; Hindu[0.045]; Hindu[0.045]; Hindu[0.045]; Hindu[0.045]; Indian[0.043]; Indian[0.043]; Delhi[0.042]; party[0.042]; Bhan[0.041]; nationalist[0.040]; Ayodhya[0.039]; parliament[0.039]; parliament[0.039]; December[0.039]; nationalists[0.039]; pro[0.039]; Babri[0.038]; work[0.038]; adjourned[0.038]; people[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rama ==> ENTITY: \u001b[32mRama\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.166:0.166[0]; log p(e|m)= -0.126:-0.126[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntemple[0.049]; Hindu[0.046]; Hindu[0.046]; Hindus[0.045]; birthplace[0.044]; Bhan[0.043]; Indian[0.043]; Indian[0.043]; India[0.042]; Ayodhya[0.042]; Lord[0.042]; underworld[0.040]; town[0.040]; northern[0.040]; people[0.040]; people[0.040]; century[0.040]; Suraj[0.040]; work[0.040]; Bombay[0.039]; killed[0.039]; killed[0.039]; months[0.039]; believe[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = New Delhi\u001b[39m ==> ENTITY: \u001b[32mNew Delhi\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.112:0.112[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.049]; India[0.049]; Hindu[0.044]; Hindu[0.044]; Hindu[0.044]; Hindu[0.044]; Hindu[0.044]; Janata[0.043]; mosque[0.041]; mosque[0.041]; parliament[0.040]; parliament[0.040]; Bharatiya[0.040]; Bhan[0.040]; party[0.040]; reconvened[0.040]; Suraj[0.040]; house[0.040]; house[0.040]; house[0.040]; Friday[0.039]; protest[0.039]; adjourned[0.039]; Babri[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ayodhya\u001b[39m ==> ENTITY: \u001b[32mAyodhya\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntemple[0.045]; Hindu[0.044]; Hindu[0.044]; Hindu[0.044]; Hindu[0.044]; Hindu[0.044]; Hindus[0.044]; birthplace[0.042]; Babri[0.042]; Rama[0.042]; India[0.041]; mosque[0.041]; mosque[0.041]; mosque[0.041]; mosque[0.041]; Indian[0.041]; Indian[0.041]; razed[0.040]; site[0.040]; Bhan[0.040]; city[0.039]; party[0.039]; Bharatiya[0.039]; Bombay[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = India ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.140:0.140[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.052]; Hindu[0.046]; Hindu[0.046]; Hindu[0.046]; Hindu[0.046]; Hindu[0.046]; Delhi[0.044]; century[0.042]; Bhan[0.042]; Janata[0.041]; Monday[0.041]; minute[0.040]; forces[0.040]; Bharatiya[0.040]; Friday[0.039]; mosque[0.039]; mosque[0.039]; mosque[0.039]; December[0.039]; forced[0.039]; Babri[0.038]; resumed[0.038]; reconvened[0.038]; razed[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = India ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.168:0.168[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndian[0.047]; Indian[0.047]; Hindu[0.046]; Hindu[0.046]; Hindus[0.044]; Bombay[0.044]; Moslem[0.043]; Moslem[0.043]; century[0.042]; northern[0.042]; Bhan[0.042]; birthplace[0.041]; months[0.041]; Monday[0.040]; temple[0.040]; Ayodhya[0.040]; people[0.039]; people[0.039]; Lord[0.039]; mosque[0.039]; mosque[0.039]; town[0.039]; officials[0.039]; Rama[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indian ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.618:-0.618[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.050]; Indian[0.049]; Hindu[0.045]; Hindu[0.045]; Hindu[0.045]; Hindu[0.045]; Hindu[0.045]; Hindus[0.044]; Bombay[0.042]; northern[0.042]; temple[0.041]; century[0.041]; birthplace[0.040]; Bhan[0.040]; people[0.039]; people[0.039]; Moslem[0.039]; Moslem[0.039]; months[0.039]; work[0.039]; city[0.038]; town[0.038]; believe[0.038]; Ayodhya[0.037]; \t\n\r [====================================>.........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 4s1ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3702/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1281testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = St Louis ==> ENTITY: \u001b[32mSt. Louis\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.094:0.094[0]; log p(e|m)= -0.092:-0.092[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLouis[0.045]; south[0.044]; Thursday[0.043]; Mississippi[0.043]; Friday[0.042]; barge[0.042]; barge[0.042]; Ohio[0.042]; barges[0.041]; barges[0.041]; barges[0.041]; barges[0.041]; barges[0.041]; January[0.041]; Exchange[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; week[0.041]; week[0.041]; week[0.041]; day[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = St. Louis ==> ENTITY: \u001b[32mSt. Louis\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.285:-0.285[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLouis[0.051]; Thursday[0.049]; Friday[0.049]; south[0.049]; Mississippi[0.048]; rates[0.048]; percent[0.047]; percent[0.047]; percent[0.047]; percent[0.047]; percent[0.047]; percent[0.047]; percent[0.047]; barge[0.047]; barge[0.047]; week[0.047]; week[0.047]; week[0.047]; points[0.047]; points[0.047]; December[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Illinois ==> ENTITY: \u001b[32mIllinois\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.170:-0.170[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIllinois[0.049]; Illinois[0.049]; January[0.049]; offered[0.048]; offered[0.048]; offered[0.048]; offered[0.048]; offered[0.048]; week[0.047]; week[0.047]; week[0.047]; week[0.047]; week[0.047]; week[0.047]; August[0.047]; August[0.047]; Mississippi[0.047]; Mississippi[0.047]; Ohio[0.047]; December[0.047]; Thursday[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Illinois ==> ENTITY: \u001b[32mIllinois\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.170:-0.170[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIllinois[0.049]; Illinois[0.049]; January[0.049]; offered[0.048]; offered[0.048]; offered[0.048]; offered[0.048]; offered[0.048]; week[0.047]; week[0.047]; week[0.047]; week[0.047]; week[0.047]; week[0.047]; August[0.047]; August[0.047]; Mississippi[0.047]; Mississippi[0.047]; Ohio[0.047]; December[0.047]; Thursday[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.093:0.093[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbid[0.046]; bid[0.046]; bid[0.046]; bid[0.046]; bid[0.046]; Illinois[0.042]; Illinois[0.042]; week[0.041]; week[0.041]; week[0.041]; week[0.041]; August[0.041]; August[0.041]; January[0.041]; mid[0.040]; mid[0.040]; percent[0.040]; percent[0.040]; percent[0.040]; percent[0.040]; percent[0.040]; percent[0.040]; percent[0.040]; percent[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.081:0.081[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nweek[0.044]; week[0.044]; week[0.044]; percent[0.042]; percent[0.042]; percent[0.042]; percent[0.042]; percent[0.042]; day[0.042]; January[0.042]; Ohio[0.042]; bid[0.042]; bid[0.042]; bid[0.042]; mid[0.042]; versus[0.041]; December[0.041]; higher[0.041]; Exchange[0.041]; session[0.040]; session[0.040]; Thursday[0.040]; offered[0.040]; offered[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ohio ==> ENTITY: \u001b[32mOhio\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.191:-0.191[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsouth[0.055]; Mississippi[0.054]; Mississippi[0.054]; Thursday[0.054]; Friday[0.054]; January[0.054]; Illinois[0.053]; Illinois[0.053]; Illinois[0.053]; December[0.052]; August[0.052]; August[0.052]; week[0.051]; week[0.051]; week[0.051]; week[0.051]; week[0.051]; week[0.051]; open[0.051]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Illinois ==> ENTITY: \u001b[32mIllinois\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.170:-0.170[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsouth[0.043]; January[0.043]; Illinois[0.043]; Illinois[0.043]; Mississippi[0.042]; week[0.042]; week[0.042]; week[0.042]; week[0.042]; week[0.042]; barge[0.042]; barge[0.042]; August[0.041]; August[0.041]; Ohio[0.041]; December[0.041]; Thursday[0.041]; barges[0.041]; barges[0.041]; barges[0.041]; barges[0.041]; barges[0.041]; barges[0.041]; barges[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nweek[0.046]; week[0.046]; week[0.046]; percent[0.044]; percent[0.044]; percent[0.044]; percent[0.044]; percent[0.044]; day[0.044]; January[0.044]; Ohio[0.043]; bid[0.043]; bid[0.043]; bid[0.043]; versus[0.043]; December[0.043]; higher[0.042]; Exchange[0.042]; session[0.042]; session[0.042]; Thursday[0.042]; offered[0.042]; offered[0.042]; \t\n\r [=====================================>........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 3s956ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3711/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1255testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bratislava\u001b[39m ==> ENTITY: \u001b[32mBratislava\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.142:0.142[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.047]; western[0.046]; Slovakia[0.045]; Slovakia[0.045]; outside[0.044]; outside[0.044]; Slovak[0.043]; Slovak[0.043]; Frantisek[0.042]; entrance[0.042]; Vladimir[0.041]; main[0.041]; Wednesday[0.041]; Friday[0.041]; leaving[0.040]; agency[0.040]; Galanta[0.040]; official[0.040]; Meciar[0.039]; party[0.039]; party[0.039]; home[0.039]; home[0.039]; home[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Galanta\u001b[39m ==> ENTITY: \u001b[32mGalanta\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwestern[0.045]; Frantisek[0.044]; Bratislava[0.043]; Slovak[0.043]; Slovak[0.043]; said[0.042]; said[0.042]; said[0.042]; Slovakia[0.042]; Slovakia[0.042]; country[0.041]; country[0.041]; damaged[0.041]; children[0.041]; family[0.041]; family[0.041]; wife[0.041]; main[0.041]; Minister[0.041]; lack[0.041]; lack[0.041]; expelled[0.041]; expelled[0.041]; Prime[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Movement for a Democratic Slovakia\u001b[39m ==> ENTITY: \u001b[32mPeople's Party – Movement for a Democratic Slovakia\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nparliamentary[0.046]; Meciar[0.046]; Slovak[0.045]; Slovak[0.045]; parliament[0.044]; party[0.044]; party[0.044]; Slovakia[0.043]; member[0.043]; Bratislava[0.042]; quit[0.041]; Frantisek[0.041]; country[0.041]; country[0.041]; Vladimir[0.041]; Prime[0.040]; Minister[0.040]; Galanta[0.040]; outside[0.039]; outside[0.039]; official[0.039]; expelled[0.039]; expelled[0.039]; main[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Slovak ==> ENTITY: \u001b[32mSlovakia\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.158:0.158[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.048]; country[0.048]; Slovakia[0.047]; Slovakia[0.047]; western[0.047]; Slovak[0.046]; Bratislava[0.045]; Meciar[0.044]; Frantisek[0.043]; Movement[0.043]; family[0.042]; Vladimir[0.042]; lack[0.042]; lack[0.042]; month[0.042]; home[0.042]; home[0.042]; home[0.042]; parliament[0.042]; parliamentary[0.042]; news[0.041]; expelled[0.041]; expelled[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Vladimir Meciar\u001b[39m ==> ENTITY: \u001b[32mVladimír Mečiar\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nparliamentary[0.045]; quit[0.043]; country[0.043]; country[0.043]; parliament[0.042]; Frantisek[0.042]; Slovak[0.042]; Slovak[0.042]; democracy[0.042]; democracy[0.042]; Slovakia[0.042]; Slovakia[0.042]; Minister[0.042]; party[0.042]; party[0.042]; expelled[0.041]; expelled[0.041]; Prime[0.041]; Bratislava[0.041]; Galanta[0.041]; Movement[0.040]; member[0.040]; leaving[0.040]; mandate[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Slovakia ==> ENTITY: \u001b[32mSlovakia\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.154:0.154[0]; log p(e|m)= -0.367:-0.367[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSlovakia[0.055]; country[0.049]; country[0.049]; western[0.048]; Slovak[0.048]; Slovak[0.048]; Bratislava[0.048]; parliamentary[0.047]; month[0.046]; member[0.045]; parliament[0.045]; official[0.044]; expelled[0.043]; expelled[0.043]; Frantisek[0.043]; Wednesday[0.043]; Friday[0.043]; protest[0.042]; news[0.042]; Meciar[0.042]; Vladimir[0.042]; explosion[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Slovak ==> ENTITY: \u001b[32mSlovakia\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.146:0.146[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.048]; Slovakia[0.047]; Slovakia[0.047]; western[0.047]; Slovak[0.047]; Bratislava[0.046]; Meciar[0.045]; parliament[0.044]; parliamentary[0.044]; Frantisek[0.043]; Movement[0.043]; news[0.043]; family[0.042]; Vladimir[0.042]; lack[0.042]; home[0.042]; home[0.042]; home[0.042]; expelled[0.041]; expelled[0.041]; official[0.041]; democracy[0.041]; member[0.041]; \t\n\r [=====================================>........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 3s918ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3718/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1261testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexico City ==> ENTITY: \u001b[32mMexico City\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.100:0.100[0]; log p(e|m)= -0.028:-0.028[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nregion[0.048]; Mexico[0.047]; area[0.045]; southern[0.045]; Mexican[0.044]; waters[0.043]; place[0.042]; local[0.042]; Tampico[0.042]; Gulf[0.041]; coast[0.041]; coast[0.041]; deep[0.041]; authorities[0.040]; adjacent[0.040]; Baja[0.040]; Transportation[0.040]; open[0.040]; said[0.040]; said[0.040]; Caribbean[0.040]; major[0.040]; California[0.040]; rains[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPacific Coast\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPacific Ocean\u001b[39m <---> \u001b[32mPacific coast\u001b[39m\t\nSCORES: global= 0.261:0.260[\u001b[31m0.001\u001b[39m]; local()= 0.206:0.200[\u001b[31m0.007\u001b[39m]; log p(e|m)= -3.540:-0.377[\u001b[32m3.164\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPacific[0.046]; coast[0.046]; coast[0.046]; coast[0.046]; Coast[0.046]; Caribbean[0.044]; waters[0.044]; region[0.044]; Mexico[0.043]; Mexico[0.043]; Mexico[0.043]; peninsula[0.043]; area[0.043]; north[0.043]; southern[0.043]; rest[0.042]; rest[0.042]; Baja[0.042]; Baja[0.042]; Mexican[0.042]; Gulf[0.042]; Gulf[0.042]; Gulf[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Quintana Roo\u001b[39m ==> ENTITY: \u001b[32mQuintana Roo\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.046]; Mexico[0.046]; state[0.044]; state[0.044]; coast[0.043]; coast[0.043]; peninsula[0.043]; Gulf[0.042]; Gulf[0.042]; City[0.042]; Baja[0.042]; Tamaulipas[0.041]; northeast[0.041]; northeast[0.041]; Caribbean[0.041]; miles[0.041]; miles[0.041]; kilometers[0.041]; kilometers[0.041]; north[0.040]; feet[0.040]; feet[0.040]; Sonora[0.039]; California[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tamaulipas\u001b[39m ==> ENTITY: \u001b[32mTamaulipas\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.233:0.233[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.046]; Mexico[0.046]; Mexico[0.046]; coast[0.042]; coast[0.042]; coast[0.042]; coast[0.042]; southern[0.042]; state[0.042]; state[0.042]; northeast[0.042]; northeast[0.042]; northeast[0.042]; Sinaloa[0.041]; north[0.040]; Gulf[0.040]; Gulf[0.040]; Gulf[0.040]; kilometers[0.040]; kilometers[0.040]; miles[0.040]; miles[0.040]; Baja[0.040]; Baja[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baja California\u001b[39m ==> ENTITY: \u001b[32mBaja California\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.232:0.232[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncoast[0.047]; coast[0.047]; coast[0.047]; California[0.044]; Baja[0.043]; region[0.042]; state[0.042]; Mexico[0.042]; Mexico[0.042]; Mexico[0.042]; peninsula[0.041]; southern[0.041]; north[0.041]; miles[0.041]; waters[0.041]; Gulf[0.040]; Gulf[0.040]; Gulf[0.040]; kilometers[0.040]; kilometers[0.040]; area[0.040]; northeast[0.040]; northeast[0.040]; Tamaulipas[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Caribbean ==> ENTITY: \u001b[32mCaribbean\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.178:0.178[0]; log p(e|m)= -0.095:-0.095[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncoast[0.045]; coast[0.045]; coast[0.045]; Mexico[0.044]; Mexico[0.044]; Mexico[0.044]; region[0.042]; waters[0.042]; Gulf[0.042]; Gulf[0.042]; Gulf[0.042]; northeast[0.041]; Mexican[0.041]; southern[0.041]; peninsula[0.041]; north[0.041]; Tampico[0.040]; area[0.040]; Pacific[0.040]; Pacific[0.040]; waves[0.040]; miles[0.040]; ship[0.039]; Cold[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gulf ==> ENTITY: \u001b[32mGulf of Mexico\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.219:0.219[0]; log p(e|m)= -1.570:-1.570[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncoast[0.048]; coast[0.048]; coast[0.048]; coast[0.048]; Gulf[0.046]; California[0.045]; California[0.045]; feet[0.044]; feet[0.044]; northeast[0.043]; northeast[0.043]; northeast[0.043]; north[0.042]; Mexico[0.042]; Mexico[0.042]; southern[0.042]; miles[0.042]; miles[0.042]; peninsula[0.041]; Caribbean[0.041]; Caribbean[0.041]; Baja[0.041]; Baja[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sinaloa\u001b[39m ==> ENTITY: \u001b[32mSinaloa\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.044]; Mexico[0.044]; Mexico[0.044]; southern[0.043]; region[0.042]; Gulf[0.042]; Gulf[0.042]; Gulf[0.042]; Tamaulipas[0.042]; waters[0.042]; coast[0.041]; coast[0.041]; coast[0.041]; Tampico[0.041]; state[0.041]; California[0.041]; California[0.041]; Baja[0.041]; Baja[0.041]; north[0.040]; conditions[0.040]; northeast[0.040]; northeast[0.040]; adjacent[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexican ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.397:-0.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.053]; Mexico[0.053]; Caribbean[0.047]; region[0.046]; Tampico[0.045]; area[0.045]; restrictions[0.043]; local[0.043]; Pacific[0.042]; Pacific[0.042]; Coast[0.042]; Coast[0.042]; Gulf[0.042]; deep[0.042]; said[0.042]; said[0.042]; place[0.042]; authorities[0.042]; City[0.041]; major[0.041]; ship[0.041]; ports[0.041]; conditions[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tampico ==> ENTITY: \u001b[32mTampico\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.180:0.180[0]; log p(e|m)= -0.218:-0.218[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexican[0.047]; Mexico[0.046]; Mexico[0.046]; Sinaloa[0.045]; place[0.043]; coast[0.043]; coast[0.043]; area[0.041]; port[0.041]; port[0.041]; port[0.041]; port[0.041]; Gulf[0.041]; southern[0.041]; region[0.041]; fishing[0.040]; local[0.040]; miles[0.040]; waters[0.040]; Baja[0.040]; northeast[0.040]; ship[0.039]; California[0.039]; Caribbean[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Caribbean ==> ENTITY: \u001b[32mCaribbean\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.095:-0.095[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncoast[0.046]; coast[0.046]; coast[0.046]; Mexico[0.045]; Mexico[0.045]; Gulf[0.043]; Gulf[0.043]; Tropical[0.043]; northeast[0.042]; northeast[0.042]; northeast[0.042]; peninsula[0.042]; north[0.041]; waves[0.041]; waves[0.041]; miles[0.040]; miles[0.040]; Cold[0.040]; Baja[0.040]; kilometers[0.039]; kilometers[0.039]; Tamaulipas[0.038]; feet[0.038]; feet[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baja California\u001b[39m ==> ENTITY: \u001b[32mBaja California\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.236:0.236[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncoast[0.046]; coast[0.046]; coast[0.046]; coast[0.046]; California[0.043]; Baja[0.042]; region[0.041]; state[0.041]; state[0.041]; Mexico[0.041]; Mexico[0.041]; Mexico[0.041]; peninsula[0.041]; southern[0.041]; north[0.041]; miles[0.040]; miles[0.040]; waters[0.040]; Gulf[0.040]; Gulf[0.040]; Gulf[0.040]; kilometers[0.040]; kilometers[0.040]; area[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPacific Coast\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPacific Ocean\u001b[39m <---> \u001b[32mPacific coast\u001b[39m\t\nSCORES: global= 0.261:0.260[\u001b[31m0.001\u001b[39m]; local()= 0.206:0.200[\u001b[31m0.007\u001b[39m]; log p(e|m)= -3.540:-0.377[\u001b[32m3.164\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPacific[0.046]; coast[0.046]; coast[0.046]; coast[0.046]; Coast[0.046]; Caribbean[0.044]; waters[0.044]; region[0.044]; Mexico[0.043]; Mexico[0.043]; Mexico[0.043]; peninsula[0.043]; area[0.043]; north[0.043]; southern[0.043]; rest[0.042]; rest[0.042]; Baja[0.042]; Baja[0.042]; Mexican[0.042]; Gulf[0.042]; Gulf[0.042]; Gulf[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexico City ==> ENTITY: \u001b[32mMexico City\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.028:-0.028[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.047]; state[0.044]; state[0.044]; northeast[0.043]; northeast[0.043]; north[0.043]; kilometers[0.042]; miles[0.041]; feet[0.041]; feet[0.041]; Gulf[0.041]; Gulf[0.041]; coast[0.041]; coast[0.041]; high[0.041]; Tamaulipas[0.041]; peninsula[0.041]; Quintana[0.041]; hour[0.040]; Baja[0.040]; Sonora[0.040]; Caribbean[0.040]; California[0.040]; Chris[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = GMT\u001b[39m ==> ENTITY: \u001b[32mGreenwich Mean Time\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.087:0.087[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nupdated[0.044]; Transportation[0.043]; Caribbean[0.043]; region[0.043]; restrictions[0.043]; area[0.042]; Gulf[0.042]; Tampico[0.042]; Mexico[0.042]; Mexico[0.042]; Pacific[0.042]; Pacific[0.042]; Communications[0.042]; Dec[0.041]; major[0.041]; southern[0.041]; update[0.041]; update[0.041]; daily[0.041]; daily[0.041]; Coast[0.041]; Coast[0.041]; said[0.041]; said[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sonora ==> ENTITY: \u001b[32mSonora\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.202:0.202[0]; log p(e|m)= -0.108:-0.108[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.047]; Mexico[0.047]; Mexico[0.047]; southern[0.046]; north[0.045]; region[0.044]; Baja[0.043]; Baja[0.043]; Tropical[0.043]; coast[0.043]; coast[0.043]; coast[0.043]; coast[0.043]; peninsula[0.043]; miles[0.043]; miles[0.043]; City[0.042]; northeast[0.042]; northeast[0.042]; northeast[0.042]; California[0.042]; California[0.042]; waters[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gulf of Mexico\u001b[39m ==> ENTITY: \u001b[32mGulf of Mexico\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncoast[0.047]; coast[0.047]; coast[0.047]; Gulf[0.045]; waters[0.043]; Mexico[0.042]; Mexico[0.042]; region[0.042]; southern[0.041]; conditions[0.041]; peninsula[0.041]; miles[0.041]; area[0.041]; north[0.041]; waves[0.041]; Caribbean[0.040]; deep[0.040]; ship[0.040]; Coast[0.040]; Coast[0.040]; northeast[0.040]; temperatures[0.040]; fishing[0.039]; Tampico[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gulf of Mexico\u001b[39m ==> ENTITY: \u001b[32mGulf of Mexico\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.225:0.225[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncoast[0.047]; coast[0.047]; coast[0.047]; coast[0.047]; Gulf[0.045]; Gulf[0.045]; Mexico[0.042]; Mexico[0.042]; southern[0.041]; conditions[0.040]; peninsula[0.040]; miles[0.040]; miles[0.040]; north[0.040]; waves[0.040]; waves[0.040]; Caribbean[0.040]; Caribbean[0.040]; Coast[0.039]; Coast[0.039]; northeast[0.039]; northeast[0.039]; northeast[0.039]; temperatures[0.039]; \t\n\r [=====================================>........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 3s839ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3736/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1304testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ankara\u001b[39m ==> ENTITY: \u001b[32mAnkara\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.048]; Turkey[0.048]; Kurdish[0.045]; Kurdistan[0.044]; Syria[0.044]; Friday[0.043]; military[0.042]; Syrian[0.042]; weeks[0.041]; forces[0.041]; border[0.040]; border[0.040]; border[0.040]; mines[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; security[0.040]; parliament[0.040]; parliamentarians[0.040]; landmines[0.040]; commission[0.040]; commission[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syria ==> ENTITY: \u001b[32mSyria\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.202:0.202[0]; log p(e|m)= -0.261:-0.261[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.049]; Turkey[0.049]; Turkey[0.049]; Iraq[0.045]; Kurdistan[0.045]; Damascus[0.044]; Kurdish[0.043]; northern[0.043]; forces[0.041]; forces[0.041]; forces[0.041]; border[0.041]; security[0.040]; security[0.040]; bases[0.039]; guerrillas[0.039]; conflict[0.039]; rule[0.039]; weeks[0.039]; southeast[0.039]; parliament[0.038]; said[0.038]; said[0.038]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = PKK\u001b[39m ==> ENTITY: \u001b[32mKurdistan Workers' Party\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.046]; Turkey[0.046]; Turkey[0.046]; Kurdish[0.045]; Kurdistan[0.044]; Syria[0.043]; forces[0.043]; forces[0.043]; forces[0.043]; bases[0.042]; Iraq[0.042]; conflict[0.041]; northern[0.041]; guerrillas[0.041]; border[0.040]; security[0.040]; security[0.040]; fighting[0.040]; said[0.039]; said[0.039]; said[0.039]; Damascus[0.039]; Friday[0.039]; parliament[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kurdistan Workers Party\u001b[39m ==> ENTITY: \u001b[32mKurdistan Workers' Party\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.045]; Turkey[0.045]; Turkey[0.045]; Turkey[0.045]; Kurdish[0.044]; Kurdish[0.044]; Syria[0.044]; Syria[0.044]; Syrian[0.043]; military[0.042]; Iraq[0.042]; Ankara[0.041]; Damascus[0.041]; guerrillas[0.040]; parliament[0.040]; northern[0.040]; forces[0.040]; forces[0.040]; forces[0.040]; said[0.039]; said[0.039]; said[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mKurdish\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mKurdish people\u001b[39m <---> \u001b[32mKurdistan\u001b[39m\t\nSCORES: global= 0.253:0.243[\u001b[31m0.009\u001b[39m]; local()= 0.167:0.197[\u001b[32m0.030\u001b[39m]; log p(e|m)= -0.713:-2.797[\u001b[31m2.084\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.048]; Iraq[0.046]; Syria[0.045]; Turkey[0.045]; Turkey[0.045]; Turkey[0.045]; northern[0.045]; forces[0.042]; forces[0.042]; forces[0.042]; border[0.041]; spokesman[0.040]; Damascus[0.040]; guerrillas[0.040]; southeast[0.040]; people[0.040]; said[0.040]; said[0.040]; said[0.040]; bases[0.040]; security[0.039]; security[0.039]; denies[0.039]; conflict[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syrian ==> ENTITY: \u001b[32mSyria\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.188:0.188[0]; log p(e|m)= -0.609:-0.609[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.051]; Turkey[0.047]; Turkey[0.047]; military[0.046]; Kurdistan[0.046]; Kurdish[0.044]; border[0.043]; border[0.043]; border[0.043]; forces[0.042]; security[0.041]; rebels[0.040]; map[0.040]; Ankara[0.040]; say[0.039]; Reuters[0.039]; seen[0.039]; minefields[0.039]; minefields[0.039]; said[0.038]; said[0.038]; said[0.038]; said[0.038]; parliament[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syria ==> ENTITY: \u001b[32mSyria\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.189:0.189[0]; log p(e|m)= -0.261:-0.261[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.050]; Turkey[0.050]; Syrian[0.047]; military[0.045]; Ankara[0.045]; Kurdistan[0.045]; Kurdish[0.044]; forces[0.042]; border[0.042]; border[0.042]; border[0.042]; border[0.042]; security[0.041]; guerrillas[0.039]; parliamentarians[0.039]; map[0.039]; weeks[0.039]; parliament[0.039]; said[0.039]; said[0.039]; said[0.039]; said[0.039]; efforts[0.038]; Officials[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mDamascus\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mDamascus\u001b[39m <---> \u001b[32mSyria\u001b[39m\t\nSCORES: global= 0.268:0.266[\u001b[31m0.002\u001b[39m]; local()= 0.154:0.206[\u001b[32m0.052\u001b[39m]; log p(e|m)= -0.120:-3.170[\u001b[31m3.050\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.050]; Turkey[0.047]; Turkey[0.047]; Turkey[0.047]; Iraq[0.045]; Kurdistan[0.045]; Kurdish[0.044]; northern[0.044]; border[0.042]; areas[0.040]; areas[0.040]; security[0.040]; security[0.040]; forces[0.040]; forces[0.040]; forces[0.040]; southeast[0.039]; mountains[0.039]; conflict[0.039]; people[0.039]; Terrorists[0.038]; said[0.038]; said[0.038]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = PKK\u001b[39m ==> ENTITY: \u001b[32mKurdistan Workers' Party\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.047]; Turkey[0.047]; Turkey[0.047]; Turkey[0.047]; Kurdish[0.046]; Kurdish[0.046]; Kurdistan[0.045]; military[0.044]; Syria[0.044]; Syria[0.044]; forces[0.043]; forces[0.043]; forces[0.043]; bases[0.042]; Iraq[0.042]; northern[0.042]; guerrillas[0.041]; border[0.041]; border[0.041]; border[0.041]; security[0.041]; security[0.041]; fighting[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = PKK\u001b[39m ==> ENTITY: \u001b[32mKurdistan Workers' Party\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.046]; Turkey[0.046]; Turkey[0.046]; Kurdish[0.045]; Kurdistan[0.044]; Syria[0.043]; forces[0.043]; forces[0.043]; forces[0.043]; bases[0.042]; Iraq[0.042]; conflict[0.041]; northern[0.041]; guerrillas[0.041]; border[0.041]; security[0.040]; security[0.040]; fighting[0.040]; said[0.040]; said[0.040]; said[0.040]; Damascus[0.040]; areas[0.039]; areas[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.063:0.063[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nreport[0.043]; report[0.043]; map[0.043]; areas[0.042]; areas[0.042]; Syrian[0.042]; border[0.042]; border[0.042]; border[0.042]; border[0.042]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; mines[0.041]; mines[0.041]; parliament[0.041]; Syria[0.041]; Syria[0.041]; sown[0.041]; available[0.041]; efforts[0.041]; says[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.168:0.168[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.047]; Kurdistan[0.047]; Kurdish[0.045]; Turkey[0.045]; Turkey[0.045]; Turkey[0.045]; Damascus[0.044]; security[0.044]; northern[0.043]; forces[0.043]; forces[0.043]; bases[0.042]; border[0.041]; said[0.041]; conflict[0.039]; areas[0.039]; areas[0.039]; fighting[0.039]; guerrillas[0.039]; rebels[0.039]; southeast[0.038]; obstacle[0.038]; people[0.038]; sponsors[0.037]; \t\n\r [=====================================>........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 3s763ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3749/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1324testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mJohn Lewis\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mJohn Lewis (department store)\u001b[39m <---> \u001b[32mJohn Lewis Partnership\u001b[39m\t\nSCORES: global= 0.259:0.255[\u001b[31m0.004\u001b[39m]; local()= 0.188:0.180[\u001b[31m0.008\u001b[39m]; log p(e|m)= -1.754:-2.526[\u001b[31m0.771\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLewis[0.047]; John[0.047]; store[0.045]; store[0.045]; Bennett[0.044]; London[0.043]; London[0.043]; Partnership[0.042]; supermarket[0.042]; chain[0.041]; said[0.041]; week[0.041]; week[0.041]; week[0.041]; week[0.041]; week[0.041]; November[0.040]; November[0.040]; Waitrose[0.040]; department[0.040]; sales[0.040]; sales[0.040]; sales[0.040]; sales[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Waitrose\u001b[39m ==> ENTITY: \u001b[32mWaitrose\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsupermarket[0.051]; store[0.050]; store[0.050]; chain[0.046]; sales[0.043]; sales[0.043]; sales[0.043]; sales[0.043]; November[0.043]; November[0.043]; department[0.043]; Partnership[0.043]; London[0.042]; London[0.042]; week[0.042]; week[0.042]; week[0.042]; week[0.042]; week[0.042]; Total[0.042]; Newsroom[0.041]; John[0.041]; John[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.081:0.081[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.050]; John[0.043]; John[0.043]; week[0.042]; week[0.042]; week[0.042]; week[0.042]; week[0.042]; store[0.042]; store[0.042]; earlier[0.042]; sales[0.041]; sales[0.041]; sales[0.041]; sales[0.041]; November[0.041]; November[0.041]; Partnership[0.040]; Bennett[0.040]; chain[0.040]; rose[0.040]; rose[0.040]; said[0.040]; weeks[0.040]; \t\n\r [=====================================>........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 3s748ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3752/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1331testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australian ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.491:-0.491[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralian[0.047]; Australian[0.047]; Territory[0.045]; Canberra[0.045]; John[0.043]; John[0.043]; John[0.043]; today[0.043]; Mr[0.042]; lower[0.042]; possible[0.041]; statement[0.041]; Bob[0.041]; said[0.040]; said[0.040]; Friday[0.040]; place[0.040]; seat[0.040]; House[0.039]; House[0.039]; Fraser[0.039]; announced[0.039]; considering[0.039]; parliamentarian[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Halverson\u001b[39m ==> ENTITY: \u001b[32mBob Halverson\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.046]; John[0.046]; John[0.046]; John[0.046]; resigning[0.044]; Australian[0.044]; Australian[0.044]; Australian[0.044]; resigned[0.044]; resigns[0.044]; Canberra[0.044]; election[0.044]; Territory[0.044]; resign[0.043]; position[0.043]; seat[0.043]; Bob[0.042]; November[0.042]; Representatives[0.042]; Representatives[0.042]; speaker[0.041]; senior[0.041]; parliamentarian[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paul Keating\u001b[39m ==> ENTITY: \u001b[32mPaul Keating\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; Labor[0.046]; Canberra[0.045]; policy[0.043]; economic[0.043]; economic[0.043]; minister[0.042]; played[0.042]; employment[0.042]; private[0.042]; office[0.041]; office[0.041]; parliament[0.041]; Inge[0.041]; announced[0.041]; conference[0.041]; said[0.041]; senior[0.040]; senior[0.040]; secretary[0.040]; campaigner[0.040]; national[0.040]; York[0.040]; position[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.108:0.108[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstatement[0.046]; today[0.043]; national[0.043]; policy[0.043]; announced[0.042]; announced[0.042]; considering[0.042]; election[0.041]; office[0.041]; office[0.041]; office[0.041]; said[0.041]; said[0.041]; said[0.041]; Territory[0.041]; letter[0.041]; economic[0.041]; economic[0.041]; representative[0.041]; position[0.041]; electoral[0.041]; November[0.041]; director[0.041]; articles[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTerritory[0.044]; Australian[0.044]; national[0.044]; played[0.043]; Nations[0.043]; office[0.042]; office[0.042]; office[0.042]; office[0.042]; considering[0.042]; Representatives[0.042]; Representatives[0.042]; today[0.041]; formally[0.041]; representative[0.041]; United[0.041]; Papua[0.041]; resigned[0.041]; Keating[0.040]; parliament[0.040]; position[0.040]; November[0.040]; resigning[0.040]; headquarters[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = John Vance Langmore\u001b[39m ==> ENTITY: \u001b[32mJohn Langmore\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.045]; Canberra[0.045]; Australian[0.045]; Australian[0.045]; Australian[0.045]; John[0.043]; John[0.043]; member[0.042]; United[0.041]; representative[0.041]; Representatives[0.041]; Representatives[0.041]; Inge[0.041]; senior[0.041]; Bob[0.041]; resigned[0.041]; resigning[0.041]; election[0.040]; Territory[0.040]; Langmore[0.040]; Langmore[0.040]; Langmore[0.040]; Mr[0.040]; Nations[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = John Langmore\u001b[39m ==> ENTITY: \u001b[32mJohn Langmore\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.126:0.126[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanberra[0.047]; Australian[0.047]; Australian[0.047]; Australian[0.047]; John[0.045]; John[0.045]; member[0.044]; Representatives[0.044]; Representatives[0.044]; Bob[0.043]; resigned[0.043]; resigning[0.043]; election[0.042]; Territory[0.042]; Langmore[0.042]; Langmore[0.042]; Langmore[0.042]; Mr[0.042]; Fraser[0.042]; parliamentarian[0.042]; November[0.042]; speaker[0.042]; resigns[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.109:0.109[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstatement[0.046]; today[0.043]; national[0.043]; policy[0.043]; announced[0.043]; considering[0.042]; election[0.041]; office[0.041]; office[0.041]; office[0.041]; said[0.041]; said[0.041]; Territory[0.041]; letter[0.041]; economic[0.041]; economic[0.041]; representative[0.041]; position[0.041]; electoral[0.041]; November[0.041]; director[0.041]; articles[0.041]; conference[0.041]; previous[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = United Nations\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.101:0.101[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJohn[0.043]; secretary[0.043]; resigning[0.043]; policy[0.043]; national[0.043]; office[0.042]; office[0.042]; office[0.042]; office[0.042]; economic[0.042]; economic[0.042]; dates[0.041]; advisor[0.041]; representative[0.041]; conference[0.041]; senior[0.041]; senior[0.041]; November[0.041]; Territory[0.041]; development[0.041]; development[0.041]; announced[0.041]; announced[0.041]; Paul[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canberra ==> ENTITY: \u001b[32mCanberra\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.092:-0.092[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayed[0.046]; national[0.044]; Labor[0.044]; Fraser[0.044]; treasurer[0.043]; Bureau[0.043]; Keating[0.042]; Paul[0.042]; secretary[0.042]; assistant[0.042]; senior[0.042]; member[0.042]; representative[0.041]; conference[0.041]; previous[0.041]; office[0.041]; Papua[0.041]; planning[0.040]; policy[0.040]; Nations[0.040]; director[0.040]; advisor[0.040]; York[0.040]; headquarters[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mRTRS\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRadio Televizija Republike Srpske\u001b[39m <---> \u001b[32mRadio and Television of Bosnia and Herzegovina\u001b[39m\t\nSCORES: global= 0.279:0.265[\u001b[31m0.014\u001b[39m]; local()= 0.034:0.069[\u001b[32m0.035\u001b[39m]; log p(e|m)= 0.000:-0.377[\u001b[31m0.377\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmember[0.046]; announced[0.044]; Territory[0.043]; considering[0.042]; Canberra[0.042]; Mr[0.042]; today[0.042]; received[0.042]; place[0.042]; office[0.042]; office[0.042]; Halverson[0.041]; Halverson[0.041]; Halverson[0.041]; Friday[0.041]; resigning[0.041]; division[0.041]; resigned[0.041]; formally[0.041]; formally[0.041]; said[0.041]; said[0.041]; resigns[0.040]; speaker[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canberra ==> ENTITY: \u001b[32mCanberra\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.092:-0.092[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralian[0.049]; Australian[0.049]; Australian[0.049]; seat[0.047]; Territory[0.046]; election[0.044]; today[0.044]; Fraser[0.043]; electoral[0.042]; November[0.042]; division[0.042]; John[0.042]; John[0.042]; John[0.042]; formally[0.042]; formally[0.042]; Mr[0.042]; considering[0.041]; Representatives[0.041]; Representatives[0.041]; Capital[0.041]; Friday[0.041]; member[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.045:0.045[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRepresentatives[0.043]; Representatives[0.043]; John[0.043]; dates[0.042]; director[0.042]; representative[0.042]; office[0.042]; office[0.042]; office[0.042]; announced[0.042]; announced[0.042]; treasurer[0.041]; Vance[0.041]; considering[0.041]; member[0.041]; member[0.041]; national[0.041]; economic[0.041]; economic[0.041]; resigning[0.041]; Capital[0.041]; position[0.041]; November[0.041]; Paul[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Australian Capital Territory\u001b[39m ==> ENTITY: \u001b[32mAustralian Capital Territory\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.045]; Canberra[0.044]; seat[0.043]; Australian[0.042]; Australian[0.042]; statement[0.042]; today[0.042]; headquarters[0.042]; formally[0.042]; formally[0.042]; member[0.041]; member[0.041]; parliament[0.041]; Labor[0.041]; division[0.041]; Mr[0.041]; Fraser[0.041]; Fraser[0.041]; policy[0.041]; Representatives[0.041]; Representatives[0.041]; economic[0.041]; considering[0.041]; election[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bob Halverson\u001b[39m ==> ENTITY: \u001b[32mBob Halverson\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.044]; John[0.044]; John[0.044]; John[0.044]; resigning[0.043]; Australian[0.043]; Australian[0.043]; Australian[0.043]; resigned[0.043]; resigns[0.042]; Canberra[0.042]; election[0.042]; Territory[0.042]; resign[0.042]; position[0.041]; seat[0.041]; November[0.040]; Representatives[0.040]; Representatives[0.040]; speaker[0.040]; senior[0.040]; parliamentarian[0.040]; formally[0.039]; formally[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Copenhagen ==> ENTITY: \u001b[32mCopenhagen\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.086:0.086[0]; log p(e|m)= -0.119:-0.119[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnational[0.045]; statement[0.044]; economic[0.043]; economic[0.043]; played[0.043]; John[0.043]; today[0.043]; conference[0.042]; Australia[0.042]; Paul[0.041]; policy[0.041]; office[0.041]; office[0.041]; office[0.041]; campaigner[0.041]; November[0.041]; place[0.041]; Canberra[0.041]; Nations[0.041]; Australian[0.041]; headquarters[0.041]; development[0.040]; development[0.040]; division[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = John Langmore\u001b[39m ==> ENTITY: \u001b[32mJohn Langmore\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanberra[0.045]; Australian[0.045]; Australian[0.045]; Australian[0.045]; John[0.043]; John[0.043]; member[0.043]; Representatives[0.042]; Representatives[0.042]; Bob[0.041]; resigned[0.041]; resigning[0.041]; election[0.041]; Territory[0.041]; Langmore[0.040]; Langmore[0.040]; Mr[0.040]; Fraser[0.040]; parliamentarian[0.040]; speaker[0.040]; resigns[0.040]; office[0.040]; office[0.040]; office[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Halverson\u001b[39m ==> ENTITY: \u001b[32mBob Halverson\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.044]; John[0.044]; John[0.044]; John[0.044]; resigning[0.043]; Australian[0.042]; Australian[0.042]; Australian[0.042]; resigned[0.042]; resigns[0.042]; Canberra[0.042]; election[0.042]; Territory[0.042]; resign[0.042]; position[0.041]; seat[0.041]; officer[0.041]; Labor[0.040]; Bob[0.040]; November[0.040]; Representatives[0.040]; Representatives[0.040]; Inge[0.040]; speaker[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Papua New Guinea ==> ENTITY: \u001b[32mPapua New Guinea\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.107:0.107[0]; log p(e|m)= -0.265:-0.265[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.051]; Canberra[0.048]; member[0.044]; national[0.043]; played[0.043]; Nations[0.043]; conference[0.043]; active[0.042]; representative[0.042]; employment[0.041]; senior[0.041]; senior[0.041]; economic[0.041]; economic[0.041]; persistent[0.040]; office[0.040]; previous[0.040]; campaigner[0.040]; minister[0.040]; Keating[0.040]; position[0.039]; development[0.039]; development[0.039]; director[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australian ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.491:-0.491[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralian[0.049]; Australian[0.049]; Territory[0.047]; Canberra[0.047]; John[0.045]; John[0.045]; John[0.045]; today[0.044]; Mr[0.044]; lower[0.043]; possible[0.043]; November[0.042]; statement[0.042]; Bob[0.042]; said[0.042]; said[0.042]; said[0.042]; Friday[0.042]; place[0.042]; seat[0.042]; House[0.041]; House[0.041]; Fraser[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Langmore\u001b[39m ==> ENTITY: \u001b[32mJohn Langmore\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLabor[0.046]; Australia[0.044]; Canberra[0.044]; Australian[0.044]; director[0.043]; member[0.042]; member[0.042]; assistant[0.042]; secretary[0.042]; United[0.041]; treasurer[0.041]; representative[0.041]; Representatives[0.041]; campaigner[0.041]; Inge[0.041]; senior[0.041]; senior[0.041]; interventionist[0.041]; relations[0.041]; minister[0.040]; resigning[0.040]; advisor[0.040]; Paul[0.040]; national[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Langmore\u001b[39m ==> ENTITY: \u001b[32mJohn Langmore\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLabor[0.046]; Australia[0.044]; Canberra[0.044]; Australian[0.044]; Australian[0.044]; John[0.042]; John[0.042]; member[0.042]; member[0.042]; secretary[0.041]; United[0.041]; treasurer[0.041]; representative[0.041]; Representatives[0.041]; Representatives[0.041]; campaigner[0.041]; Inge[0.041]; senior[0.041]; senior[0.041]; interventionist[0.041]; relations[0.040]; minister[0.040]; Bob[0.040]; resigned[0.040]; \t\n\r [=====================================>........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 3s637ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3776/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1273testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fed\u001b[39m ==> ENTITY: \u001b[32mFederal Reserve System\u001b[39m\t\nSCORES: global= 0.215:0.215[0]; local()= 0.008:0.008[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncattle[0.046]; Calves[0.045]; livestock[0.045]; Hogs[0.045]; Friday[0.045]; ago[0.045]; ago[0.045]; Week[0.044]; Week[0.044]; Week[0.044]; Saturday[0.044]; Cattle[0.043]; date[0.042]; date[0.042]; date[0.042]; est[0.042]; est[0.042]; est[0.042]; est[0.042]; est[0.042]; Dec[0.042]; Period[0.042]; Period[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = USDA\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Agriculture\u001b[39m\t\nSCORES: global= 0.235:0.235[0]; local()= 0.071:0.071[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlivestock[0.053]; hog[0.052]; slaughter[0.051]; slaughter[0.051]; act[0.051]; act[0.051]; cattle[0.051]; Calves[0.050]; Hogs[0.050]; Friday[0.050]; daily[0.049]; daily[0.049]; Cattle[0.049]; Week[0.049]; Week[0.049]; Week[0.049]; inspection[0.049]; date[0.048]; date[0.048]; date[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = USDA\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Agriculture\u001b[39m\t\nSCORES: global= 0.234:0.234[0]; local()= 0.037:0.037[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlivestock[0.045]; hog[0.044]; slaughter[0.043]; slaughter[0.043]; act[0.043]; act[0.043]; cattle[0.043]; Calves[0.042]; Hogs[0.042]; Friday[0.042]; daily[0.041]; daily[0.041]; Cattle[0.041]; Week[0.041]; Week[0.041]; Week[0.041]; inspection[0.041]; date[0.041]; date[0.041]; est[0.040]; est[0.040]; est[0.040]; est[0.040]; est[0.040]; \t\n\r [=====================================>........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 3s660ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3779/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1385testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Phoenix ==> ENTITY: \u001b[32mPhoenix Coyotes\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.251:0.251[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.046]; Anaheim[0.046]; Edmonton[0.044]; Buffalo[0.044]; Ottawa[0.043]; Montreal[0.043]; Islanders[0.043]; Phoenix[0.042]; Toronto[0.042]; Toronto[0.042]; Calgary[0.041]; Calgary[0.041]; Vancouver[0.041]; Vancouver[0.041]; Pittsburgh[0.041]; Boston[0.041]; Tampa[0.041]; Louis[0.040]; Dallas[0.040]; Rangers[0.040]; Colorado[0.040]; Colorado[0.040]; Chicago[0.039]; Chicago[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Jersey ==> ENTITY: \u001b[32mNew Jersey Devils\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.235:0.235[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.046]; Anaheim[0.046]; Islanders[0.045]; Montreal[0.043]; Edmonton[0.043]; Ottawa[0.043]; Buffalo[0.043]; Rangers[0.042]; Toronto[0.042]; Toronto[0.042]; Boston[0.042]; Tampa[0.041]; Vancouver[0.041]; Vancouver[0.041]; Pittsburgh[0.041]; Calgary[0.040]; Calgary[0.040]; Division[0.040]; Hartford[0.040]; Colorado[0.040]; Colorado[0.040]; Saturday[0.040]; Phoenix[0.040]; Phoenix[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington Capitals\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.215:0.215[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; Hockey[0.045]; Islanders[0.044]; Ottawa[0.044]; Philadelphia[0.044]; Montreal[0.043]; Buffalo[0.043]; teams[0.042]; League[0.042]; Toronto[0.042]; Pittsburgh[0.042]; Boston[0.041]; Tampa[0.041]; Rangers[0.041]; played[0.041]; Vancouver[0.041]; Dallas[0.040]; won[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Chicago[0.040]; Phoenix[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colorado ==> ENTITY: \u001b[32mColorado Avalanche\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.245:0.245[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nColorado[0.052]; Anaheim[0.045]; Anaheim[0.045]; Edmonton[0.044]; Buffalo[0.043]; Montreal[0.043]; Ottawa[0.043]; Toronto[0.042]; Vancouver[0.042]; Vancouver[0.042]; Islanders[0.042]; Calgary[0.041]; Calgary[0.041]; Pittsburgh[0.041]; Boston[0.041]; Phoenix[0.040]; Phoenix[0.040]; Tampa[0.040]; Rangers[0.040]; Washington[0.039]; Chicago[0.039]; Jersey[0.038]; Saturday[0.038]; Division[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Los Angeles ==> ENTITY: \u001b[32mLos Angeles Kings\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.234:0.234[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.047]; Anaheim[0.047]; Edmonton[0.043]; Islanders[0.043]; Montreal[0.042]; Buffalo[0.042]; Bay[0.042]; Ottawa[0.042]; Vancouver[0.042]; Vancouver[0.042]; Toronto[0.042]; Phoenix[0.042]; Phoenix[0.042]; Boston[0.041]; Pittsburgh[0.041]; Calgary[0.041]; Calgary[0.041]; Tampa[0.040]; Colorado[0.040]; Colorado[0.040]; Rangers[0.040]; Washington[0.040]; Chicago[0.039]; Hartford[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Montreal ==> ENTITY: \u001b[32mMontreal Canadiens\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.201:0.201[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.045]; Hockey[0.045]; games[0.045]; Ottawa[0.044]; League[0.043]; Islanders[0.043]; Philadelphia[0.042]; Boston[0.042]; teams[0.042]; Buffalo[0.042]; Detroit[0.042]; Pittsburgh[0.042]; Rangers[0.042]; Tampa[0.041]; won[0.041]; played[0.041]; Friday[0.040]; Friday[0.040]; Games[0.040]; Ice[0.040]; York[0.040]; Washington[0.040]; Florida[0.040]; lost[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dallas ==> ENTITY: \u001b[32mDallas Stars\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.209:0.209[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.046]; Washington[0.044]; Conference[0.044]; Edmonton[0.044]; Ottawa[0.043]; Islanders[0.043]; Toronto[0.042]; Vancouver[0.042]; Philadelphia[0.042]; Rangers[0.042]; Pittsburgh[0.042]; Boston[0.041]; Calgary[0.041]; Detroit[0.041]; Tampa[0.041]; Phoenix[0.041]; Phoenix[0.041]; Central[0.041]; Colorado[0.040]; Chicago[0.040]; Florida[0.040]; Saturday[0.040]; Western[0.039]; December[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vancouver ==> ENTITY: \u001b[32mVancouver Canucks\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.252:0.252[0]; log p(e|m)= -3.912:-3.912[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVancouver[0.046]; Anaheim[0.045]; Anaheim[0.045]; Edmonton[0.044]; Islanders[0.043]; Islanders[0.043]; Ottawa[0.043]; Toronto[0.043]; Toronto[0.043]; Montreal[0.042]; Calgary[0.042]; Calgary[0.042]; Buffalo[0.041]; Rangers[0.041]; Rangers[0.041]; Bay[0.040]; Bay[0.040]; Detroit[0.040]; Boston[0.039]; Washington[0.039]; Washington[0.039]; Pittsburgh[0.039]; Colorado[0.039]; Colorado[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Florida ==> ENTITY: \u001b[32mFlorida Panthers\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.204:0.204[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; Hockey[0.044]; Hockey[0.044]; Buffalo[0.043]; Ottawa[0.043]; Tampa[0.042]; Islanders[0.042]; teams[0.042]; Conference[0.042]; Conference[0.042]; Philadelphia[0.042]; Montreal[0.042]; League[0.041]; Pittsburgh[0.041]; Toronto[0.041]; won[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; played[0.040]; Phoenix[0.040]; Dallas[0.040]; lost[0.040]; Boston[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Calgary ==> ENTITY: \u001b[32mCalgary Flames\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.239:0.239[0]; log p(e|m)= -3.147:-3.147[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.046]; Anaheim[0.046]; Edmonton[0.045]; Ottawa[0.043]; Calgary[0.043]; Islanders[0.043]; Montreal[0.043]; Vancouver[0.043]; Vancouver[0.043]; Buffalo[0.042]; Toronto[0.042]; Toronto[0.042]; Phoenix[0.041]; Phoenix[0.041]; Pts[0.041]; Colorado[0.040]; Colorado[0.040]; Pittsburgh[0.040]; Tampa[0.040]; Boston[0.040]; Rangers[0.040]; Division[0.039]; Bay[0.039]; Louis[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Boston ==> ENTITY: \u001b[32mBoston Bruins\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.237:0.237[0]; log p(e|m)= -4.343:-4.343[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.045]; Anaheim[0.045]; Islanders[0.044]; Buffalo[0.044]; Ottawa[0.044]; Montreal[0.043]; Edmonton[0.043]; Pittsburgh[0.042]; Rangers[0.042]; Toronto[0.042]; Toronto[0.042]; Vancouver[0.042]; Vancouver[0.042]; Bay[0.041]; Calgary[0.041]; Calgary[0.041]; Washington[0.040]; Tampa[0.040]; Louis[0.040]; Colorado[0.040]; Colorado[0.040]; Chicago[0.039]; Chicago[0.039]; Hartford[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tampa Bay ==> ENTITY: \u001b[32mTampa Bay Lightning\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.253:0.253[0]; log p(e|m)= -2.120:-2.120[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.046]; Anaheim[0.046]; Edmonton[0.044]; Ottawa[0.044]; Islanders[0.043]; Montreal[0.043]; Buffalo[0.043]; Toronto[0.042]; Vancouver[0.042]; Vancouver[0.042]; Pittsburgh[0.042]; Boston[0.042]; Calgary[0.041]; Calgary[0.041]; Rangers[0.041]; Jersey[0.041]; Division[0.040]; Phoenix[0.040]; Phoenix[0.040]; Washington[0.040]; Chicago[0.040]; Colorado[0.039]; Colorado[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Jose ==> ENTITY: \u001b[32mSan Jose Sharks\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.241:0.241[0]; log p(e|m)= -3.124:-3.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.047]; Anaheim[0.047]; Ottawa[0.044]; Edmonton[0.043]; Buffalo[0.043]; Islanders[0.043]; Montreal[0.042]; Vancouver[0.042]; Vancouver[0.042]; Toronto[0.042]; Calgary[0.041]; Calgary[0.041]; Colorado[0.041]; Colorado[0.041]; Boston[0.041]; Jose[0.041]; Pittsburgh[0.041]; Phoenix[0.040]; San[0.040]; Tampa[0.040]; Rangers[0.039]; Saturday[0.039]; Division[0.039]; Chicago[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = St Louis ==> ENTITY: \u001b[32mSt. Louis Blues\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.214:0.214[0]; log p(e|m)= -2.564:-2.564[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.046]; Edmonton[0.044]; Ottawa[0.044]; Buffalo[0.044]; Islanders[0.043]; Toronto[0.043]; Philadelphia[0.043]; Boston[0.042]; Pittsburgh[0.042]; Vancouver[0.042]; Calgary[0.042]; Calgary[0.042]; Colorado[0.041]; Detroit[0.041]; Phoenix[0.041]; Phoenix[0.041]; Washington[0.041]; Rangers[0.041]; Tampa[0.041]; Chicago[0.040]; Dallas[0.040]; Pts[0.039]; Pts[0.039]; Pts[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pittsburgh ==> ENTITY: \u001b[32mPittsburgh Penguins\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.199:0.199[0]; log p(e|m)= -3.689:-3.689[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.046]; Hockey[0.046]; games[0.045]; League[0.043]; Ottawa[0.043]; Buffalo[0.043]; Philadelphia[0.043]; Montreal[0.042]; Islanders[0.042]; teams[0.042]; Boston[0.041]; Florida[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Ice[0.041]; played[0.041]; Washington[0.040]; Tampa[0.040]; Detroit[0.040]; won[0.040]; Rangers[0.040]; Dallas[0.040]; National[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Edmonton ==> ENTITY: \u001b[32mEdmonton Oilers\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.229:0.229[0]; log p(e|m)= -3.270:-3.270[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOttawa[0.045]; Anaheim[0.045]; Anaheim[0.045]; Calgary[0.044]; Calgary[0.044]; Buffalo[0.043]; Vancouver[0.043]; Vancouver[0.043]; Toronto[0.042]; Toronto[0.042]; Montreal[0.042]; Islanders[0.042]; Islanders[0.042]; Pittsburgh[0.041]; Rangers[0.040]; Rangers[0.040]; Tampa[0.040]; Tampa[0.040]; Louis[0.040]; Phoenix[0.040]; Phoenix[0.040]; Detroit[0.039]; Boston[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Buffalo ==> ENTITY: \u001b[32mBuffalo Sabres\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.227:0.227[0]; log p(e|m)= -3.540:-3.540[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.047]; Anaheim[0.047]; Islanders[0.045]; Edmonton[0.045]; Ottawa[0.045]; Montreal[0.045]; Toronto[0.044]; Toronto[0.044]; Vancouver[0.044]; Vancouver[0.044]; Pittsburgh[0.044]; Tampa[0.043]; Boston[0.043]; Calgary[0.043]; Calgary[0.043]; Rangers[0.042]; Washington[0.042]; Colorado[0.042]; Colorado[0.042]; Phoenix[0.042]; Phoenix[0.042]; Hartford[0.041]; Saturday[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hartford ==> ENTITY: \u001b[32mHartford Whalers\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.238:0.238[0]; log p(e|m)= -3.689:-3.689[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIslanders[0.045]; Buffalo[0.044]; Boston[0.044]; Anaheim[0.044]; Anaheim[0.044]; Montreal[0.043]; Ottawa[0.043]; Edmonton[0.043]; Rangers[0.043]; Toronto[0.043]; Toronto[0.043]; Pittsburgh[0.042]; Tampa[0.041]; Vancouver[0.041]; Vancouver[0.041]; Washington[0.040]; Calgary[0.040]; Calgary[0.040]; Phoenix[0.040]; Phoenix[0.040]; Jersey[0.040]; Colorado[0.039]; Colorado[0.039]; Division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NY RANGERS\u001b[39m ==> ENTITY: \u001b[32mNew York Rangers\u001b[39m\t\nSCORES: global= 0.318:0.318[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIslanders[0.047]; Montreal[0.045]; Ottawa[0.044]; Edmonton[0.044]; Buffalo[0.044]; Boston[0.044]; Toronto[0.043]; Philadelphia[0.043]; Pittsburgh[0.042]; Vancouver[0.042]; won[0.042]; Detroit[0.041]; Louis[0.041]; Tampa[0.041]; Phoenix[0.040]; Colorado[0.040]; Pts[0.040]; Pts[0.040]; Pts[0.040]; Pts[0.040]; Hartford[0.040]; Chicago[0.040]; Jersey[0.040]; Eastern[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ottawa ==> ENTITY: \u001b[32mOttawa Senators\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.224:0.224[0]; log p(e|m)= -3.689:-3.689[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.047]; Anaheim[0.047]; Buffalo[0.046]; Islanders[0.046]; Edmonton[0.046]; Montreal[0.045]; Rangers[0.045]; Pittsburgh[0.045]; Toronto[0.044]; Vancouver[0.044]; Vancouver[0.044]; Tampa[0.043]; Boston[0.043]; Calgary[0.043]; Calgary[0.043]; Phoenix[0.042]; Washington[0.042]; Hartford[0.041]; Chicago[0.041]; Bay[0.041]; Jersey[0.040]; Division[0.040]; Saturday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vancouver ==> ENTITY: \u001b[32mVancouver Canucks\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.241:0.241[0]; log p(e|m)= -3.912:-3.912[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVancouver[0.047]; Anaheim[0.046]; Anaheim[0.046]; Edmonton[0.044]; Islanders[0.044]; Ottawa[0.044]; Toronto[0.044]; Montreal[0.043]; Calgary[0.043]; Calgary[0.043]; Buffalo[0.042]; Rangers[0.042]; Bay[0.041]; Boston[0.040]; Washington[0.040]; Pittsburgh[0.040]; Colorado[0.040]; Colorado[0.040]; Tampa[0.040]; Chicago[0.039]; Phoenix[0.039]; Saturday[0.039]; Pacific[0.038]; Division[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago Blackhawks\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.244:0.244[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.046]; Ottawa[0.044]; Islanders[0.043]; Islanders[0.043]; Edmonton[0.043]; Montreal[0.043]; Buffalo[0.043]; Chicago[0.042]; Vancouver[0.042]; Toronto[0.042]; Philadelphia[0.041]; Conference[0.041]; Boston[0.041]; Calgary[0.041]; Calgary[0.041]; Rangers[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Detroit[0.040]; Phoenix[0.040]; Phoenix[0.040]; Louis[0.040]; Tampa[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Los Angeles ==> ENTITY: \u001b[32mLos Angeles Kings\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.238:0.238[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.046]; Anaheim[0.046]; Edmonton[0.043]; Islanders[0.043]; Islanders[0.043]; Conference[0.043]; Montreal[0.042]; Buffalo[0.042]; Bay[0.042]; Bay[0.042]; Ottawa[0.042]; Vancouver[0.041]; Vancouver[0.041]; Toronto[0.041]; Toronto[0.041]; Boston[0.041]; Pittsburgh[0.041]; Phoenix[0.040]; Phoenix[0.040]; Calgary[0.040]; Calgary[0.040]; Detroit[0.040]; Tampa[0.040]; Tampa[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Philadelphia ==> ENTITY: \u001b[32mPhiladelphia Flyers\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.219:0.219[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; Hockey[0.045]; Hockey[0.045]; teams[0.044]; Montreal[0.043]; Ottawa[0.043]; Buffalo[0.042]; Islanders[0.042]; Toronto[0.042]; League[0.042]; Boston[0.041]; Pittsburgh[0.041]; played[0.041]; won[0.041]; Washington[0.041]; Conference[0.040]; Conference[0.040]; Rangers[0.040]; Tampa[0.040]; Ice[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington Capitals\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.224:0.224[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.044]; Anaheim[0.044]; Islanders[0.044]; Ottawa[0.044]; Edmonton[0.044]; Montreal[0.043]; Buffalo[0.043]; Toronto[0.042]; Toronto[0.042]; Pittsburgh[0.042]; Boston[0.042]; Tampa[0.042]; Rangers[0.041]; Calgary[0.041]; Calgary[0.041]; Vancouver[0.041]; Vancouver[0.041]; Chicago[0.040]; Chicago[0.040]; Phoenix[0.040]; Phoenix[0.040]; Colorado[0.039]; Colorado[0.039]; Hartford[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NY ISLANDERS\u001b[39m ==> ENTITY: \u001b[32mNew York Islanders\u001b[39m\t\nSCORES: global= 0.319:0.319[0]; local()= 0.244:0.244[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.048]; Anaheim[0.048]; Montreal[0.046]; Buffalo[0.046]; Edmonton[0.045]; Ottawa[0.045]; Boston[0.044]; Toronto[0.044]; Toronto[0.044]; Vancouver[0.044]; Vancouver[0.044]; Pittsburgh[0.043]; Rangers[0.043]; Tampa[0.043]; Colorado[0.042]; Colorado[0.042]; Calgary[0.042]; Calgary[0.042]; Phoenix[0.042]; Phoenix[0.042]; Jersey[0.041]; Pts[0.041]; Saturday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Buffalo ==> ENTITY: \u001b[32mBuffalo Sabres\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.189:0.189[0]; log p(e|m)= -3.540:-3.540[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; teams[0.045]; Hockey[0.044]; Hockey[0.044]; Islanders[0.043]; Ottawa[0.043]; League[0.042]; National[0.042]; Montreal[0.042]; won[0.042]; Division[0.041]; Division[0.041]; Division[0.041]; played[0.041]; Pittsburgh[0.041]; York[0.041]; Philadelphia[0.041]; Central[0.040]; Tampa[0.040]; Boston[0.040]; Rangers[0.040]; Northeast[0.040]; Washington[0.040]; Atlantic[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew York\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew York\u001b[39m <---> \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.240:0.239[\u001b[31m0.001\u001b[39m]; local()= 0.076:0.074[\u001b[31m0.002\u001b[39m]; log p(e|m)= -0.289:-1.790[\u001b[31m1.501\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; League[0.045]; Philadelphia[0.044]; Boston[0.044]; teams[0.043]; Conference[0.043]; won[0.043]; played[0.043]; Friday[0.043]; Friday[0.043]; Hockey[0.041]; Hockey[0.041]; National[0.041]; Buffalo[0.041]; lost[0.041]; Jersey[0.041]; Hartford[0.040]; Pittsburgh[0.040]; Division[0.040]; Division[0.040]; Games[0.040]; Eastern[0.039]; Atlantic[0.039]; points[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Toronto ==> ENTITY: \u001b[32mToronto Maple Leafs\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.220:0.220[0]; log p(e|m)= -4.075:-4.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nToronto[0.046]; Anaheim[0.045]; Anaheim[0.045]; Edmonton[0.044]; Montreal[0.044]; Buffalo[0.043]; Islanders[0.042]; Islanders[0.042]; Vancouver[0.042]; Calgary[0.042]; Calgary[0.042]; Louis[0.042]; Rangers[0.041]; Rangers[0.041]; Philadelphia[0.041]; Boston[0.041]; Detroit[0.041]; Pittsburgh[0.040]; Tampa[0.040]; Conference[0.040]; Phoenix[0.039]; Phoenix[0.039]; December[0.039]; Western[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ottawa ==> ENTITY: \u001b[32mOttawa Senators\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.204:0.204[0]; log p(e|m)= -3.689:-3.689[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.049]; Hockey[0.049]; games[0.049]; teams[0.048]; Buffalo[0.047]; Islanders[0.047]; Montreal[0.046]; League[0.046]; Rangers[0.046]; Pittsburgh[0.046]; won[0.045]; Philadelphia[0.045]; Tampa[0.044]; Boston[0.044]; Central[0.044]; played[0.044]; Detroit[0.044]; Ice[0.043]; Louis[0.043]; Conference[0.043]; Conference[0.043]; Washington[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Jose ==> ENTITY: \u001b[32mSan Jose Sharks\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.247:0.247[0]; log p(e|m)= -3.124:-3.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.047]; Anaheim[0.047]; Ottawa[0.044]; Edmonton[0.043]; Buffalo[0.042]; Islanders[0.042]; Montreal[0.042]; Vancouver[0.042]; Vancouver[0.042]; Toronto[0.042]; Toronto[0.042]; Calgary[0.041]; Calgary[0.041]; Colorado[0.041]; Colorado[0.041]; Boston[0.041]; Jose[0.041]; Pittsburgh[0.040]; Phoenix[0.040]; Phoenix[0.040]; San[0.040]; Tampa[0.040]; Tampa[0.040]; Conference[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago Blackhawks\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.253:0.253[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.046]; Anaheim[0.046]; Ottawa[0.044]; Islanders[0.043]; Edmonton[0.043]; Montreal[0.043]; Buffalo[0.042]; Chicago[0.042]; Vancouver[0.042]; Vancouver[0.042]; Toronto[0.042]; Toronto[0.042]; Boston[0.041]; Calgary[0.041]; Calgary[0.041]; Rangers[0.041]; Pittsburgh[0.041]; Division[0.040]; Phoenix[0.040]; Phoenix[0.040]; Tampa[0.040]; Colorado[0.040]; Colorado[0.040]; Washington[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Montreal ==> ENTITY: \u001b[32mMontreal Canadiens\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.226:0.226[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.046]; Anaheim[0.046]; Ottawa[0.046]; Toronto[0.046]; Toronto[0.046]; Edmonton[0.045]; Islanders[0.045]; Boston[0.044]; Vancouver[0.044]; Vancouver[0.044]; Buffalo[0.044]; Pittsburgh[0.043]; Rangers[0.043]; Calgary[0.043]; Calgary[0.043]; Tampa[0.043]; Washington[0.042]; Saturday[0.041]; Chicago[0.041]; Chicago[0.041]; Phoenix[0.041]; Phoenix[0.041]; Division[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NY ISLANDERS\u001b[39m ==> ENTITY: \u001b[32mNew York Islanders\u001b[39m\t\nSCORES: global= 0.318:0.318[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMontreal[0.045]; Buffalo[0.045]; Edmonton[0.044]; Ottawa[0.044]; Philadelphia[0.043]; Boston[0.043]; Toronto[0.043]; Vancouver[0.043]; Pittsburgh[0.042]; Rangers[0.042]; Tampa[0.042]; Colorado[0.042]; Phoenix[0.041]; Detroit[0.041]; Florida[0.041]; Jersey[0.040]; Louis[0.040]; Pts[0.040]; Pts[0.040]; Pts[0.040]; Pts[0.040]; Chicago[0.040]; Hartford[0.040]; Dallas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Boston ==> ENTITY: \u001b[32mBoston Bruins\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.216:0.216[0]; log p(e|m)= -4.343:-4.343[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; Hockey[0.045]; Hockey[0.045]; York[0.044]; Islanders[0.043]; Buffalo[0.043]; Ottawa[0.043]; Montreal[0.043]; Philadelphia[0.042]; Pittsburgh[0.042]; teams[0.042]; League[0.042]; Rangers[0.041]; Detroit[0.041]; played[0.040]; Northeast[0.040]; Bay[0.040]; Conference[0.040]; Conference[0.040]; won[0.040]; Washington[0.040]; Ice[0.040]; Tampa[0.040]; Western[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Phoenix ==> ENTITY: \u001b[32mPhoenix Coyotes\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.248:0.248[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.045]; Anaheim[0.045]; Edmonton[0.044]; Buffalo[0.044]; Montreal[0.043]; Islanders[0.042]; Islanders[0.042]; Phoenix[0.042]; Toronto[0.041]; Toronto[0.041]; Conference[0.041]; Calgary[0.041]; Calgary[0.041]; Vancouver[0.041]; Philadelphia[0.041]; Pittsburgh[0.041]; Boston[0.041]; Tampa[0.041]; Tampa[0.041]; Louis[0.040]; Dallas[0.040]; Rangers[0.040]; Rangers[0.040]; Detroit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = National Hockey League\u001b[39m ==> ENTITY: \u001b[32mNational Hockey League\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; Hockey[0.047]; teams[0.045]; Ottawa[0.045]; Montreal[0.044]; won[0.043]; Buffalo[0.043]; played[0.043]; Philadelphia[0.042]; Boston[0.042]; Pittsburgh[0.041]; goals[0.041]; goals[0.041]; Ice[0.040]; Jersey[0.040]; Florida[0.040]; Hartford[0.040]; York[0.040]; tied[0.039]; Games[0.039]; points[0.039]; Division[0.039]; Division[0.039]; Conference[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Anaheim ==> ENTITY: \u001b[32mAnaheim Ducks\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.248:0.248[0]; log p(e|m)= -2.189:-2.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.047]; Edmonton[0.044]; Islanders[0.044]; Vancouver[0.043]; Vancouver[0.043]; Montreal[0.042]; Buffalo[0.042]; Ottawa[0.042]; Conference[0.042]; Calgary[0.041]; Calgary[0.041]; Western[0.041]; Phoenix[0.041]; Phoenix[0.041]; Toronto[0.041]; Toronto[0.041]; Pittsburgh[0.041]; Rangers[0.041]; Division[0.040]; Division[0.040]; Boston[0.040]; Colorado[0.040]; Colorado[0.040]; Detroit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Jersey ==> ENTITY: \u001b[32mNew Jersey Devils\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.222:0.222[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; Hockey[0.044]; Islanders[0.044]; League[0.043]; Philadelphia[0.042]; Montreal[0.042]; played[0.042]; Conference[0.042]; Conference[0.042]; Ottawa[0.042]; Buffalo[0.042]; teams[0.042]; York[0.042]; Rangers[0.041]; Toronto[0.041]; Standings[0.041]; Boston[0.041]; won[0.041]; Atlantic[0.041]; Tampa[0.040]; Pittsburgh[0.040]; Northeast[0.040]; Detroit[0.040]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NY RANGERS\u001b[39m ==> ENTITY: \u001b[32mNew York Rangers\u001b[39m\t\nSCORES: global= 0.319:0.319[0]; local()= 0.239:0.239[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIslanders[0.046]; Anaheim[0.046]; Anaheim[0.046]; Montreal[0.044]; Ottawa[0.044]; Edmonton[0.044]; Buffalo[0.043]; Boston[0.043]; Toronto[0.043]; Toronto[0.043]; Pittsburgh[0.042]; Vancouver[0.041]; Vancouver[0.041]; Calgary[0.041]; Calgary[0.041]; Tampa[0.040]; Phoenix[0.040]; Phoenix[0.040]; Saturday[0.040]; Colorado[0.039]; Colorado[0.039]; Pts[0.039]; Hartford[0.039]; Chicago[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Toronto ==> ENTITY: \u001b[32mToronto Maple Leafs\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.220:0.220[0]; log p(e|m)= -4.075:-4.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nToronto[0.046]; Ottawa[0.045]; Anaheim[0.045]; Anaheim[0.045]; Edmonton[0.044]; Montreal[0.044]; Buffalo[0.043]; Islanders[0.043]; Vancouver[0.043]; Vancouver[0.043]; Calgary[0.042]; Calgary[0.042]; Rangers[0.042]; Boston[0.041]; Pittsburgh[0.041]; Tampa[0.040]; Phoenix[0.039]; Phoenix[0.039]; December[0.039]; Washington[0.039]; Colorado[0.039]; Colorado[0.039]; Division[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tampa Bay ==> ENTITY: \u001b[32mTampa Bay Lightning\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.229:0.229[0]; log p(e|m)= -2.120:-2.120[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFlorida[0.045]; Edmonton[0.044]; Ottawa[0.044]; Islanders[0.044]; Montreal[0.044]; Philadelphia[0.043]; Buffalo[0.043]; Toronto[0.043]; Vancouver[0.042]; Pittsburgh[0.042]; Dallas[0.042]; Boston[0.042]; Detroit[0.042]; Rangers[0.041]; Phoenix[0.041]; Washington[0.040]; Chicago[0.040]; Conference[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Louis[0.040]; Colorado[0.040]; Hartford[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pacific ==> ENTITY: \u001b[32mPacific Division (NHL)\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.240:0.240[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.045]; Anaheim[0.045]; Edmonton[0.043]; Vancouver[0.043]; Vancouver[0.043]; Division[0.043]; Division[0.043]; Calgary[0.042]; Calgary[0.042]; Western[0.041]; Phoenix[0.041]; Phoenix[0.041]; Dallas[0.041]; Conference[0.041]; Bay[0.041]; Bay[0.041]; Ottawa[0.041]; Saturday[0.041]; Toronto[0.041]; Toronto[0.041]; Islanders[0.041]; Islanders[0.041]; Buffalo[0.040]; Detroit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlantic ==> ENTITY: \u001b[32mAtlantic Division (NHL)\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.240:0.240[0]; log p(e|m)= -3.381:-3.381[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; teams[0.044]; League[0.044]; Hockey[0.043]; Hockey[0.043]; won[0.043]; Division[0.043]; Division[0.043]; Division[0.043]; Boston[0.042]; Pittsburgh[0.042]; Philadelphia[0.042]; Northeast[0.041]; Conference[0.041]; Conference[0.041]; Ottawa[0.041]; Tampa[0.041]; Detroit[0.040]; Buffalo[0.040]; Islanders[0.040]; Florida[0.040]; Toronto[0.040]; Bay[0.040]; lost[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Calgary ==> ENTITY: \u001b[32mCalgary Flames\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.242:0.242[0]; log p(e|m)= -3.147:-3.147[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.046]; Anaheim[0.046]; Edmonton[0.045]; Ottawa[0.043]; Calgary[0.043]; Islanders[0.043]; Montreal[0.043]; Vancouver[0.042]; Vancouver[0.042]; Buffalo[0.042]; Toronto[0.042]; Toronto[0.042]; Phoenix[0.041]; Phoenix[0.041]; Pts[0.040]; Pts[0.040]; Colorado[0.040]; Colorado[0.040]; Pittsburgh[0.040]; Dallas[0.040]; Tampa[0.040]; Boston[0.040]; Rangers[0.040]; Detroit[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NHL ==> ENTITY: \u001b[32mNational Hockey League\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.190:0.190[0]; log p(e|m)= -0.036:-0.036[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.048]; Hockey[0.048]; games[0.048]; teams[0.046]; Ottawa[0.046]; won[0.045]; Montreal[0.045]; League[0.044]; Buffalo[0.044]; Division[0.043]; Division[0.043]; Pittsburgh[0.043]; played[0.043]; Boston[0.042]; Conference[0.042]; Standings[0.042]; Standings[0.042]; tied[0.041]; National[0.041]; Hartford[0.041]; Ice[0.041]; Atlantic[0.041]; lost[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mHartford\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHartford Whalers\u001b[39m <---> \u001b[32mHartford Wolf Pack\u001b[39m\t\nSCORES: global= 0.269:0.255[\u001b[31m0.014\u001b[39m]; local()= 0.219:0.208[\u001b[31m0.011\u001b[39m]; log p(e|m)= -3.689:-3.863[\u001b[31m0.174\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.045]; Hockey[0.045]; games[0.045]; Islanders[0.044]; League[0.044]; Buffalo[0.043]; Boston[0.043]; Montreal[0.043]; Ottawa[0.043]; teams[0.043]; Rangers[0.043]; Philadelphia[0.042]; Pittsburgh[0.041]; Tampa[0.041]; played[0.040]; Conference[0.040]; Ice[0.040]; Washington[0.040]; won[0.040]; Division[0.040]; Division[0.040]; Florida[0.039]; Jersey[0.039]; Northeast[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Anaheim ==> ENTITY: \u001b[32mAnaheim Ducks\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.240:0.240[0]; log p(e|m)= -2.189:-2.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.048]; Edmonton[0.044]; Islanders[0.044]; Vancouver[0.043]; Vancouver[0.043]; Montreal[0.043]; Buffalo[0.042]; Ottawa[0.042]; Calgary[0.042]; Calgary[0.042]; Phoenix[0.041]; Phoenix[0.041]; Toronto[0.041]; Toronto[0.041]; Pittsburgh[0.041]; Rangers[0.041]; Boston[0.041]; Division[0.041]; Colorado[0.040]; Colorado[0.040]; Tampa[0.040]; Hartford[0.040]; Los[0.040]; Los[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pittsburgh ==> ENTITY: \u001b[32mPittsburgh Penguins\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.229:0.229[0]; log p(e|m)= -3.689:-3.689[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.045]; Anaheim[0.045]; Ottawa[0.044]; Buffalo[0.043]; Edmonton[0.043]; Montreal[0.043]; Islanders[0.043]; Boston[0.042]; Toronto[0.042]; Toronto[0.042]; Vancouver[0.042]; Vancouver[0.042]; Rangers[0.041]; Washington[0.041]; Tampa[0.041]; Calgary[0.041]; Calgary[0.041]; Phoenix[0.040]; Phoenix[0.040]; Colorado[0.040]; Colorado[0.040]; Hartford[0.039]; Division[0.039]; Chicago[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Detroit ==> ENTITY: \u001b[32mDetroit Red Wings\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.215:0.215[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnaheim[0.046]; Ottawa[0.045]; Islanders[0.044]; Edmonton[0.044]; Montreal[0.043]; Conference[0.043]; Philadelphia[0.043]; Vancouver[0.043]; Toronto[0.042]; Rangers[0.042]; Pittsburgh[0.042]; Boston[0.042]; Calgary[0.041]; Tampa[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Phoenix[0.040]; Chicago[0.040]; Washington[0.040]; Colorado[0.040]; Louis[0.040]; Dallas[0.039]; Western[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colorado ==> ENTITY: \u001b[32mColorado Avalanche\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.258:0.258[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nColorado[0.052]; Anaheim[0.044]; Anaheim[0.044]; Edmonton[0.043]; Buffalo[0.043]; Montreal[0.043]; Ottawa[0.042]; Toronto[0.042]; Toronto[0.042]; Vancouver[0.042]; Vancouver[0.042]; Islanders[0.041]; Islanders[0.041]; Calgary[0.041]; Calgary[0.041]; Pittsburgh[0.040]; Boston[0.040]; Phoenix[0.040]; Phoenix[0.040]; Tampa[0.040]; Tampa[0.040]; Rangers[0.040]; Rangers[0.040]; Dallas[0.039]; \t\n\r [======================================>.......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 3s352ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3830/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1333testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bangkok\u001b[39m ==> ENTITY: \u001b[32mBangkok\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.118:0.118[0]; log p(e|m)= -0.001:-0.001[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThai[0.049]; Thai[0.049]; Thai[0.049]; Singapore[0.048]; Malaysia[0.046]; Philippines[0.046]; Japan[0.045]; Japan[0.045]; Indonesia[0.043]; Indonesia[0.043]; detailed[0.043]; ace[0.042]; Dec[0.042]; King[0.041]; Commerce[0.041]; Iran[0.041]; Iran[0.041]; Yang[0.041]; Destination[0.041]; Ministry[0.041]; follows[0.041]; ports[0.040]; Date[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.052]; Japan[0.052]; Singapore[0.045]; Philippines[0.045]; Vietnam[0.045]; Malaysia[0.042]; Indonesia[0.041]; Indonesia[0.041]; Ace[0.041]; Thai[0.040]; Thai[0.040]; Thai[0.040]; vessels[0.040]; Yang[0.040]; Bangkok[0.040]; Bangkok[0.040]; Burma[0.040]; Day[0.040]; Dec[0.039]; Glory[0.039]; Ministry[0.039]; King[0.039]; Commerce[0.039]; Tong[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bangkok\u001b[39m ==> ENTITY: \u001b[32mBangkok\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.102:0.102[0]; log p(e|m)= -0.001:-0.001[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.047]; Malaysia[0.045]; Philippines[0.045]; Japan[0.044]; Japan[0.044]; Japan[0.044]; Vietnam[0.044]; Indonesia[0.043]; Indonesia[0.043]; Burma[0.042]; Tong[0.042]; ace[0.041]; Yang[0.041]; King[0.041]; Iran[0.040]; Iran[0.040]; Destination[0.040]; Arrival[0.040]; Ace[0.039]; Africa[0.039]; Day[0.039]; Shun[0.039]; Princess[0.038]; Lucky[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.052]; Japan[0.052]; Singapore[0.045]; Philippines[0.045]; Vietnam[0.045]; Malaysia[0.043]; Indonesia[0.041]; Indonesia[0.041]; Ace[0.041]; Thai[0.041]; Thai[0.041]; Yang[0.040]; Bangkok[0.040]; Burma[0.040]; Day[0.040]; Glory[0.039]; Ministry[0.039]; King[0.039]; Commerce[0.039]; Tong[0.039]; ace[0.039]; follows[0.039]; Africa[0.039]; Destination[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.052]; Japan[0.052]; Singapore[0.045]; Philippines[0.045]; Vietnam[0.045]; Malaysia[0.042]; Indonesia[0.041]; Indonesia[0.041]; Ace[0.041]; Thai[0.040]; Thai[0.040]; Thai[0.040]; vessels[0.040]; Yang[0.040]; Bangkok[0.040]; Bangkok[0.040]; Burma[0.040]; Day[0.040]; Dec[0.039]; Glory[0.039]; Ministry[0.039]; King[0.039]; Commerce[0.039]; Tong[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Philippines ==> ENTITY: \u001b[32mPhilippines\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.129:0.129[0]; log p(e|m)= -0.180:-0.180[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMalaysia[0.047]; Singapore[0.045]; Indonesia[0.045]; Indonesia[0.045]; Vietnam[0.044]; Japan[0.044]; Japan[0.044]; Japan[0.044]; Thai[0.043]; Thai[0.043]; Thai[0.043]; Bangkok[0.043]; Bangkok[0.043]; vessels[0.041]; Africa[0.039]; Iran[0.039]; Iran[0.039]; Commerce[0.039]; Burma[0.039]; Dec[0.038]; Vessel[0.038]; Day[0.038]; Tong[0.038]; Yang[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vietnam ==> ENTITY: \u001b[32mVietnam\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.149:0.149[0]; log p(e|m)= -0.390:-0.390[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhilippines[0.047]; Malaysia[0.046]; Singapore[0.046]; Indonesia[0.045]; Indonesia[0.045]; Thai[0.045]; Japan[0.045]; Japan[0.045]; Japan[0.045]; Burma[0.044]; Bangkok[0.043]; Yang[0.042]; Africa[0.041]; Tong[0.040]; Day[0.040]; King[0.039]; Glory[0.039]; Iran[0.038]; Iran[0.038]; Shun[0.038]; Quantity[0.037]; Destination[0.037]; Arrival[0.037]; follows[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Africa ==> ENTITY: \u001b[32mAfrica\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDay[0.056]; Malaysia[0.048]; Indonesia[0.048]; Indonesia[0.048]; Japan[0.047]; Japan[0.047]; Japan[0.047]; Philippines[0.047]; King[0.046]; Vietnam[0.046]; Singapore[0.046]; Iran[0.044]; Iran[0.044]; follows[0.044]; Destination[0.044]; Bangkok[0.043]; Bangkok[0.043]; Ministry[0.043]; Burma[0.043]; movements[0.042]; Commerce[0.042]; Glory[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Thai ==> ENTITY: \u001b[32mThailand\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.168:0.168[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThai[0.050]; Thai[0.050]; Bangkok[0.048]; Indonesia[0.046]; Indonesia[0.046]; Philippines[0.045]; Singapore[0.045]; Japan[0.043]; Japan[0.043]; rice[0.041]; rice[0.041]; Yang[0.041]; Date[0.040]; Dec[0.039]; Princess[0.039]; Day[0.039]; Africa[0.039]; vessels[0.039]; detailed[0.038]; follows[0.038]; Commerce[0.038]; Destination[0.037]; Ministry[0.037]; movements[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Thai ==> ENTITY: \u001b[32mThailand\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.199:0.199[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThai[0.049]; Thai[0.049]; Malaysia[0.047]; Bangkok[0.046]; Indonesia[0.045]; Indonesia[0.045]; Vietnam[0.044]; Philippines[0.044]; Singapore[0.044]; Japan[0.042]; Japan[0.042]; Japan[0.042]; King[0.040]; Yang[0.040]; rice[0.039]; rice[0.039]; Tong[0.039]; Date[0.039]; Dec[0.038]; Princess[0.038]; Day[0.038]; Africa[0.038]; Shun[0.038]; vessels[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burma ==> ENTITY: \u001b[32mBurma\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.125:0.125[0]; log p(e|m)= -0.202:-0.202[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMalaysia[0.047]; Singapore[0.046]; Indonesia[0.046]; Indonesia[0.046]; Philippines[0.045]; Vietnam[0.044]; King[0.043]; Japan[0.043]; Japan[0.043]; Japan[0.043]; Bangkok[0.043]; Africa[0.041]; Iran[0.041]; Iran[0.041]; Day[0.040]; Tong[0.040]; Glory[0.039]; Princess[0.039]; Yang[0.039]; Destination[0.038]; Shun[0.038]; Arrival[0.038]; Date[0.038]; Lucky[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.174:0.174[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMalaysia[0.051]; Indonesia[0.046]; Indonesia[0.046]; Philippines[0.046]; Vietnam[0.045]; Japan[0.045]; Japan[0.045]; Japan[0.045]; ports[0.045]; vessels[0.045]; Thai[0.044]; Thai[0.044]; Thai[0.044]; Bangkok[0.044]; Bangkok[0.044]; tonnes[0.042]; rice[0.041]; rice[0.041]; Ministry[0.040]; Burma[0.040]; Tong[0.039]; Dec[0.039]; detailed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.216:0.216[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.051]; Malaysia[0.050]; Singapore[0.047]; Japan[0.047]; Japan[0.047]; Japan[0.047]; Philippines[0.047]; Vietnam[0.046]; Thai[0.045]; Thai[0.045]; Thai[0.045]; Bangkok[0.043]; Bangkok[0.043]; Africa[0.042]; rice[0.041]; rice[0.041]; Burma[0.041]; follows[0.040]; tonnes[0.039]; Day[0.039]; Tong[0.039]; Yang[0.038]; Destination[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Thai ==> ENTITY: \u001b[32mThailand\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.191:0.191[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThai[0.049]; Thai[0.049]; Malaysia[0.047]; Bangkok[0.047]; Indonesia[0.045]; Indonesia[0.045]; Philippines[0.044]; Singapore[0.044]; Japan[0.042]; Japan[0.042]; Japan[0.042]; King[0.040]; Yang[0.040]; rice[0.040]; rice[0.040]; Tong[0.039]; Date[0.039]; Dec[0.039]; Princess[0.038]; Day[0.038]; Africa[0.038]; Shun[0.038]; vessels[0.038]; detailed[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.216:0.216[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.051]; Malaysia[0.050]; Singapore[0.047]; Japan[0.047]; Japan[0.047]; Japan[0.047]; Philippines[0.047]; Vietnam[0.046]; Thai[0.045]; Thai[0.045]; Thai[0.045]; Bangkok[0.043]; Bangkok[0.043]; Africa[0.042]; rice[0.041]; rice[0.041]; Burma[0.041]; follows[0.040]; tonnes[0.039]; Day[0.039]; Tong[0.039]; Yang[0.038]; Destination[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iran ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.085:0.085[0]; log p(e|m)= -0.138:-0.138[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.052]; Japan[0.045]; Japan[0.045]; Japan[0.045]; Malaysia[0.044]; Singapore[0.044]; Indonesia[0.042]; Indonesia[0.042]; Glory[0.042]; Vietnam[0.042]; Philippines[0.041]; Day[0.041]; Dec[0.041]; Burma[0.040]; Princess[0.040]; King[0.040]; Ministry[0.040]; Africa[0.040]; Lucky[0.040]; detailed[0.040]; ace[0.039]; Sabr[0.039]; Bangkok[0.039]; Bangkok[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Malaysia ==> ENTITY: \u001b[32mMalaysia\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.172:0.172[0]; log p(e|m)= -0.202:-0.202[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.049]; Indonesia[0.048]; Indonesia[0.048]; Yang[0.044]; Philippines[0.044]; Japan[0.044]; Japan[0.044]; Japan[0.044]; Vietnam[0.043]; Tong[0.043]; Thai[0.043]; Thai[0.043]; Bangkok[0.042]; Bangkok[0.042]; Day[0.040]; Burma[0.039]; Dec[0.038]; detailed[0.038]; Africa[0.038]; Shun[0.038]; King[0.037]; ace[0.037]; Ministry[0.037]; follows[0.037]; \t\n\r [======================================>.......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 3s261ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3847/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1244testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.116:0.116[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nproposing[0.045]; States[0.043]; Secretary[0.043]; announced[0.043]; President[0.043]; discussed[0.043]; Africa[0.042]; proposes[0.042]; foreign[0.042]; foreign[0.042]; said[0.042]; candidate[0.041]; General[0.041]; Egyptian[0.041]; post[0.041]; post[0.041]; Friday[0.041]; United[0.041]; United[0.041]; Nations[0.041]; general[0.041]; secretary[0.041]; Senegal[0.040]; Senegal[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Diouf\u001b[39m ==> ENTITY: \u001b[32mAbdou Diouf\u001b[39m\t\nSCORES: global= 0.312:0.312[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.046]; Senegal[0.046]; Senegal[0.046]; general[0.045]; general[0.045]; Secretary[0.045]; secretary[0.045]; term[0.044]; summit[0.044]; General[0.044]; candidate[0.043]; minister[0.043]; minister[0.043]; second[0.043]; Dakar[0.043]; Niasse[0.042]; Niasse[0.042]; given[0.042]; States[0.042]; announced[0.042]; Africa[0.042]; Diouf[0.041]; Diouf[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Moustapha Niasse\u001b[39m ==> ENTITY: \u001b[32mMoustapha Niasse\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNiasse[0.045]; summit[0.043]; Secretary[0.043]; candidate[0.043]; secretary[0.042]; nominate[0.042]; African[0.042]; African[0.042]; African[0.042]; endorsement[0.042]; second[0.042]; United[0.042]; United[0.042]; United[0.042]; Diouf[0.041]; Diouf[0.041]; term[0.041]; General[0.041]; discussed[0.041]; general[0.041]; general[0.041]; Moustapha[0.040]; intention[0.040]; announced[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = African ==> ENTITY: \u001b[32mAfrica\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.125:0.125[0]; log p(e|m)= -0.911:-0.911[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrica[0.050]; African[0.049]; African[0.049]; Senegal[0.046]; Senegal[0.046]; President[0.045]; United[0.043]; United[0.043]; United[0.043]; Abdou[0.043]; Secretary[0.043]; summit[0.042]; endorsement[0.042]; States[0.042]; General[0.042]; announced[0.042]; Nations[0.042]; Nations[0.042]; secretary[0.041]; Egyptian[0.041]; general[0.041]; general[0.041]; Burkina[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burkina Faso ==> ENTITY: \u001b[32mBurkina Faso\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.179:0.179[0]; log p(e|m)= -0.246:-0.246[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSenegal[0.048]; Senegal[0.048]; Africa[0.046]; general[0.045]; general[0.045]; African[0.044]; African[0.044]; African[0.044]; term[0.044]; States[0.044]; Nations[0.044]; Nations[0.044]; Dakar[0.043]; state[0.043]; candidate[0.042]; Abdou[0.042]; President[0.042]; endorsement[0.041]; General[0.041]; foreign[0.041]; foreign[0.041]; heads[0.041]; Ghali[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Moustapha Niasse\u001b[39m ==> ENTITY: \u001b[32mMoustapha Niasse\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.046]; Niasse[0.044]; Senegal[0.043]; Senegal[0.043]; summit[0.043]; Secretary[0.043]; candidate[0.042]; secretary[0.042]; nominate[0.041]; African[0.041]; African[0.041]; African[0.041]; second[0.041]; United[0.041]; United[0.041]; Diouf[0.041]; Diouf[0.041]; Diouf[0.041]; term[0.041]; General[0.041]; discussed[0.041]; general[0.040]; minister[0.040]; minister[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Africa ==> ENTITY: \u001b[32mAfrica\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrican[0.046]; African[0.046]; African[0.046]; Senegal[0.045]; state[0.043]; States[0.042]; returned[0.042]; Nations[0.041]; Nations[0.041]; given[0.041]; foreign[0.041]; foreign[0.041]; second[0.041]; Dakar[0.041]; endorsement[0.041]; Secretary[0.041]; Franco[0.040]; act[0.040]; Burkina[0.040]; said[0.040]; said[0.040]; announced[0.040]; Egyptian[0.040]; General[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = African ==> ENTITY: \u001b[32mAfrica\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.125:0.125[0]; log p(e|m)= -0.911:-0.911[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrica[0.050]; African[0.049]; African[0.049]; Senegal[0.046]; Senegal[0.046]; President[0.045]; United[0.043]; United[0.043]; United[0.043]; Abdou[0.043]; Secretary[0.043]; summit[0.042]; endorsement[0.042]; States[0.042]; General[0.042]; announced[0.042]; Nations[0.042]; Nations[0.042]; secretary[0.041]; Egyptian[0.041]; general[0.041]; general[0.041]; Burkina[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Senegal ==> ENTITY: \u001b[32mSenegal\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.154:0.154[0]; log p(e|m)= -0.403:-0.403[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSenegal[0.049]; President[0.044]; minister[0.043]; minister[0.043]; Dakar[0.043]; Nations[0.043]; General[0.043]; African[0.042]; African[0.042]; African[0.042]; second[0.042]; term[0.042]; Niasse[0.041]; Burkina[0.041]; States[0.041]; general[0.040]; Abdou[0.040]; secretary[0.040]; Secretary[0.040]; discussed[0.040]; said[0.040]; candidate[0.040]; Diouf[0.040]; Diouf[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Senegal ==> ENTITY: \u001b[32mSenegal\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.166:0.166[0]; log p(e|m)= -0.403:-0.403[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSenegal[0.049]; Africa[0.045]; President[0.044]; minister[0.043]; minister[0.043]; Dakar[0.042]; Nations[0.042]; General[0.042]; African[0.042]; African[0.042]; African[0.042]; second[0.042]; term[0.041]; Niasse[0.041]; Burkina[0.041]; States[0.041]; general[0.040]; Abdou[0.040]; secretary[0.040]; Secretary[0.040]; said[0.040]; candidate[0.040]; Diouf[0.039]; Diouf[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Diouf\u001b[39m ==> ENTITY: \u001b[32mAbdou Diouf\u001b[39m\t\nSCORES: global= 0.312:0.312[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngeneral[0.044]; general[0.044]; Secretary[0.044]; secretary[0.043]; term[0.043]; summit[0.043]; General[0.043]; endorsement[0.042]; candidate[0.042]; minister[0.042]; second[0.042]; Niasse[0.041]; Niasse[0.041]; given[0.041]; States[0.041]; announced[0.041]; Africa[0.041]; Diouf[0.040]; Diouf[0.040]; foreign[0.040]; Burkina[0.040]; nominate[0.040]; post[0.040]; Faso[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Boutros Boutros-Ghali\u001b[39m ==> ENTITY: \u001b[32mBoutros Boutros-Ghali\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSecretary[0.046]; Ghali[0.045]; Senegal[0.045]; Senegal[0.045]; General[0.044]; President[0.044]; Moustapha[0.044]; Moustapha[0.044]; secretary[0.043]; said[0.043]; said[0.043]; discussed[0.043]; Egyptian[0.043]; foreign[0.043]; foreign[0.043]; general[0.043]; general[0.043]; vetoed[0.043]; Nations[0.043]; Nations[0.043]; United[0.042]; United[0.042]; United[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = United Nations\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.117:0.117[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStates[0.044]; Senegal[0.044]; Senegal[0.044]; Secretary[0.043]; discussed[0.042]; Africa[0.042]; Egyptian[0.042]; President[0.042]; foreign[0.041]; foreign[0.041]; proposes[0.041]; African[0.041]; African[0.041]; African[0.041]; United[0.041]; United[0.041]; announced[0.041]; General[0.041]; vetoed[0.041]; reporters[0.041]; post[0.040]; post[0.040]; summit[0.040]; Friday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = United Nations\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStates[0.047]; Secretary[0.046]; endorsement[0.045]; discussed[0.045]; Egyptian[0.045]; vetoed[0.045]; Africa[0.044]; United[0.044]; announced[0.044]; reporters[0.043]; General[0.043]; Nations[0.043]; summit[0.043]; Moustapha[0.043]; secretary[0.043]; returned[0.043]; said[0.042]; Diouf[0.042]; Diouf[0.042]; intention[0.042]; African[0.042]; African[0.042]; African[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dakar\u001b[39m ==> ENTITY: \u001b[32mDakar\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.008:-0.008[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSenegal[0.047]; Senegal[0.047]; Friday[0.043]; Egyptian[0.043]; African[0.043]; African[0.043]; African[0.043]; Africa[0.042]; term[0.042]; general[0.041]; Burkina[0.041]; President[0.041]; Abdou[0.041]; announced[0.041]; open[0.041]; summit[0.041]; said[0.041]; second[0.041]; returned[0.040]; Moustapha[0.040]; secretary[0.040]; Nations[0.040]; Niasse[0.040]; intention[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.050:0.050[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.045]; returned[0.043]; Senegal[0.043]; Senegal[0.043]; Nations[0.043]; Nations[0.043]; endorsement[0.042]; Friday[0.042]; African[0.042]; African[0.042]; African[0.042]; United[0.042]; United[0.042]; profile[0.042]; does[0.041]; state[0.041]; Africa[0.041]; lose[0.041]; foreign[0.041]; foreign[0.041]; given[0.040]; said[0.040]; said[0.040]; term[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Abdou Diouf\u001b[39m ==> ENTITY: \u001b[32mAbdou Diouf\u001b[39m\t\nSCORES: global= 0.312:0.312[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.046]; Senegal[0.046]; Senegal[0.046]; general[0.045]; Secretary[0.045]; secretary[0.045]; term[0.045]; summit[0.045]; General[0.044]; candidate[0.044]; minister[0.043]; minister[0.043]; second[0.043]; Dakar[0.043]; Niasse[0.043]; States[0.042]; announced[0.042]; Africa[0.042]; Diouf[0.042]; Diouf[0.042]; foreign[0.041]; foreign[0.041]; Burkina[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Egyptian ==> ENTITY: \u001b[32mEgypt\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.125:0.125[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSenegal[0.046]; Senegal[0.046]; second[0.045]; Boutros[0.043]; Boutros[0.043]; African[0.042]; African[0.042]; African[0.042]; Africa[0.042]; foreign[0.041]; foreign[0.041]; Abdou[0.041]; Nations[0.041]; Nations[0.041]; Ghali[0.041]; Moustapha[0.040]; Moustapha[0.040]; said[0.040]; said[0.040]; profile[0.040]; States[0.040]; returned[0.040]; state[0.040]; lose[0.039]; \t\n\r [======================================>.......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 3s161ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3865/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1245testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bangui\u001b[39m ==> ENTITY: \u001b[32mBangui\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBangui[0.051]; Bangui[0.051]; Burkina[0.045]; Central[0.044]; Central[0.044]; Africa[0.042]; mutineers[0.042]; colony[0.041]; political[0.041]; ethnic[0.041]; government[0.041]; government[0.041]; Republic[0.041]; capital[0.041]; African[0.040]; African[0.040]; army[0.039]; abducted[0.039]; home[0.039]; cabinet[0.039]; Jacques[0.039]; Friday[0.039]; France[0.039]; France[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Blaise Compaore\u001b[39m ==> ENTITY: \u001b[32mBlaise Compaoré\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.045]; Bangui[0.043]; Bangui[0.043]; Bangui[0.043]; elections[0.042]; Burkina[0.042]; spokesman[0.042]; party[0.042]; Patasse[0.042]; Patasse[0.042]; Patasse[0.042]; civil[0.041]; Gabon[0.041]; meeting[0.041]; said[0.041]; deadlock[0.041]; Baya[0.041]; Africa[0.041]; Central[0.041]; Mali[0.041]; rebel[0.041]; rebel[0.041]; April[0.041]; seek[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrench[0.052]; French[0.052]; French[0.052]; Christophe[0.047]; Bangui[0.046]; Bangui[0.046]; central[0.045]; fought[0.045]; palace[0.045]; ethnic[0.044]; home[0.044]; said[0.044]; said[0.044]; men[0.044]; forces[0.044]; people[0.044]; Thursday[0.044]; Thursday[0.044]; President[0.044]; tribes[0.044]; Ange[0.043]; areas[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burkina Faso ==> ENTITY: \u001b[32mBurkina Faso\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.177:0.177[0]; log p(e|m)= -0.246:-0.246[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnational[0.045]; political[0.043]; France[0.043]; France[0.043]; Central[0.043]; African[0.043]; African[0.043]; Republic[0.043]; government[0.042]; government[0.042]; pacts[0.042]; President[0.041]; President[0.041]; conference[0.041]; ethnic[0.041]; communities[0.041]; cabinet[0.041]; colony[0.040]; head[0.040]; Jacques[0.040]; Bangui[0.040]; Bangui[0.040]; Bangui[0.040]; foreign[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Codos\u001b[39m ==> ENTITY: \u001b[32mCodos\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.100:0.100[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntown[0.045]; city[0.045]; city[0.045]; forces[0.044]; troops[0.042]; troops[0.042]; Rebels[0.042]; civilian[0.042]; rebels[0.042]; rebels[0.042]; rebels[0.042]; looted[0.041]; Hospital[0.041]; Sudan[0.041]; reported[0.041]; tribal[0.041]; people[0.041]; people[0.041]; civilians[0.041]; rebel[0.040]; strongholds[0.040]; arming[0.040]; amnesty[0.040]; concessions[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bangui\u001b[39m ==> ENTITY: \u001b[32mBangui\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBangui[0.051]; Bangui[0.051]; communities[0.046]; Burkina[0.045]; Central[0.043]; Central[0.043]; Africa[0.042]; mutineers[0.041]; colony[0.041]; political[0.041]; ethnic[0.041]; government[0.041]; government[0.041]; Republic[0.041]; capital[0.041]; African[0.040]; African[0.040]; order[0.040]; army[0.039]; abducted[0.039]; national[0.039]; home[0.039]; cabinet[0.039]; Jacques[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bangui\u001b[39m ==> ENTITY: \u001b[32mBangui\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBangui[0.049]; Bangui[0.049]; Burkina[0.043]; districts[0.043]; districts[0.043]; Mali[0.043]; Patasse[0.042]; Patasse[0.042]; Patasse[0.042]; Gabon[0.042]; tribes[0.042]; Central[0.042]; Compaore[0.041]; rebels[0.040]; Africa[0.040]; central[0.040]; mutineers[0.040]; areas[0.040]; Chad[0.039]; state[0.039]; elections[0.039]; Church[0.039]; people[0.039]; African[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Central Africa ==> ENTITY: \u001b[32mCentral African Republic\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.989:-0.989[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrican[0.048]; Sudan[0.047]; Church[0.046]; Bangui[0.045]; Bangui[0.045]; Chad[0.045]; Chad[0.045]; Mali[0.045]; troops[0.044]; mission[0.044]; mission[0.044]; Gabon[0.043]; Burkina[0.043]; civilians[0.042]; uprising[0.042]; Patasse[0.042]; Patasse[0.042]; Patasse[0.042]; Patasse[0.042]; war[0.042]; civil[0.042]; state[0.042]; helicopters[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.110:0.110[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrench[0.048]; French[0.048]; French[0.048]; African[0.043]; soldiers[0.042]; soldiers[0.042]; Bangui[0.042]; Bangui[0.042]; Bangui[0.042]; Franco[0.041]; troops[0.041]; troops[0.041]; troops[0.041]; fought[0.041]; palace[0.041]; ethnic[0.040]; home[0.040]; central[0.040]; week[0.040]; said[0.040]; men[0.040]; people[0.040]; Thursday[0.040]; tribes[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.115:0.115[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnews[0.047]; conference[0.044]; told[0.043]; told[0.043]; major[0.043]; pay[0.042]; soldiers[0.042]; soldiers[0.042]; based[0.041]; troops[0.041]; troops[0.041]; troops[0.041]; government[0.041]; said[0.041]; said[0.041]; said[0.041]; domestic[0.041]; Thursday[0.041]; national[0.041]; mutiny[0.040]; foreign[0.040]; help[0.040]; debate[0.040]; Service[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.050]; French[0.045]; Republic[0.044]; government[0.042]; government[0.042]; colony[0.042]; Africa[0.042]; national[0.042]; Jacques[0.041]; African[0.041]; African[0.041]; army[0.041]; Bangui[0.041]; Bangui[0.041]; Bangui[0.041]; Chirac[0.041]; cabinet[0.041]; Christophe[0.040]; major[0.040]; combat[0.040]; communities[0.040]; political[0.040]; domestic[0.040]; Franco[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bangui\u001b[39m ==> ENTITY: \u001b[32mBangui\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBangui[0.051]; Bangui[0.051]; Burkina[0.045]; districts[0.045]; districts[0.045]; Mali[0.045]; Patasse[0.044]; Patasse[0.044]; Patasse[0.044]; Gabon[0.044]; tribes[0.044]; Central[0.044]; Compaore[0.043]; rebels[0.042]; Africa[0.042]; central[0.042]; areas[0.041]; Chad[0.041]; state[0.041]; elections[0.041]; Church[0.041]; people[0.041]; African[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Central African Republic ==> ENTITY: \u001b[32mCentral African Republic\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.308:-0.308[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrica[0.046]; government[0.045]; government[0.045]; Bangui[0.045]; Bangui[0.045]; Bangui[0.045]; African[0.045]; Central[0.044]; violence[0.044]; violence[0.044]; colony[0.043]; troops[0.043]; troops[0.043]; Burkina[0.043]; unrest[0.043]; political[0.043]; Ex[0.043]; capital[0.042]; army[0.042]; President[0.042]; home[0.042]; pacts[0.042]; help[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.071:0.071[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSudan[0.044]; Chad[0.044]; rebellion[0.044]; troops[0.043]; number[0.043]; civilian[0.043]; fighting[0.043]; fighting[0.043]; said[0.042]; people[0.042]; people[0.042]; Thursday[0.042]; Thursday[0.042]; foreign[0.041]; groups[0.041]; town[0.041]; forces[0.041]; weeks[0.041]; rebels[0.040]; rebels[0.040]; reported[0.040]; Patasse[0.040]; city[0.040]; city[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chirac\u001b[39m ==> ENTITY: \u001b[32mJacques Chirac\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.109:0.109[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.045]; French[0.044]; French[0.044]; civil[0.043]; visit[0.042]; week[0.042]; said[0.042]; party[0.042]; meeting[0.041]; Central[0.041]; rebel[0.041]; rebel[0.041]; rebels[0.041]; elections[0.041]; Bangui[0.041]; Bangui[0.041]; Bangui[0.041]; miss[0.041]; Burkina[0.041]; war[0.041]; summit[0.041]; summit[0.041]; resign[0.040]; seek[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yakoma\u001b[39m ==> ENTITY: \u001b[32mYakoma people\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.106:0.106[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrican[0.046]; Burkina[0.045]; ethnic[0.044]; communities[0.044]; capital[0.043]; campaign[0.042]; Bangui[0.042]; foreign[0.042]; tribe[0.042]; national[0.041]; said[0.041]; said[0.041]; said[0.041]; rebel[0.041]; November[0.041]; troops[0.041]; troops[0.041]; troops[0.041]; troops[0.041]; Thursday[0.040]; Thursday[0.040]; belong[0.040]; pay[0.040]; conference[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jacques Chirac\u001b[39m ==> ENTITY: \u001b[32mJacques Chirac\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.047]; France[0.047]; President[0.045]; President[0.045]; cabinet[0.045]; French[0.045]; government[0.044]; government[0.044]; political[0.044]; Christophe[0.044]; national[0.043]; minister[0.042]; minister[0.042]; minister[0.042]; minister[0.042]; order[0.042]; conference[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; elected[0.042]; involved[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.074:0.074[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrica[0.044]; Sudan[0.043]; Chad[0.043]; Chad[0.043]; Bangui[0.043]; rebellion[0.043]; troops[0.042]; number[0.042]; civilian[0.042]; fighting[0.042]; Central[0.042]; Gabon[0.041]; establish[0.041]; said[0.041]; April[0.041]; state[0.041]; Mali[0.041]; forces[0.041]; Blaise[0.041]; Thursday[0.041]; President[0.041]; Soldiers[0.041]; people[0.041]; foreign[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bangui\u001b[39m ==> ENTITY: \u001b[32mBangui\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBangui[0.051]; Bangui[0.051]; Bangui[0.051]; Burkina[0.045]; districts[0.045]; districts[0.045]; Mali[0.045]; Patasse[0.044]; Patasse[0.044]; Gabon[0.044]; tribes[0.044]; Compaore[0.043]; rebels[0.042]; central[0.042]; mutineers[0.041]; areas[0.041]; Chad[0.041]; state[0.041]; people[0.041]; African[0.040]; rebel[0.040]; soldiers[0.040]; stronghold[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gabon ==> ENTITY: \u001b[32mGabon\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.302:-0.302[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrica[0.045]; Mali[0.044]; elections[0.044]; Chad[0.043]; party[0.043]; African[0.042]; rebel[0.042]; rebel[0.042]; Patasse[0.042]; Patasse[0.042]; Patasse[0.042]; President[0.042]; state[0.041]; Bangui[0.041]; Bangui[0.041]; Bangui[0.041]; Burkina[0.041]; week[0.041]; rebels[0.041]; people[0.041]; won[0.041]; civil[0.041]; central[0.040]; visit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bangui\u001b[39m ==> ENTITY: \u001b[32mBangui\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.112:0.112[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBangui[0.051]; districts[0.045]; districts[0.045]; Patasse[0.044]; Patasse[0.044]; Patasse[0.044]; tribes[0.044]; tribe[0.042]; central[0.041]; mutineers[0.041]; areas[0.041]; presidential[0.041]; ethnic[0.041]; capital[0.041]; people[0.040]; rebel[0.040]; rebel[0.040]; palace[0.040]; soldiers[0.039]; soldiers[0.039]; soldiers[0.039]; stronghold[0.039]; city[0.039]; city[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.050]; France[0.050]; combat[0.044]; Republic[0.043]; Jacques[0.042]; government[0.042]; government[0.042]; Christophe[0.042]; army[0.042]; colony[0.042]; African[0.041]; African[0.041]; soldiers[0.040]; soldiers[0.040]; Bangui[0.040]; Bangui[0.040]; national[0.040]; Chirac[0.040]; communities[0.040]; Franco[0.040]; political[0.040]; troops[0.040]; troops[0.040]; troops[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Central Africa ==> ENTITY: \u001b[32mCentral African Republic\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.989:-0.989[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrican[0.047]; Bangui[0.044]; Bangui[0.044]; Bangui[0.044]; colony[0.044]; government[0.043]; government[0.043]; troops[0.043]; Central[0.043]; Republic[0.041]; France[0.041]; France[0.041]; ethnic[0.041]; army[0.041]; capital[0.041]; domestic[0.040]; political[0.040]; unrest[0.040]; elected[0.040]; President[0.040]; said[0.040]; said[0.040]; cabinet[0.040]; combat[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bangui\u001b[39m ==> ENTITY: \u001b[32mBangui\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBangui[0.051]; Bangui[0.051]; communities[0.046]; Burkina[0.045]; Central[0.043]; Central[0.043]; Africa[0.042]; mutineers[0.041]; colony[0.041]; political[0.041]; ethnic[0.041]; government[0.041]; government[0.041]; Republic[0.041]; capital[0.041]; African[0.040]; African[0.040]; order[0.040]; army[0.039]; abducted[0.039]; home[0.039]; cabinet[0.039]; Jacques[0.039]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burkina Faso ==> ENTITY: \u001b[32mBurkina Faso\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.162:0.162[0]; log p(e|m)= -0.246:-0.246[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrica[0.047]; Mali[0.046]; Gabon[0.046]; Bangui[0.045]; Bangui[0.045]; Bangui[0.045]; Compaore[0.045]; Central[0.044]; African[0.044]; Chad[0.043]; central[0.043]; Patasse[0.043]; Patasse[0.043]; Patasse[0.043]; areas[0.043]; party[0.043]; tribes[0.043]; dialogue[0.042]; won[0.042]; civil[0.042]; districts[0.042]; districts[0.042]; Chirac[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sudan ==> ENTITY: \u001b[32mSudan\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.285:-0.285[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrica[0.053]; Patasse[0.046]; Patasse[0.046]; Patasse[0.046]; rebels[0.044]; rebels[0.044]; rebels[0.044]; rebel[0.043]; civilians[0.042]; tribalism[0.041]; Chad[0.041]; tribal[0.040]; troops[0.040]; troops[0.040]; foreign[0.040]; groups[0.040]; mutinies[0.039]; forces[0.039]; uprising[0.039]; won[0.039]; rebellion[0.039]; civilian[0.039]; people[0.038]; people[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.118:0.118[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrench[0.047]; French[0.047]; French[0.047]; civil[0.043]; African[0.043]; Chad[0.042]; Bangui[0.042]; Bangui[0.042]; Bangui[0.042]; Bangui[0.042]; war[0.041]; establish[0.041]; tribes[0.041]; Franco[0.041]; central[0.041]; fought[0.040]; open[0.040]; Gabon[0.040]; week[0.040]; said[0.040]; said[0.040]; state[0.040]; men[0.040]; Mali[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ange-Felix Patasse\u001b[39m ==> ENTITY: \u001b[32mAnge-Félix Patassé\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.049]; President[0.049]; presidential[0.048]; political[0.047]; November[0.047]; Patasse[0.047]; Patasse[0.047]; Bangui[0.046]; national[0.045]; order[0.045]; conference[0.045]; rebel[0.045]; backed[0.045]; loyalist[0.044]; avoid[0.044]; Thursday[0.044]; Thursday[0.044]; violence[0.044]; summit[0.044]; Burkina[0.044]; African[0.044]; field[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mali ==> ENTITY: \u001b[32mMali\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.167:0.167[0]; log p(e|m)= -0.378:-0.378[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrica[0.045]; Gabon[0.045]; tribes[0.044]; Burkina[0.044]; African[0.043]; central[0.042]; Chad[0.042]; Compaore[0.042]; seek[0.041]; rebels[0.041]; Bangui[0.041]; Bangui[0.041]; Bangui[0.041]; Central[0.041]; week[0.041]; districts[0.041]; won[0.041]; Patasse[0.041]; Patasse[0.041]; Patasse[0.041]; said[0.041]; city[0.040]; state[0.040]; civil[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.120:0.120[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.053]; French[0.046]; Republic[0.046]; government[0.044]; government[0.044]; colony[0.044]; Africa[0.044]; national[0.043]; Jacques[0.043]; African[0.043]; African[0.043]; Bangui[0.043]; Bangui[0.043]; Bangui[0.043]; Chirac[0.042]; elected[0.042]; cabinet[0.042]; major[0.042]; combat[0.042]; communities[0.042]; political[0.042]; domestic[0.042]; Franco[0.042]; \t\n\r [======================================>.......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 2s993ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3895/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1174testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Harare\u001b[39m ==> ENTITY: \u001b[32mHarare\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZimbabwe[0.046]; Zimbabwe[0.046]; Zimbabwe[0.046]; Zimbabwe[0.046]; Zimbabwe[0.046]; South[0.044]; African[0.043]; Club[0.042]; Scores[0.041]; Andrew[0.041]; Andrew[0.041]; Friday[0.041]; Brett[0.040]; Park[0.040]; scores[0.040]; Mark[0.040]; Mark[0.040]; Mark[0.040]; Cayeux[0.040]; Golf[0.040]; Golf[0.040]; round[0.039]; Schalk[0.039]; Merwe[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Steve van Vuuren\u001b[39m ==> ENTITY: \u001b[32mSteve van Vuuren\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCayeux[0.048]; Murless[0.048]; Hennie[0.048]; Golf[0.047]; Golf[0.047]; van[0.047]; Swart[0.046]; Mark[0.045]; Mark[0.045]; Mark[0.045]; South[0.045]; Fourie[0.045]; Schalk[0.045]; Namibia[0.044]; Namibia[0.044]; Steve[0.044]; Terblanche[0.044]; Andrew[0.044]; Andrew[0.044]; African[0.044]; second[0.044]; Nick[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Namibia ==> ENTITY: \u001b[32mNamibia\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.368:-0.368[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZimbabwe[0.047]; Zimbabwe[0.047]; Zimbabwe[0.047]; Namibia[0.047]; South[0.045]; African[0.045]; Mark[0.041]; Mark[0.041]; Mark[0.041]; Brazil[0.041]; Steve[0.041]; Steve[0.041]; Andrew[0.041]; Andrew[0.041]; Cayeux[0.041]; Brett[0.040]; Fourie[0.040]; Swart[0.040]; Botes[0.039]; Schalk[0.039]; Ian[0.039]; Justin[0.039]; unless[0.039]; Greg[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Brett Liddle\u001b[39m ==> ENTITY: \u001b[32mBrett Liddle\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.047]; Cayeux[0.045]; African[0.045]; Murless[0.043]; Hennie[0.042]; Golf[0.042]; Terblanche[0.042]; Ian[0.042]; Nick[0.042]; Trevor[0.041]; Mark[0.041]; Mark[0.041]; Mark[0.041]; Andrew[0.041]; Andrew[0.041]; round[0.041]; Fourie[0.041]; Botes[0.041]; Schalk[0.041]; Swart[0.040]; Open[0.040]; Park[0.040]; Merwe[0.040]; Justin[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Namibia ==> ENTITY: \u001b[32mNamibia\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.368:-0.368[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZimbabwe[0.047]; Zimbabwe[0.047]; Zimbabwe[0.047]; Zimbabwe[0.047]; Zimbabwe[0.047]; Namibia[0.047]; South[0.045]; African[0.045]; round[0.043]; scores[0.043]; second[0.043]; Brazil[0.042]; Terblanche[0.042]; Harare[0.042]; Round[0.042]; Fourie[0.042]; Mark[0.041]; Mark[0.041]; Mark[0.041]; Swart[0.041]; Merwe[0.041]; Steve[0.041]; Steve[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nick Price\u001b[39m ==> ENTITY: \u001b[32mNick Price\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.213:0.213[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.044]; Open[0.044]; round[0.044]; Golf[0.044]; Golf[0.044]; Murless[0.043]; par[0.043]; Greg[0.043]; Cayeux[0.042]; Mark[0.042]; Mark[0.042]; Mark[0.042]; Steve[0.042]; Steve[0.042]; second[0.041]; Brett[0.041]; Trevor[0.040]; Justin[0.040]; African[0.040]; South[0.040]; scores[0.040]; Andrew[0.039]; Andrew[0.039]; Schalk[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mark McNulty\u001b[39m ==> ENTITY: \u001b[32mMark McNulty\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCayeux[0.044]; Open[0.044]; Open[0.044]; Zimbabwe[0.043]; Zimbabwe[0.043]; Zimbabwe[0.043]; Zimbabwe[0.043]; Zimbabwe[0.043]; round[0.043]; Golf[0.043]; Golf[0.043]; second[0.043]; African[0.041]; Round[0.041]; Harare[0.040]; Murless[0.040]; par[0.040]; Schalk[0.040]; Merwe[0.040]; der[0.040]; Scores[0.039]; Namibia[0.039]; Namibia[0.039]; South[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Adilson da Silva\u001b[39m ==> ENTITY: \u001b[32mAdilson da Silva\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.046]; Open[0.045]; Murless[0.044]; South[0.044]; Golf[0.044]; Cayeux[0.043]; African[0.043]; par[0.042]; Schalk[0.041]; Merwe[0.041]; Club[0.041]; der[0.041]; Hennie[0.041]; Namibia[0.041]; Namibia[0.041]; Botes[0.041]; Zimbabwe[0.041]; Zimbabwe[0.041]; Zimbabwe[0.041]; Zimbabwe[0.041]; Des[0.040]; Fourie[0.040]; Brett[0.040]; Daniels[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Desvonde Botes\u001b[39m ==> ENTITY: \u001b[32mDesvonde Botes\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.046]; Open[0.046]; Golf[0.046]; second[0.044]; Murless[0.044]; Cayeux[0.044]; African[0.044]; round[0.043]; South[0.042]; Schalk[0.041]; Merwe[0.041]; Hennie[0.041]; der[0.040]; par[0.040]; Fourie[0.040]; Des[0.040]; Brett[0.040]; Brazil[0.040]; Namibia[0.040]; Namibia[0.040]; Swart[0.040]; scores[0.040]; Scores[0.040]; Club[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zimbabwe ==> ENTITY: \u001b[32mZimbabwe\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.189:0.189[0]; log p(e|m)= -0.367:-0.367[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZimbabwe[0.049]; Zimbabwe[0.049]; Zimbabwe[0.049]; Zimbabwe[0.049]; South[0.043]; Harare[0.043]; African[0.043]; Namibia[0.042]; Namibia[0.042]; Swart[0.041]; Ian[0.040]; second[0.040]; Brazil[0.040]; Cayeux[0.040]; Golf[0.039]; Golf[0.039]; scores[0.039]; Brett[0.039]; Friday[0.039]; Fourie[0.039]; Steve[0.039]; Steve[0.039]; Botes[0.039]; Trevor[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zimbabwe Open\u001b[39m ==> ENTITY: \u001b[32mZimbabwe Open\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.046]; Golf[0.044]; Golf[0.044]; Cayeux[0.044]; Murless[0.043]; round[0.042]; Steve[0.041]; Zimbabwe[0.041]; Zimbabwe[0.041]; Zimbabwe[0.041]; Zimbabwe[0.041]; Mark[0.041]; Mark[0.041]; Mark[0.041]; Hennie[0.041]; Nick[0.041]; African[0.041]; Andrew[0.041]; Andrew[0.041]; Greg[0.040]; Justin[0.040]; Schalk[0.040]; Merwe[0.040]; South[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South African ==> ENTITY: \u001b[32mSouth Africa\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.044]; Open[0.044]; Zimbabwe[0.044]; Zimbabwe[0.044]; Zimbabwe[0.044]; Zimbabwe[0.044]; Zimbabwe[0.044]; round[0.043]; second[0.042]; Namibia[0.042]; Namibia[0.042]; Golf[0.042]; Golf[0.042]; Murless[0.042]; Cayeux[0.041]; Park[0.041]; Brett[0.041]; Friday[0.040]; Brazil[0.040]; Club[0.040]; Harare[0.040]; Schalk[0.040]; scores[0.039]; Swart[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Clinton Whitelaw\u001b[39m ==> ENTITY: \u001b[32mClinton Whitelaw\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrican[0.046]; Cayeux[0.045]; Open[0.045]; round[0.044]; McNulty[0.043]; South[0.043]; Murless[0.042]; second[0.042]; Friday[0.042]; Golf[0.041]; par[0.041]; Harare[0.041]; Zimbabwe[0.041]; Zimbabwe[0.041]; Zimbabwe[0.041]; Zimbabwe[0.041]; Schalk[0.041]; Merwe[0.041]; Namibia[0.041]; Namibia[0.041]; der[0.040]; Hugh[0.040]; Des[0.040]; Brett[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Justin Hobday\u001b[39m ==> ENTITY: \u001b[32mJustin Hobday\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCayeux[0.046]; round[0.045]; Open[0.044]; Open[0.044]; African[0.043]; Golf[0.043]; Golf[0.043]; South[0.043]; Nick[0.042]; second[0.042]; Trevor[0.042]; Murless[0.041]; Ian[0.041]; Steve[0.041]; Steve[0.041]; Mark[0.040]; Mark[0.040]; Mark[0.040]; McNulty[0.040]; scores[0.040]; Greg[0.040]; Schalk[0.040]; Merwe[0.040]; par[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zimbabwe ==> ENTITY: \u001b[32mZimbabwe\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.190:0.190[0]; log p(e|m)= -0.367:-0.367[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZimbabwe[0.051]; Zimbabwe[0.051]; Zimbabwe[0.051]; Zimbabwe[0.051]; South[0.045]; Harare[0.045]; African[0.045]; Namibia[0.043]; Namibia[0.043]; Swart[0.042]; second[0.042]; Brazil[0.042]; Cayeux[0.042]; Golf[0.041]; Golf[0.041]; scores[0.041]; Brett[0.041]; Friday[0.041]; Fourie[0.040]; Steve[0.040]; Steve[0.040]; Botes[0.040]; Trevor[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Trevor Dodds\u001b[39m ==> ENTITY: \u001b[32mTrevor Dodds\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCayeux[0.048]; Zimbabwe[0.044]; Zimbabwe[0.044]; Zimbabwe[0.044]; Namibia[0.043]; Namibia[0.043]; Botes[0.043]; Murless[0.043]; African[0.042]; Schalk[0.042]; Merwe[0.042]; der[0.041]; Ian[0.041]; Des[0.041]; South[0.041]; Brett[0.041]; Greg[0.040]; Hugh[0.040]; Hennie[0.040]; da[0.040]; Swart[0.040]; Vuuren[0.039]; Don[0.039]; stated[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hugh Baiocchi\u001b[39m ==> ENTITY: \u001b[32mHugh Baiocchi\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.127:0.127[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nround[0.046]; Golf[0.045]; Cayeux[0.045]; second[0.044]; South[0.044]; African[0.043]; Murless[0.042]; Open[0.042]; Schalk[0.042]; Merwe[0.042]; der[0.041]; Des[0.041]; Brett[0.041]; McNulty[0.041]; Justin[0.041]; Hennie[0.041]; par[0.041]; scores[0.041]; Club[0.041]; Park[0.040]; da[0.040]; Trevor[0.040]; Botes[0.040]; Vuuren[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zimbabwe Open\u001b[39m ==> ENTITY: \u001b[32mZimbabwe Open\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.046]; Golf[0.044]; Golf[0.044]; Cayeux[0.044]; Murless[0.043]; round[0.042]; Steve[0.041]; Zimbabwe[0.041]; Zimbabwe[0.041]; Zimbabwe[0.041]; Zimbabwe[0.041]; Mark[0.041]; Mark[0.041]; Mark[0.041]; Hennie[0.041]; Nick[0.041]; African[0.041]; Andrew[0.041]; Andrew[0.041]; Greg[0.040]; Justin[0.040]; Schalk[0.040]; Merwe[0.040]; South[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.052:0.052[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.048]; second[0.046]; South[0.045]; Steve[0.044]; Steve[0.044]; Mark[0.044]; Mark[0.044]; Mark[0.044]; round[0.044]; Park[0.044]; Price[0.043]; Adilson[0.043]; Club[0.043]; Greg[0.043]; Round[0.043]; Open[0.043]; Open[0.043]; Chapman[0.043]; Scores[0.042]; Friday[0.042]; Reid[0.042]; African[0.042]; Dion[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.032:0.032[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.047]; Steve[0.044]; Steve[0.044]; Mark[0.043]; Mark[0.043]; Mark[0.043]; Park[0.043]; Price[0.043]; Adilson[0.042]; Greg[0.042]; Reid[0.041]; Dion[0.041]; Ian[0.041]; stated[0.041]; McNulty[0.040]; Don[0.040]; Andrew[0.040]; Andrew[0.040]; Nick[0.040]; Robertson[0.040]; Pitts[0.040]; Sammy[0.040]; Justin[0.040]; Liddle[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Des Terblanche\u001b[39m ==> ENTITY: \u001b[32mDes Terblanche\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCayeux[0.050]; Murless[0.049]; South[0.048]; Golf[0.048]; Golf[0.048]; African[0.048]; Open[0.048]; Open[0.048]; Hennie[0.045]; Namibia[0.044]; Namibia[0.044]; Botes[0.044]; van[0.044]; van[0.044]; Schalk[0.044]; Merwe[0.044]; Trevor[0.044]; der[0.043]; round[0.043]; McNulty[0.043]; second[0.043]; Brett[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazil ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.029:0.029[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAdilson[0.046]; Silva[0.045]; South[0.043]; African[0.043]; Mark[0.043]; Mark[0.043]; Mark[0.043]; Park[0.042]; Club[0.042]; Don[0.042]; Nick[0.042]; Steve[0.041]; Steve[0.041]; Andrew[0.041]; Andrew[0.041]; Greg[0.041]; Justin[0.041]; Friday[0.041]; Brett[0.041]; da[0.040]; Golf[0.040]; Clinton[0.040]; Vuuren[0.040]; Ian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zimbabwe ==> ENTITY: \u001b[32mZimbabwe\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.189:0.189[0]; log p(e|m)= -0.367:-0.367[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZimbabwe[0.049]; Zimbabwe[0.049]; Zimbabwe[0.049]; Zimbabwe[0.049]; South[0.043]; Harare[0.043]; African[0.043]; Namibia[0.042]; Namibia[0.042]; Swart[0.041]; Ian[0.040]; second[0.040]; Brazil[0.040]; Cayeux[0.040]; Golf[0.039]; Golf[0.039]; scores[0.039]; Brett[0.039]; Friday[0.039]; Fourie[0.039]; Steve[0.039]; Steve[0.039]; Botes[0.039]; Trevor[0.038]; \t\n\r [=======================================>......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 2s892ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3918/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1366testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Glasgow ==> ENTITY: \u001b[32mGlasgow\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.151:0.151[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScottish[0.047]; Scottish[0.047]; Aberdeen[0.047]; Dundee[0.045]; matches[0.044]; Ross[0.043]; Dunfermline[0.043]; Kilmarnock[0.042]; played[0.042]; Celtic[0.042]; Motherwell[0.041]; Rangers[0.041]; Hibernian[0.041]; Saturday[0.041]; penalty[0.040]; Soccer[0.040]; Premier[0.040]; Raith[0.040]; Hearts[0.039]; premier[0.039]; McGinlay[0.039]; division[0.039]; Attendance[0.039]; Ferguson[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hearts ==> ENTITY: \u001b[32mHeart of Midlothian F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.208:0.208[0]; log p(e|m)= -0.298:-0.298[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAberdeen[0.044]; Dundee[0.044]; matches[0.044]; Glasgow[0.044]; Hibernian[0.043]; Rangers[0.043]; Celtic[0.043]; played[0.043]; Scottish[0.043]; Scottish[0.043]; Dunfermline[0.042]; Motherwell[0.042]; Kilmarnock[0.042]; Soccer[0.041]; Raith[0.041]; Ross[0.041]; McGinlay[0.040]; McCoist[0.040]; Premier[0.040]; Ferguson[0.040]; Saturday[0.040]; United[0.040]; premier[0.039]; Division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Celtic ==> ENTITY: \u001b[32mCeltic F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.173:0.173[0]; log p(e|m)= -0.978:-0.978[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGlasgow[0.045]; Dunfermline[0.044]; matches[0.044]; Scottish[0.044]; Scottish[0.044]; Dundee[0.044]; Aberdeen[0.044]; Kilmarnock[0.043]; Rangers[0.043]; played[0.042]; Hibernian[0.042]; Soccer[0.042]; Motherwell[0.042]; Hearts[0.040]; McCoist[0.040]; Raith[0.040]; Ferguson[0.040]; Division[0.040]; Ross[0.040]; division[0.040]; McGinlay[0.040]; United[0.039]; Saturday[0.039]; Millar[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Raith\u001b[39m ==> ENTITY: \u001b[32mRaith Rovers F.C.\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.223:0.223[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDundee[0.045]; Dunfermline[0.044]; matches[0.044]; Aberdeen[0.044]; Kilmarnock[0.043]; Rangers[0.043]; Glasgow[0.043]; Scottish[0.043]; Scottish[0.043]; Celtic[0.043]; Motherwell[0.042]; played[0.042]; Hibernian[0.042]; Soccer[0.041]; McGinlay[0.041]; Hearts[0.041]; McCoist[0.041]; Ferguson[0.041]; Ross[0.040]; Division[0.040]; division[0.039]; Halftime[0.039]; United[0.038]; Saturday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Laudrup ==> ENTITY: \u001b[32mBrian Laudrup\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.118:0.118[0]; log p(e|m)= -0.732:-0.732[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.053]; McCoist[0.046]; matches[0.046]; played[0.046]; Rangers[0.044]; Aberdeen[0.042]; Celtic[0.042]; Scottish[0.041]; Scottish[0.041]; Halftime[0.041]; Hearts[0.041]; Glasgow[0.041]; Hibernian[0.040]; Raith[0.040]; Ferguson[0.040]; Windass[0.040]; McGinlay[0.040]; Ross[0.040]; Dundee[0.040]; Motherwell[0.039]; Kilmarnock[0.039]; Summaries[0.039]; Summaries[0.039]; penalty[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kilmarnock ==> ENTITY: \u001b[32mKilmarnock F.C.\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.223:0.223[0]; log p(e|m)= -1.431:-1.431[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAberdeen[0.045]; Dundee[0.045]; played[0.044]; matches[0.044]; Celtic[0.044]; Dunfermline[0.043]; Glasgow[0.043]; Hibernian[0.043]; Scottish[0.043]; Scottish[0.043]; Motherwell[0.043]; Raith[0.042]; Rangers[0.042]; Hearts[0.041]; McCoist[0.041]; Ross[0.041]; Soccer[0.041]; McGinlay[0.040]; Ferguson[0.040]; United[0.040]; Premier[0.039]; Saturday[0.038]; Miller[0.038]; Division[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scottish ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.480:-0.480[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScottish[0.045]; Aberdeen[0.043]; Hibernian[0.043]; matches[0.043]; Celtic[0.043]; Dundee[0.043]; Dunfermline[0.043]; Rangers[0.043]; played[0.043]; Hearts[0.043]; Kilmarnock[0.042]; Glasgow[0.042]; Ross[0.042]; division[0.041]; Motherwell[0.041]; Soccer[0.041]; Raith[0.041]; Ferguson[0.041]; McCoist[0.040]; McGinlay[0.040]; Division[0.040]; Davies[0.039]; premier[0.039]; Premier[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dunfermline ==> ENTITY: \u001b[32mDunfermline Athletic F.C.\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.219:0.219[0]; log p(e|m)= -1.995:-1.995[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAberdeen[0.046]; Dundee[0.045]; played[0.044]; Kilmarnock[0.043]; matches[0.043]; Celtic[0.043]; Rangers[0.043]; Scottish[0.043]; Scottish[0.043]; Glasgow[0.043]; Hibernian[0.043]; Motherwell[0.043]; Hearts[0.042]; Raith[0.042]; Soccer[0.042]; Ross[0.040]; United[0.040]; McGinlay[0.040]; Division[0.039]; McCoist[0.039]; Premier[0.039]; Ferguson[0.039]; division[0.038]; Saturday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rangers ==> ENTITY: \u001b[32mRangers F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.193:0.193[0]; log p(e|m)= -0.302:-0.302[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; Soccer[0.044]; Aberdeen[0.044]; Glasgow[0.044]; Scottish[0.044]; Scottish[0.044]; Dundee[0.044]; Celtic[0.043]; played[0.043]; Dunfermline[0.042]; Hearts[0.042]; Hibernian[0.042]; Kilmarnock[0.042]; McCoist[0.041]; Motherwell[0.041]; Raith[0.041]; Ferguson[0.040]; Saturday[0.040]; Ross[0.040]; United[0.040]; McGinlay[0.039]; Premier[0.039]; division[0.039]; Division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dundee United\u001b[39m ==> ENTITY: \u001b[32mDundee United F.C.\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAberdeen[0.045]; matches[0.043]; Hibernian[0.043]; Glasgow[0.043]; Celtic[0.043]; Dunfermline[0.043]; Kilmarnock[0.043]; Rangers[0.043]; Scottish[0.042]; Scottish[0.042]; Hearts[0.042]; Ferguson[0.042]; played[0.042]; Ross[0.042]; Raith[0.041]; Motherwell[0.041]; Soccer[0.041]; Premier[0.040]; Division[0.040]; McCoist[0.040]; Summaries[0.040]; Summaries[0.040]; Miller[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rowson\u001b[39m ==> ENTITY: \u001b[32mDavid Rowson\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRangers[0.044]; Aberdeen[0.044]; Glasgow[0.043]; Kilmarnock[0.043]; Dundee[0.043]; Hibernian[0.042]; Halftime[0.042]; played[0.042]; Motherwell[0.042]; matches[0.042]; Celtic[0.042]; Scottish[0.042]; Scottish[0.042]; Dunfermline[0.042]; Windass[0.041]; Ross[0.041]; Soccer[0.041]; Division[0.041]; Summaries[0.041]; Summaries[0.041]; Wright[0.040]; Saturday[0.040]; Raith[0.040]; Hearts[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = McCoist\u001b[39m ==> ENTITY: \u001b[32mAlly McCoist\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.132:0.132[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; Rangers[0.044]; played[0.044]; Aberdeen[0.043]; Ross[0.042]; Soccer[0.042]; Dundee[0.042]; Hibernian[0.042]; Dunfermline[0.042]; Celtic[0.042]; Scottish[0.042]; Scottish[0.042]; Ferguson[0.042]; Halftime[0.042]; Kilmarnock[0.041]; Glasgow[0.041]; Motherwell[0.041]; Hearts[0.041]; McGinlay[0.041]; Raith[0.041]; Summaries[0.040]; Summaries[0.040]; United[0.040]; Saturday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scottish ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.480:-0.480[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScottish[0.045]; Aberdeen[0.043]; Hibernian[0.043]; matches[0.043]; Celtic[0.043]; Dundee[0.043]; Dunfermline[0.043]; Rangers[0.043]; played[0.043]; Hearts[0.043]; Kilmarnock[0.042]; Glasgow[0.042]; Ross[0.042]; division[0.041]; Motherwell[0.041]; Soccer[0.041]; Raith[0.041]; Ferguson[0.041]; McCoist[0.040]; McGinlay[0.040]; Division[0.040]; Davies[0.039]; premier[0.039]; Premier[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Motherwell ==> ENTITY: \u001b[32mMotherwell F.C.\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.210:0.210[0]; log p(e|m)= -0.335:-0.335[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScottish[0.044]; Scottish[0.044]; Aberdeen[0.044]; Dundee[0.044]; Premier[0.043]; Glasgow[0.043]; Hibernian[0.043]; Kilmarnock[0.043]; Dunfermline[0.043]; matches[0.043]; played[0.042]; Celtic[0.042]; Rangers[0.042]; Hearts[0.041]; Division[0.041]; Ferguson[0.041]; McCoist[0.041]; McGinlay[0.040]; division[0.040]; Raith[0.040]; Ross[0.040]; Soccer[0.040]; United[0.039]; Saturday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = McGinlay ==> ENTITY: \u001b[32mPat McGinlay\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.182:0.182[0]; log p(e|m)= -0.962:-0.962[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCeltic[0.045]; Hibernian[0.044]; Hearts[0.044]; matches[0.043]; Motherwell[0.043]; played[0.043]; Dundee[0.043]; Aberdeen[0.042]; Scottish[0.042]; Scottish[0.042]; Rangers[0.042]; Kilmarnock[0.042]; Dunfermline[0.042]; Glasgow[0.042]; Soccer[0.041]; McCoist[0.041]; United[0.041]; Ferguson[0.041]; Raith[0.041]; Division[0.040]; Ross[0.040]; Halftime[0.039]; division[0.038]; Davies[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Aberdeen ==> ENTITY: \u001b[32mAberdeen F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.184:0.184[0]; log p(e|m)= -1.448:-1.448[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScottish[0.045]; Scottish[0.045]; matches[0.044]; Dundee[0.044]; played[0.044]; Glasgow[0.044]; Dunfermline[0.043]; Hibernian[0.042]; Motherwell[0.042]; Kilmarnock[0.042]; Rangers[0.042]; Celtic[0.042]; Hearts[0.041]; Ross[0.041]; Soccer[0.041]; Raith[0.041]; Ferguson[0.041]; McGinlay[0.040]; Premier[0.040]; McCoist[0.040]; United[0.039]; Saturday[0.039]; Hay[0.038]; penalty[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hibernian\u001b[39m ==> ENTITY: \u001b[32mHibernian F.C.\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAberdeen[0.045]; Dundee[0.044]; Celtic[0.044]; Glasgow[0.043]; matches[0.043]; Dunfermline[0.043]; Kilmarnock[0.043]; Scottish[0.043]; Scottish[0.043]; Rangers[0.043]; Motherwell[0.042]; played[0.042]; Hearts[0.042]; McCoist[0.042]; Raith[0.041]; Soccer[0.041]; Ross[0.040]; McGinlay[0.040]; division[0.040]; Ferguson[0.040]; Division[0.039]; United[0.039]; Premier[0.039]; Laudrup[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Davies ==> ENTITY: \u001b[32mBilly Davies\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.107:0.107[0]; log p(e|m)= -3.352:-3.352[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUnited[0.045]; Soccer[0.044]; matches[0.044]; played[0.044]; Saturday[0.044]; Miller[0.043]; Ferguson[0.042]; Aberdeen[0.042]; penalty[0.042]; Premier[0.041]; division[0.041]; Windass[0.041]; Rangers[0.041]; Wright[0.041]; Scottish[0.041]; Scottish[0.041]; Division[0.041]; Ross[0.041]; Celtic[0.040]; Rowson[0.040]; Jackson[0.040]; McGinlay[0.040]; Dundee[0.040]; Glasgow[0.040]; \t\n\r [=======================================>......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 2s779ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3941/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1232testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.237:0.237[0]; local()= 0.036:0.036[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYork[0.122]; York[0.122]; York[0.122]; fixtures[0.111]; fixtures[0.111]; reported[0.108]; Desk[0.108]; Dec[0.100]; Commodities[0.098]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.237:0.237[0]; local()= 0.036:0.036[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYork[0.122]; York[0.122]; York[0.122]; fixtures[0.111]; fixtures[0.111]; reported[0.108]; Desk[0.108]; Dec[0.100]; Commodities[0.098]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.237:0.237[0]; local()= 0.036:0.036[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYork[0.122]; York[0.122]; York[0.122]; fixtures[0.111]; fixtures[0.111]; reported[0.108]; Desk[0.108]; Dec[0.100]; Commodities[0.098]; \t\n\r [=======================================>......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 2s754ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3944/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1341testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = WTO\u001b[39m ==> ENTITY: \u001b[32mWorld Trade Organization\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrade[0.046]; trade[0.046]; trade[0.046]; global[0.043]; global[0.043]; World[0.043]; meeting[0.043]; Organisation[0.042]; economies[0.042]; Trade[0.042]; key[0.041]; Singapore[0.041]; Singapore[0.041]; Singapore[0.041]; Singapore[0.041]; Singapore[0.041]; economy[0.041]; world[0.040]; hosting[0.040]; said[0.040]; said[0.040]; growth[0.040]; lifeblood[0.039]; accounting[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.180:0.180[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.050]; Singapore[0.050]; Singapore[0.050]; Singapore[0.050]; Singapore[0.050]; tourism[0.042]; Singaporean[0.041]; government[0.040]; region[0.040]; play[0.040]; ahead[0.040]; public[0.040]; industry[0.040]; economy[0.040]; economy[0.040]; Tan[0.040]; Tan[0.040]; probably[0.039]; workforce[0.039]; commercial[0.039]; gain[0.039]; electorate[0.039]; educated[0.039]; quite[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = WTO\u001b[39m ==> ENTITY: \u001b[32mWorld Trade Organization\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.246:0.246[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninternational[0.047]; trade[0.046]; trade[0.046]; trade[0.046]; trade[0.046]; countries[0.045]; services[0.045]; national[0.044]; implementation[0.043]; European[0.043]; policies[0.043]; global[0.043]; Agreement[0.043]; meeting[0.043]; meeting[0.043]; Policy[0.042]; Policy[0.042]; developing[0.042]; conference[0.042]; sector[0.042]; Trade[0.042]; Tariffs[0.041]; General[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.163:0.163[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.050]; Singapore[0.050]; Singapore[0.050]; Singapore[0.050]; tourism[0.042]; Singaporean[0.042]; government[0.041]; region[0.041]; play[0.041]; ahead[0.040]; public[0.040]; industry[0.040]; economy[0.040]; economy[0.040]; Tan[0.040]; Tan[0.040]; probably[0.039]; workforce[0.039]; commercial[0.039]; gain[0.039]; electorate[0.039]; educated[0.039]; quite[0.039]; gains[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = WTO\u001b[39m ==> ENTITY: \u001b[32mWorld Trade Organization\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.225:0.225[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninternational[0.046]; trade[0.044]; trade[0.044]; trade[0.044]; trade[0.044]; countries[0.044]; national[0.043]; global[0.042]; global[0.042]; meeting[0.042]; meeting[0.042]; Policy[0.041]; conference[0.041]; Trade[0.041]; domestic[0.040]; Singapore[0.040]; Singapore[0.040]; Singapore[0.040]; Singapore[0.040]; Singapore[0.040]; Geneva[0.040]; support[0.040]; fostering[0.040]; hosting[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.174:0.174[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.050]; Singapore[0.050]; Singapore[0.050]; Singapore[0.050]; Singapore[0.050]; global[0.041]; global[0.041]; play[0.040]; domestic[0.040]; house[0.040]; countries[0.040]; key[0.039]; nearly[0.039]; World[0.039]; growth[0.039]; expected[0.039]; trade[0.039]; trade[0.039]; trade[0.039]; economy[0.039]; going[0.039]; city[0.039]; caning[0.039]; winning[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.208:0.208[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.049]; Singapore[0.049]; Singapore[0.049]; Singapore[0.049]; Singapore[0.049]; Singapore[0.049]; Singaporean[0.041]; region[0.041]; Tan[0.041]; Tan[0.041]; tourism[0.040]; services[0.040]; commercial[0.039]; public[0.039]; countries[0.039]; probably[0.039]; market[0.038]; trade[0.038]; trade[0.038]; economy[0.038]; battle[0.038]; West[0.038]; additional[0.038]; policies[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = WTO\u001b[39m ==> ENTITY: \u001b[32mWorld Trade Organization\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrade[0.046]; trade[0.046]; trade[0.046]; global[0.043]; global[0.043]; World[0.043]; meeting[0.042]; Organisation[0.042]; economies[0.042]; Trade[0.042]; key[0.041]; domestic[0.041]; Singapore[0.041]; Singapore[0.041]; Singapore[0.041]; Singapore[0.041]; Singapore[0.041]; economy[0.041]; world[0.040]; hosting[0.040]; said[0.040]; said[0.040]; growth[0.040]; lifeblood[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.177:0.177[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.050]; Singapore[0.050]; Singapore[0.050]; Singapore[0.050]; national[0.044]; National[0.041]; global[0.041]; bid[0.041]; international[0.041]; caning[0.040]; Tan[0.040]; countries[0.040]; head[0.040]; said[0.039]; said[0.039]; said[0.039]; market[0.039]; expected[0.039]; trade[0.039]; trade[0.039]; trade[0.039]; trade[0.039]; commitment[0.039]; going[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = European Union\u001b[39m ==> ENTITY: \u001b[32mEuropean Union\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.184:0.184[0]; log p(e|m)= -0.006:-0.006[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountries[0.044]; trade[0.044]; trade[0.044]; trade[0.044]; trade[0.044]; trade[0.044]; international[0.043]; global[0.042]; policies[0.042]; commitments[0.042]; developing[0.042]; sector[0.042]; national[0.042]; implementation[0.041]; investment[0.041]; Agreement[0.041]; Policy[0.040]; Trade[0.040]; ministers[0.040]; talks[0.040]; powers[0.040]; meeting[0.040]; meeting[0.040]; meeting[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.154:0.154[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.053]; Singapore[0.053]; Singapore[0.053]; national[0.046]; National[0.044]; bid[0.043]; international[0.043]; caning[0.043]; domestic[0.042]; Tan[0.042]; countries[0.042]; head[0.042]; nearly[0.042]; said[0.042]; said[0.042]; said[0.042]; commitment[0.041]; going[0.041]; hosting[0.041]; conference[0.041]; ought[0.041]; senior[0.041]; symbolism[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.191:0.191[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.052]; Singapore[0.052]; Singapore[0.052]; Singapore[0.052]; Singapore[0.052]; international[0.043]; caning[0.042]; domestic[0.042]; countries[0.041]; nearly[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; going[0.041]; ought[0.040]; senior[0.040]; bid[0.040]; held[0.040]; symbolism[0.040]; trade[0.040]; trade[0.040]; trade[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = WTO\u001b[39m ==> ENTITY: \u001b[32mWorld Trade Organization\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntourism[0.046]; trade[0.045]; trade[0.045]; economic[0.044]; economic[0.044]; meeting[0.042]; meeting[0.042]; meeting[0.042]; agenda[0.042]; issues[0.042]; issues[0.042]; issues[0.042]; region[0.041]; concerned[0.041]; view[0.041]; Singapore[0.040]; Singapore[0.040]; Singapore[0.040]; Singapore[0.040]; Singapore[0.040]; economy[0.040]; economy[0.040]; commercial[0.040]; freer[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.175:0.175[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.050]; Singapore[0.050]; Singapore[0.050]; Singapore[0.050]; Singapore[0.050]; global[0.041]; global[0.041]; play[0.040]; domestic[0.040]; house[0.040]; countries[0.040]; key[0.039]; nearly[0.039]; World[0.039]; growth[0.039]; expected[0.039]; trade[0.039]; trade[0.039]; trade[0.039]; economy[0.039]; city[0.039]; caning[0.039]; winning[0.039]; officials[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Uruguay ==> ENTITY: \u001b[32mUruguay\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.446:-0.446[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnational[0.048]; international[0.048]; trade[0.044]; trade[0.044]; trade[0.044]; trade[0.044]; trade[0.044]; European[0.042]; countries[0.042]; free[0.041]; free[0.041]; free[0.041]; conference[0.041]; global[0.040]; June[0.040]; National[0.040]; sector[0.040]; head[0.040]; state[0.040]; open[0.040]; open[0.040]; opening[0.039]; host[0.039]; hosting[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.146:0.146[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.051]; Singapore[0.051]; Singapore[0.051]; Singaporean[0.043]; government[0.042]; play[0.042]; ahead[0.041]; gain[0.041]; public[0.041]; economic[0.041]; economic[0.041]; Tan[0.041]; probably[0.040]; trade[0.040]; economy[0.040]; economy[0.040]; electorate[0.040]; educated[0.040]; important[0.039]; April[0.039]; general[0.039]; workforce[0.039]; run[0.039]; gains[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.138:0.138[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.051]; Singapore[0.051]; Singapore[0.051]; Singapore[0.051]; global[0.042]; global[0.042]; play[0.041]; domestic[0.041]; key[0.040]; nearly[0.040]; World[0.040]; world[0.040]; growth[0.040]; trade[0.040]; trade[0.040]; trade[0.040]; economy[0.040]; winning[0.039]; hosting[0.039]; stands[0.039]; campaign[0.039]; Hussain[0.039]; prestige[0.039]; times[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.187:0.187[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.050]; Singapore[0.050]; Singapore[0.050]; Singapore[0.050]; Singapore[0.050]; Singaporean[0.042]; region[0.041]; Tan[0.041]; Tan[0.041]; tourism[0.041]; ahead[0.040]; commercial[0.040]; economic[0.040]; countries[0.039]; probably[0.039]; trade[0.039]; trade[0.039]; economy[0.039]; battle[0.039]; West[0.039]; educated[0.039]; additional[0.038]; quite[0.038]; important[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.182:0.182[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.050]; Singapore[0.050]; Singapore[0.050]; Singapore[0.050]; Singapore[0.050]; national[0.044]; National[0.041]; global[0.041]; international[0.041]; caning[0.040]; Tan[0.040]; head[0.039]; said[0.039]; said[0.039]; said[0.039]; market[0.039]; market[0.039]; trade[0.039]; trade[0.039]; trade[0.039]; trade[0.039]; commitment[0.039]; going[0.039]; Uruguay[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mickey Kantor\u001b[39m ==> ENTITY: \u001b[32mMickey Kantor\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninternational[0.045]; American[0.044]; trade[0.044]; trade[0.044]; trade[0.044]; trade[0.044]; national[0.044]; domestic[0.042]; Policy[0.042]; global[0.042]; global[0.042]; Trade[0.042]; senior[0.042]; conference[0.042]; economist[0.041]; Studies[0.040]; said[0.040]; said[0.040]; said[0.040]; December[0.040]; trading[0.040]; Representative[0.040]; commitment[0.040]; accounting[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = WTO\u001b[39m ==> ENTITY: \u001b[32mWorld Trade Organization\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.225:0.225[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrade[0.045]; trade[0.045]; trade[0.045]; countries[0.045]; global[0.042]; global[0.042]; World[0.042]; meeting[0.042]; meeting[0.042]; meeting[0.042]; Organisation[0.042]; economies[0.041]; Trade[0.041]; Trade[0.041]; key[0.041]; domestic[0.041]; Singapore[0.040]; Singapore[0.040]; Singapore[0.040]; Singapore[0.040]; Singapore[0.040]; Singapore[0.040]; economy[0.040]; world[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.050]; Singapore[0.050]; Singapore[0.050]; Singapore[0.050]; national[0.044]; implementation[0.043]; region[0.042]; global[0.041]; review[0.040]; services[0.040]; additional[0.040]; commercial[0.040]; Tan[0.040]; countries[0.039]; head[0.039]; speed[0.039]; National[0.039]; market[0.039]; market[0.039]; trade[0.039]; trade[0.039]; trade[0.039]; trade[0.039]; trade[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = WTO\u001b[39m ==> ENTITY: \u001b[32mWorld Trade Organization\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.216:0.216[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninternational[0.046]; trade[0.045]; trade[0.045]; trade[0.045]; countries[0.044]; national[0.043]; global[0.042]; meeting[0.042]; meeting[0.042]; Policy[0.042]; Policy[0.042]; conference[0.041]; Trade[0.041]; domestic[0.040]; Singapore[0.040]; Singapore[0.040]; Singapore[0.040]; Singapore[0.040]; Singapore[0.040]; Geneva[0.040]; support[0.040]; fostering[0.040]; National[0.040]; Business[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = WTO\u001b[39m ==> ENTITY: \u001b[32mWorld Trade Organization\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.249:0.249[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninternational[0.045]; trade[0.044]; trade[0.044]; trade[0.044]; trade[0.044]; trade[0.044]; countries[0.043]; services[0.043]; national[0.042]; implementation[0.041]; European[0.041]; policies[0.041]; global[0.041]; Agreement[0.041]; meeting[0.041]; meeting[0.041]; meeting[0.041]; Policy[0.041]; rights[0.040]; developing[0.040]; sector[0.040]; Trade[0.040]; Tariffs[0.040]; General[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Geneva ==> ENTITY: \u001b[32mGeneva\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.092:0.092[0]; log p(e|m)= -0.247:-0.247[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngoing[0.043]; national[0.043]; international[0.043]; National[0.043]; nearly[0.043]; American[0.043]; expected[0.042]; state[0.042]; state[0.042]; city[0.042]; city[0.042]; ought[0.042]; officials[0.042]; stand[0.041]; continued[0.041]; host[0.041]; great[0.041]; domestic[0.041]; said[0.041]; said[0.041]; said[0.041]; held[0.041]; ago[0.041]; da[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.077:0.077[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhost[0.045]; nearly[0.044]; American[0.044]; held[0.044]; hosting[0.043]; domestic[0.043]; play[0.042]; bid[0.042]; ought[0.042]; World[0.042]; conference[0.042]; world[0.041]; key[0.041]; officials[0.041]; expected[0.041]; countries[0.041]; Singapore[0.040]; Singapore[0.040]; Singapore[0.040]; Singapore[0.040]; Singapore[0.040]; trade[0.040]; trade[0.040]; trade[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.192:0.192[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.050]; Singapore[0.050]; Singapore[0.050]; Singapore[0.050]; Singapore[0.050]; Singaporean[0.042]; region[0.041]; Tan[0.041]; Tan[0.041]; tourism[0.041]; ahead[0.040]; commercial[0.040]; economic[0.040]; countries[0.039]; probably[0.039]; trade[0.039]; trade[0.039]; trade[0.039]; economy[0.039]; economy[0.039]; battle[0.039]; West[0.039]; educated[0.039]; additional[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.051]; Singapore[0.051]; Singapore[0.051]; Singapore[0.051]; global[0.042]; global[0.042]; play[0.041]; key[0.040]; World[0.040]; world[0.040]; growth[0.040]; trade[0.040]; trade[0.040]; trade[0.040]; economy[0.040]; winning[0.040]; hosting[0.039]; stands[0.039]; campaign[0.039]; Hussain[0.039]; prestige[0.039]; analysts[0.039]; economies[0.039]; role[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.138:0.138[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.051]; Singapore[0.051]; Singapore[0.051]; Singapore[0.051]; global[0.042]; global[0.042]; play[0.041]; domestic[0.040]; world[0.040]; key[0.040]; nearly[0.040]; World[0.040]; times[0.040]; growth[0.040]; trade[0.040]; trade[0.040]; trade[0.040]; economy[0.040]; winning[0.039]; stands[0.039]; campaign[0.039]; hosting[0.039]; Hussain[0.039]; prestige[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = National University of Singapore\u001b[39m ==> ENTITY: \u001b[32mNational University of Singapore\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.046]; Singapore[0.046]; Singapore[0.046]; Singapore[0.046]; Singapore[0.046]; Singapore[0.046]; Business[0.044]; international[0.044]; Institute[0.042]; national[0.042]; prestige[0.040]; Tan[0.040]; Studies[0.040]; European[0.040]; senior[0.040]; services[0.040]; commitments[0.039]; global[0.039]; globe[0.039]; hosting[0.039]; review[0.039]; Kong[0.039]; General[0.038]; conference[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.183:0.183[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.050]; Singapore[0.050]; Singapore[0.050]; Singapore[0.050]; Singapore[0.050]; region[0.042]; National[0.041]; tourism[0.041]; services[0.040]; commercial[0.040]; Tan[0.040]; Tan[0.040]; countries[0.039]; head[0.039]; market[0.039]; market[0.039]; trade[0.039]; trade[0.039]; trade[0.039]; trade[0.039]; trade[0.039]; battle[0.039]; Uruguay[0.039]; review[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Institute of Policy Studies ==> ENTITY: \u001b[32mInstitute of Policy Studies (Singapore)\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.177:0.177[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninternational[0.044]; Policy[0.044]; Singapore[0.044]; Singapore[0.044]; Singapore[0.044]; Singapore[0.044]; Singapore[0.044]; American[0.043]; global[0.043]; national[0.042]; support[0.041]; review[0.041]; commitment[0.041]; fostering[0.041]; University[0.041]; implementation[0.041]; Business[0.040]; members[0.040]; National[0.040]; meeting[0.040]; meeting[0.040]; meeting[0.040]; ought[0.039]; hosting[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = American ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.088:0.088[0]; log p(e|m)= -0.130:-0.130[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhost[0.046]; countries[0.044]; national[0.044]; international[0.044]; times[0.044]; nearly[0.043]; domestic[0.042]; disproportionately[0.042]; Mickey[0.041]; world[0.041]; said[0.041]; said[0.041]; said[0.041]; state[0.041]; state[0.041]; prestige[0.041]; opposition[0.041]; December[0.041]; trade[0.041]; trade[0.041]; trade[0.041]; conference[0.040]; officials[0.040]; hosting[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.167:0.167[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.052]; Singapore[0.052]; Singapore[0.052]; Singapore[0.052]; Singapore[0.052]; national[0.045]; implementation[0.045]; National[0.043]; international[0.042]; review[0.042]; services[0.042]; Tan[0.041]; countries[0.041]; head[0.041]; speed[0.041]; commitment[0.040]; battle[0.040]; Uruguay[0.040]; hosting[0.040]; policies[0.040]; quite[0.040]; European[0.040]; sector[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = World Trade Organisation\u001b[39m ==> ENTITY: \u001b[32mWorld Trade Organization\u001b[39m\t\nSCORES: global= 0.313:0.313[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrade[0.047]; trade[0.047]; trade[0.047]; global[0.044]; global[0.044]; economies[0.043]; domestic[0.042]; economy[0.042]; key[0.041]; growth[0.041]; meeting[0.041]; accounting[0.041]; Singapore[0.040]; Singapore[0.040]; Singapore[0.040]; Singapore[0.040]; Singapore[0.040]; bid[0.040]; campaign[0.040]; free[0.040]; oriented[0.039]; ministerial[0.039]; economist[0.039]; hosting[0.039]; \t\n\r [=======================================>......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 2s560ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3979/4485 \r\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.154:0.154[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.051]; Soccer[0.044]; Javier[0.044]; Madrid[0.044]; Madrid[0.044]; Jose[0.043]; Alvarez[0.042]; Luis[0.042]; qualifier[0.042]; squad[0.042]; Clemente[0.041]; Cup[0.041]; Cup[0.041]; Yugoslavia[0.041]; World[0.040]; World[0.040]; Armando[0.040]; Armando[0.040]; Armando[0.040]; Atletico[0.040]; Deportivo[0.040]; added[0.039]; Coruna[0.039]; midfielder[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jose Luis Caminero\u001b[39m ==> ENTITY: \u001b[32mJosé Luis Caminero\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.047]; Spain[0.047]; Javier[0.044]; Madrid[0.044]; Madrid[0.044]; Soccer[0.043]; Armando[0.042]; Armando[0.042]; Armando[0.042]; Clemente[0.042]; squad[0.041]; Deportivo[0.041]; Alvarez[0.041]; Coruna[0.041]; Atletico[0.041]; midfielder[0.040]; qualifier[0.040]; added[0.040]; playmaker[0.040]; Cup[0.040]; Cup[0.040]; World[0.039]; World[0.039]; Yugoslavia[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Atletico Madrid\u001b[39m ==> ENTITY: \u001b[32mAtlético Madrid\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.046]; Jose[0.044]; Madrid[0.044]; Javier[0.044]; Luis[0.044]; Coruna[0.044]; Deportivo[0.044]; Spain[0.043]; Spain[0.043]; Cup[0.042]; Cup[0.042]; Armando[0.041]; Armando[0.041]; Armando[0.041]; squad[0.041]; qualifier[0.040]; added[0.040]; midfielder[0.040]; Clemente[0.040]; playmaker[0.039]; Alvarez[0.039]; December[0.039]; Yugoslavia[0.039]; coach[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Madrid ==> ENTITY: \u001b[32mMadrid\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMadrid[0.046]; Spain[0.046]; Spain[0.046]; Luis[0.044]; Jose[0.044]; Soccer[0.044]; squad[0.043]; Javier[0.043]; midfielder[0.042]; Coruna[0.042]; Clemente[0.041]; Cup[0.041]; Cup[0.041]; qualifier[0.041]; Deportivo[0.040]; Armando[0.040]; Armando[0.040]; Armando[0.040]; playmaker[0.040]; Alvarez[0.040]; Atletico[0.039]; added[0.039]; said[0.039]; coach[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Cup ==> ENTITY: \u001b[32mFIFA World Cup\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.144:0.144[0]; log p(e|m)= -1.002:-1.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.047]; squad[0.047]; Spain[0.046]; Spain[0.046]; qualifier[0.045]; Cup[0.044]; World[0.044]; coach[0.042]; Yugoslavia[0.042]; December[0.041]; Madrid[0.041]; Madrid[0.041]; midfielder[0.041]; added[0.041]; Jose[0.040]; Luis[0.040]; Clash[0.040]; Javier[0.040]; playmaker[0.039]; uncapped[0.039]; Clemente[0.039]; Alvarez[0.039]; Uncapped[0.039]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Javier Clemente\u001b[39m ==> ENTITY: \u001b[32mJavier Clemente\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.045]; Spain[0.045]; Luis[0.044]; Jose[0.043]; coach[0.043]; squad[0.043]; Soccer[0.042]; Madrid[0.042]; Madrid[0.042]; qualifier[0.042]; replaces[0.041]; Cup[0.041]; Cup[0.041]; playmaker[0.041]; Atletico[0.041]; Deportivo[0.041]; Coruna[0.041]; Armando[0.041]; Armando[0.041]; Armando[0.041]; Yugoslavia[0.040]; added[0.040]; Alvarez[0.040]; uncapped[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.154:0.154[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.051]; Soccer[0.044]; Javier[0.044]; Madrid[0.044]; Madrid[0.044]; Jose[0.043]; Alvarez[0.042]; Luis[0.042]; qualifier[0.042]; squad[0.042]; Clemente[0.041]; Cup[0.041]; Cup[0.041]; Yugoslavia[0.041]; World[0.040]; World[0.040]; Armando[0.040]; Armando[0.040]; Armando[0.040]; Atletico[0.040]; Deportivo[0.040]; added[0.039]; Coruna[0.039]; midfielder[0.039]; \t\n\r [=======================================>......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 2s542ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3987/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1253testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Albright\u001b[39m ==> ENTITY: \u001b[32mMadeleine Albright\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSecretary[0.046]; President[0.045]; President[0.045]; Clinton[0.044]; Clinton[0.044]; Clinton[0.044]; policy[0.044]; summit[0.042]; Foreign[0.042]; Albright[0.042]; Albright[0.042]; nominated[0.041]; State[0.041]; Bill[0.041]; quoted[0.041]; leaders[0.040]; said[0.040]; saying[0.039]; noted[0.039]; meet[0.039]; meet[0.039]; countries[0.039]; elected[0.039]; Minister[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.064:0.064[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nforeign[0.046]; Russia[0.045]; Washington[0.044]; Foreign[0.043]; Europe[0.042]; policy[0.042]; countries[0.042]; countries[0.042]; Soviet[0.042]; included[0.042]; noted[0.041]; Moscow[0.041]; Albright[0.041]; Albright[0.041]; ambassador[0.041]; outlining[0.041]; Russian[0.041]; Russian[0.041]; saying[0.041]; saying[0.041]; agreed[0.040]; March[0.040]; March[0.040]; Nations[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Clinton\u001b[39m ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npolicy[0.046]; President[0.045]; President[0.045]; Clinton[0.044]; Clinton[0.044]; Washington[0.043]; agreed[0.042]; news[0.042]; Secretary[0.042]; saying[0.042]; nominated[0.041]; supporting[0.041]; Bill[0.041]; elected[0.041]; ambassador[0.041]; told[0.041]; noted[0.040]; biography[0.040]; Albright[0.040]; Albright[0.040]; Albright[0.040]; Albright[0.040]; outlining[0.040]; relationship[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Igor Ivanov\u001b[39m ==> ENTITY: \u001b[32mIgor Ivanov\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.046]; Russia[0.046]; Moscow[0.046]; Moscow[0.046]; Russian[0.046]; Russian[0.046]; Ivanov[0.046]; Secretary[0.045]; active[0.044]; President[0.043]; President[0.043]; Minister[0.043]; agreed[0.042]; ambassador[0.042]; Boris[0.042]; nominated[0.042]; Foreign[0.042]; foreign[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; policy[0.041]; elected[0.041]; Deputy[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.058:0.058[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnews[0.043]; policy[0.043]; expected[0.043]; foreign[0.043]; noted[0.042]; countries[0.042]; Albright[0.042]; Albright[0.042]; Albright[0.042]; saying[0.042]; meet[0.042]; meet[0.042]; Friday[0.042]; agreed[0.041]; nominated[0.041]; March[0.041]; March[0.041]; worked[0.041]; Bill[0.041]; Russia[0.041]; Russia[0.041]; Foreign[0.041]; directions[0.040]; believes[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Interfax\u001b[39m ==> ENTITY: \u001b[32mInterfax\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.235:0.235[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInterfax[0.047]; Interfax[0.047]; news[0.046]; Moscow[0.046]; Russian[0.045]; Russian[0.045]; Warsaw[0.043]; Ivanov[0.042]; Ivanov[0.042]; Russia[0.042]; agency[0.042]; quoted[0.041]; Foreign[0.040]; foreign[0.040]; Soviet[0.040]; Washington[0.040]; told[0.040]; Europe[0.040]; saying[0.039]; saying[0.039]; Igor[0.039]; countries[0.039]; countries[0.039]; State[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NATO\u001b[39m ==> ENTITY: \u001b[32mNATO\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.135:0.135[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.045]; Nato[0.045]; Ivanov[0.043]; relations[0.043]; Russian[0.042]; Russian[0.042]; Warsaw[0.042]; included[0.042]; saying[0.042]; Europe[0.042]; plans[0.042]; plans[0.042]; President[0.041]; Soviet[0.041]; supporting[0.041]; security[0.041]; opposes[0.041]; Interfax[0.041]; Interfax[0.041]; policy[0.040]; countries[0.040]; countries[0.040]; defended[0.040]; agreed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.192:0.192[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.047]; Russian[0.047]; Russia[0.047]; Russia[0.047]; Moscow[0.044]; Moscow[0.044]; Warsaw[0.042]; Europe[0.042]; Boris[0.042]; Interfax[0.041]; Interfax[0.041]; Interfax[0.041]; Ivanov[0.040]; Ivanov[0.040]; eastern[0.040]; ambassador[0.040]; Igor[0.040]; biography[0.040]; Yeltsin[0.039]; Yeltsin[0.039]; President[0.039]; President[0.039]; relations[0.039]; Washington[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Albright\u001b[39m ==> ENTITY: \u001b[32mMadeleine Albright\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSecretary[0.046]; President[0.045]; President[0.045]; Clinton[0.044]; Clinton[0.044]; Clinton[0.044]; policy[0.043]; ambassador[0.042]; Washington[0.042]; summit[0.042]; Foreign[0.042]; Albright[0.041]; Albright[0.041]; Albright[0.041]; foreign[0.041]; nominated[0.041]; State[0.041]; Bill[0.040]; quoted[0.040]; told[0.040]; leaders[0.040]; directions[0.040]; said[0.039]; Nato[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bill Clinton\u001b[39m ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npolicy[0.046]; President[0.045]; President[0.045]; Clinton[0.045]; Clinton[0.045]; agreed[0.043]; news[0.042]; Secretary[0.042]; saying[0.042]; nominated[0.042]; elected[0.041]; told[0.041]; noted[0.041]; said[0.040]; Albright[0.040]; Albright[0.040]; Albright[0.040]; outlining[0.040]; relationship[0.040]; agency[0.040]; believes[0.040]; Foreign[0.040]; leaders[0.040]; foreign[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = United Nations\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountries[0.045]; countries[0.045]; active[0.044]; ambassador[0.043]; security[0.043]; Europe[0.043]; Washington[0.043]; agreed[0.042]; actively[0.042]; policy[0.042]; saying[0.041]; saying[0.041]; Albright[0.041]; Albright[0.041]; foreign[0.041]; outlining[0.041]; March[0.041]; March[0.041]; supporting[0.041]; President[0.040]; central[0.040]; summit[0.040]; leaders[0.040]; included[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.048]; Russia[0.048]; Russia[0.048]; Russian[0.046]; Boris[0.043]; State[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Igor[0.042]; Ivanov[0.042]; March[0.041]; March[0.041]; Friday[0.041]; Interfax[0.041]; Secretary[0.040]; countries[0.040]; Minister[0.040]; leaders[0.039]; relationship[0.039]; Bill[0.039]; said[0.039]; Clinton[0.039]; Clinton[0.039]; Clinton[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Albright\u001b[39m ==> ENTITY: \u001b[32mMadeleine Albright\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSecretary[0.046]; President[0.045]; Clinton[0.044]; Clinton[0.044]; policy[0.044]; ambassador[0.043]; Washington[0.043]; summit[0.042]; Foreign[0.042]; Albright[0.042]; foreign[0.041]; State[0.041]; quoted[0.041]; told[0.041]; Warsaw[0.041]; leaders[0.040]; directions[0.040]; Soviet[0.040]; Nato[0.040]; Nato[0.040]; saying[0.040]; saying[0.040]; noted[0.040]; defended[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Boris Yeltsin\u001b[39m ==> ENTITY: \u001b[32mBoris Yeltsin\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.045]; President[0.045]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Russian[0.043]; Russian[0.043]; Moscow[0.043]; Moscow[0.043]; Yeltsin[0.043]; elected[0.042]; nominated[0.042]; Interfax[0.041]; Interfax[0.041]; Interfax[0.041]; Igor[0.040]; ambassador[0.040]; Ivanov[0.040]; Ivanov[0.040]; leaders[0.040]; defended[0.040]; Secretary[0.039]; State[0.039]; supporting[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Interfax\u001b[39m ==> ENTITY: \u001b[32mInterfax\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.241:0.241[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInterfax[0.049]; Interfax[0.049]; news[0.048]; Moscow[0.047]; Moscow[0.047]; Russian[0.046]; Russian[0.046]; Ivanov[0.044]; Ivanov[0.044]; Russia[0.043]; Russia[0.043]; agency[0.043]; quoted[0.043]; Foreign[0.042]; foreign[0.042]; told[0.041]; saying[0.041]; Igor[0.041]; countries[0.040]; State[0.040]; Friday[0.040]; March[0.040]; March[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Europe ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.091:0.091[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountries[0.045]; countries[0.045]; Russia[0.043]; eastern[0.043]; Soviet[0.043]; Nato[0.042]; Nato[0.042]; foreign[0.042]; included[0.042]; Warsaw[0.042]; United[0.041]; Russian[0.041]; Russian[0.041]; ambassador[0.041]; central[0.041]; agreed[0.041]; summit[0.041]; relations[0.041]; development[0.040]; plans[0.040]; plans[0.040]; supporting[0.040]; security[0.040]; Pact[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ivanov\u001b[39m ==> ENTITY: \u001b[32mIgor Ivanov\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.045]; Moscow[0.045]; Russian[0.044]; Russian[0.044]; Ivanov[0.044]; Secretary[0.044]; active[0.042]; President[0.042]; President[0.042]; Igor[0.042]; Minister[0.042]; agreed[0.041]; ambassador[0.041]; Boris[0.041]; nominated[0.041]; Foreign[0.040]; Soviet[0.040]; foreign[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Europe[0.040]; policy[0.040]; elected[0.040]; security[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NATO\u001b[39m ==> ENTITY: \u001b[32mNATO\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.045]; Nato[0.045]; Ivanov[0.043]; relations[0.043]; Russian[0.043]; Russian[0.043]; included[0.042]; constructive[0.042]; saying[0.042]; Europe[0.042]; President[0.042]; Soviet[0.041]; supporting[0.041]; security[0.041]; Warsaw[0.041]; opposes[0.041]; Interfax[0.041]; Interfax[0.041]; policy[0.041]; Clinton[0.040]; Clinton[0.040]; agreed[0.040]; summit[0.040]; believes[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yeltsin\u001b[39m ==> ENTITY: \u001b[32mBoris Yeltsin\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.045]; President[0.045]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Russian[0.043]; Russian[0.043]; Moscow[0.043]; Moscow[0.043]; Yeltsin[0.043]; elected[0.042]; nominated[0.041]; Interfax[0.041]; Interfax[0.041]; Interfax[0.041]; Boris[0.040]; Igor[0.040]; ambassador[0.040]; Ivanov[0.040]; Ivanov[0.040]; leaders[0.040]; defended[0.040]; Secretary[0.039]; State[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMoscow\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMoscow\u001b[39m <---> \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.262:0.241[\u001b[31m0.021\u001b[39m]; local()= 0.181:0.188[\u001b[32m0.008\u001b[39m]; log p(e|m)= -0.106:-4.269[\u001b[31m4.162\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.047]; Russia[0.047]; Russia[0.047]; Russian[0.045]; Russian[0.045]; Washington[0.044]; Boris[0.042]; State[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Igor[0.041]; Ivanov[0.041]; Ivanov[0.041]; ambassador[0.041]; Friday[0.040]; Interfax[0.040]; Interfax[0.040]; Interfax[0.040]; biography[0.040]; defended[0.040]; countries[0.039]; March[0.039]; March[0.039]; worked[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.125:0.125[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.049]; Moscow[0.046]; Moscow[0.046]; Interfax[0.044]; President[0.044]; President[0.044]; Ivanov[0.042]; Igor[0.042]; Boris[0.042]; Secretary[0.041]; agency[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; said[0.041]; Minister[0.041]; nominated[0.041]; countries[0.040]; news[0.040]; March[0.040]; elected[0.040]; policy[0.040]; relationship[0.039]; noted[0.039]; leaders[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.048]; Russian[0.046]; Moscow[0.045]; Moscow[0.045]; Interfax[0.043]; President[0.043]; President[0.043]; Ivanov[0.042]; Igor[0.041]; Boris[0.041]; Secretary[0.041]; agreed[0.041]; agency[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; said[0.040]; foreign[0.040]; Minister[0.040]; ministry[0.040]; nominated[0.040]; countries[0.040]; news[0.040]; March[0.039]; March[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Warsaw Pact\u001b[39m ==> ENTITY: \u001b[32mWarsaw Pact\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.049]; Russia[0.048]; Russian[0.046]; Russian[0.046]; countries[0.042]; countries[0.042]; ambassador[0.042]; Europe[0.042]; security[0.042]; Nato[0.042]; Nato[0.042]; Ivanov[0.041]; supporting[0.040]; Washington[0.040]; opposes[0.040]; relations[0.040]; defended[0.040]; biography[0.040]; agreed[0.040]; Interfax[0.040]; Interfax[0.040]; directions[0.040]; saying[0.039]; leaders[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Interfax\u001b[39m ==> ENTITY: \u001b[32mInterfax\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.235:0.235[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInterfax[0.047]; Interfax[0.047]; news[0.046]; Moscow[0.046]; Russian[0.045]; Russian[0.045]; Warsaw[0.043]; Ivanov[0.042]; Ivanov[0.042]; Russia[0.042]; agency[0.042]; quoted[0.041]; Foreign[0.040]; foreign[0.040]; Soviet[0.040]; Washington[0.040]; told[0.040]; Europe[0.040]; saying[0.039]; saying[0.039]; Igor[0.039]; countries[0.039]; countries[0.039]; State[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Clinton\u001b[39m ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npolicy[0.046]; President[0.045]; President[0.045]; Clinton[0.044]; Clinton[0.044]; Washington[0.043]; agreed[0.042]; news[0.042]; Secretary[0.042]; saying[0.042]; nominated[0.041]; supporting[0.041]; Bill[0.041]; elected[0.041]; ambassador[0.041]; told[0.041]; noted[0.040]; biography[0.040]; Albright[0.040]; Albright[0.040]; Albright[0.040]; Albright[0.040]; outlining[0.040]; relationship[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.050]; Russian[0.048]; Russian[0.048]; Warsaw[0.045]; Europe[0.045]; Interfax[0.045]; Interfax[0.045]; President[0.045]; Boris[0.043]; eastern[0.043]; ambassador[0.043]; opposes[0.043]; Yeltsin[0.042]; Yeltsin[0.042]; agreed[0.042]; United[0.042]; Ivanov[0.042]; supporting[0.042]; plans[0.042]; plans[0.042]; foreign[0.042]; ministry[0.041]; Washington[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Madeleine Albright\u001b[39m ==> ENTITY: \u001b[32mMadeleine Albright\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSecretary[0.046]; President[0.045]; President[0.045]; Clinton[0.044]; Clinton[0.044]; Clinton[0.044]; policy[0.043]; summit[0.042]; Foreign[0.042]; Albright[0.042]; Albright[0.042]; foreign[0.041]; nominated[0.041]; State[0.041]; Bill[0.041]; quoted[0.041]; leaders[0.040]; directions[0.040]; said[0.040]; saying[0.039]; noted[0.039]; meet[0.039]; meet[0.039]; countries[0.039]; \t\n\r [========================================>.....] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 2s401ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4015/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1209testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brussels ==> ENTITY: \u001b[32mBrussels\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.057:0.057[0]; log p(e|m)= -0.025:-0.025[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.048]; match[0.045]; basketball[0.044]; Basketball[0.044]; Euroleague[0.044]; Thursday[0.043]; Madrid[0.043]; Spain[0.043]; Charleroi[0.042]; Charleroi[0.042]; Charleroi[0.042]; Belgrade[0.042]; Belgrade[0.042]; EuroLeague[0.042]; Italy[0.041]; Yugoslavia[0.040]; Estudiantes[0.040]; Estudiantes[0.040]; Jacques[0.039]; Kinder[0.039]; Kinder[0.039]; Partizan[0.039]; Partizan[0.039]; Stas[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.068:0.068[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJuan[0.046]; Italy[0.046]; match[0.045]; Madrid[0.045]; EuroLeague[0.045]; Euroleague[0.042]; Basketball[0.042]; Estudiantes[0.042]; Estudiantes[0.042]; Yugoslavia[0.042]; Belgium[0.042]; Thursday[0.042]; basketball[0.041]; Williams[0.040]; Eric[0.040]; Ellis[0.040]; Dejan[0.040]; Belgrade[0.040]; Belgrade[0.040]; Thompson[0.040]; Group[0.040]; Group[0.040]; Savic[0.039]; halftime[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dejan Koturovic\u001b[39m ==> ENTITY: \u001b[32mDejan Koturović\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuroLeague[0.047]; Euroleague[0.047]; Basketball[0.046]; Partizan[0.045]; Partizan[0.045]; Belgrade[0.045]; Belgrade[0.045]; basketball[0.045]; Yugoslavia[0.044]; Bologna[0.041]; Aisa[0.040]; Italy[0.040]; halftime[0.040]; Savic[0.040]; Estudiantes[0.039]; Estudiantes[0.039]; Zoran[0.039]; Charleroi[0.039]; Charleroi[0.039]; Charleroi[0.039]; Madrid[0.039]; scorers[0.039]; scorers[0.039]; Brussels[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kinder Bologna\u001b[39m ==> ENTITY: \u001b[32mVirtus Pallacanestro Bologna\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuroLeague[0.052]; Euroleague[0.051]; Italy[0.051]; Basketball[0.048]; Partizan[0.046]; Partizan[0.046]; basketball[0.044]; Belgrade[0.042]; Belgrade[0.042]; Spain[0.042]; Yugoslavia[0.042]; Aisa[0.042]; Thursday[0.042]; halftime[0.042]; Savic[0.042]; Dejan[0.041]; Estudiantes[0.041]; Estudiantes[0.041]; Charleroi[0.041]; Charleroi[0.041]; Charleroi[0.041]; Belgium[0.040]; Madrid[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Charleroi ==> ENTITY: \u001b[32mSpirou Charleroi\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.193:0.193[0]; log p(e|m)= -1.938:-1.938[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.046]; EuroLeague[0.045]; Euroleague[0.045]; match[0.044]; Charleroi[0.044]; Charleroi[0.044]; Spain[0.043]; basketball[0.043]; Partizan[0.043]; Partizan[0.043]; Basketball[0.042]; Brussels[0.042]; Thursday[0.042]; Italy[0.042]; Jacques[0.042]; Madrid[0.041]; Yugoslavia[0.040]; Belgrade[0.040]; Belgrade[0.040]; Result[0.039]; Result[0.039]; Dejan[0.039]; Group[0.038]; Group[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgium ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.092:0.092[0]; log p(e|m)= -0.242:-0.242[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.048]; match[0.046]; Euroleague[0.045]; basketball[0.045]; EuroLeague[0.045]; Spain[0.045]; Italy[0.044]; Yugoslavia[0.043]; Jacques[0.041]; Eric[0.041]; Belgrade[0.040]; Belgrade[0.040]; Partizan[0.040]; Partizan[0.040]; Brussels[0.040]; Stas[0.040]; Savic[0.040]; Dejan[0.040]; Charleroi[0.040]; Charleroi[0.040]; Charleroi[0.040]; Thursday[0.040]; Madrid[0.039]; Ron[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuroLeague[0.049]; Euroleague[0.049]; Basketball[0.049]; match[0.044]; Spain[0.044]; Yugoslavia[0.044]; basketball[0.044]; Bologna[0.043]; Belgium[0.041]; Estudiantes[0.041]; Estudiantes[0.041]; Thursday[0.040]; Group[0.040]; Group[0.040]; Partizan[0.040]; Partizan[0.040]; Ellis[0.040]; Madrid[0.039]; Charleroi[0.039]; Charleroi[0.039]; Charleroi[0.039]; Belgrade[0.039]; Belgrade[0.039]; halftime[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Estudiantes ==> ENTITY: \u001b[32mCB Estudiantes\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.147:0.147[0]; log p(e|m)= -1.784:-1.784[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEstudiantes[0.049]; Juan[0.048]; Spain[0.048]; EuroLeague[0.047]; match[0.047]; Euroleague[0.046]; basketball[0.045]; Basketball[0.044]; Madrid[0.044]; Partizan[0.042]; Partizan[0.042]; scorers[0.042]; scorers[0.042]; Aisa[0.042]; Thursday[0.042]; halftime[0.041]; Savic[0.041]; Yugoslavia[0.041]; Dejan[0.041]; Group[0.041]; Group[0.041]; Belgium[0.041]; Ellis[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Euroleague\u001b[39m ==> ENTITY: \u001b[32mEuroleague Basketball\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuroLeague[0.054]; Basketball[0.051]; basketball[0.048]; Belgrade[0.048]; Belgrade[0.048]; Partizan[0.048]; Partizan[0.048]; Italy[0.046]; Madrid[0.046]; Spain[0.046]; match[0.045]; Thursday[0.045]; Harper[0.044]; Belgium[0.043]; Yugoslavia[0.043]; Brussels[0.043]; Bologna[0.042]; Estudiantes[0.042]; Estudiantes[0.042]; Group[0.042]; Group[0.042]; Thompson[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCharleroi\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSpirou Charleroi\u001b[39m <---> \u001b[32mCharleroi\u001b[39m\t\nSCORES: global= 0.259:0.240[\u001b[31m0.019\u001b[39m]; local()= 0.193:0.086[\u001b[31m0.107\u001b[39m]; log p(e|m)= -1.938:-0.255[\u001b[32m1.683\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.046]; EuroLeague[0.045]; Euroleague[0.045]; match[0.044]; Charleroi[0.044]; Charleroi[0.044]; Spain[0.043]; basketball[0.043]; Partizan[0.043]; Partizan[0.043]; Basketball[0.042]; Brussels[0.042]; Thursday[0.042]; Italy[0.042]; Jacques[0.042]; Madrid[0.041]; Yugoslavia[0.040]; Belgrade[0.040]; Belgrade[0.040]; Result[0.039]; Result[0.039]; Dejan[0.039]; Group[0.038]; Group[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Partizan ==> ENTITY: \u001b[32mKK Partizan\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.139:0.139[0]; log p(e|m)= -1.687:-1.687[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuroLeague[0.047]; Basketball[0.046]; Euroleague[0.046]; Partizan[0.046]; Yugoslavia[0.045]; basketball[0.045]; Belgrade[0.044]; Belgrade[0.044]; match[0.043]; Dejan[0.042]; Thursday[0.040]; Zoran[0.040]; Savic[0.040]; Juan[0.040]; Italy[0.040]; Eric[0.040]; Group[0.040]; Group[0.040]; Ron[0.039]; Belgium[0.039]; Stas[0.039]; Charleroi[0.038]; Charleroi[0.038]; Charleroi[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgrade ==> ENTITY: \u001b[32mBelgrade\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.094:0.094[0]; log p(e|m)= -0.052:-0.052[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYugoslavia[0.046]; Belgrade[0.046]; Belgium[0.045]; Basketball[0.044]; Partizan[0.043]; Partizan[0.043]; EuroLeague[0.043]; Italy[0.043]; match[0.043]; Dejan[0.042]; basketball[0.042]; Spain[0.041]; Thursday[0.041]; Euroleague[0.040]; Eric[0.040]; scorers[0.040]; scorers[0.040]; Charleroi[0.040]; Charleroi[0.040]; Charleroi[0.040]; Madrid[0.040]; Brussels[0.040]; Savic[0.040]; Juan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Estudiantes Madrid\u001b[39m ==> ENTITY: \u001b[32mCB Estudiantes\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.120:0.120[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.047]; EuroLeague[0.046]; Euroleague[0.046]; basketball[0.044]; Estudiantes[0.044]; Basketball[0.044]; Juan[0.042]; Partizan[0.042]; Partizan[0.042]; Aisa[0.041]; halftime[0.041]; Savic[0.041]; Yugoslavia[0.041]; Dejan[0.041]; Thursday[0.040]; Belgium[0.040]; Charleroi[0.040]; Charleroi[0.040]; Charleroi[0.040]; scorers[0.040]; scorers[0.040]; Belgrade[0.040]; Belgrade[0.040]; Italy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zoran Savic\u001b[39m ==> ENTITY: \u001b[32mZoran Savić\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuroLeague[0.049]; Basketball[0.049]; Euroleague[0.047]; basketball[0.046]; Yugoslavia[0.046]; Italy[0.046]; Eric[0.044]; Belgrade[0.044]; Belgrade[0.044]; Dejan[0.043]; Thompson[0.043]; match[0.043]; Ron[0.042]; Partizan[0.042]; Partizan[0.042]; Spain[0.042]; Aisa[0.042]; Ellis[0.042]; halftime[0.041]; Williams[0.041]; Thursday[0.041]; Estudiantes[0.041]; Estudiantes[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = EuroLeague\u001b[39m ==> ENTITY: \u001b[32mEuroleague Basketball\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuroleague[0.054]; Basketball[0.051]; basketball[0.048]; Belgrade[0.048]; Belgrade[0.048]; Partizan[0.048]; Partizan[0.048]; Italy[0.046]; Madrid[0.046]; Spain[0.046]; match[0.045]; Thursday[0.045]; Harper[0.044]; Belgium[0.043]; Yugoslavia[0.043]; Brussels[0.043]; Bologna[0.042]; Estudiantes[0.042]; Estudiantes[0.042]; Group[0.042]; Group[0.042]; Thompson[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Partizan Belgrade ==> ENTITY: \u001b[32mKK Partizan\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.135:0.135[0]; log p(e|m)= -0.909:-0.909[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuroLeague[0.047]; Basketball[0.047]; Euroleague[0.045]; Dejan[0.045]; basketball[0.045]; Belgrade[0.045]; Partizan[0.044]; match[0.044]; Yugoslavia[0.043]; Italy[0.042]; Savic[0.042]; Thursday[0.041]; Zoran[0.041]; Group[0.040]; Group[0.040]; Ron[0.040]; Eric[0.039]; scorers[0.039]; scorers[0.039]; Estudiantes[0.039]; Estudiantes[0.039]; Charleroi[0.039]; Charleroi[0.039]; Charleroi[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Charleroi ==> ENTITY: \u001b[32mSpirou Charleroi\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.193:0.193[0]; log p(e|m)= -1.938:-1.938[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.046]; EuroLeague[0.045]; Euroleague[0.045]; match[0.044]; Charleroi[0.044]; Charleroi[0.044]; Spain[0.043]; basketball[0.043]; Partizan[0.043]; Partizan[0.043]; Basketball[0.042]; Brussels[0.042]; Thursday[0.042]; Italy[0.042]; Jacques[0.042]; Madrid[0.041]; Yugoslavia[0.040]; Belgrade[0.040]; Belgrade[0.040]; Result[0.039]; Result[0.039]; Dejan[0.039]; Group[0.038]; Group[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Yugoslavia ==> ENTITY: \u001b[32mYugoslavia\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.654:-0.654[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuroLeague[0.047]; match[0.046]; Italy[0.046]; Belgrade[0.045]; Belgrade[0.045]; Spain[0.043]; Basketball[0.043]; Euroleague[0.043]; Belgium[0.042]; basketball[0.042]; Stas[0.042]; Dejan[0.042]; Savic[0.041]; Madrid[0.041]; Zoran[0.040]; Group[0.039]; Group[0.039]; Thursday[0.039]; Partizan[0.039]; Partizan[0.039]; halftime[0.039]; Charleroi[0.038]; Charleroi[0.038]; Charleroi[0.038]; \t\n\r [========================================>.....] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 2s329ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4033/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1167testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Al Ain\u001b[39m ==> ENTITY: \u001b[32mAl Ain\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEmirates[0.047]; won[0.046]; Arab[0.045]; Soccer[0.045]; Cup[0.045]; Cup[0.045]; matches[0.044]; Hassan[0.044]; Syria[0.044]; Syria[0.044]; Syria[0.044]; goals[0.043]; goals[0.043]; Ain[0.043]; goal[0.043]; Uzbekistan[0.043]; Uzbekistan[0.043]; United[0.042]; played[0.042]; played[0.042]; lost[0.042]; Asian[0.041]; Asian[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hassan Abbas\u001b[39m ==> ENTITY: \u001b[32mHassan Abbas\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.223:0.223[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.046]; Cup[0.046]; Syria[0.045]; Syria[0.045]; Syria[0.045]; played[0.043]; played[0.043]; Emirates[0.043]; Soccer[0.042]; Al[0.042]; Arab[0.042]; won[0.042]; matches[0.041]; United[0.040]; goals[0.040]; goals[0.040]; Uzbekistan[0.040]; Uzbekistan[0.040]; Nader[0.040]; goal[0.039]; Japan[0.039]; Japan[0.039]; Japan[0.039]; Group[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan national football team\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.209:0.209[0]; log p(e|m)= -3.730:-3.730[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.049]; Japan[0.049]; matches[0.044]; Soccer[0.044]; Cup[0.042]; Cup[0.042]; goal[0.042]; won[0.042]; China[0.042]; China[0.042]; Asian[0.042]; Asian[0.042]; Uzbekistan[0.041]; Uzbekistan[0.041]; played[0.041]; played[0.041]; goals[0.041]; goals[0.041]; Takuya[0.039]; group[0.039]; Group[0.039]; points[0.039]; Emirates[0.039]; United[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Takuya Takagi\u001b[39m ==> ENTITY: \u001b[32mTakuya Takagi\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.046]; Soccer[0.045]; Japan[0.044]; Japan[0.044]; Japan[0.044]; played[0.043]; played[0.043]; goals[0.041]; goals[0.041]; won[0.041]; Asian[0.041]; Asian[0.041]; lost[0.041]; Emirates[0.041]; goal[0.041]; Uzbekistan[0.040]; Uzbekistan[0.040]; Cup[0.040]; Cup[0.040]; halftime[0.040]; halftime[0.040]; Scorers[0.040]; Scorers[0.040]; Attendance[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syria ==> ENTITY: \u001b[32mSyria national football team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.204:0.204[0]; log p(e|m)= -3.058:-3.058[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.050]; Syria[0.050]; Asian[0.045]; Asian[0.045]; Arab[0.045]; Cup[0.043]; Cup[0.043]; China[0.042]; China[0.042]; won[0.041]; matches[0.041]; Uzbekistan[0.041]; Uzbekistan[0.041]; Soccer[0.040]; Al[0.040]; Japan[0.040]; Japan[0.040]; Japan[0.040]; Hassan[0.040]; Emirates[0.039]; Abbas[0.039]; Attendance[0.038]; played[0.038]; played[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syria ==> ENTITY: \u001b[32mSyria national football team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.204:0.204[0]; log p(e|m)= -3.058:-3.058[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.050]; Syria[0.050]; Asian[0.045]; Asian[0.045]; Arab[0.045]; Cup[0.043]; Cup[0.043]; China[0.042]; China[0.042]; won[0.041]; matches[0.041]; Uzbekistan[0.041]; Uzbekistan[0.041]; Soccer[0.040]; Al[0.040]; Japan[0.040]; Japan[0.040]; Japan[0.040]; Hassan[0.040]; Emirates[0.039]; Abbas[0.039]; Attendance[0.038]; played[0.038]; played[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syria ==> ENTITY: \u001b[32mSyria national football team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.204:0.204[0]; log p(e|m)= -3.058:-3.058[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.050]; Syria[0.050]; Asian[0.045]; Asian[0.045]; Arab[0.045]; Cup[0.043]; Cup[0.043]; China[0.042]; China[0.042]; won[0.041]; matches[0.041]; Uzbekistan[0.041]; Uzbekistan[0.041]; Soccer[0.040]; Al[0.040]; Japan[0.040]; Japan[0.040]; Japan[0.040]; Hassan[0.040]; Emirates[0.039]; Abbas[0.039]; Attendance[0.038]; played[0.038]; played[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Asian Cup ==> ENTITY: \u001b[32mAFC Asian Cup\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.209:0.209[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.046]; Asian[0.045]; China[0.045]; China[0.045]; Syria[0.045]; Syria[0.045]; Syria[0.045]; Uzbekistan[0.045]; Uzbekistan[0.045]; Soccer[0.044]; matches[0.044]; Japan[0.044]; Japan[0.044]; Japan[0.044]; Friday[0.044]; won[0.042]; Emirates[0.042]; played[0.041]; played[0.041]; Results[0.041]; Results[0.041]; Arab[0.041]; goal[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Asian Cup ==> ENTITY: \u001b[32mAFC Asian Cup\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.209:0.209[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.046]; Asian[0.045]; China[0.045]; China[0.045]; Syria[0.045]; Syria[0.045]; Syria[0.045]; Uzbekistan[0.045]; Uzbekistan[0.045]; Soccer[0.044]; matches[0.044]; Japan[0.044]; Japan[0.044]; Japan[0.044]; Friday[0.044]; won[0.042]; Emirates[0.042]; played[0.041]; played[0.041]; Results[0.041]; Results[0.041]; Arab[0.041]; goal[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina national football team\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.200:0.200[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.049]; Japan[0.044]; Japan[0.044]; Japan[0.044]; Asian[0.044]; Asian[0.044]; matches[0.043]; Uzbekistan[0.043]; Uzbekistan[0.043]; won[0.043]; Cup[0.042]; Cup[0.042]; played[0.040]; played[0.040]; group[0.040]; goal[0.040]; Emirates[0.040]; goals[0.039]; goals[0.039]; Syria[0.039]; Syria[0.039]; Syria[0.039]; Arab[0.039]; lost[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina national football team\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.203:0.203[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.049]; Japan[0.044]; Japan[0.044]; Japan[0.044]; Soccer[0.043]; matches[0.043]; Uzbekistan[0.043]; Uzbekistan[0.043]; won[0.043]; Asian[0.042]; Asian[0.042]; Cup[0.042]; Cup[0.042]; played[0.040]; played[0.040]; group[0.040]; goal[0.040]; Emirates[0.040]; goals[0.039]; goals[0.039]; points[0.039]; Arab[0.039]; lost[0.039]; United[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan national football team\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.209:0.209[0]; log p(e|m)= -3.730:-3.730[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.049]; Japan[0.049]; matches[0.044]; Soccer[0.044]; Cup[0.042]; Cup[0.042]; goal[0.042]; won[0.042]; China[0.042]; China[0.042]; Asian[0.042]; Asian[0.042]; Uzbekistan[0.041]; Uzbekistan[0.041]; played[0.041]; played[0.041]; goals[0.041]; goals[0.041]; Takuya[0.039]; group[0.039]; Group[0.039]; points[0.039]; Emirates[0.039]; United[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United Arab Emirates ==> ENTITY: \u001b[32mUnited Arab Emirates\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.162:0.162[0]; log p(e|m)= -0.250:-0.250[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAsian[0.046]; Asian[0.046]; matches[0.045]; Uzbekistan[0.043]; Uzbekistan[0.043]; Cup[0.043]; Cup[0.043]; won[0.043]; China[0.042]; China[0.042]; Syria[0.042]; Syria[0.042]; Syria[0.042]; Soccer[0.041]; played[0.041]; played[0.041]; Japan[0.041]; Japan[0.041]; Japan[0.041]; Al[0.040]; Hassan[0.040]; Standings[0.040]; lost[0.039]; Group[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Uzbekistan ==> ENTITY: \u001b[32mUzbekistan national football team\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.202:0.202[0]; log p(e|m)= -2.226:-2.226[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUzbekistan[0.050]; Asian[0.048]; Asian[0.048]; China[0.044]; China[0.044]; matches[0.044]; won[0.043]; Soccer[0.042]; Japan[0.042]; Japan[0.042]; Japan[0.042]; played[0.041]; played[0.041]; Cup[0.040]; Cup[0.040]; Standings[0.040]; goals[0.039]; goals[0.039]; Syria[0.039]; Syria[0.039]; Syria[0.039]; goal[0.039]; Shatskikh[0.038]; Arab[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan national football team\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.209:0.209[0]; log p(e|m)= -3.730:-3.730[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.049]; Japan[0.049]; matches[0.044]; Soccer[0.044]; Cup[0.042]; Cup[0.042]; goal[0.042]; won[0.042]; China[0.042]; China[0.042]; Asian[0.042]; Asian[0.042]; Uzbekistan[0.041]; Uzbekistan[0.041]; played[0.041]; played[0.041]; goals[0.041]; goals[0.041]; Takuya[0.039]; group[0.039]; Group[0.039]; points[0.039]; Emirates[0.039]; United[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Uzbekistan ==> ENTITY: \u001b[32mUzbekistan national football team\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.199:0.199[0]; log p(e|m)= -2.226:-2.226[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUzbekistan[0.050]; Asian[0.048]; Asian[0.048]; China[0.044]; China[0.044]; matches[0.044]; won[0.043]; Japan[0.042]; Japan[0.042]; Japan[0.042]; played[0.041]; played[0.041]; Soccer[0.041]; Friday[0.040]; Standings[0.040]; Cup[0.039]; Cup[0.039]; goals[0.039]; goals[0.039]; Group[0.039]; Syria[0.039]; Syria[0.039]; Syria[0.039]; goal[0.039]; \t\n\r [========================================>.....] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 2s244ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4049/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1286testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = USDA\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Agriculture\u001b[39m\t\nSCORES: global= 0.227:0.227[0]; local()= 0.035:0.035[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMinn[0.048]; Steers[0.048]; Thursday[0.047]; Chicago[0.047]; cattle[0.047]; Fri[0.047]; market[0.047]; Moines[0.046]; Iowa[0.046]; tested[0.046]; tested[0.046]; Reported[0.046]; feedlot[0.046]; Des[0.045]; heifers[0.045]; Week[0.045]; said[0.045]; sales[0.045]; compared[0.043]; close[0.042]; newsdesk[0.041]; Date[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Des Moines ==> ENTITY: \u001b[32mDes Moines, Iowa\u001b[39m\t\nSCORES: global= 0.227:0.227[0]; local()= 0.042:0.042[0]; log p(e|m)= -0.158:-0.158[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIowa[0.057]; Thursday[0.053]; Minn[0.053]; Steers[0.053]; cattle[0.052]; Fri[0.052]; Chicago[0.050]; Reported[0.050]; feedlot[0.050]; heifers[0.050]; compared[0.050]; Week[0.050]; said[0.049]; close[0.049]; market[0.048]; sales[0.047]; tested[0.047]; tested[0.047]; Date[0.047]; newsdesk[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = USDA\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Agriculture\u001b[39m\t\nSCORES: global= 0.227:0.227[0]; local()= 0.035:0.035[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMinn[0.048]; Steers[0.048]; Thursday[0.047]; Chicago[0.047]; cattle[0.047]; Fri[0.047]; market[0.047]; Moines[0.046]; Iowa[0.046]; tested[0.046]; tested[0.046]; Reported[0.046]; feedlot[0.046]; Des[0.045]; heifers[0.045]; Week[0.045]; said[0.045]; sales[0.045]; compared[0.043]; close[0.042]; newsdesk[0.041]; Date[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.005:0.005[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThursday[0.059]; Week[0.058]; close[0.056]; Iowa[0.055]; sales[0.055]; Fri[0.054]; said[0.054]; market[0.053]; Date[0.053]; Steers[0.052]; newsdesk[0.052]; Des[0.052]; compared[0.052]; Reported[0.050]; Minn[0.050]; heifers[0.049]; tested[0.049]; tested[0.049]; cattle[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Minn\u001b[39m ==> ENTITY: \u001b[32mMinnesota\u001b[39m\t\nSCORES: global= 0.213:0.213[0]; local()= 0.024:0.024[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSteers[0.054]; Iowa[0.053]; cattle[0.053]; Fri[0.053]; Moines[0.052]; Reported[0.051]; feedlot[0.051]; Des[0.051]; heifers[0.051]; Week[0.051]; Chicago[0.050]; Thursday[0.050]; said[0.049]; market[0.049]; sales[0.048]; close[0.047]; compared[0.047]; newsdesk[0.046]; tested[0.046]; tested[0.046]; \t\n\r [========================================>.....] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 2s222ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4054/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1370testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Birmingham ==> ENTITY: \u001b[32mBirmingham City F.C.\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.171:0.171[0]; log p(e|m)= -3.101:-3.101[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOxford[0.045]; Charlton[0.045]; Nottingham[0.044]; Wolverhampton[0.044]; Southampton[0.043]; Sheffield[0.043]; Coventry[0.043]; Blackburn[0.043]; Manchester[0.043]; City[0.042]; Norwich[0.042]; Barnsley[0.041]; Swindon[0.041]; Portsmouth[0.041]; Reading[0.041]; Ipswich[0.041]; Tranmere[0.041]; Bolton[0.040]; Stoke[0.040]; Park[0.040]; Huddersfield[0.040]; West[0.039]; United[0.039]; Queens[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oxford ==> ENTITY: \u001b[32mOxford United F.C.\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.209:0.209[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthampton[0.045]; Coventry[0.045]; Sheffield[0.044]; Charlton[0.044]; Wolverhampton[0.043]; Birmingham[0.043]; City[0.043]; Reading[0.042]; Nottingham[0.042]; Norwich[0.042]; Swindon[0.042]; Portsmouth[0.042]; Ipswich[0.042]; Manchester[0.041]; Park[0.041]; Barnsley[0.041]; Huddersfield[0.041]; Bolton[0.040]; Tranmere[0.040]; United[0.040]; Queens[0.040]; Division[0.040]; Stoke[0.040]; West[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Peterborough ==> ENTITY: \u001b[32mPeterborough United F.C.\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.204:0.204[0]; log p(e|m)= -3.170:-3.170[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.046]; Bristol[0.044]; Bristol[0.044]; Swansea[0.043]; Shrewsbury[0.043]; Walsall[0.043]; Colchester[0.042]; Carlisle[0.042]; Rotherham[0.042]; Plymouth[0.042]; Gillingham[0.042]; Cardiff[0.042]; Wycombe[0.042]; Bournemouth[0.041]; Scunthorpe[0.041]; Division[0.041]; Wrexham[0.041]; York[0.041]; Blackpool[0.041]; City[0.041]; Crewe[0.040]; Barnet[0.040]; Cambridge[0.039]; Rovers[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tottenham ==> ENTITY: \u001b[32mTottenham Hotspur F.C.\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.207:0.207[0]; log p(e|m)= -0.997:-0.997[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; league[0.044]; matches[0.043]; Southampton[0.043]; Manchester[0.043]; Blackburn[0.043]; Sheffield[0.042]; Sheffield[0.042]; London[0.042]; Sunderland[0.042]; Bolton[0.042]; Coventry[0.042]; Middlesbrough[0.042]; soccer[0.042]; Leeds[0.041]; Liverpool[0.041]; Chelsea[0.040]; Newcastle[0.040]; Everton[0.040]; played[0.040]; Leicester[0.040]; United[0.040]; Arsenal[0.040]; Nottingham[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Grimsby ==> ENTITY: \u001b[32mGrimsby Town F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.214:0.214[0]; log p(e|m)= -2.604:-2.604[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIpswich[0.043]; Southend[0.043]; Swindon[0.043]; Burnley[0.042]; Portsmouth[0.042]; Charlton[0.042]; Chesterfield[0.042]; Brentford[0.042]; Huddersfield[0.042]; West[0.042]; Stockport[0.042]; Watford[0.042]; Bury[0.042]; Oldham[0.042]; Manchester[0.042]; Bradford[0.042]; Wrexham[0.041]; Luton[0.041]; Millwall[0.041]; City[0.041]; Crewe[0.041]; Reading[0.041]; Park[0.040]; Division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bradford ==> ENTITY: \u001b[32mBradford City A.F.C.\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.233:0.233[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.043]; Burnley[0.043]; Bristol[0.043]; Watford[0.043]; Huddersfield[0.043]; Stockport[0.042]; Manchester[0.042]; Millwall[0.042]; Reading[0.042]; Chesterfield[0.042]; City[0.042]; City[0.042]; Brentford[0.042]; Oldham[0.042]; Ipswich[0.041]; Bury[0.041]; Southend[0.041]; Portsmouth[0.041]; Division[0.041]; Luton[0.041]; Crewe[0.040]; Wrexham[0.040]; Park[0.040]; Rangers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wimbledon ==> ENTITY: \u001b[32mWimbledon F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.204:0.204[0]; log p(e|m)= -1.890:-1.890[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; matches[0.045]; league[0.044]; league[0.044]; Sheffield[0.044]; Sunderland[0.043]; London[0.043]; Manchester[0.042]; Leeds[0.042]; Saturday[0.042]; played[0.042]; Soccer[0.041]; soccer[0.041]; Newcastle[0.041]; Liverpool[0.041]; Leicester[0.041]; United[0.040]; Tottenham[0.040]; Wednesday[0.040]; won[0.040]; Aston[0.040]; Derby[0.040]; Chelsea[0.040]; Everton[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Ham ==> ENTITY: \u001b[32mWest Ham United F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.192:0.192[0]; log p(e|m)= -0.512:-0.512[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthampton[0.044]; Blackburn[0.044]; Sheffield[0.044]; Sheffield[0.044]; Manchester[0.043]; Sunderland[0.043]; Middlesbrough[0.043]; Bolton[0.043]; Coventry[0.043]; Chelsea[0.042]; Tottenham[0.042]; Leeds[0.042]; Liverpool[0.042]; Newcastle[0.042]; Nottingham[0.041]; Barnsley[0.041]; United[0.040]; United[0.040]; Wolverhampton[0.040]; Aston[0.040]; Everton[0.040]; Tranmere[0.040]; Derby[0.039]; Leicester[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bristol Rovers\u001b[39m ==> ENTITY: \u001b[32mBristol Rovers F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.213:0.213[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBristol[0.051]; Shrewsbury[0.044]; Plymouth[0.044]; Watford[0.044]; Burnley[0.043]; Rotherham[0.043]; Stockport[0.043]; Blackpool[0.042]; Bournemouth[0.042]; Millwall[0.042]; Peterborough[0.042]; Gillingham[0.041]; Wrexham[0.041]; Walsall[0.041]; Crewe[0.041]; Wycombe[0.041]; Brentford[0.041]; Luton[0.040]; Chesterfield[0.040]; Bury[0.040]; Preston[0.040]; York[0.038]; Notts[0.038]; Oldham[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Shrewsbury ==> ENTITY: \u001b[32mShrewsbury Town F.C.\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.214:0.214[0]; log p(e|m)= -3.244:-3.244[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Bristol[0.043]; Bristol[0.043]; Watford[0.043]; Chesterfield[0.043]; Stockport[0.043]; Plymouth[0.042]; Walsall[0.042]; City[0.042]; Peterborough[0.042]; Gillingham[0.042]; Rotherham[0.042]; Millwall[0.041]; Crewe[0.041]; Burnley[0.041]; Brentford[0.041]; Luton[0.041]; Bournemouth[0.041]; Wycombe[0.041]; Bury[0.040]; Wrexham[0.040]; York[0.040]; Blackpool[0.040]; Rovers[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wycombe ==> ENTITY: \u001b[32mWycombe Wanderers F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.208:0.208[0]; log p(e|m)= -1.146:-1.146[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Walsall[0.043]; Swansea[0.043]; Carlisle[0.043]; Shrewsbury[0.043]; Colchester[0.042]; Bristol[0.042]; Bristol[0.042]; Cardiff[0.042]; Northampton[0.042]; Rotherham[0.042]; Peterborough[0.042]; Scunthorpe[0.042]; Gillingham[0.042]; Plymouth[0.042]; City[0.041]; Bournemouth[0.041]; Notts[0.041]; Barnet[0.041]; Blackpool[0.040]; Rovers[0.040]; Scarborough[0.040]; Cambridge[0.040]; Division[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sheffield United\u001b[39m ==> ENTITY: \u001b[32mSheffield United F.C.\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCoventry[0.045]; Southampton[0.044]; Charlton[0.044]; Birmingham[0.043]; Sunderland[0.043]; Leeds[0.043]; Blackburn[0.043]; Middlesbrough[0.043]; Barnsley[0.042]; Norwich[0.042]; Park[0.042]; Swindon[0.042]; Tranmere[0.041]; Leicester[0.041]; Wolverhampton[0.041]; Division[0.041]; Huddersfield[0.041]; Bolton[0.041]; Nottingham[0.040]; Stoke[0.040]; Rangers[0.040]; Queens[0.039]; Palace[0.039]; Tottenham[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hull ==> ENTITY: \u001b[32mHull City A.F.C.\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.183:0.183[0]; log p(e|m)= -3.772:-3.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwansea[0.047]; Cardiff[0.047]; Rochdale[0.047]; Colchester[0.046]; Scunthorpe[0.046]; Carlisle[0.045]; Hereford[0.045]; Scarborough[0.045]; Hartlepool[0.044]; Northampton[0.044]; Exeter[0.044]; Doncaster[0.044]; Darlington[0.043]; Wigan[0.043]; Barnet[0.043]; Cambridge[0.043]; Brighton[0.042]; Chester[0.042]; Torquay[0.042]; Mansfield[0.042]; Lincoln[0.041]; Orient[0.040]; Leyton[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chelsea ==> ENTITY: \u001b[32mChelsea F.C.\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.196:0.196[0]; log p(e|m)= -0.518:-0.518[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.045]; league[0.044]; league[0.044]; Manchester[0.044]; London[0.043]; matches[0.043]; Sheffield[0.042]; Soccer[0.042]; Southampton[0.042]; soccer[0.042]; Blackburn[0.042]; Coventry[0.042]; Middlesbrough[0.042]; Leeds[0.041]; Liverpool[0.041]; Sunderland[0.041]; Tottenham[0.041]; played[0.041]; United[0.040]; Newcastle[0.040]; Arsenal[0.040]; Everton[0.040]; West[0.040]; Leicester[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Stoke ==> ENTITY: \u001b[32mStoke City F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.196:0.196[0]; log p(e|m)= -0.341:-0.341[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCoventry[0.044]; Charlton[0.044]; Wolverhampton[0.044]; Sheffield[0.044]; Manchester[0.043]; Portsmouth[0.043]; Norwich[0.043]; City[0.043]; Birmingham[0.043]; Southend[0.043]; Ipswich[0.043]; Nottingham[0.042]; Barnsley[0.042]; Swindon[0.041]; Reading[0.041]; Oxford[0.041]; United[0.040]; Bolton[0.040]; Park[0.040]; Huddersfield[0.040]; Queens[0.040]; Tranmere[0.040]; Vale[0.038]; Rangers[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swindon ==> ENTITY: \u001b[32mSwindon Town F.C.\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.208:0.208[0]; log p(e|m)= -2.489:-2.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.044]; Sheffield[0.044]; Reading[0.043]; Southend[0.043]; Portsmouth[0.043]; Barnsley[0.042]; Wolverhampton[0.042]; Ipswich[0.042]; Manchester[0.042]; Norwich[0.042]; Nottingham[0.042]; Tranmere[0.042]; Birmingham[0.042]; City[0.042]; Grimsby[0.041]; Oxford[0.041]; Huddersfield[0.041]; Bolton[0.041]; Division[0.040]; United[0.040]; Park[0.040]; Stoke[0.040]; Rangers[0.039]; West[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Fulham ==> ENTITY: \u001b[32mFulham F.C.\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.154:0.154[0]; log p(e|m)= -0.322:-0.322[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.046]; Cardiff[0.045]; Rotherham[0.044]; Wigan[0.043]; Northampton[0.043]; Colchester[0.042]; Swansea[0.042]; Peterborough[0.042]; Blackpool[0.042]; Walsall[0.042]; Shrewsbury[0.041]; Barnet[0.041]; Plymouth[0.041]; Wycombe[0.041]; Bournemouth[0.041]; Carlisle[0.041]; Scunthorpe[0.041]; Division[0.041]; Gillingham[0.041]; York[0.040]; Cambridge[0.040]; Notts[0.039]; Chester[0.039]; Scarborough[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wigan ==> ENTITY: \u001b[32mWigan Athletic F.C.\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.148:0.148[0]; log p(e|m)= -2.323:-2.323[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCardiff[0.045]; Hull[0.045]; Preston[0.044]; Northampton[0.044]; Colchester[0.044]; Blackpool[0.043]; Swansea[0.043]; Plymouth[0.043]; Carlisle[0.042]; Rotherham[0.042]; Peterborough[0.042]; Walsall[0.042]; Bournemouth[0.041]; Wycombe[0.041]; Fulham[0.041]; Scunthorpe[0.041]; Gillingham[0.040]; Chester[0.040]; Scarborough[0.040]; Barnet[0.040]; Torquay[0.040]; Cambridge[0.040]; Notts[0.039]; County[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oldham ==> ENTITY: \u001b[32mOldham Athletic A.F.C.\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.207:0.207[0]; log p(e|m)= -2.865:-2.865[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBradford[0.045]; Huddersfield[0.044]; Watford[0.043]; Bristol[0.043]; Bristol[0.043]; Chesterfield[0.043]; Stockport[0.043]; Bury[0.042]; Millwall[0.042]; Burnley[0.042]; Southend[0.042]; Brentford[0.041]; Rovers[0.041]; Ipswich[0.041]; City[0.041]; City[0.041]; Manchester[0.041]; Park[0.041]; Reading[0.041]; Portsmouth[0.041]; Wrexham[0.040]; Crewe[0.040]; Luton[0.040]; Division[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brentford ==> ENTITY: \u001b[32mBrentford F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.202:0.202[0]; log p(e|m)= -0.361:-0.361[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWatford[0.044]; Millwall[0.043]; Burnley[0.043]; Stockport[0.043]; Southend[0.043]; Luton[0.042]; Reading[0.042]; Bury[0.042]; Portsmouth[0.042]; Manchester[0.042]; Park[0.042]; Shrewsbury[0.042]; Oldham[0.042]; Bristol[0.041]; Bristol[0.041]; Chesterfield[0.041]; City[0.041]; City[0.041]; Grimsby[0.041]; Ipswich[0.041]; Crewe[0.041]; West[0.041]; Wrexham[0.041]; Rovers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Middlesbrough ==> ENTITY: \u001b[32mMiddlesbrough F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.183:0.183[0]; log p(e|m)= -0.518:-0.518[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthampton[0.045]; Leeds[0.044]; Sunderland[0.044]; Blackburn[0.043]; Tottenham[0.043]; Manchester[0.043]; Sheffield[0.043]; Sheffield[0.043]; Coventry[0.043]; United[0.042]; United[0.042]; Nottingham[0.042]; Barnsley[0.042]; Norwich[0.041]; Bolton[0.041]; Wolverhampton[0.041]; Newcastle[0.041]; Aston[0.040]; Chelsea[0.040]; Division[0.040]; Tranmere[0.040]; Everton[0.040]; Villa[0.039]; Leicester[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Blackpool ==> ENTITY: \u001b[32mBlackpool F.C.\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.192:0.192[0]; log p(e|m)= -0.853:-0.853[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBury[0.044]; Preston[0.044]; Wycombe[0.044]; Rotherham[0.043]; Rovers[0.043]; Walsall[0.043]; Watford[0.043]; Wrexham[0.043]; Stockport[0.042]; Gillingham[0.042]; Peterborough[0.042]; Bournemouth[0.041]; Shrewsbury[0.041]; Notts[0.041]; Luton[0.041]; Burnley[0.041]; Wigan[0.041]; Bristol[0.041]; Bristol[0.041]; Crewe[0.040]; Plymouth[0.040]; Chesterfield[0.040]; City[0.040]; Division[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Aston Villa\u001b[39m ==> ENTITY: \u001b[32mAston Villa F.C.\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.045]; league[0.044]; league[0.044]; matches[0.043]; Middlesbrough[0.043]; Sunderland[0.043]; Manchester[0.043]; Sheffield[0.042]; Soccer[0.042]; soccer[0.042]; Leeds[0.042]; goals[0.041]; goals[0.041]; won[0.041]; Tottenham[0.041]; Newcastle[0.041]; Liverpool[0.041]; played[0.041]; Chelsea[0.040]; Everton[0.040]; United[0.040]; Leicester[0.040]; Premier[0.040]; Arsenal[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBarnsley\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBarnsley F.C.\u001b[39m <---> \u001b[32mBurnley F.C.\u001b[39m\t\nSCORES: global= 0.273:0.258[\u001b[31m0.015\u001b[39m]; local()= 0.192:0.192[\u001b[32m0.000\u001b[39m]; log p(e|m)= -0.713:-2.830[\u001b[31m2.117\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwindon[0.045]; Stoke[0.045]; Middlesbrough[0.044]; Huddersfield[0.044]; Sunderland[0.044]; Leeds[0.043]; Charlton[0.043]; Wolverhampton[0.043]; Coventry[0.043]; Southampton[0.042]; Tranmere[0.042]; Blackburn[0.042]; Sheffield[0.042]; Norwich[0.042]; Birmingham[0.041]; Leicester[0.041]; Nottingham[0.041]; Bolton[0.040]; Park[0.040]; United[0.039]; Vale[0.039]; Oxford[0.039]; Rangers[0.038]; West[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rochdale ==> ENTITY: \u001b[32mRochdale A.F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.205:0.205[0]; log p(e|m)= -0.952:-0.952[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.051]; Carlisle[0.050]; Colchester[0.050]; Swansea[0.050]; Scunthorpe[0.049]; Darlington[0.049]; Chester[0.049]; Northampton[0.048]; Exeter[0.048]; Mansfield[0.048]; Doncaster[0.048]; Cardiff[0.048]; Scarborough[0.047]; Hartlepool[0.047]; Hereford[0.047]; Barnet[0.047]; Torquay[0.047]; Brighton[0.047]; Lincoln[0.045]; Orient[0.043]; Leyton[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Doncaster ==> ENTITY: \u001b[32mDoncaster Rovers F.C.\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.198:0.198[0]; log p(e|m)= -2.957:-2.957[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.058]; Swansea[0.056]; Rochdale[0.056]; Hereford[0.055]; Scunthorpe[0.055]; Exeter[0.054]; Colchester[0.054]; Scarborough[0.053]; Darlington[0.053]; Northampton[0.053]; Torquay[0.052]; Hartlepool[0.052]; Chester[0.052]; Brighton[0.051]; Barnet[0.051]; Mansfield[0.051]; Lincoln[0.049]; Orient[0.048]; Leyton[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Charlton ==> ENTITY: \u001b[32mCharlton Athletic F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.192:0.192[0]; log p(e|m)= -2.104:-2.104[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSheffield[0.044]; Norwich[0.044]; Manchester[0.044]; Portsmouth[0.043]; Bolton[0.043]; Reading[0.043]; Southend[0.043]; City[0.042]; Birmingham[0.042]; Barnsley[0.042]; Ipswich[0.042]; Park[0.042]; Wolverhampton[0.042]; Swindon[0.041]; Oxford[0.041]; United[0.041]; Stoke[0.041]; Tranmere[0.041]; Bradford[0.040]; Division[0.040]; Grimsby[0.040]; Huddersfield[0.040]; Rangers[0.040]; West[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Huddersfield ==> ENTITY: \u001b[32mHuddersfield Town F.C.\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.195:0.195[0]; log p(e|m)= -2.283:-2.283[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBradford[0.045]; Sheffield[0.044]; Wolverhampton[0.044]; Charlton[0.043]; Manchester[0.043]; Birmingham[0.043]; Park[0.042]; Oldham[0.042]; Barnsley[0.042]; Portsmouth[0.042]; City[0.042]; Swindon[0.042]; Southend[0.042]; Norwich[0.042]; Reading[0.041]; Tranmere[0.041]; Ipswich[0.041]; West[0.041]; Grimsby[0.040]; Rangers[0.040]; Bolton[0.040]; Oxford[0.040]; United[0.040]; Stoke[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cardiff ==> ENTITY: \u001b[32mCardiff City F.C.\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.203:0.203[0]; log p(e|m)= -3.474:-3.474[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.046]; Hull[0.045]; Swansea[0.045]; Carlisle[0.044]; Rotherham[0.043]; Exeter[0.043]; Plymouth[0.042]; Bournemouth[0.042]; Northampton[0.042]; Colchester[0.042]; Rochdale[0.042]; Scunthorpe[0.042]; Wycombe[0.041]; Peterborough[0.041]; Wigan[0.041]; Gillingham[0.041]; Scarborough[0.040]; Chester[0.040]; Barnet[0.040]; Fulham[0.040]; Torquay[0.039]; Cambridge[0.039]; Notts[0.039]; Division[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Derby ==> ENTITY: \u001b[32mDerby County F.C.\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.196:0.196[0]; log p(e|m)= -3.689:-3.689[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthampton[0.045]; league[0.044]; Sheffield[0.044]; Sheffield[0.044]; Coventry[0.044]; Blackburn[0.043]; matches[0.043]; Middlesbrough[0.043]; Sunderland[0.043]; Nottingham[0.043]; Manchester[0.043]; Leeds[0.042]; Bolton[0.041]; played[0.041]; Newcastle[0.041]; Leicester[0.040]; soccer[0.040]; Liverpool[0.040]; Aston[0.040]; Tottenham[0.040]; United[0.040]; United[0.040]; Chelsea[0.039]; Everton[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Millwall\u001b[39m ==> ENTITY: \u001b[32mMillwall F.C.\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWatford[0.044]; Stockport[0.043]; Brentford[0.043]; Portsmouth[0.042]; Manchester[0.042]; Shrewsbury[0.042]; Reading[0.042]; Chesterfield[0.042]; Burnley[0.042]; Southend[0.042]; Luton[0.042]; City[0.042]; City[0.042]; Crewe[0.042]; Bristol[0.042]; Bristol[0.042]; Oldham[0.041]; Ipswich[0.041]; Bury[0.041]; Rovers[0.041]; Wrexham[0.041]; Division[0.040]; Grimsby[0.040]; Bradford[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bury ==> ENTITY: \u001b[32mBury F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.722:-0.722[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBradford[0.044]; Stockport[0.043]; Grimsby[0.043]; Burnley[0.042]; Watford[0.042]; Ipswich[0.042]; Chesterfield[0.042]; Reading[0.042]; Southend[0.042]; Oldham[0.042]; City[0.042]; City[0.042]; Luton[0.042]; Brentford[0.042]; Blackpool[0.042]; Wrexham[0.041]; Millwall[0.041]; Manchester[0.041]; Shrewsbury[0.041]; Portsmouth[0.041]; Crewe[0.041]; Bristol[0.040]; Bristol[0.040]; Division[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Blackburn ==> ENTITY: \u001b[32mBlackburn Rovers F.C.\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.190:0.190[0]; log p(e|m)= -2.733:-2.733[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWolverhampton[0.044]; Sheffield[0.044]; Sheffield[0.044]; Sunderland[0.043]; Southampton[0.043]; Middlesbrough[0.043]; Manchester[0.043]; Norwich[0.042]; Leeds[0.042]; Coventry[0.042]; Birmingham[0.042]; Newcastle[0.042]; Derby[0.042]; Nottingham[0.042]; Bolton[0.042]; Barnsley[0.041]; Tranmere[0.041]; Leicester[0.041]; West[0.040]; Everton[0.040]; Tottenham[0.040]; United[0.040]; United[0.040]; Forest[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Walsall ==> ENTITY: \u001b[32mWalsall F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.620:-0.620[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Stockport[0.043]; Bristol[0.043]; Bristol[0.043]; Luton[0.042]; City[0.042]; Peterborough[0.042]; Burnley[0.042]; Watford[0.042]; Rotherham[0.042]; Shrewsbury[0.042]; Bournemouth[0.042]; Gillingham[0.041]; Crewe[0.041]; Wycombe[0.041]; Wrexham[0.041]; Carlisle[0.041]; Division[0.041]; Notts[0.041]; Blackpool[0.041]; Chesterfield[0.041]; Plymouth[0.040]; Rovers[0.040]; Wigan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Preston ==> ENTITY: \u001b[32mPreston North End F.C.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.168:0.168[0]; log p(e|m)= -3.194:-3.194[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCarlisle[0.044]; Watford[0.044]; Walsall[0.043]; Bristol[0.043]; Bristol[0.043]; Stockport[0.043]; Swansea[0.043]; Chesterfield[0.042]; City[0.042]; Gillingham[0.042]; Peterborough[0.042]; Cardiff[0.042]; Shrewsbury[0.042]; Blackpool[0.041]; Rotherham[0.041]; Bournemouth[0.041]; Plymouth[0.041]; Wycombe[0.041]; York[0.041]; Crewe[0.041]; Wrexham[0.040]; Cambridge[0.040]; Rovers[0.040]; County[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barnet ==> ENTITY: \u001b[32mBarnet F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.199:0.199[0]; log p(e|m)= -0.679:-0.679[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBournemouth[0.045]; Wycombe[0.044]; Colchester[0.044]; Hull[0.044]; Rochdale[0.043]; Rotherham[0.043]; Swansea[0.043]; Carlisle[0.043]; Peterborough[0.042]; Northampton[0.042]; Cambridge[0.042]; Scunthorpe[0.042]; Cardiff[0.042]; Notts[0.041]; Exeter[0.041]; Plymouth[0.041]; Mansfield[0.041]; Fulham[0.040]; Chester[0.040]; Torquay[0.040]; Doncaster[0.040]; Scarborough[0.039]; Wigan[0.039]; Division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Crystal Palace ==> ENTITY: \u001b[32mCrystal Palace F.C.\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.188:0.188[0]; log p(e|m)= -0.511:-0.511[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthampton[0.045]; Coventry[0.044]; Charlton[0.044]; Sheffield[0.044]; Sunderland[0.043]; Birmingham[0.043]; Blackburn[0.043]; Middlesbrough[0.043]; Bolton[0.042]; Leeds[0.042]; Ipswich[0.042]; Park[0.042]; Norwich[0.042]; Barnsley[0.042]; Tranmere[0.041]; Swindon[0.041]; Wolverhampton[0.041]; Stoke[0.040]; Huddersfield[0.040]; United[0.040]; Nottingham[0.040]; Queens[0.039]; Division[0.039]; West[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Leyton Orient\u001b[39m ==> ENTITY: \u001b[32mLeyton Orient F.C.\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.067]; Colchester[0.067]; Rochdale[0.065]; Northampton[0.065]; Scunthorpe[0.064]; Brighton[0.063]; Torquay[0.063]; Mansfield[0.062]; Exeter[0.062]; Darlington[0.061]; Chester[0.061]; Hartlepool[0.061]; Doncaster[0.061]; Hereford[0.060]; Scarborough[0.060]; Lincoln[0.058]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Darlington ==> ENTITY: \u001b[32mDarlington F.C.\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.188:0.188[0]; log p(e|m)= -1.871:-1.871[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.073]; Northampton[0.071]; Exeter[0.070]; Hartlepool[0.069]; Doncaster[0.069]; Mansfield[0.069]; Rochdale[0.068]; Scarborough[0.067]; Torquay[0.066]; Hereford[0.066]; Chester[0.065]; Brighton[0.064]; Lincoln[0.063]; Orient[0.061]; Leyton[0.057]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Everton ==> ENTITY: \u001b[32mEverton F.C.\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.199:0.199[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.045]; league[0.044]; league[0.044]; Blackburn[0.043]; Manchester[0.043]; Coventry[0.043]; soccer[0.042]; Southampton[0.042]; matches[0.042]; Sheffield[0.042]; Middlesbrough[0.042]; Nottingham[0.042]; Soccer[0.042]; Leeds[0.042]; Sunderland[0.042]; Forest[0.041]; played[0.041]; Aston[0.041]; Tottenham[0.040]; Newcastle[0.040]; United[0.040]; Liverpool[0.040]; Chelsea[0.039]; Leicester[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Queens Park Rangers\u001b[39m ==> ENTITY: \u001b[32mQueens Park Rangers F.C.\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.045]; Sheffield[0.044]; Manchester[0.043]; Brentford[0.043]; Southend[0.043]; Barnsley[0.042]; Birmingham[0.042]; Portsmouth[0.042]; City[0.042]; Reading[0.042]; Oldham[0.042]; Stoke[0.041]; Ipswich[0.041]; Norwich[0.041]; Tranmere[0.041]; Swindon[0.041]; Division[0.041]; Wolverhampton[0.041]; Bradford[0.041]; Huddersfield[0.040]; Grimsby[0.040]; United[0.040]; Vale[0.040]; Palace[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Port Vale\u001b[39m ==> ENTITY: \u001b[32mPort Vale F.C.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBury[0.043]; Southend[0.043]; Charlton[0.043]; Barnsley[0.043]; Portsmouth[0.042]; City[0.042]; Norwich[0.042]; Ipswich[0.042]; Swindon[0.042]; Oldham[0.042]; Brentford[0.042]; Park[0.042]; Manchester[0.042]; Reading[0.042]; Millwall[0.041]; Grimsby[0.041]; Birmingham[0.041]; Tranmere[0.041]; Bradford[0.041]; Huddersfield[0.041]; Division[0.040]; Wolverhampton[0.040]; Stoke[0.040]; Rangers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Arsenal ==> ENTITY: \u001b[32mArsenal F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.180:0.180[0]; log p(e|m)= -0.158:-0.158[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; league[0.045]; league[0.045]; matches[0.045]; Manchester[0.043]; Chelsea[0.043]; soccer[0.042]; won[0.042]; Sheffield[0.042]; Soccer[0.042]; Liverpool[0.042]; Tottenham[0.042]; Villa[0.041]; Leeds[0.041]; played[0.041]; Newcastle[0.040]; goals[0.040]; goals[0.040]; United[0.040]; Everton[0.040]; Aston[0.040]; Leicester[0.039]; lost[0.039]; Derby[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scunthorpe ==> ENTITY: \u001b[32mScunthorpe United F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.195:0.195[0]; log p(e|m)= -1.510:-1.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.047]; Swansea[0.044]; Cardiff[0.044]; Rotherham[0.044]; Rochdale[0.043]; Colchester[0.043]; Carlisle[0.043]; Peterborough[0.043]; Northampton[0.043]; Wycombe[0.042]; Doncaster[0.042]; Hereford[0.041]; Exeter[0.041]; Mansfield[0.041]; Chester[0.041]; Torquay[0.041]; Wigan[0.040]; Scarborough[0.040]; Barnet[0.040]; Division[0.039]; Fulham[0.039]; Notts[0.039]; Cambridge[0.039]; Lincoln[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Luton ==> ENTITY: \u001b[32mLuton Town F.C.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.213:0.213[0]; log p(e|m)= -2.551:-2.551[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWalsall[0.043]; Watford[0.043]; Stockport[0.042]; Chesterfield[0.042]; Shrewsbury[0.042]; Burnley[0.042]; Brentford[0.042]; Manchester[0.042]; Bristol[0.042]; Bristol[0.042]; Millwall[0.042]; Portsmouth[0.042]; Crewe[0.042]; Southend[0.042]; Blackpool[0.041]; Oldham[0.041]; Ipswich[0.041]; Bury[0.041]; Wrexham[0.041]; Bradford[0.041]; Grimsby[0.041]; City[0.040]; City[0.040]; Rovers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burnley ==> ENTITY: \u001b[32mBurnley F.C.\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.192:0.192[0]; log p(e|m)= -0.992:-0.992[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChesterfield[0.046]; Watford[0.043]; Stockport[0.043]; Walsall[0.043]; Oldham[0.043]; Bradford[0.042]; Gillingham[0.042]; Brentford[0.042]; Bury[0.042]; Millwall[0.042]; Southend[0.042]; Portsmouth[0.041]; Bristol[0.041]; Bristol[0.041]; Shrewsbury[0.041]; Wrexham[0.041]; Manchester[0.041]; City[0.041]; City[0.041]; Luton[0.041]; Reading[0.040]; Grimsby[0.040]; Crewe[0.040]; Blackpool[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.054:0.054[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.048]; matches[0.047]; league[0.047]; league[0.047]; Manchester[0.046]; Sheffield[0.045]; Soccer[0.044]; soccer[0.044]; Liverpool[0.044]; Tottenham[0.043]; Newcastle[0.043]; Arsenal[0.043]; Chelsea[0.043]; played[0.043]; Everton[0.042]; won[0.042]; United[0.042]; Wednesday[0.041]; Aston[0.041]; goals[0.041]; goals[0.041]; Premier[0.041]; Villa[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Southampton ==> ENTITY: \u001b[32mSouthampton F.C.\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.183:0.183[0]; log p(e|m)= -1.149:-1.149[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCoventry[0.047]; Norwich[0.047]; Sunderland[0.045]; Manchester[0.044]; Middlesbrough[0.044]; Barnsley[0.043]; Nottingham[0.043]; Oxford[0.043]; Chelsea[0.042]; Blackburn[0.042]; Wolverhampton[0.042]; Birmingham[0.041]; Sheffield[0.041]; Sheffield[0.041]; Leeds[0.041]; Tottenham[0.041]; Bolton[0.040]; Everton[0.040]; Crystal[0.040]; Tranmere[0.039]; Leicester[0.038]; United[0.038]; United[0.038]; Derby[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Plymouth ==> ENTITY: \u001b[32mPlymouth Argyle F.C.\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.207:0.207[0]; log p(e|m)= -4.075:-4.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.046]; Watford[0.044]; Walsall[0.043]; Swansea[0.043]; Colchester[0.043]; Carlisle[0.042]; Cardiff[0.042]; Bristol[0.042]; Bristol[0.042]; Gillingham[0.042]; City[0.042]; Peterborough[0.042]; Rotherham[0.042]; Shrewsbury[0.042]; Blackpool[0.041]; Crewe[0.041]; Wycombe[0.041]; Bournemouth[0.041]; Division[0.040]; Rovers[0.040]; County[0.040]; Wrexham[0.040]; Cambridge[0.040]; Barnet[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Notts County\u001b[39m ==> ENTITY: \u001b[32mNotts County F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.216:0.216[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Walsall[0.043]; Division[0.043]; Colchester[0.043]; Gillingham[0.043]; Scunthorpe[0.043]; Swansea[0.042]; Peterborough[0.042]; Rotherham[0.042]; Shrewsbury[0.042]; Bournemouth[0.042]; Carlisle[0.041]; Cardiff[0.041]; Wycombe[0.041]; Plymouth[0.041]; Crewe[0.041]; Bristol[0.041]; Bristol[0.041]; Barnet[0.041]; City[0.041]; Northampton[0.041]; Rovers[0.041]; Blackpool[0.040]; Fulham[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Newcastle ==> ENTITY: \u001b[32mNewcastle United F.C.\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.192:0.192[0]; log p(e|m)= -2.489:-2.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.045]; league[0.045]; League[0.044]; played[0.044]; Leeds[0.044]; matches[0.044]; Sunderland[0.043]; Manchester[0.043]; Blackburn[0.043]; Sheffield[0.043]; Leicester[0.043]; Middlesbrough[0.043]; soccer[0.041]; Soccer[0.041]; Liverpool[0.041]; London[0.040]; United[0.040]; Tottenham[0.040]; Chelsea[0.040]; goals[0.039]; goals[0.039]; Everton[0.039]; West[0.039]; won[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wolverhampton ==> ENTITY: \u001b[32mWolverhampton Wanderers F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.191:0.191[0]; log p(e|m)= -2.273:-2.273[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNottingham[0.044]; Charlton[0.044]; Southampton[0.043]; Sheffield[0.043]; Coventry[0.043]; Blackburn[0.043]; Birmingham[0.043]; Middlesbrough[0.043]; Sunderland[0.043]; Oxford[0.042]; Manchester[0.042]; Bolton[0.042]; Park[0.042]; Barnsley[0.041]; Swindon[0.041]; City[0.041]; Tranmere[0.041]; Norwich[0.041]; Ipswich[0.041]; Stoke[0.040]; Huddersfield[0.040]; United[0.039]; West[0.039]; Division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brighton ==> ENTITY: \u001b[32mBrighton & Hove Albion F.C.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.175:0.175[0]; log p(e|m)= -3.124:-3.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.084]; Exeter[0.082]; Chester[0.080]; Rochdale[0.080]; Hereford[0.078]; Torquay[0.077]; Darlington[0.076]; Mansfield[0.076]; Hartlepool[0.076]; Doncaster[0.075]; Leyton[0.073]; Orient[0.071]; Lincoln[0.071]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scarborough ==> ENTITY: \u001b[32mScarborough F.C.\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.201:0.201[0]; log p(e|m)= -1.814:-1.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRochdale[0.044]; Rotherham[0.044]; Carlisle[0.044]; Hull[0.043]; Swansea[0.043]; Darlington[0.043]; Hartlepool[0.043]; Cardiff[0.043]; Colchester[0.042]; Doncaster[0.042]; Exeter[0.042]; Wycombe[0.042]; Torquay[0.042]; Cambridge[0.042]; Scunthorpe[0.042]; Chester[0.042]; Division[0.042]; Hereford[0.041]; Barnet[0.041]; Northampton[0.041]; Mansfield[0.041]; Wigan[0.039]; Lincoln[0.037]; Leyton[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ipswich ==> ENTITY: \u001b[32mIpswich Town F.C.\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.194:0.194[0]; log p(e|m)= -2.688:-2.688[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.045]; Manchester[0.044]; Portsmouth[0.043]; Birmingham[0.043]; Norwich[0.043]; Millwall[0.043]; Brentford[0.042]; Reading[0.042]; Southend[0.042]; Tranmere[0.042]; Oldham[0.042]; Swindon[0.042]; Luton[0.042]; City[0.041]; Bury[0.041]; Stoke[0.041]; Park[0.041]; Wolverhampton[0.041]; Huddersfield[0.041]; Bradford[0.041]; Oxford[0.041]; Grimsby[0.040]; Division[0.040]; Rangers[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Manchester City\u001b[39m ==> ENTITY: \u001b[32mManchester City F.C.\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.046]; Portsmouth[0.045]; Birmingham[0.044]; Division[0.043]; Burnley[0.043]; Millwall[0.043]; Norwich[0.042]; Ipswich[0.042]; Brentford[0.042]; Southend[0.042]; Reading[0.042]; Park[0.041]; Stoke[0.041]; Luton[0.041]; Swindon[0.041]; Bury[0.041]; Tranmere[0.041]; Wolverhampton[0.041]; Grimsby[0.040]; Oldham[0.040]; Huddersfield[0.040]; Rangers[0.040]; Bradford[0.040]; Queens[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chester ==> ENTITY: \u001b[32mChester City F.C.\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.192:0.192[0]; log p(e|m)= -3.219:-3.219[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.046]; Carlisle[0.044]; Swansea[0.044]; Colchester[0.043]; Darlington[0.043]; Scunthorpe[0.043]; Rochdale[0.043]; Northampton[0.043]; Cardiff[0.043]; Exeter[0.042]; Torquay[0.042]; Barnet[0.042]; Hartlepool[0.041]; Hereford[0.041]; Doncaster[0.041]; Brighton[0.041]; Division[0.041]; Scarborough[0.041]; Cambridge[0.041]; Mansfield[0.040]; Fulham[0.039]; Orient[0.039]; Wigan[0.039]; Lincoln[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leicester ==> ENTITY: \u001b[32mLeicester City F.C.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.195:0.195[0]; log p(e|m)= -3.219:-3.219[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCoventry[0.045]; Nottingham[0.044]; Manchester[0.044]; league[0.044]; Southampton[0.043]; Blackburn[0.043]; Leeds[0.043]; Sunderland[0.043]; Sheffield[0.042]; Sheffield[0.042]; Middlesbrough[0.042]; Liverpool[0.042]; Bolton[0.041]; Barnsley[0.041]; Everton[0.041]; United[0.041]; United[0.041]; Arsenal[0.040]; Newcastle[0.040]; Chelsea[0.040]; West[0.040]; goals[0.040]; goals[0.040]; Division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chesterfield ==> ENTITY: \u001b[32mChesterfield F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.200:0.200[0]; log p(e|m)= -1.269:-1.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Stockport[0.043]; Bradford[0.043]; Bristol[0.043]; Bristol[0.043]; Walsall[0.042]; Oldham[0.042]; Shrewsbury[0.042]; York[0.042]; Portsmouth[0.042]; Watford[0.042]; Burnley[0.042]; Bury[0.042]; Millwall[0.042]; Southend[0.041]; Luton[0.041]; Brentford[0.041]; Gillingham[0.041]; Reading[0.041]; Wrexham[0.041]; City[0.040]; Crewe[0.040]; Blackpool[0.040]; Grimsby[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Stockport ==> ENTITY: \u001b[32mStockport County F.C.\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.221:0.221[0]; log p(e|m)= -2.526:-2.526[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.044]; York[0.043]; Bradford[0.043]; Walsall[0.042]; Bury[0.042]; Chesterfield[0.042]; Southend[0.042]; Watford[0.042]; Bristol[0.042]; Bristol[0.042]; Bournemouth[0.042]; Burnley[0.042]; Gillingham[0.042]; Brentford[0.041]; Shrewsbury[0.041]; Wrexham[0.041]; Luton[0.041]; Crewe[0.041]; Oldham[0.041]; Millwall[0.041]; City[0.041]; Blackpool[0.041]; Grimsby[0.040]; Reading[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Liverpool ==> ENTITY: \u001b[32mLiverpool F.C.\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.182:0.182[0]; log p(e|m)= -1.171:-1.171[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; league[0.044]; league[0.044]; Manchester[0.044]; matches[0.044]; Sunderland[0.043]; Leeds[0.042]; Tottenham[0.042]; Sheffield[0.042]; Soccer[0.042]; London[0.042]; soccer[0.042]; Newcastle[0.041]; Chelsea[0.041]; Leicester[0.041]; Premier[0.041]; Everton[0.041]; United[0.041]; played[0.040]; Arsenal[0.040]; Villa[0.040]; West[0.040]; English[0.040]; English[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Coventry ==> ENTITY: \u001b[32mCoventry City F.C.\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.186:0.186[0]; log p(e|m)= -3.124:-3.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeeds[0.045]; Blackburn[0.044]; Southampton[0.044]; Leicester[0.044]; Sheffield[0.044]; Sheffield[0.044]; Nottingham[0.044]; Birmingham[0.043]; Middlesbrough[0.042]; Sunderland[0.042]; Bolton[0.042]; Oxford[0.041]; Barnsley[0.041]; Tranmere[0.041]; Norwich[0.041]; United[0.041]; Wolverhampton[0.040]; Tottenham[0.040]; Stoke[0.040]; West[0.040]; Chelsea[0.039]; Everton[0.039]; Palace[0.039]; Wednesday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rotherham ==> ENTITY: \u001b[32mRotherham United F.C.\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.207:0.207[0]; log p(e|m)= -2.645:-2.645[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.048]; Bristol[0.044]; Walsall[0.043]; Colchester[0.043]; Swansea[0.043]; Cardiff[0.043]; Gillingham[0.043]; Carlisle[0.043]; Northampton[0.043]; Scunthorpe[0.043]; Rovers[0.042]; Shrewsbury[0.042]; Wycombe[0.042]; Peterborough[0.042]; Plymouth[0.041]; Bournemouth[0.041]; Blackpool[0.041]; Division[0.040]; Barnet[0.040]; York[0.040]; Notts[0.039]; Wigan[0.039]; Scarborough[0.039]; Fulham[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = English ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.095:0.095[0]; log p(e|m)= -1.355:-1.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEnglish[0.050]; League[0.050]; league[0.049]; league[0.049]; matches[0.047]; Sheffield[0.045]; played[0.045]; Premier[0.043]; London[0.043]; Soccer[0.043]; Manchester[0.043]; soccer[0.043]; Newcastle[0.042]; won[0.042]; United[0.042]; Aston[0.042]; Saturday[0.042]; Liverpool[0.041]; Wednesday[0.041]; Standings[0.040]; Standings[0.040]; points[0.040]; lost[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Torquay ==> ENTITY: \u001b[32mTorquay United F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.185:0.185[0]; log p(e|m)= -2.017:-2.017[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.050]; Carlisle[0.049]; Swansea[0.048]; Colchester[0.048]; Northampton[0.047]; Rochdale[0.047]; Scunthorpe[0.047]; Exeter[0.047]; Cardiff[0.047]; Chester[0.046]; Hereford[0.046]; Darlington[0.046]; Brighton[0.045]; Mansfield[0.045]; Hartlepool[0.045]; Doncaster[0.044]; Barnet[0.044]; Scarborough[0.044]; Wigan[0.043]; Lincoln[0.042]; Orient[0.041]; Leyton[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tranmere ==> ENTITY: \u001b[32mTranmere Rovers F.C.\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.219:0.219[0]; log p(e|m)= -0.600:-0.600[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.045]; Southampton[0.045]; Blackburn[0.044]; Coventry[0.043]; Middlesbrough[0.043]; Portsmouth[0.043]; Barnsley[0.043]; Sheffield[0.042]; Manchester[0.042]; Birmingham[0.042]; Ipswich[0.042]; Swindon[0.042]; Norwich[0.042]; Huddersfield[0.041]; Park[0.041]; Bolton[0.041]; City[0.041]; Wolverhampton[0.041]; Stoke[0.040]; United[0.040]; Nottingham[0.039]; Port[0.039]; Division[0.039]; Vale[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sheffield Wednesday\u001b[39m ==> ENTITY: \u001b[32mSheffield Wednesday F.C.\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.047]; league[0.045]; league[0.045]; matches[0.043]; Coventry[0.043]; Leeds[0.043]; Soccer[0.043]; Southampton[0.042]; Blackburn[0.042]; Division[0.042]; Sunderland[0.042]; Manchester[0.041]; Middlesbrough[0.041]; Standings[0.041]; Standings[0.041]; played[0.041]; Leicester[0.040]; soccer[0.040]; Saturday[0.040]; Newcastle[0.040]; goals[0.040]; goals[0.040]; won[0.040]; Bolton[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portsmouth ==> ENTITY: \u001b[32mPortsmouth F.C.\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.199:0.199[0]; log p(e|m)= -1.231:-1.231[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.046]; Millwall[0.043]; Burnley[0.043]; Manchester[0.043]; Birmingham[0.043]; Norwich[0.043]; Ipswich[0.042]; Brentford[0.042]; Southend[0.042]; City[0.042]; Reading[0.042]; Chesterfield[0.042]; Tranmere[0.041]; Park[0.041]; Luton[0.041]; Bury[0.041]; Stoke[0.041]; Swindon[0.041]; Grimsby[0.041]; Oldham[0.041]; Division[0.040]; Huddersfield[0.040]; Bradford[0.040]; West[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Reading ==> ENTITY: \u001b[32mReading F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.175:0.175[0]; log p(e|m)= -1.201:-1.201[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.044]; Oxford[0.043]; Swindon[0.042]; Millwall[0.042]; Division[0.042]; Birmingham[0.042]; Manchester[0.042]; Burnley[0.042]; Norwich[0.042]; Brentford[0.042]; Stockport[0.042]; Portsmouth[0.042]; City[0.042]; Ipswich[0.042]; West[0.041]; Luton[0.041]; Chesterfield[0.041]; Park[0.041]; Bury[0.041]; Southend[0.041]; Bradford[0.041]; Oldham[0.041]; Huddersfield[0.040]; Rangers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Exeter ==> ENTITY: \u001b[32mExeter City F.C.\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.209:0.209[0]; log p(e|m)= -3.650:-3.650[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.054]; Cardiff[0.053]; Colchester[0.053]; Swansea[0.052]; Northampton[0.052]; Rochdale[0.052]; Chester[0.052]; Scunthorpe[0.051]; Hereford[0.051]; Brighton[0.050]; Darlington[0.050]; Mansfield[0.050]; Barnet[0.050]; Torquay[0.050]; Scarborough[0.049]; Hartlepool[0.048]; Doncaster[0.048]; Orient[0.046]; Lincoln[0.045]; Leyton[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leeds ==> ENTITY: \u001b[32mLeeds United A.F.C.\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.202:0.202[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNottingham[0.044]; Blackburn[0.044]; Sheffield[0.044]; Sheffield[0.044]; Coventry[0.044]; Southampton[0.044]; league[0.043]; Manchester[0.043]; Middlesbrough[0.043]; Sunderland[0.043]; Leicester[0.042]; Barnsley[0.042]; Newcastle[0.042]; Bolton[0.042]; West[0.041]; United[0.041]; United[0.041]; Liverpool[0.040]; Derby[0.040]; Tottenham[0.040]; Everton[0.039]; Aston[0.039]; Chelsea[0.039]; Wednesday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Watford ==> ENTITY: \u001b[32mWatford F.C.\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.192:0.192[0]; log p(e|m)= -0.960:-0.960[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Brentford[0.043]; Burnley[0.043]; Walsall[0.043]; Millwall[0.042]; Bristol[0.042]; Bristol[0.042]; Stockport[0.042]; Shrewsbury[0.042]; Bury[0.042]; Chesterfield[0.042]; Luton[0.042]; Blackpool[0.042]; Southend[0.042]; Bournemouth[0.042]; Gillingham[0.042]; Crewe[0.041]; Oldham[0.041]; City[0.041]; Bradford[0.040]; Wrexham[0.040]; Plymouth[0.040]; Grimsby[0.040]; Rovers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wrexham ==> ENTITY: \u001b[32mWrexham F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.205:0.205[0]; log p(e|m)= -0.882:-0.882[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.047]; Stockport[0.043]; Walsall[0.043]; Chesterfield[0.042]; Bristol[0.042]; Bristol[0.042]; Shrewsbury[0.042]; Peterborough[0.042]; Watford[0.042]; Bournemouth[0.042]; Gillingham[0.042]; Bury[0.042]; Oldham[0.042]; Grimsby[0.041]; Luton[0.041]; Plymouth[0.041]; Brentford[0.041]; Millwall[0.041]; Burnley[0.041]; Bradford[0.041]; Southend[0.041]; Crewe[0.041]; City[0.040]; Blackpool[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nottingham Forest\u001b[39m ==> ENTITY: \u001b[32mNottingham Forest F.C.\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthampton[0.044]; Coventry[0.044]; Charlton[0.044]; Blackburn[0.043]; Birmingham[0.043]; Leeds[0.043]; Sunderland[0.043]; Middlesbrough[0.043]; Sheffield[0.043]; Sheffield[0.043]; Bolton[0.042]; Barnsley[0.042]; Norwich[0.041]; Swindon[0.041]; Leicester[0.041]; Division[0.041]; Tranmere[0.041]; United[0.041]; Stoke[0.040]; Wolverhampton[0.040]; Tottenham[0.040]; Everton[0.039]; West[0.039]; Oxford[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Carlisle ==> ENTITY: \u001b[32mCarlisle United F.C.\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.180:0.180[0]; log p(e|m)= -3.612:-3.612[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Hull[0.044]; Rochdale[0.044]; Peterborough[0.044]; Cardiff[0.043]; Colchester[0.043]; Walsall[0.043]; Swansea[0.042]; Bournemouth[0.042]; Wycombe[0.042]; Rotherham[0.042]; Northampton[0.042]; Gillingham[0.042]; Plymouth[0.042]; Scarborough[0.042]; Scunthorpe[0.042]; Division[0.041]; Chester[0.041]; Cambridge[0.040]; Torquay[0.040]; Barnet[0.040]; Notts[0.039]; County[0.038]; Fulham[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swansea ==> ENTITY: \u001b[32mSwansea City A.F.C.\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.174:0.174[0]; log p(e|m)= -2.604:-2.604[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCardiff[0.047]; Preston[0.047]; Hull[0.044]; Rotherham[0.043]; Carlisle[0.043]; Northampton[0.043]; Colchester[0.043]; Wycombe[0.042]; Rochdale[0.042]; Scunthorpe[0.042]; Peterborough[0.042]; Fulham[0.041]; Chester[0.041]; Barnet[0.041]; Plymouth[0.041]; Bournemouth[0.041]; Division[0.040]; Exeter[0.040]; Scarborough[0.040]; Wigan[0.040]; Torquay[0.040]; Doncaster[0.039]; Notts[0.039]; Cambridge[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Manchester United\u001b[39m ==> ENTITY: \u001b[32mManchester United F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.049]; league[0.049]; League[0.044]; matches[0.042]; Leeds[0.042]; Middlesbrough[0.042]; Liverpool[0.042]; Sheffield[0.042]; Southampton[0.042]; Blackburn[0.042]; Sunderland[0.041]; soccer[0.041]; Soccer[0.041]; Newcastle[0.041]; English[0.041]; English[0.041]; Tottenham[0.041]; Chelsea[0.040]; Arsenal[0.040]; goals[0.040]; goals[0.040]; Everton[0.039]; Leicester[0.039]; played[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norwich ==> ENTITY: \u001b[32mNorwich City F.C.\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.199:0.199[0]; log p(e|m)= -2.733:-2.733[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthampton[0.045]; Charlton[0.044]; Coventry[0.043]; Wolverhampton[0.043]; Blackburn[0.043]; Sheffield[0.043]; Birmingham[0.043]; Portsmouth[0.043]; Manchester[0.043]; Middlesbrough[0.042]; Barnsley[0.042]; Ipswich[0.042]; Bolton[0.042]; Swindon[0.041]; City[0.041]; Reading[0.041]; Park[0.041]; Nottingham[0.040]; Oxford[0.040]; Tranmere[0.040]; Stoke[0.040]; United[0.040]; Division[0.039]; Queens[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = English ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.095:0.095[0]; log p(e|m)= -1.355:-1.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEnglish[0.050]; League[0.050]; league[0.049]; league[0.049]; matches[0.047]; Sheffield[0.045]; played[0.045]; Premier[0.043]; London[0.043]; Soccer[0.043]; Manchester[0.043]; soccer[0.043]; Newcastle[0.042]; won[0.042]; United[0.042]; Aston[0.042]; Saturday[0.042]; Liverpool[0.041]; Wednesday[0.041]; Standings[0.040]; Standings[0.040]; points[0.040]; lost[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Premier league\u001b[39m ==> ENTITY: \u001b[32mPremier League\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.048]; league[0.046]; matches[0.045]; Soccer[0.045]; Manchester[0.045]; Sheffield[0.045]; Leeds[0.044]; soccer[0.044]; won[0.044]; Tottenham[0.044]; Liverpool[0.044]; Standings[0.043]; Standings[0.043]; played[0.043]; Newcastle[0.043]; Everton[0.043]; Leicester[0.042]; United[0.042]; English[0.042]; English[0.042]; Arsenal[0.042]; Chelsea[0.042]; Villa[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gillingham ==> ENTITY: \u001b[32mGillingham F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.201:0.201[0]; log p(e|m)= -0.257:-0.257[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Watford[0.043]; Walsall[0.043]; Stockport[0.043]; Peterborough[0.043]; City[0.042]; Chesterfield[0.042]; Burnley[0.042]; Notts[0.042]; Rotherham[0.042]; Wycombe[0.042]; Carlisle[0.042]; Bristol[0.042]; Bristol[0.042]; Cardiff[0.041]; Blackpool[0.041]; Shrewsbury[0.041]; Bournemouth[0.041]; Crewe[0.041]; Plymouth[0.041]; Wrexham[0.041]; County[0.040]; Rovers[0.040]; Wigan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Bromwich ==> ENTITY: \u001b[32mWest Bromwich Albion F.C.\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.199:0.199[0]; log p(e|m)= -1.501:-1.501[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.044]; Watford[0.044]; Birmingham[0.044]; City[0.043]; Bradford[0.043]; Millwall[0.042]; Oxford[0.042]; Manchester[0.042]; Portsmouth[0.042]; Burnley[0.042]; Swindon[0.042]; Bury[0.041]; Oldham[0.041]; Brentford[0.041]; Reading[0.041]; Stockport[0.041]; Chesterfield[0.041]; Ipswich[0.041]; Southend[0.041]; Luton[0.041]; Park[0.040]; Division[0.040]; Stoke[0.040]; Rangers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Southend ==> ENTITY: \u001b[32mSouthend United F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.225:0.225[0]; log p(e|m)= -1.772:-1.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.044]; Chesterfield[0.043]; Millwall[0.043]; Stockport[0.043]; Oldham[0.043]; Watford[0.043]; Reading[0.043]; Brentford[0.042]; Burnley[0.042]; Portsmouth[0.042]; Manchester[0.042]; Bury[0.042]; City[0.042]; Huddersfield[0.042]; Swindon[0.041]; Ipswich[0.041]; Luton[0.041]; Park[0.041]; Wrexham[0.041]; Grimsby[0.041]; Bradford[0.041]; Queens[0.039]; Stoke[0.039]; Division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hereford ==> ENTITY: \u001b[32mHereford United F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.188:0.188[0]; log p(e|m)= -2.765:-2.765[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.068]; Rochdale[0.067]; Scunthorpe[0.066]; Northampton[0.066]; Exeter[0.064]; Chester[0.064]; Torquay[0.064]; Darlington[0.063]; Hartlepool[0.063]; Doncaster[0.061]; Brighton[0.061]; Scarborough[0.061]; Mansfield[0.061]; Lincoln[0.057]; Orient[0.057]; Leyton[0.056]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sunderland ==> ENTITY: \u001b[32mSunderland A.F.C.\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.396:-0.396[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthampton[0.044]; Coventry[0.044]; Middlesbrough[0.044]; Sheffield[0.044]; Sheffield[0.044]; Blackburn[0.043]; Newcastle[0.043]; Barnsley[0.043]; Manchester[0.043]; Leeds[0.043]; Wolverhampton[0.042]; Everton[0.042]; Leicester[0.041]; Nottingham[0.041]; Bolton[0.041]; Liverpool[0.041]; Tottenham[0.041]; United[0.040]; United[0.040]; Chelsea[0.040]; West[0.039]; Division[0.039]; Derby[0.039]; Aston[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Crewe ==> ENTITY: \u001b[32mCrewe Alexandra F.C.\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.191:0.191[0]; log p(e|m)= -2.830:-2.830[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Watford[0.043]; Bristol[0.043]; Bristol[0.043]; Shrewsbury[0.042]; Chesterfield[0.042]; Walsall[0.042]; Plymouth[0.042]; Burnley[0.042]; Stockport[0.042]; Peterborough[0.041]; Oldham[0.041]; Brentford[0.041]; Gillingham[0.041]; Bury[0.041]; Grimsby[0.041]; Bradford[0.041]; Bournemouth[0.041]; Millwall[0.041]; Luton[0.041]; York[0.041]; Blackpool[0.041]; Wrexham[0.040]; City[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bristol City\u001b[39m ==> ENTITY: \u001b[32mBristol City F.C.\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBristol[0.045]; Preston[0.044]; Watford[0.043]; Walsall[0.043]; Gillingham[0.042]; Chesterfield[0.042]; Stockport[0.042]; Brentford[0.042]; Millwall[0.042]; Shrewsbury[0.042]; Burnley[0.042]; Bournemouth[0.041]; Peterborough[0.041]; Plymouth[0.041]; Crewe[0.041]; Bury[0.041]; Wycombe[0.041]; Blackpool[0.041]; Luton[0.041]; Oldham[0.041]; Rovers[0.040]; Bradford[0.040]; Division[0.040]; Wrexham[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bournemouth ==> ENTITY: \u001b[32mA.F.C. Bournemouth\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.194:0.194[0]; log p(e|m)= -1.058:-1.058[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBlackpool[0.045]; Stockport[0.044]; Walsall[0.043]; Shrewsbury[0.043]; Wycombe[0.043]; Wrexham[0.043]; Swansea[0.043]; Preston[0.043]; Notts[0.043]; Gillingham[0.043]; Rovers[0.042]; Rotherham[0.042]; Peterborough[0.042]; Crewe[0.041]; Cardiff[0.041]; Watford[0.041]; Barnet[0.041]; Bristol[0.041]; Bristol[0.041]; Wigan[0.041]; City[0.041]; Plymouth[0.039]; Carlisle[0.038]; Fulham[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bolton ==> ENTITY: \u001b[32mBolton Wanderers F.C.\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.178:0.178[0]; log p(e|m)= -2.283:-2.283[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.044]; Blackburn[0.044]; Sunderland[0.043]; Southampton[0.043]; Coventry[0.043]; Leeds[0.043]; Leicester[0.043]; Middlesbrough[0.042]; Sheffield[0.042]; Norwich[0.042]; Tranmere[0.042]; Wolverhampton[0.042]; Birmingham[0.042]; Barnsley[0.041]; Nottingham[0.041]; Huddersfield[0.041]; United[0.041]; Swindon[0.041]; Tottenham[0.041]; Oxford[0.040]; Stoke[0.040]; Wednesday[0.040]; West[0.040]; Division[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colchester ==> ENTITY: \u001b[32mColchester United F.C.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.181:0.181[0]; log p(e|m)= -2.865:-2.865[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.046]; Swansea[0.044]; Peterborough[0.044]; Exeter[0.044]; Carlisle[0.043]; Cambridge[0.043]; Northampton[0.043]; Rochdale[0.043]; Plymouth[0.043]; Scunthorpe[0.042]; Rotherham[0.042]; Wycombe[0.042]; Cardiff[0.042]; Barnet[0.042]; Mansfield[0.041]; Chester[0.041]; Torquay[0.040]; Scarborough[0.040]; Doncaster[0.040]; Division[0.040]; Fulham[0.040]; Notts[0.039]; Wigan[0.039]; Leyton[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Northampton ==> ENTITY: \u001b[32mNorthampton Town F.C.\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.217:0.217[0]; log p(e|m)= -3.244:-3.244[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.045]; Cardiff[0.044]; Colchester[0.044]; Swansea[0.044]; Wycombe[0.044]; Rochdale[0.043]; Scunthorpe[0.043]; Carlisle[0.043]; Rotherham[0.042]; Chester[0.042]; Cambridge[0.041]; Barnet[0.041]; Exeter[0.041]; Torquay[0.041]; Hereford[0.041]; Darlington[0.041]; Doncaster[0.040]; Mansfield[0.040]; Division[0.040]; Notts[0.040]; Wigan[0.040]; Scarborough[0.039]; Orient[0.039]; Fulham[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mansfield ==> ENTITY: \u001b[32mMansfield Town F.C.\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.215:0.215[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.061]; Colchester[0.059]; Rochdale[0.058]; Northampton[0.058]; Hereford[0.057]; Exeter[0.057]; Scunthorpe[0.057]; Chester[0.057]; Darlington[0.056]; Hartlepool[0.055]; Doncaster[0.055]; Brighton[0.055]; Barnet[0.055]; Torquay[0.055]; Scarborough[0.055]; Lincoln[0.051]; Orient[0.050]; Leyton[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hartlepool ==> ENTITY: \u001b[32mHartlepool United F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.165:0.165[0]; log p(e|m)= -1.754:-1.754[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.078]; Hereford[0.075]; Rochdale[0.074]; Darlington[0.073]; Brighton[0.073]; Doncaster[0.072]; Chester[0.072]; Exeter[0.072]; Scarborough[0.072]; Mansfield[0.071]; Torquay[0.071]; Lincoln[0.066]; Leyton[0.065]; Orient[0.065]; \t\n\r [=========================================>....] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s698ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4150/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1274testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hartford ==> ENTITY: \u001b[32mHartford, Connecticut\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= -0.033:-0.033[0]; log p(e|m)= -0.580:-0.580[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHartford[0.060]; Connecticut[0.059]; City[0.052]; Conn[0.051]; General[0.049]; List[0.048]; Date[0.048]; A1[0.047]; Obligation[0.047]; Delivery[0.047]; Insured[0.046]; Moody[0.046]; mln[0.045]; Balance[0.045]; Balance[0.045]; Coupon[0.045]; Maturity[0.044]; Aaa[0.044]; Aaa[0.044]; Aa[0.044]; Bonds[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = S&P\u001b[39m ==> ENTITY: \u001b[32mStandard & Poor's\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.073:0.073[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoody[0.055]; Aaa[0.049]; Aaa[0.049]; Hartford[0.047]; Hartford[0.047]; List[0.047]; A1[0.047]; Connecticut[0.046]; Aa[0.044]; Balance[0.044]; Balance[0.044]; General[0.044]; Insured[0.044]; Obligation[0.044]; City[0.044]; mln[0.044]; Maturity[0.044]; Delivery[0.043]; Date[0.043]; Conn[0.043]; Bonds[0.043]; Coupon[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = FSA ==> ENTITY: \u001b[32mFinancial Services Authority\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.030:0.030[0]; log p(e|m)= -1.155:-1.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInsured[0.051]; List[0.050]; Obligation[0.049]; Aaa[0.048]; Aaa[0.048]; General[0.048]; Aa[0.048]; Balance[0.048]; Balance[0.048]; Conn[0.048]; Delivery[0.047]; Hartford[0.047]; Hartford[0.047]; mln[0.047]; Maturity[0.047]; Coupon[0.047]; Connecticut[0.047]; City[0.047]; Moody[0.046]; Date[0.046]; A1[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hartford ==> ENTITY: \u001b[32mHartford, Connecticut\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= -0.033:-0.033[0]; log p(e|m)= -0.580:-0.580[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHartford[0.060]; Connecticut[0.059]; City[0.052]; Conn[0.051]; General[0.049]; List[0.048]; Date[0.048]; A1[0.047]; Obligation[0.047]; Delivery[0.047]; Insured[0.046]; Moody[0.046]; mln[0.045]; Balance[0.045]; Balance[0.045]; Coupon[0.045]; Maturity[0.044]; Aaa[0.044]; Aaa[0.044]; Aa[0.044]; Bonds[0.043]; \t\n\r [=========================================>....] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s690ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4154/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1219testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.064:0.064[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnumber[0.047]; London[0.045]; latest[0.045]; date[0.044]; date[0.044]; date[0.044]; deal[0.043]; Department[0.042]; service[0.042]; Uk[0.042]; Uk[0.042]; set[0.041]; set[0.041]; Friday[0.041]; ended[0.041]; market[0.041]; concluding[0.041]; transatlantic[0.040]; flight[0.040]; progress[0.039]; said[0.039]; agreement[0.039]; open[0.039]; open[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.088:0.088[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nservice[0.045]; told[0.045]; London[0.045]; market[0.044]; latest[0.043]; deal[0.043]; air[0.043]; said[0.042]; date[0.042]; date[0.042]; date[0.042]; spokesman[0.041]; Friday[0.041]; number[0.041]; Department[0.041]; produce[0.040]; useful[0.040]; issues[0.040]; progress[0.040]; transatlantic[0.040]; agreement[0.040]; ended[0.040]; Transport[0.040]; flight[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.077:0.077[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nflight[0.047]; service[0.045]; date[0.043]; date[0.043]; date[0.043]; transatlantic[0.043]; number[0.043]; Friday[0.043]; latest[0.043]; skies[0.042]; skies[0.042]; deal[0.042]; ended[0.041]; set[0.041]; set[0.041]; air[0.041]; round[0.041]; Transport[0.041]; agreement[0.040]; said[0.040]; Department[0.040]; progress[0.039]; spokesman[0.039]; told[0.039]; \t\n\r [=========================================>....] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s682ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4157/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1269testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.062:0.062[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.047]; Asia[0.046]; Indian[0.042]; Indian[0.042]; Indian[0.042]; tonnes[0.042]; tonnes[0.042]; tonnes[0.042]; tonnes[0.042]; tonnes[0.042]; demand[0.042]; demand[0.042]; demand[0.042]; season[0.042]; season[0.042]; projected[0.041]; reached[0.040]; March[0.040]; April[0.040]; gap[0.040]; local[0.040]; added[0.040]; Friday[0.040]; century[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indian ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.618:-0.618[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.061]; Indian[0.056]; Indian[0.056]; Indian[0.056]; Asia[0.051]; Singapore[0.050]; hit[0.049]; expected[0.049]; expected[0.049]; operated[0.048]; century[0.048]; production[0.048]; production[0.048]; production[0.048]; production[0.048]; production[0.048]; rise[0.047]; rise[0.047]; April[0.047]; local[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = India ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.055:0.055[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nindustry[0.046]; Singapore[0.045]; domestic[0.045]; million[0.043]; reach[0.043]; reach[0.043]; time[0.043]; land[0.042]; demand[0.042]; possibly[0.042]; widening[0.042]; official[0.041]; increase[0.041]; development[0.041]; gap[0.041]; Board[0.041]; said[0.040]; said[0.040]; natural[0.040]; possible[0.040]; estimates[0.040]; imports[0.040]; view[0.040]; extent[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indian ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.114:0.114[0]; log p(e|m)= -0.618:-0.618[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.053]; Indian[0.048]; Indian[0.048]; Asia[0.044]; Singapore[0.043]; season[0.043]; season[0.043]; local[0.042]; century[0.041]; April[0.041]; March[0.041]; Friday[0.040]; said[0.040]; said[0.040]; hit[0.040]; added[0.040]; natural[0.039]; production[0.039]; production[0.039]; production[0.039]; production[0.039]; seen[0.039]; seen[0.039]; persist[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rubber Board\u001b[39m ==> ENTITY: \u001b[32mRubber Board\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncultivation[0.052]; plantation[0.050]; rubber[0.050]; rubber[0.050]; rubber[0.050]; rubber[0.050]; development[0.049]; domestic[0.049]; production[0.047]; production[0.047]; production[0.047]; India[0.047]; industry[0.046]; consumption[0.046]; consumption[0.046]; Newsroom[0.046]; Singapore[0.046]; season[0.045]; season[0.045]; demand[0.045]; demand[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indian ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.087:0.087[0]; log p(e|m)= -0.618:-0.618[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.062]; Indian[0.057]; Asia[0.052]; hit[0.050]; expected[0.050]; expected[0.050]; widening[0.049]; operated[0.049]; century[0.049]; production[0.049]; production[0.049]; production[0.049]; production[0.049]; production[0.049]; million[0.048]; rise[0.048]; rise[0.048]; rise[0.048]; rise[0.048]; domestic[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indian ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.110:0.110[0]; log p(e|m)= -0.618:-0.618[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndia[0.053]; Indian[0.049]; Indian[0.049]; Asia[0.044]; Singapore[0.044]; season[0.043]; season[0.043]; century[0.041]; April[0.041]; March[0.041]; Friday[0.040]; local[0.040]; said[0.040]; said[0.040]; natural[0.040]; production[0.039]; production[0.039]; production[0.039]; production[0.039]; seen[0.039]; seen[0.039]; persist[0.039]; Board[0.039]; meeting[0.039]; \t\n\r [=========================================>....] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s635ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4164/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1191testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew York\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew York City\u001b[39m <---> \u001b[32mNew York\u001b[39m\t\nSCORES: global= 0.241:0.237[\u001b[31m0.004\u001b[39m]; local()= 0.082:0.068[\u001b[31m0.014\u001b[39m]; log p(e|m)= -1.790:-0.289[\u001b[32m1.501\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.087]; League[0.082]; game[0.082]; Philadelphia[0.081]; Thursday[0.081]; Thursday[0.081]; Football[0.075]; Football[0.075]; American[0.074]; National[0.074]; home[0.072]; Indianapolis[0.070]; Caps[0.068]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIndianapolis\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIndianapolis\u001b[39m <---> \u001b[32mIndianapolis Colts\u001b[39m\t\nSCORES: global= 0.242:0.241[\u001b[31m0.001\u001b[39m]; local()= 0.083:0.141[\u001b[32m0.057\u001b[39m]; log p(e|m)= -0.302:-3.411[\u001b[31m3.109\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.074]; game[0.074]; League[0.072]; Football[0.070]; Football[0.070]; Philadelphia[0.068]; Thursday[0.066]; Thursday[0.066]; American[0.066]; home[0.064]; National[0.064]; York[0.064]; Caps[0.061]; Result[0.060]; Result[0.060]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPhiladelphia\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPhiladelphia\u001b[39m <---> \u001b[32mPhiladelphia Eagles\u001b[39m\t\nSCORES: global= 0.242:0.235[\u001b[31m0.006\u001b[39m]; local()= 0.069:0.155[\u001b[32m0.086\u001b[39m]; log p(e|m)= -0.211:-4.510[\u001b[31m4.299\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.079]; League[0.077]; game[0.071]; Football[0.069]; Football[0.069]; Caps[0.069]; National[0.068]; York[0.067]; Thursday[0.066]; Thursday[0.066]; American[0.066]; Indianapolis[0.064]; home[0.062]; Result[0.053]; Result[0.053]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NFL ==> ENTITY: \u001b[32mNational Football League\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.141:0.141[0]; log p(e|m)= -0.019:-0.019[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.072]; team[0.068]; game[0.067]; Football[0.066]; Football[0.066]; Philadelphia[0.065]; Indianapolis[0.063]; National[0.063]; York[0.062]; Thursday[0.062]; Thursday[0.062]; American[0.061]; home[0.058]; Caps[0.056]; Result[0.054]; Result[0.054]; \t\n\r [=========================================>....] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s610ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4169/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1303testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Netanyahu\u001b[39m ==> ENTITY: \u001b[32mBenjamin Netanyahu\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.053]; Israeli[0.049]; Israeli[0.049]; Netanyahu[0.048]; Netanyahu[0.048]; Palestinian[0.047]; Jerusalem[0.046]; Hebron[0.044]; Prime[0.043]; President[0.042]; President[0.042]; President[0.042]; meeting[0.041]; said[0.041]; said[0.041]; hosted[0.041]; Minister[0.041]; Plo[0.040]; Plo[0.040]; Plo[0.040]; Plo[0.040]; official[0.040]; told[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moslem ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.148:0.148[0]; log p(e|m)= -1.214:-1.214[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPalestine[0.044]; Ahmed[0.043]; Rahman[0.042]; Jerusalem[0.042]; agreed[0.042]; Cairo[0.042]; Cairo[0.042]; Abdel[0.042]; understood[0.042]; Osama[0.042]; successful[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; Israel[0.041]; Israel[0.041]; held[0.041]; Palestinian[0.041]; Palestinian[0.041]; September[0.041]; rule[0.041]; met[0.041]; leaders[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hebron ==> ENTITY: \u001b[32mHebron\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.166:0.166[0]; log p(e|m)= -0.390:-0.390[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPalestine[0.049]; Hebron[0.048]; Israel[0.048]; Israel[0.048]; Jerusalem[0.048]; Israelis[0.046]; Israeli[0.046]; Palestinians[0.045]; Abu[0.045]; Cairo[0.043]; Cairo[0.043]; near[0.042]; Moslem[0.041]; position[0.041]; official[0.041]; Netanyahu[0.041]; Netanyahu[0.041]; Netanyahu[0.041]; Saturday[0.041]; Saturday[0.041]; Nabil[0.041]; delayed[0.041]; September[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jerusalem ==> ENTITY: \u001b[32mJerusalem\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.046]; Israel[0.046]; Palestine[0.045]; Israeli[0.043]; Moslem[0.043]; Hebron[0.043]; Hebron[0.043]; Israelis[0.043]; Cairo[0.042]; Cairo[0.042]; Palestinian[0.042]; Palestinian[0.042]; near[0.042]; Ahmed[0.041]; Palestinians[0.041]; Abdel[0.040]; Abdel[0.040]; Arafat[0.040]; Arafat[0.040]; Arafat[0.040]; Arafat[0.040]; Netanyahu[0.039]; Netanyahu[0.039]; Sunday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cairo ==> ENTITY: \u001b[32mCairo\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.152:-0.152[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEgyptian[0.051]; Nabil[0.051]; Jerusalem[0.050]; Yasser[0.049]; Israel[0.048]; President[0.048]; President[0.048]; President[0.048]; President[0.048]; President[0.048]; Abu[0.048]; said[0.047]; said[0.047]; said[0.047]; work[0.047]; Palestinian[0.047]; negotiations[0.046]; Hosni[0.046]; Benjamin[0.046]; Israeli[0.046]; Israeli[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hosni Mubarak\u001b[39m ==> ENTITY: \u001b[32mHosni Mubarak\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEgyptian[0.047]; Mubarak[0.047]; Cairo[0.046]; Israel[0.045]; Nabil[0.045]; Palestinian[0.044]; Yasser[0.044]; Israeli[0.044]; Israeli[0.044]; Arafat[0.043]; Arafat[0.043]; Arafat[0.043]; Arafat[0.043]; Arafat[0.043]; Arafat[0.043]; President[0.043]; President[0.043]; President[0.043]; President[0.043]; President[0.043]; Abu[0.042]; adviser[0.042]; official[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPalestinians\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPalestinian National Authority\u001b[39m <---> \u001b[32mPalestine\u001b[39m\t\nSCORES: global= 0.248:0.238[\u001b[31m0.010\u001b[39m]; local()= 0.212:0.162[\u001b[31m0.050\u001b[39m]; log p(e|m)= -1.754:-1.640[\u001b[32m0.115\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.047]; Israel[0.047]; Palestine[0.047]; Palestinian[0.046]; Palestinian[0.046]; Israelis[0.045]; Israeli[0.045]; Netanyahu[0.041]; Netanyahu[0.041]; Netanyahu[0.041]; Jerusalem[0.041]; Hebron[0.041]; Hebron[0.041]; Arafat[0.040]; Arafat[0.040]; Arafat[0.040]; Arafat[0.040]; Arafat[0.040]; Cairo[0.040]; agreed[0.039]; President[0.039]; President[0.039]; disputed[0.038]; Secretary[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.167:0.167[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.053]; Israeli[0.048]; Palestinian[0.046]; Jerusalem[0.045]; Hebron[0.045]; Netanyahu[0.043]; Netanyahu[0.043]; Netanyahu[0.043]; Abu[0.042]; Egyptian[0.042]; official[0.040]; Cairo[0.040]; hosted[0.040]; Yasser[0.040]; adviser[0.039]; Arafat[0.039]; Arafat[0.039]; Arafat[0.039]; Arafat[0.039]; Arafat[0.039]; said[0.039]; said[0.039]; said[0.039]; says[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arafat\u001b[39m ==> ENTITY: \u001b[32mYasser Arafat\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.048]; Israeli[0.046]; Israeli[0.046]; Palestinians[0.043]; Arafat[0.042]; Arafat[0.042]; Arafat[0.042]; Arafat[0.042]; Arafat[0.042]; Arafat[0.042]; Hebron[0.041]; Hebron[0.041]; Egyptian[0.041]; Netanyahu[0.041]; Netanyahu[0.041]; Netanyahu[0.041]; Netanyahu[0.041]; Yasser[0.041]; Abu[0.041]; Cairo[0.040]; Cairo[0.040]; Mubarak[0.039]; Mubarak[0.039]; Mubarak[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cairo ==> ENTITY: \u001b[32mCairo\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.171:0.171[0]; log p(e|m)= -0.152:-0.152[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAhmed[0.048]; Abdel[0.047]; Abdel[0.047]; Jerusalem[0.045]; Palestine[0.045]; near[0.044]; Israel[0.044]; el[0.043]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; Osama[0.043]; Mubarak[0.043]; Palestinian[0.043]; Palestinian[0.043]; religious[0.042]; Thursday[0.042]; Thursday[0.042]; Thursday[0.042]; Moslem[0.042]; Israeli[0.042]; sites[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jewish ==> ENTITY: \u001b[32mJews\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.126:0.126[0]; log p(e|m)= -0.805:-0.805[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJewish[0.050]; Israel[0.047]; religious[0.046]; Israeli[0.044]; Palestinian[0.044]; Palestinian[0.044]; Abdel[0.044]; Abdel[0.044]; Hebron[0.043]; Cairo[0.042]; observance[0.041]; Ahmed[0.041]; sabbath[0.041]; sabbath[0.041]; el[0.040]; night[0.040]; said[0.039]; said[0.039]; said[0.039]; said[0.039]; Osama[0.039]; Netanyahu[0.039]; Friday[0.038]; Baz[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arafat\u001b[39m ==> ENTITY: \u001b[32mYasser Arafat\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israel[0.050]; Israeli[0.048]; Israelis[0.045]; Palestine[0.045]; Palestinians[0.044]; Arafat[0.044]; Arafat[0.044]; Arafat[0.044]; Arafat[0.044]; Arafat[0.044]; Jerusalem[0.043]; Hebron[0.043]; Hebron[0.043]; Netanyahu[0.042]; Netanyahu[0.042]; Netanyahu[0.042]; Abu[0.042]; Cairo[0.041]; Mubarak[0.040]; Mubarak[0.040]; Nabil[0.040]; met[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhosted[0.045]; told[0.044]; Israel[0.044]; official[0.043]; deal[0.043]; Israeli[0.042]; Israeli[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; Jerusalem[0.042]; Egyptian[0.042]; date[0.041]; time[0.041]; negotiators[0.041]; work[0.041]; Palestinian[0.041]; adviser[0.041]; Friday[0.040]; Cairo[0.040]; Cairo[0.040]; says[0.040]; possible[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hebron ==> ENTITY: \u001b[32mHebron\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.390:-0.390[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.059]; Jerusalem[0.059]; Palestinian[0.057]; Israeli[0.057]; Israeli[0.057]; Abu[0.056]; Cairo[0.053]; Egyptian[0.052]; position[0.050]; official[0.050]; Yasser[0.050]; Netanyahu[0.050]; Netanyahu[0.050]; Netanyahu[0.050]; Netanyahu[0.050]; Saturday[0.050]; Saturday[0.050]; Saturday[0.050]; Nabil[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.204:0.204[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.053]; Israeli[0.048]; Palestine[0.047]; Palestinian[0.046]; Palestinian[0.046]; Jewish[0.045]; Jewish[0.045]; Israelis[0.045]; Palestinians[0.044]; Netanyahu[0.044]; Netanyahu[0.044]; Jerusalem[0.044]; Washington[0.042]; negotiations[0.041]; talks[0.041]; talks[0.041]; Secretary[0.041]; agreed[0.041]; Hebron[0.041]; Hebron[0.041]; sabbath[0.040]; sabbath[0.040]; September[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Benjamin Netanyahu\u001b[39m ==> ENTITY: \u001b[32mBenjamin Netanyahu\u001b[39m\t\nSCORES: global= 0.311:0.311[0]; local()= 0.223:0.223[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israeli[0.046]; Israeli[0.046]; Netanyahu[0.045]; Netanyahu[0.045]; Netanyahu[0.045]; Palestinian[0.044]; Jerusalem[0.044]; Hebron[0.042]; Prime[0.041]; Prime[0.041]; President[0.040]; President[0.040]; President[0.040]; President[0.040]; meeting[0.039]; meeting[0.039]; adviser[0.039]; said[0.039]; said[0.039]; said[0.039]; hosted[0.039]; Minister[0.038]; Minister[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arafat\u001b[39m ==> ENTITY: \u001b[32mYasser Arafat\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.048]; Israel[0.048]; Israeli[0.046]; Israelis[0.043]; Palestine[0.043]; Palestinians[0.043]; Arafat[0.042]; Arafat[0.042]; Arafat[0.042]; Arafat[0.042]; Arafat[0.042]; Jerusalem[0.042]; Hebron[0.041]; Hebron[0.041]; Netanyahu[0.040]; Netanyahu[0.040]; Netanyahu[0.040]; Abu[0.040]; Cairo[0.040]; Mubarak[0.039]; Mubarak[0.039]; Mubarak[0.039]; Nabil[0.039]; met[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hebron ==> ENTITY: \u001b[32mHebron\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.187:0.187[0]; log p(e|m)= -0.390:-0.390[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPalestine[0.046]; Hebron[0.045]; Israel[0.045]; Israel[0.045]; Jerusalem[0.045]; Israelis[0.044]; Jewish[0.044]; Jewish[0.044]; Palestinian[0.043]; Palestinian[0.043]; Israeli[0.043]; Palestinians[0.043]; sabbath[0.041]; sabbath[0.041]; Cairo[0.040]; Cairo[0.040]; held[0.039]; Moslem[0.039]; el[0.039]; Netanyahu[0.038]; Netanyahu[0.038]; near[0.038]; Abdel[0.038]; Abdel[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jerusalem ==> ENTITY: \u001b[32mJerusalem\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israeli[0.047]; Israeli[0.047]; Hebron[0.046]; Cairo[0.046]; Palestinian[0.046]; Egyptian[0.044]; Arafat[0.044]; Arafat[0.044]; Arafat[0.044]; Arafat[0.044]; set[0.043]; Netanyahu[0.042]; Netanyahu[0.042]; Netanyahu[0.042]; Yasser[0.042]; work[0.042]; requested[0.042]; officials[0.041]; Friday[0.041]; Mubarak[0.041]; Mubarak[0.041]; Hosni[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israelis ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.194:0.194[0]; log p(e|m)= -0.546:-0.546[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.054]; Israel[0.054]; Israeli[0.049]; Palestine[0.048]; Palestinian[0.046]; Palestinian[0.046]; Jerusalem[0.045]; Palestinians[0.044]; Netanyahu[0.044]; Netanyahu[0.044]; Netanyahu[0.044]; Hebron[0.042]; Hebron[0.042]; September[0.040]; Prime[0.040]; Cairo[0.040]; President[0.040]; Secretary[0.040]; Washington[0.040]; position[0.040]; delayed[0.040]; adviser[0.040]; agreed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Netanyahu\u001b[39m ==> ENTITY: \u001b[32mBenjamin Netanyahu\u001b[39m\t\nSCORES: global= 0.312:0.312[0]; local()= 0.231:0.231[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israel[0.050]; Israeli[0.046]; Netanyahu[0.045]; Netanyahu[0.045]; Netanyahu[0.045]; Israelis[0.044]; Hebron[0.041]; Hebron[0.041]; Palestinians[0.041]; Prime[0.041]; President[0.040]; President[0.040]; President[0.040]; President[0.040]; meeting[0.039]; meeting[0.039]; meeting[0.039]; meeting[0.039]; meeting[0.039]; adviser[0.039]; said[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Netanyahu\u001b[39m ==> ENTITY: \u001b[32mBenjamin Netanyahu\u001b[39m\t\nSCORES: global= 0.312:0.312[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.052]; Israeli[0.048]; Israeli[0.048]; Netanyahu[0.047]; Netanyahu[0.047]; Netanyahu[0.047]; Palestinian[0.046]; Jerusalem[0.045]; Hebron[0.043]; Hebron[0.043]; Prime[0.042]; Prime[0.042]; President[0.041]; President[0.041]; President[0.041]; President[0.041]; President[0.041]; meeting[0.041]; meeting[0.041]; meeting[0.041]; meeting[0.041]; meeting[0.041]; adviser[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jewish ==> ENTITY: \u001b[32mJews\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.805:-0.805[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJewish[0.050]; Israel[0.047]; religious[0.046]; Israeli[0.045]; Palestinian[0.044]; Palestinian[0.044]; Abdel[0.044]; Abdel[0.044]; Hebron[0.043]; Cairo[0.042]; observance[0.041]; Ahmed[0.041]; sabbath[0.041]; sabbath[0.041]; el[0.040]; night[0.040]; said[0.039]; said[0.039]; said[0.039]; Osama[0.039]; Netanyahu[0.039]; Friday[0.038]; Baz[0.038]; Arafat[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Egyptian ==> ENTITY: \u001b[32mEgypt\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.133:0.133[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNabil[0.048]; Cairo[0.047]; Israel[0.046]; Yasser[0.046]; Israeli[0.045]; Israeli[0.045]; Jerusalem[0.042]; Palestinian[0.042]; Abu[0.042]; Hebron[0.041]; Mubarak[0.040]; Mubarak[0.040]; officials[0.040]; meeting[0.040]; meeting[0.040]; Benjamin[0.040]; said[0.040]; said[0.040]; said[0.040]; Hosni[0.039]; official[0.039]; adviser[0.039]; requested[0.039]; says[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Netanyahu\u001b[39m ==> ENTITY: \u001b[32mBenjamin Netanyahu\u001b[39m\t\nSCORES: global= 0.312:0.312[0]; local()= 0.226:0.226[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.057]; Israel[0.057]; Israeli[0.052]; Netanyahu[0.051]; Netanyahu[0.051]; Palestine[0.050]; Israelis[0.049]; Jerusalem[0.049]; Hebron[0.047]; Hebron[0.047]; Palestinians[0.046]; Prime[0.046]; President[0.045]; President[0.045]; meeting[0.044]; meeting[0.044]; meeting[0.044]; meeting[0.044]; meeting[0.044]; adviser[0.044]; adviser[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.065]; Israeli[0.059]; Palestinian[0.056]; Jerusalem[0.055]; Hebron[0.055]; Hebron[0.055]; Netanyahu[0.052]; Netanyahu[0.052]; Netanyahu[0.052]; Netanyahu[0.052]; Netanyahu[0.052]; Abu[0.051]; Egyptian[0.051]; official[0.049]; Cairo[0.048]; Cairo[0.048]; hosted[0.048]; Yasser[0.048]; adviser[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mubarak\u001b[39m ==> ENTITY: \u001b[32mHosni Mubarak\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMubarak[0.045]; Cairo[0.044]; Cairo[0.044]; Israel[0.043]; Israel[0.043]; Ahmed[0.043]; Abdel[0.042]; Abdel[0.042]; Palestinian[0.042]; Palestinian[0.042]; Israeli[0.042]; Palestine[0.042]; Palestinians[0.041]; Arafat[0.041]; Arafat[0.041]; Arafat[0.041]; Israelis[0.040]; Osama[0.040]; adviser[0.040]; violence[0.040]; Plo[0.040]; Plo[0.040]; Netanyahu[0.039]; Netanyahu[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.213:0.213[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.053]; Israel[0.053]; Palestine[0.048]; Jewish[0.047]; Jewish[0.047]; Israelis[0.047]; religious[0.046]; Palestinian[0.045]; Palestinian[0.045]; Jerusalem[0.044]; Hebron[0.044]; Palestinians[0.044]; Netanyahu[0.042]; Moslem[0.041]; Washington[0.041]; observance[0.040]; near[0.040]; Cairo[0.039]; rule[0.039]; Thursday[0.039]; Thursday[0.039]; Thursday[0.039]; el[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cairo ==> ENTITY: \u001b[32mCairo\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.152:-0.152[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJerusalem[0.050]; Washington[0.049]; Palestine[0.049]; Nabil[0.049]; near[0.048]; Israel[0.048]; Israel[0.048]; President[0.048]; President[0.048]; President[0.048]; said[0.048]; said[0.048]; said[0.048]; said[0.048]; delayed[0.047]; Abu[0.047]; long[0.046]; negotiations[0.046]; September[0.046]; Israeli[0.046]; disputed[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mubarak\u001b[39m ==> ENTITY: \u001b[32mHosni Mubarak\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEgyptian[0.045]; Mubarak[0.044]; Hosni[0.044]; Cairo[0.044]; Cairo[0.044]; Israel[0.043]; Nabil[0.043]; Palestinian[0.041]; Yasser[0.041]; Israeli[0.041]; Israeli[0.041]; Arafat[0.041]; Arafat[0.041]; Arafat[0.041]; Arafat[0.041]; Arafat[0.041]; Arafat[0.041]; Arafat[0.041]; President[0.041]; President[0.041]; President[0.041]; President[0.041]; President[0.041]; Abu[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.142:0.142[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsraeli[0.050]; Israeli[0.050]; Palestinian[0.047]; Netanyahu[0.045]; Netanyahu[0.045]; Netanyahu[0.045]; Netanyahu[0.045]; Jerusalem[0.045]; Egyptian[0.044]; President[0.043]; President[0.043]; President[0.043]; President[0.043]; President[0.043]; negotiations[0.042]; official[0.042]; Hebron[0.042]; requested[0.041]; adviser[0.041]; Cairo[0.041]; officials[0.040]; Prime[0.040]; Prime[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.189:0.189[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.052]; Israeli[0.047]; Palestine[0.046]; Palestinian[0.044]; Palestinian[0.044]; Israelis[0.044]; Palestinians[0.043]; Netanyahu[0.042]; Netanyahu[0.042]; Netanyahu[0.042]; Jerusalem[0.042]; Washington[0.040]; negotiations[0.040]; talks[0.040]; talks[0.040]; Secretary[0.040]; agreed[0.039]; Hebron[0.039]; Hebron[0.039]; September[0.039]; Cairo[0.039]; disputed[0.038]; Moslem[0.038]; President[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arafat\u001b[39m ==> ENTITY: \u001b[32mYasser Arafat\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.055]; Israeli[0.053]; Israeli[0.053]; Palestinian[0.052]; Arafat[0.049]; Arafat[0.049]; Jerusalem[0.048]; Hebron[0.048]; Egyptian[0.047]; Netanyahu[0.047]; Netanyahu[0.047]; Netanyahu[0.047]; Yasser[0.047]; Cairo[0.046]; Mubarak[0.045]; Mubarak[0.045]; Hosni[0.045]; President[0.044]; President[0.044]; President[0.044]; meeting[0.043]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPalestinian\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPalestinian people\u001b[39m <---> \u001b[32mPalestine\u001b[39m\t\nSCORES: global= 0.245:0.236[\u001b[31m0.008\u001b[39m]; local()= 0.181:0.144[\u001b[31m0.037\u001b[39m]; log p(e|m)= -0.389:-1.911[\u001b[31m1.522\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.049]; Israeli[0.046]; Israeli[0.046]; Jerusalem[0.044]; Hebron[0.044]; Egyptian[0.043]; Abu[0.043]; Cairo[0.043]; Netanyahu[0.042]; Netanyahu[0.042]; Netanyahu[0.042]; Nabil[0.042]; Yasser[0.041]; Arafat[0.041]; Arafat[0.041]; Arafat[0.041]; Arafat[0.041]; official[0.040]; deal[0.039]; requested[0.039]; officials[0.039]; said[0.038]; said[0.038]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arafat\u001b[39m ==> ENTITY: \u001b[32mYasser Arafat\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.048]; Israel[0.048]; Israeli[0.046]; Israelis[0.043]; Palestine[0.043]; Palestinians[0.042]; Arafat[0.042]; Arafat[0.042]; Arafat[0.042]; Arafat[0.042]; Arafat[0.042]; Jerusalem[0.042]; Hebron[0.041]; Hebron[0.041]; Netanyahu[0.040]; Netanyahu[0.040]; Netanyahu[0.040]; Abu[0.040]; Cairo[0.040]; Cairo[0.040]; Mubarak[0.039]; Mubarak[0.039]; Mubarak[0.039]; Nabil[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.092:0.092[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.044]; President[0.044]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; agreed[0.042]; talks[0.042]; talks[0.042]; leaders[0.042]; Thursday[0.042]; meeting[0.041]; meeting[0.041]; meeting[0.041]; meeting[0.041]; meeting[0.041]; negotiations[0.041]; September[0.041]; negotiators[0.041]; meet[0.041]; meet[0.041]; crucial[0.040]; successful[0.040]; long[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arafat\u001b[39m ==> ENTITY: \u001b[32mYasser Arafat\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.224:0.224[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.052]; Israeli[0.050]; Israeli[0.050]; Palestinian[0.048]; Arafat[0.046]; Arafat[0.046]; Arafat[0.046]; Arafat[0.046]; Arafat[0.046]; Arafat[0.046]; Jerusalem[0.045]; Hebron[0.045]; Hebron[0.045]; Egyptian[0.044]; Netanyahu[0.044]; Netanyahu[0.044]; Netanyahu[0.044]; Netanyahu[0.044]; Yasser[0.044]; Abu[0.044]; Cairo[0.043]; Cairo[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arafat\u001b[39m ==> ENTITY: \u001b[32mYasser Arafat\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.218:0.218[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.048]; Israeli[0.046]; Israeli[0.046]; Palestinian[0.045]; Arafat[0.042]; Arafat[0.042]; Arafat[0.042]; Arafat[0.042]; Arafat[0.042]; Jerusalem[0.042]; Hebron[0.041]; Hebron[0.041]; Egyptian[0.041]; Netanyahu[0.041]; Netanyahu[0.041]; Netanyahu[0.041]; Netanyahu[0.041]; Yasser[0.041]; Abu[0.041]; Cairo[0.040]; Mubarak[0.039]; Mubarak[0.039]; Nabil[0.039]; Hosni[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPalestinian\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPalestinian people\u001b[39m <---> \u001b[32mPalestine\u001b[39m\t\nSCORES: global= 0.245:0.238[\u001b[31m0.007\u001b[39m]; local()= 0.201:0.173[\u001b[31m0.029\u001b[39m]; log p(e|m)= -0.389:-1.911[\u001b[31m1.522\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPalestine[0.049]; Israel[0.048]; Israel[0.048]; Palestinian[0.046]; Israeli[0.044]; Israelis[0.044]; Palestinians[0.044]; Jerusalem[0.043]; Hebron[0.043]; Jewish[0.042]; Jewish[0.042]; Cairo[0.041]; Netanyahu[0.040]; Abdel[0.040]; Abdel[0.040]; el[0.040]; Arafat[0.039]; Ahmed[0.039]; Liberation[0.039]; Authority[0.038]; Moslem[0.038]; sites[0.038]; Organisation[0.038]; Osama[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Palestinian Authority\u001b[39m ==> ENTITY: \u001b[32mPalestinian National Authority\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPalestine[0.049]; Israel[0.048]; Israel[0.048]; Palestinian[0.046]; Israeli[0.046]; Cairo[0.044]; Hebron[0.044]; Palestinians[0.044]; Israelis[0.044]; Jerusalem[0.043]; Abdel[0.040]; Abdel[0.040]; Netanyahu[0.040]; Jewish[0.040]; Jewish[0.040]; Arafat[0.040]; Mubarak[0.039]; el[0.039]; Ahmed[0.038]; Osama[0.038]; Organisation[0.038]; Secretary[0.038]; Plo[0.038]; Plo[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mubarak\u001b[39m ==> ENTITY: \u001b[32mHosni Mubarak\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMubarak[0.047]; Cairo[0.046]; Israel[0.046]; Israel[0.046]; Nabil[0.045]; Israeli[0.044]; Palestine[0.044]; Palestinians[0.043]; Arafat[0.043]; Arafat[0.043]; Arafat[0.043]; Arafat[0.043]; Arafat[0.043]; Arafat[0.043]; President[0.043]; President[0.043]; Abu[0.043]; Israelis[0.042]; Osama[0.042]; adviser[0.042]; adviser[0.042]; violence[0.042]; official[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yasser Arafat\u001b[39m ==> ENTITY: \u001b[32mYasser Arafat\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.215:0.215[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.048]; Israeli[0.047]; Israeli[0.047]; Palestinian[0.046]; Arafat[0.043]; Arafat[0.043]; Arafat[0.043]; Arafat[0.043]; Jerusalem[0.042]; Hebron[0.042]; Egyptian[0.041]; Netanyahu[0.041]; Netanyahu[0.041]; Netanyahu[0.041]; Abu[0.041]; Cairo[0.041]; Mubarak[0.040]; Mubarak[0.040]; Nabil[0.039]; Hosni[0.039]; President[0.038]; President[0.038]; President[0.038]; President[0.038]; \t\n\r [==========================================>...] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s382ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4211/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1338testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tokyo ==> ENTITY: \u001b[32mTokyo\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= -0.017:-0.017[0]; log p(e|m)= -0.093:-0.093[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nyen[0.047]; yen[0.047]; yen[0.047]; yen[0.047]; yen[0.047]; March[0.046]; Sales[0.044]; parent[0.043]; billions[0.043]; trader[0.041]; forecast[0.040]; Current[0.040]; specified[0.040]; Latest[0.040]; unless[0.040]; Parent[0.040]; Net[0.039]; Note[0.039]; div[0.039]; Actual[0.038]; Forecast[0.038]; Ago[0.038]; interiors[0.037]; Ord[0.037]; \t\n\r [==========================================>...] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s376ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4212/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1173testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tony Underwood\u001b[39m ==> ENTITY: \u001b[32mTony Underwood\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.045]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Scotland[0.042]; Scotland[0.042]; Leicester[0.042]; Ireland[0.042]; Twickenham[0.040]; Neil[0.040]; Barbarians[0.040]; Barbarians[0.040]; played[0.040]; Australia[0.040]; Jones[0.039]; Jones[0.039]; Scott[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = All Black\u001b[39m ==> ENTITY: \u001b[32mNew Zealand national rugby union team\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.043]; Ireland[0.043]; internationals[0.043]; Barbarians[0.042]; Barbarians[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; England[0.042]; England[0.042]; England[0.042]; England[0.042]; England[0.042]; Twickenham[0.041]; Zealand[0.040]; Zealand[0.040]; played[0.040]; Scotland[0.040]; Scotland[0.040]; game[0.040]; Cup[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Joe Roff\u001b[39m ==> ENTITY: \u001b[32mJoe Roff\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.045]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Ireland[0.043]; Barbarians[0.042]; Campese[0.041]; Scotland[0.041]; Leicester[0.040]; Rob[0.040]; Rob[0.040]; Finegan[0.040]; Neil[0.040]; Andrew[0.039]; Andrew[0.039]; Zealand[0.039]; Zealand[0.039]; Michael[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = David Campese\u001b[39m ==> ENTITY: \u001b[32mDavid Campese\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWallabies[0.046]; Wallabies[0.046]; Campese[0.046]; Australia[0.045]; Australian[0.044]; winger[0.043]; Wales[0.043]; Andrew[0.042]; Daniel[0.042]; Daniel[0.042]; captain[0.041]; Barbarians[0.041]; Barbarians[0.041]; Rob[0.041]; Queenslander[0.041]; Finegan[0.041]; match[0.040]; Owen[0.040]; Twickenham[0.040]; kicked[0.040]; Herbert[0.039]; ahead[0.039]; massive[0.039]; test[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEngland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEngland national rugby union team\u001b[39m <---> \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.245:0.238[\u001b[31m0.007\u001b[39m]; local()= 0.210:0.172[\u001b[31m0.038\u001b[39m]; log p(e|m)= -3.772:-0.190[\u001b[32m3.582\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.046]; England[0.046]; England[0.046]; England[0.046]; Wales[0.044]; Wales[0.044]; Wales[0.044]; Wales[0.044]; Wales[0.044]; Wales[0.044]; Australia[0.044]; Australia[0.044]; League[0.043]; internationals[0.043]; Ireland[0.043]; Ireland[0.043]; Scotland[0.042]; Scotland[0.042]; Barbarians[0.042]; Barbarians[0.042]; Owen[0.042]; Leicester[0.041]; Dale[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rob Andrew\u001b[39m ==> ENTITY: \u001b[32mRob Andrew\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.045]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; captain[0.043]; Leicester[0.042]; Ireland[0.041]; Ireland[0.041]; Neil[0.041]; game[0.041]; Wales[0.041]; Wales[0.041]; Wales[0.041]; Wales[0.041]; Wales[0.041]; Wales[0.041]; internationals[0.041]; scored[0.040]; Scotland[0.040]; Scotland[0.040]; recalled[0.040]; Australia[0.040]; selectorial[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barbarians ==> ENTITY: \u001b[32mBarbarian F.C.\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.165:0.165[0]; log p(e|m)= -0.171:-0.171[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.048]; Wales[0.045]; Australia[0.044]; Wallabies[0.044]; Twickenham[0.043]; fixture[0.043]; tour[0.042]; tour[0.042]; tour[0.042]; Little[0.042]; Little[0.042]; Little[0.042]; weekend[0.042]; Saturday[0.041]; Campese[0.041]; Campese[0.041]; place[0.040]; Union[0.040]; Finegan[0.040]; European[0.040]; style[0.039]; David[0.039]; row[0.039]; London[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rob Andrew\u001b[39m ==> ENTITY: \u001b[32mRob Andrew\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.046]; captain[0.045]; played[0.043]; match[0.043]; Ireland[0.042]; game[0.042]; Wales[0.042]; kicked[0.042]; internationals[0.042]; scored[0.041]; Scotland[0.041]; Australia[0.041]; Australia[0.041]; selectorial[0.041]; Barbarians[0.041]; Barbarians[0.041]; Cup[0.041]; final[0.041]; final[0.041]; final[0.041]; Twickenham[0.041]; won[0.041]; winger[0.041]; Andrew[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWales\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mWales national rugby union team\u001b[39m <---> \u001b[32mWales\u001b[39m\t\nSCORES: global= 0.254:0.245[\u001b[31m0.008\u001b[39m]; local()= 0.197:0.183[\u001b[31m0.014\u001b[39m]; log p(e|m)= -2.718:-0.373[\u001b[32m2.346\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWales[0.048]; Wales[0.048]; Wales[0.048]; Wales[0.048]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; Ireland[0.041]; Scotland[0.041]; Owen[0.040]; Barbarians[0.039]; Australia[0.039]; Jones[0.039]; Jones[0.039]; Pontypridd[0.039]; Gavin[0.039]; Leicester[0.038]; Howley[0.038]; Payne[0.038]; David[0.038]; David[0.038]; David[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Darren Garforth\u001b[39m ==> ENTITY: \u001b[32mDarren Garforth\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeicester[0.046]; captain[0.046]; Barbarians[0.045]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; Quinnell[0.044]; Quinnell[0.044]; Nigel[0.044]; Gavin[0.044]; forwards[0.043]; Pontypridd[0.043]; Neil[0.043]; Matthew[0.042]; Scotland[0.042]; Ian[0.042]; Ian[0.042]; Allan[0.042]; Howley[0.042]; David[0.042]; David[0.042]; David[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tim Horan\u001b[39m ==> ENTITY: \u001b[32mTim Horan\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.049]; Howley[0.047]; Australia[0.047]; Campese[0.046]; Gavin[0.046]; Quinnell[0.046]; Quinnell[0.046]; Burke[0.046]; Finegan[0.046]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; Daniel[0.045]; Michael[0.045]; Neil[0.045]; Owen[0.045]; Pat[0.044]; Brial[0.044]; Tony[0.044]; Craig[0.044]; Payne[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scotland ==> ENTITY: \u001b[32mScotland national rugby union team\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.222:0.222[0]; log p(e|m)= -3.381:-3.381[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.047]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; Ireland[0.044]; Wales[0.043]; Wales[0.043]; Wales[0.043]; internationals[0.042]; played[0.041]; Australia[0.041]; Australia[0.041]; Barbarians[0.041]; Barbarians[0.041]; Barbarians[0.041]; Twickenham[0.040]; League[0.040]; game[0.040]; selectorial[0.039]; Italy[0.039]; Cup[0.039]; final[0.039]; final[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wales ==> ENTITY: \u001b[32mWales national rugby union team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.161:0.161[0]; log p(e|m)= -2.718:-2.718[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.051]; fixture[0.047]; Owen[0.046]; Twickenham[0.046]; Barbarians[0.046]; match[0.045]; Australia[0.045]; final[0.044]; David[0.044]; Wallabies[0.044]; Wallabies[0.044]; London[0.043]; row[0.042]; European[0.042]; ahead[0.041]; Little[0.041]; Little[0.041]; Little[0.041]; weekend[0.041]; winger[0.041]; Daniel[0.041]; Daniel[0.041]; place[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ian Jones\u001b[39m ==> ENTITY: \u001b[32mIan Jones (rugby union)\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.203:0.203[0]; log p(e|m)= -0.002:-0.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nQuinnell[0.045]; Quinnell[0.045]; played[0.044]; Wales[0.044]; Wales[0.044]; Wales[0.044]; Wales[0.044]; Wales[0.044]; Wales[0.044]; Allan[0.044]; Tony[0.044]; Tony[0.044]; Neil[0.043]; Craig[0.043]; Barbarians[0.043]; Barbarians[0.043]; Cup[0.043]; Scott[0.043]; internationals[0.043]; forwards[0.042]; Pontypridd[0.042]; Ian[0.042]; Jones[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Norm Hewitt\u001b[39m ==> ENTITY: \u001b[32mNorm Hewitt\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Zealand[0.042]; Zealand[0.042]; England[0.042]; England[0.042]; England[0.042]; England[0.042]; England[0.042]; appearance[0.042]; Australia[0.042]; Australia[0.042]; played[0.041]; Ireland[0.041]; Ireland[0.041]; Neil[0.041]; game[0.041]; Hewitt[0.040]; Allan[0.040]; country[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pat Howard\u001b[39m ==> ENTITY: \u001b[32mPat Howard\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.045]; Ireland[0.044]; England[0.044]; England[0.044]; England[0.044]; Howley[0.042]; Finegan[0.042]; Leicester[0.042]; Campese[0.042]; Neil[0.042]; Tony[0.041]; Wales[0.041]; Wales[0.041]; Wales[0.041]; Wales[0.041]; captain[0.041]; Scotland[0.041]; Gavin[0.041]; David[0.041]; David[0.041]; David[0.041]; Allan[0.040]; Joe[0.040]; Ian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Owen Finegan\u001b[39m ==> ENTITY: \u001b[32mOwen Finegan\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWales[0.044]; Wales[0.044]; Wales[0.044]; Howley[0.044]; Ireland[0.043]; Campese[0.043]; Australia[0.043]; captain[0.043]; Horan[0.043]; Leicester[0.042]; England[0.042]; England[0.042]; England[0.042]; Quinnell[0.041]; Quinnell[0.041]; Giffin[0.041]; Pontypridd[0.040]; Burke[0.040]; Gavin[0.040]; Zealand[0.040]; Zealand[0.040]; Garforth[0.040]; Norm[0.039]; Brial[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEngland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEngland national rugby union team\u001b[39m <---> \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.245:0.238[\u001b[31m0.007\u001b[39m]; local()= 0.210:0.173[\u001b[31m0.037\u001b[39m]; log p(e|m)= -3.772:-0.190[\u001b[32m3.582\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.048]; England[0.048]; England[0.048]; England[0.048]; Wales[0.046]; Wales[0.046]; Wales[0.046]; Wales[0.046]; Wales[0.046]; Wales[0.046]; Australia[0.046]; League[0.045]; internationals[0.045]; Ireland[0.045]; Ireland[0.045]; Scotland[0.044]; Scotland[0.044]; Barbarians[0.044]; Barbarians[0.044]; Owen[0.044]; Leicester[0.043]; Dale[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rob Howley\u001b[39m ==> ENTITY: \u001b[32mRob Howley\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; England[0.047]; England[0.047]; England[0.047]; England[0.047]; Ireland[0.045]; Ireland[0.045]; Wales[0.044]; Wales[0.044]; Wales[0.044]; Wales[0.044]; Wales[0.044]; Wales[0.044]; Scotland[0.044]; Scotland[0.044]; Barbarians[0.042]; Barbarians[0.042]; internationals[0.042]; Leicester[0.041]; game[0.041]; Neil[0.040]; Australia[0.040]; selectorial[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national rugby union team\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.209:0.209[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.052]; England[0.048]; internationals[0.048]; Wallabies[0.048]; Wales[0.047]; match[0.047]; Ireland[0.046]; Barbarians[0.046]; Barbarians[0.046]; country[0.046]; League[0.045]; Australian[0.045]; Twickenham[0.044]; Campese[0.044]; Campese[0.044]; selectorial[0.044]; Cup[0.044]; captain[0.043]; Scotland[0.043]; played[0.043]; game[0.043]; Italy[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leicester ==> ENTITY: \u001b[32mLeicester Tigers\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.155:0.155[0]; log p(e|m)= -3.194:-3.194[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.044]; England[0.044]; England[0.044]; England[0.044]; Ireland[0.042]; Barbarians[0.042]; Quinnell[0.041]; Quinnell[0.041]; Howley[0.041]; David[0.041]; David[0.041]; David[0.041]; Wales[0.041]; Wales[0.041]; Wales[0.041]; Wales[0.041]; Wales[0.041]; Scott[0.041]; forwards[0.041]; Darren[0.041]; Neil[0.041]; Scotland[0.041]; captain[0.040]; Michael[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Norm Hewitt\u001b[39m ==> ENTITY: \u001b[32mNorm Hewitt\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Zealand[0.043]; Zealand[0.043]; England[0.042]; England[0.042]; England[0.042]; England[0.042]; Australia[0.042]; Ireland[0.042]; Gavin[0.041]; Neil[0.041]; Hewitt[0.041]; Allan[0.041]; David[0.040]; David[0.040]; David[0.040]; Michael[0.040]; Barbarians[0.040]; Norm[0.040]; Matthew[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mQueenslander\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mQueensland cricket team\u001b[39m <---> \u001b[32mQueensland\u001b[39m\t\nSCORES: global= 0.260:0.257[\u001b[31m0.003\u001b[39m]; local()= 0.117:0.089[\u001b[31m0.028\u001b[39m]; log p(e|m)= -1.382:-1.124[\u001b[32m0.258\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.046]; style[0.045]; Campese[0.044]; Campese[0.044]; match[0.044]; Rugby[0.044]; Wallabies[0.043]; Wallabies[0.043]; Wales[0.042]; Australian[0.042]; house[0.042]; Finegan[0.041]; large[0.041]; final[0.041]; successive[0.041]; record[0.041]; Barbarians[0.041]; took[0.040]; injury[0.040]; fixture[0.040]; miss[0.039]; row[0.039]; winger[0.039]; knee[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ireland ==> ENTITY: \u001b[32mIreland national rugby union team\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.220:0.220[0]; log p(e|m)= -3.381:-3.381[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.044]; England[0.044]; England[0.044]; England[0.044]; League[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; internationals[0.042]; Scotland[0.042]; Scotland[0.042]; Australia[0.042]; Australia[0.042]; Barbarians[0.041]; Barbarians[0.041]; Barbarians[0.041]; Twickenham[0.041]; country[0.041]; played[0.041]; Cup[0.040]; Howley[0.040]; Italy[0.040]; game[0.039]; selectorial[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barbarians ==> ENTITY: \u001b[32mBarbarian F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.223:0.223[0]; log p(e|m)= -0.171:-0.171[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; Ireland[0.043]; Ireland[0.043]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Barbarians[0.042]; Scotland[0.041]; Scotland[0.041]; Australia[0.041]; Australia[0.041]; internationals[0.040]; Twickenham[0.040]; League[0.040]; Neil[0.039]; played[0.039]; tour[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tim Stimpson\u001b[39m ==> ENTITY: \u001b[32mTim Stimpson\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; England[0.047]; England[0.047]; England[0.047]; England[0.047]; Leicester[0.045]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; scored[0.043]; Ireland[0.042]; Ireland[0.042]; final[0.042]; game[0.042]; Andrew[0.042]; Andrew[0.042]; Ian[0.042]; Ian[0.042]; Scotland[0.041]; Scotland[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dan Crowley\u001b[39m ==> ENTITY: \u001b[32mDan Crowley\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.045]; Ireland[0.045]; captain[0.044]; Quinnell[0.044]; Giffin[0.044]; England[0.043]; Campese[0.042]; Ian[0.042]; Wales[0.042]; Neil[0.042]; Zealand[0.042]; Zealand[0.042]; Gavin[0.042]; Finegan[0.041]; Payne[0.041]; Horan[0.041]; Craig[0.040]; Pat[0.040]; Tim[0.040]; Tim[0.040]; Darren[0.040]; Joe[0.039]; Andrew[0.039]; Norm[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australian ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.238:0.238[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.491:-0.491[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.049]; Wallabies[0.045]; Wallabies[0.045]; Wales[0.045]; Barbarians[0.044]; country[0.043]; match[0.043]; Campese[0.042]; Campese[0.042]; Cup[0.042]; captain[0.041]; played[0.041]; Queenslander[0.041]; David[0.041]; Finegan[0.040]; final[0.040]; final[0.040]; winger[0.040]; following[0.039]; Owen[0.039]; Andrew[0.039]; successive[0.039]; World[0.039]; Rob[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWales\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mWales national rugby union team\u001b[39m <---> \u001b[32mWales\u001b[39m\t\nSCORES: global= 0.254:0.245[\u001b[31m0.009\u001b[39m]; local()= 0.215:0.197[\u001b[31m0.018\u001b[39m]; log p(e|m)= -2.718:-0.373[\u001b[32m2.346\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWales[0.047]; Wales[0.047]; Wales[0.047]; Wales[0.047]; Wales[0.047]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; Ireland[0.041]; Scotland[0.040]; internationals[0.040]; League[0.039]; Owen[0.039]; Barbarians[0.039]; Barbarians[0.039]; Australia[0.038]; Jones[0.038]; Jones[0.038]; Pontypridd[0.038]; Gavin[0.038]; game[0.038]; Leicester[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Scott Quinnell\u001b[39m ==> ENTITY: \u001b[32mScott Quinnell\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.222:0.222[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWales[0.044]; Wales[0.044]; Wales[0.044]; Wales[0.044]; Wales[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; Scotland[0.043]; Ireland[0.042]; Pontypridd[0.041]; internationals[0.041]; Gavin[0.040]; Australia[0.040]; Quinnell[0.040]; game[0.039]; Darren[0.039]; League[0.039]; Leicester[0.039]; Howley[0.039]; Ian[0.039]; Ian[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Cup ==> ENTITY: \u001b[32mRugby World Cup\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.203:0.203[0]; log p(e|m)= -3.058:-3.058[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.045]; Australia[0.045]; Ireland[0.044]; England[0.044]; Wales[0.043]; match[0.043]; League[0.042]; final[0.042]; final[0.042]; final[0.042]; internationals[0.042]; country[0.041]; Italy[0.041]; Scotland[0.041]; Australian[0.041]; Twickenham[0.041]; Barbarians[0.040]; Barbarians[0.040]; Wallabies[0.040]; game[0.040]; played[0.040]; European[0.039]; appearance[0.039]; won[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew Zealand\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew Zealand national rugby union team\u001b[39m <---> \u001b[32mNew Zealand\u001b[39m\t\nSCORES: global= 0.240:0.239[\u001b[31m0.001\u001b[39m]; local()= 0.170:0.151[\u001b[31m0.018\u001b[39m]; log p(e|m)= -3.817:-0.238[\u001b[32m3.578\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.050]; England[0.049]; England[0.049]; England[0.049]; England[0.049]; Ireland[0.047]; Barbarians[0.046]; Wales[0.046]; Wales[0.046]; Wales[0.046]; Wales[0.046]; Wales[0.046]; Scotland[0.045]; Zealand[0.044]; Allan[0.043]; Neil[0.043]; Quinnell[0.042]; Quinnell[0.042]; Teams[0.042]; Scott[0.042]; Howard[0.042]; Campese[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIreland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIreland national rugby union team\u001b[39m <---> \u001b[32mRepublic of Ireland\u001b[39m\t\nSCORES: global= 0.244:0.238[\u001b[31m0.006\u001b[39m]; local()= 0.207:0.116[\u001b[31m0.091\u001b[39m]; log p(e|m)= -3.381:-1.505[\u001b[32m1.876\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.046]; England[0.046]; England[0.046]; England[0.046]; Wales[0.045]; Wales[0.045]; Wales[0.045]; Wales[0.045]; Wales[0.045]; Scotland[0.045]; Australia[0.045]; Barbarians[0.044]; Owen[0.043]; Leicester[0.042]; Howley[0.042]; Neil[0.042]; Horan[0.041]; Finegan[0.041]; Quinnell[0.041]; Quinnell[0.041]; Pontypridd[0.041]; Zealand[0.041]; Zealand[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wales ==> ENTITY: \u001b[32mWales national rugby union team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.207:0.207[0]; log p(e|m)= -2.718:-2.718[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWales[0.049]; Wales[0.049]; Wales[0.049]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; Ireland[0.042]; Scotland[0.041]; Scotland[0.041]; internationals[0.041]; League[0.040]; Twickenham[0.040]; Barbarians[0.040]; Barbarians[0.040]; Barbarians[0.040]; Australia[0.039]; Australia[0.039]; played[0.039]; game[0.039]; selectorial[0.038]; country[0.038]; Howley[0.038]; Cup[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wallabies\u001b[39m ==> ENTITY: \u001b[32mAustralia national rugby union team\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.049]; Australia[0.047]; Wallabies[0.046]; Wales[0.045]; Barbarians[0.044]; Barbarians[0.044]; match[0.043]; Twickenham[0.042]; Australian[0.041]; Campese[0.041]; Campese[0.041]; Campese[0.041]; captain[0.041]; fixture[0.041]; Finegan[0.041]; final[0.040]; large[0.039]; winger[0.039]; European[0.039]; tour[0.039]; tour[0.039]; tour[0.039]; successive[0.039]; weekend[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEngland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEngland national rugby union team\u001b[39m <---> \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.245:0.238[\u001b[31m0.007\u001b[39m]; local()= 0.212:0.174[\u001b[31m0.039\u001b[39m]; log p(e|m)= -3.772:-0.190[\u001b[32m3.582\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.044]; England[0.044]; England[0.044]; England[0.044]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Australia[0.042]; Australia[0.042]; League[0.041]; internationals[0.041]; Ireland[0.041]; Ireland[0.041]; Scotland[0.040]; Scotland[0.040]; Twickenham[0.040]; Barbarians[0.040]; Barbarians[0.040]; Leicester[0.039]; Dale[0.039]; Nigel[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nigel Walker\u001b[39m ==> ENTITY: \u001b[32mNigel Walker\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.045]; game[0.043]; England[0.043]; England[0.043]; England[0.043]; England[0.043]; England[0.043]; Scotland[0.043]; Scotland[0.043]; Ireland[0.042]; Ireland[0.042]; Leicester[0.041]; internationals[0.041]; Wales[0.041]; Wales[0.041]; Wales[0.041]; Wales[0.041]; Wales[0.041]; Wales[0.041]; Jones[0.041]; Jones[0.041]; Australia[0.040]; Australia[0.040]; Teams[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAustralia\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustralia national rugby union team\u001b[39m <---> \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.245:0.241[\u001b[31m0.004\u001b[39m]; local()= 0.192:0.097[\u001b[31m0.095\u001b[39m]; log p(e|m)= -4.200:-0.172[\u001b[32m4.028\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.051]; Wallabies[0.047]; Wales[0.046]; Barbarians[0.045]; Twickenham[0.044]; Campese[0.044]; fixture[0.042]; Queenslander[0.042]; Finegan[0.042]; successive[0.040]; injury[0.040]; European[0.040]; tour[0.040]; tour[0.040]; tour[0.040]; Saturday[0.040]; weekend[0.040]; row[0.040]; victory[0.040]; record[0.040]; took[0.040]; set[0.039]; place[0.039]; Owen[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWales\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mWales national rugby union team\u001b[39m <---> \u001b[32mWales\u001b[39m\t\nSCORES: global= 0.254:0.245[\u001b[31m0.009\u001b[39m]; local()= 0.210:0.187[\u001b[31m0.023\u001b[39m]; log p(e|m)= -2.718:-0.373[\u001b[32m2.346\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWales[0.048]; Wales[0.048]; Wales[0.048]; Wales[0.048]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; Ireland[0.041]; Scotland[0.041]; internationals[0.040]; League[0.040]; Owen[0.040]; Barbarians[0.039]; Barbarians[0.039]; Australia[0.039]; Jones[0.038]; Jones[0.038]; Pontypridd[0.038]; Gavin[0.038]; game[0.038]; Leicester[0.038]; Howley[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Neil Back\u001b[39m ==> ENTITY: \u001b[32mNeil Back\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.045]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; Ireland[0.045]; Leicester[0.044]; Gavin[0.044]; Scotland[0.044]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Quinnell[0.043]; Quinnell[0.043]; Howley[0.043]; internationals[0.042]; captain[0.042]; Barbarians[0.042]; Barbarians[0.042]; Andrew[0.042]; Andrew[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Campese\u001b[39m ==> ENTITY: \u001b[32mDavid Campese\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWallabies[0.047]; Australia[0.047]; Rugby[0.045]; Wales[0.044]; Daniel[0.043]; Daniel[0.043]; Barbarians[0.043]; Jason[0.042]; Queenslander[0.042]; Finegan[0.042]; Owen[0.041]; Twickenham[0.041]; Herbert[0.041]; ahead[0.041]; test[0.041]; miss[0.040]; fellow[0.040]; knee[0.040]; Robert[0.040]; injury[0.040]; ruled[0.040]; European[0.039]; successive[0.039]; took[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cape Town\u001b[39m ==> ENTITY: \u001b[32mCape Town\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.047]; England[0.046]; Ireland[0.045]; played[0.045]; Australia[0.044]; Australia[0.044]; final[0.044]; final[0.044]; final[0.044]; game[0.044]; won[0.044]; Cup[0.043]; spectators[0.043]; League[0.043]; captain[0.043]; internationals[0.043]; Scotland[0.042]; Wales[0.042]; Twickenham[0.042]; appearance[0.042]; country[0.042]; currently[0.042]; Teams[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Matthew Burke\u001b[39m ==> ENTITY: \u001b[32mMatt Burke\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.208:0.208[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.044]; Horan[0.043]; Australia[0.043]; England[0.043]; England[0.043]; England[0.043]; England[0.043]; Finegan[0.042]; Howley[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Barbarians[0.041]; Gavin[0.041]; Quinnell[0.041]; Quinnell[0.041]; Neil[0.041]; Campese[0.041]; Giffin[0.040]; Andrew[0.040]; Andrew[0.040]; Michael[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Daniel Herbert\u001b[39m ==> ENTITY: \u001b[32mDaniel Herbert\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.052]; England[0.052]; England[0.052]; England[0.052]; England[0.052]; Barbarians[0.051]; Ireland[0.051]; Andrew[0.050]; Andrew[0.050]; Scott[0.050]; Matthew[0.050]; Craig[0.050]; Horan[0.049]; captain[0.049]; David[0.049]; David[0.049]; David[0.049]; Campese[0.049]; Gavin[0.048]; Giffin[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Daniel Herbert\u001b[39m ==> ENTITY: \u001b[32mDaniel Herbert\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.045]; Wallabies[0.044]; Wallabies[0.044]; Barbarians[0.044]; Rugby[0.044]; match[0.043]; Australian[0.042]; David[0.042]; Robert[0.042]; Campese[0.042]; Campese[0.042]; Daniel[0.041]; fixture[0.041]; tour[0.041]; tour[0.041]; tour[0.041]; Wales[0.041]; Finegan[0.041]; injury[0.040]; test[0.040]; winger[0.040]; Twickenham[0.040]; Queenslander[0.040]; knee[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barbarians ==> ENTITY: \u001b[32mBarbarian F.C.\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.220:0.220[0]; log p(e|m)= -0.171:-0.171[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; Ireland[0.043]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Barbarians[0.042]; Barbarians[0.042]; Scotland[0.041]; Scotland[0.041]; Australia[0.041]; Australia[0.041]; Cup[0.041]; internationals[0.041]; Twickenham[0.040]; League[0.040]; played[0.039]; tour[0.039]; Howley[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWales\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mWales national rugby union team\u001b[39m <---> \u001b[32mWales\u001b[39m\t\nSCORES: global= 0.254:0.245[\u001b[31m0.009\u001b[39m]; local()= 0.221:0.205[\u001b[31m0.016\u001b[39m]; log p(e|m)= -2.718:-0.373[\u001b[32m2.346\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWales[0.047]; Wales[0.047]; Wales[0.047]; Wales[0.047]; Wales[0.047]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; Ireland[0.040]; Ireland[0.040]; Scotland[0.040]; Scotland[0.040]; internationals[0.040]; League[0.039]; Twickenham[0.039]; Barbarians[0.039]; Barbarians[0.039]; Australia[0.038]; Australia[0.038]; played[0.038]; Jones[0.038]; Jones[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Allan Bateman\u001b[39m ==> ENTITY: \u001b[32mAllan Bateman\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPontypridd[0.046]; Wales[0.045]; Wales[0.045]; Wales[0.045]; Wales[0.045]; Wales[0.045]; Wales[0.045]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; Twickenham[0.043]; internationals[0.042]; Scotland[0.042]; Scotland[0.042]; League[0.042]; Leicester[0.042]; Jones[0.042]; Jones[0.042]; Ireland[0.042]; Ireland[0.042]; game[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wallabies\u001b[39m ==> ENTITY: \u001b[32mAustralia national rugby union team\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.047]; Wallabies[0.047]; Wales[0.046]; Barbarians[0.044]; match[0.043]; Cup[0.042]; played[0.042]; Australian[0.042]; Campese[0.042]; Campese[0.042]; captain[0.042]; Finegan[0.041]; final[0.041]; final[0.041]; country[0.041]; large[0.040]; winger[0.040]; European[0.040]; kicked[0.039]; tour[0.039]; tour[0.039]; successive[0.039]; weekend[0.039]; row[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEngland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEngland national rugby union team\u001b[39m <---> \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.244:0.238[\u001b[31m0.006\u001b[39m]; local()= 0.191:0.164[\u001b[31m0.027\u001b[39m]; log p(e|m)= -3.772:-0.190[\u001b[32m3.582\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.045]; England[0.045]; England[0.045]; England[0.045]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Australia[0.043]; League[0.042]; internationals[0.042]; Ireland[0.042]; Scotland[0.041]; Barbarians[0.041]; Barbarians[0.041]; Owen[0.040]; Leicester[0.040]; Dale[0.040]; David[0.039]; David[0.039]; David[0.039]; Nigel[0.039]; Pontypridd[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dale McIntosh\u001b[39m ==> ENTITY: \u001b[32mDale McIntosh\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPontypridd[0.044]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Scotland[0.043]; England[0.042]; England[0.042]; England[0.042]; England[0.042]; England[0.042]; Quinnell[0.042]; Quinnell[0.042]; forwards[0.041]; Gavin[0.041]; Barbarians[0.041]; Barbarians[0.041]; Darren[0.040]; Ireland[0.040]; Payne[0.040]; Neil[0.039]; internationals[0.039]; Owen[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Twickenham ==> ENTITY: \u001b[32mTwickenham Stadium\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.191:0.191[0]; log p(e|m)= -0.594:-0.594[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.049]; Wales[0.047]; Barbarians[0.044]; Australia[0.044]; final[0.044]; fixture[0.043]; Wallabies[0.043]; London[0.042]; Saturday[0.042]; row[0.041]; Owen[0.041]; weekend[0.041]; victory[0.040]; Centre[0.040]; successive[0.040]; Campese[0.040]; Campese[0.040]; Finegan[0.040]; David[0.040]; Union[0.040]; ahead[0.040]; tour[0.040]; tour[0.040]; tour[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Twickenham ==> ENTITY: \u001b[32mTwickenham Stadium\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.206:0.206[0]; log p(e|m)= -0.594:-0.594[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.046]; England[0.046]; Wales[0.044]; Wales[0.044]; Wales[0.044]; Ireland[0.043]; internationals[0.043]; Scotland[0.042]; Scotland[0.042]; Barbarians[0.041]; Barbarians[0.041]; Barbarians[0.041]; Australia[0.041]; Australia[0.041]; game[0.041]; final[0.041]; final[0.041]; played[0.041]; League[0.040]; Cup[0.040]; captain[0.039]; won[0.039]; Rob[0.039]; Teams[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = David Campese\u001b[39m ==> ENTITY: \u001b[32mDavid Campese\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.206:0.206[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.044]; Gavin[0.043]; England[0.042]; England[0.042]; England[0.042]; Ireland[0.042]; Neil[0.042]; Quinnell[0.042]; Quinnell[0.042]; Howley[0.042]; Horan[0.042]; Michael[0.042]; Wales[0.041]; Wales[0.041]; Wales[0.041]; Wales[0.041]; Wales[0.041]; David[0.041]; David[0.041]; Matthew[0.041]; Scott[0.041]; Andrew[0.041]; Andrew[0.041]; Craig[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mDavid Wilson\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mDavid Wilson (rugby union)\u001b[39m <---> \u001b[32mDave Wilson (rugby union)\u001b[39m\t\nSCORES: global= 0.259:0.239[\u001b[31m0.021\u001b[39m]; local()= 0.146:0.170[\u001b[32m0.024\u001b[39m]; log p(e|m)= -1.917:-3.147[\u001b[31m1.229\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; England[0.044]; England[0.044]; England[0.044]; Ireland[0.043]; Craig[0.042]; Neil[0.042]; Scott[0.042]; Owen[0.042]; Ian[0.041]; Leicester[0.041]; Rob[0.041]; Rob[0.041]; Wales[0.041]; Wales[0.041]; Wales[0.041]; Dale[0.041]; captain[0.040]; Nick[0.040]; Gavin[0.040]; Andrew[0.040]; Andrew[0.040]; David[0.040]; David[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrecord[0.046]; fixture[0.044]; victory[0.043]; tour[0.043]; tour[0.043]; tour[0.043]; Saturday[0.042]; Owen[0.042]; set[0.042]; Wales[0.042]; Farewell[0.042]; European[0.041]; Robert[0.041]; Daniel[0.041]; Daniel[0.041]; injury[0.041]; Centre[0.041]; weekend[0.040]; Little[0.040]; Little[0.040]; Little[0.040]; ahead[0.040]; row[0.040]; Rugby[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEngland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEngland national rugby union team\u001b[39m <---> \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.244:0.236[\u001b[31m0.008\u001b[39m]; local()= 0.218:0.165[\u001b[31m0.053\u001b[39m]; log p(e|m)= -3.772:-0.190[\u001b[32m3.582\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.045]; England[0.045]; England[0.045]; England[0.045]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Australia[0.042]; Australia[0.042]; League[0.041]; internationals[0.041]; Ireland[0.041]; played[0.041]; Scotland[0.040]; Scotland[0.040]; Twickenham[0.040]; Barbarians[0.040]; Barbarians[0.040]; Pontypridd[0.039]; game[0.039]; selectorial[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = European ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.256:-0.256[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWorld[0.044]; fixture[0.044]; record[0.044]; match[0.044]; final[0.043]; tour[0.042]; tour[0.042]; tour[0.042]; Australia[0.042]; Australia[0.042]; place[0.042]; percent[0.041]; David[0.041]; Wales[0.041]; set[0.041]; miss[0.041]; ahead[0.041]; victory[0.041]; Australian[0.041]; took[0.040]; London[0.040]; weekend[0.040]; Saturday[0.040]; spectators[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWales\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mWales national rugby union team\u001b[39m <---> \u001b[32mWales\u001b[39m\t\nSCORES: global= 0.253:0.245[\u001b[31m0.008\u001b[39m]; local()= 0.218:0.205[\u001b[31m0.013\u001b[39m]; log p(e|m)= -2.718:-0.373[\u001b[32m2.346\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWales[0.047]; Wales[0.047]; Wales[0.047]; Wales[0.047]; Wales[0.047]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; Ireland[0.040]; Ireland[0.040]; Scotland[0.040]; Scotland[0.040]; internationals[0.040]; League[0.039]; Twickenham[0.039]; Barbarians[0.039]; Barbarians[0.039]; Australia[0.038]; Australia[0.038]; Jones[0.038]; Jones[0.038]; Pontypridd[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gregor Townsend\u001b[39m ==> ENTITY: \u001b[32mGregor Townsend\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.044]; Scotland[0.044]; England[0.043]; England[0.043]; England[0.043]; England[0.043]; England[0.043]; Scott[0.042]; Ireland[0.042]; Ireland[0.042]; Craig[0.041]; Allan[0.041]; Horan[0.041]; Howley[0.041]; Neil[0.041]; Ian[0.041]; Ian[0.041]; captain[0.041]; Barbarians[0.040]; Barbarians[0.040]; Jones[0.040]; Jones[0.040]; Walker[0.040]; Twickenham[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Campese\u001b[39m ==> ENTITY: \u001b[32mDavid Campese\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWallabies[0.045]; Wallabies[0.045]; Campese[0.045]; Australia[0.044]; Australia[0.044]; Australian[0.043]; Ireland[0.043]; winger[0.042]; Wales[0.042]; Wales[0.042]; David[0.042]; Andrew[0.041]; Andrew[0.041]; Daniel[0.041]; captain[0.041]; Barbarians[0.041]; game[0.041]; Rob[0.040]; Italy[0.040]; played[0.040]; scored[0.040]; match[0.040]; said[0.039]; Twickenham[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPontypridd\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPontypridd RFC\u001b[39m <---> \u001b[32mPontypridd\u001b[39m\t\nSCORES: global= 0.267:0.264[\u001b[31m0.004\u001b[39m]; local()= 0.182:0.168[\u001b[31m0.013\u001b[39m]; log p(e|m)= -1.820:0.000[\u001b[32m1.820\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.043]; England[0.043]; England[0.043]; England[0.043]; England[0.043]; Tony[0.043]; Tony[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Rob[0.041]; Rob[0.041]; Ireland[0.041]; Sam[0.041]; Leicester[0.040]; Neil[0.040]; Darren[0.040]; Gavin[0.040]; Nigel[0.040]; David[0.040]; David[0.040]; David[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ian Jones\u001b[39m ==> ENTITY: \u001b[32mIan Jones (rugby union)\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.002:-0.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGavin[0.043]; Quinnell[0.043]; Quinnell[0.043]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Allan[0.042]; Tony[0.042]; Tony[0.042]; Neil[0.042]; Michael[0.041]; Craig[0.041]; Barbarians[0.041]; Barbarians[0.041]; David[0.041]; David[0.041]; David[0.041]; Scott[0.041]; Finegan[0.041]; internationals[0.041]; forwards[0.041]; Pontypridd[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national rugby union team\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.201:0.201[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.047]; England[0.044]; England[0.044]; England[0.044]; internationals[0.044]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Ireland[0.042]; Barbarians[0.042]; Barbarians[0.042]; Barbarians[0.042]; country[0.042]; League[0.041]; Twickenham[0.040]; Campese[0.040]; selectorial[0.040]; Cup[0.040]; captain[0.040]; Scotland[0.040]; Scotland[0.040]; played[0.040]; game[0.039]; Italy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = David Giffin\u001b[39m ==> ENTITY: \u001b[32mDavid Giffin\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.128:0.128[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.044]; Campese[0.044]; Finegan[0.044]; Wales[0.044]; Wales[0.044]; Wales[0.044]; Pontypridd[0.043]; Quinnell[0.043]; Quinnell[0.043]; England[0.042]; England[0.042]; captain[0.041]; Andrew[0.041]; Andrew[0.041]; Howley[0.041]; David[0.041]; David[0.041]; Daniel[0.041]; Gavin[0.040]; Neil[0.040]; Horan[0.040]; Rob[0.040]; Scott[0.040]; Norm[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Owen Finegan\u001b[39m ==> ENTITY: \u001b[32mOwen Finegan\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.046]; Wales[0.044]; Wallabies[0.044]; Wallabies[0.044]; winger[0.043]; Campese[0.043]; Campese[0.043]; Barbarians[0.043]; Australia[0.043]; knee[0.042]; injury[0.042]; Australian[0.041]; currently[0.041]; Queenslander[0.041]; fixture[0.041]; took[0.040]; Centre[0.040]; row[0.040]; Union[0.040]; place[0.040]; Little[0.040]; Little[0.040]; Little[0.040]; successive[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tim Stimpson\u001b[39m ==> ENTITY: \u001b[32mTim Stimpson\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.045]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; scored[0.041]; kicked[0.041]; Ireland[0.041]; final[0.041]; final[0.041]; game[0.040]; goal[0.040]; Andrew[0.040]; Andrew[0.040]; Ian[0.040]; Ian[0.040]; Scotland[0.040]; Scotland[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rugby Union\u001b[39m ==> ENTITY: \u001b[32mRugby union\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.050]; Wales[0.049]; Wallabies[0.049]; Twickenham[0.047]; Barbarians[0.047]; fixture[0.045]; Finegan[0.044]; Campese[0.044]; weekend[0.044]; Owen[0.043]; Saturday[0.042]; row[0.042]; tour[0.041]; tour[0.041]; injury[0.041]; successive[0.041]; London[0.041]; Little[0.041]; Little[0.041]; Little[0.041]; Queenslander[0.041]; test[0.041]; place[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jason Little\u001b[39m ==> ENTITY: \u001b[32mJason Little (rugby union)\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.051]; Wales[0.050]; Rugby[0.049]; Wallabies[0.049]; Campese[0.047]; Twickenham[0.046]; Barbarians[0.046]; Finegan[0.046]; Daniel[0.045]; Daniel[0.045]; Little[0.045]; Little[0.045]; fixture[0.045]; Owen[0.044]; knee[0.044]; Robert[0.044]; Saturday[0.043]; injury[0.043]; victory[0.043]; Queenslander[0.043]; weekend[0.043]; successive[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nick Popplewell\u001b[39m ==> ENTITY: \u001b[32mNick Popplewell\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.050]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; Australia[0.045]; Scotland[0.044]; Barbarians[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Zealand[0.040]; Zealand[0.040]; Gavin[0.040]; Campese[0.039]; Pontypridd[0.039]; forwards[0.039]; captain[0.038]; Matthew[0.038]; Neil[0.038]; Leicester[0.038]; Pat[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mScotland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mScotland national rugby union team\u001b[39m <---> \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.247:0.239[\u001b[31m0.008\u001b[39m]; local()= 0.228:0.182[\u001b[31m0.046\u001b[39m]; log p(e|m)= -3.381:-0.229[\u001b[32m3.152\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.047]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; Ireland[0.044]; Ireland[0.044]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Scott[0.043]; Neil[0.043]; internationals[0.042]; Craig[0.042]; Australia[0.042]; Australia[0.042]; Allan[0.041]; Barbarians[0.041]; Barbarians[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tony Underwood\u001b[39m ==> ENTITY: \u001b[32mTony Underwood\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.045]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Scotland[0.042]; Scotland[0.042]; Leicester[0.042]; Ireland[0.042]; Ireland[0.042]; Neil[0.040]; Barbarians[0.040]; Barbarians[0.040]; Australia[0.039]; Australia[0.039]; Jones[0.039]; Jones[0.039]; Scott[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barbarians ==> ENTITY: \u001b[32mBarbarian F.C.\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.183:0.183[0]; log p(e|m)= -0.171:-0.171[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.045]; Wales[0.044]; Barbarians[0.044]; Scotland[0.043]; Australia[0.043]; Australia[0.043]; Cup[0.043]; Wallabies[0.043]; Wallabies[0.043]; internationals[0.043]; Twickenham[0.042]; League[0.042]; played[0.041]; tour[0.041]; tour[0.041]; Australian[0.041]; match[0.040]; special[0.040]; game[0.040]; Italy[0.040]; World[0.040]; country[0.039]; Campese[0.039]; Campese[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew Zealand\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew Zealand national rugby union team\u001b[39m <---> \u001b[32mNew Zealand\u001b[39m\t\nSCORES: global= 0.241:0.239[\u001b[31m0.002\u001b[39m]; local()= 0.183:0.156[\u001b[31m0.027\u001b[39m]; log p(e|m)= -3.817:-0.238[\u001b[32m3.578\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.046]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; Ireland[0.043]; internationals[0.042]; Barbarians[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Scotland[0.041]; Zealand[0.040]; Neil[0.039]; Ian[0.039]; Ian[0.039]; captain[0.039]; Quinnell[0.039]; Quinnell[0.039]; Campese[0.039]; Gavin[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Craig Quinnell\u001b[39m ==> ENTITY: \u001b[32mCraig Quinnell\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nforwards[0.043]; Quinnell[0.043]; England[0.043]; England[0.043]; England[0.043]; England[0.043]; England[0.043]; recalled[0.042]; Scotland[0.042]; Gavin[0.042]; Pontypridd[0.042]; Leicester[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; Wales[0.042]; internationals[0.042]; Darren[0.041]; Scott[0.041]; wing[0.040]; Barbarians[0.040]; Ireland[0.040]; Neil[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAustralia\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustralia\u001b[39m <---> \u001b[32mAustralia national rugby union team\u001b[39m\t\nSCORES: global= 0.243:0.241[\u001b[31m0.002\u001b[39m]; local()= 0.143:0.160[\u001b[32m0.017\u001b[39m]; log p(e|m)= -0.172:-4.200[\u001b[31m4.028\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.044]; England[0.044]; England[0.044]; England[0.044]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Ireland[0.042]; Barbarians[0.042]; Scott[0.041]; Michael[0.041]; Campese[0.041]; David[0.040]; David[0.040]; David[0.040]; Darren[0.040]; Horan[0.040]; Zealand[0.040]; Zealand[0.040]; Neil[0.040]; captain[0.040]; Scotland[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly national rugby union team\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.194:0.194[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.044]; England[0.043]; England[0.043]; England[0.043]; England[0.043]; League[0.043]; Wales[0.043]; Wales[0.043]; Wales[0.043]; Scotland[0.043]; Scotland[0.043]; internationals[0.042]; won[0.041]; Cup[0.041]; country[0.041]; Australia[0.041]; Australia[0.041]; Twickenham[0.040]; Teams[0.040]; played[0.040]; Barbarians[0.040]; Barbarians[0.040]; Barbarians[0.040]; game[0.040]; \t\n\r [==========================================>...] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 980ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4289/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1234testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= -0.027:-0.027[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYork[0.131]; Western[0.117]; tankers[0.113]; Desk[0.111]; Hemisphere[0.111]; Mobil[0.110]; Danila[0.104]; Clean[0.104]; Fixtures[0.099]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMobil\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMobil\u001b[39m <---> \u001b[32mExxonMobil\u001b[39m\t\nSCORES: global= 0.263:0.245[\u001b[31m0.019\u001b[39m]; local()= 0.010:-0.012[\u001b[31m0.022\u001b[39m]; log p(e|m)= 0.000:-1.041[\u001b[31m1.041\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntankers[0.088]; Fixtures[0.087]; Danila[0.084]; Western[0.084]; York[0.084]; York[0.084]; Commodities[0.083]; Desk[0.083]; Caribs[0.082]; fixtures[0.081]; Clean[0.081]; Hemisphere[0.080]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = GMT\u001b[39m ==> ENTITY: \u001b[32mGreenwich Mean Time\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.017:0.017[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWestern[0.082]; Fixtures[0.081]; Danila[0.078]; York[0.078]; York[0.078]; Desk[0.077]; Commodities[0.077]; Mobil[0.076]; Caribs[0.076]; fixtures[0.076]; Hemisphere[0.075]; tankers[0.074]; Clean[0.071]; \t\n\r [===========================================>..] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 969ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4292/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1171testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bowyer\u001b[39m ==> ENTITY: \u001b[32mLee Bowyer\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; Leeds[0.046]; Leeds[0.046]; Leeds[0.046]; charge[0.046]; Bowyer[0.045]; Bowyer[0.045]; Bowyer[0.045]; fined[0.045]; fined[0.045]; fined[0.045]; affray[0.043]; Lee[0.043]; fracas[0.043]; striker[0.042]; guilty[0.042]; Friday[0.042]; pay[0.042]; October[0.042]; total[0.041]; injured[0.041]; Fined[0.041]; court[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.095:0.095[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.048]; London[0.048]; England[0.046]; club[0.046]; Middlesbrough[0.044]; Soccer[0.043]; Leeds[0.043]; Leeds[0.043]; Leeds[0.043]; play[0.041]; warned[0.041]; striker[0.041]; security[0.040]; Lee[0.040]; moved[0.040]; Friday[0.040]; October[0.040]; charge[0.040]; pay[0.039]; Saturday[0.039]; August[0.039]; cameras[0.039]; east[0.039]; total[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.095:0.095[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.048]; London[0.048]; England[0.046]; club[0.046]; Middlesbrough[0.044]; Soccer[0.043]; Leeds[0.043]; Leeds[0.043]; Leeds[0.043]; play[0.041]; warned[0.041]; striker[0.041]; security[0.040]; Lee[0.040]; moved[0.040]; Friday[0.040]; October[0.040]; charge[0.040]; pay[0.039]; Saturday[0.039]; August[0.039]; cameras[0.039]; east[0.039]; total[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lee Bowyer\u001b[39m ==> ENTITY: \u001b[32mLee Bowyer\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.045]; Leeds[0.044]; Leeds[0.044]; Leeds[0.044]; charge[0.044]; Bowyer[0.043]; Bowyer[0.043]; Bowyer[0.043]; fined[0.043]; fined[0.043]; fined[0.043]; affray[0.042]; fracas[0.042]; striker[0.040]; guilty[0.040]; Friday[0.040]; pay[0.040]; October[0.040]; total[0.040]; injured[0.040]; place[0.040]; Fined[0.039]; criminal[0.039]; court[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leeds ==> ENTITY: \u001b[32mLeeds United A.F.C.\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.163:0.163[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.049]; Leeds[0.048]; Leeds[0.048]; club[0.046]; Yorkshire[0.046]; Middlesbrough[0.046]; London[0.045]; London[0.045]; London[0.045]; east[0.044]; Soccer[0.044]; play[0.042]; expected[0.042]; Saturday[0.042]; October[0.042]; charge[0.041]; Lee[0.041]; staff[0.041]; staff[0.041]; striker[0.041]; August[0.040]; Friday[0.040]; moved[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leeds ==> ENTITY: \u001b[32mLeeds United A.F.C.\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.124:0.124[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.050]; Leeds[0.049]; Leeds[0.049]; London[0.046]; London[0.046]; London[0.046]; east[0.046]; Soccer[0.046]; October[0.043]; charge[0.043]; Lee[0.043]; staff[0.043]; staff[0.043]; striker[0.042]; Friday[0.042]; ordered[0.041]; total[0.041]; pay[0.040]; guilty[0.040]; McDonald[0.040]; affray[0.040]; fracas[0.040]; security[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leeds ==> ENTITY: \u001b[32mLeeds United A.F.C.\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.123:0.123[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.048]; Leeds[0.047]; Leeds[0.047]; London[0.045]; London[0.045]; London[0.045]; east[0.044]; Soccer[0.044]; October[0.041]; charge[0.041]; Lee[0.041]; staff[0.041]; staff[0.041]; striker[0.041]; Friday[0.040]; ordered[0.039]; total[0.039]; repeat[0.039]; pay[0.039]; warned[0.039]; guilty[0.039]; McDonald[0.039]; affray[0.038]; fracas[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bowyer\u001b[39m ==> ENTITY: \u001b[32mLee Bowyer\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.044]; Leeds[0.043]; Leeds[0.043]; Leeds[0.043]; charge[0.043]; Middlesbrough[0.043]; Bowyer[0.042]; Bowyer[0.042]; Bowyer[0.042]; Bowyer[0.042]; club[0.042]; fined[0.042]; fined[0.042]; fined[0.042]; affray[0.041]; Lee[0.041]; Saturday[0.041]; fracas[0.041]; play[0.041]; striker[0.040]; August[0.040]; guilty[0.040]; Friday[0.040]; pay[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bowyer\u001b[39m ==> ENTITY: \u001b[32mLee Bowyer\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.044]; Leeds[0.043]; Leeds[0.043]; Leeds[0.043]; charge[0.043]; Middlesbrough[0.043]; Bowyer[0.042]; Bowyer[0.042]; Bowyer[0.042]; Bowyer[0.042]; club[0.042]; fined[0.042]; fined[0.042]; fined[0.042]; affray[0.041]; Lee[0.041]; fracas[0.041]; play[0.041]; striker[0.040]; August[0.040]; guilty[0.040]; Friday[0.040]; pay[0.040]; October[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.061:0.061[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.049]; London[0.049]; England[0.047]; Soccer[0.044]; Leeds[0.044]; Leeds[0.044]; Leeds[0.044]; Friday[0.042]; striker[0.042]; Lee[0.041]; October[0.041]; charge[0.041]; security[0.040]; east[0.040]; total[0.040]; members[0.040]; fined[0.040]; fined[0.040]; fined[0.040]; chairs[0.039]; McDonald[0.039]; warned[0.039]; act[0.039]; Fined[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bowyer\u001b[39m ==> ENTITY: \u001b[32mLee Bowyer\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeeds[0.044]; charge[0.044]; Middlesbrough[0.044]; Bowyer[0.044]; Bowyer[0.044]; club[0.043]; fined[0.043]; fined[0.043]; affray[0.042]; Saturday[0.042]; fracas[0.042]; play[0.042]; August[0.041]; guilty[0.041]; Friday[0.041]; pay[0.041]; October[0.041]; total[0.040]; injured[0.040]; place[0.040]; million[0.040]; million[0.040]; expected[0.040]; criminal[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEngland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEngland national football team\u001b[39m <---> \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.238:0.236[\u001b[31m0.002\u001b[39m]; local()= 0.116:0.071[\u001b[31m0.044\u001b[39m]; log p(e|m)= -2.781:-0.190[\u001b[32m2.591\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeeds[0.046]; Leeds[0.046]; Leeds[0.046]; Soccer[0.045]; striker[0.043]; London[0.042]; London[0.042]; London[0.042]; Lee[0.042]; charge[0.042]; total[0.041]; October[0.041]; repeat[0.041]; Bowyer[0.040]; Bowyer[0.040]; Bowyer[0.040]; Bowyer[0.040]; east[0.040]; fined[0.040]; fined[0.040]; fined[0.040]; Friday[0.039]; fracas[0.039]; guilty[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Middlesbrough ==> ENTITY: \u001b[32mMiddlesbrough F.C.\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.518:-0.518[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.050]; Yorkshire[0.047]; Leeds[0.047]; play[0.047]; place[0.044]; Saturday[0.044]; October[0.044]; August[0.043]; charge[0.043]; east[0.043]; London[0.043]; London[0.043]; million[0.043]; million[0.043]; moved[0.043]; expected[0.043]; staff[0.042]; total[0.042]; Bowyer[0.042]; Bowyer[0.042]; Bowyer[0.042]; injured[0.041]; repeat[0.041]; \t\n\r [===========================================>..] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 907ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4305/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1169testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Plymouth ==> ENTITY: \u001b[32mPlymouth Argyle F.C.\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.137:0.137[0]; log p(e|m)= -4.075:-4.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.068]; Exeter[0.068]; second[0.067]; Soccer[0.065]; London[0.064]; Friday[0.063]; Challenge[0.063]; Cup[0.063]; Cup[0.063]; English[0.062]; English[0.062]; Round[0.061]; round[0.060]; Second[0.059]; Result[0.056]; Result[0.056]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Exeter ==> ENTITY: \u001b[32mExeter City F.C.\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.153:0.153[0]; log p(e|m)= -3.650:-3.650[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPlymouth[0.068]; match[0.067]; second[0.066]; Soccer[0.066]; Cup[0.065]; Cup[0.065]; English[0.064]; English[0.064]; Friday[0.063]; round[0.063]; London[0.063]; Challenge[0.061]; Round[0.060]; Second[0.059]; Result[0.053]; Result[0.053]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.053:0.053[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.070]; match[0.068]; Soccer[0.067]; Cup[0.067]; Cup[0.067]; round[0.065]; Friday[0.063]; English[0.062]; English[0.062]; Plymouth[0.062]; Exeter[0.061]; Round[0.060]; Challenge[0.059]; Second[0.059]; Result[0.055]; Result[0.055]; \t\n\r [===========================================>..] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 897ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4308/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1183testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Melbourne ==> ENTITY: \u001b[32mMelbourne\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.168:-0.168[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.047]; Australia[0.047]; match[0.047]; Melbourne[0.046]; Cricket[0.043]; Cricket[0.043]; captain[0.042]; captain[0.042]; Reiffel[0.042]; Glenn[0.042]; Waugh[0.041]; winning[0.041]; Paul[0.040]; Mark[0.040]; Mark[0.040]; Stuart[0.040]; toss[0.040]; Shane[0.040]; competition[0.040]; Michael[0.040]; Greg[0.040]; McGrath[0.039]; Friday[0.039]; elected[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Courtney Walsh\u001b[39m ==> ENTITY: \u001b[32mCourtney Walsh\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMcGrath[0.046]; Stuart[0.045]; Curtly[0.044]; Australia[0.043]; Waugh[0.043]; Gillespie[0.043]; Michael[0.042]; Shivnarine[0.042]; Reiffel[0.042]; Glenn[0.042]; Chanderpaul[0.041]; Bevan[0.041]; Mark[0.041]; Mark[0.041]; Samuels[0.041]; Lara[0.041]; Warne[0.041]; Ian[0.041]; Taylor[0.040]; Ricky[0.040]; Paul[0.040]; Ponting[0.040]; Robert[0.040]; Greg[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sherwin Campbell\u001b[39m ==> ENTITY: \u001b[32mSherwin Campbell\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStuart[0.044]; toss[0.043]; match[0.043]; Curtly[0.043]; Michael[0.043]; Jason[0.042]; Glenn[0.042]; bat[0.042]; Mark[0.042]; Mark[0.042]; Ricky[0.042]; McGrath[0.042]; Australia[0.041]; Australia[0.041]; Shivnarine[0.041]; West[0.041]; West[0.041]; Robert[0.041]; Reiffel[0.041]; Samuels[0.041]; Waugh[0.041]; Ian[0.041]; Junior[0.040]; Taylor[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Roland Holder\u001b[39m ==> ENTITY: \u001b[32mRoland Holder\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncaptain[0.047]; captain[0.047]; Lara[0.046]; Reiffel[0.044]; Ricky[0.044]; Taylor[0.044]; Shivnarine[0.044]; Junior[0.044]; West[0.044]; Samuels[0.044]; Curtly[0.044]; Michael[0.043]; Kenneth[0.043]; Indies[0.043]; Benjamin[0.043]; Gillespie[0.043]; Stuart[0.043]; Robert[0.042]; Healy[0.042]; Jason[0.042]; Glenn[0.042]; Waugh[0.041]; Murray[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Stuart Law\u001b[39m ==> ENTITY: \u001b[32mStuart Law\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCurtly[0.045]; Mark[0.044]; Mark[0.044]; overs[0.043]; Shivnarine[0.043]; Ian[0.043]; Cricket[0.043]; Cricket[0.043]; bat[0.042]; Reiffel[0.042]; Michael[0.042]; Bevan[0.042]; McGrath[0.041]; Australia[0.041]; Australia[0.041]; Lara[0.041]; Greg[0.041]; Tom[0.040]; captain[0.040]; captain[0.040]; captain[0.040]; Ponting[0.040]; Shane[0.040]; toss[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mark Taylor ==> ENTITY: \u001b[32mMark Taylor (cricketer)\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.183:0.183[0]; log p(e|m)= -0.821:-0.821[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.045]; Australia[0.045]; Mark[0.044]; Michael[0.043]; Brian[0.043]; Glenn[0.042]; match[0.042]; Shivnarine[0.042]; Waugh[0.042]; McGrath[0.042]; Jason[0.042]; Robert[0.041]; Stuart[0.041]; Ian[0.041]; toss[0.041]; Reiffel[0.041]; Paul[0.041]; Curtly[0.041]; Campbell[0.040]; Cricket[0.040]; Cricket[0.040]; Greg[0.040]; Walsh[0.040]; Walsh[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nixon McLean\u001b[39m ==> ENTITY: \u001b[32mNixon McLean\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.128:0.128[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStuart[0.044]; Samuels[0.043]; Gillespie[0.043]; Ambrose[0.043]; Michael[0.043]; Curtly[0.042]; Waugh[0.042]; McGrath[0.042]; Bevan[0.042]; Campbell[0.042]; Shivnarine[0.042]; Australia[0.041]; Australia[0.041]; Lara[0.041]; Adams[0.041]; West[0.041]; Indies[0.041]; Taylor[0.041]; Junior[0.041]; Chanderpaul[0.040]; Mark[0.040]; Mark[0.040]; Robert[0.040]; Blewett[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.175:0.175[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.049]; match[0.045]; Cricket[0.044]; Cricket[0.044]; Waugh[0.043]; toss[0.042]; Stuart[0.042]; Reiffel[0.042]; Melbourne[0.042]; Melbourne[0.042]; captain[0.042]; captain[0.042]; bat[0.041]; Shivnarine[0.041]; McGrath[0.041]; Warne[0.041]; overs[0.041]; Gillespie[0.040]; Bevan[0.040]; Ponting[0.040]; Lara[0.040]; World[0.039]; Junior[0.039]; Glenn[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jason Gillespie\u001b[39m ==> ENTITY: \u001b[32mJason Gillespie\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCurtly[0.045]; Shivnarine[0.044]; Reiffel[0.043]; bat[0.043]; Ponting[0.043]; Stuart[0.043]; McGrath[0.042]; Chanderpaul[0.042]; Waugh[0.042]; overs[0.042]; toss[0.042]; Warne[0.041]; match[0.041]; Australia[0.041]; Australia[0.041]; Lara[0.041]; Mark[0.041]; Mark[0.041]; Michael[0.041]; Glenn[0.040]; captain[0.040]; captain[0.040]; captain[0.040]; Bevan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Carl Hooper\u001b[39m ==> ENTITY: \u001b[32mCarl Hooper\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReiffel[0.045]; Curtly[0.045]; Waugh[0.043]; Bevan[0.043]; Stuart[0.043]; Shivnarine[0.043]; Gillespie[0.043]; McGrath[0.042]; match[0.042]; Cricket[0.042]; Australia[0.041]; Australia[0.041]; Ponting[0.041]; Lara[0.041]; Taylor[0.041]; Warne[0.041]; toss[0.041]; captain[0.041]; captain[0.041]; Chanderpaul[0.040]; Ricky[0.040]; Glenn[0.040]; Mark[0.040]; Mark[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ian Healy\u001b[39m ==> ENTITY: \u001b[32mIan Healy\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; Australia[0.048]; Stuart[0.047]; Reiffel[0.045]; Curtly[0.045]; Glenn[0.044]; McGrath[0.044]; Waugh[0.044]; Mark[0.043]; Mark[0.043]; Bevan[0.043]; match[0.043]; captain[0.043]; captain[0.043]; captain[0.043]; Gillespie[0.042]; Michael[0.042]; Shivnarine[0.042]; Warne[0.042]; Cricket[0.042]; Cricket[0.042]; Ponting[0.041]; bat[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ricky Ponting\u001b[39m ==> ENTITY: \u001b[32mRicky Ponting\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; Australia[0.048]; Curtly[0.047]; toss[0.047]; Shivnarine[0.047]; Reiffel[0.047]; Michael[0.047]; Waugh[0.047]; Warne[0.046]; Stuart[0.045]; match[0.045]; Lara[0.045]; McGrath[0.045]; bat[0.044]; Mark[0.044]; Mark[0.044]; Bevan[0.044]; Glenn[0.044]; Cricket[0.044]; Cricket[0.044]; overs[0.044]; Brian[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Melbourne ==> ENTITY: \u001b[32mMelbourne\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.118:0.118[0]; log p(e|m)= -0.168:-0.168[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.047]; Australia[0.047]; match[0.046]; Melbourne[0.046]; Cricket[0.043]; Cricket[0.043]; Stuart[0.042]; captain[0.042]; captain[0.042]; Reiffel[0.042]; Glenn[0.041]; Waugh[0.041]; Ian[0.041]; Murray[0.040]; Mark[0.040]; Mark[0.040]; Paul[0.040]; toss[0.040]; Junior[0.040]; competition[0.040]; Greg[0.039]; McGrath[0.039]; Friday[0.039]; elected[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Robert Samuels\u001b[39m ==> ENTITY: \u001b[32mRobert Samuels\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMichael[0.044]; Curtly[0.044]; Ricky[0.043]; Stuart[0.043]; Cricket[0.043]; West[0.043]; Kenneth[0.042]; Glenn[0.042]; toss[0.042]; Lara[0.042]; Shivnarine[0.042]; Waugh[0.042]; Reiffel[0.042]; bat[0.042]; Chanderpaul[0.041]; Friday[0.041]; Mark[0.041]; Mark[0.041]; Benjamin[0.041]; Jason[0.040]; match[0.040]; Junior[0.040]; Gillespie[0.040]; Carl[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Indies ==> ENTITY: \u001b[32mWest Indies cricket team\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.170:0.170[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.044]; Australia[0.043]; Australia[0.043]; Cricket[0.043]; Cricket[0.043]; Reiffel[0.043]; toss[0.042]; Stuart[0.042]; McGrath[0.042]; bat[0.042]; overs[0.041]; Ponting[0.041]; Ricky[0.041]; Ground[0.041]; captain[0.041]; captain[0.041]; Gillespie[0.041]; Bevan[0.041]; Indies[0.041]; Indies[0.041]; Waugh[0.041]; Michael[0.040]; Mark[0.040]; Mark[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kenneth Benjamin\u001b[39m ==> ENTITY: \u001b[32mKenny Benjamin\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCurtly[0.046]; Glenn[0.044]; Stuart[0.043]; Australia[0.043]; Ricky[0.043]; Bevan[0.043]; Law[0.042]; Michael[0.042]; Mark[0.042]; Mark[0.042]; Ian[0.042]; Gillespie[0.041]; Cricket[0.041]; Shivnarine[0.041]; Waugh[0.041]; Carl[0.041]; Reiffel[0.041]; Greg[0.041]; Shane[0.041]; Holder[0.041]; Samuels[0.040]; Warne[0.040]; Tom[0.040]; Ponting[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Glenn McGrath\u001b[39m ==> ENTITY: \u001b[32mGlenn McGrath\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.229:0.229[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCurtly[0.048]; Warne[0.046]; Shivnarine[0.045]; Australia[0.045]; Australia[0.045]; toss[0.044]; Stuart[0.044]; Waugh[0.044]; overs[0.044]; match[0.044]; Ponting[0.044]; Reiffel[0.044]; Bevan[0.043]; Cricket[0.043]; bat[0.043]; Lara[0.043]; Chanderpaul[0.043]; Mark[0.042]; Mark[0.042]; Gillespie[0.042]; Ian[0.041]; Michael[0.041]; Shane[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Michael Bevan\u001b[39m ==> ENTITY: \u001b[32mMichael Bevan\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.047]; Australia[0.047]; match[0.045]; Curtly[0.045]; Stuart[0.045]; bat[0.045]; Reiffel[0.044]; toss[0.044]; overs[0.044]; McGrath[0.043]; Shivnarine[0.043]; Melbourne[0.043]; Melbourne[0.043]; Cricket[0.043]; Cricket[0.043]; Lara[0.043]; Waugh[0.043]; Warne[0.042]; limited[0.042]; captain[0.042]; captain[0.042]; captain[0.042]; Ponting[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.156:0.156[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; match[0.044]; Cricket[0.044]; Cricket[0.044]; Mark[0.043]; Mark[0.043]; Waugh[0.042]; Michael[0.042]; Curtly[0.041]; toss[0.041]; Stuart[0.041]; Reiffel[0.041]; Melbourne[0.041]; Melbourne[0.041]; Glenn[0.041]; captain[0.041]; captain[0.041]; Walsh[0.041]; Walsh[0.041]; bat[0.040]; Ian[0.040]; Shivnarine[0.040]; Taylor[0.040]; Campbell[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jimmy Adams\u001b[39m ==> ENTITY: \u001b[32mJimmy Adams\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCurtly[0.047]; Stuart[0.044]; Reiffel[0.043]; Mark[0.043]; Mark[0.043]; overs[0.042]; Ian[0.042]; Australia[0.042]; Australia[0.042]; Michael[0.042]; Taylor[0.041]; Glenn[0.041]; Bevan[0.041]; McGrath[0.041]; Gillespie[0.041]; Shivnarine[0.041]; Cricket[0.041]; Waugh[0.041]; Brian[0.041]; captain[0.040]; captain[0.040]; Shane[0.040]; Ricky[0.040]; Jason[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paul Reiffel\u001b[39m ==> ENTITY: \u001b[32mPaul Reiffel\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCurtly[0.044]; Shivnarine[0.043]; bat[0.043]; Australia[0.043]; Australia[0.043]; Cricket[0.042]; Cricket[0.042]; captain[0.042]; captain[0.042]; captain[0.042]; match[0.042]; toss[0.042]; Lara[0.041]; Carl[0.041]; Waugh[0.041]; Melbourne[0.041]; Melbourne[0.041]; Ricky[0.041]; McGrath[0.041]; Mark[0.041]; Mark[0.041]; Glenn[0.040]; Stuart[0.040]; Junior[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Brian Lara\u001b[39m ==> ENTITY: \u001b[32mBrian Lara\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.222:0.222[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nShivnarine[0.044]; McGrath[0.044]; match[0.043]; Waugh[0.043]; Curtly[0.043]; Cricket[0.043]; toss[0.043]; Stuart[0.043]; Michael[0.042]; Reiffel[0.042]; Gillespie[0.042]; Warne[0.042]; Taylor[0.042]; Chanderpaul[0.041]; bat[0.041]; Bevan[0.041]; Samuels[0.041]; Australia[0.040]; Australia[0.040]; Ricky[0.040]; Mark[0.040]; Mark[0.040]; Ponting[0.040]; overs[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tom Moody\u001b[39m ==> ENTITY: \u001b[32mTom Moody\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReiffel[0.044]; Cricket[0.043]; Cricket[0.043]; Australia[0.043]; Australia[0.043]; overs[0.043]; Shivnarine[0.043]; Curtly[0.043]; bat[0.042]; match[0.042]; McGrath[0.042]; Stuart[0.042]; Waugh[0.041]; toss[0.041]; Warne[0.041]; captain[0.041]; captain[0.041]; captain[0.041]; Gillespie[0.041]; Bevan[0.041]; Melbourne[0.041]; Melbourne[0.041]; Ricky[0.040]; Lara[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mark Waugh\u001b[39m ==> ENTITY: \u001b[32mMark Waugh\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCurtly[0.044]; Australia[0.044]; Australia[0.044]; Shivnarine[0.044]; Reiffel[0.044]; match[0.043]; Stuart[0.043]; McGrath[0.043]; Warne[0.043]; toss[0.042]; bat[0.042]; Lara[0.041]; Bevan[0.041]; Cricket[0.041]; Cricket[0.041]; overs[0.041]; Ponting[0.040]; captain[0.040]; captain[0.040]; captain[0.040]; Michael[0.040]; Mark[0.040]; Gillespie[0.040]; Ian[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Curtly Ambrose\u001b[39m ==> ENTITY: \u001b[32mCurtly Ambrose\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.209:0.209[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMcGrath[0.046]; Gillespie[0.045]; Waugh[0.044]; Shivnarine[0.043]; Stuart[0.043]; Chanderpaul[0.042]; Taylor[0.042]; Reiffel[0.042]; Glenn[0.042]; Ricky[0.042]; Lara[0.041]; Bevan[0.041]; Warne[0.041]; Samuels[0.041]; Walsh[0.041]; Mark[0.041]; Mark[0.041]; Australia[0.041]; Michael[0.041]; Ponting[0.040]; Healy[0.040]; Shane[0.040]; Ian[0.040]; Tom[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Indies ==> ENTITY: \u001b[32mWest Indies cricket team\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.164:0.164[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWest[0.045]; match[0.044]; Australia[0.043]; Australia[0.043]; Cricket[0.043]; Cricket[0.043]; Waugh[0.042]; Stuart[0.042]; Reiffel[0.042]; bat[0.042]; Indies[0.041]; overs[0.041]; Ponting[0.041]; toss[0.041]; Ricky[0.041]; Ground[0.041]; captain[0.041]; captain[0.041]; Bevan[0.041]; Gillespie[0.041]; Michael[0.040]; Mark[0.040]; Mark[0.040]; Warne[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Courtney Walsh\u001b[39m ==> ENTITY: \u001b[32mCourtney Walsh\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMcGrath[0.048]; Stuart[0.047]; Australia[0.045]; Australia[0.045]; Waugh[0.045]; Gillespie[0.045]; Michael[0.044]; overs[0.044]; Reiffel[0.043]; Glenn[0.043]; toss[0.043]; match[0.043]; Bevan[0.043]; bat[0.043]; Mark[0.043]; Mark[0.043]; Samuels[0.042]; Warne[0.042]; Ian[0.042]; Taylor[0.042]; Ricky[0.042]; Paul[0.042]; Ponting[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Indies ==> ENTITY: \u001b[32mWest Indies cricket team\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.189:0.189[0]; log p(e|m)= -1.103:-1.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLara[0.044]; Australia[0.043]; Australia[0.043]; Chanderpaul[0.043]; Stuart[0.043]; Curtly[0.043]; Shivnarine[0.043]; match[0.042]; Cricket[0.042]; Reiffel[0.042]; McGrath[0.042]; bat[0.042]; Waugh[0.041]; overs[0.041]; Ponting[0.041]; toss[0.041]; Gillespie[0.041]; Ricky[0.041]; Bevan[0.041]; Indies[0.041]; Samuels[0.040]; Michael[0.040]; Mark[0.040]; Mark[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Junior Murray\u001b[39m ==> ENTITY: \u001b[32mJunior Murray\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStuart[0.046]; Curtly[0.046]; Australia[0.043]; Australia[0.043]; Shivnarine[0.043]; Ricky[0.043]; Reiffel[0.043]; overs[0.042]; Waugh[0.042]; Chanderpaul[0.042]; Lara[0.041]; Ian[0.041]; Glenn[0.041]; Carl[0.041]; Taylor[0.041]; captain[0.041]; captain[0.041]; Jason[0.041]; Indies[0.040]; Benjamin[0.040]; Samuels[0.040]; Jimmy[0.040]; McGrath[0.040]; Cricket[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shane Warne\u001b[39m ==> ENTITY: \u001b[32mShane Warne\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.215:0.215[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReiffel[0.044]; Curtly[0.044]; Stuart[0.043]; Shivnarine[0.043]; Waugh[0.043]; McGrath[0.043]; Australia[0.042]; Australia[0.042]; Cricket[0.042]; Cricket[0.042]; toss[0.042]; Ponting[0.042]; match[0.042]; Lara[0.041]; bat[0.041]; Bevan[0.041]; Mark[0.041]; Mark[0.041]; overs[0.041]; Gillespie[0.040]; Ian[0.040]; Chanderpaul[0.040]; Michael[0.040]; Ricky[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Greg Blewett\u001b[39m ==> ENTITY: \u001b[32mGreg Blewett\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; Australia[0.048]; Curtly[0.048]; match[0.047]; Stuart[0.047]; bat[0.046]; Reiffel[0.046]; Warne[0.046]; Cricket[0.046]; Cricket[0.046]; Mark[0.045]; Mark[0.045]; McGrath[0.045]; Waugh[0.045]; Shivnarine[0.044]; Michael[0.044]; Lara[0.044]; overs[0.044]; Ponting[0.044]; Gillespie[0.044]; Glenn[0.043]; Shane[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shivnarine Chanderpaul\u001b[39m ==> ENTITY: \u001b[32mShivnarine Chanderpaul\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntoss[0.044]; Curtly[0.044]; bat[0.044]; Lara[0.043]; McGrath[0.043]; Reiffel[0.043]; Stuart[0.042]; overs[0.042]; match[0.042]; Waugh[0.042]; Ricky[0.041]; Taylor[0.041]; Cricket[0.041]; Ponting[0.041]; Gillespie[0.041]; Mark[0.041]; Mark[0.041]; Australia[0.041]; Australia[0.041]; Warne[0.040]; Michael[0.040]; captain[0.040]; captain[0.040]; Samuels[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Series ==> ENTITY: \u001b[32mWorld Series Cricket\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.167:0.167[0]; log p(e|m)= -3.324:-3.324[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.047]; Australia[0.047]; Cricket[0.044]; Cricket[0.044]; Stuart[0.043]; match[0.043]; Waugh[0.042]; McGrath[0.042]; Reiffel[0.042]; Warne[0.041]; winning[0.041]; Bevan[0.041]; Melbourne[0.041]; Melbourne[0.041]; captain[0.040]; captain[0.040]; toss[0.040]; Ponting[0.040]; Gillespie[0.040]; Carl[0.040]; Mark[0.040]; Mark[0.040]; Taylor[0.040]; Ian[0.040]; \t\n\r [===========================================>..] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 726ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4341/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1260testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Buenos Aires ==> ENTITY: \u001b[32mBuenos Aires\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.033:0.033[0]; log p(e|m)= -0.126:-0.126[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBuenos[0.052]; Buenos[0.052]; Aires[0.051]; Aires[0.051]; Argentine[0.047]; Rosario[0.046]; Bahia[0.045]; Blanca[0.045]; grain[0.042]; December[0.041]; Soybeans[0.040]; Sorghum[0.040]; Maize[0.039]; Maize[0.039]; Wheat[0.039]; Newsroom[0.038]; Flint[0.038]; Millet[0.037]; Daily[0.037]; Dent[0.036]; fix[0.036]; Avg[0.036]; price[0.036]; Oats[0.035]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Buenos Aires ==> ENTITY: \u001b[32mBuenos Aires\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.033:0.033[0]; log p(e|m)= -0.126:-0.126[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBuenos[0.052]; Buenos[0.052]; Aires[0.051]; Aires[0.051]; Argentine[0.047]; Rosario[0.046]; Bahia[0.045]; Blanca[0.045]; grain[0.042]; December[0.041]; Soybeans[0.040]; Sorghum[0.040]; Maize[0.039]; Maize[0.039]; Wheat[0.039]; Newsroom[0.038]; Flint[0.038]; Millet[0.037]; Daily[0.037]; Dent[0.036]; fix[0.036]; Avg[0.036]; price[0.036]; Oats[0.035]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rosario ==> ENTITY: \u001b[32mRosario, Santa Fe\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.072:0.072[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBlanca[0.048]; Bahia[0.046]; Buenos[0.045]; Buenos[0.045]; Buenos[0.045]; Aires[0.045]; Aires[0.045]; Aires[0.045]; Argentine[0.044]; December[0.043]; Maize[0.041]; Maize[0.041]; Soybeans[0.041]; Dent[0.040]; grain[0.040]; Flint[0.040]; Newsroom[0.039]; fixings[0.039]; Oats[0.039]; fix[0.038]; Wheat[0.038]; price[0.038]; Millet[0.038]; Sorghum[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bahia Blanca\u001b[39m ==> ENTITY: \u001b[32mBahía Blanca\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.087:0.087[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRosario[0.048]; Argentine[0.046]; Buenos[0.045]; Buenos[0.045]; Buenos[0.045]; Maize[0.044]; Maize[0.044]; Avg[0.043]; December[0.043]; Aires[0.043]; Aires[0.043]; Aires[0.043]; Newsroom[0.042]; Oats[0.041]; Sorghum[0.040]; Dent[0.040]; Soybeans[0.040]; grain[0.040]; fixings[0.038]; Wheat[0.037]; fix[0.037]; Millet[0.037]; Daily[0.036]; price[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Argentine ==> ENTITY: \u001b[32mArgentina\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.062:0.062[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBuenos[0.046]; Buenos[0.046]; Buenos[0.046]; Rosario[0.045]; Bahia[0.045]; Blanca[0.044]; Aires[0.044]; Aires[0.044]; Aires[0.044]; December[0.042]; fixings[0.042]; grain[0.042]; Flint[0.041]; Soybeans[0.040]; Daily[0.039]; Oats[0.039]; Millet[0.039]; fix[0.039]; Newsroom[0.039]; price[0.039]; Maize[0.039]; Maize[0.039]; Wheat[0.038]; Sorghum[0.038]; \t\n\r [===========================================>..] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 702ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4346/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1288testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.116:0.116[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npeacekeeping[0.048]; states[0.048]; president[0.047]; envoy[0.046]; according[0.046]; Mauritania[0.046]; Niger[0.046]; election[0.046]; diplomats[0.045]; foreign[0.045]; said[0.045]; Assembly[0.045]; undersecretary[0.045]; General[0.045]; Annan[0.045]; post[0.045]; post[0.045]; Friday[0.045]; special[0.044]; United[0.044]; Kofi[0.044]; Nations[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = African ==> ENTITY: \u001b[32mAfrica\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.080:0.080[0]; log p(e|m)= -0.911:-0.911[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.048]; Africans[0.048]; president[0.045]; Council[0.045]; Council[0.045]; endorsement[0.044]; General[0.043]; General[0.043]; General[0.043]; states[0.043]; Security[0.043]; Security[0.043]; secretary[0.043]; States[0.043]; general[0.043]; said[0.043]; said[0.043]; supported[0.042]; election[0.042]; nominated[0.042]; saying[0.042]; seek[0.042]; job[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mauritania ==> ENTITY: \u001b[32mMauritania\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.183:0.183[0]; log p(e|m)= -0.373:-0.373[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMauritania[0.050]; country[0.047]; Niger[0.046]; Niger[0.046]; Nigeria[0.044]; Ghana[0.044]; Ghana[0.044]; Burundi[0.044]; Ahmedou[0.043]; suspending[0.043]; general[0.043]; general[0.043]; general[0.043]; Ould[0.043]; Africans[0.043]; meeting[0.042]; Italy[0.042]; Abdallah[0.042]; election[0.042]; said[0.042]; said[0.042]; said[0.042]; forward[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Security Council\u001b[39m ==> ENTITY: \u001b[32mUnited Nations Security Council\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecretary[0.046]; Security[0.044]; decision[0.044]; Council[0.044]; General[0.043]; supported[0.043]; general[0.043]; States[0.042]; saying[0.042]; withdrawing[0.041]; suspending[0.041]; vetoes[0.041]; December[0.041]; Assembly[0.041]; states[0.041]; vetoed[0.041]; Nov[0.041]; said[0.040]; vote[0.040]; seek[0.040]; bid[0.040]; council[0.040]; election[0.040]; United[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Amara Essy\u001b[39m ==> ENTITY: \u001b[32mAmara Essy\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npresident[0.044]; president[0.044]; president[0.044]; minister[0.043]; secretary[0.043]; secretary[0.043]; United[0.042]; undersecretary[0.042]; African[0.041]; meeting[0.041]; Ahmedou[0.041]; general[0.041]; general[0.041]; general[0.041]; Organisation[0.041]; Conference[0.041]; General[0.041]; General[0.041]; General[0.041]; Council[0.041]; Assembly[0.041]; Assembly[0.041]; candidacy[0.040]; Burundi[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npeacekeeping[0.043]; states[0.043]; Council[0.043]; Security[0.043]; Diplomats[0.042]; president[0.042]; president[0.042]; Islamic[0.042]; envoy[0.042]; according[0.042]; Mauritania[0.042]; Mauritania[0.042]; missions[0.041]; Niger[0.041]; Niger[0.041]; Organisation[0.041]; election[0.041]; diplomats[0.041]; foreign[0.041]; said[0.041]; said[0.041]; Assembly[0.041]; undersecretary[0.041]; envoys[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Africans ==> ENTITY: \u001b[32mAfrica\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.118:0.118[0]; log p(e|m)= -1.187:-1.187[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrican[0.049]; country[0.046]; States[0.043]; president[0.042]; president[0.042]; Mauritania[0.042]; Niger[0.042]; Niger[0.042]; states[0.041]; Italy[0.041]; said[0.041]; said[0.041]; said[0.041]; saying[0.041]; supported[0.041]; suspending[0.041]; supporters[0.041]; Organisation[0.041]; decision[0.041]; job[0.041]; Ghana[0.040]; case[0.040]; Nigeria[0.040]; meant[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Boutros Boutros-Ghali\u001b[39m ==> ENTITY: \u001b[32mBoutros Boutros-Ghali\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nundersecretary[0.044]; Ghali[0.043]; Annan[0.043]; envoy[0.043]; Council[0.043]; peacekeeping[0.042]; diplomats[0.042]; General[0.042]; Burundi[0.042]; Ahmedou[0.042]; secretary[0.041]; secretary[0.041]; bid[0.041]; said[0.041]; foreign[0.041]; Security[0.041]; general[0.041]; general[0.041]; general[0.041]; Nations[0.041]; scheduled[0.041]; United[0.041]; meeting[0.041]; African[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.112:0.112[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsaying[0.045]; country[0.043]; peacekeeping[0.043]; states[0.042]; Council[0.042]; Security[0.042]; Diplomats[0.042]; president[0.042]; president[0.042]; president[0.042]; candidacy[0.042]; candidacy[0.042]; Islamic[0.042]; envoy[0.041]; according[0.041]; Mauritania[0.041]; Mauritania[0.041]; withdrawing[0.041]; missions[0.041]; Niger[0.041]; Niger[0.041]; Organisation[0.041]; election[0.041]; diplomats[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Joseph Garba\u001b[39m ==> ENTITY: \u001b[32mJoseph Nanven Garba\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrican[0.049]; Burundi[0.048]; Nigeria[0.048]; Niger[0.047]; Niger[0.047]; Ghana[0.046]; General[0.046]; General[0.046]; General[0.046]; Ahmedou[0.046]; Assembly[0.045]; Assembly[0.045]; country[0.045]; suspending[0.045]; Conference[0.044]; Security[0.044]; Security[0.044]; peacekeeping[0.044]; secretary[0.044]; nominated[0.044]; withdrawing[0.044]; general[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hamid Algabid\u001b[39m ==> ENTITY: \u001b[32mHamid Algabid\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npresident[0.043]; president[0.043]; president[0.043]; Assembly[0.042]; Assembly[0.042]; Ahmedou[0.042]; nominated[0.042]; Niger[0.042]; Niger[0.042]; election[0.042]; envoys[0.041]; African[0.041]; general[0.041]; general[0.041]; general[0.041]; Mauritania[0.041]; Mauritania[0.041]; secretary[0.041]; secretary[0.041]; General[0.041]; General[0.041]; General[0.041]; candidacy[0.041]; minister[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mauritania ==> ENTITY: \u001b[32mMauritania\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.183:0.183[0]; log p(e|m)= -0.373:-0.373[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMauritania[0.048]; Niger[0.044]; Niger[0.044]; Nations[0.042]; Nigeria[0.042]; Ghana[0.042]; Ghana[0.042]; African[0.042]; Burundi[0.042]; Ahmedou[0.041]; states[0.041]; general[0.041]; general[0.041]; general[0.041]; Ould[0.041]; Africans[0.041]; meeting[0.041]; Italy[0.040]; Abdallah[0.040]; election[0.040]; said[0.040]; said[0.040]; said[0.040]; forward[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Boutros-Ghali\u001b[39m ==> ENTITY: \u001b[32mBoutros Boutros-Ghali\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGhali[0.044]; Council[0.043]; Council[0.043]; General[0.043]; secretary[0.042]; bid[0.042]; said[0.042]; Security[0.042]; Security[0.042]; decision[0.042]; general[0.042]; vetoed[0.042]; United[0.041]; supported[0.041]; expires[0.041]; States[0.041]; week[0.041]; African[0.041]; members[0.041]; members[0.041]; term[0.041]; seek[0.041]; December[0.040]; council[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kofi Annan\u001b[39m ==> ENTITY: \u001b[32mKofi Annan\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nenvoy[0.045]; Niger[0.045]; Niger[0.045]; undersecretary[0.045]; General[0.044]; said[0.044]; said[0.044]; diplomats[0.044]; envoys[0.044]; peacekeeping[0.044]; Mauritania[0.044]; Mauritania[0.044]; according[0.043]; Council[0.043]; Burundi[0.043]; meeting[0.043]; secretary[0.043]; secretary[0.043]; foreign[0.042]; Ghali[0.042]; Italy[0.042]; states[0.042]; Security[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Niger ==> ENTITY: \u001b[32mNiger\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.465:-0.465[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.046]; Niger[0.046]; Nigeria[0.046]; Assembly[0.044]; Assembly[0.044]; Mauritania[0.044]; Mauritania[0.044]; Representatives[0.043]; Ghana[0.042]; Ghana[0.042]; peacekeeping[0.042]; Garba[0.041]; Burundi[0.041]; diplomats[0.041]; Ahmedou[0.040]; according[0.040]; envoys[0.040]; foreign[0.040]; remained[0.040]; Annan[0.040]; bid[0.039]; envoy[0.039]; Ould[0.039]; Diplomats[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = United Nations\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.126:0.126[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npresident[0.045]; undersecretary[0.044]; secretary[0.044]; secretary[0.044]; Niger[0.043]; peacekeeping[0.043]; envoy[0.042]; Representatives[0.042]; states[0.042]; diplomats[0.042]; Annan[0.042]; foreign[0.041]; Abdallah[0.041]; Organisation[0.041]; according[0.041]; temporarily[0.041]; minister[0.041]; General[0.040]; Mauritania[0.040]; said[0.040]; post[0.040]; post[0.040]; Friday[0.040]; bid[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.058:0.058[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.045]; states[0.044]; Italy[0.044]; withdrawing[0.043]; bid[0.042]; said[0.042]; said[0.042]; said[0.042]; week[0.042]; members[0.042]; members[0.042]; Wednesday[0.041]; voting[0.041]; endorsement[0.041]; remained[0.041]; African[0.041]; nominated[0.041]; saying[0.041]; decision[0.041]; contest[0.041]; vote[0.041]; Nigeria[0.040]; supported[0.040]; suspending[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ghana ==> ENTITY: \u001b[32mGhana\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.141:0.141[0]; log p(e|m)= -0.392:-0.392[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGhana[0.050]; country[0.049]; president[0.045]; president[0.045]; president[0.045]; Nigeria[0.045]; Italy[0.045]; secretary[0.043]; secretary[0.043]; Representatives[0.043]; Council[0.042]; Security[0.042]; Burundi[0.042]; Africans[0.042]; Conference[0.042]; General[0.042]; General[0.042]; General[0.042]; Kofi[0.042]; bid[0.042]; Coast[0.042]; Coast[0.042]; election[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burundi ==> ENTITY: \u001b[32mBurundi\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.374:-0.374[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGhana[0.046]; Ghana[0.046]; Nigeria[0.044]; Mauritania[0.044]; Mauritania[0.044]; Niger[0.044]; Niger[0.044]; Nations[0.043]; Garba[0.042]; Annan[0.042]; African[0.042]; peacekeeping[0.041]; Kofi[0.041]; Africans[0.041]; diplomats[0.040]; General[0.040]; General[0.040]; General[0.040]; Ould[0.040]; envoys[0.040]; Ahmedou[0.039]; Conference[0.039]; envoy[0.039]; Representatives[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ghana ==> ENTITY: \u001b[32mGhana\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.146:0.146[0]; log p(e|m)= -0.392:-0.392[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGhana[0.048]; president[0.044]; president[0.044]; African[0.044]; Italy[0.044]; Nations[0.043]; foreign[0.042]; secretary[0.042]; secretary[0.042]; Representatives[0.041]; Council[0.041]; states[0.041]; Security[0.041]; Burundi[0.041]; Africans[0.041]; General[0.041]; Kofi[0.041]; peacekeeping[0.041]; bid[0.040]; Coast[0.040]; Coast[0.040]; election[0.040]; Niger[0.040]; Niger[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nigeria ==> ENTITY: \u001b[32mNigeria\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.146:0.146[0]; log p(e|m)= -0.315:-0.315[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.050]; Ghana[0.048]; African[0.046]; Africans[0.045]; Italy[0.044]; Representatives[0.044]; bid[0.043]; foreign[0.043]; Niger[0.043]; Niger[0.043]; Coast[0.043]; Coast[0.043]; States[0.043]; Organisation[0.043]; election[0.042]; Mauritania[0.042]; Mauritania[0.042]; states[0.042]; envoy[0.042]; suspending[0.042]; general[0.042]; Council[0.042]; Council[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Niger ==> ENTITY: \u001b[32mNiger\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.191:0.191[0]; log p(e|m)= -0.465:-0.465[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.046]; Niger[0.046]; Nigeria[0.046]; Assembly[0.044]; Assembly[0.044]; Mauritania[0.043]; Mauritania[0.043]; Representatives[0.043]; Ghana[0.042]; Ghana[0.042]; peacekeeping[0.041]; African[0.041]; Garba[0.041]; Burundi[0.041]; diplomats[0.040]; Nations[0.040]; Ahmedou[0.040]; according[0.040]; envoys[0.040]; foreign[0.040]; Annan[0.039]; bid[0.039]; states[0.039]; envoy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Security Council\u001b[39m ==> ENTITY: \u001b[32mUnited Nations Security Council\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecretary[0.047]; president[0.046]; president[0.046]; president[0.046]; peacekeeping[0.044]; minister[0.044]; General[0.044]; General[0.044]; General[0.044]; Annan[0.044]; undersecretary[0.044]; general[0.043]; general[0.043]; general[0.043]; meeting[0.043]; saying[0.042]; withdrawing[0.042]; Representatives[0.042]; Wednesday[0.042]; suspending[0.042]; Assembly[0.042]; Assembly[0.042]; envoy[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ivory Coast\u001b[39m ==> ENTITY: \u001b[32mIvory Coast\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurundi[0.044]; Niger[0.043]; Niger[0.043]; Ghana[0.043]; Ghana[0.043]; Mauritania[0.043]; Mauritania[0.043]; president[0.042]; president[0.042]; president[0.042]; African[0.042]; Conference[0.042]; Friday[0.042]; Coast[0.041]; Nations[0.041]; peacekeeping[0.041]; Nigeria[0.041]; Ivory[0.041]; Italy[0.041]; Africans[0.041]; election[0.040]; candidacy[0.040]; states[0.040]; bid[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ahmedou Ould Abdallah\u001b[39m ==> ENTITY: \u001b[32mAhmedou Ould-Abdallah\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMauritania[0.044]; Mauritania[0.044]; peacekeeping[0.044]; General[0.043]; General[0.043]; Burundi[0.042]; special[0.042]; foreign[0.042]; Annan[0.042]; Representatives[0.042]; Council[0.041]; secretary[0.041]; secretary[0.041]; envoy[0.041]; envoys[0.041]; nominations[0.041]; undersecretary[0.041]; Organisation[0.041]; Niger[0.041]; Niger[0.041]; general[0.040]; general[0.040]; general[0.040]; diplomats[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Security Council\u001b[39m ==> ENTITY: \u001b[32mUnited Nations Security Council\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecretary[0.045]; president[0.044]; Council[0.043]; Security[0.043]; decision[0.043]; General[0.042]; General[0.042]; General[0.042]; supported[0.042]; general[0.042]; States[0.041]; saying[0.041]; country[0.041]; withdrawing[0.041]; states[0.041]; suspending[0.041]; vetoes[0.041]; December[0.041]; Assembly[0.041]; Assembly[0.041]; Wednesday[0.041]; week[0.040]; vetoed[0.040]; United[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Africans ==> ENTITY: \u001b[32mAfrica\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.088:0.088[0]; log p(e|m)= -1.187:-1.187[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrican[0.050]; Burundi[0.044]; president[0.043]; Mauritania[0.043]; according[0.043]; states[0.043]; foreign[0.042]; said[0.042]; Ghana[0.042]; Essy[0.041]; general[0.041]; general[0.041]; nominate[0.041]; candidates[0.041]; nominees[0.041]; Kofi[0.041]; Ivory[0.041]; Nations[0.040]; special[0.040]; Ahmedou[0.040]; secretary[0.040]; Abdallah[0.040]; peacekeeping[0.040]; minister[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = African ==> ENTITY: \u001b[32mAfrica\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.090:0.090[0]; log p(e|m)= -0.911:-0.911[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfricans[0.045]; Mauritania[0.044]; Burundi[0.044]; Ahmedou[0.044]; diplomats[0.043]; president[0.043]; United[0.042]; Niger[0.042]; peacekeeping[0.042]; Organisation[0.042]; envoy[0.042]; Representatives[0.041]; Abdallah[0.041]; General[0.041]; Nations[0.041]; special[0.041]; according[0.041]; states[0.041]; secretary[0.040]; secretary[0.040]; general[0.040]; general[0.040]; general[0.040]; Kofi[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ivory Coast\u001b[39m ==> ENTITY: \u001b[32mIvory Coast\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.046]; Burundi[0.044]; Niger[0.043]; Niger[0.043]; Ghana[0.043]; Ghana[0.043]; Mauritania[0.042]; Mauritania[0.042]; president[0.042]; president[0.042]; president[0.042]; Conference[0.041]; Friday[0.041]; Wednesday[0.041]; Coast[0.041]; peacekeeping[0.041]; supporters[0.041]; Nigeria[0.041]; Ivory[0.040]; Italy[0.040]; Africans[0.040]; election[0.040]; candidacy[0.040]; candidacy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Boutros-Ghali\u001b[39m ==> ENTITY: \u001b[32mBoutros Boutros-Ghali\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGhali[0.044]; Council[0.043]; Council[0.043]; General[0.043]; General[0.043]; General[0.043]; envoys[0.042]; secretary[0.042]; bid[0.042]; said[0.042]; said[0.042]; said[0.042]; Security[0.041]; Security[0.041]; decision[0.041]; general[0.041]; vetoed[0.041]; scheduled[0.041]; United[0.041]; supported[0.041]; meeting[0.041]; Italy[0.041]; States[0.041]; week[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Organisation of the Islamic Conference\u001b[39m ==> ENTITY: \u001b[32mOrganisation of Islamic Cooperation\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.046]; Council[0.044]; meeting[0.043]; foreign[0.043]; envoys[0.043]; Mauritania[0.042]; Mauritania[0.042]; envoy[0.042]; Nigeria[0.042]; states[0.042]; undersecretary[0.042]; diplomats[0.041]; Wednesday[0.041]; peacekeeping[0.041]; Representatives[0.041]; secretary[0.041]; secretary[0.041]; African[0.041]; General[0.041]; General[0.041]; General[0.041]; according[0.041]; Diplomats[0.041]; missions[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.119:0.119[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npeacekeeping[0.043]; states[0.043]; Council[0.043]; Security[0.043]; Diplomats[0.042]; president[0.042]; president[0.042]; Islamic[0.042]; envoy[0.042]; according[0.042]; Mauritania[0.042]; Mauritania[0.042]; missions[0.041]; Niger[0.041]; Niger[0.041]; Organisation[0.041]; election[0.041]; diplomats[0.041]; foreign[0.041]; said[0.041]; said[0.041]; said[0.041]; Assembly[0.041]; Assembly[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.075:0.075[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.046]; Paolo[0.044]; Wednesday[0.043]; Ghana[0.043]; Ghana[0.043]; meeting[0.042]; bid[0.042]; withdrawing[0.042]; temporarily[0.042]; temporarily[0.042]; contest[0.041]; suspending[0.041]; States[0.041]; United[0.041]; saying[0.041]; case[0.041]; job[0.041]; Nov[0.041]; candidacy[0.041]; candidacy[0.041]; missions[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.122:0.122[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npeacekeeping[0.043]; states[0.043]; president[0.042]; Islamic[0.042]; envoy[0.042]; according[0.042]; Mauritania[0.042]; Mauritania[0.042]; missions[0.042]; Niger[0.042]; Niger[0.042]; Organisation[0.042]; election[0.042]; diplomats[0.041]; foreign[0.041]; said[0.041]; Assembly[0.041]; undersecretary[0.041]; General[0.041]; Annan[0.041]; post[0.041]; post[0.041]; Friday[0.041]; special[0.041]; \t\n\r [===========================================>..] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 538ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4380/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1187testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = GMT\u001b[39m ==> ENTITY: \u001b[32mGreenwich Mean Time\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.054:0.054[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThursday[0.050]; Thursday[0.050]; Phoenix[0.046]; Hartford[0.045]; Boston[0.044]; Calgary[0.044]; Corrected[0.044]; Jersey[0.044]; Bay[0.043]; games[0.043]; National[0.043]; Caps[0.043]; headline[0.043]; Los[0.042]; home[0.042]; League[0.042]; second[0.042]; York[0.042]; Angeles[0.042]; Hockey[0.042]; Hockey[0.042]; Clippers[0.041]; Florida[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Boston ==> ENTITY: \u001b[32mBoston Bruins\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.215:0.215[0]; log p(e|m)= -4.343:-4.343[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; Hockey[0.046]; Hockey[0.046]; team[0.045]; team[0.045]; Islanders[0.044]; Islanders[0.044]; Caps[0.043]; League[0.042]; Calgary[0.041]; second[0.041]; Ice[0.041]; Tampa[0.040]; Clippers[0.040]; Louis[0.040]; Thursday[0.040]; Thursday[0.040]; York[0.040]; Florida[0.040]; Hartford[0.039]; Phoenix[0.039]; Jersey[0.039]; National[0.039]; home[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tampa Bay ==> ENTITY: \u001b[32mTampa Bay Lightning\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.212:0.212[0]; log p(e|m)= -2.120:-2.120[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.046]; team[0.046]; League[0.046]; games[0.046]; Florida[0.045]; Islanders[0.043]; Islanders[0.043]; Hockey[0.043]; Hockey[0.043]; Caps[0.042]; Boston[0.041]; York[0.041]; second[0.041]; Calgary[0.041]; Jersey[0.040]; Phoenix[0.040]; home[0.040]; Thursday[0.040]; Thursday[0.040]; Ice[0.040]; Louis[0.039]; Hartford[0.039]; National[0.039]; Clippers[0.035]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Florida ==> ENTITY: \u001b[32mFlorida Panthers\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.201:0.201[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; Hockey[0.046]; Hockey[0.046]; team[0.044]; team[0.044]; Tampa[0.044]; Islanders[0.044]; Islanders[0.044]; Caps[0.044]; League[0.042]; Calgary[0.042]; Phoenix[0.041]; second[0.041]; Boston[0.041]; Jersey[0.040]; Thursday[0.040]; Thursday[0.040]; Ice[0.039]; Hartford[0.039]; Bay[0.039]; Louis[0.039]; home[0.039]; National[0.039]; York[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = National Hockey League\u001b[39m ==> ENTITY: \u001b[32mNational Hockey League\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.048]; Hockey[0.047]; team[0.047]; team[0.047]; Islanders[0.045]; Islanders[0.045]; Caps[0.044]; second[0.043]; Calgary[0.042]; Boston[0.042]; Tampa[0.041]; Ice[0.041]; Jersey[0.041]; Phoenix[0.041]; Florida[0.040]; Hartford[0.040]; York[0.040]; Louis[0.039]; Thursday[0.039]; Thursday[0.039]; home[0.038]; Bay[0.038]; Clippers[0.037]; Angeles[0.034]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Los Angeles ==> ENTITY: \u001b[32mLos Angeles Kings\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.209:0.209[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; Hockey[0.045]; Hockey[0.045]; team[0.044]; team[0.044]; Islanders[0.044]; Islanders[0.044]; League[0.043]; Bay[0.042]; Caps[0.042]; Boston[0.042]; Phoenix[0.041]; Calgary[0.041]; second[0.041]; Tampa[0.041]; Clippers[0.041]; Ice[0.040]; Thursday[0.040]; Thursday[0.040]; Florida[0.040]; Hartford[0.040]; York[0.039]; National[0.039]; Jersey[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Jersey ==> ENTITY: \u001b[32mNew Jersey Devils\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.205:0.205[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; Hockey[0.045]; Hockey[0.045]; Islanders[0.045]; Islanders[0.045]; team[0.044]; team[0.044]; League[0.043]; York[0.042]; Caps[0.042]; Thursday[0.042]; Thursday[0.042]; Clippers[0.041]; Boston[0.041]; Tampa[0.041]; second[0.040]; Calgary[0.040]; Florida[0.040]; Hartford[0.040]; Ice[0.039]; Phoenix[0.039]; Louis[0.039]; National[0.039]; Bay[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NHL ==> ENTITY: \u001b[32mNational Hockey League\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.206:0.206[0]; log p(e|m)= -0.036:-0.036[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.047]; Hockey[0.047]; team[0.046]; team[0.046]; games[0.046]; Islanders[0.045]; Islanders[0.045]; League[0.043]; Caps[0.043]; second[0.042]; Boston[0.041]; Calgary[0.041]; National[0.040]; Hartford[0.040]; Ice[0.040]; Phoenix[0.039]; Jersey[0.039]; Tampa[0.039]; York[0.039]; Thursday[0.039]; Thursday[0.039]; Bay[0.039]; Louis[0.038]; Florida[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NBA\u001b[39m ==> ENTITY: \u001b[32mNational Basketball Association\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; team[0.047]; team[0.047]; League[0.046]; second[0.043]; Phoenix[0.042]; Thursday[0.042]; Thursday[0.042]; Hockey[0.042]; Hockey[0.042]; Boston[0.042]; Clippers[0.042]; National[0.041]; Caps[0.040]; Florida[0.040]; Islanders[0.040]; Islanders[0.040]; Jersey[0.040]; Hartford[0.040]; Los[0.039]; Tampa[0.039]; York[0.039]; Ice[0.039]; Angeles[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Calgary ==> ENTITY: \u001b[32mCalgary Flames\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.194:0.194[0]; log p(e|m)= -3.147:-3.147[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; Hockey[0.046]; Hockey[0.046]; team[0.044]; team[0.044]; Islanders[0.044]; Islanders[0.044]; Caps[0.043]; Phoenix[0.042]; Ice[0.042]; second[0.042]; League[0.042]; Tampa[0.041]; Boston[0.041]; Florida[0.040]; Bay[0.040]; Louis[0.040]; National[0.039]; Thursday[0.039]; Thursday[0.039]; Jersey[0.039]; York[0.039]; Hartford[0.039]; home[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hartford ==> ENTITY: \u001b[32mHartford Whalers\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.214:0.214[0]; log p(e|m)= -3.689:-3.689[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.047]; team[0.047]; Hockey[0.045]; Hockey[0.045]; games[0.045]; Islanders[0.045]; Islanders[0.045]; League[0.044]; Boston[0.044]; Caps[0.042]; Tampa[0.041]; second[0.041]; Ice[0.040]; Calgary[0.040]; Phoenix[0.040]; Florida[0.040]; Jersey[0.040]; York[0.039]; Thursday[0.039]; Thursday[0.039]; Louis[0.039]; Clippers[0.038]; home[0.037]; National[0.037]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew York\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew York\u001b[39m <---> \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.239:0.238[\u001b[31m0.000\u001b[39m]; local()= 0.081:0.082[\u001b[32m0.001\u001b[39m]; log p(e|m)= -0.289:-1.790[\u001b[31m1.501\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.048]; team[0.048]; games[0.046]; League[0.045]; Thursday[0.044]; Thursday[0.044]; Boston[0.044]; second[0.042]; Hockey[0.041]; Hockey[0.041]; Clippers[0.041]; headline[0.041]; Islanders[0.041]; Islanders[0.041]; Tampa[0.041]; National[0.041]; Jersey[0.041]; Hartford[0.040]; Phoenix[0.040]; Florida[0.040]; home[0.039]; Louis[0.039]; Bay[0.037]; Corrected[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Phoenix ==> ENTITY: \u001b[32mPhoenix Coyotes\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.201:0.201[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; team[0.045]; team[0.045]; Hockey[0.045]; Hockey[0.045]; Islanders[0.043]; Islanders[0.043]; Calgary[0.042]; Caps[0.042]; second[0.042]; League[0.042]; Boston[0.042]; Tampa[0.041]; Thursday[0.041]; Thursday[0.041]; Louis[0.041]; Clippers[0.041]; Florida[0.039]; Ice[0.039]; home[0.039]; National[0.039]; Hartford[0.038]; York[0.038]; Los[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NHL ==> ENTITY: \u001b[32mNational Hockey League\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.206:0.206[0]; log p(e|m)= -0.036:-0.036[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHockey[0.046]; Hockey[0.046]; team[0.046]; team[0.046]; games[0.046]; Islanders[0.044]; Islanders[0.044]; League[0.043]; Caps[0.043]; second[0.042]; Calgary[0.042]; Boston[0.041]; National[0.040]; Hartford[0.040]; Tampa[0.040]; Ice[0.040]; Phoenix[0.040]; Jersey[0.039]; York[0.039]; Thursday[0.039]; Thursday[0.039]; Florida[0.039]; Bay[0.038]; Louis[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = St Louis ==> ENTITY: \u001b[32mSt. Louis Blues\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.196:0.196[0]; log p(e|m)= -2.564:-2.564[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; Hockey[0.045]; Hockey[0.045]; team[0.044]; team[0.044]; Islanders[0.044]; Islanders[0.044]; League[0.043]; Boston[0.043]; Calgary[0.042]; Caps[0.042]; Phoenix[0.041]; second[0.041]; Tampa[0.041]; Thursday[0.041]; Thursday[0.041]; home[0.040]; Florida[0.040]; York[0.040]; La[0.039]; National[0.039]; Ice[0.039]; Hartford[0.038]; Jersey[0.038]; \t\n\r [============================================>.] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 459ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4395/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1215testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Prague ==> ENTITY: \u001b[32mPrague\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.090:0.090[0]; log p(e|m)= -0.053:-0.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnational[0.046]; Saturday[0.046]; Czech[0.045]; Czech[0.045]; Josef[0.045]; financing[0.045]; Friday[0.044]; executive[0.044]; coalition[0.044]; premier[0.044]; Vaclav[0.044]; government[0.043]; June[0.043]; general[0.043]; led[0.043]; chairman[0.043]; majority[0.042]; Civic[0.042]; Klaus[0.042]; Reuters[0.042]; Party[0.041]; Party[0.041]; congress[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brno ==> ENTITY: \u001b[32mBrno\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.018:-0.018[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCzech[0.045]; Czech[0.045]; second[0.045]; Josef[0.044]; winning[0.043]; Saturday[0.042]; November[0.042]; Friday[0.042]; Vaclav[0.041]; Klaus[0.041]; Klaus[0.041]; Klaus[0.041]; Klaus[0.041]; June[0.041]; summer[0.041]; premier[0.041]; British[0.041]; developed[0.041]; country[0.040]; Prime[0.040]; Prime[0.040]; economic[0.040]; chairman[0.040]; seats[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.118:0.118[0]; log p(e|m)= -1.207:-1.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.047]; Czech[0.044]; Prague[0.044]; majority[0.043]; seats[0.043]; seats[0.043]; vote[0.043]; Klaus[0.042]; Klaus[0.042]; Klaus[0.042]; Klaus[0.042]; Communist[0.042]; Social[0.041]; wing[0.041]; Margaret[0.041]; general[0.040]; place[0.040]; late[0.040]; Democrats[0.040]; June[0.040]; party[0.040]; party[0.040]; party[0.040]; November[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = British ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.102:0.102[0]; log p(e|m)= -0.489:-0.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.051]; parliament[0.046]; government[0.045]; late[0.045]; parliamentary[0.044]; Margaret[0.044]; economic[0.044]; second[0.044]; week[0.043]; developed[0.043]; June[0.043]; summer[0.043]; November[0.043]; general[0.043]; general[0.043]; politics[0.042]; Thatcher[0.042]; control[0.042]; having[0.042]; coalition[0.042]; executive[0.042]; premier[0.042]; post[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Vaclav Klaus\u001b[39m ==> ENTITY: \u001b[32mVáclav Klaus\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.209:0.209[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nparliamentary[0.043]; economic[0.043]; Czech[0.042]; Czech[0.042]; Czech[0.042]; Prague[0.042]; Klaus[0.042]; Klaus[0.042]; coalition[0.042]; party[0.041]; party[0.041]; party[0.041]; party[0.041]; party[0.041]; party[0.041]; party[0.041]; congress[0.041]; congress[0.041]; congress[0.041]; leadership[0.041]; premier[0.041]; Party[0.041]; elections[0.041]; elections[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Margaret Thatcher\u001b[39m ==> ENTITY: \u001b[32mMargaret Thatcher\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.046]; government[0.043]; proposals[0.043]; proposals[0.043]; politics[0.043]; vote[0.042]; economic[0.042]; coalition[0.042]; coalition[0.042]; party[0.042]; party[0.042]; party[0.042]; party[0.042]; party[0.042]; country[0.041]; fundamental[0.041]; reforms[0.041]; parliament[0.041]; Prime[0.041]; Prime[0.041]; criticism[0.040]; said[0.040]; said[0.040]; parliamentary[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.104:0.104[0]; log p(e|m)= -1.207:-1.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCzech[0.047]; Prague[0.045]; national[0.044]; making[0.044]; Josef[0.044]; Party[0.043]; Party[0.043]; June[0.042]; Civic[0.042]; Democratic[0.041]; Foreign[0.041]; party[0.041]; party[0.041]; party[0.041]; party[0.041]; Minister[0.041]; chairman[0.041]; vice[0.040]; vice[0.040]; Vice[0.040]; premier[0.040]; Pm[0.040]; having[0.040]; told[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zieleniec\u001b[39m ==> ENTITY: \u001b[32mJosef Zieleniec\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nchairman[0.043]; politics[0.043]; Vaclav[0.042]; parliamentary[0.042]; Josef[0.042]; congress[0.042]; congress[0.042]; congress[0.042]; Prague[0.042]; elections[0.042]; Vice[0.042]; Minister[0.042]; Minister[0.042]; Minister[0.042]; said[0.041]; said[0.041]; said[0.041]; Foreign[0.041]; Klaus[0.041]; Klaus[0.041]; Modernisation[0.041]; executive[0.041]; government[0.041]; Brno[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.133:0.133[0]; log p(e|m)= -1.207:-1.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.047]; Czech[0.045]; Czech[0.045]; style[0.043]; Klaus[0.042]; Klaus[0.042]; Klaus[0.042]; Klaus[0.042]; Klaus[0.042]; November[0.042]; Communist[0.042]; British[0.041]; June[0.041]; June[0.041]; Vaclav[0.041]; late[0.040]; house[0.040]; party[0.040]; party[0.040]; party[0.040]; party[0.040]; calls[0.040]; Margaret[0.040]; developed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.128:0.128[0]; log p(e|m)= -1.207:-1.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.048]; Czech[0.045]; Klaus[0.043]; Klaus[0.043]; Klaus[0.043]; Klaus[0.043]; Communist[0.043]; British[0.043]; Josef[0.042]; economic[0.042]; Vaclav[0.042]; late[0.042]; city[0.041]; June[0.041]; developed[0.041]; summer[0.040]; second[0.040]; diverse[0.040]; Margaret[0.040]; Brno[0.040]; premier[0.040]; present[0.040]; Friday[0.040]; style[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Josef Zieleniec\u001b[39m ==> ENTITY: \u001b[32mJosef Zieleniec\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nchairman[0.043]; Vaclav[0.043]; parliamentary[0.043]; congress[0.042]; congress[0.042]; congress[0.042]; Prague[0.042]; elections[0.042]; Vice[0.042]; Minister[0.042]; Minister[0.042]; said[0.042]; said[0.042]; said[0.042]; Foreign[0.042]; Klaus[0.041]; Klaus[0.041]; Modernisation[0.041]; executive[0.041]; government[0.041]; Brno[0.041]; Party[0.041]; Party[0.041]; premier[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zieleniec\u001b[39m ==> ENTITY: \u001b[32mJosef Zieleniec\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCommunist[0.043]; politics[0.043]; economic[0.042]; reforms[0.042]; Vaclav[0.042]; leadership[0.042]; parliamentary[0.042]; congress[0.042]; congress[0.042]; elections[0.042]; elections[0.042]; elections[0.042]; elections[0.042]; Minister[0.041]; Minister[0.041]; said[0.041]; said[0.041]; parliament[0.041]; parliament[0.041]; Klaus[0.041]; Klaus[0.041]; Klaus[0.041]; Klaus[0.041]; Klaus[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.059:0.059[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.048]; told[0.046]; fundamental[0.044]; Klaus[0.044]; Klaus[0.044]; summer[0.044]; national[0.044]; Congress[0.044]; government[0.044]; politics[0.043]; said[0.043]; said[0.043]; said[0.043]; congress[0.043]; congress[0.043]; congress[0.043]; premier[0.043]; developed[0.043]; Foreign[0.043]; general[0.042]; general[0.042]; executive[0.042]; Democratic[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = post-Communist\u001b[39m ==> ENTITY: \u001b[32mPost-communism\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.216:0.216[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \neconomic[0.047]; country[0.047]; Communist[0.045]; politics[0.043]; Prague[0.042]; reforms[0.042]; opposition[0.042]; coalition[0.042]; Czech[0.041]; Czech[0.041]; Czech[0.041]; parliament[0.041]; parliament[0.041]; elections[0.041]; elections[0.041]; elections[0.041]; party[0.041]; party[0.041]; party[0.041]; British[0.041]; Vaclav[0.040]; Social[0.040]; leadership[0.040]; vote[0.040]; \t\n\r [============================================>.] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 389ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4409/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1327testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canberra ==> ENTITY: \u001b[32mCanberra\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.092:-0.092[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralian[0.048]; Australian[0.048]; Australian[0.048]; major[0.043]; Chris[0.042]; inquiry[0.042]; inquiry[0.042]; inquiry[0.042]; service[0.042]; service[0.042]; months[0.041]; Hunt[0.041]; Court[0.041]; federal[0.041]; Friday[0.041]; left[0.040]; existing[0.040]; head[0.040]; head[0.040]; head[0.040]; high[0.040]; investigation[0.040]; investigation[0.040]; forced[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australian ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.095:0.095[0]; log p(e|m)= -0.491:-0.491[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralian[0.047]; Australian[0.047]; newspaper[0.046]; Canberra[0.045]; told[0.044]; service[0.043]; service[0.043]; Laurie[0.042]; Hunt[0.042]; major[0.042]; Chris[0.041]; Today[0.041]; said[0.041]; said[0.041]; Court[0.041]; Friday[0.041]; original[0.040]; months[0.040]; ahead[0.039]; continuing[0.039]; stand[0.039]; foreign[0.039]; affairs[0.039]; federal[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australian ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.076:0.076[0]; log p(e|m)= -0.491:-0.491[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralian[0.047]; Neil[0.046]; newspaper[0.046]; government[0.043]; government[0.043]; government[0.043]; service[0.043]; reported[0.043]; foreign[0.042]; Downer[0.042]; having[0.041]; report[0.041]; Pamela[0.041]; said[0.041]; April[0.040]; whitewash[0.040]; original[0.040]; allegations[0.040]; Department[0.040]; secure[0.040]; pressing[0.040]; commitment[0.039]; Alexander[0.039]; administrative[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alexander Downer\u001b[39m ==> ENTITY: \u001b[32mAlexander Downer\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralian[0.045]; Australian[0.045]; Minister[0.044]; appointment[0.043]; government[0.042]; government[0.042]; government[0.042]; Foreign[0.042]; Foreign[0.042]; diplomat[0.042]; inquiry[0.041]; inquiry[0.041]; inquiry[0.041]; foreign[0.041]; foreign[0.041]; diplomatic[0.041]; said[0.041]; said[0.041]; Department[0.040]; diplomats[0.040]; appointing[0.040]; allegations[0.040]; senior[0.040]; April[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cambodian ==> ENTITY: \u001b[32mCambodia\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.057:0.057[0]; log p(e|m)= -0.331:-0.331[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngovernment[0.045]; government[0.045]; foreign[0.043]; said[0.043]; administrative[0.043]; say[0.043]; children[0.042]; certain[0.042]; allegations[0.042]; Minister[0.042]; internal[0.042]; diplomats[0.041]; Downer[0.041]; Foreign[0.041]; Foreign[0.041]; influence[0.041]; spokesman[0.041]; service[0.041]; inquiry[0.041]; inquiry[0.041]; appointing[0.041]; appointment[0.041]; commitment[0.041]; pursue[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australian ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.090:0.090[0]; log p(e|m)= -0.491:-0.491[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralian[0.048]; Australian[0.048]; newspaper[0.046]; Canberra[0.046]; told[0.044]; service[0.043]; service[0.043]; Laurie[0.042]; Hunt[0.042]; major[0.042]; Chris[0.041]; Today[0.041]; said[0.041]; Court[0.041]; Friday[0.041]; months[0.040]; continuing[0.039]; stand[0.039]; foreign[0.039]; affairs[0.039]; federal[0.039]; diplomatic[0.039]; diplomatic[0.039]; high[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mForeign Affairs\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mForeign Affairs\u001b[39m <---> \u001b[32mMinister for Foreign Affairs (Australia)\u001b[39m\t\nSCORES: global= 0.251:0.245[\u001b[31m0.007\u001b[39m]; local()= 0.131:0.188[\u001b[32m0.056\u001b[39m]; log p(e|m)= -0.526:-3.442[\u001b[31m2.916\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAffairs[0.045]; Foreign[0.045]; affairs[0.044]; appointment[0.043]; Minister[0.043]; foreign[0.042]; foreign[0.042]; newspaper[0.042]; Neil[0.042]; Department[0.041]; Downer[0.041]; government[0.041]; government[0.041]; government[0.041]; Australian[0.041]; Australian[0.041]; allegations[0.041]; spokesman[0.041]; spokesman[0.041]; report[0.040]; April[0.040]; appointing[0.040]; internal[0.040]; expert[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Laurie Brereton\u001b[39m ==> ENTITY: \u001b[32mLaurie Brereton\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralian[0.045]; Australian[0.045]; Downer[0.045]; Affairs[0.043]; Affairs[0.043]; inquiry[0.042]; inquiry[0.042]; inquiry[0.042]; inquiry[0.042]; inquiry[0.042]; Minister[0.042]; Department[0.041]; government[0.041]; government[0.041]; evidence[0.041]; months[0.041]; senior[0.041]; affairs[0.040]; administrative[0.040]; federal[0.040]; appointment[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australian ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.075:0.075[0]; log p(e|m)= -0.491:-0.491[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralian[0.048]; Neil[0.047]; newspaper[0.047]; reported[0.043]; service[0.043]; having[0.042]; report[0.042]; Downer[0.041]; Pamela[0.041]; said[0.041]; foreign[0.041]; Foreign[0.041]; April[0.041]; whitewash[0.041]; children[0.041]; allegations[0.041]; Department[0.040]; diplomat[0.040]; commitment[0.040]; diplomats[0.040]; Alexander[0.040]; administrative[0.040]; Affairs[0.040]; expert[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australian ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.491:-0.491[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralian[0.048]; Australian[0.048]; newspaper[0.047]; Canberra[0.046]; told[0.044]; service[0.043]; service[0.043]; Hunt[0.042]; major[0.042]; Chris[0.042]; Today[0.041]; said[0.041]; Court[0.041]; Friday[0.041]; months[0.040]; continuing[0.039]; stand[0.039]; federal[0.039]; diplomatic[0.039]; diplomatic[0.039]; high[0.039]; left[0.039]; court[0.039]; court[0.039]; \t\n\r [============================================>.] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 339ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4419/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1223testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = American ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.073:0.073[0]; log p(e|m)= -0.130:-0.130[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.049]; Airlines[0.047]; British[0.046]; British[0.046]; airlines[0.045]; Airways[0.044]; Britain[0.043]; Britain[0.043]; Britain[0.043]; carriers[0.042]; number[0.041]; States[0.041]; United[0.040]; deal[0.040]; government[0.040]; Atlantic[0.039]; Atlantic[0.039]; Ian[0.039]; alliance[0.039]; alliance[0.039]; air[0.039]; proposed[0.039]; added[0.038]; London[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = British Airways\u001b[39m ==> ENTITY: \u001b[32mBritish Airways\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairline[0.046]; British[0.043]; British[0.043]; British[0.043]; British[0.043]; Airways[0.043]; Airways[0.043]; Airways[0.043]; Airways[0.043]; Heathrow[0.043]; London[0.041]; London[0.041]; London[0.041]; services[0.041]; similar[0.041]; USAir[0.041]; Europe[0.040]; routes[0.040]; American[0.040]; American[0.040]; American[0.040]; American[0.040]; drop[0.040]; drop[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.044]; Britain[0.044]; month[0.044]; American[0.043]; American[0.043]; American[0.043]; latest[0.042]; States[0.042]; date[0.042]; case[0.042]; terms[0.041]; British[0.041]; British[0.041]; Europe[0.041]; Office[0.041]; air[0.041]; deal[0.041]; deal[0.041]; deal[0.041]; deal[0.041]; asked[0.040]; actual[0.040]; currently[0.040]; drop[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ian Lang\u001b[39m ==> ENTITY: \u001b[32mIan Lang, Baron Lang of Monkton\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSecretary[0.045]; Britain[0.044]; Britain[0.044]; Britain[0.044]; Britain[0.044]; British[0.044]; British[0.044]; London[0.043]; Plc[0.042]; number[0.041]; Office[0.041]; Ba[0.041]; Trade[0.041]; Industry[0.041]; Commission[0.041]; Lang[0.040]; met[0.040]; said[0.040]; Mergers[0.040]; month[0.040]; warned[0.040]; Friday[0.039]; routes[0.039]; routes[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = British Airways\u001b[39m ==> ENTITY: \u001b[32mBritish Airways\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.044]; British[0.044]; Airways[0.044]; Airways[0.044]; London[0.043]; London[0.043]; services[0.043]; USAir[0.042]; does[0.042]; routes[0.042]; American[0.042]; American[0.042]; drop[0.041]; Office[0.041]; Office[0.041]; said[0.040]; said[0.040]; route[0.040]; route[0.040]; introduction[0.040]; joint[0.040]; added[0.040]; additional[0.040]; allow[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = American Airlines\u001b[39m ==> ENTITY: \u001b[32mAmerican Airlines\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairlines[0.047]; Airways[0.045]; American[0.044]; Trade[0.043]; carriers[0.043]; air[0.043]; month[0.042]; Friday[0.042]; United[0.042]; deal[0.041]; routes[0.041]; asked[0.041]; added[0.041]; Office[0.041]; create[0.040]; alliance[0.040]; alliance[0.040]; Atlantic[0.040]; Atlantic[0.040]; Britain[0.040]; Britain[0.040]; Britain[0.040]; Britain[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Europe ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.076:0.076[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairline[0.044]; competition[0.043]; government[0.043]; market[0.043]; market[0.043]; British[0.043]; British[0.043]; British[0.043]; British[0.043]; latest[0.042]; USAir[0.042]; Industry[0.041]; American[0.041]; American[0.041]; American[0.041]; terms[0.041]; Corp[0.041]; drop[0.041]; drop[0.041]; agreement[0.041]; Airways[0.040]; Airways[0.040]; Airways[0.040]; Airways[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = American ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.047:0.047[0]; log p(e|m)= -0.130:-0.130[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.048]; British[0.045]; British[0.045]; British[0.045]; British[0.045]; airline[0.045]; USAir[0.044]; Airways[0.043]; Airways[0.043]; Airways[0.043]; Airways[0.043]; Boston[0.042]; services[0.040]; agreement[0.040]; deal[0.039]; additional[0.039]; Dallas[0.039]; government[0.039]; routes[0.039]; Jan[0.039]; Worth[0.039]; Fort[0.039]; Atlantic[0.039]; suggested[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = British Airways\u001b[39m ==> ENTITY: \u001b[32mBritish Airways\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairline[0.051]; British[0.047]; British[0.047]; British[0.047]; Airways[0.047]; Airways[0.047]; Airways[0.047]; Heathrow[0.047]; London[0.045]; London[0.045]; passengers[0.045]; passengers[0.045]; similar[0.045]; Europe[0.044]; routes[0.044]; American[0.044]; American[0.044]; American[0.044]; drop[0.044]; unprecedented[0.043]; airport[0.043]; terms[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.102:0.102[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.046]; Britain[0.046]; Britain[0.046]; Britain[0.046]; final[0.046]; American[0.045]; American[0.045]; competition[0.045]; British[0.043]; British[0.043]; number[0.043]; Ian[0.043]; month[0.043]; Friday[0.042]; compete[0.042]; came[0.042]; added[0.042]; loss[0.042]; currently[0.042]; Lang[0.042]; Lang[0.042]; London[0.042]; set[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.173:0.173[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.052]; Britain[0.052]; Britain[0.052]; British[0.048]; British[0.048]; London[0.042]; Commission[0.041]; American[0.041]; American[0.041]; Atlantic[0.040]; Atlantic[0.040]; government[0.040]; month[0.040]; States[0.040]; number[0.039]; proposed[0.039]; proposed[0.039]; Ian[0.039]; Friday[0.039]; Office[0.038]; Trade[0.038]; supported[0.038]; Secretary[0.038]; met[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = British Airways\u001b[39m ==> ENTITY: \u001b[32mBritish Airways\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairline[0.048]; British[0.045]; British[0.045]; British[0.045]; British[0.045]; Airways[0.044]; Airways[0.044]; Airways[0.044]; Airways[0.044]; Heathrow[0.044]; London[0.043]; London[0.043]; London[0.043]; services[0.043]; similar[0.043]; USAir[0.042]; does[0.042]; Europe[0.042]; routes[0.042]; American[0.042]; American[0.042]; American[0.042]; American[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.046]; Britain[0.046]; Britain[0.046]; Plc[0.046]; British[0.045]; British[0.045]; routes[0.043]; traffic[0.041]; Airlines[0.041]; number[0.041]; Friday[0.041]; Secretary[0.041]; Ian[0.041]; airlines[0.040]; skies[0.040]; deal[0.040]; Airways[0.040]; air[0.040]; States[0.040]; American[0.040]; American[0.040]; proposed[0.039]; United[0.039]; final[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = American ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.046:0.046[0]; log p(e|m)= -0.130:-0.130[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.049]; British[0.046]; British[0.046]; British[0.046]; USAir[0.045]; Airways[0.044]; Airways[0.044]; Airways[0.044]; services[0.041]; Boston[0.041]; agreement[0.040]; additional[0.040]; Dallas[0.040]; Office[0.040]; Office[0.040]; routes[0.040]; Jan[0.040]; said[0.039]; said[0.039]; Worth[0.039]; responses[0.039]; Atlantic[0.039]; suggested[0.039]; government[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = American ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.062:0.062[0]; log p(e|m)= -0.130:-0.130[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.046]; American[0.046]; airline[0.045]; USAir[0.044]; British[0.043]; British[0.043]; British[0.043]; British[0.043]; Airways[0.043]; Airways[0.043]; Airways[0.043]; Airways[0.043]; Europe[0.041]; latest[0.040]; airport[0.040]; government[0.040]; agreement[0.040]; said[0.039]; said[0.039]; said[0.039]; deal[0.039]; deal[0.039]; deal[0.039]; additional[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Office of Fair Trading\u001b[39m ==> ENTITY: \u001b[32mOffice of Fair Trading\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTrade[0.047]; Britain[0.045]; Britain[0.045]; Britain[0.045]; Commission[0.045]; month[0.043]; Plc[0.042]; British[0.042]; British[0.042]; Uk[0.042]; complied[0.042]; States[0.041]; London[0.041]; Ba[0.041]; competition[0.040]; added[0.040]; date[0.040]; case[0.040]; Monopolies[0.040]; proposed[0.040]; proposed[0.040]; asked[0.040]; signed[0.040]; undertakings[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Monopolies and Mergers Commission\u001b[39m ==> ENTITY: \u001b[32mCompetition Commission\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPlc[0.044]; Britain[0.043]; Britain[0.043]; Britain[0.043]; Britain[0.043]; Office[0.043]; month[0.043]; undertakings[0.042]; British[0.042]; British[0.042]; Trading[0.042]; deal[0.042]; asked[0.041]; added[0.041]; clearance[0.041]; Trade[0.041]; London[0.041]; examine[0.041]; United[0.041]; Friday[0.040]; warned[0.040]; competition[0.040]; complied[0.040]; Secretary[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.065:0.065[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.050]; British[0.047]; British[0.047]; British[0.047]; British[0.047]; route[0.045]; route[0.045]; routes[0.045]; airline[0.043]; allow[0.043]; Office[0.042]; Office[0.042]; services[0.042]; imposed[0.042]; introduction[0.042]; Airways[0.042]; Airways[0.042]; Airways[0.042]; Airways[0.042]; Worth[0.041]; USAir[0.041]; Boston[0.041]; party[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = American ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.071:0.071[0]; log p(e|m)= -0.130:-0.130[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.046]; American[0.046]; airline[0.045]; Airways[0.043]; Airways[0.043]; Airways[0.043]; Airways[0.043]; British[0.043]; British[0.043]; British[0.043]; British[0.043]; deal[0.041]; deal[0.041]; deal[0.041]; Europe[0.041]; latest[0.040]; airport[0.040]; Atlantic[0.040]; government[0.040]; signed[0.040]; currently[0.039]; said[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = USAir\u001b[39m ==> ENTITY: \u001b[32mUS Airways\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairline[0.046]; Airways[0.045]; Airways[0.045]; Airways[0.045]; Airways[0.045]; American[0.043]; American[0.043]; American[0.043]; airport[0.042]; routes[0.042]; agreement[0.041]; services[0.041]; Heathrow[0.041]; Dallas[0.041]; deal[0.041]; added[0.040]; slots[0.040]; slots[0.040]; Atlantic[0.040]; additional[0.040]; Boston[0.040]; divestiture[0.039]; route[0.039]; route[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.047]; airport[0.046]; Heathrow[0.045]; Britain[0.045]; British[0.044]; British[0.044]; British[0.044]; route[0.043]; routes[0.042]; routes[0.042]; month[0.040]; airline[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; traffic[0.040]; signed[0.040]; date[0.040]; latest[0.040]; passengers[0.040]; passengers[0.040]; skies[0.039]; came[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.174:0.174[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.051]; Britain[0.051]; Britain[0.051]; British[0.048]; British[0.048]; London[0.042]; Commission[0.041]; American[0.041]; American[0.041]; States[0.040]; Atlantic[0.040]; Atlantic[0.040]; Atlantic[0.040]; Office[0.040]; government[0.040]; month[0.039]; Secretary[0.039]; number[0.039]; proposed[0.039]; proposed[0.039]; Ian[0.039]; Friday[0.039]; Uk[0.038]; latest[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Office of Fair Trading\u001b[39m ==> ENTITY: \u001b[32mOffice of Fair Trading\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.122:0.122[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFair[0.054]; Office[0.050]; Trade[0.048]; British[0.043]; British[0.043]; services[0.043]; London[0.042]; London[0.042]; Worth[0.041]; called[0.040]; clause[0.040]; said[0.040]; said[0.040]; routes[0.040]; equivalent[0.040]; Dallas[0.040]; Jan[0.039]; trans[0.039]; event[0.039]; Fort[0.039]; reduce[0.039]; suggested[0.039]; access[0.039]; access[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.154:0.154[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.053]; Britain[0.053]; British[0.049]; British[0.049]; London[0.043]; Commission[0.042]; American[0.042]; American[0.042]; Atlantic[0.041]; government[0.041]; States[0.041]; number[0.040]; proposed[0.040]; Ian[0.040]; Friday[0.040]; Trade[0.039]; Secretary[0.039]; met[0.039]; deal[0.039]; Plc[0.038]; added[0.038]; warned[0.038]; refer[0.037]; open[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = British ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.489:-0.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.053]; Britain[0.053]; Britain[0.053]; British[0.049]; Ian[0.043]; London[0.043]; American[0.043]; American[0.043]; Atlantic[0.041]; Atlantic[0.041]; government[0.041]; States[0.041]; number[0.040]; Secretary[0.038]; Commission[0.038]; Plc[0.038]; added[0.038]; Lang[0.038]; Lang[0.038]; proposed[0.038]; Edna[0.037]; create[0.037]; greater[0.037]; said[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = AMR Corp.\u001b[39m ==> ENTITY: \u001b[32mAMR Corporation\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairline[0.046]; American[0.044]; American[0.044]; American[0.044]; Airways[0.043]; Airways[0.043]; Airways[0.043]; Airways[0.043]; USAir[0.042]; deal[0.041]; deal[0.041]; deal[0.041]; signed[0.041]; agreement[0.041]; added[0.041]; divestiture[0.040]; alliances[0.040]; value[0.040]; additional[0.040]; currently[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Office of Fair Trade\u001b[39m ==> ENTITY: \u001b[32mOffice of Fair Trading\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.044]; British[0.044]; British[0.044]; services[0.043]; London[0.043]; London[0.043]; Office[0.043]; introduction[0.042]; competition[0.042]; added[0.042]; Trading[0.042]; market[0.041]; basis[0.041]; called[0.041]; government[0.041]; additional[0.041]; clause[0.041]; said[0.041]; said[0.041]; routes[0.041]; equivalent[0.041]; Dallas[0.040]; Fair[0.040]; Jan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = London Heathrow\u001b[39m ==> ENTITY: \u001b[32mLondon Heathrow Airport\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.048]; routes[0.047]; airport[0.046]; British[0.044]; British[0.044]; British[0.044]; British[0.044]; route[0.044]; passengers[0.042]; passengers[0.042]; busiest[0.042]; airline[0.041]; Ba[0.040]; Airways[0.039]; Airways[0.039]; Airways[0.039]; Airways[0.039]; additional[0.039]; Europe[0.039]; USAir[0.039]; American[0.039]; American[0.039]; American[0.039]; authorities[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.053]; Britain[0.053]; British[0.049]; London[0.043]; Commission[0.042]; American[0.042]; Office[0.041]; Atlantic[0.041]; Atlantic[0.041]; Atlantic[0.041]; month[0.041]; States[0.041]; number[0.040]; proposed[0.040]; proposed[0.040]; Ian[0.040]; signed[0.039]; Uk[0.039]; route[0.039]; latest[0.039]; Trade[0.039]; supported[0.039]; blamed[0.039]; Secretary[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = British Airways\u001b[39m ==> ENTITY: \u001b[32mBritish Airways\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairline[0.051]; British[0.047]; British[0.047]; British[0.047]; Airways[0.047]; Airways[0.047]; Airways[0.047]; London[0.046]; London[0.046]; services[0.046]; similar[0.045]; USAir[0.045]; does[0.045]; routes[0.044]; American[0.044]; American[0.044]; drop[0.044]; Office[0.044]; Office[0.044]; unprecedented[0.044]; terms[0.043]; deal[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = British Airways Plc\u001b[39m ==> ENTITY: \u001b[32mBritish Airways\u001b[39m\t\nSCORES: global= 0.317:0.317[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirlines[0.050]; Britain[0.047]; Britain[0.047]; Britain[0.047]; Britain[0.047]; airlines[0.047]; British[0.046]; London[0.045]; carriers[0.044]; routes[0.043]; American[0.043]; American[0.043]; month[0.043]; United[0.043]; deal[0.042]; number[0.042]; Mergers[0.041]; Office[0.040]; air[0.040]; Ian[0.040]; added[0.040]; alliance[0.040]; alliance[0.040]; \t\n\r [============================================>.] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 178ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4450/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1392testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Madrid ==> ENTITY: \u001b[32mMadrid\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.170:0.170[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMadrid[0.046]; Madrid[0.046]; Madrid[0.046]; Soccer[0.044]; games[0.043]; Spanish[0.043]; Spanish[0.043]; match[0.042]; Coruna[0.042]; played[0.042]; Bilbao[0.042]; won[0.042]; Barcelona[0.041]; Barcelona[0.041]; Deportivo[0.041]; Valladolid[0.040]; Atletico[0.040]; Sociedad[0.039]; Real[0.039]; Real[0.039]; Real[0.039]; Real[0.039]; lost[0.039]; Athletic[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sevilla ==> ENTITY: \u001b[32mSevilla FC\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.559:-0.559[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.062]; Compostela[0.058]; Zaragoza[0.058]; Valladolid[0.058]; Valencia[0.058]; Gijon[0.058]; Espanyol[0.058]; Sociedad[0.057]; Extremadura[0.056]; Celta[0.055]; Santander[0.054]; Oviedo[0.054]; Tenerife[0.054]; Hercules[0.053]; Sporting[0.053]; Rayo[0.053]; Vigo[0.052]; Racing[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rayo Vallecano\u001b[39m ==> ENTITY: \u001b[32mRayo Vallecano\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; division[0.043]; Valladolid[0.043]; Valencia[0.043]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Sevilla[0.042]; Deportivo[0.042]; Zaragoza[0.042]; match[0.042]; Standings[0.042]; Spanish[0.042]; Gijon[0.041]; Tenerife[0.041]; Sociedad[0.041]; won[0.041]; Atletico[0.041]; Oviedo[0.041]; Espanyol[0.041]; goals[0.041]; Coruna[0.040]; Betis[0.040]; Athletic[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Atletico Madrid\u001b[39m ==> ENTITY: \u001b[32mAtlético Madrid\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.228:0.228[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.046]; Soccer[0.044]; Valencia[0.043]; Madrid[0.043]; Madrid[0.043]; Madrid[0.043]; games[0.043]; Coruna[0.043]; Deportivo[0.043]; Barcelona[0.042]; Barcelona[0.042]; match[0.041]; Sociedad[0.041]; Espanyol[0.041]; Sporting[0.041]; Spanish[0.041]; Spanish[0.041]; Bilbao[0.040]; Betis[0.040]; Athletic[0.040]; Gijon[0.040]; won[0.040]; goals[0.040]; division[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Madrid\u001b[39m ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.046]; Vallecano[0.045]; Madrid[0.043]; Madrid[0.043]; Madrid[0.043]; games[0.043]; Deportivo[0.042]; Valencia[0.042]; Barcelona[0.042]; Barcelona[0.042]; match[0.042]; Atletico[0.042]; Coruna[0.041]; Division[0.041]; Real[0.041]; Real[0.041]; Real[0.041]; Athletic[0.040]; played[0.040]; won[0.040]; points[0.040]; Spanish[0.040]; Spanish[0.040]; division[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oviedo ==> ENTITY: \u001b[32mReal Oviedo\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.194:0.194[0]; log p(e|m)= -2.254:-2.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.048]; Valencia[0.045]; Madrid[0.044]; Bilbao[0.043]; Atletico[0.043]; Zaragoza[0.043]; Espanyol[0.042]; Tenerife[0.042]; Sevilla[0.042]; Sociedad[0.042]; Extremadura[0.041]; Celta[0.041]; Sporting[0.041]; Rayo[0.041]; Gijon[0.041]; Betis[0.041]; Valladolid[0.041]; Compostela[0.040]; Athletic[0.040]; Real[0.040]; Real[0.040]; Santander[0.040]; Vigo[0.040]; Hercules[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Celta Vigo\u001b[39m ==> ENTITY: \u001b[32mCelta de Vigo\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.213:0.213[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.046]; Atletico[0.044]; Valencia[0.044]; Deportivo[0.044]; Sociedad[0.043]; Zaragoza[0.043]; Espanyol[0.043]; Coruna[0.043]; Bilbao[0.043]; Sevilla[0.042]; Madrid[0.042]; Madrid[0.042]; Barcelona[0.041]; Tenerife[0.041]; Betis[0.041]; Gijon[0.041]; Sporting[0.040]; Athletic[0.040]; Real[0.040]; Real[0.040]; Real[0.040]; Valladolid[0.040]; Rayo[0.039]; Santander[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Sociedad\u001b[39m ==> ENTITY: \u001b[32mReal Sociedad\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.046]; Soccer[0.044]; Deportivo[0.043]; games[0.042]; Valencia[0.042]; Spanish[0.042]; Spanish[0.042]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Atletico[0.042]; Coruna[0.042]; division[0.041]; Zaragoza[0.041]; Sevilla[0.041]; played[0.041]; Real[0.041]; Real[0.041]; Real[0.041]; Barcelona[0.041]; Barcelona[0.041]; Sporting[0.040]; lost[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barcelona ==> ENTITY: \u001b[32mFC Barcelona\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.195:0.195[0]; log p(e|m)= -1.709:-1.709[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.045]; Vallecano[0.044]; Valencia[0.044]; Madrid[0.043]; Madrid[0.043]; Madrid[0.043]; Madrid[0.043]; games[0.043]; Bilbao[0.043]; won[0.042]; match[0.041]; Coruna[0.041]; Spanish[0.041]; Spanish[0.041]; Barcelona[0.041]; Deportivo[0.041]; Atletico[0.041]; Racing[0.041]; Rayo[0.041]; played[0.040]; Division[0.040]; division[0.040]; Athletic[0.040]; points[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Compostela ==> ENTITY: \u001b[32mSD Compostela\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.118:0.118[0]; log p(e|m)= -0.875:-0.875[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.069]; Tenerife[0.066]; Extremadura[0.065]; Zaragoza[0.065]; Valencia[0.064]; Rayo[0.064]; Santander[0.064]; Oviedo[0.064]; Valladolid[0.063]; Gijon[0.061]; Vigo[0.061]; Hercules[0.061]; Sevilla[0.060]; Espanyol[0.059]; Celta[0.058]; Sporting[0.058]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spanish ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.138:0.138[0]; log p(e|m)= -1.241:-1.241[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpanish[0.051]; games[0.044]; Soccer[0.044]; match[0.044]; won[0.043]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Division[0.042]; played[0.041]; Standings[0.041]; Standings[0.041]; Barcelona[0.041]; Barcelona[0.041]; division[0.040]; points[0.040]; goals[0.040]; Santander[0.040]; Atletico[0.040]; Real[0.040]; Real[0.040]; Real[0.040]; Real[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSporting\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSporting Clube de Portugal\u001b[39m <---> \u001b[32mSporting de Gijón\u001b[39m\t\nSCORES: global= 0.275:0.263[\u001b[31m0.012\u001b[39m]; local()= 0.152:0.198[\u001b[32m0.047\u001b[39m]; log p(e|m)= -0.121:-1.858[\u001b[31m1.737\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.049]; Valencia[0.047]; Atletico[0.046]; Athletic[0.046]; Sociedad[0.045]; Zaragoza[0.045]; Gijon[0.045]; Espanyol[0.044]; Bilbao[0.044]; Madrid[0.044]; Sevilla[0.044]; Tenerife[0.043]; Compostela[0.043]; Valladolid[0.043]; Celta[0.043]; Hercules[0.042]; Extremadura[0.042]; Vigo[0.042]; Real[0.042]; Oviedo[0.041]; Rayo[0.041]; Santander[0.041]; Racing[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Racing Santander\u001b[39m ==> ENTITY: \u001b[32mRacing de Santander\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.238:0.238[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.052]; Valencia[0.049]; Deportivo[0.049]; Zaragoza[0.049]; Sociedad[0.048]; Madrid[0.048]; Madrid[0.048]; Madrid[0.048]; Madrid[0.048]; games[0.048]; Coruna[0.047]; Bilbao[0.047]; Sevilla[0.047]; Atletico[0.047]; Espanyol[0.047]; Athletic[0.047]; Betis[0.046]; Barcelona[0.046]; Barcelona[0.046]; Spanish[0.046]; goals[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Betis\u001b[39m ==> ENTITY: \u001b[32mReal Betis\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.244:0.244[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.047]; Valencia[0.044]; Deportivo[0.043]; Atletico[0.043]; Sociedad[0.043]; games[0.042]; Espanyol[0.042]; Bilbao[0.042]; Soccer[0.042]; Valladolid[0.042]; division[0.041]; Coruna[0.041]; Celta[0.041]; Madrid[0.041]; Madrid[0.041]; Madrid[0.041]; Madrid[0.041]; match[0.041]; Barcelona[0.040]; Barcelona[0.040]; Rayo[0.040]; Spanish[0.040]; Spanish[0.040]; won[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tenerife ==> ENTITY: \u001b[32mCD Tenerife\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.188:0.188[0]; log p(e|m)= -1.565:-1.565[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaragoza[0.046]; Vallecano[0.045]; Santander[0.045]; Extremadura[0.044]; Gijon[0.044]; Madrid[0.043]; Sociedad[0.043]; Valladolid[0.043]; Deportivo[0.043]; Bilbao[0.042]; Valencia[0.042]; Atletico[0.042]; Compostela[0.041]; Espanyol[0.041]; Celta[0.041]; Vigo[0.041]; Betis[0.041]; Sevilla[0.041]; Oviedo[0.041]; Rayo[0.040]; Coruna[0.040]; Athletic[0.038]; Barcelona[0.038]; Sporting[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hercules ==> ENTITY: \u001b[32mHércules CF\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.220:0.220[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.071]; Zaragoza[0.068]; Valencia[0.068]; Sevilla[0.066]; Espanyol[0.066]; Tenerife[0.064]; Gijon[0.063]; Rayo[0.063]; Celta[0.062]; Santander[0.062]; Extremadura[0.062]; Sporting[0.060]; Compostela[0.058]; Oviedo[0.056]; Vigo[0.056]; Racing[0.054]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Valencia ==> ENTITY: \u001b[32mValencia CF\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.207:0.207[0]; log p(e|m)= -1.604:-1.604[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaragoza[0.047]; Vallecano[0.046]; Madrid[0.044]; Madrid[0.044]; Bilbao[0.044]; Gijon[0.043]; Santander[0.042]; Atletico[0.042]; Valladolid[0.042]; games[0.042]; Oviedo[0.041]; Extremadura[0.041]; Deportivo[0.041]; Coruna[0.041]; Barcelona[0.041]; Sevilla[0.041]; Compostela[0.041]; Betis[0.040]; Tenerife[0.040]; Sociedad[0.040]; Espanyol[0.040]; Real[0.039]; Real[0.039]; Real[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Athletic Bilbao\u001b[39m ==> ENTITY: \u001b[32mAthletic Bilbao\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.230:0.230[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.045]; Soccer[0.044]; Valencia[0.043]; Deportivo[0.043]; games[0.043]; Sociedad[0.042]; Atletico[0.042]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Coruna[0.042]; Spanish[0.042]; Spanish[0.042]; match[0.041]; Barcelona[0.041]; Barcelona[0.041]; Espanyol[0.041]; Sporting[0.040]; division[0.040]; Valladolid[0.040]; Standings[0.040]; Standings[0.040]; Betis[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spanish ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.134:0.134[0]; log p(e|m)= -1.241:-1.241[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpanish[0.051]; games[0.044]; Soccer[0.044]; match[0.044]; won[0.043]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Madrid[0.042]; Division[0.042]; played[0.041]; Standings[0.041]; Standings[0.041]; Barcelona[0.041]; Barcelona[0.041]; division[0.040]; points[0.040]; goals[0.040]; Atletico[0.040]; Real[0.040]; Real[0.040]; Real[0.040]; Real[0.040]; Deportivo[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Madrid\u001b[39m ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.209:0.209[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.046]; Vallecano[0.045]; Madrid[0.043]; Madrid[0.043]; Madrid[0.043]; games[0.043]; Deportivo[0.042]; Barcelona[0.042]; Barcelona[0.042]; match[0.042]; Atletico[0.042]; Coruna[0.042]; Division[0.041]; Real[0.041]; Real[0.041]; Real[0.041]; Athletic[0.040]; played[0.040]; won[0.040]; points[0.040]; Spanish[0.040]; Spanish[0.040]; division[0.040]; Betis[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zaragoza ==> ENTITY: \u001b[32mReal Zaragoza\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.157:0.157[0]; log p(e|m)= -1.609:-1.609[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nValladolid[0.060]; Sevilla[0.058]; Valencia[0.058]; Celta[0.058]; Espanyol[0.057]; Sociedad[0.057]; Gijon[0.056]; Vallecano[0.055]; Santander[0.053]; Extremadura[0.053]; Compostela[0.052]; Rayo[0.052]; Vigo[0.051]; Tenerife[0.051]; Oviedo[0.050]; Hercules[0.046]; Sporting[0.046]; Real[0.045]; Racing[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Espanyol\u001b[39m ==> ENTITY: \u001b[32mRCD Espanyol\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.046]; Sociedad[0.044]; Valencia[0.044]; Zaragoza[0.044]; Madrid[0.043]; Atletico[0.043]; Deportivo[0.043]; Coruna[0.043]; Celta[0.042]; Extremadura[0.042]; Sevilla[0.042]; Valladolid[0.042]; Sporting[0.042]; Gijon[0.041]; Vigo[0.041]; Athletic[0.041]; Betis[0.041]; Bilbao[0.040]; Tenerife[0.040]; Real[0.040]; Real[0.040]; Santander[0.039]; Rayo[0.039]; Oviedo[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Valladolid ==> ENTITY: \u001b[32mReal Valladolid\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.204:0.204[0]; log p(e|m)= -1.650:-1.650[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.044]; Madrid[0.043]; Madrid[0.043]; Madrid[0.043]; Madrid[0.043]; Soccer[0.043]; Spanish[0.042]; Spanish[0.042]; Deportivo[0.042]; Bilbao[0.042]; Sociedad[0.042]; games[0.042]; Zaragoza[0.042]; Atletico[0.041]; Compostela[0.041]; Espanyol[0.041]; Barcelona[0.041]; Barcelona[0.041]; Valencia[0.041]; Santander[0.041]; Sevilla[0.040]; Rayo[0.040]; Coruna[0.040]; Betis[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Extremadura ==> ENTITY: \u001b[32mCF Extremadura\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.167:0.167[0]; log p(e|m)= -1.211:-1.211[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVallecano[0.078]; Tenerife[0.074]; Rayo[0.074]; Valencia[0.073]; Zaragoza[0.073]; Espanyol[0.072]; Hercules[0.071]; Oviedo[0.071]; Compostela[0.071]; Celta[0.070]; Gijon[0.069]; Sevilla[0.069]; Sporting[0.068]; Vigo[0.068]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barcelona ==> ENTITY: \u001b[32mFC Barcelona\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.190:0.190[0]; log p(e|m)= -1.709:-1.709[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.045]; Vallecano[0.044]; Madrid[0.043]; Madrid[0.043]; Madrid[0.043]; Madrid[0.043]; games[0.043]; Bilbao[0.043]; won[0.042]; match[0.042]; Coruna[0.041]; Spanish[0.041]; Spanish[0.041]; Barcelona[0.041]; Deportivo[0.041]; Atletico[0.041]; Racing[0.041]; Rayo[0.041]; played[0.040]; Division[0.040]; division[0.040]; Athletic[0.040]; points[0.040]; Santander[0.040]; \t\n\r [============================================>.] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 51ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4475/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1221testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.045:0.045[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.059]; London[0.059]; recorded[0.053]; shares[0.052]; day[0.050]; day[0.050]; Thursday[0.049]; Friday[0.049]; previous[0.048]; represents[0.048]; closed[0.048]; high[0.046]; low[0.046]; close[0.045]; Indonesia[0.045]; Jakarta[0.044]; Telekomunikasi[0.043]; newsroom[0.042]; Telkom[0.042]; Telkom[0.042]; Ads[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = PT Telekomunikasi Indonesia\u001b[39m ==> ENTITY: \u001b[32mTelkom Indonesia\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.090:0.090[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nshares[0.053]; Telkom[0.051]; Telkom[0.051]; ordinary[0.051]; Jakarta[0.050]; closed[0.050]; previous[0.049]; Thursday[0.048]; high[0.048]; day[0.047]; day[0.047]; Friday[0.047]; represents[0.047]; close[0.047]; low[0.047]; newsroom[0.046]; recorded[0.046]; London[0.044]; London[0.044]; London[0.044]; Ads[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jakarta\u001b[39m ==> ENTITY: \u001b[32mJakarta\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.077:0.077[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.052]; Thursday[0.047]; London[0.046]; London[0.046]; London[0.046]; Friday[0.045]; Telekomunikasi[0.045]; low[0.045]; high[0.045]; closed[0.045]; recorded[0.044]; represents[0.043]; day[0.043]; day[0.043]; previous[0.042]; Telkom[0.042]; Telkom[0.042]; close[0.041]; newsroom[0.041]; Pt[0.041]; shares[0.040]; Ads[0.039]; ordinary[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.045:0.045[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.059]; London[0.059]; recorded[0.053]; shares[0.052]; day[0.050]; day[0.050]; Thursday[0.049]; Friday[0.049]; previous[0.048]; represents[0.048]; closed[0.048]; high[0.046]; low[0.046]; close[0.045]; Indonesia[0.045]; Jakarta[0.044]; Telekomunikasi[0.043]; newsroom[0.042]; Telkom[0.042]; Telkom[0.042]; Ads[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.045:0.045[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.059]; London[0.059]; recorded[0.053]; shares[0.052]; day[0.050]; day[0.050]; Thursday[0.049]; Friday[0.049]; previous[0.048]; represents[0.048]; closed[0.048]; high[0.046]; low[0.046]; close[0.045]; Indonesia[0.045]; Jakarta[0.044]; Telekomunikasi[0.043]; newsroom[0.042]; Telkom[0.042]; Telkom[0.042]; Ads[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mTelkom\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mTelkom (South Africa)\u001b[39m <---> \u001b[32mTelkom Indonesia\u001b[39m\t\nSCORES: global= 0.287:0.279[\u001b[31m0.008\u001b[39m]; local()= 0.099:0.092[\u001b[31m0.007\u001b[39m]; log p(e|m)= 0.000:-0.673[\u001b[31m0.673\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTelkom[0.051]; Telekomunikasi[0.049]; Indonesia[0.047]; shares[0.046]; close[0.045]; ordinary[0.044]; Jakarta[0.044]; Pt[0.044]; closed[0.044]; previous[0.043]; Thursday[0.043]; day[0.043]; day[0.043]; London[0.042]; London[0.042]; London[0.042]; Friday[0.042]; high[0.042]; low[0.041]; represents[0.041]; recorded[0.040]; newsroom[0.040]; Ads[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mTelkom\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mTelkom (South Africa)\u001b[39m <---> \u001b[32mTelkom Indonesia\u001b[39m\t\nSCORES: global= 0.287:0.279[\u001b[31m0.008\u001b[39m]; local()= 0.099:0.092[\u001b[31m0.007\u001b[39m]; log p(e|m)= 0.000:-0.673[\u001b[31m0.673\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTelkom[0.051]; Telekomunikasi[0.049]; Indonesia[0.047]; shares[0.046]; close[0.045]; ordinary[0.044]; Jakarta[0.044]; Pt[0.044]; closed[0.044]; previous[0.043]; Thursday[0.043]; day[0.043]; day[0.043]; London[0.042]; London[0.042]; London[0.042]; Friday[0.042]; high[0.042]; low[0.041]; represents[0.041]; recorded[0.040]; newsroom[0.040]; Ads[0.040]; \t\n\r [============================================>.] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4482/4485 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1239testb ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mOakland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mOakland, California\u001b[39m <---> \u001b[32mOakland, New Jersey\u001b[39m\t\nSCORES: global= 0.247:0.226[\u001b[31m0.022\u001b[39m]; local()= 0.066:0.060[\u001b[31m0.006\u001b[39m]; log p(e|m)= -0.499:-3.689[\u001b[31m3.190\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFriday[0.044]; July[0.043]; July[0.043]; July[0.043]; services[0.043]; company[0.042]; provide[0.042]; operating[0.042]; said[0.042]; said[0.042]; Cooke[0.042]; Cooke[0.042]; retire[0.042]; retire[0.042]; serve[0.042]; continue[0.041]; Curts[0.041]; Russ[0.041]; Russ[0.041]; maker[0.040]; consulting[0.040]; president[0.040]; president[0.040]; director[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = N.J.\u001b[39m ==> ENTITY: \u001b[32mNew Jersey\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.061:0.061[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJuly[0.047]; July[0.047]; July[0.047]; serve[0.046]; operating[0.045]; services[0.044]; said[0.044]; said[0.044]; director[0.044]; chief[0.044]; Friday[0.043]; Oakland[0.043]; Curts[0.043]; officer[0.043]; retire[0.043]; retire[0.043]; president[0.042]; president[0.042]; company[0.042]; consulting[0.042]; Russ[0.042]; Russ[0.042]; effective[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Russ Berrie\u001b[39m ==> ENTITY: \u001b[32mKid Brands\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.090:0.090[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncompany[0.050]; services[0.047]; gift[0.045]; toy[0.045]; Russ[0.045]; maker[0.045]; operating[0.044]; consulting[0.043]; Berrie[0.043]; said[0.043]; said[0.043]; provide[0.043]; July[0.043]; July[0.043]; July[0.043]; director[0.043]; Friday[0.043]; chief[0.043]; Curts[0.042]; officer[0.042]; retire[0.042]; retire[0.042]; serve[0.041]; \t\n\r [=============================================>] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b Tot: 22s618ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4485/4485 \n\r==> time to test 1 sample = 5.0430747045987ms\t\n==> \u001b[31maida-B\u001b[39m aida-B ; EPOCH = 307: Micro recall = 92.08% ; Micro F1 = \u001b[31m92.08%\u001b[39m\t\n ===> entity frequency stats :\t\nfreq = 11-20 : num = 4 ; correctly classified = 4 ; perc = 100.00\t\nfreq = 2-5 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\nfreq = 50+ : num = 4345 ; correctly classified = 4078 ; perc = 93.86\t\nfreq = 1 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\nfreq = 21-50 : num = 50 ; correctly classified = 44 ; perc = 88.00\t\nfreq = 0 : num = 5 ; correctly classified = 4 ; perc = 80.00\t\nfreq = 6-10 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\n\t\n ===> entity p(e|m) stats :\t\np(e|m) = <=0.001 : num = 1 ; correctly classified = 0 ; perc = 0.00\t\np(e|m) = 0.01-0.03 : num = 249 ; correctly classified = 224 ; perc = 89.96\t\np(e|m) = 0.003-0.01 : num = 36 ; correctly classified = 33 ; perc = 91.67\t\np(e|m) = 0.3+ : num = 3431 ; correctly classified = 3309 ; perc = 96.44\t\np(e|m) = 0.1-0.3 : num = 381 ; correctly classified = 324 ; perc = 85.04\t\np(e|m) = 0.03-0.1 : num = 306 ; correctly classified = 240 ; perc = 78.43\t\np(e|m) = 0.001-0.003 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\n num_mentions_w/o_gold_ent_in_candidates = 81 total num mentions in dataset = 4485\t\n percentage_mentions_w/o_gold_ent_in_candidates = 1.81%; percentage_mentions_solved : both_pem_ours = 69.88%; only_pem_not_ours = 2.10%; only_ours_not_pem = 22.21%; not_ours_not_pem = 4.01%\t\n\n===> aida-A; num mentions = 4791\t\n [..............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 0ms | Step: 0ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 0/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1154testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brussels ==> ENTITY: \u001b[32mBrussels\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.025:-0.025[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgian[0.047]; Belgian[0.047]; Saturday[0.046]; Friday[0.045]; town[0.044]; Liege[0.044]; Liege[0.044]; eastern[0.044]; Thursday[0.044]; Thursday[0.044]; official[0.044]; link[0.043]; trip[0.043]; Severine[0.042]; Severine[0.042]; shopping[0.042]; said[0.042]; said[0.042]; police[0.042]; police[0.042]; police[0.042]; Earlier[0.041]; reported[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Liege\u001b[39m ==> ENTITY: \u001b[32mLiège\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.048]; town[0.044]; Belgian[0.044]; Belgian[0.044]; Liege[0.044]; eastern[0.044]; Brussels[0.043]; Severine[0.042]; Severine[0.042]; Friday[0.042]; Thursday[0.042]; Thursday[0.042]; Marc[0.041]; weeks[0.040]; Saturday[0.040]; said[0.040]; said[0.040]; trip[0.040]; reported[0.040]; home[0.040]; return[0.040]; declined[0.039]; declined[0.039]; official[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgium ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.242:-0.242[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLiege[0.045]; Liege[0.045]; sex[0.044]; weeks[0.043]; eastern[0.043]; return[0.043]; scandal[0.042]; Marc[0.042]; Thursday[0.042]; Friday[0.042]; Saturday[0.041]; trip[0.041]; case[0.041]; official[0.041]; past[0.041]; link[0.041]; Late[0.041]; Severine[0.041]; Severine[0.041]; missing[0.040]; home[0.040]; declined[0.040]; declined[0.040]; failing[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgian ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.124:-0.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgian[0.047]; eastern[0.044]; said[0.043]; said[0.043]; Liege[0.043]; Liege[0.043]; Brussels[0.042]; home[0.042]; Marc[0.042]; official[0.042]; Severine[0.041]; Severine[0.041]; trip[0.041]; town[0.041]; return[0.041]; say[0.041]; suspected[0.041]; Friday[0.041]; Thursday[0.041]; Thursday[0.041]; teenagers[0.040]; reported[0.040]; Saturday[0.040]; gone[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marc Dutroux\u001b[39m ==> ENTITY: \u001b[32mMarc Dutroux\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nscandal[0.045]; Belgium[0.044]; abuse[0.044]; case[0.044]; Belgian[0.043]; police[0.043]; police[0.043]; police[0.043]; murder[0.041]; sex[0.041]; suspected[0.041]; teenage[0.041]; girls[0.041]; girls[0.041]; girls[0.041]; girls[0.041]; Liege[0.041]; Liege[0.041]; rocked[0.040]; Severine[0.040]; Severine[0.040]; said[0.040]; said[0.040]; home[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Liege\u001b[39m ==> ENTITY: \u001b[32mLiège\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.048]; town[0.044]; Belgian[0.044]; Belgian[0.044]; Liege[0.044]; eastern[0.044]; Brussels[0.043]; Severine[0.042]; Severine[0.042]; Friday[0.042]; Thursday[0.042]; Thursday[0.042]; Marc[0.041]; weeks[0.040]; Saturday[0.040]; said[0.040]; said[0.040]; trip[0.040]; reported[0.040]; home[0.040]; return[0.040]; declined[0.039]; declined[0.039]; official[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgian ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.141:0.141[0]; log p(e|m)= -0.124:-0.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgian[0.048]; eastern[0.045]; said[0.045]; said[0.045]; Liege[0.045]; Liege[0.045]; Thursday[0.044]; Thursday[0.044]; Friday[0.044]; Brussels[0.044]; home[0.044]; official[0.043]; Severine[0.043]; Severine[0.043]; declined[0.043]; declined[0.043]; trip[0.043]; Saturday[0.043]; return[0.042]; town[0.042]; say[0.042]; teenagers[0.041]; reported[0.041]; \t\n\r [..............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 39s870ms | Step: 8ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 7/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1085testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mLatvia\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mLatvia\u001b[39m <---> \u001b[32mLatvia national football team\u001b[39m\t\nSCORES: global= 0.262:0.242[\u001b[31m0.020\u001b[39m]; local()= 0.106:0.121[\u001b[32m0.015\u001b[39m]; log p(e|m)= -0.351:-2.577[\u001b[31m2.226\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLatvia[0.052]; Sweden[0.050]; Sweden[0.050]; soccer[0.050]; championship[0.049]; European[0.049]; European[0.049]; qualifier[0.048]; Soccer[0.048]; Qualifier[0.047]; Riga[0.047]; Andersson[0.044]; Saturday[0.044]; beat[0.043]; minute[0.043]; Beat[0.042]; Persson[0.042]; Attendance[0.041]; halftime[0.041]; Daniel[0.041]; Scorers[0.041]; Joakim[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Riga ==> ENTITY: \u001b[32mRiga\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.069:0.069[0]; log p(e|m)= -0.043:-0.043[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.051]; Latvia[0.051]; Latvia[0.051]; championship[0.050]; soccer[0.049]; Sweden[0.049]; Sweden[0.049]; qualifier[0.047]; Qualifier[0.046]; European[0.046]; European[0.046]; minute[0.045]; Attendance[0.044]; Andersson[0.044]; Saturday[0.043]; beat[0.042]; Beat[0.042]; Daniel[0.042]; halftime[0.042]; Persson[0.041]; Joakim[0.039]; Scorers[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Joakim Persson\u001b[39m ==> ENTITY: \u001b[32mJoakim Persson\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.126:0.126[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.056]; Sweden[0.056]; Andersson[0.052]; Latvia[0.051]; Latvia[0.051]; soccer[0.050]; Soccer[0.049]; qualifier[0.048]; Qualifier[0.048]; European[0.048]; European[0.048]; minute[0.047]; halftime[0.047]; Scorers[0.046]; Attendance[0.046]; championship[0.045]; Daniel[0.045]; Riga[0.045]; Saturday[0.043]; beat[0.042]; Beat[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = European ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.034:0.034[0]; log p(e|m)= -0.256:-0.256[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuropean[0.053]; soccer[0.052]; Soccer[0.051]; qualifier[0.049]; Qualifier[0.048]; Sweden[0.048]; Sweden[0.048]; Latvia[0.046]; Latvia[0.046]; championship[0.046]; Riga[0.045]; beat[0.044]; Andersson[0.044]; Saturday[0.044]; minute[0.043]; Persson[0.043]; Attendance[0.043]; Daniel[0.042]; halftime[0.041]; Scorers[0.040]; Beat[0.040]; Joakim[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Daniel Andersson\u001b[39m ==> ENTITY: \u001b[32mDaniel Andersson (footballer born 1977)\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.120:0.120[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.055]; Sweden[0.055]; soccer[0.052]; European[0.050]; European[0.050]; qualifier[0.050]; Qualifier[0.048]; Latvia[0.048]; Latvia[0.048]; Soccer[0.048]; Joakim[0.048]; championship[0.047]; halftime[0.046]; Persson[0.046]; Scorers[0.045]; minute[0.045]; Attendance[0.045]; Riga[0.045]; Saturday[0.043]; beat[0.042]; Beat[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = European ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.034:0.034[0]; log p(e|m)= -0.256:-0.256[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuropean[0.053]; soccer[0.052]; Soccer[0.051]; qualifier[0.049]; Qualifier[0.048]; Sweden[0.048]; Sweden[0.048]; Latvia[0.046]; Latvia[0.046]; championship[0.046]; Riga[0.045]; beat[0.044]; Andersson[0.044]; Saturday[0.044]; minute[0.043]; Persson[0.043]; Attendance[0.043]; Daniel[0.042]; halftime[0.041]; Scorers[0.040]; Beat[0.040]; Joakim[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mLatvia\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mLatvia\u001b[39m <---> \u001b[32mLatvia national football team\u001b[39m\t\nSCORES: global= 0.262:0.242[\u001b[31m0.020\u001b[39m]; local()= 0.106:0.121[\u001b[32m0.015\u001b[39m]; log p(e|m)= -0.351:-2.577[\u001b[31m2.226\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLatvia[0.052]; Sweden[0.050]; Sweden[0.050]; soccer[0.050]; championship[0.049]; European[0.049]; European[0.049]; qualifier[0.048]; Soccer[0.048]; Qualifier[0.047]; Riga[0.047]; Andersson[0.044]; Saturday[0.044]; beat[0.043]; minute[0.043]; Beat[0.042]; Persson[0.042]; Attendance[0.041]; halftime[0.041]; Daniel[0.041]; Scorers[0.041]; Joakim[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSweden\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSweden\u001b[39m <---> \u001b[32mSweden national football team\u001b[39m\t\nSCORES: global= 0.255:0.232[\u001b[31m0.024\u001b[39m]; local()= 0.102:0.149[\u001b[32m0.046\u001b[39m]; log p(e|m)= -0.213:-2.957[\u001b[31m2.743\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.056]; soccer[0.051]; qualifier[0.050]; Soccer[0.049]; Qualifier[0.048]; European[0.048]; European[0.048]; Latvia[0.048]; Latvia[0.048]; Andersson[0.047]; championship[0.046]; minute[0.045]; beat[0.044]; Persson[0.043]; Riga[0.043]; Joakim[0.042]; Saturday[0.042]; Beat[0.042]; halftime[0.042]; Daniel[0.041]; Attendance[0.040]; Scorers[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSweden\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSweden\u001b[39m <---> \u001b[32mSweden national football team\u001b[39m\t\nSCORES: global= 0.255:0.232[\u001b[31m0.024\u001b[39m]; local()= 0.102:0.149[\u001b[32m0.046\u001b[39m]; log p(e|m)= -0.213:-2.957[\u001b[31m2.743\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.056]; soccer[0.051]; qualifier[0.050]; Soccer[0.049]; Qualifier[0.048]; European[0.048]; European[0.048]; Latvia[0.048]; Latvia[0.048]; Andersson[0.047]; championship[0.046]; minute[0.045]; beat[0.044]; Persson[0.043]; Riga[0.043]; Joakim[0.042]; Saturday[0.042]; Beat[0.042]; halftime[0.042]; Daniel[0.041]; Attendance[0.040]; Scorers[0.040]; \t\n\r [..............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 36s725ms | Step: 7ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 16/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1138testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = S. Korea\u001b[39m ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.208:0.208[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.049]; Korea[0.049]; Korea[0.049]; Korea[0.049]; Seoul[0.045]; Seoul[0.045]; Korean[0.044]; South[0.042]; South[0.042]; North[0.041]; North[0.041]; North[0.041]; Kang[0.040]; president[0.039]; cerebral[0.039]; Kim[0.039]; Kim[0.039]; Song[0.039]; southern[0.038]; demanded[0.038]; northern[0.038]; ho[0.038]; haemorrhage[0.038]; Young[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = North Korea ==> ENTITY: \u001b[32mNorth Korea\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.252:0.252[0]; log p(e|m)= -0.309:-0.309[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.047]; Korea[0.047]; Korea[0.047]; Korea[0.047]; communist[0.044]; Korean[0.044]; Korean[0.044]; Korean[0.044]; Seoul[0.042]; Seoul[0.042]; Yonhap[0.041]; North[0.040]; North[0.040]; Kim[0.040]; Kim[0.040]; Kim[0.040]; Kang[0.040]; war[0.040]; war[0.040]; war[0.040]; southern[0.039]; southern[0.039]; Song[0.038]; northern[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Korea ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.168:0.168[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSeoul[0.046]; Korean[0.045]; Korean[0.045]; Korean[0.045]; Korean[0.045]; South[0.044]; South[0.044]; South[0.044]; Yonhap[0.042]; North[0.042]; southern[0.042]; northern[0.042]; Kim[0.042]; consider[0.040]; regular[0.040]; International[0.040]; mo[0.040]; domestic[0.039]; government[0.039]; officials[0.039]; Li[0.039]; Government[0.039]; reunions[0.038]; unification[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = North ==> ENTITY: \u001b[32mNorth Korea\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.188:0.188[0]; log p(e|m)= -2.718:-2.718[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.047]; southern[0.046]; southern[0.046]; northern[0.045]; Korean[0.045]; Korean[0.045]; Korean[0.045]; Korean[0.045]; Seoul[0.043]; Yonhap[0.042]; communist[0.041]; war[0.041]; regular[0.040]; South[0.040]; South[0.040]; South[0.040]; South[0.040]; government[0.039]; include[0.039]; repatriation[0.039]; Li[0.038]; Li[0.038]; officials[0.038]; split[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = South Korean\u001b[39m ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.261:0.261[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.046]; Korea[0.046]; Korea[0.046]; Korea[0.046]; Korean[0.044]; Korean[0.044]; Korean[0.044]; Korean[0.044]; Seoul[0.042]; Seoul[0.042]; Seoul[0.042]; Yonhap[0.040]; South[0.039]; South[0.039]; South[0.039]; South[0.039]; North[0.039]; North[0.039]; North[0.039]; North[0.039]; Kim[0.039]; Kim[0.039]; Kim[0.039]; Kang[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.241:0.241[0]; log p(e|m)= -3.381:-3.381[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.046]; Korea[0.046]; Korea[0.046]; Korea[0.046]; Korea[0.046]; northern[0.043]; southern[0.043]; Seoul[0.043]; Seoul[0.043]; Korean[0.042]; Korean[0.042]; Korean[0.042]; South[0.040]; South[0.040]; North[0.040]; North[0.040]; North[0.040]; return[0.040]; return[0.040]; war[0.039]; war[0.039]; war[0.039]; Kang[0.038]; Young[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Seoul ==> ENTITY: \u001b[32mSeoul\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.177:0.177[0]; log p(e|m)= -0.038:-0.038[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.047]; Korean[0.044]; Korean[0.044]; Korean[0.044]; Korean[0.044]; North[0.043]; government[0.042]; South[0.042]; South[0.042]; South[0.042]; South[0.042]; Kim[0.042]; Kim[0.042]; Kang[0.041]; Yonhap[0.041]; reunions[0.041]; officials[0.041]; demand[0.040]; Government[0.040]; released[0.040]; International[0.040]; communist[0.039]; domestic[0.039]; unification[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mKorean\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSouth Korea\u001b[39m <---> \u001b[32mKorea\u001b[39m\t\nSCORES: global= 0.255:0.252[\u001b[31m0.003\u001b[39m]; local()= 0.227:0.227[\u001b[31m0.000\u001b[39m]; log p(e|m)= -2.703:-1.754[\u001b[32m0.949\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.047]; Korea[0.047]; Korean[0.045]; Korean[0.045]; Korean[0.045]; Korean[0.045]; Seoul[0.043]; Seoul[0.043]; southern[0.042]; southern[0.042]; Yonhap[0.041]; families[0.041]; northern[0.040]; northern[0.040]; communist[0.040]; Kim[0.040]; Kim[0.040]; Kim[0.040]; North[0.039]; North[0.039]; North[0.039]; war[0.039]; war[0.039]; war[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = N. Korea\u001b[39m ==> ENTITY: \u001b[32mNorth Korea\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.225:0.225[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.048]; Korea[0.048]; Korea[0.048]; Korea[0.048]; Korean[0.046]; Seoul[0.044]; Seoul[0.044]; North[0.041]; North[0.041]; North[0.041]; war[0.041]; war[0.041]; war[0.041]; southern[0.040]; northern[0.040]; Kim[0.040]; Kim[0.040]; demanded[0.039]; South[0.038]; South[0.038]; unification[0.038]; said[0.038]; said[0.038]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = North Korea ==> ENTITY: \u001b[32mNorth Korea\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.235:0.235[0]; log p(e|m)= -0.309:-0.309[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.047]; Korea[0.047]; Korea[0.047]; Korea[0.047]; Korean[0.045]; Korean[0.045]; Seoul[0.043]; Seoul[0.043]; North[0.041]; North[0.041]; Kim[0.040]; Kim[0.040]; Kang[0.040]; war[0.040]; war[0.040]; war[0.040]; southern[0.040]; Song[0.039]; northern[0.039]; unification[0.039]; unification[0.039]; South[0.039]; South[0.039]; South[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mKorean\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSouth Korea\u001b[39m <---> \u001b[32mKorea\u001b[39m\t\nSCORES: global= 0.254:0.251[\u001b[31m0.003\u001b[39m]; local()= 0.183:0.180[\u001b[31m0.003\u001b[39m]; log p(e|m)= -2.703:-1.754[\u001b[32m0.949\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.053]; Korean[0.051]; Korean[0.051]; Korean[0.051]; Seoul[0.049]; southern[0.047]; Yonhap[0.046]; families[0.046]; northern[0.046]; communist[0.045]; Kim[0.045]; Kim[0.045]; North[0.044]; war[0.044]; repatriation[0.043]; government[0.043]; officials[0.042]; requests[0.042]; domestic[0.042]; repatriated[0.042]; decades[0.042]; unification[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = International Red Cross\u001b[39m ==> ENTITY: \u001b[32mInternational Committee of the Red Cross\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngovernment[0.044]; released[0.043]; repatriated[0.043]; repatriation[0.043]; repatriation[0.043]; war[0.042]; cerebral[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; agency[0.042]; families[0.042]; officials[0.041]; Government[0.041]; reunions[0.041]; accepted[0.041]; southern[0.041]; southern[0.041]; Cross[0.041]; news[0.041]; haemorrhage[0.041]; War[0.040]; War[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Seoul ==> ENTITY: \u001b[32mSeoul\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.235:0.235[0]; log p(e|m)= -0.038:-0.038[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSeoul[0.046]; Korea[0.046]; Korea[0.046]; Korea[0.046]; Korea[0.046]; Korea[0.046]; Korean[0.042]; Korean[0.042]; Korean[0.042]; North[0.041]; North[0.041]; North[0.041]; South[0.041]; South[0.041]; South[0.041]; Kim[0.040]; Kim[0.040]; Kim[0.040]; Kang[0.040]; Saturday[0.039]; return[0.038]; return[0.038]; released[0.038]; communist[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.169:0.169[0]; log p(e|m)= -3.381:-3.381[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.049]; northern[0.046]; southern[0.046]; Seoul[0.045]; Korean[0.044]; Korean[0.044]; Korean[0.044]; South[0.043]; South[0.043]; North[0.042]; regular[0.042]; war[0.042]; Yonhap[0.041]; families[0.040]; decades[0.040]; War[0.039]; War[0.039]; officials[0.039]; domestic[0.039]; spending[0.039]; repatriation[0.038]; International[0.038]; government[0.038]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Seoul ==> ENTITY: \u001b[32mSeoul\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.218:0.218[0]; log p(e|m)= -0.038:-0.038[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSeoul[0.047]; Korea[0.046]; Korea[0.046]; Korea[0.046]; Korea[0.046]; Korea[0.046]; Korean[0.043]; North[0.042]; North[0.042]; North[0.042]; South[0.041]; South[0.041]; South[0.041]; Kim[0.041]; Kim[0.041]; Kang[0.041]; Saturday[0.040]; return[0.039]; return[0.039]; Young[0.039]; demanded[0.038]; unification[0.038]; repatriation[0.038]; Song[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yonhap\u001b[39m ==> ENTITY: \u001b[32mYonhap\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnews[0.045]; Korea[0.045]; Korea[0.045]; agency[0.044]; Korean[0.044]; Korean[0.044]; Korean[0.044]; Korean[0.044]; Korean[0.044]; North[0.042]; North[0.042]; press[0.042]; Seoul[0.041]; cerebral[0.041]; International[0.040]; domestic[0.040]; government[0.040]; ho[0.040]; family[0.039]; reunions[0.039]; president[0.039]; Kim[0.039]; Kim[0.039]; requests[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Korea ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.228:0.228[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.048]; Korea[0.048]; Korea[0.048]; Korea[0.048]; Seoul[0.044]; Seoul[0.044]; Korean[0.043]; Korean[0.043]; South[0.042]; South[0.042]; North[0.040]; North[0.040]; North[0.040]; Kim[0.040]; Kim[0.040]; Kim[0.040]; Kang[0.040]; president[0.039]; return[0.039]; return[0.039]; Song[0.039]; Young[0.038]; southern[0.038]; Saturday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = South Korean\u001b[39m ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorea[0.048]; Korean[0.046]; Korean[0.046]; Korean[0.046]; Korean[0.046]; Seoul[0.044]; families[0.042]; Yonhap[0.042]; South[0.041]; South[0.041]; South[0.041]; North[0.041]; Kim[0.041]; Kim[0.041]; Kang[0.040]; president[0.040]; repatriated[0.040]; government[0.039]; Song[0.039]; reunions[0.039]; officials[0.039]; domestic[0.039]; ho[0.039]; Young[0.038]; \t\n\r [..............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 27s406ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 34/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1017testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Middle East\u001b[39m ==> ENTITY: \u001b[32mMiddle East\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArab[0.049]; Israel[0.047]; Palestinian[0.044]; Jewish[0.044]; Jerusalem[0.043]; Jerusalem[0.043]; Palestinians[0.043]; war[0.043]; Israelis[0.042]; surrounding[0.042]; Bank[0.042]; East[0.042]; East[0.042]; West[0.041]; Middle[0.040]; community[0.040]; demolition[0.040]; Yasser[0.039]; settlements[0.039]; Netanyahu[0.039]; course[0.039]; action[0.039]; announced[0.039]; Arafat[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.116:0.116[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGerman[0.053]; German[0.053]; Jewish[0.048]; Israel[0.047]; Volker[0.047]; David[0.045]; Klaus[0.045]; government[0.045]; conservative[0.045]; West[0.045]; foreign[0.044]; September[0.044]; September[0.044]; month[0.044]; Bonn[0.044]; Bonn[0.044]; President[0.044]; Bank[0.044]; relations[0.044]; East[0.043]; leader[0.043]; cabinet[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.209:0.209[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.051]; Israel[0.051]; Israeli[0.046]; Israeli[0.046]; Israeli[0.046]; Jewish[0.046]; Jerusalem[0.043]; Arab[0.042]; David[0.042]; West[0.042]; Netanyahu[0.041]; Bank[0.041]; settlement[0.040]; Levy[0.039]; Levy[0.039]; Levy[0.039]; Levy[0.039]; German[0.039]; German[0.039]; German[0.039]; German[0.039]; Union[0.038]; establishment[0.038]; month[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Volker Ruehe\u001b[39m ==> ENTITY: \u001b[32mVolker Rühe\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nminister[0.044]; cabinet[0.044]; President[0.043]; ministry[0.042]; ministry[0.042]; ministry[0.042]; ministry[0.042]; ministry[0.042]; Kinkel[0.042]; Kinkel[0.042]; Defence[0.042]; Peres[0.042]; government[0.041]; September[0.041]; September[0.041]; Minister[0.041]; Minister[0.041]; Minister[0.041]; Bonn[0.041]; Bonn[0.041]; Prime[0.041]; visit[0.041]; visit[0.041]; visit[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Benjamin Netanyahu\u001b[39m ==> ENTITY: \u001b[32mBenjamin Netanyahu\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.227:0.227[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israeli[0.046]; Israeli[0.046]; Jewish[0.044]; Palestinian[0.044]; Israelis[0.044]; Jerusalem[0.044]; Jerusalem[0.044]; Jerusalem[0.044]; Arab[0.041]; Arab[0.041]; Palestinians[0.041]; Prime[0.041]; President[0.040]; Bank[0.040]; announced[0.039]; said[0.039]; said[0.039]; said[0.039]; said[0.039]; policy[0.039]; talks[0.039]; West[0.038]; Minister[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jewish ==> ENTITY: \u001b[32mJews\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.805:-0.805[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJerusalem[0.049]; Jerusalem[0.049]; occupied[0.046]; Israel[0.046]; Israeli[0.045]; demolition[0.045]; settlements[0.045]; Arab[0.044]; Israelis[0.044]; Palestinian[0.044]; community[0.043]; Palestinians[0.043]; West[0.043]; City[0.043]; settlement[0.043]; Bank[0.042]; surrounding[0.042]; war[0.042]; erected[0.042]; Martin[0.041]; German[0.041]; German[0.041]; Benjamin[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.233:0.233[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.049]; Israel[0.049]; Israeli[0.045]; Israeli[0.045]; Israeli[0.045]; Israeli[0.045]; Israeli[0.045]; Jewish[0.044]; Yitzhak[0.044]; Shimon[0.042]; Weizman[0.042]; Jerusalem[0.041]; Peres[0.041]; Mordechai[0.041]; David[0.041]; West[0.039]; Ezer[0.038]; Bank[0.038]; Levy[0.038]; Levy[0.038]; Levy[0.038]; Levy[0.038]; Netanyahu[0.038]; Netanyahu[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Palestinians ==> ENTITY: \u001b[32mState of Palestine\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.249:0.249[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.049]; Palestinian[0.047]; Israeli[0.047]; Israelis[0.046]; Arab[0.045]; Netanyahu[0.043]; Jerusalem[0.043]; Jerusalem[0.043]; peace[0.042]; peace[0.042]; West[0.042]; Jewish[0.041]; Bank[0.041]; policy[0.040]; Arafat[0.040]; settlements[0.040]; Yasser[0.040]; settlement[0.039]; region[0.039]; government[0.039]; government[0.039]; community[0.038]; occupied[0.038]; war[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = European Union\u001b[39m ==> ENTITY: \u001b[32mEuropean Union\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.006:-0.006[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npolicy[0.044]; government[0.043]; government[0.043]; renewed[0.043]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; Israel[0.042]; German[0.041]; German[0.041]; German[0.041]; Foreign[0.041]; Foreign[0.041]; foreign[0.041]; visit[0.041]; visit[0.041]; efforts[0.041]; concerned[0.041]; meeting[0.041]; peace[0.041]; peace[0.041]; peace[0.041]; spokesman[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shimon Peres\u001b[39m ==> ENTITY: \u001b[32mShimon Peres\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.221:0.221[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israeli[0.048]; Israeli[0.048]; Israeli[0.048]; Israeli[0.048]; Yitzhak[0.046]; Netanyahu[0.045]; Netanyahu[0.045]; Weizman[0.043]; Mordechai[0.041]; cabinet[0.041]; President[0.040]; leader[0.039]; Ezer[0.039]; Prime[0.039]; meeting[0.038]; minister[0.038]; Bonn[0.038]; visit[0.038]; visit[0.038]; talks[0.038]; said[0.038]; said[0.038]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Klaus Kinkel\u001b[39m ==> ENTITY: \u001b[32mKlaus Kinkel\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nForeign[0.044]; minister[0.043]; foreign[0.043]; government[0.043]; cabinet[0.042]; German[0.042]; German[0.042]; German[0.042]; Defence[0.042]; Kinkel[0.042]; Minister[0.042]; Minister[0.042]; Bonn[0.041]; Bonn[0.041]; Bonn[0.041]; ministry[0.041]; ministry[0.041]; ministry[0.041]; ministry[0.041]; ministry[0.041]; President[0.041]; European[0.040]; September[0.040]; September[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.081:0.081[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nregion[0.044]; European[0.044]; Jewish[0.043]; Jewish[0.043]; war[0.043]; occupied[0.043]; Israel[0.043]; Israel[0.043]; Benjamin[0.043]; policy[0.042]; government[0.042]; Martin[0.042]; foreign[0.041]; Bank[0.041]; Bank[0.041]; West[0.041]; West[0.041]; Bonn[0.040]; East[0.040]; East[0.040]; establishment[0.040]; settlements[0.040]; President[0.040]; supported[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = East Jerusalem\u001b[39m ==> ENTITY: \u001b[32mEast Jerusalem\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.216:0.216[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.047]; Jerusalem[0.045]; Palestinian[0.045]; Palestinians[0.045]; Israeli[0.044]; Israelis[0.044]; Arab[0.043]; Jewish[0.043]; West[0.043]; Bank[0.042]; Netanyahu[0.042]; occupied[0.041]; settlements[0.041]; settlement[0.040]; Arafat[0.040]; Yasser[0.040]; peace[0.040]; peace[0.040]; peace[0.040]; peace[0.040]; policy[0.039]; tantamount[0.039]; establishment[0.039]; City[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsraeli[0.049]; Israeli[0.049]; Israeli[0.049]; Israeli[0.049]; Israeli[0.049]; Peres[0.045]; Netanyahu[0.044]; Netanyahu[0.044]; foreign[0.043]; Yitzhak[0.042]; talks[0.042]; talks[0.042]; Foreign[0.042]; announced[0.041]; David[0.041]; Levy[0.041]; Levy[0.041]; Levy[0.041]; Levy[0.041]; discuss[0.041]; government[0.041]; Weizman[0.041]; month[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Middle East\u001b[39m ==> ENTITY: \u001b[32mMiddle East\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.142:0.142[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArab[0.049]; Israel[0.046]; Palestinian[0.044]; region[0.044]; Jewish[0.043]; Israeli[0.043]; Jerusalem[0.043]; Jerusalem[0.043]; Palestinians[0.042]; war[0.042]; Israelis[0.042]; surrounding[0.041]; Bank[0.041]; East[0.041]; East[0.041]; occupied[0.041]; West[0.040]; Middle[0.040]; community[0.040]; Foreign[0.040]; demolition[0.040]; Yasser[0.038]; settlements[0.038]; Netanyahu[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = David Levy ==> ENTITY: \u001b[32mDavid Levy (Israeli politician)\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.193:0.193[0]; log p(e|m)= -1.143:-1.143[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.045]; Israel[0.045]; Jewish[0.044]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; Netanyahu[0.042]; Levy[0.042]; Levy[0.042]; Levy[0.042]; Jerusalem[0.041]; cabinet[0.041]; Arab[0.041]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; visit[0.039]; visit[0.039]; visit[0.039]; Bank[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.188:0.188[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.055]; Israeli[0.050]; Israeli[0.050]; Israeli[0.050]; Yitzhak[0.050]; Shimon[0.047]; Weizman[0.047]; Peres[0.046]; Mordechai[0.046]; David[0.046]; Netanyahu[0.045]; Netanyahu[0.045]; Ezer[0.043]; Levy[0.043]; Levy[0.043]; Levy[0.043]; Levy[0.043]; German[0.042]; German[0.042]; German[0.042]; expected[0.042]; month[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bonn\u001b[39m ==> ENTITY: \u001b[32mBonn\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWest[0.049]; Bonn[0.045]; East[0.044]; East[0.044]; East[0.044]; German[0.043]; Friday[0.043]; week[0.043]; announced[0.042]; erected[0.042]; Benjamin[0.042]; government[0.042]; government[0.042]; expansion[0.041]; President[0.040]; Bank[0.040]; process[0.040]; Arab[0.040]; Jewish[0.040]; region[0.040]; war[0.039]; said[0.039]; said[0.039]; surrounding[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Netanyahu\u001b[39m ==> ENTITY: \u001b[32mBenjamin Netanyahu\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.240:0.240[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.055]; Israel[0.055]; Peres[0.050]; Israeli[0.050]; Israeli[0.050]; Israeli[0.050]; Israeli[0.050]; Israeli[0.050]; Netanyahu[0.049]; Jewish[0.048]; Jerusalem[0.048]; Yitzhak[0.048]; Weizman[0.046]; cabinet[0.046]; Shimon[0.045]; Prime[0.045]; David[0.044]; President[0.044]; minister[0.043]; leader[0.043]; Bank[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.198:0.198[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsraeli[0.050]; Israeli[0.050]; Israeli[0.050]; Israeli[0.050]; Israeli[0.050]; Arab[0.050]; Jewish[0.047]; Netanyahu[0.045]; Jerusalem[0.045]; David[0.044]; West[0.044]; foreign[0.044]; policy[0.043]; talks[0.043]; talks[0.043]; Bank[0.043]; Foreign[0.043]; Foreign[0.043]; European[0.043]; visit[0.042]; visit[0.042]; announced[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yasser Arafat\u001b[39m ==> ENTITY: \u001b[32mYasser Arafat\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.049]; Israeli[0.048]; Israeli[0.048]; Palestinian[0.046]; Israelis[0.045]; Palestinians[0.044]; Jerusalem[0.043]; Jerusalem[0.043]; Arab[0.043]; Arab[0.043]; Netanyahu[0.042]; Jewish[0.040]; settlements[0.040]; Bank[0.039]; President[0.039]; West[0.039]; government[0.039]; government[0.039]; said[0.038]; said[0.038]; said[0.038]; said[0.038]; Bonn[0.038]; Bonn[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.191:0.191[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsraeli[0.052]; Arab[0.051]; Jewish[0.050]; Palestinian[0.049]; Jerusalem[0.048]; Jerusalem[0.048]; Israelis[0.048]; Palestinians[0.047]; Netanyahu[0.047]; West[0.045]; policy[0.044]; Bank[0.044]; Foreign[0.044]; announced[0.043]; government[0.043]; government[0.043]; week[0.043]; Martin[0.043]; war[0.043]; Ministry[0.042]; establishment[0.042]; officials[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kinkel\u001b[39m ==> ENTITY: \u001b[32mKlaus Kinkel\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nForeign[0.046]; minister[0.045]; foreign[0.044]; government[0.044]; cabinet[0.044]; German[0.044]; German[0.044]; German[0.044]; Defence[0.044]; Kinkel[0.044]; Minister[0.043]; Minister[0.043]; Minister[0.043]; Bonn[0.043]; Bonn[0.043]; Bonn[0.043]; ministry[0.043]; ministry[0.043]; ministry[0.043]; ministry[0.043]; ministry[0.043]; President[0.043]; European[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jerusalem ==> ENTITY: \u001b[32mJerusalem\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.180:0.180[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.049]; Jewish[0.049]; Israeli[0.046]; Israeli[0.046]; Israeli[0.046]; Israeli[0.046]; Israeli[0.046]; occupied[0.046]; West[0.044]; Arab[0.044]; David[0.043]; Bank[0.043]; settlement[0.042]; settlement[0.042]; Ministry[0.042]; Netanyahu[0.041]; Netanyahu[0.041]; Martin[0.041]; month[0.041]; Union[0.041]; renewed[0.041]; discuss[0.040]; region[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yitzhak Mordechai\u001b[39m ==> ENTITY: \u001b[32mYitzhak Mordechai\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.046]; Weizman[0.044]; Israeli[0.044]; Israeli[0.044]; Israeli[0.044]; Israeli[0.044]; Netanyahu[0.043]; Netanyahu[0.043]; cabinet[0.042]; Peres[0.042]; Kinkel[0.042]; Kinkel[0.042]; Shimon[0.041]; Bonn[0.041]; Bonn[0.041]; relations[0.041]; minister[0.040]; Defence[0.040]; leader[0.040]; opposition[0.040]; month[0.040]; visit[0.040]; visit[0.040]; visit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.193:0.193[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.052]; Israeli[0.047]; Israeli[0.047]; Israeli[0.047]; Yitzhak[0.046]; Shimon[0.044]; Weizman[0.044]; Peres[0.043]; Mordechai[0.043]; Netanyahu[0.042]; Netanyahu[0.042]; Ezer[0.040]; Levy[0.040]; Levy[0.040]; Levy[0.040]; German[0.039]; German[0.039]; German[0.039]; expected[0.039]; month[0.039]; said[0.038]; said[0.038]; said[0.038]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bonn\u001b[39m ==> ENTITY: \u001b[32mBonn\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWest[0.048]; Bonn[0.044]; Bonn[0.044]; David[0.044]; East[0.043]; German[0.043]; German[0.043]; German[0.043]; European[0.041]; announced[0.041]; government[0.041]; Foreign[0.041]; month[0.041]; September[0.041]; September[0.041]; meeting[0.040]; visit[0.040]; visit[0.040]; visit[0.040]; Martin[0.040]; Defence[0.040]; cabinet[0.040]; Klaus[0.040]; Volker[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.240:0.240[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israel[0.050]; Israeli[0.045]; Jewish[0.045]; Jewish[0.045]; Israelis[0.044]; Palestinian[0.043]; Jerusalem[0.042]; Jerusalem[0.042]; Jerusalem[0.042]; settlements[0.042]; Palestinians[0.042]; Arab[0.041]; Arab[0.041]; West[0.041]; Netanyahu[0.040]; Bank[0.040]; settlement[0.039]; settlement[0.039]; occupied[0.038]; German[0.038]; German[0.038]; community[0.038]; Union[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jerusalem ==> ENTITY: \u001b[32mJerusalem\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.172:0.172[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJerusalem[0.048]; Israel[0.047]; Jewish[0.047]; Israeli[0.044]; occupied[0.044]; Israelis[0.044]; Palestinian[0.043]; West[0.043]; Arab[0.042]; Palestinians[0.042]; settlements[0.042]; Bank[0.041]; erected[0.041]; surrounding[0.041]; week[0.040]; settlement[0.040]; Friday[0.040]; Ministry[0.040]; Netanyahu[0.040]; Martin[0.039]; community[0.039]; officials[0.039]; Yasser[0.039]; region[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bonn\u001b[39m ==> ENTITY: \u001b[32mBonn\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWest[0.050]; Bonn[0.046]; David[0.046]; East[0.045]; German[0.044]; German[0.044]; German[0.044]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; ministry[0.043]; ministry[0.043]; ministry[0.043]; ministry[0.043]; ministry[0.043]; announced[0.043]; month[0.042]; September[0.042]; September[0.042]; government[0.042]; Klaus[0.041]; Volker[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bonn\u001b[39m ==> ENTITY: \u001b[32mBonn\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.107:0.107[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWest[0.049]; Bonn[0.045]; East[0.044]; East[0.044]; East[0.044]; German[0.043]; Friday[0.043]; week[0.043]; announced[0.042]; erected[0.042]; Benjamin[0.042]; government[0.042]; government[0.042]; expansion[0.041]; President[0.040]; Bank[0.040]; process[0.040]; Arab[0.040]; Jewish[0.040]; said[0.039]; said[0.039]; surrounding[0.039]; City[0.039]; Palestinian[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.103:0.103[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nregion[0.045]; Jewish[0.044]; war[0.044]; Israel[0.044]; Benjamin[0.044]; government[0.043]; government[0.043]; Bank[0.042]; community[0.042]; West[0.042]; East[0.041]; East[0.041]; East[0.041]; settlements[0.041]; said[0.041]; said[0.041]; Middle[0.041]; Middle[0.041]; expansion[0.041]; surrounding[0.040]; Bonn[0.040]; Bonn[0.040]; course[0.040]; officials[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jewish ==> ENTITY: \u001b[32mJews\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.805:-0.805[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.049]; Jerusalem[0.049]; Israeli[0.047]; Israeli[0.047]; Israeli[0.047]; Israeli[0.047]; Israeli[0.047]; David[0.046]; Arab[0.046]; efforts[0.045]; European[0.045]; Levy[0.045]; Levy[0.045]; Levy[0.045]; Levy[0.045]; occupied[0.044]; discuss[0.044]; West[0.044]; supported[0.044]; conservative[0.044]; month[0.044]; Martin[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Middle East\u001b[39m ==> ENTITY: \u001b[32mMiddle East\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArab[0.047]; Israel[0.045]; Israel[0.045]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; European[0.042]; Jewish[0.042]; West[0.042]; relations[0.042]; foreign[0.042]; government[0.040]; Jerusalem[0.040]; Bank[0.040]; President[0.040]; Defence[0.039]; Foreign[0.039]; David[0.039]; announced[0.039]; month[0.039]; peace[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bonn\u001b[39m ==> ENTITY: \u001b[32mBonn\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWest[0.047]; establishment[0.045]; Bonn[0.044]; David[0.044]; East[0.043]; German[0.042]; German[0.042]; German[0.042]; European[0.041]; said[0.041]; said[0.041]; said[0.041]; ministry[0.041]; ministry[0.041]; announced[0.041]; Benjamin[0.041]; government[0.041]; government[0.041]; Foreign[0.041]; Foreign[0.041]; Ministry[0.040]; month[0.040]; September[0.040]; meeting[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPalestinian\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPalestinian people\u001b[39m <---> \u001b[32mState of Palestine\u001b[39m\t\nSCORES: global= 0.238:0.238[\u001b[31m0.000\u001b[39m]; local()= 0.197:0.267[\u001b[32m0.070\u001b[39m]; log p(e|m)= -0.389:-2.847[\u001b[31m2.459\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.048]; Arab[0.045]; Arab[0.045]; Israeli[0.045]; Israelis[0.045]; Palestinians[0.044]; Jerusalem[0.043]; Jerusalem[0.043]; West[0.042]; Jewish[0.042]; Bank[0.041]; Netanyahu[0.041]; Yasser[0.040]; settlement[0.040]; region[0.040]; policy[0.040]; peace[0.040]; peace[0.040]; peace[0.040]; peace[0.040]; establishment[0.039]; Arafat[0.039]; government[0.039]; government[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ezer Weizman\u001b[39m ==> ENTITY: \u001b[32mEzer Weizman\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israeli[0.048]; Israeli[0.048]; Israeli[0.048]; Israeli[0.048]; Peres[0.047]; Yitzhak[0.046]; Netanyahu[0.044]; Netanyahu[0.044]; cabinet[0.044]; Shimon[0.043]; President[0.042]; minister[0.042]; Kinkel[0.041]; Kinkel[0.041]; Defence[0.041]; Mordechai[0.041]; Bonn[0.041]; leader[0.040]; government[0.040]; visit[0.040]; visit[0.040]; month[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Netanyahu\u001b[39m ==> ENTITY: \u001b[32mBenjamin Netanyahu\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.230:0.230[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Peres[0.046]; Israeli[0.046]; Israeli[0.046]; Israeli[0.046]; Israeli[0.046]; Netanyahu[0.045]; Yitzhak[0.043]; Weizman[0.042]; cabinet[0.042]; Shimon[0.041]; Prime[0.041]; President[0.040]; minister[0.040]; leader[0.040]; Levy[0.039]; Levy[0.039]; Mordechai[0.039]; meeting[0.039]; said[0.039]; said[0.039]; said[0.039]; visit[0.039]; visit[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.117:0.117[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGerman[0.048]; German[0.048]; Jewish[0.044]; Israel[0.043]; Israel[0.043]; European[0.043]; Volker[0.043]; David[0.041]; Klaus[0.041]; government[0.041]; conservative[0.041]; West[0.041]; foreign[0.041]; Union[0.041]; September[0.040]; September[0.040]; month[0.040]; Bonn[0.040]; Bonn[0.040]; Bonn[0.040]; President[0.040]; Bank[0.040]; supported[0.040]; relations[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIsraelis\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIsrael\u001b[39m <---> \u001b[32mIsraelis\u001b[39m\t\nSCORES: global= 0.269:0.263[\u001b[31m0.006\u001b[39m]; local()= 0.195:0.192[\u001b[31m0.003\u001b[39m]; log p(e|m)= -0.546:-0.009[\u001b[32m0.537\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.053]; Israeli[0.048]; Jewish[0.047]; Arab[0.045]; Palestinian[0.045]; Jerusalem[0.044]; Jerusalem[0.044]; Palestinians[0.044]; West[0.042]; Bank[0.041]; Netanyahu[0.040]; settlements[0.040]; community[0.040]; Middle[0.040]; Middle[0.040]; President[0.039]; government[0.039]; government[0.039]; officials[0.039]; policy[0.039]; announced[0.038]; week[0.038]; said[0.038]; said[0.038]; \t\n\r [..............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 21s555ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 74/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 961testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.040:0.040[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayer[0.047]; Thursday[0.046]; Tennis[0.045]; Washington[0.044]; Washington[0.044]; Larry[0.043]; night[0.043]; finalist[0.042]; reprising[0.042]; fell[0.041]; Jeff[0.041]; Alex[0.041]; Stadium[0.041]; connection[0.041]; lost[0.041]; lost[0.041]; comeback[0.040]; defeated[0.040]; Andre[0.040]; Fine[0.040]; short[0.040]; bad[0.040]; seed[0.039]; seeded[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rios\u001b[39m ==> ENTITY: \u001b[32mMarcelo Ríos\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntennis[0.048]; tennis[0.048]; tennis[0.048]; tennis[0.048]; match[0.045]; match[0.045]; college[0.044]; won[0.043]; lost[0.041]; tiebreaker[0.041]; having[0.040]; played[0.040]; played[0.040]; played[0.040]; day[0.040]; day[0.040]; day[0.040]; set[0.039]; ball[0.039]; play[0.039]; play[0.039]; play[0.039]; play[0.039]; play[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Washington\u001b[39m ==> ENTITY: \u001b[32mMaliVai Washington\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.050]; match[0.050]; collegiate[0.050]; Wimbledon[0.049]; game[0.047]; play[0.047]; ranked[0.047]; fifth[0.046]; set[0.046]; set[0.046]; sixth[0.045]; eighth[0.045]; win[0.044]; national[0.044]; champion[0.044]; victory[0.043]; serve[0.043]; serve[0.043]; backhand[0.043]; backhand[0.043]; couple[0.043]; forehand[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tarango\u001b[39m ==> ENTITY: \u001b[32mJeff Tarango\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.101:0.101[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntennis[0.049]; match[0.047]; Tarango[0.043]; Tarango[0.043]; played[0.043]; court[0.042]; won[0.042]; crowd[0.042]; play[0.042]; play[0.042]; play[0.042]; play[0.042]; ball[0.041]; really[0.040]; really[0.040]; lost[0.040]; blatantly[0.040]; American[0.040]; Rios[0.040]; Rios[0.040]; Rios[0.040]; fun[0.040]; fans[0.040]; time[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Leander Paes\u001b[39m ==> ENTITY: \u001b[32mLeander Paes\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWimbledon[0.046]; Wimbledon[0.046]; match[0.045]; player[0.043]; semifinals[0.043]; wildcard[0.042]; champion[0.042]; Martin[0.042]; sixth[0.042]; comeback[0.042]; comeback[0.042]; Agassi[0.042]; title[0.041]; ranked[0.041]; India[0.041]; games[0.041]; seeded[0.041]; win[0.041]; winning[0.040]; defeated[0.040]; seed[0.039]; hander[0.039]; climbed[0.039]; winner[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rios\u001b[39m ==> ENTITY: \u001b[32mMarcelo Ríos\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.046]; Open[0.046]; Agassi[0.045]; Agassi[0.045]; Wimbledon[0.045]; Wimbledon[0.045]; Wimbledon[0.045]; Wimbledon[0.045]; Alex[0.044]; Martin[0.044]; Andre[0.043]; champion[0.043]; Marcelo[0.043]; player[0.043]; Paes[0.043]; winning[0.042]; Leander[0.042]; Larry[0.042]; semifinals[0.042]; wildcard[0.042]; Chilean[0.042]; sixth[0.041]; games[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tarango\u001b[39m ==> ENTITY: \u001b[32mJeff Tarango\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntennis[0.047]; tennis[0.047]; tennis[0.047]; tennis[0.047]; match[0.045]; college[0.043]; player[0.042]; Tarango[0.041]; Tarango[0.041]; spectators[0.041]; played[0.041]; played[0.041]; played[0.041]; tiebreaker[0.040]; having[0.040]; won[0.040]; crowd[0.040]; ranked[0.040]; play[0.040]; play[0.040]; play[0.040]; set[0.039]; Chilean[0.039]; booed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marcelo Rios\u001b[39m ==> ENTITY: \u001b[32mMarcelo Ríos\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.046]; Open[0.045]; Agassi[0.045]; Wimbledon[0.044]; Wimbledon[0.044]; Wimbledon[0.044]; Alex[0.043]; Andre[0.043]; player[0.042]; Larry[0.041]; wildcard[0.041]; Chilean[0.041]; sixth[0.041]; lost[0.041]; lost[0.041]; Jeff[0.041]; finalist[0.040]; comeback[0.040]; upsets[0.040]; defeated[0.040]; day[0.040]; Rios[0.039]; seeded[0.039]; seeded[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Washington\u001b[39m ==> ENTITY: \u001b[32mMaliVai Washington\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.045]; collegiate[0.045]; Wimbledon[0.045]; game[0.043]; ranked[0.043]; games[0.043]; semifinals[0.042]; pro[0.042]; fifth[0.042]; sets[0.042]; set[0.041]; set[0.041]; set[0.041]; sixth[0.041]; sixth[0.041]; Agassi[0.041]; Leander[0.040]; win[0.040]; title[0.040]; national[0.040]; seeded[0.040]; champion[0.040]; champion[0.040]; Todd[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wimbledon ==> ENTITY: \u001b[32mThe Championships, Wimbledon\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.141:0.141[0]; log p(e|m)= -0.851:-0.851[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.045]; match[0.045]; match[0.045]; England[0.045]; played[0.044]; play[0.043]; victory[0.043]; set[0.042]; won[0.042]; hour[0.042]; player[0.042]; eighth[0.041]; ago[0.041]; serve[0.041]; American[0.040]; American[0.040]; spectators[0.040]; cheering[0.040]; netted[0.040]; fans[0.040]; ranked[0.040]; moments[0.040]; winner[0.040]; couple[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Todd Martin\u001b[39m ==> ENTITY: \u001b[32mTodd Martin\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.045]; Wimbledon[0.044]; Wimbledon[0.044]; Agassi[0.044]; win[0.043]; player[0.043]; champion[0.042]; ranked[0.042]; title[0.042]; Jeff[0.041]; defeated[0.041]; Leander[0.041]; set[0.041]; set[0.041]; sets[0.041]; comeback[0.041]; games[0.041]; semifinals[0.041]; winner[0.041]; sixth[0.040]; winning[0.040]; seeded[0.040]; seed[0.040]; Paes[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jeff Tarango\u001b[39m ==> ENTITY: \u001b[32mJeff Tarango\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsixth[0.044]; Open[0.044]; Agassi[0.044]; Agassi[0.044]; Martin[0.043]; champion[0.043]; Wimbledon[0.042]; Wimbledon[0.042]; Wimbledon[0.042]; Wimbledon[0.042]; games[0.042]; player[0.042]; semifinals[0.042]; Alex[0.041]; Larry[0.041]; Andre[0.041]; pro[0.041]; upsets[0.041]; court[0.040]; Todd[0.040]; sets[0.040]; finalist[0.040]; set[0.040]; Chilean[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rios\u001b[39m ==> ENTITY: \u001b[32mMarcelo Ríos\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntennis[0.050]; match[0.048]; won[0.045]; lost[0.043]; played[0.042]; day[0.042]; ball[0.041]; time[0.041]; play[0.041]; play[0.041]; play[0.041]; play[0.041]; Rios[0.041]; really[0.041]; really[0.041]; said[0.041]; said[0.041]; American[0.041]; court[0.040]; crowd[0.040]; fans[0.039]; points[0.039]; blew[0.039]; lose[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. Open ==> ENTITY: \u001b[32mUS Open (tennis)\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.201:0.201[0]; log p(e|m)= -1.238:-1.238[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.048]; Open[0.045]; Wimbledon[0.044]; Wimbledon[0.044]; Wimbledon[0.044]; upsets[0.042]; Agassi[0.042]; Agassi[0.042]; champion[0.042]; finalist[0.042]; Larry[0.041]; Jeff[0.041]; York[0.041]; wildcard[0.041]; Andre[0.041]; player[0.041]; sixth[0.041]; hander[0.040]; seeded[0.040]; seeded[0.040]; Alex[0.040]; Upsets[0.040]; defeated[0.039]; Thursday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Washington\u001b[39m ==> ENTITY: \u001b[32mMaliVai Washington\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.051]; match[0.050]; collegiate[0.050]; Wimbledon[0.050]; game[0.048]; play[0.047]; fifth[0.047]; set[0.046]; sixth[0.046]; eighth[0.045]; national[0.044]; champion[0.044]; England[0.044]; victory[0.044]; serve[0.043]; serve[0.043]; fans[0.043]; Club[0.043]; backhand[0.043]; backhand[0.043]; couple[0.043]; forehand[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tarango\u001b[39m ==> ENTITY: \u001b[32mJeff Tarango\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.105:0.105[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntennis[0.050]; match[0.047]; Tarango[0.044]; played[0.043]; court[0.042]; won[0.042]; crowd[0.042]; play[0.042]; play[0.042]; play[0.042]; play[0.042]; ball[0.041]; really[0.041]; really[0.041]; lost[0.041]; blatantly[0.040]; American[0.040]; Rios[0.040]; fans[0.040]; time[0.040]; said[0.040]; cheering[0.039]; jeering[0.039]; blew[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andre Agassi\u001b[39m ==> ENTITY: \u001b[32mAndre Agassi\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.051]; Open[0.047]; Wimbledon[0.047]; Wimbledon[0.047]; Wimbledon[0.047]; finalist[0.045]; player[0.045]; Larry[0.045]; Jeff[0.044]; Alex[0.044]; defeated[0.043]; upsets[0.042]; seeded[0.042]; wildcard[0.042]; court[0.041]; lost[0.041]; lost[0.041]; seed[0.041]; comeback[0.041]; Marcelo[0.041]; Thursday[0.040]; hander[0.040]; Upsets[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rios\u001b[39m ==> ENTITY: \u001b[32mMarcelo Ríos\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntennis[0.054]; tennis[0.054]; tennis[0.054]; tennis[0.054]; match[0.051]; match[0.051]; college[0.049]; won[0.048]; player[0.047]; ranked[0.046]; lost[0.046]; tiebreaker[0.046]; having[0.045]; played[0.045]; played[0.045]; played[0.045]; day[0.045]; day[0.045]; day[0.045]; set[0.044]; ball[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.068:0.068[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwinner[0.046]; winner[0.046]; play[0.045]; game[0.044]; win[0.044]; semifinals[0.044]; eighth[0.044]; fifth[0.044]; Washington[0.044]; Washington[0.044]; Washington[0.044]; Washington[0.044]; national[0.044]; sixth[0.043]; title[0.043]; weeks[0.043]; champion[0.043]; serving[0.042]; served[0.042]; break[0.042]; victory[0.042]; ranked[0.042]; collegiate[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = India ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.241:0.241[0]; local()= 0.049:0.049[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwinning[0.044]; games[0.044]; match[0.044]; champion[0.043]; day[0.043]; semifinals[0.043]; came[0.043]; win[0.042]; sixth[0.042]; player[0.042]; lost[0.042]; lost[0.042]; Stadium[0.042]; winner[0.042]; went[0.041]; comeback[0.041]; comeback[0.041]; title[0.040]; chance[0.040]; ranked[0.040]; hander[0.040]; defeated[0.040]; minute[0.040]; Paes[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wimbledon ==> ENTITY: \u001b[32mThe Championships, Wimbledon\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.851:-0.851[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.047]; Wimbledon[0.047]; Wimbledon[0.047]; Open[0.044]; champion[0.043]; Agassi[0.043]; Agassi[0.043]; finalist[0.043]; comeback[0.042]; wildcard[0.042]; sixth[0.041]; upsets[0.041]; Andre[0.041]; games[0.041]; hour[0.041]; player[0.040]; defeated[0.040]; seeded[0.040]; seeded[0.040]; Alex[0.039]; Thursday[0.039]; winning[0.039]; York[0.038]; seed[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rios\u001b[39m ==> ENTITY: \u001b[32mMarcelo Ríos\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntennis[0.048]; tennis[0.048]; tournament[0.046]; match[0.045]; match[0.045]; Wimbledon[0.044]; college[0.044]; won[0.042]; player[0.042]; eighth[0.041]; Chilean[0.041]; ranked[0.041]; tiebreaker[0.040]; couple[0.040]; victory[0.040]; ago[0.040]; played[0.040]; played[0.040]; day[0.039]; set[0.039]; backhand[0.039]; play[0.039]; play[0.039]; fraternities[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wimbledon ==> ENTITY: \u001b[32mThe Championships, Wimbledon\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.169:0.169[0]; log p(e|m)= -0.851:-0.851[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWimbledon[0.048]; match[0.045]; Agassi[0.044]; champion[0.044]; sixth[0.043]; comeback[0.043]; Martin[0.042]; semifinals[0.042]; title[0.042]; win[0.041]; games[0.041]; ranked[0.041]; player[0.041]; set[0.041]; set[0.041]; defeated[0.041]; sets[0.040]; ago[0.040]; seeded[0.040]; charging[0.040]; winning[0.040]; weeks[0.040]; winner[0.040]; seed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Washington\u001b[39m ==> ENTITY: \u001b[32mMaliVai Washington\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.046]; match[0.045]; match[0.045]; player[0.045]; Wimbledon[0.045]; played[0.045]; game[0.043]; play[0.043]; ranked[0.043]; fifth[0.042]; set[0.041]; set[0.041]; sixth[0.041]; eighth[0.040]; American[0.040]; tiebreaker[0.040]; England[0.040]; victory[0.039]; serve[0.039]; serve[0.039]; fans[0.039]; Club[0.039]; backhand[0.039]; backhand[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Washington\u001b[39m ==> ENTITY: \u001b[32mMaliVai Washington\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.045]; player[0.045]; Wimbledon[0.045]; Wimbledon[0.045]; ranked[0.043]; games[0.043]; semifinals[0.042]; pro[0.042]; sets[0.042]; set[0.041]; set[0.041]; sixth[0.041]; defeated[0.041]; Agassi[0.041]; Leander[0.040]; win[0.040]; Jeff[0.040]; title[0.040]; seeded[0.040]; champion[0.040]; Todd[0.040]; comeback[0.040]; court[0.040]; Martin[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wimbledon ==> ENTITY: \u001b[32mThe Championships, Wimbledon\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.851:-0.851[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWimbledon[0.049]; Wimbledon[0.049]; Wimbledon[0.049]; Open[0.046]; Agassi[0.045]; Agassi[0.045]; champion[0.044]; sixth[0.044]; finalist[0.044]; wildcard[0.043]; Martin[0.043]; upsets[0.042]; semifinals[0.042]; games[0.042]; hour[0.042]; player[0.042]; set[0.042]; Andre[0.042]; sets[0.041]; ago[0.041]; winner[0.041]; seeded[0.041]; seeded[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tarango\u001b[39m ==> ENTITY: \u001b[32mJeff Tarango\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntennis[0.050]; tennis[0.050]; tennis[0.050]; tennis[0.050]; match[0.048]; match[0.048]; college[0.047]; player[0.045]; Tarango[0.045]; Tarango[0.045]; Tarango[0.045]; spectators[0.044]; played[0.044]; played[0.044]; played[0.044]; tiebreaker[0.043]; having[0.043]; court[0.043]; won[0.043]; crowd[0.043]; crowd[0.043]; ranked[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tarango\u001b[39m ==> ENTITY: \u001b[32mJeff Tarango\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.048]; Open[0.045]; Agassi[0.045]; Wimbledon[0.044]; Wimbledon[0.044]; player[0.043]; Alex[0.042]; Larry[0.042]; Andre[0.042]; upsets[0.042]; court[0.042]; finalist[0.041]; Thursday[0.041]; wildcard[0.041]; Upsets[0.041]; seeded[0.040]; lost[0.040]; lost[0.040]; Rios[0.040]; Rios[0.040]; Stadium[0.040]; comeback[0.039]; fortunate[0.039]; grandstand[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Washington\u001b[39m ==> ENTITY: \u001b[32mMaliVai Washington\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.048]; match[0.048]; collegiate[0.048]; game[0.046]; play[0.045]; ranked[0.045]; pro[0.045]; fifth[0.044]; set[0.044]; set[0.044]; sixth[0.044]; eighth[0.043]; win[0.043]; title[0.042]; national[0.042]; champion[0.042]; victory[0.042]; serve[0.041]; serve[0.041]; backhand[0.041]; backhand[0.041]; couple[0.041]; forehand[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tarango\u001b[39m ==> ENTITY: \u001b[32mJeff Tarango\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntennis[0.047]; tennis[0.047]; tennis[0.047]; tournament[0.045]; match[0.044]; college[0.043]; Wimbledon[0.042]; player[0.041]; Tarango[0.041]; Tarango[0.041]; Tarango[0.041]; spectators[0.041]; played[0.041]; played[0.041]; tiebreaker[0.040]; having[0.040]; won[0.040]; crowd[0.040]; ranked[0.040]; play[0.040]; play[0.040]; set[0.039]; Chilean[0.039]; booed[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew Haven\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew Haven Open at Yale\u001b[39m <---> \u001b[32mNew Haven, Connecticut\u001b[39m\t\nSCORES: global= 0.252:0.249[\u001b[31m0.003\u001b[39m]; local()= 0.167:0.091[\u001b[31m0.076\u001b[39m]; log p(e|m)= -2.919:-0.462[\u001b[32m2.457\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; Agassi[0.044]; match[0.044]; Wimbledon[0.043]; Wimbledon[0.043]; home[0.042]; Martin[0.042]; champion[0.042]; champion[0.042]; semifinals[0.041]; Washington[0.041]; Washington[0.041]; Washington[0.041]; Jeff[0.041]; seeded[0.041]; national[0.041]; win[0.041]; fifth[0.041]; defeated[0.041]; ranked[0.041]; title[0.040]; night[0.040]; sixth[0.040]; folded[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Agassi\u001b[39m ==> ENTITY: \u001b[32mAndre Agassi\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMartin[0.046]; Open[0.044]; Wimbledon[0.044]; Wimbledon[0.044]; Wimbledon[0.044]; match[0.044]; champion[0.043]; player[0.042]; Jeff[0.042]; Alex[0.042]; sixth[0.042]; semifinals[0.041]; sets[0.041]; title[0.041]; Todd[0.040]; games[0.040]; defeated[0.040]; upsets[0.040]; set[0.040]; set[0.040]; seeded[0.040]; seeded[0.040]; wildcard[0.040]; Paes[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tarango\u001b[39m ==> ENTITY: \u001b[32mJeff Tarango\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.046]; match[0.046]; match[0.046]; eighth[0.044]; Wimbledon[0.043]; player[0.042]; Tarango[0.042]; played[0.042]; serve[0.042]; backhand[0.041]; backhand[0.041]; game[0.041]; tiebreaker[0.041]; won[0.041]; ranked[0.041]; play[0.040]; set[0.040]; Chilean[0.040]; booed[0.040]; served[0.040]; blasted[0.040]; aces[0.040]; forehand[0.040]; victory[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = MaliVai Washington\u001b[39m ==> ENTITY: \u001b[32mMaliVai Washington\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.046]; player[0.046]; Wimbledon[0.045]; Wimbledon[0.045]; Wimbledon[0.045]; Open[0.045]; wildcard[0.043]; finalist[0.042]; defeated[0.042]; Agassi[0.041]; Jeff[0.041]; Andre[0.041]; seeded[0.041]; comeback[0.041]; court[0.040]; hander[0.040]; fell[0.040]; upsets[0.040]; lost[0.040]; lost[0.040]; Alex[0.039]; Washington[0.039]; Stadium[0.039]; seed[0.039]; \t\n\r [>.............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 19s883ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 109/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1021testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bombay ==> ENTITY: \u001b[32mMumbai\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.103:0.103[0]; log p(e|m)= -0.242:-0.242[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndian[0.048]; Indian[0.048]; rupees[0.044]; rupees[0.044]; rupees[0.044]; relating[0.043]; state[0.043]; state[0.043]; Hindustan[0.042]; Hindustan[0.042]; firm[0.042]; announcement[0.041]; awaited[0.041]; owned[0.041]; gains[0.041]; trade[0.041]; supply[0.040]; rates[0.039]; expect[0.039]; base[0.039]; makers[0.039]; buying[0.039]; rose[0.039]; dealers[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indian ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.618:-0.618[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndian[0.049]; Bombay[0.046]; Hindustan[0.044]; Hindustan[0.044]; state[0.042]; state[0.042]; rupees[0.042]; rupees[0.042]; rupees[0.042]; producer[0.042]; awaited[0.042]; relating[0.041]; said[0.041]; said[0.041]; owned[0.040]; narrow[0.040]; rates[0.040]; expect[0.040]; trade[0.040]; cut[0.040]; cut[0.040]; falls[0.040]; extended[0.040]; gains[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indian ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.100:0.100[0]; log p(e|m)= -0.618:-0.618[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndian[0.056]; Bombay[0.048]; Hindustan[0.046]; Hindustan[0.046]; rupees[0.045]; rupees[0.045]; state[0.044]; state[0.044]; producer[0.043]; awaited[0.043]; relating[0.043]; said[0.042]; said[0.042]; owned[0.042]; trade[0.042]; rates[0.041]; announcement[0.041]; firm[0.041]; base[0.041]; extended[0.041]; narrow[0.041]; fell[0.040]; fell[0.040]; \t\n\r [>.............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 21s279ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 112/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 986testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Frank Bruno\u001b[39m ==> ENTITY: \u001b[32mFrank Bruno\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.045]; heavyweight[0.045]; title[0.045]; Boxing[0.044]; Boxing[0.044]; champion[0.044]; Briton[0.043]; World[0.042]; London[0.042]; ring[0.042]; medical[0.042]; Tyson[0.041]; boxed[0.041]; Bruno[0.041]; Bruno[0.041]; Bruno[0.041]; lost[0.040]; doctor[0.040]; Mike[0.040]; quit[0.040]; wanted[0.040]; quits[0.039]; glad[0.039]; specialist[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bruno\u001b[39m ==> ENTITY: \u001b[32mFrank Bruno\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfights[0.051]; Britain[0.050]; Britain[0.050]; heavyweight[0.050]; title[0.049]; title[0.049]; title[0.049]; Boxing[0.049]; champion[0.049]; Oliver[0.048]; Briton[0.047]; David[0.047]; finally[0.047]; World[0.046]; ring[0.046]; ring[0.046]; medical[0.046]; Professor[0.045]; Tyson[0.045]; Frank[0.045]; boxed[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Briton ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.049:0.049[0]; log p(e|m)= -0.970:-0.970[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBriton[0.047]; world[0.045]; world[0.045]; Oliver[0.045]; American[0.045]; American[0.045]; round[0.042]; beating[0.042]; fight[0.041]; fought[0.041]; Wembley[0.041]; referee[0.040]; Bruno[0.040]; Bruno[0.040]; Bruno[0.040]; flurry[0.040]; crown[0.040]; September[0.040]; unanimous[0.040]; McCall[0.040]; hold[0.040]; lost[0.040]; lost[0.040]; opening[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bruno\u001b[39m ==> ENTITY: \u001b[32mFrank Bruno\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfights[0.047]; heavyweight[0.046]; title[0.045]; title[0.045]; title[0.045]; fight[0.045]; fought[0.044]; Oliver[0.044]; Briton[0.044]; Briton[0.044]; David[0.044]; unanimous[0.044]; round[0.043]; finally[0.043]; ring[0.042]; punches[0.042]; Professor[0.042]; Tyson[0.042]; stopped[0.042]; Bruno[0.042]; Bruno[0.042]; Bruno[0.042]; Wembley[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Oliver McCall\u001b[39m ==> ENTITY: \u001b[32mOliver McCall\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTyson[0.046]; heavyweight[0.045]; fights[0.045]; fight[0.043]; David[0.043]; unanimous[0.042]; stopped[0.042]; lost[0.042]; lost[0.042]; ring[0.042]; round[0.041]; title[0.041]; title[0.041]; title[0.041]; title[0.041]; referee[0.041]; decision[0.041]; fought[0.041]; punches[0.041]; Vegas[0.041]; Professor[0.040]; finally[0.040]; beating[0.040]; flurry[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.054]; Briton[0.044]; Oliver[0.043]; David[0.043]; American[0.043]; quit[0.042]; test[0.042]; world[0.041]; world[0.041]; World[0.041]; Council[0.041]; day[0.041]; unanimous[0.041]; title[0.041]; title[0.041]; title[0.041]; fights[0.040]; right[0.040]; said[0.040]; said[0.040]; specialist[0.040]; reported[0.040]; March[0.040]; popular[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Briton ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.970:-0.970[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.053]; Britain[0.053]; David[0.045]; London[0.043]; sportsmen[0.043]; world[0.043]; world[0.043]; newspaper[0.042]; newspaper[0.042]; champion[0.040]; day[0.040]; reported[0.040]; specialist[0.040]; Council[0.040]; popular[0.040]; World[0.040]; examined[0.039]; really[0.039]; Bruno[0.039]; Bruno[0.039]; Bruno[0.039]; Bruno[0.039]; Bruno[0.039]; Frank[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tyson\u001b[39m ==> ENTITY: \u001b[32mMike Tyson\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfights[0.045]; heavyweight[0.044]; punches[0.044]; ring[0.043]; fight[0.043]; referee[0.042]; stopped[0.042]; eye[0.042]; eye[0.042]; eye[0.042]; lost[0.041]; lost[0.041]; title[0.041]; title[0.041]; title[0.041]; round[0.041]; fought[0.041]; unanimous[0.041]; Oliver[0.041]; Bruno[0.040]; Bruno[0.040]; Bruno[0.040]; Bruno[0.040]; Vegas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bruno\u001b[39m ==> ENTITY: \u001b[32mFrank Bruno\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfights[0.051]; Britain[0.050]; heavyweight[0.050]; title[0.049]; title[0.049]; title[0.049]; title[0.049]; title[0.049]; Boxing[0.048]; fought[0.047]; Oliver[0.047]; Briton[0.047]; Briton[0.047]; David[0.047]; unanimous[0.047]; finally[0.046]; World[0.046]; ring[0.046]; Professor[0.045]; Tyson[0.045]; Tyson[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.081:0.081[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.048]; champion[0.045]; title[0.044]; said[0.043]; said[0.043]; Friday[0.043]; Council[0.043]; day[0.042]; told[0.042]; told[0.042]; Briton[0.042]; Frank[0.041]; boxed[0.041]; March[0.041]; Mike[0.041]; quits[0.041]; World[0.041]; wanted[0.041]; reported[0.040]; lost[0.040]; medical[0.040]; Bruno[0.040]; Bruno[0.040]; Bruno[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sun ==> ENTITY: \u001b[32mThe Sun (United Kingdom)\u001b[39m\t\nSCORES: global= 0.238:0.238[0]; local()= 0.135:0.135[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnewspaper[0.046]; newspaper[0.046]; Britain[0.044]; Britain[0.044]; told[0.044]; told[0.044]; medical[0.043]; reported[0.042]; quit[0.042]; wanted[0.042]; Mike[0.041]; London[0.041]; boxed[0.041]; Frank[0.041]; day[0.040]; soon[0.040]; World[0.040]; quits[0.040]; popular[0.040]; doctor[0.040]; world[0.040]; Briton[0.040]; March[0.039]; title[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Briton ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.065:0.065[0]; log p(e|m)= -0.970:-0.970[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBriton[0.046]; David[0.046]; world[0.044]; world[0.044]; world[0.044]; Oliver[0.044]; American[0.044]; American[0.044]; round[0.042]; beating[0.042]; fight[0.041]; fought[0.041]; examined[0.041]; Wembley[0.040]; referee[0.040]; Bruno[0.040]; Bruno[0.040]; Bruno[0.040]; Bruno[0.040]; flurry[0.040]; crown[0.040]; September[0.040]; unanimous[0.040]; training[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sun ==> ENTITY: \u001b[32mThe Sun (United Kingdom)\u001b[39m\t\nSCORES: global= 0.237:0.237[0]; local()= 0.092:0.092[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbeating[0.045]; Britain[0.045]; told[0.044]; told[0.044]; David[0.042]; wanted[0.042]; Mike[0.042]; title[0.041]; title[0.041]; title[0.041]; title[0.041]; title[0.041]; point[0.041]; lost[0.041]; lost[0.041]; Oliver[0.041]; finally[0.041]; day[0.041]; soon[0.041]; points[0.041]; Briton[0.040]; World[0.040]; landing[0.040]; popular[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mike Tyson\u001b[39m ==> ENTITY: \u001b[32mMike Tyson\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.128:0.128[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nheavyweight[0.044]; ring[0.044]; Boxing[0.044]; Boxing[0.044]; boxed[0.043]; eye[0.042]; eye[0.042]; eye[0.042]; eye[0.042]; lost[0.042]; title[0.042]; title[0.042]; medical[0.041]; Bruno[0.041]; Bruno[0.041]; Bruno[0.041]; Bruno[0.041]; Bruno[0.041]; champion[0.041]; March[0.041]; Professor[0.041]; training[0.040]; World[0.040]; Frank[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bruno\u001b[39m ==> ENTITY: \u001b[32mFrank Bruno\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfights[0.046]; heavyweight[0.044]; title[0.044]; title[0.044]; title[0.044]; fight[0.044]; fought[0.042]; Oliver[0.042]; Briton[0.042]; Briton[0.042]; unanimous[0.042]; round[0.042]; finally[0.042]; ring[0.041]; punches[0.041]; Tyson[0.040]; stopped[0.040]; Bruno[0.040]; Bruno[0.040]; Bruno[0.040]; Wembley[0.040]; lost[0.040]; lost[0.040]; crown[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = American ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.047:0.047[0]; log p(e|m)= -0.130:-0.130[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.050]; David[0.044]; Oliver[0.044]; Briton[0.043]; Briton[0.043]; world[0.042]; world[0.042]; September[0.041]; Bruno[0.041]; Bruno[0.041]; Bruno[0.041]; Bruno[0.041]; lost[0.041]; lost[0.041]; McCall[0.041]; points[0.041]; title[0.040]; title[0.040]; title[0.040]; flurry[0.040]; finally[0.040]; round[0.040]; heavyweight[0.040]; fight[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.055]; Briton[0.047]; London[0.045]; quit[0.043]; champion[0.043]; world[0.042]; World[0.042]; specialist[0.041]; Council[0.041]; title[0.041]; quits[0.041]; popular[0.041]; sportsmen[0.041]; March[0.041]; heavyweight[0.040]; lost[0.040]; Frank[0.040]; day[0.040]; medical[0.040]; Mike[0.040]; Tyson[0.040]; wanted[0.039]; old[0.039]; really[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bruno\u001b[39m ==> ENTITY: \u001b[32mFrank Bruno\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfights[0.051]; Britain[0.050]; heavyweight[0.050]; title[0.049]; title[0.049]; title[0.049]; title[0.049]; fight[0.049]; fought[0.048]; Oliver[0.048]; Briton[0.047]; Briton[0.047]; David[0.047]; unanimous[0.047]; round[0.047]; finally[0.047]; ring[0.046]; punches[0.045]; Professor[0.045]; Tyson[0.045]; stopped[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bruno\u001b[39m ==> ENTITY: \u001b[32mFrank Bruno\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.045]; Britain[0.045]; heavyweight[0.045]; title[0.044]; title[0.044]; Boxing[0.044]; Boxing[0.044]; champion[0.043]; Briton[0.042]; World[0.042]; London[0.041]; ring[0.041]; medical[0.041]; Tyson[0.041]; Frank[0.040]; boxed[0.040]; Bruno[0.040]; Bruno[0.040]; Bruno[0.040]; lost[0.040]; sportsmen[0.040]; doctor[0.040]; Mike[0.040]; quit[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Las Vegas ==> ENTITY: \u001b[32mLas Vegas\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.496:-0.496[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstopped[0.044]; points[0.043]; title[0.042]; title[0.042]; title[0.042]; Briton[0.042]; Briton[0.042]; round[0.042]; American[0.042]; American[0.042]; finally[0.042]; world[0.041]; world[0.041]; Bruno[0.041]; Bruno[0.041]; Bruno[0.041]; Bruno[0.041]; beating[0.041]; opening[0.041]; crown[0.041]; flurry[0.041]; McCall[0.041]; David[0.041]; September[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = American ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.068:0.068[0]; log p(e|m)= -0.130:-0.130[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.053]; Britain[0.049]; David[0.047]; Oliver[0.047]; Briton[0.046]; Briton[0.046]; popular[0.046]; world[0.045]; world[0.045]; world[0.045]; innings[0.045]; September[0.045]; Bruno[0.044]; Bruno[0.044]; Bruno[0.044]; Bruno[0.044]; Bruno[0.044]; lost[0.044]; lost[0.044]; McCall[0.044]; points[0.044]; Monday[0.044]; \t\n\r [>.............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 21s684ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 134/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1145testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baghdad\u001b[39m ==> ENTITY: \u001b[32mBaghdad\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraqi[0.046]; Iraqi[0.046]; Iraqi[0.046]; Baghdad[0.046]; Kurdistan[0.045]; troops[0.044]; military[0.043]; Arbil[0.043]; Arbil[0.043]; Kurds[0.042]; official[0.042]; official[0.042]; official[0.042]; city[0.042]; Barzani[0.041]; Massoud[0.040]; Party[0.039]; aid[0.039]; headquarters[0.039]; leaders[0.038]; Reuters[0.038]; Democratic[0.038]; artillery[0.037]; confirmed[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arbil\u001b[39m ==> ENTITY: \u001b[32mErbil\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.048]; Kurds[0.047]; Arbil[0.046]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; troops[0.044]; Baghdad[0.044]; Baghdad[0.044]; Barzani[0.043]; military[0.042]; city[0.041]; headquarters[0.041]; official[0.040]; official[0.040]; official[0.040]; Massoud[0.039]; senior[0.039]; taking[0.038]; leaders[0.038]; backed[0.038]; Democratic[0.038]; got[0.038]; information[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = KDP\u001b[39m ==> ENTITY: \u001b[32mKurdistan Democratic Party\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.050]; Kurds[0.047]; Barzani[0.046]; Arbil[0.046]; Arbil[0.046]; Iraqi[0.046]; Iraqi[0.046]; Iraqi[0.046]; Baghdad[0.044]; Baghdad[0.044]; Party[0.043]; backed[0.043]; troops[0.042]; senior[0.042]; military[0.042]; leaders[0.041]; Massoud[0.041]; official[0.041]; official[0.041]; official[0.041]; headquarters[0.041]; artillery[0.041]; Democratic[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.085:0.085[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Baghdad[0.043]; Baghdad[0.043]; military[0.043]; Kurdistan[0.042]; confirms[0.042]; official[0.041]; official[0.041]; official[0.041]; leaders[0.041]; said[0.041]; said[0.041]; backed[0.041]; troops[0.041]; Party[0.041]; Arbil[0.041]; Arbil[0.041]; confirmed[0.040]; told[0.040]; headquarters[0.040]; Reuters[0.040]; information[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.173:0.173[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.048]; Iraqi[0.048]; Iraqi[0.048]; Kurds[0.046]; Baghdad[0.045]; Baghdad[0.045]; troops[0.044]; military[0.043]; Arbil[0.042]; Arbil[0.042]; Barzani[0.042]; aid[0.041]; told[0.040]; said[0.040]; said[0.040]; Massoud[0.040]; official[0.039]; official[0.039]; official[0.039]; identified[0.038]; city[0.038]; confirmed[0.038]; Reuters[0.038]; artillery[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.105:0.105[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntold[0.045]; official[0.043]; official[0.043]; official[0.043]; troops[0.043]; said[0.042]; said[0.042]; headquarters[0.042]; Iraqi[0.042]; Iraqi[0.042]; Iraqi[0.042]; aid[0.042]; military[0.041]; confirms[0.041]; backed[0.041]; information[0.041]; Democratic[0.041]; attack[0.041]; Baghdad[0.041]; Baghdad[0.041]; leaders[0.040]; city[0.040]; confirmed[0.040]; rival[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baghdad\u001b[39m ==> ENTITY: \u001b[32mBaghdad\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraqi[0.046]; Iraqi[0.046]; Iraqi[0.046]; Baghdad[0.046]; Kurdistan[0.045]; troops[0.044]; military[0.043]; Arbil[0.043]; Arbil[0.043]; Kurds[0.042]; official[0.042]; official[0.042]; official[0.042]; city[0.042]; Barzani[0.041]; Massoud[0.040]; Party[0.039]; aid[0.039]; headquarters[0.039]; leaders[0.038]; Reuters[0.038]; Democratic[0.038]; artillery[0.037]; confirmed[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.178:0.178[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.048]; Iraqi[0.047]; Iraqi[0.047]; Kurds[0.046]; military[0.046]; Baghdad[0.044]; Baghdad[0.044]; troops[0.044]; Arbil[0.042]; Arbil[0.042]; Barzani[0.042]; aid[0.041]; artillery[0.041]; said[0.040]; said[0.040]; Massoud[0.039]; official[0.039]; official[0.039]; official[0.039]; backed[0.038]; city[0.038]; Reuters[0.038]; attack[0.037]; identified[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = KDP\u001b[39m ==> ENTITY: \u001b[32mKurdistan Democratic Party\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.050]; Kurds[0.047]; Barzani[0.046]; Arbil[0.046]; Arbil[0.046]; Iraqi[0.046]; Iraqi[0.046]; Iraqi[0.046]; Baghdad[0.044]; Baghdad[0.044]; Party[0.043]; backed[0.043]; troops[0.042]; senior[0.042]; military[0.042]; leaders[0.041]; Massoud[0.041]; official[0.041]; official[0.041]; official[0.041]; headquarters[0.041]; artillery[0.041]; Democratic[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.178:0.178[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.048]; Iraqi[0.047]; Iraqi[0.047]; Kurds[0.046]; military[0.046]; Baghdad[0.044]; Baghdad[0.044]; troops[0.044]; Arbil[0.042]; Arbil[0.042]; Barzani[0.042]; aid[0.041]; artillery[0.041]; said[0.040]; said[0.040]; Massoud[0.039]; official[0.039]; official[0.039]; official[0.039]; backed[0.038]; city[0.038]; Reuters[0.038]; attack[0.037]; identified[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kurdistan Democratic Party\u001b[39m ==> ENTITY: \u001b[32mKurdistan Democratic Party\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurds[0.046]; Barzani[0.045]; Arbil[0.044]; Arbil[0.044]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; military[0.043]; Baghdad[0.043]; Baghdad[0.043]; backed[0.041]; official[0.041]; official[0.041]; official[0.041]; troops[0.041]; senior[0.040]; leaders[0.040]; Massoud[0.040]; headquarters[0.040]; artillery[0.040]; said[0.040]; said[0.040]; rival[0.038]; Reuters[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arbil\u001b[39m ==> ENTITY: \u001b[32mErbil\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.048]; Kurds[0.047]; Arbil[0.046]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; troops[0.044]; Baghdad[0.044]; Baghdad[0.044]; Barzani[0.043]; military[0.042]; city[0.041]; headquarters[0.041]; official[0.040]; official[0.040]; official[0.040]; Massoud[0.039]; senior[0.039]; taking[0.038]; leaders[0.038]; backed[0.038]; Democratic[0.038]; got[0.038]; information[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Massoud Barzani\u001b[39m ==> ENTITY: \u001b[32mMasoud Barzani\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.047]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Kurds[0.044]; Arbil[0.044]; Arbil[0.044]; Baghdad[0.043]; Baghdad[0.043]; troops[0.043]; military[0.042]; leaders[0.041]; backed[0.041]; senior[0.041]; official[0.040]; official[0.040]; official[0.040]; said[0.040]; said[0.040]; Party[0.039]; Democratic[0.039]; artillery[0.038]; aid[0.038]; told[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = KDP\u001b[39m ==> ENTITY: \u001b[32mKurdistan Democratic Party\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.050]; Kurds[0.047]; Barzani[0.046]; Arbil[0.046]; Arbil[0.046]; Iraqi[0.046]; Iraqi[0.046]; Iraqi[0.046]; Baghdad[0.044]; Baghdad[0.044]; Party[0.043]; backed[0.043]; troops[0.042]; senior[0.042]; military[0.042]; leaders[0.041]; Massoud[0.041]; official[0.041]; official[0.041]; official[0.041]; headquarters[0.041]; artillery[0.041]; Democratic[0.040]; \t\n\r [>.............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 22s13ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 149/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1053testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Faroes ==> ENTITY: \u001b[32mFaroe Islands national football team\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.170:0.170[0]; log p(e|m)= -1.336:-1.336[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSlovakia[0.047]; Slovakia[0.047]; Slovakia[0.047]; match[0.045]; Toftir[0.044]; soccer[0.043]; European[0.042]; qualifying[0.042]; Soccer[0.041]; Qualifier[0.041]; Faroe[0.040]; Faroe[0.040]; Faroe[0.040]; Islands[0.040]; Islands[0.040]; Islands[0.040]; World[0.040]; World[0.040]; Cup[0.040]; Cup[0.040]; minute[0.040]; beat[0.039]; halftime[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Faroe Islands ==> ENTITY: \u001b[32mFaroe Islands national football team\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.176:0.176[0]; log p(e|m)= -0.872:-0.872[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSlovakia[0.046]; Slovakia[0.046]; Slovakia[0.046]; match[0.044]; Faroe[0.043]; Faroe[0.043]; Faroes[0.043]; Toftir[0.042]; qualifying[0.042]; Cup[0.042]; Cup[0.042]; soccer[0.041]; European[0.041]; group[0.041]; Islands[0.041]; Islands[0.041]; Soccer[0.040]; minute[0.040]; Scorers[0.040]; Qualifier[0.040]; World[0.039]; World[0.039]; Saturday[0.039]; beat[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Slovakia ==> ENTITY: \u001b[32mSlovakia national football team\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.172:0.172[0]; log p(e|m)= -2.129:-2.129[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSlovakia[0.049]; Slovakia[0.049]; soccer[0.047]; Qualifier[0.044]; match[0.044]; European[0.044]; Cup[0.043]; Cup[0.043]; qualifying[0.043]; Soccer[0.043]; Toftir[0.042]; Lubomir[0.041]; World[0.040]; World[0.040]; Faroes[0.040]; Mueller[0.040]; Jan[0.039]; beat[0.039]; Beat[0.039]; group[0.038]; minute[0.038]; Saturday[0.038]; Scorers[0.037]; Allan[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Slovakia ==> ENTITY: \u001b[32mSlovakia national football team\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.172:0.172[0]; log p(e|m)= -2.129:-2.129[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSlovakia[0.049]; Slovakia[0.049]; soccer[0.047]; Qualifier[0.045]; match[0.044]; Cup[0.044]; Cup[0.044]; Soccer[0.043]; European[0.043]; qualifying[0.043]; Toftir[0.042]; Lubomir[0.041]; World[0.041]; World[0.041]; Faroes[0.040]; Mueller[0.040]; Jan[0.039]; beat[0.039]; Beat[0.039]; group[0.038]; minute[0.038]; Saturday[0.038]; Scorers[0.037]; Allan[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Slovakia ==> ENTITY: \u001b[32mSlovakia national football team\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.172:0.172[0]; log p(e|m)= -2.129:-2.129[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSlovakia[0.049]; Slovakia[0.049]; soccer[0.047]; Qualifier[0.044]; match[0.044]; European[0.044]; Cup[0.043]; Cup[0.043]; qualifying[0.043]; Soccer[0.043]; Toftir[0.042]; Lubomir[0.041]; World[0.040]; World[0.040]; Faroes[0.040]; Mueller[0.040]; Jan[0.039]; beat[0.039]; Beat[0.039]; group[0.038]; minute[0.038]; Saturday[0.038]; Scorers[0.037]; Allan[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Toftir\u001b[39m ==> ENTITY: \u001b[32mToftir\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.142:0.142[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.046]; soccer[0.045]; Cup[0.044]; Cup[0.044]; Soccer[0.044]; qualifying[0.043]; beat[0.043]; Faroes[0.043]; Qualifier[0.042]; Faroe[0.041]; Faroe[0.041]; Faroe[0.041]; Islands[0.041]; Islands[0.041]; Islands[0.041]; Saturday[0.041]; halftime[0.040]; Attendance[0.040]; Slovakia[0.040]; Slovakia[0.040]; Slovakia[0.040]; Scorers[0.040]; group[0.040]; Lubomir[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lubomir Moravcik\u001b[39m ==> ENTITY: \u001b[32mĽubomír Moravčík\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.050]; match[0.049]; Soccer[0.049]; European[0.047]; Cup[0.047]; Cup[0.047]; Qualifier[0.046]; Slovakia[0.046]; Slovakia[0.046]; Slovakia[0.046]; halftime[0.045]; minute[0.045]; Scorers[0.045]; Attendance[0.045]; Faroes[0.045]; Jan[0.044]; Toftir[0.044]; qualifying[0.043]; Saturday[0.043]; beat[0.043]; World[0.043]; World[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Faroe Islands ==> ENTITY: \u001b[32mFaroe Islands national football team\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.176:0.176[0]; log p(e|m)= -0.872:-0.872[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSlovakia[0.046]; Slovakia[0.046]; Slovakia[0.046]; match[0.044]; Faroe[0.043]; Faroe[0.043]; Faroes[0.043]; Toftir[0.042]; qualifying[0.042]; Cup[0.042]; Cup[0.042]; soccer[0.041]; European[0.041]; group[0.041]; Islands[0.041]; Islands[0.041]; Soccer[0.040]; minute[0.040]; Scorers[0.040]; Qualifier[0.040]; World[0.039]; World[0.039]; Saturday[0.039]; beat[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Peter Dubovsky\u001b[39m ==> ENTITY: \u001b[32mPeter Dubovský\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.101:0.101[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.049]; Slovakia[0.049]; Slovakia[0.049]; Slovakia[0.049]; match[0.047]; Qualifier[0.046]; Soccer[0.046]; halftime[0.046]; European[0.045]; Scorers[0.045]; Attendance[0.045]; Lubomir[0.045]; Toftir[0.044]; group[0.044]; Saturday[0.044]; beat[0.044]; qualifying[0.044]; Faroes[0.044]; Jan[0.044]; Cup[0.044]; Cup[0.044]; minute[0.043]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mFaroe Islands\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFaroe Islands national football team\u001b[39m <---> \u001b[32mFaroe Islands\u001b[39m\t\nSCORES: global= 0.269:0.240[\u001b[31m0.029\u001b[39m]; local()= 0.176:0.082[\u001b[31m0.094\u001b[39m]; log p(e|m)= -0.872:-0.596[\u001b[32m0.276\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSlovakia[0.046]; Slovakia[0.046]; Slovakia[0.046]; match[0.044]; Faroe[0.043]; Faroe[0.043]; Faroes[0.043]; Toftir[0.042]; qualifying[0.042]; Cup[0.042]; Cup[0.042]; soccer[0.041]; European[0.041]; group[0.041]; Islands[0.041]; Islands[0.041]; Soccer[0.040]; minute[0.040]; Scorers[0.040]; Qualifier[0.040]; World[0.039]; World[0.039]; Saturday[0.039]; beat[0.038]; \t\n\r [>.............................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 22s298ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 161/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 967testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sven Nylander\u001b[39m ==> ENTITY: \u001b[32mSven Nylander\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.109:0.109[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMatete[0.044]; Barsosio[0.044]; seconds[0.043]; seconds[0.043]; seconds[0.043]; Sweden[0.043]; Norway[0.043]; Anja[0.043]; Loroupe[0.043]; Gunhild[0.041]; Finland[0.041]; Tegla[0.041]; Gete[0.041]; Women[0.041]; Women[0.041]; Ruecker[0.041]; Wami[0.041]; Karin[0.040]; Halle[0.040]; Derrick[0.040]; minutes[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jamaica ==> ENTITY: \u001b[32mJamaica\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.333:-0.333[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGreece[0.045]; Britain[0.044]; seconds[0.044]; Australia[0.043]; Canada[0.043]; Cuba[0.042]; Cuba[0.042]; Johnson[0.042]; George[0.042]; relay[0.042]; Charles[0.041]; javelin[0.041]; Fredericks[0.041]; Michael[0.041]; Michael[0.041]; Michael[0.041]; Lithuania[0.041]; Nigeria[0.041]; Nigeria[0.041]; Nigeria[0.041]; Green[0.041]; Belarus[0.040]; Portugal[0.040]; race[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = William Tanui\u001b[39m ==> ENTITY: \u001b[32mWilliam Tanui\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.099:0.099[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKenya[0.045]; Kenya[0.045]; Devers[0.044]; Rotich[0.043]; seconds[0.043]; seconds[0.043]; Morceli[0.043]; Juergen[0.043]; Lars[0.042]; discus[0.042]; Alekna[0.042]; Geir[0.041]; Boldon[0.041]; Andreas[0.041]; Washington[0.041]; shot[0.041]; Mayock[0.041]; Johnson[0.040]; Lithuania[0.040]; minutes[0.040]; Norway[0.040]; Patrick[0.040]; Denmark[0.040]; mile[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Geir Moen\u001b[39m ==> ENTITY: \u001b[32mGeir Moen\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.098:0.098[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.046]; Denmark[0.045]; Britain[0.042]; Florian[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Busemann[0.042]; Thorsett[0.042]; Johnson[0.042]; Belgium[0.042]; Netherlands[0.042]; Morceli[0.041]; seconds[0.041]; seconds[0.041]; seconds[0.041]; Men[0.040]; Men[0.040]; Men[0.040]; hurdles[0.040]; Ivan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jearl Miles\u001b[39m ==> ENTITY: \u001b[32mJearl Miles Clark\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.058:0.058[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBarsosio[0.047]; Matete[0.046]; Anja[0.044]; Donovan[0.044]; Robinson[0.044]; Geir[0.044]; hurdles[0.044]; Jamaica[0.044]; Kipketer[0.043]; Australia[0.043]; Torrance[0.043]; Davidson[0.043]; Mitchell[0.043]; Dennis[0.043]; Thomas[0.043]; Gunhild[0.043]; Derrick[0.043]; Smith[0.043]; Michael[0.043]; seconds[0.042]; seconds[0.042]; Bailey[0.042]; Ruecker[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.050]; Russia[0.050]; Mark[0.049]; Steve[0.048]; Canada[0.048]; Slovenia[0.047]; Switzerland[0.046]; Sweden[0.046]; Russell[0.045]; Portugal[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Britain[0.044]; Athletics[0.044]; Tony[0.043]; Svetlana[0.043]; France[0.043]; Jack[0.043]; Cuba[0.042]; Cuba[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Heli Rantanen\u001b[39m ==> ENTITY: \u001b[32mHeli Rantanen\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \njavelin[0.051]; Finland[0.049]; Trine[0.048]; McPaul[0.047]; Louise[0.047]; jump[0.046]; Sweden[0.046]; Tanja[0.046]; Norway[0.045]; Norway[0.045]; Florian[0.045]; race[0.045]; Isel[0.045]; relay[0.045]; seconds[0.045]; Australia[0.044]; Johnson[0.044]; Greece[0.044]; Fita[0.044]; Ethiopia[0.044]; Kenny[0.043]; Canada[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jamaica ==> ENTITY: \u001b[32mJamaica\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.091:0.091[0]; log p(e|m)= -0.333:-0.333[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.044]; Britain[0.044]; seconds[0.044]; Hungary[0.043]; Australia[0.043]; Canada[0.043]; Smith[0.043]; Cuba[0.043]; Mitchell[0.042]; Denmark[0.042]; Kipketer[0.041]; Michael[0.041]; Robinson[0.041]; Kiptoo[0.041]; David[0.041]; Nigeria[0.041]; Nigeria[0.041]; Nigeria[0.041]; Nigeria[0.041]; Green[0.041]; Craig[0.040]; Phylis[0.040]; Drummond[0.040]; Langat[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sally Barsosio\u001b[39m ==> ENTITY: \u001b[32mSally Barsosio\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPrivalova[0.045]; Loroupe[0.045]; Tegla[0.043]; Kenya[0.043]; Kenya[0.043]; Kenya[0.043]; Kenya[0.043]; Matete[0.042]; Phylis[0.042]; Wami[0.042]; Cheruiyot[0.041]; Germany[0.041]; Anja[0.041]; Ottey[0.041]; Jearl[0.041]; Romania[0.041]; Chryste[0.041]; Hungary[0.041]; Onyali[0.041]; Ethiopia[0.040]; Halle[0.040]; Chandra[0.040]; Australia[0.040]; Gunhild[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belarus ==> ENTITY: \u001b[32mBelarus\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.422:-0.422[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMen[0.048]; Men[0.048]; Belarus[0.046]; Russia[0.043]; Russia[0.043]; Lithuania[0.043]; Romania[0.043]; Ireland[0.042]; Women[0.041]; Women[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Privalova[0.041]; Alekna[0.041]; Netherlands[0.040]; Vladimir[0.040]; Britain[0.040]; Algeria[0.040]; Spain[0.039]; Jamaica[0.039]; Devers[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Donovan Bailey\u001b[39m ==> ENTITY: \u001b[32mDonovan Bailey\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.111:0.111[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrelay[0.044]; seconds[0.044]; Canada[0.043]; Trine[0.043]; Jamaica[0.043]; Michael[0.043]; Michael[0.043]; Michael[0.043]; Kenny[0.042]; George[0.042]; javelin[0.042]; Mike[0.042]; Tanja[0.042]; Johnson[0.041]; race[0.041]; Charles[0.041]; Florian[0.041]; Harrison[0.041]; Isel[0.040]; McPaul[0.040]; Davidson[0.040]; Peter[0.040]; Norway[0.039]; Norway[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.157:0.157[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Denmark[0.044]; Portugal[0.043]; France[0.042]; Britain[0.042]; Belgium[0.041]; Canada[0.041]; Norway[0.041]; Johnson[0.040]; Men[0.040]; Men[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; minutes[0.039]; Trinidad[0.039]; Michael[0.038]; Women[0.038]; Women[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Isaac Viciosa\u001b[39m ==> ENTITY: \u001b[32mIsaac Viciosa\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.137:0.137[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMorceli[0.059]; Spain[0.057]; Juergen[0.054]; Lars[0.053]; discus[0.053]; Devers[0.053]; Alekna[0.052]; seconds[0.052]; seconds[0.052]; Onyali[0.052]; Rotich[0.052]; Andreas[0.052]; Ottey[0.052]; Merlene[0.051]; Kenya[0.051]; Kenya[0.051]; Mayock[0.051]; shot[0.051]; Garcia[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.047]; Donovan[0.044]; Belarus[0.044]; Britain[0.043]; Russia[0.043]; Germany[0.042]; Germany[0.042]; Portugal[0.042]; Finland[0.042]; Norway[0.041]; Norway[0.041]; Johnson[0.041]; Cuba[0.041]; Cuba[0.041]; Cuba[0.041]; Men[0.041]; Men[0.041]; Men[0.041]; Australia[0.041]; Michael[0.040]; Michael[0.040]; Lithuania[0.040]; Charles[0.040]; Carlos[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dmitri Markov\u001b[39m ==> ENTITY: \u001b[32mDmitri Markov\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Andrei[0.043]; Pyotr[0.043]; Igor[0.042]; Igor[0.042]; jump[0.042]; vault[0.042]; Tarasov[0.042]; Lobinger[0.041]; Ukraine[0.041]; Nele[0.040]; Tatyana[0.040]; Kazakstan[0.040]; Natalya[0.040]; Yelena[0.039]; Paul[0.039]; Maksim[0.039]; Stefka[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Karin Janke\u001b[39m ==> ENTITY: \u001b[32mKarin Janke\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.109:0.109[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Matete[0.043]; Anja[0.042]; Geir[0.042]; Sven[0.041]; Sweden[0.041]; Kipketer[0.041]; Rodal[0.041]; Linda[0.041]; Norway[0.040]; Jearl[0.040]; Ruecker[0.040]; Eric[0.040]; Phylis[0.040]; Denmark[0.040]; Nylander[0.040]; Men[0.040]; Men[0.040]; Nico[0.040]; Dusan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.046]; Belarus[0.046]; Russia[0.046]; Steve[0.044]; Denmark[0.043]; Johnson[0.042]; Ireland[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Michael[0.041]; Norway[0.041]; Vladimir[0.041]; Spain[0.041]; Britain[0.041]; Belgium[0.040]; Netherlands[0.040]; Irina[0.040]; William[0.039]; John[0.039]; Sullivan[0.039]; Algeria[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Denmark ==> ENTITY: \u001b[32mDenmark\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.162:0.162[0]; log p(e|m)= -0.268:-0.268[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.047]; Norway[0.047]; Kenya[0.043]; Kenya[0.043]; Men[0.043]; Men[0.043]; Men[0.043]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Britain[0.042]; Britain[0.042]; Canada[0.040]; Algeria[0.040]; Jamaica[0.040]; France[0.039]; Nigeria[0.038]; Nigeria[0.038]; Nigeria[0.038]; Lobinger[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Slovenia ==> ENTITY: \u001b[32mSlovenia\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.074:0.074[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.050]; Switzerland[0.049]; France[0.049]; Men[0.048]; Canada[0.048]; Germany[0.048]; Portugal[0.048]; Russia[0.046]; Russia[0.046]; Russia[0.046]; Women[0.045]; Women[0.045]; minutes[0.044]; seconds[0.044]; seconds[0.044]; seconds[0.044]; hurdles[0.043]; hurdles[0.043]; Results[0.043]; Athletics[0.043]; athletics[0.043]; Engquist[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.093:0.093[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.046]; Romania[0.044]; Women[0.043]; Women[0.043]; Hungary[0.043]; Men[0.043]; Men[0.043]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Britain[0.042]; Barsosio[0.042]; Kenya[0.041]; Kenya[0.041]; Kenya[0.041]; Kenya[0.041]; Sweden[0.041]; Jamaica[0.040]; Nigeria[0.040]; Nigeria[0.040]; Nigeria[0.040]; Norway[0.040]; Tegla[0.040]; Fabrizio[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrace[0.047]; Canada[0.045]; Australia[0.044]; George[0.043]; Charles[0.042]; Norway[0.042]; Norway[0.042]; Jamaica[0.042]; Sweden[0.042]; Johnson[0.042]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Michael[0.041]; Michael[0.041]; Michael[0.041]; Lithuania[0.040]; Jesse[0.040]; Peter[0.040]; seconds[0.040]; Davidson[0.040]; Finland[0.040]; Carlos[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Frank Busemann\u001b[39m ==> ENTITY: \u001b[32mFrank Busemann\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.137:0.137[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Fredericks[0.043]; Geir[0.042]; Florian[0.042]; Thorsett[0.042]; Claus[0.041]; Jon[0.041]; Norway[0.041]; hurdles[0.041]; seconds[0.040]; seconds[0.040]; seconds[0.040]; Men[0.040]; Men[0.040]; Ivan[0.040]; Astrid[0.040]; Denmark[0.040]; Tony[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kenya ==> ENTITY: \u001b[32mKenya\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKenya[0.056]; Kenya[0.056]; Kenya[0.056]; Ethiopia[0.047]; Zambia[0.046]; Nigeria[0.043]; Tegla[0.042]; Loroupe[0.041]; Jamaica[0.040]; Cheruiyot[0.040]; Barsosio[0.040]; Australia[0.040]; Norway[0.039]; Russia[0.039]; Wami[0.039]; Sweden[0.038]; Konga[0.038]; Finland[0.038]; Matete[0.037]; Italy[0.037]; Bahamas[0.037]; Nylander[0.036]; Hungary[0.036]; Privalova[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.103:0.103[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.047]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Portugal[0.044]; Canada[0.043]; Britain[0.043]; Sweden[0.042]; Slovenia[0.042]; Mark[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Steve[0.041]; Ludmila[0.040]; Michelle[0.040]; Julie[0.040]; Carla[0.040]; Sara[0.040]; Pierce[0.040]; Svetlana[0.039]; Yekaterina[0.039]; minutes[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nele Zilinskiene\u001b[39m ==> ENTITY: \u001b[32mNelė Žilinskienė\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.105:0.105[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLithuania[0.048]; high[0.043]; jump[0.042]; jump[0.042]; Bulgaria[0.042]; Lobinger[0.042]; Belarus[0.042]; Stefka[0.042]; Hanne[0.042]; Cuba[0.041]; Inga[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Norway[0.041]; Norway[0.041]; Haugland[0.041]; Ukraine[0.041]; Rodal[0.041]; Dmitri[0.041]; Kostadinova[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cuba ==> ENTITY: \u001b[32mCuba\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.140:0.140[0]; log p(e|m)= -0.240:-0.240[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCuba[0.053]; Cuba[0.053]; Jamaica[0.048]; Lithuania[0.047]; Donovan[0.047]; Canada[0.046]; Carlos[0.045]; Portugal[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Men[0.044]; Men[0.044]; Belarus[0.044]; Morocco[0.044]; Sweden[0.044]; Lopez[0.044]; Britain[0.044]; Britain[0.044]; Norway[0.043]; Norway[0.043]; Greece[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chandra Sturrup\u001b[39m ==> ENTITY: \u001b[32mChandra Sturrup\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJamaica[0.045]; Bahamas[0.044]; seconds[0.043]; seconds[0.043]; seconds[0.043]; Women[0.043]; Women[0.043]; Matete[0.043]; Juergen[0.042]; Lars[0.042]; discus[0.042]; Ottey[0.042]; Alekna[0.041]; Chryste[0.041]; Onyali[0.041]; Devers[0.041]; Andreas[0.041]; Merlene[0.041]; Barsosio[0.040]; Miller[0.040]; Privalova[0.040]; Gunhild[0.040]; hurdles[0.040]; Gete[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Algeria ==> ENTITY: \u001b[32mAlgeria\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.337:-0.337[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.044]; Spain[0.043]; Kenya[0.043]; Kenya[0.043]; Belgium[0.043]; Burundi[0.042]; Lithuania[0.042]; Noureddine[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Men[0.041]; Men[0.041]; Men[0.041]; Netherlands[0.041]; Ireland[0.041]; Trinidad[0.041]; Morceli[0.040]; Britain[0.040]; Belarus[0.040]; Belarus[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPaul Koech\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPaul Kipsiele Koech\u001b[39m <---> \u001b[32mPaul Koech\u001b[39m\t\nSCORES: global= 0.279:0.277[\u001b[31m0.002\u001b[39m]; local()= 0.099:0.108[\u001b[32m0.010\u001b[39m]; log p(e|m)= -0.693:0.000[\u001b[32m0.693\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKenya[0.046]; Kenya[0.046]; Kenya[0.046]; Kenya[0.046]; Nele[0.043]; javelin[0.042]; Fita[0.042]; high[0.042]; Ethiopia[0.042]; Stefka[0.041]; Hanne[0.041]; Men[0.041]; Men[0.041]; Haugland[0.041]; Norway[0.040]; Norway[0.040]; seconds[0.040]; Charles[0.040]; Kostadinova[0.040]; minutes[0.040]; Michael[0.040]; Shem[0.040]; Kenny[0.040]; Komen[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.103:0.103[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.047]; Denmark[0.045]; France[0.045]; Mark[0.045]; Norway[0.044]; Jamaica[0.044]; Jamaica[0.044]; Jarrett[0.044]; Johnson[0.044]; Michael[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Switzerland[0.043]; Jon[0.043]; Belgium[0.042]; Tony[0.042]; seconds[0.042]; seconds[0.042]; seconds[0.042]; Russell[0.042]; Patrick[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mike Conley\u001b[39m ==> ENTITY: \u001b[32mMike Conley, Sr.\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.094:0.094[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \njump[0.044]; Trine[0.044]; triple[0.043]; Michael[0.043]; Michael[0.043]; Canada[0.043]; Harrison[0.042]; Daniel[0.042]; javelin[0.042]; Paul[0.042]; Johnson[0.042]; Tanja[0.042]; Kenny[0.042]; relay[0.042]; Jesse[0.041]; seconds[0.041]; Isel[0.041]; Charles[0.041]; McPaul[0.041]; Australia[0.041]; Jonathan[0.040]; Donovan[0.040]; Fita[0.040]; race[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.223:0.223[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.045]; Russia[0.045]; Russia[0.045]; Russia[0.045]; Russia[0.045]; Russia[0.045]; Lithuania[0.043]; Ukraine[0.043]; Bulgaria[0.042]; Andrei[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Denmark[0.040]; Norway[0.040]; Norway[0.040]; Yelena[0.040]; Britain[0.040]; Natalya[0.040]; Maksim[0.040]; Tarasov[0.040]; Kazakstan[0.039]; Tatyana[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.151:0.151[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Hungary[0.044]; Denmark[0.044]; Italy[0.043]; Britain[0.042]; Sweden[0.042]; Canada[0.042]; Norway[0.041]; Sven[0.040]; Algeria[0.040]; Dusan[0.040]; Donovan[0.040]; Men[0.040]; Men[0.040]; Australia[0.039]; Eric[0.039]; Nigeria[0.039]; Nigeria[0.039]; Nigeria[0.039]; Nigeria[0.039]; Thomas[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.059:0.059[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.048]; Men[0.046]; Finland[0.044]; Privalova[0.044]; Italy[0.043]; Barsosio[0.043]; Sven[0.042]; Women[0.042]; Women[0.042]; Romania[0.042]; Hungary[0.041]; Russia[0.041]; Eric[0.041]; Australia[0.040]; Kenya[0.040]; Kenya[0.040]; Kenya[0.040]; Kenya[0.040]; seconds[0.040]; seconds[0.040]; seconds[0.040]; seconds[0.040]; Thomas[0.040]; hurdles[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.044]; Trinidad[0.044]; Denmark[0.044]; Britain[0.043]; Russia[0.042]; Russia[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Norway[0.041]; Johnson[0.041]; Netherlands[0.041]; William[0.041]; Algeria[0.041]; Cuba[0.041]; Cuba[0.041]; Men[0.040]; Men[0.040]; Men[0.040]; John[0.040]; Michael[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Michelle Freeman\u001b[39m ==> ENTITY: \u001b[32mMichelle Freeman\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.101:0.101[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nathletics[0.047]; hurdles[0.047]; Athletics[0.047]; Dionne[0.045]; Thorsett[0.044]; Engquist[0.044]; seconds[0.044]; seconds[0.044]; Rose[0.043]; Women[0.043]; Women[0.043]; Canada[0.043]; Jamaica[0.043]; Jamaica[0.043]; Jamaica[0.043]; Leah[0.043]; Germany[0.042]; Grand[0.042]; Grand[0.042]; Brigita[0.042]; Baumann[0.042]; Slovenia[0.042]; minutes[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Inger Miller\u001b[39m ==> ENTITY: \u001b[32mInger Miller\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.083:0.083[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMatete[0.048]; Juergen[0.048]; Lars[0.047]; Alekna[0.046]; Chryste[0.046]; Onyali[0.046]; Andreas[0.046]; Merlene[0.046]; Barsosio[0.045]; Gunhild[0.045]; Devers[0.045]; Kenya[0.045]; Kenya[0.045]; Kenya[0.045]; Kenya[0.045]; Gete[0.045]; Torrance[0.045]; Ottey[0.045]; Norway[0.044]; Privalova[0.044]; Jamaica[0.044]; Halle[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kenya ==> ENTITY: \u001b[32mKenya\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKenya[0.062]; Burundi[0.048]; Rotich[0.046]; Tanui[0.045]; Algeria[0.045]; Ato[0.044]; Netherlands[0.044]; Denmark[0.044]; Norway[0.043]; Russia[0.043]; Russia[0.043]; Belarus[0.042]; Belarus[0.042]; Lithuania[0.042]; Niyongabo[0.042]; Britain[0.042]; Morceli[0.042]; Cuba[0.041]; Trinidad[0.041]; Belgium[0.040]; Ireland[0.040]; Spain[0.039]; Devers[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kenya ==> ENTITY: \u001b[32mKenya\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.182:0.182[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKenya[0.057]; Kenya[0.057]; Kenya[0.057]; Ethiopia[0.048]; Morocco[0.042]; Morocco[0.042]; Koech[0.041]; Australia[0.040]; Norway[0.040]; Norway[0.040]; Norway[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Bulgaria[0.039]; Lithuania[0.039]; Finland[0.038]; Britain[0.038]; Portugal[0.038]; Ukraine[0.038]; Cuba[0.037]; Cuba[0.037]; Komen[0.037]; Armenia[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Samuel Matete\u001b[39m ==> ENTITY: \u001b[32mSamuel Matete\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhurdles[0.045]; Barsosio[0.044]; Zambia[0.044]; Kenya[0.043]; Kenya[0.043]; Kenya[0.043]; Kenya[0.043]; Tegla[0.043]; Loroupe[0.042]; Sturrup[0.042]; seconds[0.042]; seconds[0.042]; seconds[0.042]; Ethiopia[0.042]; Nigeria[0.040]; Nigeria[0.040]; Nigeria[0.040]; Wami[0.040]; Australia[0.040]; Anja[0.040]; Ogunkoya[0.040]; Jearl[0.040]; Cheruiyot[0.040]; Privalova[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Florian Schwarthoff\u001b[39m ==> ENTITY: \u001b[32mFlorian Schwarthoff\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.049]; Germany[0.049]; Germany[0.049]; Germany[0.049]; Germany[0.049]; Johnson[0.047]; Trinidad[0.046]; Claus[0.045]; Switzerland[0.045]; Denmark[0.045]; Norway[0.045]; Baumann[0.045]; hurdles[0.044]; Fredericks[0.044]; Geir[0.044]; Canada[0.044]; Busemann[0.044]; Jon[0.044]; Russell[0.044]; Thorsett[0.044]; Cuba[0.043]; Cuba[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.082:0.082[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMen[0.052]; Men[0.052]; Germany[0.051]; Women[0.047]; Lithuania[0.047]; Portugal[0.046]; Fita[0.046]; javelin[0.046]; Britain[0.045]; Britain[0.045]; relay[0.045]; Norway[0.045]; Norway[0.045]; Finland[0.044]; Johnson[0.043]; jump[0.043]; Donovan[0.043]; Belarus[0.043]; Canada[0.043]; Australia[0.043]; seconds[0.043]; race[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.048]; Germany[0.048]; Lithuania[0.044]; France[0.043]; Britain[0.042]; Norway[0.042]; Norway[0.042]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Bulgaria[0.041]; Algeria[0.041]; jump[0.041]; Belarus[0.041]; Ukraine[0.040]; Men[0.040]; Men[0.040]; Men[0.040]; Maksim[0.040]; minutes[0.040]; Markov[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.094:0.094[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; Jamaica[0.046]; Denmark[0.045]; Sweden[0.044]; Mitchell[0.044]; Women[0.044]; Kenya[0.044]; Kenya[0.044]; Craig[0.044]; Norway[0.043]; Norway[0.043]; Donovan[0.043]; David[0.043]; Cuba[0.043]; Marc[0.043]; Britain[0.043]; Britain[0.043]; Michael[0.043]; Algeria[0.042]; Hungary[0.042]; Men[0.042]; Men[0.042]; Men[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kenya ==> ENTITY: \u001b[32mKenya\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.217:0.217[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKenya[0.055]; Kenya[0.055]; Kenya[0.055]; Ethiopia[0.047]; Zambia[0.045]; Nigeria[0.042]; Nigeria[0.042]; Nigeria[0.042]; Nigeria[0.042]; Tegla[0.042]; Loroupe[0.040]; Cheruiyot[0.040]; Barsosio[0.039]; Australia[0.039]; Norway[0.039]; Russia[0.038]; Wami[0.038]; Sweden[0.037]; Konga[0.037]; Yusuf[0.037]; Finland[0.037]; Britain[0.037]; Matete[0.037]; Italy[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ivan Garcia\u001b[39m ==> ENTITY: \u001b[32mIván García (athlete)\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTrinidad[0.045]; Spain[0.044]; Cuba[0.043]; Cuba[0.043]; Rotich[0.043]; Morceli[0.043]; Women[0.042]; seconds[0.042]; seconds[0.042]; seconds[0.042]; Geir[0.042]; Florian[0.041]; Busemann[0.041]; Men[0.041]; Men[0.041]; Fredericks[0.041]; Mayock[0.041]; Valle[0.041]; Britain[0.041]; Britain[0.041]; Norway[0.040]; Ireland[0.040]; Tanui[0.040]; mile[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hungary ==> ENTITY: \u001b[32mHungary\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.326:-0.326[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMen[0.046]; Men[0.046]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Women[0.044]; Italy[0.044]; Canada[0.043]; Britain[0.043]; Sweden[0.042]; Norway[0.042]; Norway[0.042]; Jamaica[0.041]; Barsosio[0.040]; Finland[0.040]; Marc[0.040]; Australia[0.040]; Eric[0.040]; Michael[0.040]; Ethiopia[0.039]; hurdles[0.039]; Dusan[0.039]; Derrick[0.039]; Sven[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cuba ==> ENTITY: \u001b[32mCuba\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.072:0.072[0]; log p(e|m)= -0.240:-0.240[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJamaica[0.045]; Jamaica[0.045]; Jamaica[0.045]; Men[0.045]; Switzerland[0.044]; Canada[0.043]; Slovenia[0.043]; Athletics[0.043]; athletics[0.042]; Portugal[0.042]; Germany[0.042]; Russia[0.041]; Russia[0.041]; Russia[0.041]; France[0.041]; Sweden[0.041]; Lopez[0.041]; seconds[0.040]; seconds[0.040]; Women[0.040]; Women[0.040]; Berlin[0.039]; Berlin[0.039]; Berlin[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rose Cheruiyot\u001b[39m ==> ENTITY: \u001b[32mRose Cheruiyot\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.121:0.121[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKenya[0.049]; Kenya[0.049]; Kenya[0.049]; Kenya[0.049]; Matete[0.047]; Wami[0.046]; Ethiopia[0.046]; Juergen[0.046]; Devers[0.046]; Barsosio[0.046]; Michael[0.045]; Chryste[0.045]; Onyali[0.045]; Tegla[0.045]; Andreas[0.045]; Merlene[0.044]; Gete[0.044]; Gunhild[0.044]; Norway[0.043]; hurdles[0.043]; Privalova[0.043]; Halle[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.087:0.087[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.049]; Britain[0.046]; Australia[0.046]; Lithuania[0.045]; Sweden[0.043]; Canada[0.043]; Jamaica[0.042]; Finland[0.042]; Johnson[0.042]; Donovan[0.041]; Belarus[0.041]; Nigeria[0.041]; Nigeria[0.041]; Nigeria[0.041]; Namibia[0.041]; Greece[0.040]; Michael[0.040]; Michael[0.040]; race[0.040]; Men[0.039]; Florian[0.039]; Cuba[0.039]; Falk[0.039]; Davidson[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Armenia ==> ENTITY: \u001b[32mArmenia\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.114:0.114[0]; log p(e|m)= -0.431:-0.431[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.046]; Armen[0.046]; Lithuania[0.044]; Ethiopia[0.043]; jump[0.043]; Britain[0.042]; Britain[0.042]; Carlos[0.042]; Germany[0.042]; Germany[0.042]; Natalya[0.041]; Cuba[0.041]; Cuba[0.041]; Cuba[0.041]; Morocco[0.041]; Morocco[0.041]; Portugal[0.041]; Fita[0.040]; Lopez[0.040]; Donovan[0.040]; Johnson[0.040]; Norway[0.040]; Norway[0.040]; Finland[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.071:0.071[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.044]; Belarus[0.044]; Britain[0.044]; Russia[0.044]; Washington[0.043]; Devers[0.043]; Finland[0.042]; Jamaica[0.042]; Norway[0.042]; Miller[0.042]; Ireland[0.042]; Men[0.041]; Men[0.041]; John[0.041]; Michael[0.041]; Lithuania[0.041]; Sullivan[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Privalova[0.040]; Bahamas[0.040]; discus[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belarus ==> ENTITY: \u001b[32mBelarus\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.118:0.118[0]; log p(e|m)= -0.422:-0.422[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.045]; Lithuania[0.045]; Men[0.043]; Men[0.043]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Sweden[0.042]; Finland[0.042]; Greece[0.041]; Armenia[0.041]; Norway[0.041]; Norway[0.041]; Donovan[0.041]; Britain[0.041]; Britain[0.041]; Portugal[0.041]; Johnson[0.041]; Jamaica[0.041]; Women[0.040]; Cuba[0.040]; Cuba[0.040]; Cuba[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Svetlana Masterkova\u001b[39m ==> ENTITY: \u001b[32mSvetlana Masterkova\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYekaterina[0.051]; athletics[0.048]; Ludmila[0.047]; seconds[0.047]; seconds[0.047]; seconds[0.047]; Russia[0.047]; Russia[0.047]; Russia[0.047]; Athletics[0.046]; Florian[0.045]; Margarita[0.045]; Thorsett[0.045]; Engquist[0.045]; hurdles[0.043]; hurdles[0.043]; Berlin[0.043]; Berlin[0.043]; Berlin[0.043]; minutes[0.043]; Women[0.043]; Women[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cuba ==> ENTITY: \u001b[32mCuba\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.240:-0.240[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJamaica[0.045]; Men[0.044]; Men[0.044]; Men[0.044]; Donovan[0.044]; Canada[0.042]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Denmark[0.041]; Algeria[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Belarus[0.041]; France[0.041]; Britain[0.040]; Britain[0.040]; Norway[0.040]; Norway[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nigeria ==> ENTITY: \u001b[32mNigeria\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.129:0.129[0]; log p(e|m)= -0.315:-0.315[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNigeria[0.047]; Nigeria[0.047]; Men[0.046]; javelin[0.045]; Women[0.044]; Australia[0.043]; Lithuania[0.042]; relay[0.042]; seconds[0.042]; Britain[0.041]; Jamaica[0.041]; Portugal[0.041]; Sweden[0.041]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Johnson[0.040]; Norway[0.040]; Canada[0.040]; Belarus[0.039]; Donovan[0.039]; Finland[0.039]; George[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bob Kennedy ==> ENTITY: \u001b[32mBob Kennedy (athlete)\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.077:0.077[0]; log p(e|m)= -0.607:-0.607[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHarrison[0.045]; Paul[0.045]; Michael[0.043]; Mike[0.043]; Nele[0.042]; Tom[0.042]; Carlos[0.042]; Tim[0.041]; seconds[0.041]; Charles[0.041]; Lobinger[0.041]; Edwards[0.041]; Daniel[0.041]; Jonathan[0.041]; Stefka[0.041]; jump[0.041]; jump[0.041]; Hanne[0.041]; high[0.041]; Kenny[0.041]; Haugland[0.040]; triple[0.040]; Kostadinova[0.040]; Fita[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Juergen Schult\u001b[39m ==> ENTITY: \u001b[32mJürgen Schult\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.142:0.142[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAndreas[0.050]; Germany[0.048]; Germany[0.048]; Germany[0.048]; Germany[0.048]; Alekna[0.047]; Lars[0.046]; discus[0.046]; Chryste[0.045]; Onyali[0.045]; Privalova[0.045]; Merlene[0.045]; Vladimir[0.045]; Mayock[0.044]; Belarus[0.044]; Belarus[0.044]; Gete[0.044]; Irina[0.044]; Marko[0.043]; Men[0.043]; Men[0.043]; Tanui[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.178:0.178[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Denmark[0.045]; Spain[0.043]; Belgium[0.043]; Netherlands[0.043]; Ireland[0.043]; Norway[0.042]; Britain[0.041]; Algeria[0.039]; Johnson[0.039]; Namibia[0.039]; Russia[0.039]; Russia[0.039]; minutes[0.038]; Marko[0.038]; Trinidad[0.038]; Michael[0.038]; Kenya[0.038]; Kenya[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Igor Trandenkov\u001b[39m ==> ENTITY: \u001b[32mIgor Trandenkov\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIgor[0.046]; pole[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Dmitri[0.043]; Andrei[0.042]; Pyotr[0.042]; Natalya[0.041]; Nele[0.041]; Cuba[0.041]; Tatyana[0.041]; jump[0.041]; Tarasov[0.041]; Yelena[0.041]; Lobinger[0.040]; Ukraine[0.040]; Stefka[0.040]; Geir[0.040]; Hanne[0.040]; Belarus[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.046]; Switzerland[0.044]; Trinidad[0.044]; Britain[0.043]; Sweden[0.043]; Russia[0.042]; Russia[0.042]; Russia[0.042]; Slovenia[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Portugal[0.042]; Norway[0.041]; Jamaica[0.041]; Jamaica[0.041]; Jamaica[0.041]; Johnson[0.041]; Cuba[0.040]; Cuba[0.040]; Michael[0.040]; Belgium[0.039]; Mark[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alina Astafei\u001b[39m ==> ENTITY: \u001b[32mAlina Astafei\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.119:0.119[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \njump[0.046]; Kostadinova[0.045]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Nele[0.043]; Bulgaria[0.042]; Lobinger[0.042]; high[0.042]; Stefka[0.042]; Hanne[0.042]; Igor[0.041]; Igor[0.041]; Haugland[0.041]; Lithuania[0.041]; Dmitri[0.041]; Rodal[0.041]; Andrei[0.041]; Fita[0.040]; Yelena[0.040]; seconds[0.040]; Norway[0.040]; Norway[0.040]; Belarus[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Donovan Bailey\u001b[39m ==> ENTITY: \u001b[32mDonovan Bailey\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.113:0.113[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKipketer[0.046]; seconds[0.046]; Canada[0.046]; Jamaica[0.046]; Michael[0.045]; David[0.045]; Craig[0.044]; Anja[0.044]; Eric[0.044]; Jearl[0.043]; Geir[0.043]; Jon[0.043]; Marc[0.043]; Dennis[0.043]; Kenya[0.042]; Kenya[0.042]; Thomas[0.042]; Mitchell[0.042]; Rodal[0.042]; Wilson[0.042]; Davidson[0.042]; Smith[0.042]; Ruecker[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bahamas ==> ENTITY: \u001b[32mThe Bahamas\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.024:0.024[0]; log p(e|m)= -0.351:-0.351[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJamaica[0.045]; Ireland[0.044]; Sturrup[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Ottey[0.043]; Devers[0.042]; seconds[0.042]; seconds[0.042]; seconds[0.042]; Privalova[0.042]; Alekna[0.042]; Michael[0.042]; Barsosio[0.041]; Derrick[0.041]; Onyali[0.040]; Anthony[0.040]; Vladimir[0.040]; Samuel[0.040]; hurdles[0.040]; Norway[0.039]; Zambia[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.104:0.104[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.046]; Trinidad[0.044]; Denmark[0.044]; Britain[0.043]; Russia[0.042]; Russia[0.042]; Russia[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Norway[0.041]; Johnson[0.041]; William[0.041]; Algeria[0.041]; Cuba[0.040]; Cuba[0.040]; Michael[0.040]; Belgium[0.040]; Mark[0.039]; Brown[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gail Devers\u001b[39m ==> ENTITY: \u001b[32mGail Devers\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.065:0.065[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSturrup[0.051]; Juergen[0.050]; Ottey[0.049]; Lars[0.049]; discus[0.049]; Alekna[0.048]; Chryste[0.048]; Torrence[0.048]; Onyali[0.048]; Andreas[0.048]; Privalova[0.047]; Merlene[0.047]; Mayock[0.047]; Gunhild[0.047]; seconds[0.047]; seconds[0.047]; Gete[0.047]; Norway[0.046]; Anthony[0.046]; Halle[0.046]; Washington[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Armen Martirosyan ==> ENTITY: \u001b[32mArmen Martirosyan (athlete)\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.154:0.154[0]; log p(e|m)= -0.304:-0.304[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \njump[0.049]; Armenia[0.048]; triple[0.047]; Trine[0.045]; javelin[0.045]; Belarus[0.045]; Lithuania[0.044]; Ethiopia[0.043]; Tanja[0.043]; Portugal[0.043]; Natalya[0.043]; Russia[0.043]; Fita[0.043]; Britain[0.042]; Britain[0.042]; Canada[0.042]; Linford[0.042]; relay[0.042]; Isel[0.042]; McPaul[0.042]; Rivero[0.042]; Morocco[0.042]; Morocco[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belarus ==> ENTITY: \u001b[32mBelarus\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.182:0.182[0]; log p(e|m)= -0.422:-0.422[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMen[0.047]; Men[0.047]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Lithuania[0.043]; Bulgaria[0.042]; Ukraine[0.042]; Women[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Norway[0.040]; Norway[0.040]; Britain[0.040]; France[0.040]; Algeria[0.040]; Cuba[0.039]; Tarasov[0.039]; Markov[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lars Riedel\u001b[39m ==> ENTITY: \u001b[32mLars Riedel\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAndreas[0.044]; Juergen[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; discus[0.043]; Privalova[0.042]; Alekna[0.042]; Chryste[0.042]; Onyali[0.042]; seconds[0.041]; seconds[0.041]; Merlene[0.041]; Mayock[0.041]; Marcus[0.040]; Tanui[0.040]; Rotich[0.040]; Men[0.039]; Men[0.039]; Astrid[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kenya ==> ENTITY: \u001b[32mKenya\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.168:0.168[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKenya[0.057]; Kenya[0.057]; Kenya[0.057]; Ethiopia[0.048]; Morocco[0.042]; Morocco[0.042]; Koech[0.041]; Norway[0.040]; Norway[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Bulgaria[0.039]; Belarus[0.039]; Lithuania[0.039]; Britain[0.038]; Ukraine[0.037]; Cuba[0.037]; Komen[0.037]; Armenia[0.036]; Shem[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Norberto Tellez\u001b[39m ==> ENTITY: \u001b[32mNorberto Téllez\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.040:0.040[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCuba[0.044]; Anja[0.044]; Lobinger[0.043]; Geir[0.043]; vault[0.043]; Rodal[0.042]; Canada[0.042]; Kipketer[0.042]; Ruecker[0.042]; Norway[0.041]; Norway[0.041]; Karin[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Men[0.041]; Men[0.041]; Men[0.041]; Nico[0.041]; Kiptoo[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cuba ==> ENTITY: \u001b[32mCuba\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.125:0.125[0]; log p(e|m)= -0.240:-0.240[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCuba[0.049]; Trinidad[0.045]; Jamaica[0.044]; Men[0.044]; Men[0.044]; Canada[0.042]; Portugal[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Denmark[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; France[0.041]; Britain[0.040]; Belgium[0.040]; Norway[0.040]; Margarita[0.040]; Garcia[0.039]; Michael[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Laban Rotich\u001b[39m ==> ENTITY: \u001b[32mLaban Rotich\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseconds[0.058]; seconds[0.058]; Juergen[0.057]; Kenya[0.057]; Kenya[0.057]; mile[0.057]; Lars[0.056]; discus[0.056]; minutes[0.056]; Alekna[0.055]; Geir[0.055]; Andreas[0.055]; Merlene[0.054]; Mayock[0.054]; Tanui[0.054]; Women[0.054]; Women[0.054]; Norway[0.053]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ethiopia ==> ENTITY: \u001b[32mEthiopia\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.236:-0.236[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKenya[0.047]; Kenya[0.047]; Kenya[0.047]; Kenya[0.047]; Men[0.046]; Men[0.046]; Australia[0.046]; Morocco[0.044]; Morocco[0.044]; Women[0.044]; Britain[0.043]; Lithuania[0.043]; Fita[0.042]; Russia[0.042]; Russia[0.042]; Russia[0.042]; Norway[0.042]; Norway[0.042]; Norway[0.042]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Armenia[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.226:0.226[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.045]; Russia[0.045]; Russia[0.045]; Russia[0.045]; Russia[0.045]; Russia[0.045]; Lithuania[0.043]; Ukraine[0.043]; Denmark[0.042]; Bulgaria[0.042]; Andrei[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Norway[0.040]; Norway[0.040]; Norway[0.040]; Britain[0.040]; Tarasov[0.040]; Kazakstan[0.039]; Markov[0.039]; France[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Anja Ruecker\u001b[39m ==> ENTITY: \u001b[32mAnja Rücker\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.116:0.116[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBarsosio[0.044]; Matete[0.044]; seconds[0.044]; seconds[0.044]; Kipketer[0.043]; Geir[0.042]; Linda[0.042]; hurdles[0.042]; Men[0.042]; Men[0.042]; Men[0.042]; Jearl[0.042]; Sally[0.041]; Women[0.041]; Norway[0.041]; Norway[0.041]; Karin[0.041]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Nylander[0.040]; Eric[0.040]; Dusan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = George Panayiotopoulos\u001b[39m ==> ENTITY: \u001b[32mGeorgios Panagiotopoulos\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.129:0.129[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGreece[0.049]; Jamaica[0.044]; seconds[0.044]; Fredericks[0.042]; Michael[0.042]; Michael[0.042]; Florian[0.042]; Lithuania[0.042]; Isel[0.042]; Finland[0.042]; McPaul[0.042]; relay[0.042]; Germany[0.041]; Germany[0.041]; Britain[0.041]; Donovan[0.041]; Norway[0.041]; Men[0.040]; Sweden[0.040]; Australia[0.040]; Belarus[0.040]; Heli[0.040]; Christie[0.040]; Peter[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.047]; Russia[0.046]; Steve[0.044]; Denmark[0.043]; Johnson[0.042]; Ireland[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Michael[0.042]; Norway[0.042]; Spain[0.041]; Britain[0.041]; Belgium[0.041]; Netherlands[0.041]; Irina[0.040]; William[0.040]; John[0.040]; Sullivan[0.040]; Algeria[0.040]; Jack[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.046]; Donovan[0.044]; Denmark[0.044]; Britain[0.043]; Britain[0.043]; Russia[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Norway[0.041]; Norway[0.041]; Jamaica[0.041]; Hungary[0.041]; Algeria[0.041]; Cuba[0.040]; Thomas[0.040]; Men[0.040]; Men[0.040]; Men[0.040]; Australia[0.040]; David[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Torrance Zellner\u001b[39m ==> ENTITY: \u001b[32mTorrance Zellner\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.064:0.064[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhurdles[0.046]; seconds[0.044]; seconds[0.044]; seconds[0.044]; Matete[0.044]; Barsosio[0.043]; Anja[0.043]; Privalova[0.042]; Chryste[0.042]; Onyali[0.042]; Gunhild[0.041]; Gete[0.041]; Gaines[0.041]; Jearl[0.041]; Ruecker[0.040]; Norway[0.040]; Sturrup[0.040]; Karin[0.040]; Halle[0.040]; Australia[0.040]; Germany[0.040]; Germany[0.040]; Nylander[0.040]; Men[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Michael Green ==> ENTITY: \u001b[32mMichael Green (sprinter)\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.090:0.090[0]; log p(e|m)= -2.033:-2.033[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMichael[0.052]; Michael[0.052]; Peter[0.047]; George[0.044]; Charles[0.044]; Lithuania[0.042]; Jamaica[0.041]; Trine[0.041]; Canada[0.041]; Linford[0.041]; Donovan[0.041]; Carlos[0.040]; Johnson[0.040]; Men[0.040]; Jesse[0.040]; Bailey[0.040]; Women[0.040]; Sweden[0.040]; javelin[0.040]; Davidson[0.039]; Tanja[0.039]; Australia[0.039]; seconds[0.039]; Florian[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Romania ==> ENTITY: \u001b[32mRomania\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.106:0.106[0]; log p(e|m)= -0.212:-0.212[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMen[0.045]; Women[0.045]; Women[0.045]; Italy[0.044]; Lithuania[0.044]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Belarus[0.042]; Belarus[0.042]; Russia[0.042]; Privalova[0.042]; Barsosio[0.042]; Finland[0.041]; Sweden[0.041]; Norway[0.040]; Jamaica[0.040]; Alekna[0.040]; minutes[0.039]; seconds[0.039]; seconds[0.039]; seconds[0.039]; Bahamas[0.039]; Nigeria[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.113:0.113[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.050]; Britain[0.048]; Britain[0.048]; Men[0.046]; Men[0.046]; Germany[0.045]; Germany[0.045]; Finland[0.044]; Canada[0.043]; Portugal[0.043]; Morocco[0.042]; Morocco[0.042]; race[0.042]; Fita[0.042]; Women[0.042]; relay[0.042]; Lithuania[0.042]; Belarus[0.042]; Australia[0.041]; Kenya[0.041]; Kenya[0.041]; Kenya[0.041]; Kenya[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Inga Babakova\u001b[39m ==> ENTITY: \u001b[32mInha Babakova\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUkraine[0.044]; Belarus[0.043]; jump[0.043]; Kazakstan[0.043]; Natalya[0.043]; Kostadinova[0.043]; high[0.043]; Maksim[0.043]; Tatyana[0.042]; Russia[0.042]; Russia[0.042]; Russia[0.042]; Russia[0.042]; Russia[0.042]; Russia[0.042]; Nele[0.041]; Lithuania[0.041]; vault[0.041]; Bulgaria[0.041]; Lobinger[0.040]; Stefka[0.040]; Hanne[0.040]; Yelena[0.040]; seconds[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mary Onyali\u001b[39m ==> ENTITY: \u001b[32mMary Onyali-Omagbemi\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNigeria[0.043]; discus[0.043]; Juergen[0.043]; Loroupe[0.042]; Sturrup[0.042]; Lars[0.042]; Cheruiyot[0.042]; Jamaica[0.042]; Wami[0.042]; Alekna[0.042]; Chryste[0.042]; Barsosio[0.042]; Kenya[0.041]; Kenya[0.041]; Kenya[0.041]; Kenya[0.041]; Andreas[0.041]; hurdles[0.041]; Devers[0.041]; Merlene[0.041]; Women[0.041]; Women[0.041]; Mayock[0.041]; Gunhild[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Stephanie Storp\u001b[39m ==> ENTITY: \u001b[32mStephanie Storp\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndiscus[0.045]; Alekna[0.045]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Astrid[0.043]; seconds[0.042]; seconds[0.042]; Lars[0.042]; Irina[0.041]; Geir[0.041]; Valentina[0.041]; Busemann[0.041]; Claudia[0.040]; Belarus[0.040]; Belarus[0.040]; Mayock[0.040]; shot[0.040]; Norway[0.040]; Russia[0.039]; Russia[0.039]; Morceli[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lithuania ==> ENTITY: \u001b[32mLithuania\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.397:-0.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBulgaria[0.045]; Russia[0.044]; Russia[0.044]; Russia[0.044]; Russia[0.044]; Russia[0.044]; Russia[0.044]; Ukraine[0.043]; Belarus[0.043]; Cuba[0.042]; Norway[0.042]; Men[0.042]; Men[0.042]; Men[0.042]; Kazakstan[0.040]; Natalya[0.040]; Britain[0.040]; Britain[0.040]; Women[0.039]; Maksim[0.039]; Germany[0.039]; Germany[0.039]; Germany[0.039]; Dmitri[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.129:0.129[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.048]; Finland[0.047]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Canada[0.044]; Britain[0.043]; Belarus[0.043]; Lithuania[0.042]; Nigeria[0.042]; Nigeria[0.042]; Nigeria[0.042]; Donovan[0.041]; Greece[0.040]; Australia[0.040]; Karlsson[0.040]; Men[0.040]; seconds[0.039]; Cuba[0.039]; Cuba[0.039]; Johnson[0.039]; Michael[0.039]; Michael[0.039]; Jamaica[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tony Jarrett\u001b[39m ==> ENTITY: \u001b[32mTony Jarrett\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhurdles[0.049]; Russell[0.046]; Johnson[0.046]; Trinidad[0.045]; Jamaica[0.045]; Jamaica[0.045]; Michael[0.044]; seconds[0.044]; seconds[0.044]; seconds[0.044]; Geir[0.043]; Boldon[0.042]; Britain[0.042]; Denmark[0.042]; Florian[0.042]; Busemann[0.042]; Switzerland[0.042]; Thorsett[0.042]; Mark[0.042]; Stevens[0.042]; shot[0.042]; Pierce[0.042]; Jack[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fita Bayissa\u001b[39m ==> ENTITY: \u001b[32mFita Bayisa\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.111:0.111[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEthiopia[0.046]; Trine[0.044]; Nele[0.043]; Kenya[0.043]; Kenya[0.043]; Kenya[0.043]; Kenya[0.043]; javelin[0.042]; Tanja[0.042]; Hanne[0.042]; jump[0.042]; Koech[0.041]; Isel[0.041]; McPaul[0.041]; Haugland[0.041]; Cuba[0.041]; Cuba[0.041]; Cuba[0.041]; Morocco[0.041]; Morocco[0.041]; Norway[0.040]; Norway[0.040]; Norway[0.040]; Shem[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Maksim Tarasov\u001b[39m ==> ENTITY: \u001b[32mMaksim Tarasov\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.083:0.083[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.043]; jump[0.043]; Nele[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; pole[0.042]; Natalya[0.042]; Lobinger[0.042]; Stefka[0.042]; Geir[0.042]; Hanne[0.042]; Haugland[0.041]; Ukraine[0.041]; Rodal[0.041]; Kostadinova[0.041]; Kazakstan[0.040]; Norway[0.040]; Norway[0.040]; Norway[0.040]; Pyotr[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nigeria ==> ENTITY: \u001b[32mNigeria\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.165:0.165[0]; log p(e|m)= -0.315:-0.315[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNigeria[0.045]; Nigeria[0.045]; Nigeria[0.045]; Men[0.045]; Men[0.045]; Men[0.045]; Women[0.043]; hurdles[0.042]; Australia[0.042]; Kenya[0.042]; Kenya[0.042]; Denmark[0.041]; Barsosio[0.040]; Italy[0.040]; Britain[0.040]; Jamaica[0.040]; Sweden[0.040]; Afolabi[0.040]; Germany[0.039]; Germany[0.039]; Germany[0.039]; Germany[0.039]; Norway[0.039]; Norway[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Finland ==> ENTITY: \u001b[32mFinland\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.245:-0.245[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.050]; Norway[0.048]; Hungary[0.044]; Germany[0.044]; Germany[0.044]; Romania[0.043]; Italy[0.043]; Men[0.042]; Russia[0.042]; Andreas[0.041]; Sven[0.041]; Dusan[0.041]; Women[0.041]; Women[0.041]; Women[0.041]; Jamaica[0.040]; Privalova[0.040]; Bahamas[0.040]; Sandell[0.039]; Eric[0.039]; Sturrup[0.039]; Barsosio[0.039]; minutes[0.039]; Michael[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgium ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.242:-0.242[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.046]; Netherlands[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Britain[0.045]; Spain[0.044]; Norway[0.042]; Algeria[0.041]; Fredericks[0.040]; Marko[0.040]; minutes[0.040]; Men[0.039]; Men[0.039]; Men[0.039]; Burundi[0.039]; Johnson[0.039]; seconds[0.039]; seconds[0.039]; seconds[0.039]; Claus[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Irina Privalova\u001b[39m ==> ENTITY: \u001b[32mIrina Privalova\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.117:0.117[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMatete[0.044]; Juergen[0.043]; Lars[0.043]; discus[0.043]; Merlene[0.042]; Alekna[0.042]; Chryste[0.042]; Onyali[0.042]; Barsosio[0.042]; hurdles[0.042]; Andreas[0.042]; Russia[0.042]; Ottey[0.042]; Finland[0.041]; Gunhild[0.041]; seconds[0.041]; seconds[0.041]; seconds[0.041]; Gete[0.041]; Devers[0.041]; Norway[0.040]; Sally[0.040]; Belarus[0.040]; Belarus[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Linford Christie\u001b[39m ==> ENTITY: \u001b[32mLinford Christie\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.099:0.099[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJohnson[0.046]; Michael[0.045]; Michael[0.045]; Michael[0.045]; Trine[0.043]; race[0.042]; Britain[0.042]; George[0.042]; javelin[0.042]; Peter[0.042]; Tanja[0.042]; Bailey[0.041]; Jamaica[0.041]; Fredericks[0.041]; relay[0.041]; Florian[0.041]; Isel[0.041]; McPaul[0.040]; Charles[0.040]; Davidson[0.040]; Canada[0.040]; Norway[0.040]; Norway[0.040]; Louise[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kenya ==> ENTITY: \u001b[32mKenya\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.176:0.176[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKenya[0.057]; Kenya[0.057]; Kenya[0.057]; Ethiopia[0.048]; Morocco[0.042]; Morocco[0.042]; Koech[0.041]; Norway[0.040]; Norway[0.040]; Norway[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Bulgaria[0.039]; Belarus[0.039]; Lithuania[0.039]; Britain[0.038]; Portugal[0.038]; Ukraine[0.038]; Cuba[0.038]; Cuba[0.038]; Komen[0.037]; Armenia[0.037]; Shem[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMen[0.051]; Germany[0.050]; Germany[0.050]; Germany[0.050]; Spain[0.047]; Romania[0.047]; Women[0.046]; Women[0.046]; Lithuania[0.046]; Privalova[0.046]; Ireland[0.045]; Britain[0.045]; Netherlands[0.044]; Andreas[0.044]; discus[0.043]; Finland[0.043]; Algeria[0.043]; Juergen[0.043]; Belarus[0.043]; Belarus[0.043]; Miller[0.043]; Russia[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sigurd Njerve\u001b[39m ==> ENTITY: \u001b[32mSigurd Njerve\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.086:0.086[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTrine[0.047]; jump[0.047]; Norway[0.046]; Norway[0.046]; javelin[0.045]; Tanja[0.045]; triple[0.044]; Isel[0.044]; McPaul[0.044]; relay[0.043]; Fita[0.043]; Lithuania[0.043]; Shem[0.043]; Kenny[0.042]; Women[0.042]; Germany[0.042]; Germany[0.042]; Men[0.042]; Men[0.042]; Daniel[0.042]; Quesada[0.042]; Australia[0.042]; Heli[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jamaica ==> ENTITY: \u001b[32mJamaica\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.104:0.104[0]; log p(e|m)= -0.333:-0.333[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJamaica[0.050]; Jamaica[0.050]; Britain[0.043]; seconds[0.043]; seconds[0.043]; seconds[0.043]; Canada[0.042]; Cuba[0.042]; Cuba[0.042]; Russell[0.042]; hurdles[0.041]; hurdles[0.041]; Athletics[0.041]; Switzerland[0.041]; Dionne[0.040]; Yekaterina[0.040]; Gillian[0.040]; Mark[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Portugal[0.040]; Sweden[0.039]; Slovenia[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.046]; Donovan[0.044]; Denmark[0.044]; Britain[0.043]; Britain[0.043]; Sweden[0.043]; Italy[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Norway[0.041]; Norway[0.041]; Jamaica[0.041]; Hungary[0.041]; Algeria[0.041]; Cuba[0.040]; Thomas[0.040]; Men[0.040]; Men[0.040]; Australia[0.040]; David[0.040]; Wilson[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bulgaria ==> ENTITY: \u001b[32mBulgaria\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.173:0.173[0]; log p(e|m)= -0.317:-0.317[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLithuania[0.044]; Belarus[0.044]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Ukraine[0.043]; Britain[0.042]; Dmitri[0.042]; Igor[0.042]; Igor[0.042]; Andrei[0.041]; Maksim[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Men[0.041]; Men[0.041]; Norway[0.040]; Norway[0.040]; Kostadinova[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.180:0.180[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Spain[0.044]; Netherlands[0.043]; Ireland[0.043]; Lithuania[0.042]; Britain[0.042]; Andreas[0.040]; Algeria[0.040]; Juergen[0.039]; Belarus[0.039]; Belarus[0.039]; Miller[0.039]; Russia[0.039]; Russia[0.039]; Russia[0.039]; minutes[0.038]; Nigeria[0.038]; Marko[0.038]; Anthony[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Denmark ==> ENTITY: \u001b[32mDenmark\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.157:0.157[0]; log p(e|m)= -0.268:-0.268[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.047]; Kenya[0.044]; Kenya[0.044]; Men[0.043]; Men[0.043]; Netherlands[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Britain[0.042]; Britain[0.042]; Women[0.042]; Belgium[0.042]; Trinidad[0.041]; Spain[0.040]; Algeria[0.040]; William[0.040]; John[0.039]; Namibia[0.039]; Johnson[0.039]; Geir[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Geir Moen\u001b[39m ==> ENTITY: \u001b[32mGeir Moen\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.045]; Norway[0.045]; Denmark[0.044]; Canada[0.043]; Anja[0.042]; Lobinger[0.042]; vault[0.042]; Kipketer[0.042]; Britain[0.041]; Britain[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Australia[0.041]; Rodal[0.041]; Kiptoo[0.041]; Ruecker[0.040]; Karin[0.040]; Mitchell[0.040]; seconds[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fabrizio Mori\u001b[39m ==> ENTITY: \u001b[32mFabrizio Mori\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.094:0.094[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMatete[0.048]; Italy[0.047]; hurdles[0.047]; Anja[0.047]; Jearl[0.047]; Barsosio[0.046]; Michael[0.046]; Jamaica[0.046]; Gunhild[0.045]; Hungary[0.045]; Gete[0.045]; Samuel[0.045]; seconds[0.045]; seconds[0.045]; seconds[0.045]; Ruecker[0.045]; Australia[0.045]; Norway[0.045]; Ogunkoya[0.045]; Karin[0.044]; Halle[0.044]; Romania[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gabriela Szabo\u001b[39m ==> ENTITY: \u001b[32mGabriela Szabo\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.065:0.065[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMatete[0.044]; Juergen[0.044]; Alekna[0.043]; Chryste[0.043]; Onyali[0.043]; Andreas[0.043]; Romania[0.042]; Merlene[0.042]; Barsosio[0.042]; Gunhild[0.042]; Devers[0.042]; Gete[0.041]; Norway[0.041]; hurdles[0.041]; Privalova[0.041]; Belarus[0.041]; Belarus[0.041]; Halle[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Nylander[0.041]; Men[0.041]; Eric[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Namibia ==> ENTITY: \u001b[32mNamibia\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.368:-0.368[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Australia[0.045]; Men[0.043]; seconds[0.042]; Canada[0.042]; javelin[0.041]; Sweden[0.041]; relay[0.041]; Finland[0.041]; Women[0.041]; Portugal[0.041]; Nigeria[0.041]; Nigeria[0.041]; Nigeria[0.041]; Fredericks[0.041]; Lithuania[0.040]; Norway[0.040]; Norway[0.040]; race[0.040]; Jamaica[0.040]; Britain[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.044]; Charles[0.043]; Norway[0.043]; Norway[0.043]; Norway[0.043]; Edwards[0.043]; Michael[0.042]; Germany[0.042]; Germany[0.042]; Kenya[0.042]; Kenya[0.042]; Kenya[0.042]; Kenya[0.042]; Jonathan[0.041]; Lithuania[0.041]; Lithuania[0.041]; seconds[0.041]; Paul[0.041]; Finland[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Tom[0.040]; Daniel[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.074:0.074[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.046]; Belarus[0.046]; Britain[0.046]; Russia[0.046]; Washington[0.045]; Devers[0.044]; Finland[0.044]; Jamaica[0.044]; Norway[0.043]; Miller[0.043]; Ireland[0.043]; Men[0.043]; John[0.043]; Michael[0.043]; Lithuania[0.043]; Sullivan[0.043]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Privalova[0.042]; Bahamas[0.041]; discus[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Isel Lopez\u001b[39m ==> ENTITY: \u001b[32mIsel López\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.102:0.102[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCuba[0.046]; Cuba[0.046]; Cuba[0.046]; javelin[0.044]; Trine[0.044]; Carlos[0.043]; Tanja[0.042]; Rivero[0.042]; McPaul[0.041]; Women[0.041]; Men[0.041]; Men[0.041]; Xiomara[0.041]; Fita[0.041]; Frankie[0.040]; Norway[0.040]; Norway[0.040]; relay[0.040]; Shem[0.040]; Kenny[0.040]; Morocco[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.110:0.110[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.048]; Britain[0.046]; Britain[0.046]; Men[0.044]; Men[0.044]; Men[0.044]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Morocco[0.041]; Morocco[0.041]; Fita[0.041]; Women[0.040]; Belarus[0.040]; Lithuania[0.040]; Bulgaria[0.040]; France[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Ukraine[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nigeria ==> ENTITY: \u001b[32mNigeria\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.180:0.180[0]; log p(e|m)= -0.315:-0.315[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNigeria[0.047]; Nigeria[0.047]; Nigeria[0.047]; Men[0.047]; Men[0.047]; Men[0.047]; Women[0.045]; Australia[0.044]; Kenya[0.044]; Kenya[0.044]; Algeria[0.043]; Denmark[0.043]; Britain[0.042]; Britain[0.042]; Jamaica[0.042]; Afolabi[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Vladimir Dubrovshchik\u001b[39m ==> ENTITY: \u001b[32mVladimir Dubrovshchik\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.048]; Belarus[0.048]; Russia[0.045]; Russia[0.045]; Juergen[0.045]; Lithuania[0.044]; discus[0.044]; Lars[0.044]; Alekna[0.044]; Chryste[0.043]; Onyali[0.043]; Andreas[0.043]; Romania[0.043]; Irina[0.043]; Merlene[0.043]; Mayock[0.042]; Privalova[0.042]; Gete[0.042]; Women[0.042]; Women[0.042]; Britain[0.042]; Ireland[0.042]; Michael[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cuba ==> ENTITY: \u001b[32mCuba\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.111:0.111[0]; log p(e|m)= -0.240:-0.240[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCuba[0.049]; Spain[0.047]; Trinidad[0.045]; Men[0.044]; Men[0.044]; Men[0.044]; Netherlands[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Denmark[0.041]; Algeria[0.041]; Russia[0.041]; Russia[0.041]; Britain[0.040]; Britain[0.040]; Belgium[0.040]; Norway[0.040]; Garcia[0.039]; Michael[0.039]; Johnson[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kenya ==> ENTITY: \u001b[32mKenya\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.164:0.164[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKenya[0.067]; Nigeria[0.051]; Nigeria[0.051]; Kipketer[0.049]; Algeria[0.049]; Jamaica[0.048]; Langat[0.048]; Denmark[0.047]; Norway[0.047]; Norway[0.047]; Russia[0.046]; Russia[0.046]; Russia[0.046]; Kiptoo[0.046]; Canada[0.046]; Bulgaria[0.045]; Belarus[0.045]; Britain[0.045]; Britain[0.045]; Cuba[0.044]; France[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.046]; Russia[0.046]; Mark[0.045]; Steve[0.044]; Canada[0.044]; Slovenia[0.043]; Switzerland[0.043]; Sweden[0.042]; Russell[0.042]; Johnson[0.042]; Michael[0.041]; Portugal[0.041]; Norway[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Britain[0.041]; Tony[0.040]; Svetlana[0.039]; France[0.039]; Jack[0.039]; Cuba[0.039]; Cuba[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jamaica ==> ENTITY: \u001b[32mJamaica\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.086:0.086[0]; log p(e|m)= -0.333:-0.333[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJamaica[0.051]; Jamaica[0.051]; seconds[0.044]; seconds[0.044]; Canada[0.043]; Cuba[0.042]; Russell[0.042]; hurdles[0.042]; Athletics[0.042]; Switzerland[0.042]; Dionne[0.041]; Yekaterina[0.041]; Gillian[0.041]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Portugal[0.040]; Sweden[0.040]; Slovenia[0.040]; Friday[0.040]; Michelle[0.039]; Svetlana[0.039]; Ludmila[0.039]; Germany[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Linda Kisabaka\u001b[39m ==> ENTITY: \u001b[32mLinda Kisabaka\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.091:0.091[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMatete[0.044]; seconds[0.044]; seconds[0.044]; Anja[0.044]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Geir[0.042]; Phylis[0.042]; Kipketer[0.041]; Jearl[0.041]; Ruecker[0.041]; Norway[0.041]; Karin[0.041]; hurdles[0.041]; Men[0.040]; Men[0.040]; Men[0.040]; Nylander[0.040]; Eric[0.040]; Dusan[0.040]; Australia[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nigeria ==> ENTITY: \u001b[32mNigeria\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.110:0.110[0]; log p(e|m)= -0.315:-0.315[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMen[0.049]; Men[0.049]; discus[0.049]; Women[0.047]; Women[0.047]; hurdles[0.046]; Kenya[0.046]; Kenya[0.046]; Kenya[0.046]; Kenya[0.046]; Lithuania[0.045]; seconds[0.045]; seconds[0.045]; Ireland[0.044]; Barsosio[0.044]; Spain[0.044]; Romania[0.044]; Ethiopia[0.044]; Britain[0.044]; Jamaica[0.044]; Onyali[0.043]; Anthony[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.044]; Lithuania[0.044]; Algeria[0.044]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Belarus[0.043]; Britain[0.042]; Norway[0.041]; Norway[0.041]; Markov[0.041]; Tim[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Bulgaria[0.040]; Cuba[0.040]; Ukraine[0.040]; David[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Namibia ==> ENTITY: \u001b[32mNamibia\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.368:-0.368[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlgeria[0.044]; Men[0.043]; Men[0.043]; Men[0.043]; seconds[0.043]; seconds[0.043]; seconds[0.043]; hurdles[0.043]; Canada[0.042]; Denmark[0.042]; Women[0.041]; Michael[0.041]; Fredericks[0.041]; Burundi[0.041]; Belgium[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Jon[0.041]; Mark[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andreas Seelig\u001b[39m ==> ENTITY: \u001b[32mAndreas Seelig\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndiscus[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Men[0.043]; Lithuania[0.043]; Juergen[0.042]; Alekna[0.042]; Netherlands[0.042]; Privalova[0.042]; Lars[0.042]; Romania[0.041]; Chryste[0.041]; Onyali[0.041]; Spain[0.041]; Britain[0.040]; Ireland[0.040]; Belarus[0.040]; Belarus[0.040]; Merlene[0.040]; Devers[0.040]; Mayock[0.040]; Jamaica[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Florian Schwarthoff\u001b[39m ==> ENTITY: \u001b[32mFlorian Schwarthoff\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; Germany[0.047]; Johnson[0.045]; Sweden[0.043]; Fredericks[0.042]; Finland[0.042]; Canada[0.042]; Cuba[0.041]; Cuba[0.041]; seconds[0.041]; McPaul[0.041]; Linford[0.041]; Belarus[0.041]; race[0.041]; Lithuania[0.041]; Nigeria[0.041]; Nigeria[0.041]; Nigeria[0.041]; Balzer[0.041]; Jamaica[0.040]; Australia[0.040]; Falk[0.040]; relay[0.040]; Rantanen[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jonathan Edwards ==> ENTITY: \u001b[32mJonathan Edwards (athlete)\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.069:0.069[0]; log p(e|m)= -1.415:-1.415[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.046]; Paul[0.043]; Trine[0.043]; Charles[0.043]; Tom[0.043]; Michael[0.043]; jump[0.042]; Nele[0.042]; Bob[0.042]; Australia[0.042]; javelin[0.042]; Harrison[0.042]; Tanja[0.042]; Daniel[0.041]; Hanne[0.041]; Kenny[0.041]; Mike[0.041]; Isel[0.040]; McPaul[0.040]; triple[0.040]; Haugland[0.040]; seconds[0.040]; Kennedy[0.040]; Fita[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tim Lobinger\u001b[39m ==> ENTITY: \u001b[32mTim Lobinger\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.126:0.126[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \njump[0.046]; Nele[0.045]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Russia[0.044]; Russia[0.044]; Russia[0.044]; Russia[0.044]; Russia[0.044]; Russia[0.044]; Stefka[0.043]; Geir[0.043]; Hanne[0.043]; Dmitri[0.043]; vault[0.043]; Kipketer[0.043]; Haugland[0.042]; Kazakstan[0.042]; Rodal[0.042]; Kostadinova[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.135:0.135[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.044]; Russia[0.044]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Algeria[0.043]; Ireland[0.042]; Denmark[0.042]; Britain[0.042]; Cuba[0.041]; Netherlands[0.041]; Belarus[0.041]; Belarus[0.041]; Belgium[0.041]; Garcia[0.041]; Lithuania[0.041]; Men[0.040]; Men[0.040]; Nigeria[0.039]; Anthony[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hanne Haugland\u001b[39m ==> ENTITY: \u001b[32mHanne Haugland\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.119:0.119[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.045]; Norway[0.045]; jump[0.045]; jump[0.045]; Lobinger[0.044]; Nele[0.043]; Fita[0.043]; high[0.043]; Stefka[0.042]; Women[0.042]; Kostadinova[0.042]; vault[0.042]; Rodal[0.041]; triple[0.041]; Inga[0.040]; seconds[0.040]; Shem[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Men[0.040]; Men[0.040]; Men[0.040]; Daniel[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Switzerland ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.103:0.103[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.051]; Sweden[0.049]; Germany[0.047]; Germany[0.047]; Russia[0.047]; Russia[0.047]; Russia[0.047]; Slovenia[0.047]; Men[0.046]; France[0.046]; Women[0.045]; Women[0.045]; Portugal[0.045]; Mark[0.045]; Ludmila[0.045]; minutes[0.044]; Britain[0.044]; seconds[0.043]; seconds[0.043]; seconds[0.043]; Russell[0.043]; Friday[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Davidson Ezinwa\u001b[39m ==> ENTITY: \u001b[32mDavidson Ezinwa\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJamaica[0.045]; Kenya[0.045]; Kenya[0.045]; Kipketer[0.044]; Langat[0.043]; Nigeria[0.042]; Nigeria[0.042]; Nigeria[0.042]; Nigeria[0.042]; Kiptoo[0.042]; Ogunkoya[0.042]; Lobinger[0.042]; Afolabi[0.041]; Anja[0.041]; Jearl[0.041]; Geir[0.040]; Wilson[0.040]; seconds[0.040]; Norway[0.040]; Norway[0.040]; Smith[0.040]; Sammy[0.040]; vault[0.040]; Rodal[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Noureddine Morceli\u001b[39m ==> ENTITY: \u001b[32mNoureddine Morceli\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseconds[0.045]; seconds[0.045]; Rotich[0.043]; Tanui[0.043]; minutes[0.043]; Juergen[0.043]; Lars[0.042]; discus[0.042]; Michael[0.042]; Spain[0.042]; Kenya[0.042]; Kenya[0.042]; Trinidad[0.042]; Alekna[0.041]; Geir[0.041]; Andreas[0.041]; Mayock[0.040]; Denmark[0.040]; Algeria[0.040]; Fredericks[0.040]; Anthony[0.040]; Garcia[0.040]; Norway[0.040]; Cuba[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ato Boldon\u001b[39m ==> ENTITY: \u001b[32mAto Boldon\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.077:0.077[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTrinidad[0.047]; Women[0.044]; Johnson[0.044]; Michael[0.043]; Kenya[0.043]; Kenya[0.043]; Britain[0.043]; Cuba[0.042]; Cuba[0.042]; Geir[0.042]; Florian[0.042]; Busemann[0.041]; Thorsett[0.041]; Fredericks[0.041]; Brown[0.041]; hurdles[0.041]; Steve[0.040]; Mark[0.040]; Morceli[0.040]; Tony[0.040]; Norway[0.040]; Men[0.040]; Men[0.040]; Men[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.046]; Finland[0.044]; Men[0.043]; Men[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Romania[0.043]; Italy[0.043]; Canada[0.043]; Hungary[0.042]; Barsosio[0.042]; Women[0.041]; Women[0.041]; Sven[0.041]; Nigeria[0.040]; Nigeria[0.040]; Nigeria[0.040]; Britain[0.040]; seconds[0.040]; seconds[0.040]; seconds[0.040]; Australia[0.039]; Karin[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yekaterina Podkopayeva\u001b[39m ==> ENTITY: \u001b[32mYekaterina Podkopayeva\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.095:0.095[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.047]; Russia[0.047]; Russia[0.047]; athletics[0.047]; Athletics[0.045]; Svetlana[0.045]; Ludmila[0.044]; Florian[0.044]; Busemann[0.043]; Canada[0.043]; Thorsett[0.043]; Engquist[0.043]; Russell[0.042]; seconds[0.042]; seconds[0.042]; seconds[0.042]; hurdles[0.042]; hurdles[0.042]; Britain[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Carlos Calado\u001b[39m ==> ENTITY: \u001b[32mCarlos Calado\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \njump[0.049]; Portugal[0.049]; javelin[0.046]; triple[0.046]; Trine[0.046]; Fita[0.045]; Tanja[0.044]; race[0.044]; Rivero[0.043]; seconds[0.043]; seconds[0.043]; Isel[0.043]; McPaul[0.043]; Lithuania[0.043]; Norway[0.042]; Norway[0.042]; relay[0.042]; Shem[0.042]; Kenny[0.042]; Ethiopia[0.042]; Cuba[0.042]; Cuba[0.042]; Cuba[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.110:0.110[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.045]; Australia[0.044]; Denmark[0.043]; Marc[0.042]; David[0.042]; Norway[0.042]; Jamaica[0.042]; Sweden[0.042]; Michael[0.041]; Italy[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Jon[0.041]; Wilson[0.041]; Kenya[0.041]; Kenya[0.041]; Kenya[0.041]; Hungary[0.041]; seconds[0.040]; seconds[0.040]; Davidson[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gunhild Halle\u001b[39m ==> ENTITY: \u001b[32mGunhild Haugen\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBarsosio[0.047]; Privalova[0.045]; Matete[0.044]; Norway[0.044]; Chryste[0.042]; Onyali[0.042]; Tegla[0.042]; Inger[0.042]; Merlene[0.042]; Ottey[0.041]; Cheruiyot[0.041]; Sweden[0.041]; Devers[0.041]; Gete[0.041]; Sturrup[0.041]; seconds[0.041]; seconds[0.041]; seconds[0.041]; seconds[0.041]; Loroupe[0.041]; hurdles[0.041]; Hungary[0.040]; Jearl[0.040]; Nylander[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Anthony Washington\u001b[39m ==> ENTITY: \u001b[32mAnthony Washington\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.080:0.080[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMarcus[0.046]; discus[0.045]; John[0.044]; Michael[0.044]; William[0.044]; Juergen[0.042]; Torrence[0.042]; Lars[0.042]; Gaines[0.041]; Alekna[0.041]; Chryste[0.041]; Miller[0.041]; Onyali[0.041]; Andreas[0.041]; Men[0.041]; Men[0.041]; Women[0.041]; Women[0.041]; Merlene[0.040]; Isaac[0.040]; Ireland[0.040]; Mayock[0.040]; Jamaica[0.040]; Devers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Finland ==> ENTITY: \u001b[32mFinland\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.165:0.165[0]; log p(e|m)= -0.245:-0.245[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.048]; Norway[0.047]; Norway[0.047]; Lithuania[0.044]; Canada[0.043]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Belarus[0.042]; Britain[0.041]; Britain[0.041]; Rantanen[0.041]; Men[0.041]; Men[0.041]; Portugal[0.040]; Greece[0.040]; Armenia[0.040]; Women[0.039]; Karlsson[0.039]; Jamaica[0.039]; Tanja[0.039]; Fita[0.039]; Johnson[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Falk Balzer\u001b[39m ==> ENTITY: \u001b[32mFalk Balzer\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Canada[0.042]; Florian[0.042]; Britain[0.041]; Belgium[0.040]; Norway[0.040]; Claus[0.040]; Denmark[0.040]; Geir[0.040]; Mark[0.040]; France[0.039]; Busemann[0.039]; Russia[0.039]; Russia[0.039]; Russia[0.039]; Russia[0.039]; Thorsett[0.039]; Michael[0.039]; Johnson[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.049]; Belarus[0.049]; Lithuania[0.047]; Romania[0.047]; Finland[0.045]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Norway[0.044]; Vladimir[0.044]; Irina[0.042]; Privalova[0.042]; Nigeria[0.042]; Men[0.042]; Men[0.042]; Alekna[0.041]; Women[0.041]; Women[0.041]; Washington[0.041]; Miller[0.041]; Jamaica[0.041]; Vasily[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.120:0.120[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.047]; Russia[0.047]; Mark[0.045]; Steve[0.045]; Canada[0.044]; Slovenia[0.043]; Switzerland[0.043]; Sweden[0.043]; Russell[0.042]; Portugal[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Britain[0.041]; Athletics[0.041]; Tony[0.040]; Svetlana[0.040]; France[0.040]; Cuba[0.039]; Cuba[0.039]; Jamaica[0.039]; Jamaica[0.039]; Jamaica[0.039]; athletics[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kazakstan\u001b[39m ==> ENTITY: \u001b[32mKazakhstan\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.046]; Russia[0.046]; Russia[0.046]; Russia[0.046]; Russia[0.046]; Russia[0.046]; Belarus[0.044]; Ukraine[0.044]; Lithuania[0.041]; Britain[0.041]; Nele[0.041]; Bulgaria[0.040]; Tarasov[0.040]; Norway[0.040]; Norway[0.040]; Norway[0.040]; Lobinger[0.040]; Stefka[0.040]; Hanne[0.040]; Cuba[0.039]; Haugland[0.039]; Algeria[0.039]; Rodal[0.039]; Kostadinova[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cuba ==> ENTITY: \u001b[32mCuba\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.240:-0.240[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCuba[0.049]; Cuba[0.049]; Carlos[0.045]; Lithuania[0.044]; Lithuania[0.044]; Portugal[0.041]; Germany[0.041]; Germany[0.041]; Lopez[0.041]; Men[0.041]; Men[0.041]; Men[0.041]; Russia[0.041]; Russia[0.041]; Belarus[0.041]; Morocco[0.041]; Morocco[0.041]; Britain[0.040]; Rivero[0.040]; Norway[0.040]; Norway[0.040]; Norway[0.040]; Daniel[0.039]; Michael[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Merlene Ottey\u001b[39m ==> ENTITY: \u001b[32mMerlene Ottey\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.103:0.103[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJamaica[0.046]; Sturrup[0.046]; Bahamas[0.043]; Juergen[0.043]; Lars[0.042]; Britain[0.042]; discus[0.042]; Devers[0.042]; Norway[0.042]; Alekna[0.042]; Lithuania[0.041]; Chryste[0.041]; Onyali[0.041]; Finland[0.041]; Andreas[0.041]; Privalova[0.041]; seconds[0.041]; seconds[0.041]; Michael[0.041]; Mayock[0.040]; Gunhild[0.040]; Belarus[0.040]; Belarus[0.040]; Gete[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Valentina Fedyushina\u001b[39m ==> ENTITY: \u001b[32mValentina Fedyushina\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.106:0.106[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVladimir[0.048]; Vasily[0.048]; discus[0.048]; Belarus[0.047]; Belarus[0.047]; Lars[0.047]; Irina[0.047]; Geir[0.046]; Busemann[0.046]; Russia[0.045]; Russia[0.045]; Mayock[0.045]; Michael[0.044]; Norway[0.044]; seconds[0.044]; seconds[0.044]; shot[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ukraine ==> ENTITY: \u001b[32mUkraine\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.199:0.199[0]; log p(e|m)= -0.354:-0.354[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.048]; Lithuania[0.045]; Russia[0.045]; Russia[0.045]; Russia[0.045]; Russia[0.045]; Russia[0.045]; Russia[0.045]; Bulgaria[0.042]; Kazakstan[0.041]; Norway[0.040]; Norway[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Andrei[0.040]; Dmitri[0.039]; Tatyana[0.039]; Algeria[0.039]; Igor[0.039]; Igor[0.039]; Pyotr[0.039]; Tarasov[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.104:0.104[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.046]; Switzerland[0.044]; Trinidad[0.044]; Denmark[0.044]; Britain[0.043]; Russia[0.042]; Russia[0.042]; Russia[0.042]; Slovenia[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Portugal[0.042]; Norway[0.041]; Jamaica[0.041]; Jamaica[0.041]; Johnson[0.041]; Cuba[0.040]; Cuba[0.040]; Michael[0.040]; Belgium[0.039]; Mark[0.039]; Brown[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kenya ==> ENTITY: \u001b[32mKenya\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.171:0.171[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKenya[0.057]; Kenya[0.057]; Kenya[0.057]; Ethiopia[0.048]; Morocco[0.042]; Morocco[0.042]; Koech[0.041]; Norway[0.040]; Norway[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Bulgaria[0.039]; Belarus[0.039]; Lithuania[0.039]; Britain[0.039]; Portugal[0.038]; Ukraine[0.038]; Cuba[0.038]; Komen[0.037]; Armenia[0.037]; Shem[0.036]; France[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.158:0.158[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Canada[0.043]; Denmark[0.043]; Norway[0.041]; Norway[0.041]; France[0.041]; Britain[0.041]; Britain[0.041]; Kenya[0.041]; Kenya[0.041]; Men[0.040]; Men[0.040]; Men[0.040]; Bulgaria[0.040]; Algeria[0.040]; jump[0.039]; Nigeria[0.039]; Nigeria[0.039]; Donovan[0.039]; Belarus[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rohan Robinson\u001b[39m ==> ENTITY: \u001b[32mRohan Robinson\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.047]; Jamaica[0.045]; seconds[0.044]; seconds[0.044]; seconds[0.044]; Matete[0.043]; Barsosio[0.043]; Anja[0.042]; minutes[0.042]; Jearl[0.041]; Kenya[0.041]; Kenya[0.041]; Kenya[0.041]; Kenya[0.041]; hurdles[0.041]; Canada[0.041]; Thomas[0.041]; Gunhild[0.040]; Gete[0.040]; Ruecker[0.040]; Norway[0.040]; Derrick[0.040]; Mitchell[0.040]; Michael[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ethiopia ==> ENTITY: \u001b[32mEthiopia\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.114:0.114[0]; log p(e|m)= -0.236:-0.236[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKenya[0.045]; Kenya[0.045]; Kenya[0.045]; Kenya[0.045]; Men[0.044]; Australia[0.044]; Italy[0.043]; Barsosio[0.043]; Women[0.042]; Women[0.042]; Zambia[0.042]; Nigeria[0.042]; Jamaica[0.041]; Lithuania[0.041]; Russia[0.040]; Tegla[0.040]; Sweden[0.040]; Norway[0.040]; Hungary[0.040]; Wami[0.040]; Privalova[0.039]; Romania[0.039]; Bahamas[0.039]; Loroupe[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.048]; Britain[0.045]; Australia[0.045]; Canada[0.044]; George[0.044]; Lithuania[0.044]; Sweden[0.042]; Michael[0.042]; Michael[0.042]; Norway[0.042]; Jamaica[0.042]; Johnson[0.041]; Donovan[0.041]; Belarus[0.041]; Finland[0.040]; Nigeria[0.040]; Nigeria[0.040]; Nigeria[0.040]; Namibia[0.040]; Peter[0.040]; Greece[0.040]; race[0.039]; Cuba[0.039]; Cuba[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.047]; Britain[0.046]; Men[0.044]; Men[0.044]; Men[0.044]; Netherlands[0.044]; Belgium[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Women[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Geir[0.039]; Kenya[0.039]; Kenya[0.039]; seconds[0.039]; seconds[0.039]; seconds[0.039]; Trinidad[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chryste Gaines\u001b[39m ==> ENTITY: \u001b[32mChryste Gaines\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.126:0.126[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDevers[0.046]; Ottey[0.043]; Torrence[0.043]; Juergen[0.043]; Sturrup[0.043]; Miller[0.043]; Privalova[0.043]; Lars[0.042]; discus[0.042]; Barsosio[0.042]; Alekna[0.042]; Onyali[0.041]; Inger[0.041]; Andreas[0.041]; Chandra[0.041]; Gail[0.041]; Sally[0.041]; Gwen[0.041]; Merlene[0.041]; Mayock[0.040]; Gunhild[0.040]; Gete[0.040]; Norway[0.040]; John[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Xiomara Rivero\u001b[39m ==> ENTITY: \u001b[32mXiomara Rivero\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.075:0.075[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTrine[0.045]; Carlos[0.044]; Cuba[0.043]; Cuba[0.043]; Cuba[0.043]; javelin[0.043]; Tanja[0.043]; Florian[0.042]; relay[0.042]; Isel[0.042]; McPaul[0.042]; Fita[0.041]; Norway[0.041]; Norway[0.041]; jump[0.041]; Kenny[0.041]; Lopez[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Men[0.040]; Men[0.040]; Quesada[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Daniel Komen\u001b[39m ==> ENTITY: \u001b[32mDaniel Komen\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.116:0.116[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKenya[0.046]; Kenya[0.046]; Kenya[0.046]; Kenya[0.046]; seconds[0.044]; Koech[0.044]; Ethiopia[0.043]; minutes[0.043]; Nele[0.042]; high[0.042]; Lobinger[0.041]; Stefka[0.041]; Hanne[0.041]; jump[0.040]; jump[0.040]; Kenny[0.040]; Haugland[0.040]; Kostadinova[0.039]; Morocco[0.039]; Morocco[0.039]; Fita[0.039]; Tim[0.039]; Norway[0.039]; Norway[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lithuania ==> ENTITY: \u001b[32mLithuania\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.397:-0.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.045]; Romania[0.045]; Belarus[0.045]; Belarus[0.045]; Men[0.043]; Men[0.043]; Ireland[0.043]; Spain[0.042]; Britain[0.041]; Privalova[0.041]; Jamaica[0.041]; Women[0.041]; Women[0.041]; Devers[0.041]; Alekna[0.040]; Vladimir[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Bahamas[0.040]; Sturrup[0.040]; Ottey[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yoelvis Quesada\u001b[39m ==> ENTITY: \u001b[32mYoelbi Quesada\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCuba[0.047]; Cuba[0.047]; Cuba[0.047]; jump[0.045]; Carlos[0.044]; Rivero[0.044]; triple[0.042]; Lopez[0.042]; Trine[0.042]; javelin[0.041]; Lithuania[0.041]; Lithuania[0.041]; Natalya[0.041]; Natalya[0.041]; Nele[0.041]; Belarus[0.040]; Tanja[0.040]; Xiomara[0.040]; relay[0.040]; Hanne[0.039]; Russia[0.039]; Russia[0.039]; Isel[0.039]; McPaul[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Igor Potapovich\u001b[39m ==> ENTITY: \u001b[32mIgor Potapovich\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Kazakstan[0.043]; Ukraine[0.043]; Belarus[0.042]; pole[0.042]; vault[0.042]; Kipketer[0.042]; Igor[0.041]; Nele[0.041]; Alina[0.041]; Kostadinova[0.041]; Dmitri[0.041]; Lithuania[0.041]; jump[0.040]; Lobinger[0.040]; Natalya[0.040]; Stefka[0.040]; Geir[0.040]; Hanne[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nigeria ==> ENTITY: \u001b[32mNigeria\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.162:0.162[0]; log p(e|m)= -0.315:-0.315[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNigeria[0.045]; Nigeria[0.045]; Nigeria[0.045]; Men[0.045]; Men[0.045]; Men[0.045]; Women[0.043]; hurdles[0.042]; Australia[0.042]; Kenya[0.042]; Kenya[0.042]; Denmark[0.041]; Sweden[0.041]; Barsosio[0.040]; Italy[0.040]; Britain[0.040]; Jamaica[0.040]; Norway[0.040]; Germany[0.039]; Germany[0.039]; Germany[0.039]; Germany[0.039]; Germany[0.039]; Canada[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = John Mayock\u001b[39m ==> ENTITY: \u001b[32mJohn Mayock\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.117:0.117[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDevers[0.045]; Britain[0.044]; Juergen[0.043]; Lars[0.043]; discus[0.043]; Denmark[0.042]; Kenya[0.042]; Kenya[0.042]; Alekna[0.042]; Chryste[0.042]; Morceli[0.042]; Onyali[0.042]; Ireland[0.042]; Andreas[0.042]; Merlene[0.041]; mile[0.041]; Michael[0.041]; Jamaica[0.041]; Anthony[0.040]; Ottey[0.040]; Rotich[0.040]; seconds[0.040]; seconds[0.040]; Tanui[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jon Drummond\u001b[39m ==> ENTITY: \u001b[32mJon Drummond\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.047:0.047[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.043]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Britain[0.042]; Boldon[0.042]; Mark[0.042]; Norway[0.042]; Geir[0.042]; seconds[0.042]; seconds[0.042]; seconds[0.042]; hurdles[0.041]; Michael[0.041]; Steve[0.041]; Trinidad[0.041]; Johnson[0.041]; Rotich[0.041]; Florian[0.041]; Tony[0.041]; Busemann[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Michael Green ==> ENTITY: \u001b[32mMichael Green (sprinter)\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.116:0.116[0]; log p(e|m)= -2.033:-2.033[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDavid[0.049]; Marc[0.046]; Dennis[0.045]; Thomas[0.045]; Jon[0.045]; Craig[0.044]; Eric[0.044]; Mitchell[0.043]; Miles[0.042]; Smith[0.041]; Jamaica[0.041]; Canada[0.041]; Robinson[0.040]; Donovan[0.040]; Wilson[0.040]; Langat[0.040]; Kipketer[0.040]; Bailey[0.040]; Jearl[0.039]; Davidson[0.039]; Sweden[0.039]; Anja[0.039]; Kenya[0.039]; Kenya[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.094:0.094[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJamaica[0.045]; Jamaica[0.045]; Jamaica[0.045]; Steve[0.043]; Mark[0.043]; Sweden[0.043]; Women[0.043]; Women[0.043]; Britain[0.042]; Switzerland[0.042]; Michael[0.042]; France[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Men[0.041]; Men[0.041]; Johnson[0.041]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Russell[0.040]; Tony[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nigeria ==> ENTITY: \u001b[32mNigeria\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.315:-0.315[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNigeria[0.049]; Nigeria[0.049]; Men[0.048]; javelin[0.047]; Women[0.046]; Australia[0.045]; Lithuania[0.044]; Belarus[0.043]; Britain[0.043]; Jamaica[0.043]; Sweden[0.043]; relay[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Johnson[0.042]; Norway[0.042]; Canada[0.042]; seconds[0.041]; Donovan[0.041]; Finland[0.041]; George[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nico Motchebon\u001b[39m ==> ENTITY: \u001b[32mNico Motchebon\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.132:0.132[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKipketer[0.047]; Lobinger[0.046]; Kiptoo[0.043]; Michael[0.043]; Stefka[0.042]; Geir[0.042]; Kostadinova[0.042]; jump[0.042]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Jamaica[0.041]; Marc[0.041]; Rodal[0.041]; David[0.041]; Cuba[0.041]; Sammy[0.041]; Norway[0.040]; Norway[0.040]; Donovan[0.040]; Karin[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.081:0.081[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.047]; Donovan[0.045]; Britain[0.044]; Sweden[0.043]; Italy[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Romania[0.042]; Finland[0.042]; Norway[0.042]; Jamaica[0.042]; Hungary[0.042]; Thomas[0.041]; Men[0.041]; Men[0.041]; Australia[0.041]; Michael[0.041]; Smith[0.040]; Nigeria[0.039]; Nigeria[0.039]; Nigeria[0.039]; Torrance[0.039]; Samuel[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.124:0.124[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMen[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Spain[0.044]; Romania[0.043]; Women[0.043]; Women[0.043]; Lithuania[0.043]; Privalova[0.042]; Ireland[0.042]; Britain[0.041]; Netherlands[0.041]; Andreas[0.040]; discus[0.040]; Finland[0.040]; Juergen[0.039]; Belarus[0.039]; Belarus[0.039]; Miller[0.039]; Russia[0.039]; seconds[0.039]; seconds[0.039]; Alekna[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.188:0.188[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.048]; Germany[0.048]; Germany[0.048]; Germany[0.048]; Germany[0.048]; Germany[0.048]; Denmark[0.045]; Canada[0.045]; Britain[0.044]; Portugal[0.044]; France[0.043]; Belgium[0.043]; Norway[0.042]; Michael[0.042]; Johnson[0.042]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Trinidad[0.040]; Claus[0.039]; Mark[0.039]; Brown[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.081:0.081[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.051]; Slovenia[0.049]; Men[0.048]; Germany[0.048]; Canada[0.048]; Portugal[0.047]; Athletics[0.046]; Women[0.046]; Women[0.046]; Russia[0.046]; Russia[0.046]; Russia[0.046]; Ludmila[0.045]; seconds[0.044]; seconds[0.044]; France[0.044]; Svetlana[0.044]; athletics[0.043]; meeting[0.043]; minutes[0.043]; hurdles[0.043]; Yekaterina[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.090:0.090[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.044]; Belarus[0.043]; Belarus[0.043]; Britain[0.043]; Ireland[0.043]; Russia[0.042]; Russia[0.042]; Russia[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Washington[0.042]; Devers[0.041]; Jamaica[0.041]; Netherlands[0.041]; William[0.041]; Algeria[0.041]; Miller[0.040]; John[0.040]; Michael[0.040]; Lithuania[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.204:0.204[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.046]; Russia[0.046]; Russia[0.046]; Russia[0.046]; Russia[0.046]; Russia[0.046]; Lithuania[0.043]; Ukraine[0.043]; Bulgaria[0.043]; Andrei[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Norway[0.041]; Britain[0.040]; Tarasov[0.040]; Kazakstan[0.040]; Markov[0.039]; France[0.039]; Maksim[0.039]; Dmitri[0.039]; triple[0.039]; Men[0.039]; Men[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wilson Kipketer\u001b[39m ==> ENTITY: \u001b[32mWilson Kipketer\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKiptoo[0.045]; Langat[0.044]; Kenya[0.044]; Kenya[0.044]; seconds[0.043]; Denmark[0.043]; Lobinger[0.042]; Anja[0.042]; Jamaica[0.042]; Canada[0.041]; Geir[0.041]; Britain[0.041]; Britain[0.041]; Michael[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Rodal[0.040]; Norway[0.040]; Norway[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.107:0.107[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.045]; Britain[0.043]; Britain[0.043]; Lithuania[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Women[0.042]; Nigeria[0.042]; Nigeria[0.042]; Nigeria[0.042]; Jamaica[0.042]; Norway[0.042]; Norway[0.042]; Belarus[0.042]; Portugal[0.041]; Davidson[0.041]; Donovan[0.041]; Sweden[0.041]; Johnson[0.041]; Kenya[0.040]; Ethiopia[0.040]; George[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Davidson Ezinwa\u001b[39m ==> ENTITY: \u001b[32mDavidson Ezinwa\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJamaica[0.046]; Trine[0.044]; Nigeria[0.044]; Nigeria[0.044]; Nigeria[0.044]; javelin[0.044]; relay[0.043]; Osmond[0.042]; Tanja[0.042]; Fredericks[0.042]; Florian[0.041]; seconds[0.041]; Norway[0.041]; Isel[0.041]; race[0.041]; McPaul[0.041]; Johnson[0.041]; Namibia[0.040]; Sweden[0.040]; Bailey[0.040]; Frankie[0.040]; Canada[0.040]; Aliu[0.040]; Linford[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.111:0.111[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMen[0.047]; Men[0.047]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Switzerland[0.044]; Women[0.042]; Women[0.042]; Portugal[0.042]; Friday[0.041]; Slovenia[0.041]; hurdles[0.041]; hurdles[0.041]; France[0.041]; Britain[0.041]; Sweden[0.041]; Canada[0.041]; Russia[0.039]; Russia[0.039]; Russia[0.039]; seconds[0.038]; seconds[0.038]; seconds[0.038]; seconds[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.226:0.226[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.045]; Russia[0.045]; Russia[0.045]; Russia[0.045]; Russia[0.045]; Russia[0.045]; Lithuania[0.043]; Ukraine[0.043]; Denmark[0.042]; Bulgaria[0.042]; Andrei[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Norway[0.040]; Norway[0.040]; Norway[0.040]; Britain[0.040]; Tarasov[0.040]; Kazakstan[0.039]; Markov[0.039]; France[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fatima Yusuf\u001b[39m ==> ENTITY: \u001b[32mFatima Yusuf\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNigeria[0.046]; Nigeria[0.046]; Nigeria[0.046]; Nigeria[0.046]; Barsosio[0.046]; Kipketer[0.045]; Kenya[0.045]; Kenya[0.045]; Ogunkoya[0.045]; Afolabi[0.045]; Matete[0.044]; Olabisi[0.044]; Women[0.044]; Anja[0.043]; Geir[0.042]; Australia[0.042]; Torrance[0.042]; Gunhild[0.041]; Italy[0.041]; Men[0.041]; Men[0.041]; Men[0.041]; Ruecker[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jamaica ==> ENTITY: \u001b[32mJamaica\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.078:0.078[0]; log p(e|m)= -0.333:-0.333[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.045]; Sturrup[0.045]; Norway[0.044]; Bahamas[0.044]; Britain[0.043]; seconds[0.043]; seconds[0.043]; Belarus[0.043]; Belarus[0.043]; Lithuania[0.042]; Finland[0.042]; Ottey[0.042]; Devers[0.041]; Ireland[0.041]; Spain[0.041]; Barsosio[0.041]; Privalova[0.041]; John[0.041]; Michael[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Alekna[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.091:0.091[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.046]; Trinidad[0.044]; Denmark[0.043]; Britain[0.043]; Russia[0.042]; Russia[0.042]; Russia[0.042]; Russia[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Portugal[0.041]; Norway[0.041]; Johnson[0.041]; Algeria[0.040]; Cuba[0.040]; Cuba[0.040]; Men[0.040]; Men[0.040]; Men[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Virgilijus Alekna\u001b[39m ==> ENTITY: \u001b[32mVirgilijus Alekna\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.057:0.057[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJuergen[0.044]; Lars[0.044]; discus[0.044]; Devers[0.043]; Chryste[0.043]; Onyali[0.043]; Andreas[0.043]; Merlene[0.042]; Ottey[0.042]; Mayock[0.042]; Privalova[0.042]; Sturrup[0.041]; Gete[0.041]; Lithuania[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Men[0.041]; Men[0.041]; Tanui[0.040]; Marcus[0.040]; Vladimir[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Frankie Fredericks\u001b[39m ==> ENTITY: \u001b[32mFrankie Fredericks\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.117:0.117[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJohnson[0.045]; race[0.045]; Jamaica[0.045]; Bailey[0.044]; Trine[0.043]; Australia[0.042]; javelin[0.042]; relay[0.042]; seconds[0.042]; Tanja[0.042]; Donovan[0.041]; Florian[0.041]; Nigeria[0.041]; Nigeria[0.041]; Nigeria[0.041]; Linford[0.041]; Michael[0.041]; Michael[0.041]; Michael[0.041]; Isel[0.041]; McPaul[0.040]; Namibia[0.040]; Davidson[0.040]; George[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shem Kororia\u001b[39m ==> ENTITY: \u001b[32mShem Kororia\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.132:0.132[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEthiopia[0.045]; Kenya[0.044]; Kenya[0.044]; Kenya[0.044]; Kenya[0.044]; Trine[0.043]; Koech[0.042]; Nele[0.042]; minutes[0.042]; Women[0.042]; Women[0.042]; javelin[0.042]; Tanja[0.041]; Stefka[0.041]; Hanne[0.041]; seconds[0.041]; Men[0.041]; Men[0.041]; Isel[0.040]; Haugland[0.040]; Kostadinova[0.040]; Fita[0.040]; Morocco[0.039]; Morocco[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Algeria ==> ENTITY: \u001b[32mAlgeria\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.146:0.146[0]; log p(e|m)= -0.337:-0.337[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.050]; Denmark[0.045]; Kenya[0.045]; Kenya[0.045]; Nigeria[0.044]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Men[0.043]; Men[0.043]; Men[0.043]; Canada[0.043]; Russia[0.042]; Russia[0.042]; Russia[0.042]; Russia[0.042]; Belarus[0.042]; Britain[0.042]; Jean[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kenya ==> ENTITY: \u001b[32mKenya\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKenya[0.062]; Burundi[0.048]; Rotich[0.046]; Tanui[0.045]; Algeria[0.045]; Ato[0.044]; Netherlands[0.044]; Denmark[0.044]; Norway[0.043]; Russia[0.043]; Russia[0.043]; Belarus[0.042]; Belarus[0.042]; Lithuania[0.042]; Niyongabo[0.042]; Britain[0.042]; Morceli[0.042]; Cuba[0.041]; Trinidad[0.041]; Belgium[0.040]; Ireland[0.040]; Spain[0.039]; Devers[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Stefka Kostadinova\u001b[39m ==> ENTITY: \u001b[32mStefka Kostadinova\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \njump[0.046]; Yelena[0.043]; Nele[0.043]; Tatyana[0.043]; Andrei[0.042]; Natalya[0.042]; Bulgaria[0.042]; seconds[0.042]; Lithuania[0.042]; Lobinger[0.042]; Belarus[0.042]; Hanne[0.042]; Dmitri[0.041]; high[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Haugland[0.041]; Rodal[0.041]; Men[0.040]; Men[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dennis Mitchell\u001b[39m ==> ENTITY: \u001b[32mDennis Mitchell\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.110:0.110[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWilson[0.046]; Derrick[0.046]; Robinson[0.045]; Miles[0.045]; Smith[0.044]; Craig[0.044]; Bailey[0.044]; David[0.044]; Jamaica[0.044]; Anja[0.043]; Thomas[0.043]; Women[0.043]; Davidson[0.043]; Michael[0.043]; Jearl[0.043]; Donovan[0.043]; Sammy[0.043]; Geir[0.043]; Linda[0.043]; seconds[0.042]; Kipketer[0.042]; Canada[0.042]; Rodal[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Michael Johnson ==> ENTITY: \u001b[32mMichael Johnson (sprinter)\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.361:-0.361[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSteve[0.045]; Brown[0.045]; Jarrett[0.044]; Mark[0.044]; Trinidad[0.044]; Canada[0.042]; Jon[0.042]; Fredericks[0.042]; Tony[0.042]; Drummond[0.041]; hurdles[0.041]; Stephanie[0.041]; seconds[0.041]; seconds[0.041]; seconds[0.041]; Patrick[0.041]; Boldon[0.041]; Jack[0.041]; Geir[0.041]; Britain[0.041]; Florian[0.040]; William[0.040]; minutes[0.040]; Busemann[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.048:0.048[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.045]; Russia[0.045]; Sweden[0.045]; Finland[0.043]; Norway[0.043]; Miller[0.043]; Thomas[0.043]; Men[0.042]; Australia[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; hurdles[0.041]; Privalova[0.041]; Smith[0.041]; Bahamas[0.041]; Hungary[0.040]; Italy[0.040]; Miles[0.040]; Women[0.040]; Women[0.040]; seconds[0.040]; seconds[0.040]; seconds[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portugal ==> ENTITY: \u001b[32mPortugal\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.309:-0.309[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSlovenia[0.047]; Britain[0.045]; France[0.045]; Sweden[0.045]; Switzerland[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Canada[0.044]; Cuba[0.043]; Cuba[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Jamaica[0.043]; Jamaica[0.043]; Jamaica[0.043]; Men[0.042]; Men[0.042]; Athletics[0.042]; Lopez[0.041]; Tony[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kenny Harrison\u001b[39m ==> ENTITY: \u001b[32mKenny Harrison\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.074:0.074[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \njump[0.049]; triple[0.046]; Trine[0.046]; relay[0.045]; Nele[0.045]; Daniel[0.045]; Carlos[0.044]; javelin[0.044]; Tanja[0.044]; seconds[0.044]; Michael[0.043]; Edwards[0.043]; Jonathan[0.043]; Charles[0.043]; Isel[0.043]; Bob[0.043]; McPaul[0.043]; Fita[0.042]; Norway[0.042]; Norway[0.042]; Shem[0.041]; Paul[0.041]; Kennedy[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.094:0.094[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.048]; Donovan[0.046]; Denmark[0.046]; Britain[0.045]; Sweden[0.045]; Italy[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Norway[0.043]; Norway[0.043]; Jamaica[0.043]; Hungary[0.043]; Thomas[0.042]; Men[0.042]; Men[0.042]; Men[0.042]; Australia[0.042]; Wilson[0.042]; Michael[0.042]; Davidson[0.041]; Smith[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Annemari Sandell\u001b[39m ==> ENTITY: \u001b[32mAnnemari Sandell-Hyvärinen\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBarsosio[0.048]; hurdles[0.046]; Finland[0.045]; Loroupe[0.045]; Devers[0.044]; Onyali[0.044]; Matete[0.044]; Privalova[0.044]; Ottey[0.044]; Sweden[0.044]; Wami[0.043]; Norway[0.043]; Chryste[0.043]; seconds[0.043]; seconds[0.043]; seconds[0.043]; seconds[0.043]; Sturrup[0.042]; Andreas[0.042]; Ethiopia[0.042]; Merlene[0.042]; Gunhild[0.042]; Cheruiyot[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.114:0.114[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.047]; Denmark[0.043]; William[0.043]; Washington[0.043]; Jamaica[0.042]; Spain[0.042]; Michael[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Kenya[0.041]; Kenya[0.041]; Lithuania[0.041]; John[0.041]; seconds[0.041]; seconds[0.041]; Netherlands[0.040]; Russia[0.040]; Russia[0.040]; Cuba[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portugal ==> ENTITY: \u001b[32mPortugal\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.309:-0.309[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLithuania[0.048]; Canada[0.045]; Britain[0.045]; Britain[0.045]; Australia[0.045]; Morocco[0.044]; Morocco[0.044]; Germany[0.044]; Germany[0.044]; Belarus[0.043]; Carlos[0.043]; Namibia[0.043]; Cuba[0.043]; Cuba[0.043]; Cuba[0.043]; Finland[0.043]; Norway[0.043]; Norway[0.043]; Jamaica[0.042]; Donovan[0.042]; Men[0.042]; Men[0.042]; Nigeria[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Emilio Valle\u001b[39m ==> ENTITY: \u001b[32mEmilio Valle\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.120:0.120[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCuba[0.051]; Cuba[0.051]; seconds[0.050]; seconds[0.050]; seconds[0.050]; hurdles[0.049]; Margarita[0.048]; Geir[0.048]; Florian[0.047]; Busemann[0.047]; Trinidad[0.047]; Thorsett[0.047]; Garcia[0.047]; Jamaica[0.047]; Svetlana[0.046]; Yekaterina[0.046]; Drummond[0.046]; Balzer[0.046]; Norway[0.046]; Johnson[0.046]; minutes[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ludmila Engquist\u001b[39m ==> ENTITY: \u001b[32mLudmila Engquist\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.105:0.105[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.045]; Yekaterina[0.045]; Russia[0.045]; Russia[0.045]; Russia[0.045]; athletics[0.044]; Svetlana[0.043]; Athletics[0.043]; Germany[0.042]; Thorsett[0.041]; Berlin[0.041]; Berlin[0.041]; Berlin[0.041]; hurdles[0.041]; Grand[0.041]; Grand[0.041]; Canada[0.040]; Switzerland[0.040]; Slovenia[0.040]; Men[0.040]; Prix[0.040]; Prix[0.040]; Brigita[0.040]; Baumann[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Venuste Niyongabo\u001b[39m ==> ENTITY: \u001b[32mVénuste Niyongabo\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.132:0.132[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMorceli[0.050]; Kenya[0.048]; Kenya[0.048]; Burundi[0.047]; Juergen[0.047]; Lars[0.046]; discus[0.046]; seconds[0.046]; seconds[0.046]; Alekna[0.046]; Geir[0.045]; Andreas[0.045]; Rotich[0.045]; Tanui[0.045]; Noureddine[0.045]; Mayock[0.044]; Fredericks[0.044]; Women[0.044]; Norway[0.044]; Algeria[0.043]; mile[0.043]; Johnson[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kenya ==> ENTITY: \u001b[32mKenya\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKenya[0.067]; Nigeria[0.052]; Kipketer[0.049]; Algeria[0.049]; Jamaica[0.048]; Langat[0.048]; Denmark[0.047]; Norway[0.047]; Norway[0.047]; Russia[0.047]; Russia[0.047]; Russia[0.047]; Kiptoo[0.046]; Canada[0.046]; Bulgaria[0.046]; Belarus[0.046]; Britain[0.045]; Britain[0.045]; Ukraine[0.044]; Cuba[0.044]; France[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.091:0.091[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.044]; Michael[0.044]; Britain[0.043]; Italy[0.043]; Dennis[0.042]; Miles[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Women[0.042]; Nigeria[0.042]; Nigeria[0.042]; Nigeria[0.042]; Nigeria[0.042]; Jamaica[0.042]; Norway[0.041]; Hungary[0.041]; Davidson[0.041]; Mitchell[0.041]; Donovan[0.041]; Sweden[0.041]; Smith[0.040]; Sally[0.040]; Jon[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Morocco ==> ENTITY: \u001b[32mMorocco\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.296:-0.296[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMorocco[0.049]; Britain[0.043]; Portugal[0.042]; Cuba[0.042]; Kenya[0.042]; Kenya[0.042]; Kenya[0.042]; Kenya[0.042]; Men[0.041]; Men[0.041]; Bulgaria[0.041]; Norway[0.041]; Norway[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Ethiopia[0.041]; Fita[0.041]; seconds[0.041]; Lithuania[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Michael Johnson ==> ENTITY: \u001b[32mMichael Johnson (sprinter)\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.152:0.152[0]; log p(e|m)= -0.361:-0.361[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBailey[0.044]; Michael[0.044]; Michael[0.044]; race[0.044]; Jamaica[0.044]; Jesse[0.043]; Australia[0.043]; Donovan[0.043]; Davidson[0.042]; relay[0.042]; Canada[0.042]; Fredericks[0.041]; Lopez[0.041]; Carlos[0.041]; Mike[0.041]; George[0.041]; Charles[0.041]; seconds[0.040]; Men[0.040]; Green[0.040]; Britain[0.040]; Peter[0.040]; Louise[0.040]; Owens[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.159:0.159[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Denmark[0.045]; Spain[0.044]; Belgium[0.044]; Netherlands[0.043]; Ireland[0.043]; Norway[0.042]; Britain[0.042]; Algeria[0.040]; Johnson[0.039]; Namibia[0.039]; Russia[0.039]; Russia[0.039]; minutes[0.038]; Marko[0.038]; Trinidad[0.038]; Anthony[0.038]; Michael[0.038]; Kenya[0.038]; Kenya[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.157:0.157[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Hungary[0.044]; Denmark[0.044]; Italy[0.043]; Britain[0.042]; Sweden[0.042]; Canada[0.041]; Norway[0.041]; Norway[0.041]; Sven[0.040]; Algeria[0.040]; Dusan[0.040]; Donovan[0.040]; Men[0.040]; Men[0.040]; Australia[0.039]; Eric[0.039]; Nigeria[0.039]; Nigeria[0.039]; Nigeria[0.039]; Nigeria[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pauline Konga\u001b[39m ==> ENTITY: \u001b[32mPauline Konga\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.228:0.228[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKenya[0.044]; Kenya[0.044]; Kenya[0.044]; Kenya[0.044]; Barsosio[0.044]; Cheruiyot[0.043]; Tegla[0.042]; Sturrup[0.042]; Wami[0.041]; Matete[0.041]; Loroupe[0.041]; Ogunkoya[0.041]; Privalova[0.041]; Mary[0.041]; Ethiopia[0.041]; Italy[0.041]; Onyali[0.041]; Ottey[0.041]; Irina[0.041]; Jamaica[0.041]; Norway[0.040]; Men[0.040]; Anja[0.040]; Zambia[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cuba ==> ENTITY: \u001b[32mCuba\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.240:-0.240[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCuba[0.053]; Cuba[0.053]; Jamaica[0.048]; Lithuania[0.047]; Donovan[0.047]; Canada[0.046]; Carlos[0.045]; Portugal[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Men[0.044]; Men[0.044]; Belarus[0.044]; Sweden[0.044]; Lopez[0.044]; Britain[0.043]; Britain[0.043]; Norway[0.043]; Norway[0.043]; Greece[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Derrick Adkins\u001b[39m ==> ENTITY: \u001b[32mDerrick Adkins\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.098:0.098[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhurdles[0.045]; Jearl[0.044]; Matete[0.043]; Bahamas[0.043]; Anja[0.042]; Barsosio[0.042]; Sturrup[0.042]; Kenya[0.042]; Kenya[0.042]; Kenya[0.042]; Kenya[0.042]; Smith[0.042]; seconds[0.042]; seconds[0.042]; seconds[0.042]; Privalova[0.041]; Gunhild[0.041]; Ethiopia[0.041]; Gete[0.040]; Miller[0.040]; Dennis[0.040]; Zambia[0.040]; Robinson[0.040]; Ruecker[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.057:0.057[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.049]; Belarus[0.049]; Russia[0.048]; Washington[0.047]; Devers[0.047]; Finland[0.047]; Jamaica[0.046]; Norway[0.046]; Miller[0.046]; Men[0.046]; Michael[0.045]; Lithuania[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; hurdles[0.044]; Privalova[0.044]; Bahamas[0.044]; Italy[0.043]; Women[0.043]; Women[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Greece ==> ENTITY: \u001b[32mGreece\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.276:-0.276[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLithuania[0.045]; Germany[0.044]; Germany[0.044]; Finland[0.044]; Sweden[0.044]; Britain[0.044]; Canada[0.043]; Belarus[0.042]; Nigeria[0.042]; Nigeria[0.042]; Nigeria[0.042]; Peter[0.041]; Australia[0.041]; Michael[0.041]; Michael[0.041]; George[0.041]; Donovan[0.041]; Men[0.040]; Johnson[0.040]; Namibia[0.040]; Cuba[0.039]; Cuba[0.039]; Jamaica[0.039]; Panayiotopoulos[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.046]; Jamaica[0.044]; Sweden[0.043]; Finland[0.043]; Belarus[0.043]; Women[0.042]; Norway[0.042]; Norway[0.042]; Cuba[0.042]; Cuba[0.042]; George[0.042]; Britain[0.042]; Donovan[0.042]; Lithuania[0.041]; Michael[0.041]; Michael[0.041]; Michael[0.041]; Men[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; race[0.040]; Charles[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Trinidad ==> ENTITY: \u001b[32mTrinidad\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.056:0.056[0]; log p(e|m)= -0.341:-0.341[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCuba[0.044]; Cuba[0.044]; Garcia[0.044]; Johnson[0.043]; Margarita[0.043]; William[0.043]; Valle[0.043]; Britain[0.042]; Jon[0.042]; Ivan[0.042]; Emilio[0.041]; Michael[0.041]; mile[0.041]; Tony[0.041]; Frank[0.041]; Frankie[0.041]; Steve[0.041]; Kenya[0.041]; Kenya[0.041]; Mark[0.040]; Jarrett[0.040]; Florian[0.040]; shot[0.040]; Norway[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Aliuska Lopez\u001b[39m ==> ENTITY: \u001b[32mAliuska López\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.108:0.108[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhurdles[0.046]; hurdles[0.046]; athletics[0.045]; Cuba[0.044]; Athletics[0.043]; Jamaica[0.043]; Jamaica[0.043]; Jamaica[0.043]; Thorsett[0.042]; Engquist[0.042]; Canada[0.042]; Germany[0.040]; Grand[0.040]; Grand[0.040]; Men[0.040]; Brigita[0.040]; Portugal[0.040]; Margarita[0.040]; Slovenia[0.040]; Baumann[0.040]; Berlin[0.040]; Berlin[0.040]; Berlin[0.040]; Switzerland[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gwen Torrence\u001b[39m ==> ENTITY: \u001b[32mGwen Torrence\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSturrup[0.046]; Devers[0.044]; Gaines[0.043]; Juergen[0.043]; Barsosio[0.043]; Lars[0.042]; discus[0.042]; Jamaica[0.042]; Ottey[0.042]; Alekna[0.042]; Chryste[0.042]; Onyali[0.042]; Andreas[0.041]; Merlene[0.041]; Marcus[0.041]; Mayock[0.041]; Gunhild[0.041]; Miller[0.041]; Women[0.040]; Women[0.040]; Gete[0.040]; Norway[0.040]; seconds[0.040]; seconds[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.101:0.101[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.050]; Lithuania[0.046]; Portugal[0.045]; Britain[0.045]; Britain[0.045]; Sweden[0.044]; Canada[0.044]; Norway[0.044]; Norway[0.044]; relay[0.043]; Finland[0.043]; jump[0.043]; Johnson[0.043]; Donovan[0.043]; Belarus[0.043]; Men[0.042]; Men[0.042]; race[0.042]; Australia[0.042]; Nigeria[0.042]; Nigeria[0.042]; Nigeria[0.042]; Fita[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.222:0.222[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.045]; Russia[0.045]; Russia[0.045]; Russia[0.045]; Russia[0.045]; Russia[0.045]; Lithuania[0.043]; Ukraine[0.043]; Bulgaria[0.042]; Andrei[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Norway[0.041]; Norway[0.041]; Britain[0.040]; Britain[0.040]; Tarasov[0.040]; Kazakstan[0.040]; Markov[0.039]; France[0.039]; Maksim[0.039]; Dmitri[0.039]; Algeria[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belarus ==> ENTITY: \u001b[32mBelarus\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.149:0.149[0]; log p(e|m)= -0.422:-0.422[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMen[0.047]; Men[0.047]; Belarus[0.046]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Lithuania[0.043]; Romania[0.043]; Ireland[0.042]; Women[0.041]; Women[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Privalova[0.041]; Alekna[0.040]; Netherlands[0.040]; Vladimir[0.040]; Britain[0.039]; Algeria[0.039]; Spain[0.039]; Jamaica[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Brahim Lahlafi\u001b[39m ==> ENTITY: \u001b[32mBrahim Lahlafi\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.105:0.105[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTrine[0.044]; Nele[0.043]; Ethiopia[0.043]; javelin[0.043]; Morocco[0.043]; Morocco[0.043]; Tanja[0.042]; Kenya[0.042]; Kenya[0.042]; Kenya[0.042]; Kenya[0.042]; Stefka[0.042]; Hanne[0.042]; Paul[0.042]; Isel[0.041]; Fita[0.041]; jump[0.041]; jump[0.041]; Haugland[0.041]; Kostadinova[0.041]; Norway[0.040]; Norway[0.040]; Norway[0.040]; Shem[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gillian Russell\u001b[39m ==> ENTITY: \u001b[32mGillian Russell\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.064:0.064[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nathletics[0.050]; Athletics[0.048]; hurdles[0.047]; hurdles[0.047]; Florian[0.047]; Thorsett[0.046]; Engquist[0.046]; Jamaica[0.046]; Jamaica[0.046]; Jamaica[0.046]; Women[0.045]; Women[0.045]; Germany[0.045]; Germany[0.045]; Canada[0.045]; Grand[0.045]; Grand[0.045]; Men[0.044]; Brigita[0.044]; Tony[0.044]; Baumann[0.044]; Slovenia[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.045]; Denmark[0.043]; France[0.043]; pole[0.043]; Marc[0.042]; David[0.042]; Norway[0.042]; Norway[0.042]; Norway[0.042]; Jamaica[0.042]; Michael[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Jon[0.041]; Wilson[0.041]; Kenya[0.041]; Kenya[0.041]; Lithuania[0.040]; Davidson[0.040]; Kipketer[0.040]; Dennis[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nigeria ==> ENTITY: \u001b[32mNigeria\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.315:-0.315[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNigeria[0.047]; Nigeria[0.047]; Nigeria[0.047]; Men[0.047]; Men[0.047]; Men[0.047]; Women[0.045]; hurdles[0.044]; Australia[0.044]; Kenya[0.044]; Kenya[0.044]; Kenya[0.044]; seconds[0.043]; seconds[0.043]; Barsosio[0.042]; Italy[0.042]; Britain[0.041]; Jamaica[0.041]; Sweden[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marko Koers\u001b[39m ==> ENTITY: \u001b[32mMarko Koers\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.044]; Belgium[0.044]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Juergen[0.042]; Denmark[0.042]; Lars[0.042]; Andreas[0.042]; Norway[0.042]; Ottey[0.042]; discus[0.041]; Alekna[0.041]; Claus[0.041]; Geir[0.040]; Merlene[0.040]; Mayock[0.040]; Kenya[0.039]; Kenya[0.039]; Spain[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Patrick Stevens\u001b[39m ==> ENTITY: \u001b[32mPatrick Stevens\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.125:0.125[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.044]; Johnson[0.044]; Belgium[0.043]; Norway[0.043]; Denmark[0.043]; Michael[0.042]; Geir[0.042]; Kenya[0.042]; Kenya[0.042]; Trinidad[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Florian[0.041]; Busemann[0.041]; Thorsett[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Astrid[0.041]; Jon[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Phylis Smith\u001b[39m ==> ENTITY: \u001b[32mPhylis Smith\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.098:0.098[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.044]; Jamaica[0.044]; Matete[0.044]; hurdles[0.043]; Bailey[0.043]; Wilson[0.043]; Kipketer[0.042]; Canada[0.042]; Anja[0.042]; Geir[0.042]; Australia[0.042]; Barsosio[0.041]; Drummond[0.041]; Robinson[0.041]; Nigeria[0.041]; Nigeria[0.041]; Nigeria[0.041]; Nigeria[0.041]; Miles[0.041]; Sally[0.041]; seconds[0.040]; seconds[0.040]; Ruecker[0.040]; Norway[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Leah Pells\u001b[39m ==> ENTITY: \u001b[32mLeah Pells\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.046]; Jamaica[0.044]; Jamaica[0.044]; Jamaica[0.044]; athletics[0.044]; hurdles[0.042]; hurdles[0.042]; Michelle[0.042]; Jack[0.042]; Ludmila[0.041]; Svetlana[0.041]; Dionne[0.041]; Gillian[0.041]; Florian[0.041]; Busemann[0.041]; Thorsett[0.041]; Sara[0.041]; Engquist[0.041]; Women[0.040]; Women[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Julie[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jesse Owens\u001b[39m ==> ENTITY: \u001b[32mJesse Owens\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.072:0.072[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; George[0.043]; Michael[0.043]; Michael[0.043]; Michael[0.043]; Johnson[0.042]; Trine[0.042]; Canada[0.041]; relay[0.041]; Fredericks[0.041]; Harrison[0.041]; Norway[0.041]; Norway[0.041]; Britain[0.041]; javelin[0.040]; Charles[0.040]; Sweden[0.040]; Tanja[0.040]; Christie[0.040]; Australia[0.040]; race[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.146:0.146[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.047]; Denmark[0.047]; Britain[0.045]; Britain[0.045]; Men[0.043]; Men[0.043]; Men[0.043]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Canada[0.041]; Women[0.040]; Russia[0.039]; Russia[0.039]; Hungary[0.039]; Geir[0.039]; Kenya[0.038]; Kenya[0.038]; seconds[0.038]; Lobinger[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zambia ==> ENTITY: \u001b[32mZambia\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.157:0.157[0]; log p(e|m)= -0.320:-0.320[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKenya[0.046]; Kenya[0.046]; Kenya[0.046]; Kenya[0.046]; Men[0.044]; Men[0.044]; Ethiopia[0.044]; Australia[0.042]; Britain[0.042]; Barsosio[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Nigeria[0.041]; Nigeria[0.041]; Nigeria[0.041]; Bahamas[0.040]; Women[0.040]; Women[0.040]; Tegla[0.039]; Privalova[0.039]; Romania[0.038]; Sturrup[0.038]; Italy[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMen[0.051]; Men[0.051]; Germany[0.050]; Germany[0.050]; Germany[0.050]; Spain[0.047]; Romania[0.046]; Women[0.046]; Women[0.046]; Lithuania[0.046]; Privalova[0.046]; Ireland[0.045]; Britain[0.044]; Netherlands[0.044]; Andreas[0.043]; discus[0.043]; Algeria[0.043]; Juergen[0.042]; Belarus[0.042]; Belarus[0.042]; Miller[0.042]; Russia[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Falk Balzer\u001b[39m ==> ENTITY: \u001b[32mFalk Balzer\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.049]; Germany[0.049]; Germany[0.049]; Sweden[0.043]; Australia[0.043]; Trine[0.043]; Canada[0.043]; Florian[0.043]; Finland[0.042]; Britain[0.041]; Norway[0.041]; race[0.040]; Nigeria[0.040]; Nigeria[0.040]; Nigeria[0.040]; Isel[0.040]; McPaul[0.040]; Michael[0.040]; Michael[0.040]; Louise[0.040]; Belarus[0.039]; Linford[0.039]; Greece[0.039]; Peter[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.096:0.096[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.046]; Trinidad[0.044]; Denmark[0.043]; Britain[0.043]; Russia[0.042]; Russia[0.042]; Russia[0.042]; Russia[0.042]; Russia[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Portugal[0.042]; Norway[0.041]; Johnson[0.041]; Cuba[0.040]; Cuba[0.040]; Men[0.040]; Men[0.040]; Michael[0.040]; Belgium[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIreland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIreland\u001b[39m <---> \u001b[32mRepublic of Ireland\u001b[39m\t\nSCORES: global= 0.261:0.259[\u001b[31m0.003\u001b[39m]; local()= 0.113:0.090[\u001b[31m0.023\u001b[39m]; log p(e|m)= -0.566:-1.505[\u001b[31m0.939\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.049]; Spain[0.045]; Michael[0.044]; Burundi[0.044]; Kenya[0.044]; Kenya[0.044]; Jamaica[0.044]; William[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; John[0.043]; Netherlands[0.043]; Men[0.042]; Men[0.042]; Algeria[0.042]; Lithuania[0.042]; Onyali[0.042]; Sullivan[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tanja Damaske\u001b[39m ==> ENTITY: \u001b[32mTanja Damaske\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \njavelin[0.045]; Germany[0.045]; Germany[0.045]; Fredericks[0.044]; Trine[0.044]; Tom[0.042]; Lithuania[0.041]; Rantanen[0.041]; Britain[0.041]; Britain[0.041]; Women[0.041]; Fita[0.041]; Isel[0.041]; Sweden[0.041]; Owens[0.041]; McPaul[0.041]; relay[0.041]; jump[0.041]; Mike[0.041]; Belarus[0.040]; Natalya[0.040]; race[0.040]; Norway[0.040]; Norway[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tatyana Motkova\u001b[39m ==> ENTITY: \u001b[32mTatyana Babashkina\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \njump[0.046]; jump[0.046]; Belarus[0.044]; Lithuania[0.044]; high[0.043]; Lobinger[0.042]; Yelena[0.042]; Nele[0.042]; Kostadinova[0.041]; Natalya[0.041]; Ukraine[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; seconds[0.041]; Andrei[0.041]; Stefka[0.041]; Hanne[0.040]; Bulgaria[0.040]; Alina[0.040]; Haugland[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jon Drummond\u001b[39m ==> ENTITY: \u001b[32mJon Drummond\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.043:0.043[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.046]; Jamaica[0.043]; Canada[0.043]; Denmark[0.042]; Sweden[0.042]; Kipketer[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Craig[0.042]; Britain[0.042]; Britain[0.042]; Norway[0.041]; Norway[0.041]; Geir[0.041]; seconds[0.041]; David[0.041]; Anja[0.041]; Michael[0.041]; Bailey[0.040]; Marc[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.151:0.151[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Denmark[0.045]; Spain[0.044]; Belgium[0.044]; Netherlands[0.043]; Ireland[0.043]; Lithuania[0.043]; Norway[0.042]; Britain[0.042]; Algeria[0.040]; Johnson[0.040]; Namibia[0.040]; Belarus[0.039]; Belarus[0.039]; Russia[0.039]; Russia[0.039]; minutes[0.039]; Marko[0.038]; Trinidad[0.038]; Anthony[0.038]; Michael[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.162:0.162[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.047]; Norway[0.047]; Denmark[0.047]; Britain[0.045]; Men[0.043]; Men[0.043]; Men[0.043]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Canada[0.041]; Women[0.040]; Bulgaria[0.039]; Russia[0.039]; Russia[0.039]; Russia[0.039]; Russia[0.039]; Belarus[0.039]; France[0.039]; Ukraine[0.039]; Geir[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Falilat Ogunkoya\u001b[39m ==> ENTITY: \u001b[32mFalilat Ogunkoya\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNigeria[0.043]; Nigeria[0.043]; Nigeria[0.043]; Nigeria[0.043]; Matete[0.043]; Loroupe[0.043]; Barsosio[0.043]; Jamaica[0.042]; Kenya[0.042]; Kenya[0.042]; Kenya[0.042]; Anja[0.042]; Canada[0.042]; Australia[0.041]; Geir[0.041]; Tegla[0.041]; Zambia[0.041]; Jearl[0.041]; Women[0.040]; Gunhild[0.040]; seconds[0.040]; seconds[0.040]; Phylis[0.040]; Ruecker[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jack Pierce ==> ENTITY: \u001b[32mJack Pierce (athlete)\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.094:0.094[0]; log p(e|m)= -0.899:-0.899[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJohnson[0.044]; seconds[0.044]; seconds[0.044]; Canada[0.042]; Garcia[0.042]; Sacramento[0.042]; Morceli[0.042]; Patrick[0.042]; Geir[0.042]; Tony[0.042]; Brown[0.042]; Trinidad[0.041]; Florian[0.041]; Steve[0.041]; Britain[0.041]; Busemann[0.041]; Thorsett[0.041]; Michael[0.041]; hurdles[0.041]; Ivan[0.041]; Stevens[0.041]; shot[0.041]; Frankie[0.041]; Valle[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kenya ==> ENTITY: \u001b[32mKenya\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.206:0.206[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKenya[0.056]; Kenya[0.056]; Kenya[0.056]; Ethiopia[0.047]; Zambia[0.046]; Nigeria[0.043]; Nigeria[0.043]; Tegla[0.042]; Loroupe[0.041]; Jamaica[0.040]; Cheruiyot[0.040]; Barsosio[0.039]; Australia[0.039]; Norway[0.039]; Russia[0.039]; Wami[0.038]; Sweden[0.038]; Konga[0.038]; Finland[0.038]; Matete[0.037]; Italy[0.037]; Bahamas[0.036]; Nylander[0.036]; Hungary[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lithuania ==> ENTITY: \u001b[32mLithuania\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.397:-0.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.045]; Cuba[0.044]; Cuba[0.044]; Cuba[0.044]; Men[0.044]; Sweden[0.043]; Greece[0.042]; Finland[0.042]; Norway[0.042]; Norway[0.042]; Portugal[0.042]; Natalya[0.041]; Britain[0.041]; Britain[0.041]; Australia[0.041]; Canada[0.041]; Jamaica[0.041]; Women[0.041]; Armenia[0.041]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Carlos[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.066:0.066[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.044]; Britain[0.044]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Germany[0.043]; Germany[0.043]; Portugal[0.043]; Norway[0.042]; Norway[0.042]; Cuba[0.041]; Men[0.041]; Men[0.041]; Michael[0.041]; Markov[0.041]; Lithuania[0.041]; Charles[0.040]; Carlos[0.040]; Bulgaria[0.040]; Ukraine[0.040]; triple[0.040]; Morocco[0.040]; Morocco[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burundi ==> ENTITY: \u001b[32mBurundi\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.374:-0.374[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKenya[0.047]; Kenya[0.047]; Belgium[0.044]; Namibia[0.044]; Netherlands[0.043]; Algeria[0.042]; Denmark[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Norway[0.042]; Spain[0.041]; Ireland[0.040]; Lithuania[0.040]; Britain[0.040]; Belarus[0.040]; Belarus[0.040]; Niyongabo[0.039]; Cuba[0.039]; minutes[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.178:0.178[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.051]; Germany[0.051]; Germany[0.051]; Germany[0.051]; Denmark[0.047]; Lithuania[0.046]; France[0.045]; Britain[0.045]; Norway[0.045]; Norway[0.045]; Norway[0.045]; Russia[0.044]; Russia[0.044]; Russia[0.044]; Russia[0.044]; Russia[0.044]; Russia[0.044]; Bulgaria[0.044]; Algeria[0.043]; jump[0.043]; Belarus[0.043]; Ukraine[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jamaica ==> ENTITY: \u001b[32mJamaica\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.333:-0.333[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJamaica[0.049]; Jamaica[0.049]; Russia[0.045]; Russia[0.045]; Russia[0.045]; Canada[0.043]; Britain[0.042]; seconds[0.042]; seconds[0.042]; seconds[0.042]; Switzerland[0.041]; France[0.041]; Cuba[0.041]; hurdles[0.040]; hurdles[0.040]; Athletics[0.040]; Sweden[0.040]; Russell[0.040]; Dionne[0.039]; Yekaterina[0.039]; Germany[0.039]; Slovenia[0.039]; Mark[0.039]; Portugal[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Olabisi Afolabi\u001b[39m ==> ENTITY: \u001b[32mOlabisi Afolabi\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNigeria[0.049]; Nigeria[0.049]; Nigeria[0.049]; Nigeria[0.049]; Ogunkoya[0.046]; Kipketer[0.046]; Barsosio[0.045]; Matete[0.044]; Kenya[0.044]; Kenya[0.044]; Kenya[0.044]; Anja[0.043]; Jamaica[0.043]; Geir[0.042]; Canada[0.042]; Langat[0.041]; Norway[0.041]; Women[0.041]; Sweden[0.041]; Ruecker[0.041]; Konga[0.040]; Karin[0.040]; hurdles[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Frankie Fredericks\u001b[39m ==> ENTITY: \u001b[32mFrankie Fredericks\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.094:0.094[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJohnson[0.052]; Morceli[0.050]; seconds[0.049]; seconds[0.049]; seconds[0.049]; Geir[0.048]; Trinidad[0.048]; Brown[0.048]; Florian[0.048]; Michael[0.047]; Busemann[0.047]; Tony[0.047]; Thorsett[0.047]; Namibia[0.047]; Stevens[0.047]; Boldon[0.047]; Patrick[0.046]; Britain[0.046]; hurdles[0.046]; Norway[0.046]; Canada[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kenya ==> ENTITY: \u001b[32mKenya\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.218:0.218[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKenya[0.055]; Kenya[0.055]; Kenya[0.055]; Ethiopia[0.047]; Zambia[0.046]; Nigeria[0.043]; Nigeria[0.043]; Nigeria[0.043]; Tegla[0.042]; Loroupe[0.040]; Jamaica[0.040]; Cheruiyot[0.040]; Barsosio[0.039]; Australia[0.039]; Norway[0.039]; Russia[0.039]; Wami[0.038]; Sweden[0.038]; Konga[0.037]; Yusuf[0.037]; Finland[0.037]; Matete[0.037]; Italy[0.036]; Bahamas[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Carla Sacramento\u001b[39m ==> ENTITY: \u001b[32mCarla Sacramento\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.092:0.092[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPortugal[0.048]; athletics[0.047]; Athletics[0.045]; minutes[0.044]; Florian[0.044]; Busemann[0.044]; Thorsett[0.044]; Jamaica[0.044]; Jamaica[0.044]; Jamaica[0.044]; Engquist[0.044]; Women[0.043]; Women[0.043]; seconds[0.043]; seconds[0.043]; seconds[0.043]; hurdles[0.043]; hurdles[0.043]; Emilio[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tom Nyariki\u001b[39m ==> ENTITY: \u001b[32mTom Nyariki\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.104:0.104[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKenya[0.049]; Kenya[0.049]; Kenya[0.049]; Kenya[0.049]; Trine[0.045]; Koech[0.045]; Nele[0.044]; javelin[0.043]; Tanja[0.043]; Stefka[0.043]; Hanne[0.043]; Fita[0.042]; Ethiopia[0.042]; Isel[0.042]; McPaul[0.042]; Bob[0.042]; Haugland[0.042]; Kenny[0.041]; Kostadinova[0.041]; Norway[0.041]; Norway[0.041]; Norway[0.041]; Shem[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Eric Thomas ==> ENTITY: \u001b[32mEric Thomas (athlete)\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.060:0.060[0]; log p(e|m)= -1.005:-1.005[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMichael[0.045]; Derrick[0.044]; Matete[0.044]; Britain[0.043]; Sweden[0.043]; Jon[0.043]; Anja[0.043]; Italy[0.042]; Torrance[0.042]; Donovan[0.042]; Barsosio[0.041]; Australia[0.041]; Smith[0.041]; Gunhild[0.041]; Romania[0.041]; Gete[0.041]; Samuel[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Canada[0.041]; Norway[0.040]; Dennis[0.040]; Ruecker[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gete Wami\u001b[39m ==> ENTITY: \u001b[32mGete Wami\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.108:0.108[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEthiopia[0.049]; Barsosio[0.048]; Cheruiyot[0.047]; Matete[0.047]; Juergen[0.047]; seconds[0.046]; seconds[0.046]; seconds[0.046]; Alekna[0.045]; Privalova[0.045]; Chryste[0.045]; Onyali[0.045]; Loroupe[0.045]; Andreas[0.045]; Michael[0.045]; Devers[0.045]; Merlene[0.045]; hurdles[0.044]; Gunhild[0.044]; minutes[0.044]; Ottey[0.044]; Norway[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Netherlands ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.180:0.180[0]; log p(e|m)= -0.250:-0.250[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.045]; Denmark[0.045]; Cuba[0.045]; Lithuania[0.044]; Norway[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Kenya[0.044]; Kenya[0.044]; Spain[0.044]; Ireland[0.043]; Britain[0.042]; Belarus[0.042]; Belarus[0.042]; Russia[0.042]; Russia[0.042]; Algeria[0.041]; Jamaica[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.096:0.096[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.047]; Donovan[0.044]; Belarus[0.043]; Britain[0.043]; Russia[0.043]; Russia[0.043]; Germany[0.042]; Germany[0.042]; Portugal[0.042]; Finland[0.042]; Norway[0.041]; Norway[0.041]; Cuba[0.041]; Cuba[0.041]; Cuba[0.041]; Men[0.041]; Men[0.041]; Men[0.041]; Australia[0.040]; Michael[0.040]; Michael[0.040]; Lithuania[0.040]; Lithuania[0.040]; Charles[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.151:0.151[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Switzerland[0.045]; Denmark[0.044]; Portugal[0.043]; France[0.042]; Britain[0.042]; Belgium[0.042]; Canada[0.042]; Norway[0.041]; Johnson[0.040]; Men[0.040]; Men[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; minutes[0.039]; Trinidad[0.039]; Michael[0.039]; Women[0.039]; Women[0.039]; Claus[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Louise McPaul\u001b[39m ==> ENTITY: \u001b[32mLouise Currey\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.121:0.121[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; javelin[0.048]; race[0.047]; relay[0.046]; jump[0.046]; Trine[0.046]; Fita[0.044]; Tanja[0.044]; seconds[0.044]; Finland[0.043]; Florian[0.043]; Fredericks[0.043]; Isel[0.043]; Jamaica[0.042]; George[0.042]; Linford[0.042]; Peter[0.042]; Britain[0.042]; Britain[0.042]; Norway[0.042]; Norway[0.042]; Canada[0.041]; Kenny[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.221:0.221[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.047]; Russia[0.047]; Russia[0.047]; Russia[0.047]; Russia[0.047]; Russia[0.047]; Lithuania[0.045]; Ukraine[0.045]; Bulgaria[0.044]; Andrei[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Norway[0.042]; Britain[0.042]; Britain[0.042]; Tarasov[0.042]; Kazakstan[0.041]; Markov[0.041]; France[0.040]; Maksim[0.040]; Dmitri[0.040]; triple[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Germany[0.046]; Britain[0.044]; Britain[0.044]; Denmark[0.043]; France[0.041]; Norway[0.040]; Norway[0.040]; Algeria[0.040]; Canada[0.040]; Jamaica[0.040]; Kenya[0.040]; Kenya[0.040]; Donovan[0.039]; Russia[0.039]; Russia[0.039]; Nigeria[0.039]; Nigeria[0.039]; Nigeria[0.039]; Nigeria[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.061:0.061[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.044]; Sweden[0.044]; Russia[0.044]; Italy[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Romania[0.043]; Finland[0.042]; Norway[0.042]; Hungary[0.042]; Miller[0.042]; Thomas[0.042]; Men[0.041]; Men[0.041]; Australia[0.041]; Smith[0.040]; Privalova[0.040]; Nigeria[0.040]; Nigeria[0.040]; Nigeria[0.040]; Torrance[0.040]; Samuel[0.039]; Robinson[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.110:0.110[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.046]; Donovan[0.044]; Belarus[0.043]; Britain[0.043]; Sweden[0.043]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Portugal[0.042]; Finland[0.041]; George[0.041]; Norway[0.041]; Norway[0.041]; Jamaica[0.041]; Johnson[0.041]; Cuba[0.041]; Cuba[0.041]; Men[0.040]; Australia[0.040]; Michael[0.040]; Michael[0.040]; Michael[0.040]; Lithuania[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Astrid Kumbernuss\u001b[39m ==> ENTITY: \u001b[32mAstrid Kumbernuss\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.051:0.051[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLars[0.044]; discus[0.044]; shot[0.043]; Geir[0.043]; Florian[0.043]; Busemann[0.043]; Mayock[0.042]; Norway[0.042]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Men[0.041]; Men[0.041]; Men[0.041]; Ivan[0.041]; Tanui[0.041]; Marcus[0.041]; Noureddine[0.041]; Michael[0.041]; Denmark[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nigeria ==> ENTITY: \u001b[32mNigeria\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.315:-0.315[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNigeria[0.048]; Nigeria[0.048]; Men[0.048]; javelin[0.047]; Women[0.046]; Australia[0.045]; Lithuania[0.044]; relay[0.044]; seconds[0.044]; Britain[0.043]; Jamaica[0.043]; Sweden[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Johnson[0.042]; Norway[0.042]; Canada[0.042]; Belarus[0.041]; Donovan[0.041]; Finland[0.041]; George[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.050]; Germany[0.050]; Germany[0.050]; Hungary[0.046]; Denmark[0.046]; Sweden[0.046]; Italy[0.045]; Britain[0.044]; Norway[0.044]; Sven[0.043]; Dusan[0.042]; Women[0.042]; Donovan[0.042]; Canada[0.042]; Australia[0.041]; Eric[0.041]; Nigeria[0.041]; Nigeria[0.041]; Nigeria[0.041]; Nigeria[0.041]; Thomas[0.041]; Michael[0.041]; Marc[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tegla Loroupe\u001b[39m ==> ENTITY: \u001b[32mTegla Loroupe\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEthiopia[0.046]; Kenya[0.045]; Kenya[0.045]; Kenya[0.045]; Kenya[0.045]; Barsosio[0.045]; Matete[0.043]; Wami[0.041]; Chryste[0.041]; Onyali[0.041]; Cheruiyot[0.041]; Australia[0.041]; Merlene[0.041]; Bahamas[0.041]; Jamaica[0.040]; Gunhild[0.040]; Norway[0.040]; Sweden[0.040]; Gete[0.040]; Ottey[0.040]; Zambia[0.040]; Women[0.039]; Women[0.039]; Women[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.126:0.126[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.048]; Sweden[0.046]; Britain[0.046]; Britain[0.046]; Men[0.044]; Men[0.044]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Finland[0.042]; Canada[0.041]; Portugal[0.041]; Morocco[0.040]; race[0.040]; Fita[0.040]; Women[0.040]; Belarus[0.040]; relay[0.040]; Lithuania[0.040]; Australia[0.039]; Kenya[0.039]; Kenya[0.039]; seconds[0.039]; Armenia[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.062:0.062[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.044]; Belarus[0.044]; Britain[0.044]; Russia[0.044]; Washington[0.043]; Devers[0.043]; Finland[0.042]; Jamaica[0.042]; Norway[0.042]; Miller[0.042]; Ireland[0.042]; Men[0.041]; Men[0.041]; Michael[0.041]; Lithuania[0.041]; Sullivan[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; hurdles[0.040]; Privalova[0.040]; Bahamas[0.040]; discus[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Berlin ==> ENTITY: \u001b[32mBerlin\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.061:0.061[0]; log p(e|m)= -0.154:-0.154[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBerlin[0.048]; Berlin[0.048]; Switzerland[0.045]; Russell[0.043]; Athletics[0.042]; Jamaica[0.042]; Jamaica[0.042]; Jamaica[0.042]; Cuba[0.042]; Sweden[0.041]; athletics[0.041]; Friday[0.041]; Russia[0.041]; Russia[0.041]; France[0.041]; Prix[0.041]; Prix[0.041]; Slovenia[0.041]; Baumann[0.040]; Dionne[0.040]; Lopez[0.040]; Grand[0.040]; Grand[0.040]; Freeman[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Osmond Ezinwa\u001b[39m ==> ENTITY: \u001b[32mOsmond Ezinwa\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.112:0.112[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDavidson[0.046]; relay[0.045]; Trine[0.044]; Nigeria[0.043]; Nigeria[0.043]; Nigeria[0.043]; Sweden[0.042]; javelin[0.042]; Jamaica[0.042]; Tanja[0.042]; Norway[0.042]; Johnson[0.041]; Florian[0.041]; Fredericks[0.041]; Isel[0.041]; McPaul[0.041]; Canada[0.041]; race[0.040]; Rantanen[0.040]; Cuba[0.040]; Cuba[0.040]; Australia[0.040]; Finland[0.040]; Namibia[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Morocco ==> ENTITY: \u001b[32mMorocco\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.296:-0.296[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMorocco[0.049]; Britain[0.043]; Portugal[0.042]; Cuba[0.042]; Cuba[0.042]; Kenya[0.042]; Kenya[0.042]; Kenya[0.042]; Kenya[0.042]; Men[0.041]; Men[0.041]; Bulgaria[0.041]; Norway[0.041]; Norway[0.041]; Norway[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Ethiopia[0.041]; Fita[0.041]; seconds[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.167:0.167[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; Germany[0.047]; Germany[0.047]; Germany[0.047]; Denmark[0.043]; Lithuania[0.043]; Canada[0.042]; France[0.041]; Britain[0.041]; Norway[0.041]; Norway[0.041]; Norway[0.041]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Bulgaria[0.040]; Algeria[0.040]; jump[0.040]; Donovan[0.039]; Belarus[0.039]; Ukraine[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Michael Moellenbeck\u001b[39m ==> ENTITY: \u001b[32mMichael Möllenbeck\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.106:0.106[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAndreas[0.045]; Juergen[0.045]; Riedel[0.044]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; discus[0.043]; Lars[0.043]; Alekna[0.042]; Chryste[0.042]; Onyali[0.042]; Merlene[0.041]; Schult[0.041]; Mayock[0.041]; Gete[0.041]; Ireland[0.040]; Privalova[0.040]; Men[0.040]; Tanui[0.040]; Marcus[0.040]; Vladimir[0.040]; Romania[0.039]; Netherlands[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Irina Korzhanenko\u001b[39m ==> ENTITY: \u001b[32mIrina Korzhanenko\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.054:0.054[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.044]; Russia[0.044]; Lars[0.044]; discus[0.044]; Belarus[0.043]; Geir[0.043]; Women[0.042]; Busemann[0.042]; Mayock[0.042]; Norway[0.041]; Britain[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Men[0.041]; Men[0.041]; Men[0.041]; Ivan[0.041]; Astrid[0.041]; Tanui[0.041]; Marcus[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jean Galfione\u001b[39m ==> ENTITY: \u001b[32mJean Galfione\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLobinger[0.049]; Marc[0.048]; Kipketer[0.047]; vault[0.047]; Nele[0.047]; pole[0.046]; Craig[0.046]; David[0.046]; Inga[0.045]; Stefka[0.045]; Hanne[0.045]; France[0.045]; Daniel[0.045]; Tim[0.045]; Paul[0.044]; Haugland[0.044]; Dmitri[0.044]; Lithuania[0.044]; Kiptoo[0.044]; Rodal[0.044]; Belarus[0.044]; Kostadinova[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mark Crear\u001b[39m ==> ENTITY: \u001b[32mMark Crear\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.065:0.065[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhurdles[0.045]; Trinidad[0.044]; Tony[0.043]; Geir[0.043]; Florian[0.042]; Busemann[0.042]; Michael[0.042]; Thorsett[0.042]; Johnson[0.042]; seconds[0.042]; seconds[0.042]; seconds[0.042]; Steve[0.042]; Norway[0.041]; Jamaica[0.041]; Jamaica[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Men[0.041]; Men[0.041]; Ivan[0.041]; Brigita[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yelena Gulyayeva\u001b[39m ==> ENTITY: \u001b[32mYelena Gulyayeva\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.102:0.102[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \njump[0.044]; jump[0.044]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Russia[0.043]; Belarus[0.043]; high[0.043]; Nele[0.042]; Lobinger[0.041]; Stefka[0.041]; Hanne[0.041]; Ukraine[0.041]; Lithuania[0.041]; Tatyana[0.040]; Haugland[0.040]; Kazakstan[0.040]; Kostadinova[0.040]; Britain[0.040]; Britain[0.040]; Fita[0.040]; Norway[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marc Blume\u001b[39m ==> ENTITY: \u001b[32mMarc Blume\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; seconds[0.043]; Lobinger[0.043]; Anja[0.042]; Geir[0.042]; Rodal[0.041]; Kipketer[0.040]; Jon[0.040]; Ruecker[0.040]; Norway[0.040]; Norway[0.040]; Karin[0.040]; Men[0.040]; Men[0.040]; Men[0.040]; Jearl[0.040]; Nico[0.039]; Kiptoo[0.039]; \t\n\r [===>..........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 16s99ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 438/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 966testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bailey\u001b[39m ==> ENTITY: \u001b[32mDonovan Bailey\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.047]; Olympic[0.047]; record[0.045]; Games[0.045]; champion[0.044]; sprinting[0.044]; sprint[0.043]; Athletics[0.043]; despite[0.042]; seconds[0.042]; time[0.042]; Champions[0.041]; Bailey[0.040]; Bailey[0.040]; Bailey[0.040]; finish[0.040]; Adrian[0.040]; world[0.040]; world[0.040]; champions[0.039]; Dennis[0.039]; start[0.039]; Atlanta[0.039]; win[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mitchell\u001b[39m ==> ENTITY: \u001b[32mDennis Mitchell\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.046]; Olympic[0.046]; Olympic[0.046]; Olympics[0.044]; series[0.044]; Games[0.044]; champion[0.043]; champion[0.043]; sport[0.043]; medals[0.043]; American[0.042]; Jamaican[0.041]; Bailey[0.041]; Bailey[0.041]; relay[0.040]; world[0.040]; gold[0.039]; champions[0.039]; champions[0.039]; month[0.039]; second[0.039]; Merlene[0.039]; Devers[0.039]; women[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brussels ==> ENTITY: \u001b[32mBrussels\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.086:0.086[0]; log p(e|m)= -0.025:-0.025[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstadium[0.047]; Brussels[0.045]; record[0.044]; month[0.043]; sport[0.043]; Berlin[0.042]; Berlin[0.042]; Berlin[0.042]; Friday[0.041]; Friday[0.041]; champions[0.041]; champions[0.041]; Olympic[0.041]; Olympic[0.041]; Olympic[0.041]; set[0.041]; grand[0.041]; Olympics[0.041]; win[0.040]; win[0.040]; American[0.040]; American[0.040]; Games[0.040]; second[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gail Devers\u001b[39m ==> ENTITY: \u001b[32mGail Devers\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.047]; Olympic[0.047]; Olympics[0.045]; champion[0.043]; champion[0.043]; Despite[0.043]; medals[0.043]; second[0.042]; Ottey[0.042]; relay[0.042]; sport[0.041]; Mitchell[0.041]; beat[0.041]; win[0.041]; Merlene[0.041]; month[0.040]; Jamaican[0.040]; stadium[0.040]; mark[0.040]; world[0.040]; Earlier[0.040]; said[0.040]; watch[0.039]; Today[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = American ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.130:-0.130[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseries[0.050]; American[0.049]; Michael[0.045]; record[0.045]; second[0.045]; fourth[0.045]; Dennis[0.045]; win[0.044]; win[0.044]; major[0.043]; start[0.043]; Donovan[0.042]; Warner[0.042]; Atlanta[0.042]; Atlanta[0.042]; Mitchell[0.042]; Mitchell[0.042]; Mitchell[0.042]; finished[0.042]; strong[0.042]; Jamaica[0.042]; lucrative[0.042]; month[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Berlin ==> ENTITY: \u001b[32mBerlin\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.038:0.038[0]; log p(e|m)= -0.154:-0.154[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.051]; Olympic[0.051]; Olympic[0.051]; Games[0.049]; Games[0.049]; Berlin[0.049]; Warner[0.046]; Michael[0.045]; Dennis[0.045]; American[0.044]; American[0.044]; Mitchell[0.044]; Mitchell[0.044]; Mitchell[0.044]; Mitchell[0.044]; Donovan[0.043]; record[0.043]; champion[0.043]; Athletics[0.043]; win[0.043]; win[0.043]; second[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jesse Owens\u001b[39m ==> ENTITY: \u001b[32mJesse Owens\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympics[0.049]; Olympic[0.048]; Olympic[0.048]; Olympic[0.048]; Games[0.045]; American[0.043]; sport[0.043]; Berlin[0.042]; Berlin[0.042]; medals[0.041]; Michael[0.041]; champion[0.040]; champion[0.040]; relay[0.040]; win[0.040]; world[0.040]; fourth[0.039]; upstaged[0.039]; stadium[0.039]; Zurich[0.039]; Zurich[0.039]; month[0.039]; second[0.039]; second[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Atlanta Games\u001b[39m ==> ENTITY: \u001b[32m1996 Summer Olympics\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.049]; Olympic[0.049]; Olympic[0.049]; Olympics[0.048]; sport[0.043]; Jamaica[0.043]; medals[0.042]; held[0.042]; fourth[0.041]; second[0.040]; start[0.040]; champion[0.040]; finished[0.040]; American[0.040]; record[0.039]; finishing[0.039]; stadium[0.039]; despite[0.039]; month[0.039]; champions[0.039]; champions[0.039]; champions[0.039]; series[0.039]; gold[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brussels ==> ENTITY: \u001b[32mBrussels\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.025:-0.025[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstadium[0.049]; Brussels[0.047]; record[0.046]; month[0.045]; sport[0.044]; Berlin[0.044]; Berlin[0.044]; clocking[0.043]; Friday[0.043]; Friday[0.043]; champions[0.043]; champions[0.043]; Olympic[0.043]; Olympic[0.043]; Olympic[0.043]; set[0.043]; Olympics[0.042]; win[0.042]; win[0.042]; American[0.042]; American[0.042]; Games[0.042]; second[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Michael Green ==> ENTITY: \u001b[32mMichael Green (sprinter)\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.156:0.156[0]; log p(e|m)= -2.033:-2.033[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrecord[0.045]; series[0.045]; American[0.044]; Jamaica[0.043]; sport[0.042]; time[0.042]; fourth[0.042]; Games[0.042]; Games[0.042]; second[0.041]; world[0.041]; world[0.041]; world[0.041]; Bailey[0.041]; Bailey[0.041]; Bailey[0.041]; Bailey[0.041]; Bailey[0.041]; Bailey[0.041]; month[0.041]; Olympic[0.041]; Olympic[0.041]; Olympic[0.041]; sprint[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bailey\u001b[39m ==> ENTITY: \u001b[32mDonovan Bailey\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.213:0.213[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.052]; Olympic[0.052]; Olympic[0.052]; record[0.050]; Games[0.049]; Games[0.049]; champion[0.049]; champion[0.049]; sprinting[0.048]; sprint[0.048]; Athletics[0.047]; despite[0.047]; seconds[0.047]; Jamaica[0.046]; Michael[0.046]; time[0.046]; Champions[0.045]; finishing[0.045]; beat[0.045]; second[0.045]; fourth[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jamaican ==> ENTITY: \u001b[32mJamaica\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.047:0.047[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.047]; Olympic[0.047]; Olympics[0.047]; medals[0.044]; American[0.044]; sport[0.043]; clocking[0.043]; champion[0.043]; said[0.042]; second[0.042]; relay[0.042]; Despite[0.041]; Mitchell[0.041]; Devers[0.040]; Zurich[0.040]; Zurich[0.040]; belong[0.040]; form[0.040]; women[0.040]; Berlin[0.039]; Berlin[0.039]; gold[0.039]; lucrative[0.039]; mark[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Donovan Bailey\u001b[39m ==> ENTITY: \u001b[32mDonovan Bailey\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.046]; Olympic[0.046]; record[0.045]; Games[0.044]; champion[0.044]; sprinting[0.044]; sprint[0.043]; Athletics[0.042]; despite[0.042]; seconds[0.042]; Jamaica[0.042]; Michael[0.041]; time[0.041]; Champions[0.041]; Bailey[0.040]; Bailey[0.040]; Bailey[0.040]; finish[0.040]; finish[0.040]; Adrian[0.039]; world[0.039]; world[0.039]; champions[0.039]; Dennis[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bailey\u001b[39m ==> ENTITY: \u001b[32mDonovan Bailey\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.050]; Olympic[0.050]; Olympic[0.050]; Olympics[0.048]; record[0.048]; Games[0.047]; champion[0.047]; champion[0.047]; relay[0.046]; despite[0.045]; seconds[0.045]; Jamaica[0.045]; sport[0.044]; Michael[0.044]; clocking[0.043]; Devers[0.043]; medals[0.043]; finishing[0.043]; Despite[0.043]; beat[0.043]; second[0.043]; fourth[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Merlene Ottey\u001b[39m ==> ENTITY: \u001b[32mMerlene Ottey\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.050]; Olympic[0.050]; Olympics[0.049]; Jamaican[0.046]; sport[0.044]; medals[0.043]; clocking[0.042]; champion[0.042]; Devers[0.042]; relay[0.041]; mark[0.041]; women[0.041]; second[0.041]; American[0.040]; gold[0.040]; Berlin[0.040]; Mitchell[0.039]; Brussels[0.039]; said[0.039]; Zurich[0.039]; champions[0.038]; going[0.038]; crowd[0.038]; concentration[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Berlin ==> ENTITY: \u001b[32mBerlin\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.017:0.017[0]; log p(e|m)= -0.154:-0.154[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.048]; Olympic[0.048]; Games[0.046]; Berlin[0.046]; Warner[0.043]; Dennis[0.042]; American[0.041]; American[0.041]; Mitchell[0.041]; Mitchell[0.041]; Mitchell[0.041]; Donovan[0.041]; record[0.041]; champion[0.040]; Athletics[0.040]; win[0.040]; Bailey[0.040]; Bailey[0.040]; Bailey[0.040]; Bailey[0.040]; world[0.040]; world[0.040]; despite[0.039]; Champions[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mOlympics\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mOlympic Games\u001b[39m <---> \u001b[32mSummer Olympic Games\u001b[39m\t\nSCORES: global= 0.251:0.234[\u001b[31m0.017\u001b[39m]; local()= 0.203:0.209[\u001b[32m0.006\u001b[39m]; log p(e|m)= -0.821:-2.733[\u001b[31m1.912\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.050]; Olympic[0.050]; Olympic[0.050]; Games[0.047]; sport[0.045]; medals[0.042]; American[0.041]; relay[0.041]; champion[0.041]; champion[0.041]; world[0.041]; second[0.040]; win[0.040]; stadium[0.040]; Jamaican[0.040]; fourth[0.040]; lucrative[0.039]; month[0.039]; Berlin[0.039]; Berlin[0.039]; gold[0.039]; champions[0.039]; champions[0.039]; Devers[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlanta ==> ENTITY: \u001b[32mAtlanta\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.157:0.157[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.050]; Olympic[0.050]; Olympic[0.050]; Olympic[0.050]; Games[0.046]; Games[0.046]; Atlanta[0.045]; Jamaica[0.044]; sport[0.042]; fourth[0.042]; second[0.042]; Champions[0.042]; series[0.041]; start[0.041]; champion[0.041]; champion[0.041]; Friday[0.041]; Friday[0.041]; Friday[0.041]; American[0.041]; American[0.041]; win[0.041]; win[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zurich ==> ENTITY: \u001b[32mZürich\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.089:-0.089[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrecord[0.046]; Olympic[0.043]; Olympic[0.043]; Olympic[0.043]; Zurich[0.042]; clocking[0.042]; Berlin[0.042]; Berlin[0.042]; fourth[0.042]; Olympics[0.041]; Games[0.041]; win[0.041]; win[0.041]; Friday[0.041]; Friday[0.041]; stadium[0.041]; relay[0.041]; second[0.041]; meeting[0.041]; champion[0.041]; champion[0.041]; set[0.041]; victory[0.040]; month[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bailey\u001b[39m ==> ENTITY: \u001b[32mDonovan Bailey\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.051]; Olympic[0.051]; Olympic[0.051]; Olympics[0.050]; record[0.050]; Games[0.049]; Games[0.049]; champion[0.049]; sprinting[0.048]; sprint[0.048]; relay[0.047]; despite[0.047]; seconds[0.047]; Jamaica[0.046]; sport[0.046]; Michael[0.046]; medals[0.045]; finishing[0.045]; beat[0.045]; second[0.044]; fourth[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zurich ==> ENTITY: \u001b[32mZürich\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.089:-0.089[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrecord[0.046]; Olympic[0.043]; Olympic[0.043]; Olympic[0.043]; Zurich[0.042]; clocking[0.042]; Berlin[0.042]; Berlin[0.042]; fourth[0.042]; Olympics[0.041]; Games[0.041]; win[0.041]; win[0.041]; Friday[0.041]; Friday[0.041]; stadium[0.041]; relay[0.041]; second[0.041]; meeting[0.041]; champion[0.041]; champion[0.041]; set[0.041]; victory[0.040]; month[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jamaica ==> ENTITY: \u001b[32mJamaica\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.333:-0.333[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.047]; Olympic[0.047]; Olympic[0.047]; Games[0.043]; Games[0.043]; seconds[0.043]; second[0.042]; fourth[0.042]; sprint[0.042]; series[0.041]; finished[0.041]; record[0.041]; finishing[0.041]; relay[0.041]; American[0.040]; time[0.040]; Michael[0.040]; finish[0.040]; finish[0.040]; start[0.040]; champion[0.040]; champion[0.040]; despite[0.040]; sprinting[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mitchell\u001b[39m ==> ENTITY: \u001b[32mDennis Mitchell\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.045]; Olympic[0.045]; Olympic[0.045]; Mitchell[0.043]; Mitchell[0.043]; Mitchell[0.043]; Athletics[0.043]; Games[0.042]; Games[0.042]; champion[0.042]; record[0.042]; Dennis[0.042]; Adrian[0.041]; American[0.041]; American[0.041]; Champions[0.041]; time[0.040]; sprint[0.040]; Bailey[0.040]; Bailey[0.040]; Bailey[0.040]; Bailey[0.040]; Bailey[0.040]; Bailey[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Berlin ==> ENTITY: \u001b[32mBerlin\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.085:0.085[0]; log p(e|m)= -0.154:-0.154[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympics[0.049]; Olympic[0.048]; Olympic[0.048]; Olympic[0.048]; Games[0.045]; Berlin[0.045]; sport[0.043]; medals[0.042]; American[0.041]; champion[0.040]; champion[0.040]; relay[0.040]; win[0.040]; second[0.040]; stadium[0.039]; world[0.039]; Zurich[0.039]; Zurich[0.039]; trio[0.039]; concentration[0.039]; month[0.039]; Today[0.039]; Friday[0.039]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mitchell\u001b[39m ==> ENTITY: \u001b[32mDennis Mitchell\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.046]; Olympic[0.046]; Mitchell[0.044]; Mitchell[0.044]; Athletics[0.043]; Games[0.043]; champion[0.043]; record[0.043]; Dennis[0.042]; Adrian[0.042]; American[0.042]; Champions[0.042]; time[0.041]; sprint[0.041]; Bailey[0.040]; Bailey[0.040]; Bailey[0.040]; Bailey[0.040]; Donovan[0.040]; world[0.040]; world[0.040]; Warner[0.039]; champions[0.039]; despite[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = American ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.060:0.060[0]; log p(e|m)= -0.130:-0.130[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.048]; record[0.044]; Adrian[0.044]; Dennis[0.044]; Athletics[0.043]; win[0.043]; Champions[0.043]; major[0.042]; start[0.042]; Donovan[0.041]; Warner[0.041]; Atlanta[0.041]; Mitchell[0.041]; Mitchell[0.041]; Mitchell[0.041]; Olympic[0.041]; Olympic[0.041]; Friday[0.040]; world[0.040]; world[0.040]; champion[0.040]; time[0.040]; victory[0.040]; champions[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bailey\u001b[39m ==> ENTITY: \u001b[32mDonovan Bailey\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.049]; Olympic[0.049]; Olympic[0.049]; Olympic[0.049]; Olympics[0.048]; record[0.048]; Games[0.047]; champion[0.047]; champion[0.047]; sprint[0.045]; relay[0.045]; despite[0.045]; seconds[0.045]; Jamaica[0.044]; sport[0.044]; Michael[0.044]; medals[0.043]; finishing[0.043]; Despite[0.043]; beat[0.043]; second[0.042]; fourth[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mitchell\u001b[39m ==> ENTITY: \u001b[32mDennis Mitchell\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.050]; Olympic[0.050]; Olympic[0.050]; Mitchell[0.047]; Olympics[0.047]; series[0.047]; Games[0.047]; champion[0.046]; record[0.046]; sport[0.046]; medals[0.046]; American[0.045]; sprint[0.044]; Bailey[0.044]; Bailey[0.044]; Bailey[0.044]; Bailey[0.044]; Bailey[0.044]; fourth[0.044]; Jamaica[0.043]; Michael[0.043]; relay[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bailey\u001b[39m ==> ENTITY: \u001b[32mDonovan Bailey\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.049]; Olympic[0.049]; Olympic[0.049]; record[0.048]; Games[0.047]; Games[0.047]; champion[0.047]; champion[0.047]; sprinting[0.046]; sprint[0.045]; Athletics[0.045]; despite[0.045]; seconds[0.045]; Jamaica[0.044]; sport[0.044]; Michael[0.044]; time[0.044]; Champions[0.043]; finishing[0.043]; beat[0.043]; second[0.043]; fourth[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Berlin ==> ENTITY: \u001b[32mBerlin\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.018:0.018[0]; log p(e|m)= -0.154:-0.154[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympics[0.049]; Olympic[0.047]; Olympic[0.047]; Olympic[0.047]; Games[0.045]; Berlin[0.045]; sport[0.043]; Michael[0.041]; medals[0.041]; American[0.040]; American[0.040]; Mitchell[0.040]; Mitchell[0.040]; record[0.040]; champion[0.040]; champion[0.040]; relay[0.039]; win[0.039]; win[0.039]; Bailey[0.039]; Bailey[0.039]; Bailey[0.039]; Bailey[0.039]; Bailey[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dennis Mitchell\u001b[39m ==> ENTITY: \u001b[32mDennis Mitchell\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.046]; Olympic[0.046]; Mitchell[0.044]; Mitchell[0.044]; Athletics[0.043]; Games[0.043]; champion[0.043]; record[0.043]; Adrian[0.042]; American[0.042]; American[0.042]; Champions[0.042]; time[0.041]; sprint[0.041]; Bailey[0.041]; Bailey[0.041]; Bailey[0.041]; Bailey[0.041]; Donovan[0.040]; world[0.040]; world[0.040]; Warner[0.039]; champions[0.039]; despite[0.039]; \t\n\r [===>..........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 16s110ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 473/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 949testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.049]; England[0.047]; matches[0.047]; matches[0.047]; matches[0.047]; British[0.046]; Cricket[0.044]; Cricket[0.044]; play[0.044]; play[0.044]; tour[0.043]; tour[0.043]; Counties[0.043]; Scotland[0.043]; Friday[0.042]; Lord[0.042]; month[0.041]; internationals[0.041]; start[0.041]; English[0.041]; Test[0.041]; sides[0.041]; county[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mScotland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mScotland national cricket team\u001b[39m <---> \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.253:0.241[\u001b[31m0.012\u001b[39m]; local()= 0.175:0.054[\u001b[31m0.121\u001b[39m]; log p(e|m)= -3.689:-0.229[\u001b[32m3.459\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOval[0.054]; match[0.053]; Nottingham[0.052]; Hampshire[0.051]; British[0.051]; Headingley[0.051]; Edgbaston[0.051]; Leicestershire[0.050]; Middlesex[0.050]; Surrey[0.050]; Manchester[0.050]; Gloucestershire[0.049]; Sussex[0.049]; Glamorgan[0.049]; Somerset[0.049]; international[0.049]; Birmingham[0.049]; day[0.048]; Derbyshire[0.048]; Kent[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Surrey ==> ENTITY: \u001b[32mSurrey County Cricket Club\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.174:0.174[0]; log p(e|m)= -1.796:-1.796[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHampshire[0.046]; Middlesex[0.045]; Leicestershire[0.045]; Sussex[0.044]; Gloucestershire[0.044]; Worcestershire[0.043]; Durham[0.043]; Glamorgan[0.042]; Derbyshire[0.042]; Oval[0.041]; matches[0.041]; matches[0.041]; Edgbaston[0.041]; Universities[0.041]; Universities[0.041]; Oxford[0.041]; match[0.041]; Headingley[0.041]; Norfolk[0.040]; play[0.040]; Northampton[0.040]; Birmingham[0.039]; county[0.039]; county[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Glamorgan ==> ENTITY: \u001b[32mGlamorgan County Cricket Club\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.154:0.154[0]; log p(e|m)= -0.790:-0.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSomerset[0.044]; Edgbaston[0.044]; match[0.044]; Hampshire[0.044]; Lord[0.043]; Lord[0.043]; Sussex[0.043]; Middlesex[0.043]; Derbyshire[0.042]; Oval[0.042]; Universities[0.042]; Gloucestershire[0.042]; Leicestershire[0.042]; Nottingham[0.042]; Headingley[0.042]; Surrey[0.041]; class[0.040]; Kent[0.040]; Sixth[0.040]; British[0.040]; Oxford[0.040]; Counties[0.039]; county[0.039]; Birmingham[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Durham ==> ENTITY: \u001b[32mDurham County Cricket Club\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.167:0.167[0]; log p(e|m)= -2.397:-2.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOxford[0.046]; Hampshire[0.046]; Sussex[0.045]; Worcestershire[0.045]; Leicestershire[0.044]; Gloucestershire[0.044]; matches[0.044]; matches[0.044]; matches[0.044]; Derbyshire[0.044]; match[0.044]; Test[0.043]; Edgbaston[0.043]; Universities[0.043]; Universities[0.043]; Cricket[0.043]; Surrey[0.042]; Headingley[0.042]; Northampton[0.042]; Leeds[0.042]; British[0.041]; British[0.041]; Scotland[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mScotland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mScotland national cricket team\u001b[39m <---> \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.250:0.237[\u001b[31m0.013\u001b[39m]; local()= 0.207:0.057[\u001b[31m0.150\u001b[39m]; log p(e|m)= -3.689:-0.229[\u001b[32m3.459\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.045]; matches[0.044]; matches[0.044]; matches[0.044]; Oval[0.044]; match[0.043]; Test[0.043]; internationals[0.042]; Australia[0.042]; Australia[0.042]; British[0.041]; Headingley[0.041]; Edgbaston[0.041]; series[0.041]; Cricket[0.041]; Cricket[0.041]; Surrey[0.040]; Worcestershire[0.040]; Gloucestershire[0.040]; Sussex[0.040]; international[0.040]; international[0.040]; international[0.040]; Durham[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hampshire ==> ENTITY: \u001b[32mHampshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.193:0.193[0]; log p(e|m)= -1.924:-1.924[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSomerset[0.045]; Middlesex[0.045]; Sussex[0.044]; Leicestershire[0.044]; Durham[0.043]; match[0.043]; Glamorgan[0.043]; Kent[0.043]; Surrey[0.043]; Gloucestershire[0.043]; Nottingham[0.042]; Derbyshire[0.042]; county[0.041]; Counties[0.041]; Headingley[0.041]; Headingley[0.041]; Edgbaston[0.040]; Oval[0.040]; Oval[0.040]; class[0.040]; Old[0.039]; day[0.039]; day[0.039]; day[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.201:0.201[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; Test[0.045]; England[0.045]; internationals[0.044]; matches[0.044]; matches[0.044]; matches[0.044]; Cricket[0.043]; Cricket[0.043]; series[0.041]; Ashes[0.040]; Ashes[0.040]; Scotland[0.040]; day[0.040]; day[0.040]; day[0.040]; day[0.040]; day[0.040]; Headingley[0.040]; international[0.040]; international[0.040]; play[0.039]; play[0.039]; Worcestershire[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.101:0.101[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.046]; London[0.046]; Surrey[0.044]; Hampshire[0.044]; Sussex[0.043]; Durham[0.042]; British[0.042]; British[0.042]; Gloucestershire[0.042]; Leicestershire[0.042]; matches[0.041]; matches[0.041]; matches[0.041]; Norfolk[0.041]; Edgbaston[0.041]; Worcestershire[0.041]; Derbyshire[0.040]; Northampton[0.040]; Oxford[0.040]; match[0.040]; play[0.040]; Manchester[0.040]; tour[0.040]; Trafford[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = The Oval\u001b[39m ==> ENTITY: \u001b[32mThe Oval\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHampshire[0.050]; Middlesex[0.049]; Somerset[0.049]; Glamorgan[0.047]; Scotland[0.047]; Nottingham[0.047]; Headingley[0.047]; Manchester[0.046]; Old[0.046]; Kent[0.046]; Universities[0.046]; Fifth[0.045]; Sixth[0.045]; Oxford[0.044]; London[0.044]; Lord[0.043]; August[0.043]; August[0.043]; August[0.043]; August[0.043]; August[0.043]; Counties[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNottingham[0.053]; Hampshire[0.052]; British[0.050]; Somerset[0.050]; Middlesex[0.048]; Manchester[0.048]; Oxford[0.048]; Kent[0.047]; Scotland[0.047]; Glamorgan[0.047]; Headingley[0.047]; June[0.046]; June[0.046]; July[0.046]; July[0.046]; July[0.046]; July[0.046]; July[0.046]; July[0.046]; July[0.046]; Oval[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leicestershire ==> ENTITY: \u001b[32mLeicestershire County Cricket Club\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.146:0.146[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDurham[0.045]; Hampshire[0.043]; Nottingham[0.043]; Worcestershire[0.042]; Lord[0.042]; Lord[0.042]; Lord[0.042]; Somerset[0.042]; Middlesex[0.042]; match[0.042]; Glamorgan[0.042]; Gloucestershire[0.042]; Sussex[0.041]; Surrey[0.041]; Derbyshire[0.041]; Headingley[0.041]; Headingley[0.041]; Oval[0.041]; Northampton[0.041]; Kent[0.041]; Oxford[0.041]; Birmingham[0.041]; Edgbaston[0.041]; Leeds[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leeds ==> ENTITY: \u001b[32mLeeds\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.255:-0.255[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOval[0.044]; Headingley[0.044]; Norfolk[0.043]; Hampshire[0.043]; Birmingham[0.043]; Leicestershire[0.043]; match[0.043]; Edgbaston[0.042]; Gloucestershire[0.042]; Worcestershire[0.042]; matches[0.042]; matches[0.042]; matches[0.042]; Northampton[0.041]; Sussex[0.041]; Derbyshire[0.041]; Cricket[0.041]; internationals[0.041]; London[0.041]; London[0.041]; London[0.041]; Oxford[0.040]; Durham[0.040]; Surrey[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Arundel ==> ENTITY: \u001b[32mArundel\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.122:-0.122[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSussex[0.049]; England[0.049]; Norfolk[0.048]; County[0.047]; Gloucestershire[0.046]; Northampton[0.046]; county[0.046]; county[0.046]; Lord[0.045]; Lord[0.045]; Lord[0.045]; Surrey[0.045]; Durham[0.045]; Leicestershire[0.045]; Scotland[0.045]; Counties[0.044]; tour[0.044]; tour[0.044]; British[0.044]; Worcestershire[0.044]; Duke[0.043]; English[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Middlesex ==> ENTITY: \u001b[32mMiddlesex County Cricket Club\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.164:0.164[0]; log p(e|m)= -1.570:-1.570[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHampshire[0.045]; Surrey[0.045]; Sussex[0.045]; Somerset[0.044]; match[0.044]; Leicestershire[0.043]; Nottingham[0.043]; Glamorgan[0.043]; Kent[0.043]; Oval[0.043]; Oxford[0.043]; Gloucestershire[0.043]; Derbyshire[0.042]; Edgbaston[0.042]; Headingley[0.040]; London[0.040]; Birmingham[0.040]; class[0.039]; Counties[0.039]; Scotland[0.039]; Lord[0.039]; British[0.039]; county[0.039]; Manchester[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = English ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.113:0.113[0]; log p(e|m)= -1.355:-1.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.046]; Test[0.044]; Australia[0.043]; Australia[0.043]; British[0.043]; Leeds[0.043]; internationals[0.043]; matches[0.042]; matches[0.042]; matches[0.042]; Scotland[0.042]; Oval[0.042]; Cricket[0.042]; Cricket[0.042]; Northampton[0.041]; Worcestershire[0.041]; Gloucestershire[0.041]; Sussex[0.040]; Surrey[0.040]; class[0.040]; Durham[0.040]; series[0.040]; Headingley[0.040]; Counties[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Derbyshire ==> ENTITY: \u001b[32mDerbyshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.182:0.182[0]; log p(e|m)= -1.483:-1.483[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHampshire[0.044]; Worcestershire[0.043]; Lord[0.043]; Lord[0.043]; Lord[0.043]; Leicestershire[0.042]; Sussex[0.042]; Glamorgan[0.042]; Gloucestershire[0.042]; Surrey[0.042]; Middlesex[0.042]; Durham[0.042]; matches[0.042]; June[0.041]; June[0.041]; June[0.041]; June[0.041]; June[0.041]; June[0.041]; June[0.041]; June[0.041]; Headingley[0.041]; Headingley[0.041]; Northampton[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mHeadingley\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHeadingley Stadium\u001b[39m <---> \u001b[32mHeadingley\u001b[39m\t\nSCORES: global= 0.271:0.260[\u001b[31m0.011\u001b[39m]; local()= 0.149:0.165[\u001b[32m0.016\u001b[39m]; log p(e|m)= -0.302:-0.571[\u001b[31m0.268\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOval[0.052]; Hampshire[0.051]; Leicestershire[0.051]; match[0.050]; Somerset[0.050]; Edgbaston[0.050]; Glamorgan[0.050]; Middlesex[0.049]; Nottingham[0.049]; Derbyshire[0.048]; Kent[0.048]; Scotland[0.046]; Counties[0.046]; Trafford[0.046]; Birmingham[0.046]; Oxford[0.046]; Trent[0.046]; Lord[0.045]; Manchester[0.045]; British[0.044]; class[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.191:0.191[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; Test[0.045]; England[0.045]; internationals[0.045]; matches[0.045]; matches[0.045]; matches[0.045]; Cricket[0.044]; Cricket[0.044]; series[0.041]; Ashes[0.041]; Ashes[0.041]; Scotland[0.041]; day[0.040]; day[0.040]; day[0.040]; play[0.040]; play[0.040]; month[0.039]; starting[0.038]; tour[0.038]; tour[0.038]; British[0.038]; start[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Old Trafford ==> ENTITY: \u001b[32mOld Trafford Cricket Ground\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.138:0.138[0]; log p(e|m)= -0.644:-0.644[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nManchester[0.045]; Hampshire[0.044]; Glamorgan[0.044]; Leicestershire[0.043]; Leeds[0.043]; match[0.043]; Somerset[0.043]; Middlesex[0.043]; Gloucestershire[0.043]; Sussex[0.042]; Birmingham[0.042]; Derbyshire[0.042]; Nottingham[0.042]; Surrey[0.041]; Kent[0.041]; Edgbaston[0.041]; Headingley[0.041]; Oval[0.040]; Oval[0.040]; Oxford[0.040]; day[0.040]; day[0.040]; British[0.039]; class[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gloucestershire ==> ENTITY: \u001b[32mGloucestershire County Cricket Club\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.173:0.173[0]; log p(e|m)= -1.378:-1.378[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMiddlesex[0.044]; Hampshire[0.044]; Leicestershire[0.044]; Worcestershire[0.043]; Glamorgan[0.043]; Sussex[0.043]; Durham[0.042]; Oxford[0.042]; matches[0.042]; matches[0.042]; match[0.042]; Derbyshire[0.042]; Surrey[0.041]; Northampton[0.041]; Oval[0.041]; Universities[0.041]; Universities[0.041]; Norfolk[0.041]; Edgbaston[0.041]; Headingley[0.040]; Lord[0.040]; Lord[0.040]; Lord[0.040]; Leeds[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Worcestershire ==> ENTITY: \u001b[32mWorcestershire County Cricket Club\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.169:0.169[0]; log p(e|m)= -0.916:-0.916[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncounty[0.048]; county[0.048]; Leicestershire[0.047]; Gloucestershire[0.047]; County[0.046]; Sussex[0.046]; Counties[0.046]; Durham[0.046]; Surrey[0.045]; Derbyshire[0.045]; play[0.045]; play[0.045]; Lord[0.045]; Lord[0.045]; Lord[0.045]; Edgbaston[0.045]; Test[0.045]; Cricket[0.045]; match[0.044]; Headingley[0.044]; English[0.044]; Northampton[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = British Universities\u001b[39m ==> ENTITY: \u001b[32mBritish Universities cricket team\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.046]; Test[0.043]; Oval[0.043]; matches[0.042]; matches[0.042]; matches[0.042]; Leeds[0.042]; Worcestershire[0.042]; class[0.042]; Cricket[0.042]; Cricket[0.042]; Scotland[0.042]; Headingley[0.041]; Ashes[0.041]; Ashes[0.041]; international[0.041]; international[0.041]; international[0.041]; internationals[0.041]; Gloucestershire[0.041]; Durham[0.041]; Australia[0.041]; Australia[0.041]; Sussex[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = British Universities\u001b[39m ==> ENTITY: \u001b[32mBritish Universities cricket team\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHampshire[0.043]; Oval[0.043]; Oval[0.043]; Leicestershire[0.043]; Glamorgan[0.042]; Leeds[0.042]; Oxford[0.042]; Middlesex[0.042]; Worcestershire[0.042]; class[0.042]; Edgbaston[0.042]; Scotland[0.042]; match[0.041]; Headingley[0.041]; Headingley[0.041]; international[0.041]; international[0.041]; international[0.041]; Nottingham[0.041]; Gloucestershire[0.041]; Durham[0.041]; Sussex[0.040]; Somerset[0.040]; Surrey[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = The Oval\u001b[39m ==> ENTITY: \u001b[32mThe Oval\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.044]; matches[0.044]; matches[0.044]; Hampshire[0.044]; Surrey[0.043]; match[0.043]; Edgbaston[0.043]; Sussex[0.042]; Gloucestershire[0.042]; Leicestershire[0.042]; Scotland[0.041]; Headingley[0.041]; Derbyshire[0.041]; Worcestershire[0.041]; Manchester[0.041]; Leeds[0.041]; Old[0.040]; day[0.040]; day[0.040]; day[0.040]; day[0.040]; day[0.040]; day[0.040]; internationals[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Somerset ==> ENTITY: \u001b[32mSomerset County Cricket Club\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.170:0.170[0]; log p(e|m)= -2.419:-2.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHampshire[0.052]; Middlesex[0.052]; Leicestershire[0.051]; match[0.049]; Nottingham[0.049]; Glamorgan[0.049]; Kent[0.048]; Lord[0.048]; Headingley[0.048]; Edgbaston[0.048]; Oval[0.048]; Counties[0.047]; county[0.047]; class[0.046]; Oxford[0.046]; August[0.045]; August[0.045]; August[0.045]; August[0.045]; August[0.045]; Trent[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Trent Bridge\u001b[39m ==> ENTITY: \u001b[32mTrent Bridge\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHampshire[0.047]; Leicestershire[0.046]; Somerset[0.046]; Headingley[0.045]; Oval[0.045]; Edgbaston[0.045]; Glamorgan[0.045]; Nottingham[0.044]; Middlesex[0.043]; Kent[0.041]; Manchester[0.041]; Oxford[0.040]; Counties[0.040]; Lord[0.039]; test[0.039]; test[0.039]; test[0.039]; test[0.039]; test[0.039]; August[0.039]; August[0.039]; August[0.039]; August[0.039]; August[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEdgbaston\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEdgbaston Cricket Ground\u001b[39m <---> \u001b[32mEdgbaston\u001b[39m\t\nSCORES: global= 0.273:0.251[\u001b[31m0.022\u001b[39m]; local()= 0.147:0.123[\u001b[31m0.025\u001b[39m]; log p(e|m)= -0.451:-0.309[\u001b[32m0.142\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; match[0.044]; Hampshire[0.043]; Oval[0.043]; Leicestershire[0.043]; Somerset[0.042]; Oxford[0.042]; Headingley[0.042]; Headingley[0.042]; Worcestershire[0.042]; Glamorgan[0.042]; Birmingham[0.041]; Durham[0.041]; Trent[0.041]; Surrey[0.041]; day[0.041]; day[0.041]; day[0.041]; day[0.041]; Gloucestershire[0.041]; Sussex[0.040]; London[0.040]; London[0.040]; London[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mHeadingley\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHeadingley Stadium\u001b[39m <---> \u001b[32mHeadingley\u001b[39m\t\nSCORES: global= 0.271:0.259[\u001b[31m0.013\u001b[39m]; local()= 0.167:0.165[\u001b[31m0.002\u001b[39m]; log p(e|m)= -0.302:-0.571[\u001b[31m0.268\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOval[0.046]; Hampshire[0.045]; Leeds[0.045]; Leicestershire[0.045]; match[0.045]; Edgbaston[0.045]; Northampton[0.045]; Gloucestershire[0.044]; Worcestershire[0.044]; matches[0.044]; matches[0.044]; matches[0.044]; Sussex[0.043]; Derbyshire[0.043]; Cricket[0.043]; internationals[0.043]; Durham[0.042]; Surrey[0.042]; Australia[0.041]; Scotland[0.041]; Counties[0.041]; Birmingham[0.041]; Oxford[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.103:0.103[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.045]; London[0.045]; London[0.045]; Surrey[0.044]; England[0.044]; Sussex[0.042]; Durham[0.042]; British[0.042]; Gloucestershire[0.042]; matches[0.042]; matches[0.042]; matches[0.042]; match[0.041]; Norfolk[0.041]; Edgbaston[0.041]; Worcestershire[0.040]; Derbyshire[0.040]; Northampton[0.040]; Cricket[0.040]; Cricket[0.040]; play[0.040]; play[0.040]; tour[0.040]; tour[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Manchester ==> ENTITY: \u001b[32mManchester\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.120:0.120[0]; log p(e|m)= -0.322:-0.322[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.044]; Hampshire[0.043]; London[0.043]; London[0.043]; London[0.043]; Nottingham[0.043]; Birmingham[0.042]; Edgbaston[0.042]; Somerset[0.042]; Trafford[0.042]; Sussex[0.042]; Oval[0.042]; Oval[0.042]; Scotland[0.041]; county[0.041]; Leicestershire[0.041]; Glamorgan[0.041]; Gloucestershire[0.041]; Headingley[0.041]; Middlesex[0.040]; Oxford[0.040]; day[0.040]; day[0.040]; British[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ashes ==> ENTITY: \u001b[32mThe Ashes\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.192:0.192[0]; log p(e|m)= -0.170:-0.170[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.044]; Test[0.044]; Australia[0.044]; Australia[0.044]; matches[0.043]; matches[0.043]; matches[0.043]; Cricket[0.043]; Cricket[0.043]; internationals[0.043]; play[0.042]; play[0.042]; Ashes[0.041]; start[0.041]; series[0.041]; day[0.040]; day[0.040]; day[0.040]; Friday[0.040]; tour[0.040]; tour[0.040]; month[0.040]; Scotland[0.040]; starting[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ashes ==> ENTITY: \u001b[32mThe Ashes\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.192:0.192[0]; log p(e|m)= -0.170:-0.170[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.044]; Test[0.044]; Australia[0.044]; Australia[0.044]; matches[0.043]; matches[0.043]; matches[0.043]; Cricket[0.043]; Cricket[0.043]; internationals[0.043]; play[0.042]; play[0.042]; Ashes[0.041]; start[0.041]; series[0.041]; day[0.040]; day[0.040]; day[0.040]; Friday[0.040]; tour[0.040]; tour[0.040]; month[0.040]; Scotland[0.040]; starting[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sussex ==> ENTITY: \u001b[32mSussex County Cricket Club\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.163:0.163[0]; log p(e|m)= -1.556:-1.556[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHampshire[0.045]; Middlesex[0.044]; Durham[0.043]; match[0.043]; Surrey[0.043]; Oxford[0.043]; Leicestershire[0.042]; Oval[0.042]; matches[0.042]; matches[0.042]; Gloucestershire[0.042]; Worcestershire[0.042]; Glamorgan[0.041]; British[0.041]; British[0.041]; Norfolk[0.041]; Edgbaston[0.041]; Derbyshire[0.041]; London[0.040]; London[0.040]; London[0.040]; Universities[0.040]; Universities[0.040]; Leeds[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oxford ==> ENTITY: \u001b[32mOxford\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.124:0.124[0]; log p(e|m)= -0.681:-0.681[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHampshire[0.044]; Birmingham[0.043]; Leeds[0.043]; Nottingham[0.043]; Worcestershire[0.042]; Durham[0.042]; Surrey[0.042]; Universities[0.042]; Manchester[0.042]; London[0.042]; London[0.042]; London[0.042]; Leicestershire[0.042]; Somerset[0.042]; Northampton[0.042]; match[0.041]; Gloucestershire[0.041]; Bridge[0.041]; Trafford[0.041]; British[0.041]; Sussex[0.041]; Kent[0.040]; Middlesex[0.040]; Scotland[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.101:0.101[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.045]; London[0.045]; Surrey[0.044]; Hampshire[0.044]; Sussex[0.042]; Durham[0.042]; British[0.042]; British[0.042]; Gloucestershire[0.042]; Leicestershire[0.042]; matches[0.042]; matches[0.042]; matches[0.042]; match[0.041]; Norfolk[0.041]; Edgbaston[0.041]; Worcestershire[0.040]; Derbyshire[0.040]; Northampton[0.040]; Oxford[0.040]; play[0.040]; Manchester[0.040]; tour[0.040]; Trafford[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Northampton ==> ENTITY: \u001b[32mNorthamptonshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.190:0.190[0]; log p(e|m)= -3.650:-3.650[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.044]; Leicestershire[0.044]; matches[0.043]; matches[0.043]; matches[0.043]; Sussex[0.043]; Worcestershire[0.043]; Gloucestershire[0.042]; match[0.042]; Durham[0.042]; Derbyshire[0.042]; Test[0.042]; Surrey[0.042]; Leeds[0.042]; internationals[0.041]; Cricket[0.041]; Edgbaston[0.041]; Oxford[0.041]; Scotland[0.041]; Headingley[0.040]; Oval[0.040]; Counties[0.040]; play[0.040]; play[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nottingham ==> ENTITY: \u001b[32mNottingham\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.292:-0.292[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOxford[0.058]; Hampshire[0.056]; Leicestershire[0.056]; Somerset[0.055]; Headingley[0.054]; Universities[0.054]; Oval[0.054]; Glamorgan[0.053]; British[0.053]; Manchester[0.052]; Middlesex[0.052]; Trafford[0.052]; London[0.052]; Scotland[0.051]; Kent[0.050]; county[0.050]; Bridge[0.050]; Trent[0.049]; class[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kent ==> ENTITY: \u001b[32mKent County Cricket Club\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.128:0.128[0]; log p(e|m)= -2.513:-2.513[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMiddlesex[0.047]; Hampshire[0.047]; Somerset[0.046]; Leicestershire[0.045]; Oxford[0.045]; Nottingham[0.045]; Universities[0.044]; Glamorgan[0.044]; British[0.043]; Headingley[0.042]; Oval[0.042]; London[0.042]; county[0.041]; Trafford[0.040]; Counties[0.040]; Scotland[0.040]; Manchester[0.040]; Lord[0.040]; Trent[0.038]; August[0.038]; August[0.038]; August[0.038]; August[0.038]; August[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Birmingham ==> ENTITY: \u001b[32mBirmingham\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.141:0.141[0]; log p(e|m)= -0.402:-0.402[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEdgbaston[0.043]; Hampshire[0.043]; matches[0.043]; match[0.043]; Manchester[0.042]; Oval[0.042]; Leicestershire[0.042]; Somerset[0.042]; Leeds[0.042]; Norfolk[0.042]; Headingley[0.042]; Headingley[0.042]; London[0.042]; London[0.042]; London[0.042]; Oxford[0.042]; Worcestershire[0.042]; Glamorgan[0.041]; Durham[0.041]; Gloucestershire[0.040]; Northampton[0.040]; Sussex[0.040]; British[0.040]; Surrey[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland cricket team\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.179:0.179[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTest[0.045]; Australia[0.045]; Australia[0.045]; internationals[0.044]; matches[0.043]; matches[0.043]; matches[0.043]; Scotland[0.043]; Cricket[0.043]; Cricket[0.043]; British[0.042]; Northampton[0.041]; series[0.041]; day[0.041]; day[0.041]; day[0.041]; play[0.040]; play[0.040]; Ashes[0.040]; Ashes[0.040]; Norfolk[0.039]; month[0.039]; London[0.039]; London[0.039]; \t\n\r [===>..........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s965ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 514/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1095testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oakland ==> ENTITY: \u001b[32mOakland Athletics\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.202:0.202[0]; log p(e|m)= -3.912:-3.912[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPittsburgh[0.044]; Oakland[0.044]; League[0.043]; Cleveland[0.043]; Cleveland[0.043]; Houston[0.042]; California[0.042]; California[0.042]; Cincinnati[0.042]; Philadelphia[0.042]; Baltimore[0.042]; Boston[0.041]; Boston[0.041]; Detroit[0.041]; Detroit[0.041]; Central[0.041]; Central[0.041]; Toronto[0.041]; Toronto[0.041]; Texas[0.040]; Texas[0.040]; Florida[0.040]; City[0.040]; City[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York Yankees\u001b[39m\t\nSCORES: global= 0.238:0.238[0]; local()= 0.202:0.202[0]; log p(e|m)= -5.116:-5.116[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.044]; Cleveland[0.043]; Cleveland[0.043]; Boston[0.043]; Boston[0.043]; League[0.043]; City[0.043]; City[0.043]; Baltimore[0.042]; Pittsburgh[0.042]; Cincinnati[0.042]; Houston[0.042]; Toronto[0.042]; Toronto[0.042]; Oakland[0.041]; Oakland[0.041]; Detroit[0.041]; Detroit[0.041]; York[0.040]; Central[0.040]; Central[0.040]; Atlanta[0.040]; Saturday[0.040]; Saturday[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMilwaukee\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAtlanta Braves\u001b[39m <---> \u001b[32mMilwaukee Brewers\u001b[39m\t\nSCORES: global= 0.253:0.251[\u001b[31m0.002\u001b[39m]; local()= 0.245:0.216[\u001b[31m0.029\u001b[39m]; log p(e|m)= -3.912:-3.411[\u001b[32m0.501\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.044]; games[0.043]; games[0.043]; Cleveland[0.043]; Cleveland[0.043]; Baseball[0.043]; Baseball[0.043]; League[0.042]; League[0.042]; League[0.042]; League[0.042]; Boston[0.042]; Boston[0.042]; Baltimore[0.041]; Baltimore[0.041]; Detroit[0.041]; Detroit[0.041]; Oakland[0.041]; Oakland[0.041]; Atlanta[0.041]; Central[0.040]; played[0.040]; Milwaukee[0.040]; winning[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Western Division ==> ENTITY: \u001b[32mNational League West\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.177:0.177[0]; log p(e|m)= -1.295:-1.295[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; Division[0.043]; Division[0.043]; Houston[0.043]; Houston[0.043]; Oakland[0.043]; Texas[0.043]; Cleveland[0.042]; Cincinnati[0.042]; Cincinnati[0.042]; Pittsburgh[0.042]; Pittsburgh[0.042]; Philadelphia[0.041]; Philadelphia[0.041]; Toronto[0.041]; Eastern[0.041]; Colorado[0.041]; Colorado[0.041]; Detroit[0.040]; York[0.040]; York[0.040]; York[0.040]; National[0.040]; Baltimore[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Diego ==> ENTITY: \u001b[32mSan Diego Padres\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.221:0.221[0]; log p(e|m)= -3.912:-3.912[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; Houston[0.045]; Houston[0.045]; Oakland[0.045]; California[0.045]; Cincinnati[0.045]; Cincinnati[0.045]; Cleveland[0.044]; Philadelphia[0.044]; Philadelphia[0.044]; August[0.043]; August[0.043]; Baltimore[0.043]; Pittsburgh[0.043]; Pittsburgh[0.043]; Boston[0.043]; Toronto[0.042]; Detroit[0.042]; Texas[0.042]; Florida[0.042]; Florida[0.042]; National[0.042]; City[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Baltimore ==> ENTITY: \u001b[32mBaltimore Orioles\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.218:0.218[0]; log p(e|m)= -4.075:-4.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBaltimore[0.047]; games[0.045]; games[0.045]; Cleveland[0.045]; Cleveland[0.045]; League[0.044]; League[0.044]; League[0.044]; won[0.044]; Baseball[0.044]; Baseball[0.044]; Boston[0.043]; Boston[0.043]; played[0.043]; Oakland[0.043]; Detroit[0.043]; Detroit[0.043]; Toronto[0.042]; Toronto[0.042]; winning[0.042]; lost[0.041]; York[0.041]; York[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oakland ==> ENTITY: \u001b[32mOakland Athletics\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.221:0.221[0]; log p(e|m)= -3.912:-3.912[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOakland[0.045]; games[0.045]; games[0.045]; League[0.044]; League[0.044]; League[0.044]; Cleveland[0.044]; Cleveland[0.044]; Baseball[0.044]; Houston[0.044]; California[0.044]; California[0.044]; Philadelphia[0.043]; Baltimore[0.043]; Baltimore[0.043]; Boston[0.043]; Boston[0.043]; Detroit[0.043]; Detroit[0.043]; Central[0.042]; Central[0.042]; Toronto[0.042]; Toronto[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Philadelphia ==> ENTITY: \u001b[32mPhiladelphia Phillies\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.185:0.185[0]; log p(e|m)= -4.510:-4.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.047]; Houston[0.044]; Houston[0.044]; Cincinnati[0.044]; Cincinnati[0.044]; Pittsburgh[0.044]; Pittsburgh[0.044]; Atlanta[0.042]; Florida[0.041]; Florida[0.041]; York[0.041]; York[0.041]; Chicago[0.041]; Chicago[0.041]; Division[0.041]; Division[0.041]; Montreal[0.041]; Saturday[0.040]; Colorado[0.040]; Colorado[0.040]; Central[0.040]; Western[0.040]; Louis[0.039]; Louis[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Diego ==> ENTITY: \u001b[32mSan Diego Padres\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.174:0.174[0]; log p(e|m)= -3.912:-3.912[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDiego[0.048]; San[0.048]; San[0.048]; San[0.048]; Cincinnati[0.044]; Cincinnati[0.044]; Houston[0.043]; Houston[0.043]; Philadelphia[0.043]; Philadelphia[0.043]; August[0.043]; Los[0.043]; Los[0.043]; Saturday[0.043]; Pittsburgh[0.042]; Pittsburgh[0.042]; Francisco[0.042]; Francisco[0.042]; Louis[0.042]; Louis[0.042]; York[0.041]; York[0.041]; Atlanta[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mToronto\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mToronto Blue Jays\u001b[39m <---> \u001b[32mToronto\u001b[39m\t\nSCORES: global= 0.249:0.237[\u001b[31m0.012\u001b[39m]; local()= 0.211:0.084[\u001b[31m0.127\u001b[39m]; log p(e|m)= -4.423:-0.123[\u001b[32m4.300\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nToronto[0.046]; Montreal[0.044]; League[0.044]; Boston[0.043]; Boston[0.043]; Cleveland[0.042]; Cleveland[0.042]; Cincinnati[0.042]; Pittsburgh[0.042]; Philadelphia[0.042]; Detroit[0.041]; Detroit[0.041]; Houston[0.041]; Baltimore[0.041]; Baltimore[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Atlanta[0.041]; Louis[0.040]; Oakland[0.040]; Oakland[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cleveland ==> ENTITY: \u001b[32mCleveland Indians\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.236:0.236[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.050]; games[0.048]; games[0.048]; Baseball[0.046]; Baseball[0.046]; League[0.046]; League[0.046]; League[0.046]; League[0.046]; Baltimore[0.045]; Baltimore[0.045]; Boston[0.045]; Boston[0.045]; Oakland[0.044]; Oakland[0.044]; Detroit[0.044]; Detroit[0.044]; American[0.044]; Toronto[0.044]; Toronto[0.044]; played[0.044]; Atlanta[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = National League Eastern Division\u001b[39m ==> ENTITY: \u001b[32mNational League East\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.044]; Cleveland[0.044]; Philadelphia[0.044]; Cincinnati[0.043]; Pittsburgh[0.043]; Houston[0.042]; Houston[0.042]; Oakland[0.042]; Oakland[0.042]; Atlanta[0.042]; Atlanta[0.042]; Baltimore[0.042]; Boston[0.041]; Detroit[0.041]; Detroit[0.041]; Toronto[0.041]; Toronto[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Montreal[0.040]; Kansas[0.040]; Kansas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colorado ==> ENTITY: \u001b[32mColorado Rockies\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.189:0.189[0]; log p(e|m)= -4.510:-4.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nColorado[0.053]; Montreal[0.043]; Montreal[0.043]; Toronto[0.042]; California[0.042]; Philadelphia[0.042]; Philadelphia[0.042]; Cincinnati[0.041]; Cincinnati[0.041]; Pittsburgh[0.041]; Pittsburgh[0.041]; Houston[0.041]; Houston[0.041]; Florida[0.041]; Florida[0.041]; Boston[0.041]; Cleveland[0.041]; Texas[0.041]; League[0.041]; Atlanta[0.040]; Atlanta[0.040]; Oakland[0.040]; Louis[0.040]; Louis[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Montreal ==> ENTITY: \u001b[32mMontreal Expos\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.193:0.193[0]; log p(e|m)= -4.510:-4.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.044]; Houston[0.044]; Cincinnati[0.044]; Cincinnati[0.044]; Philadelphia[0.043]; Philadelphia[0.043]; Louis[0.043]; Louis[0.043]; Pittsburgh[0.043]; Pittsburgh[0.043]; Atlanta[0.042]; York[0.041]; York[0.041]; Florida[0.041]; Saturday[0.041]; Chicago[0.041]; Chicago[0.041]; Division[0.040]; Division[0.040]; August[0.040]; Francisco[0.040]; Francisco[0.040]; Colorado[0.040]; Colorado[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colorado ==> ENTITY: \u001b[32mColorado Rockies\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.165:0.165[0]; log p(e|m)= -4.510:-4.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nColorado[0.054]; Montreal[0.045]; Philadelphia[0.043]; Philadelphia[0.043]; Cincinnati[0.043]; Cincinnati[0.043]; Pittsburgh[0.042]; Pittsburgh[0.042]; Houston[0.042]; Houston[0.042]; Florida[0.042]; Atlanta[0.041]; Louis[0.041]; Louis[0.041]; Central[0.040]; Chicago[0.040]; Chicago[0.040]; Western[0.040]; Saturday[0.039]; Division[0.039]; Division[0.039]; August[0.039]; Francisco[0.038]; Francisco[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Toronto ==> ENTITY: \u001b[32mToronto Blue Jays\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.204:0.204[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nToronto[0.045]; games[0.044]; games[0.044]; League[0.043]; League[0.043]; League[0.043]; League[0.043]; Baseball[0.043]; Baseball[0.043]; played[0.042]; Boston[0.042]; Boston[0.042]; Cleveland[0.041]; Cleveland[0.041]; Detroit[0.040]; Detroit[0.040]; won[0.040]; Baltimore[0.040]; Baltimore[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; winning[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Minnesota ==> ENTITY: \u001b[32mMinnesota Twins\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.217:0.217[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.044]; Cleveland[0.044]; Minnesota[0.044]; League[0.043]; Oakland[0.043]; Oakland[0.043]; Detroit[0.042]; Detroit[0.042]; Baltimore[0.042]; Baltimore[0.042]; Philadelphia[0.042]; Cincinnati[0.042]; Houston[0.042]; Pittsburgh[0.041]; National[0.041]; Boston[0.041]; Boston[0.041]; Texas[0.040]; Texas[0.040]; Saturday[0.040]; Central[0.040]; Central[0.040]; Toronto[0.040]; Toronto[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlanta ==> ENTITY: \u001b[32mAtlanta Braves\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.220:0.220[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.046]; Cleveland[0.045]; Cleveland[0.045]; League[0.045]; Atlanta[0.045]; Cincinnati[0.045]; Houston[0.044]; Houston[0.044]; Pittsburgh[0.044]; Pittsburgh[0.044]; Diego[0.044]; Boston[0.044]; Baltimore[0.043]; Oakland[0.043]; Oakland[0.043]; Detroit[0.043]; Detroit[0.043]; Los[0.042]; Toronto[0.042]; Western[0.042]; Western[0.042]; Francisco[0.041]; Francisco[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Montreal ==> ENTITY: \u001b[32mMontreal Expos\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.229:0.229[0]; log p(e|m)= -4.510:-4.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nToronto[0.044]; Baltimore[0.043]; League[0.043]; Houston[0.043]; Houston[0.043]; Cincinnati[0.042]; Cleveland[0.042]; Cleveland[0.042]; Philadelphia[0.042]; Boston[0.042]; Louis[0.042]; Detroit[0.042]; Detroit[0.042]; Pittsburgh[0.042]; Pittsburgh[0.042]; Oakland[0.041]; Oakland[0.041]; Atlanta[0.041]; Atlanta[0.041]; York[0.040]; York[0.040]; York[0.040]; Florida[0.040]; Florida[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kansas City ==> ENTITY: \u001b[32mKansas City Royals\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.225:0.225[0]; log p(e|m)= -3.576:-3.576[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; games[0.044]; Baseball[0.043]; Baseball[0.043]; Houston[0.042]; League[0.042]; League[0.042]; League[0.042]; League[0.042]; Cleveland[0.042]; Cleveland[0.042]; Toronto[0.042]; Toronto[0.042]; Baltimore[0.041]; Baltimore[0.041]; Philadelphia[0.041]; Oakland[0.041]; Oakland[0.041]; City[0.041]; won[0.040]; Detroit[0.040]; Detroit[0.040]; played[0.040]; winning[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Francisco ==> ENTITY: \u001b[32mSan Francisco Giants\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.179:0.179[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; Philadelphia[0.046]; Philadelphia[0.046]; Cleveland[0.045]; Houston[0.045]; Houston[0.045]; California[0.044]; Oakland[0.044]; Cincinnati[0.043]; Cincinnati[0.043]; Diego[0.043]; Diego[0.043]; Francisco[0.043]; San[0.043]; San[0.043]; San[0.043]; Pittsburgh[0.043]; Pittsburgh[0.043]; Boston[0.042]; Texas[0.042]; Los[0.042]; Los[0.042]; National[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlanta ==> ENTITY: \u001b[32mAtlanta Braves\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.187:0.187[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.046]; Philadelphia[0.046]; League[0.045]; Atlanta[0.045]; Cincinnati[0.045]; Cincinnati[0.045]; Houston[0.045]; Houston[0.045]; Pittsburgh[0.045]; Pittsburgh[0.045]; Diego[0.044]; Diego[0.044]; Los[0.043]; Los[0.043]; Western[0.042]; Francisco[0.042]; Francisco[0.042]; Chicago[0.042]; Chicago[0.042]; Division[0.041]; Division[0.041]; Division[0.041]; Eastern[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kansas City ==> ENTITY: \u001b[32mKansas City Royals\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.232:0.232[0]; log p(e|m)= -3.576:-3.576[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; Houston[0.043]; Cleveland[0.043]; Cleveland[0.043]; Baltimore[0.042]; Baltimore[0.042]; Philadelphia[0.042]; League[0.042]; League[0.042]; Oakland[0.042]; Oakland[0.042]; City[0.042]; Cincinnati[0.042]; Pittsburgh[0.041]; Detroit[0.041]; Detroit[0.041]; Boston[0.041]; Boston[0.041]; Toronto[0.040]; Toronto[0.040]; Texas[0.040]; Texas[0.040]; Milwaukee[0.040]; Milwaukee[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMilwaukee\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAtlanta Braves\u001b[39m <---> \u001b[32mMilwaukee Brewers\u001b[39m\t\nSCORES: global= 0.254:0.251[\u001b[31m0.003\u001b[39m]; local()= 0.243:0.205[\u001b[31m0.038\u001b[39m]; log p(e|m)= -3.912:-3.411[\u001b[32m0.501\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.044]; Cleveland[0.044]; Cleveland[0.044]; Cincinnati[0.043]; Pittsburgh[0.043]; League[0.042]; Boston[0.042]; Boston[0.042]; Baltimore[0.042]; Baltimore[0.042]; Houston[0.042]; Detroit[0.042]; Detroit[0.042]; Oakland[0.041]; Oakland[0.041]; Atlanta[0.041]; Central[0.041]; Central[0.041]; Milwaukee[0.041]; Toronto[0.040]; Toronto[0.040]; York[0.040]; York[0.040]; York[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Florida ==> ENTITY: \u001b[32mMiami Marlins\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.225:0.225[0]; log p(e|m)= -4.828:-4.828[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFlorida[0.046]; Cincinnati[0.043]; Cincinnati[0.043]; Philadelphia[0.043]; Philadelphia[0.043]; Montreal[0.042]; League[0.042]; Pittsburgh[0.042]; Pittsburgh[0.042]; Toronto[0.042]; Texas[0.042]; Texas[0.042]; Cleveland[0.042]; Cleveland[0.042]; Houston[0.041]; Houston[0.041]; California[0.041]; California[0.041]; Milwaukee[0.041]; Milwaukee[0.041]; Oakland[0.040]; Oakland[0.040]; Baltimore[0.040]; Boston[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York Yankees\u001b[39m\t\nSCORES: global= 0.238:0.238[0]; local()= 0.204:0.204[0]; log p(e|m)= -5.116:-5.116[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.046]; Philadelphia[0.046]; Cleveland[0.045]; Boston[0.045]; League[0.045]; City[0.044]; City[0.044]; Baltimore[0.044]; Pittsburgh[0.044]; Pittsburgh[0.044]; Cincinnati[0.044]; Cincinnati[0.044]; Houston[0.043]; Houston[0.043]; Toronto[0.043]; Oakland[0.043]; Oakland[0.043]; Detroit[0.043]; York[0.042]; York[0.042]; Central[0.041]; Atlanta[0.041]; Atlanta[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York Yankees\u001b[39m\t\nSCORES: global= 0.239:0.239[0]; local()= 0.200:0.200[0]; log p(e|m)= -5.116:-5.116[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; games[0.047]; Baseball[0.046]; Baseball[0.046]; Cleveland[0.045]; Boston[0.044]; League[0.044]; League[0.044]; League[0.044]; City[0.044]; City[0.044]; Baltimore[0.043]; Baltimore[0.043]; Toronto[0.043]; Toronto[0.043]; Oakland[0.043]; played[0.042]; Detroit[0.042]; Detroit[0.042]; York[0.042]; American[0.041]; won[0.041]; Central[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Houston ==> ENTITY: \u001b[32mHouston Astros\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.162:0.162[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.049]; Cincinnati[0.044]; Cincinnati[0.044]; Philadelphia[0.044]; Philadelphia[0.044]; Pittsburgh[0.043]; Pittsburgh[0.043]; Atlanta[0.042]; Atlanta[0.042]; Saturday[0.041]; Western[0.041]; Colorado[0.041]; Colorado[0.041]; Florida[0.041]; Florida[0.041]; Chicago[0.041]; Chicago[0.041]; August[0.040]; Central[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; York[0.039]; York[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew York\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew York Yankees\u001b[39m <---> \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.239:0.232[\u001b[31m0.006\u001b[39m]; local()= 0.186:0.096[\u001b[31m0.089\u001b[39m]; log p(e|m)= -5.116:-1.790[\u001b[32m3.326\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; games[0.045]; Baseball[0.044]; Baseball[0.044]; Cleveland[0.043]; Boston[0.043]; League[0.043]; League[0.043]; League[0.043]; City[0.042]; Baltimore[0.042]; Toronto[0.041]; played[0.041]; Detroit[0.041]; York[0.040]; American[0.040]; won[0.040]; Central[0.040]; Friday[0.040]; Friday[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; winning[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Detroit ==> ENTITY: \u001b[32mDetroit Tigers\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.227:0.227[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDetroit[0.044]; Cleveland[0.043]; Cleveland[0.043]; Cincinnati[0.043]; Montreal[0.043]; Philadelphia[0.043]; League[0.042]; League[0.042]; Toronto[0.042]; Toronto[0.042]; Oakland[0.042]; Oakland[0.042]; Pittsburgh[0.041]; Boston[0.041]; Boston[0.041]; Houston[0.041]; Baltimore[0.041]; Baltimore[0.041]; Division[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Detroit ==> ENTITY: \u001b[32mDetroit Tigers\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.231:0.231[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.048]; games[0.048]; Detroit[0.045]; Baseball[0.045]; Baseball[0.045]; Cleveland[0.044]; Cleveland[0.044]; Montreal[0.044]; League[0.043]; League[0.043]; League[0.043]; League[0.043]; Toronto[0.043]; Toronto[0.043]; Boston[0.042]; Boston[0.042]; played[0.042]; Oakland[0.042]; Oakland[0.042]; Division[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cleveland ==> ENTITY: \u001b[32mCleveland Indians\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.222:0.222[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.048]; Philadelphia[0.045]; Cincinnati[0.045]; League[0.045]; Houston[0.044]; Baltimore[0.044]; Baltimore[0.044]; Boston[0.044]; Boston[0.044]; Pittsburgh[0.043]; Detroit[0.043]; Detroit[0.043]; Oakland[0.043]; Oakland[0.043]; Toronto[0.043]; Toronto[0.043]; Atlanta[0.042]; York[0.042]; York[0.042]; City[0.042]; City[0.042]; Texas[0.042]; Texas[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Philadelphia ==> ENTITY: \u001b[32mPhiladelphia Phillies\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.220:0.220[0]; log p(e|m)= -4.510:-4.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.046]; Cleveland[0.045]; Houston[0.043]; Houston[0.043]; League[0.043]; Cincinnati[0.043]; Cincinnati[0.043]; Baltimore[0.043]; Pittsburgh[0.042]; Pittsburgh[0.042]; Boston[0.042]; Oakland[0.042]; Oakland[0.042]; Detroit[0.041]; Atlanta[0.040]; Atlanta[0.040]; Florida[0.040]; Florida[0.040]; York[0.040]; York[0.040]; York[0.040]; Toronto[0.040]; Milwaukee[0.040]; Milwaukee[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Central Division ==> ENTITY: \u001b[32mNational League Central\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.184:0.184[0]; log p(e|m)= -0.989:-0.989[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPittsburgh[0.044]; Pittsburgh[0.044]; League[0.044]; Division[0.044]; Division[0.044]; Division[0.044]; Cincinnati[0.043]; Cincinnati[0.043]; Cleveland[0.043]; Philadelphia[0.042]; Philadelphia[0.042]; Boston[0.042]; Detroit[0.041]; Baltimore[0.041]; Western[0.041]; Western[0.041]; Houston[0.040]; Houston[0.040]; Toronto[0.040]; Oakland[0.040]; Oakland[0.040]; Eastern[0.039]; Atlanta[0.039]; Atlanta[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Western Division ==> ENTITY: \u001b[32mNational League West\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.212:0.212[0]; log p(e|m)= -1.295:-1.295[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; League[0.044]; League[0.044]; League[0.044]; Boston[0.042]; Boston[0.042]; games[0.042]; games[0.042]; Philadelphia[0.041]; Cleveland[0.041]; Cleveland[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; winning[0.041]; won[0.041]; Detroit[0.041]; Detroit[0.041]; Texas[0.041]; Texas[0.041]; standings[0.041]; Oakland[0.041]; Oakland[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pittsburgh ==> ENTITY: \u001b[32mPittsburgh Pirates\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.200:0.200[0]; log p(e|m)= -4.017:-4.017[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.047]; Philadelphia[0.047]; Pittsburgh[0.047]; Cincinnati[0.046]; Cincinnati[0.046]; Montreal[0.045]; Montreal[0.045]; Florida[0.044]; Florida[0.044]; Houston[0.044]; Houston[0.044]; Louis[0.043]; Louis[0.043]; Atlanta[0.043]; Atlanta[0.043]; Colorado[0.042]; Colorado[0.042]; York[0.042]; York[0.042]; Chicago[0.041]; Chicago[0.041]; Saturday[0.041]; Central[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Francisco ==> ENTITY: \u001b[32mSan Francisco Giants\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.163:0.163[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.045]; Philadelphia[0.045]; Houston[0.044]; Houston[0.044]; Cincinnati[0.042]; Cincinnati[0.042]; Diego[0.042]; Diego[0.042]; Francisco[0.042]; San[0.042]; San[0.042]; San[0.042]; Pittsburgh[0.042]; Pittsburgh[0.042]; Los[0.041]; Los[0.041]; Florida[0.040]; Florida[0.040]; Atlanta[0.040]; Atlanta[0.040]; York[0.040]; York[0.040]; Chicago[0.040]; Chicago[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Boston ==> ENTITY: \u001b[32mBoston Red Sox\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.226:0.226[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBoston[0.046]; games[0.046]; games[0.046]; Cleveland[0.043]; Cleveland[0.043]; Baseball[0.042]; Baseball[0.042]; League[0.042]; League[0.042]; League[0.042]; Toronto[0.042]; Toronto[0.042]; Detroit[0.041]; Detroit[0.041]; Baltimore[0.041]; Baltimore[0.041]; played[0.041]; won[0.040]; Oakland[0.040]; Oakland[0.040]; Central[0.040]; Texas[0.040]; Texas[0.040]; winning[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Baltimore ==> ENTITY: \u001b[32mBaltimore Orioles\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.221:0.221[0]; log p(e|m)= -4.075:-4.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBaltimore[0.045]; Philadelphia[0.044]; Cleveland[0.043]; Cleveland[0.043]; Houston[0.043]; Cincinnati[0.043]; Pittsburgh[0.043]; Oakland[0.042]; Oakland[0.042]; Boston[0.042]; Boston[0.042]; League[0.041]; League[0.041]; National[0.041]; Detroit[0.041]; Detroit[0.041]; Toronto[0.041]; Toronto[0.041]; City[0.040]; City[0.040]; Montreal[0.040]; Florida[0.040]; Seattle[0.040]; Seattle[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Major League Baseball\u001b[39m ==> ENTITY: \u001b[32mMajor League Baseball\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBaseball[0.046]; games[0.044]; games[0.044]; League[0.044]; League[0.044]; won[0.044]; winning[0.043]; Cleveland[0.043]; played[0.042]; Oakland[0.042]; Detroit[0.042]; Baltimore[0.041]; standings[0.041]; Standings[0.041]; Boston[0.041]; York[0.040]; York[0.040]; Toronto[0.040]; Friday[0.040]; Friday[0.040]; percentage[0.040]; Texas[0.040]; Milwaukee[0.039]; Chicago[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Texas ==> ENTITY: \u001b[32mTexas Rangers (baseball)\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.207:0.207[0]; log p(e|m)= -4.962:-4.962[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTexas[0.045]; Cleveland[0.043]; Cleveland[0.043]; Houston[0.043]; League[0.043]; Baltimore[0.042]; Baltimore[0.042]; Cincinnati[0.042]; Division[0.042]; Division[0.042]; Division[0.042]; Division[0.042]; Division[0.042]; Oakland[0.041]; Oakland[0.041]; Philadelphia[0.041]; Detroit[0.041]; Detroit[0.041]; Pittsburgh[0.041]; Saturday[0.040]; August[0.040]; Boston[0.040]; Boston[0.040]; Atlanta[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Minnesota ==> ENTITY: \u001b[32mMinnesota Twins\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.228:0.228[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; games[0.044]; Baseball[0.043]; Baseball[0.043]; Cleveland[0.043]; Cleveland[0.043]; Minnesota[0.043]; League[0.042]; League[0.042]; League[0.042]; League[0.042]; Oakland[0.042]; Oakland[0.042]; played[0.042]; Detroit[0.041]; Detroit[0.041]; Baltimore[0.041]; Baltimore[0.041]; National[0.040]; Boston[0.040]; Boston[0.040]; Texas[0.039]; Texas[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Houston ==> ENTITY: \u001b[32mHouston Astros\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.194:0.194[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.048]; League[0.044]; Texas[0.043]; Texas[0.043]; Cleveland[0.043]; Baltimore[0.042]; Cincinnati[0.042]; Cincinnati[0.042]; Philadelphia[0.042]; Philadelphia[0.042]; Oakland[0.042]; Oakland[0.042]; Pittsburgh[0.041]; Pittsburgh[0.041]; Boston[0.041]; City[0.041]; City[0.041]; Atlanta[0.040]; Atlanta[0.040]; National[0.040]; Toronto[0.040]; Saturday[0.040]; Saturday[0.040]; Detroit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Texas ==> ENTITY: \u001b[32mTexas Rangers (baseball)\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.192:0.192[0]; log p(e|m)= -4.962:-4.962[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTexas[0.047]; Baseball[0.045]; Cleveland[0.044]; Cleveland[0.044]; Houston[0.044]; games[0.044]; games[0.044]; League[0.044]; League[0.044]; League[0.044]; League[0.044]; Baltimore[0.043]; Baltimore[0.043]; Division[0.043]; Division[0.043]; Division[0.043]; Division[0.043]; Division[0.043]; Standings[0.043]; Oakland[0.043]; Oakland[0.043]; Philadelphia[0.042]; standings[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York Yankees\u001b[39m\t\nSCORES: global= 0.238:0.238[0]; local()= 0.173:0.173[0]; log p(e|m)= -5.116:-5.116[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.047]; Philadelphia[0.047]; Pittsburgh[0.045]; Pittsburgh[0.045]; Cincinnati[0.045]; Cincinnati[0.045]; Houston[0.045]; Houston[0.045]; York[0.044]; Central[0.043]; Atlanta[0.043]; Saturday[0.043]; Division[0.043]; Division[0.043]; Montreal[0.043]; Montreal[0.043]; Chicago[0.042]; Chicago[0.042]; Florida[0.042]; Florida[0.042]; Colorado[0.041]; Colorado[0.041]; Western[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Florida ==> ENTITY: \u001b[32mMiami Marlins\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.204:0.204[0]; log p(e|m)= -4.828:-4.828[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFlorida[0.047]; Cincinnati[0.043]; Cincinnati[0.043]; Philadelphia[0.043]; Philadelphia[0.043]; Montreal[0.043]; Montreal[0.043]; Pittsburgh[0.043]; Pittsburgh[0.043]; Houston[0.042]; Houston[0.042]; Atlanta[0.041]; Colorado[0.041]; Colorado[0.041]; Chicago[0.041]; Chicago[0.041]; Division[0.040]; Division[0.040]; Francisco[0.040]; Francisco[0.040]; Central[0.040]; Saturday[0.040]; Louis[0.039]; Louis[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Seattle ==> ENTITY: \u001b[32mSeattle Mariners\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.183:0.183[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSeattle[0.044]; Cleveland[0.043]; Cleveland[0.043]; League[0.043]; Houston[0.042]; Cincinnati[0.042]; Oakland[0.042]; Oakland[0.042]; Baltimore[0.042]; Baltimore[0.042]; Pittsburgh[0.042]; Philadelphia[0.041]; Western[0.041]; Western[0.041]; Boston[0.041]; Boston[0.041]; Toronto[0.041]; Toronto[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Central Division ==> ENTITY: \u001b[32mNational League Central\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.185:0.185[0]; log p(e|m)= -0.989:-0.989[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; League[0.044]; League[0.044]; League[0.044]; games[0.043]; games[0.043]; Division[0.043]; Division[0.043]; Division[0.043]; Cleveland[0.042]; Cleveland[0.042]; Boston[0.041]; Boston[0.041]; winning[0.041]; won[0.041]; Detroit[0.041]; Detroit[0.041]; Western[0.040]; Baltimore[0.040]; Baltimore[0.040]; Baseball[0.040]; Baseball[0.040]; Toronto[0.040]; Toronto[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Boston ==> ENTITY: \u001b[32mBoston Red Sox\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.226:0.226[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBoston[0.049]; York[0.047]; York[0.047]; Montreal[0.045]; Cleveland[0.044]; Cleveland[0.044]; Philadelphia[0.044]; Pittsburgh[0.044]; League[0.043]; City[0.043]; City[0.043]; Toronto[0.043]; Toronto[0.043]; Cincinnati[0.042]; Detroit[0.042]; Detroit[0.042]; Baltimore[0.042]; Baltimore[0.042]; Houston[0.042]; Oakland[0.042]; Oakland[0.042]; Louis[0.041]; Colorado[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Major League ==> ENTITY: \u001b[32mMajor League Baseball\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.171:0.171[0]; log p(e|m)= -0.478:-0.478[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.047]; League[0.047]; games[0.044]; games[0.044]; Baseball[0.044]; Baseball[0.044]; won[0.043]; Cleveland[0.043]; winning[0.042]; Major[0.042]; played[0.042]; Baltimore[0.041]; Detroit[0.041]; Boston[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; American[0.040]; Texas[0.040]; Toronto[0.040]; Chicago[0.039]; Standings[0.039]; Milwaukee[0.039]; standings[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = American League Eastern Division\u001b[39m ==> ENTITY: \u001b[32mAmerican League East\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.230:0.230[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; League[0.044]; Boston[0.044]; Cleveland[0.043]; winning[0.043]; won[0.043]; Detroit[0.043]; Detroit[0.043]; Baltimore[0.042]; Baltimore[0.042]; games[0.042]; games[0.042]; Toronto[0.042]; Oakland[0.042]; Baseball[0.041]; Baseball[0.041]; standings[0.040]; Division[0.040]; Division[0.040]; lost[0.040]; played[0.040]; Chicago[0.040]; York[0.039]; York[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cincinnati ==> ENTITY: \u001b[32mCincinnati Reds\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.202:0.202[0]; log p(e|m)= -3.772:-3.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCincinnati[0.045]; Cleveland[0.044]; League[0.043]; Philadelphia[0.043]; Philadelphia[0.043]; Oakland[0.043]; Houston[0.043]; Houston[0.043]; Pittsburgh[0.042]; Pittsburgh[0.042]; Baltimore[0.042]; Louis[0.041]; Louis[0.041]; Detroit[0.041]; Boston[0.041]; Toronto[0.041]; Saturday[0.040]; Saturday[0.040]; Milwaukee[0.040]; National[0.040]; Minnesota[0.040]; Texas[0.040]; Atlanta[0.040]; Atlanta[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pittsburgh ==> ENTITY: \u001b[32mPittsburgh Pirates\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.223:0.223[0]; log p(e|m)= -4.017:-4.017[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.044]; Philadelphia[0.044]; Pittsburgh[0.044]; Cleveland[0.043]; Cincinnati[0.043]; Cincinnati[0.043]; League[0.043]; Montreal[0.042]; Montreal[0.042]; Baltimore[0.041]; Boston[0.041]; Toronto[0.041]; Oakland[0.041]; Florida[0.041]; Florida[0.041]; Houston[0.041]; Houston[0.041]; Detroit[0.041]; Louis[0.040]; Louis[0.040]; Atlanta[0.040]; Atlanta[0.040]; Colorado[0.040]; Colorado[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Seattle ==> ENTITY: \u001b[32mSeattle Mariners\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.219:0.219[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.050]; games[0.050]; Seattle[0.050]; Baseball[0.049]; Cleveland[0.048]; Cleveland[0.048]; League[0.048]; League[0.048]; League[0.048]; Houston[0.048]; Oakland[0.047]; Oakland[0.047]; Baltimore[0.047]; Baltimore[0.047]; Philadelphia[0.047]; played[0.047]; Western[0.047]; Boston[0.047]; Boston[0.047]; Toronto[0.046]; Toronto[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cincinnati ==> ENTITY: \u001b[32mCincinnati Reds\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.174:0.174[0]; log p(e|m)= -3.772:-3.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCincinnati[0.047]; Philadelphia[0.044]; Philadelphia[0.044]; Houston[0.044]; Houston[0.044]; Pittsburgh[0.043]; Pittsburgh[0.043]; Louis[0.042]; Louis[0.042]; Saturday[0.041]; Atlanta[0.041]; Montreal[0.041]; Florida[0.041]; Florida[0.041]; Chicago[0.041]; Chicago[0.041]; Division[0.040]; Division[0.040]; Central[0.040]; Colorado[0.040]; Colorado[0.040]; August[0.040]; Francisco[0.040]; Francisco[0.040]; \t\n\r [====>.........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 16s272ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 577/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1065testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lynagh\u001b[39m ==> ENTITY: \u001b[32mMichael Lynagh\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.047]; Wallaby[0.045]; Rugby[0.045]; Lynagh[0.045]; Lynagh[0.045]; career[0.043]; Saracens[0.042]; league[0.042]; coaching[0.042]; Dwyer[0.041]; Dwyer[0.041]; club[0.041]; club[0.041]; captain[0.041]; Michael[0.041]; Leicester[0.041]; Leicester[0.041]; return[0.041]; win[0.040]; kicked[0.040]; Bob[0.040]; big[0.040]; coach[0.039]; Victory[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Laurent Benezech\u001b[39m ==> ENTITY: \u001b[32mLaurent Bénézech\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.047]; Saracens[0.045]; Harlequins[0.044]; Wasps[0.044]; club[0.043]; England[0.043]; England[0.043]; game[0.042]; Football[0.042]; league[0.042]; scrum[0.041]; Leicester[0.041]; season[0.041]; Bath[0.041]; clubs[0.041]; Laurent[0.041]; team[0.040]; Philippe[0.040]; Cabannes[0.040]; scored[0.040]; French[0.040]; French[0.040]; debut[0.040]; captain[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEnglish[0.048]; French[0.048]; England[0.046]; England[0.046]; team[0.045]; league[0.044]; league[0.044]; national[0.044]; Philippe[0.044]; Laurent[0.043]; Laurent[0.043]; club[0.043]; club[0.043]; Football[0.043]; clubs[0.043]; game[0.043]; Rugby[0.043]; Carling[0.042]; Michael[0.041]; season[0.041]; early[0.041]; win[0.041]; debut[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bob Dwyer\u001b[39m ==> ENTITY: \u001b[32mBob Dwyer\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.047]; Rugby[0.045]; team[0.044]; Wallaby[0.043]; Harlequins[0.043]; England[0.043]; Leicester[0.042]; Leicester[0.042]; scrum[0.042]; coaching[0.042]; coach[0.042]; Dwyer[0.042]; captain[0.042]; club[0.041]; club[0.041]; league[0.041]; Saracens[0.040]; debut[0.040]; alongside[0.040]; game[0.040]; enjoyed[0.039]; Kyran[0.039]; win[0.039]; Nigel[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Northampton ==> ENTITY: \u001b[32mNorthampton Saints\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.209:0.209[0]; log p(e|m)= -3.147:-3.147[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.048]; England[0.045]; England[0.045]; Harlequins[0.045]; league[0.045]; Bath[0.043]; team[0.043]; Wasps[0.043]; season[0.042]; Gloucester[0.042]; Bristol[0.042]; Carling[0.041]; Football[0.041]; champions[0.041]; clubs[0.041]; captain[0.040]; scrum[0.040]; day[0.040]; Bracken[0.040]; start[0.039]; Orrell[0.039]; alongside[0.039]; home[0.038]; scoring[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wallaby ==> ENTITY: \u001b[32mAustralia national rugby union team\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.562:-0.562[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.051]; Rugby[0.051]; league[0.045]; coaching[0.043]; Saracens[0.042]; game[0.042]; national[0.042]; captain[0.042]; coach[0.042]; club[0.041]; club[0.041]; career[0.041]; win[0.040]; debut[0.040]; Leicester[0.040]; Leicester[0.040]; kicked[0.040]; Lynagh[0.040]; Lynagh[0.040]; Lynagh[0.040]; Saturday[0.040]; Union[0.039]; early[0.039]; home[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.114:0.114[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrench[0.046]; England[0.044]; England[0.044]; team[0.043]; league[0.043]; league[0.043]; national[0.043]; Philippe[0.042]; Laurent[0.042]; Laurent[0.042]; club[0.042]; Football[0.042]; clubs[0.042]; game[0.042]; Rugby[0.041]; Carling[0.041]; champions[0.041]; season[0.040]; win[0.040]; debut[0.040]; captain[0.040]; coaching[0.040]; goal[0.039]; early[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rugby Football Union\u001b[39m ==> ENTITY: \u001b[32mRugby Football Union\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.048]; England[0.048]; Bath[0.047]; league[0.047]; club[0.046]; Harlequins[0.046]; clubs[0.046]; team[0.046]; Gloucester[0.045]; scrum[0.045]; Bristol[0.045]; Wasps[0.043]; Northampton[0.043]; game[0.043]; captain[0.041]; kicking[0.041]; Carling[0.041]; Lynagh[0.040]; Bracken[0.040]; London[0.040]; season[0.040]; champions[0.040]; Orrell[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lynagh\u001b[39m ==> ENTITY: \u001b[32mMichael Lynagh\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.046]; Rugby[0.044]; Lynagh[0.044]; Lynagh[0.044]; Harlequins[0.043]; career[0.042]; season[0.042]; Bath[0.042]; England[0.042]; England[0.042]; game[0.042]; scrum[0.041]; Saracens[0.041]; league[0.041]; league[0.041]; Gloucester[0.041]; coaching[0.041]; Dwyer[0.040]; try[0.040]; try[0.040]; club[0.040]; club[0.040]; captain[0.040]; Michael[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lynagh\u001b[39m ==> ENTITY: \u001b[32mMichael Lynagh\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.046]; Wallaby[0.045]; Rugby[0.044]; Lynagh[0.044]; Lynagh[0.044]; Lynagh[0.044]; Harlequins[0.043]; career[0.042]; England[0.042]; game[0.041]; scrum[0.041]; Saracens[0.041]; league[0.041]; coaching[0.041]; Dwyer[0.040]; Dwyer[0.040]; try[0.040]; try[0.040]; club[0.040]; club[0.040]; captain[0.040]; Michael[0.040]; Leicester[0.040]; Leicester[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Michael Lynagh\u001b[39m ==> ENTITY: \u001b[32mMichael Lynagh\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.047]; Wallaby[0.045]; Rugby[0.045]; Lynagh[0.044]; Lynagh[0.044]; Lynagh[0.044]; career[0.042]; game[0.042]; Saracens[0.042]; league[0.042]; coaching[0.041]; Dwyer[0.041]; Dwyer[0.041]; club[0.041]; club[0.041]; captain[0.041]; Leicester[0.040]; Leicester[0.040]; return[0.040]; win[0.039]; kicked[0.039]; Bob[0.039]; big[0.039]; coach[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gloucester ==> ENTITY: \u001b[32mGloucester Rugby\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.233:0.233[0]; log p(e|m)= -3.058:-3.058[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.048]; England[0.045]; England[0.045]; Bristol[0.045]; Harlequins[0.045]; Bath[0.044]; Saracens[0.044]; Northampton[0.043]; Wasps[0.043]; club[0.042]; league[0.042]; scrum[0.042]; team[0.042]; Carling[0.041]; clubs[0.040]; season[0.040]; national[0.039]; north[0.039]; London[0.039]; captain[0.039]; Nigel[0.039]; game[0.038]; home[0.038]; home[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dwyer\u001b[39m ==> ENTITY: \u001b[32mBob Dwyer\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.048]; Rugby[0.046]; Wallaby[0.044]; Leicester[0.043]; Leicester[0.043]; coaching[0.043]; coach[0.043]; Dwyer[0.043]; captain[0.043]; club[0.042]; club[0.042]; league[0.042]; Saracens[0.041]; debut[0.041]; Bob[0.040]; win[0.040]; Nigel[0.040]; career[0.040]; kicked[0.039]; Michael[0.039]; return[0.039]; Lynagh[0.039]; Lynagh[0.039]; Lynagh[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wasps ==> ENTITY: \u001b[32mLondon Wasps\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.214:0.214[0]; log p(e|m)= -0.070:-0.070[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.048]; Northampton[0.046]; Harlequins[0.045]; England[0.044]; England[0.044]; Gloucester[0.044]; Bristol[0.044]; league[0.044]; Bath[0.044]; scrum[0.043]; team[0.042]; Carling[0.042]; clubs[0.041]; season[0.040]; Football[0.040]; champions[0.040]; kicking[0.040]; captain[0.039]; start[0.039]; Orrell[0.039]; try[0.039]; try[0.039]; scored[0.038]; scoring[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Harlequins\u001b[39m ==> ENTITY: \u001b[32mHarlequin F.C.\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.232:0.232[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.046]; England[0.044]; England[0.044]; Leicester[0.043]; team[0.043]; league[0.043]; league[0.043]; club[0.043]; season[0.043]; Bath[0.042]; Saracens[0.042]; Wasps[0.042]; Bristol[0.041]; coaching[0.041]; Northampton[0.041]; coach[0.041]; Football[0.041]; game[0.040]; Gloucester[0.040]; scrum[0.040]; clubs[0.040]; Carling[0.039]; London[0.039]; captain[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.046]; England[0.045]; England[0.045]; club[0.044]; club[0.044]; league[0.044]; league[0.044]; team[0.043]; clubs[0.042]; debut[0.042]; game[0.041]; season[0.041]; Football[0.041]; Carling[0.041]; win[0.040]; Rugby[0.040]; Gloucester[0.040]; start[0.040]; career[0.040]; Bracken[0.040]; Leicester[0.039]; Leicester[0.039]; return[0.039]; Michael[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = English ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.102:0.102[0]; log p(e|m)= -1.355:-1.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.047]; rugby[0.047]; league[0.046]; club[0.043]; club[0.043]; team[0.043]; Leicester[0.042]; Leicester[0.042]; Saracens[0.042]; Wallaby[0.041]; national[0.041]; Nigel[0.041]; game[0.041]; career[0.041]; captain[0.041]; London[0.040]; London[0.040]; debut[0.040]; coaching[0.040]; win[0.040]; Michael[0.040]; millionaire[0.040]; Saturday[0.039]; alongside[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bath ==> ENTITY: \u001b[32mBath Rugby\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.207:0.207[0]; log p(e|m)= -2.604:-2.604[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBristol[0.052]; Gloucester[0.051]; Northampton[0.050]; England[0.050]; England[0.050]; Rugby[0.045]; Harlequins[0.044]; Wasps[0.043]; league[0.042]; Orrell[0.041]; team[0.040]; clubs[0.040]; scrum[0.040]; game[0.040]; season[0.039]; Bracken[0.038]; Football[0.038]; captain[0.038]; Carling[0.038]; Lynagh[0.037]; home[0.037]; kicking[0.036]; start[0.036]; French[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland national rugby union team\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.184:0.184[0]; log p(e|m)= -3.772:-3.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.046]; Rugby[0.046]; rugby[0.045]; Harlequins[0.043]; team[0.043]; Bath[0.043]; Bristol[0.043]; league[0.042]; league[0.042]; season[0.042]; Saracens[0.042]; club[0.042]; club[0.042]; Leicester[0.041]; Northampton[0.041]; scrum[0.040]; Gloucester[0.040]; game[0.040]; Carling[0.040]; clubs[0.040]; Wasps[0.040]; Football[0.039]; captain[0.039]; debut[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland national rugby union team\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.177:0.177[0]; log p(e|m)= -3.772:-3.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; Rugby[0.047]; Harlequins[0.044]; team[0.044]; Bath[0.044]; Bristol[0.044]; league[0.043]; season[0.043]; Saracens[0.042]; club[0.042]; Northampton[0.041]; scrum[0.041]; Gloucester[0.041]; game[0.041]; Carling[0.040]; clubs[0.040]; Wasps[0.040]; Football[0.040]; captain[0.040]; win[0.040]; Bracken[0.039]; scoring[0.039]; London[0.039]; points[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Philippe Sella\u001b[39m ==> ENTITY: \u001b[32mPhilippe Sella\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.213:0.213[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nscrum[0.047]; rugby[0.046]; Rugby[0.046]; Lynagh[0.045]; Lynagh[0.045]; Lynagh[0.045]; Wallaby[0.044]; league[0.044]; league[0.044]; Harlequins[0.043]; club[0.043]; club[0.043]; game[0.043]; Saracens[0.043]; career[0.043]; Laurent[0.042]; Laurent[0.042]; Michael[0.042]; coaching[0.042]; French[0.042]; French[0.042]; team[0.042]; goal[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leicester ==> ENTITY: \u001b[32mLeicester Tigers\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.211:0.211[0]; log p(e|m)= -3.194:-3.194[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.048]; rugby[0.047]; Leicester[0.044]; league[0.044]; England[0.043]; Harlequins[0.043]; club[0.043]; club[0.043]; scrum[0.042]; Saracens[0.042]; London[0.041]; London[0.041]; team[0.041]; win[0.041]; game[0.041]; kicked[0.040]; Saturday[0.040]; captain[0.039]; Michael[0.039]; Nigel[0.039]; debut[0.039]; Wallaby[0.039]; return[0.039]; career[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bristol ==> ENTITY: \u001b[32mBristol Rugby\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.192:0.192[0]; log p(e|m)= -3.772:-3.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.046]; England[0.046]; England[0.046]; Harlequins[0.046]; league[0.045]; Northampton[0.045]; Bath[0.045]; Wasps[0.044]; Gloucester[0.043]; clubs[0.042]; season[0.042]; scrum[0.041]; half[0.040]; day[0.040]; scoring[0.040]; captain[0.039]; victory[0.039]; Football[0.039]; start[0.039]; finished[0.039]; victories[0.039]; Carling[0.039]; scored[0.038]; Bracken[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rugby Union\u001b[39m ==> ENTITY: \u001b[32mRugby union\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.056]; Wallaby[0.046]; league[0.045]; club[0.044]; Saracens[0.043]; Leicester[0.043]; Leicester[0.043]; coaching[0.042]; national[0.042]; captain[0.041]; win[0.041]; debut[0.041]; coach[0.041]; career[0.041]; Lynagh[0.040]; Lynagh[0.040]; Lynagh[0.040]; Saturday[0.040]; return[0.039]; Dwyer[0.039]; Dwyer[0.039]; kicked[0.039]; London[0.039]; London[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leicester ==> ENTITY: \u001b[32mLeicester Tigers\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.191:0.191[0]; log p(e|m)= -3.194:-3.194[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.049]; rugby[0.048]; Leicester[0.044]; league[0.044]; club[0.044]; club[0.044]; Saracens[0.042]; London[0.042]; London[0.042]; win[0.042]; kicked[0.041]; Saturday[0.041]; captain[0.040]; Michael[0.040]; Nigel[0.040]; debut[0.040]; Wallaby[0.040]; return[0.040]; career[0.040]; home[0.040]; north[0.040]; goal[0.039]; big[0.039]; coaching[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Laurent Cabannes\u001b[39m ==> ENTITY: \u001b[32mLaurent Cabannes\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.047]; club[0.043]; England[0.043]; England[0.043]; clubs[0.042]; Saracens[0.042]; league[0.042]; league[0.042]; champions[0.042]; victories[0.042]; Harlequins[0.042]; penalties[0.041]; Wasps[0.041]; win[0.041]; Football[0.041]; points[0.041]; French[0.041]; French[0.041]; Orrell[0.041]; game[0.040]; season[0.040]; team[0.040]; scrum[0.040]; Laurent[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Orrell ==> ENTITY: \u001b[32mOrrell R.U.F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.194:0.194[0]; log p(e|m)= -0.931:-0.931[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.045]; Northampton[0.045]; Rugby[0.045]; clubs[0.044]; England[0.044]; England[0.044]; Bristol[0.044]; Gloucester[0.044]; Bath[0.043]; Harlequins[0.043]; league[0.043]; scrum[0.042]; game[0.041]; season[0.040]; start[0.040]; Carling[0.040]; champions[0.040]; Wasps[0.040]; captain[0.040]; Football[0.039]; Bracken[0.039]; scored[0.038]; Kyran[0.038]; half[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kyran Bracken\u001b[39m ==> ENTITY: \u001b[32mKyran Bracken\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBath[0.046]; England[0.046]; England[0.046]; Bristol[0.045]; team[0.044]; Rugby[0.044]; Leicester[0.043]; Harlequins[0.042]; scrum[0.041]; captain[0.041]; Northampton[0.041]; Gloucester[0.041]; Saracens[0.041]; season[0.041]; Nigel[0.041]; game[0.040]; Wasps[0.040]; clubs[0.040]; club[0.040]; coaching[0.040]; enjoyed[0.039]; day[0.039]; league[0.039]; league[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Saracens ==> ENTITY: \u001b[32mSaracens F.C.\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.220:0.220[0]; log p(e|m)= -0.567:-0.567[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.048]; rugby[0.048]; Harlequins[0.045]; England[0.044]; England[0.044]; team[0.043]; Leicester[0.043]; Leicester[0.043]; league[0.043]; club[0.042]; club[0.042]; Gloucester[0.042]; Wallaby[0.041]; scrum[0.040]; coaching[0.040]; Saturday[0.039]; debut[0.039]; game[0.039]; London[0.039]; London[0.039]; alongside[0.039]; career[0.039]; captain[0.039]; captain[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Will Carling\u001b[39m ==> ENTITY: \u001b[32mWill Carling\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; England[0.047]; Rugby[0.044]; team[0.044]; Bath[0.044]; Bristol[0.042]; captain[0.042]; scrum[0.042]; game[0.042]; Gloucester[0.041]; Kyran[0.041]; Harlequins[0.041]; Northampton[0.041]; Nigel[0.041]; Bracken[0.041]; early[0.040]; Saracens[0.040]; Wasps[0.040]; league[0.040]; club[0.040]; half[0.040]; enjoyed[0.040]; successes[0.040]; big[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.049]; club[0.046]; club[0.046]; league[0.046]; debut[0.043]; win[0.042]; Rugby[0.042]; career[0.041]; Leicester[0.041]; Leicester[0.041]; return[0.041]; Michael[0.041]; early[0.041]; Saturday[0.041]; English[0.040]; goal[0.040]; national[0.040]; coaching[0.040]; began[0.040]; home[0.040]; Bob[0.040]; Victory[0.040]; rugby[0.040]; enjoyed[0.040]; \t\n\r [====>.........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 16s65ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 608/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1125testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South African ==> ENTITY: \u001b[32mSouth Africa\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrican[0.048]; South[0.046]; Pretoria[0.045]; Africa[0.044]; Afrikaners[0.043]; Afrikaners[0.043]; Afrikaners[0.043]; Cape[0.043]; Cape[0.043]; Saturday[0.043]; Afrikaner[0.041]; ethnic[0.041]; said[0.041]; government[0.041]; Northern[0.040]; society[0.040]; Town[0.040]; campaign[0.040]; territory[0.040]; home[0.040]; place[0.039]; province[0.039]; rule[0.039]; told[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Afrikaner\u001b[39m ==> ENTITY: \u001b[32mAfrikaner\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrikaners[0.046]; Afrikaners[0.046]; Afrikaners[0.046]; African[0.044]; African[0.044]; country[0.043]; South[0.042]; South[0.042]; South[0.042]; province[0.042]; Africa[0.042]; Cape[0.041]; Cape[0.041]; Mandela[0.041]; whites[0.041]; Pretoria[0.041]; Northern[0.040]; National[0.040]; territory[0.039]; territory[0.039]; regions[0.039]; provincial[0.039]; Town[0.039]; race[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.092:0.092[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.046]; country[0.046]; territory[0.044]; Africa[0.043]; war[0.043]; Dutch[0.043]; descendants[0.042]; settlers[0.042]; South[0.042]; African[0.042]; people[0.042]; French[0.042]; speak[0.042]; rule[0.041]; rule[0.041]; Northern[0.041]; state[0.040]; sovereignty[0.040]; saying[0.040]; saying[0.040]; provinces[0.040]; National[0.040]; Coloureds[0.039]; April[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Africa ==> ENTITY: \u001b[32mSouth Africa\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.263:-0.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.046]; country[0.046]; country[0.046]; African[0.046]; Africa[0.045]; Cape[0.042]; Mandela[0.042]; world[0.042]; National[0.041]; came[0.041]; April[0.041]; President[0.041]; constitution[0.041]; provinces[0.040]; regard[0.040]; Afrikaners[0.040]; Northern[0.040]; parts[0.040]; taking[0.040]; Congress[0.040]; state[0.040]; saying[0.040]; saying[0.040]; Afrikaner[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Africa ==> ENTITY: \u001b[32mSouth Africa\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.162:0.162[0]; log p(e|m)= -0.263:-0.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.047]; South[0.047]; country[0.047]; African[0.047]; African[0.047]; Pretoria[0.046]; Africa[0.046]; Saturday[0.044]; Cape[0.043]; Mandela[0.043]; world[0.043]; National[0.042]; government[0.042]; level[0.042]; level[0.042]; came[0.042]; home[0.042]; April[0.042]; President[0.041]; place[0.041]; constitution[0.041]; regard[0.041]; campaign[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pretoria\u001b[39m ==> ENTITY: \u001b[32mPretoria\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.045]; South[0.045]; African[0.045]; African[0.045]; African[0.045]; government[0.044]; Saturday[0.044]; country[0.044]; Africa[0.044]; place[0.042]; Mandela[0.041]; President[0.041]; National[0.041]; area[0.040]; home[0.040]; conference[0.040]; delegated[0.040]; seek[0.040]; Northern[0.040]; April[0.039]; majority[0.039]; Cape[0.039]; Cape[0.039]; regard[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Viljoen\u001b[39m ==> ENTITY: \u001b[32mConstand Viljoen\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.129:0.129[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.044]; elections[0.043]; elections[0.043]; National[0.043]; President[0.043]; African[0.042]; war[0.042]; saying[0.042]; saying[0.042]; threats[0.042]; Northern[0.042]; Mandela[0.042]; Viljoen[0.042]; Afrikaans[0.041]; Afrikaners[0.041]; According[0.041]; demanded[0.041]; people[0.041]; right[0.041]; conference[0.041]; Africa[0.041]; Coloureds[0.040]; cooperating[0.040]; constitution[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Afrikaners\u001b[39m ==> ENTITY: \u001b[32mAfrikaner\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrican[0.045]; Coloureds[0.044]; Cape[0.044]; President[0.044]; settlers[0.043]; Afrikaans[0.043]; South[0.043]; country[0.043]; country[0.043]; National[0.043]; territory[0.042]; Mandela[0.041]; whites[0.041]; whites[0.041]; Africa[0.041]; descendants[0.041]; elections[0.040]; elections[0.040]; regard[0.040]; conference[0.040]; people[0.040]; speak[0.039]; majority[0.039]; majority[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Viljoen\u001b[39m ==> ENTITY: \u001b[32mConstand Viljoen\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.044]; South[0.044]; elections[0.043]; elections[0.043]; National[0.042]; Pretoria[0.042]; President[0.042]; party[0.042]; African[0.042]; war[0.042]; saying[0.042]; saying[0.042]; campaign[0.042]; threats[0.041]; wing[0.041]; said[0.041]; Northern[0.041]; Mandela[0.041]; Viljoen[0.041]; Viljoen[0.041]; Freedom[0.041]; government[0.041]; Afrikaners[0.041]; Afrikaners[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cape Town\u001b[39m ==> ENTITY: \u001b[32mCape Town\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.047]; African[0.045]; African[0.045]; area[0.045]; Cape[0.044]; Saturday[0.044]; Pretoria[0.043]; place[0.043]; Northern[0.042]; home[0.041]; provincial[0.041]; said[0.041]; Freedom[0.041]; government[0.040]; province[0.040]; territory[0.040]; local[0.040]; regions[0.040]; Afrikaners[0.040]; Afrikaners[0.040]; Afrikaners[0.040]; campaign[0.039]; Afrikaner[0.039]; conference[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Constand Viljoen\u001b[39m ==> ENTITY: \u001b[32mConstand Viljoen\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.044]; South[0.044]; elections[0.043]; Pretoria[0.043]; party[0.042]; African[0.042]; African[0.042]; saying[0.042]; campaign[0.042]; wing[0.042]; wing[0.042]; said[0.042]; Northern[0.041]; Viljoen[0.041]; Freedom[0.041]; government[0.041]; Afrikaners[0.041]; Afrikaners[0.041]; Afrikaners[0.041]; leader[0.041]; taking[0.041]; Afrikaner[0.041]; right[0.040]; right[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dutch ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.794:-0.794[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.045]; country[0.045]; French[0.045]; German[0.043]; territory[0.043]; Afrikaans[0.043]; settlers[0.042]; descendants[0.042]; Africa[0.042]; South[0.042]; Cape[0.041]; African[0.041]; provinces[0.041]; sovereignty[0.041]; people[0.041]; war[0.041]; state[0.041]; rule[0.041]; rule[0.041]; Northern[0.040]; speak[0.040]; form[0.040]; whites[0.040]; whites[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nelson Mandela\u001b[39m ==> ENTITY: \u001b[32mNelson Mandela\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.092:0.092[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.045]; country[0.045]; African[0.044]; South[0.044]; South[0.044]; Pretoria[0.043]; local[0.042]; majority[0.041]; majority[0.041]; Africa[0.041]; Africa[0.041]; National[0.041]; President[0.041]; government[0.041]; Afrikaners[0.041]; Afrikaners[0.041]; state[0.041]; model[0.041]; model[0.041]; society[0.041]; world[0.040]; Cape[0.040]; Afrikaner[0.040]; television[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Afrikaners\u001b[39m ==> ENTITY: \u001b[32mAfrikaner\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrikaners[0.045]; Afrikaners[0.045]; Afrikaner[0.044]; African[0.044]; African[0.044]; African[0.044]; Cape[0.043]; Cape[0.043]; President[0.043]; South[0.042]; South[0.042]; country[0.042]; National[0.042]; territory[0.041]; Mandela[0.040]; whites[0.040]; Pretoria[0.040]; Africa[0.040]; populated[0.040]; society[0.039]; leader[0.039]; said[0.039]; elections[0.039]; regard[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Afrikaans\u001b[39m ==> ENTITY: \u001b[32mAfrikaans\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCape[0.047]; Afrikaners[0.047]; Coloureds[0.044]; Northern[0.044]; South[0.043]; country[0.043]; state[0.043]; Africa[0.042]; territory[0.042]; settlers[0.042]; whites[0.041]; provinces[0.041]; speak[0.041]; television[0.041]; introduced[0.041]; German[0.041]; people[0.040]; parts[0.040]; news[0.040]; French[0.040]; majority[0.039]; rule[0.039]; rule[0.039]; Dutch[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.090:0.090[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.046]; country[0.046]; South[0.043]; African[0.042]; Africa[0.042]; provinces[0.042]; territory[0.042]; state[0.042]; constitution[0.042]; sovereignty[0.042]; race[0.042]; race[0.042]; war[0.041]; German[0.041]; settlers[0.041]; descendants[0.041]; majority[0.041]; majority[0.041]; whites[0.040]; whites[0.040]; parts[0.040]; rule[0.040]; rule[0.040]; speak[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Northern Cape\u001b[39m ==> ENTITY: \u001b[32mNorthern Cape\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCape[0.046]; South[0.044]; South[0.044]; province[0.044]; Afrikaners[0.043]; Afrikaners[0.043]; Afrikaners[0.043]; African[0.042]; African[0.042]; regions[0.042]; area[0.042]; country[0.042]; provincial[0.041]; Africa[0.041]; form[0.041]; Afrikaner[0.041]; territory[0.041]; whites[0.040]; regard[0.040]; Pretoria[0.040]; April[0.039]; level[0.039]; level[0.039]; local[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = African National Congress\u001b[39m ==> ENTITY: \u001b[32mAfrican National Congress\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.137:0.137[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.044]; South[0.044]; Mandela[0.044]; President[0.043]; Pretoria[0.043]; majority[0.043]; majority[0.043]; country[0.042]; country[0.042]; Africa[0.042]; Africa[0.042]; constitution[0.042]; Afrikaners[0.041]; Afrikaners[0.041]; elections[0.041]; elections[0.041]; making[0.041]; Afrikaner[0.041]; government[0.041]; April[0.040]; Northern[0.040]; sovereignty[0.040]; taking[0.040]; broke[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Freedom Front\u001b[39m ==> ENTITY: \u001b[32mFreedom Front Plus\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleader[0.044]; elections[0.043]; Afrikaners[0.042]; Afrikaners[0.042]; Afrikaners[0.042]; Afrikaner[0.042]; African[0.042]; African[0.042]; party[0.042]; President[0.042]; government[0.042]; said[0.042]; South[0.041]; South[0.041]; province[0.041]; wing[0.041]; wing[0.041]; Northern[0.041]; Pretoria[0.041]; campaign[0.041]; country[0.041]; Viljoen[0.040]; Viljoen[0.040]; provincial[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Northern Cape\u001b[39m ==> ENTITY: \u001b[32mNorthern Cape\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.046]; Afrikaans[0.045]; Afrikaners[0.044]; state[0.044]; Coloureds[0.044]; African[0.044]; country[0.043]; provinces[0.043]; Africa[0.043]; form[0.042]; territory[0.042]; whites[0.041]; Mandela[0.041]; Dutch[0.041]; settlers[0.040]; war[0.040]; French[0.040]; National[0.040]; people[0.040]; rule[0.040]; rule[0.040]; introduced[0.039]; making[0.039]; conference[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Viljoen\u001b[39m ==> ENTITY: \u001b[32mConstand Viljoen\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.044]; South[0.044]; elections[0.043]; elections[0.043]; National[0.042]; President[0.042]; African[0.042]; war[0.042]; saying[0.042]; saying[0.042]; threats[0.041]; wing[0.041]; said[0.041]; Northern[0.041]; Mandela[0.041]; Viljoen[0.041]; Viljoen[0.041]; Afrikaans[0.041]; Afrikaners[0.041]; taking[0.041]; According[0.041]; Afrikaner[0.041]; demanded[0.041]; people[0.040]; \t\n\r [=====>........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s992ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 629/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 979testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = St. Petersburg ==> ENTITY: \u001b[32mSt. Petersburg, Florida\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.057:0.057[0]; log p(e|m)= -1.313:-1.313[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclosed[0.046]; Louis[0.045]; Friday[0.044]; Friday[0.044]; Fla[0.044]; said[0.043]; Chicago[0.043]; completed[0.042]; common[0.041]; completes[0.041]; Jones[0.041]; Jones[0.041]; Jones[0.041]; Jones[0.041]; Daniels[0.041]; Daniels[0.041]; Medical[0.040]; Medical[0.040]; largest[0.040]; Pharmaceuticals[0.040]; Pharmaceuticals[0.040]; stock[0.040]; stock[0.040]; treating[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Levoxyl\u001b[39m ==> ENTITY: \u001b[32mLevothyroxine\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.119:0.119[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nprescription[0.051]; hormone[0.049]; thyroid[0.047]; hypothyroidism[0.047]; Medical[0.046]; Medical[0.046]; products[0.046]; treating[0.046]; pharmaceutical[0.045]; Pharmaceuticals[0.044]; Pharmaceuticals[0.044]; synthetic[0.044]; common[0.042]; Fla[0.041]; completed[0.041]; Petersburg[0.041]; said[0.041]; Industries[0.040]; closed[0.040]; acquisition[0.040]; acquisition[0.040]; Chicago[0.040]; manufactures[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.044:0.044[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFriday[0.045]; Friday[0.045]; said[0.044]; largest[0.043]; Jones[0.043]; Jones[0.043]; Jones[0.043]; Jones[0.043]; Louis[0.043]; closed[0.042]; Daniels[0.041]; Daniels[0.041]; completes[0.041]; completed[0.041]; Medical[0.040]; Medical[0.040]; stock[0.040]; stock[0.040]; acquisition[0.040]; acquisition[0.040]; shares[0.040]; Petersburg[0.039]; pharmaceutical[0.039]; common[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = St Louis ==> ENTITY: \u001b[32mSt. Louis\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.042:0.042[0]; log p(e|m)= -0.092:-0.092[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMedical[0.048]; Medical[0.048]; Chicago[0.047]; Friday[0.045]; Friday[0.045]; thyroid[0.044]; completed[0.043]; closed[0.043]; Fla[0.043]; Petersburg[0.043]; Jones[0.043]; Jones[0.043]; Jones[0.043]; Jones[0.043]; Pharmaceuticals[0.043]; Pharmaceuticals[0.043]; said[0.043]; largest[0.042]; treating[0.042]; prescription[0.042]; Industries[0.042]; products[0.042]; hypothyroidism[0.042]; \t\n\r [=====>........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 16s242ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 633/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 996testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tavildara\u001b[39m ==> ENTITY: \u001b[32mTavildara\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTavildara[0.048]; Tavildara[0.048]; Tajikistan[0.045]; Tajikistan[0.045]; Tajik[0.044]; Tajik[0.044]; mountains[0.042]; remote[0.041]; town[0.041]; town[0.041]; town[0.041]; government[0.041]; surrounding[0.040]; rebels[0.040]; Islamic[0.040]; Pamir[0.040]; strategically[0.040]; commander[0.040]; opposition[0.040]; devastated[0.040]; devastated[0.040]; sporadic[0.039]; chief[0.039]; General[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tajikistan ==> ENTITY: \u001b[32mTajikistan\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.217:0.217[0]; log p(e|m)= -0.368:-0.368[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTajikistan[0.048]; Islamic[0.044]; Tavildara[0.044]; Tavildara[0.044]; Tavildara[0.044]; Tavildara[0.044]; Tavildara[0.044]; Tajik[0.044]; Tajik[0.044]; held[0.041]; mountains[0.041]; mountains[0.041]; Pamir[0.041]; Dushanbe[0.041]; east[0.041]; east[0.041]; government[0.041]; capital[0.040]; cases[0.040]; miles[0.040]; miles[0.040]; troops[0.039]; troops[0.039]; troops[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tavildara\u001b[39m ==> ENTITY: \u001b[32mTavildara\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.104:0.104[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTajikistan[0.046]; Tajik[0.045]; borders[0.043]; territory[0.043]; straddles[0.042]; long[0.042]; school[0.042]; Afghanistan[0.042]; fallen[0.042]; inhabited[0.042]; rebel[0.041]; town[0.041]; town[0.041]; government[0.041]; government[0.041]; government[0.041]; Islamic[0.041]; inhabitants[0.041]; killed[0.041]; strategically[0.041]; road[0.040]; displaced[0.040]; widespread[0.040]; conflict[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Afghanistan ==> ENTITY: \u001b[32mAfghanistan\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.223:-0.223[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoldiers[0.045]; Tavildara[0.044]; Tavildara[0.044]; Tavildara[0.044]; coalition[0.044]; war[0.044]; war[0.044]; displaced[0.043]; government[0.043]; Tajikistan[0.043]; Islamic[0.042]; people[0.041]; borders[0.041]; groups[0.041]; help[0.040]; territory[0.040]; fighters[0.040]; fighting[0.040]; thousands[0.040]; February[0.039]; children[0.039]; China[0.039]; conflict[0.039]; mountains[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tavildara\u001b[39m ==> ENTITY: \u001b[32mTavildara\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTavildara[0.046]; Tavildara[0.046]; Tavildara[0.046]; Tavildara[0.046]; Tajikistan[0.043]; Tajikistan[0.043]; capital[0.042]; miles[0.042]; miles[0.042]; Tajik[0.042]; Tajik[0.042]; km[0.042]; km[0.042]; east[0.040]; east[0.040]; mountains[0.040]; mountains[0.040]; territory[0.040]; straddles[0.040]; Dushanbe[0.039]; remote[0.039]; fallen[0.039]; village[0.039]; rebel[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Islamic ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.087:0.087[0]; log p(e|m)= -0.212:-0.212[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTavildara[0.046]; Tavildara[0.046]; Tavildara[0.046]; Tavildara[0.046]; Tavildara[0.046]; forces[0.042]; forces[0.042]; nearby[0.042]; group[0.042]; Tajikistan[0.041]; Tajikistan[0.041]; said[0.040]; said[0.040]; Tajik[0.040]; Tajik[0.040]; government[0.040]; fighters[0.040]; east[0.040]; town[0.040]; town[0.040]; town[0.040]; troops[0.040]; troops[0.040]; troops[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.021:0.021[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfighting[0.043]; fighting[0.043]; war[0.043]; white[0.043]; soldiers[0.043]; soldiers[0.043]; soldiers[0.043]; forces[0.043]; town[0.042]; town[0.042]; inhabitants[0.042]; homes[0.042]; school[0.042]; declined[0.041]; destroyed[0.041]; holds[0.041]; women[0.041]; fight[0.041]; fighters[0.041]; hospital[0.041]; doubling[0.040]; elderly[0.040]; elderly[0.040]; aid[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.094:0.094[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nterritory[0.045]; war[0.044]; war[0.044]; began[0.043]; Afghanistan[0.043]; men[0.042]; government[0.042]; Nations[0.042]; women[0.042]; women[0.042]; February[0.042]; civil[0.042]; soldiers[0.041]; Tajikistan[0.041]; United[0.041]; communists[0.041]; fighting[0.041]; held[0.041]; capital[0.041]; groups[0.040]; conflict[0.040]; Tavildara[0.040]; Tavildara[0.040]; Tavildara[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tavildara\u001b[39m ==> ENTITY: \u001b[32mTavildara\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTavildara[0.048]; Tavildara[0.048]; Tavildara[0.048]; Tajikistan[0.045]; Tajikistan[0.045]; capital[0.045]; miles[0.044]; miles[0.044]; km[0.044]; km[0.044]; borders[0.043]; east[0.042]; east[0.042]; mountains[0.042]; mountains[0.042]; territory[0.042]; straddles[0.042]; Dushanbe[0.042]; Afghanistan[0.041]; remote[0.041]; fallen[0.041]; village[0.041]; rebel[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tajik ==> ENTITY: \u001b[32mTajikistan\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.185:0.185[0]; log p(e|m)= -2.180:-2.180[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTavildara[0.048]; Tavildara[0.048]; Tavildara[0.048]; Tajikistan[0.048]; Tajikistan[0.048]; Tajik[0.045]; Islamic[0.043]; troops[0.042]; troops[0.042]; troops[0.042]; government[0.041]; mountains[0.041]; Pamir[0.041]; remote[0.040]; forces[0.039]; forces[0.039]; said[0.039]; casualties[0.039]; armed[0.039]; town[0.038]; town[0.038]; town[0.038]; held[0.038]; commander[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tajik ==> ENTITY: \u001b[32mTajikistan\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.176:0.176[0]; log p(e|m)= -2.180:-2.180[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTavildara[0.047]; Tavildara[0.047]; Tavildara[0.047]; Tajikistan[0.047]; Tajikistan[0.047]; Tajik[0.045]; troops[0.043]; troops[0.043]; troops[0.043]; Islamic[0.042]; government[0.040]; mountains[0.040]; mountains[0.040]; Pamir[0.040]; forces[0.039]; forces[0.039]; remote[0.039]; commander[0.039]; fighters[0.038]; said[0.038]; said[0.038]; revealed[0.038]; casualties[0.038]; armed[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Islamic ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.102:0.102[0]; log p(e|m)= -0.212:-0.212[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTavildara[0.047]; Tavildara[0.047]; Tavildara[0.047]; Afghanistan[0.044]; groups[0.043]; earth[0.043]; people[0.043]; coalition[0.042]; government[0.042]; linking[0.042]; war[0.041]; war[0.041]; Tajikistan[0.041]; said[0.041]; apparently[0.040]; barefoot[0.040]; territory[0.040]; important[0.040]; thousands[0.040]; inhabited[0.040]; humanitarian[0.040]; conflict[0.039]; liberal[0.039]; trees[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tavildara\u001b[39m ==> ENTITY: \u001b[32mTavildara\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTavildara[0.047]; Tavildara[0.047]; Tajikistan[0.044]; Tajikistan[0.044]; capital[0.043]; miles[0.043]; miles[0.043]; km[0.042]; km[0.042]; borders[0.041]; east[0.041]; east[0.041]; mountains[0.041]; mountains[0.041]; territory[0.041]; straddles[0.040]; Dushanbe[0.040]; Afghanistan[0.040]; remote[0.040]; fallen[0.040]; village[0.039]; rebel[0.039]; town[0.039]; town[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pamir ==> ENTITY: \u001b[32mPamir Mountains\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.217:0.217[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTavildara[0.047]; Tavildara[0.047]; Tavildara[0.047]; Tavildara[0.047]; Tavildara[0.047]; mountains[0.046]; mountains[0.046]; Tajikistan[0.043]; Tajikistan[0.043]; east[0.043]; east[0.043]; Tajik[0.043]; Tajik[0.043]; miles[0.042]; miles[0.042]; village[0.042]; town[0.042]; town[0.042]; town[0.042]; surrounding[0.041]; remote[0.041]; km[0.041]; km[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tajik ==> ENTITY: \u001b[32mTajikistan\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.116:0.116[0]; log p(e|m)= -2.180:-2.180[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTavildara[0.050]; soldiers[0.044]; soldiers[0.044]; soldiers[0.044]; inhabited[0.043]; government[0.042]; government[0.042]; government[0.042]; refugees[0.042]; war[0.042]; forces[0.041]; homes[0.041]; inhabitants[0.041]; aid[0.040]; fighters[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; women[0.040]; women[0.040]; killed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tajikistan ==> ENTITY: \u001b[32mTajikistan\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.178:0.178[0]; log p(e|m)= -0.368:-0.368[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTajikistan[0.048]; Islamic[0.045]; Tavildara[0.045]; Tavildara[0.045]; Tavildara[0.045]; Tajik[0.044]; Tajik[0.044]; held[0.041]; mountains[0.041]; mountains[0.041]; Pamir[0.041]; government[0.041]; troops[0.040]; troops[0.040]; troops[0.040]; took[0.040]; took[0.040]; vital[0.040]; August[0.040]; control[0.040]; control[0.040]; remote[0.040]; forces[0.039]; forces[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dushanbe\u001b[39m ==> ENTITY: \u001b[32mDushanbe\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTajikistan[0.047]; Tajikistan[0.047]; Tavildara[0.047]; Tavildara[0.047]; Tavildara[0.047]; Afghanistan[0.045]; capital[0.044]; territory[0.044]; Islamic[0.043]; troops[0.043]; fighters[0.043]; August[0.043]; east[0.043]; east[0.043]; village[0.042]; forces[0.042]; forces[0.042]; town[0.042]; town[0.042]; miles[0.042]; miles[0.042]; Pamir[0.042]; Friday[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tavildara\u001b[39m ==> ENTITY: \u001b[32mTavildara\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTavildara[0.047]; Tavildara[0.047]; Tavildara[0.047]; Tajikistan[0.044]; Tajikistan[0.044]; miles[0.043]; Tajik[0.043]; Tajik[0.043]; km[0.043]; east[0.041]; mountains[0.041]; mountains[0.041]; remote[0.040]; village[0.040]; town[0.040]; town[0.040]; town[0.040]; government[0.040]; surrounding[0.039]; rebels[0.039]; rebels[0.039]; Islamic[0.039]; Pamir[0.039]; strategically[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tajikistan ==> ENTITY: \u001b[32mTajikistan\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.184:0.184[0]; log p(e|m)= -0.368:-0.368[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfghanistan[0.046]; Islamic[0.044]; Tavildara[0.044]; Tavildara[0.044]; Tavildara[0.044]; territory[0.043]; China[0.043]; borders[0.043]; displaced[0.042]; groups[0.041]; held[0.041]; mountains[0.041]; Dushanbe[0.041]; east[0.041]; east[0.041]; government[0.041]; important[0.041]; capital[0.040]; cases[0.040]; women[0.040]; women[0.040]; straddles[0.040]; miles[0.040]; miles[0.040]; \t\n\r [=====>........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 16s265ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 652/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 975testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colombo ==> ENTITY: \u001b[32mColombo\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.165:0.165[0]; log p(e|m)= -0.194:-0.194[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.047]; Cricket[0.046]; Lanka[0.045]; Lanka[0.045]; series[0.043]; match[0.043]; Australia[0.043]; Australia[0.043]; Australia[0.043]; tournament[0.042]; Sri[0.041]; Sri[0.041]; Teams[0.040]; day[0.040]; Ranatunga[0.040]; Jayasuriya[0.040]; Michael[0.040]; Michael[0.040]; overs[0.040]; world[0.040]; toss[0.040]; limited[0.039]; McGrath[0.039]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Roshan Mahanama\u001b[39m ==> ENTITY: \u001b[32mRoshan Mahanama\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.046]; Dharmasena[0.044]; Kaluwitharana[0.044]; Hashan[0.043]; Australia[0.043]; Aravinda[0.043]; Waugh[0.042]; Waugh[0.042]; match[0.042]; Ranatunga[0.042]; Sanath[0.042]; overs[0.042]; Muralitharan[0.041]; Upul[0.041]; Ponting[0.041]; Asanka[0.041]; Chaminda[0.041]; Chandana[0.040]; captain[0.040]; captain[0.040]; day[0.040]; Vaas[0.040]; Lanka[0.040]; Lanka[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mark Waugh\u001b[39m ==> ENTITY: \u001b[32mMark Waugh\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWaugh[0.044]; cricket[0.044]; Australia[0.044]; Australia[0.044]; Australia[0.044]; match[0.043]; Stuart[0.043]; McGrath[0.043]; toss[0.043]; bat[0.042]; Kaluwitharana[0.041]; Bevan[0.041]; Cricket[0.041]; overs[0.041]; series[0.041]; Ponting[0.041]; captain[0.040]; captain[0.040]; Michael[0.040]; Michael[0.040]; Gillespie[0.040]; Darren[0.040]; Mahanama[0.040]; Muralitharan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Stuart Law\u001b[39m ==> ENTITY: \u001b[32mStuart Law\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMark[0.044]; cricket[0.044]; overs[0.043]; Darren[0.043]; Ian[0.043]; Cricket[0.043]; bat[0.043]; Michael[0.042]; Michael[0.042]; Steve[0.042]; Bevan[0.042]; McGrath[0.042]; Australia[0.041]; Australia[0.041]; Australia[0.041]; Damien[0.041]; captain[0.041]; captain[0.041]; Ponting[0.040]; Kaluwitharana[0.040]; toss[0.040]; Ricky[0.040]; series[0.040]; match[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ian Healy\u001b[39m ==> ENTITY: \u001b[32mIan Healy\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.046]; Australia[0.046]; Australia[0.046]; Stuart[0.045]; cricket[0.043]; Glenn[0.042]; McGrath[0.042]; Waugh[0.042]; Waugh[0.042]; Mark[0.042]; Bevan[0.041]; match[0.041]; captain[0.041]; captain[0.041]; Steve[0.041]; Gillespie[0.041]; Michael[0.041]; Michael[0.041]; Cricket[0.040]; series[0.040]; Darren[0.040]; Kaluwitharana[0.040]; Ponting[0.039]; Mahanama[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Romesh Kaluwitharana\u001b[39m ==> ENTITY: \u001b[32mRomesh Kaluwitharana\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.044]; Upul[0.044]; match[0.043]; Muralitharan[0.043]; Australia[0.043]; Australia[0.043]; Dharmasena[0.042]; Vaas[0.042]; bat[0.042]; Chandana[0.042]; Mahanama[0.042]; overs[0.042]; Jayasuriya[0.041]; Aravinda[0.041]; Lanka[0.041]; Lanka[0.041]; Asanka[0.041]; Chaminda[0.041]; Sanath[0.041]; Ranatunga[0.041]; Hashan[0.041]; Ponting[0.040]; toss[0.040]; tournament[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sri Lanka ==> ENTITY: \u001b[32mSri Lanka national cricket team\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.212:0.212[0]; log p(e|m)= -2.263:-2.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.045]; Cricket[0.044]; Kaluwitharana[0.043]; Australia[0.043]; Australia[0.043]; Australia[0.043]; Colombo[0.043]; Lanka[0.043]; Muralitharan[0.042]; match[0.042]; overs[0.042]; Mahanama[0.042]; tournament[0.041]; Upul[0.041]; Hashan[0.041]; Jayasuriya[0.041]; Ponting[0.041]; series[0.040]; Vaas[0.040]; day[0.040]; McGrath[0.040]; world[0.040]; Dharmasena[0.040]; Chandana[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ricky Ponting\u001b[39m ==> ENTITY: \u001b[32mRicky Ponting\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.044]; Australia[0.044]; Australia[0.044]; toss[0.043]; cricket[0.043]; Michael[0.043]; Michael[0.043]; Waugh[0.043]; Waugh[0.043]; Stuart[0.042]; match[0.041]; McGrath[0.041]; Kaluwitharana[0.041]; bat[0.041]; Singer[0.041]; Mark[0.041]; Bevan[0.041]; Glenn[0.041]; Cricket[0.041]; overs[0.040]; day[0.040]; captain[0.040]; captain[0.040]; Muralitharan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Michael Bevan\u001b[39m ==> ENTITY: \u001b[32mMichael Bevan\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.045]; Australia[0.045]; Australia[0.045]; match[0.044]; Stuart[0.044]; bat[0.043]; cricket[0.043]; toss[0.042]; tournament[0.042]; overs[0.042]; Kaluwitharana[0.042]; series[0.042]; McGrath[0.042]; Cricket[0.041]; Waugh[0.041]; Waugh[0.041]; day[0.040]; Muralitharan[0.040]; limited[0.040]; captain[0.040]; captain[0.040]; Ponting[0.040]; won[0.040]; Mark[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chaminda Vaas\u001b[39m ==> ENTITY: \u001b[32mChaminda Vaas\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKaluwitharana[0.045]; Mahanama[0.044]; cricket[0.043]; overs[0.043]; Muralitharan[0.043]; Upul[0.042]; Aravinda[0.042]; Jayasuriya[0.042]; Hashan[0.042]; McGrath[0.042]; Chandana[0.042]; Dharmasena[0.041]; Ricky[0.041]; Australia[0.041]; Waugh[0.041]; Waugh[0.041]; Lanka[0.041]; Sanath[0.041]; Asanka[0.041]; Gillespie[0.041]; Mark[0.040]; match[0.040]; Stuart[0.040]; Ponting[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.190:0.190[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; Australia[0.048]; cricket[0.046]; match[0.044]; Cricket[0.044]; tournament[0.042]; Waugh[0.042]; Waugh[0.042]; series[0.041]; Colombo[0.041]; toss[0.041]; Stuart[0.041]; captain[0.041]; captain[0.041]; bat[0.040]; day[0.040]; Mark[0.040]; McGrath[0.040]; won[0.040]; overs[0.040]; Ian[0.040]; world[0.039]; Gillespie[0.039]; Bevan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Asanka Gurusinha\u001b[39m ==> ENTITY: \u001b[32mAsanka Gurusinha\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKaluwitharana[0.044]; cricket[0.044]; Aravinda[0.043]; Waugh[0.043]; Waugh[0.043]; Australia[0.042]; match[0.042]; Sanath[0.042]; Upul[0.042]; Dharmasena[0.042]; Muralitharan[0.042]; Chandana[0.042]; Mahanama[0.042]; Jayasuriya[0.041]; Bevan[0.041]; Ranatunga[0.041]; Ponting[0.041]; Vaas[0.041]; overs[0.041]; toss[0.041]; bat[0.041]; Lanka[0.040]; Lanka[0.040]; Chaminda[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Darren Lehmann\u001b[39m ==> ENTITY: \u001b[32mDarren Lehmann\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.045]; Australia[0.044]; Australia[0.044]; Australia[0.044]; Cricket[0.044]; match[0.043]; Stuart[0.041]; Waugh[0.041]; Waugh[0.041]; Bevan[0.041]; captain[0.041]; captain[0.041]; Ponting[0.041]; series[0.041]; toss[0.041]; Michael[0.041]; Michael[0.041]; Kaluwitharana[0.041]; McGrath[0.041]; Mark[0.041]; bat[0.041]; tournament[0.041]; overs[0.041]; day[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jason Gillespie\u001b[39m ==> ENTITY: \u001b[32mJason Gillespie\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbat[0.044]; Ponting[0.043]; Kaluwitharana[0.043]; Stuart[0.043]; McGrath[0.043]; Waugh[0.042]; Waugh[0.042]; overs[0.042]; toss[0.042]; cricket[0.042]; match[0.042]; Australia[0.042]; Australia[0.042]; Australia[0.042]; Mark[0.041]; Michael[0.041]; Michael[0.041]; Glenn[0.041]; Darren[0.041]; Muralitharan[0.041]; captain[0.041]; captain[0.041]; Vaas[0.040]; Steve[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arjuna Ranatunga\u001b[39m ==> ENTITY: \u001b[32mArjuna Ranatunga\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.216:0.216[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMahanama[0.045]; Hashan[0.044]; cricket[0.044]; Colombo[0.044]; Kaluwitharana[0.043]; Muralitharan[0.043]; Lanka[0.043]; Lanka[0.043]; Upul[0.043]; Australia[0.042]; Australia[0.042]; Dharmasena[0.042]; match[0.041]; Jayasuriya[0.041]; Sri[0.041]; Sri[0.041]; Aravinda[0.041]; Asanka[0.040]; toss[0.040]; captain[0.040]; captain[0.040]; Waugh[0.040]; Waugh[0.040]; Chandana[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Muthiah Muralitharan\u001b[39m ==> ENTITY: \u001b[32mMuttiah Muralitharan\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.221:0.221[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKaluwitharana[0.046]; Mahanama[0.044]; McGrath[0.044]; Australia[0.044]; Upul[0.042]; Aravinda[0.042]; Hashan[0.042]; Ponting[0.042]; Jayasuriya[0.042]; Chandana[0.042]; Lanka[0.042]; Waugh[0.041]; Waugh[0.041]; series[0.041]; Chaminda[0.041]; match[0.041]; Dharmasena[0.041]; Stuart[0.041]; Kumara[0.041]; Ranatunga[0.041]; Vaas[0.040]; Bevan[0.040]; Asanka[0.040]; Sanath[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Steve Waugh\u001b[39m ==> ENTITY: \u001b[32mSteve Waugh\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.223:0.223[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWaugh[0.045]; Australia[0.045]; Australia[0.045]; Australia[0.045]; McGrath[0.045]; cricket[0.044]; Stuart[0.043]; Gillespie[0.042]; Bevan[0.042]; Cricket[0.042]; match[0.041]; Michael[0.041]; Michael[0.041]; Glenn[0.041]; series[0.041]; Mark[0.040]; Healy[0.040]; toss[0.040]; Ian[0.040]; Kaluwitharana[0.040]; Ponting[0.040]; captain[0.039]; captain[0.039]; Darren[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Glenn McGrath\u001b[39m ==> ENTITY: \u001b[32mGlenn McGrath\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.217:0.217[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.046]; Kaluwitharana[0.045]; Australia[0.045]; Australia[0.045]; Australia[0.045]; toss[0.044]; Stuart[0.044]; Waugh[0.044]; Waugh[0.044]; overs[0.044]; match[0.044]; Ponting[0.044]; Bevan[0.044]; Cricket[0.043]; bat[0.043]; Muralitharan[0.043]; Darren[0.042]; Mark[0.042]; Vaas[0.042]; Gillespie[0.042]; Jayasuriya[0.041]; Steve[0.041]; Ian[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.191:0.191[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; Australia[0.048]; cricket[0.046]; match[0.044]; Cricket[0.044]; tournament[0.042]; Waugh[0.042]; Waugh[0.042]; series[0.041]; Colombo[0.041]; toss[0.041]; Stuart[0.041]; captain[0.041]; bat[0.041]; day[0.040]; Mark[0.040]; McGrath[0.040]; won[0.040]; overs[0.040]; Ian[0.040]; world[0.040]; Gillespie[0.039]; Bevan[0.039]; Ponting[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sri Lanka ==> ENTITY: \u001b[32mSri Lanka national cricket team\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.207:0.207[0]; log p(e|m)= -2.263:-2.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.045]; Cricket[0.044]; Kaluwitharana[0.044]; Australia[0.044]; Australia[0.044]; Australia[0.044]; Colombo[0.043]; Lanka[0.043]; match[0.043]; overs[0.042]; tournament[0.042]; Hashan[0.041]; Jayasuriya[0.041]; Ponting[0.041]; series[0.041]; day[0.041]; McGrath[0.040]; world[0.040]; Ranatunga[0.040]; bat[0.040]; toss[0.040]; Aravinda[0.039]; Waugh[0.039]; Waugh[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Aravinda de Silva\u001b[39m ==> ENTITY: \u001b[32mAravinda de Silva\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.047]; Muralitharan[0.044]; Ranatunga[0.043]; Australia[0.043]; Kaluwitharana[0.043]; Mahanama[0.042]; McGrath[0.042]; Ponting[0.042]; Waugh[0.042]; Waugh[0.042]; bat[0.042]; match[0.042]; Upul[0.042]; Bevan[0.041]; Dharmasena[0.041]; overs[0.041]; Lanka[0.041]; Lanka[0.041]; Jayasuriya[0.040]; Sanath[0.040]; Hashan[0.040]; series[0.040]; Ricky[0.040]; Stuart[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sanath Jayasuriya\u001b[39m ==> ENTITY: \u001b[32mSanath Jayasuriya\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.045]; Kaluwitharana[0.045]; Muralitharan[0.043]; Hashan[0.043]; Australia[0.043]; Australia[0.043]; Colombo[0.043]; Upul[0.042]; overs[0.042]; match[0.042]; toss[0.041]; Mahanama[0.041]; Chandana[0.041]; McGrath[0.041]; Chaminda[0.041]; bat[0.041]; Ranatunga[0.041]; Dharmasena[0.041]; Waugh[0.041]; Waugh[0.041]; series[0.040]; Vaas[0.040]; Aravinda[0.040]; Ponting[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Michael Slater\u001b[39m ==> ENTITY: \u001b[32mMichael Slater\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.045]; Australia[0.045]; Australia[0.045]; cricket[0.045]; series[0.043]; Stuart[0.042]; Waugh[0.042]; Waugh[0.042]; captain[0.042]; captain[0.042]; match[0.041]; Michael[0.041]; Darren[0.041]; Cricket[0.041]; Mark[0.041]; Ponting[0.040]; day[0.040]; Bevan[0.040]; Ricky[0.040]; McGrath[0.040]; Asanka[0.040]; Glenn[0.040]; Steve[0.040]; Ian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Upul Chandana\u001b[39m ==> ENTITY: \u001b[32mUpul Chandana\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMuralitharan[0.045]; Kaluwitharana[0.045]; Australia[0.044]; Hashan[0.043]; series[0.043]; Stuart[0.042]; Jayasuriya[0.042]; Chaminda[0.042]; Vaas[0.042]; Lanka[0.042]; Kumara[0.041]; Dharmasena[0.041]; Ponting[0.041]; Asanka[0.041]; McGrath[0.041]; Ranatunga[0.041]; Darren[0.041]; Mahanama[0.040]; Sanath[0.040]; captain[0.040]; captain[0.040]; Romesh[0.040]; Bevan[0.040]; Damien[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.200:0.200[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; Australia[0.048]; cricket[0.046]; match[0.044]; Cricket[0.043]; Waugh[0.042]; Waugh[0.042]; Mahanama[0.041]; series[0.041]; tournament[0.041]; Colombo[0.041]; toss[0.041]; Stuart[0.041]; captain[0.041]; captain[0.041]; Hashan[0.040]; bat[0.040]; day[0.040]; Muralitharan[0.040]; McGrath[0.040]; overs[0.040]; Kaluwitharana[0.040]; Mark[0.040]; Darren[0.040]; \t\n\r [=====>........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 16s58ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 677/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1012testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NationsBank Corp.\u001b[39m ==> ENTITY: \u001b[32mNationsBank\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNationsBank[0.046]; bank[0.045]; acquisition[0.045]; Bancshares[0.043]; company[0.043]; Bancorp[0.043]; banking[0.043]; banking[0.043]; Ancell[0.042]; Mercantile[0.042]; Mercantile[0.042]; based[0.042]; deal[0.041]; assets[0.041]; franchise[0.041]; holding[0.040]; Missouri[0.040]; Missouri[0.040]; Midwest[0.040]; Midwest[0.040]; billion[0.040]; billion[0.040]; analysts[0.040]; banks[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mKansas City\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mKansas City metropolitan area\u001b[39m <---> \u001b[32mKansas City, Missouri\u001b[39m\t\nSCORES: global= 0.257:0.254[\u001b[31m0.003\u001b[39m]; local()= 0.117:0.074[\u001b[31m0.042\u001b[39m]; log p(e|m)= -1.448:-0.805[\u001b[32m0.643\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfranchise[0.044]; James[0.043]; Missouri[0.043]; Missouri[0.043]; Jones[0.042]; Michael[0.042]; bank[0.042]; bank[0.042]; bank[0.042]; Edwards[0.041]; regional[0.041]; Mercantile[0.041]; Mercantile[0.041]; Mercantile[0.041]; Mercantile[0.041]; Mercantile[0.041]; Commerce[0.041]; Commerce[0.041]; Commerce[0.041]; Commerce[0.041]; return[0.041]; Midwest[0.041]; Midwest[0.041]; Midwest[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NationsBank\u001b[39m ==> ENTITY: \u001b[32mNationsBank\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBank[0.050]; NationsBank[0.046]; NationsBank[0.046]; banking[0.045]; KeyCorp[0.043]; acquisitions[0.042]; Ancell[0.042]; banks[0.042]; banks[0.042]; Banc[0.041]; stock[0.041]; deal[0.041]; deal[0.041]; deal[0.041]; assets[0.040]; Norwest[0.040]; Missouri[0.040]; Missouri[0.040]; Missouri[0.040]; seeking[0.040]; shareholder[0.040]; bid[0.040]; bid[0.040]; bid[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = KeyCorp\u001b[39m ==> ENTITY: \u001b[32mKeyBank\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBank[0.045]; Bank[0.045]; Financial[0.045]; Group[0.043]; Mercantile[0.043]; Mercantile[0.043]; KeyCorp[0.043]; bank[0.042]; Norwest[0.042]; Norwest[0.042]; banking[0.041]; NationsBank[0.041]; Ohio[0.041]; Corp[0.041]; Corp[0.041]; Corp[0.041]; bid[0.040]; acquisitions[0.040]; Minneapolis[0.040]; Missouri[0.040]; Missouri[0.040]; Missouri[0.040]; Missouri[0.040]; assets[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Norwest\u001b[39m ==> ENTITY: \u001b[32mNorwest\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.239:0.239[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbanking[0.045]; Bank[0.044]; Bank[0.044]; Norwest[0.044]; bank[0.043]; banks[0.043]; banks[0.043]; Financial[0.042]; Midwest[0.042]; Mercantile[0.041]; Mercantile[0.041]; Mercantile[0.041]; Corp[0.041]; Corp[0.041]; Corp[0.041]; Commerce[0.041]; Commerce[0.041]; Illinois[0.041]; KeyCorp[0.041]; KeyCorp[0.041]; Chicago[0.041]; Chicago[0.041]; billion[0.041]; billion[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Commerce Bancshares Inc.\u001b[39m ==> ENTITY: \u001b[32mCommerce Bancshares\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMercantile[0.044]; Mercantile[0.044]; Mercantile[0.044]; Mercantile[0.044]; Mercantile[0.044]; Financial[0.043]; Commerce[0.043]; Commerce[0.043]; Commerce[0.043]; Bank[0.043]; Ancell[0.041]; banking[0.041]; banking[0.041]; City[0.041]; bank[0.041]; bank[0.041]; bank[0.041]; company[0.040]; Kansas[0.040]; NationsBank[0.040]; thrift[0.040]; Bancorp[0.040]; Missouri[0.040]; Missouri[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Midwest ==> ENTITY: \u001b[32mMidwestern United States\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.125:0.125[0]; log p(e|m)= -0.031:-0.031[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMidwest[0.047]; Midwest[0.047]; state[0.046]; regional[0.045]; franchise[0.044]; Chicago[0.042]; industry[0.042]; Missouri[0.042]; Missouri[0.042]; City[0.041]; Kansas[0.041]; involving[0.041]; size[0.041]; strong[0.040]; landscape[0.040]; outside[0.040]; James[0.040]; added[0.040]; proposed[0.040]; Jones[0.040]; Michael[0.040]; Brad[0.040]; position[0.040]; big[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Missouri ==> ENTITY: \u001b[32mMissouri\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.114:0.114[0]; log p(e|m)= -0.270:-0.270[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMissouri[0.046]; Missouri[0.046]; Missouri[0.046]; Ohio[0.044]; states[0.044]; Illinois[0.043]; central[0.042]; federal[0.042]; Bank[0.041]; Bank[0.041]; banks[0.041]; banks[0.041]; Midwest[0.041]; outside[0.040]; brought[0.040]; based[0.040]; based[0.040]; said[0.040]; said[0.040]; said[0.040]; having[0.040]; Corp[0.040]; Corp[0.040]; Corp[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Banc One\u001b[39m ==> ENTITY: \u001b[32mBank One Corporation\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.226:0.226[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBank[0.045]; Bank[0.045]; Financial[0.044]; NationsBank[0.044]; Corp[0.043]; Corp[0.043]; Corp[0.043]; bank[0.042]; banking[0.042]; Mercantile[0.042]; Mercantile[0.042]; Mercantile[0.042]; Banc[0.041]; KeyCorp[0.041]; KeyCorp[0.041]; buying[0.041]; Norwest[0.040]; Norwest[0.040]; Group[0.040]; bid[0.040]; Chicago[0.040]; Chicago[0.040]; banks[0.040]; banks[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NationsBank\u001b[39m ==> ENTITY: \u001b[32mNationsBank\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.125:0.125[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNationsBank[0.048]; NationsBank[0.048]; Ancell[0.044]; stock[0.043]; deal[0.043]; deal[0.043]; shareholder[0.042]; bid[0.042]; bid[0.042]; bid[0.042]; McDonald[0.041]; percent[0.041]; analysts[0.041]; deals[0.041]; Michael[0.041]; initially[0.041]; analyst[0.040]; buyer[0.040]; Boatmen[0.040]; Boatmen[0.040]; Boatmen[0.040]; large[0.040]; large[0.040]; large[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Midwest ==> ENTITY: \u001b[32mMidwestern United States\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.126:0.126[0]; log p(e|m)= -0.031:-0.031[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMidwest[0.048]; regional[0.047]; franchise[0.046]; Ohio[0.045]; Chicago[0.044]; Financial[0.044]; seeking[0.044]; company[0.043]; buying[0.043]; Illinois[0.043]; Missouri[0.043]; Missouri[0.043]; Bank[0.043]; billion[0.043]; billion[0.043]; billion[0.043]; based[0.042]; based[0.042]; based[0.042]; based[0.042]; based[0.042]; size[0.042]; strong[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.071:0.071[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMichael[0.048]; deal[0.046]; based[0.045]; outside[0.045]; said[0.045]; Jones[0.044]; Midwest[0.044]; Midwest[0.044]; Louis[0.043]; Commerce[0.043]; market[0.043]; Edward[0.043]; industry[0.043]; banking[0.043]; banking[0.043]; company[0.043]; trophy[0.043]; Brad[0.043]; Mercantile[0.043]; Mercantile[0.043]; acquisition[0.042]; biggest[0.042]; bank[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NationsBank\u001b[39m ==> ENTITY: \u001b[32mNationsBank\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbank[0.045]; NationsBank[0.045]; acquisition[0.044]; banking[0.044]; banking[0.044]; Corp[0.042]; Bancshares[0.042]; Bancshares[0.042]; company[0.042]; Bancorp[0.042]; franchise[0.041]; Mercantile[0.041]; Mercantile[0.041]; Ancell[0.041]; based[0.041]; based[0.041]; banks[0.041]; Commerce[0.040]; Commerce[0.040]; deal[0.040]; Kansas[0.040]; assets[0.040]; assets[0.040]; holding[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = First Chicago\u001b[39m ==> ENTITY: \u001b[32mFirst Chicago Bank\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.225:0.225[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBank[0.046]; Bank[0.046]; banking[0.044]; bank[0.044]; Financial[0.043]; Chicago[0.042]; seeking[0.042]; seeking[0.042]; Mercantile[0.041]; Mercantile[0.041]; Mercantile[0.041]; banks[0.041]; banks[0.041]; Corp[0.041]; Corp[0.041]; Corp[0.041]; KeyCorp[0.041]; KeyCorp[0.041]; NationsBank[0.041]; Banc[0.040]; Banc[0.040]; Commerce[0.040]; Commerce[0.040]; Group[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = A.G. Edwards & Sons\u001b[39m ==> ENTITY: \u001b[32mA. G. Edwards\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.218:0.218[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFinancial[0.046]; company[0.043]; Group[0.042]; KeyCorp[0.042]; KeyCorp[0.042]; Mercantile[0.042]; Mercantile[0.042]; Mercantile[0.042]; Mercantile[0.042]; Mercantile[0.042]; Bank[0.042]; Bank[0.042]; Louis[0.041]; Louis[0.041]; bank[0.041]; bank[0.041]; bank[0.041]; Bancorp[0.041]; assets[0.041]; assets[0.041]; Minneapolis[0.041]; regional[0.040]; asset[0.040]; Bancshares[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Missouri ==> ENTITY: \u001b[32mMissouri\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.270:-0.270[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMissouri[0.046]; Kansas[0.044]; Jones[0.044]; state[0.043]; City[0.042]; Midwest[0.042]; Midwest[0.042]; Midwest[0.042]; James[0.042]; regional[0.042]; establish[0.042]; Commerce[0.041]; Commerce[0.041]; Commerce[0.041]; Commerce[0.041]; Mercantile[0.041]; Mercantile[0.041]; Mercantile[0.041]; Mercantile[0.041]; Mercantile[0.041]; franchise[0.041]; bank[0.041]; bank[0.041]; bank[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Missouri ==> ENTITY: \u001b[32mMissouri\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.096:0.096[0]; log p(e|m)= -0.270:-0.270[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMissouri[0.049]; Missouri[0.049]; Ohio[0.046]; states[0.046]; Illinois[0.045]; central[0.044]; federal[0.044]; Bank[0.043]; banks[0.043]; banks[0.043]; Midwest[0.043]; outside[0.042]; brought[0.042]; based[0.042]; said[0.042]; said[0.042]; said[0.042]; Corp[0.042]; Corp[0.042]; Corp[0.042]; wanted[0.042]; Weber[0.042]; beat[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Missouri ==> ENTITY: \u001b[32mMissouri\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.125:0.125[0]; log p(e|m)= -0.270:-0.270[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMissouri[0.046]; Missouri[0.046]; Missouri[0.046]; Ohio[0.044]; states[0.043]; Illinois[0.043]; central[0.042]; federal[0.042]; Mercantile[0.041]; Mercantile[0.041]; bank[0.041]; Bank[0.041]; Bank[0.041]; banks[0.041]; banks[0.041]; return[0.041]; Midwest[0.040]; outside[0.040]; brought[0.040]; based[0.040]; based[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Missouri ==> ENTITY: \u001b[32mMissouri\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.086:0.086[0]; log p(e|m)= -0.270:-0.270[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMissouri[0.047]; Kansas[0.045]; state[0.043]; City[0.043]; Midwest[0.043]; Midwest[0.043]; Michael[0.042]; Mercantile[0.041]; Mercantile[0.041]; franchise[0.041]; bank[0.041]; industry[0.041]; banks[0.041]; involving[0.041]; Jones[0.041]; Louis[0.041]; outside[0.041]; based[0.041]; said[0.041]; Corp[0.041]; Edward[0.040]; protected[0.040]; market[0.040]; Bancorp[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Missouri ==> ENTITY: \u001b[32mMissouri\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.270:-0.270[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMissouri[0.046]; Missouri[0.046]; Representatives[0.045]; James[0.045]; Edwards[0.044]; Ohio[0.043]; states[0.043]; Illinois[0.043]; seeking[0.042]; federal[0.041]; return[0.041]; Mercantile[0.040]; Mercantile[0.040]; Mercantile[0.040]; bank[0.040]; bank[0.040]; Bank[0.040]; Bank[0.040]; banks[0.040]; Commerce[0.040]; Commerce[0.040]; Commerce[0.040]; Midwest[0.040]; Louis[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NationsBank\u001b[39m ==> ENTITY: \u001b[32mNationsBank\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNationsBank[0.049]; NationsBank[0.049]; Ancell[0.045]; banks[0.045]; stock[0.044]; deal[0.044]; deal[0.044]; deal[0.044]; Missouri[0.043]; seeking[0.043]; shareholder[0.043]; bid[0.043]; bid[0.043]; bid[0.043]; Midwest[0.043]; McDonald[0.043]; percent[0.042]; analysts[0.042]; deals[0.042]; Michael[0.042]; initially[0.042]; analyst[0.041]; buyer[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Midwest ==> ENTITY: \u001b[32mMidwestern United States\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.118:0.118[0]; log p(e|m)= -0.031:-0.031[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMidwest[0.047]; regional[0.046]; franchise[0.045]; Chicago[0.043]; Missouri[0.042]; Missouri[0.042]; state[0.042]; City[0.041]; Kansas[0.041]; involving[0.041]; based[0.041]; based[0.041]; size[0.041]; Mercantile[0.041]; Mercantile[0.041]; landscape[0.041]; Jones[0.041]; Michael[0.041]; Brad[0.040]; position[0.040]; big[0.040]; flurry[0.040]; holding[0.040]; deal[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Midwest ==> ENTITY: \u001b[32mMidwestern United States\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.031:-0.031[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstock[0.044]; states[0.044]; Ohio[0.043]; outside[0.043]; Chicago[0.043]; Chicago[0.043]; central[0.043]; seeking[0.042]; winning[0.042]; Illinois[0.042]; Missouri[0.041]; Missouri[0.041]; Missouri[0.041]; Bank[0.041]; Bank[0.041]; bid[0.041]; bid[0.041]; federal[0.041]; based[0.041]; based[0.041]; large[0.040]; large[0.040]; representing[0.040]; expect[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Illinois ==> ENTITY: \u001b[32mIllinois\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.170:-0.170[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstates[0.043]; Missouri[0.043]; Missouri[0.043]; Missouri[0.043]; Missouri[0.043]; Weber[0.042]; Weber[0.042]; Weber[0.042]; Midwest[0.042]; Midwest[0.042]; Ohio[0.042]; James[0.042]; seeking[0.042]; seeking[0.042]; federal[0.041]; Bank[0.041]; Bank[0.041]; Chicago[0.041]; Chicago[0.041]; based[0.041]; based[0.041]; based[0.041]; central[0.041]; prize[0.040]; \t\n\r [=====>........................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 16s179ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 701/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1058testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland cricket team\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.190:0.190[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.046]; England[0.046]; England[0.046]; England[0.046]; England[0.046]; cricket[0.044]; batsman[0.043]; Birmingham[0.043]; matches[0.042]; Cricket[0.041]; Lancashire[0.041]; Edgbaston[0.041]; Hollioake[0.040]; international[0.040]; seamer[0.040]; Surrey[0.040]; Pakistan[0.040]; Pakistan[0.040]; Pakistan[0.040]; overs[0.040]; Atherton[0.040]; rounder[0.039]; day[0.039]; second[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Salim Malik\u001b[39m ==> ENTITY: \u001b[32mSaleem Malik\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.044]; Pakistan[0.044]; Akram[0.044]; Saqlain[0.043]; batsman[0.043]; Ijaz[0.043]; Wasim[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Khan[0.042]; matches[0.042]; Younis[0.042]; Sohail[0.042]; Waqar[0.042]; Hollioake[0.041]; seamer[0.041]; Moin[0.041]; Inzamam[0.040]; Saeed[0.040]; England[0.040]; England[0.040]; captain[0.040]; captain[0.040]; Rehman[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Matthew Maynard\u001b[39m ==> ENTITY: \u001b[32mMatthew Maynard\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.045]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; batsman[0.043]; Surrey[0.042]; cricket[0.042]; captain[0.042]; captain[0.042]; Hollioake[0.041]; Hollioake[0.041]; Lancashire[0.041]; rounder[0.041]; seamer[0.041]; bat[0.040]; Atherton[0.040]; Edgbaston[0.040]; toss[0.040]; Gough[0.039]; wickets[0.039]; Darren[0.039]; Saqlain[0.039]; overs[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Inzamam-ul-Haq\u001b[39m ==> ENTITY: \u001b[32mInzamam-ul-Haq\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.044]; Pakistan[0.044]; Wasim[0.044]; Mushtaq[0.043]; Mushtaq[0.043]; Akram[0.043]; Saqlain[0.043]; Waqar[0.042]; Younis[0.042]; Ijaz[0.042]; England[0.042]; England[0.042]; Sohail[0.042]; seamer[0.042]; Malik[0.042]; Moin[0.041]; matches[0.040]; Khan[0.040]; Hollioake[0.040]; captain[0.040]; captain[0.040]; wickets[0.040]; Rehman[0.039]; Saeed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pakistan ==> ENTITY: \u001b[32mPakistan national cricket team\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.195:0.195[0]; log p(e|m)= -2.749:-2.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.048]; Pakistan[0.048]; Pakistan[0.048]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; cricket[0.043]; Akram[0.042]; batsman[0.042]; matches[0.041]; Wasim[0.041]; Malik[0.040]; won[0.039]; Ijaz[0.039]; Inzamam[0.039]; Cricket[0.039]; Edgbaston[0.039]; seamer[0.038]; Saeed[0.038]; Ahmed[0.038]; rounder[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Adam Hollioake\u001b[39m ==> ENTITY: \u001b[32mAdam Hollioake\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.044]; Cricket[0.044]; cricket[0.043]; rounder[0.043]; won[0.042]; Darren[0.042]; Hollioake[0.042]; England[0.042]; England[0.042]; England[0.042]; England[0.042]; England[0.042]; England[0.042]; Teams[0.041]; Alan[0.041]; Martin[0.041]; day[0.041]; batsman[0.041]; Mike[0.041]; Graham[0.041]; Graham[0.041]; second[0.041]; captain[0.041]; Surrey[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pakistan ==> ENTITY: \u001b[32mPakistan national cricket team\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.201:0.201[0]; log p(e|m)= -2.749:-2.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.052]; England[0.047]; England[0.047]; England[0.047]; Akram[0.045]; batsman[0.045]; Saqlain[0.044]; Wasim[0.044]; Waqar[0.043]; Malik[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; matches[0.043]; Ijaz[0.042]; Inzamam[0.042]; Khan[0.042]; seamer[0.042]; Rehman[0.041]; Saeed[0.041]; Younis[0.041]; Ahmed[0.041]; Ahmed[0.041]; rounder[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland cricket team\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.180:0.180[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.046]; England[0.046]; England[0.046]; England[0.046]; England[0.046]; cricket[0.045]; batsman[0.043]; matches[0.042]; Cricket[0.041]; Lancashire[0.041]; Edgbaston[0.041]; Birmingham[0.040]; Hollioake[0.040]; Hollioake[0.040]; seamer[0.040]; Surrey[0.040]; Pakistan[0.040]; Pakistan[0.040]; Pakistan[0.040]; overs[0.040]; Atherton[0.040]; Graham[0.040]; Graham[0.040]; rounder[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nick Knight ==> ENTITY: \u001b[32mNick Knight\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.176:0.176[0]; log p(e|m)= -0.550:-0.550[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.044]; England[0.043]; England[0.043]; England[0.043]; England[0.043]; England[0.043]; England[0.043]; Edgbaston[0.043]; Martin[0.042]; matches[0.042]; Hollioake[0.042]; Hollioake[0.042]; overs[0.042]; Atherton[0.041]; wickets[0.041]; cricket[0.041]; Lancashire[0.041]; Robert[0.040]; Peter[0.040]; Gough[0.040]; Mushtaq[0.040]; Mushtaq[0.040]; bat[0.040]; Surrey[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Darren Gough\u001b[39m ==> ENTITY: \u001b[32mDarren Gough\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.045]; Lancashire[0.043]; England[0.043]; England[0.043]; England[0.043]; cricket[0.043]; Hollioake[0.043]; Hollioake[0.043]; Croft[0.042]; overs[0.042]; Edgbaston[0.042]; rounder[0.041]; matches[0.041]; seamer[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Saqlain[0.041]; Atherton[0.041]; wickets[0.041]; Surrey[0.041]; day[0.040]; Pakistan[0.040]; Pakistan[0.040]; Matthew[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Old Trafford ==> ENTITY: \u001b[32mOld Trafford Cricket Ground\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.160:0.160[0]; log p(e|m)= -0.644:-0.644[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLancashire[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; batsman[0.043]; cricket[0.042]; matches[0.042]; second[0.042]; Birmingham[0.041]; Surrey[0.041]; Hollioake[0.041]; Hollioake[0.041]; seamer[0.041]; Edgbaston[0.040]; Saturday[0.040]; Gough[0.040]; Cricket[0.040]; toss[0.040]; Saqlain[0.040]; Croft[0.040]; Thursday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lancashire ==> ENTITY: \u001b[32mLancashire County Cricket Club\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.176:0.176[0]; log p(e|m)= -1.732:-1.732[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; batsman[0.044]; cricket[0.043]; matches[0.042]; Surrey[0.042]; Hollioake[0.041]; Hollioake[0.041]; seamer[0.041]; Edgbaston[0.041]; Gough[0.041]; Cricket[0.041]; Birmingham[0.040]; Croft[0.040]; overs[0.040]; Atherton[0.040]; rounder[0.040]; second[0.040]; captain[0.040]; wickets[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wasim Akram\u001b[39m ==> ENTITY: \u001b[32mWasim Akram\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMushtaq[0.045]; Mushtaq[0.045]; Ijaz[0.044]; Saqlain[0.044]; Pakistan[0.044]; Pakistan[0.044]; Waqar[0.043]; Malik[0.043]; Sohail[0.042]; Khan[0.042]; Moin[0.041]; Younis[0.041]; Rehman[0.041]; Saeed[0.041]; England[0.041]; England[0.041]; matches[0.041]; Inzamam[0.041]; Ahmed[0.040]; Ahmed[0.040]; wickets[0.040]; Hollioake[0.040]; captain[0.039]; captain[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEdgbaston\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEdgbaston Cricket Ground\u001b[39m <---> \u001b[32mEdgbaston\u001b[39m\t\nSCORES: global= 0.274:0.247[\u001b[31m0.027\u001b[39m]; local()= 0.166:0.131[\u001b[31m0.035\u001b[39m]; log p(e|m)= -0.451:-0.309[\u001b[32m0.142\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; cricket[0.043]; batsman[0.043]; second[0.042]; Cricket[0.041]; Hollioake[0.041]; Hollioake[0.041]; toss[0.041]; overs[0.041]; Martin[0.040]; Birmingham[0.040]; Gough[0.040]; Lancashire[0.040]; Surrey[0.040]; Croft[0.040]; day[0.040]; Atherton[0.040]; bat[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Adam Hollioake\u001b[39m ==> ENTITY: \u001b[32mAdam Hollioake\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; cricket[0.043]; rounder[0.043]; won[0.043]; Darren[0.043]; Hollioake[0.042]; England[0.042]; England[0.042]; England[0.042]; England[0.042]; Teams[0.042]; Alan[0.041]; Martin[0.041]; day[0.041]; batsman[0.041]; Mike[0.041]; Graham[0.041]; Graham[0.041]; second[0.041]; captain[0.041]; captain[0.041]; Surrey[0.041]; Lancashire[0.041]; Matthew[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Moin Khan\u001b[39m ==> ENTITY: \u001b[32mMoin Khan\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaqlain[0.046]; Akram[0.046]; Waqar[0.046]; Wasim[0.046]; Pakistan[0.045]; Pakistan[0.045]; Mushtaq[0.045]; Mushtaq[0.045]; Ijaz[0.044]; captain[0.044]; captain[0.044]; Malik[0.043]; Younis[0.042]; matches[0.042]; wickets[0.042]; Hollioake[0.042]; England[0.042]; England[0.042]; Haq[0.042]; Inzamam[0.042]; Sohail[0.041]; Rehman[0.041]; kept[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Birmingham ==> ENTITY: \u001b[32mBirmingham\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.402:-0.402[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; cricket[0.044]; batsman[0.043]; Edgbaston[0.043]; second[0.042]; matches[0.042]; Cricket[0.041]; Hollioake[0.041]; overs[0.041]; Lancashire[0.041]; Atherton[0.040]; Martin[0.040]; bat[0.040]; toss[0.040]; wickets[0.039]; rounder[0.039]; Trafford[0.039]; day[0.039]; Surrey[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saeed Anwar\u001b[39m ==> ENTITY: \u001b[32mSaeed Anwar\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.230:0.230[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.045]; Pakistan[0.045]; Wasim[0.045]; Saqlain[0.044]; Akram[0.043]; Waqar[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; batsman[0.042]; Ijaz[0.042]; Younis[0.041]; Moin[0.041]; Malik[0.041]; Inzamam[0.041]; Sohail[0.041]; Rehman[0.040]; matches[0.040]; Ahmed[0.040]; Ahmed[0.040]; seamer[0.040]; Khan[0.040]; England[0.040]; England[0.040]; England[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alec Stewart\u001b[39m ==> ENTITY: \u001b[32mAlec Stewart\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.044]; England[0.043]; England[0.043]; England[0.043]; England[0.043]; England[0.043]; England[0.043]; rounder[0.042]; Surrey[0.042]; bat[0.042]; Gough[0.042]; cricket[0.042]; Akram[0.041]; day[0.041]; matches[0.041]; toss[0.041]; Edgbaston[0.041]; Atherton[0.041]; seamer[0.041]; Wasim[0.041]; Hollioake[0.041]; Hollioake[0.041]; international[0.040]; Croft[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Peter Martin ==> ENTITY: \u001b[32mPeter Martin (cricketer)\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.176:0.176[0]; log p(e|m)= -1.231:-1.231[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.046]; Dean[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; Alan[0.044]; Robert[0.044]; cricket[0.044]; seamer[0.044]; rounder[0.044]; Nick[0.043]; Cricket[0.043]; matches[0.043]; Matthew[0.042]; Surrey[0.042]; Mike[0.042]; bat[0.042]; Lancashire[0.042]; Adam[0.042]; Adam[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Surrey ==> ENTITY: \u001b[32mSurrey County Cricket Club\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.187:0.187[0]; log p(e|m)= -1.796:-1.796[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; Lancashire[0.044]; batsman[0.043]; cricket[0.043]; matches[0.041]; Atherton[0.041]; Edgbaston[0.041]; Hollioake[0.041]; Hollioake[0.041]; rounder[0.040]; Cricket[0.040]; Gough[0.040]; Birmingham[0.040]; seamer[0.040]; Croft[0.040]; second[0.040]; bat[0.040]; overs[0.039]; wickets[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dean Headley\u001b[39m ==> ENTITY: \u001b[32mDean Headley\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; England[0.047]; England[0.047]; Lancashire[0.044]; batsman[0.043]; Surrey[0.043]; matches[0.042]; Atherton[0.042]; Croft[0.042]; Edgbaston[0.041]; seamer[0.041]; Thorpe[0.041]; Darren[0.041]; Saqlain[0.040]; Matthew[0.040]; Mushtaq[0.040]; Mushtaq[0.040]; Akram[0.040]; Gough[0.040]; captain[0.040]; captain[0.040]; Maynard[0.040]; Graham[0.040]; Graham[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland cricket team\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.181:0.181[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.046]; England[0.046]; England[0.046]; England[0.046]; England[0.046]; cricket[0.044]; batsman[0.043]; matches[0.042]; Cricket[0.041]; Lancashire[0.041]; Edgbaston[0.041]; Saqlain[0.040]; Birmingham[0.040]; Akram[0.040]; Hollioake[0.040]; Hollioake[0.040]; seamer[0.040]; Surrey[0.040]; Pakistan[0.040]; Pakistan[0.040]; Pakistan[0.040]; Pakistan[0.040]; Mushtaq[0.040]; Mushtaq[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saqlain Mushtaq\u001b[39m ==> ENTITY: \u001b[32mSaqlain Mushtaq\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWasim[0.045]; Akram[0.045]; Mushtaq[0.044]; Pakistan[0.044]; England[0.043]; England[0.043]; Hollioake[0.043]; matches[0.043]; Ijaz[0.043]; wickets[0.042]; Waqar[0.042]; Sohail[0.041]; Malik[0.041]; Inzamam[0.041]; Khan[0.041]; Younis[0.041]; Rehman[0.040]; Atherton[0.040]; day[0.040]; Moin[0.039]; Croft[0.039]; Gough[0.039]; Ahmed[0.039]; Ahmed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mushtaq Ahmed\u001b[39m ==> ENTITY: \u001b[32mMushtaq Ahmed\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.228:0.228[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMushtaq[0.048]; Akram[0.047]; Wasim[0.047]; Saqlain[0.047]; Hollioake[0.045]; Waqar[0.045]; Ijaz[0.045]; Malik[0.044]; Sohail[0.044]; Younis[0.043]; wickets[0.043]; England[0.043]; Pakistan[0.043]; Khan[0.043]; Inzamam[0.043]; Ahmed[0.042]; Saeed[0.042]; matches[0.042]; Rehman[0.042]; Atherton[0.041]; Moin[0.041]; Gough[0.040]; Haq[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ronnie Irani\u001b[39m ==> ENTITY: \u001b[32mRonnie Irani\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMushtaq[0.046]; Mushtaq[0.046]; batsman[0.045]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; Hollioake[0.044]; Hollioake[0.044]; Saqlain[0.044]; seamer[0.044]; Lancashire[0.044]; Akram[0.044]; Wasim[0.043]; cricket[0.042]; Darren[0.042]; rounder[0.042]; Younis[0.042]; Surrey[0.042]; captain[0.042]; captain[0.042]; Malik[0.042]; Ijaz[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland cricket team\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.184:0.184[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.046]; England[0.046]; England[0.046]; England[0.046]; England[0.046]; cricket[0.044]; batsman[0.043]; matches[0.042]; Cricket[0.041]; Lancashire[0.041]; Edgbaston[0.041]; Saqlain[0.040]; Birmingham[0.040]; Akram[0.040]; Hollioake[0.040]; Hollioake[0.040]; seamer[0.040]; Surrey[0.040]; Pakistan[0.040]; Pakistan[0.040]; Pakistan[0.040]; Pakistan[0.040]; Mushtaq[0.040]; overs[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland cricket team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.176:0.176[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.046]; England[0.046]; England[0.046]; England[0.046]; England[0.046]; cricket[0.045]; batsman[0.043]; matches[0.042]; Cricket[0.041]; Lancashire[0.041]; Edgbaston[0.041]; Birmingham[0.040]; Hollioake[0.040]; Hollioake[0.040]; seamer[0.040]; Surrey[0.040]; Pakistan[0.040]; Pakistan[0.040]; Pakistan[0.040]; Pakistan[0.040]; overs[0.040]; Atherton[0.040]; Graham[0.040]; Graham[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ijaz Ahmed\u001b[39m ==> ENTITY: \u001b[32mIjaz Ahmed (cricketer)\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.224:0.224[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWasim[0.045]; Pakistan[0.044]; Pakistan[0.044]; Mushtaq[0.043]; Mushtaq[0.043]; Waqar[0.043]; Saqlain[0.043]; Khan[0.043]; Rehman[0.042]; Sohail[0.042]; Ahmed[0.042]; Akram[0.042]; Malik[0.042]; Saeed[0.041]; batsman[0.041]; Inzamam[0.041]; Younis[0.040]; Aamir[0.040]; Moin[0.040]; seamer[0.040]; matches[0.040]; England[0.039]; England[0.039]; day[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pakistan ==> ENTITY: \u001b[32mPakistan national cricket team\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.181:0.181[0]; log p(e|m)= -2.749:-2.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.049]; Pakistan[0.049]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; cricket[0.045]; batsman[0.043]; matches[0.043]; won[0.041]; Cricket[0.040]; Edgbaston[0.040]; seamer[0.040]; rounder[0.038]; Hollioake[0.038]; Lancashire[0.038]; overs[0.038]; bat[0.038]; Surrey[0.038]; toss[0.038]; day[0.038]; debut[0.037]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEngland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEngland cricket team\u001b[39m <---> \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.254:0.239[\u001b[31m0.015\u001b[39m]; local()= 0.182:0.122[\u001b[31m0.060\u001b[39m]; log p(e|m)= -3.297:-0.190[\u001b[32m3.107\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.046]; England[0.046]; England[0.046]; England[0.046]; England[0.046]; cricket[0.045]; batsman[0.043]; matches[0.042]; Cricket[0.041]; Lancashire[0.041]; Edgbaston[0.041]; Birmingham[0.040]; Hollioake[0.040]; seamer[0.040]; Surrey[0.040]; Pakistan[0.040]; Pakistan[0.040]; Pakistan[0.040]; overs[0.040]; Atherton[0.040]; Graham[0.040]; Graham[0.040]; rounder[0.040]; day[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Graham Thorpe\u001b[39m ==> ENTITY: \u001b[32mGraham Thorpe\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.213:0.213[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.045]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; batsman[0.044]; seamer[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Gough[0.042]; cricket[0.041]; Atherton[0.041]; Akram[0.041]; Saqlain[0.041]; rounder[0.041]; toss[0.040]; Martin[0.040]; Surrey[0.040]; Wasim[0.040]; wickets[0.040]; Hollioake[0.040]; Hollioake[0.040]; bat[0.040]; Edgbaston[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alan Mullally\u001b[39m ==> ENTITY: \u001b[32mAlan Mullally\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbatsman[0.045]; England[0.044]; England[0.044]; England[0.044]; Saqlain[0.043]; Lancashire[0.043]; Hollioake[0.042]; Hollioake[0.042]; Gough[0.042]; Atherton[0.042]; seamer[0.042]; Mushtaq[0.041]; Mushtaq[0.041]; debut[0.041]; Surrey[0.041]; wickets[0.041]; Croft[0.041]; Pakistan[0.040]; Pakistan[0.040]; Wasim[0.040]; Akram[0.040]; matches[0.040]; rounder[0.040]; Darren[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mike Atherton\u001b[39m ==> ENTITY: \u001b[32mMichael Atherton\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.048]; England[0.048]; England[0.048]; England[0.048]; England[0.048]; England[0.048]; cricket[0.047]; batsman[0.046]; Lancashire[0.045]; Cricket[0.045]; Saqlain[0.045]; Surrey[0.045]; seamer[0.044]; Akram[0.044]; Edgbaston[0.044]; matches[0.044]; Gough[0.044]; Wasim[0.044]; Alan[0.044]; bat[0.044]; Mushtaq[0.044]; Mushtaq[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Graham Lloyd\u001b[39m ==> ENTITY: \u001b[32mGraham Lloyd\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; Lancashire[0.042]; Graham[0.042]; seamer[0.042]; cricket[0.041]; matches[0.041]; Cricket[0.041]; Birmingham[0.041]; batsman[0.041]; rounder[0.041]; Alan[0.041]; Adam[0.041]; Adam[0.041]; Edgbaston[0.041]; bat[0.040]; Darren[0.040]; Surrey[0.040]; Croft[0.040]; captain[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Waqar Younis\u001b[39m ==> ENTITY: \u001b[32mWaqar Younis\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.229:0.229[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWasim[0.048]; Akram[0.047]; Saqlain[0.045]; Pakistan[0.045]; Mushtaq[0.044]; Mushtaq[0.044]; Malik[0.044]; Ijaz[0.044]; Moin[0.042]; Khan[0.042]; matches[0.042]; Sohail[0.042]; Inzamam[0.042]; Rehman[0.040]; Hollioake[0.040]; Saeed[0.040]; Ahmed[0.040]; Ahmed[0.040]; England[0.039]; captain[0.039]; captain[0.039]; Aamir[0.038]; Darren[0.037]; day[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ata-ur-Rehman\u001b[39m ==> ENTITY: \u001b[32mAta-ur-Rehman\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAkram[0.045]; Ijaz[0.045]; Pakistan[0.044]; Wasim[0.044]; Saqlain[0.043]; Younis[0.043]; Malik[0.043]; Saeed[0.042]; Khan[0.042]; Sohail[0.042]; Waqar[0.042]; Mushtaq[0.041]; Mushtaq[0.041]; Inzamam[0.041]; Ahmed[0.041]; Ahmed[0.041]; Moin[0.041]; Headley[0.041]; Hollioake[0.041]; Rehman[0.040]; England[0.040]; Haq[0.040]; matches[0.039]; Anwar[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Robert Croft\u001b[39m ==> ENTITY: \u001b[32mRobert Croft\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHollioake[0.044]; Hollioake[0.044]; batsman[0.044]; cricket[0.043]; England[0.043]; England[0.043]; England[0.043]; Lancashire[0.043]; overs[0.043]; seamer[0.042]; Edgbaston[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Gough[0.042]; Saqlain[0.042]; wickets[0.041]; rounder[0.040]; Surrey[0.040]; day[0.040]; Wasim[0.040]; second[0.040]; Darren[0.039]; Akram[0.039]; Malik[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pakistan ==> ENTITY: \u001b[32mPakistan national cricket team\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.181:0.181[0]; log p(e|m)= -2.749:-2.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.049]; Pakistan[0.049]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; England[0.045]; cricket[0.045]; batsman[0.043]; matches[0.043]; won[0.041]; Cricket[0.040]; Edgbaston[0.040]; seamer[0.040]; rounder[0.038]; Hollioake[0.038]; Lancashire[0.038]; overs[0.038]; bat[0.038]; Surrey[0.038]; toss[0.038]; day[0.038]; debut[0.037]; \t\n\r [======>.......................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s928ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 740/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1071testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mINDYCAR\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAmerican Championship car racing\u001b[39m <---> \u001b[32mIndyCar Series\u001b[39m\t\nSCORES: global= 0.266:0.259[\u001b[31m0.007\u001b[39m]; local()= 0.223:0.209[\u001b[31m0.014\u001b[39m]; log p(e|m)= -0.119:-1.386[\u001b[31m1.268\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRacing[0.046]; race[0.045]; IndyCar[0.044]; qualifying[0.044]; Motor[0.043]; drivers[0.042]; Penske[0.042]; Ford[0.042]; Honda[0.042]; Honda[0.042]; Vasser[0.042]; driver[0.042]; Rahal[0.042]; Qualifiers[0.041]; lap[0.041]; Andretti[0.041]; Race[0.041]; Tracy[0.040]; Sunday[0.040]; Bobby[0.040]; Herta[0.040]; Bryan[0.040]; Canada[0.040]; Saturday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mauricio Gugelmin\u001b[39m ==> ENTITY: \u001b[32mMaurício Gugelmin\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.242:0.242[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZanardi[0.044]; lap[0.043]; driver[0.043]; Honda[0.043]; Honda[0.043]; Honda[0.043]; Honda[0.043]; motor[0.042]; Mercedes[0.042]; Mercedes[0.042]; Mercedes[0.042]; Mercedes[0.042]; Penske[0.041]; Penske[0.041]; Vasser[0.041]; Andretti[0.041]; Michael[0.041]; Ford[0.041]; Ford[0.041]; Cosworth[0.040]; Cosworth[0.040]; Alex[0.040]; chassis[0.040]; Rahal[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jimmy Vasser\u001b[39m ==> ENTITY: \u001b[32mJimmy Vasser\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.235:0.235[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndycar[0.044]; drivers[0.043]; Racing[0.043]; Penske[0.043]; Penske[0.043]; Rahal[0.043]; Tracy[0.042]; IndyCar[0.042]; driver[0.042]; race[0.041]; Honda[0.041]; Honda[0.041]; Honda[0.041]; Honda[0.041]; Andretti[0.041]; lap[0.041]; Herta[0.041]; Zanardi[0.041]; Michael[0.041]; qualifying[0.041]; Unser[0.041]; Qualifiers[0.041]; Ford[0.040]; Ford[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.072:0.072[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.048]; final[0.046]; Qualifiers[0.046]; qualifying[0.045]; country[0.045]; Vancouver[0.044]; Vancouver[0.044]; Vancouver[0.044]; Bryan[0.043]; Michael[0.041]; Saturday[0.040]; Brazil[0.040]; Brazil[0.040]; Brazil[0.040]; times[0.040]; Bobby[0.040]; Italy[0.040]; Sunday[0.040]; seconds[0.039]; Paul[0.039]; Gil[0.039]; Alex[0.039]; Ford[0.039]; Ford[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alex Zanardi\u001b[39m ==> ENTITY: \u001b[32mAlex Zanardi\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.253:0.253[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrace[0.045]; Indycar[0.044]; qualifying[0.044]; Racing[0.044]; Honda[0.044]; Honda[0.044]; Honda[0.044]; lap[0.043]; Ford[0.042]; Ford[0.042]; drivers[0.041]; driver[0.041]; Mercedes[0.041]; Mercedes[0.041]; Mercedes[0.041]; Mercedes[0.041]; Vasser[0.041]; Qualifiers[0.041]; Race[0.040]; Michael[0.040]; IndyCar[0.040]; Lola[0.040]; Lola[0.040]; Mauricio[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfinal[0.047]; qualifying[0.046]; country[0.046]; Canada[0.045]; Bryan[0.044]; times[0.043]; Michael[0.042]; Jimmy[0.042]; Bobby[0.041]; Brazil[0.041]; Brazil[0.041]; Brazil[0.041]; Gil[0.041]; Italy[0.041]; Tracy[0.041]; Saturday[0.040]; seconds[0.040]; Paul[0.040]; Sunday[0.040]; lap[0.040]; Alex[0.040]; Ford[0.040]; Ford[0.040]; race[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paul Tracy\u001b[39m ==> ENTITY: \u001b[32mPaul Tracy\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.236:0.236[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrace[0.051]; lap[0.050]; drivers[0.050]; Indycar[0.049]; Racing[0.049]; qualifying[0.049]; Honda[0.048]; Honda[0.048]; Honda[0.048]; Honda[0.048]; Penske[0.048]; Penske[0.048]; IndyCar[0.048]; Vasser[0.047]; driver[0.047]; Race[0.046]; Qualifiers[0.046]; Ford[0.045]; Ford[0.045]; Rahal[0.045]; Herta[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.072:0.072[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.048]; final[0.046]; Qualifiers[0.046]; qualifying[0.045]; country[0.045]; Vancouver[0.044]; Vancouver[0.044]; Vancouver[0.044]; Bryan[0.043]; Michael[0.041]; Saturday[0.040]; Brazil[0.040]; Brazil[0.040]; Brazil[0.040]; times[0.040]; Bobby[0.040]; Italy[0.040]; Sunday[0.040]; seconds[0.039]; Paul[0.039]; Gil[0.039]; Alex[0.039]; Ford[0.039]; Ford[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nqualifying[0.050]; times[0.050]; country[0.049]; Qualifiers[0.049]; final[0.048]; Brazil[0.048]; Brazil[0.048]; Brazil[0.048]; Canada[0.047]; Bryan[0.045]; Mauricio[0.045]; Saturday[0.044]; Sunday[0.044]; race[0.044]; Bobby[0.043]; Andre[0.043]; Gil[0.043]; Michael[0.042]; Ferran[0.042]; Vancouver[0.042]; Vancouver[0.042]; Vancouver[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazil ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.050]; Brazil[0.050]; country[0.046]; Italy[0.045]; Mauricio[0.045]; Bryan[0.044]; Canada[0.043]; Gil[0.042]; Alex[0.042]; Ribeiro[0.041]; Bobby[0.040]; Mercedes[0.040]; Mercedes[0.040]; Mercedes[0.040]; Mercedes[0.040]; Michael[0.040]; Jr[0.040]; times[0.040]; Andre[0.039]; Ferran[0.039]; Jimmy[0.039]; seconds[0.039]; Tracy[0.038]; Paul[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vancouver ==> ENTITY: \u001b[32mVancouver\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.094:0.094[0]; log p(e|m)= -0.121:-0.121[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVancouver[0.050]; Vancouver[0.050]; Canada[0.049]; Bryan[0.043]; qualifying[0.043]; race[0.043]; Race[0.042]; country[0.042]; Indycar[0.041]; Saturday[0.041]; Sunday[0.041]; Racing[0.040]; lap[0.040]; Qualifiers[0.040]; tabulate[0.040]; final[0.040]; Vasser[0.040]; IndyCar[0.040]; driver[0.040]; Bobby[0.039]; Andretti[0.039]; Paul[0.039]; Tracy[0.039]; Michael[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andre Ribeiro\u001b[39m ==> ENTITY: \u001b[32mAndré Ribeiro\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.279:0.279[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPenske[0.050]; Penske[0.050]; Vasser[0.049]; Brazil[0.049]; Brazil[0.049]; Brazil[0.049]; Andretti[0.048]; chassis[0.048]; race[0.048]; Rahal[0.047]; qualifying[0.047]; Honda[0.047]; Honda[0.047]; Honda[0.047]; Honda[0.047]; Canada[0.047]; driver[0.047]; Ford[0.047]; Ford[0.047]; Cosworth[0.045]; Cosworth[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.071:0.071[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfinal[0.047]; Qualifiers[0.047]; qualifying[0.046]; country[0.046]; Canada[0.044]; Bryan[0.044]; times[0.043]; Michael[0.042]; Jimmy[0.041]; Bobby[0.041]; Brazil[0.041]; Brazil[0.041]; Brazil[0.041]; Gil[0.041]; Italy[0.040]; Tracy[0.040]; Saturday[0.040]; seconds[0.040]; Paul[0.040]; Sunday[0.039]; lap[0.039]; Alex[0.039]; Ford[0.039]; Ford[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Michael Andretti\u001b[39m ==> ENTITY: \u001b[32mMichael Andretti\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.230:0.230[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrace[0.044]; drivers[0.043]; lap[0.043]; driver[0.043]; qualifying[0.043]; Indycar[0.042]; Racing[0.042]; Rahal[0.042]; Vasser[0.042]; Honda[0.042]; Honda[0.042]; Honda[0.042]; Honda[0.042]; Penske[0.042]; Penske[0.042]; IndyCar[0.041]; Qualifiers[0.041]; Ford[0.040]; Ford[0.040]; Mercedes[0.040]; Mercedes[0.040]; Mercedes[0.040]; Mercedes[0.040]; Herta[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.075:0.075[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrace[0.046]; qualifying[0.044]; Vancouver[0.044]; Vancouver[0.044]; Vancouver[0.044]; country[0.043]; Qualifiers[0.043]; Race[0.042]; Indycar[0.042]; Brazil[0.042]; Brazil[0.042]; Brazil[0.042]; lap[0.041]; driver[0.041]; Ford[0.041]; Ford[0.041]; final[0.040]; Penske[0.040]; Penske[0.040]; Italy[0.040]; Bryan[0.040]; drivers[0.040]; Saturday[0.040]; IndyCar[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gil de Ferran\u001b[39m ==> ENTITY: \u001b[32mGil de Ferran\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.209:0.209[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHonda[0.044]; Honda[0.044]; Honda[0.044]; Honda[0.044]; Vasser[0.042]; Herta[0.042]; Andretti[0.042]; Lola[0.042]; Lola[0.042]; Zanardi[0.042]; Penske[0.041]; Penske[0.041]; Rahal[0.041]; Brazil[0.041]; Brazil[0.041]; Brazil[0.041]; Mauricio[0.041]; Mercedes[0.041]; Mercedes[0.041]; Mercedes[0.041]; Mercedes[0.041]; Ford[0.041]; Ford[0.041]; seconds[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.071:0.071[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfinal[0.047]; Qualifiers[0.047]; qualifying[0.046]; country[0.046]; Canada[0.044]; Bryan[0.044]; times[0.043]; Michael[0.042]; Jimmy[0.041]; Bobby[0.041]; Brazil[0.041]; Brazil[0.041]; Brazil[0.041]; Gil[0.041]; Italy[0.040]; Tracy[0.040]; Saturday[0.040]; seconds[0.040]; Paul[0.040]; Sunday[0.039]; lap[0.039]; Alex[0.039]; Ford[0.039]; Ford[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bobby Rahal\u001b[39m ==> ENTITY: \u001b[32mBobby Rahal\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.215:0.215[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrace[0.044]; qualifying[0.044]; driver[0.043]; lap[0.043]; Vasser[0.043]; drivers[0.043]; Indycar[0.042]; Penske[0.042]; Penske[0.042]; Honda[0.042]; Honda[0.042]; Honda[0.042]; Honda[0.042]; Andretti[0.042]; Racing[0.041]; Tracy[0.041]; Qualifiers[0.041]; Race[0.041]; IndyCar[0.041]; final[0.040]; Zanardi[0.040]; Herta[0.040]; Michael[0.040]; Unser[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vancouver ==> ENTITY: \u001b[32mVancouver\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.094:0.094[0]; log p(e|m)= -0.121:-0.121[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVancouver[0.050]; Vancouver[0.050]; Canada[0.049]; Bryan[0.043]; qualifying[0.043]; race[0.043]; Race[0.042]; country[0.042]; Indycar[0.041]; Saturday[0.041]; Sunday[0.041]; Racing[0.040]; lap[0.040]; Qualifiers[0.040]; tabulate[0.040]; final[0.040]; Vasser[0.040]; IndyCar[0.040]; driver[0.040]; Bobby[0.039]; Andretti[0.039]; Paul[0.039]; Tracy[0.039]; Michael[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vancouver ==> ENTITY: \u001b[32mVancouver\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.094:0.094[0]; log p(e|m)= -0.121:-0.121[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVancouver[0.050]; Vancouver[0.050]; Canada[0.049]; Bryan[0.043]; qualifying[0.043]; race[0.043]; Race[0.042]; country[0.042]; Indycar[0.041]; Saturday[0.041]; Sunday[0.041]; Racing[0.040]; lap[0.040]; Qualifiers[0.040]; tabulate[0.040]; final[0.040]; Vasser[0.040]; IndyCar[0.040]; driver[0.040]; Bobby[0.039]; Andretti[0.039]; Paul[0.039]; Tracy[0.039]; Michael[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazil ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.049]; Brazil[0.049]; qualifying[0.046]; country[0.045]; Italy[0.045]; Mauricio[0.045]; Bryan[0.043]; Canada[0.042]; Gil[0.042]; Alex[0.042]; Ribeiro[0.040]; Saturday[0.040]; Bobby[0.040]; Mercedes[0.039]; Mercedes[0.039]; Mercedes[0.039]; Mercedes[0.039]; Michael[0.039]; Jr[0.039]; times[0.039]; Andre[0.039]; Ferran[0.039]; Jimmy[0.039]; seconds[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAl Unser\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAl Unser\u001b[39m <---> \u001b[32mAl Unser, Jr.\u001b[39m\t\nSCORES: global= 0.290:0.283[\u001b[31m0.008\u001b[39m]; local()= 0.201:0.195[\u001b[31m0.006\u001b[39m]; log p(e|m)= 0.000:-1.100[\u001b[31m1.100\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndrivers[0.046]; race[0.046]; Vasser[0.045]; driver[0.045]; qualifying[0.045]; lap[0.045]; Rahal[0.044]; Penske[0.044]; Penske[0.044]; IndyCar[0.044]; Ford[0.043]; Ford[0.043]; Gil[0.043]; Bobby[0.043]; Mauricio[0.043]; Andretti[0.043]; Bryan[0.043]; Tracy[0.043]; Michael[0.042]; Jimmy[0.042]; Herta[0.042]; final[0.041]; Alex[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bryan Herta\u001b[39m ==> ENTITY: \u001b[32mBryan Herta\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.244:0.244[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrace[0.048]; Racing[0.047]; Indycar[0.045]; drivers[0.045]; Race[0.044]; Honda[0.044]; Honda[0.044]; Honda[0.044]; Honda[0.044]; Penske[0.044]; Penske[0.044]; qualifying[0.044]; driver[0.044]; IndyCar[0.043]; lap[0.043]; Rahal[0.043]; Vasser[0.042]; Qualifiers[0.042]; Andretti[0.042]; Ford[0.042]; Ford[0.042]; Sunday[0.041]; Saturday[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazil ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.055]; Brazil[0.055]; Italy[0.050]; Mauricio[0.050]; Bryan[0.048]; Canada[0.047]; Gil[0.046]; Alex[0.046]; Ribeiro[0.044]; Bobby[0.044]; Mercedes[0.044]; Mercedes[0.044]; Mercedes[0.044]; Mercedes[0.044]; Michael[0.044]; Jr[0.043]; Andre[0.043]; Ferran[0.043]; Jimmy[0.043]; Tracy[0.042]; Paul[0.042]; Al[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIndyCar\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAmerican Championship car racing\u001b[39m <---> \u001b[32mIndyCar Series\u001b[39m\t\nSCORES: global= 0.266:0.259[\u001b[31m0.007\u001b[39m]; local()= 0.229:0.220[\u001b[31m0.009\u001b[39m]; log p(e|m)= -0.119:-1.386[\u001b[31m1.268\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndycar[0.046]; Racing[0.046]; race[0.044]; qualifying[0.044]; Motor[0.042]; drivers[0.042]; Penske[0.042]; Penske[0.042]; Ford[0.042]; Honda[0.042]; Honda[0.042]; Vasser[0.042]; driver[0.041]; Rahal[0.041]; Unser[0.041]; Qualifiers[0.041]; lap[0.040]; Andretti[0.040]; Race[0.040]; Tracy[0.040]; Sunday[0.040]; Bobby[0.040]; Herta[0.040]; Bryan[0.040]; \t\n\r [======>.......................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s882ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 765/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 963testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wisconsin ==> ENTITY: \u001b[32mWisconsin\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.061:0.061[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayers[0.046]; round[0.043]; Thursday[0.043]; Mike[0.043]; Milwaukee[0.043]; Milwaukee[0.043]; Milwaukee[0.043]; Dave[0.042]; Brown[0.042]; scores[0.042]; Canada[0.042]; Brian[0.042]; Neal[0.041]; Park[0.041]; Sullivan[0.041]; Golf[0.041]; Golf[0.041]; Nolan[0.041]; Steve[0.041]; Bob[0.041]; Payne[0.040]; Mark[0.040]; Billy[0.040]; Willie[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Billy Mayfair\u001b[39m ==> ENTITY: \u001b[32mBilly Mayfair\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBryan[0.044]; Golf[0.043]; Brian[0.043]; Brian[0.043]; round[0.043]; Bob[0.043]; Billy[0.042]; Steve[0.042]; Steve[0.042]; Ken[0.042]; Mike[0.042]; Mike[0.042]; Estes[0.041]; Mark[0.041]; Parnevik[0.041]; Tim[0.041]; Open[0.041]; Dave[0.041]; Woods[0.041]; Willie[0.041]; Jerry[0.041]; players[0.040]; Course[0.040]; Calcavecchia[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.069:0.069[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayers[0.046]; Steve[0.046]; Steve[0.046]; Bryan[0.046]; Mark[0.046]; Sweden[0.045]; Brian[0.044]; Brian[0.044]; Mike[0.044]; Mike[0.044]; Dave[0.043]; Neal[0.043]; Thursday[0.043]; Park[0.043]; Bob[0.043]; Sullivan[0.042]; round[0.042]; Nolan[0.042]; Tim[0.042]; scores[0.042]; Russ[0.042]; Brown[0.042]; Ken[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Duffy Waldorf\u001b[39m ==> ENTITY: \u001b[32mDuffy Waldorf\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.116:0.116[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.046]; Golf[0.046]; round[0.045]; Open[0.045]; Open[0.045]; players[0.045]; Woods[0.043]; Bryan[0.042]; Parnevik[0.041]; Course[0.041]; Estes[0.041]; Kelly[0.041]; Mark[0.041]; Wisconsin[0.040]; Stricker[0.040]; Jesper[0.040]; Stewart[0.040]; scores[0.040]; Jurgenson[0.040]; Deer[0.040]; Olin[0.040]; Tim[0.040]; yard[0.040]; Sweden[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tiger Woods\u001b[39m ==> ENTITY: \u001b[32mTiger Woods\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayers[0.045]; Stricker[0.045]; Parnevik[0.044]; Calcavecchia[0.043]; Bryan[0.043]; Jerry[0.042]; Kelly[0.042]; Mark[0.042]; Sweden[0.042]; Canada[0.042]; Steve[0.041]; Steve[0.041]; Bob[0.041]; Brian[0.041]; Brian[0.041]; Tim[0.041]; Mayfair[0.041]; Neal[0.040]; Billy[0.040]; Billy[0.040]; Loren[0.040]; Mike[0.040]; Mike[0.040]; Dave[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nolan Henke\u001b[39m ==> ENTITY: \u001b[32mNolan Henke\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.046]; Golf[0.046]; Open[0.046]; Open[0.046]; round[0.043]; Estes[0.043]; Course[0.043]; Parnevik[0.043]; par[0.042]; players[0.041]; Bryan[0.041]; Woods[0.041]; Calcavecchia[0.041]; Bob[0.041]; Waldorf[0.040]; Stricker[0.040]; scores[0.040]; Green[0.040]; Park[0.040]; Mike[0.040]; Mike[0.040]; Wisconsin[0.040]; Payne[0.040]; Mayfair[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Russ Cochran\u001b[39m ==> ENTITY: \u001b[32mRuss Cochran\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.049]; Open[0.048]; Open[0.048]; round[0.046]; Bob[0.045]; Estes[0.044]; Bryan[0.044]; Tiger[0.043]; Stewart[0.043]; Course[0.043]; Woods[0.043]; Steve[0.043]; Steve[0.043]; Brian[0.042]; Brian[0.042]; Mike[0.042]; Mike[0.042]; Parnevik[0.042]; Calcavecchia[0.042]; Mark[0.042]; players[0.042]; Ken[0.042]; Dave[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Olin Browne\u001b[39m ==> ENTITY: \u001b[32mOlin Browne\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.110:0.110[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nround[0.048]; Golf[0.044]; Course[0.043]; Bryan[0.043]; Woods[0.042]; Deer[0.042]; Mayfair[0.042]; Stricker[0.042]; Waldorf[0.042]; Simpson[0.042]; Estes[0.042]; players[0.042]; Parnevik[0.042]; Calcavecchia[0.041]; Tiger[0.041]; Jesper[0.041]; Jurgenson[0.041]; Tim[0.041]; Sweden[0.040]; Thursday[0.040]; Henninger[0.040]; Wood[0.040]; Brian[0.040]; Brian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Willie Wood ==> ENTITY: \u001b[32mWillie Wood (golfer)\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.868:-0.868[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.045]; Golf[0.045]; round[0.044]; Billy[0.043]; Billy[0.043]; players[0.042]; Open[0.042]; Open[0.042]; Estes[0.042]; Green[0.042]; Canada[0.042]; Park[0.041]; Bryan[0.041]; Jerry[0.041]; Tiger[0.041]; Deer[0.041]; Bob[0.041]; Woods[0.041]; Calcavecchia[0.041]; yard[0.041]; Mike[0.040]; Mike[0.040]; Dave[0.040]; Ken[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.053:0.053[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayers[0.046]; Canada[0.045]; Golf[0.044]; Golf[0.044]; Bryan[0.043]; Jurgenson[0.043]; Park[0.043]; Jesper[0.042]; Open[0.042]; Open[0.042]; Estes[0.042]; round[0.041]; Dave[0.041]; Steve[0.041]; Steve[0.041]; Tim[0.040]; Mark[0.040]; Payne[0.040]; Brian[0.040]; Brian[0.040]; Billy[0.040]; Billy[0.040]; Bob[0.040]; Calcavecchia[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bob Estes\u001b[39m ==> ENTITY: \u001b[32mBob Estes\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCourse[0.045]; round[0.044]; Golf[0.043]; Golf[0.043]; Bryan[0.043]; Jerry[0.043]; Open[0.042]; Open[0.042]; par[0.042]; Billy[0.042]; Billy[0.042]; Dave[0.042]; Steve[0.042]; Steve[0.042]; Ken[0.041]; Mike[0.041]; Mike[0.041]; Brian[0.041]; Brian[0.041]; Parnevik[0.040]; Mark[0.040]; players[0.040]; Waldorf[0.040]; Woods[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ken Green\u001b[39m ==> ENTITY: \u001b[32mKen Green (golfer)\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.108:0.108[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nround[0.045]; Golf[0.044]; Steve[0.043]; Steve[0.043]; players[0.043]; Sweden[0.043]; Brian[0.042]; Brian[0.042]; Bryan[0.042]; Bob[0.042]; Tiger[0.042]; Mayfair[0.041]; Mark[0.041]; Open[0.041]; Park[0.041]; Deer[0.041]; Tim[0.041]; Thursday[0.041]; Woods[0.041]; Parnevik[0.041]; Waldorf[0.040]; Lancaster[0.040]; Willie[0.040]; Course[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Greater Milwaukee Open\u001b[39m ==> ENTITY: \u001b[32mU.S. Bank Championship in Milwaukee\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.047]; Golf[0.047]; Golf[0.047]; round[0.046]; Park[0.043]; Calcavecchia[0.042]; Estes[0.042]; players[0.042]; Thursday[0.042]; Mayfair[0.041]; Stricker[0.041]; Deer[0.041]; Course[0.041]; Parnevik[0.041]; Waldorf[0.040]; Greater[0.040]; Milwaukee[0.040]; Milwaukee[0.040]; Wisconsin[0.040]; million[0.040]; par[0.040]; Wood[0.040]; Jesper[0.039]; Payne[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mark Calcavecchia\u001b[39m ==> ENTITY: \u001b[32mMark Calcavecchia\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.209:0.209[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.045]; Open[0.045]; Woods[0.045]; round[0.044]; Tiger[0.042]; Parnevik[0.042]; Brian[0.042]; Brian[0.042]; Bob[0.042]; par[0.042]; Course[0.041]; Jerry[0.041]; Stricker[0.041]; Golf[0.041]; players[0.041]; Estes[0.041]; Mayfair[0.041]; Ken[0.041]; Steve[0.041]; Steve[0.041]; Tim[0.040]; Bryan[0.040]; Mike[0.040]; Mike[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dave Barr ==> ENTITY: \u001b[32mDave Barr (golfer)\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.138:0.138[0]; log p(e|m)= -0.300:-0.300[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBryan[0.044]; Golf[0.043]; Golf[0.043]; round[0.043]; Canada[0.043]; players[0.042]; Mike[0.042]; Mike[0.042]; Mark[0.042]; Steve[0.042]; Steve[0.042]; Bob[0.042]; Brian[0.042]; Brian[0.042]; Open[0.041]; Open[0.041]; Tim[0.041]; Lancaster[0.041]; Jerry[0.040]; Green[0.040]; Ken[0.040]; Billy[0.040]; Billy[0.040]; scores[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Billy Andrade\u001b[39m ==> ENTITY: \u001b[32mBilly Andrade\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.046]; Golf[0.046]; Bryan[0.044]; round[0.043]; Billy[0.043]; players[0.042]; Estes[0.042]; Brian[0.042]; Brian[0.042]; Open[0.042]; Open[0.042]; Bob[0.041]; Woods[0.041]; Mark[0.041]; Tiger[0.041]; Neal[0.041]; Mike[0.040]; Mike[0.040]; Dave[0.040]; Parnevik[0.040]; Calcavecchia[0.040]; Payne[0.040]; Green[0.040]; Sweden[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mike Hulbert\u001b[39m ==> ENTITY: \u001b[32mMike Hulbert\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.135:0.135[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nround[0.049]; Golf[0.048]; players[0.045]; Course[0.044]; Jerry[0.044]; Calcavecchia[0.044]; Parnevik[0.043]; Woods[0.043]; Mayfair[0.043]; Mike[0.043]; Stricker[0.043]; Mark[0.043]; Bryan[0.043]; Bob[0.043]; Billy[0.043]; Billy[0.043]; Steve[0.043]; Steve[0.043]; Willie[0.043]; Tiger[0.042]; Estes[0.042]; Waldorf[0.042]; Dave[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Loren Roberts\u001b[39m ==> ENTITY: \u001b[32mLoren Roberts\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.047]; Open[0.047]; Golf[0.047]; Golf[0.047]; round[0.045]; Woods[0.043]; Course[0.042]; Parnevik[0.042]; Stricker[0.042]; Bryan[0.042]; Tiger[0.041]; par[0.041]; Calcavecchia[0.041]; scores[0.041]; Mayfair[0.040]; Estes[0.040]; Green[0.040]; Jerry[0.040]; Waldorf[0.039]; players[0.039]; Payne[0.039]; Mark[0.039]; Brian[0.039]; Brian[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jerry Kelly\u001b[39m ==> ENTITY: \u001b[32mJerry Kelly\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nround[0.047]; players[0.045]; Open[0.045]; Golf[0.044]; Stricker[0.043]; par[0.043]; Bryan[0.042]; Woods[0.042]; Calcavecchia[0.041]; Tiger[0.041]; Waldorf[0.041]; Deer[0.041]; Course[0.041]; Estes[0.041]; Parnevik[0.041]; Brian[0.041]; Brian[0.041]; Thursday[0.041]; Mayfair[0.040]; Sweden[0.040]; Milwaukee[0.040]; Tim[0.040]; Mark[0.040]; Willie[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mike Sullivan ==> ENTITY: \u001b[32mMike Sullivan (golfer)\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.134:0.134[0]; log p(e|m)= -1.143:-1.143[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBryan[0.046]; Thursday[0.045]; players[0.045]; Billy[0.044]; Billy[0.044]; Brian[0.044]; Brian[0.044]; Mike[0.044]; Neal[0.044]; Steve[0.044]; Steve[0.044]; Golf[0.044]; Golf[0.044]; Shane[0.044]; Mark[0.043]; Bob[0.043]; Tim[0.043]; round[0.042]; Dave[0.042]; Jerry[0.042]; Brown[0.042]; Duffy[0.042]; Woods[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jesper Parnevik\u001b[39m ==> ENTITY: \u001b[32mJesper Parnevik\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.046]; Golf[0.046]; Open[0.046]; Open[0.046]; round[0.043]; Woods[0.043]; Calcavecchia[0.042]; Mark[0.042]; Sweden[0.041]; Jerry[0.041]; Bryan[0.041]; Park[0.041]; Bob[0.041]; Mayfair[0.041]; Dave[0.041]; players[0.040]; Stricker[0.040]; Tiger[0.040]; Willie[0.040]; Steve[0.040]; Steve[0.040]; Tim[0.040]; Billy[0.040]; Billy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Milwaukee ==> ENTITY: \u001b[32mMilwaukee\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.113:0.113[0]; log p(e|m)= -0.282:-0.282[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayers[0.044]; Mark[0.044]; Park[0.044]; Milwaukee[0.044]; Milwaukee[0.044]; Thursday[0.044]; Neal[0.042]; Wisconsin[0.042]; Steve[0.042]; Mike[0.042]; round[0.042]; Brian[0.041]; Dave[0.041]; Russ[0.041]; million[0.041]; Willie[0.041]; Brown[0.041]; Billy[0.041]; Bob[0.040]; Wood[0.040]; Payne[0.040]; Greater[0.040]; Greater[0.040]; yard[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Brown Deer Park Golf Course\u001b[39m ==> ENTITY: \u001b[32mBrown Deer Park Golf Course\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.046]; Calcavecchia[0.044]; Stricker[0.043]; Estes[0.043]; Open[0.043]; Open[0.043]; Jerry[0.043]; round[0.042]; par[0.042]; Parnevik[0.042]; Ken[0.041]; Cochran[0.041]; Milwaukee[0.041]; Milwaukee[0.041]; Milwaukee[0.041]; Willie[0.041]; Waldorf[0.041]; Loren[0.041]; Thursday[0.040]; Simpson[0.040]; Bob[0.040]; Payne[0.040]; players[0.040]; Tim[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Payne Stewart\u001b[39m ==> ENTITY: \u001b[32mPayne Stewart\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.045]; Woods[0.043]; Thursday[0.043]; Calcavecchia[0.043]; Steve[0.042]; Steve[0.042]; Mark[0.042]; Jerry[0.042]; Bob[0.042]; round[0.042]; par[0.042]; Parnevik[0.042]; Brian[0.041]; Brian[0.041]; Cochran[0.041]; Course[0.041]; players[0.041]; Mike[0.041]; Mike[0.041]; Dave[0.041]; Billy[0.041]; Billy[0.041]; Tim[0.041]; Tiger[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Brian Henninger\u001b[39m ==> ENTITY: \u001b[32mBrian Henninger\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nround[0.045]; Bob[0.043]; Woods[0.043]; Ken[0.042]; Stricker[0.042]; Brian[0.042]; players[0.042]; Calcavecchia[0.042]; Mike[0.042]; Mike[0.042]; Tim[0.042]; Claar[0.042]; Steve[0.041]; Steve[0.041]; Parnevik[0.041]; Bryan[0.041]; Russ[0.041]; Dave[0.041]; Payne[0.041]; Mark[0.041]; Waldorf[0.040]; Estes[0.040]; Hulbert[0.040]; Jerry[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Steve Stricker\u001b[39m ==> ENTITY: \u001b[32mSteve Stricker\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.046]; Golf[0.046]; Open[0.045]; Open[0.045]; round[0.045]; Woods[0.043]; Mark[0.042]; Bryan[0.042]; par[0.041]; scores[0.041]; Calcavecchia[0.041]; players[0.041]; Brian[0.041]; Brian[0.041]; Steve[0.040]; Course[0.040]; Parnevik[0.040]; Tiger[0.040]; Tim[0.040]; Mayfair[0.040]; Thursday[0.040]; Mike[0.040]; Mike[0.040]; Scores[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Neal Lancaster\u001b[39m ==> ENTITY: \u001b[32mNeal Lancaster\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.048]; Open[0.048]; round[0.047]; Woods[0.046]; Golf[0.045]; Golf[0.045]; Bryan[0.045]; Calcavecchia[0.044]; Stricker[0.044]; Estes[0.044]; Parnevik[0.043]; par[0.043]; Course[0.042]; Brian[0.042]; Brian[0.042]; Deer[0.042]; Tiger[0.042]; Cochran[0.042]; Mark[0.041]; Andrade[0.041]; Mike[0.041]; Mike[0.041]; Willie[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tim Simpson\u001b[39m ==> ENTITY: \u001b[32mTim Simpson\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.047]; round[0.045]; players[0.043]; Parnevik[0.042]; Billy[0.042]; Billy[0.042]; Jerry[0.042]; Tiger[0.042]; Bryan[0.041]; Brian[0.041]; Brian[0.041]; Bob[0.041]; Willie[0.041]; Woods[0.041]; Cochran[0.041]; Dave[0.041]; Mike[0.041]; Mike[0.041]; Steve[0.041]; Steve[0.041]; Waldorf[0.041]; Sweden[0.041]; Course[0.040]; Payne[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.066:0.066[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.048]; players[0.047]; Sweden[0.044]; Golf[0.044]; Golf[0.044]; scores[0.041]; Sullivan[0.041]; round[0.041]; Brian[0.041]; Brian[0.041]; Dave[0.041]; Brown[0.041]; million[0.041]; Steve[0.041]; Park[0.041]; Kelly[0.041]; Bob[0.040]; Mark[0.040]; Jerry[0.040]; Stewart[0.040]; Open[0.040]; Open[0.040]; Mike[0.040]; Mike[0.040]; \t\n\r [======>.......................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s807ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 794/4791 \r\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.219:0.219[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israeli[0.045]; Israeli[0.045]; Jewish[0.045]; Israelis[0.044]; Palestinian[0.043]; Jerusalem[0.042]; Jerusalem[0.042]; Jerusalem[0.042]; Jerusalem[0.042]; Palestinians[0.042]; Palestinians[0.042]; Arab[0.042]; Jordan[0.042]; Jordan[0.042]; West[0.041]; Bank[0.040]; Abu[0.039]; holy[0.039]; pray[0.038]; pray[0.038]; pray[0.038]; peace[0.037]; peace[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.280:0.280[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.048]; Israel[0.048]; Israel[0.048]; Israeli[0.043]; Israeli[0.043]; Jewish[0.043]; Jewish[0.043]; Hebrew[0.043]; Palestinian[0.041]; Palestinian[0.041]; Palestinian[0.041]; Jerusalem[0.040]; Jerusalem[0.040]; Jerusalem[0.040]; Jerusalem[0.040]; settlements[0.040]; Palestinians[0.040]; Palestinians[0.040]; Arab[0.040]; Gaza[0.040]; Jordan[0.040]; Arabs[0.040]; West[0.039]; West[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = al-Ram\u001b[39m ==> ENTITY: \u001b[32mAl-Ram\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPalestinian[0.044]; Palestinians[0.043]; Palestinians[0.043]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; Israel[0.043]; West[0.043]; Abu[0.043]; Israelis[0.043]; Bank[0.042]; Jerusalem[0.041]; Jerusalem[0.041]; Jerusalem[0.041]; Jerusalem[0.041]; Jamil[0.041]; Ram[0.041]; Mustafa[0.041]; al[0.040]; al[0.040]; Arab[0.040]; Aqsa[0.039]; Aqsa[0.039]; security[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jerusalem ==> ENTITY: \u001b[32mJerusalem\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.262:0.262[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJerusalem[0.048]; Jerusalem[0.048]; Jerusalem[0.048]; Israel[0.048]; Israel[0.048]; Israel[0.048]; Israel[0.048]; Jewish[0.047]; Jewish[0.047]; Hebrew[0.046]; mosque[0.045]; Israeli[0.044]; Israeli[0.044]; Israeli[0.044]; Moslem[0.044]; Arabs[0.044]; prayer[0.043]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; Arab[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hebrew ==> ENTITY: \u001b[32mHebrew language\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.208:0.208[0]; log p(e|m)= -0.080:-0.080[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJewish[0.050]; Jewish[0.050]; Jerusalem[0.047]; Jerusalem[0.047]; Jerusalem[0.047]; Jerusalem[0.047]; Israel[0.046]; Israel[0.046]; Israel[0.046]; Arabs[0.046]; worship[0.045]; Israeli[0.045]; Israeli[0.045]; Israeli[0.045]; holy[0.044]; prayer[0.044]; Pilgrims[0.043]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; Moslem[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = al-Aqsa\u001b[39m ==> ENTITY: \u001b[32mAl-Aqsa Mosque\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJerusalem[0.044]; Jerusalem[0.044]; Jerusalem[0.044]; Jerusalem[0.044]; Jerusalem[0.044]; Abu[0.043]; al[0.042]; al[0.042]; Israel[0.042]; Israel[0.042]; Jamil[0.041]; Arab[0.041]; Jordan[0.041]; Jordan[0.041]; Palestinian[0.041]; Hijazi[0.041]; Aqsa[0.040]; Aqsa[0.040]; Palestinians[0.040]; Palestinians[0.040]; Palestinians[0.040]; Arafat[0.040]; Moslem[0.040]; West[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jerusalem ==> ENTITY: \u001b[32mJerusalem\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.233:0.233[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJerusalem[0.045]; Jerusalem[0.045]; Israel[0.045]; Israel[0.045]; Israel[0.045]; Jewish[0.044]; Hebrew[0.044]; mosque[0.042]; Israeli[0.042]; Israeli[0.042]; Moslem[0.041]; Arabs[0.041]; prayer[0.041]; Palestinian[0.040]; Palestinian[0.040]; Palestinian[0.040]; Palestinian[0.040]; West[0.040]; West[0.040]; West[0.040]; holy[0.040]; Palestinians[0.040]; worship[0.040]; settlements[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jerusalem ==> ENTITY: \u001b[32mJerusalem\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.255:0.255[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJerusalem[0.044]; Jerusalem[0.044]; Jerusalem[0.044]; Jerusalem[0.044]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Jewish[0.044]; Jewish[0.044]; Hebrew[0.043]; mosque[0.041]; Israeli[0.041]; Israeli[0.041]; Moslem[0.041]; Arabs[0.040]; Palestinian[0.040]; Palestinian[0.040]; Palestinian[0.040]; Arab[0.040]; West[0.040]; West[0.040]; Abu[0.039]; Ammar[0.039]; holy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Abu Ammar\u001b[39m ==> ENTITY: \u001b[32mYasser Arafat\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsraeli[0.047]; Jordan[0.045]; Arafat[0.045]; Israelis[0.044]; Israelis[0.044]; Palestinians[0.044]; Jerusalem[0.043]; Jerusalem[0.043]; Khaled[0.043]; Mustafa[0.042]; Hijazi[0.042]; security[0.042]; Ahmed[0.041]; Ali[0.040]; West[0.040]; said[0.040]; said[0.040]; said[0.040]; peace[0.040]; peace[0.040]; peace[0.040]; Bank[0.039]; Qalandia[0.038]; forces[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arafat\u001b[39m ==> ENTITY: \u001b[32mYasser Arafat\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsraeli[0.052]; Israelis[0.049]; Israelis[0.049]; Palestinians[0.048]; Jordan[0.048]; Jerusalem[0.047]; Jerusalem[0.047]; Ammar[0.047]; Abu[0.046]; Qalandia[0.045]; Khaled[0.045]; Mustafa[0.045]; security[0.045]; forces[0.045]; Hijazi[0.044]; Ahmed[0.043]; Bank[0.043]; Ali[0.043]; West[0.042]; said[0.042]; said[0.042]; said[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gaza ==> ENTITY: \u001b[32mGaza Strip\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.258:0.258[0]; log p(e|m)= -1.374:-1.374[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.045]; Israel[0.045]; Israel[0.045]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; Palestinians[0.042]; Palestinians[0.042]; Jerusalem[0.040]; Jerusalem[0.040]; Jerusalem[0.040]; Jerusalem[0.040]; Jerusalem[0.040]; Arab[0.040]; Bank[0.040]; Bank[0.040]; Bank[0.040]; West[0.040]; West[0.040]; West[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arafat\u001b[39m ==> ENTITY: \u001b[32mYasser Arafat\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.046]; Israel[0.046]; Israel[0.046]; Israel[0.046]; Israeli[0.044]; Israeli[0.044]; Israeli[0.044]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; Gaza[0.042]; Palestinians[0.041]; Arafat[0.041]; Jerusalem[0.040]; Jerusalem[0.040]; Jerusalem[0.040]; Jerusalem[0.040]; Yasser[0.039]; Hebrew[0.038]; militants[0.038]; Jewish[0.038]; Jewish[0.038]; settlements[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jerusalem ==> ENTITY: \u001b[32mJerusalem\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.212:0.212[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJerusalem[0.046]; Jerusalem[0.046]; Jerusalem[0.046]; Israel[0.046]; Israeli[0.043]; Israeli[0.043]; Israelis[0.042]; Mustafa[0.042]; Arab[0.041]; West[0.041]; Abu[0.041]; Abu[0.041]; Palestinian[0.041]; Ammar[0.041]; Ammar[0.041]; holy[0.041]; Palestinians[0.040]; Palestinians[0.040]; Jordan[0.040]; Jordan[0.040]; Bank[0.040]; al[0.040]; al[0.040]; al[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = al-Aqsa\u001b[39m ==> ENTITY: \u001b[32mAl-Aqsa Mosque\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJerusalem[0.048]; Jerusalem[0.048]; Jerusalem[0.048]; mosque[0.047]; Al[0.046]; Israel[0.046]; Israel[0.046]; Israel[0.046]; Jamil[0.045]; Jamil[0.045]; Palestinian[0.045]; Palestinian[0.045]; Palestinian[0.045]; Palestinian[0.045]; Arabs[0.045]; Gaza[0.044]; Aqsa[0.044]; Palestinians[0.044]; holy[0.044]; Arafat[0.044]; Arafat[0.044]; Moslem[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jerusalem ==> ENTITY: \u001b[32mJerusalem\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.256:0.256[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJerusalem[0.046]; Jerusalem[0.046]; Jerusalem[0.046]; Jerusalem[0.046]; Israel[0.046]; Israel[0.046]; Israel[0.046]; Jewish[0.045]; Jewish[0.045]; Hebrew[0.045]; Israeli[0.043]; Israeli[0.043]; Moslem[0.042]; Arabs[0.042]; Palestinian[0.041]; Palestinian[0.041]; Palestinian[0.041]; Arab[0.041]; West[0.041]; West[0.041]; Abu[0.041]; Ammar[0.041]; holy[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jordan ==> ENTITY: \u001b[32mJordan\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.231:0.231[0]; log p(e|m)= -0.536:-0.536[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArab[0.047]; Jordan[0.047]; Palestinian[0.045]; Israeli[0.045]; Israeli[0.045]; Jerusalem[0.044]; Jerusalem[0.044]; Jerusalem[0.044]; Palestinians[0.044]; Palestinians[0.044]; holy[0.044]; Hijazi[0.043]; Bank[0.043]; Israelis[0.043]; Israelis[0.043]; West[0.043]; Mustafa[0.042]; Ammar[0.042]; Ammar[0.042]; security[0.042]; al[0.042]; al[0.042]; al[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.228:0.228[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israel[0.050]; Israeli[0.045]; Jewish[0.045]; Hebrew[0.044]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; Jerusalem[0.042]; Jerusalem[0.042]; settlements[0.042]; Gaza[0.041]; Arabs[0.041]; West[0.041]; West[0.041]; Bank[0.040]; Bank[0.040]; Al[0.038]; worship[0.038]; Yasser[0.038]; mosque[0.037]; Aqsa[0.037]; holy[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = al-Aqsa\u001b[39m ==> ENTITY: \u001b[32mAl-Aqsa Mosque\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJerusalem[0.044]; Jerusalem[0.044]; Jerusalem[0.044]; Jerusalem[0.044]; Jerusalem[0.044]; Abu[0.043]; al[0.042]; al[0.042]; Israel[0.042]; Israel[0.042]; Jamil[0.041]; Arab[0.041]; Jordan[0.041]; Jordan[0.041]; Palestinian[0.041]; Hijazi[0.041]; Aqsa[0.040]; Aqsa[0.040]; Palestinians[0.040]; Palestinians[0.040]; Arafat[0.040]; Moslem[0.040]; West[0.040]; Israelis[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moslem ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.196:0.196[0]; log p(e|m)= -1.214:-1.214[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npray[0.045]; pray[0.045]; Arabs[0.045]; holy[0.045]; Arab[0.044]; prayers[0.042]; mosque[0.042]; people[0.042]; al[0.041]; al[0.041]; al[0.041]; eternal[0.041]; Jerusalem[0.041]; Jerusalem[0.041]; Jerusalem[0.041]; Jerusalem[0.041]; Jerusalem[0.041]; Jamil[0.041]; Jamil[0.041]; Jewish[0.040]; Jewish[0.040]; Hebrew[0.040]; Pilgrims[0.040]; Ammar[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jewish ==> ENTITY: \u001b[32mJews\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.193:0.193[0]; log p(e|m)= -0.805:-0.805[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJewish[0.046]; Hebrew[0.044]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Israel[0.043]; Jerusalem[0.043]; Jerusalem[0.043]; Jerusalem[0.043]; Jerusalem[0.043]; Arabs[0.042]; Israeli[0.041]; Israeli[0.041]; Israeli[0.041]; Palestinian[0.040]; Palestinian[0.040]; Palestinian[0.040]; Palestinian[0.040]; Moslem[0.040]; al[0.040]; Gaza[0.039]; worship[0.039]; settlements[0.039]; Palestinians[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israelis ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.546:-0.546[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsraeli[0.050]; Israelis[0.049]; Jerusalem[0.046]; Palestinians[0.045]; Jordan[0.045]; West[0.044]; security[0.043]; Bank[0.043]; forces[0.042]; refugee[0.040]; said[0.040]; said[0.040]; said[0.040]; asked[0.040]; Abu[0.040]; Arafat[0.040]; Hijazi[0.040]; feel[0.040]; peace[0.040]; peace[0.040]; camp[0.039]; ground[0.039]; Qalandia[0.039]; patrol[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.280:0.280[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.052]; Israel[0.052]; Israel[0.052]; Israeli[0.047]; Israeli[0.047]; Jewish[0.047]; Jewish[0.047]; Hebrew[0.046]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Jerusalem[0.044]; Jerusalem[0.044]; Jerusalem[0.044]; settlements[0.044]; Palestinians[0.043]; Gaza[0.043]; Arabs[0.043]; West[0.043]; West[0.043]; West[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Abu Ammar\u001b[39m ==> ENTITY: \u001b[32mYasser Arafat\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.215:0.215[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.046]; Israel[0.046]; Palestinian[0.044]; Israeli[0.044]; Israeli[0.044]; Israeli[0.044]; Jordan[0.042]; Jordan[0.042]; Arafat[0.042]; Arab[0.042]; Israelis[0.042]; Palestinians[0.041]; Palestinians[0.041]; Palestinians[0.041]; Jerusalem[0.040]; Jerusalem[0.040]; Jerusalem[0.040]; Jerusalem[0.040]; Jerusalem[0.040]; Khaled[0.040]; Hijazi[0.040]; security[0.039]; Jamil[0.039]; Jewish[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jerusalem ==> ENTITY: \u001b[32mJerusalem\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.182:0.182[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJerusalem[0.049]; Israeli[0.045]; Israeli[0.045]; Israelis[0.045]; Israelis[0.045]; Mustafa[0.044]; Qalandia[0.044]; West[0.044]; Abu[0.043]; Abu[0.043]; Palestinian[0.043]; Ammar[0.043]; Ammar[0.043]; holy[0.043]; Ahmed[0.043]; Palestinians[0.043]; Palestinians[0.043]; Jordan[0.042]; Jordan[0.042]; Bank[0.042]; al[0.042]; al[0.042]; al[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.262:0.262[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.051]; Israel[0.051]; Israeli[0.046]; Israeli[0.046]; Arab[0.045]; Jewish[0.045]; Jewish[0.045]; Palestinian[0.044]; Palestinian[0.044]; Jerusalem[0.043]; Jerusalem[0.043]; Jerusalem[0.043]; Jerusalem[0.043]; Jerusalem[0.043]; Gaza[0.042]; Palestinians[0.042]; Palestinians[0.042]; Jordan[0.042]; Arabs[0.041]; West[0.040]; West[0.040]; policy[0.040]; holy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jordan ==> ENTITY: \u001b[32mJordan\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.236:0.236[0]; log p(e|m)= -0.536:-0.536[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.047]; Arab[0.045]; Jordan[0.045]; Palestinian[0.043]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; Jerusalem[0.042]; Jerusalem[0.042]; Jerusalem[0.042]; Jerusalem[0.042]; Palestinians[0.042]; Palestinians[0.042]; holy[0.042]; Hijazi[0.041]; Israelis[0.041]; Mustafa[0.040]; Ammar[0.040]; Jewish[0.040]; al[0.040]; al[0.040]; al[0.040]; Abu[0.040]; Khaled[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.207:0.207[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.052]; Hebrew[0.046]; Israeli[0.046]; Israeli[0.046]; Jewish[0.046]; Jerusalem[0.044]; Jerusalem[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; West[0.041]; Bank[0.040]; Wednesday[0.039]; worship[0.039]; prayer[0.039]; settlements[0.038]; legislature[0.038]; Al[0.038]; al[0.038]; pilgrims[0.038]; Friday[0.037]; prayers[0.037]; today[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jewish ==> ENTITY: \u001b[32mJews\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.191:0.191[0]; log p(e|m)= -0.805:-0.805[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJewish[0.048]; Hebrew[0.046]; Arab[0.045]; Israel[0.045]; Israel[0.045]; Israel[0.045]; Jerusalem[0.045]; Jerusalem[0.045]; Jerusalem[0.045]; Jerusalem[0.045]; Arabs[0.044]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; Palestinian[0.042]; Palestinian[0.042]; Palestinian[0.042]; Moslem[0.041]; al[0.041]; al[0.041]; Gaza[0.041]; settlements[0.041]; Jordan[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arafat\u001b[39m ==> ENTITY: \u001b[32mYasser Arafat\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.047]; Israel[0.047]; Israeli[0.045]; Israeli[0.045]; Israeli[0.045]; Palestinian[0.044]; Israelis[0.042]; Palestinians[0.042]; Palestinians[0.042]; Palestinians[0.042]; Jordan[0.041]; Jordan[0.041]; Jerusalem[0.041]; Jerusalem[0.041]; Jerusalem[0.041]; Jerusalem[0.041]; Jerusalem[0.041]; Arab[0.041]; Ammar[0.041]; Abu[0.040]; Khaled[0.039]; security[0.039]; militants[0.039]; forces[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.270:0.270[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.048]; Israel[0.048]; Israeli[0.044]; Israeli[0.044]; Arab[0.043]; Hebrew[0.043]; Jewish[0.043]; Jewish[0.043]; Palestinian[0.041]; Palestinian[0.041]; Palestinian[0.041]; Palestinian[0.041]; Jerusalem[0.041]; Jerusalem[0.041]; Jerusalem[0.041]; Jerusalem[0.041]; Jerusalem[0.041]; Gaza[0.040]; Palestinians[0.040]; Palestinians[0.040]; Arabs[0.039]; West[0.038]; West[0.038]; policy[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.279:0.279[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.048]; Israel[0.048]; Israel[0.048]; Israeli[0.044]; Israeli[0.044]; Israeli[0.044]; Hebrew[0.043]; Jewish[0.043]; Jewish[0.043]; Palestinian[0.041]; Palestinian[0.041]; Palestinian[0.041]; Palestinian[0.041]; Jerusalem[0.041]; Jerusalem[0.041]; Jerusalem[0.041]; Jerusalem[0.041]; Gaza[0.040]; Palestinians[0.040]; Arabs[0.039]; West[0.038]; West[0.038]; West[0.038]; policy[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.163:0.163[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsraeli[0.049]; Israelis[0.048]; Israelis[0.048]; Jerusalem[0.046]; Jerusalem[0.046]; Palestinians[0.046]; Jordan[0.045]; Jordan[0.045]; West[0.045]; Qalandia[0.044]; Bank[0.044]; Abu[0.043]; Abu[0.043]; holy[0.043]; pray[0.041]; pray[0.041]; pray[0.041]; peace[0.040]; peace[0.040]; peace[0.040]; refugee[0.040]; Arafat[0.040]; Arafat[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israelis ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.546:-0.546[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsraeli[0.049]; Israeli[0.049]; Israelis[0.047]; Jerusalem[0.045]; Jerusalem[0.045]; Palestinians[0.044]; Jordan[0.044]; Jordan[0.044]; West[0.043]; security[0.042]; Bank[0.042]; soldiers[0.039]; refugee[0.039]; said[0.039]; said[0.039]; said[0.039]; said[0.039]; forces[0.039]; asked[0.039]; Abu[0.039]; Abu[0.039]; Arafat[0.039]; Arafat[0.039]; Hijazi[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yasser Arafat\u001b[39m ==> ENTITY: \u001b[32mYasser Arafat\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.049]; Israel[0.049]; Israel[0.049]; Israeli[0.047]; Israeli[0.047]; Palestinian[0.046]; Palestinian[0.046]; Palestinian[0.046]; Palestinian[0.046]; Gaza[0.045]; Palestinians[0.044]; Arafat[0.043]; Jerusalem[0.043]; Jerusalem[0.043]; Jerusalem[0.043]; Hebrew[0.041]; Jewish[0.040]; settlements[0.040]; Jamil[0.039]; Jamil[0.039]; Arabs[0.039]; Aqsa[0.039]; Al[0.039]; \t\n\r [=======>......................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s893ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 836/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1056testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Salim Malik\u001b[39m ==> ENTITY: \u001b[32mSaleem Malik\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.206:0.206[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAkram[0.043]; Akram[0.043]; Saqlain[0.043]; Saqlain[0.043]; Ijaz[0.042]; Wasim[0.042]; Wasim[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Khan[0.042]; Younis[0.041]; Younis[0.041]; Sohail[0.041]; Sohail[0.041]; Waqar[0.041]; Waqar[0.041]; Hollioake[0.040]; Hollioake[0.040]; Hollioake[0.040]; Hollioake[0.040]; Hollioake[0.040]; Moin[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Waqar Younis\u001b[39m ==> ENTITY: \u001b[32mWaqar Younis\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.255:0.255[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWasim[0.046]; Akram[0.045]; Waqar[0.044]; Saqlain[0.043]; Saqlain[0.043]; Pakistan[0.043]; Pakistan[0.043]; Mushtaq[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; cricket[0.042]; match[0.041]; Cricket[0.041]; Moin[0.040]; Moin[0.040]; Khan[0.040]; Khan[0.040]; Sohail[0.040]; Sohail[0.040]; Younis[0.040]; Rehman[0.039]; Rehman[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gough\u001b[39m ==> ENTITY: \u001b[32mDarren Gough\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHollioake[0.043]; Hollioake[0.043]; Hollioake[0.043]; Hollioake[0.043]; Hollioake[0.043]; Croft[0.043]; Croft[0.043]; Croft[0.043]; Croft[0.043]; overs[0.043]; Gough[0.043]; Gough[0.043]; Gough[0.043]; Mushtaq[0.041]; Mushtaq[0.041]; Saqlain[0.041]; wickets[0.041]; lbw[0.040]; lbw[0.040]; Akram[0.040]; Wasim[0.039]; Thorpe[0.039]; Inzamam[0.039]; Bowling[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mullally ==> ENTITY: \u001b[32mAlan Mullally\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.163:0.163[0]; log p(e|m)= -0.740:-0.740[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaqlain[0.043]; Hollioake[0.043]; Hollioake[0.043]; Hollioake[0.043]; Hollioake[0.043]; Hollioake[0.043]; Gough[0.043]; Gough[0.043]; Gough[0.043]; Mullally[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; overs[0.042]; wickets[0.041]; Croft[0.041]; Croft[0.041]; Croft[0.041]; Croft[0.041]; Stewart[0.041]; Wasim[0.041]; Akram[0.041]; Waqar[0.040]; Ijaz[0.040]; Younis[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pakistan ==> ENTITY: \u001b[32mPakistan national cricket team\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.180:0.180[0]; log p(e|m)= -2.749:-2.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.050]; England[0.046]; England[0.046]; England[0.046]; England[0.046]; cricket[0.045]; Saqlain[0.043]; match[0.042]; Waqar[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Cricket[0.041]; Khan[0.040]; Rehman[0.040]; Younis[0.039]; Ahmed[0.039]; Ahmed[0.039]; Hollioake[0.039]; Moin[0.039]; day[0.038]; wickets[0.038]; second[0.038]; Headley[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mushtaq Ahmed\u001b[39m ==> ENTITY: \u001b[32mMushtaq Ahmed\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.256:0.256[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMushtaq[0.044]; Mushtaq[0.044]; Mushtaq[0.044]; Akram[0.043]; Akram[0.043]; Wasim[0.043]; Wasim[0.043]; Saqlain[0.043]; Saqlain[0.043]; Hollioake[0.041]; Hollioake[0.041]; Hollioake[0.041]; Hollioake[0.041]; Hollioake[0.041]; Waqar[0.041]; Waqar[0.041]; Waqar[0.041]; Ijaz[0.041]; Malik[0.041]; Sohail[0.040]; Sohail[0.040]; Younis[0.040]; Younis[0.040]; Younis[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mullally ==> ENTITY: \u001b[32mAlan Mullally\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.158:0.158[0]; log p(e|m)= -0.740:-0.740[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaqlain[0.043]; Saqlain[0.043]; Hollioake[0.042]; Hollioake[0.042]; Hollioake[0.042]; Hollioake[0.042]; Gough[0.042]; Gough[0.042]; Gough[0.042]; Gough[0.042]; Mullally[0.042]; Mullally[0.042]; Mushtaq[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; overs[0.041]; wickets[0.041]; wickets[0.041]; Croft[0.041]; Croft[0.041]; Croft[0.041]; Stewart[0.040]; Stewart[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mushtaq Ahmed\u001b[39m ==> ENTITY: \u001b[32mMushtaq Ahmed\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.231:0.231[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMushtaq[0.045]; Mushtaq[0.045]; Mushtaq[0.045]; Akram[0.044]; Wasim[0.044]; Saqlain[0.043]; Saqlain[0.043]; cricket[0.042]; Hollioake[0.042]; Waqar[0.042]; Waqar[0.042]; Younis[0.041]; Younis[0.041]; wickets[0.040]; England[0.040]; England[0.040]; England[0.040]; England[0.040]; Pakistan[0.040]; Pakistan[0.040]; Khan[0.040]; Cricket[0.040]; match[0.040]; innings[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hollioake\u001b[39m ==> ENTITY: \u001b[32mAdam Hollioake\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.065:0.065[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHollioake[0.047]; Hollioake[0.047]; Hollioake[0.047]; Hollioake[0.047]; wickets[0.044]; Wasim[0.044]; Wasim[0.044]; overs[0.044]; Saqlain[0.044]; Saqlain[0.044]; Inzamam[0.043]; Croft[0.043]; Croft[0.043]; Croft[0.043]; Croft[0.043]; Waqar[0.042]; Waqar[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; lbw[0.042]; lbw[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wasim Akram\u001b[39m ==> ENTITY: \u001b[32mWasim Akram\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.230:0.230[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAkram[0.048]; Wasim[0.047]; Mushtaq[0.045]; Mushtaq[0.045]; Mushtaq[0.045]; Mushtaq[0.045]; Ijaz[0.045]; Saqlain[0.045]; Saqlain[0.045]; Waqar[0.044]; Waqar[0.044]; Malik[0.044]; Sohail[0.043]; Sohail[0.043]; Khan[0.042]; Moin[0.042]; Younis[0.042]; Younis[0.042]; Rehman[0.042]; Rehman[0.042]; Saeed[0.042]; Inzamam[0.041]; overs[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Irani ==> ENTITY: \u001b[32mRonnie Irani\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.167:0.167[0]; log p(e|m)= -2.526:-2.526[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMushtaq[0.045]; Mushtaq[0.045]; Saqlain[0.045]; Wasim[0.044]; Hollioake[0.044]; Hollioake[0.044]; Hollioake[0.044]; Hollioake[0.044]; Hollioake[0.044]; Akram[0.044]; Waqar[0.042]; Younis[0.041]; Malik[0.041]; Ijaz[0.041]; Inzamam[0.041]; wickets[0.041]; overs[0.040]; lbw[0.040]; Gough[0.039]; Gough[0.039]; Rehman[0.039]; Croft[0.039]; Croft[0.039]; Croft[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saeed Anwar\u001b[39m ==> ENTITY: \u001b[32mSaeed Anwar\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.249:0.249[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWasim[0.044]; Wasim[0.044]; Saqlain[0.043]; Saqlain[0.043]; Akram[0.043]; Akram[0.043]; Waqar[0.042]; Waqar[0.042]; Waqar[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Ijaz[0.042]; Younis[0.041]; Younis[0.041]; Younis[0.041]; Moin[0.041]; Malik[0.041]; Inzamam[0.040]; Sohail[0.040]; Sohail[0.040]; Rehman[0.040]; Rehman[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Moin Khan\u001b[39m ==> ENTITY: \u001b[32mMoin Khan\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaqlain[0.045]; Akram[0.044]; Waqar[0.044]; Wasim[0.044]; Pakistan[0.043]; Pakistan[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; cricket[0.043]; Cricket[0.041]; second[0.041]; innings[0.041]; Younis[0.041]; wickets[0.040]; Hollioake[0.040]; England[0.040]; England[0.040]; England[0.040]; England[0.040]; match[0.040]; Rehman[0.040]; Ahmed[0.039]; Ahmed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ijaz Ahmed\u001b[39m ==> ENTITY: \u001b[32mIjaz Ahmed (cricketer)\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.238:0.238[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWasim[0.044]; Wasim[0.044]; Mushtaq[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Waqar[0.042]; Waqar[0.042]; Saqlain[0.042]; Saqlain[0.042]; Khan[0.042]; Rehman[0.042]; Rehman[0.042]; Sohail[0.041]; Sohail[0.041]; Ahmed[0.041]; Ahmed[0.041]; Akram[0.041]; Akram[0.041]; Malik[0.041]; Saeed[0.040]; Inzamam[0.040]; Younis[0.040]; Younis[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hollioake\u001b[39m ==> ENTITY: \u001b[32mAdam Hollioake\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.065:0.065[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHollioake[0.047]; Hollioake[0.047]; Hollioake[0.047]; Hollioake[0.047]; wickets[0.044]; Wasim[0.044]; Wasim[0.044]; overs[0.044]; Saqlain[0.044]; Saqlain[0.044]; Inzamam[0.043]; Croft[0.043]; Croft[0.043]; Croft[0.043]; Croft[0.043]; Waqar[0.042]; Waqar[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; lbw[0.042]; lbw[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wasim Akram\u001b[39m ==> ENTITY: \u001b[32mWasim Akram\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.234:0.234[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAkram[0.047]; Wasim[0.046]; Mushtaq[0.044]; Mushtaq[0.044]; Mushtaq[0.044]; Mushtaq[0.044]; Mushtaq[0.044]; Mushtaq[0.044]; Mushtaq[0.044]; Ijaz[0.044]; Saqlain[0.044]; Saqlain[0.044]; Saqlain[0.044]; Waqar[0.043]; Waqar[0.043]; Waqar[0.043]; Malik[0.043]; Sohail[0.042]; Sohail[0.042]; Khan[0.042]; Khan[0.042]; Moin[0.041]; Moin[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Croft ==> ENTITY: \u001b[32mRobert Croft\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.144:0.144[0]; log p(e|m)= -3.170:-3.170[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHollioake[0.045]; Hollioake[0.045]; Hollioake[0.045]; Hollioake[0.045]; Hollioake[0.045]; Croft[0.044]; overs[0.043]; Gough[0.042]; Gough[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Saqlain[0.042]; wickets[0.042]; lbw[0.041]; Bowling[0.041]; Thorpe[0.040]; Wasim[0.040]; Akram[0.040]; Malik[0.040]; Younis[0.039]; Waqar[0.039]; Inzamam[0.039]; Stewart[0.039]; Headley[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hollioake\u001b[39m ==> ENTITY: \u001b[32mAdam Hollioake\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.066:0.066[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHollioake[0.046]; Hollioake[0.046]; Hollioake[0.046]; Hollioake[0.046]; wickets[0.043]; Wasim[0.043]; overs[0.043]; Saqlain[0.043]; Croft[0.042]; Croft[0.042]; Waqar[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; lbw[0.041]; Akram[0.041]; Gough[0.040]; Gough[0.040]; lb[0.040]; Younis[0.040]; Fall[0.040]; Bowling[0.040]; Total[0.040]; Thorpe[0.039]; Malik[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Croft ==> ENTITY: \u001b[32mRobert Croft\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.165:0.165[0]; log p(e|m)= -3.170:-3.170[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHollioake[0.044]; Hollioake[0.044]; Hollioake[0.044]; Hollioake[0.044]; Croft[0.044]; Croft[0.044]; overs[0.042]; Gough[0.041]; Gough[0.041]; Gough[0.041]; Gough[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Saqlain[0.041]; Saqlain[0.041]; wickets[0.041]; wickets[0.041]; innings[0.040]; lbw[0.040]; lbw[0.040]; Bowling[0.040]; Bowling[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ata-ur-Rehman\u001b[39m ==> ENTITY: \u001b[32mAta-ur-Rehman\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAkram[0.044]; Pakistan[0.044]; Pakistan[0.044]; Wasim[0.043]; cricket[0.043]; Saqlain[0.043]; Saqlain[0.043]; Younis[0.043]; Younis[0.043]; Khan[0.041]; Waqar[0.041]; Waqar[0.041]; Cricket[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Ahmed[0.041]; Ahmed[0.041]; Ahmed[0.041]; Moin[0.040]; Headley[0.040]; match[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mushtaq Ahmed\u001b[39m ==> ENTITY: \u001b[32mMushtaq Ahmed\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.256:0.256[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMushtaq[0.045]; Mushtaq[0.045]; Mushtaq[0.045]; Akram[0.043]; Akram[0.043]; Wasim[0.043]; Wasim[0.043]; Saqlain[0.043]; Saqlain[0.043]; Hollioake[0.041]; Hollioake[0.041]; Hollioake[0.041]; Hollioake[0.041]; Hollioake[0.041]; Waqar[0.041]; Waqar[0.041]; Ijaz[0.041]; Malik[0.041]; Sohail[0.040]; Sohail[0.040]; Younis[0.040]; Younis[0.040]; wickets[0.040]; Khan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saqlain Mushtaq\u001b[39m ==> ENTITY: \u001b[32mSaqlain Mushtaq\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaqlain[0.045]; Wasim[0.044]; Wasim[0.044]; Akram[0.043]; Akram[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Hollioake[0.042]; Hollioake[0.042]; Hollioake[0.042]; Hollioake[0.042]; Hollioake[0.042]; Ijaz[0.041]; wickets[0.041]; Waqar[0.041]; Waqar[0.041]; overs[0.041]; Sohail[0.040]; Sohail[0.040]; Malik[0.040]; Inzamam[0.040]; lbw[0.040]; lbw[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland cricket team\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.171:0.171[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; England[0.047]; England[0.047]; cricket[0.046]; match[0.043]; Cricket[0.043]; Saqlain[0.042]; Birmingham[0.042]; Hollioake[0.041]; Pakistan[0.041]; Pakistan[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Atherton[0.041]; Croft[0.041]; day[0.041]; second[0.040]; Gough[0.040]; wickets[0.040]; Waqar[0.040]; Saturday[0.039]; innings[0.039]; Thorpe[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pakistan ==> ENTITY: \u001b[32mPakistan national cricket team\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.180:0.180[0]; log p(e|m)= -2.749:-2.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.050]; England[0.046]; England[0.046]; England[0.046]; England[0.046]; cricket[0.045]; Saqlain[0.043]; match[0.042]; Waqar[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Cricket[0.041]; Khan[0.040]; Rehman[0.040]; Younis[0.039]; Ahmed[0.039]; Ahmed[0.039]; Hollioake[0.039]; Moin[0.039]; day[0.038]; wickets[0.038]; second[0.038]; Headley[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gough\u001b[39m ==> ENTITY: \u001b[32mDarren Gough\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHollioake[0.045]; Hollioake[0.045]; Hollioake[0.045]; Hollioake[0.045]; Croft[0.044]; Croft[0.044]; Croft[0.044]; Croft[0.044]; overs[0.044]; Gough[0.044]; Gough[0.044]; Gough[0.044]; Gough[0.044]; Mushtaq[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Saqlain[0.043]; Saqlain[0.043]; wickets[0.042]; wickets[0.042]; lbw[0.042]; lbw[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saqlain Mushtaq\u001b[39m ==> ENTITY: \u001b[32mSaqlain Mushtaq\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.044]; Wasim[0.044]; Akram[0.044]; Mushtaq[0.043]; Mushtaq[0.043]; match[0.043]; Pakistan[0.043]; Pakistan[0.043]; England[0.042]; England[0.042]; England[0.042]; England[0.042]; Hollioake[0.042]; Cricket[0.042]; wickets[0.041]; Waqar[0.041]; Waqar[0.041]; lbw[0.040]; lbw[0.040]; Khan[0.040]; Younis[0.040]; Younis[0.040]; Rehman[0.039]; Atherton[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Inzamam-ul-Haq\u001b[39m ==> ENTITY: \u001b[32mInzamam-ul-Haq\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.208:0.208[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWasim[0.043]; Wasim[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Akram[0.043]; Akram[0.043]; Saqlain[0.042]; Saqlain[0.042]; Waqar[0.042]; Waqar[0.042]; Younis[0.042]; Younis[0.042]; Ijaz[0.042]; Sohail[0.041]; Sohail[0.041]; Malik[0.041]; Moin[0.040]; Khan[0.040]; Hollioake[0.040]; Hollioake[0.040]; Hollioake[0.040]; Hollioake[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hollioake\u001b[39m ==> ENTITY: \u001b[32mAdam Hollioake\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.065:0.065[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHollioake[0.047]; Hollioake[0.047]; Hollioake[0.047]; Hollioake[0.047]; wickets[0.044]; Wasim[0.044]; Wasim[0.044]; overs[0.044]; Saqlain[0.044]; Saqlain[0.044]; Inzamam[0.043]; Croft[0.043]; Croft[0.043]; Croft[0.043]; Croft[0.043]; Waqar[0.042]; Waqar[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; Mushtaq[0.042]; lbw[0.042]; lbw[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Waqar Younis\u001b[39m ==> ENTITY: \u001b[32mWaqar Younis\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.271:0.271[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWasim[0.053]; Wasim[0.053]; Akram[0.051]; Akram[0.051]; Waqar[0.049]; Saqlain[0.049]; Saqlain[0.049]; Mushtaq[0.048]; Mushtaq[0.048]; Mushtaq[0.048]; Mushtaq[0.048]; Malik[0.048]; Ijaz[0.047]; Moin[0.046]; Khan[0.045]; Sohail[0.045]; Sohail[0.045]; Inzamam[0.045]; Younis[0.045]; Rehman[0.044]; Rehman[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Waqar Younis\u001b[39m ==> ENTITY: \u001b[32mWaqar Younis\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.277:0.277[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWasim[0.047]; Wasim[0.047]; Akram[0.046]; Akram[0.046]; Waqar[0.045]; Waqar[0.045]; Saqlain[0.044]; Saqlain[0.044]; Saqlain[0.044]; Mushtaq[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Malik[0.043]; Ijaz[0.043]; Moin[0.041]; Khan[0.041]; Sohail[0.041]; Sohail[0.041]; Inzamam[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland cricket team\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.173:0.173[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; England[0.047]; England[0.047]; cricket[0.046]; Birmingham[0.044]; match[0.043]; Cricket[0.042]; Saqlain[0.041]; Hollioake[0.041]; Pakistan[0.041]; Pakistan[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Atherton[0.041]; Croft[0.040]; Wasim[0.040]; day[0.040]; second[0.040]; Gough[0.040]; wickets[0.040]; Waqar[0.039]; Saturday[0.039]; innings[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Croft ==> ENTITY: \u001b[32mRobert Croft\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.176:0.176[0]; log p(e|m)= -3.170:-3.170[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHollioake[0.046]; Hollioake[0.046]; Hollioake[0.046]; Hollioake[0.046]; Croft[0.045]; Croft[0.045]; Croft[0.045]; overs[0.044]; Gough[0.043]; Gough[0.043]; Gough[0.043]; Gough[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Saqlain[0.043]; Saqlain[0.043]; wickets[0.042]; wickets[0.042]; innings[0.042]; lbw[0.042]; lbw[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Birmingham ==> ENTITY: \u001b[32mBirmingham\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.402:-0.402[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; England[0.047]; England[0.047]; England[0.047]; cricket[0.047]; second[0.045]; match[0.045]; innings[0.045]; Cricket[0.044]; Hollioake[0.044]; Gough[0.044]; Croft[0.043]; Atherton[0.043]; wickets[0.042]; day[0.042]; Mushtaq[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; run[0.041]; run[0.041]; run[0.041]; Saturday[0.041]; Waqar[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Thorpe ==> ENTITY: \u001b[32mGraham Thorpe\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.200:0.200[0]; log p(e|m)= -2.976:-2.976[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMushtaq[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Mushtaq[0.043]; Gough[0.043]; Gough[0.043]; Gough[0.043]; Gough[0.043]; Akram[0.042]; Akram[0.042]; Saqlain[0.042]; Saqlain[0.042]; Wasim[0.041]; Wasim[0.041]; wickets[0.041]; wickets[0.041]; Hollioake[0.041]; Hollioake[0.041]; Hollioake[0.041]; Hollioake[0.041]; Hollioake[0.041]; innings[0.041]; lbw[0.041]; lbw[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = D. Gough\u001b[39m ==> ENTITY: \u001b[32mDarren Gough\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.043]; England[0.043]; England[0.043]; England[0.043]; cricket[0.043]; Hollioake[0.043]; Croft[0.043]; Gough[0.042]; match[0.042]; Mushtaq[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Saqlain[0.041]; Saqlain[0.041]; Atherton[0.041]; wickets[0.041]; day[0.041]; Cricket[0.041]; Pakistan[0.041]; Pakistan[0.041]; lbw[0.040]; lbw[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gough\u001b[39m ==> ENTITY: \u001b[32mDarren Gough\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHollioake[0.043]; Hollioake[0.043]; Hollioake[0.043]; Hollioake[0.043]; Hollioake[0.043]; Croft[0.042]; Croft[0.042]; Croft[0.042]; Croft[0.042]; overs[0.042]; Gough[0.042]; Gough[0.042]; Gough[0.042]; Mushtaq[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Saqlain[0.041]; Saqlain[0.041]; wickets[0.041]; lbw[0.040]; lbw[0.040]; Akram[0.040]; Akram[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ata-ur-Rehman\u001b[39m ==> ENTITY: \u001b[32mAta-ur-Rehman\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAkram[0.044]; Akram[0.044]; Ijaz[0.044]; Wasim[0.043]; Saqlain[0.043]; Saqlain[0.043]; Younis[0.043]; Younis[0.043]; Malik[0.042]; Saeed[0.041]; Khan[0.041]; Sohail[0.041]; Sohail[0.041]; Waqar[0.041]; Waqar[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Inzamam[0.040]; Ahmed[0.040]; Ahmed[0.040]; Ahmed[0.040]; Moin[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ata-ur-Rehman\u001b[39m ==> ENTITY: \u001b[32mAta-ur-Rehman\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAkram[0.044]; Akram[0.044]; Ijaz[0.044]; Wasim[0.043]; Wasim[0.043]; Saqlain[0.042]; Saqlain[0.042]; Younis[0.042]; Younis[0.042]; Younis[0.042]; Malik[0.042]; Saeed[0.041]; Khan[0.041]; Sohail[0.041]; Sohail[0.041]; Waqar[0.041]; Waqar[0.041]; Waqar[0.041]; Mushtaq[0.040]; Mushtaq[0.040]; Mushtaq[0.040]; Mushtaq[0.040]; Mushtaq[0.040]; Mushtaq[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gough\u001b[39m ==> ENTITY: \u001b[32mDarren Gough\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHollioake[0.043]; Hollioake[0.043]; Hollioake[0.043]; Hollioake[0.043]; Croft[0.042]; Croft[0.042]; Croft[0.042]; Croft[0.042]; overs[0.042]; Gough[0.042]; Gough[0.042]; Gough[0.042]; Mushtaq[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Saqlain[0.041]; Saqlain[0.041]; wickets[0.041]; wickets[0.041]; lbw[0.040]; lbw[0.040]; Akram[0.040]; Akram[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mushtaq Ahmed\u001b[39m ==> ENTITY: \u001b[32mMushtaq Ahmed\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.229:0.229[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMushtaq[0.048]; Mushtaq[0.048]; Akram[0.046]; Wasim[0.046]; Saqlain[0.046]; cricket[0.044]; Hollioake[0.044]; Waqar[0.044]; Waqar[0.044]; Younis[0.043]; Younis[0.043]; wickets[0.043]; England[0.042]; England[0.042]; England[0.042]; England[0.042]; Pakistan[0.042]; Pakistan[0.042]; Khan[0.042]; Cricket[0.042]; match[0.042]; innings[0.042]; Ahmed[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Croft ==> ENTITY: \u001b[32mRobert Croft\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.177:0.177[0]; log p(e|m)= -3.170:-3.170[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHollioake[0.044]; Hollioake[0.044]; Hollioake[0.044]; Hollioake[0.044]; Hollioake[0.044]; Croft[0.043]; Croft[0.043]; Croft[0.043]; overs[0.042]; Gough[0.041]; Gough[0.041]; Gough[0.041]; Gough[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Saqlain[0.041]; Saqlain[0.041]; wickets[0.041]; wickets[0.041]; innings[0.040]; lbw[0.040]; lbw[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland cricket team\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.171:0.171[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; England[0.047]; England[0.047]; cricket[0.046]; match[0.043]; Cricket[0.043]; Birmingham[0.042]; Saqlain[0.042]; Hollioake[0.041]; Pakistan[0.041]; Pakistan[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Atherton[0.041]; Croft[0.041]; day[0.041]; second[0.040]; Gough[0.040]; wickets[0.040]; Waqar[0.039]; Saturday[0.039]; innings[0.039]; Thorpe[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEngland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEngland cricket team\u001b[39m <---> \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.257:0.239[\u001b[31m0.017\u001b[39m]; local()= 0.171:0.077[\u001b[31m0.094\u001b[39m]; log p(e|m)= -3.297:-0.190[\u001b[32m3.107\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; England[0.047]; England[0.047]; cricket[0.046]; match[0.043]; Cricket[0.043]; Saqlain[0.042]; Birmingham[0.042]; Hollioake[0.041]; Pakistan[0.041]; Pakistan[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Mushtaq[0.041]; Atherton[0.041]; Croft[0.041]; day[0.041]; second[0.040]; Gough[0.040]; wickets[0.040]; Waqar[0.040]; Saturday[0.039]; innings[0.039]; Thorpe[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hollioake\u001b[39m ==> ENTITY: \u001b[32mAdam Hollioake\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.069:0.069[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHollioake[0.049]; Hollioake[0.049]; Hollioake[0.049]; Hollioake[0.049]; wickets[0.046]; Wasim[0.046]; overs[0.046]; Saqlain[0.046]; Saqlain[0.046]; Inzamam[0.045]; Croft[0.045]; Croft[0.045]; Croft[0.045]; Croft[0.045]; Waqar[0.044]; Mushtaq[0.044]; Mushtaq[0.044]; Mushtaq[0.044]; Mushtaq[0.044]; lbw[0.044]; lbw[0.044]; Akram[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saqlain Mushtaq\u001b[39m ==> ENTITY: \u001b[32mSaqlain Mushtaq\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaqlain[0.046]; Wasim[0.045]; Wasim[0.045]; Akram[0.045]; Akram[0.045]; Mushtaq[0.044]; Mushtaq[0.044]; Mushtaq[0.044]; Mushtaq[0.044]; Hollioake[0.043]; Hollioake[0.043]; Hollioake[0.043]; Hollioake[0.043]; Hollioake[0.043]; Ijaz[0.043]; wickets[0.042]; Waqar[0.042]; Waqar[0.042]; Waqar[0.042]; Sohail[0.042]; Sohail[0.042]; Malik[0.041]; Inzamam[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Moin Khan\u001b[39m ==> ENTITY: \u001b[32mMoin Khan\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaqlain[0.050]; Saqlain[0.050]; Akram[0.049]; Akram[0.049]; Waqar[0.049]; Waqar[0.049]; Waqar[0.049]; Wasim[0.049]; Wasim[0.049]; Mushtaq[0.048]; Mushtaq[0.048]; Mushtaq[0.048]; Mushtaq[0.048]; Ijaz[0.047]; Malik[0.046]; innings[0.046]; Younis[0.045]; Younis[0.045]; Younis[0.045]; wickets[0.045]; wickets[0.045]; \t\n\r [=======>......................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s441ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 887/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1061testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.047]; British[0.045]; British[0.045]; won[0.044]; Masters[0.043]; Mark[0.042]; Raymond[0.042]; Padraig[0.042]; Saturday[0.042]; Russell[0.041]; winners[0.041]; Colin[0.041]; European[0.041]; European[0.041]; Andrew[0.041]; Paul[0.041]; Paul[0.041]; Robert[0.041]; Robert[0.041]; Ian[0.041]; Lee[0.040]; Stephen[0.040]; Frank[0.040]; Wayne[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.129:0.129[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.053]; Italy[0.045]; Ireland[0.044]; Ireland[0.044]; Trinidad[0.044]; Miguel[0.044]; Miguel[0.044]; Martin[0.042]; Jimenez[0.042]; Mark[0.041]; British[0.041]; Robert[0.041]; Stephen[0.041]; Jonathan[0.040]; Frank[0.040]; Raymond[0.040]; Saturday[0.040]; Australia[0.040]; Australia[0.040]; South[0.040]; Andrew[0.039]; pounds[0.039]; Angel[0.039]; Angel[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mTrinidad\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mTrinidad and Tobago\u001b[39m <---> \u001b[32mTrinidad\u001b[39m\t\nSCORES: global= 0.253:0.253[\u001b[31m0.000\u001b[39m]; local()= 0.126:0.113[\u001b[31m0.014\u001b[39m]; log p(e|m)= -2.996:-0.341[\u001b[32m2.655\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMiguel[0.047]; Miguel[0.047]; Spain[0.046]; Spain[0.046]; Robert[0.045]; Robert[0.045]; Stephen[0.044]; South[0.044]; Lee[0.043]; Martin[0.043]; Raymond[0.043]; Australia[0.043]; Australia[0.043]; Mark[0.042]; Colin[0.042]; won[0.042]; Andrew[0.042]; Jimenez[0.042]; Ireland[0.042]; Ireland[0.042]; British[0.042]; British[0.042]; Saturday[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Peter Mitchell ==> ENTITY: \u001b[32mPeter Mitchell (golfer)\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.123:0.123[0]; log p(e|m)= -1.204:-1.204[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.044]; Mark[0.043]; Ian[0.043]; Martin[0.043]; Australia[0.042]; Australia[0.042]; Paul[0.042]; Paul[0.042]; Paul[0.042]; Colin[0.042]; Stephen[0.041]; Andrew[0.041]; Russell[0.041]; Montgomerie[0.041]; Saturday[0.041]; Nobilo[0.041]; Ames[0.041]; Woosnam[0.041]; Spain[0.041]; Spain[0.041]; Retief[0.041]; Westwood[0.041]; Coltart[0.041]; Goosen[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mark McNulty\u001b[39m ==> ENTITY: \u001b[32mMark McNulty\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTour[0.045]; Tour[0.045]; Zimbabwe[0.044]; Masters[0.044]; Coltart[0.044]; Golf[0.043]; European[0.042]; European[0.042]; Ireland[0.042]; Ireland[0.042]; Africa[0.042]; Goosen[0.042]; Nobilo[0.041]; Lawrie[0.041]; Padraig[0.041]; Retief[0.041]; British[0.040]; British[0.040]; McGinley[0.040]; Harrington[0.040]; Colin[0.040]; Australia[0.040]; Australia[0.040]; Jonathan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Robert Allenby\u001b[39m ==> ENTITY: \u001b[32mRobert Allenby\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwon[0.044]; Tour[0.044]; Tour[0.044]; Masters[0.043]; Golf[0.043]; Mark[0.042]; Lawrie[0.042]; Allenby[0.042]; Colin[0.042]; Martin[0.042]; Lee[0.042]; Montgomerie[0.041]; Nobilo[0.041]; Australia[0.041]; Australia[0.041]; Westwood[0.041]; Ian[0.041]; South[0.041]; Harrington[0.041]; McGinley[0.041]; Andrew[0.040]; Raymond[0.040]; Jonathan[0.040]; Robert[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Retief Goosen\u001b[39m ==> ENTITY: \u001b[32mRetief Goosen\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMasters[0.045]; Tour[0.045]; won[0.045]; Nobilo[0.044]; Montgomerie[0.042]; Coltart[0.042]; Andrew[0.042]; Harrington[0.042]; Mark[0.042]; Lawrie[0.042]; Colin[0.041]; Allenby[0.041]; Allenby[0.041]; Martin[0.041]; European[0.041]; Woosnam[0.041]; Westwood[0.041]; Ian[0.041]; Robert[0.040]; Robert[0.040]; Peter[0.040]; Jonathan[0.040]; McGinley[0.040]; Africa[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = British ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.489:-0.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.048]; Ireland[0.045]; Ireland[0.045]; Paul[0.043]; Paul[0.043]; Stephen[0.042]; Russell[0.042]; Ian[0.042]; Peter[0.042]; Mark[0.042]; Robert[0.042]; Robert[0.042]; Australia[0.042]; Australia[0.042]; London[0.041]; Colin[0.041]; Wayne[0.041]; Andrew[0.040]; European[0.040]; European[0.040]; Lawrie[0.040]; Lee[0.040]; Italy[0.039]; South[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Raymond Russell\u001b[39m ==> ENTITY: \u001b[32mRaymond Russell\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMasters[0.047]; Tour[0.046]; Tour[0.046]; Lawrie[0.042]; Spain[0.042]; Spain[0.042]; won[0.042]; Montgomerie[0.042]; Harrington[0.042]; Golf[0.041]; European[0.041]; European[0.041]; British[0.041]; British[0.041]; Nobilo[0.041]; Robert[0.041]; Robert[0.041]; Westwood[0.040]; Coltart[0.040]; Goosen[0.040]; Colin[0.040]; Padraig[0.040]; Ireland[0.040]; Ireland[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIreland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIreland\u001b[39m <---> \u001b[32mRepublic of Ireland\u001b[39m\t\nSCORES: global= 0.256:0.253[\u001b[31m0.003\u001b[39m]; local()= 0.155:0.134[\u001b[31m0.021\u001b[39m]; log p(e|m)= -0.566:-1.505[\u001b[31m0.939\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.054]; Australia[0.044]; Australia[0.044]; British[0.043]; British[0.043]; Italy[0.042]; Martin[0.042]; Colin[0.041]; Stephen[0.041]; won[0.041]; Russell[0.041]; Zealand[0.041]; Ian[0.041]; Zimbabwe[0.041]; Lawrie[0.041]; Padraig[0.041]; Paul[0.040]; Paul[0.040]; Paul[0.040]; Trinidad[0.040]; European[0.040]; McGinley[0.040]; Andrew[0.040]; Mark[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andrew Coltart\u001b[39m ==> ENTITY: \u001b[32mAndrew Coltart\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMasters[0.044]; Lawrie[0.044]; Golf[0.043]; Tour[0.043]; Tour[0.043]; European[0.043]; European[0.043]; Montgomerie[0.042]; British[0.042]; British[0.042]; won[0.042]; Colin[0.042]; Westwood[0.041]; Ian[0.041]; Woosnam[0.041]; Martin[0.041]; Broadhurst[0.041]; Peter[0.041]; McGinley[0.040]; Harrington[0.040]; Ireland[0.040]; Ireland[0.040]; Trinidad[0.040]; South[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lee Westwood\u001b[39m ==> ENTITY: \u001b[32mLee Westwood\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.047]; Masters[0.047]; Tour[0.045]; Tour[0.045]; Montgomerie[0.044]; Goosen[0.044]; European[0.044]; European[0.044]; Woosnam[0.044]; McGinley[0.044]; Lawrie[0.043]; Harrington[0.043]; Martin[0.043]; Padraig[0.043]; Ireland[0.043]; Ireland[0.043]; Spain[0.043]; Spain[0.043]; won[0.043]; Mark[0.042]; Coltart[0.042]; British[0.042]; British[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Miguel Angel Martin\u001b[39m ==> ENTITY: \u001b[32mMiguel Ángel Martín (golfer)\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.125:0.125[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.045]; Spain[0.045]; Miguel[0.044]; Retief[0.043]; Goosen[0.043]; Nobilo[0.043]; Woosnam[0.043]; Trinidad[0.042]; Coltart[0.042]; Padraig[0.042]; Montgomerie[0.042]; Lawrie[0.042]; Harrington[0.041]; Peter[0.041]; Ireland[0.041]; Ireland[0.041]; Westwood[0.040]; Angel[0.040]; Broadhurst[0.040]; Australia[0.040]; Australia[0.040]; South[0.040]; Colin[0.040]; Italy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ian Woosnam\u001b[39m ==> ENTITY: \u001b[32mIan Woosnam\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.046]; Montgomerie[0.044]; Tour[0.044]; Tour[0.044]; won[0.044]; Masters[0.044]; Westwood[0.043]; Goosen[0.042]; Lawrie[0.041]; European[0.041]; European[0.041]; McGinley[0.041]; Allenby[0.041]; Allenby[0.041]; Coltart[0.041]; Padraig[0.041]; British[0.040]; British[0.040]; Mark[0.040]; Colin[0.040]; Harrington[0.040]; South[0.040]; Nobilo[0.040]; Retief[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Frank Nobilo\u001b[39m ==> ENTITY: \u001b[32mFrank Nobilo\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTour[0.045]; Tour[0.045]; Golf[0.044]; Masters[0.044]; Australia[0.043]; Australia[0.043]; South[0.042]; Mark[0.041]; Peter[0.041]; Andrew[0.041]; Spain[0.041]; Spain[0.041]; Colin[0.041]; European[0.041]; European[0.041]; Woosnam[0.041]; Westwood[0.041]; Ireland[0.041]; Ireland[0.041]; Lawrie[0.041]; Italy[0.041]; Montgomerie[0.040]; Martin[0.040]; Stephen[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.124:0.124[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.053]; Italy[0.045]; Ireland[0.044]; Ireland[0.044]; Trinidad[0.044]; Miguel[0.044]; Miguel[0.044]; Martin[0.042]; Jimenez[0.042]; Mark[0.041]; Robert[0.041]; Stephen[0.041]; Jonathan[0.040]; Frank[0.040]; Raymond[0.040]; Australia[0.040]; Australia[0.040]; South[0.040]; Andrew[0.039]; pounds[0.039]; Angel[0.039]; Angel[0.039]; Africa[0.039]; Russell[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = European Tour\u001b[39m ==> ENTITY: \u001b[32mPGA European Tour\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTour[0.045]; Masters[0.045]; European[0.044]; British[0.044]; British[0.044]; Golf[0.043]; Mark[0.042]; Ireland[0.042]; Ireland[0.042]; won[0.042]; Lawrie[0.042]; Coltart[0.041]; winners[0.041]; Australia[0.041]; Australia[0.041]; Winners[0.041]; Money[0.041]; Italy[0.041]; Montgomerie[0.041]; Colin[0.040]; Woosnam[0.040]; Robert[0.040]; Robert[0.040]; Saturday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Africa ==> ENTITY: \u001b[32mSouth Africa\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.263:-0.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.046]; Australia[0.046]; Ireland[0.045]; Ireland[0.045]; Italy[0.044]; Zimbabwe[0.043]; Spain[0.043]; Spain[0.043]; Zealand[0.042]; won[0.042]; Saturday[0.042]; Colin[0.041]; Tour[0.041]; Ian[0.041]; Andrew[0.041]; Mark[0.041]; Martin[0.040]; Trinidad[0.040]; Stephen[0.040]; Russell[0.040]; Jonathan[0.040]; Coltart[0.040]; Wayne[0.039]; Mitchell[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paul Lawrie\u001b[39m ==> ENTITY: \u001b[32mPaul Lawrie\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.046]; Masters[0.045]; Tour[0.044]; Tour[0.044]; Coltart[0.043]; Montgomerie[0.042]; Goosen[0.042]; won[0.042]; McGinley[0.042]; Colin[0.042]; Harrington[0.042]; Woosnam[0.042]; British[0.041]; British[0.041]; Westwood[0.041]; Ian[0.041]; Ireland[0.041]; Ireland[0.041]; European[0.041]; European[0.041]; Mark[0.040]; Martin[0.040]; Jimenez[0.040]; Padraig[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.125:0.125[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.046]; Ireland[0.046]; Australia[0.046]; Australia[0.046]; Spain[0.044]; Spain[0.044]; won[0.044]; European[0.043]; European[0.043]; British[0.042]; British[0.042]; Zealand[0.041]; Golf[0.041]; Trinidad[0.040]; Rocca[0.040]; South[0.040]; Tour[0.040]; Tour[0.040]; Saturday[0.040]; Wayne[0.039]; Africa[0.039]; Martin[0.039]; Zimbabwe[0.039]; Nobilo[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Colin Montgomerie\u001b[39m ==> ENTITY: \u001b[32mColin Montgomerie\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.046]; Masters[0.043]; won[0.043]; Woosnam[0.043]; European[0.043]; European[0.043]; Tour[0.042]; Tour[0.042]; Lawrie[0.042]; McGinley[0.042]; British[0.042]; British[0.042]; Westwood[0.042]; Coltart[0.041]; Ian[0.041]; Ireland[0.041]; Ireland[0.041]; Goosen[0.041]; Harrington[0.040]; Padraig[0.040]; Spain[0.040]; Spain[0.040]; Mark[0.040]; Andrew[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Padraig Harrington\u001b[39m ==> ENTITY: \u001b[32mPádraig Harrington\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLawrie[0.046]; McGinley[0.044]; Martin[0.044]; Masters[0.044]; Westwood[0.043]; Tour[0.042]; Montgomerie[0.042]; Ireland[0.042]; Ireland[0.042]; Woosnam[0.042]; Colin[0.042]; Mark[0.041]; McNulty[0.041]; Ian[0.041]; won[0.041]; Coltart[0.041]; Allenby[0.041]; Allenby[0.041]; Lee[0.040]; Paul[0.040]; Paul[0.040]; Paul[0.040]; Broadhurst[0.040]; Andrew[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wayne Riley\u001b[39m ==> ENTITY: \u001b[32mWayne Riley\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTour[0.046]; Tour[0.046]; Masters[0.045]; Australia[0.043]; Australia[0.043]; Golf[0.042]; Zealand[0.042]; Nobilo[0.042]; Ireland[0.042]; Ireland[0.042]; Harrington[0.041]; European[0.041]; European[0.041]; Spain[0.041]; Spain[0.041]; South[0.041]; Coltart[0.041]; Retief[0.041]; Montgomerie[0.040]; Westwood[0.040]; won[0.040]; Padraig[0.040]; Colin[0.040]; Italy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jonathan Lomas\u001b[39m ==> ENTITY: \u001b[32mJonathan Lomas\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMartin[0.047]; Colin[0.046]; Spain[0.045]; Spain[0.045]; Padraig[0.045]; Nobilo[0.044]; Andrew[0.044]; Robert[0.044]; Montgomerie[0.044]; Lee[0.044]; Coltart[0.044]; Mark[0.043]; Broadhurst[0.043]; Westwood[0.043]; Retief[0.043]; Harrington[0.043]; Trinidad[0.042]; Raymond[0.042]; Ames[0.042]; Goosen[0.042]; Peter[0.042]; Lawrie[0.042]; Stephen[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paul Broadhurst\u001b[39m ==> ENTITY: \u001b[32mPaul Broadhurst\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.129:0.129[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWestwood[0.044]; Lawrie[0.044]; Martin[0.043]; Padraig[0.043]; Lomas[0.043]; Paul[0.042]; Paul[0.042]; Andrew[0.042]; Spain[0.042]; Spain[0.042]; Nobilo[0.042]; Mark[0.042]; Coltart[0.042]; Harrington[0.042]; Peter[0.041]; Retief[0.041]; Jimenez[0.041]; Goosen[0.041]; Allenby[0.041]; Miguel[0.041]; Miguel[0.041]; Australia[0.040]; Australia[0.040]; Raymond[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Zealand ==> ENTITY: \u001b[32mNew Zealand\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.238:-0.238[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; Australia[0.048]; Ireland[0.045]; Ireland[0.045]; Trinidad[0.044]; Italy[0.042]; Martin[0.041]; Zimbabwe[0.041]; Stephen[0.041]; Mark[0.041]; British[0.041]; British[0.041]; Spain[0.041]; Spain[0.041]; Russell[0.040]; South[0.040]; won[0.040]; Ian[0.040]; European[0.040]; European[0.040]; Saturday[0.040]; Wayne[0.040]; Colin[0.039]; Andrew[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zimbabwe ==> ENTITY: \u001b[32mZimbabwe\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.367:-0.367[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.047]; Australia[0.047]; South[0.047]; Africa[0.046]; Ireland[0.045]; Ireland[0.045]; won[0.044]; Coltart[0.044]; Ian[0.044]; Trinidad[0.043]; Saturday[0.043]; Wayne[0.043]; Zealand[0.042]; Golf[0.042]; Spain[0.042]; Spain[0.042]; Retief[0.042]; European[0.042]; European[0.042]; Martin[0.042]; Colin[0.042]; Italy[0.042]; Jonathan[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.051]; Ireland[0.046]; Ireland[0.046]; Martin[0.045]; Mark[0.045]; Stephen[0.044]; Colin[0.043]; Zealand[0.043]; South[0.043]; Ian[0.043]; Italy[0.043]; Andrew[0.043]; Lee[0.043]; Russell[0.043]; won[0.042]; Jonathan[0.042]; Spain[0.042]; Spain[0.042]; Masters[0.042]; Peter[0.042]; Trinidad[0.042]; Lawrie[0.042]; Zimbabwe[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Miguel Angel Jimenez\u001b[39m ==> ENTITY: \u001b[32mMiguel Ángel Jiménez\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMasters[0.046]; Spain[0.045]; Spain[0.045]; Nobilo[0.042]; Miguel[0.042]; Martin[0.042]; Padraig[0.042]; Goosen[0.042]; Montgomerie[0.042]; Woosnam[0.042]; Retief[0.041]; won[0.041]; Harrington[0.041]; British[0.041]; Lawrie[0.041]; Colin[0.041]; Ireland[0.040]; Ireland[0.040]; Lee[0.040]; Coltart[0.040]; Westwood[0.040]; McGinley[0.040]; Mark[0.040]; Andrew[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = European Tour\u001b[39m ==> ENTITY: \u001b[32mPGA European Tour\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTour[0.045]; Masters[0.045]; European[0.044]; British[0.044]; British[0.044]; Golf[0.043]; Mark[0.042]; Ireland[0.042]; won[0.042]; Lawrie[0.042]; Coltart[0.041]; winners[0.041]; Australia[0.041]; Australia[0.041]; Winners[0.041]; Money[0.041]; Italy[0.041]; Montgomerie[0.041]; Colin[0.041]; Woosnam[0.040]; Robert[0.040]; Robert[0.040]; Saturday[0.040]; Stephen[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = British Masters\u001b[39m ==> ENTITY: \u001b[32mBritish Masters\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.045]; Ireland[0.045]; Ireland[0.045]; Harrington[0.043]; McGinley[0.043]; won[0.042]; Padraig[0.042]; British[0.042]; Lawrie[0.042]; Raymond[0.042]; Tour[0.042]; Tour[0.042]; Robert[0.042]; Robert[0.042]; European[0.041]; European[0.041]; Andrew[0.041]; Mark[0.040]; Colin[0.040]; Ian[0.040]; Westwood[0.040]; Stephen[0.040]; Paul[0.040]; Paul[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Stephen Ames\u001b[39m ==> ENTITY: \u001b[32mStephen Ames\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.049]; Masters[0.047]; Martin[0.046]; Trinidad[0.045]; Tour[0.045]; Tour[0.045]; won[0.044]; Westwood[0.044]; Mark[0.044]; Lawrie[0.043]; Colin[0.043]; Raymond[0.043]; Harrington[0.043]; Peter[0.043]; Coltart[0.042]; Nobilo[0.042]; Ireland[0.042]; Ireland[0.042]; Andrew[0.042]; Padraig[0.042]; Robert[0.042]; Robert[0.042]; Ian[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIreland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIreland\u001b[39m <---> \u001b[32mRepublic of Ireland\u001b[39m\t\nSCORES: global= 0.256:0.253[\u001b[31m0.003\u001b[39m]; local()= 0.155:0.134[\u001b[31m0.021\u001b[39m]; log p(e|m)= -0.566:-1.505[\u001b[31m0.939\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.054]; Australia[0.044]; Australia[0.044]; British[0.043]; British[0.043]; Italy[0.042]; Martin[0.042]; Colin[0.041]; Stephen[0.041]; won[0.041]; Russell[0.041]; Zealand[0.041]; Ian[0.041]; Zimbabwe[0.041]; Lawrie[0.041]; Padraig[0.041]; Paul[0.040]; Paul[0.040]; Paul[0.040]; Trinidad[0.040]; European[0.040]; McGinley[0.040]; Andrew[0.040]; Mark[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.051]; Ireland[0.046]; Ireland[0.046]; Martin[0.045]; Mark[0.045]; Stephen[0.044]; Colin[0.043]; Zealand[0.043]; South[0.043]; Ian[0.043]; Italy[0.043]; Andrew[0.043]; Lee[0.043]; Russell[0.043]; won[0.042]; Jonathan[0.042]; Spain[0.042]; Spain[0.042]; Masters[0.042]; Peter[0.042]; Trinidad[0.042]; Lawrie[0.042]; Zimbabwe[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Costantino Rocca\u001b[39m ==> ENTITY: \u001b[32mCostantino Rocca\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.045]; Tour[0.044]; Tour[0.044]; McGinley[0.044]; Westwood[0.044]; Masters[0.044]; Montgomerie[0.043]; Woosnam[0.043]; European[0.042]; European[0.042]; Ireland[0.041]; Ireland[0.041]; Padraig[0.041]; British[0.041]; British[0.041]; Nobilo[0.041]; Goosen[0.041]; Allenby[0.041]; Allenby[0.041]; won[0.040]; Harrington[0.040]; Lawrie[0.040]; Broadhurst[0.040]; Italy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paul McGinley\u001b[39m ==> ENTITY: \u001b[32mPaul McGinley\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.044]; Ireland[0.044]; Masters[0.044]; Padraig[0.043]; Lawrie[0.043]; Tour[0.043]; Tour[0.043]; Montgomerie[0.043]; Westwood[0.042]; won[0.042]; Harrington[0.042]; Woosnam[0.041]; Nobilo[0.041]; Mark[0.041]; Colin[0.041]; Peter[0.041]; Goosen[0.041]; South[0.040]; European[0.040]; European[0.040]; Saturday[0.040]; British[0.040]; British[0.040]; Stephen[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Robert Allenby\u001b[39m ==> ENTITY: \u001b[32mRobert Allenby\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwon[0.045]; Tour[0.044]; Tour[0.044]; Masters[0.043]; Golf[0.043]; Mark[0.042]; Lawrie[0.042]; Allenby[0.042]; Colin[0.042]; Lee[0.042]; Montgomerie[0.041]; Nobilo[0.041]; Australia[0.041]; Australia[0.041]; Westwood[0.041]; Ian[0.041]; South[0.041]; Harrington[0.041]; McGinley[0.041]; Andrew[0.041]; Raymond[0.040]; Robert[0.040]; European[0.040]; European[0.040]; \t\n\r [=======>......................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s157ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 924/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 991testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lebed\u001b[39m ==> ENTITY: \u001b[32mAlexander Lebed\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebed[0.044]; Lebed[0.044]; Lebed[0.044]; Russian[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; President[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chernomyrdin[0.042]; Chernomyrdin[0.042]; Moscow[0.041]; Moscow[0.041]; saying[0.040]; Boris[0.040]; Boris[0.040]; Interfax[0.040]; Interfax[0.040]; security[0.040]; backed[0.039]; Viktor[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yeltsin\u001b[39m ==> ENTITY: \u001b[32mBoris Yeltsin\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.045]; Russian[0.045]; Moscow[0.045]; Yeltsin[0.045]; Yeltsin[0.045]; Federation[0.042]; Chernomyrdin[0.042]; Tass[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; office[0.041]; officials[0.041]; discussed[0.041]; meeting[0.041]; press[0.040]; sent[0.040]; Lebed[0.040]; Lebed[0.040]; Lebed[0.040]; Lebed[0.040]; gave[0.040]; news[0.040]; late[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lebed\u001b[39m ==> ENTITY: \u001b[32mAlexander Lebed\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebed[0.049]; Lebed[0.049]; Lebed[0.049]; Lebed[0.049]; Russian[0.049]; military[0.048]; Yeltsin[0.048]; Yeltsin[0.048]; Yeltsin[0.048]; Russia[0.047]; Federation[0.047]; Chechnya[0.047]; Chechnya[0.047]; Chechnya[0.047]; Chechnya[0.047]; Chechnya[0.047]; Maskhadov[0.047]; Chernomyrdin[0.047]; separatist[0.047]; Moscow[0.046]; month[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yeltsin\u001b[39m ==> ENTITY: \u001b[32mBoris Yeltsin\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.045]; Russian[0.045]; Moscow[0.045]; Yeltsin[0.045]; Yeltsin[0.045]; Federation[0.042]; Chernomyrdin[0.042]; Tass[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; office[0.041]; officials[0.041]; discussed[0.041]; meeting[0.041]; press[0.040]; sent[0.040]; Lebed[0.040]; Lebed[0.040]; Lebed[0.040]; Lebed[0.040]; gave[0.040]; news[0.040]; late[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Interfax\u001b[39m ==> ENTITY: \u001b[32mInterfax\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.230:0.230[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInterfax[0.050]; news[0.049]; Moscow[0.049]; Moscow[0.049]; Russian[0.047]; agency[0.044]; agency[0.044]; quoted[0.044]; yesterday[0.043]; Maskhadov[0.042]; saying[0.042]; month[0.042]; Chernomyrdin[0.042]; Chernomyrdin[0.042]; week[0.041]; Viktor[0.041]; Prime[0.041]; Friday[0.041]; Friday[0.041]; President[0.041]; document[0.041]; Boris[0.041]; Boris[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alexander Lebed\u001b[39m ==> ENTITY: \u001b[32mAlexander Lebed\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebed[0.043]; Lebed[0.043]; Lebed[0.043]; Russian[0.043]; military[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; President[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Maskhadov[0.042]; Chernomyrdin[0.042]; Chernomyrdin[0.042]; separatist[0.041]; Moscow[0.041]; Moscow[0.041]; political[0.040]; saying[0.040]; Boris[0.040]; Boris[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lebed\u001b[39m ==> ENTITY: \u001b[32mAlexander Lebed\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebed[0.044]; Lebed[0.044]; Lebed[0.044]; Lebed[0.044]; Russian[0.043]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Russia[0.042]; Federation[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Maskhadov[0.042]; Chernomyrdin[0.042]; Moscow[0.041]; month[0.040]; political[0.040]; political[0.040]; political[0.040]; political[0.040]; saying[0.040]; troops[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Viktor Chernomyrdin\u001b[39m ==> ENTITY: \u001b[32mViktor Chernomyrdin\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChernomyrdin[0.046]; President[0.045]; Moscow[0.044]; Moscow[0.044]; Russian[0.044]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Interfax[0.042]; Interfax[0.042]; security[0.042]; Boris[0.041]; Boris[0.041]; Chechnya[0.041]; Chechnya[0.041]; Chechnya[0.041]; Chechnya[0.041]; Prime[0.040]; Minister[0.040]; news[0.040]; agency[0.039]; agency[0.039]; military[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.047]; Chechnya[0.047]; Chechnya[0.047]; Chechnya[0.047]; Chechnya[0.047]; military[0.042]; region[0.042]; region[0.042]; Russian[0.041]; Russian[0.041]; troops[0.040]; President[0.040]; separatist[0.040]; political[0.040]; political[0.040]; political[0.040]; security[0.040]; conflict[0.040]; war[0.040]; backed[0.040]; Moscow[0.040]; Moscow[0.040]; Moscow[0.040]; fighting[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yeltsin\u001b[39m ==> ENTITY: \u001b[32mBoris Yeltsin\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.051]; Russian[0.051]; Moscow[0.051]; Yeltsin[0.051]; Yeltsin[0.051]; Federation[0.048]; Chernomyrdin[0.048]; Tass[0.047]; Chechnya[0.047]; Chechnya[0.047]; Chechnya[0.047]; Chechnya[0.047]; month[0.047]; office[0.047]; officials[0.046]; discussed[0.046]; discussed[0.046]; Maskhadov[0.046]; meeting[0.046]; press[0.046]; sent[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yeltsin\u001b[39m ==> ENTITY: \u001b[32mBoris Yeltsin\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.045]; Russian[0.044]; Moscow[0.044]; Moscow[0.044]; Moscow[0.044]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Interfax[0.041]; Interfax[0.041]; Chernomyrdin[0.041]; Chernomyrdin[0.041]; Boris[0.041]; Boris[0.041]; Chechnya[0.041]; Chechnya[0.041]; Chechnya[0.041]; Chechnya[0.041]; Chechnya[0.041]; Chechnya[0.041]; Viktor[0.040]; month[0.040]; security[0.040]; Alexander[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Interfax\u001b[39m ==> ENTITY: \u001b[32mInterfax\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.235:0.235[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInterfax[0.054]; news[0.053]; Moscow[0.053]; Moscow[0.053]; Russian[0.052]; agency[0.048]; agency[0.048]; quoted[0.048]; yesterday[0.047]; Maskhadov[0.046]; saying[0.046]; month[0.046]; Chernomyrdin[0.046]; Chernomyrdin[0.046]; week[0.045]; Viktor[0.045]; Prime[0.045]; Friday[0.045]; President[0.045]; Boris[0.045]; Boris[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chernomyrdin\u001b[39m ==> ENTITY: \u001b[32mViktor Chernomyrdin\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.045]; Moscow[0.045]; Russian[0.045]; Tass[0.044]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Federation[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; political[0.042]; political[0.042]; press[0.041]; conflict[0.041]; office[0.041]; news[0.041]; independence[0.040]; agency[0.040]; officials[0.040]; territorial[0.040]; rebels[0.040]; quoted[0.039]; decision[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Boris Yeltsin\u001b[39m ==> ENTITY: \u001b[32mBoris Yeltsin\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.052]; Russian[0.050]; Moscow[0.050]; Moscow[0.050]; Yeltsin[0.050]; Yeltsin[0.050]; Yeltsin[0.050]; Interfax[0.047]; Interfax[0.047]; Chernomyrdin[0.047]; Chernomyrdin[0.047]; Boris[0.047]; Chechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; Viktor[0.046]; security[0.046]; Alexander[0.046]; backed[0.046]; discussed[0.045]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMoscow\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMoscow\u001b[39m <---> \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.257:0.238[\u001b[31m0.019\u001b[39m]; local()= 0.125:0.162[\u001b[32m0.038\u001b[39m]; log p(e|m)= -0.106:-4.269[\u001b[31m4.162\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.047]; Russian[0.045]; military[0.043]; Federation[0.042]; region[0.042]; region[0.042]; Boris[0.042]; territorial[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; fighting[0.041]; ending[0.040]; war[0.040]; Friday[0.040]; month[0.040]; Tass[0.040]; troops[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lebed\u001b[39m ==> ENTITY: \u001b[32mAlexander Lebed\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebed[0.045]; Lebed[0.045]; Lebed[0.045]; Russian[0.044]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Russia[0.043]; Federation[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chernomyrdin[0.042]; political[0.041]; saying[0.041]; troops[0.040]; Tass[0.040]; December[0.040]; quell[0.039]; rebels[0.039]; officials[0.039]; sent[0.039]; independence[0.039]; territorial[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.049]; Chechnya[0.049]; Chechnya[0.049]; Russian[0.043]; President[0.042]; security[0.042]; backed[0.042]; Moscow[0.041]; Moscow[0.041]; peace[0.041]; peace[0.041]; peace[0.041]; Interfax[0.041]; Interfax[0.041]; saying[0.041]; news[0.040]; quoted[0.040]; said[0.039]; said[0.039]; Yeltsin[0.039]; Yeltsin[0.039]; Yeltsin[0.039]; Yeltsin[0.039]; outside[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.048]; Chechnya[0.048]; Chechnya[0.048]; region[0.044]; Russian[0.042]; troops[0.042]; Russia[0.042]; political[0.041]; political[0.041]; political[0.041]; Tass[0.041]; conflict[0.041]; war[0.041]; Moscow[0.041]; peace[0.041]; independence[0.040]; saying[0.040]; month[0.040]; news[0.040]; press[0.040]; December[0.039]; officials[0.039]; rebels[0.039]; rebels[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Aslan Maskhadov\u001b[39m ==> ENTITY: \u001b[32mAslan Maskhadov\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecurity[0.044]; Chechnya[0.044]; Chechnya[0.044]; Chechnya[0.044]; Chechnya[0.044]; Chechnya[0.044]; military[0.044]; chief[0.043]; chief[0.043]; troops[0.042]; backed[0.041]; month[0.041]; conflict[0.041]; peace[0.041]; peace[0.041]; separatist[0.041]; Tass[0.041]; political[0.040]; political[0.040]; political[0.040]; fighting[0.040]; December[0.040]; war[0.040]; sent[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.047]; Chechnya[0.047]; Chechnya[0.047]; Chechnya[0.047]; military[0.043]; region[0.043]; Russian[0.042]; President[0.041]; separatist[0.041]; political[0.041]; political[0.041]; war[0.041]; backed[0.040]; Moscow[0.040]; Moscow[0.040]; fighting[0.040]; peace[0.040]; peace[0.040]; peace[0.040]; security[0.040]; Interfax[0.040]; Interfax[0.040]; saying[0.040]; month[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.142:0.142[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.050]; Russian[0.047]; military[0.045]; region[0.044]; Boris[0.044]; Boris[0.044]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Viktor[0.043]; outside[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; fighting[0.043]; ending[0.042]; Friday[0.042]; Interfax[0.042]; Interfax[0.042]; Alexander[0.042]; Nikolayevich[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFederation[0.046]; Russian[0.046]; Moscow[0.045]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; region[0.043]; region[0.043]; territorial[0.043]; month[0.041]; Tass[0.041]; late[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; officials[0.040]; agency[0.040]; bid[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; plans[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lebed\u001b[39m ==> ENTITY: \u001b[32mAlexander Lebed\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebed[0.045]; Lebed[0.045]; Lebed[0.045]; Lebed[0.045]; Russian[0.044]; Russian[0.044]; military[0.044]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; President[0.043]; Federation[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Maskhadov[0.043]; Chernomyrdin[0.043]; Chernomyrdin[0.043]; separatist[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lebed\u001b[39m ==> ENTITY: \u001b[32mAlexander Lebed\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebed[0.045]; Lebed[0.045]; Lebed[0.045]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Russia[0.043]; Federation[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chernomyrdin[0.043]; political[0.042]; saying[0.041]; troops[0.041]; Tass[0.040]; December[0.040]; quell[0.040]; rebels[0.040]; officials[0.040]; sent[0.040]; independence[0.040]; territorial[0.039]; office[0.039]; discussed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.121:0.121[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.050]; Russia[0.043]; political[0.043]; Tass[0.043]; Thursday[0.043]; Thursday[0.043]; peace[0.042]; independence[0.042]; saying[0.042]; news[0.042]; officials[0.041]; December[0.041]; rebels[0.041]; territorial[0.041]; said[0.041]; said[0.041]; said[0.041]; quoted[0.041]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; press[0.040]; office[0.039]; Chernomyrdin[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.047]; Chechnya[0.047]; Chechnya[0.047]; military[0.043]; region[0.043]; region[0.043]; Russian[0.042]; troops[0.041]; Russia[0.041]; separatist[0.041]; political[0.041]; political[0.041]; political[0.041]; political[0.041]; Tass[0.041]; conflict[0.041]; war[0.041]; Moscow[0.040]; fighting[0.040]; peace[0.040]; Maskhadov[0.040]; independence[0.040]; saying[0.040]; month[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chernomyrdin\u001b[39m ==> ENTITY: \u001b[32mViktor Chernomyrdin\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChernomyrdin[0.046]; President[0.044]; Moscow[0.044]; Moscow[0.044]; Russian[0.044]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Interfax[0.042]; Interfax[0.042]; security[0.041]; Viktor[0.041]; Boris[0.041]; Boris[0.041]; Chechnya[0.041]; Chechnya[0.041]; Chechnya[0.041]; Chechnya[0.041]; Chechnya[0.041]; political[0.041]; political[0.041]; Maskhadov[0.040]; Prime[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.048]; Chechnya[0.048]; Chechnya[0.048]; military[0.043]; region[0.043]; Russian[0.042]; President[0.041]; separatist[0.041]; political[0.041]; war[0.041]; backed[0.041]; Moscow[0.041]; Moscow[0.041]; fighting[0.041]; peace[0.040]; peace[0.040]; peace[0.040]; security[0.040]; Interfax[0.040]; Interfax[0.040]; saying[0.040]; month[0.040]; Maskhadov[0.040]; rebels[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lebed\u001b[39m ==> ENTITY: \u001b[32mAlexander Lebed\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebed[0.044]; Lebed[0.044]; Lebed[0.044]; Lebed[0.044]; Russian[0.043]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Russia[0.042]; Federation[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chernomyrdin[0.042]; Moscow[0.041]; political[0.041]; political[0.041]; political[0.041]; saying[0.040]; troops[0.040]; Tass[0.039]; December[0.039]; quell[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yeltsin\u001b[39m ==> ENTITY: \u001b[32mBoris Yeltsin\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.046]; Russian[0.044]; Moscow[0.044]; Moscow[0.044]; Yeltsin[0.044]; Yeltsin[0.044]; Yeltsin[0.044]; Interfax[0.042]; Interfax[0.042]; Chernomyrdin[0.041]; Chernomyrdin[0.041]; Boris[0.041]; Boris[0.041]; Chechnya[0.041]; Chechnya[0.041]; Chechnya[0.041]; Viktor[0.041]; security[0.040]; Alexander[0.040]; backed[0.040]; Lebed[0.039]; Lebed[0.039]; Lebed[0.039]; Lebed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.142:0.142[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.048]; Russian[0.046]; military[0.043]; Boris[0.043]; Boris[0.043]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Viktor[0.042]; outside[0.042]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Friday[0.041]; Interfax[0.041]; Interfax[0.041]; Alexander[0.040]; Nikolayevich[0.040]; Lebed[0.040]; Lebed[0.040]; Lebed[0.040]; Lebed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.047]; Chechnya[0.047]; Chechnya[0.047]; Chechnya[0.047]; military[0.043]; region[0.043]; region[0.043]; Russian[0.041]; troops[0.041]; Russia[0.041]; separatist[0.040]; political[0.040]; political[0.040]; political[0.040]; political[0.040]; Tass[0.040]; conflict[0.040]; war[0.040]; Moscow[0.040]; fighting[0.040]; peace[0.040]; Maskhadov[0.040]; Interfax[0.040]; independence[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Itar-Tass\u001b[39m ==> ENTITY: \u001b[32mInformation Telegraph Agency of Russia\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nagency[0.045]; news[0.045]; Tass[0.044]; press[0.044]; Russian[0.044]; Russia[0.044]; Federation[0.044]; Moscow[0.043]; document[0.043]; officials[0.042]; office[0.042]; saying[0.042]; quoted[0.040]; Chernomyrdin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; staff[0.040]; political[0.040]; political[0.040]; political[0.040]; political[0.040]; day[0.040]; rebel[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yeltsin\u001b[39m ==> ENTITY: \u001b[32mBoris Yeltsin\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.045]; Russian[0.043]; Russian[0.043]; Moscow[0.043]; Moscow[0.043]; Moscow[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Interfax[0.041]; Interfax[0.041]; Federation[0.041]; Chernomyrdin[0.041]; Chernomyrdin[0.041]; Boris[0.041]; Boris[0.041]; Chechnya[0.040]; Chechnya[0.040]; Chechnya[0.040]; Chechnya[0.040]; Chechnya[0.040]; Chechnya[0.040]; Viktor[0.040]; month[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lebed\u001b[39m ==> ENTITY: \u001b[32mAlexander Lebed\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebed[0.045]; Lebed[0.045]; Lebed[0.045]; Lebed[0.045]; Russian[0.045]; military[0.044]; Yeltsin[0.044]; Yeltsin[0.044]; Yeltsin[0.044]; Yeltsin[0.044]; President[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Maskhadov[0.043]; Chernomyrdin[0.043]; Chernomyrdin[0.043]; separatist[0.042]; Moscow[0.042]; Moscow[0.042]; month[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Russian Federation\u001b[39m ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.050]; Moscow[0.048]; Chernomyrdin[0.048]; Yeltsin[0.047]; Yeltsin[0.047]; Yeltsin[0.047]; Yeltsin[0.047]; Chechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; month[0.046]; region[0.045]; region[0.045]; territorial[0.045]; office[0.045]; Tass[0.043]; meeting[0.043]; military[0.043]; press[0.043]; bid[0.043]; chief[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.151:0.151[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.045]; Moscow[0.045]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Boris[0.043]; Boris[0.043]; Interfax[0.042]; Interfax[0.042]; Viktor[0.042]; Alexander[0.042]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Nikolayevich[0.040]; President[0.040]; Lebed[0.040]; Lebed[0.040]; Lebed[0.040]; Lebed[0.040]; military[0.040]; separatist[0.040]; \t\n\r [========>.....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s980ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 961/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1083testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Switzerland ==> ENTITY: \u001b[32mSwitzerland national football team\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.157:0.157[0]; log p(e|m)= -2.645:-2.645[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.056]; match[0.054]; soccer[0.053]; Qualifier[0.052]; qualifying[0.051]; Cup[0.051]; Cup[0.051]; Soccer[0.051]; European[0.049]; World[0.047]; World[0.047]; beat[0.047]; Saturday[0.045]; Scorer[0.044]; Azerbaijan[0.044]; Azerbaijan[0.044]; group[0.044]; halftime[0.044]; Attendance[0.043]; Beat[0.041]; Baku[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBaku\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBaku FC\u001b[39m <---> \u001b[32mBaku\u001b[39m\t\nSCORES: global= 0.256:0.255[\u001b[31m0.001\u001b[39m]; local()= 0.125:0.073[\u001b[31m0.052\u001b[39m]; log p(e|m)= -2.538:-0.064[\u001b[32m2.474\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAzerbaijan[0.050]; Azerbaijan[0.050]; soccer[0.047]; Cup[0.046]; Cup[0.046]; match[0.044]; qualifying[0.044]; Rzayev[0.044]; European[0.043]; Soccer[0.043]; Saturday[0.043]; Attendance[0.043]; Switzerland[0.042]; Switzerland[0.042]; Qualifier[0.042]; Vidadi[0.042]; World[0.042]; World[0.042]; beat[0.042]; group[0.041]; Scorer[0.041]; Beat[0.040]; halftime[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEuropean\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUEFA\u001b[39m <---> \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.243:0.240[\u001b[31m0.003\u001b[39m]; local()= 0.134:0.053[\u001b[31m0.081\u001b[39m]; log p(e|m)= -3.576:-0.256[\u001b[32m3.319\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.053]; Soccer[0.053]; qualifying[0.051]; match[0.051]; Cup[0.051]; Cup[0.051]; Qualifier[0.050]; Switzerland[0.049]; Switzerland[0.049]; Azerbaijan[0.048]; Azerbaijan[0.048]; group[0.048]; beat[0.046]; Saturday[0.046]; World[0.045]; World[0.045]; Baku[0.045]; Attendance[0.044]; Scorer[0.043]; Beat[0.042]; halftime[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Vidadi Rzayev\u001b[39m ==> ENTITY: \u001b[32mVidadi Rzayev\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.052]; Azerbaijan[0.052]; Azerbaijan[0.052]; qualifying[0.050]; Qualifier[0.050]; match[0.050]; Switzerland[0.050]; Switzerland[0.050]; Soccer[0.049]; Baku[0.048]; European[0.048]; beat[0.048]; halftime[0.046]; Attendance[0.046]; Cup[0.046]; Cup[0.046]; Scorer[0.045]; Beat[0.043]; Saturday[0.043]; World[0.042]; World[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Switzerland ==> ENTITY: \u001b[32mSwitzerland national football team\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.157:0.157[0]; log p(e|m)= -2.645:-2.645[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.056]; match[0.054]; soccer[0.053]; Qualifier[0.052]; qualifying[0.051]; Cup[0.051]; Cup[0.051]; Soccer[0.051]; European[0.049]; World[0.047]; World[0.047]; beat[0.047]; Saturday[0.045]; Scorer[0.044]; Azerbaijan[0.044]; Azerbaijan[0.044]; group[0.044]; halftime[0.044]; Attendance[0.043]; Beat[0.041]; Baku[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Azerbaijan ==> ENTITY: \u001b[32mAzerbaijan national football team\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.143:0.143[0]; log p(e|m)= -2.617:-2.617[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAzerbaijan[0.051]; Soccer[0.048]; soccer[0.047]; match[0.047]; qualifying[0.046]; World[0.046]; World[0.046]; Baku[0.045]; Qualifier[0.044]; Vidadi[0.044]; European[0.043]; Switzerland[0.043]; Switzerland[0.043]; Rzayev[0.043]; Saturday[0.042]; Cup[0.042]; Cup[0.042]; group[0.042]; halftime[0.040]; beat[0.040]; Beat[0.039]; Attendance[0.038]; Scorer[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Azerbaijan ==> ENTITY: \u001b[32mAzerbaijan national football team\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.143:0.143[0]; log p(e|m)= -2.617:-2.617[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAzerbaijan[0.051]; Soccer[0.048]; soccer[0.047]; match[0.047]; qualifying[0.046]; World[0.046]; World[0.046]; Baku[0.045]; Qualifier[0.044]; Vidadi[0.044]; European[0.043]; Switzerland[0.043]; Switzerland[0.043]; Rzayev[0.043]; Saturday[0.042]; Cup[0.042]; Cup[0.042]; group[0.042]; halftime[0.040]; beat[0.040]; Beat[0.039]; Attendance[0.038]; Scorer[0.038]; \t\n\r [========>.....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s57ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 970/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1035testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = El Al\u001b[39m ==> ENTITY: \u001b[32mEl Al\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.126:0.126[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \naircraft[0.047]; Aviv[0.046]; crash[0.043]; transport[0.043]; transport[0.043]; freighter[0.043]; El[0.042]; El[0.042]; Al[0.042]; Al[0.042]; said[0.042]; pressure[0.041]; office[0.041]; Tel[0.041]; suburb[0.041]; Amsterdam[0.041]; Annemarie[0.040]; reopen[0.040]; say[0.040]; investigate[0.040]; information[0.040]; country[0.040]; second[0.040]; head[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = El Al\u001b[39m ==> ENTITY: \u001b[32mEl Al\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \naircraft[0.046]; aircraft[0.046]; aircraft[0.046]; Aviv[0.045]; Air[0.043]; crash[0.042]; transport[0.042]; transport[0.042]; freighter[0.041]; El[0.041]; El[0.041]; El[0.041]; Al[0.041]; Al[0.041]; Al[0.041]; Cargo[0.041]; dangerous[0.041]; said[0.040]; said[0.040]; said[0.040]; pressure[0.040]; office[0.040]; Tel[0.040]; Tel[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tel Aviv\u001b[39m ==> ENTITY: \u001b[32mTel Aviv\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.111:0.111[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTel[0.048]; aircraft[0.048]; aircraft[0.048]; aircraft[0.048]; Air[0.047]; Al[0.046]; Al[0.046]; Al[0.046]; Al[0.046]; Cargo[0.046]; incident[0.045]; office[0.045]; reopen[0.045]; transport[0.045]; transport[0.045]; second[0.045]; crash[0.044]; carrying[0.044]; said[0.044]; said[0.044]; said[0.044]; suburb[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = El Al\u001b[39m ==> ENTITY: \u001b[32mEl Al\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \naircraft[0.046]; aircraft[0.046]; aircraft[0.046]; Aviv[0.045]; crash[0.042]; transport[0.042]; transport[0.042]; freighter[0.041]; El[0.041]; El[0.041]; El[0.041]; Al[0.041]; Al[0.041]; Al[0.041]; dangerous[0.041]; said[0.040]; said[0.040]; said[0.040]; pressure[0.040]; office[0.040]; Tel[0.040]; suburb[0.040]; Amsterdam[0.040]; carrying[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dutch ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.794:-0.794[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDutch[0.049]; Dutch[0.049]; Dutch[0.049]; country[0.046]; Amsterdam[0.043]; freighter[0.042]; aircraft[0.041]; aircraft[0.041]; aircraft[0.041]; Air[0.041]; parliaments[0.041]; Hague[0.041]; second[0.041]; believed[0.040]; office[0.040]; said[0.040]; said[0.040]; said[0.040]; come[0.040]; documentation[0.039]; thought[0.039]; carrying[0.039]; reopen[0.039]; latest[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dutch ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.794:-0.794[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDutch[0.051]; Dutch[0.051]; Dutch[0.051]; country[0.047]; Amsterdam[0.045]; Hague[0.042]; second[0.042]; freighter[0.041]; said[0.041]; come[0.041]; office[0.041]; say[0.040]; reopen[0.040]; documentation[0.040]; transport[0.039]; transport[0.039]; reason[0.039]; hold[0.039]; Jorritsma[0.039]; pressure[0.039]; left[0.039]; section[0.039]; suburb[0.039]; need[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dutch ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.794:-0.794[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDutch[0.050]; Dutch[0.050]; Dutch[0.050]; country[0.047]; Amsterdam[0.044]; parliaments[0.042]; Hague[0.041]; second[0.041]; believed[0.041]; office[0.041]; freighter[0.041]; said[0.040]; hold[0.040]; come[0.040]; head[0.040]; reopen[0.039]; documentation[0.039]; carrying[0.039]; say[0.039]; transport[0.039]; transport[0.039]; Jorritsma[0.039]; pressure[0.039]; members[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dutch ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.794:-0.794[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDutch[0.050]; Dutch[0.050]; Dutch[0.050]; country[0.046]; Amsterdam[0.044]; parliaments[0.041]; freighter[0.041]; Hague[0.041]; second[0.041]; believed[0.041]; office[0.041]; said[0.040]; said[0.040]; said[0.040]; come[0.040]; documentation[0.040]; thought[0.040]; added[0.039]; reopen[0.039]; goods[0.039]; investigate[0.039]; transport[0.039]; transport[0.039]; say[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Amsterdam ==> ENTITY: \u001b[32mAmsterdam\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.124:0.124[0]; log p(e|m)= -0.064:-0.064[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDutch[0.044]; Dutch[0.044]; Dutch[0.044]; Dutch[0.044]; aircraft[0.043]; aircraft[0.043]; aircraft[0.043]; transport[0.043]; transport[0.043]; office[0.042]; section[0.042]; reopen[0.042]; country[0.042]; airfreight[0.041]; freighter[0.041]; suburb[0.041]; way[0.040]; Hague[0.040]; told[0.040]; cross[0.039]; said[0.039]; said[0.039]; said[0.039]; left[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = The Hague\u001b[39m ==> ENTITY: \u001b[32mThe Hague\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.115:0.115[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDutch[0.045]; Dutch[0.045]; Dutch[0.045]; Dutch[0.045]; second[0.044]; Amsterdam[0.044]; country[0.043]; reopen[0.043]; transport[0.043]; transport[0.043]; Jorritsma[0.041]; parliaments[0.041]; office[0.041]; section[0.040]; Al[0.040]; Al[0.040]; Al[0.040]; Tel[0.040]; suburb[0.040]; chamber[0.039]; Annemarie[0.039]; cross[0.039]; said[0.039]; left[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jorritsma\u001b[39m ==> ENTITY: \u001b[32mAnnemarie Jorritsma\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.090:0.090[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDutch[0.046]; Dutch[0.046]; ministry[0.043]; ministry[0.043]; Newsroom[0.043]; office[0.043]; Tel[0.042]; Tel[0.042]; Amsterdam[0.042]; airfreight[0.041]; head[0.041]; transport[0.041]; Fax[0.041]; aircraft[0.041]; aircraft[0.041]; aircraft[0.041]; said[0.041]; said[0.041]; said[0.041]; suburb[0.040]; added[0.040]; Cargo[0.040]; parliaments[0.040]; Air[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Annemarie Jorritsma\u001b[39m ==> ENTITY: \u001b[32mAnnemarie Jorritsma\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.135:0.135[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDutch[0.046]; Dutch[0.046]; Dutch[0.046]; Dutch[0.046]; Hague[0.044]; minister[0.043]; ministry[0.042]; ministry[0.042]; office[0.042]; country[0.042]; Tel[0.042]; Amsterdam[0.041]; head[0.041]; transport[0.040]; transport[0.040]; aircraft[0.040]; aircraft[0.040]; said[0.040]; chamber[0.040]; suburb[0.040]; parliaments[0.039]; freighter[0.039]; reopen[0.039]; information[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = El Al\u001b[39m ==> ENTITY: \u001b[32mEl Al\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \naircraft[0.046]; aircraft[0.046]; aircraft[0.046]; Aviv[0.045]; Air[0.043]; transport[0.042]; freighter[0.042]; El[0.041]; El[0.041]; Al[0.041]; Al[0.041]; Cargo[0.041]; dangerous[0.041]; said[0.041]; said[0.041]; said[0.041]; pressure[0.040]; office[0.040]; Tel[0.040]; Tel[0.040]; suburb[0.040]; Newsroom[0.040]; Amsterdam[0.040]; carrying[0.040]; \t\n\r [========>.....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s110ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 983/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1128testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.083:0.083[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsigned[0.044]; field[0.043]; peace[0.042]; peace[0.042]; Saturday[0.042]; Saturday[0.042]; deal[0.042]; deal[0.042]; thousand[0.042]; town[0.042]; Chechens[0.042]; Chechens[0.042]; independence[0.042]; withdraw[0.042]; crowd[0.042]; celebrations[0.041]; uncertainty[0.041]; north[0.041]; pro[0.041]; held[0.041]; gathered[0.041]; miles[0.040]; fired[0.040]; Grozny[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.116:0.116[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.048]; Russian[0.048]; military[0.044]; military[0.044]; weapons[0.044]; region[0.043]; Chechnya[0.042]; Alexander[0.042]; officials[0.042]; Chechen[0.042]; Caucasus[0.041]; troops[0.041]; North[0.041]; Troops[0.040]; Lebed[0.040]; war[0.040]; early[0.040]; peace[0.040]; peace[0.040]; women[0.040]; support[0.039]; independence[0.039]; withdraw[0.039]; December[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Islamic ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.212:-0.212[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIslamic[0.050]; law[0.044]; law[0.044]; law[0.044]; impose[0.042]; Allahu[0.042]; strict[0.042]; Akhbar[0.042]; war[0.042]; war[0.042]; war[0.042]; Chechnya[0.042]; Freedom[0.041]; peace[0.041]; conflict[0.040]; Chechen[0.040]; said[0.040]; said[0.040]; God[0.040]; God[0.040]; rebel[0.040]; rebels[0.040]; believe[0.040]; carrying[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = God ==> ENTITY: \u001b[32mGod\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.142:0.142[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGod[0.052]; sons[0.044]; Islamic[0.043]; Islamic[0.043]; believe[0.043]; Allahu[0.042]; peace[0.042]; peace[0.042]; mother[0.042]; mother[0.042]; places[0.042]; said[0.042]; said[0.042]; united[0.041]; people[0.041]; joked[0.041]; come[0.041]; works[0.040]; like[0.040]; outside[0.040]; town[0.040]; warned[0.040]; sister[0.040]; sister[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZavgayev[0.045]; Russian[0.044]; troops[0.043]; people[0.043]; separatist[0.043]; separatists[0.043]; war[0.042]; war[0.042]; war[0.042]; backed[0.042]; Moscow[0.042]; peace[0.042]; peace[0.042]; support[0.042]; town[0.041]; civil[0.040]; rebels[0.040]; rebels[0.040]; said[0.040]; said[0.040]; said[0.040]; welcomed[0.040]; united[0.040]; leader[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechen[0.045]; military[0.044]; military[0.044]; region[0.044]; Caucasus[0.043]; Russian[0.042]; Russian[0.042]; Russian[0.042]; troops[0.042]; people[0.042]; separatists[0.041]; separatists[0.041]; conflict[0.041]; war[0.041]; Moslem[0.041]; groups[0.041]; declared[0.041]; peace[0.041]; peace[0.041]; support[0.041]; Dudayev[0.040]; independence[0.040]; Grozny[0.040]; withdrawal[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.115:0.115[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.048]; Russian[0.048]; military[0.044]; military[0.044]; weapons[0.044]; region[0.043]; Chechnya[0.042]; Alexander[0.042]; officials[0.042]; Chechen[0.042]; Caucasus[0.041]; troops[0.041]; North[0.041]; Troops[0.040]; Lebed[0.040]; war[0.040]; early[0.040]; peace[0.040]; peace[0.040]; women[0.040]; women[0.040]; support[0.039]; thousand[0.039]; independence[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = God ==> ENTITY: \u001b[32mGod\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGod[0.053]; believe[0.044]; Islamic[0.044]; Islamic[0.044]; Allahu[0.043]; said[0.042]; said[0.042]; united[0.042]; peace[0.042]; outside[0.042]; joked[0.041]; come[0.041]; works[0.041]; strict[0.041]; man[0.040]; black[0.040]; Akhbar[0.040]; law[0.040]; law[0.040]; law[0.040]; welcomed[0.040]; responded[0.040]; columns[0.039]; impose[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Doku Zavgayev\u001b[39m ==> ENTITY: \u001b[32mDoku Zavgayev\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.082:0.082[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.047]; Russian[0.045]; Moscow[0.044]; separatists[0.042]; troops[0.042]; separatist[0.042]; leader[0.042]; rebels[0.042]; rebels[0.042]; war[0.041]; war[0.041]; war[0.041]; Martan[0.041]; Martan[0.041]; backed[0.041]; really[0.041]; really[0.041]; rushed[0.041]; air[0.041]; Urus[0.041]; Urus[0.041]; support[0.041]; pocket[0.041]; sidelined[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = North Caucasus\u001b[39m ==> ENTITY: \u001b[32mNorth Caucasus\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nregion[0.049]; Chechnya[0.046]; Chechen[0.044]; separatists[0.043]; separatists[0.043]; Russian[0.043]; Russian[0.043]; Russian[0.043]; military[0.043]; military[0.043]; Moslem[0.042]; Dudayev[0.041]; troops[0.041]; conflict[0.041]; officials[0.040]; people[0.040]; withdrawal[0.040]; mainly[0.040]; children[0.040]; Troops[0.040]; declared[0.039]; war[0.039]; war[0.039]; December[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Urus Martan\u001b[39m ==> ENTITY: \u001b[32mUrus-Martan\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMartan[0.047]; Martan[0.047]; Grozny[0.045]; Urus[0.045]; town[0.044]; Chechens[0.043]; Chechens[0.043]; rebels[0.043]; held[0.043]; groups[0.041]; capital[0.041]; air[0.041]; air[0.041]; field[0.041]; north[0.041]; Russia[0.041]; celebrations[0.040]; gathered[0.040]; fired[0.039]; Adults[0.039]; km[0.039]; children[0.039]; wait[0.039]; firing[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dzhokhar Dudayev\u001b[39m ==> ENTITY: \u001b[32mDzhokhar Dudayev\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.044]; military[0.044]; Chechnya[0.044]; Chechens[0.043]; Chechen[0.042]; separatists[0.042]; separatists[0.042]; Russian[0.042]; Russian[0.042]; Russian[0.042]; Caucasus[0.042]; leader[0.042]; troops[0.041]; conflict[0.041]; Martan[0.041]; Lebed[0.041]; independence[0.041]; independence[0.041]; envoy[0.040]; declared[0.040]; Grozny[0.040]; fired[0.040]; sent[0.040]; officials[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.097:0.097[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.048]; Chechnya[0.044]; troops[0.043]; late[0.043]; Zavgayev[0.043]; war[0.042]; war[0.042]; peace[0.042]; peace[0.042]; guns[0.042]; people[0.041]; town[0.041]; firing[0.041]; ended[0.041]; united[0.041]; outside[0.041]; civil[0.041]; leader[0.041]; backed[0.040]; said[0.040]; said[0.040]; said[0.040]; woman[0.040]; separatist[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dudayeva\u001b[39m ==> ENTITY: \u001b[32mAlla Dudayeva\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.093:0.093[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.046]; Russian[0.044]; Chechnya[0.044]; Caucasus[0.043]; Zavgayev[0.042]; leader[0.042]; rebels[0.042]; rebels[0.042]; old[0.042]; old[0.042]; separatists[0.041]; really[0.041]; really[0.041]; troops[0.041]; woman[0.041]; rushed[0.041]; separatist[0.041]; air[0.041]; Urus[0.041]; Urus[0.041]; North[0.041]; pocket[0.041]; late[0.041]; leave[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alexander Lebed\u001b[39m ==> ENTITY: \u001b[32mAlexander Lebed\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.142:0.142[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.045]; Russian[0.045]; Russian[0.045]; Dudayev[0.045]; military[0.044]; military[0.044]; Chechnya[0.043]; Caucasus[0.042]; Chechen[0.042]; troops[0.041]; leader[0.041]; signed[0.041]; separatists[0.041]; separatists[0.041]; war[0.041]; envoy[0.040]; carrying[0.040]; December[0.040]; conflict[0.040]; rebels[0.040]; rebels[0.040]; rebels[0.040]; officials[0.040]; sent[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Islamic ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.212:-0.212[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIslamic[0.050]; God[0.044]; God[0.044]; impose[0.042]; Allahu[0.042]; Akhbar[0.042]; believe[0.042]; war[0.042]; war[0.042]; war[0.042]; Chechnya[0.042]; strict[0.042]; Freedom[0.041]; law[0.041]; law[0.041]; law[0.041]; peace[0.040]; conflict[0.040]; Chechen[0.040]; late[0.040]; said[0.040]; said[0.040]; rebel[0.040]; rebels[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chechen ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.146:0.146[0]; log p(e|m)= -0.863:-0.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.050]; people[0.043]; separatists[0.043]; conflict[0.043]; war[0.043]; war[0.043]; war[0.043]; Islamic[0.043]; Islamic[0.043]; peace[0.042]; adopted[0.042]; rebels[0.041]; rebels[0.041]; said[0.041]; said[0.041]; outside[0.041]; welcomed[0.041]; united[0.040]; rebel[0.040]; hold[0.040]; Saturday[0.040]; late[0.039]; town[0.039]; Freedom[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Grozny\u001b[39m ==> ENTITY: \u001b[32mGrozny\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.044]; military[0.044]; Chechnya[0.043]; Chechen[0.043]; Chechens[0.043]; Chechens[0.043]; Russian[0.042]; Russian[0.042]; Russian[0.042]; troops[0.042]; Troops[0.042]; weapons[0.042]; north[0.041]; miles[0.041]; officials[0.041]; rebels[0.041]; rebels[0.041]; km[0.041]; Dudayev[0.041]; Russia[0.041]; capital[0.040]; carried[0.040]; firing[0.040]; carrying[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Urus-Martan\u001b[39m ==> ENTITY: \u001b[32mUrus-Martan\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMartan[0.048]; Martan[0.048]; Zavgayev[0.047]; Chechnya[0.046]; separatists[0.046]; Urus[0.046]; town[0.045]; separatist[0.044]; rebels[0.044]; rebels[0.044]; carrying[0.043]; civil[0.043]; Russian[0.042]; region[0.042]; really[0.042]; really[0.042]; Moscow[0.042]; rushed[0.042]; adding[0.042]; air[0.042]; pocket[0.041]; turned[0.041]; North[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.114:0.114[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.048]; Russian[0.048]; military[0.044]; military[0.044]; weapons[0.044]; region[0.043]; Chechnya[0.042]; Alexander[0.042]; officials[0.042]; Chechen[0.042]; Caucasus[0.041]; troops[0.041]; North[0.041]; Troops[0.040]; Lebed[0.040]; war[0.040]; early[0.040]; peace[0.040]; women[0.040]; women[0.040]; guns[0.040]; support[0.039]; thousand[0.039]; independence[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechen[0.047]; separatists[0.044]; conflict[0.044]; war[0.044]; war[0.044]; war[0.044]; peace[0.043]; Islamic[0.042]; Islamic[0.042]; rebels[0.042]; rebels[0.042]; said[0.041]; said[0.041]; welcomed[0.041]; adopted[0.041]; united[0.041]; rebel[0.041]; joked[0.040]; Freedom[0.040]; ended[0.040]; responded[0.040]; late[0.040]; hold[0.039]; impose[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Urus-Martan\u001b[39m ==> ENTITY: \u001b[32mUrus-Martan\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMartan[0.048]; Martan[0.048]; Zavgayev[0.047]; Chechnya[0.046]; separatists[0.046]; Urus[0.046]; town[0.045]; separatist[0.044]; rebels[0.044]; rebels[0.044]; carrying[0.043]; civil[0.043]; Russian[0.042]; region[0.042]; really[0.042]; really[0.042]; Moscow[0.042]; rushed[0.042]; adding[0.042]; air[0.042]; pocket[0.041]; turned[0.041]; North[0.041]; \t\n\r [========>.....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s77ms | Step: 3ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1005/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1075testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nouakchott\u001b[39m ==> ENTITY: \u001b[32mNouakchott\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.137:0.137[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMauritania[0.048]; Mauritania[0.048]; Mauritania[0.048]; domestic[0.044]; Mohamed[0.043]; leg[0.043]; Friday[0.043]; Lemine[0.042]; held[0.041]; African[0.041]; country[0.041]; Saturday[0.041]; national[0.040]; season[0.040]; eliminated[0.040]; Africans[0.040]; Exit[0.040]; failure[0.040]; North[0.040]; National[0.040]; draw[0.039]; Soccer[0.039]; federation[0.039]; federation[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mauritania ==> ENTITY: \u001b[32mMauritania\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.373:-0.373[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMauritania[0.047]; Mauritania[0.047]; national[0.045]; country[0.044]; federation[0.044]; federation[0.044]; team[0.044]; soccer[0.043]; Cup[0.042]; Cup[0.042]; Nations[0.042]; African[0.041]; Benin[0.041]; Soccer[0.041]; qualifying[0.041]; Nouakchott[0.041]; Africans[0.040]; Mohamed[0.040]; held[0.039]; Team[0.039]; qualify[0.039]; domestic[0.039]; eliminated[0.039]; Saturday[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMauritania\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMauritania\u001b[39m <---> \u001b[32mMauritania national football team\u001b[39m\t\nSCORES: global= 0.274:0.241[\u001b[31m0.032\u001b[39m]; local()= 0.141:0.188[\u001b[32m0.046\u001b[39m]; log p(e|m)= -0.373:-2.146[\u001b[31m1.773\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMauritania[0.047]; Mauritania[0.047]; country[0.044]; team[0.044]; national[0.043]; soccer[0.043]; Cup[0.042]; Cup[0.042]; Nations[0.042]; African[0.042]; Benin[0.042]; Soccer[0.041]; federation[0.041]; federation[0.041]; qualifying[0.041]; Nouakchott[0.041]; Africans[0.040]; Mohamed[0.040]; commitments[0.040]; president[0.039]; Team[0.039]; qualify[0.039]; domestic[0.039]; eliminated[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBenin\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBenin\u001b[39m <---> \u001b[32mBenin national football team\u001b[39m\t\nSCORES: global= 0.264:0.224[\u001b[31m0.039\u001b[39m]; local()= 0.139:0.190[\u001b[32m0.051\u001b[39m]; log p(e|m)= -0.420:-2.163[\u001b[31m1.743\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.046]; Soccer[0.045]; team[0.044]; soccer[0.044]; federation[0.043]; federation[0.043]; national[0.043]; president[0.043]; Cup[0.042]; Cup[0.042]; National[0.042]; suspended[0.041]; African[0.041]; season[0.041]; Nations[0.041]; qualifying[0.040]; Mauritania[0.040]; Mauritania[0.040]; Mauritania[0.040]; North[0.040]; Team[0.040]; held[0.039]; dissolved[0.039]; qualify[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMauritania\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMauritania\u001b[39m <---> \u001b[32mMauritania national football team\u001b[39m\t\nSCORES: global= 0.275:0.246[\u001b[31m0.029\u001b[39m]; local()= 0.143:0.204[\u001b[32m0.060\u001b[39m]; log p(e|m)= -0.373:-2.146[\u001b[31m1.773\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMauritania[0.047]; Mauritania[0.047]; country[0.044]; team[0.044]; national[0.043]; soccer[0.043]; Cup[0.042]; Cup[0.042]; Nations[0.042]; African[0.042]; Benin[0.042]; Soccer[0.041]; Africans[0.041]; federation[0.041]; federation[0.041]; qualifying[0.041]; Nouakchott[0.041]; Mohamed[0.040]; Team[0.039]; qualify[0.039]; domestic[0.039]; eliminated[0.039]; Saturday[0.039]; season[0.039]; \t\n\r [========>.....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s218ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1010/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1136testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Suu Kyi\u001b[39m ==> ENTITY: \u001b[32mAung San Suu Kyi\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurma[0.047]; Burma[0.047]; military[0.043]; military[0.043]; government[0.042]; democracy[0.042]; democracy[0.042]; democracy[0.042]; jailing[0.042]; Kyi[0.042]; Kyi[0.042]; leader[0.041]; activists[0.041]; Rangoon[0.041]; Rangoon[0.041]; reporters[0.041]; Aung[0.041]; Suu[0.041]; Suu[0.041]; world[0.040]; conference[0.040]; press[0.039]; movement[0.039]; movement[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burma ==> ENTITY: \u001b[32mBurma\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.202:-0.202[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurma[0.048]; government[0.047]; movement[0.043]; movement[0.043]; repression[0.042]; said[0.042]; said[0.042]; Kyi[0.042]; Kyi[0.042]; August[0.042]; party[0.042]; democracy[0.042]; democracy[0.042]; increased[0.041]; efforts[0.041]; serving[0.041]; early[0.041]; arrests[0.040]; close[0.040]; stop[0.040]; Suu[0.040]; Suu[0.040]; came[0.040]; members[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Insein Prison\u001b[39m ==> ENTITY: \u001b[32mInsein Prison\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \njail[0.046]; Rangoon[0.046]; torture[0.046]; activists[0.046]; Insein[0.046]; Insein[0.046]; sleep[0.046]; prisoners[0.046]; prisoners[0.046]; prisoners[0.046]; prison[0.046]; prison[0.046]; prison[0.046]; prison[0.046]; prison[0.046]; held[0.045]; Kyi[0.045]; Kyi[0.045]; Kyi[0.045]; rights[0.044]; August[0.044]; conditions[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Suu Kyi\u001b[39m ==> ENTITY: \u001b[32mAung San Suu Kyi\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurma[0.047]; Burmese[0.045]; military[0.043]; government[0.042]; government[0.042]; government[0.042]; government[0.042]; Democracy[0.042]; democracy[0.042]; democracy[0.042]; National[0.042]; Kyi[0.041]; Kyi[0.041]; Kyi[0.041]; peace[0.041]; activists[0.041]; activists[0.041]; supporters[0.041]; Suu[0.040]; Suu[0.040]; Suu[0.040]; party[0.040]; party[0.040]; state[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Suu Kyi\u001b[39m ==> ENTITY: \u001b[32mAung San Suu Kyi\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurmese[0.045]; military[0.043]; military[0.043]; government[0.042]; government[0.042]; government[0.042]; Democracy[0.042]; democracy[0.042]; democracy[0.042]; democracy[0.042]; democracy[0.042]; democracy[0.042]; National[0.041]; Kyi[0.041]; Kyi[0.041]; Kyi[0.041]; Kyi[0.041]; peace[0.041]; activists[0.041]; activists[0.041]; activists[0.041]; Rangoon[0.041]; supporters[0.041]; supporters[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Suu Kyi\u001b[39m ==> ENTITY: \u001b[32mAung San Suu Kyi\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurma[0.050]; Burma[0.050]; government[0.045]; democracy[0.045]; democracy[0.045]; political[0.045]; Kyi[0.044]; Kyi[0.044]; Suu[0.043]; Suu[0.043]; party[0.043]; parliament[0.043]; serving[0.042]; arrest[0.042]; prisoners[0.042]; movement[0.042]; movement[0.042]; released[0.042]; popularity[0.042]; said[0.041]; said[0.041]; said[0.041]; member[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Suu Kyi\u001b[39m ==> ENTITY: \u001b[32mAung San Suu Kyi\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurma[0.046]; Burma[0.046]; Burmese[0.044]; military[0.042]; military[0.042]; military[0.042]; government[0.041]; government[0.041]; government[0.041]; Democracy[0.041]; democracy[0.041]; democracy[0.041]; democracy[0.041]; democracy[0.041]; democracy[0.041]; democracy[0.041]; jailing[0.041]; National[0.041]; Kyi[0.041]; Kyi[0.041]; Kyi[0.041]; Kyi[0.041]; Kyi[0.041]; leader[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NLD ==> ENTITY: \u001b[32mNational League for Democracy\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.179:0.179[0]; log p(e|m)= -1.217:-1.217[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nparty[0.044]; supporters[0.043]; activists[0.043]; Kyi[0.043]; Kyi[0.043]; Kyi[0.043]; released[0.043]; arrest[0.043]; Burma[0.042]; won[0.042]; demonstrations[0.042]; play[0.041]; landslide[0.041]; government[0.041]; peace[0.041]; campaign[0.041]; rights[0.041]; election[0.041]; Suu[0.041]; Suu[0.041]; Suu[0.041]; July[0.040]; poll[0.040]; democracy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Suu Kyi\u001b[39m ==> ENTITY: \u001b[32mAung San Suu Kyi\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurma[0.050]; Burma[0.050]; government[0.045]; democracy[0.045]; democracy[0.045]; political[0.045]; political[0.045]; Kyi[0.044]; Kyi[0.044]; Rangoon[0.043]; Suu[0.043]; Suu[0.043]; held[0.043]; party[0.043]; parliament[0.042]; serving[0.042]; arrest[0.042]; prisoners[0.042]; prisoners[0.042]; movement[0.041]; movement[0.041]; released[0.041]; popularity[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Suu Kyi\u001b[39m ==> ENTITY: \u001b[32mAung San Suu Kyi\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurma[0.048]; military[0.044]; government[0.043]; government[0.043]; democracy[0.043]; democracy[0.043]; Kyi[0.042]; Kyi[0.042]; peace[0.042]; activists[0.042]; activists[0.042]; supporters[0.042]; Suu[0.041]; Suu[0.041]; party[0.041]; State[0.040]; state[0.040]; arrest[0.040]; Htein[0.040]; crackdown[0.040]; won[0.040]; secretive[0.040]; politicians[0.040]; released[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NLD ==> ENTITY: \u001b[32mNational League for Democracy\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.199:0.199[0]; log p(e|m)= -1.217:-1.217[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDemocracy[0.044]; crackdown[0.043]; crackdown[0.043]; party[0.043]; supporters[0.043]; supporters[0.043]; week[0.042]; activists[0.042]; activists[0.042]; National[0.042]; Kyi[0.042]; Kyi[0.042]; Kyi[0.042]; Kyi[0.042]; Kyi[0.042]; released[0.042]; Burmese[0.041]; military[0.041]; government[0.040]; government[0.040]; government[0.040]; recent[0.040]; peace[0.040]; world[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rangoon\u001b[39m ==> ENTITY: \u001b[32mYangon\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRangoon[0.045]; military[0.045]; military[0.045]; military[0.045]; Burmese[0.044]; Burma[0.044]; government[0.041]; government[0.041]; government[0.041]; government[0.041]; crackdown[0.041]; crackdown[0.041]; authorities[0.041]; activists[0.041]; activists[0.041]; National[0.041]; saying[0.041]; reporters[0.040]; prison[0.040]; Democracy[0.040]; sentenced[0.040]; ahead[0.040]; arrested[0.040]; arrested[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Suu Kyi\u001b[39m ==> ENTITY: \u001b[32mAung San Suu Kyi\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurma[0.050]; government[0.045]; democracy[0.045]; democracy[0.045]; political[0.045]; political[0.045]; Kyi[0.044]; Kyi[0.044]; Rangoon[0.044]; Suu[0.043]; Suu[0.043]; held[0.043]; rights[0.043]; party[0.043]; parliament[0.043]; serving[0.042]; prisoners[0.042]; prisoners[0.042]; prisoners[0.042]; movement[0.042]; movement[0.042]; released[0.042]; popularity[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rangoon\u001b[39m ==> ENTITY: \u001b[32mYangon\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRangoon[0.046]; military[0.046]; military[0.046]; Burma[0.045]; Burma[0.045]; government[0.042]; authorities[0.042]; activists[0.042]; saying[0.042]; reporters[0.042]; arrested[0.041]; order[0.041]; Aung[0.041]; Saturday[0.040]; supporters[0.040]; jailing[0.040]; told[0.040]; recent[0.040]; home[0.040]; said[0.040]; hit[0.040]; arrests[0.040]; leader[0.040]; main[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NLD ==> ENTITY: \u001b[32mNational League for Democracy\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.169:0.169[0]; log p(e|m)= -1.217:-1.217[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nparty[0.046]; political[0.046]; Kyi[0.045]; Kyi[0.045]; Kyi[0.045]; arrest[0.045]; popularity[0.044]; Burma[0.044]; Burma[0.044]; weeks[0.043]; government[0.043]; arrests[0.043]; Suu[0.043]; Suu[0.043]; Suu[0.043]; parliament[0.043]; early[0.043]; August[0.042]; movement[0.042]; movement[0.042]; growing[0.042]; democracy[0.042]; democracy[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Suu Kyi\u001b[39m ==> ENTITY: \u001b[32mAung San Suu Kyi\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndemocracy[0.044]; political[0.044]; political[0.044]; activists[0.043]; Rangoon[0.043]; held[0.042]; rights[0.042]; party[0.042]; State[0.042]; won[0.041]; prisoners[0.041]; prisoners[0.041]; prisoners[0.041]; released[0.041]; released[0.041]; said[0.041]; said[0.041]; said[0.041]; landslide[0.041]; demonstrations[0.041]; people[0.041]; people[0.041]; Insein[0.041]; election[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Suu Kyi\u001b[39m ==> ENTITY: \u001b[32mAung San Suu Kyi\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurma[0.046]; Burmese[0.045]; military[0.042]; government[0.042]; government[0.042]; government[0.042]; government[0.042]; Democracy[0.042]; democracy[0.042]; democracy[0.042]; democracy[0.042]; democracy[0.042]; National[0.041]; Kyi[0.041]; Kyi[0.041]; Kyi[0.041]; Kyi[0.041]; Kyi[0.041]; peace[0.041]; activists[0.041]; activists[0.041]; Rangoon[0.041]; supporters[0.040]; supporters[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Suu Kyi\u001b[39m ==> ENTITY: \u001b[32mAung San Suu Kyi\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurma[0.048]; Burmese[0.046]; military[0.044]; military[0.044]; military[0.044]; government[0.043]; government[0.043]; government[0.043]; government[0.043]; Democracy[0.043]; democracy[0.043]; democracy[0.043]; democracy[0.043]; democracy[0.043]; democracy[0.043]; democracy[0.043]; jailing[0.043]; National[0.043]; Kyi[0.043]; Kyi[0.043]; Kyi[0.043]; Kyi[0.043]; leader[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Danish ==> ENTITY: \u001b[32mDenmark\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.083:0.083[0]; log p(e|m)= -0.650:-0.650[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nparty[0.045]; political[0.044]; political[0.044]; popularity[0.043]; government[0.043]; member[0.043]; James[0.042]; elected[0.042]; parliament[0.042]; came[0.041]; like[0.041]; serving[0.041]; early[0.041]; friend[0.041]; bring[0.041]; August[0.041]; movement[0.041]; movement[0.041]; held[0.041]; death[0.040]; death[0.040]; death[0.040]; growing[0.040]; physical[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBurma\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSaffron Revolution\u001b[39m <---> \u001b[32mBurma\u001b[39m\t\nSCORES: global= 0.247:0.246[\u001b[31m0.002\u001b[39m]; local()= 0.156:0.098[\u001b[31m0.058\u001b[39m]; log p(e|m)= -2.996:-0.202[\u001b[32m2.794\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncrackdown[0.044]; activists[0.043]; military[0.043]; government[0.043]; government[0.043]; Htein[0.043]; people[0.042]; rule[0.042]; supporters[0.042]; saying[0.041]; election[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; Kyi[0.041]; Kyi[0.041]; Kyi[0.041]; demonstrations[0.041]; party[0.041]; July[0.041]; landslide[0.041]; democracy[0.041]; peace[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NLD ==> ENTITY: \u001b[32mNational League for Democracy\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.174:0.174[0]; log p(e|m)= -1.217:-1.217[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nparty[0.046]; political[0.045]; political[0.045]; Kyi[0.045]; Kyi[0.045]; Kyi[0.045]; released[0.044]; arrest[0.044]; popularity[0.044]; Burma[0.044]; held[0.043]; weeks[0.043]; government[0.043]; arrests[0.043]; Suu[0.042]; Suu[0.042]; Suu[0.042]; parliament[0.042]; early[0.042]; August[0.042]; movement[0.042]; movement[0.042]; growing[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NLD ==> ENTITY: \u001b[32mNational League for Democracy\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.206:0.206[0]; log p(e|m)= -1.217:-1.217[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDemocracy[0.044]; crackdown[0.043]; party[0.043]; party[0.043]; supporters[0.042]; week[0.042]; activists[0.042]; activists[0.042]; National[0.042]; Kyi[0.042]; Kyi[0.042]; Kyi[0.042]; Kyi[0.042]; released[0.042]; released[0.042]; arrest[0.042]; Burma[0.041]; won[0.041]; Burmese[0.041]; military[0.040]; landslide[0.040]; government[0.040]; government[0.040]; government[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = National League for Democracy\u001b[39m ==> ENTITY: \u001b[32mNational League for Democracy\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nparty[0.049]; crackdown[0.047]; crackdown[0.047]; supporters[0.047]; supporters[0.047]; activists[0.046]; activists[0.046]; Kyi[0.046]; Kyi[0.046]; Kyi[0.046]; Kyi[0.046]; Kyi[0.046]; Burmese[0.045]; military[0.044]; military[0.044]; government[0.044]; government[0.044]; government[0.044]; recent[0.044]; peace[0.044]; week[0.044]; night[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burma ==> ENTITY: \u001b[32mBurma\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.202:-0.202[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurma[0.048]; government[0.047]; movement[0.042]; movement[0.042]; parliament[0.042]; repression[0.042]; said[0.042]; said[0.042]; said[0.042]; Kyi[0.042]; Kyi[0.042]; Kyi[0.042]; August[0.042]; party[0.041]; democracy[0.041]; democracy[0.041]; increased[0.041]; efforts[0.041]; serving[0.041]; early[0.040]; Hla[0.040]; arrests[0.040]; close[0.040]; stop[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burma ==> ENTITY: \u001b[32mBurma\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.202:-0.202[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurma[0.047]; activists[0.044]; military[0.043]; military[0.043]; government[0.043]; rulers[0.043]; Rangoon[0.042]; leader[0.042]; Aung[0.042]; movement[0.042]; movement[0.042]; saying[0.041]; Kyi[0.041]; Kyi[0.041]; Kyi[0.041]; democracy[0.041]; democracy[0.041]; democracy[0.041]; recent[0.041]; jailing[0.041]; world[0.040]; order[0.040]; told[0.039]; arrests[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burma ==> ENTITY: \u001b[32mBurma\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.111:0.111[0]; log p(e|m)= -0.202:-0.202[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurma[0.047]; government[0.046]; activists[0.043]; military[0.043]; military[0.043]; rulers[0.043]; reporters[0.042]; Rangoon[0.042]; Rangoon[0.042]; leader[0.041]; Aung[0.041]; movement[0.041]; movement[0.041]; supporters[0.041]; saying[0.041]; said[0.041]; Kyi[0.041]; Kyi[0.041]; Kyi[0.041]; Kyi[0.041]; democracy[0.040]; democracy[0.040]; democracy[0.040]; democracy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rangoon\u001b[39m ==> ENTITY: \u001b[32mYangon\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.118:0.118[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nserving[0.048]; authorities[0.047]; activists[0.047]; prison[0.046]; prison[0.046]; prison[0.046]; prison[0.046]; prison[0.046]; sentenced[0.046]; start[0.045]; parliament[0.045]; Insein[0.045]; Insein[0.045]; held[0.045]; jail[0.045]; said[0.045]; said[0.045]; said[0.045]; said[0.045]; sentence[0.045]; sentence[0.045]; sentence[0.045]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mInsein\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mInsein Prison\u001b[39m <---> \u001b[32mInsein Township\u001b[39m\t\nSCORES: global= 0.278:0.270[\u001b[31m0.008\u001b[39m]; local()= 0.152:0.092[\u001b[31m0.059\u001b[39m]; log p(e|m)= -2.198:0.000[\u001b[32m2.198\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInsein[0.052]; Burma[0.050]; Burma[0.050]; Prison[0.043]; instances[0.043]; Kyi[0.043]; Kyi[0.043]; Kyi[0.043]; prisoners[0.042]; torture[0.041]; sleep[0.040]; prison[0.040]; prison[0.040]; arrest[0.040]; Suu[0.039]; Suu[0.039]; Suu[0.039]; Hla[0.039]; August[0.039]; arrests[0.039]; free[0.039]; repression[0.039]; custody[0.038]; weeks[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NLD ==> ENTITY: \u001b[32mNational League for Democracy\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.179:0.179[0]; log p(e|m)= -1.217:-1.217[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nparty[0.044]; supporters[0.043]; activists[0.043]; activists[0.043]; Kyi[0.043]; Kyi[0.043]; Kyi[0.043]; released[0.042]; arrest[0.042]; Burma[0.042]; won[0.042]; demonstrations[0.042]; play[0.041]; landslide[0.041]; government[0.041]; peace[0.041]; campaign[0.041]; election[0.041]; Suu[0.041]; Suu[0.041]; Suu[0.041]; July[0.040]; poll[0.040]; democracy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burmese ==> ENTITY: \u001b[32mBurma\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.073:0.073[0]; log p(e|m)= -1.228:-1.228[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.043]; government[0.042]; government[0.042]; government[0.042]; movement[0.042]; movement[0.042]; Suu[0.042]; Suu[0.042]; Suu[0.042]; Suu[0.042]; Suu[0.042]; crackdown[0.042]; crackdown[0.042]; people[0.042]; Democracy[0.041]; recent[0.041]; Kyi[0.041]; Kyi[0.041]; Kyi[0.041]; Kyi[0.041]; Kyi[0.041]; Rangoon[0.041]; dozen[0.041]; peace[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mInsein\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mInsein Prison\u001b[39m <---> \u001b[32mInsein Township\u001b[39m\t\nSCORES: global= 0.277:0.269[\u001b[31m0.008\u001b[39m]; local()= 0.163:0.113[\u001b[31m0.050\u001b[39m]; log p(e|m)= -2.198:0.000[\u001b[32m2.198\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInsein[0.051]; Insein[0.051]; Burma[0.049]; Burma[0.049]; Rangoon[0.048]; Prison[0.042]; instances[0.042]; Kyi[0.042]; Kyi[0.042]; Kyi[0.042]; prisoners[0.040]; prisoners[0.040]; jail[0.039]; torture[0.039]; sleep[0.039]; prison[0.039]; prison[0.039]; prison[0.039]; held[0.039]; arrest[0.039]; people[0.038]; Suu[0.038]; Suu[0.038]; Suu[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Aung San Suu Kyi\u001b[39m ==> ENTITY: \u001b[32mAung San Suu Kyi\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurma[0.047]; Burma[0.047]; military[0.043]; military[0.043]; government[0.042]; government[0.042]; democracy[0.042]; democracy[0.042]; democracy[0.042]; democracy[0.042]; jailing[0.041]; Kyi[0.041]; Kyi[0.041]; Kyi[0.041]; leader[0.041]; activists[0.041]; Rangoon[0.041]; Rangoon[0.041]; supporters[0.041]; reporters[0.041]; Suu[0.040]; Suu[0.040]; Suu[0.040]; world[0.039]; \t\n\r [=========>....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s99ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1042/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1081testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lithuania ==> ENTITY: \u001b[32mLithuania national football team\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.158:0.158[0]; log p(e|m)= -2.129:-2.129[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLithuania[0.047]; Cup[0.045]; Cup[0.045]; Romania[0.045]; Romania[0.045]; Romania[0.045]; Qualifier[0.044]; soccer[0.044]; European[0.043]; qualifier[0.043]; Soccer[0.043]; World[0.042]; World[0.042]; Moldovan[0.040]; beat[0.040]; Bucharest[0.040]; minute[0.040]; Scorers[0.039]; Beat[0.038]; group[0.038]; Petrescu[0.038]; Viorel[0.038]; halftime[0.037]; Saturday[0.037]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEuropean\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUEFA\u001b[39m <---> \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.243:0.238[\u001b[31m0.004\u001b[39m]; local()= 0.110:0.042[\u001b[31m0.068\u001b[39m]; log p(e|m)= -3.576:-0.256[\u001b[32m3.319\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.047]; Soccer[0.047]; Cup[0.045]; Cup[0.045]; qualifier[0.045]; Qualifier[0.044]; Romania[0.043]; Romania[0.043]; Romania[0.043]; group[0.043]; Lithuania[0.042]; Lithuania[0.042]; Bucharest[0.042]; Petrescu[0.041]; beat[0.040]; Saturday[0.040]; World[0.040]; World[0.040]; minute[0.040]; Moldovan[0.039]; Beat[0.037]; Dan[0.037]; halftime[0.037]; Viorel[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Romania ==> ENTITY: \u001b[32mRomania national football team\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.159:0.159[0]; log p(e|m)= -3.058:-3.058[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRomania[0.048]; Romania[0.048]; soccer[0.045]; Lithuania[0.045]; Lithuania[0.045]; Soccer[0.045]; qualifier[0.043]; Qualifier[0.042]; Cup[0.042]; Cup[0.042]; European[0.042]; Bucharest[0.042]; halftime[0.041]; beat[0.041]; Moldovan[0.040]; minute[0.040]; World[0.039]; World[0.039]; Scorers[0.039]; Petrescu[0.039]; group[0.039]; Viorel[0.039]; Saturday[0.038]; Dan[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lithuania ==> ENTITY: \u001b[32mLithuania national football team\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.158:0.158[0]; log p(e|m)= -2.129:-2.129[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLithuania[0.047]; Cup[0.045]; Cup[0.045]; Romania[0.045]; Romania[0.045]; Romania[0.045]; Qualifier[0.044]; soccer[0.044]; European[0.043]; qualifier[0.043]; Soccer[0.043]; World[0.042]; World[0.042]; Moldovan[0.040]; beat[0.040]; Bucharest[0.040]; minute[0.040]; Scorers[0.039]; Beat[0.038]; group[0.038]; Petrescu[0.038]; Viorel[0.038]; halftime[0.037]; Saturday[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Viorel Moldovan\u001b[39m ==> ENTITY: \u001b[32mViorel Moldovan\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.047]; Soccer[0.045]; Romania[0.044]; Romania[0.044]; Romania[0.044]; Cup[0.044]; Cup[0.044]; Petrescu[0.044]; qualifier[0.043]; Qualifier[0.042]; European[0.042]; halftime[0.041]; Scorers[0.041]; minute[0.041]; beat[0.040]; Constantin[0.040]; Bucharest[0.040]; Dan[0.039]; Lithuania[0.039]; Lithuania[0.039]; Saturday[0.039]; Beat[0.038]; World[0.038]; World[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Romania ==> ENTITY: \u001b[32mRomania national football team\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.159:0.159[0]; log p(e|m)= -3.058:-3.058[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRomania[0.048]; Romania[0.048]; soccer[0.045]; Lithuania[0.045]; Lithuania[0.045]; Soccer[0.045]; qualifier[0.043]; Qualifier[0.042]; Cup[0.042]; Cup[0.042]; European[0.042]; Bucharest[0.042]; halftime[0.041]; beat[0.041]; Moldovan[0.040]; minute[0.040]; World[0.039]; World[0.039]; Scorers[0.039]; Petrescu[0.039]; group[0.039]; Viorel[0.039]; Saturday[0.038]; Dan[0.037]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBucharest\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFC Steaua București\u001b[39m <---> \u001b[32mBucharest\u001b[39m\t\nSCORES: global= 0.268:0.266[\u001b[31m0.002\u001b[39m]; local()= 0.152:0.112[\u001b[31m0.040\u001b[39m]; log p(e|m)= -2.564:-0.018[\u001b[32m2.546\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.046]; Cup[0.046]; qualifier[0.045]; Qualifier[0.044]; Romania[0.044]; Romania[0.044]; Romania[0.044]; soccer[0.043]; World[0.043]; World[0.043]; Soccer[0.043]; Petrescu[0.042]; European[0.041]; Moldovan[0.041]; Saturday[0.040]; Lithuania[0.040]; Lithuania[0.040]; minute[0.040]; Viorel[0.039]; beat[0.039]; Dan[0.038]; group[0.038]; Constantin[0.038]; Beat[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dan Petrescu\u001b[39m ==> ENTITY: \u001b[32mDan Petrescu\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRomania[0.047]; Romania[0.047]; Romania[0.047]; soccer[0.046]; Soccer[0.046]; European[0.045]; Cup[0.044]; Cup[0.044]; Viorel[0.044]; Bucharest[0.044]; beat[0.043]; qualifier[0.043]; Moldovan[0.043]; Qualifier[0.043]; Lithuania[0.042]; Lithuania[0.042]; halftime[0.042]; Scorers[0.042]; minute[0.042]; Saturday[0.041]; Beat[0.041]; Constantin[0.041]; group[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Constantin Galca\u001b[39m ==> ENTITY: \u001b[32mConstantin Gâlcă\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.127:0.127[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRomania[0.045]; Romania[0.045]; Romania[0.045]; qualifier[0.044]; soccer[0.043]; European[0.043]; Qualifier[0.042]; Cup[0.042]; Cup[0.042]; Soccer[0.042]; Bucharest[0.042]; halftime[0.041]; Scorers[0.041]; Lithuania[0.041]; Lithuania[0.041]; minute[0.041]; Petrescu[0.040]; Saturday[0.040]; World[0.040]; World[0.040]; group[0.040]; Moldovan[0.040]; beat[0.039]; Dan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Romania ==> ENTITY: \u001b[32mRomania national football team\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.159:0.159[0]; log p(e|m)= -3.058:-3.058[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRomania[0.048]; Romania[0.048]; soccer[0.045]; Lithuania[0.045]; Lithuania[0.045]; Soccer[0.045]; qualifier[0.043]; Qualifier[0.042]; Cup[0.042]; Cup[0.042]; European[0.042]; Bucharest[0.042]; halftime[0.041]; beat[0.041]; Moldovan[0.040]; minute[0.040]; World[0.039]; World[0.039]; Scorers[0.039]; Petrescu[0.039]; group[0.039]; Viorel[0.039]; Saturday[0.038]; Dan[0.037]; \t\n\r [=========>....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s148ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1054/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 985testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = European ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.102:0.102[0]; log p(e|m)= -0.256:-0.256[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.047]; trading[0.045]; trading[0.045]; Options[0.045]; price[0.043]; Germany[0.043]; business[0.043]; economic[0.042]; world[0.042]; September[0.042]; day[0.041]; day[0.041]; settlement[0.041]; International[0.041]; struggled[0.040]; major[0.040]; lower[0.040]; extended[0.040]; Futures[0.040]; higher[0.040]; higher[0.040]; late[0.040]; London[0.040]; London[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmillion[0.045]; million[0.045]; million[0.045]; dollar[0.045]; week[0.044]; week[0.044]; issued[0.042]; European[0.041]; International[0.041]; day[0.041]; day[0.041]; September[0.041]; stock[0.041]; Electric[0.041]; employment[0.041]; Japanese[0.041]; Japanese[0.041]; Day[0.040]; price[0.040]; set[0.040]; launched[0.040]; Thursday[0.040]; raise[0.040]; Monday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Labour Day ==> ENTITY: \u001b[32mLabor Day\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.134:0.134[0]; log p(e|m)= -1.328:-1.328[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlate[0.045]; weekend[0.044]; day[0.044]; ahead[0.043]; strong[0.042]; Thursday[0.042]; Thursday[0.042]; Thursday[0.042]; rates[0.042]; industrial[0.042]; levels[0.042]; data[0.041]; data[0.041]; expected[0.041]; output[0.041]; output[0.041]; benchmark[0.041]; September[0.041]; September[0.041]; States[0.040]; afternoon[0.040]; following[0.040]; sharply[0.040]; held[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Paris ==> ENTITY: \u001b[32mParis\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.129:0.129[0]; log p(e|m)= -0.115:-0.115[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.048]; Thursday[0.046]; Thursday[0.046]; Thursday[0.046]; late[0.045]; Monday[0.045]; held[0.044]; near[0.044]; Italy[0.043]; struggled[0.043]; August[0.043]; despite[0.043]; settled[0.043]; drop[0.042]; September[0.042]; September[0.042]; September[0.042]; afternoon[0.042]; ended[0.042]; ended[0.042]; day[0.042]; day[0.042]; day[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.S. Treasuries\u001b[39m ==> ENTITY: \u001b[32mUnited States Treasury security\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nExchange[0.045]; trading[0.043]; trading[0.043]; trading[0.043]; notional[0.042]; Financial[0.042]; Bond[0.042]; Bond[0.042]; bond[0.042]; bond[0.042]; market[0.042]; futures[0.042]; futures[0.042]; lower[0.041]; Treasuries[0.041]; September[0.041]; September[0.041]; September[0.041]; higher[0.041]; higher[0.041]; higher[0.041]; Futures[0.041]; markets[0.040]; losses[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tokyo Stock Exchange\u001b[39m ==> ENTITY: \u001b[32mTokyo Stock Exchange\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.224:0.224[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrading[0.045]; trading[0.045]; Tokyo[0.045]; Japan[0.045]; stock[0.045]; Japanese[0.043]; Japanese[0.043]; market[0.042]; launched[0.042]; closed[0.042]; following[0.042]; day[0.041]; day[0.041]; launch[0.041]; institution[0.041]; September[0.040]; Thursday[0.040]; Monday[0.040]; investors[0.040]; issued[0.040]; fell[0.039]; million[0.039]; million[0.039]; million[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kansai International Airport\u001b[39m ==> ENTITY: \u001b[32mKansai International Airport\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.120:0.120[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKansai[0.047]; Tokyo[0.047]; million[0.044]; million[0.044]; million[0.044]; launch[0.043]; Japanese[0.043]; launched[0.042]; Thursday[0.042]; Monday[0.042]; Exchange[0.042]; sharply[0.041]; Electric[0.041]; Eurobonds[0.040]; Liffe[0.040]; dollar[0.040]; lower[0.040]; market[0.040]; week[0.040]; week[0.040]; Primary[0.040]; deals[0.040]; discount[0.039]; wound[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.104:0.104[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapanese[0.049]; Tokyo[0.048]; Tokyo[0.048]; ahead[0.047]; following[0.045]; Exchange[0.043]; lower[0.043]; United[0.043]; States[0.043]; weekend[0.043]; September[0.043]; September[0.043]; September[0.043]; cash[0.043]; closed[0.042]; day[0.042]; day[0.042]; low[0.042]; rates[0.042]; helped[0.042]; long[0.042]; long[0.042]; strong[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Treasuries\u001b[39m ==> ENTITY: \u001b[32mUnited States Treasury security\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.227:0.227[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTreasury[0.044]; Treasury[0.044]; trading[0.043]; trading[0.043]; trading[0.043]; trading[0.043]; securities[0.042]; notional[0.042]; Thursday[0.042]; Thursday[0.042]; Thursday[0.042]; Thursday[0.042]; Monday[0.041]; Bond[0.041]; Bond[0.041]; bond[0.041]; bond[0.041]; bond[0.041]; bond[0.041]; market[0.041]; futures[0.041]; futures[0.041]; lower[0.041]; lower[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.113:0.113[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.047]; despite[0.045]; Paris[0.044]; Britain[0.044]; ahead[0.043]; States[0.042]; late[0.042]; ended[0.042]; ended[0.042]; struggled[0.042]; settled[0.042]; settlement[0.041]; settlement[0.041]; settlement[0.041]; settlement[0.041]; held[0.041]; said[0.041]; Futures[0.040]; September[0.040]; September[0.040]; September[0.040]; pulled[0.040]; United[0.040]; August[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japanese ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.992:-0.992[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapanese[0.052]; sharply[0.048]; Tokyo[0.047]; Tokyo[0.047]; employment[0.045]; European[0.044]; million[0.043]; million[0.043]; million[0.043]; investors[0.043]; lower[0.043]; day[0.042]; day[0.042]; September[0.042]; dollar[0.042]; players[0.042]; prices[0.042]; gains[0.042]; market[0.042]; Monday[0.042]; launched[0.041]; week[0.041]; week[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.054:0.054[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.047]; European[0.045]; International[0.045]; world[0.045]; London[0.045]; London[0.045]; major[0.042]; September[0.042]; economic[0.041]; ended[0.041]; losses[0.041]; rates[0.041]; lower[0.040]; price[0.040]; bond[0.040]; bond[0.040]; Exchange[0.040]; gilts[0.040]; data[0.040]; rise[0.040]; day[0.040]; day[0.040]; Bund[0.040]; shock[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.078:0.078[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmillion[0.045]; million[0.045]; million[0.045]; dollar[0.045]; week[0.044]; week[0.044]; following[0.042]; issued[0.042]; European[0.041]; International[0.041]; expected[0.041]; day[0.041]; day[0.041]; September[0.041]; stock[0.041]; Electric[0.040]; employment[0.040]; Japanese[0.040]; Japanese[0.040]; Day[0.040]; price[0.040]; set[0.040]; launched[0.040]; Thursday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.S. Treasury\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of the Treasury\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTreasury[0.047]; Treasury[0.047]; late[0.043]; States[0.042]; Chicago[0.042]; Chicago[0.042]; September[0.042]; September[0.042]; September[0.042]; helped[0.042]; securities[0.042]; report[0.042]; Board[0.041]; Thursday[0.041]; Thursday[0.041]; Thursday[0.041]; Thursday[0.041]; Bond[0.040]; Bond[0.040]; fell[0.040]; following[0.040]; Trade[0.040]; cash[0.040]; Pibor[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.072:0.072[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.045]; States[0.043]; late[0.043]; August[0.042]; September[0.042]; September[0.042]; September[0.042]; report[0.042]; despite[0.042]; expected[0.042]; day[0.042]; day[0.042]; day[0.042]; settled[0.041]; drop[0.041]; price[0.041]; price[0.041]; price[0.041]; price[0.041]; helped[0.041]; securities[0.041]; higher[0.040]; Italy[0.040]; losses[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = BOJ\u001b[39m ==> ENTITY: \u001b[32mBank of Japan\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.045]; Japanese[0.043]; launched[0.043]; Tokyo[0.042]; Tokyo[0.042]; markets[0.042]; September[0.042]; September[0.042]; expected[0.042]; week[0.042]; Exchange[0.042]; hours[0.042]; rates[0.041]; benchmark[0.041]; Monday[0.041]; Treasury[0.041]; Treasury[0.041]; million[0.041]; market[0.041]; day[0.041]; day[0.041]; prices[0.041]; Thursday[0.041]; Thursday[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norwegian ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.072:0.072[0]; log p(e|m)= -0.607:-0.607[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nweek[0.047]; week[0.047]; European[0.046]; International[0.043]; placed[0.043]; day[0.043]; day[0.043]; players[0.042]; Airport[0.042]; hours[0.042]; discount[0.041]; employment[0.041]; launched[0.041]; September[0.041]; lost[0.041]; Monday[0.040]; holiday[0.040]; ahead[0.040]; Day[0.040]; price[0.040]; deals[0.040]; launch[0.040]; pre[0.040]; Thursday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chicago Board of Trade\u001b[39m ==> ENTITY: \u001b[32mChicago Board of Trade\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrading[0.045]; trading[0.045]; trading[0.045]; trading[0.045]; futures[0.045]; futures[0.045]; stock[0.044]; Chicago[0.043]; markets[0.042]; traders[0.042]; benchmark[0.041]; securities[0.040]; gains[0.040]; price[0.040]; price[0.040]; cash[0.040]; data[0.040]; data[0.040]; Thursday[0.040]; Thursday[0.040]; Liffe[0.040]; Liffe[0.040]; closed[0.040]; prices[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.102:0.102[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.045]; despite[0.044]; August[0.043]; September[0.043]; September[0.043]; September[0.043]; September[0.043]; held[0.042]; helped[0.042]; Monday[0.042]; Thursday[0.042]; Thursday[0.042]; Thursday[0.042]; weekend[0.041]; expected[0.041]; stronger[0.041]; day[0.041]; day[0.041]; drop[0.041]; notional[0.040]; States[0.040]; United[0.040]; ahead[0.040]; ahead[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.088:0.088[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.049]; Britain[0.046]; trading[0.045]; trading[0.045]; Financial[0.043]; Exchange[0.043]; business[0.042]; European[0.042]; extended[0.042]; International[0.041]; late[0.041]; market[0.041]; day[0.041]; day[0.041]; rates[0.041]; Thursday[0.041]; markets[0.041]; Friday[0.040]; September[0.040]; ended[0.040]; major[0.039]; session[0.039]; Late[0.039]; price[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.081:0.081[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInternational[0.049]; Britain[0.047]; major[0.047]; European[0.045]; Germany[0.045]; late[0.044]; world[0.044]; day[0.043]; day[0.043]; September[0.043]; September[0.043]; August[0.043]; ended[0.043]; price[0.043]; Futures[0.042]; Friday[0.042]; ahead[0.042]; Late[0.042]; higher[0.042]; higher[0.042]; Thursday[0.042]; losses[0.042]; Monday[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Labour Day ==> ENTITY: \u001b[32mLabor Day\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.110:0.110[0]; log p(e|m)= -1.328:-1.328[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nemployment[0.050]; ahead[0.043]; holiday[0.043]; day[0.043]; day[0.043]; Monday[0.042]; Thursday[0.042]; rates[0.042]; industrial[0.042]; week[0.041]; week[0.041]; data[0.041]; data[0.041]; data[0.041]; output[0.041]; output[0.041]; activity[0.041]; following[0.041]; hours[0.041]; September[0.040]; sharply[0.040]; issued[0.040]; European[0.040]; future[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tokyo ==> ENTITY: \u001b[32mTokyo\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.093:-0.093[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTokyo[0.046]; Japan[0.045]; trading[0.044]; trading[0.044]; trading[0.044]; stock[0.043]; Japanese[0.042]; Japanese[0.042]; September[0.042]; September[0.042]; institution[0.041]; Exchange[0.041]; employment[0.041]; Thursday[0.041]; Thursday[0.041]; market[0.041]; launched[0.041]; closed[0.041]; following[0.040]; crown[0.040]; Monday[0.040]; day[0.040]; day[0.040]; Stock[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.090:0.090[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.044]; day[0.044]; despite[0.043]; France[0.043]; weekend[0.042]; Japan[0.042]; Thursday[0.042]; Thursday[0.042]; Thursday[0.042]; Thursday[0.042]; surprisingly[0.042]; ahead[0.041]; Day[0.041]; held[0.041]; drop[0.041]; September[0.041]; September[0.041]; September[0.041]; following[0.041]; helped[0.041]; losses[0.041]; fell[0.041]; Chicago[0.041]; Chicago[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEuropean\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEurope\u001b[39m <---> \u001b[32mEuropean Union\u001b[39m\t\nSCORES: global= 0.258:0.241[\u001b[31m0.016\u001b[39m]; local()= 0.082:0.104[\u001b[32m0.022\u001b[39m]; log p(e|m)= -0.256:-2.937[\u001b[31m2.681\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nweek[0.044]; week[0.044]; players[0.044]; September[0.043]; market[0.042]; million[0.042]; million[0.042]; million[0.042]; issued[0.042]; likely[0.042]; Japanese[0.042]; settlement[0.042]; International[0.042]; placed[0.042]; employment[0.042]; Airport[0.041]; Norwegian[0.041]; Monday[0.041]; ahead[0.041]; pre[0.040]; day[0.040]; sharply[0.040]; launch[0.040]; discount[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nday[0.045]; day[0.045]; day[0.045]; ahead[0.045]; London[0.045]; London[0.045]; session[0.045]; France[0.044]; European[0.044]; despite[0.044]; Monday[0.043]; September[0.043]; September[0.043]; September[0.043]; Thursday[0.043]; Thursday[0.043]; Germany[0.043]; late[0.042]; world[0.042]; revived[0.042]; Friday[0.042]; major[0.042]; August[0.042]; \t\n\r [=========>....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s331ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1080/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1132testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Albright\u001b[39m ==> ENTITY: \u001b[32mMadeleine Albright\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.049]; ambassador[0.048]; Ambassador[0.048]; Foreign[0.047]; Albright[0.047]; Albright[0.047]; States[0.046]; statement[0.046]; discuss[0.046]; debate[0.046]; tour[0.045]; Council[0.045]; nation[0.044]; said[0.044]; Security[0.044]; officials[0.044]; talks[0.044]; talks[0.044]; meet[0.044]; Monday[0.044]; Minister[0.044]; embassy[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Latin American ==> ENTITY: \u001b[32mLatin America\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.175:0.175[0]; log p(e|m)= -0.137:-0.137[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLatin[0.052]; American[0.046]; Bolivia[0.045]; Guatemala[0.044]; Jose[0.044]; Miguel[0.043]; Honduras[0.042]; Chile[0.042]; Chile[0.042]; Chile[0.042]; nation[0.042]; States[0.041]; local[0.041]; non[0.041]; Foreign[0.040]; include[0.039]; Monday[0.039]; Monday[0.039]; tour[0.039]; tour[0.039]; currently[0.039]; Uruguay[0.039]; city[0.039]; rest[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chile ==> ENTITY: \u001b[32mChile\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.203:0.203[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChile[0.046]; Chile[0.046]; Chile[0.046]; Chile[0.046]; Honduras[0.043]; Chilean[0.043]; Guatemala[0.043]; Bolivia[0.042]; States[0.042]; Uruguay[0.042]; Foreign[0.041]; ambassador[0.041]; Monday[0.040]; Monday[0.040]; Ambassador[0.040]; Miguel[0.040]; statement[0.040]; Jose[0.040]; city[0.040]; Latin[0.040]; Latin[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Security Council\u001b[39m ==> ENTITY: \u001b[32mUnited Nations Security Council\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.045]; Albright[0.045]; Albright[0.045]; Albright[0.045]; Albright[0.045]; Minister[0.044]; Foreign[0.044]; official[0.044]; statement[0.043]; States[0.043]; ambassador[0.043]; Ambassador[0.043]; discuss[0.043]; permanent[0.043]; officials[0.043]; officials[0.043]; officials[0.043]; relations[0.043]; talks[0.043]; talks[0.043]; Monday[0.043]; Monday[0.043]; member[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = United Nations\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.118:0.118[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCouncil[0.044]; ambassador[0.044]; American[0.044]; Ambassador[0.043]; Albright[0.043]; Albright[0.043]; Albright[0.043]; Albright[0.043]; Foreign[0.042]; States[0.042]; statement[0.042]; Security[0.042]; officials[0.041]; officials[0.041]; Monday[0.041]; United[0.041]; nation[0.041]; Jose[0.041]; debate[0.040]; tour[0.040]; Santiago[0.040]; Miguel[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Uruguay ==> ENTITY: \u001b[32mUruguay\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.180:0.180[0]; log p(e|m)= -0.446:-0.446[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChile[0.046]; Chile[0.046]; Chile[0.046]; Chile[0.046]; Guatemala[0.043]; Chilean[0.043]; Honduras[0.043]; American[0.043]; American[0.043]; Bolivia[0.043]; Latin[0.043]; Latin[0.043]; relations[0.041]; Jose[0.040]; Miguel[0.040]; States[0.039]; nation[0.039]; debate[0.039]; issues[0.039]; issues[0.039]; issues[0.039]; Council[0.039]; Security[0.039]; currently[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.102:0.102[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHonduras[0.045]; Uruguay[0.043]; Guatemala[0.043]; American[0.043]; American[0.043]; day[0.042]; Chile[0.042]; Chile[0.042]; Chile[0.042]; Chile[0.042]; Chile[0.042]; tour[0.041]; tour[0.041]; Nations[0.041]; weekend[0.041]; nation[0.041]; member[0.041]; Latin[0.041]; Latin[0.041]; Friday[0.040]; United[0.040]; Monday[0.040]; Monday[0.040]; Jose[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Latin American ==> ENTITY: \u001b[32mLatin America\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.170:0.170[0]; log p(e|m)= -0.137:-0.137[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLatin[0.052]; American[0.045]; Bolivia[0.044]; Jose[0.043]; Chilean[0.043]; States[0.043]; Miguel[0.043]; Chile[0.042]; Chile[0.042]; Chile[0.042]; Chile[0.042]; Chile[0.042]; Foreign[0.041]; local[0.041]; nation[0.041]; Security[0.040]; late[0.040]; Uruguay[0.039]; permanent[0.039]; United[0.039]; United[0.039]; Santiago[0.039]; Nations[0.039]; currently[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Madeleine Albright\u001b[39m ==> ENTITY: \u001b[32mMadeleine Albright\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.045]; ambassador[0.044]; Ambassador[0.044]; Foreign[0.043]; Albright[0.043]; Albright[0.043]; Albright[0.043]; States[0.043]; statement[0.042]; discuss[0.042]; debate[0.042]; tour[0.041]; Council[0.041]; nation[0.041]; said[0.041]; said[0.041]; Security[0.041]; officials[0.041]; officials[0.041]; talks[0.040]; talks[0.040]; meet[0.040]; Monday[0.040]; Minister[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cochabamba ==> ENTITY: \u001b[32mCochabamba\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.103:-0.103[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBolivia[0.055]; Chile[0.052]; Chile[0.052]; Chile[0.052]; Guatemala[0.051]; city[0.048]; Honduras[0.048]; Latin[0.047]; Miguel[0.046]; relations[0.046]; Jose[0.046]; Uruguay[0.044]; Insulza[0.043]; local[0.042]; affecting[0.042]; issues[0.041]; issues[0.041]; Monday[0.041]; Monday[0.041]; council[0.041]; States[0.041]; leave[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chile ==> ENTITY: \u001b[32mChile\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.154:0.154[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChile[0.049]; Chile[0.049]; Chile[0.049]; Chilean[0.046]; States[0.045]; Foreign[0.044]; ambassador[0.043]; Monday[0.043]; Ambassador[0.043]; Miguel[0.043]; statement[0.043]; Jose[0.043]; Latin[0.043]; said[0.042]; local[0.042]; Friday[0.042]; officials[0.042]; Security[0.042]; Santiago[0.042]; currently[0.042]; United[0.042]; United[0.042]; American[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chile ==> ENTITY: \u001b[32mChile\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.191:0.191[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChile[0.048]; Chile[0.048]; Chile[0.048]; Honduras[0.045]; Chilean[0.045]; Guatemala[0.045]; Bolivia[0.045]; Uruguay[0.044]; Foreign[0.043]; Monday[0.043]; Monday[0.043]; Miguel[0.042]; statement[0.042]; Jose[0.042]; city[0.042]; Latin[0.042]; Latin[0.042]; said[0.042]; said[0.042]; said[0.042]; official[0.042]; local[0.041]; Friday[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jose Miguel Insulza\u001b[39m ==> ENTITY: \u001b[32mJosé Miguel Insulza\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHonduras[0.043]; Guatemala[0.043]; Chile[0.043]; Chile[0.043]; Chile[0.043]; Chile[0.043]; Chile[0.043]; Foreign[0.043]; Chilean[0.043]; Latin[0.042]; Latin[0.042]; Bolivia[0.042]; Ambassador[0.042]; ambassador[0.042]; Council[0.041]; States[0.041]; statement[0.041]; American[0.040]; American[0.040]; said[0.040]; said[0.040]; said[0.040]; Minister[0.040]; relations[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chile ==> ENTITY: \u001b[32mChile\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.169:0.169[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChile[0.046]; Chile[0.046]; Chile[0.046]; Chile[0.046]; Chilean[0.044]; States[0.043]; Uruguay[0.042]; Foreign[0.042]; ambassador[0.041]; Monday[0.041]; Ambassador[0.041]; Miguel[0.041]; statement[0.041]; Jose[0.040]; Latin[0.040]; said[0.040]; said[0.040]; local[0.040]; Friday[0.040]; officials[0.040]; officials[0.040]; Security[0.040]; Santiago[0.040]; currently[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chile ==> ENTITY: \u001b[32mChile\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChile[0.046]; Chile[0.046]; Chile[0.046]; Chile[0.046]; Honduras[0.043]; Chilean[0.043]; Guatemala[0.043]; Bolivia[0.043]; Uruguay[0.042]; Foreign[0.041]; Monday[0.041]; Monday[0.041]; Miguel[0.040]; statement[0.040]; Jose[0.040]; city[0.040]; Latin[0.040]; Latin[0.040]; said[0.040]; said[0.040]; said[0.040]; official[0.040]; local[0.040]; Friday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Honduras ==> ENTITY: \u001b[32mHonduras\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.163:0.163[0]; log p(e|m)= -0.307:-0.307[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGuatemala[0.047]; Bolivia[0.046]; American[0.045]; Chile[0.044]; Chile[0.044]; Chile[0.044]; Uruguay[0.044]; Latin[0.043]; official[0.043]; Cochabamba[0.042]; Miguel[0.041]; Jose[0.041]; Monday[0.041]; Monday[0.041]; summit[0.041]; statement[0.040]; member[0.040]; States[0.039]; officials[0.039]; officials[0.039]; city[0.039]; United[0.039]; tour[0.039]; non[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHonduras[0.045]; Uruguay[0.043]; American[0.043]; American[0.043]; Guatemala[0.043]; Chile[0.042]; Chile[0.042]; Chile[0.042]; Chile[0.042]; Chile[0.042]; various[0.042]; various[0.042]; States[0.041]; day[0.041]; late[0.041]; nation[0.041]; officials[0.041]; officials[0.041]; officials[0.041]; tour[0.040]; tour[0.040]; United[0.040]; United[0.040]; currently[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chilean ==> ENTITY: \u001b[32mChile\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.196:0.196[0]; log p(e|m)= -0.257:-0.257[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChile[0.046]; Chile[0.046]; Chile[0.046]; Chile[0.046]; Chile[0.046]; American[0.045]; Bolivia[0.043]; Latin[0.043]; Uruguay[0.042]; Santiago[0.042]; Jose[0.042]; Miguel[0.041]; States[0.041]; United[0.041]; United[0.041]; nation[0.040]; currently[0.039]; ambassador[0.039]; non[0.038]; said[0.038]; said[0.038]; said[0.038]; local[0.038]; Insulza[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstatement[0.046]; Albright[0.044]; Albright[0.044]; Albright[0.044]; ambassador[0.044]; Ambassador[0.043]; Council[0.043]; Security[0.042]; American[0.042]; officials[0.042]; Foreign[0.042]; said[0.041]; Monday[0.041]; talks[0.041]; talks[0.041]; issues[0.041]; issues[0.041]; issues[0.041]; debate[0.040]; discuss[0.040]; Friday[0.040]; United[0.040]; nation[0.040]; Nations[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Santiago ==> ENTITY: \u001b[32mSantiago\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.163:0.163[0]; log p(e|m)= -0.465:-0.465[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMiguel[0.046]; Chilean[0.046]; Jose[0.046]; Chile[0.044]; Chile[0.044]; Chile[0.044]; Chile[0.044]; Monday[0.043]; nation[0.042]; Friday[0.041]; local[0.041]; Latin[0.041]; arrived[0.041]; late[0.040]; tour[0.040]; American[0.040]; currently[0.040]; officials[0.040]; officials[0.040]; said[0.040]; Council[0.039]; ambassador[0.039]; Insulza[0.039]; embassy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bolivia ==> ENTITY: \u001b[32mBolivia\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.178:0.178[0]; log p(e|m)= -0.218:-0.218[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGuatemala[0.046]; United[0.045]; Chile[0.044]; Chile[0.044]; Chile[0.044]; States[0.044]; Honduras[0.044]; Uruguay[0.043]; Latin[0.042]; Latin[0.042]; nation[0.042]; Miguel[0.041]; Chilean[0.041]; Jose[0.041]; American[0.041]; American[0.041]; said[0.040]; said[0.040]; said[0.040]; Cochabamba[0.040]; day[0.039]; officials[0.039]; officials[0.039]; officials[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Albright\u001b[39m ==> ENTITY: \u001b[32mMadeleine Albright\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.046]; American[0.046]; summit[0.045]; Foreign[0.045]; Albright[0.044]; Albright[0.044]; attend[0.044]; States[0.044]; meeting[0.044]; statement[0.044]; discuss[0.044]; debate[0.044]; tour[0.043]; tour[0.043]; Council[0.043]; nation[0.042]; said[0.042]; said[0.042]; said[0.042]; Security[0.042]; officials[0.042]; officials[0.042]; officials[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Albright\u001b[39m ==> ENTITY: \u001b[32mMadeleine Albright\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.048]; American[0.048]; ambassador[0.047]; Ambassador[0.047]; summit[0.046]; Foreign[0.046]; Albright[0.046]; Albright[0.046]; Albright[0.046]; attend[0.046]; States[0.046]; meeting[0.046]; statement[0.046]; discuss[0.045]; debate[0.045]; tour[0.044]; Council[0.044]; nation[0.044]; said[0.044]; said[0.044]; said[0.044]; Security[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Guatemala ==> ENTITY: \u001b[32mGuatemala\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.161:0.161[0]; log p(e|m)= -0.245:-0.245[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMonday[0.045]; Monday[0.045]; Honduras[0.045]; Latin[0.044]; American[0.044]; Bolivia[0.044]; Chile[0.044]; Chile[0.044]; Chile[0.044]; Uruguay[0.043]; official[0.042]; Miguel[0.042]; member[0.042]; statement[0.041]; Insulza[0.040]; Cochabamba[0.040]; States[0.040]; day[0.039]; weekend[0.039]; city[0.039]; attend[0.039]; various[0.038]; summit[0.038]; begin[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.080:0.080[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.045]; Chile[0.043]; Chile[0.043]; Chile[0.043]; Chile[0.043]; States[0.043]; late[0.042]; nation[0.042]; tour[0.042]; United[0.041]; United[0.041]; currently[0.041]; Jose[0.041]; statement[0.041]; Miguel[0.041]; Latin[0.041]; Monday[0.041]; Council[0.041]; officials[0.041]; officials[0.041]; Nations[0.041]; Albright[0.041]; Albright[0.041]; Albright[0.041]; \t\n\r [=========>....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s178ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1105/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1105testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Madrid ==> ENTITY: \u001b[32mMadrid\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMadrid[0.060]; Soccer[0.057]; Carlos[0.056]; Spanish[0.056]; Spanish[0.056]; Roberto[0.056]; Coruna[0.055]; game[0.054]; played[0.054]; Deportivo[0.053]; Real[0.051]; Summary[0.051]; Summary[0.051]; division[0.049]; Saturday[0.049]; Division[0.049]; Martins[0.048]; minute[0.048]; Halftime[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spanish ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.113:0.113[0]; log p(e|m)= -1.241:-1.241[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpanish[0.067]; Soccer[0.058]; Carlos[0.055]; Roberto[0.055]; Madrid[0.055]; Madrid[0.055]; game[0.054]; Division[0.054]; played[0.053]; division[0.053]; Real[0.052]; Deportivo[0.052]; Coruna[0.051]; minute[0.051]; Martins[0.048]; Saturday[0.048]; Halftime[0.048]; Summary[0.045]; Summary[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Madrid\u001b[39m ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.063]; Carlos[0.060]; Madrid[0.060]; Deportivo[0.058]; game[0.058]; Roberto[0.058]; Coruna[0.057]; Division[0.057]; minute[0.057]; played[0.056]; Spanish[0.055]; Spanish[0.055]; division[0.055]; Halftime[0.051]; Saturday[0.051]; Martins[0.049]; Summary[0.049]; Summary[0.049]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mRoberto Carlos\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRoberto Carlos (footballer)\u001b[39m <---> \u001b[32mRoberto Carlos (Spanish footballer)\u001b[39m\t\nSCORES: global= 0.268:0.241[\u001b[31m0.028\u001b[39m]; local()= 0.142:0.110[\u001b[31m0.032\u001b[39m]; log p(e|m)= -0.057:-1.599[\u001b[31m1.543\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.063]; Spanish[0.059]; Spanish[0.059]; Madrid[0.058]; Madrid[0.058]; game[0.057]; Coruna[0.057]; Real[0.057]; Halftime[0.056]; Deportivo[0.055]; minute[0.055]; Martins[0.055]; played[0.054]; Division[0.052]; Saturday[0.052]; division[0.052]; Summary[0.050]; Summary[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spanish ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.113:0.113[0]; log p(e|m)= -1.241:-1.241[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpanish[0.067]; Soccer[0.058]; Carlos[0.055]; Roberto[0.055]; Madrid[0.055]; Madrid[0.055]; game[0.054]; Division[0.054]; played[0.053]; division[0.053]; Real[0.052]; Deportivo[0.052]; Coruna[0.051]; minute[0.051]; Martins[0.048]; Saturday[0.048]; Halftime[0.048]; Summary[0.045]; Summary[0.045]; \t\n\r [=========>....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s265ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1110/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 989testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Johannesburg ==> ENTITY: \u001b[32mJohannesburg\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.095:-0.095[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJohannesburg[0.050]; South[0.049]; Wednesday[0.045]; Friday[0.044]; east[0.044]; East[0.044]; Africa[0.043]; Fields[0.043]; Fields[0.043]; Consolidated[0.043]; Consolidated[0.043]; company[0.043]; night[0.043]; month[0.043]; failed[0.043]; gave[0.043]; managers[0.043]; statement[0.042]; Kloof[0.042]; sparked[0.042]; said[0.042]; said[0.042]; employee[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Johannesburg ==> ENTITY: \u001b[32mJohannesburg\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.095:-0.095[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJohannesburg[0.050]; South[0.049]; Wednesday[0.045]; Friday[0.044]; east[0.044]; East[0.044]; Africa[0.043]; Fields[0.043]; Fields[0.043]; Consolidated[0.043]; Consolidated[0.043]; company[0.043]; night[0.043]; month[0.043]; failed[0.043]; gave[0.043]; managers[0.043]; statement[0.042]; Kloof[0.042]; sparked[0.042]; said[0.042]; said[0.042]; employee[0.041]; \t\n\r [=========>....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s432ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1112/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1001testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Budapest Stock Exchange\u001b[39m ==> ENTITY: \u001b[32mBudapest Stock Exchange\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbourse[0.045]; conference[0.044]; company[0.043]; opening[0.043]; opened[0.043]; Budapest[0.043]; Friday[0.043]; Friday[0.043]; Thursday[0.043]; annual[0.042]; plunged[0.042]; profit[0.041]; profit[0.041]; Vienna[0.041]; following[0.041]; GDRs[0.041]; close[0.040]; H1[0.040]; million[0.040]; Global[0.040]; forints[0.040]; Emese[0.040]; registered[0.040]; release[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hungary ==> ENTITY: \u001b[32mHungary\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.120:0.120[0]; log p(e|m)= -0.326:-0.326[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrepresents[0.053]; Poland[0.047]; foreign[0.044]; Hungarian[0.043]; conference[0.043]; trading[0.042]; Republic[0.042]; resignation[0.042]; general[0.041]; week[0.041]; Petofi[0.041]; Petofi[0.041]; Petofi[0.041]; Petofi[0.041]; Budapest[0.041]; Czech[0.041]; group[0.040]; meeting[0.040]; markets[0.040]; markets[0.040]; January[0.040]; months[0.040]; months[0.040]; directors[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kecskemet\u001b[39m ==> ENTITY: \u001b[32mKecskemét\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.110:0.110[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHungarian[0.045]; Budapest[0.043]; January[0.043]; manager[0.042]; manager[0.042]; Net[0.042]; performance[0.042]; June[0.042]; Petofi[0.041]; Petofi[0.041]; Petofi[0.041]; Petofi[0.041]; Petofi[0.041]; Petofi[0.041]; Petofi[0.041]; board[0.041]; units[0.041]; months[0.041]; months[0.041]; Laszlo[0.041]; site[0.041]; directors[0.041]; francs[0.041]; francs[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHungary[0.046]; Republic[0.043]; months[0.043]; months[0.043]; week[0.043]; Hungarian[0.043]; Czech[0.043]; half[0.043]; January[0.042]; Petofi[0.042]; Petofi[0.042]; Petofi[0.042]; Petofi[0.042]; June[0.042]; Budapest[0.042]; regarded[0.041]; manager[0.040]; group[0.040]; represents[0.040]; Laszlo[0.039]; think[0.039]; changed[0.039]; figures[0.039]; foreign[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncompany[0.044]; company[0.044]; second[0.043]; week[0.043]; goal[0.042]; following[0.042]; conference[0.042]; conference[0.042]; despite[0.042]; million[0.041]; expects[0.041]; Stephen[0.041]; Thursday[0.041]; said[0.041]; said[0.041]; said[0.041]; opening[0.041]; represents[0.041]; Friday[0.041]; Friday[0.041]; foreign[0.041]; release[0.041]; group[0.041]; group[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Budapest ==> ENTITY: \u001b[32mBudapest\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.030:-0.030[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPetofi[0.044]; Petofi[0.044]; Petofi[0.044]; Petofi[0.044]; Petofi[0.044]; Petofi[0.044]; Petofi[0.044]; Hungarian[0.044]; June[0.043]; January[0.043]; Kecskemet[0.041]; months[0.041]; house[0.041]; site[0.041]; leaving[0.040]; units[0.040]; French[0.040]; said[0.040]; said[0.040]; period[0.040]; serve[0.039]; submitted[0.039]; Laszlo[0.039]; francs[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech Republic ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.047]; Hungary[0.046]; January[0.044]; June[0.043]; represents[0.043]; million[0.042]; million[0.042]; million[0.042]; million[0.042]; foreign[0.041]; conference[0.041]; months[0.041]; months[0.041]; Hungarian[0.041]; opportunity[0.041]; opportunity[0.041]; Laszlo[0.041]; Petofi[0.040]; Petofi[0.040]; Petofi[0.040]; Petofi[0.040]; Petofi[0.040]; changed[0.040]; group[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Budapest ==> ENTITY: \u001b[32mBudapest\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.118:0.118[0]; log p(e|m)= -0.030:-0.030[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.047]; Budapest[0.046]; despite[0.045]; opening[0.045]; Vienna[0.043]; Thursday[0.042]; Friday[0.042]; Friday[0.042]; close[0.042]; opened[0.041]; forints[0.041]; registered[0.041]; France[0.041]; annual[0.041]; H1[0.041]; Emese[0.041]; following[0.040]; results[0.040]; results[0.040]; conference[0.040]; half[0.040]; half[0.040]; group[0.040]; group[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = BSE ==> ENTITY: \u001b[32mBudapest Stock Exchange\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.188:0.188[0]; log p(e|m)= -2.477:-2.477[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nExchange[0.045]; Depositary[0.044]; bourse[0.043]; GDRs[0.042]; Thursday[0.042]; conference[0.042]; company[0.042]; opening[0.042]; Friday[0.042]; Friday[0.042]; Shares[0.042]; opened[0.042]; following[0.042]; Budapest[0.042]; Budapest[0.042]; Stock[0.041]; Global[0.041]; annual[0.041]; forints[0.041]; Stephen[0.041]; million[0.040]; plunged[0.040]; sharply[0.040]; declined[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.071:0.071[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrench[0.048]; Republic[0.045]; fundamental[0.043]; fundamental[0.043]; issue[0.043]; Czech[0.042]; Hungarian[0.042]; fact[0.041]; think[0.041]; general[0.041]; general[0.041]; half[0.041]; Poland[0.041]; period[0.041]; said[0.041]; said[0.041]; said[0.041]; regarded[0.041]; months[0.041]; months[0.041]; January[0.041]; June[0.041]; position[0.040]; represents[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Budapest ==> ENTITY: \u001b[32mBudapest\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.092:0.092[0]; log p(e|m)= -0.030:-0.030[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.050]; ahead[0.049]; Hungary[0.048]; Poland[0.048]; opening[0.048]; operate[0.047]; Vienna[0.046]; million[0.046]; position[0.046]; operates[0.046]; Friday[0.045]; Czech[0.045]; represents[0.045]; week[0.044]; forints[0.044]; annual[0.044]; regarded[0.043]; point[0.043]; figures[0.043]; goal[0.043]; French[0.043]; net[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = BSE ==> ENTITY: \u001b[32mBudapest Stock Exchange\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.188:0.188[0]; log p(e|m)= -2.477:-2.477[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nExchange[0.047]; Depositary[0.046]; GDRs[0.046]; company[0.045]; bourse[0.045]; opened[0.044]; Shares[0.044]; Thursday[0.044]; conference[0.044]; opening[0.043]; Stock[0.043]; Friday[0.043]; Friday[0.043]; following[0.043]; Budapest[0.043]; Budapest[0.043]; Global[0.042]; annual[0.042]; forints[0.042]; Stephen[0.042]; million[0.042]; plunged[0.042]; sharply[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.072:0.072[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrench[0.049]; Republic[0.046]; fundamental[0.043]; fundamental[0.043]; Czech[0.042]; Hungarian[0.042]; fact[0.042]; serve[0.042]; think[0.042]; Poland[0.041]; period[0.041]; months[0.041]; months[0.041]; January[0.041]; June[0.041]; general[0.041]; general[0.041]; position[0.041]; represents[0.041]; said[0.040]; said[0.040]; said[0.040]; appointed[0.040]; group[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vienna ==> ENTITY: \u001b[32mVienna\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.118:0.118[0]; log p(e|m)= -0.126:-0.126[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBudapest[0.043]; Budapest[0.043]; Budapest[0.043]; opened[0.043]; release[0.043]; annual[0.043]; Thursday[0.043]; Stephen[0.042]; ahead[0.042]; released[0.042]; Friday[0.042]; Friday[0.042]; second[0.042]; week[0.041]; declined[0.041]; opening[0.041]; said[0.041]; close[0.041]; list[0.041]; expects[0.040]; group[0.040]; group[0.040]; million[0.040]; France[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hungarian ==> ENTITY: \u001b[32mHungary\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.155:0.155[0]; log p(e|m)= -1.033:-1.033[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHungary[0.049]; Poland[0.049]; Republic[0.046]; Budapest[0.044]; Petofi[0.044]; Petofi[0.044]; Petofi[0.044]; Petofi[0.044]; Petofi[0.044]; Petofi[0.044]; Petofi[0.044]; Czech[0.044]; million[0.042]; million[0.042]; million[0.042]; million[0.042]; months[0.042]; months[0.042]; Kecskemet[0.042]; said[0.041]; said[0.041]; compared[0.041]; reported[0.041]; \t\n\r [=========>....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s375ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1127/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1152testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dallas ==> ENTITY: \u001b[32mDallas\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.096:0.096[0]; log p(e|m)= -0.462:-0.462[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDallas[0.048]; Texas[0.047]; arrested[0.044]; traffic[0.044]; allegedly[0.044]; late[0.044]; Friday[0.043]; Saturday[0.042]; time[0.042]; police[0.042]; said[0.041]; old[0.041]; old[0.041]; killing[0.041]; ran[0.040]; group[0.040]; illegal[0.040]; vehicle[0.040]; man[0.040]; tried[0.040]; woman[0.039]; street[0.039]; avoid[0.039]; avoid[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dallas ==> ENTITY: \u001b[32mDallas\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.084:0.084[0]; log p(e|m)= -0.462:-0.462[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDallas[0.048]; Texas[0.046]; charged[0.045]; arrested[0.044]; traffic[0.044]; allegedly[0.044]; late[0.043]; counts[0.042]; Friday[0.042]; Saturday[0.042]; time[0.042]; police[0.041]; police[0.041]; said[0.040]; old[0.040]; old[0.040]; killing[0.040]; manslaughter[0.040]; ran[0.040]; group[0.040]; illegal[0.040]; vehicle[0.040]; man[0.039]; tried[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dallas ==> ENTITY: \u001b[32mDallas\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.073:0.073[0]; log p(e|m)= -0.462:-0.462[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncharged[0.045]; arrested[0.044]; traffic[0.044]; allegedly[0.044]; late[0.043]; straight[0.043]; counts[0.043]; Friday[0.043]; Saturday[0.042]; time[0.042]; police[0.042]; police[0.042]; illegally[0.041]; road[0.041]; stretch[0.041]; said[0.040]; said[0.040]; old[0.040]; old[0.040]; manslaughter[0.040]; ran[0.040]; group[0.040]; man[0.040]; tried[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Texas ==> ENTITY: \u001b[32mTexas\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.077:0.077[0]; log p(e|m)= -0.161:-0.161[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCar[0.045]; race[0.044]; race[0.044]; race[0.044]; race[0.044]; Dallas[0.044]; Dallas[0.044]; late[0.043]; Saturday[0.043]; Friday[0.043]; cars[0.042]; driver[0.042]; driver[0.042]; ran[0.041]; time[0.041]; group[0.041]; said[0.040]; slow[0.040]; injuring[0.039]; avoid[0.039]; avoid[0.039]; people[0.039]; preparing[0.039]; dozen[0.038]; \t\n\r [=========>....................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s488ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1131/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1050testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEschen\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUSV Eschen/Mauren\u001b[39m <---> \u001b[32mEschen\u001b[39m\t\nSCORES: global= 0.272:0.264[\u001b[31m0.008\u001b[39m]; local()= 0.113:0.086[\u001b[31m0.027\u001b[39m]; log p(e|m)= -2.323:0.000[\u001b[32m2.323\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.046]; Cup[0.045]; win[0.045]; qualifier[0.045]; biggest[0.044]; Liechtenstein[0.044]; Liechtenstein[0.044]; fifth[0.043]; Saturday[0.042]; second[0.042]; old[0.041]; beating[0.041]; Andy[0.041]; Keith[0.041]; Alpine[0.041]; Niall[0.040]; Win[0.040]; manager[0.040]; forward[0.040]; Ian[0.040]; Neill[0.040]; teenager[0.040]; McCarthy[0.040]; rout[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mick McCarthy\u001b[39m ==> ENTITY: \u001b[32mMick McCarthy\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.045]; Ireland[0.045]; Ireland[0.045]; Ireland[0.045]; Sunderland[0.043]; manager[0.043]; Irish[0.043]; Irish[0.043]; Irish[0.043]; Niall[0.042]; Keith[0.041]; added[0.041]; striker[0.041]; Neill[0.041]; Harte[0.040]; Republic[0.040]; Ian[0.040]; Andy[0.040]; Quinn[0.040]; Quinn[0.040]; Norwich[0.040]; Cup[0.040]; qualifier[0.040]; captain[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIreland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRepublic of Ireland\u001b[39m <---> \u001b[32mRepublic of Ireland national football team\u001b[39m\t\nSCORES: global= 0.242:0.235[\u001b[31m0.007\u001b[39m]; local()= 0.136:0.187[\u001b[32m0.051\u001b[39m]; log p(e|m)= -1.505:-4.423[\u001b[31m2.918\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.054]; Irish[0.048]; Irish[0.048]; Sunderland[0.043]; Charlton[0.043]; Norwich[0.042]; Cup[0.041]; Neill[0.041]; win[0.041]; qualifier[0.041]; McCarthy[0.041]; finals[0.041]; European[0.040]; Ian[0.040]; stadium[0.040]; second[0.040]; captain[0.040]; Keith[0.040]; Quinn[0.040]; Quinn[0.040]; place[0.040]; months[0.039]; fifth[0.039]; draw[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Quinn\u001b[39m ==> ENTITY: \u001b[32mNiall Quinn\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.045]; Ireland[0.045]; Ireland[0.045]; Sunderland[0.043]; Quinn[0.043]; Neill[0.042]; Ian[0.042]; Keith[0.042]; Irish[0.042]; Irish[0.042]; Irish[0.042]; McCarthy[0.042]; Charlton[0.041]; Andy[0.041]; second[0.041]; Norwich[0.041]; frustrating[0.041]; manager[0.041]; fifth[0.040]; World[0.040]; Saturday[0.040]; place[0.040]; forward[0.040]; added[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andy Townsend\u001b[39m ==> ENTITY: \u001b[32mAndy Townsend\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSunderland[0.045]; Norwich[0.044]; Charlton[0.044]; Ireland[0.044]; Ireland[0.044]; Ireland[0.044]; Ian[0.043]; Soccer[0.042]; Cup[0.042]; striker[0.042]; second[0.042]; Neill[0.041]; European[0.041]; manager[0.041]; finals[0.041]; Keith[0.041]; qualifier[0.041]; win[0.041]; World[0.040]; months[0.040]; Niall[0.040]; fifth[0.040]; helped[0.040]; draw[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jack Charlton\u001b[39m ==> ENTITY: \u001b[32mJack Charlton\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.047]; Ireland[0.047]; Andy[0.043]; Irish[0.043]; Irish[0.043]; manager[0.042]; Sunderland[0.042]; European[0.042]; Keith[0.042]; McCarthy[0.042]; Norwich[0.041]; Ian[0.041]; draw[0.041]; Neill[0.041]; months[0.041]; Quinn[0.040]; Quinn[0.040]; took[0.040]; qualifier[0.040]; forward[0.040]; Cup[0.040]; finals[0.040]; Niall[0.040]; second[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIreland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRepublic of Ireland\u001b[39m <---> \u001b[32mRepublic of Ireland national football team\u001b[39m\t\nSCORES: global= 0.244:0.235[\u001b[31m0.009\u001b[39m]; local()= 0.169:0.200[\u001b[32m0.032\u001b[39m]; log p(e|m)= -1.505:-4.423[\u001b[31m2.918\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.052]; Ireland[0.052]; Irish[0.046]; Irish[0.046]; Irish[0.046]; Sunderland[0.042]; Charlton[0.042]; Republic[0.040]; Norwich[0.040]; Soccer[0.040]; Cup[0.040]; Neill[0.040]; win[0.040]; qualifier[0.040]; McCarthy[0.040]; finals[0.040]; European[0.039]; Ian[0.039]; stadium[0.039]; second[0.039]; captain[0.039]; Keith[0.039]; Quinn[0.039]; Quinn[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norwich ==> ENTITY: \u001b[32mNorwich City F.C.\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.167:0.167[0]; log p(e|m)= -2.733:-2.733[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.047]; Sunderland[0.046]; manager[0.043]; draw[0.043]; Soccer[0.042]; second[0.042]; Cup[0.042]; Neill[0.042]; win[0.042]; ago[0.042]; Keith[0.042]; Andy[0.041]; months[0.041]; striker[0.041]; Win[0.041]; Ian[0.041]; took[0.041]; Saturday[0.040]; added[0.040]; championship[0.040]; minutes[0.040]; McCarthy[0.040]; stadium[0.040]; helped[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIrish\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIreland\u001b[39m <---> \u001b[32mRepublic of Ireland national football team\u001b[39m\t\nSCORES: global= 0.238:0.235[\u001b[31m0.003\u001b[39m]; local()= 0.141:0.191[\u001b[32m0.050\u001b[39m]; log p(e|m)= -1.457:-4.605[\u001b[31m3.148\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.054]; Ireland[0.054]; Irish[0.048]; Sunderland[0.043]; Cup[0.043]; Republic[0.042]; Niall[0.042]; Soccer[0.041]; qualifier[0.041]; Norwich[0.041]; McCarthy[0.041]; Neill[0.040]; Quinn[0.040]; Quinn[0.040]; second[0.040]; captain[0.040]; Saturday[0.040]; manager[0.039]; Liechtenstein[0.039]; Liechtenstein[0.039]; World[0.039]; Harte[0.039]; Mick[0.038]; added[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Republic of Ireland ==> ENTITY: \u001b[32mRepublic of Ireland national football team\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.188:0.188[0]; log p(e|m)= -1.461:-1.461[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.054]; Irish[0.047]; Irish[0.047]; Irish[0.047]; Cup[0.046]; Soccer[0.045]; Sunderland[0.044]; win[0.044]; beating[0.043]; Quinn[0.043]; Quinn[0.043]; McCarthy[0.043]; qualifier[0.043]; Norwich[0.042]; Saturday[0.042]; Neill[0.042]; second[0.042]; Niall[0.041]; manager[0.041]; fifth[0.041]; Harte[0.040]; captain[0.040]; Andy[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIrish\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIreland\u001b[39m <---> \u001b[32mRepublic of Ireland national football team\u001b[39m\t\nSCORES: global= 0.241:0.236[\u001b[31m0.005\u001b[39m]; local()= 0.169:0.204[\u001b[32m0.035\u001b[39m]; log p(e|m)= -1.457:-4.605[\u001b[31m3.148\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.053]; Ireland[0.053]; Ireland[0.053]; Irish[0.047]; Sunderland[0.042]; Republic[0.041]; Charlton[0.041]; Niall[0.041]; Cup[0.040]; win[0.040]; qualifier[0.040]; Norwich[0.040]; McCarthy[0.040]; finals[0.040]; European[0.040]; Neill[0.040]; Quinn[0.039]; Quinn[0.039]; second[0.039]; Jack[0.039]; captain[0.039]; place[0.039]; months[0.039]; stadium[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sunderland ==> ENTITY: \u001b[32mSunderland A.F.C.\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.396:-0.396[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.045]; second[0.044]; Norwich[0.043]; fifth[0.042]; Soccer[0.042]; Neill[0.042]; Ireland[0.042]; Ireland[0.042]; Ireland[0.042]; win[0.042]; finals[0.042]; draw[0.042]; Cup[0.042]; Ian[0.041]; Irish[0.041]; Irish[0.041]; Irish[0.041]; championship[0.041]; months[0.041]; minutes[0.041]; Quinn[0.041]; Quinn[0.041]; Keith[0.040]; forward[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIrish\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIreland\u001b[39m <---> \u001b[32mRepublic of Ireland national football team\u001b[39m\t\nSCORES: global= 0.239:0.235[\u001b[31m0.004\u001b[39m]; local()= 0.175:0.211[\u001b[32m0.036\u001b[39m]; log p(e|m)= -1.457:-4.605[\u001b[31m3.148\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.052]; Ireland[0.052]; Ireland[0.052]; Irish[0.047]; Irish[0.047]; Sunderland[0.042]; Cup[0.042]; Republic[0.040]; Charlton[0.040]; Niall[0.040]; Soccer[0.040]; qualifier[0.040]; Norwich[0.040]; McCarthy[0.040]; Neill[0.039]; Quinn[0.039]; Quinn[0.039]; second[0.039]; Jack[0.039]; captain[0.039]; stadium[0.039]; win[0.039]; Saturday[0.038]; months[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEschen\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUSV Eschen/Mauren\u001b[39m <---> \u001b[32mEschen\u001b[39m\t\nSCORES: global= 0.273:0.264[\u001b[31m0.008\u001b[39m]; local()= 0.122:0.093[\u001b[31m0.029\u001b[39m]; log p(e|m)= -2.323:0.000[\u001b[32m2.323\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfinals[0.046]; draw[0.045]; championship[0.045]; Cup[0.044]; win[0.044]; qualifier[0.044]; biggest[0.044]; Liechtenstein[0.043]; stadium[0.043]; fifth[0.042]; Jack[0.042]; Saturday[0.041]; second[0.041]; ago[0.040]; old[0.040]; beating[0.040]; Andy[0.040]; Keith[0.040]; Alpine[0.040]; Niall[0.040]; result[0.040]; manager[0.040]; forward[0.039]; Ian[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mLiechtenstein\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mLiechtenstein national football team\u001b[39m <---> \u001b[32mLiechtenstein\u001b[39m\t\nSCORES: global= 0.263:0.256[\u001b[31m0.008\u001b[39m]; local()= 0.176:0.109[\u001b[31m0.068\u001b[39m]; log p(e|m)= -1.995:-0.470[\u001b[32m1.525\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLiechtenstein[0.048]; Ireland[0.047]; Ireland[0.047]; qualifier[0.045]; Soccer[0.044]; Cup[0.044]; second[0.043]; Republic[0.042]; Saturday[0.041]; win[0.041]; Eschen[0.041]; Irish[0.041]; Irish[0.041]; Irish[0.041]; McCarthy[0.040]; fifth[0.040]; Alpine[0.040]; World[0.040]; minutes[0.040]; Win[0.039]; forward[0.039]; captain[0.039]; away[0.039]; beating[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Niall Quinn\u001b[39m ==> ENTITY: \u001b[32mNiall Quinn\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.045]; Ireland[0.045]; Ireland[0.045]; Sunderland[0.043]; Quinn[0.043]; Neill[0.042]; Ian[0.042]; Keith[0.042]; Irish[0.042]; Irish[0.042]; Irish[0.042]; McCarthy[0.042]; Charlton[0.041]; Andy[0.041]; second[0.041]; Norwich[0.041]; frustrating[0.041]; manager[0.041]; fifth[0.040]; World[0.040]; Saturday[0.040]; place[0.040]; forward[0.040]; added[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mLiechtenstein\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mLiechtenstein national football team\u001b[39m <---> \u001b[32mLiechtenstein\u001b[39m\t\nSCORES: global= 0.267:0.258[\u001b[31m0.008\u001b[39m]; local()= 0.189:0.123[\u001b[31m0.066\u001b[39m]; log p(e|m)= -1.995:-0.470[\u001b[32m1.525\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLiechtenstein[0.048]; Ireland[0.047]; Ireland[0.047]; Ireland[0.047]; qualifier[0.044]; Soccer[0.043]; second[0.042]; stadium[0.042]; Republic[0.042]; months[0.041]; win[0.040]; Cup[0.040]; Eschen[0.040]; Eschen[0.040]; Irish[0.040]; Irish[0.040]; Irish[0.040]; McCarthy[0.040]; fifth[0.040]; Alpine[0.040]; World[0.039]; minutes[0.039]; helped[0.039]; forward[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ian Harte\u001b[39m ==> ENTITY: \u001b[32mIan Harte\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.044]; Ireland[0.044]; Ireland[0.044]; Charlton[0.043]; months[0.043]; added[0.043]; Sunderland[0.042]; Andy[0.042]; Keith[0.042]; manager[0.042]; second[0.041]; finals[0.041]; took[0.041]; McCarthy[0.041]; helped[0.041]; qualifier[0.041]; Neill[0.041]; Saturday[0.041]; Irish[0.041]; Irish[0.041]; Irish[0.041]; Quinn[0.041]; Quinn[0.041]; Soccer[0.041]; \t\n\r [==========>...................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s475ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1150/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1108testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scotland ==> ENTITY: \u001b[32mScotland national football team\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.141:0.141[0]; log p(e|m)= -2.703:-2.703[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.069]; soccer[0.058]; Soccer[0.058]; qualifier[0.057]; Cup[0.057]; Cup[0.057]; World[0.057]; World[0.057]; European[0.056]; Qualifier[0.056]; Saturday[0.055]; Austria[0.054]; Austria[0.054]; group[0.052]; drew[0.052]; Attendance[0.051]; Draw[0.049]; Vienna[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vienna ==> ENTITY: \u001b[32mVienna\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.065:0.065[0]; log p(e|m)= -0.126:-0.126[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.061]; Austria[0.061]; Qualifier[0.060]; qualifier[0.058]; Cup[0.057]; Cup[0.057]; World[0.055]; World[0.055]; drew[0.055]; Saturday[0.054]; Scotland[0.054]; Scotland[0.054]; group[0.054]; soccer[0.054]; European[0.054]; Soccer[0.054]; Attendance[0.052]; Draw[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = European ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.081:0.081[0]; log p(e|m)= -0.256:-0.256[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.062]; Soccer[0.061]; Cup[0.059]; Cup[0.059]; qualifier[0.059]; Qualifier[0.057]; Scotland[0.057]; Scotland[0.057]; group[0.056]; Austria[0.055]; Austria[0.055]; drew[0.055]; Saturday[0.053]; World[0.052]; World[0.052]; Attendance[0.051]; Draw[0.051]; Vienna[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Cup ==> ENTITY: \u001b[32mFIFA World Cup\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.133:0.133[0]; log p(e|m)= -1.002:-1.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.066]; Soccer[0.064]; Cup[0.063]; World[0.063]; Austria[0.062]; Austria[0.062]; qualifier[0.060]; European[0.060]; Scotland[0.060]; Scotland[0.060]; Qualifier[0.060]; Draw[0.056]; Saturday[0.054]; drew[0.054]; group[0.054]; Attendance[0.052]; Vienna[0.051]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scotland ==> ENTITY: \u001b[32mScotland national football team\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.141:0.141[0]; log p(e|m)= -2.703:-2.703[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.069]; soccer[0.058]; Soccer[0.058]; qualifier[0.057]; Cup[0.057]; Cup[0.057]; World[0.057]; World[0.057]; European[0.056]; Qualifier[0.056]; Saturday[0.055]; Austria[0.054]; Austria[0.054]; group[0.052]; drew[0.052]; Attendance[0.051]; Draw[0.049]; Vienna[0.048]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAustria\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustria\u001b[39m <---> \u001b[32mAustria national football team\u001b[39m\t\nSCORES: global= 0.255:0.252[\u001b[31m0.002\u001b[39m]; local()= 0.117:0.161[\u001b[32m0.044\u001b[39m]; log p(e|m)= -0.252:-2.937[\u001b[31m2.685\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.064]; soccer[0.059]; Cup[0.058]; Cup[0.058]; qualifier[0.058]; World[0.057]; World[0.057]; Qualifier[0.057]; Soccer[0.057]; European[0.057]; Scotland[0.056]; Scotland[0.056]; Vienna[0.053]; Saturday[0.052]; drew[0.051]; group[0.051]; Draw[0.050]; Attendance[0.048]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAustria\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustria\u001b[39m <---> \u001b[32mAustria national football team\u001b[39m\t\nSCORES: global= 0.255:0.252[\u001b[31m0.002\u001b[39m]; local()= 0.117:0.161[\u001b[32m0.044\u001b[39m]; log p(e|m)= -0.252:-2.937[\u001b[31m2.685\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.064]; soccer[0.059]; Cup[0.058]; Cup[0.058]; qualifier[0.058]; World[0.057]; World[0.057]; Qualifier[0.057]; Soccer[0.057]; European[0.057]; Scotland[0.056]; Scotland[0.056]; Vienna[0.053]; Saturday[0.052]; drew[0.051]; group[0.051]; Draw[0.050]; Attendance[0.048]; \t\n\r [==========>...................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s477ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1158/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1079testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Benetton ==> ENTITY: \u001b[32mTreviso Basket\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.118:0.118[0]; log p(e|m)= -2.180:-2.180[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.051]; Italy[0.051]; basketball[0.049]; tournament[0.048]; Beograd[0.046]; Beograd[0.046]; Tournament[0.045]; international[0.045]; Yugoslavia[0.044]; Dinamo[0.043]; halftime[0.043]; Saturday[0.043]; place[0.043]; place[0.043]; Alba[0.042]; Belgrade[0.042]; Germany[0.042]; Red[0.041]; Russia[0.041]; Star[0.040]; Fifth[0.039]; Results[0.039]; Results[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.039:0.039[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.049]; Basketball[0.048]; Yugoslavia[0.047]; Belgrade[0.047]; international[0.047]; Germany[0.046]; basketball[0.046]; Italy[0.045]; Dinamo[0.045]; Tournament[0.045]; Saturday[0.044]; place[0.043]; place[0.043]; Beograd[0.042]; Beograd[0.042]; halftime[0.042]; Star[0.041]; Benetton[0.041]; Red[0.041]; Results[0.040]; Results[0.040]; Fifth[0.039]; Alba[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Yugoslavia ==> ENTITY: \u001b[32mYugoslavia\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.084:0.084[0]; log p(e|m)= -0.654:-0.654[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.049]; tournament[0.048]; Belgrade[0.048]; international[0.048]; Germany[0.047]; Basketball[0.046]; Russia[0.045]; basketball[0.045]; Beograd[0.044]; Beograd[0.044]; place[0.043]; place[0.043]; Fifth[0.043]; Tournament[0.042]; Saturday[0.042]; Dinamo[0.042]; Red[0.041]; halftime[0.041]; Star[0.040]; Results[0.040]; Results[0.040]; Benetton[0.039]; Alba[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.066:0.066[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.050]; international[0.048]; Yugoslavia[0.048]; Tournament[0.047]; Basketball[0.047]; Germany[0.047]; place[0.045]; place[0.045]; Saturday[0.045]; Russia[0.044]; basketball[0.044]; Dinamo[0.044]; Belgrade[0.042]; Benetton[0.042]; halftime[0.042]; Beograd[0.041]; Beograd[0.041]; Star[0.040]; Results[0.040]; Results[0.040]; Red[0.040]; Alba[0.040]; Fifth[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mRed Star\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRed Star Belgrade\u001b[39m <---> \u001b[32mKK Crvena zvezda\u001b[39m\t\nSCORES: global= 0.263:0.255[\u001b[31m0.008\u001b[39m]; local()= 0.111:0.140[\u001b[32m0.029\u001b[39m]; log p(e|m)= -1.036:-2.465[\u001b[31m1.429\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.053]; Belgrade[0.052]; tournament[0.050]; basketball[0.050]; Beograd[0.050]; Beograd[0.050]; Tournament[0.048]; Dinamo[0.047]; Yugoslavia[0.047]; international[0.047]; Italy[0.045]; Saturday[0.045]; place[0.044]; place[0.044]; Russia[0.044]; Germany[0.043]; halftime[0.042]; Benetton[0.041]; Alba[0.041]; Fifth[0.040]; Results[0.039]; Results[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.045:0.045[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.051]; Basketball[0.051]; Yugoslavia[0.048]; Italy[0.047]; basketball[0.046]; international[0.046]; Tournament[0.046]; Russia[0.045]; place[0.045]; place[0.045]; Belgrade[0.044]; Beograd[0.042]; Beograd[0.042]; Dinamo[0.042]; halftime[0.042]; Saturday[0.041]; Red[0.041]; Results[0.040]; Results[0.040]; Star[0.040]; Benetton[0.039]; Fifth[0.039]; Alba[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgrade ==> ENTITY: \u001b[32mBelgrade\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.090:0.090[0]; log p(e|m)= -0.052:-0.052[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYugoslavia[0.049]; tournament[0.049]; Beograd[0.047]; Beograd[0.047]; Basketball[0.047]; Germany[0.046]; basketball[0.045]; Saturday[0.045]; Tournament[0.045]; place[0.045]; place[0.045]; Italy[0.044]; international[0.044]; Red[0.043]; halftime[0.042]; Russia[0.041]; Dinamo[0.041]; Star[0.041]; Fifth[0.041]; Benetton[0.040]; Alba[0.038]; Results[0.038]; Results[0.038]; \t\n\r [==========>...................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s554ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1165/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1093testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.056:0.056[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.048]; Cup[0.044]; Cup[0.044]; Tennis[0.044]; Saturday[0.043]; championships[0.042]; Tournament[0.042]; finals[0.042]; Team[0.042]; tennis[0.042]; semifinals[0.042]; round[0.041]; round[0.041]; March[0.041]; March[0.041]; World[0.041]; World[0.041]; July[0.040]; South[0.040]; South[0.040]; Women[0.040]; Korea[0.040]; States[0.039]; September[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. Open ==> ENTITY: \u001b[32mUS Open (tennis)\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.179:0.179[0]; log p(e|m)= -1.238:-1.238[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.051]; tennis[0.051]; Open[0.046]; round[0.045]; round[0.045]; semifinals[0.044]; finals[0.044]; South[0.044]; South[0.044]; Women[0.043]; championships[0.043]; Switzerland[0.043]; Tournament[0.042]; Cup[0.042]; Cup[0.042]; Australia[0.042]; World[0.042]; World[0.042]; Germany[0.041]; team[0.041]; York[0.041]; Argentina[0.041]; women[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = 1997 Fed Cup\u001b[39m ==> ENTITY: \u001b[32m1997 Fed Cup\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWorld[0.043]; World[0.043]; Argentina[0.043]; Cup[0.043]; Australia[0.043]; Spain[0.043]; Group[0.043]; Group[0.043]; Switzerland[0.042]; Croatia[0.042]; Belgium[0.041]; Austria[0.041]; France[0.041]; Open[0.041]; round[0.041]; round[0.041]; semifinals[0.041]; Netherlands[0.041]; finals[0.041]; Germany[0.040]; Japan[0.040]; Africa[0.040]; South[0.040]; South[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mUnited States\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUnited States\u001b[39m <---> \u001b[32mUnited States Fed Cup team\u001b[39m\t\nSCORES: global= 0.261:0.215[\u001b[31m0.046\u001b[39m]; local()= 0.081:0.177[\u001b[32m0.096\u001b[39m]; log p(e|m)= -0.074:-5.298[\u001b[31m5.225\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.045]; championships[0.045]; Argentina[0.044]; Australia[0.044]; Cup[0.043]; Cup[0.043]; Switzerland[0.043]; Spain[0.043]; finals[0.043]; Team[0.042]; Tennis[0.041]; Austria[0.041]; Open[0.041]; semifinals[0.041]; Korea[0.041]; Germany[0.041]; tennis[0.040]; South[0.040]; South[0.040]; Saturday[0.040]; France[0.040]; Netherlands[0.040]; World[0.040]; World[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = 1997 Fed Cup\u001b[39m ==> ENTITY: \u001b[32m1997 Fed Cup\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWorld[0.043]; World[0.043]; Argentina[0.043]; Cup[0.043]; Australia[0.043]; Spain[0.043]; Group[0.043]; Group[0.043]; Switzerland[0.042]; Croatia[0.042]; Belgium[0.041]; Austria[0.041]; France[0.041]; Open[0.041]; round[0.041]; round[0.041]; semifinals[0.041]; Netherlands[0.041]; finals[0.041]; Germany[0.040]; Japan[0.040]; Africa[0.040]; South[0.040]; South[0.040]; \t\n\r [==========>...................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s628ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1172/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1084testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgrade ==> ENTITY: \u001b[32mBelgrade\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.101:0.101[0]; log p(e|m)= -0.052:-0.052[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.063]; play[0.060]; Beograd[0.060]; Basketball[0.060]; basketball[0.058]; Saturday[0.057]; place[0.057]; Italy[0.057]; international[0.056]; fifth[0.055]; beat[0.054]; halftime[0.053]; Russia[0.053]; Dinamo[0.053]; Dinamo[0.053]; Benetton[0.051]; Benetton[0.051]; Beat[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Benetton ==> ENTITY: \u001b[32mTreviso Basket\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.165:0.165[0]; log p(e|m)= -2.180:-2.180[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBenetton[0.064]; Basketball[0.063]; Italy[0.063]; basketball[0.060]; tournament[0.059]; play[0.058]; fifth[0.056]; international[0.055]; beat[0.054]; Dinamo[0.053]; Dinamo[0.053]; halftime[0.053]; Saturday[0.053]; place[0.053]; Belgrade[0.052]; Beograd[0.051]; Russia[0.050]; Beat[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.075:0.075[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.066]; tournament[0.060]; basketball[0.059]; fifth[0.058]; international[0.057]; play[0.056]; Benetton[0.056]; Benetton[0.056]; place[0.055]; Saturday[0.055]; beat[0.054]; Dinamo[0.054]; Dinamo[0.054]; Russia[0.053]; Belgrade[0.053]; Beograd[0.052]; halftime[0.052]; Beat[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.053:0.053[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfifth[0.061]; tournament[0.060]; Basketball[0.059]; play[0.058]; Belgrade[0.058]; international[0.058]; basketball[0.057]; Italy[0.056]; Dinamo[0.056]; Dinamo[0.056]; beat[0.056]; Saturday[0.054]; place[0.054]; Beograd[0.052]; Beat[0.052]; halftime[0.052]; Benetton[0.051]; Benetton[0.051]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Benetton ==> ENTITY: \u001b[32mTreviso Basket\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.165:0.165[0]; log p(e|m)= -2.180:-2.180[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBenetton[0.064]; Basketball[0.063]; Italy[0.063]; basketball[0.060]; tournament[0.059]; play[0.058]; fifth[0.056]; international[0.055]; beat[0.054]; Dinamo[0.053]; Dinamo[0.053]; halftime[0.053]; Saturday[0.053]; place[0.053]; Belgrade[0.052]; Beograd[0.051]; Russia[0.050]; Beat[0.050]; \t\n\r [==========>...................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s673ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1177/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1047testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = La Coruna\u001b[39m ==> ENTITY: \u001b[32mA Coruña\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.050]; Coruna[0.049]; Spanish[0.048]; Roberto[0.047]; Carlos[0.046]; Madrid[0.044]; Deportivo[0.041]; Deportivo[0.041]; peseta[0.041]; playmaker[0.041]; Saturday[0.041]; coach[0.040]; Auxerre[0.040]; taking[0.040]; home[0.040]; headed[0.039]; opening[0.039]; effort[0.039]; half[0.039]; started[0.039]; Fabio[0.039]; Soccer[0.039]; Scrape[0.039]; game[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Real ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.187:0.187[0]; log p(e|m)= -3.352:-3.352[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.052]; Brazilian[0.050]; Lendoiro[0.047]; players[0.047]; Real[0.046]; Roberto[0.046]; Deportivo[0.046]; Deportivo[0.046]; Deportivo[0.046]; match[0.046]; match[0.046]; Luis[0.045]; squad[0.045]; Carlos[0.045]; midfielder[0.044]; playing[0.044]; Europe[0.044]; Alvarez[0.044]; chances[0.044]; national[0.043]; Augusto[0.043]; break[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Donato ==> ENTITY: \u001b[32mDonato Gama da Silva\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.227:0.227[0]; log p(e|m)= -0.865:-0.865[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRoberto[0.043]; Roberto[0.043]; Deportivo[0.042]; Deportivo[0.042]; Deportivo[0.042]; Deportivo[0.042]; international[0.042]; Brazilian[0.042]; Brazilian[0.042]; Donato[0.042]; midfielder[0.042]; Fabio[0.042]; past[0.042]; Luis[0.041]; Real[0.041]; Real[0.041]; Real[0.041]; Real[0.041]; Madrid[0.041]; Spanish[0.041]; Spanish[0.041]; playmaker[0.041]; started[0.041]; shortly[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Real ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.204:0.204[0]; log p(e|m)= -3.352:-3.352[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazilian[0.045]; Brazilian[0.045]; Soccer[0.044]; Madrid[0.043]; Spain[0.043]; billion[0.043]; peseta[0.042]; Real[0.042]; Real[0.042]; Real[0.042]; Roberto[0.042]; Coruna[0.041]; Coruna[0.041]; Deportivo[0.041]; Deportivo[0.041]; Deportivo[0.041]; Luis[0.041]; Carlos[0.040]; game[0.040]; Spanish[0.040]; Spanish[0.040]; draw[0.040]; defender[0.040]; midfielder[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazil ==> ENTITY: \u001b[32mBrazil national football team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.164:0.164[0]; log p(e|m)= -2.590:-2.590[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayers[0.045]; match[0.045]; match[0.045]; national[0.045]; Brazilian[0.044]; Fifa[0.044]; playing[0.044]; squad[0.043]; Augusto[0.043]; Europe[0.042]; final[0.042]; coach[0.042]; event[0.041]; decided[0.040]; Real[0.040]; chances[0.040]; midfielder[0.040]; minutes[0.039]; Mauro[0.039]; kick[0.039]; Silva[0.039]; Silva[0.039]; tour[0.039]; Lendoiro[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fabio Capello\u001b[39m ==> ENTITY: \u001b[32mFabio Capello\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.044]; coach[0.044]; Spain[0.043]; game[0.043]; international[0.043]; Match[0.043]; defender[0.042]; midfielder[0.042]; draw[0.042]; signed[0.042]; Real[0.042]; Real[0.042]; Real[0.042]; Real[0.042]; Madrid[0.041]; playmaker[0.041]; Coruna[0.040]; Coruna[0.040]; Roberto[0.040]; shortly[0.040]; Deportivo[0.040]; Deportivo[0.040]; goal[0.040]; earn[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Real ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.198:0.198[0]; log p(e|m)= -3.352:-3.352[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazilian[0.045]; Brazilian[0.045]; Lendoiro[0.043]; players[0.042]; Real[0.042]; Real[0.042]; Roberto[0.042]; Coruna[0.042]; Deportivo[0.042]; Deportivo[0.042]; Deportivo[0.042]; Deportivo[0.042]; Deportivo[0.042]; match[0.042]; match[0.042]; Luis[0.041]; Carlos[0.041]; Spanish[0.041]; midfielder[0.041]; midfielder[0.041]; born[0.040]; playing[0.040]; Europe[0.040]; Alvarez[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mauro Silva\u001b[39m ==> ENTITY: \u001b[32mMauro Silva\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayers[0.044]; Deportivo[0.044]; Deportivo[0.044]; Lendoiro[0.044]; Roberto[0.044]; Carlos[0.044]; Silva[0.043]; Brazilian[0.043]; squad[0.043]; Brazil[0.042]; midfielder[0.042]; Real[0.041]; Real[0.041]; Augusto[0.041]; Fifa[0.041]; match[0.041]; match[0.041]; playing[0.041]; Armando[0.040]; minutes[0.040]; minutes[0.040]; coach[0.040]; Songo[0.040]; Alvarez[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Luis Milla\u001b[39m ==> ENTITY: \u001b[32mLuis Milla\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayers[0.044]; Roberto[0.043]; Fabio[0.043]; final[0.043]; Armando[0.043]; game[0.043]; title[0.042]; Deportivo[0.042]; Deportivo[0.042]; Deportivo[0.042]; Deportivo[0.042]; Capello[0.042]; Spanish[0.042]; Spanish[0.042]; championship[0.042]; draw[0.042]; Carlos[0.041]; midfielder[0.041]; started[0.041]; Songo[0.040]; playmaker[0.040]; bookable[0.040]; Coruna[0.040]; peseta[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Auxerre ==> ENTITY: \u001b[32mAJ Auxerre\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.305:-0.305[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnumerical[0.043]; home[0.043]; home[0.043]; Jacques[0.043]; game[0.043]; draw[0.042]; Deportivo[0.042]; Deportivo[0.042]; Deportivo[0.042]; Deportivo[0.042]; Coruna[0.042]; Coruna[0.042]; line[0.041]; Match[0.041]; midfielder[0.041]; championship[0.041]; defender[0.041]; time[0.041]; minutes[0.041]; minutes[0.041]; advantage[0.041]; sent[0.040]; sent[0.040]; title[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spanish ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.148:0.148[0]; log p(e|m)= -1.241:-1.241[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.051]; Spanish[0.051]; Soccer[0.044]; international[0.042]; Luis[0.042]; Carlos[0.042]; Roberto[0.042]; Madrid[0.042]; game[0.042]; draw[0.041]; Brazilian[0.041]; Brazilian[0.041]; peseta[0.041]; Match[0.041]; opening[0.040]; La[0.040]; signed[0.040]; Fabio[0.040]; Capello[0.040]; championship[0.040]; Real[0.039]; Real[0.039]; Real[0.039]; Real[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Deportivo\u001b[39m ==> ENTITY: \u001b[32mDeportivo de La Coruña\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDeportivo[0.046]; Deportivo[0.046]; Roberto[0.044]; Carlos[0.043]; players[0.043]; Lendoiro[0.043]; playing[0.043]; Brazilian[0.042]; match[0.042]; match[0.042]; Brazil[0.042]; Augusto[0.041]; national[0.041]; squad[0.041]; Armando[0.041]; Real[0.041]; Real[0.041]; time[0.040]; Silva[0.040]; Silva[0.040]; midfielder[0.040]; minutes[0.040]; minutes[0.040]; minutes[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.241:0.241[0]; local()= 0.111:0.111[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpanish[0.045]; Soccer[0.044]; Madrid[0.044]; Carlos[0.043]; Coruna[0.043]; Coruna[0.043]; Deportivo[0.042]; Deportivo[0.042]; Roberto[0.042]; championship[0.042]; game[0.042]; draw[0.042]; point[0.042]; Match[0.041]; earn[0.041]; Real[0.041]; Real[0.041]; Real[0.041]; Fabio[0.040]; Capello[0.040]; Brazilian[0.040]; opening[0.040]; goal[0.040]; started[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Deportivo\u001b[39m ==> ENTITY: \u001b[32mDeportivo de La Coruña\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.235:0.235[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDeportivo[0.045]; Deportivo[0.045]; Luis[0.044]; Roberto[0.043]; Carlos[0.043]; players[0.043]; Spanish[0.042]; Lendoiro[0.042]; playing[0.042]; Brazilian[0.042]; Brazilian[0.042]; match[0.042]; match[0.042]; international[0.041]; Brazil[0.041]; Augusto[0.041]; national[0.040]; squad[0.040]; Armando[0.040]; Real[0.040]; Real[0.040]; Real[0.040]; Auxerre[0.040]; time[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spanish ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.147:0.147[0]; log p(e|m)= -1.241:-1.241[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpanish[0.051]; peseta[0.043]; international[0.043]; Luis[0.043]; Carlos[0.043]; Carlos[0.043]; Roberto[0.042]; Roberto[0.042]; Madrid[0.042]; game[0.042]; draw[0.042]; Brazilian[0.041]; Brazilian[0.041]; opening[0.041]; Alvarez[0.041]; Fabio[0.040]; past[0.040]; Capello[0.040]; Real[0.040]; Real[0.040]; Real[0.040]; Real[0.040]; championship[0.040]; Armando[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazilian ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.152:0.152[0]; log p(e|m)= -0.273:-0.273[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpanish[0.048]; Spanish[0.048]; Brazilian[0.047]; Soccer[0.046]; Roberto[0.046]; Carlos[0.046]; Luis[0.045]; Spain[0.044]; Match[0.044]; game[0.043]; born[0.043]; international[0.042]; Deportivo[0.042]; Deportivo[0.042]; Deportivo[0.042]; Fabio[0.042]; Real[0.042]; Real[0.042]; Real[0.042]; Real[0.042]; draw[0.041]; title[0.041]; championship[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Europe ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.056:0.056[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnational[0.046]; tour[0.045]; Fifa[0.045]; match[0.044]; match[0.044]; players[0.043]; squad[0.043]; John[0.042]; final[0.042]; playing[0.042]; event[0.041]; Deportivo[0.041]; Deportivo[0.041]; Brazil[0.041]; Real[0.040]; Real[0.040]; Lendoiro[0.040]; past[0.040]; decision[0.040]; banning[0.040]; Toshack[0.040]; failing[0.040]; decided[0.040]; time[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Deportivo\u001b[39m ==> ENTITY: \u001b[32mDeportivo de La Coruña\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.255:0.255[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDeportivo[0.044]; Deportivo[0.044]; Deportivo[0.044]; Deportivo[0.044]; Luis[0.043]; Roberto[0.043]; Coruna[0.043]; Carlos[0.042]; game[0.042]; players[0.042]; Spanish[0.042]; Spanish[0.042]; Lendoiro[0.041]; Brazilian[0.041]; Brazilian[0.041]; match[0.041]; international[0.041]; Augusto[0.040]; Saturday[0.040]; Armando[0.040]; Real[0.039]; Real[0.039]; Real[0.039]; Real[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = FIFA\u001b[39m ==> ENTITY: \u001b[32mFIFA\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.118:0.118[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsquad[0.046]; playing[0.045]; players[0.045]; Brazil[0.044]; national[0.044]; match[0.043]; match[0.043]; Real[0.043]; Real[0.043]; Europe[0.043]; final[0.041]; Brazilian[0.041]; John[0.040]; event[0.040]; Deportivo[0.040]; Deportivo[0.040]; Deportivo[0.040]; tour[0.040]; coach[0.040]; midfielder[0.040]; Roberto[0.040]; chances[0.040]; time[0.040]; kick[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Real ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.206:0.206[0]; log p(e|m)= -3.352:-3.352[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazilian[0.045]; Brazilian[0.045]; Madrid[0.044]; billion[0.043]; peseta[0.043]; Real[0.042]; Real[0.042]; Real[0.042]; Roberto[0.042]; Coruna[0.042]; Deportivo[0.041]; Deportivo[0.041]; Deportivo[0.041]; Deportivo[0.041]; Luis[0.041]; Carlos[0.041]; Carlos[0.041]; game[0.041]; Spanish[0.040]; Spanish[0.040]; draw[0.040]; midfielder[0.040]; born[0.040]; Fabio[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Deportivo\u001b[39m ==> ENTITY: \u001b[32mDeportivo de La Coruña\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.254:0.254[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDeportivo[0.050]; Deportivo[0.050]; Soccer[0.050]; La[0.049]; Luis[0.049]; Spain[0.048]; Roberto[0.048]; Roberto[0.048]; Madrid[0.048]; Coruna[0.048]; Coruna[0.048]; Carlos[0.048]; Carlos[0.048]; game[0.047]; Spanish[0.047]; Spanish[0.047]; Brazilian[0.046]; Brazilian[0.046]; international[0.046]; Saturday[0.045]; Armando[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Real ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.197:0.197[0]; log p(e|m)= -3.352:-3.352[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.046]; Brazilian[0.045]; Real[0.044]; Real[0.044]; Madrid[0.043]; Soccer[0.043]; Spanish[0.043]; billion[0.043]; Deportivo[0.043]; Deportivo[0.043]; peseta[0.042]; Roberto[0.042]; Coruna[0.042]; Coruna[0.042]; Carlos[0.041]; game[0.040]; draw[0.040]; defender[0.040]; Fabio[0.040]; goal[0.039]; Match[0.039]; started[0.039]; title[0.039]; signed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Roberto Carlos ==> ENTITY: \u001b[32mRoberto Carlos (footballer)\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.196:0.196[0]; log p(e|m)= -0.057:-0.057[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazilian[0.046]; Brazilian[0.046]; Soccer[0.045]; born[0.044]; defender[0.043]; Fabio[0.043]; Spanish[0.043]; Spanish[0.043]; Madrid[0.041]; game[0.041]; Spain[0.041]; Coruna[0.041]; Coruna[0.041]; goal[0.041]; Match[0.041]; Real[0.041]; Real[0.041]; Real[0.041]; Real[0.041]; playmaker[0.041]; midfielder[0.040]; La[0.040]; Deportivo[0.040]; Deportivo[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Donato ==> ENTITY: \u001b[32mDonato Gama da Silva\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.218:0.218[0]; log p(e|m)= -0.865:-0.865[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnational[0.046]; players[0.045]; Mauro[0.044]; coach[0.044]; Brazil[0.044]; playing[0.044]; Roberto[0.044]; match[0.044]; match[0.044]; Deportivo[0.043]; Deportivo[0.043]; Deportivo[0.043]; international[0.043]; Brazilian[0.043]; Brazilian[0.043]; Donato[0.043]; Lendoiro[0.043]; squad[0.043]; midfielder[0.043]; midfielder[0.043]; past[0.043]; Augusto[0.042]; Luis[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = John Toshack\u001b[39m ==> ENTITY: \u001b[32mJohn Toshack\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsquad[0.044]; players[0.044]; Carlos[0.043]; match[0.042]; match[0.042]; Brazil[0.042]; Lendoiro[0.042]; playing[0.042]; coach[0.042]; national[0.042]; Europe[0.042]; minutes[0.042]; Songo[0.042]; booked[0.041]; midfielder[0.041]; Deportivo[0.041]; Deportivo[0.041]; kick[0.041]; Brazilian[0.040]; lost[0.040]; Real[0.040]; Real[0.040]; decision[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Roberto Carlos ==> ENTITY: \u001b[32mRoberto Carlos (footballer)\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.057:-0.057[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.049]; Augusto[0.046]; Armando[0.045]; Brazilian[0.045]; Brazilian[0.045]; born[0.043]; players[0.043]; Lendoiro[0.042]; Spanish[0.042]; match[0.042]; match[0.042]; Mauro[0.040]; Coruna[0.040]; Fifa[0.040]; Real[0.040]; Real[0.040]; Real[0.040]; playmaker[0.040]; John[0.040]; midfielder[0.040]; midfielder[0.040]; kick[0.040]; playing[0.039]; keeper[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Brazilian-born\u001b[39m ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.132:0.132[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazilian[0.046]; international[0.043]; Roberto[0.043]; Roberto[0.043]; Spanish[0.043]; Spanish[0.043]; Carlos[0.042]; Carlos[0.042]; Luis[0.042]; Songo[0.041]; playmaker[0.041]; bookable[0.041]; Fabio[0.041]; Coruna[0.041]; Deportivo[0.041]; Deportivo[0.041]; Deportivo[0.041]; Deportivo[0.041]; game[0.041]; coach[0.041]; born[0.040]; Auxerre[0.040]; Madrid[0.040]; midfielder[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazilian ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.140:0.140[0]; log p(e|m)= -0.273:-0.273[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.047]; Roberto[0.044]; Carlos[0.044]; players[0.043]; match[0.043]; match[0.043]; Augusto[0.042]; Fifa[0.042]; national[0.042]; playing[0.042]; Mauro[0.042]; squad[0.041]; Deportivo[0.041]; Deportivo[0.041]; Silva[0.041]; Silva[0.041]; Armando[0.041]; Real[0.041]; Real[0.041]; Europe[0.040]; final[0.039]; past[0.039]; Lendoiro[0.039]; numerical[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Madrid\u001b[39m ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.046]; Carlos[0.044]; Spain[0.044]; Deportivo[0.043]; Deportivo[0.043]; game[0.043]; championship[0.042]; Roberto[0.042]; Coruna[0.042]; Coruna[0.042]; minute[0.041]; goal[0.041]; Real[0.041]; Real[0.041]; Real[0.041]; home[0.041]; started[0.041]; draw[0.041]; Spanish[0.040]; Spanish[0.040]; peseta[0.040]; Fabio[0.040]; midfielder[0.040]; signed[0.040]; \t\n\r [==========>...................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s598ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1206/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 951testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mRed Star\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRed Star Belgrade\u001b[39m <---> \u001b[32mKK Crvena zvezda\u001b[39m\t\nSCORES: global= 0.265:0.246[\u001b[31m0.019\u001b[39m]; local()= 0.117:0.152[\u001b[32m0.035\u001b[39m]; log p(e|m)= -1.036:-2.465[\u001b[31m1.429\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBasketball[0.076]; Belgrade[0.074]; tournament[0.072]; basketball[0.072]; Tournament[0.068]; Dinamo[0.068]; Yugoslavia[0.068]; international[0.067]; International[0.064]; Friday[0.064]; beat[0.063]; Russia[0.063]; halftime[0.061]; Result[0.060]; Result[0.060]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Yugoslavia ==> ENTITY: \u001b[32mYugoslavia\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.076:0.076[0]; log p(e|m)= -0.654:-0.654[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.070]; Belgrade[0.070]; international[0.069]; Basketball[0.066]; Russia[0.065]; basketball[0.065]; beat[0.063]; Tournament[0.061]; International[0.061]; Dinamo[0.061]; Red[0.060]; halftime[0.059]; Friday[0.059]; Star[0.058]; Result[0.056]; Result[0.056]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.044:0.044[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.069]; Basketball[0.068]; Yugoslavia[0.066]; Belgrade[0.066]; international[0.066]; basketball[0.065]; Dinamo[0.064]; Tournament[0.064]; beat[0.064]; International[0.061]; Friday[0.060]; halftime[0.059]; Star[0.058]; Red[0.057]; Result[0.056]; Result[0.056]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgrade ==> ENTITY: \u001b[32mBelgrade\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.080:0.080[0]; log p(e|m)= -0.052:-0.052[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYugoslavia[0.071]; tournament[0.071]; Basketball[0.067]; basketball[0.065]; Tournament[0.064]; Friday[0.064]; international[0.063]; International[0.063]; Red[0.062]; beat[0.061]; halftime[0.060]; Russia[0.060]; Dinamo[0.060]; Star[0.059]; Result[0.055]; Result[0.055]; \t\n\r [==========>...................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s661ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1210/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1025testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nasdaq\u001b[39m ==> ENTITY: \u001b[32mNASDAQ\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncompany[0.048]; shares[0.045]; corporate[0.044]; Financial[0.044]; provider[0.044]; markets[0.043]; Stock[0.043]; symbol[0.042]; offered[0.041]; offered[0.041]; price[0.041]; Total[0.041]; Filing[0.041]; outstanding[0.040]; Business[0.040]; Data[0.040]; Net[0.040]; Hospitality[0.040]; Hospitality[0.040]; extended[0.040]; extended[0.040]; Company[0.040]; Company[0.040]; ipo[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.045:0.045[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnational[0.048]; selected[0.047]; Washington[0.046]; goal[0.044]; Bear[0.043]; extended[0.042]; extended[0.042]; outstanding[0.042]; stay[0.042]; stay[0.042]; provider[0.041]; offered[0.041]; offered[0.041]; general[0.041]; strategically[0.041]; high[0.040]; Stearns[0.040]; corporate[0.040]; Stock[0.040]; range[0.040]; finance[0.040]; Lead[0.040]; located[0.039]; Business[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.035:0.035[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnational[0.045]; Estimated[0.045]; States[0.045]; United[0.043]; located[0.043]; Business[0.043]; hotels[0.043]; hotels[0.043]; general[0.042]; selected[0.042]; development[0.041]; high[0.041]; extended[0.041]; extended[0.041]; Stearns[0.041]; Use[0.041]; additional[0.041]; finance[0.040]; Total[0.040]; offered[0.040]; offered[0.040]; Financial[0.040]; company[0.040]; allotment[0.040]; \t\n\r [==========>...................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s771ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1213/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1129testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Warsaw ==> ENTITY: \u001b[32mWarsaw\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.086:0.086[0]; log p(e|m)= -0.143:-0.143[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPolish[0.045]; Polish[0.045]; Polish[0.045]; Polish[0.045]; plans[0.043]; bid[0.043]; water[0.043]; percent[0.042]; percent[0.042]; percent[0.042]; Warszawy[0.042]; Warszawy[0.042]; state[0.041]; state[0.041]; Saturday[0.041]; firm[0.041]; distribution[0.040]; million[0.040]; Solorz[0.040]; press[0.040]; Bank[0.040]; rival[0.039]; Sa[0.039]; Sa[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zycie Warszawy\u001b[39m ==> ENTITY: \u001b[32mŻycie Warszawy\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarszawy[0.048]; Warsaw[0.047]; Warsaw[0.047]; publisher[0.046]; Polish[0.046]; Polish[0.046]; Polish[0.046]; Polish[0.046]; quoted[0.045]; press[0.043]; press[0.043]; daily[0.043]; paper[0.043]; paper[0.043]; newsstand[0.042]; Sa[0.042]; Sa[0.042]; owned[0.042]; saying[0.040]; said[0.040]; said[0.040]; Hachette[0.040]; division[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swiss ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwiss[0.046]; Monday[0.045]; million[0.043]; sum[0.043]; offering[0.043]; listed[0.043]; group[0.043]; Poland[0.042]; March[0.042]; Polish[0.042]; Polish[0.042]; plans[0.042]; change[0.041]; bid[0.041]; said[0.041]; said[0.041]; extend[0.041]; included[0.041]; holding[0.040]; bids[0.040]; invest[0.040]; initially[0.040]; increase[0.040]; investment[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.049]; Poland[0.049]; Polish[0.045]; Polish[0.045]; Warsaw[0.045]; country[0.045]; communism[0.043]; Monday[0.041]; support[0.041]; news[0.040]; President[0.040]; said[0.040]; said[0.040]; conditions[0.040]; Wieslaw[0.040]; September[0.040]; total[0.040]; press[0.039]; press[0.039]; press[0.039]; saying[0.039]; distribution[0.039]; stands[0.039]; visits[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.080:0.080[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.046]; Jacques[0.044]; Polish[0.044]; Polish[0.044]; Chirac[0.042]; press[0.042]; press[0.042]; press[0.042]; state[0.042]; Poland[0.041]; Poland[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; initiators[0.041]; included[0.041]; support[0.041]; said[0.041]; said[0.041]; conditions[0.041]; President[0.041]; saying[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Polish ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.814:-0.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPolish[0.049]; Polish[0.049]; Polish[0.049]; Polish[0.049]; Warsaw[0.046]; Warsaw[0.046]; state[0.041]; state[0.041]; million[0.040]; distribution[0.040]; Bank[0.040]; Warszawy[0.040]; Warszawy[0.040]; group[0.040]; group[0.040]; group[0.040]; total[0.039]; conditions[0.039]; market[0.039]; said[0.039]; said[0.039]; France[0.039]; businessman[0.039]; exchange[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Polish ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.141:0.141[0]; log p(e|m)= -0.814:-0.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPolish[0.051]; Polish[0.051]; Polish[0.051]; Warsaw[0.048]; Warsaw[0.048]; state[0.043]; state[0.043]; million[0.042]; distribution[0.042]; Bank[0.042]; Warszawy[0.042]; Warszawy[0.042]; group[0.042]; group[0.042]; total[0.041]; said[0.041]; said[0.041]; France[0.041]; businessman[0.041]; exchange[0.041]; plans[0.041]; press[0.041]; wanted[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Polish ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.814:-0.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.049]; Poland[0.049]; Poland[0.049]; Polish[0.048]; Polish[0.048]; country[0.044]; communism[0.043]; zlotys[0.041]; zlotys[0.041]; President[0.040]; news[0.040]; Wieslaw[0.040]; September[0.040]; minority[0.039]; French[0.039]; million[0.039]; million[0.039]; Monday[0.039]; distribution[0.039]; initial[0.039]; distributor[0.039]; says[0.039]; March[0.039]; Warszawy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jacques Chirac\u001b[39m ==> ENTITY: \u001b[32mJacques Chirac\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.132:0.132[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncase[0.045]; President[0.044]; French[0.044]; press[0.043]; press[0.043]; press[0.043]; support[0.043]; Warszawy[0.043]; saying[0.042]; country[0.042]; total[0.041]; said[0.041]; said[0.041]; Warsaw[0.041]; sole[0.041]; zlotys[0.040]; zlotys[0.040]; Monday[0.040]; Minister[0.040]; news[0.040]; visits[0.040]; group[0.040]; communism[0.040]; controls[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Anthony Barker\u001b[39m ==> ENTITY: \u001b[32mAnthony Raine Barker\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.045:0.045[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninitially[0.047]; consortium[0.044]; jointly[0.044]; said[0.043]; equity[0.042]; minority[0.042]; extend[0.042]; Swiss[0.042]; Swiss[0.042]; unhappy[0.042]; firms[0.041]; pulled[0.041]; increase[0.041]; shares[0.041]; consortia[0.041]; bids[0.041]; option[0.041]; option[0.041]; conditions[0.041]; Ruch[0.041]; Ruch[0.041]; bid[0.041]; offered[0.041]; Polish[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zygmunt Solorz\u001b[39m ==> ENTITY: \u001b[32mZygmunt Solorz-Żak\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.122:0.122[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbid[0.044]; Warszawy[0.043]; Warszawy[0.043]; owned[0.043]; Bank[0.042]; Polish[0.042]; Polish[0.042]; Polish[0.042]; Polish[0.042]; Polish[0.042]; Sa[0.042]; Sa[0.042]; control[0.042]; Warsaw[0.041]; Warsaw[0.041]; total[0.041]; million[0.041]; exchange[0.041]; bidding[0.041]; capital[0.040]; finance[0.040]; businessman[0.040]; firm[0.040]; rival[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Polish ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.125:0.125[0]; log p(e|m)= -0.814:-0.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPolish[0.050]; Polish[0.050]; Polish[0.050]; Warsaw[0.047]; state[0.042]; million[0.041]; distribution[0.041]; Bank[0.041]; Warszawy[0.041]; Warszawy[0.041]; group[0.041]; group[0.041]; said[0.040]; said[0.040]; France[0.040]; businessman[0.040]; plans[0.040]; press[0.040]; wanted[0.040]; daily[0.039]; publisher[0.039]; bid[0.039]; Saturday[0.039]; firm[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngroup[0.044]; group[0.044]; wanted[0.043]; Saturday[0.043]; bid[0.043]; state[0.042]; state[0.042]; Warsaw[0.042]; Warsaw[0.042]; press[0.042]; division[0.042]; said[0.041]; said[0.041]; million[0.041]; publisher[0.041]; offered[0.041]; Polish[0.041]; Polish[0.041]; Polish[0.041]; Polish[0.041]; plans[0.041]; Bank[0.040]; forming[0.040]; Solorz[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.125:0.125[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.050]; Polish[0.046]; Polish[0.046]; Warsaw[0.045]; country[0.045]; communism[0.043]; Monday[0.042]; support[0.041]; news[0.041]; President[0.041]; said[0.041]; said[0.041]; conditions[0.040]; Wieslaw[0.040]; September[0.040]; total[0.040]; press[0.040]; press[0.040]; press[0.040]; saying[0.040]; distribution[0.040]; stands[0.040]; visits[0.040]; included[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Warsaw ==> ENTITY: \u001b[32mWarsaw\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.143:-0.143[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.046]; Poland[0.046]; Polish[0.045]; Polish[0.045]; Polish[0.045]; water[0.042]; September[0.042]; percent[0.042]; percent[0.042]; percent[0.042]; percent[0.042]; percent[0.042]; state[0.041]; state[0.041]; total[0.041]; firm[0.040]; Warszawy[0.040]; Warszawy[0.040]; Saturday[0.040]; distribution[0.040]; distribution[0.040]; conditions[0.040]; million[0.040]; capital[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zycie Warszawy\u001b[39m ==> ENTITY: \u001b[32mŻycie Warszawy\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarszawy[0.049]; Warsaw[0.048]; Warsaw[0.048]; publisher[0.046]; Polish[0.046]; Polish[0.046]; Polish[0.046]; Polish[0.046]; quoted[0.045]; press[0.044]; daily[0.043]; paper[0.043]; paper[0.043]; newsstand[0.042]; Sa[0.042]; Sa[0.042]; owned[0.042]; said[0.040]; said[0.040]; Hachette[0.040]; division[0.040]; Bre[0.040]; Bre[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Polish ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.101:0.101[0]; log p(e|m)= -0.814:-0.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.051]; Polish[0.049]; country[0.045]; Anthony[0.043]; communism[0.042]; group[0.042]; March[0.041]; jointly[0.041]; Wieslaw[0.041]; minority[0.041]; initially[0.041]; million[0.041]; Monday[0.041]; plans[0.040]; initial[0.040]; distributor[0.040]; says[0.040]; Warszawy[0.040]; consortium[0.040]; consortium[0.040]; sole[0.040]; included[0.040]; government[0.040]; stands[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zycie Warszawy\u001b[39m ==> ENTITY: \u001b[32mŻycie Warszawy\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPolish[0.045]; Polish[0.045]; Polish[0.045]; Poland[0.043]; Poland[0.043]; Poland[0.043]; press[0.043]; press[0.043]; press[0.043]; daily[0.042]; September[0.042]; March[0.042]; fall[0.041]; News[0.041]; news[0.041]; zlotys[0.041]; zlotys[0.041]; says[0.040]; Initially[0.040]; Wieslaw[0.040]; communism[0.040]; invest[0.040]; Minister[0.040]; saying[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Polish ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.814:-0.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPolish[0.050]; Polish[0.050]; Polish[0.050]; Warsaw[0.047]; state[0.042]; state[0.042]; million[0.041]; distribution[0.041]; Bank[0.041]; Warszawy[0.040]; Warszawy[0.040]; group[0.040]; group[0.040]; said[0.040]; said[0.040]; France[0.040]; businessman[0.040]; plans[0.040]; press[0.040]; wanted[0.039]; percent[0.039]; percent[0.039]; percent[0.039]; daily[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mHachette\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mLouis Christophe François Hachette\u001b[39m <---> \u001b[32mHachette Filipacchi Médias\u001b[39m\t\nSCORES: global= 0.247:0.238[\u001b[31m0.009\u001b[39m]; local()= 0.129:0.123[\u001b[31m0.006\u001b[39m]; log p(e|m)= -1.609:-2.048[\u001b[31m0.439\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npublisher[0.048]; French[0.043]; distributor[0.043]; distributor[0.043]; group[0.042]; group[0.042]; publishers[0.042]; chain[0.042]; division[0.042]; firm[0.041]; Jacques[0.041]; sales[0.041]; distribution[0.041]; News[0.041]; consortium[0.041]; consortium[0.041]; consortium[0.041]; consortium[0.041]; million[0.041]; million[0.041]; support[0.041]; control[0.040]; open[0.040]; September[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swiss ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.091:0.091[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwiss[0.046]; Monday[0.045]; million[0.043]; sum[0.043]; offering[0.043]; listed[0.043]; group[0.043]; Poland[0.042]; March[0.042]; Polish[0.042]; Polish[0.042]; plans[0.042]; change[0.041]; bid[0.041]; said[0.041]; extend[0.041]; included[0.041]; holding[0.040]; bids[0.040]; invest[0.040]; initially[0.040]; increase[0.040]; investment[0.040]; stake[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mHachette\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mLouis Christophe François Hachette\u001b[39m <---> \u001b[32mHachette Filipacchi Médias\u001b[39m\t\nSCORES: global= 0.246:0.238[\u001b[31m0.008\u001b[39m]; local()= 0.101:0.113[\u001b[32m0.012\u001b[39m]; log p(e|m)= -1.609:-2.048[\u001b[31m0.439\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npublisher[0.049]; France[0.044]; group[0.043]; group[0.043]; chain[0.043]; chain[0.043]; division[0.043]; owned[0.043]; firm[0.042]; distribution[0.042]; consortium[0.042]; consortium[0.042]; consortium[0.042]; million[0.042]; newsstand[0.041]; finance[0.040]; offered[0.040]; shares[0.040]; rival[0.040]; provided[0.039]; press[0.039]; plans[0.039]; forming[0.039]; daily[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.050]; Polish[0.046]; Polish[0.046]; country[0.045]; communism[0.044]; Monday[0.042]; government[0.042]; March[0.041]; bid[0.041]; initially[0.041]; news[0.041]; said[0.041]; said[0.041]; zlotys[0.041]; zlotys[0.041]; conditions[0.040]; group[0.040]; Wieslaw[0.040]; included[0.040]; distribution[0.040]; initial[0.040]; press[0.039]; press[0.039]; Warszawy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Polish ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.814:-0.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.049]; Poland[0.049]; Polish[0.048]; Polish[0.048]; Warsaw[0.045]; country[0.045]; communism[0.043]; support[0.041]; group[0.041]; group[0.041]; President[0.040]; news[0.040]; September[0.040]; state[0.040]; French[0.040]; million[0.039]; million[0.039]; distribution[0.039]; wanted[0.039]; total[0.039]; Warszawy[0.039]; included[0.039]; stands[0.038]; conditions[0.038]; \t\n\r [==========>...................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s914ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1237/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1045testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nablus\u001b[39m ==> ENTITY: \u001b[32mNablus\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRamallah[0.046]; Nablus[0.045]; Nablus[0.045]; Nablus[0.045]; Israeli[0.044]; Palestinian[0.044]; Palestinian[0.044]; Jerusalem[0.043]; Palestinians[0.042]; West[0.042]; West[0.042]; forces[0.041]; Bank[0.041]; Bank[0.041]; city[0.041]; Yasser[0.041]; holy[0.041]; enclave[0.039]; ahead[0.039]; Arafat[0.039]; Arafat[0.039]; Arafat[0.039]; Arafat[0.039]; Authority[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arafat\u001b[39m ==> ENTITY: \u001b[32mYasser Arafat\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.047]; Israel[0.047]; Israeli[0.046]; Palestinian[0.044]; Palestinian[0.044]; Palestinians[0.042]; Palestinians[0.042]; Ramallah[0.042]; Arafat[0.042]; Arafat[0.042]; Arafat[0.042]; Jerusalem[0.041]; Jerusalem[0.041]; Arab[0.041]; Nablus[0.041]; Nablus[0.041]; Nablus[0.041]; Nablus[0.041]; Yasser[0.040]; security[0.039]; forces[0.039]; Arabs[0.038]; cabinet[0.038]; cabinet[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.215:0.215[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.051]; Israel[0.051]; Israel[0.051]; Jerusalem[0.043]; Jerusalem[0.043]; Jerusalem[0.043]; Palestinians[0.043]; Palestinians[0.043]; Arab[0.043]; Arabs[0.043]; West[0.042]; Jews[0.042]; Bank[0.041]; settlement[0.040]; expected[0.038]; Moslems[0.038]; occupation[0.038]; holy[0.038]; peace[0.038]; peace[0.038]; rule[0.038]; policy[0.038]; Arafat[0.038]; Arafat[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.222:0.222[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israel[0.050]; Israeli[0.045]; Israeli[0.045]; Arab[0.045]; Palestinian[0.043]; Jerusalem[0.042]; Jerusalem[0.042]; Jerusalem[0.042]; Jerusalem[0.042]; Jews[0.042]; Palestinians[0.041]; Palestinians[0.041]; Arabs[0.041]; security[0.040]; West[0.040]; policy[0.039]; holy[0.039]; negotiations[0.039]; Bank[0.039]; pray[0.038]; week[0.038]; peace[0.037]; peace[0.037]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mArabs\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPalestinian territories\u001b[39m <---> \u001b[32mArab people\u001b[39m\t\nSCORES: global= 0.250:0.249[\u001b[31m0.001\u001b[39m]; local()= 0.256:0.207[\u001b[31m0.049\u001b[39m]; log p(e|m)= -3.170:-1.877[\u001b[32m1.293\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israel[0.050]; Israel[0.050]; Arab[0.048]; Palestinian[0.048]; Palestinian[0.048]; Israeli[0.047]; Israeli[0.047]; Palestinians[0.045]; Palestinians[0.045]; Jerusalem[0.044]; Jerusalem[0.044]; Jerusalem[0.044]; Jerusalem[0.044]; Ramallah[0.044]; West[0.044]; West[0.044]; West[0.044]; Bank[0.043]; Bank[0.043]; Bank[0.043]; Yasser[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nablus\u001b[39m ==> ENTITY: \u001b[32mNablus\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRamallah[0.049]; Nablus[0.047]; Nablus[0.047]; Nablus[0.047]; Israeli[0.047]; Palestinian[0.046]; Palestinian[0.046]; Arabs[0.046]; Arab[0.045]; Jerusalem[0.045]; Jerusalem[0.045]; Jerusalem[0.045]; Israel[0.045]; Israel[0.045]; Palestinians[0.045]; Palestinians[0.045]; West[0.044]; West[0.044]; West[0.044]; Jews[0.044]; forces[0.044]; state[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nablus\u001b[39m ==> ENTITY: \u001b[32mNablus\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRamallah[0.045]; Nablus[0.044]; Nablus[0.044]; Nablus[0.044]; Israeli[0.043]; Palestinian[0.043]; Palestinian[0.043]; Arabs[0.043]; Arab[0.042]; Jerusalem[0.042]; Jerusalem[0.042]; Israel[0.041]; Palestinians[0.041]; Palestinians[0.041]; West[0.041]; West[0.041]; West[0.041]; Jews[0.041]; forces[0.040]; Bank[0.040]; Bank[0.040]; Bank[0.040]; city[0.040]; Yasser[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arafat\u001b[39m ==> ENTITY: \u001b[32mYasser Arafat\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.048]; Israel[0.048]; Israel[0.048]; Israeli[0.046]; Israeli[0.046]; Palestinians[0.043]; Palestinians[0.043]; Arafat[0.042]; Jerusalem[0.042]; Jerusalem[0.042]; Jerusalem[0.042]; Jerusalem[0.042]; Arab[0.042]; Nablus[0.041]; security[0.040]; forces[0.040]; occupation[0.039]; visit[0.039]; Arabs[0.039]; cabinet[0.038]; Bank[0.038]; West[0.038]; peace[0.037]; peace[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jews ==> ENTITY: \u001b[32mJews\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.161:0.161[0]; log p(e|m)= -0.126:-0.126[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.046]; Israel[0.046]; Israel[0.046]; Arabs[0.046]; Israeli[0.045]; Israeli[0.045]; Arab[0.044]; Palestinian[0.043]; Palestinian[0.043]; Jerusalem[0.042]; Jerusalem[0.042]; Jerusalem[0.042]; Jerusalem[0.042]; Palestinians[0.041]; Palestinians[0.041]; enclave[0.039]; West[0.039]; West[0.039]; West[0.039]; Nablus[0.039]; Nablus[0.039]; Nablus[0.039]; holy[0.037]; Moslems[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jerusalem ==> ENTITY: \u001b[32mJerusalem\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.216:0.216[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJerusalem[0.048]; Jerusalem[0.048]; Israel[0.047]; Israel[0.047]; Israel[0.047]; Jews[0.044]; Israeli[0.044]; Arabs[0.043]; Palestinian[0.043]; Palestinian[0.043]; Arab[0.043]; West[0.042]; West[0.042]; West[0.042]; Nablus[0.042]; Nablus[0.042]; Nablus[0.042]; Nablus[0.042]; holy[0.042]; Palestinians[0.042]; Palestinians[0.042]; enclave[0.042]; Ramallah[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arafat\u001b[39m ==> ENTITY: \u001b[32mYasser Arafat\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsraeli[0.049]; Palestinian[0.048]; Palestinian[0.048]; Palestinians[0.045]; Ramallah[0.045]; Arafat[0.045]; Arafat[0.045]; Arafat[0.045]; Jerusalem[0.044]; Nablus[0.044]; Nablus[0.044]; Nablus[0.044]; Nablus[0.044]; Yasser[0.043]; security[0.042]; forces[0.042]; cabinet[0.041]; cabinet[0.041]; Bank[0.041]; Bank[0.041]; President[0.040]; West[0.040]; West[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.242:0.242[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.054]; Israel[0.054]; Israeli[0.049]; Israeli[0.049]; Arab[0.048]; Palestinian[0.046]; Palestinian[0.046]; Jerusalem[0.045]; Jerusalem[0.045]; Jerusalem[0.045]; Jerusalem[0.045]; Jews[0.045]; Palestinians[0.044]; Palestinians[0.044]; Arabs[0.044]; security[0.043]; West[0.043]; West[0.043]; West[0.043]; policy[0.042]; holy[0.042]; negotiations[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.246:0.246[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.054]; Israel[0.054]; Israel[0.054]; Palestinian[0.046]; Palestinian[0.046]; Jerusalem[0.045]; Jerusalem[0.045]; Jerusalem[0.045]; Palestinians[0.045]; Palestinians[0.045]; Arab[0.045]; Arabs[0.045]; West[0.044]; West[0.044]; West[0.044]; Jews[0.044]; Bank[0.043]; Bank[0.043]; Bank[0.043]; settlement[0.042]; Ramallah[0.042]; enclave[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ramallah\u001b[39m ==> ENTITY: \u001b[32mRamallah\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.226:0.226[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPalestinian[0.048]; Palestinian[0.048]; Israeli[0.048]; Israel[0.048]; Arab[0.046]; Jerusalem[0.046]; Jerusalem[0.046]; Nablus[0.046]; Nablus[0.046]; Nablus[0.046]; Nablus[0.046]; West[0.045]; West[0.045]; West[0.045]; Bank[0.045]; Bank[0.045]; Bank[0.045]; Palestinians[0.045]; Palestinians[0.045]; Yasser[0.043]; Arabs[0.043]; enclave[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jerusalem ==> ENTITY: \u001b[32mJerusalem\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.212:0.212[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJerusalem[0.047]; Jerusalem[0.047]; Israel[0.046]; Israel[0.046]; Israel[0.046]; Jews[0.043]; Israeli[0.043]; Arabs[0.043]; Arab[0.042]; West[0.042]; holy[0.041]; Palestinians[0.041]; Palestinians[0.041]; occupation[0.040]; Bank[0.040]; Arafat[0.040]; Arafat[0.040]; pray[0.040]; set[0.039]; week[0.039]; settlement[0.039]; city[0.039]; city[0.039]; Moslems[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = East Jerusalem\u001b[39m ==> ENTITY: \u001b[32mEast Jerusalem\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.234:0.234[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.045]; Israel[0.045]; Israel[0.045]; Jerusalem[0.044]; Jerusalem[0.044]; Jerusalem[0.044]; Palestinian[0.044]; Palestinians[0.043]; Palestinians[0.043]; Arabs[0.043]; Israeli[0.043]; Israeli[0.043]; Arab[0.042]; West[0.041]; Bank[0.041]; occupation[0.040]; Nablus[0.040]; Jews[0.039]; settlement[0.039]; Arafat[0.039]; Arafat[0.039]; peace[0.038]; peace[0.038]; state[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nablus\u001b[39m ==> ENTITY: \u001b[32mNablus\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRamallah[0.052]; Nablus[0.050]; Nablus[0.050]; Nablus[0.050]; Israeli[0.050]; Palestinian[0.049]; Palestinian[0.049]; Arabs[0.049]; Jerusalem[0.048]; Palestinians[0.048]; West[0.047]; West[0.047]; Jews[0.047]; forces[0.046]; Bank[0.046]; Bank[0.046]; city[0.046]; Yasser[0.046]; holy[0.046]; enclave[0.044]; ahead[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arafat\u001b[39m ==> ENTITY: \u001b[32mYasser Arafat\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.216:0.216[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.046]; Israel[0.046]; Israel[0.046]; Israeli[0.044]; Israeli[0.044]; Palestinian[0.043]; Palestinian[0.043]; Palestinians[0.041]; Palestinians[0.041]; Ramallah[0.041]; Arafat[0.041]; Arafat[0.041]; Arafat[0.041]; Arafat[0.041]; Jerusalem[0.040]; Jerusalem[0.040]; Jerusalem[0.040]; Jerusalem[0.040]; Arab[0.040]; Nablus[0.040]; Nablus[0.040]; Nablus[0.040]; Nablus[0.040]; Yasser[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yasser Arafat\u001b[39m ==> ENTITY: \u001b[32mYasser Arafat\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israeli[0.049]; Palestinian[0.047]; Palestinian[0.047]; Palestinians[0.045]; Ramallah[0.045]; Arafat[0.045]; Arafat[0.045]; Arafat[0.045]; Jerusalem[0.044]; Nablus[0.043]; Nablus[0.043]; Nablus[0.043]; Nablus[0.043]; security[0.042]; forces[0.042]; Arabs[0.041]; cabinet[0.040]; cabinet[0.040]; Bank[0.040]; Bank[0.040]; Bank[0.040]; President[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.219:0.219[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israel[0.050]; Israeli[0.046]; Israeli[0.046]; Arab[0.045]; Jerusalem[0.043]; Jerusalem[0.043]; Jerusalem[0.043]; Jerusalem[0.043]; Jews[0.043]; Palestinians[0.042]; Palestinians[0.042]; Arabs[0.041]; security[0.040]; West[0.040]; policy[0.040]; holy[0.039]; negotiations[0.039]; Bank[0.039]; pray[0.038]; week[0.038]; peace[0.037]; peace[0.037]; visit[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Moslems\u001b[39m ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArabs[0.047]; Arab[0.046]; holy[0.043]; Jerusalem[0.042]; Jerusalem[0.042]; Jerusalem[0.042]; Jerusalem[0.042]; Jews[0.042]; enclave[0.042]; said[0.041]; said[0.041]; peace[0.041]; peace[0.041]; ahead[0.041]; week[0.041]; Israel[0.041]; Israel[0.041]; Israel[0.041]; city[0.041]; city[0.041]; Palestinian[0.041]; Palestinian[0.041]; expected[0.041]; state[0.040]; \t\n\r [===========>..................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 15s769ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1259/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1059testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andy Pettitte\u001b[39m ==> ENTITY: \u001b[32mAndy Pettitte\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.222:0.222[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYankees[0.044]; Yankees[0.044]; Randy[0.043]; game[0.042]; game[0.042]; game[0.042]; seven[0.042]; seven[0.042]; Roger[0.042]; Tino[0.042]; league[0.042]; games[0.042]; Jeff[0.041]; Yankee[0.041]; Cleveland[0.041]; Cleveland[0.041]; ninth[0.041]; League[0.041]; League[0.041]; gave[0.041]; fourth[0.041]; Martinez[0.041]; homers[0.040]; Brian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York Yankees ==> ENTITY: \u001b[32mNew York Yankees\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.201:0.201[0]; log p(e|m)= -0.143:-0.143[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; Yankees[0.044]; League[0.044]; League[0.044]; Incaviglia[0.044]; season[0.042]; sixth[0.042]; game[0.042]; game[0.042]; game[0.042]; games[0.042]; Cleveland[0.042]; Cleveland[0.042]; fourth[0.042]; York[0.041]; losing[0.041]; ninth[0.041]; Yankee[0.041]; beat[0.040]; Sox[0.040]; runs[0.040]; runs[0.040]; following[0.040]; losses[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mark McGwire\u001b[39m ==> ENTITY: \u001b[32mMark McGwire\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTartabull[0.044]; game[0.043]; game[0.043]; games[0.043]; hit[0.042]; hit[0.042]; hit[0.042]; Oakland[0.042]; league[0.042]; career[0.042]; run[0.042]; run[0.042]; run[0.042]; sixth[0.042]; seventh[0.041]; seventh[0.041]; ninth[0.041]; home[0.041]; Dave[0.041]; Detroit[0.041]; League[0.040]; seven[0.040]; McGwire[0.040]; Tony[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indians ==> ENTITY: \u001b[32mCleveland Indians\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.234:0.234[0]; log p(e|m)= -2.386:-2.386[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.044]; team[0.044]; Cleveland[0.044]; League[0.043]; seasons[0.043]; season[0.042]; season[0.042]; season[0.042]; season[0.042]; season[0.042]; games[0.042]; games[0.042]; games[0.042]; game[0.042]; game[0.042]; game[0.042]; game[0.042]; American[0.041]; history[0.041]; players[0.040]; Gonzalez[0.039]; Gonzalez[0.039]; Gonzalez[0.039]; Gonzalez[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Incaviglia\u001b[39m ==> ENTITY: \u001b[32mPete Incaviglia\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIncaviglia[0.051]; game[0.046]; League[0.046]; season[0.045]; homers[0.045]; Kevin[0.044]; Yankees[0.044]; Jim[0.043]; sixth[0.043]; sixth[0.043]; Randy[0.043]; Jay[0.043]; Sox[0.043]; Sox[0.043]; career[0.042]; career[0.042]; Darryl[0.042]; Yankee[0.042]; professional[0.042]; improved[0.042]; Toronto[0.042]; ninth[0.042]; Thomas[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Baltimore ==> ENTITY: \u001b[32mBaltimore Orioles\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.220:0.220[0]; log p(e|m)= -4.075:-4.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBaltimore[0.046]; Incaviglia[0.044]; Incaviglia[0.044]; League[0.043]; Yankees[0.043]; game[0.043]; season[0.042]; Randy[0.042]; Orioles[0.042]; Orioles[0.042]; Buhner[0.042]; Jays[0.042]; Toronto[0.041]; Sox[0.041]; Sox[0.041]; seven[0.041]; runs[0.041]; runs[0.041]; runs[0.041]; sixth[0.040]; sixth[0.040]; pitched[0.040]; Jim[0.040]; Yankee[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ivan Rodriguez\u001b[39m ==> ENTITY: \u001b[32mIván Rodríguez\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.247:0.247[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRodriguez[0.047]; Gonzalez[0.047]; Gonzalez[0.047]; Gonzalez[0.047]; Gonzalez[0.047]; Gonzalez[0.047]; Juan[0.045]; fifth[0.044]; season[0.044]; Baseball[0.043]; homered[0.043]; second[0.042]; hitting[0.042]; games[0.042]; games[0.042]; game[0.042]; game[0.042]; players[0.041]; Rangers[0.041]; Rangers[0.041]; Cleveland[0.041]; hit[0.041]; hit[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = American League\u001b[39m ==> ENTITY: \u001b[32mAmerican League\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.231:0.231[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.049]; League[0.049]; Cleveland[0.048]; Cleveland[0.048]; Cleveland[0.048]; Yankees[0.048]; team[0.048]; team[0.048]; game[0.048]; game[0.048]; game[0.048]; game[0.048]; games[0.048]; games[0.048]; run[0.047]; run[0.047]; fourth[0.047]; Sox[0.047]; homered[0.047]; runs[0.046]; beat[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Blue Jays ==> ENTITY: \u001b[32mToronto Blue Jays\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.199:0.199[0]; log p(e|m)= -0.473:-0.473[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTartabull[0.044]; Toronto[0.043]; league[0.043]; game[0.043]; game[0.043]; wild[0.043]; Red[0.043]; League[0.042]; sixth[0.042]; season[0.042]; runs[0.041]; runs[0.041]; runs[0.041]; runs[0.041]; seven[0.041]; major[0.041]; Boston[0.041]; ninth[0.041]; Oakland[0.041]; Sox[0.041]; Sox[0.041]; Sox[0.041]; victory[0.041]; homered[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Thomas ==> ENTITY: \u001b[32mFrank Thomas (designated hitter)\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.214:0.214[0]; log p(e|m)= -0.552:-0.552[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTartabull[0.046]; league[0.046]; American[0.045]; League[0.044]; Frank[0.044]; Jay[0.044]; Robin[0.044]; career[0.044]; career[0.044]; season[0.044]; Sox[0.044]; Sox[0.044]; Sox[0.044]; game[0.043]; game[0.043]; Toronto[0.043]; Mark[0.043]; seven[0.043]; Dave[0.042]; Thomas[0.042]; Kevin[0.042]; Oakland[0.042]; ninth[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Telgheder\u001b[39m ==> ENTITY: \u001b[32mDave Telgheder\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; games[0.047]; game[0.046]; Royals[0.044]; Tigers[0.044]; Jays[0.044]; Kansas[0.044]; Dave[0.044]; career[0.044]; cap[0.043]; Harold[0.043]; pitched[0.043]; pitched[0.043]; Todd[0.043]; innings[0.043]; Mark[0.043]; seven[0.043]; Oakland[0.042]; Sox[0.042]; Sox[0.042]; major[0.042]; Robin[0.041]; Blue[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = American League Central\u001b[39m ==> ENTITY: \u001b[32mAmerican League Central\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.232:0.232[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSox[0.045]; season[0.043]; season[0.043]; season[0.043]; season[0.043]; Yankees[0.043]; Yankees[0.043]; Cleveland[0.042]; Cleveland[0.042]; Cleveland[0.042]; fourth[0.042]; games[0.042]; games[0.042]; homers[0.042]; team[0.041]; League[0.041]; league[0.041]; game[0.040]; game[0.040]; game[0.040]; lead[0.040]; lead[0.040]; Angels[0.040]; American[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jose Parra\u001b[39m ==> ENTITY: \u001b[32mJosé Parra (baseball)\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; games[0.044]; Royals[0.044]; Royals[0.044]; Bob[0.043]; Tigers[0.043]; Matt[0.042]; Mike[0.042]; hitter[0.041]; hitter[0.041]; Brewers[0.041]; Brewers[0.041]; Kansas[0.041]; Kansas[0.041]; Lawton[0.041]; Dave[0.041]; Todd[0.041]; Twins[0.041]; homered[0.041]; homered[0.041]; career[0.040]; Minnesota[0.040]; Minnesota[0.040]; Tony[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Minnesota Twins ==> ENTITY: \u001b[32mMinnesota Twins\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.171:0.171[0]; log p(e|m)= -0.052:-0.052[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; Royals[0.044]; Royals[0.044]; second[0.044]; second[0.044]; Tigers[0.043]; won[0.042]; Dave[0.042]; Mike[0.042]; Matt[0.041]; win[0.041]; ninth[0.041]; ninth[0.041]; homered[0.041]; homered[0.041]; Jose[0.041]; closer[0.040]; victory[0.040]; Lawton[0.040]; hitter[0.040]; hitter[0.040]; Minnesota[0.040]; innings[0.040]; Bob[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mickey Rivers\u001b[39m ==> ENTITY: \u001b[32mMickey Rivers\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJohnny[0.043]; Gonzalez[0.042]; Gonzalez[0.042]; Gonzalez[0.042]; Gonzalez[0.042]; Gonzalez[0.042]; Gonzalez[0.042]; Cleveland[0.042]; Cleveland[0.042]; seasons[0.042]; games[0.042]; games[0.042]; games[0.042]; game[0.042]; game[0.042]; game[0.042]; team[0.041]; team[0.041]; record[0.041]; Rangers[0.041]; Rangers[0.041]; Ruben[0.041]; Rodriguez[0.040]; Rodriguez[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Angels ==> ENTITY: \u001b[32mLos Angeles Angels of Anaheim\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.219:0.219[0]; log p(e|m)= -1.754:-1.754[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIncaviglia[0.043]; game[0.043]; game[0.043]; game[0.043]; season[0.043]; league[0.043]; games[0.043]; sixth[0.042]; League[0.042]; League[0.042]; homers[0.042]; seven[0.042]; seven[0.042]; ninth[0.042]; Yankees[0.041]; Yankees[0.041]; homered[0.041]; Cleveland[0.041]; Cleveland[0.041]; fourth[0.041]; Sox[0.041]; Martinez[0.040]; run[0.040]; run[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York Yankees\u001b[39m\t\nSCORES: global= 0.239:0.239[0]; local()= 0.214:0.214[0]; log p(e|m)= -5.116:-5.116[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; Yankees[0.044]; Yankees[0.044]; league[0.044]; game[0.043]; game[0.043]; game[0.043]; Incaviglia[0.043]; Cleveland[0.043]; Cleveland[0.043]; League[0.042]; League[0.042]; season[0.042]; Baltimore[0.041]; Yankee[0.041]; fourth[0.040]; professional[0.040]; sixth[0.040]; York[0.040]; Sox[0.040]; American[0.039]; American[0.039]; beat[0.039]; ninth[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = McGwire\u001b[39m ==> ENTITY: \u001b[32mMark McGwire\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTartabull[0.044]; game[0.043]; games[0.043]; hit[0.042]; hit[0.042]; Oakland[0.042]; league[0.042]; career[0.042]; run[0.042]; run[0.042]; run[0.042]; run[0.042]; sixth[0.042]; seventh[0.041]; seventh[0.041]; Mark[0.041]; ninth[0.041]; home[0.041]; second[0.041]; Dave[0.041]; Detroit[0.041]; seven[0.041]; McGwire[0.040]; time[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mDetroit\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mDetroit Tigers\u001b[39m <---> \u001b[32mDetroit\u001b[39m\t\nSCORES: global= 0.251:0.240[\u001b[31m0.012\u001b[39m]; local()= 0.208:0.108[\u001b[31m0.100\u001b[39m]; log p(e|m)= -4.135:-0.217[\u001b[32m3.918\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.049]; games[0.049]; league[0.047]; game[0.045]; Tigers[0.045]; Royals[0.044]; Royals[0.044]; second[0.043]; Boston[0.043]; Boston[0.043]; sixth[0.043]; scoring[0.043]; ninth[0.043]; Oakland[0.043]; Sox[0.043]; seventh[0.042]; seventh[0.042]; seven[0.042]; walked[0.041]; City[0.041]; City[0.041]; hit[0.041]; hit[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Eddie Murray\u001b[39m ==> ENTITY: \u001b[32mEddie Murray\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIncaviglia[0.048]; Incaviglia[0.048]; game[0.043]; sixth[0.043]; sixth[0.043]; ninth[0.043]; season[0.042]; Yankees[0.042]; homer[0.042]; homer[0.042]; homer[0.042]; homer[0.042]; homer[0.042]; Randy[0.041]; League[0.041]; Jays[0.041]; Orioles[0.040]; Orioles[0.040]; hit[0.040]; hit[0.040]; hit[0.040]; Sox[0.040]; Sox[0.040]; Buhner[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jay Buhner\u001b[39m ==> ENTITY: \u001b[32mJay Buhner\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.208:0.208[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIncaviglia[0.044]; Incaviglia[0.044]; Tartabull[0.044]; career[0.043]; career[0.043]; sixth[0.043]; sixth[0.043]; season[0.043]; Yankees[0.042]; game[0.042]; ninth[0.042]; Randy[0.041]; Mariners[0.041]; Mariners[0.041]; homer[0.041]; homer[0.041]; homer[0.041]; homer[0.041]; homer[0.041]; homered[0.040]; seven[0.040]; Danny[0.040]; Jays[0.040]; Kevin[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jeff Russell\u001b[39m ==> ENTITY: \u001b[32mJeff Russell\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.229:0.229[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.047]; Cleveland[0.047]; Cleveland[0.047]; League[0.047]; League[0.047]; team[0.047]; team[0.047]; season[0.046]; season[0.046]; season[0.046]; season[0.046]; Texas[0.045]; Texas[0.045]; league[0.045]; games[0.045]; games[0.045]; Sox[0.045]; American[0.044]; American[0.044]; Brian[0.043]; Jim[0.043]; Angels[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Harold Baines\u001b[39m ==> ENTITY: \u001b[32mHarold Baines\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTartabull[0.046]; league[0.045]; League[0.044]; season[0.043]; game[0.043]; game[0.043]; career[0.043]; career[0.043]; Sox[0.042]; Sox[0.042]; Sox[0.042]; Jays[0.041]; ninth[0.041]; seven[0.041]; Buhner[0.041]; Orioles[0.041]; Baltimore[0.040]; home[0.040]; homered[0.040]; runs[0.040]; runs[0.040]; runs[0.040]; runs[0.040]; Mariners[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Robin Ventura\u001b[39m ==> ENTITY: \u001b[32mRobin Ventura\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.217:0.217[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTartabull[0.046]; League[0.044]; game[0.043]; game[0.043]; league[0.043]; ninth[0.042]; season[0.042]; Oakland[0.041]; hit[0.041]; hit[0.041]; hit[0.041]; hit[0.041]; homered[0.041]; career[0.041]; career[0.041]; homer[0.041]; homer[0.041]; homer[0.041]; homer[0.041]; seven[0.041]; Sox[0.041]; Sox[0.041]; Sox[0.041]; Martinez[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cleveland Indians ==> ENTITY: \u001b[32mCleveland Indians\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.212:0.212[0]; log p(e|m)= -0.088:-0.088[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndians[0.044]; seasons[0.044]; fifth[0.044]; games[0.044]; games[0.044]; game[0.043]; game[0.043]; Baseball[0.043]; season[0.042]; season[0.042]; second[0.041]; matchup[0.041]; Friday[0.041]; homered[0.041]; straight[0.040]; Gonzalez[0.040]; Gonzalez[0.040]; Gonzalez[0.040]; Gonzalez[0.040]; Gonzalez[0.040]; Gonzalez[0.040]; Rangers[0.040]; Rangers[0.040]; players[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Texas ==> ENTITY: \u001b[32mTexas Rangers (baseball)\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.219:0.219[0]; log p(e|m)= -4.962:-4.962[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; team[0.043]; Cleveland[0.043]; Cleveland[0.043]; Cleveland[0.043]; game[0.043]; game[0.043]; game[0.043]; games[0.042]; games[0.042]; League[0.042]; League[0.042]; fourth[0.041]; season[0.041]; season[0.041]; season[0.041]; season[0.041]; Yankees[0.041]; Yankees[0.041]; Brian[0.040]; seven[0.040]; seven[0.040]; seven[0.040]; play[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bob Wickman\u001b[39m ==> ENTITY: \u001b[32mBob Wickman\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nninth[0.044]; ninth[0.044]; games[0.043]; Mike[0.043]; Royals[0.042]; Royals[0.042]; Dave[0.042]; Matt[0.042]; second[0.042]; second[0.042]; Lawton[0.042]; Brewers[0.042]; Brewers[0.042]; Tigers[0.041]; closer[0.041]; Kansas[0.041]; Kansas[0.041]; pitched[0.041]; threw[0.041]; Milwaukee[0.040]; Milwaukee[0.040]; Milwaukee[0.040]; Twins[0.040]; homered[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Matt Lawton\u001b[39m ==> ENTITY: \u001b[32mMatt Lawton\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.049]; Royals[0.048]; homered[0.048]; ninth[0.047]; ninth[0.047]; Mike[0.046]; second[0.046]; second[0.046]; Twins[0.046]; pitches[0.045]; homer[0.045]; Dave[0.045]; scored[0.045]; Jose[0.045]; Kansas[0.045]; Brewers[0.044]; Brewers[0.044]; hit[0.044]; double[0.044]; run[0.044]; run[0.044]; run[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mariano Duncan\u001b[39m ==> ENTITY: \u001b[32mMariano Duncan\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIncaviglia[0.047]; Incaviglia[0.047]; league[0.047]; games[0.046]; game[0.045]; game[0.045]; Cleveland[0.045]; Cleveland[0.045]; League[0.044]; season[0.043]; homers[0.043]; homered[0.043]; Brian[0.043]; Giles[0.042]; Yankees[0.042]; Yankees[0.042]; sixth[0.042]; sixth[0.042]; seven[0.042]; Texas[0.042]; Jim[0.042]; Jim[0.042]; Baltimore[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Darryl Strawberry\u001b[39m ==> ENTITY: \u001b[32mDarryl Strawberry\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIncaviglia[0.045]; Incaviglia[0.045]; game[0.043]; game[0.043]; Cleveland[0.043]; Cleveland[0.043]; Yankees[0.042]; Yankees[0.042]; games[0.042]; league[0.042]; League[0.042]; sixth[0.042]; sixth[0.042]; hit[0.041]; Yankee[0.041]; ninth[0.041]; homers[0.041]; Giles[0.041]; season[0.040]; seven[0.040]; run[0.040]; run[0.040]; Baltimore[0.040]; Baltimore[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tony Clark\u001b[39m ==> ENTITY: \u001b[32mTony Clark\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.052]; game[0.046]; Mark[0.045]; homer[0.044]; homer[0.044]; games[0.044]; games[0.044]; home[0.044]; homered[0.044]; homered[0.044]; rallied[0.043]; won[0.043]; career[0.042]; Tigers[0.042]; Kansas[0.042]; Kansas[0.042]; run[0.042]; run[0.042]; run[0.042]; run[0.042]; run[0.042]; hit[0.042]; losing[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Texas ==> ENTITY: \u001b[32mTexas Rangers (baseball)\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.230:0.230[0]; log p(e|m)= -4.962:-4.962[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTexas[0.046]; game[0.045]; game[0.045]; game[0.045]; game[0.045]; team[0.044]; team[0.044]; fourth[0.044]; Cleveland[0.044]; Cleveland[0.044]; Cleveland[0.044]; games[0.043]; games[0.043]; games[0.043]; seasons[0.043]; second[0.043]; League[0.043]; League[0.043]; season[0.042]; season[0.042]; season[0.042]; season[0.042]; season[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Edgar Martinez\u001b[39m ==> ENTITY: \u001b[32mEdgar Martínez\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTartabull[0.048]; season[0.046]; Incaviglia[0.046]; Incaviglia[0.046]; League[0.046]; Buhner[0.045]; game[0.044]; Mariners[0.044]; Mariners[0.044]; sixth[0.044]; sixth[0.044]; hit[0.043]; hit[0.043]; hit[0.043]; homered[0.043]; Danny[0.042]; ninth[0.042]; got[0.042]; career[0.042]; career[0.042]; Jays[0.042]; cruised[0.041]; Frank[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Seattle ==> ENTITY: \u001b[32mSeattle Mariners\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.226:0.226[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIncaviglia[0.045]; Incaviglia[0.045]; Seattle[0.044]; Seattle[0.044]; season[0.044]; Tartabull[0.043]; game[0.043]; League[0.042]; Baltimore[0.041]; Baltimore[0.041]; career[0.041]; career[0.041]; Yankees[0.041]; Toronto[0.041]; Jays[0.041]; Mariners[0.041]; Mariners[0.041]; sixth[0.041]; sixth[0.041]; Orioles[0.040]; Orioles[0.040]; ninth[0.040]; professional[0.040]; past[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Brian Giles\u001b[39m ==> ENTITY: \u001b[32mBrian Giles\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplay[0.044]; league[0.043]; games[0.043]; games[0.043]; game[0.043]; game[0.043]; game[0.043]; game[0.043]; League[0.042]; League[0.042]; homers[0.042]; Cleveland[0.042]; Cleveland[0.042]; Cleveland[0.042]; fourth[0.042]; seven[0.041]; seven[0.041]; Sox[0.041]; Yankees[0.041]; homered[0.040]; Darryl[0.040]; team[0.040]; Russell[0.040]; runs[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baines\u001b[39m ==> ENTITY: \u001b[32mHarold Baines\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.208:0.208[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTartabull[0.046]; league[0.045]; League[0.044]; season[0.043]; game[0.043]; game[0.043]; career[0.043]; sixth[0.043]; Sox[0.042]; Sox[0.042]; Sox[0.042]; Jays[0.041]; ninth[0.041]; seven[0.041]; Buhner[0.041]; Orioles[0.041]; home[0.040]; homered[0.040]; runs[0.040]; runs[0.040]; runs[0.040]; runs[0.040]; Mariners[0.040]; Oakland[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kansas City Royals ==> ENTITY: \u001b[32mKansas City Royals\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.172:0.172[0]; log p(e|m)= -0.029:-0.029[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.043]; games[0.043]; game[0.043]; Royals[0.043]; Sox[0.043]; sixth[0.043]; seventh[0.042]; seventh[0.042]; ninth[0.042]; won[0.042]; Jose[0.041]; second[0.041]; second[0.041]; Detroit[0.041]; Tigers[0.041]; innings[0.041]; win[0.041]; Boston[0.041]; Boston[0.041]; homered[0.041]; homered[0.041]; loss[0.041]; blanked[0.040]; blanked[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Seattle ==> ENTITY: \u001b[32mSeattle Mariners\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.216:0.216[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; games[0.043]; games[0.043]; season[0.043]; season[0.043]; season[0.043]; season[0.043]; game[0.042]; game[0.042]; game[0.042]; Cleveland[0.042]; Cleveland[0.042]; Cleveland[0.042]; League[0.042]; League[0.042]; regular[0.041]; fourth[0.040]; Yankees[0.040]; Yankees[0.040]; seven[0.040]; seven[0.040]; seven[0.040]; homers[0.039]; play[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Minnesota ==> ENTITY: \u001b[32mMinnesota Twins\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.175:0.175[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; Minnesota[0.045]; Twins[0.044]; second[0.043]; win[0.043]; Mike[0.043]; Lawton[0.042]; Dave[0.042]; Matt[0.041]; victory[0.041]; won[0.041]; closer[0.041]; Brewers[0.041]; Brewers[0.041]; Kansas[0.041]; Milwaukee[0.041]; Milwaukee[0.041]; Milwaukee[0.041]; Bob[0.041]; ninth[0.041]; ninth[0.041]; tied[0.040]; City[0.040]; hitter[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dave Nilsson\u001b[39m ==> ENTITY: \u001b[32mDave Nilsson\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.047]; ninth[0.047]; ninth[0.047]; won[0.046]; win[0.045]; games[0.045]; hit[0.045]; homered[0.044]; hitter[0.044]; victory[0.043]; closer[0.043]; Matt[0.043]; innings[0.043]; Twins[0.042]; pitched[0.042]; double[0.042]; Lawton[0.042]; Mike[0.042]; Minnesota[0.042]; Minnesota[0.042]; leading[0.042]; rallied[0.042]; hitless[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Texas ==> ENTITY: \u001b[32mTexas Rangers (baseball)\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.223:0.223[0]; log p(e|m)= -4.962:-4.962[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTexas[0.050]; Texas[0.050]; Texas[0.050]; Texas[0.050]; game[0.049]; game[0.049]; game[0.049]; team[0.048]; team[0.048]; Cleveland[0.047]; Cleveland[0.047]; twice[0.047]; games[0.047]; games[0.047]; games[0.047]; seasons[0.047]; second[0.047]; season[0.046]; season[0.046]; season[0.046]; season[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jim Leyritz\u001b[39m ==> ENTITY: \u001b[32mJim Leyritz\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngame[0.049]; game[0.049]; sixth[0.047]; sixth[0.047]; Incaviglia[0.047]; Incaviglia[0.047]; homered[0.047]; ninth[0.046]; Yankees[0.046]; Yankees[0.046]; Cleveland[0.045]; Cleveland[0.045]; hit[0.045]; games[0.045]; Brian[0.044]; Andy[0.044]; allowed[0.044]; season[0.044]; league[0.044]; Yankee[0.044]; following[0.044]; seven[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Yankees ==> ENTITY: \u001b[32mNew York Yankees\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.219:0.219[0]; log p(e|m)= -0.429:-0.429[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIncaviglia[0.045]; Incaviglia[0.045]; Yankees[0.045]; League[0.043]; league[0.043]; sixth[0.042]; sixth[0.042]; game[0.042]; game[0.042]; games[0.042]; season[0.042]; season[0.042]; Yankee[0.041]; homers[0.041]; Orioles[0.041]; Orioles[0.041]; Baltimore[0.041]; Baltimore[0.041]; ninth[0.041]; Angels[0.041]; runs[0.040]; seven[0.040]; Mariners[0.040]; beat[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Juan Gonzalez ==> ENTITY: \u001b[32mJuan González (baseball)\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.246:0.246[0]; log p(e|m)= -1.076:-1.076[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRodriguez[0.047]; Rodriguez[0.047]; Gonzalez[0.046]; Gonzalez[0.046]; Gonzalez[0.046]; Gonzalez[0.046]; games[0.044]; games[0.044]; season[0.043]; Texas[0.043]; Texas[0.043]; Texas[0.043]; Texas[0.043]; Baseball[0.043]; game[0.043]; fifth[0.042]; Rangers[0.042]; Rangers[0.042]; players[0.042]; twice[0.042]; second[0.041]; Ivan[0.041]; homered[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kansas City ==> ENTITY: \u001b[32mKansas City Royals\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.169:0.169[0]; log p(e|m)= -3.576:-3.576[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; games[0.044]; Royals[0.043]; Royals[0.043]; sixth[0.043]; career[0.042]; City[0.042]; seventh[0.042]; seventh[0.042]; ninth[0.042]; ninth[0.042]; won[0.042]; second[0.042]; second[0.042]; Detroit[0.041]; win[0.041]; Boston[0.041]; blanked[0.041]; Tigers[0.041]; Tony[0.040]; week[0.040]; victory[0.040]; Lawton[0.040]; Mike[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Orioles ==> ENTITY: \u001b[32mBaltimore Orioles\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.233:0.233[0]; log p(e|m)= -0.182:-0.182[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBaltimore[0.048]; Baltimore[0.048]; Incaviglia[0.047]; Incaviglia[0.047]; Tartabull[0.047]; Yankees[0.046]; game[0.046]; League[0.046]; Randy[0.045]; Orioles[0.045]; season[0.045]; Buhner[0.045]; runs[0.045]; runs[0.045]; runs[0.045]; Jays[0.045]; Oakland[0.045]; Toronto[0.044]; Yankee[0.044]; Eddie[0.044]; Sox[0.044]; Sox[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = White Sox ==> ENTITY: \u001b[32mChicago White Sox\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.208:0.208[0]; log p(e|m)= -0.080:-0.080[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTartabull[0.045]; League[0.045]; Incaviglia[0.043]; Incaviglia[0.043]; game[0.043]; sixth[0.042]; sixth[0.042]; White[0.042]; season[0.042]; victory[0.042]; Sox[0.041]; Oakland[0.041]; ninth[0.041]; Baltimore[0.041]; Baltimore[0.041]; homered[0.041]; career[0.041]; career[0.041]; hit[0.040]; hit[0.040]; hit[0.040]; hit[0.040]; American[0.040]; tie[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indians ==> ENTITY: \u001b[32mCleveland Indians\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.206:0.206[0]; log p(e|m)= -2.386:-2.386[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.046]; games[0.044]; games[0.044]; Indians[0.044]; Baseball[0.043]; game[0.043]; season[0.043]; history[0.043]; players[0.042]; division[0.041]; Gonzalez[0.041]; Gonzalez[0.041]; Gonzalez[0.041]; Gonzalez[0.041]; Gonzalez[0.041]; Rangers[0.041]; Rangers[0.041]; Friday[0.040]; Texas[0.040]; Texas[0.040]; Texas[0.040]; Texas[0.040]; matchup[0.039]; straight[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jim Thome\u001b[39m ==> ENTITY: \u001b[32mJim Thome\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.243:0.243[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhomers[0.044]; game[0.044]; game[0.044]; game[0.044]; game[0.044]; games[0.043]; games[0.043]; Cleveland[0.042]; Cleveland[0.042]; Cleveland[0.042]; Yankees[0.042]; Sox[0.042]; homered[0.041]; fourth[0.041]; league[0.041]; play[0.041]; League[0.040]; League[0.040]; team[0.040]; runs[0.040]; runs[0.040]; homer[0.040]; Giles[0.040]; Jeff[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dave Telgheder\u001b[39m ==> ENTITY: \u001b[32mDave Telgheder\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.122:0.122[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.048]; games[0.047]; game[0.046]; game[0.046]; League[0.045]; Jays[0.044]; Harold[0.043]; Mariners[0.043]; pitched[0.043]; pitched[0.043]; Todd[0.043]; innings[0.043]; innings[0.043]; Mark[0.043]; seven[0.043]; Oakland[0.043]; Sox[0.043]; Sox[0.043]; Sox[0.043]; major[0.042]; Robin[0.042]; Tapani[0.041]; Blue[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Boston Red Sox ==> ENTITY: \u001b[32mBoston Red Sox\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.207:0.207[0]; log p(e|m)= -0.084:-0.084[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngame[0.045]; games[0.044]; Boston[0.043]; Tartabull[0.043]; league[0.043]; sixth[0.042]; Sox[0.042]; Tigers[0.042]; Royals[0.041]; seventh[0.041]; seventh[0.041]; Detroit[0.041]; seven[0.041]; ninth[0.041]; runs[0.041]; runs[0.041]; runs[0.041]; career[0.041]; hits[0.041]; hits[0.041]; hits[0.041]; hits[0.041]; Jays[0.041]; Oakland[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chad Ogea\u001b[39m ==> ENTITY: \u001b[32mChad Ogea\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngame[0.045]; game[0.045]; game[0.045]; homered[0.044]; run[0.044]; homer[0.044]; homer[0.044]; homer[0.044]; homer[0.044]; homer[0.044]; Indians[0.044]; Indians[0.044]; Indians[0.044]; Baseball[0.044]; fifth[0.043]; Cleveland[0.042]; games[0.042]; games[0.042]; second[0.042]; record[0.042]; hit[0.042]; hit[0.042]; players[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West ==> ENTITY: \u001b[32mAmerican League West\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.242:0.242[0]; log p(e|m)= -4.343:-4.343[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; season[0.043]; season[0.043]; season[0.043]; season[0.043]; League[0.043]; League[0.043]; fourth[0.043]; games[0.042]; games[0.042]; Texas[0.041]; Cleveland[0.041]; Cleveland[0.041]; Cleveland[0.041]; Yankees[0.041]; Yankees[0.041]; game[0.041]; game[0.041]; game[0.041]; American[0.041]; American[0.041]; Angels[0.041]; seven[0.040]; seven[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marc Newfield\u001b[39m ==> ENTITY: \u001b[32mMarc Newfield\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; games[0.046]; Todd[0.046]; Milwaukee[0.044]; Milwaukee[0.044]; Milwaukee[0.044]; Mike[0.044]; Dave[0.044]; Matt[0.044]; cap[0.044]; Detroit[0.044]; Tigers[0.043]; Kansas[0.043]; Kansas[0.043]; Brewers[0.043]; Brewers[0.043]; tied[0.043]; Twins[0.043]; Tony[0.042]; Royals[0.042]; Royals[0.042]; Minnesota[0.042]; Minnesota[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mariano Rivera\u001b[39m ==> ENTITY: \u001b[32mMariano Rivera\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.234:0.234[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIncaviglia[0.044]; Incaviglia[0.044]; ninth[0.043]; Yankees[0.043]; Yankees[0.043]; sixth[0.043]; sixth[0.043]; game[0.043]; game[0.043]; seven[0.042]; games[0.042]; homers[0.042]; career[0.041]; career[0.041]; Yankee[0.041]; league[0.041]; pitched[0.041]; innings[0.040]; runs[0.040]; Mariano[0.040]; season[0.040]; season[0.040]; Randy[0.040]; Pettitte[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Randy Velarde\u001b[39m ==> ENTITY: \u001b[32mRandy Velarde\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIncaviglia[0.049]; Incaviglia[0.049]; sixth[0.046]; sixth[0.046]; ninth[0.044]; career[0.044]; career[0.044]; homers[0.043]; game[0.043]; game[0.043]; game[0.043]; league[0.043]; professional[0.043]; Yankees[0.042]; Yankees[0.042]; seven[0.042]; season[0.042]; season[0.042]; League[0.042]; Rivera[0.042]; Orioles[0.041]; Orioles[0.041]; Jim[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brewers ==> ENTITY: \u001b[32mMilwaukee Brewers\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.196:0.196[0]; log p(e|m)= -0.248:-0.248[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwon[0.046]; Brewers[0.046]; games[0.045]; games[0.045]; scored[0.044]; ninth[0.044]; ninth[0.044]; win[0.044]; blanked[0.043]; Royals[0.043]; Royals[0.043]; career[0.043]; run[0.043]; run[0.043]; run[0.043]; run[0.043]; inning[0.043]; Milwaukee[0.043]; Milwaukee[0.043]; Milwaukee[0.043]; hit[0.042]; closer[0.042]; rallied[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Johnny Oates\u001b[39m ==> ENTITY: \u001b[32mJohnny Oates\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.213:0.213[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseasons[0.044]; team[0.043]; team[0.043]; Cleveland[0.043]; Cleveland[0.043]; Cleveland[0.043]; season[0.042]; season[0.042]; season[0.042]; season[0.042]; season[0.042]; Jim[0.041]; games[0.041]; games[0.041]; games[0.041]; games[0.041]; game[0.041]; game[0.041]; game[0.041]; game[0.041]; League[0.041]; League[0.041]; Russell[0.040]; manager[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = American League\u001b[39m ==> ENTITY: \u001b[32mAmerican League\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.221:0.221[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIncaviglia[0.044]; Incaviglia[0.044]; Tartabull[0.043]; sixth[0.043]; sixth[0.043]; game[0.042]; drove[0.042]; ninth[0.042]; homered[0.042]; Sox[0.042]; Sox[0.042]; runs[0.041]; runs[0.041]; runs[0.041]; season[0.041]; hit[0.041]; hit[0.041]; hit[0.041]; hit[0.041]; Martinez[0.041]; victory[0.041]; Jays[0.041]; seven[0.040]; scored[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mariners ==> ENTITY: \u001b[32mSeattle Mariners\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.231:0.231[0]; log p(e|m)= -0.126:-0.126[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; season[0.045]; Mariners[0.044]; game[0.043]; Incaviglia[0.043]; Incaviglia[0.043]; Tartabull[0.043]; Sox[0.043]; Sox[0.043]; Jays[0.041]; Seattle[0.041]; Seattle[0.041]; Seattle[0.041]; sixth[0.041]; sixth[0.041]; Oakland[0.040]; seven[0.040]; Orioles[0.040]; Orioles[0.040]; career[0.040]; career[0.040]; Toronto[0.040]; Baltimore[0.040]; Baltimore[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Texas Rangers ==> ENTITY: \u001b[32mTexas Rangers (baseball)\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.226:0.226[0]; log p(e|m)= -0.151:-0.151[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBaseball[0.045]; games[0.045]; games[0.045]; Rangers[0.045]; Cleveland[0.044]; season[0.044]; season[0.044]; game[0.044]; game[0.044]; seasons[0.044]; fifth[0.044]; Gonzalez[0.043]; Gonzalez[0.043]; Gonzalez[0.043]; Gonzalez[0.043]; Gonzalez[0.043]; Gonzalez[0.043]; second[0.043]; Texas[0.042]; Texas[0.042]; Texas[0.042]; Indians[0.042]; Indians[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cleveland ==> ENTITY: \u001b[32mCleveland Indians\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.224:0.224[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.047]; Cleveland[0.047]; games[0.045]; games[0.045]; league[0.045]; team[0.045]; season[0.044]; season[0.044]; season[0.044]; season[0.044]; game[0.044]; game[0.044]; game[0.044]; League[0.043]; League[0.043]; Yankees[0.043]; American[0.042]; American[0.042]; regular[0.041]; fourth[0.041]; York[0.041]; York[0.041]; play[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Buddy Groom\u001b[39m ==> ENTITY: \u001b[32mBuddy Groom\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.206:0.206[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; Royals[0.043]; Royals[0.043]; games[0.043]; seventh[0.043]; seventh[0.043]; sixth[0.042]; ninth[0.042]; Boston[0.042]; Boston[0.042]; Detroit[0.042]; career[0.042]; Tigers[0.041]; Sox[0.041]; Sox[0.041]; game[0.041]; second[0.041]; seven[0.041]; Jays[0.041]; major[0.041]; got[0.040]; Tartabull[0.040]; Chicago[0.040]; Chicago[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Boston ==> ENTITY: \u001b[32mBoston Red Sox\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.202:0.202[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBoston[0.049]; games[0.047]; games[0.047]; league[0.046]; game[0.045]; Tartabull[0.045]; Sox[0.044]; Dave[0.043]; Mark[0.043]; Royals[0.043]; Royals[0.043]; Detroit[0.043]; Todd[0.042]; Oakland[0.042]; second[0.042]; seven[0.042]; Red[0.042]; major[0.042]; Tony[0.042]; sixth[0.042]; Tigers[0.042]; seventh[0.042]; seventh[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Arlington ==> ENTITY: \u001b[32mArlington, Texas\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.159:0.159[0]; log p(e|m)= -1.295:-1.295[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTexas[0.046]; Texas[0.046]; Texas[0.046]; Texas[0.046]; Indians[0.042]; Indians[0.042]; game[0.042]; twice[0.042]; Rangers[0.042]; Rangers[0.042]; games[0.041]; games[0.041]; earlier[0.041]; Baseball[0.041]; Friday[0.041]; second[0.041]; Cleveland[0.040]; history[0.040]; matchup[0.040]; Chad[0.040]; season[0.040]; players[0.040]; run[0.039]; division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Milwaukee ==> ENTITY: \u001b[32mMilwaukee Brewers\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.197:0.197[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; Royals[0.045]; Royals[0.045]; ninth[0.044]; ninth[0.044]; second[0.044]; second[0.044]; career[0.044]; Milwaukee[0.043]; Milwaukee[0.043]; Tigers[0.043]; run[0.043]; run[0.043]; run[0.043]; run[0.043]; inning[0.043]; hit[0.043]; Brewers[0.043]; Brewers[0.043]; won[0.043]; closer[0.043]; win[0.043]; rallied[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = California ==> ENTITY: \u001b[32mCalifornia\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.084:0.084[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngame[0.045]; game[0.045]; game[0.045]; games[0.044]; games[0.044]; league[0.043]; season[0.043]; season[0.043]; season[0.043]; season[0.043]; play[0.041]; fourth[0.041]; run[0.041]; run[0.041]; regular[0.041]; League[0.040]; League[0.040]; American[0.040]; American[0.040]; keyed[0.040]; Texas[0.040]; Jeff[0.039]; gave[0.039]; post[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ruben Sierra\u001b[39m ==> ENTITY: \u001b[32mRubén Sierra\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.236:0.236[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRodriguez[0.045]; Gonzalez[0.045]; Gonzalez[0.045]; Gonzalez[0.045]; Gonzalez[0.045]; League[0.042]; seasons[0.041]; season[0.041]; season[0.041]; season[0.041]; season[0.041]; season[0.041]; Jeff[0.041]; game[0.041]; game[0.041]; game[0.041]; game[0.041]; second[0.040]; games[0.040]; games[0.040]; games[0.040]; Texas[0.040]; Texas[0.040]; Texas[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSeattle\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSeattle Mariners\u001b[39m <---> \u001b[32mSeattle\u001b[39m\t\nSCORES: global= 0.245:0.232[\u001b[31m0.012\u001b[39m]; local()= 0.229:0.122[\u001b[31m0.107\u001b[39m]; log p(e|m)= -3.863:-0.105[\u001b[32m3.758\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; Incaviglia[0.046]; Incaviglia[0.046]; Seattle[0.045]; Seattle[0.045]; season[0.045]; season[0.045]; game[0.044]; game[0.044]; game[0.044]; League[0.043]; Baltimore[0.043]; Baltimore[0.043]; career[0.042]; career[0.042]; Yankees[0.042]; Yankees[0.042]; seven[0.042]; Toronto[0.042]; Mariners[0.042]; Mariners[0.042]; sixth[0.042]; sixth[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pete Incaviglia\u001b[39m ==> ENTITY: \u001b[32mPete Incaviglia\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.209:0.209[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIncaviglia[0.048]; league[0.045]; game[0.044]; game[0.044]; game[0.044]; League[0.043]; season[0.043]; season[0.043]; homers[0.042]; Kevin[0.041]; Yankees[0.041]; Yankees[0.041]; Jim[0.041]; sixth[0.040]; sixth[0.040]; Randy[0.040]; Jay[0.040]; Sox[0.040]; career[0.040]; career[0.040]; Darryl[0.040]; Yankee[0.040]; professional[0.040]; improved[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brewers ==> ENTITY: \u001b[32mMilwaukee Brewers\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.177:0.177[0]; log p(e|m)= -0.248:-0.248[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwon[0.045]; Brewers[0.045]; games[0.044]; scored[0.043]; ninth[0.042]; ninth[0.042]; win[0.042]; innings[0.042]; Royals[0.042]; run[0.041]; run[0.041]; run[0.041]; Milwaukee[0.041]; Milwaukee[0.041]; Milwaukee[0.041]; victory[0.041]; hit[0.041]; closer[0.041]; rallied[0.041]; second[0.041]; second[0.041]; defeating[0.040]; homered[0.040]; Jose[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cleveland ==> ENTITY: \u001b[32mCleveland Indians\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.240:0.240[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.045]; Cleveland[0.045]; games[0.043]; games[0.043]; games[0.043]; team[0.043]; team[0.043]; seasons[0.043]; season[0.042]; season[0.042]; season[0.042]; season[0.042]; season[0.042]; game[0.041]; game[0.041]; game[0.041]; game[0.041]; League[0.041]; League[0.041]; players[0.040]; regular[0.040]; Gonzalez[0.039]; Gonzalez[0.039]; fourth[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mToronto\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mToronto Blue Jays\u001b[39m <---> \u001b[32mToronto\u001b[39m\t\nSCORES: global= 0.248:0.232[\u001b[31m0.015\u001b[39m]; local()= 0.218:0.045[\u001b[31m0.173\u001b[39m]; log p(e|m)= -4.423:-0.123[\u001b[32m4.300\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; League[0.045]; Incaviglia[0.044]; Incaviglia[0.044]; Tartabull[0.043]; game[0.043]; season[0.043]; Jays[0.042]; major[0.041]; Baltimore[0.041]; Baltimore[0.041]; Orioles[0.041]; Orioles[0.041]; sixth[0.041]; sixth[0.041]; seven[0.041]; Murray[0.040]; runs[0.040]; runs[0.040]; runs[0.040]; tie[0.040]; Oakland[0.040]; Sox[0.040]; Sox[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kevin Tapani\u001b[39m ==> ENTITY: \u001b[32mKevin Tapani\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.230:0.230[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; League[0.044]; Tartabull[0.044]; career[0.044]; career[0.044]; season[0.044]; Incaviglia[0.043]; Incaviglia[0.043]; seven[0.042]; game[0.041]; Sox[0.041]; Sox[0.041]; Mariners[0.041]; Mariners[0.041]; Martinez[0.041]; sixth[0.041]; sixth[0.041]; Orioles[0.040]; Orioles[0.040]; Jays[0.040]; Oakland[0.040]; Eddie[0.040]; Buhner[0.040]; Dave[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cleveland ==> ENTITY: \u001b[32mCleveland Indians\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.224:0.224[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.047]; Cleveland[0.047]; games[0.045]; games[0.045]; league[0.045]; team[0.045]; season[0.044]; season[0.044]; season[0.044]; season[0.044]; game[0.044]; game[0.044]; game[0.044]; League[0.043]; League[0.043]; Yankees[0.043]; American[0.042]; American[0.042]; regular[0.041]; fourth[0.041]; York[0.041]; York[0.041]; play[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Todd Van Poppel\u001b[39m ==> ENTITY: \u001b[32mTodd Van Poppel\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; games[0.044]; league[0.043]; career[0.043]; sixth[0.043]; seventh[0.043]; seventh[0.043]; ninth[0.042]; seven[0.042]; Mark[0.041]; game[0.041]; Royals[0.041]; Royals[0.041]; major[0.041]; Tigers[0.041]; Kansas[0.041]; Kansas[0.041]; Dave[0.041]; Milwaukee[0.041]; pitched[0.041]; pitched[0.041]; Sox[0.041]; starters[0.040]; Brewers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tigers ==> ENTITY: \u001b[32mDetroit Tigers\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.216:0.216[0]; log p(e|m)= -1.146:-1.146[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; game[0.046]; games[0.045]; games[0.045]; Detroit[0.044]; Royals[0.044]; Royals[0.044]; Sox[0.042]; week[0.041]; Boston[0.041]; Boston[0.041]; sixth[0.041]; ninth[0.040]; Twins[0.040]; second[0.040]; seventh[0.040]; seventh[0.040]; career[0.040]; won[0.040]; homered[0.040]; homered[0.040]; Tony[0.040]; Brewers[0.040]; cap[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Texas ==> ENTITY: \u001b[32mTexas Rangers (baseball)\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.222:0.222[0]; log p(e|m)= -4.962:-4.962[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTexas[0.044]; Texas[0.044]; Texas[0.044]; Baseball[0.043]; team[0.043]; seasons[0.042]; Cleveland[0.042]; game[0.042]; game[0.042]; game[0.042]; games[0.042]; games[0.042]; season[0.041]; season[0.041]; history[0.041]; Gonzalez[0.040]; Gonzalez[0.040]; Gonzalez[0.040]; Gonzalez[0.040]; Gonzalez[0.040]; Gonzalez[0.040]; Rangers[0.040]; Rangers[0.040]; Chad[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tino Martinez\u001b[39m ==> ENTITY: \u001b[32mTino Martinez\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.206:0.206[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYankees[0.044]; Yankees[0.044]; game[0.043]; game[0.043]; game[0.043]; homered[0.042]; fourth[0.042]; play[0.042]; Cleveland[0.041]; Cleveland[0.041]; League[0.041]; League[0.041]; games[0.041]; games[0.041]; homer[0.041]; homers[0.041]; run[0.041]; run[0.041]; league[0.041]; Sox[0.041]; ninth[0.041]; Randy[0.041]; Giles[0.040]; Yankee[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rangers ==> ENTITY: \u001b[32mTexas Rangers (baseball)\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.210:0.210[0]; log p(e|m)= -2.688:-2.688[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.048]; games[0.048]; Rangers[0.046]; Baseball[0.046]; game[0.045]; Cleveland[0.045]; season[0.045]; players[0.044]; Gonzalez[0.043]; Gonzalez[0.043]; Gonzalez[0.043]; Gonzalez[0.043]; Texas[0.043]; Texas[0.043]; Texas[0.043]; Texas[0.043]; Rodriguez[0.042]; Rodriguez[0.042]; twice[0.042]; Friday[0.042]; matchup[0.041]; joined[0.041]; Juan[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mTexas\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mTexas Rangers (baseball)\u001b[39m <---> \u001b[32mTexas\u001b[39m\t\nSCORES: global= 0.245:0.236[\u001b[31m0.009\u001b[39m]; local()= 0.208:0.099[\u001b[31m0.109\u001b[39m]; log p(e|m)= -4.962:-0.161[\u001b[32m4.801\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTexas[0.045]; Texas[0.045]; Texas[0.045]; Baseball[0.044]; Cleveland[0.043]; game[0.043]; games[0.042]; games[0.042]; season[0.041]; history[0.041]; Gonzalez[0.041]; Gonzalez[0.041]; Gonzalez[0.041]; Gonzalez[0.041]; Gonzalez[0.041]; Rangers[0.041]; Rangers[0.041]; Chad[0.041]; second[0.040]; hit[0.040]; hit[0.040]; Friday[0.040]; players[0.040]; matchup[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Danny Tartabull\u001b[39m ==> ENTITY: \u001b[32mDanny Tartabull\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; game[0.043]; game[0.043]; Toronto[0.043]; Sox[0.042]; Sox[0.042]; Sox[0.042]; Oakland[0.042]; League[0.042]; career[0.042]; Jays[0.042]; Buhner[0.042]; home[0.041]; Orioles[0.041]; seven[0.041]; Boston[0.041]; Boston[0.041]; major[0.041]; sixth[0.041]; hit[0.041]; hit[0.041]; hit[0.041]; hit[0.041]; Martinez[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Baltimore Orioles ==> ENTITY: \u001b[32mBaltimore Orioles\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.213:0.213[0]; log p(e|m)= -0.218:-0.218[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYankees[0.044]; Incaviglia[0.044]; Incaviglia[0.044]; Baltimore[0.044]; League[0.043]; homers[0.043]; game[0.043]; game[0.043]; season[0.042]; season[0.042]; Randy[0.042]; Orioles[0.042]; Sox[0.042]; Buhner[0.042]; Yankee[0.041]; Toronto[0.041]; seven[0.041]; runs[0.040]; runs[0.040]; runs[0.040]; sixth[0.040]; sixth[0.040]; American[0.040]; pitched[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Seattle Mariners\u001b[39m ==> ENTITY: \u001b[32mSeattle Mariners\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.205:0.205[0]; log p(e|m)= -0.008:-0.008[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.047]; homers[0.045]; Incaviglia[0.045]; Incaviglia[0.045]; seven[0.045]; season[0.044]; game[0.044]; game[0.044]; sixth[0.044]; sixth[0.044]; Yankees[0.044]; ninth[0.044]; strikeouts[0.043]; Mariners[0.042]; innings[0.042]; innings[0.042]; Orioles[0.042]; Orioles[0.042]; Baltimore[0.042]; Baltimore[0.042]; runs[0.042]; runs[0.042]; runs[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Milwaukee ==> ENTITY: \u001b[32mMilwaukee Brewers\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.183:0.183[0]; log p(e|m)= -3.411:-3.411[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; ninth[0.043]; ninth[0.043]; innings[0.042]; second[0.042]; Milwaukee[0.042]; Milwaukee[0.042]; Brewers[0.042]; Brewers[0.042]; win[0.042]; run[0.042]; run[0.042]; run[0.042]; victory[0.041]; hit[0.041]; won[0.041]; closer[0.041]; rallied[0.041]; City[0.041]; Minnesota[0.040]; Minnesota[0.040]; Mike[0.040]; homer[0.040]; homered[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Roger Pavlik\u001b[39m ==> ENTITY: \u001b[32mRoger Pavlik\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; team[0.046]; team[0.046]; League[0.045]; League[0.045]; season[0.044]; season[0.044]; season[0.044]; season[0.044]; Texas[0.043]; Texas[0.043]; game[0.043]; game[0.043]; game[0.043]; game[0.043]; Pettitte[0.043]; seasons[0.043]; record[0.042]; winner[0.042]; winner[0.042]; second[0.042]; Thome[0.042]; Sox[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Frank Thomas ==> ENTITY: \u001b[32mFrank Thomas (designated hitter)\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.221:0.221[0]; log p(e|m)= -0.552:-0.552[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIncaviglia[0.046]; Tartabull[0.046]; league[0.045]; American[0.045]; League[0.044]; Jay[0.044]; Robin[0.044]; career[0.044]; career[0.044]; season[0.043]; Sox[0.043]; Sox[0.043]; Sox[0.043]; game[0.043]; Toronto[0.043]; sixth[0.043]; sixth[0.043]; Mark[0.043]; seven[0.043]; Dave[0.042]; Thomas[0.042]; Kevin[0.042]; Oakland[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago White Sox ==> ENTITY: \u001b[32mChicago White Sox\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.217:0.217[0]; log p(e|m)= -0.079:-0.079[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTartabull[0.047]; League[0.047]; league[0.046]; Incaviglia[0.045]; game[0.045]; game[0.045]; sixth[0.044]; season[0.044]; victory[0.044]; Sox[0.043]; Sox[0.043]; Oakland[0.043]; White[0.043]; Baltimore[0.043]; homered[0.043]; career[0.042]; career[0.042]; runs[0.042]; runs[0.042]; runs[0.042]; runs[0.042]; American[0.042]; Boston[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mOakland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mOakland Athletics\u001b[39m <---> \u001b[32mOakland, California\u001b[39m\t\nSCORES: global= 0.262:0.248[\u001b[31m0.014\u001b[39m]; local()= 0.224:0.115[\u001b[31m0.109\u001b[39m]; log p(e|m)= -3.912:-0.499[\u001b[32m3.413\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.048]; league[0.048]; League[0.048]; Tartabull[0.047]; game[0.047]; game[0.047]; Boston[0.046]; Boston[0.046]; Detroit[0.045]; major[0.045]; Sox[0.045]; Sox[0.045]; Sox[0.045]; sixth[0.045]; Toronto[0.045]; Orioles[0.045]; seventh[0.044]; seventh[0.044]; ninth[0.044]; Jays[0.044]; seven[0.044]; American[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = White Sox ==> ENTITY: \u001b[32mChicago White Sox\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.224:0.224[0]; log p(e|m)= -0.080:-0.080[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; League[0.044]; league[0.043]; games[0.043]; games[0.043]; Cleveland[0.042]; Cleveland[0.042]; Cleveland[0.042]; game[0.042]; game[0.042]; game[0.042]; homers[0.041]; Yankees[0.041]; season[0.041]; season[0.041]; season[0.041]; season[0.041]; fourth[0.041]; team[0.041]; gave[0.041]; homered[0.040]; beat[0.040]; American[0.040]; American[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Royals ==> ENTITY: \u001b[32mKansas City Royals\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.165:0.165[0]; log p(e|m)= -0.311:-0.311[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.048]; games[0.048]; game[0.045]; Royals[0.045]; sixth[0.045]; seventh[0.044]; seventh[0.044]; ninth[0.044]; won[0.044]; innings[0.043]; Jose[0.043]; second[0.043]; second[0.043]; Detroit[0.043]; win[0.043]; Boston[0.043]; homered[0.042]; homered[0.042]; loss[0.042]; Tigers[0.042]; blanked[0.042]; Mike[0.042]; losing[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMilwaukee\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMilwaukee Brewers\u001b[39m <---> \u001b[32mMilwaukee\u001b[39m\t\nSCORES: global= 0.258:0.234[\u001b[31m0.024\u001b[39m]; local()= 0.207:0.100[\u001b[31m0.107\u001b[39m]; log p(e|m)= -3.411:-0.282[\u001b[32m3.129\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; games[0.045]; ninth[0.042]; ninth[0.042]; seventh[0.042]; seventh[0.042]; innings[0.042]; second[0.042]; second[0.042]; Royals[0.041]; Royals[0.041]; career[0.041]; Milwaukee[0.041]; Milwaukee[0.041]; Brewers[0.041]; Brewers[0.041]; Boston[0.041]; win[0.041]; run[0.041]; run[0.041]; run[0.041]; run[0.041]; run[0.041]; inning[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mike Fetters\u001b[39m ==> ENTITY: \u001b[32mMike Fetters\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.126:0.126[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npitched[0.045]; games[0.045]; Dave[0.044]; Kansas[0.044]; ninth[0.044]; ninth[0.044]; Twins[0.044]; hit[0.044]; Bob[0.043]; pitches[0.043]; Matt[0.043]; homered[0.043]; time[0.043]; Brewers[0.043]; Brewers[0.043]; Minnesota[0.043]; Minnesota[0.043]; Royals[0.043]; Milwaukee[0.043]; Milwaukee[0.043]; Milwaukee[0.043]; second[0.043]; second[0.043]; \t\n\r [===========>..................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s996ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1353/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1097testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cubs ==> ENTITY: \u001b[32mChicago Cubs\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.183:0.183[0]; log p(e|m)= -0.288:-0.288[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngame[0.044]; game[0.044]; games[0.043]; league[0.043]; eighth[0.043]; ninth[0.043]; seventh[0.043]; Phillies[0.042]; Phillies[0.042]; singled[0.042]; tying[0.042]; second[0.042]; second[0.042]; infield[0.041]; won[0.041]; Dodgers[0.041]; Chicago[0.041]; inning[0.041]; homer[0.041]; win[0.041]; win[0.041]; Cubs[0.041]; doubleheader[0.040]; hit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ryne Sandberg\u001b[39m ==> ENTITY: \u001b[32mRyne Sandberg\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.045]; games[0.045]; game[0.044]; game[0.044]; eighth[0.042]; seventh[0.042]; ninth[0.042]; second[0.042]; second[0.042]; Billy[0.042]; Braves[0.041]; Braves[0.041]; Worrell[0.041]; Dodgers[0.041]; overall[0.041]; home[0.041]; home[0.041]; win[0.041]; win[0.041]; Jones[0.040]; Jones[0.040]; Cubs[0.040]; Cubs[0.040]; Kevin[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bobby Valentine\u001b[39m ==> ENTITY: \u001b[32mBobby Valentine\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; Mets[0.043]; Mets[0.043]; season[0.043]; season[0.043]; manager[0.043]; Ray[0.043]; Giants[0.042]; Pittsburgh[0.042]; Cardinals[0.042]; Cardinals[0.042]; second[0.042]; homers[0.042]; won[0.041]; career[0.041]; field[0.041]; eighth[0.041]; pitch[0.040]; seventh[0.040]; win[0.040]; Eric[0.040]; ninth[0.040]; Colorado[0.040]; Colorado[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = John Boles ==> ENTITY: \u001b[32mJohn Boles (baseball)\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.178:0.178[0]; log p(e|m)= -0.812:-0.812[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; league[0.044]; Cardinals[0.044]; Ray[0.044]; League[0.043]; Marlins[0.043]; Cincinnati[0.043]; Reds[0.043]; Reds[0.043]; Reds[0.043]; manager[0.042]; manager[0.042]; season[0.041]; major[0.041]; Todd[0.041]; game[0.041]; Florida[0.041]; Florida[0.041]; won[0.040]; fifth[0.040]; fifth[0.040]; ninth[0.039]; seventh[0.039]; seventh[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scott Sanders ==> ENTITY: \u001b[32mScott Sanders (baseball)\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.695:-0.695[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; game[0.045]; Sanders[0.044]; David[0.042]; right[0.042]; majors[0.042]; record[0.042]; hitter[0.042]; Braves[0.042]; homered[0.041]; Padres[0.041]; Jones[0.041]; Jones[0.041]; won[0.041]; won[0.041]; inning[0.041]; sixth[0.041]; errors[0.041]; Cubs[0.041]; hit[0.041]; hit[0.041]; hit[0.041]; Foster[0.041]; Chicago[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wally Joyner\u001b[39m ==> ENTITY: \u001b[32mWally Joyner\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngame[0.048]; games[0.048]; games[0.048]; Atlanta[0.047]; Andruw[0.046]; Expos[0.046]; Braves[0.046]; Padres[0.046]; majors[0.045]; David[0.045]; Jones[0.045]; Jones[0.045]; Mike[0.045]; Mike[0.045]; doubled[0.045]; second[0.045]; second[0.045]; second[0.045]; second[0.045]; Jose[0.045]; Scott[0.045]; sixth[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Devon White\u001b[39m ==> ENTITY: \u001b[32mDevon White (baseball)\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; league[0.046]; fifth[0.045]; fifth[0.045]; seventh[0.045]; League[0.044]; game[0.044]; Marlins[0.044]; Marlins[0.044]; Baseball[0.043]; Cincinnati[0.043]; Cincinnati[0.043]; run[0.043]; Reds[0.043]; Reds[0.043]; Reds[0.043]; inning[0.043]; beat[0.043]; Cardinals[0.042]; double[0.042]; threw[0.042]; threw[0.042]; season[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Delino DeShields\u001b[39m ==> ENTITY: \u001b[32mDelino DeShields\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; league[0.044]; games[0.044]; Philadelphia[0.043]; Phillies[0.043]; Phillies[0.043]; Worrell[0.042]; seventh[0.042]; seventh[0.042]; Dodgers[0.042]; Reds[0.042]; ninth[0.042]; ball[0.041]; overall[0.041]; manager[0.041]; Billy[0.041]; eighth[0.041]; young[0.041]; pitches[0.041]; Ray[0.040]; starting[0.040]; Chad[0.040]; home[0.040]; run[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mets ==> ENTITY: \u001b[32mNew York Mets\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.191:0.191[0]; log p(e|m)= -0.264:-0.264[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMets[0.047]; Mets[0.047]; games[0.043]; second[0.042]; second[0.042]; eighth[0.042]; seventh[0.042]; seventh[0.042]; seventh[0.042]; ninth[0.042]; ninth[0.042]; ninth[0.042]; game[0.042]; season[0.041]; Bobby[0.041]; Giants[0.041]; Giants[0.041]; Shawn[0.040]; Francisco[0.040]; straight[0.040]; straight[0.040]; won[0.040]; hit[0.040]; hit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pete Harnisch\u001b[39m ==> ENTITY: \u001b[32mPete Harnisch\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.043]; game[0.043]; games[0.043]; career[0.042]; Mets[0.042]; Mets[0.042]; Mets[0.042]; Giants[0.042]; Giants[0.042]; Bobby[0.042]; seventh[0.042]; seventh[0.042]; seventh[0.042]; Mike[0.042]; eighth[0.041]; manager[0.041]; Shawn[0.041]; time[0.041]; time[0.041]; second[0.041]; second[0.041]; ninth[0.041]; ninth[0.041]; ninth[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tom Pagnozzi\u001b[39m ==> ENTITY: \u001b[32mTom Pagnozzi\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.046]; season[0.046]; games[0.046]; Cardinals[0.045]; Cardinals[0.045]; Bobby[0.044]; homers[0.044]; field[0.044]; Giants[0.044]; Eric[0.044]; hit[0.043]; hit[0.043]; manager[0.043]; Rod[0.043]; Colorado[0.043]; Colorado[0.043]; Mets[0.042]; Mets[0.042]; Rockies[0.042]; Rockies[0.042]; doubles[0.042]; September[0.042]; Brent[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Todd Worrell\u001b[39m ==> ENTITY: \u001b[32mTodd Worrell\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; league[0.046]; Chad[0.045]; Curtis[0.045]; Jeff[0.045]; games[0.045]; Ryan[0.044]; allowing[0.044]; game[0.043]; Ryne[0.043]; pitcher[0.043]; Billy[0.043]; eighth[0.043]; starting[0.043]; doubleheader[0.043]; Kevin[0.043]; run[0.042]; run[0.042]; innings[0.042]; seventh[0.042]; Braves[0.042]; Braves[0.042]; picked[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Houston Astros ==> ENTITY: \u001b[32mHouston Astros\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.180:0.180[0]; log p(e|m)= -0.049:-0.049[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.046]; season[0.046]; league[0.044]; Cardinals[0.044]; Cardinals[0.044]; game[0.043]; game[0.043]; games[0.042]; second[0.042]; second[0.042]; straight[0.041]; time[0.041]; time[0.041]; time[0.041]; major[0.041]; win[0.041]; win[0.041]; homers[0.040]; Louis[0.040]; seven[0.040]; won[0.040]; Pirates[0.040]; career[0.040]; tied[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Berry\u001b[39m ==> ENTITY: \u001b[32mSean Berry\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.044]; Sean[0.044]; Ray[0.043]; league[0.043]; Astros[0.043]; Eric[0.043]; Berry[0.043]; Rockies[0.042]; second[0.042]; second[0.042]; Cardinals[0.042]; Pirates[0.042]; Colorado[0.042]; season[0.041]; season[0.041]; games[0.041]; doubles[0.041]; homers[0.041]; hit[0.041]; Anthony[0.040]; Pittsburgh[0.040]; career[0.040]; Louis[0.040]; went[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = National League ==> ENTITY: \u001b[32mNational League\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.207:0.207[0]; log p(e|m)= -0.075:-0.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; league[0.046]; won[0.044]; season[0.043]; Baseball[0.043]; game[0.043]; win[0.043]; fifth[0.042]; fifth[0.042]; Cincinnati[0.042]; Cincinnati[0.042]; seventh[0.041]; Marlins[0.041]; Marlins[0.041]; tie[0.041]; Reds[0.041]; Reds[0.041]; Reds[0.041]; earned[0.041]; Cardinals[0.040]; Friday[0.040]; early[0.040]; early[0.040]; beat[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kevin Foster ==> ENTITY: \u001b[32mKevin Foster (baseball)\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.644:-0.644[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; games[0.044]; second[0.044]; second[0.044]; game[0.043]; game[0.043]; seventh[0.042]; Cubs[0.042]; Cubs[0.042]; win[0.041]; win[0.041]; Todd[0.041]; eighth[0.041]; doubleheader[0.041]; Mike[0.041]; straight[0.041]; innings[0.041]; ninth[0.041]; overall[0.041]; Foster[0.041]; leading[0.040]; victory[0.040]; victory[0.040]; committed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlanta ==> ENTITY: \u001b[32mAtlanta Braves\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.198:0.198[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; games[0.045]; league[0.045]; game[0.044]; game[0.044]; Atlanta[0.044]; Braves[0.042]; Braves[0.042]; Phillies[0.042]; Worrell[0.041]; Los[0.041]; Los[0.041]; won[0.041]; won[0.041]; Jose[0.041]; seventh[0.041]; second[0.041]; second[0.041]; Jones[0.040]; Jones[0.040]; Chicago[0.040]; Chicago[0.040]; ninth[0.040]; overall[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cardinals ==> ENTITY: \u001b[32mSt. Louis Cardinals\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.203:0.203[0]; log p(e|m)= -1.127:-1.127[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; games[0.044]; Cardinals[0.044]; Giants[0.044]; season[0.044]; Pittsburgh[0.043]; eighth[0.042]; Houston[0.042]; seven[0.042]; ninth[0.041]; Louis[0.041]; Louis[0.041]; homers[0.041]; career[0.041]; career[0.041]; second[0.041]; Mets[0.041]; major[0.041]; Pagnozzi[0.041]; Ray[0.041]; win[0.040]; September[0.040]; went[0.040]; Bobby[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Benes\u001b[39m ==> ENTITY: \u001b[32mAlan Benes\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; season[0.046]; Cardinals[0.046]; Cardinals[0.046]; seven[0.045]; league[0.044]; homers[0.044]; innings[0.043]; ninth[0.043]; Louis[0.043]; Louis[0.043]; career[0.043]; career[0.043]; eighth[0.043]; Eric[0.043]; hits[0.043]; hits[0.043]; Alan[0.043]; Benes[0.042]; runs[0.042]; win[0.042]; beat[0.042]; run[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Diego Padres ==> ENTITY: \u001b[32mSan Diego Padres\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.174:0.174[0]; log p(e|m)= -0.050:-0.050[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; games[0.044]; fourth[0.043]; tying[0.043]; win[0.042]; win[0.042]; Expos[0.042]; majors[0.042]; second[0.042]; second[0.042]; tie[0.041]; ninth[0.041]; ninth[0.041]; ninth[0.041]; victory[0.041]; sixth[0.041]; Hernandez[0.041]; seventh[0.041]; seventh[0.041]; ahead[0.041]; home[0.041]; straight[0.041]; straight[0.041]; record[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Expos ==> ENTITY: \u001b[32mMontreal Expos\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.193:0.193[0]; log p(e|m)= -0.035:-0.035[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; games[0.046]; Montreal[0.045]; Montreal[0.045]; Montreal[0.045]; majors[0.044]; sixth[0.044]; seventh[0.044]; seventh[0.044]; ninth[0.043]; ninth[0.043]; ninth[0.043]; win[0.043]; win[0.043]; Atlanta[0.043]; Scott[0.043]; second[0.043]; second[0.043]; David[0.043]; fourth[0.042]; record[0.042]; time[0.042]; time[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jeff Parrett\u001b[39m ==> ENTITY: \u001b[32mJeff Parrett\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.213:0.213[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; Philadelphia[0.044]; won[0.043]; Braves[0.043]; league[0.043]; league[0.043]; innings[0.042]; innings[0.042]; Reds[0.042]; win[0.042]; Phillies[0.042]; Phillies[0.042]; Dodgers[0.041]; Chad[0.041]; Ryan[0.041]; Koosman[0.041]; Tom[0.041]; victory[0.041]; seventh[0.041]; seventh[0.041]; doubleheader[0.041]; Chicago[0.040]; dropped[0.040]; Cubs[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Giants ==> ENTITY: \u001b[32mSan Francisco Giants\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.185:0.185[0]; log p(e|m)= -0.911:-0.911[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGiants[0.049]; seventh[0.047]; seventh[0.047]; eighth[0.047]; game[0.047]; ninth[0.047]; ninth[0.047]; ninth[0.047]; second[0.047]; games[0.046]; Cardinals[0.046]; season[0.046]; field[0.045]; straight[0.044]; pitch[0.044]; Francisco[0.044]; beat[0.044]; beat[0.044]; dropped[0.043]; hit[0.043]; hit[0.043]; Bobby[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Estes\u001b[39m ==> ENTITY: \u001b[32mShawn Estes\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngame[0.044]; Giants[0.044]; Giants[0.044]; games[0.043]; season[0.043]; career[0.043]; Shawn[0.042]; Rockies[0.042]; second[0.042]; seventh[0.042]; seventh[0.042]; ninth[0.041]; ninth[0.041]; ninth[0.041]; eighth[0.041]; pitch[0.041]; Mets[0.041]; Mets[0.041]; Mets[0.041]; field[0.041]; Cardinals[0.040]; Colorado[0.040]; Estes[0.040]; dropped[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cincinnati Reds ==> ENTITY: \u001b[32mCincinnati Reds\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; league[0.046]; season[0.044]; Cincinnati[0.043]; game[0.043]; Baseball[0.043]; League[0.043]; fifth[0.042]; fifth[0.042]; Reds[0.042]; Reds[0.042]; Cardinals[0.042]; earned[0.042]; seventh[0.041]; run[0.041]; major[0.041]; Friday[0.040]; win[0.040]; innings[0.040]; Marlins[0.040]; Marlins[0.040]; National[0.040]; inning[0.039]; pitches[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Phillies ==> ENTITY: \u001b[32mPhiladelphia Phillies\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.211:0.211[0]; log p(e|m)= -0.118:-0.118[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; seventh[0.043]; Philadelphia[0.043]; eighth[0.043]; game[0.042]; game[0.042]; games[0.042]; second[0.042]; second[0.042]; inning[0.042]; win[0.042]; win[0.042]; Phillies[0.042]; won[0.041]; straight[0.041]; Worrell[0.041]; innings[0.041]; run[0.041]; run[0.041]; run[0.041]; Dodgers[0.041]; victory[0.041]; Braves[0.041]; Braves[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tom Glavine\u001b[39m ==> ENTITY: \u001b[32mTom Glavine\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.233:0.233[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngame[0.045]; game[0.045]; league[0.044]; Braves[0.043]; Braves[0.043]; games[0.043]; games[0.043]; Atlanta[0.042]; Atlanta[0.042]; win[0.041]; win[0.041]; second[0.041]; second[0.041]; ninth[0.041]; eighth[0.041]; seventh[0.041]; Worrell[0.041]; inning[0.041]; Mike[0.040]; tying[0.040]; straight[0.040]; homered[0.040]; hitter[0.040]; innings[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Braves ==> ENTITY: \u001b[32mAtlanta Braves\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.209:0.209[0]; log p(e|m)= -0.285:-0.285[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; game[0.043]; game[0.043]; league[0.043]; second[0.043]; second[0.043]; eighth[0.042]; Braves[0.042]; majors[0.042]; Atlanta[0.042]; Atlanta[0.042]; ninth[0.042]; Phillies[0.041]; record[0.041]; Worrell[0.041]; straight[0.041]; doubleheader[0.041]; tying[0.041]; innings[0.041]; innings[0.041]; win[0.040]; homered[0.040]; Chicago[0.040]; Chicago[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ray Lankford\u001b[39m ==> ENTITY: \u001b[32mRay Lankford\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; second[0.044]; second[0.044]; career[0.043]; career[0.043]; eighth[0.043]; games[0.042]; homers[0.042]; game[0.042]; game[0.042]; season[0.042]; season[0.042]; Houston[0.042]; Cardinals[0.042]; Cardinals[0.042]; Louis[0.041]; Louis[0.041]; Pagnozzi[0.040]; Anthony[0.040]; Pirates[0.040]; Colorado[0.040]; Colorado[0.040]; seven[0.040]; won[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Reds ==> ENTITY: \u001b[32mCincinnati Reds\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.171:0.171[0]; log p(e|m)= -1.024:-1.024[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.048]; games[0.046]; season[0.046]; Philadelphia[0.045]; eighth[0.044]; seventh[0.044]; seventh[0.044]; pitch[0.044]; pitch[0.044]; Reds[0.044]; ninth[0.043]; Phillies[0.043]; Dodgers[0.043]; starting[0.043]; win[0.042]; Curtis[0.042]; won[0.042]; Delino[0.042]; John[0.042]; Ray[0.042]; victory[0.042]; pitcher[0.042]; yesterday[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = David Segui\u001b[39m ==> ENTITY: \u001b[32mDavid Segui\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nExpos[0.047]; Shawn[0.046]; homered[0.044]; Jose[0.044]; Montreal[0.044]; Montreal[0.044]; Montreal[0.044]; game[0.044]; Mets[0.044]; Mets[0.044]; pitch[0.044]; Giants[0.044]; season[0.043]; Scott[0.043]; Mike[0.043]; Mike[0.043]; majors[0.043]; games[0.043]; games[0.043]; homer[0.042]; Francisco[0.042]; Padres[0.042]; retired[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mike Oquist\u001b[39m ==> ENTITY: \u001b[32mMike Oquist\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; games[0.046]; Giants[0.045]; Giants[0.045]; game[0.045]; Expos[0.045]; York[0.045]; Padres[0.045]; majors[0.043]; David[0.043]; Mets[0.043]; Mets[0.043]; season[0.043]; San[0.042]; San[0.042]; won[0.042]; Montreal[0.042]; Montreal[0.042]; Montreal[0.042]; pitch[0.042]; run[0.042]; run[0.042]; run[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chad Curtis\u001b[39m ==> ENTITY: \u001b[32mChad Curtis\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngame[0.045]; Jeff[0.044]; games[0.044]; Todd[0.043]; Billy[0.043]; Jerry[0.043]; Ryan[0.043]; doubleheader[0.042]; Tom[0.042]; Dodgers[0.041]; league[0.041]; league[0.041]; pitches[0.041]; run[0.041]; time[0.041]; time[0.041]; hit[0.041]; scored[0.041]; Ryne[0.041]; home[0.041]; Philadelphia[0.040]; innings[0.040]; innings[0.040]; win[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rod Beck\u001b[39m ==> ENTITY: \u001b[32mRod Beck\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.043]; game[0.043]; eighth[0.043]; second[0.042]; ninth[0.042]; ninth[0.042]; ninth[0.042]; games[0.042]; career[0.042]; dropped[0.042]; Shawn[0.042]; Giants[0.042]; Giants[0.042]; seventh[0.042]; seventh[0.042]; closer[0.042]; hitting[0.041]; straight[0.041]; tie[0.041]; field[0.041]; hit[0.041]; hit[0.041]; Francisco[0.041]; Mike[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alvaro Espinoza\u001b[39m ==> ENTITY: \u001b[32mÁlvaro Espinoza\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npitch[0.046]; Francisco[0.046]; hitting[0.045]; game[0.045]; games[0.044]; career[0.044]; win[0.044]; right[0.044]; manager[0.044]; ninth[0.043]; ninth[0.043]; field[0.043]; homer[0.043]; Pagnozzi[0.043]; Colorado[0.043]; Colorado[0.043]; hit[0.043]; seventh[0.043]; Cardinals[0.043]; Rockies[0.043]; eighth[0.042]; hits[0.042]; hits[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Eric Anthony\u001b[39m ==> ENTITY: \u001b[32mEric Anthony\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; games[0.044]; game[0.043]; game[0.043]; career[0.043]; Houston[0.043]; season[0.043]; season[0.043]; homers[0.043]; Pirates[0.042]; seven[0.041]; Astros[0.041]; major[0.041]; second[0.041]; second[0.041]; hit[0.041]; doubles[0.041]; Cardinals[0.040]; Cardinals[0.040]; Colorado[0.040]; Colorado[0.040]; Rockies[0.040]; Rockies[0.040]; straight[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Billy Ashley\u001b[39m ==> ENTITY: \u001b[32mBilly Ashley\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; game[0.044]; game[0.044]; games[0.043]; seventh[0.042]; triple[0.042]; second[0.042]; second[0.042]; home[0.042]; home[0.042]; ninth[0.042]; scored[0.042]; tying[0.042]; eighth[0.042]; doubleheader[0.041]; starting[0.041]; hit[0.041]; Dodgers[0.041]; single[0.040]; single[0.040]; Chicago[0.040]; time[0.040]; Ryne[0.040]; scoreless[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Reds ==> ENTITY: \u001b[32mCincinnati Reds\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.194:0.194[0]; log p(e|m)= -1.024:-1.024[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.045]; league[0.045]; game[0.044]; Cincinnati[0.043]; season[0.043]; League[0.043]; fifth[0.042]; fifth[0.042]; eighth[0.042]; seventh[0.041]; seventh[0.041]; pitch[0.041]; pitch[0.041]; Cardinals[0.041]; Reds[0.041]; Reds[0.041]; run[0.041]; major[0.041]; ninth[0.041]; earned[0.041]; Friday[0.040]; win[0.040]; Marlins[0.040]; won[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chipper Jones\u001b[39m ==> ENTITY: \u001b[32mChipper Jones\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.218:0.218[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngame[0.052]; game[0.052]; tying[0.051]; league[0.051]; majors[0.050]; Braves[0.050]; Braves[0.050]; ninth[0.050]; Andruw[0.050]; hit[0.050]; hit[0.050]; eighth[0.050]; games[0.050]; second[0.049]; second[0.049]; singled[0.049]; singled[0.049]; shortstop[0.049]; win[0.049]; inning[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andruw Jones\u001b[39m ==> ENTITY: \u001b[32mAndruw Jones\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.236:0.236[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmajors[0.044]; Jose[0.043]; games[0.042]; game[0.042]; game[0.042]; hit[0.042]; hit[0.042]; sixth[0.042]; homered[0.042]; Braves[0.042]; Braves[0.042]; doubled[0.042]; ninth[0.042]; eighth[0.042]; Hernandez[0.042]; singled[0.041]; singled[0.041]; homer[0.041]; run[0.041]; run[0.041]; run[0.041]; run[0.041]; run[0.041]; run[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kevin Brown ==> ENTITY: \u001b[32mKevin Brown (right-handed pitcher)\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.190:0.190[0]; log p(e|m)= -0.263:-0.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBaseball[0.045]; league[0.045]; league[0.045]; League[0.044]; game[0.043]; fifth[0.042]; fifth[0.042]; seventh[0.042]; major[0.042]; season[0.042]; Kevin[0.042]; Brown[0.041]; Brown[0.041]; Brown[0.041]; lead[0.041]; Friday[0.041]; Marlins[0.041]; Marlins[0.041]; Todd[0.040]; John[0.040]; won[0.040]; win[0.040]; run[0.040]; starts[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPittsburgh\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPittsburgh Pirates\u001b[39m <---> \u001b[32mPittsburgh\u001b[39m\t\nSCORES: global= 0.252:0.233[\u001b[31m0.019\u001b[39m]; local()= 0.191:0.084[\u001b[31m0.107\u001b[39m]; log p(e|m)= -4.017:-0.298[\u001b[32m3.719\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; league[0.045]; game[0.044]; game[0.044]; Cardinals[0.044]; Cardinals[0.044]; season[0.043]; season[0.043]; Pirates[0.042]; Houston[0.041]; major[0.041]; Louis[0.041]; Louis[0.041]; won[0.041]; Astros[0.041]; career[0.040]; win[0.040]; win[0.040]; Colorado[0.040]; Colorado[0.040]; second[0.040]; second[0.040]; tied[0.040]; homers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Montreal ==> ENTITY: \u001b[32mMontreal Expos\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.214:0.214[0]; log p(e|m)= -4.510:-4.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; Montreal[0.045]; Montreal[0.045]; Expos[0.045]; season[0.043]; game[0.043]; majors[0.042]; Shawn[0.042]; record[0.041]; Mets[0.041]; Mets[0.041]; won[0.041]; fourth[0.041]; Giants[0.041]; Giants[0.041]; second[0.041]; second[0.041]; sixth[0.040]; Padres[0.040]; seventh[0.040]; seventh[0.040]; seventh[0.040]; pitch[0.040]; David[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Montreal ==> ENTITY: \u001b[32mMontreal Expos\u001b[39m\t\nSCORES: global= 0.241:0.241[0]; local()= 0.213:0.213[0]; log p(e|m)= -4.510:-4.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.048]; Montreal[0.047]; Montreal[0.047]; Expos[0.047]; season[0.045]; game[0.045]; Shawn[0.044]; Mets[0.043]; Mets[0.043]; won[0.043]; fourth[0.043]; Giants[0.043]; Giants[0.043]; second[0.042]; second[0.042]; sixth[0.042]; Padres[0.042]; seventh[0.042]; seventh[0.042]; seventh[0.042]; pitch[0.042]; David[0.042]; York[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Marlins ==> ENTITY: \u001b[32mMiami Marlins\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.203:0.203[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseventh[0.044]; league[0.044]; league[0.044]; fifth[0.043]; fifth[0.043]; game[0.043]; Cincinnati[0.043]; Cincinnati[0.043]; Baseball[0.043]; League[0.042]; Marlins[0.042]; win[0.041]; straight[0.041]; major[0.041]; Reds[0.041]; Reds[0.041]; beat[0.041]; run[0.040]; Cardinals[0.040]; Florida[0.040]; inning[0.040]; earned[0.040]; batter[0.039]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Los Angeles Dodgers ==> ENTITY: \u001b[32mLos Angeles Dodgers\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.170:0.170[0]; log p(e|m)= -0.105:-0.105[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngame[0.044]; league[0.043]; league[0.043]; Philadelphia[0.043]; games[0.043]; run[0.042]; eighth[0.042]; seventh[0.042]; starting[0.042]; win[0.042]; Jeff[0.041]; homer[0.041]; manager[0.041]; won[0.041]; time[0.041]; time[0.041]; Worrell[0.041]; victory[0.041]; tie[0.041]; doubleheader[0.041]; leading[0.041]; Phillies[0.041]; Phillies[0.041]; home[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mets ==> ENTITY: \u001b[32mNew York Mets\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.187:0.187[0]; log p(e|m)= -0.264:-0.264[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMets[0.048]; Mets[0.048]; second[0.043]; eighth[0.043]; seventh[0.043]; ninth[0.043]; game[0.043]; season[0.042]; season[0.042]; win[0.042]; Bobby[0.041]; Giants[0.041]; Giants[0.041]; straight[0.041]; won[0.041]; hit[0.040]; runs[0.040]; beat[0.040]; beat[0.040]; field[0.040]; Cardinals[0.040]; Cardinals[0.040]; Ray[0.040]; dropped[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSt Louis\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSt. Louis Cardinals\u001b[39m <---> \u001b[32mSt. Louis\u001b[39m\t\nSCORES: global= 0.256:0.253[\u001b[31m0.002\u001b[39m]; local()= 0.205:0.083[\u001b[31m0.122\u001b[39m]; log p(e|m)= -4.828:-0.092[\u001b[32m4.736\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; seventh[0.044]; Cardinals[0.044]; Cardinals[0.044]; eighth[0.043]; ninth[0.043]; Giants[0.043]; Louis[0.043]; season[0.043]; season[0.043]; Pittsburgh[0.042]; Colorado[0.042]; Colorado[0.042]; second[0.041]; won[0.041]; Mets[0.041]; Mets[0.041]; homers[0.040]; career[0.040]; Pagnozzi[0.039]; single[0.039]; scattered[0.039]; win[0.039]; manager[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Todd Stottlemyre\u001b[39m ==> ENTITY: \u001b[32mTodd Stottlemyre\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npitch[0.043]; pitch[0.043]; Cardinals[0.042]; run[0.042]; Marlins[0.042]; Marlins[0.042]; Baseball[0.042]; pitches[0.042]; game[0.042]; league[0.042]; league[0.042]; earned[0.042]; Kevin[0.042]; Kevin[0.042]; win[0.042]; gave[0.041]; Cincinnati[0.041]; Cincinnati[0.041]; John[0.041]; Reds[0.041]; Reds[0.041]; Reds[0.041]; fifth[0.041]; fifth[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Los Angeles ==> ENTITY: \u001b[32mLos Angeles Dodgers\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.191:0.191[0]; log p(e|m)= -4.711:-4.711[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; league[0.045]; Philadelphia[0.044]; game[0.044]; game[0.044]; Dodgers[0.042]; home[0.041]; home[0.041]; second[0.041]; second[0.041]; won[0.041]; Worrell[0.041]; hit[0.041]; Phillies[0.041]; Phillies[0.041]; eighth[0.041]; Braves[0.041]; Braves[0.041]; seventh[0.041]; Atlanta[0.041]; Curtis[0.041]; infield[0.041]; ninth[0.040]; Chicago[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Donne Wall\u001b[39m ==> ENTITY: \u001b[32mDonne Wall\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.045]; games[0.044]; game[0.044]; game[0.044]; Ray[0.043]; career[0.042]; homers[0.042]; seven[0.042]; won[0.042]; Sean[0.042]; season[0.042]; season[0.042]; win[0.041]; win[0.041]; straight[0.041]; hit[0.041]; Houston[0.041]; Cardinals[0.041]; Cardinals[0.041]; Eric[0.040]; doubles[0.040]; Astros[0.040]; Wall[0.040]; tied[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cubs ==> ENTITY: \u001b[32mChicago Cubs\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.170:0.170[0]; log p(e|m)= -0.288:-0.288[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngame[0.044]; game[0.044]; games[0.043]; eighth[0.043]; ninth[0.043]; leadoff[0.043]; sixth[0.042]; fourth[0.042]; homered[0.042]; singled[0.042]; singled[0.042]; tying[0.042]; second[0.041]; second[0.041]; second[0.041]; majors[0.041]; record[0.041]; infield[0.041]; won[0.041]; Expos[0.041]; shortstop[0.041]; Jose[0.041]; doubled[0.041]; Chicago[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jose Hernandez ==> ENTITY: \u001b[32mJosé Hernández (baseball)\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.161:0.161[0]; log p(e|m)= -0.849:-0.849[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSanders[0.044]; Sanders[0.044]; majors[0.044]; games[0.043]; Joyner[0.043]; Scott[0.043]; David[0.042]; Diego[0.042]; game[0.042]; game[0.042]; home[0.041]; Mike[0.041]; shortstop[0.041]; record[0.041]; Chicago[0.041]; hit[0.041]; hit[0.041]; Kevin[0.041]; Jones[0.041]; Jones[0.041]; second[0.041]; second[0.041]; second[0.041]; doubleheader[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pirates ==> ENTITY: \u001b[32mPittsburgh Pirates\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.198:0.198[0]; log p(e|m)= -1.474:-1.474[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; game[0.045]; game[0.045]; league[0.044]; season[0.044]; season[0.044]; Pittsburgh[0.044]; straight[0.042]; Cardinals[0.042]; Astros[0.042]; Houston[0.041]; major[0.041]; second[0.041]; second[0.041]; Ray[0.041]; seven[0.040]; Eric[0.040]; won[0.040]; win[0.040]; win[0.040]; Rockies[0.040]; Rockies[0.040]; career[0.040]; homers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ray Knight\u001b[39m ==> ENTITY: \u001b[32mRay Knight\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDeShields[0.043]; eighth[0.043]; seventh[0.042]; seventh[0.042]; Curtis[0.042]; Jeff[0.042]; pitch[0.042]; pitch[0.042]; Phillies[0.042]; ninth[0.042]; Dodgers[0.042]; gave[0.042]; Ryan[0.042]; won[0.042]; win[0.042]; Philadelphia[0.042]; Seaver[0.041]; Brown[0.041]; league[0.041]; Chad[0.041]; games[0.041]; victory[0.041]; scored[0.040]; pitcher[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew York\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew York Yankees\u001b[39m <---> \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.237:0.229[\u001b[31m0.008\u001b[39m]; local()= 0.192:0.071[\u001b[31m0.121\u001b[39m]; log p(e|m)= -5.116:-1.790[\u001b[32m3.326\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; game[0.046]; Giants[0.045]; Giants[0.045]; Mets[0.045]; Mets[0.045]; season[0.045]; Expos[0.045]; David[0.044]; career[0.043]; second[0.043]; second[0.043]; fourth[0.043]; straight[0.043]; straight[0.043]; sixth[0.042]; win[0.042]; win[0.042]; seventh[0.042]; seventh[0.042]; seventh[0.042]; beat[0.042]; Padres[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marvin Benard\u001b[39m ==> ENTITY: \u001b[32mMarvin Benard\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsixth[0.043]; Benard[0.042]; seventh[0.042]; seventh[0.042]; seventh[0.042]; fourth[0.042]; homer[0.042]; ninth[0.042]; ninth[0.042]; ninth[0.042]; second[0.042]; second[0.042]; games[0.041]; career[0.041]; hit[0.041]; hit[0.041]; hit[0.041]; game[0.041]; leadoff[0.041]; season[0.041]; Giants[0.041]; Giants[0.041]; double[0.041]; double[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Florida Marlins\u001b[39m ==> ENTITY: \u001b[32mMiami Marlins\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMarlins[0.044]; league[0.043]; league[0.043]; season[0.043]; fifth[0.043]; fifth[0.043]; game[0.043]; Cincinnati[0.043]; Cincinnati[0.043]; Baseball[0.043]; seventh[0.042]; League[0.042]; win[0.042]; won[0.041]; earned[0.041]; starts[0.041]; straight[0.041]; major[0.041]; Reds[0.040]; Reds[0.040]; Reds[0.040]; run[0.040]; Cardinals[0.040]; innings[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Phillies ==> ENTITY: \u001b[32mPhiladelphia Phillies\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.203:0.203[0]; log p(e|m)= -0.118:-0.118[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; league[0.044]; seventh[0.044]; Philadelphia[0.044]; eighth[0.043]; game[0.043]; games[0.043]; inning[0.042]; win[0.042]; Phillies[0.042]; won[0.042]; Worrell[0.041]; innings[0.041]; innings[0.041]; run[0.041]; Dodgers[0.041]; victory[0.041]; Braves[0.041]; doubleheader[0.040]; Jeff[0.040]; homer[0.040]; triple[0.040]; infield[0.040]; Reds[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alan Benes\u001b[39m ==> ENTITY: \u001b[32mAlan Benes\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; season[0.044]; Cardinals[0.044]; Cardinals[0.044]; seven[0.043]; homers[0.042]; seventh[0.042]; innings[0.042]; ninth[0.041]; Louis[0.041]; Louis[0.041]; career[0.041]; career[0.041]; eighth[0.041]; Eric[0.041]; hits[0.041]; hits[0.041]; Benes[0.041]; runs[0.041]; win[0.040]; beat[0.040]; run[0.040]; struck[0.040]; struck[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colorado Rockies ==> ENTITY: \u001b[32mColorado Rockies\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.102:-0.102[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; league[0.045]; season[0.044]; Colorado[0.043]; win[0.042]; Pittsburgh[0.042]; Giants[0.042]; homers[0.042]; hitter[0.041]; won[0.041]; Cardinals[0.041]; Cardinals[0.041]; Houston[0.041]; eighth[0.041]; major[0.041]; ninth[0.041]; Rockies[0.041]; Pirates[0.041]; seven[0.041]; Astros[0.041]; second[0.041]; career[0.040]; career[0.040]; innings[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Reds ==> ENTITY: \u001b[32mCincinnati Reds\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.192:0.192[0]; log p(e|m)= -1.024:-1.024[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; league[0.046]; game[0.044]; Cincinnati[0.043]; Cincinnati[0.043]; Baseball[0.043]; League[0.043]; fifth[0.043]; fifth[0.043]; seventh[0.042]; Cardinals[0.041]; Reds[0.041]; run[0.041]; major[0.041]; earned[0.041]; Friday[0.041]; win[0.040]; Marlins[0.040]; Marlins[0.040]; won[0.040]; inning[0.040]; National[0.039]; double[0.039]; batter[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wall\u001b[39m ==> ENTITY: \u001b[32mDonne Wall\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; games[0.046]; game[0.046]; game[0.046]; Ray[0.045]; career[0.044]; homers[0.044]; seven[0.044]; won[0.044]; Sean[0.043]; season[0.043]; season[0.043]; win[0.043]; win[0.043]; straight[0.043]; hit[0.043]; Houston[0.042]; Cardinals[0.042]; Eric[0.042]; doubles[0.042]; Astros[0.042]; Wall[0.042]; tied[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.066:0.066[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; league[0.046]; game[0.044]; game[0.044]; Curtis[0.043]; won[0.043]; Todd[0.042]; second[0.042]; second[0.042]; Jeff[0.042]; Billy[0.042]; Kevin[0.041]; Tom[0.041]; win[0.041]; win[0.041]; Jones[0.041]; Jones[0.041]; overall[0.040]; seventh[0.040]; Atlanta[0.040]; single[0.040]; single[0.040]; victory[0.040]; ninth[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Florida ==> ENTITY: \u001b[32mMiami Marlins\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.187:0.187[0]; log p(e|m)= -4.828:-4.828[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFlorida[0.048]; game[0.046]; season[0.045]; league[0.045]; Cincinnati[0.044]; fifth[0.044]; fifth[0.044]; League[0.044]; Marlins[0.044]; win[0.044]; won[0.043]; eighth[0.043]; seventh[0.043]; seventh[0.043]; ninth[0.043]; straight[0.042]; major[0.042]; tie[0.042]; Kevin[0.042]; Cardinals[0.042]; Todd[0.042]; threw[0.042]; threw[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mike Mordecai\u001b[39m ==> ENTITY: \u001b[32mMike Mordecai\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.215:0.215[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nExpos[0.043]; David[0.043]; Jose[0.043]; majors[0.043]; homered[0.042]; sixth[0.042]; shortstop[0.042]; ninth[0.042]; seventh[0.042]; infield[0.042]; games[0.041]; singled[0.041]; singled[0.041]; game[0.041]; second[0.041]; second[0.041]; second[0.041]; fourth[0.041]; home[0.041]; eighth[0.041]; tying[0.041]; Braves[0.041]; leadoff[0.041]; Kevin[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nolan Ryan\u001b[39m ==> ENTITY: \u001b[32mNolan Ryan\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.048]; pitcher[0.048]; eighth[0.046]; seventh[0.046]; seventh[0.046]; ninth[0.046]; pitch[0.046]; pitch[0.046]; games[0.046]; Jeff[0.046]; won[0.046]; win[0.045]; pitches[0.045]; Seaver[0.045]; Dodgers[0.045]; Jerry[0.045]; Worrell[0.044]; DeShields[0.044]; hit[0.044]; Reds[0.044]; Ray[0.044]; Todd[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shawn Estes\u001b[39m ==> ENTITY: \u001b[32mShawn Estes\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.215:0.215[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngame[0.044]; Giants[0.044]; Giants[0.044]; games[0.043]; season[0.043]; career[0.043]; Padres[0.042]; sixth[0.042]; second[0.041]; second[0.041]; seventh[0.041]; seventh[0.041]; seventh[0.041]; ninth[0.041]; ninth[0.041]; ninth[0.041]; pitch[0.041]; fourth[0.041]; Mets[0.041]; Mets[0.041]; Expos[0.041]; field[0.041]; Estes[0.040]; homer[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Francisco Giants ==> ENTITY: \u001b[32mSan Francisco Giants\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.178:0.178[0]; log p(e|m)= -0.076:-0.076[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; seventh[0.043]; seventh[0.043]; seventh[0.043]; eighth[0.043]; fourth[0.043]; ninth[0.043]; ninth[0.043]; ninth[0.043]; game[0.042]; second[0.042]; second[0.042]; win[0.041]; win[0.041]; Giants[0.041]; won[0.041]; straight[0.041]; straight[0.041]; straight[0.041]; season[0.040]; right[0.040]; right[0.040]; pitch[0.040]; field[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Los Angeles ==> ENTITY: \u001b[32mLos Angeles Dodgers\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.182:0.182[0]; log p(e|m)= -4.711:-4.711[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; league[0.044]; league[0.044]; game[0.044]; Philadelphia[0.044]; Dodgers[0.042]; home[0.042]; won[0.041]; Worrell[0.041]; hit[0.041]; Phillies[0.041]; Phillies[0.041]; eighth[0.041]; Braves[0.041]; seventh[0.041]; Atlanta[0.041]; infield[0.041]; win[0.041]; Curtis[0.040]; overall[0.040]; long[0.040]; Los[0.040]; Los[0.040]; manager[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colorado ==> ENTITY: \u001b[32mColorado Rockies\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.207:0.207[0]; log p(e|m)= -4.510:-4.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngame[0.045]; games[0.045]; Colorado[0.044]; league[0.044]; season[0.043]; second[0.042]; Houston[0.042]; career[0.042]; career[0.042]; loss[0.042]; Cardinals[0.041]; Cardinals[0.041]; Pittsburgh[0.041]; dropped[0.041]; major[0.041]; homers[0.041]; Pirates[0.041]; seven[0.041]; Astros[0.041]; win[0.041]; Rockies[0.040]; Rockies[0.040]; went[0.040]; Mets[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cardinals ==> ENTITY: \u001b[32mSt. Louis Cardinals\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.209:0.209[0]; log p(e|m)= -1.127:-1.127[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.044]; games[0.044]; Cardinals[0.044]; game[0.044]; game[0.044]; season[0.043]; season[0.043]; Pittsburgh[0.042]; Houston[0.042]; seven[0.041]; Pirates[0.041]; Louis[0.041]; Louis[0.041]; career[0.041]; second[0.041]; second[0.041]; Mets[0.041]; major[0.041]; Astros[0.041]; Ray[0.040]; homers[0.040]; innings[0.040]; eighth[0.040]; September[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cardinals ==> ENTITY: \u001b[32mSt. Louis Cardinals\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.215:0.215[0]; log p(e|m)= -1.127:-1.127[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; league[0.046]; Baseball[0.046]; game[0.046]; season[0.045]; League[0.045]; Cincinnati[0.045]; Cincinnati[0.045]; seventh[0.043]; fifth[0.043]; fifth[0.043]; Reds[0.043]; Reds[0.043]; Reds[0.043]; major[0.042]; John[0.042]; earned[0.042]; Marlins[0.042]; Marlins[0.042]; win[0.042]; starts[0.042]; Friday[0.042]; won[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Braves ==> ENTITY: \u001b[32mAtlanta Braves\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.214:0.214[0]; log p(e|m)= -0.285:-0.285[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; game[0.044]; game[0.044]; league[0.043]; second[0.043]; second[0.043]; eighth[0.042]; Braves[0.042]; Dodgers[0.042]; seventh[0.042]; Atlanta[0.042]; ninth[0.042]; Phillies[0.042]; Phillies[0.042]; Worrell[0.041]; straight[0.041]; doubleheader[0.041]; tying[0.041]; innings[0.041]; win[0.040]; win[0.040]; Chicago[0.040]; home[0.040]; home[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kevin Brown ==> ENTITY: \u001b[32mKevin Brown (right-handed pitcher)\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.182:0.182[0]; log p(e|m)= -0.263:-0.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBaseball[0.047]; league[0.047]; league[0.047]; League[0.046]; game[0.045]; fifth[0.044]; fifth[0.044]; seventh[0.044]; major[0.044]; Kevin[0.044]; Brown[0.043]; Brown[0.043]; lead[0.043]; Friday[0.043]; Marlins[0.042]; Marlins[0.042]; Todd[0.042]; win[0.042]; run[0.042]; average[0.041]; Cardinals[0.041]; earned[0.041]; White[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCincinnati\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mCincinnati Reds\u001b[39m <---> \u001b[32mCincinnati\u001b[39m\t\nSCORES: global= 0.261:0.239[\u001b[31m0.022\u001b[39m]; local()= 0.185:0.087[\u001b[31m0.098\u001b[39m]; log p(e|m)= -3.772:-0.460[\u001b[32m3.312\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCincinnati[0.045]; game[0.045]; league[0.044]; league[0.044]; Baseball[0.043]; League[0.043]; won[0.042]; fifth[0.042]; fifth[0.042]; win[0.042]; straight[0.042]; Todd[0.041]; Cardinals[0.041]; seventh[0.041]; major[0.041]; Reds[0.041]; Reds[0.041]; earned[0.041]; Kevin[0.040]; Kevin[0.040]; beat[0.040]; Marlins[0.040]; Marlins[0.040]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Beck\u001b[39m ==> ENTITY: \u001b[32mRod Beck\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.043]; game[0.043]; eighth[0.043]; ninth[0.042]; ninth[0.042]; ninth[0.042]; games[0.042]; career[0.042]; dropped[0.042]; Rod[0.042]; Shawn[0.042]; Giants[0.042]; Giants[0.042]; seventh[0.042]; closer[0.042]; hitting[0.041]; straight[0.041]; tie[0.041]; field[0.041]; win[0.041]; hit[0.041]; Francisco[0.041]; run[0.041]; run[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlanta ==> ENTITY: \u001b[32mAtlanta Braves\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.187:0.187[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; game[0.044]; Atlanta[0.044]; Expos[0.043]; Diego[0.043]; Braves[0.042]; majors[0.042]; Sanders[0.042]; Sanders[0.042]; won[0.041]; won[0.041]; Padres[0.041]; sixth[0.041]; Jose[0.041]; fourth[0.041]; seventh[0.041]; second[0.041]; second[0.041]; second[0.041]; Jones[0.041]; Jones[0.041]; Scott[0.041]; Chicago[0.040]; ninth[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMontreal\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMontreal Expos\u001b[39m <---> \u001b[32mMontreal\u001b[39m\t\nSCORES: global= 0.239:0.233[\u001b[31m0.006\u001b[39m]; local()= 0.200:0.129[\u001b[31m0.071\u001b[39m]; log p(e|m)= -4.510:-0.110[\u001b[32m4.400\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.048]; Montreal[0.047]; Montreal[0.047]; Expos[0.047]; game[0.045]; majors[0.044]; record[0.043]; Atlanta[0.043]; won[0.043]; won[0.043]; fourth[0.043]; Braves[0.043]; second[0.043]; second[0.043]; second[0.043]; second[0.043]; sixth[0.042]; Padres[0.042]; seventh[0.042]; seventh[0.042]; David[0.042]; win[0.042]; win[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPhiladelphia\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPhiladelphia Phillies\u001b[39m <---> \u001b[32mPhiladelphia\u001b[39m\t\nSCORES: global= 0.253:0.238[\u001b[31m0.015\u001b[39m]; local()= 0.193:0.062[\u001b[31m0.131\u001b[39m]; log p(e|m)= -4.510:-0.211[\u001b[32m4.299\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; league[0.045]; league[0.045]; Phillies[0.043]; Phillies[0.043]; Worrell[0.043]; Dodgers[0.042]; Curtis[0.042]; Todd[0.042]; seventh[0.041]; seventh[0.041]; starting[0.041]; Reds[0.041]; won[0.041]; ninth[0.041]; Ray[0.041]; eighth[0.041]; pitcher[0.041]; Billy[0.041]; early[0.040]; win[0.040]; overall[0.040]; Ryan[0.040]; Jerry[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mets ==> ENTITY: \u001b[32mNew York Mets\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.189:0.189[0]; log p(e|m)= -0.264:-0.264[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMets[0.046]; Mets[0.046]; games[0.043]; fourth[0.042]; second[0.042]; second[0.042]; season[0.042]; eighth[0.042]; seventh[0.042]; seventh[0.042]; seventh[0.042]; game[0.042]; ninth[0.042]; ninth[0.042]; ninth[0.042]; Giants[0.041]; Giants[0.041]; win[0.041]; Shawn[0.040]; straight[0.040]; straight[0.040]; won[0.040]; hit[0.039]; hit[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Benard\u001b[39m ==> ENTITY: \u001b[32mMarvin Benard\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \neighth[0.043]; Benard[0.043]; seventh[0.042]; seventh[0.042]; seventh[0.042]; fourth[0.042]; homer[0.042]; ninth[0.042]; ninth[0.042]; ninth[0.042]; second[0.042]; second[0.042]; games[0.041]; career[0.041]; hit[0.041]; hit[0.041]; game[0.041]; leadoff[0.041]; season[0.041]; Giants[0.041]; Giants[0.041]; double[0.041]; double[0.041]; Marvin[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Brent Mayne\u001b[39m ==> ENTITY: \u001b[32mBrent Mayne\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.218:0.218[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseventh[0.043]; game[0.043]; eighth[0.043]; pitch[0.043]; second[0.042]; runs[0.042]; Shawn[0.042]; season[0.042]; season[0.042]; went[0.042]; right[0.042]; win[0.042]; homer[0.042]; career[0.042]; ninth[0.041]; hit[0.041]; longest[0.041]; run[0.041]; run[0.041]; straight[0.041]; hitting[0.041]; Bobby[0.041]; Rockies[0.040]; plus[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Darren Dreifort\u001b[39m ==> ENTITY: \u001b[32mDarren Dreifort\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; league[0.045]; league[0.045]; seventh[0.044]; game[0.043]; eighth[0.043]; starting[0.043]; doubleheader[0.042]; pitcher[0.042]; Philadelphia[0.042]; home[0.042]; young[0.041]; Dodgers[0.041]; Worrell[0.041]; dropped[0.041]; tie[0.040]; straight[0.040]; innings[0.040]; innings[0.040]; split[0.040]; Angeles[0.040]; Angeles[0.040]; Angeles[0.040]; scoreless[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rockies ==> ENTITY: \u001b[32mColorado Rockies\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.914:-0.914[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; game[0.044]; game[0.044]; league[0.043]; hits[0.042]; hits[0.042]; season[0.042]; season[0.042]; homers[0.042]; Colorado[0.042]; Colorado[0.042]; win[0.042]; win[0.042]; innings[0.042]; runs[0.041]; Cardinals[0.041]; Cardinals[0.041]; Houston[0.041]; Pittsburgh[0.041]; major[0.040]; Rockies[0.040]; Pirates[0.040]; won[0.040]; seven[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sean Berry\u001b[39m ==> ENTITY: \u001b[32mSean Berry\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.043]; Ray[0.043]; league[0.043]; Astros[0.043]; Eric[0.043]; Berry[0.042]; Rockies[0.042]; Rockies[0.042]; second[0.042]; second[0.042]; Cardinals[0.042]; Cardinals[0.042]; Tom[0.042]; Pirates[0.041]; Colorado[0.041]; Colorado[0.041]; season[0.041]; season[0.041]; Pagnozzi[0.041]; games[0.041]; doubles[0.041]; Benes[0.041]; Benes[0.041]; homers[0.041]; \t\n\r [============>.................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s425ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1440/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1102testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWorld Cup\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFIFA World Cup\u001b[39m <---> \u001b[32m1998 FIFA World Cup\u001b[39m\t\nSCORES: global= 0.253:0.232[\u001b[31m0.021\u001b[39m]; local()= 0.169:0.183[\u001b[32m0.013\u001b[39m]; log p(e|m)= -1.002:-4.200[\u001b[31m3.197\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKarembeu[0.045]; team[0.043]; European[0.043]; Lamouchi[0.043]; finals[0.043]; finals[0.043]; France[0.043]; France[0.043]; Djorkaeff[0.042]; Djorkaeff[0.042]; Djorkaeff[0.042]; Mexico[0.042]; playing[0.041]; Thuram[0.041]; play[0.041]; hosting[0.041]; Desailly[0.041]; Milan[0.041]; Bixente[0.041]; month[0.040]; injury[0.040]; Zidane[0.040]; Zidane[0.040]; championship[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance national football team\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.195:0.195[0]; log p(e|m)= -3.381:-3.381[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.048]; Lamouchi[0.045]; team[0.044]; Karembeu[0.044]; finals[0.044]; finals[0.044]; Deschamps[0.043]; Cup[0.042]; Djorkaeff[0.041]; Djorkaeff[0.041]; championship[0.041]; European[0.041]; Thuram[0.041]; Nantes[0.041]; Desailly[0.041]; playing[0.040]; Bixente[0.040]; Zidane[0.040]; Laurent[0.040]; Didier[0.040]; Blanc[0.040]; midfielder[0.040]; Duilio[0.039]; Youri[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Franck Leboeuf\u001b[39m ==> ENTITY: \u001b[32mFrank Leboeuf\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuro[0.044]; Euro[0.044]; matches[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; Djorkaeff[0.042]; Djorkaeff[0.042]; Juventus[0.042]; Leboeuf[0.042]; having[0.041]; Soccer[0.041]; play[0.041]; international[0.041]; Jacquet[0.041]; second[0.041]; second[0.041]; Ricardo[0.040]; substitutions[0.040]; Cup[0.040]; Cup[0.040]; Youri[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nicolas Ouedec\u001b[39m ==> ENTITY: \u001b[32mNicolas Ouédec\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBixente[0.045]; Deschamps[0.044]; Karembeu[0.043]; team[0.043]; Blanc[0.043]; France[0.043]; France[0.043]; Thuram[0.043]; Desailly[0.042]; Laurent[0.042]; finals[0.042]; Cup[0.042]; Pedros[0.041]; Lamouchi[0.041]; Patrice[0.041]; Didier[0.041]; Marcel[0.041]; Franck[0.040]; Reynald[0.040]; Youri[0.040]; Lizarazu[0.040]; Djorkaeff[0.039]; Loko[0.039]; Florian[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Florian Maurice\u001b[39m ==> ENTITY: \u001b[32mFlorian Maurice\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLamouchi[0.045]; Deschamps[0.044]; Bixente[0.044]; Djorkaeff[0.043]; team[0.043]; France[0.043]; France[0.043]; Cup[0.043]; Didier[0.042]; Pedros[0.042]; Thuram[0.042]; Marcel[0.042]; Franck[0.042]; Bernard[0.041]; Karembeu[0.041]; Youri[0.041]; Patrice[0.041]; finals[0.040]; Laurent[0.040]; Lizarazu[0.040]; Desailly[0.040]; month[0.040]; Pavel[0.040]; Nicolas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Youri Djorkaeff\u001b[39m ==> ENTITY: \u001b[32mYouri Djorkaeff\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.045]; Nicolas[0.044]; international[0.043]; France[0.043]; France[0.043]; France[0.043]; matches[0.043]; Chelsea[0.043]; Cup[0.042]; Cup[0.042]; Euro[0.041]; Euro[0.041]; defender[0.041]; sending[0.041]; shown[0.041]; goals[0.040]; Ricardo[0.040]; World[0.040]; World[0.040]; bookable[0.040]; second[0.040]; Mexico[0.040]; scored[0.040]; hosts[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance national football team\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.173:0.173[0]; log p(e|m)= -3.381:-3.381[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.048]; France[0.048]; matches[0.044]; Euro[0.043]; Saturday[0.042]; international[0.042]; second[0.042]; second[0.042]; Djorkaeff[0.042]; Djorkaeff[0.042]; European[0.041]; Nantes[0.041]; finals[0.041]; championship[0.041]; defender[0.041]; defender[0.041]; Juventus[0.041]; playing[0.040]; Zidane[0.040]; play[0.040]; goals[0.040]; midfielder[0.040]; score[0.040]; coach[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jesus Arellano\u001b[39m ==> ENTITY: \u001b[32mJesús Arellano\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.047]; Enrique[0.045]; Luis[0.044]; Osvaldo[0.043]; Francisco[0.043]; Bixente[0.042]; Alberto[0.042]; Ricardo[0.042]; Pedros[0.042]; Cuauhtemoc[0.042]; Joaquin[0.041]; Djorkaeff[0.041]; Raul[0.041]; Rodrigo[0.041]; Sanchez[0.041]; Suarez[0.041]; Karembeu[0.041]; Nicolas[0.041]; Blanco[0.040]; del[0.040]; Lizarazu[0.040]; Florian[0.040]; Pavel[0.040]; Zidane[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Enrique Alfaro\u001b[39m ==> ENTITY: \u001b[32mEnrique Alfaro\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.046]; Raul[0.045]; Bixente[0.045]; Ricardo[0.044]; Suarez[0.043]; Luis[0.043]; Pedros[0.042]; Blanco[0.042]; Gomez[0.042]; Francisco[0.041]; Osvaldo[0.041]; Alfonso[0.041]; Sanchez[0.041]; Djorkaeff[0.041]; Pires[0.041]; Garcia[0.040]; Garcia[0.040]; Lizarazu[0.040]; Alberto[0.040]; Palencia[0.040]; Cuauhtemoc[0.040]; Florian[0.040]; Rodrigo[0.040]; Pavel[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nantes ==> ENTITY: \u001b[32mFC Nantes\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.214:0.214[0]; log p(e|m)= -1.457:-1.457[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLamouchi[0.046]; France[0.045]; France[0.045]; France[0.045]; France[0.045]; Cup[0.043]; play[0.043]; team[0.042]; hosting[0.042]; championship[0.041]; playing[0.041]; Djorkaeff[0.041]; Djorkaeff[0.041]; Juventus[0.040]; finals[0.040]; finals[0.040]; second[0.040]; second[0.040]; European[0.040]; midfielder[0.040]; defender[0.040]; Barcelona[0.040]; Thuram[0.040]; forward[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Aime Jacquet\u001b[39m ==> ENTITY: \u001b[32mAimé Jacquet\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.045]; France[0.045]; France[0.045]; France[0.045]; France[0.045]; Soccer[0.044]; international[0.044]; matches[0.043]; Cup[0.042]; Cup[0.042]; Euro[0.042]; Euro[0.042]; coach[0.041]; midfield[0.041]; bookable[0.041]; Djorkaeff[0.040]; semifinal[0.040]; substitutions[0.040]; Mexico[0.040]; victory[0.040]; came[0.039]; second[0.039]; second[0.039]; Youri[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bernard Lama\u001b[39m ==> ENTITY: \u001b[32mBernard Lama\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLaurent[0.044]; Deschamps[0.044]; team[0.044]; Pedros[0.042]; Lamouchi[0.042]; Nicolas[0.042]; Didier[0.042]; Franck[0.042]; month[0.042]; Marcel[0.042]; Djorkaeff[0.042]; Djorkaeff[0.042]; Blanc[0.042]; France[0.042]; France[0.042]; Youri[0.041]; playing[0.041]; injury[0.041]; finals[0.041]; finals[0.041]; Karembeu[0.041]; Desailly[0.041]; Lizarazu[0.040]; championship[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lilian Thuram\u001b[39m ==> ENTITY: \u001b[32mLilian Thuram\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLamouchi[0.046]; France[0.044]; France[0.044]; Karembeu[0.044]; Milan[0.043]; Djorkaeff[0.042]; Djorkaeff[0.042]; team[0.042]; Zidane[0.042]; Desailly[0.042]; Lizarazu[0.041]; Pires[0.041]; Bixente[0.041]; Youri[0.041]; playing[0.041]; Didier[0.041]; Deschamps[0.041]; European[0.041]; Pedros[0.040]; Franck[0.040]; Zinedine[0.040]; Internazionale[0.040]; Claudio[0.040]; month[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Patrice Loko\u001b[39m ==> ENTITY: \u001b[32mPatrice Loko\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.121:0.121[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.047]; finals[0.046]; Pedros[0.043]; goalkeeper[0.043]; Lamouchi[0.043]; Cup[0.043]; hosting[0.042]; Marcel[0.042]; Nicolas[0.042]; Bixente[0.042]; Lizarazu[0.041]; Robert[0.041]; Djorkaeff[0.041]; Florian[0.041]; Pavel[0.041]; Duilio[0.041]; Teams[0.040]; Youri[0.040]; Karembeu[0.040]; Laurent[0.040]; Osvaldo[0.040]; Osvaldo[0.040]; Thuram[0.040]; diving[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zinedine Zidane\u001b[39m ==> ENTITY: \u001b[32mZinedine Zidane\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLamouchi[0.045]; Karembeu[0.044]; France[0.044]; France[0.044]; Djorkaeff[0.044]; Thuram[0.042]; Didier[0.042]; Desailly[0.042]; Nicolas[0.042]; Youri[0.042]; team[0.042]; Bixente[0.042]; Laurent[0.041]; Pedros[0.041]; Marcel[0.041]; month[0.041]; Villa[0.040]; goalkeeper[0.040]; Ricardo[0.040]; finals[0.040]; Enrique[0.040]; Cup[0.040]; Alberto[0.040]; Deschamps[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance national football team\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.163:0.163[0]; log p(e|m)= -3.381:-3.381[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.049]; France[0.049]; Euro[0.043]; Djorkaeff[0.042]; Djorkaeff[0.042]; European[0.042]; Nantes[0.042]; defender[0.041]; defender[0.041]; Juventus[0.041]; playing[0.041]; championship[0.041]; Zidane[0.041]; play[0.041]; finals[0.041]; second[0.040]; second[0.040]; penalty[0.040]; semifinal[0.040]; goals[0.040]; midfielder[0.040]; elimination[0.040]; coach[0.040]; substitutions[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Francisco Palencia\u001b[39m ==> ENTITY: \u001b[32mFrancisco Palencia\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.046]; Luis[0.045]; Enrique[0.044]; Raul[0.044]; Ricardo[0.044]; Alberto[0.043]; Osvaldo[0.042]; Rodrigo[0.042]; Karembeu[0.042]; Suarez[0.042]; Pedros[0.041]; Joaquin[0.041]; del[0.041]; Blanco[0.041]; Djorkaeff[0.041]; Gomez[0.040]; Rios[0.040]; Cuauhtemoc[0.040]; Garcia[0.040]; Garcia[0.040]; Claudio[0.040]; Nicolas[0.040]; Lizarazu[0.040]; Florian[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Luis Garcia ==> ENTITY: \u001b[32mLuis García Postigo\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.212:0.212[0]; log p(e|m)= -1.625:-1.625[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEnrique[0.044]; Mexico[0.044]; Ramon[0.043]; Joaquin[0.043]; Rodrigo[0.043]; Raul[0.042]; Garcia[0.042]; Osvaldo[0.042]; Suarez[0.042]; Francisco[0.042]; Ricardo[0.042]; Alberto[0.042]; Gomez[0.042]; Nicolas[0.041]; del[0.041]; Sanchez[0.041]; Rios[0.041]; Robert[0.040]; Karembeu[0.040]; Claudio[0.040]; Pardo[0.040]; Palencia[0.040]; Cuauhtemoc[0.040]; Becerril[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance national football team\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.166:0.166[0]; log p(e|m)= -3.381:-3.381[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.049]; France[0.049]; matches[0.046]; Soccer[0.044]; Euro[0.043]; Euro[0.043]; Saturday[0.042]; Cup[0.042]; Cup[0.042]; international[0.042]; second[0.042]; Paris[0.042]; Djorkaeff[0.042]; penalty[0.040]; semifinal[0.040]; elimination[0.040]; Mexico[0.040]; coach[0.039]; World[0.039]; World[0.039]; time[0.039]; central[0.039]; having[0.039]; early[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sabri Lamouchi\u001b[39m ==> ENTITY: \u001b[32mSabri Lamouchi\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmonth[0.044]; Karembeu[0.044]; team[0.044]; Djorkaeff[0.043]; Djorkaeff[0.043]; Pedros[0.043]; Milan[0.042]; Internazionale[0.042]; France[0.042]; France[0.042]; minutes[0.041]; Zinedine[0.041]; Nantes[0.041]; Bixente[0.041]; World[0.041]; finals[0.041]; finals[0.041]; midfielder[0.041]; Desailly[0.041]; Nicolas[0.041]; Zidane[0.041]; Thuram[0.041]; Alberto[0.041]; Lizarazu[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Laurent Blanc\u001b[39m ==> ENTITY: \u001b[32mLaurent Blanc\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLamouchi[0.048]; Karembeu[0.046]; team[0.046]; Didier[0.045]; Deschamps[0.045]; Cup[0.044]; Djorkaeff[0.044]; Djorkaeff[0.044]; Bixente[0.044]; Desailly[0.044]; Youri[0.043]; Thuram[0.043]; France[0.043]; France[0.043]; Nicolas[0.043]; Pedros[0.043]; Marcel[0.042]; month[0.042]; Zidane[0.042]; Internazionale[0.042]; midfielder[0.042]; European[0.042]; forward[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Leboeuf\u001b[39m ==> ENTITY: \u001b[32mFrank Leboeuf\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuro[0.044]; Euro[0.044]; matches[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; Djorkaeff[0.042]; Djorkaeff[0.042]; Juventus[0.042]; Leboeuf[0.042]; having[0.041]; Soccer[0.041]; play[0.041]; international[0.041]; Jacquet[0.041]; second[0.041]; second[0.041]; Ricardo[0.040]; substitutions[0.040]; Cup[0.040]; Cup[0.040]; Youri[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexico ==> ENTITY: \u001b[32mMexico national football team\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.216:0.216[0]; log p(e|m)= -2.865:-2.865[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.045]; Enrique[0.043]; Garcia[0.043]; Garcia[0.043]; Luis[0.042]; Raul[0.042]; Ramirez[0.042]; Lamouchi[0.042]; Sanchez[0.042]; Sanchez[0.042]; Joaquin[0.042]; Karembeu[0.042]; Djorkaeff[0.042]; Rios[0.041]; Alberto[0.041]; Ramon[0.041]; Francisco[0.041]; Ricardo[0.041]; Osvaldo[0.041]; Osvaldo[0.041]; Bixente[0.040]; Cup[0.040]; Suarez[0.040]; Rodrigo[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pavel Pardo\u001b[39m ==> ENTITY: \u001b[32mPável Pardo\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.044]; Lamouchi[0.043]; Raul[0.043]; Ricardo[0.043]; Alberto[0.043]; Luis[0.043]; Cup[0.043]; Enrique[0.043]; Mexico[0.042]; Karembeu[0.042]; Rodrigo[0.042]; Gomez[0.041]; Francisco[0.041]; Osvaldo[0.041]; Bixente[0.041]; Djorkaeff[0.041]; Suarez[0.041]; del[0.041]; Joaquin[0.041]; Pedros[0.041]; Nicolas[0.040]; Sanchez[0.040]; Garcia[0.039]; Garcia[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Juventus\u001b[39m ==> ENTITY: \u001b[32mJuventus F.C.\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMilan[0.045]; Djorkaeff[0.044]; Djorkaeff[0.044]; Ricardo[0.043]; Internazionale[0.043]; Osvaldo[0.042]; Zidane[0.042]; goals[0.042]; midfielder[0.042]; month[0.042]; Chelsea[0.042]; Chelsea[0.042]; defender[0.041]; defender[0.041]; Cup[0.041]; penalty[0.041]; play[0.041]; substitute[0.041]; playing[0.041]; European[0.041]; second[0.041]; second[0.041]; Barcelona[0.041]; joined[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ricardo Pelaez\u001b[39m ==> ENTITY: \u001b[32mRicardo Peláez\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.046]; Enrique[0.045]; Rodrigo[0.044]; Pedros[0.044]; Cuauhtemoc[0.043]; Luis[0.043]; Suarez[0.042]; Alberto[0.042]; Osvaldo[0.042]; Djorkaeff[0.041]; Florian[0.041]; Pavel[0.041]; Palencia[0.041]; Zinedine[0.041]; Francisco[0.041]; Raul[0.041]; Garcia[0.040]; Garcia[0.040]; Youri[0.040]; Claudio[0.040]; Alfaro[0.040]; del[0.040]; Joaquin[0.040]; Ramon[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Joaquin del Olmo\u001b[39m ==> ENTITY: \u001b[32mJoaquín del Olmo\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.045]; Karembeu[0.044]; Mexicans[0.043]; Djorkaeff[0.043]; Djorkaeff[0.043]; Cup[0.043]; Lamouchi[0.043]; Bixente[0.042]; midfielder[0.042]; championship[0.041]; finals[0.041]; finals[0.041]; playing[0.041]; defender[0.041]; Ricardo[0.041]; returning[0.041]; Barcelona[0.041]; World[0.041]; France[0.040]; France[0.040]; France[0.040]; France[0.040]; past[0.040]; Blanc[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Euro 96\u001b[39m ==> ENTITY: \u001b[32mUEFA Euro 1996\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.048]; Euro[0.044]; second[0.044]; second[0.044]; international[0.043]; Soccer[0.043]; World[0.043]; World[0.043]; Cup[0.042]; Cup[0.042]; semifinal[0.041]; France[0.040]; France[0.040]; France[0.040]; France[0.040]; France[0.040]; goals[0.040]; goals[0.040]; Djorkaeff[0.040]; hosts[0.040]; Mexico[0.040]; Chelsea[0.040]; Chelsea[0.040]; marred[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marcel Desailly\u001b[39m ==> ENTITY: \u001b[32mMarcel Desailly\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.225:0.225[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLamouchi[0.045]; France[0.044]; France[0.044]; Djorkaeff[0.044]; Djorkaeff[0.044]; Thuram[0.044]; Laurent[0.043]; Zidane[0.043]; Didier[0.042]; Karembeu[0.042]; Nicolas[0.042]; Deschamps[0.042]; Youri[0.042]; team[0.041]; Blanc[0.041]; Cup[0.041]; Zinedine[0.041]; midfielder[0.040]; Lizarazu[0.040]; Franck[0.040]; Pedros[0.040]; playing[0.040]; Patrice[0.040]; Pires[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Franck Leboeuf\u001b[39m ==> ENTITY: \u001b[32mFrank Leboeuf\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLamouchi[0.044]; playing[0.044]; France[0.044]; France[0.044]; team[0.043]; Djorkaeff[0.043]; Djorkaeff[0.043]; Bernard[0.042]; Pedros[0.042]; Laurent[0.042]; Teams[0.041]; Milan[0.041]; Jacquet[0.041]; Alberto[0.041]; Desailly[0.041]; Cup[0.041]; Youri[0.041]; Bixente[0.041]; finals[0.041]; Karembeu[0.041]; Suarez[0.041]; Lizarazu[0.040]; Didier[0.040]; Florian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nicolas Ouedec\u001b[39m ==> ENTITY: \u001b[32mNicolas Ouédec\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.051]; Soccer[0.047]; France[0.046]; France[0.046]; France[0.046]; goals[0.045]; Cup[0.045]; Cup[0.045]; scored[0.044]; Franck[0.043]; Youri[0.043]; bookable[0.042]; Paris[0.042]; Djorkaeff[0.042]; Aime[0.042]; coach[0.042]; semifinal[0.042]; Jacquet[0.042]; Euro[0.041]; Euro[0.041]; unbeaten[0.041]; international[0.041]; substitute[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Claudio Suarez\u001b[39m ==> ENTITY: \u001b[32mClaudio Suárez\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.045]; Luis[0.045]; Alberto[0.044]; Enrique[0.044]; team[0.043]; Raul[0.043]; Osvaldo[0.043]; Cuauhtemoc[0.042]; Pedros[0.042]; Rodrigo[0.041]; Palencia[0.041]; Djorkaeff[0.041]; Ricardo[0.041]; Francisco[0.041]; del[0.041]; Garcia[0.041]; Garcia[0.041]; Lamouchi[0.041]; Karembeu[0.041]; Bixente[0.040]; Rios[0.040]; finals[0.040]; World[0.040]; Nicolas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Duilio Davino\u001b[39m ==> ENTITY: \u001b[32mDuilio Davino\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.046]; Mexico[0.044]; Alberto[0.044]; Enrique[0.043]; team[0.043]; Ricardo[0.043]; Francisco[0.043]; Gomez[0.042]; del[0.041]; Raul[0.041]; Pedros[0.041]; Blanco[0.041]; Nicolas[0.041]; Osvaldo[0.041]; Sanchez[0.041]; Ramon[0.041]; Cuauhtemoc[0.041]; Joaquin[0.041]; Rodrigo[0.041]; Didier[0.040]; Lamouchi[0.040]; Karembeu[0.040]; Bixente[0.040]; Youri[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zinedine Zidane\u001b[39m ==> ENTITY: \u001b[32mZinedine Zidane\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.044]; France[0.044]; France[0.044]; Djorkaeff[0.044]; Djorkaeff[0.044]; midfielder[0.042]; team[0.042]; second[0.042]; second[0.042]; defender[0.042]; defender[0.042]; joined[0.041]; playing[0.041]; injury[0.041]; month[0.041]; penalty[0.041]; play[0.041]; score[0.041]; European[0.041]; substitute[0.041]; goalkeeper[0.040]; Milan[0.040]; Ricardo[0.040]; Juventus[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexico ==> ENTITY: \u001b[32mMexico national football team\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.171:0.171[0]; log p(e|m)= -2.865:-2.865[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.046]; Mexican[0.045]; matches[0.045]; Cup[0.045]; Cup[0.045]; Djorkaeff[0.043]; semifinal[0.043]; Ricardo[0.042]; international[0.042]; World[0.041]; World[0.041]; second[0.041]; second[0.041]; having[0.041]; France[0.040]; France[0.040]; France[0.040]; elimination[0.040]; friendly[0.040]; Euro[0.040]; Euro[0.040]; scored[0.039]; defender[0.039]; defender[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Djorkaeff\u001b[39m ==> ENTITY: \u001b[32mYouri Djorkaeff\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.222:0.222[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLamouchi[0.045]; Karembeu[0.045]; Djorkaeff[0.044]; Djorkaeff[0.044]; Youri[0.044]; Didier[0.042]; team[0.042]; Juventus[0.042]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; Desailly[0.041]; Laurent[0.041]; Milan[0.041]; Chelsea[0.041]; Internazionale[0.040]; midfielder[0.040]; Cup[0.040]; Osvaldo[0.040]; Deschamps[0.040]; Zidane[0.040]; Lizarazu[0.040]; Thuram[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bixente Lizarazu\u001b[39m ==> ENTITY: \u001b[32mBixente Lizarazu\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.224:0.224[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLamouchi[0.046]; Djorkaeff[0.044]; Djorkaeff[0.044]; Youri[0.043]; France[0.043]; France[0.043]; playing[0.043]; Karembeu[0.043]; Didier[0.043]; Laurent[0.042]; midfielder[0.042]; team[0.041]; Nicolas[0.041]; Pedros[0.040]; Alberto[0.040]; Zidane[0.040]; Thuram[0.040]; Osvaldo[0.040]; Osvaldo[0.040]; Pires[0.040]; Bernard[0.040]; Marcel[0.040]; Cup[0.040]; forward[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Christian Karembeu\u001b[39m ==> ENTITY: \u001b[32mChristian Karembeu\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.043]; France[0.043]; team[0.043]; Lamouchi[0.043]; Cup[0.043]; playing[0.042]; Djorkaeff[0.042]; Djorkaeff[0.042]; Thuram[0.042]; Desailly[0.042]; Deschamps[0.042]; Pedros[0.042]; Bixente[0.042]; Didier[0.041]; Nicolas[0.041]; Lizarazu[0.041]; Zinedine[0.041]; finals[0.040]; month[0.040]; Zidane[0.040]; Laurent[0.040]; Blanc[0.040]; Franck[0.040]; Youri[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMexican\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMexico national football team\u001b[39m <---> \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.251:0.247[\u001b[31m0.004\u001b[39m]; local()= 0.182:0.111[\u001b[31m0.071\u001b[39m]; log p(e|m)= -3.730:-0.397[\u001b[32m3.332\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.049]; matches[0.045]; Mexicans[0.044]; championship[0.044]; Djorkaeff[0.043]; Djorkaeff[0.043]; Ricardo[0.042]; international[0.042]; second[0.042]; second[0.042]; Cup[0.041]; Cup[0.041]; finals[0.041]; play[0.041]; victory[0.040]; having[0.040]; European[0.040]; early[0.040]; France[0.040]; France[0.040]; France[0.040]; France[0.040]; Barcelona[0.040]; score[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = German Villa\u001b[39m ==> ENTITY: \u001b[32mGermán Villa\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBixente[0.045]; Enrique[0.044]; Mexico[0.043]; Osvaldo[0.043]; Karembeu[0.043]; Lamouchi[0.042]; Ricardo[0.042]; Raul[0.042]; team[0.042]; Pedros[0.042]; Luis[0.042]; Rodrigo[0.041]; Cuauhtemoc[0.041]; Joaquin[0.041]; Alberto[0.041]; Lizarazu[0.041]; Youri[0.041]; Francisco[0.041]; Duilio[0.041]; del[0.041]; Djorkaeff[0.040]; Suarez[0.040]; Thuram[0.040]; Ramon[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gomez ==> ENTITY: \u001b[32mHerculez Gomez\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.151:0.151[0]; log p(e|m)= -2.551:-2.551[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.045]; Alberto[0.043]; team[0.043]; Osvaldo[0.043]; Enrique[0.043]; Ricardo[0.043]; Ramirez[0.043]; Garcia[0.042]; Garcia[0.042]; Rios[0.042]; Rodrigo[0.042]; Raul[0.042]; del[0.041]; Alfonso[0.041]; Bixente[0.041]; Sanchez[0.041]; Francisco[0.041]; Alfaro[0.041]; Mexico[0.040]; Cuauhtemoc[0.040]; Claudio[0.040]; Suarez[0.040]; Becerril[0.040]; Joaquin[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Cup ==> ENTITY: \u001b[32mFIFA World Cup\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.147:0.147[0]; log p(e|m)= -1.002:-1.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.048]; Soccer[0.047]; Cup[0.045]; Euro[0.045]; Euro[0.045]; international[0.045]; World[0.045]; France[0.044]; France[0.044]; France[0.044]; second[0.044]; Mexico[0.044]; Djorkaeff[0.043]; coach[0.043]; hosts[0.042]; friendly[0.042]; time[0.042]; having[0.041]; defender[0.041]; defender[0.041]; penalty[0.041]; scored[0.041]; Ricardo[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Djorkaeff\u001b[39m ==> ENTITY: \u001b[32mYouri Djorkaeff\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDjorkaeff[0.045]; team[0.043]; Juventus[0.043]; France[0.043]; France[0.043]; France[0.043]; Milan[0.042]; Chelsea[0.042]; Chelsea[0.042]; Internazionale[0.042]; midfielder[0.042]; Cup[0.042]; Osvaldo[0.041]; Zidane[0.041]; European[0.041]; joined[0.041]; Joaquin[0.041]; Barcelona[0.041]; del[0.041]; finals[0.041]; finals[0.041]; defender[0.040]; defender[0.040]; sending[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Paris ==> ENTITY: \u001b[32mParis\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.103:0.103[0]; log p(e|m)= -0.115:-0.115[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.046]; France[0.045]; France[0.045]; France[0.045]; Nicolas[0.044]; Soccer[0.044]; semifinal[0.044]; Cup[0.043]; Cup[0.043]; second[0.042]; Saturday[0.041]; Ricardo[0.041]; victory[0.041]; World[0.041]; World[0.041]; beat[0.040]; Franck[0.040]; marred[0.040]; coach[0.040]; having[0.039]; hosts[0.039]; central[0.039]; dismissed[0.039]; time[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Internazionale Milan\u001b[39m ==> ENTITY: \u001b[32mInter Milan\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLamouchi[0.048]; Juventus[0.046]; Karembeu[0.045]; Cup[0.042]; team[0.042]; month[0.042]; Djorkaeff[0.042]; Djorkaeff[0.042]; Djorkaeff[0.042]; Barcelona[0.042]; second[0.041]; Bixente[0.041]; Pedros[0.041]; goals[0.041]; Osvaldo[0.041]; playing[0.041]; European[0.041]; joined[0.041]; Deschamps[0.041]; midfielder[0.041]; finals[0.040]; finals[0.040]; defender[0.040]; Youri[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chelsea ==> ENTITY: \u001b[32mChelsea F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.518:-0.518[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; Soccer[0.045]; Chelsea[0.045]; Cup[0.043]; Cup[0.043]; play[0.042]; Mexican[0.042]; goals[0.042]; goals[0.042]; second[0.042]; second[0.042]; defender[0.041]; defender[0.041]; early[0.041]; Saturday[0.041]; Djorkaeff[0.041]; Djorkaeff[0.041]; Mexicans[0.040]; Juventus[0.040]; Ricardo[0.040]; substitutions[0.040]; coach[0.040]; having[0.040]; international[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Didier Deschamps\u001b[39m ==> ENTITY: \u001b[32mDidier Deschamps\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.215:0.215[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLamouchi[0.047]; Karembeu[0.045]; Bixente[0.043]; France[0.043]; France[0.043]; Internazionale[0.042]; Milan[0.042]; Desailly[0.042]; Djorkaeff[0.041]; Djorkaeff[0.041]; Alberto[0.041]; team[0.041]; Laurent[0.041]; Claudio[0.041]; Lizarazu[0.041]; Nicolas[0.041]; Thuram[0.041]; Pedros[0.041]; Enrique[0.041]; Blanc[0.041]; Youri[0.041]; Zidane[0.040]; Franck[0.040]; Zinedine[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Youri Djorkaeff\u001b[39m ==> ENTITY: \u001b[32mYouri Djorkaeff\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLamouchi[0.046]; Karembeu[0.045]; Djorkaeff[0.044]; Didier[0.043]; team[0.042]; Nicolas[0.042]; Claudio[0.042]; Enrique[0.042]; France[0.042]; France[0.042]; Desailly[0.042]; Laurent[0.041]; Milan[0.041]; Alberto[0.041]; Luis[0.041]; Internazionale[0.041]; Cup[0.041]; Osvaldo[0.041]; Osvaldo[0.041]; Deschamps[0.040]; Zidane[0.040]; Lizarazu[0.040]; Pedros[0.040]; Thuram[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ricardo Pelaez\u001b[39m ==> ENTITY: \u001b[32mRicardo Peláez\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.047]; Mexico[0.045]; international[0.044]; scored[0.043]; time[0.042]; time[0.042]; Mexican[0.042]; midfield[0.041]; finals[0.041]; second[0.041]; second[0.041]; bookable[0.041]; goals[0.041]; goals[0.041]; play[0.041]; Djorkaeff[0.041]; Djorkaeff[0.041]; Mexicans[0.041]; Zinedine[0.041]; semifinal[0.041]; World[0.041]; coach[0.041]; Cup[0.040]; Espanyol[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Joaquin del Olmo\u001b[39m ==> ENTITY: \u001b[32mJoaquín del Olmo\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.213:0.213[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.049]; Karembeu[0.043]; Enrique[0.043]; Djorkaeff[0.043]; Luis[0.043]; Raul[0.043]; Lamouchi[0.042]; Alberto[0.042]; Bixente[0.042]; Francisco[0.042]; Ramon[0.042]; Pedros[0.042]; Ricardo[0.041]; Arellano[0.041]; Rodrigo[0.041]; Claudio[0.041]; Suarez[0.041]; Rios[0.040]; Cuauhtemoc[0.040]; France[0.040]; Blanc[0.040]; Davino[0.040]; Ramirez[0.040]; Sanchez[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cuauhtemoc Blanco\u001b[39m ==> ENTITY: \u001b[32mCuauhtémoc Blanco\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.229:0.229[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEnrique[0.045]; Luis[0.045]; Mexico[0.045]; Ricardo[0.044]; Rodrigo[0.043]; Raul[0.043]; Alberto[0.043]; Osvaldo[0.043]; Joaquin[0.043]; Francisco[0.043]; Djorkaeff[0.041]; Lamouchi[0.041]; Pedros[0.040]; Gomez[0.040]; Suarez[0.040]; Alfonso[0.040]; Nicolas[0.040]; del[0.040]; Garcia[0.040]; Garcia[0.040]; Sanchez[0.040]; Bixente[0.040]; Claudio[0.040]; Youri[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Espanyol\u001b[39m ==> ENTITY: \u001b[32mRCD Espanyol\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexican[0.046]; team[0.046]; Djorkaeff[0.043]; Djorkaeff[0.043]; Barcelona[0.042]; Osvaldo[0.042]; Cup[0.042]; Juventus[0.042]; Mexicans[0.042]; month[0.042]; Joaquin[0.041]; Milan[0.041]; came[0.041]; Ricardo[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; Internazionale[0.041]; joined[0.041]; playing[0.041]; del[0.041]; second[0.040]; second[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance national football team\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.181:0.181[0]; log p(e|m)= -3.381:-3.381[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.050]; France[0.050]; France[0.050]; matches[0.047]; Soccer[0.045]; Euro[0.044]; Euro[0.044]; Saturday[0.044]; Cup[0.043]; Cup[0.043]; international[0.043]; second[0.043]; second[0.043]; Paris[0.043]; Djorkaeff[0.043]; penalty[0.041]; semifinal[0.041]; elimination[0.041]; Mexico[0.041]; coach[0.040]; substitutions[0.040]; World[0.040]; World[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mFrance\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFrance national football team\u001b[39m <---> \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.245:0.241[\u001b[31m0.004\u001b[39m]; local()= 0.209:0.123[\u001b[31m0.086\u001b[39m]; log p(e|m)= -3.381:-0.163[\u001b[32m3.219\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.048]; France[0.048]; Lamouchi[0.045]; team[0.044]; Karembeu[0.043]; Deschamps[0.043]; Cup[0.042]; Djorkaeff[0.041]; Djorkaeff[0.041]; Djorkaeff[0.041]; European[0.041]; Thuram[0.041]; Nantes[0.041]; Desailly[0.041]; Juventus[0.040]; playing[0.040]; championship[0.040]; Bixente[0.040]; Zidane[0.040]; Zidane[0.040]; Laurent[0.040]; play[0.040]; finals[0.040]; finals[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance national football team\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.167:0.167[0]; log p(e|m)= -3.381:-3.381[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.049]; France[0.049]; matches[0.046]; Soccer[0.044]; Euro[0.043]; Euro[0.043]; Saturday[0.042]; Cup[0.042]; Cup[0.042]; international[0.042]; second[0.042]; Paris[0.042]; Djorkaeff[0.042]; penalty[0.040]; semifinal[0.040]; elimination[0.040]; Mexico[0.040]; coach[0.039]; World[0.039]; World[0.039]; time[0.039]; central[0.039]; having[0.039]; hosts[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Euro 96\u001b[39m ==> ENTITY: \u001b[32mUEFA Euro 1996\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.048]; Euro[0.044]; second[0.044]; international[0.044]; Soccer[0.044]; World[0.043]; World[0.043]; Cup[0.043]; Cup[0.043]; semifinal[0.041]; France[0.041]; France[0.041]; France[0.041]; goals[0.041]; Djorkaeff[0.041]; hosts[0.041]; Mexico[0.040]; Chelsea[0.040]; marred[0.040]; having[0.040]; penalty[0.040]; scored[0.040]; bookable[0.040]; friendly[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reynald Pedros\u001b[39m ==> ENTITY: \u001b[32mReynald Pedros\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.102:0.102[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYouri[0.043]; Nicolas[0.043]; Laurent[0.043]; Marcel[0.043]; Karembeu[0.043]; France[0.042]; France[0.042]; team[0.042]; Bixente[0.042]; Lamouchi[0.042]; Cup[0.042]; Franck[0.042]; finals[0.041]; Deschamps[0.041]; Desailly[0.041]; Lizarazu[0.041]; hosting[0.041]; Djorkaeff[0.041]; Djorkaeff[0.041]; Didier[0.041]; Florian[0.041]; Pavel[0.041]; Thuram[0.041]; Patrice[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Robert Pires\u001b[39m ==> ENTITY: \u001b[32mRobert Pirès\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLamouchi[0.047]; Djorkaeff[0.044]; Desailly[0.043]; Karembeu[0.043]; finals[0.042]; Didier[0.042]; Nicolas[0.042]; Cup[0.042]; Bixente[0.042]; Lizarazu[0.042]; Villa[0.041]; team[0.041]; Internazionale[0.041]; Pedros[0.041]; Enrique[0.041]; Zidane[0.041]; France[0.041]; France[0.041]; Thuram[0.041]; forward[0.041]; Ricardo[0.041]; Deschamps[0.041]; Laurent[0.040]; Youri[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chelsea ==> ENTITY: \u001b[32mChelsea F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.146:0.146[0]; log p(e|m)= -0.518:-0.518[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; Chelsea[0.045]; Cup[0.043]; play[0.042]; playing[0.042]; Mexican[0.042]; goals[0.042]; goals[0.042]; second[0.042]; second[0.042]; finals[0.042]; defender[0.042]; defender[0.042]; midfielder[0.041]; early[0.041]; Saturday[0.041]; joined[0.041]; injury[0.041]; Djorkaeff[0.041]; Djorkaeff[0.041]; Mexicans[0.040]; Juventus[0.040]; Ricardo[0.040]; substitutions[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alberto Garcia Aspe\u001b[39m ==> ENTITY: \u001b[32mAlberto García Aspe\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.045]; Enrique[0.045]; Rodrigo[0.044]; Luis[0.043]; Suarez[0.043]; Ricardo[0.042]; Raul[0.042]; Francisco[0.042]; Djorkaeff[0.042]; Ramon[0.042]; Lamouchi[0.042]; Pedros[0.042]; Joaquin[0.041]; del[0.041]; Claudio[0.041]; Bixente[0.041]; Nicolas[0.041]; Osvaldo[0.041]; Gomez[0.040]; Garcia[0.040]; Rios[0.040]; Karembeu[0.040]; Cuauhtemoc[0.040]; Palencia[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mRamon Ramirez\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRamón Ramírez (Dominican pitcher)\u001b[39m <---> \u001b[32mRamón Ramírez (footballer)\u001b[39m\t\nSCORES: global= 0.272:0.250[\u001b[31m0.021\u001b[39m]; local()= 0.105:0.146[\u001b[32m0.041\u001b[39m]; log p(e|m)= 0.000:-4.510[\u001b[31m4.510\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.044]; Mexico[0.044]; Francisco[0.044]; Enrique[0.044]; Alberto[0.043]; Pedros[0.043]; Osvaldo[0.042]; Karembeu[0.042]; Bixente[0.042]; Ricardo[0.042]; Rios[0.041]; Djorkaeff[0.041]; Joaquin[0.041]; Rodrigo[0.041]; Claudio[0.041]; Sanchez[0.041]; Suarez[0.041]; Garcia[0.041]; Garcia[0.041]; Raul[0.041]; Lizarazu[0.041]; del[0.040]; Florian[0.040]; Pavel[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBarcelona\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFC Barcelona\u001b[39m <---> \u001b[32mBarcelona\u001b[39m\t\nSCORES: global= 0.261:0.244[\u001b[31m0.018\u001b[39m]; local()= 0.153:0.157[\u001b[32m0.004\u001b[39m]; log p(e|m)= -1.709:-0.371[\u001b[32m1.338\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOsvaldo[0.047]; team[0.046]; Cup[0.045]; del[0.044]; Ricardo[0.044]; playing[0.044]; championship[0.043]; Djorkaeff[0.043]; Djorkaeff[0.043]; Mexican[0.043]; World[0.043]; second[0.043]; second[0.043]; Joaquin[0.043]; finals[0.043]; finals[0.043]; Juventus[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; Espanyol[0.042]; month[0.042]; \t\n\r [=============>................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 14s23ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1504/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 948testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.052:0.052[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.046]; Surrey[0.045]; matches[0.045]; Durham[0.043]; Somerset[0.043]; Somerset[0.043]; Somerset[0.043]; Warwickshire[0.043]; Leicestershire[0.043]; Leicestershire[0.043]; Leicestershire[0.043]; Championship[0.043]; Nottinghamshire[0.041]; Cricket[0.041]; play[0.041]; Kent[0.040]; Glamorgan[0.040]; championship[0.040]; Leicester[0.040]; Oval[0.040]; runs[0.039]; beat[0.039]; Chester[0.039]; Johnson[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Northamptonshire ==> ENTITY: \u001b[32mNorthamptonshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.152:0.152[0]; log p(e|m)= -1.207:-1.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.047]; Hampshire[0.046]; Middlesex[0.046]; Lancashire[0.045]; Worcestershire[0.045]; Gloucestershire[0.045]; Sussex[0.044]; Bristol[0.043]; Derbyshire[0.043]; Surrey[0.043]; Oval[0.043]; Ramprakash[0.042]; Chesterfield[0.041]; Drakes[0.041]; Adams[0.041]; Hove[0.041]; Gatting[0.040]; Smith[0.040]; Bicknell[0.039]; Russell[0.039]; London[0.038]; Lewis[0.038]; Portsmouth[0.037]; Pooley[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sussex ==> ENTITY: \u001b[32mSussex County Cricket Club\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.195:0.195[0]; log p(e|m)= -1.556:-1.556[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHampshire[0.044]; Warwickshire[0.044]; Middlesex[0.043]; Somerset[0.043]; Somerset[0.043]; Somerset[0.043]; Durham[0.042]; Surrey[0.042]; Leicestershire[0.042]; Leicestershire[0.042]; Leicestershire[0.042]; Lancashire[0.042]; Oval[0.042]; matches[0.042]; Gloucestershire[0.041]; Worcestershire[0.041]; Kent[0.041]; Glamorgan[0.041]; Northamptonshire[0.041]; Nottinghamshire[0.040]; Derbyshire[0.040]; Hove[0.040]; London[0.040]; Bristol[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tunbridge Wells ==> ENTITY: \u001b[32mRoyal Tunbridge Wells\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.126:0.126[0]; log p(e|m)= -0.260:-0.260[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHampshire[0.044]; Kent[0.044]; Cricket[0.043]; Somerset[0.043]; Somerset[0.043]; Somerset[0.043]; Oval[0.042]; Adams[0.042]; Middlesex[0.042]; Sussex[0.042]; Leicestershire[0.042]; Leicestershire[0.042]; Leicestershire[0.042]; Surrey[0.041]; Warwickshire[0.041]; matches[0.041]; Bristol[0.040]; Hove[0.040]; London[0.040]; London[0.040]; Barnett[0.040]; Morris[0.040]; Lancashire[0.040]; Durham[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leicestershire ==> ENTITY: \u001b[32mLeicestershire County Cricket Club\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.179:0.179[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; Leicestershire[0.044]; Leicestershire[0.044]; English[0.042]; English[0.042]; Ramprakash[0.042]; Lancashire[0.042]; matches[0.042]; Somerset[0.042]; Somerset[0.042]; Somerset[0.042]; Middlesex[0.042]; county[0.042]; Nottinghamshire[0.041]; Durham[0.041]; County[0.041]; Glamorgan[0.041]; Drakes[0.041]; Sussex[0.041]; Cricket[0.041]; Championship[0.041]; Surrey[0.041]; play[0.040]; innings[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Derbyshire ==> ENTITY: \u001b[32mDerbyshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.196:0.196[0]; log p(e|m)= -1.483:-1.483[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.046]; Hampshire[0.045]; Lancashire[0.045]; Worcestershire[0.044]; Sussex[0.044]; Gloucestershire[0.043]; Nottinghamshire[0.043]; Surrey[0.043]; Middlesex[0.043]; Durham[0.043]; Northamptonshire[0.042]; Kent[0.042]; Bristol[0.042]; Oval[0.041]; Drakes[0.041]; Hove[0.040]; Adams[0.040]; Ramprakash[0.040]; Chesterfield[0.040]; Gatting[0.039]; Smith[0.039]; Ratcliffe[0.039]; Morris[0.038]; Bicknell[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Somerset ==> ENTITY: \u001b[32mSomerset County Cricket Club\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.205:0.205[0]; log p(e|m)= -2.419:-2.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSomerset[0.045]; Somerset[0.045]; Warwickshire[0.044]; Sussex[0.043]; Leicestershire[0.043]; Leicestershire[0.043]; Leicestershire[0.043]; Lancashire[0.043]; Surrey[0.042]; matches[0.042]; Durham[0.042]; Glamorgan[0.042]; Nottinghamshire[0.042]; Kent[0.041]; Cricket[0.041]; Oval[0.041]; Drakes[0.041]; County[0.040]; county[0.040]; innings[0.040]; Hove[0.039]; runs[0.039]; scores[0.039]; play[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Warwickshire ==> ENTITY: \u001b[32mWarwickshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.189:0.189[0]; log p(e|m)= -1.109:-1.109[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHampshire[0.044]; Leicestershire[0.043]; Leicestershire[0.043]; Leicestershire[0.043]; Lancashire[0.043]; Durham[0.042]; Somerset[0.042]; Somerset[0.042]; Somerset[0.042]; Middlesex[0.042]; Worcestershire[0.042]; Glamorgan[0.042]; Gloucestershire[0.042]; Sussex[0.042]; Northamptonshire[0.041]; Nottinghamshire[0.041]; Derbyshire[0.041]; matches[0.041]; Cricket[0.041]; Kent[0.040]; Surrey[0.040]; innings[0.040]; Bristol[0.040]; scores[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hampshire ==> ENTITY: \u001b[32mHampshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.189:0.189[0]; log p(e|m)= -1.924:-1.924[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.045]; Lancashire[0.044]; Somerset[0.044]; Middlesex[0.043]; Glamorgan[0.043]; Gloucestershire[0.043]; Durham[0.043]; Worcestershire[0.043]; Nottinghamshire[0.043]; Sussex[0.042]; Derbyshire[0.042]; Northamptonshire[0.042]; Kent[0.042]; Ramprakash[0.042]; Surrey[0.042]; Bristol[0.041]; Oval[0.041]; Drakes[0.040]; Hove[0.039]; Gatting[0.039]; Adams[0.039]; Athey[0.039]; Chesterfield[0.038]; Smith[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kent ==> ENTITY: \u001b[32mKent County Cricket Club\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.202:0.202[0]; log p(e|m)= -2.513:-2.513[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; Middlesex[0.044]; Hampshire[0.043]; Somerset[0.043]; Somerset[0.043]; Somerset[0.043]; Sussex[0.043]; Surrey[0.043]; Leicestershire[0.042]; Leicestershire[0.042]; Leicestershire[0.042]; Lancashire[0.042]; Worcestershire[0.042]; Gloucestershire[0.042]; Durham[0.042]; Glamorgan[0.041]; Derbyshire[0.041]; Nottinghamshire[0.040]; Bristol[0.040]; Hove[0.040]; matches[0.040]; Cricket[0.040]; Ramprakash[0.039]; Drakes[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mChester-le-Street\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRiverside Ground\u001b[39m <---> \u001b[32mChester-le-Street\u001b[39m\t\nSCORES: global= 0.273:0.270[\u001b[31m0.003\u001b[39m]; local()= 0.185:0.166[\u001b[31m0.019\u001b[39m]; log p(e|m)= -1.452:-0.030[\u001b[32m1.422\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; Somerset[0.043]; Somerset[0.043]; Somerset[0.043]; Durham[0.043]; Warwickshire[0.043]; Leicestershire[0.043]; Leicestershire[0.043]; Leicestershire[0.043]; Hampshire[0.042]; Lancashire[0.042]; Oval[0.042]; county[0.042]; Sussex[0.041]; Middlesex[0.041]; Street[0.041]; Cricket[0.041]; Nottinghamshire[0.041]; County[0.040]; Kent[0.040]; Glamorgan[0.040]; Championship[0.040]; Surrey[0.040]; play[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Somerset ==> ENTITY: \u001b[32mSomerset County Cricket Club\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.211:0.211[0]; log p(e|m)= -2.419:-2.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSomerset[0.044]; Somerset[0.044]; Middlesex[0.044]; Warwickshire[0.044]; Sussex[0.043]; Leicestershire[0.043]; Leicestershire[0.043]; Leicestershire[0.043]; Lancashire[0.043]; Surrey[0.042]; matches[0.042]; Durham[0.042]; Glamorgan[0.042]; Nottinghamshire[0.041]; Kent[0.041]; Cricket[0.041]; Oval[0.040]; Drakes[0.040]; County[0.040]; county[0.040]; innings[0.039]; Hove[0.039]; runs[0.039]; scores[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lancashire ==> ENTITY: \u001b[32mLancashire County Cricket Club\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.189:0.189[0]; log p(e|m)= -1.732:-1.732[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWorcestershire[0.044]; Warwickshire[0.044]; Northamptonshire[0.043]; Hampshire[0.043]; Leicestershire[0.043]; Leicestershire[0.043]; Glamorgan[0.043]; Gloucestershire[0.043]; Somerset[0.042]; Somerset[0.042]; Somerset[0.042]; Middlesex[0.042]; Derbyshire[0.041]; Sussex[0.041]; Durham[0.041]; Nottinghamshire[0.041]; Surrey[0.041]; Bristol[0.041]; Kent[0.040]; Ramprakash[0.040]; Chesterfield[0.040]; Oval[0.039]; Hove[0.039]; Gatting[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = English ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.140:0.140[0]; log p(e|m)= -1.355:-1.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEnglish[0.047]; matches[0.043]; Warwickshire[0.043]; Leicestershire[0.043]; Leicestershire[0.043]; Leicestershire[0.043]; Oval[0.043]; Cricket[0.043]; Somerset[0.042]; Somerset[0.042]; Somerset[0.042]; Leicester[0.042]; Glamorgan[0.041]; Sussex[0.041]; Surrey[0.041]; Championship[0.041]; Durham[0.041]; Nottinghamshire[0.040]; Kent[0.040]; play[0.040]; London[0.040]; London[0.040]; Hove[0.039]; Chester[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bristol ==> ENTITY: \u001b[32mBristol\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.119:0.119[0]; log p(e|m)= -0.435:-0.435[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.045]; Lancashire[0.044]; Hampshire[0.044]; Oval[0.044]; Worcestershire[0.043]; Gloucestershire[0.043]; Portsmouth[0.042]; Nottinghamshire[0.042]; Hove[0.042]; Northamptonshire[0.042]; Sussex[0.042]; Chesterfield[0.041]; Middlesex[0.041]; Derbyshire[0.041]; Surrey[0.041]; Johnson[0.041]; Gatting[0.041]; Kent[0.041]; Ramprakash[0.041]; London[0.040]; Adams[0.040]; Russell[0.040]; Drakes[0.040]; Curran[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Surrey ==> ENTITY: \u001b[32mSurrey County Cricket Club\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.217:0.217[0]; log p(e|m)= -1.796:-1.796[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHampshire[0.044]; Warwickshire[0.044]; Middlesex[0.043]; Leicestershire[0.043]; Leicestershire[0.043]; Leicestershire[0.043]; Lancashire[0.043]; Somerset[0.042]; Somerset[0.042]; Somerset[0.042]; Sussex[0.042]; Gloucestershire[0.042]; Worcestershire[0.042]; Durham[0.042]; Northamptonshire[0.041]; Glamorgan[0.041]; Nottinghamshire[0.041]; Derbyshire[0.040]; Kent[0.040]; Oval[0.040]; matches[0.040]; Bristol[0.040]; Cricket[0.039]; Ramprakash[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Durham ==> ENTITY: \u001b[32mDurham County Cricket Club\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.221:0.221[0]; log p(e|m)= -2.397:-2.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.045]; Middlesex[0.044]; Hampshire[0.044]; Somerset[0.043]; Somerset[0.043]; Somerset[0.043]; Lancashire[0.043]; Sussex[0.043]; Worcestershire[0.043]; Glamorgan[0.042]; Nottinghamshire[0.042]; Leicestershire[0.042]; Leicestershire[0.042]; Leicestershire[0.042]; matches[0.042]; Derbyshire[0.042]; Cricket[0.040]; Surrey[0.040]; Chester[0.039]; scores[0.039]; Kent[0.039]; Ramprakash[0.039]; play[0.039]; Oval[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leicestershire ==> ENTITY: \u001b[32mLeicestershire County Cricket Club\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.179:0.179[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; Leicestershire[0.044]; Leicestershire[0.044]; English[0.042]; English[0.042]; Ramprakash[0.042]; Lancashire[0.042]; matches[0.042]; Somerset[0.042]; Somerset[0.042]; Somerset[0.042]; Middlesex[0.042]; county[0.042]; Nottinghamshire[0.041]; Durham[0.041]; County[0.041]; Glamorgan[0.041]; Drakes[0.041]; Sussex[0.041]; Cricket[0.041]; Championship[0.041]; Surrey[0.041]; play[0.040]; innings[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chesterfield ==> ENTITY: \u001b[32mChesterfield\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.129:0.129[0]; log p(e|m)= -0.939:-0.939[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBristol[0.045]; Portsmouth[0.044]; Hampshire[0.044]; Oval[0.044]; Lancashire[0.043]; Warwickshire[0.043]; Middlesex[0.043]; Derbyshire[0.043]; Nottinghamshire[0.042]; Durham[0.042]; Sussex[0.042]; Surrey[0.041]; Northamptonshire[0.041]; Glamorgan[0.041]; Worcestershire[0.041]; Kent[0.041]; Adams[0.041]; Gloucestershire[0.041]; Tunbridge[0.041]; Ramprakash[0.040]; Johnson[0.040]; Hove[0.039]; Wells[0.039]; London[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.063:0.063[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.047]; Surrey[0.046]; Hampshire[0.046]; Sussex[0.044]; Durham[0.044]; Gloucestershire[0.044]; Somerset[0.044]; Somerset[0.044]; Somerset[0.044]; Warwickshire[0.043]; Leicestershire[0.043]; Leicestershire[0.043]; Leicestershire[0.043]; Bristol[0.043]; Northamptonshire[0.043]; matches[0.043]; Lancashire[0.043]; Nottinghamshire[0.042]; Worcestershire[0.042]; Middlesex[0.042]; Derbyshire[0.042]; Hove[0.042]; Cricket[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portsmouth ==> ENTITY: \u001b[32mPortsmouth\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.755:-0.755[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBristol[0.045]; Chesterfield[0.043]; Lancashire[0.043]; Hampshire[0.043]; Worcestershire[0.043]; Northamptonshire[0.042]; Warwickshire[0.042]; Kent[0.042]; Durham[0.042]; Leicestershire[0.042]; Leicestershire[0.042]; Smith[0.041]; Russell[0.041]; Sussex[0.041]; Johnson[0.041]; Somerset[0.041]; Somerset[0.041]; Gloucestershire[0.041]; Surrey[0.041]; Chester[0.041]; London[0.041]; Derbyshire[0.040]; Street[0.040]; Nottinghamshire[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nottinghamshire ==> ENTITY: \u001b[32mNottinghamshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.197:0.197[0]; log p(e|m)= -1.106:-1.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; Hampshire[0.044]; Somerset[0.043]; Somerset[0.043]; Somerset[0.043]; Middlesex[0.043]; Leicestershire[0.043]; Leicestershire[0.043]; Leicestershire[0.043]; Lancashire[0.042]; Sussex[0.042]; Glamorgan[0.042]; Worcestershire[0.042]; matches[0.042]; Gloucestershire[0.042]; Durham[0.041]; Ramprakash[0.041]; Derbyshire[0.041]; Surrey[0.041]; Cricket[0.040]; Kent[0.040]; Oval[0.039]; Adams[0.039]; Gatting[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Somerset ==> ENTITY: \u001b[32mSomerset County Cricket Club\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.212:0.212[0]; log p(e|m)= -2.419:-2.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHampshire[0.044]; Somerset[0.044]; Somerset[0.044]; Middlesex[0.044]; Warwickshire[0.044]; Sussex[0.043]; Leicestershire[0.043]; Leicestershire[0.043]; Leicestershire[0.043]; Lancashire[0.042]; Surrey[0.042]; matches[0.042]; Durham[0.042]; Glamorgan[0.042]; Nottinghamshire[0.041]; Kent[0.041]; Cricket[0.041]; Oval[0.040]; Ramprakash[0.040]; Drakes[0.040]; County[0.040]; county[0.039]; innings[0.039]; Hove[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gloucestershire ==> ENTITY: \u001b[32mGloucestershire County Cricket Club\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.172:0.172[0]; log p(e|m)= -1.378:-1.378[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.046]; Middlesex[0.045]; Hampshire[0.045]; Worcestershire[0.044]; Sussex[0.044]; Lancashire[0.044]; Ramprakash[0.043]; Bristol[0.043]; Northamptonshire[0.043]; Derbyshire[0.043]; Nottinghamshire[0.043]; Surrey[0.042]; Kent[0.042]; Oval[0.042]; Chesterfield[0.040]; Adams[0.040]; Drakes[0.040]; Portsmouth[0.040]; Gatting[0.039]; Hove[0.039]; Smith[0.039]; Johnson[0.039]; Bicknell[0.039]; London[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = English County Championship\u001b[39m ==> ENTITY: \u001b[32mCounty Championship\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.046]; Somerset[0.044]; Somerset[0.044]; Somerset[0.044]; Leicestershire[0.044]; Leicestershire[0.044]; Leicestershire[0.044]; matches[0.044]; Glamorgan[0.044]; Durham[0.044]; Nottinghamshire[0.043]; Surrey[0.043]; Cricket[0.042]; Kent[0.041]; Oval[0.040]; play[0.039]; county[0.039]; runs[0.039]; innings[0.039]; championship[0.039]; Wells[0.039]; scores[0.039]; Johnson[0.038]; Chester[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leicester ==> ENTITY: \u001b[32mLeicester\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.263:-0.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; Leicestershire[0.044]; Leicestershire[0.044]; Leicestershire[0.044]; Championship[0.043]; Warwickshire[0.043]; Somerset[0.042]; Somerset[0.042]; Somerset[0.042]; Durham[0.042]; Oval[0.042]; Glamorgan[0.041]; Nottinghamshire[0.041]; Cricket[0.041]; Sussex[0.041]; county[0.041]; London[0.040]; London[0.040]; County[0.040]; Kent[0.040]; play[0.040]; Johnson[0.040]; beat[0.040]; Drakes[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hove ==> ENTITY: \u001b[32mHove\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.175:0.175[0]; log p(e|m)= -0.041:-0.041[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDurham[0.044]; Sussex[0.043]; Hampshire[0.043]; Somerset[0.043]; Somerset[0.043]; Somerset[0.043]; matches[0.043]; Oval[0.042]; Middlesex[0.042]; Warwickshire[0.042]; Leicestershire[0.042]; Leicestershire[0.042]; Leicestershire[0.042]; Nottinghamshire[0.042]; Surrey[0.041]; Lancashire[0.041]; Bristol[0.041]; Worcestershire[0.041]; Derbyshire[0.041]; Gloucestershire[0.040]; Kent[0.040]; Northamptonshire[0.040]; Glamorgan[0.040]; London[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Worcestershire ==> ENTITY: \u001b[32mWorcestershire County Cricket Club\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.172:0.172[0]; log p(e|m)= -0.916:-0.916[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.045]; Ramprakash[0.044]; Hampshire[0.044]; Lancashire[0.044]; Gloucestershire[0.044]; Middlesex[0.044]; Nottinghamshire[0.043]; Sussex[0.043]; Durham[0.043]; Surrey[0.043]; Northamptonshire[0.043]; Derbyshire[0.043]; Drakes[0.041]; Gatting[0.041]; Oval[0.041]; Kent[0.041]; Bristol[0.041]; Hove[0.040]; Chesterfield[0.039]; Bicknell[0.039]; Adams[0.038]; Wells[0.038]; Portsmouth[0.038]; Smith[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = The Oval\u001b[39m ==> ENTITY: \u001b[32mThe Oval\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.044]; Hampshire[0.044]; Surrey[0.043]; Warwickshire[0.043]; Middlesex[0.043]; Somerset[0.043]; Somerset[0.043]; Somerset[0.043]; Lancashire[0.042]; Sussex[0.042]; Gloucestershire[0.042]; Leicestershire[0.042]; Leicestershire[0.042]; Leicestershire[0.042]; Glamorgan[0.041]; Cricket[0.041]; Ramprakash[0.041]; Street[0.041]; Derbyshire[0.041]; Worcestershire[0.041]; Northamptonshire[0.041]; Nottinghamshire[0.040]; Hove[0.040]; innings[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Glamorgan ==> ENTITY: \u001b[32mGlamorgan County Cricket Club\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.193:0.193[0]; log p(e|m)= -0.790:-0.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; Leicestershire[0.044]; Leicestershire[0.044]; Leicestershire[0.044]; Somerset[0.043]; Somerset[0.043]; Somerset[0.043]; Hampshire[0.043]; Ramprakash[0.043]; Lancashire[0.042]; matches[0.042]; Middlesex[0.042]; innings[0.042]; Nottinghamshire[0.041]; Durham[0.041]; Sussex[0.041]; Oval[0.041]; Surrey[0.041]; runs[0.040]; Kent[0.040]; Cricket[0.040]; Drakes[0.040]; Championship[0.040]; Gatting[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Middlesex ==> ENTITY: \u001b[32mMiddlesex County Cricket Club\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.184:0.184[0]; log p(e|m)= -1.570:-1.570[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHampshire[0.044]; Warwickshire[0.044]; Surrey[0.044]; Sussex[0.044]; Somerset[0.043]; Somerset[0.043]; Durham[0.043]; Lancashire[0.042]; Leicestershire[0.042]; Leicestershire[0.042]; Glamorgan[0.042]; Kent[0.042]; Oval[0.042]; Gloucestershire[0.042]; Worcestershire[0.041]; Bristol[0.041]; Nottinghamshire[0.041]; Derbyshire[0.041]; Northamptonshire[0.041]; Hove[0.040]; Ramprakash[0.040]; London[0.039]; Chester[0.039]; Portsmouth[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leicestershire ==> ENTITY: \u001b[32mLeicestershire County Cricket Club\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.174:0.174[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; Leicestershire[0.044]; Leicestershire[0.044]; English[0.042]; English[0.042]; matches[0.042]; Somerset[0.042]; Somerset[0.042]; Somerset[0.042]; county[0.042]; Nottinghamshire[0.042]; Durham[0.041]; County[0.041]; Glamorgan[0.041]; Drakes[0.041]; Sussex[0.041]; Cricket[0.041]; Championship[0.041]; Surrey[0.041]; play[0.041]; innings[0.041]; Oval[0.040]; Kent[0.040]; runs[0.039]; \t\n\r [=============>................................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s999ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1536/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1036testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Azeris ==> ENTITY: \u001b[32mAzerbaijan\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.179:0.179[0]; log p(e|m)= -1.743:-1.743[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAzerbaijan[0.049]; Armenia[0.046]; Armenian[0.045]; Nagorno[0.045]; Nagorno[0.045]; region[0.045]; Armenians[0.045]; Azerbaijani[0.044]; Vafa[0.043]; Karabakh[0.042]; Karabakh[0.042]; enclave[0.042]; spoke[0.041]; hold[0.039]; flown[0.039]; said[0.039]; said[0.039]; held[0.039]; disputed[0.039]; disputed[0.039]; uneasy[0.039]; Representatives[0.038]; diplomatic[0.038]; home[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.094:0.094[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.054]; Europe[0.050]; region[0.048]; Finland[0.046]; declined[0.046]; Azerbaijan[0.046]; Russia[0.045]; Russia[0.045]; enclave[0.045]; place[0.045]; countries[0.045]; Group[0.045]; Nagorno[0.045]; Nagorno[0.045]; face[0.044]; face[0.044]; independence[0.044]; Armenia[0.044]; took[0.044]; conflict[0.044]; reported[0.044]; forces[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Azeris ==> ENTITY: \u001b[32mAzerbaijan\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.239:0.239[0]; log p(e|m)= -1.743:-1.743[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAzerbaijan[0.046]; Azerbaijan[0.046]; Armenia[0.043]; Armenian[0.043]; Armenian[0.043]; ethnic[0.042]; Nagorno[0.042]; Nagorno[0.042]; Nagorno[0.042]; Nagorno[0.042]; region[0.042]; region[0.042]; Armenians[0.042]; Azerbaijani[0.042]; Russia[0.041]; territory[0.041]; Azeri[0.041]; Vafa[0.041]; Karabakh[0.040]; Karabakh[0.040]; Karabakh[0.040]; Karabakh[0.040]; enclave[0.039]; enclave[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Azerbaijani ==> ENTITY: \u001b[32mAzerbaijan\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.211:0.211[0]; log p(e|m)= -1.041:-1.041[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAzerbaijan[0.047]; Armenian[0.046]; Azeris[0.044]; Azeris[0.044]; Armenia[0.044]; Azeri[0.043]; Nagorno[0.043]; Nagorno[0.043]; Nagorno[0.043]; region[0.042]; region[0.042]; ethnic[0.042]; territory[0.042]; Armenians[0.041]; Armenians[0.041]; Karabakh[0.040]; Karabakh[0.040]; Karabakh[0.040]; enclave[0.040]; diplomatic[0.039]; Zhirayr[0.039]; Representatives[0.038]; conflict[0.038]; spoke[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Azerbaijan ==> ENTITY: \u001b[32mAzerbaijan\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.190:0.190[0]; log p(e|m)= -0.337:-0.337[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArmenian[0.047]; Azerbaijani[0.046]; Armenia[0.045]; Azeri[0.044]; Nagorno[0.044]; Nagorno[0.044]; region[0.043]; territory[0.042]; held[0.042]; Karabakh[0.041]; Karabakh[0.041]; Armenians[0.041]; Armenians[0.041]; presidential[0.041]; Azeris[0.041]; Azeris[0.041]; hold[0.041]; diplomatic[0.040]; Zhirayr[0.040]; disputed[0.039]; disputed[0.039]; week[0.039]; Germany[0.039]; Germany[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Armenia ==> ENTITY: \u001b[32mArmenia\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.199:0.199[0]; log p(e|m)= -0.431:-0.431[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArmenian[0.046]; Azerbaijan[0.046]; Nagorno[0.045]; Nagorno[0.045]; Nagorno[0.045]; region[0.044]; territory[0.043]; Russia[0.042]; Russia[0.042]; Armenians[0.041]; independence[0.041]; Europe[0.041]; autonomy[0.041]; location[0.040]; Azeri[0.040]; Karabakh[0.040]; Karabakh[0.040]; Karabakh[0.040]; Azeris[0.040]; ethnic[0.040]; conflict[0.040]; Germany[0.040]; Germany[0.040]; enclave[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.048]; Finland[0.044]; Armenian[0.044]; Europe[0.044]; Armenia[0.043]; Germany[0.043]; Interfax[0.042]; Interfax[0.042]; Nagorno[0.042]; Minsk[0.042]; December[0.042]; held[0.042]; Karabakh[0.041]; enclave[0.041]; took[0.041]; countries[0.040]; peace[0.040]; Tuesday[0.040]; place[0.040]; independence[0.040]; accept[0.040]; autonomy[0.039]; led[0.039]; Group[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.084:0.084[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.049]; Europe[0.046]; territory[0.045]; region[0.044]; declined[0.042]; Azerbaijan[0.042]; Russia[0.042]; enclave[0.041]; enclave[0.041]; home[0.041]; place[0.041]; Group[0.041]; Nagorno[0.041]; Nagorno[0.041]; Nagorno[0.041]; face[0.040]; face[0.040]; independence[0.040]; Armenia[0.040]; took[0.040]; disputed[0.040]; conflict[0.040]; ethnic[0.040]; reported[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nagorno-Karabakh ==> ENTITY: \u001b[32mNagorno-Karabakh Republic\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.255:0.255[0]; log p(e|m)= -1.044:-1.044[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNagorno[0.046]; Nagorno[0.046]; Nagorno[0.046]; Azerbaijani[0.045]; Armenian[0.044]; Armenian[0.044]; Azerbaijan[0.044]; Azerbaijan[0.044]; Armenia[0.044]; Armenia[0.044]; Armenians[0.043]; Armenians[0.043]; territory[0.043]; region[0.043]; region[0.043]; Azeris[0.042]; Azeris[0.042]; Karabakh[0.042]; Karabakh[0.042]; Karabakh[0.042]; Karabakh[0.042]; independence[0.041]; conflict[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.103:0.103[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.051]; territory[0.045]; region[0.044]; region[0.044]; week[0.044]; Wednesday[0.042]; Azerbaijan[0.042]; earlier[0.042]; began[0.041]; Friday[0.041]; enclave[0.041]; home[0.041]; Nagorno[0.041]; Nagorno[0.041]; disputed[0.040]; disputed[0.040]; conflict[0.040]; ethnic[0.040]; represents[0.040]; said[0.040]; said[0.040]; Armenia[0.040]; claimed[0.039]; Bonn[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Amsterdam ==> ENTITY: \u001b[32mAmsterdam\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.064:-0.064[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurope[0.045]; Germany[0.045]; Germany[0.045]; Tuesday[0.042]; exact[0.042]; officials[0.042]; news[0.042]; location[0.042]; said[0.042]; said[0.042]; said[0.042]; subject[0.041]; departed[0.041]; place[0.041]; based[0.041]; Organisation[0.041]; Group[0.041]; countries[0.041]; reveal[0.041]; agency[0.041]; Russia[0.041]; Russia[0.041]; enclave[0.040]; prepared[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.095:0.095[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.051]; territory[0.046]; region[0.046]; Azerbaijan[0.043]; week[0.043]; enclave[0.042]; home[0.042]; Nagorno[0.042]; Nagorno[0.042]; disputed[0.041]; disputed[0.041]; Bonn[0.041]; said[0.041]; said[0.041]; Armenia[0.041]; diplomatic[0.040]; Friday[0.040]; Wednesday[0.040]; Armenian[0.040]; met[0.040]; bringing[0.040]; source[0.040]; source[0.040]; lasting[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.168:0.168[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.049]; Azerbaijan[0.046]; territory[0.046]; Finland[0.045]; Armenian[0.045]; Europe[0.044]; Interfax[0.044]; Interfax[0.044]; region[0.044]; Armenia[0.044]; Germany[0.044]; Germany[0.044]; Nagorno[0.043]; Nagorno[0.043]; enclave[0.042]; officials[0.042]; agency[0.042]; Minsk[0.042]; Azeri[0.041]; declined[0.041]; said[0.041]; said[0.041]; said[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Armenians ==> ENTITY: \u001b[32mArmenia\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.231:0.231[0]; log p(e|m)= -2.526:-2.526[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArmenian[0.046]; Armenian[0.046]; Armenia[0.043]; Armenia[0.043]; Nagorno[0.043]; Nagorno[0.043]; Nagorno[0.043]; Nagorno[0.043]; Azerbaijan[0.042]; Azerbaijan[0.042]; Azeris[0.042]; region[0.042]; region[0.042]; territory[0.041]; Russia[0.041]; Karabakh[0.041]; Karabakh[0.041]; Karabakh[0.041]; Karabakh[0.041]; ethnic[0.040]; Azerbaijani[0.040]; independence[0.039]; autonomy[0.039]; Azeri[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nagorno-Karabakh ==> ENTITY: \u001b[32mNagorno-Karabakh Republic\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.223:0.223[0]; log p(e|m)= -1.044:-1.044[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNagorno[0.046]; Azerbaijani[0.044]; Armenian[0.044]; Azerbaijan[0.044]; Armenia[0.044]; Armenians[0.043]; Armenians[0.043]; territory[0.043]; region[0.043]; region[0.043]; Azeris[0.042]; Azeris[0.042]; Karabakh[0.042]; Karabakh[0.042]; conflict[0.041]; disputed[0.040]; disputed[0.040]; enclave[0.040]; Azeri[0.040]; peace[0.039]; peace[0.039]; ethnic[0.039]; peaceful[0.039]; held[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Finland ==> ENTITY: \u001b[32mFinland\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.245:-0.245[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurope[0.045]; Germany[0.044]; Russia[0.044]; Russia[0.044]; Cooperation[0.044]; negotiations[0.043]; peace[0.042]; countries[0.042]; Armenia[0.042]; held[0.042]; December[0.041]; Nagorno[0.041]; Organisation[0.041]; Security[0.041]; took[0.041]; autonomy[0.041]; independence[0.040]; place[0.040]; reported[0.040]; talks[0.040]; talks[0.040]; discussions[0.040]; Interfax[0.040]; Interfax[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Azerbaijan ==> ENTITY: \u001b[32mAzerbaijan\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.209:0.209[0]; log p(e|m)= -0.337:-0.337[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArmenian[0.046]; Armenian[0.046]; Azeris[0.044]; Armenia[0.044]; Azeri[0.043]; Nagorno[0.043]; Nagorno[0.043]; Nagorno[0.043]; region[0.042]; Russia[0.042]; ethnic[0.042]; forces[0.041]; territory[0.041]; held[0.041]; Armenians[0.041]; Karabakh[0.040]; Karabakh[0.040]; Karabakh[0.040]; presidential[0.040]; Europe[0.040]; countries[0.040]; took[0.039]; Interfax[0.039]; Interfax[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Interfax\u001b[39m ==> ENTITY: \u001b[32mInterfax\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInterfax[0.048]; news[0.047]; Azerbaijan[0.045]; Minsk[0.044]; Russia[0.043]; Russia[0.043]; Armenian[0.043]; agency[0.043]; reported[0.042]; Armenia[0.042]; based[0.042]; officials[0.041]; Europe[0.041]; Tuesday[0.041]; countries[0.040]; December[0.040]; reveal[0.040]; details[0.040]; source[0.039]; Group[0.039]; Security[0.039]; Nagorno[0.039]; Nagorno[0.039]; held[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nagorno-Karabakh ==> ENTITY: \u001b[32mNagorno-Karabakh Republic\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.226:0.226[0]; log p(e|m)= -1.044:-1.044[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNagorno[0.045]; Nagorno[0.045]; Azerbaijani[0.044]; Armenian[0.043]; Armenian[0.043]; Azerbaijan[0.043]; Armenia[0.043]; Armenians[0.042]; territory[0.042]; region[0.042]; region[0.042]; Azeris[0.042]; Karabakh[0.041]; Karabakh[0.041]; Karabakh[0.041]; Interfax[0.041]; Interfax[0.041]; independence[0.040]; conflict[0.040]; disputed[0.040]; officials[0.040]; enclave[0.040]; enclave[0.040]; Russia[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Armenian ==> ENTITY: \u001b[32mArmenia\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.216:0.216[0]; log p(e|m)= -1.698:-1.698[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNagorno[0.048]; Nagorno[0.048]; Nagorno[0.048]; Armenia[0.045]; Azerbaijan[0.045]; Azerbaijan[0.045]; Armenians[0.045]; Armenians[0.045]; Azeris[0.044]; Azeris[0.044]; territory[0.044]; region[0.044]; region[0.044]; Karabakh[0.042]; Karabakh[0.042]; Karabakh[0.042]; Azerbaijani[0.042]; ethnic[0.041]; enclave[0.041]; Azeri[0.041]; conflict[0.040]; disputed[0.040]; disputed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Armenians ==> ENTITY: \u001b[32mArmenia\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.179:0.179[0]; log p(e|m)= -2.526:-2.526[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArmenian[0.049]; Armenia[0.046]; Nagorno[0.045]; Nagorno[0.045]; Azerbaijan[0.045]; Azeris[0.045]; region[0.044]; Azerbaijani[0.042]; Karabakh[0.041]; Vafa[0.041]; diplomatic[0.040]; disputed[0.040]; disputed[0.040]; enclave[0.040]; said[0.040]; said[0.040]; met[0.040]; spoke[0.040]; peace[0.039]; peace[0.039]; bringing[0.039]; Germany[0.039]; Germany[0.039]; Representatives[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bonn\u001b[39m ==> ENTITY: \u001b[32mBonn\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.119:0.119[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.048]; Germany[0.048]; Friday[0.045]; Wednesday[0.044]; week[0.044]; close[0.043]; earlier[0.043]; held[0.041]; Nagorno[0.041]; Nagorno[0.041]; met[0.041]; flown[0.041]; region[0.041]; discuss[0.041]; anonymity[0.040]; said[0.040]; said[0.040]; spoke[0.040]; ceasefire[0.040]; adviser[0.040]; home[0.040]; percent[0.040]; peace[0.040]; peace[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Interfax\u001b[39m ==> ENTITY: \u001b[32mInterfax\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInterfax[0.048]; news[0.047]; Azerbaijan[0.045]; Minsk[0.044]; Azeri[0.043]; Russia[0.043]; Russia[0.043]; Armenian[0.043]; agency[0.043]; reported[0.042]; Armenia[0.041]; based[0.041]; officials[0.041]; Europe[0.041]; Tuesday[0.041]; claimed[0.040]; countries[0.040]; December[0.040]; reveal[0.039]; details[0.039]; source[0.039]; Group[0.039]; Azeris[0.039]; Security[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nagorno-Karabakh ==> ENTITY: \u001b[32mNagorno-Karabakh Republic\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.220:0.220[0]; log p(e|m)= -1.044:-1.044[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNagorno[0.046]; Nagorno[0.046]; Armenian[0.044]; Azerbaijan[0.044]; Armenia[0.043]; Armenians[0.042]; territory[0.042]; region[0.042]; Azeris[0.042]; Karabakh[0.042]; Karabakh[0.042]; Karabakh[0.042]; Interfax[0.041]; Interfax[0.041]; independence[0.041]; conflict[0.040]; disputed[0.040]; officials[0.040]; enclave[0.040]; enclave[0.040]; Russia[0.040]; Russia[0.040]; Azeri[0.040]; autonomy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Azeri ==> ENTITY: \u001b[32mAzerbaijan\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.239:0.239[0]; log p(e|m)= -2.263:-2.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAzerbaijan[0.046]; Azerbaijan[0.046]; Azeris[0.043]; Armenia[0.043]; Armenia[0.043]; Azerbaijani[0.043]; Armenian[0.043]; Armenian[0.043]; Nagorno[0.042]; Nagorno[0.042]; Nagorno[0.042]; Nagorno[0.042]; region[0.042]; region[0.042]; Russia[0.041]; ethnic[0.041]; territory[0.041]; Vafa[0.041]; Armenians[0.040]; enclave[0.039]; enclave[0.039]; disputed[0.039]; disputed[0.039]; spoke[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Minsk Group\u001b[39m ==> ENTITY: \u001b[32mOSCE Minsk Group\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAzerbaijan[0.045]; Armenia[0.044]; Nagorno[0.044]; Russia[0.043]; Russia[0.043]; discussions[0.043]; Armenian[0.043]; Karabakh[0.043]; negotiations[0.042]; Finland[0.042]; Security[0.042]; talks[0.041]; talks[0.041]; Europe[0.041]; peace[0.041]; Germany[0.041]; Germany[0.041]; countries[0.041]; Cooperation[0.040]; Organisation[0.040]; held[0.040]; Interfax[0.040]; Interfax[0.040]; Tuesday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Armenia ==> ENTITY: \u001b[32mArmenia\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.190:0.190[0]; log p(e|m)= -0.431:-0.431[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArmenian[0.047]; Azerbaijan[0.046]; Nagorno[0.045]; Nagorno[0.045]; region[0.044]; territory[0.044]; Azerbaijani[0.042]; Armenians[0.042]; Armenians[0.042]; Azeri[0.041]; Karabakh[0.041]; Karabakh[0.041]; Azeris[0.041]; ethnic[0.041]; Zhirayr[0.041]; enclave[0.041]; disputed[0.040]; disputed[0.040]; Germany[0.040]; Germany[0.040]; home[0.039]; said[0.039]; said[0.039]; bringing[0.039]; \t\n\r [==============>...............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s850ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1563/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1064testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Northern Ireland ==> ENTITY: \u001b[32mNorthern Ireland national football team\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.129:0.129[0]; log p(e|m)= -2.207:-2.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.051]; Belfast[0.049]; qualifier[0.047]; Cup[0.045]; Cup[0.045]; soccer[0.045]; Soccer[0.044]; Qualifier[0.044]; European[0.044]; Northern[0.042]; beat[0.041]; Saturday[0.041]; Ukraine[0.040]; Ukraine[0.040]; World[0.040]; World[0.040]; group[0.040]; Rebrov[0.039]; minute[0.039]; Scorer[0.038]; Attendance[0.037]; halftime[0.037]; Beat[0.037]; Sergei[0.034]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belfast ==> ENTITY: \u001b[32mBelfast\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.069:0.069[0]; log p(e|m)= -0.131:-0.131[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.052]; Ireland[0.052]; soccer[0.049]; Cup[0.047]; Cup[0.047]; Soccer[0.047]; Saturday[0.046]; Qualifier[0.045]; qualifier[0.045]; group[0.043]; Northern[0.043]; Northern[0.043]; Attendance[0.042]; European[0.042]; beat[0.041]; halftime[0.041]; minute[0.041]; Scorer[0.041]; World[0.040]; World[0.040]; Beat[0.038]; Rebrov[0.038]; Sergei[0.037]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEuropean\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUEFA\u001b[39m <---> \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.244:0.240[\u001b[31m0.004\u001b[39m]; local()= 0.112:0.054[\u001b[31m0.058\u001b[39m]; log p(e|m)= -3.576:-0.256[\u001b[32m3.319\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.047]; Soccer[0.047]; Cup[0.045]; Cup[0.045]; qualifier[0.045]; Ireland[0.044]; Ireland[0.044]; Qualifier[0.044]; group[0.043]; Ukraine[0.042]; Ukraine[0.042]; Rebrov[0.041]; Northern[0.040]; Northern[0.040]; beat[0.040]; Saturday[0.040]; World[0.040]; World[0.040]; Belfast[0.040]; minute[0.040]; Attendance[0.039]; Scorer[0.038]; Beat[0.037]; halftime[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Cup ==> ENTITY: \u001b[32mFIFA World Cup\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.105:0.105[0]; log p(e|m)= -1.002:-1.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.049]; Ireland[0.048]; Ireland[0.048]; Soccer[0.047]; Cup[0.047]; qualifier[0.045]; European[0.044]; World[0.044]; Qualifier[0.044]; Ukraine[0.041]; Ukraine[0.041]; Saturday[0.040]; Belfast[0.040]; minute[0.040]; group[0.040]; beat[0.039]; Scorer[0.039]; Beat[0.039]; Rebrov[0.039]; halftime[0.038]; Northern[0.037]; Northern[0.037]; Attendance[0.037]; Sergei[0.035]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sergei Rebrov\u001b[39m ==> ENTITY: \u001b[32mSerhiy Rebrov\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.089:0.089[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUkraine[0.047]; Ukraine[0.047]; Soccer[0.045]; soccer[0.045]; European[0.045]; qualifier[0.043]; Qualifier[0.043]; Northern[0.042]; Northern[0.042]; Cup[0.042]; Cup[0.042]; halftime[0.042]; Attendance[0.041]; beat[0.041]; Scorer[0.041]; World[0.041]; World[0.041]; minute[0.040]; Belfast[0.040]; Beat[0.039]; Saturday[0.039]; Ireland[0.038]; Ireland[0.038]; group[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Northern Ireland ==> ENTITY: \u001b[32mNorthern Ireland national football team\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.129:0.129[0]; log p(e|m)= -2.207:-2.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.051]; Belfast[0.049]; qualifier[0.047]; Cup[0.045]; Cup[0.045]; soccer[0.045]; Soccer[0.044]; Qualifier[0.044]; European[0.044]; Northern[0.042]; beat[0.041]; Saturday[0.041]; Ukraine[0.040]; Ukraine[0.040]; World[0.040]; World[0.040]; group[0.040]; Rebrov[0.039]; minute[0.039]; Scorer[0.038]; Attendance[0.037]; halftime[0.037]; Beat[0.037]; Sergei[0.034]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ukraine ==> ENTITY: \u001b[32mUkraine national football team\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.146:0.146[0]; log p(e|m)= -3.101:-3.101[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUkraine[0.056]; soccer[0.049]; qualifier[0.048]; Soccer[0.047]; Qualifier[0.046]; Cup[0.046]; Cup[0.046]; European[0.046]; Sergei[0.044]; group[0.042]; Rebrov[0.042]; Beat[0.042]; Ireland[0.042]; Ireland[0.042]; Saturday[0.041]; beat[0.041]; halftime[0.041]; World[0.041]; World[0.041]; Scorer[0.041]; minute[0.040]; Attendance[0.039]; Belfast[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ukraine ==> ENTITY: \u001b[32mUkraine national football team\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.146:0.146[0]; log p(e|m)= -3.101:-3.101[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUkraine[0.056]; soccer[0.049]; qualifier[0.048]; Soccer[0.047]; Qualifier[0.046]; Cup[0.046]; Cup[0.046]; European[0.046]; Sergei[0.044]; group[0.042]; Rebrov[0.042]; Beat[0.042]; Ireland[0.042]; Ireland[0.042]; Saturday[0.041]; beat[0.041]; halftime[0.041]; World[0.041]; World[0.041]; Scorer[0.041]; minute[0.040]; Attendance[0.039]; Belfast[0.039]; \t\n\r [==============>...............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s874ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1572/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1009testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Midwest ==> ENTITY: \u001b[32mMidwestern United States\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.092:0.092[0]; log p(e|m)= -0.031:-0.031[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \narea[0.046]; second[0.043]; economic[0.043]; Chicago[0.043]; day[0.042]; industrial[0.042]; nearly[0.042]; expected[0.042]; straight[0.042]; Washington[0.042]; sector[0.041]; Glenn[0.041]; Friday[0.041]; financial[0.041]; billion[0.041]; managers[0.041]; July[0.041]; July[0.041]; twice[0.041]; increase[0.041]; August[0.040]; percent[0.040]; order[0.040]; separate[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fed\u001b[39m ==> ENTITY: \u001b[32mFederal Reserve System\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.121:0.121[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrate[0.045]; revised[0.044]; inflationary[0.042]; economy[0.042]; following[0.042]; June[0.042]; June[0.042]; University[0.042]; paying[0.042]; considers[0.042]; said[0.041]; said[0.041]; manufacturers[0.041]; July[0.041]; July[0.041]; August[0.041]; confidence[0.041]; providing[0.041]; sentiment[0.041]; Friday[0.041]; annual[0.041]; production[0.040]; durable[0.040]; durable[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = CNBC\u001b[39m ==> ENTITY: \u001b[32mCNBC\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.108:0.108[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntelevision[0.049]; interview[0.045]; network[0.045]; Friday[0.043]; Friday[0.043]; subscribers[0.043]; growth[0.043]; growth[0.043]; August[0.042]; signals[0.042]; cable[0.042]; July[0.041]; Sept[0.041]; economy[0.041]; expansion[0.041]; expansion[0.041]; paying[0.040]; consumer[0.040]; consumer[0.040]; Market[0.040]; spending[0.040]; added[0.039]; available[0.039]; strategy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = FOMC\u001b[39m ==> ENTITY: \u001b[32mFederal Open Market Committee\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npolicy[0.045]; Committee[0.043]; rate[0.043]; rate[0.043]; rate[0.043]; economic[0.043]; economic[0.043]; economic[0.043]; economic[0.043]; rates[0.043]; Federal[0.042]; Federal[0.042]; revised[0.042]; months[0.041]; adopted[0.041]; moderating[0.041]; spending[0.041]; Philadelphia[0.040]; Market[0.040]; growth[0.040]; growth[0.040]; economy[0.040]; strategy[0.039]; demand[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fed\u001b[39m ==> ENTITY: \u001b[32mFederal Reserve System\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npolicy[0.048]; Banks[0.046]; Federal[0.045]; Street[0.045]; rates[0.045]; Commerce[0.045]; month[0.044]; bank[0.044]; central[0.044]; report[0.044]; Treasury[0.043]; inflation[0.043]; spending[0.043]; Department[0.042]; economy[0.042]; economy[0.042]; Thursday[0.042]; Thursday[0.042]; June[0.042]; paid[0.042]; nation[0.042]; temporary[0.042]; growth[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Federal Reserve\u001b[39m ==> ENTITY: \u001b[32mFederal Reserve System\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBanks[0.046]; policy[0.046]; bank[0.046]; month[0.045]; central[0.044]; Treasury[0.044]; orders[0.044]; report[0.044]; nation[0.044]; months[0.043]; worries[0.043]; economy[0.043]; economy[0.043]; rates[0.043]; June[0.043]; Chicago[0.043]; temporary[0.042]; separate[0.042]; Barnett[0.042]; purchasing[0.042]; said[0.042]; said[0.042]; paid[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lawrence Lindsey\u001b[39m ==> ENTITY: \u001b[32mLawrence B. Lindsey\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUniversity[0.044]; policy[0.044]; Committee[0.043]; spending[0.042]; economic[0.042]; economic[0.042]; economic[0.042]; economic[0.042]; rate[0.042]; Federal[0.042]; Federal[0.042]; economy[0.042]; August[0.041]; growth[0.041]; growth[0.041]; July[0.041]; July[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; rates[0.041]; Michigan[0.041]; Reserve[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Federal Open Market Committee\u001b[39m ==> ENTITY: \u001b[32mFederal Open Market Committee\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.209:0.209[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npolicy[0.046]; rate[0.044]; rate[0.044]; rate[0.044]; economic[0.043]; economic[0.043]; economic[0.043]; economic[0.043]; rates[0.043]; Federal[0.042]; revised[0.042]; months[0.042]; adopted[0.041]; moderating[0.041]; spending[0.041]; Philadelphia[0.041]; growth[0.040]; growth[0.040]; economy[0.040]; economy[0.040]; strategy[0.040]; demand[0.040]; bank[0.040]; Governor[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Treasury ==> ENTITY: \u001b[32mUnited States Department of the Treasury\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.133:0.133[0]; log p(e|m)= -1.013:-1.013[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDepartment[0.045]; policy[0.044]; Federal[0.044]; economy[0.043]; month[0.043]; month[0.043]; bank[0.042]; Commerce[0.042]; Banks[0.042]; rate[0.042]; central[0.042]; consumer[0.042]; rates[0.041]; spending[0.041]; spending[0.041]; spending[0.041]; bond[0.041]; report[0.041]; raise[0.040]; lower[0.040]; growth[0.040]; yield[0.040]; sharply[0.040]; rose[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.081:0.081[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \neconomic[0.044]; Department[0.044]; nearly[0.043]; Glenn[0.043]; said[0.043]; said[0.043]; second[0.043]; July[0.042]; July[0.042]; straight[0.042]; Somerville[0.042]; order[0.042]; coming[0.041]; Commerce[0.041]; August[0.041]; Friday[0.041]; months[0.041]; financial[0.041]; twice[0.041]; climbed[0.040]; orders[0.040]; orders[0.040]; day[0.040]; sign[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.082:0.082[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \neconomy[0.045]; area[0.044]; area[0.044]; policy[0.044]; nearly[0.044]; businesses[0.043]; production[0.043]; late[0.042]; financial[0.041]; said[0.041]; said[0.041]; said[0.041]; sector[0.041]; region[0.041]; Midwest[0.041]; Midwest[0.041]; June[0.041]; August[0.040]; August[0.040]; Friday[0.040]; wanted[0.040]; Federal[0.040]; Commerce[0.040]; temporary[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = FOMC\u001b[39m ==> ENTITY: \u001b[32mFederal Open Market Committee\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npolicy[0.046]; Committee[0.044]; rate[0.044]; economic[0.043]; economic[0.043]; economic[0.043]; economic[0.043]; rates[0.043]; Federal[0.043]; Federal[0.043]; months[0.042]; adopted[0.041]; moderating[0.041]; spending[0.041]; Philadelphia[0.041]; Market[0.040]; growth[0.040]; growth[0.040]; economy[0.040]; economy[0.040]; strategy[0.040]; demand[0.040]; bank[0.040]; Governor[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mDow Jones\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mDow Jones & Company\u001b[39m <---> \u001b[32mDow Jones Industrial Average\u001b[39m\t\nSCORES: global= 0.267:0.264[\u001b[31m0.004\u001b[39m]; local()= 0.102:0.229[\u001b[32m0.127\u001b[39m]; log p(e|m)= 0.000:-0.423[\u001b[31m0.423\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndecline[0.044]; gained[0.044]; stock[0.044]; Banks[0.042]; sharply[0.042]; month[0.042]; month[0.042]; lower[0.042]; reports[0.042]; dropping[0.042]; rose[0.041]; analysts[0.041]; point[0.041]; point[0.041]; point[0.041]; rate[0.041]; rate[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; price[0.041]; late[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jacksonville ==> ENTITY: \u001b[32mJacksonville, Florida\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.562:-0.562[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \narea[0.046]; Lynn[0.045]; Street[0.043]; late[0.042]; late[0.042]; July[0.042]; Department[0.042]; June[0.042]; Thursday[0.042]; Thursday[0.042]; Banks[0.042]; second[0.041]; month[0.041]; driving[0.041]; nation[0.041]; businesses[0.041]; August[0.041]; moving[0.041]; manufacturing[0.041]; manufacturing[0.041]; rates[0.040]; running[0.040]; showing[0.040]; Commerce[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = University of Michigan\u001b[39m ==> ENTITY: \u001b[32mUniversity of Michigan\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.038:0.038[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.048]; economic[0.047]; rate[0.046]; Federal[0.046]; percent[0.046]; percent[0.046]; percent[0.046]; percent[0.046]; considers[0.046]; annual[0.045]; Friday[0.045]; Friday[0.045]; consumer[0.045]; consumer[0.045]; consumer[0.045]; expansion[0.045]; expansion[0.045]; index[0.045]; added[0.044]; spending[0.044]; rose[0.044]; rose[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Federal Reserve\u001b[39m ==> ENTITY: \u001b[32mFederal Reserve System\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npolicy[0.047]; Federal[0.046]; Committee[0.046]; economic[0.044]; economic[0.044]; economic[0.044]; economic[0.044]; rate[0.044]; economy[0.044]; rates[0.044]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; July[0.043]; July[0.043]; spending[0.043]; August[0.043]; added[0.042]; Market[0.042]; Lawrence[0.042]; help[0.042]; adopted[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Commerce Department\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Commerce\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.109:0.109[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \neconomic[0.047]; report[0.046]; financial[0.044]; months[0.044]; manufacturing[0.044]; manufacturing[0.044]; manufacturing[0.044]; manufacturing[0.044]; reports[0.044]; reports[0.044]; Midwest[0.044]; sector[0.043]; adjusted[0.043]; Washington[0.043]; orders[0.043]; orders[0.043]; orders[0.043]; July[0.043]; July[0.043]; July[0.043]; August[0.043]; August[0.043]; backlogs[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Commerce Department\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Commerce\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nreport[0.047]; January[0.045]; Treasury[0.045]; months[0.045]; month[0.044]; month[0.044]; reports[0.044]; adjusted[0.044]; July[0.043]; July[0.043]; June[0.043]; June[0.043]; spending[0.043]; spending[0.043]; spending[0.043]; said[0.043]; showed[0.043]; Thursday[0.043]; Thursday[0.043]; rates[0.042]; growth[0.042]; revised[0.042]; industrial[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Midwest ==> ENTITY: \u001b[32mMidwestern United States\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.031:-0.031[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nregion[0.047]; area[0.045]; area[0.045]; nearly[0.042]; Chicago[0.042]; industrial[0.042]; expected[0.042]; businesses[0.042]; economy[0.042]; economy[0.042]; sector[0.041]; group[0.041]; Reserve[0.041]; late[0.041]; billion[0.041]; policy[0.041]; indicates[0.041]; manufacturing[0.041]; manufacturing[0.041]; manufacturing[0.041]; manufacturing[0.041]; point[0.040]; twice[0.040]; increase[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lindsey\u001b[39m ==> ENTITY: \u001b[32mLawrence B. Lindsey\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUniversity[0.044]; policy[0.044]; spending[0.043]; economic[0.042]; economic[0.042]; rate[0.042]; Federal[0.042]; economy[0.042]; economy[0.042]; August[0.042]; growth[0.041]; July[0.041]; July[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; Lawrence[0.041]; rates[0.041]; Michigan[0.041]; Reserve[0.041]; interview[0.041]; Lindsey[0.041]; subscribers[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Philadelphia ==> ENTITY: \u001b[32mPhiladelphia\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.074:0.074[0]; log p(e|m)= -0.211:-0.211[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJuly[0.043]; month[0.043]; nation[0.043]; economic[0.042]; economic[0.042]; economic[0.042]; second[0.042]; overall[0.042]; gained[0.042]; months[0.042]; large[0.042]; June[0.042]; June[0.042]; January[0.042]; Market[0.042]; policy[0.041]; forecasting[0.041]; half[0.041]; half[0.041]; Joel[0.041]; gain[0.041]; Federal[0.040]; Union[0.040]; written[0.040]; \t\n\r [==============>...............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s781ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1594/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1100testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Markus Schopp\u001b[39m ==> ENTITY: \u001b[32mMarkus Schopp\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.047]; team[0.043]; team[0.043]; team[0.043]; players[0.042]; Dietmar[0.042]; Dietmar[0.042]; Rapid[0.042]; game[0.042]; Austrian[0.042]; Austrian[0.042]; Dieter[0.042]; Wolfgang[0.042]; coach[0.042]; Sk[0.041]; Goram[0.041]; Scotland[0.041]; Andreas[0.041]; Andreas[0.041]; Andreas[0.041]; play[0.040]; Anton[0.040]; Anton[0.040]; McCoist[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Peter Schoettel\u001b[39m ==> ENTITY: \u001b[32mPeter Schöttel\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.045]; players[0.044]; Rapid[0.043]; Dietmar[0.043]; Dietmar[0.043]; game[0.043]; Sk[0.042]; team[0.042]; team[0.042]; team[0.042]; Markus[0.042]; Austrian[0.041]; Austrian[0.041]; Andreas[0.041]; Andreas[0.041]; Andreas[0.041]; won[0.041]; coach[0.041]; Dieter[0.041]; play[0.040]; Anton[0.040]; Anton[0.040]; Michael[0.040]; Wolfgang[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Markus Schopp\u001b[39m ==> ENTITY: \u001b[32mMarkus Schopp\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.128:0.128[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.047]; match[0.044]; Sc[0.043]; Freiburg[0.042]; referee[0.042]; players[0.042]; Dietmar[0.042]; qualifier[0.042]; Rapid[0.042]; Austrian[0.042]; Austrian[0.042]; coach[0.042]; Sk[0.041]; captain[0.041]; Scotland[0.041]; Scotland[0.041]; Scotland[0.041]; Scotland[0.041]; Andreas[0.041]; play[0.040]; goalless[0.040]; goalkeeper[0.040]; ball[0.040]; Everton[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Craig Burley\u001b[39m ==> ENTITY: \u001b[32mCraig Burley\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFerguson[0.046]; Scotland[0.045]; Goram[0.043]; McCoist[0.043]; players[0.043]; team[0.042]; team[0.042]; team[0.042]; Dietmar[0.042]; Calderwood[0.042]; McAllister[0.042]; pitch[0.042]; Michael[0.041]; Austria[0.041]; Gordon[0.041]; Gary[0.041]; started[0.041]; McCall[0.041]; game[0.041]; Andreas[0.040]; Andreas[0.040]; Andreas[0.040]; Colin[0.040]; Colin[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Thomas Boyd ==> ENTITY: \u001b[32mTom Boyd (footballer)\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.134:0.134[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThomas[0.045]; Scotland[0.044]; John[0.044]; Gary[0.043]; Herbert[0.042]; players[0.042]; Craig[0.042]; won[0.042]; McCall[0.042]; game[0.042]; Colin[0.042]; Colin[0.042]; Peter[0.041]; Scots[0.041]; McCoist[0.041]; Duncan[0.041]; Michael[0.041]; McAllister[0.041]; team[0.041]; team[0.041]; Gordon[0.041]; Andrew[0.040]; lacked[0.040]; Goram[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Michael Konsel\u001b[39m ==> ENTITY: \u001b[32mMichael Konsel\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.132:0.132[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.047]; Austrian[0.044]; Austrian[0.044]; Belarus[0.043]; match[0.043]; Sc[0.043]; Freiburg[0.042]; team[0.042]; Dietmar[0.042]; Rapid[0.042]; players[0.042]; players[0.042]; Sk[0.041]; game[0.041]; repeat[0.041]; goalkeeper[0.041]; Andreas[0.040]; Markus[0.040]; draw[0.040]; coach[0.040]; pitch[0.040]; play[0.040]; Everton[0.040]; ball[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAustrian\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustria\u001b[39m <---> \u001b[32mAustria national football team\u001b[39m\t\nSCORES: global= 0.252:0.245[\u001b[31m0.007\u001b[39m]; local()= 0.157:0.159[\u001b[32m0.001\u001b[39m]; log p(e|m)= -0.406:-3.817[\u001b[31m3.410\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.048]; players[0.044]; players[0.044]; Austrian[0.043]; Austrian[0.043]; team[0.043]; team[0.043]; team[0.043]; Scotland[0.042]; Scotland[0.042]; Herbert[0.041]; play[0.041]; Markus[0.041]; Markus[0.041]; won[0.041]; Wolfgang[0.040]; arguing[0.040]; Michael[0.040]; Michael[0.040]; Andreas[0.040]; Andreas[0.040]; game[0.040]; draw[0.040]; Scots[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Stephan Marasek\u001b[39m ==> ENTITY: \u001b[32mStephan Marasek\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.113:0.113[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayers[0.048]; captain[0.045]; Soccer[0.045]; Austrian[0.045]; Sc[0.045]; Austria[0.045]; Austria[0.045]; Austria[0.045]; Freiburg[0.044]; Gary[0.044]; Steve[0.044]; Markus[0.044]; qualifier[0.043]; Qualifier[0.043]; goalkeeper[0.042]; Scotland[0.042]; Scotland[0.042]; Scotland[0.042]; Scotland[0.042]; Scotland[0.042]; Vienna[0.042]; Michael[0.042]; ball[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Anton Pfeffer\u001b[39m ==> ENTITY: \u001b[32mAnton Pfeffer\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.045]; won[0.043]; players[0.042]; Austrian[0.042]; Austrian[0.042]; Dietmar[0.042]; Dietmar[0.042]; game[0.042]; Wolfgang[0.042]; team[0.042]; team[0.042]; team[0.042]; Andreas[0.042]; Andreas[0.042]; Andreas[0.042]; play[0.042]; Sk[0.042]; draw[0.041]; Markus[0.041]; Anton[0.040]; coach[0.040]; points[0.040]; Dieter[0.040]; pitch[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wolfgang Feiersinger\u001b[39m ==> ENTITY: \u001b[32mWolfgang Feiersinger\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.047]; players[0.043]; team[0.043]; team[0.043]; team[0.043]; Austrian[0.042]; Austrian[0.042]; won[0.042]; game[0.042]; play[0.042]; Andreas[0.042]; Andreas[0.042]; Andreas[0.042]; Markus[0.042]; started[0.041]; Dietmar[0.041]; Dietmar[0.041]; draw[0.041]; Sk[0.040]; Anton[0.040]; Anton[0.040]; Rapid[0.040]; Scotland[0.040]; minutes[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Colin Calderwood\u001b[39m ==> ENTITY: \u001b[32mColin Calderwood\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.045]; team[0.045]; Burley[0.044]; McAllister[0.043]; McCall[0.043]; players[0.043]; Craig[0.043]; Ferguson[0.043]; Gary[0.043]; Colin[0.042]; Stuart[0.042]; Andrew[0.042]; Michael[0.041]; Scotland[0.041]; Duncan[0.040]; game[0.040]; Dietmar[0.040]; coach[0.040]; won[0.040]; Boyd[0.040]; Hendry[0.040]; Collins[0.040]; Herbert[0.040]; pitch[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Colin Hendry\u001b[39m ==> ENTITY: \u001b[32mColin Hendry\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.045]; team[0.045]; Craig[0.044]; Colin[0.043]; Stuart[0.043]; Gary[0.043]; Goram[0.043]; McAllister[0.042]; Scotland[0.042]; players[0.042]; Burley[0.042]; Ferguson[0.042]; Michael[0.041]; won[0.041]; John[0.041]; Gordon[0.041]; Andrew[0.040]; Dietmar[0.040]; Duncan[0.040]; Calderwood[0.040]; pitch[0.040]; Boyd[0.040]; lacked[0.040]; coach[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mScotland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mScotland national football team\u001b[39m <---> \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.245:0.239[\u001b[31m0.006\u001b[39m]; local()= 0.165:0.138[\u001b[31m0.027\u001b[39m]; log p(e|m)= -2.703:-0.229[\u001b[32m2.474\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.049]; Scottish[0.045]; Scottish[0.045]; match[0.045]; Scots[0.044]; team[0.043]; players[0.043]; players[0.043]; McAllister[0.042]; Ferguson[0.041]; Ferguson[0.041]; captain[0.041]; play[0.041]; game[0.041]; Craig[0.041]; coach[0.041]; Duncan[0.040]; Gary[0.040]; draw[0.039]; Michael[0.039]; Brown[0.039]; pitch[0.039]; Austria[0.039]; Everton[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scotland ==> ENTITY: \u001b[32mScotland national football team\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.164:0.164[0]; log p(e|m)= -2.703:-2.703[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.050]; Scotland[0.050]; Scotland[0.050]; Scottish[0.046]; match[0.045]; Soccer[0.042]; Cup[0.042]; Cup[0.042]; Ferguson[0.041]; qualifier[0.041]; Duncan[0.040]; Saturday[0.040]; Belarus[0.040]; Qualifier[0.040]; World[0.040]; World[0.040]; draw[0.040]; Steve[0.039]; opening[0.039]; Austria[0.039]; Austria[0.039]; Austria[0.039]; group[0.039]; group[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dietmar Kuehbauer\u001b[39m ==> ENTITY: \u001b[32mDietmar Kühbauer\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.045]; Dieter[0.044]; team[0.043]; team[0.043]; team[0.043]; coach[0.042]; coach[0.042]; Dietmar[0.042]; Markus[0.041]; Markus[0.041]; players[0.041]; players[0.041]; game[0.041]; Sk[0.041]; Andreas[0.041]; Andreas[0.041]; Andreas[0.041]; Andreas[0.041]; Austrian[0.041]; Austrian[0.041]; Austrian[0.041]; Stephan[0.041]; Wolfgang[0.040]; Rapid[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Stephan Marasek\u001b[39m ==> ENTITY: \u001b[32mStephan Marasek\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncoach[0.047]; players[0.045]; Austrian[0.043]; Austrian[0.043]; team[0.042]; team[0.042]; team[0.042]; Austria[0.042]; Gary[0.042]; Herbert[0.042]; Dietmar[0.042]; Dietmar[0.042]; Markus[0.042]; Andreas[0.041]; Andreas[0.041]; Andreas[0.041]; Sk[0.041]; impressive[0.040]; John[0.040]; Colin[0.040]; Colin[0.040]; Dieter[0.040]; game[0.040]; play[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dieter Ramusch\u001b[39m ==> ENTITY: \u001b[32mDieter Ramusch\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustrian[0.044]; Austrian[0.044]; coach[0.044]; Wolfgang[0.044]; Dietmar[0.044]; Dietmar[0.044]; players[0.043]; Markus[0.043]; Andreas[0.043]; Andreas[0.043]; Andreas[0.043]; Austria[0.042]; Gary[0.041]; team[0.041]; team[0.041]; team[0.041]; game[0.040]; play[0.040]; Sk[0.039]; Anton[0.039]; Anton[0.039]; John[0.039]; Stephan[0.039]; Peter[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Everton ==> ENTITY: \u001b[32mEverton F.C.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.046]; area[0.044]; match[0.044]; Michael[0.043]; Gary[0.043]; play[0.043]; players[0.042]; close[0.042]; close[0.042]; draw[0.042]; draw[0.042]; group[0.042]; Craig[0.041]; captain[0.041]; Ferguson[0.041]; Ferguson[0.041]; ball[0.041]; opening[0.041]; advantage[0.040]; goalkeeper[0.040]; Scotland[0.040]; Scotland[0.040]; Scotland[0.040]; diving[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Duncan Ferguson\u001b[39m ==> ENTITY: \u001b[32mDuncan Ferguson\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCraig[0.046]; Gary[0.045]; Colin[0.045]; Colin[0.045]; Andrew[0.044]; team[0.044]; team[0.044]; Michael[0.044]; Burley[0.044]; Scotland[0.044]; Calderwood[0.044]; McCall[0.044]; Goram[0.043]; Stuart[0.043]; John[0.043]; McAllister[0.043]; Gordon[0.043]; McCoist[0.043]; Boyd[0.043]; Collins[0.042]; Dietmar[0.042]; Herbert[0.042]; Peter[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Stuart McCall\u001b[39m ==> ENTITY: \u001b[32mStuart McCall\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.044]; team[0.044]; Burley[0.044]; Ferguson[0.043]; McAllister[0.043]; Craig[0.043]; Scotland[0.043]; Gary[0.043]; Calderwood[0.042]; Gordon[0.042]; Michael[0.042]; Collins[0.041]; Colin[0.041]; Colin[0.041]; Andrew[0.041]; McCoist[0.041]; Duncan[0.041]; Dietmar[0.040]; Boyd[0.040]; lacked[0.040]; John[0.040]; Goram[0.040]; Herbert[0.039]; won[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gary McAllister\u001b[39m ==> ENTITY: \u001b[32mGary McAllister\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.043]; team[0.043]; Craig[0.043]; Ferguson[0.043]; Ferguson[0.043]; players[0.043]; players[0.043]; Scotland[0.042]; Scotland[0.042]; Sc[0.042]; pitch[0.042]; game[0.041]; play[0.041]; Michael[0.041]; Freiburg[0.041]; draw[0.041]; minute[0.041]; minute[0.041]; minute[0.041]; Dietmar[0.041]; penalty[0.040]; advantage[0.040]; minutes[0.040]; saw[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austrian ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.406:-0.406[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.049]; Austrian[0.044]; Scotland[0.043]; players[0.042]; Thomas[0.042]; Thomas[0.042]; Herbert[0.042]; team[0.042]; team[0.042]; team[0.042]; Markus[0.041]; won[0.041]; Wolfgang[0.041]; arguing[0.041]; Michael[0.041]; Andreas[0.041]; Andreas[0.041]; Andreas[0.041]; draw[0.041]; Herzog[0.041]; Scots[0.041]; performance[0.040]; play[0.040]; pitch[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = SC Freiburg\u001b[39m ==> ENTITY: \u001b[32mSC Freiburg\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.045]; match[0.045]; Stephan[0.045]; draw[0.045]; players[0.044]; Markus[0.043]; Andreas[0.042]; advantage[0.041]; captain[0.041]; group[0.041]; group[0.041]; qualifier[0.041]; Saturday[0.041]; came[0.041]; Steve[0.041]; Qualifier[0.041]; penalty[0.041]; taking[0.041]; saw[0.040]; chance[0.040]; scramble[0.040]; opening[0.040]; goalkeeper[0.040]; fall[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andreas Ogris\u001b[39m ==> ENTITY: \u001b[32mAndreas Ogris\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.047]; Austrian[0.043]; Austrian[0.043]; players[0.043]; Scotland[0.043]; coach[0.042]; Dietmar[0.042]; Dietmar[0.042]; Andreas[0.042]; Andreas[0.042]; Wolfgang[0.042]; game[0.041]; Dieter[0.041]; Sk[0.041]; McCoist[0.041]; Markus[0.041]; Calderwood[0.041]; pitch[0.041]; Goram[0.041]; team[0.040]; team[0.040]; team[0.040]; performance[0.040]; Burley[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dietmar Kuehbauer\u001b[39m ==> ENTITY: \u001b[32mDietmar Kühbauer\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.046]; Dieter[0.044]; team[0.044]; team[0.044]; team[0.044]; coach[0.043]; Dietmar[0.042]; Markus[0.042]; players[0.042]; game[0.042]; Sk[0.041]; Andreas[0.041]; Andreas[0.041]; Andreas[0.041]; Austrian[0.041]; Austrian[0.041]; Stephan[0.041]; Wolfgang[0.041]; Rapid[0.040]; won[0.040]; Anton[0.040]; Anton[0.040]; really[0.039]; started[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria national football team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.178:0.178[0]; log p(e|m)= -2.937:-2.937[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.050]; Austria[0.050]; Belarus[0.046]; Cup[0.045]; Cup[0.045]; qualifier[0.045]; World[0.045]; World[0.045]; Scotland[0.044]; Scotland[0.044]; Scotland[0.044]; Qualifier[0.044]; match[0.043]; Soccer[0.042]; draw[0.042]; Vienna[0.042]; went[0.042]; dominated[0.041]; Saturday[0.040]; Scottish[0.040]; thrashed[0.040]; came[0.040]; opening[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alistair McCoist\u001b[39m ==> ENTITY: \u001b[32mAlly McCoist\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCraig[0.047]; team[0.046]; team[0.046]; Goram[0.045]; Gary[0.045]; Scotland[0.045]; Michael[0.045]; Colin[0.044]; Colin[0.044]; Stuart[0.044]; Gordon[0.043]; Burley[0.043]; John[0.043]; Ferguson[0.043]; Calderwood[0.043]; McAllister[0.042]; Dietmar[0.042]; McCall[0.042]; Andrew[0.042]; Boyd[0.042]; Peter[0.041]; coach[0.041]; won[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gordon Durie\u001b[39m ==> ENTITY: \u001b[32mGordon Durie\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.078:0.078[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMcCoist[0.045]; Burley[0.045]; team[0.044]; Scotland[0.043]; Goram[0.043]; Ferguson[0.043]; Colin[0.043]; Colin[0.043]; Gary[0.042]; Dietmar[0.042]; Calderwood[0.042]; Craig[0.042]; Stuart[0.041]; coach[0.041]; Andrew[0.041]; Duncan[0.041]; McCall[0.040]; Andreas[0.040]; Andreas[0.040]; Andreas[0.040]; John[0.040]; Markus[0.040]; Hendry[0.040]; Michael[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scotland ==> ENTITY: \u001b[32mScotland national football team\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.159:0.159[0]; log p(e|m)= -2.703:-2.703[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.050]; Scotland[0.050]; Scottish[0.046]; match[0.046]; Soccer[0.043]; Cup[0.042]; Cup[0.042]; qualifier[0.041]; Saturday[0.041]; Belarus[0.041]; Qualifier[0.040]; World[0.040]; World[0.040]; draw[0.040]; Steve[0.040]; opening[0.040]; Austria[0.040]; Austria[0.040]; Austria[0.040]; group[0.040]; group[0.040]; chance[0.039]; thrashed[0.039]; performance[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andreas Heraf\u001b[39m ==> ENTITY: \u001b[32mAndreas Heraf\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayers[0.044]; players[0.044]; game[0.044]; play[0.043]; Sc[0.043]; Freiburg[0.043]; Dietmar[0.043]; won[0.042]; Rapid[0.042]; started[0.042]; Markus[0.042]; Sk[0.042]; team[0.041]; team[0.041]; Austrian[0.041]; Austrian[0.041]; taking[0.040]; captain[0.040]; coach[0.040]; ball[0.040]; minutes[0.040]; points[0.040]; goalkeeper[0.040]; draw[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belarus ==> ENTITY: \u001b[32mBelarus national football team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.164:0.164[0]; log p(e|m)= -2.025:-2.025[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.047]; qualifier[0.046]; match[0.045]; Austria[0.045]; Austria[0.045]; Austria[0.045]; Scotland[0.045]; Scotland[0.045]; Scotland[0.045]; Scotland[0.045]; Qualifier[0.043]; draw[0.043]; Cup[0.042]; Cup[0.042]; Scottish[0.042]; left[0.042]; Saturday[0.042]; group[0.042]; group[0.042]; World[0.041]; World[0.041]; close[0.041]; close[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Herbert Prohaska\u001b[39m ==> ENTITY: \u001b[32mHerbert Prohaska\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.045]; team[0.044]; team[0.044]; team[0.044]; Austrian[0.043]; Austrian[0.043]; Dietmar[0.042]; Dietmar[0.042]; players[0.041]; McCall[0.041]; Gary[0.041]; won[0.041]; Craig[0.041]; Craig[0.041]; Markus[0.041]; Rapid[0.041]; draw[0.041]; Ferguson[0.041]; Wolfgang[0.041]; Schopp[0.041]; play[0.041]; Sk[0.041]; Thomas[0.040]; Thomas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Anton Polster\u001b[39m ==> ENTITY: \u001b[32mToni Polster\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.044]; team[0.043]; team[0.043]; team[0.043]; Markus[0.043]; Andreas[0.043]; Andreas[0.043]; Andreas[0.043]; Austrian[0.042]; Dieter[0.042]; Wolfgang[0.042]; Dietmar[0.042]; Dietmar[0.042]; won[0.042]; coach[0.042]; Sk[0.041]; players[0.041]; Stephan[0.040]; Herzog[0.040]; game[0.040]; started[0.040]; Anton[0.040]; slip[0.039]; really[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Herfried Sabitzer\u001b[39m ==> ENTITY: \u001b[32mHerfried Sabitzer\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.044]; team[0.043]; team[0.043]; team[0.043]; players[0.043]; won[0.043]; points[0.043]; Dietmar[0.043]; Dietmar[0.043]; game[0.042]; Austrian[0.042]; Sk[0.042]; lost[0.042]; started[0.041]; Andreas[0.041]; Andreas[0.041]; Andreas[0.041]; Markus[0.041]; coach[0.040]; Rapid[0.040]; Anton[0.040]; Anton[0.040]; Dieter[0.040]; Wolfgang[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Duncan Ferguson\u001b[39m ==> ENTITY: \u001b[32mDuncan Ferguson\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.137:0.137[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCraig[0.044]; Ferguson[0.044]; Gary[0.044]; match[0.043]; team[0.043]; Michael[0.043]; players[0.042]; Scotland[0.042]; Scotland[0.042]; Scotland[0.042]; play[0.042]; McAllister[0.042]; Everton[0.042]; Sc[0.042]; Freiburg[0.041]; Dietmar[0.041]; captain[0.041]; Scottish[0.040]; Scottish[0.040]; penalty[0.040]; draw[0.040]; draw[0.040]; Sk[0.040]; hoped[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gary McAllister\u001b[39m ==> ENTITY: \u001b[32mGary McAllister\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.119:0.119[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.044]; Craig[0.044]; Calderwood[0.044]; Burley[0.044]; Ferguson[0.043]; Colin[0.043]; Colin[0.043]; Scotland[0.043]; Michael[0.042]; Dietmar[0.042]; Stuart[0.042]; lacked[0.042]; Goram[0.041]; Andrew[0.041]; Boyd[0.041]; Gordon[0.041]; John[0.040]; Duncan[0.040]; McCall[0.040]; Andreas[0.040]; Andreas[0.040]; Andreas[0.040]; Markus[0.040]; Alistair[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = John Collins ==> ENTITY: \u001b[32mJohn Collins (footballer born 1968)\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.124:0.124[0]; log p(e|m)= -1.619:-1.619[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nColin[0.044]; Colin[0.044]; Gary[0.043]; team[0.043]; Scotland[0.043]; Thomas[0.043]; Thomas[0.043]; Michael[0.043]; Ferguson[0.042]; Andrew[0.042]; McAllister[0.042]; Stuart[0.042]; Craig[0.042]; Duncan[0.042]; Gordon[0.041]; Peter[0.041]; Calderwood[0.040]; Burley[0.040]; Boyd[0.040]; McCoist[0.040]; Goram[0.040]; McCall[0.040]; Alistair[0.039]; great[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scotland ==> ENTITY: \u001b[32mScotland national football team\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.159:0.159[0]; log p(e|m)= -2.703:-2.703[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.050]; Scotland[0.050]; Scottish[0.046]; match[0.046]; Soccer[0.043]; Cup[0.042]; Cup[0.042]; qualifier[0.041]; Saturday[0.041]; Belarus[0.041]; Qualifier[0.040]; World[0.040]; World[0.040]; draw[0.040]; Steve[0.040]; opening[0.040]; Austria[0.040]; Austria[0.040]; Austria[0.040]; group[0.040]; group[0.040]; chance[0.039]; thrashed[0.039]; performance[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ferguson\u001b[39m ==> ENTITY: \u001b[32mDuncan Ferguson\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCraig[0.044]; Ferguson[0.044]; Gary[0.043]; team[0.042]; team[0.042]; game[0.042]; Duncan[0.042]; Michael[0.042]; players[0.042]; players[0.042]; Scotland[0.042]; Scotland[0.042]; pitch[0.042]; play[0.042]; McAllister[0.042]; Everton[0.041]; Freiburg[0.041]; Dietmar[0.041]; captain[0.040]; Scottish[0.040]; Scottish[0.040]; penalty[0.040]; draw[0.040]; Sk[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAustria\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustria\u001b[39m <---> \u001b[32mAustria national football team\u001b[39m\t\nSCORES: global= 0.252:0.252[\u001b[31m0.000\u001b[39m]; local()= 0.159:0.147[\u001b[31m0.012\u001b[39m]; log p(e|m)= -0.252:-2.937[\u001b[31m2.685\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.047]; team[0.047]; team[0.047]; Austrian[0.043]; Austrian[0.043]; won[0.042]; Markus[0.042]; Wolfgang[0.042]; Scotland[0.042]; Andreas[0.041]; Andreas[0.041]; Andreas[0.041]; players[0.041]; points[0.041]; Herzog[0.041]; lost[0.041]; draw[0.041]; game[0.040]; Dieter[0.040]; Peter[0.040]; Anton[0.039]; Anton[0.039]; Scots[0.039]; play[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mScottish\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mScotland\u001b[39m <---> \u001b[32mScotland national football team\u001b[39m\t\nSCORES: global= 0.248:0.244[\u001b[31m0.004\u001b[39m]; local()= 0.153:0.175[\u001b[32m0.022\u001b[39m]; log p(e|m)= -0.480:-3.912[\u001b[31m3.432\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.053]; Scotland[0.053]; Scotland[0.053]; Scotland[0.053]; Scotland[0.053]; players[0.045]; Soccer[0.045]; match[0.045]; Cup[0.044]; Cup[0.044]; Ferguson[0.044]; Ferguson[0.044]; McAllister[0.043]; qualifier[0.043]; Duncan[0.043]; Belarus[0.043]; Gary[0.042]; Qualifier[0.042]; Michael[0.042]; Steve[0.042]; dominated[0.042]; draw[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andreas Herzog\u001b[39m ==> ENTITY: \u001b[32mAndreas Herzog\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.045]; Dieter[0.044]; Andreas[0.044]; Andreas[0.044]; Markus[0.042]; team[0.042]; team[0.042]; team[0.042]; players[0.042]; Austrian[0.042]; Rapid[0.042]; Stephan[0.042]; Dietmar[0.041]; Dietmar[0.041]; coach[0.041]; Polster[0.041]; Michael[0.041]; Wolfgang[0.041]; Gary[0.041]; Sk[0.041]; game[0.040]; pitch[0.040]; Craig[0.040]; won[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andreas Heraf\u001b[39m ==> ENTITY: \u001b[32mAndreas Heraf\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayers[0.045]; Austria[0.044]; game[0.044]; Dietmar[0.043]; Dietmar[0.043]; won[0.043]; started[0.042]; Markus[0.042]; team[0.042]; team[0.042]; team[0.042]; Austrian[0.042]; Andreas[0.041]; Andreas[0.041]; coach[0.041]; Dieter[0.041]; Anton[0.041]; Anton[0.041]; Wolfgang[0.041]; minutes[0.040]; points[0.040]; really[0.040]; Stephan[0.040]; lost[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scotland ==> ENTITY: \u001b[32mScotland national football team\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.169:0.169[0]; log p(e|m)= -2.703:-2.703[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.049]; Scotland[0.049]; Scottish[0.045]; Scottish[0.045]; match[0.045]; team[0.043]; players[0.043]; McAllister[0.041]; Ferguson[0.041]; Ferguson[0.041]; captain[0.041]; play[0.041]; game[0.041]; Craig[0.041]; coach[0.041]; Duncan[0.040]; Belarus[0.040]; Gary[0.039]; draw[0.039]; draw[0.039]; Michael[0.039]; Brown[0.039]; opening[0.039]; Austria[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scottish ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.480:-0.480[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.050]; Scotland[0.050]; Scots[0.045]; team[0.043]; team[0.043]; team[0.043]; players[0.042]; players[0.042]; Ferguson[0.041]; Ferguson[0.041]; won[0.041]; play[0.041]; McAllister[0.041]; Craig[0.041]; Duncan[0.041]; game[0.041]; Gary[0.040]; Michael[0.040]; Herbert[0.039]; draw[0.039]; captain[0.039]; Everton[0.039]; coach[0.039]; coach[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria national football team\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.176:0.176[0]; log p(e|m)= -2.937:-2.937[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.048]; Austria[0.048]; Belarus[0.044]; qualifier[0.043]; players[0.043]; Markus[0.042]; Scotland[0.042]; Scotland[0.042]; Scotland[0.042]; Scotland[0.042]; Scotland[0.042]; match[0.042]; draw[0.041]; Austrian[0.041]; World[0.040]; World[0.040]; Vienna[0.040]; went[0.040]; went[0.040]; Soccer[0.040]; Cup[0.040]; Cup[0.040]; Qualifier[0.039]; Michael[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Michael Konsel\u001b[39m ==> ENTITY: \u001b[32mMichael Konsel\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.050]; Austrian[0.046]; Austrian[0.046]; team[0.044]; team[0.044]; team[0.044]; Dietmar[0.044]; Dietmar[0.044]; Rapid[0.044]; players[0.044]; Sk[0.043]; game[0.043]; won[0.043]; Andreas[0.042]; Andreas[0.042]; Andreas[0.042]; Markus[0.042]; points[0.042]; draw[0.042]; coach[0.042]; pitch[0.042]; play[0.042]; Dieter[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria national football team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.175:0.175[0]; log p(e|m)= -2.937:-2.937[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.048]; Austria[0.048]; Belarus[0.045]; Cup[0.044]; Cup[0.044]; qualifier[0.043]; World[0.043]; World[0.043]; Scotland[0.042]; Scotland[0.042]; Scotland[0.042]; Qualifier[0.042]; match[0.042]; Soccer[0.041]; draw[0.040]; Vienna[0.040]; dominated[0.039]; Saturday[0.039]; Scottish[0.039]; thrashed[0.039]; came[0.039]; opening[0.038]; group[0.038]; group[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vienna ==> ENTITY: \u001b[32mVienna\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.086:0.086[0]; log p(e|m)= -0.126:-0.126[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.045]; Austria[0.045]; Austria[0.045]; Austria[0.045]; Qualifier[0.045]; qualifier[0.043]; draw[0.042]; Cup[0.042]; Cup[0.042]; Saturday[0.041]; Steve[0.041]; came[0.041]; opening[0.041]; area[0.041]; World[0.041]; World[0.041]; Scotland[0.040]; Scotland[0.040]; Scotland[0.040]; group[0.040]; group[0.040]; Stephan[0.040]; taking[0.040]; Soccer[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scotland ==> ENTITY: \u001b[32mScotland national football team\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.153:0.153[0]; log p(e|m)= -2.703:-2.703[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDuncan[0.044]; Scots[0.043]; Burley[0.043]; Stuart[0.043]; Craig[0.043]; Goram[0.043]; team[0.043]; team[0.043]; team[0.043]; Gordon[0.042]; players[0.042]; Andrew[0.042]; Colin[0.041]; Colin[0.041]; McAllister[0.041]; Calderwood[0.041]; Ferguson[0.041]; game[0.041]; coach[0.040]; McCall[0.040]; Boyd[0.040]; won[0.040]; John[0.040]; Alistair[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andrew Goram\u001b[39m ==> ENTITY: \u001b[32mAndy Goram\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.046]; Craig[0.044]; Colin[0.044]; Colin[0.044]; Stuart[0.043]; Gary[0.043]; McCoist[0.043]; Burley[0.042]; team[0.042]; team[0.042]; team[0.042]; Calderwood[0.042]; won[0.042]; players[0.041]; Herbert[0.041]; game[0.041]; Dietmar[0.040]; Gordon[0.040]; Michael[0.040]; Alistair[0.040]; Ferguson[0.040]; John[0.040]; coach[0.040]; Scots[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austrian ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.118:0.118[0]; log p(e|m)= -0.406:-0.406[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.049]; players[0.045]; players[0.045]; Austrian[0.043]; team[0.043]; Belarus[0.043]; Scotland[0.043]; Scotland[0.043]; Scotland[0.043]; match[0.042]; play[0.041]; Markus[0.041]; arguing[0.041]; Michael[0.041]; Andreas[0.041]; game[0.041]; draw[0.041]; performance[0.040]; performance[0.040]; decision[0.040]; coach[0.039]; Ferguson[0.039]; Ferguson[0.039]; started[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Craig Brown ==> ENTITY: \u001b[32mCraig Brown (footballer born 1940)\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.167:0.167[0]; log p(e|m)= -1.973:-1.973[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.044]; Scotland[0.044]; Ferguson[0.043]; Ferguson[0.043]; team[0.043]; team[0.043]; team[0.043]; players[0.043]; players[0.043]; coach[0.043]; coach[0.043]; Duncan[0.042]; won[0.042]; Schopp[0.041]; Scottish[0.041]; McAllister[0.041]; captain[0.040]; Herbert[0.040]; Teams[0.040]; Michael[0.040]; Gary[0.040]; game[0.040]; play[0.040]; started[0.040]; \t\n\r [==============>...............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s414ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1649/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1060testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kent ==> ENTITY: \u001b[32mKent County Cricket Club\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.200:0.200[0]; log p(e|m)= -2.513:-2.513[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEssex[0.047]; Essex[0.047]; Essex[0.047]; Surrey[0.045]; Kent[0.044]; Kent[0.044]; Derbyshire[0.043]; match[0.043]; Kettleborough[0.043]; Nottinghamshire[0.043]; Cricket[0.042]; Yorkshire[0.041]; Richard[0.040]; wickets[0.040]; wickets[0.040]; day[0.040]; day[0.040]; county[0.040]; English[0.040]; class[0.040]; century[0.039]; final[0.038]; maiden[0.038]; Saturday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tim Munton\u001b[39m ==> ENTITY: \u001b[32mTim Munton\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWorcestershire[0.047]; Derbyshire[0.045]; Derbyshire[0.045]; Warwickshire[0.045]; Warwickshire[0.045]; Leicestershire[0.045]; wicket[0.044]; class[0.044]; wickets[0.044]; wickets[0.044]; season[0.044]; match[0.043]; match[0.043]; match[0.043]; innings[0.043]; Nottinghamshire[0.043]; Richard[0.043]; Kettleborough[0.042]; Yorkshire[0.042]; runs[0.041]; Somerset[0.041]; struggling[0.041]; catches[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tunbridge Wells ==> ENTITY: \u001b[32mRoyal Tunbridge Wells\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.260:-0.260[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.044]; match[0.044]; match[0.044]; Essex[0.044]; Essex[0.044]; Kent[0.043]; Kent[0.043]; Somerset[0.043]; Leicestershire[0.042]; Surrey[0.042]; Surrey[0.042]; Surrey[0.042]; Warwickshire[0.041]; Warwickshire[0.041]; class[0.041]; season[0.041]; Dave[0.041]; run[0.040]; Saturday[0.040]; Worcestershire[0.040]; Richard[0.040]; day[0.040]; day[0.040]; day[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Essex ==> ENTITY: \u001b[32mEssex County Cricket Club\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.219:0.219[0]; log p(e|m)= -2.343:-2.343[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; Warwickshire[0.044]; Essex[0.044]; Essex[0.044]; Leicestershire[0.044]; Worcestershire[0.043]; Somerset[0.042]; match[0.042]; match[0.042]; Kettleborough[0.042]; Surrey[0.042]; Surrey[0.042]; Nottinghamshire[0.042]; Derbyshire[0.041]; Derbyshire[0.041]; wicket[0.041]; Cricket[0.040]; Kent[0.040]; Kent[0.040]; Kent[0.040]; Yorkshire[0.040]; wickets[0.040]; wickets[0.040]; innings[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Derbyshire ==> ENTITY: \u001b[32mDerbyshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.220:0.220[0]; log p(e|m)= -1.483:-1.483[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.046]; Essex[0.044]; Essex[0.044]; Essex[0.044]; Worcestershire[0.044]; Nottinghamshire[0.043]; Derbyshire[0.043]; Surrey[0.043]; Surrey[0.043]; Cricket[0.042]; Kent[0.042]; Kent[0.042]; Kent[0.042]; Kettleborough[0.041]; wicket[0.041]; match[0.041]; match[0.041]; wickets[0.041]; wickets[0.041]; Yorkshire[0.040]; Richard[0.039]; runs[0.038]; innings[0.038]; English[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leicestershire ==> ENTITY: \u001b[32mLeicestershire County Cricket Club\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.188:0.188[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.045]; Warwickshire[0.045]; wicket[0.044]; wickets[0.043]; wickets[0.043]; Essex[0.042]; Worcestershire[0.042]; match[0.042]; match[0.042]; match[0.042]; Somerset[0.042]; Nottinghamshire[0.042]; Kettleborough[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Surrey[0.041]; Surrey[0.041]; innings[0.041]; Kent[0.041]; season[0.040]; Richard[0.040]; win[0.040]; runs[0.040]; Yorkshire[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Derbyshire ==> ENTITY: \u001b[32mDerbyshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.208:0.208[0]; log p(e|m)= -1.483:-1.483[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.046]; Warwickshire[0.046]; Worcestershire[0.044]; Somerset[0.044]; Leicestershire[0.044]; Nottinghamshire[0.044]; Derbyshire[0.043]; Surrey[0.043]; Surrey[0.043]; Kent[0.042]; wicket[0.042]; match[0.042]; match[0.042]; wickets[0.041]; Munton[0.040]; season[0.040]; captain[0.040]; captain[0.040]; Dean[0.040]; win[0.039]; final[0.039]; runs[0.039]; innings[0.039]; Dave[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Worcestershire ==> ENTITY: \u001b[32mWorcestershire County Cricket Club\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.916:-0.916[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwicket[0.044]; Warwickshire[0.044]; Warwickshire[0.044]; Essex[0.044]; Leicestershire[0.043]; wickets[0.043]; wickets[0.043]; Somerset[0.043]; Nottinghamshire[0.042]; innings[0.042]; Surrey[0.041]; Surrey[0.041]; Surrey[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; match[0.041]; match[0.041]; match[0.041]; runs[0.040]; Kettleborough[0.040]; Richard[0.040]; Kent[0.040]; final[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Richard Kettleborough\u001b[39m ==> ENTITY: \u001b[32mRichard Kettleborough\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.045]; Warwickshire[0.045]; Worcestershire[0.044]; match[0.043]; match[0.043]; Cricket[0.043]; Leicestershire[0.043]; Nottinghamshire[0.042]; day[0.042]; day[0.042]; day[0.042]; Essex[0.041]; Essex[0.041]; Essex[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; English[0.041]; Yorkshire[0.041]; class[0.040]; Monday[0.040]; Kent[0.040]; Kent[0.040]; Kent[0.040]; Saturday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Surrey ==> ENTITY: \u001b[32mSurrey County Cricket Club\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.206:0.206[0]; log p(e|m)= -1.796:-1.796[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.045]; Essex[0.045]; Essex[0.045]; Essex[0.045]; Surrey[0.044]; Worcestershire[0.044]; Nottinghamshire[0.042]; Derbyshire[0.042]; Derbyshire[0.042]; Kent[0.042]; Kent[0.042]; Kent[0.042]; Kettleborough[0.042]; match[0.041]; match[0.041]; Yorkshire[0.041]; Cricket[0.040]; wicket[0.039]; wickets[0.039]; wickets[0.039]; county[0.039]; London[0.039]; day[0.039]; day[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Surrey ==> ENTITY: \u001b[32mSurrey County Cricket Club\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.203:0.203[0]; log p(e|m)= -1.796:-1.796[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.046]; Warwickshire[0.046]; Leicestershire[0.045]; Surrey[0.045]; Somerset[0.044]; Worcestershire[0.044]; Nottinghamshire[0.043]; Derbyshire[0.042]; Derbyshire[0.042]; Kent[0.042]; Kettleborough[0.042]; match[0.041]; match[0.041]; match[0.041]; Yorkshire[0.041]; wicket[0.040]; wickets[0.040]; day[0.039]; day[0.039]; Wells[0.039]; captain[0.039]; captain[0.039]; runs[0.039]; class[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.065:0.065[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSurrey[0.046]; Essex[0.045]; Essex[0.045]; Essex[0.045]; match[0.044]; match[0.044]; final[0.043]; Nottinghamshire[0.042]; Derbyshire[0.042]; Cricket[0.042]; Kent[0.041]; Kent[0.041]; Kent[0.041]; Yorkshire[0.041]; championship[0.041]; ahead[0.040]; hopefuls[0.040]; Richard[0.040]; Kettleborough[0.040]; beat[0.040]; Monday[0.040]; Saturday[0.040]; English[0.039]; title[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Somerset ==> ENTITY: \u001b[32mSomerset County Cricket Club\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.209:0.209[0]; log p(e|m)= -2.419:-2.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEssex[0.047]; Warwickshire[0.046]; Warwickshire[0.046]; Leicestershire[0.045]; Worcestershire[0.045]; Surrey[0.044]; Surrey[0.044]; match[0.044]; match[0.044]; match[0.044]; Nottinghamshire[0.044]; Derbyshire[0.043]; Derbyshire[0.043]; Kent[0.043]; wicket[0.043]; wickets[0.042]; wickets[0.042]; Richard[0.042]; Kettleborough[0.042]; county[0.042]; innings[0.041]; class[0.041]; Yorkshire[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nottinghamshire ==> ENTITY: \u001b[32mNottinghamshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.202:0.202[0]; log p(e|m)= -1.106:-1.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; Warwickshire[0.044]; Essex[0.044]; Essex[0.044]; Essex[0.044]; Somerset[0.043]; Leicestershire[0.043]; Worcestershire[0.042]; wicket[0.042]; match[0.041]; match[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; wickets[0.041]; wickets[0.041]; Surrey[0.041]; Surrey[0.041]; Surrey[0.041]; Cricket[0.040]; Kent[0.040]; Kent[0.040]; Kent[0.040]; Kettleborough[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = English ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.128:0.128[0]; log p(e|m)= -1.355:-1.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.043]; Leicestershire[0.043]; match[0.043]; match[0.043]; Cricket[0.043]; Essex[0.043]; Essex[0.043]; Essex[0.043]; Kettleborough[0.042]; Worcestershire[0.042]; Richard[0.042]; Somerset[0.042]; Surrey[0.041]; Surrey[0.041]; class[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; wicket[0.041]; Nottinghamshire[0.040]; wickets[0.040]; wickets[0.040]; Kent[0.040]; Kent[0.040]; Kent[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAustralian\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustralia national cricket team\u001b[39m <---> \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.253:0.243[\u001b[31m0.011\u001b[39m]; local()= 0.147:0.053[\u001b[31m0.094\u001b[39m]; log p(e|m)= -3.612:-0.491[\u001b[32m3.121\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralian[0.046]; match[0.046]; match[0.046]; captain[0.044]; captain[0.044]; Warwickshire[0.044]; Warwickshire[0.044]; Leicestershire[0.044]; day[0.044]; day[0.044]; season[0.043]; coach[0.043]; Surrey[0.043]; Surrey[0.043]; final[0.043]; Somerset[0.043]; wickets[0.043]; Worcestershire[0.043]; wicket[0.042]; Kent[0.042]; win[0.042]; Nottinghamshire[0.042]; Tim[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Surrey ==> ENTITY: \u001b[32mSurrey County Cricket Club\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.209:0.209[0]; log p(e|m)= -1.796:-1.796[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.049]; Warwickshire[0.049]; Essex[0.049]; Leicestershire[0.048]; Surrey[0.048]; Surrey[0.048]; Somerset[0.047]; Worcestershire[0.047]; Nottinghamshire[0.045]; Derbyshire[0.045]; Derbyshire[0.045]; Derbyshire[0.045]; Kent[0.045]; Kettleborough[0.045]; match[0.044]; match[0.044]; match[0.044]; Yorkshire[0.044]; wicket[0.042]; wickets[0.042]; wickets[0.042]; county[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAustralian\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustralia national cricket team\u001b[39m <---> \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.254:0.242[\u001b[31m0.011\u001b[39m]; local()= 0.153:0.053[\u001b[31m0.100\u001b[39m]; log p(e|m)= -3.612:-0.491[\u001b[32m3.121\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralian[0.044]; match[0.043]; match[0.043]; match[0.043]; captain[0.042]; captain[0.042]; Warwickshire[0.042]; Warwickshire[0.042]; Leicestershire[0.042]; day[0.042]; day[0.042]; Kettleborough[0.042]; season[0.041]; coach[0.041]; Richard[0.041]; Surrey[0.041]; Surrey[0.041]; final[0.041]; Somerset[0.041]; wickets[0.041]; Worcestershire[0.041]; wicket[0.040]; Kent[0.040]; win[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Derbyshire ==> ENTITY: \u001b[32mDerbyshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.226:0.226[0]; log p(e|m)= -1.483:-1.483[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.047]; Warwickshire[0.047]; Essex[0.045]; Worcestershire[0.045]; Somerset[0.045]; Leicestershire[0.045]; Nottinghamshire[0.044]; Derbyshire[0.044]; Derbyshire[0.044]; Surrey[0.044]; Surrey[0.044]; Surrey[0.044]; Kent[0.043]; Kettleborough[0.043]; wicket[0.043]; match[0.042]; match[0.042]; match[0.042]; wickets[0.042]; wickets[0.042]; Yorkshire[0.042]; Munton[0.041]; season[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Yorkshire ==> ENTITY: \u001b[32mYorkshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.218:0.218[0]; log p(e|m)= -1.778:-1.778[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.046]; Warwickshire[0.046]; Leicestershire[0.046]; Essex[0.045]; Essex[0.045]; Essex[0.045]; Worcestershire[0.045]; Somerset[0.044]; Nottinghamshire[0.044]; Derbyshire[0.044]; Derbyshire[0.044]; Surrey[0.043]; Surrey[0.043]; Surrey[0.043]; match[0.043]; match[0.043]; wicket[0.042]; class[0.042]; Kettleborough[0.042]; Cricket[0.041]; wickets[0.041]; wickets[0.041]; championship[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Essex ==> ENTITY: \u001b[32mEssex County Cricket Club\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.206:0.206[0]; log p(e|m)= -2.343:-2.343[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEssex[0.046]; Essex[0.046]; match[0.044]; Kettleborough[0.044]; Surrey[0.044]; Nottinghamshire[0.043]; Derbyshire[0.043]; Cricket[0.042]; Kent[0.042]; Kent[0.042]; Kent[0.042]; Yorkshire[0.042]; wickets[0.041]; wickets[0.041]; class[0.041]; day[0.041]; day[0.041]; Monday[0.040]; English[0.039]; maiden[0.039]; final[0.039]; Richard[0.039]; county[0.039]; run[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Essex ==> ENTITY: \u001b[32mEssex County Cricket Club\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.210:0.210[0]; log p(e|m)= -2.343:-2.343[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEssex[0.045]; Essex[0.045]; match[0.044]; match[0.044]; Kettleborough[0.044]; Surrey[0.043]; Nottinghamshire[0.043]; Derbyshire[0.043]; Cricket[0.042]; Kent[0.042]; Kent[0.042]; Kent[0.042]; Yorkshire[0.041]; wickets[0.041]; wickets[0.041]; class[0.041]; day[0.041]; day[0.041]; Monday[0.040]; English[0.039]; maiden[0.039]; final[0.039]; Richard[0.039]; county[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Warwickshire ==> ENTITY: \u001b[32mWarwickshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.185:0.185[0]; log p(e|m)= -1.109:-1.109[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.047]; Leicestershire[0.046]; Worcestershire[0.045]; Essex[0.045]; Somerset[0.045]; wicket[0.044]; Nottinghamshire[0.044]; Derbyshire[0.043]; Derbyshire[0.043]; match[0.043]; match[0.043]; match[0.043]; Kent[0.043]; Kettleborough[0.043]; Surrey[0.043]; Surrey[0.043]; final[0.043]; win[0.043]; innings[0.042]; wickets[0.042]; wickets[0.042]; Yorkshire[0.042]; runs[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dean Jones ==> ENTITY: \u001b[32mDean Jones (cricketer)\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.149:0.149[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.043]; Warwickshire[0.043]; match[0.043]; match[0.043]; day[0.042]; day[0.042]; Leicestershire[0.042]; Nottinghamshire[0.042]; Worcestershire[0.042]; captain[0.042]; captain[0.042]; Tim[0.042]; Surrey[0.041]; Surrey[0.041]; Dave[0.041]; Australian[0.041]; Australian[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Somerset[0.041]; took[0.041]; season[0.041]; wicket[0.040]; wickets[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Warwickshire ==> ENTITY: \u001b[32mWarwickshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.184:0.184[0]; log p(e|m)= -1.109:-1.109[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.047]; Leicestershire[0.045]; Worcestershire[0.045]; Essex[0.045]; county[0.045]; Somerset[0.044]; wicket[0.044]; English[0.044]; Nottinghamshire[0.044]; Derbyshire[0.043]; Derbyshire[0.043]; Derbyshire[0.043]; match[0.043]; match[0.043]; match[0.043]; Kent[0.043]; Kettleborough[0.043]; Surrey[0.043]; Surrey[0.043]; Surrey[0.043]; final[0.042]; win[0.042]; innings[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kent ==> ENTITY: \u001b[32mKent County Cricket Club\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.205:0.205[0]; log p(e|m)= -2.513:-2.513[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEssex[0.044]; Essex[0.044]; Essex[0.044]; Warwickshire[0.044]; Warwickshire[0.044]; Somerset[0.043]; Surrey[0.043]; Surrey[0.043]; Surrey[0.043]; Leicestershire[0.042]; Worcestershire[0.042]; Kent[0.041]; Kent[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; match[0.041]; match[0.041]; Kettleborough[0.040]; Nottinghamshire[0.040]; Cricket[0.040]; Yorkshire[0.039]; wicket[0.039]; Richard[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dave Gilbert ==> ENTITY: \u001b[32mDave Gilbert (cricketer)\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.872:-0.872[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.043]; Warwickshire[0.043]; captain[0.043]; captain[0.043]; win[0.043]; match[0.042]; match[0.042]; match[0.042]; final[0.042]; took[0.042]; Leicestershire[0.042]; Jones[0.042]; coach[0.042]; Somerset[0.041]; title[0.041]; beat[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; conceding[0.041]; Australian[0.041]; Australian[0.041]; Surrey[0.041]; Surrey[0.041]; Worcestershire[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kent ==> ENTITY: \u001b[32mKent County Cricket Club\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.196:0.196[0]; log p(e|m)= -2.513:-2.513[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEssex[0.046]; Essex[0.046]; Essex[0.046]; Surrey[0.045]; Kent[0.043]; Kent[0.043]; Derbyshire[0.043]; match[0.043]; match[0.043]; Kettleborough[0.042]; Nottinghamshire[0.042]; Cricket[0.042]; Yorkshire[0.041]; Richard[0.040]; wickets[0.040]; wickets[0.040]; day[0.040]; day[0.040]; county[0.040]; English[0.039]; class[0.039]; Wells[0.039]; century[0.039]; Tunbridge[0.038]; \t\n\r [===============>..............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s323ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1677/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1040testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesian ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.165:0.165[0]; log p(e|m)= -0.742:-0.742[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesian[0.047]; Indonesian[0.047]; Philippines[0.046]; country[0.046]; Jakarta[0.044]; Jakarta[0.044]; southern[0.043]; businessmen[0.043]; newspapers[0.042]; newspapers[0.042]; experiences[0.042]; Kompas[0.042]; stories[0.041]; stories[0.041]; business[0.040]; Reuters[0.039]; Reuters[0.039]; state[0.039]; social[0.039]; coming[0.039]; major[0.039]; Representatives[0.039]; asked[0.039]; Digest[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.220:0.220[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.050]; Indonesia[0.050]; Indonesian[0.045]; Philippines[0.045]; Korea[0.043]; Jakarta[0.042]; Jakarta[0.042]; region[0.042]; Philippine[0.041]; national[0.041]; Timor[0.041]; Timor[0.041]; Sukarnoputri[0.041]; leader[0.040]; ousted[0.040]; ousted[0.040]; government[0.040]; government[0.040]; government[0.040]; South[0.039]; Putra[0.039]; Suharto[0.039]; start[0.039]; Central[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = MNLF\u001b[39m ==> ENTITY: \u001b[32mMoro National Liberation Front\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhilippines[0.049]; Moro[0.049]; Philippine[0.048]; peace[0.047]; national[0.045]; government[0.045]; region[0.044]; country[0.044]; conflict[0.044]; Liberation[0.043]; National[0.043]; efforts[0.043]; President[0.042]; President[0.042]; effort[0.042]; Indonesia[0.042]; Indonesia[0.042]; Indonesia[0.042]; Representatives[0.042]; bring[0.041]; agreement[0.041]; southern[0.041]; Putra[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jakarta Post\u001b[39m ==> ENTITY: \u001b[32mThe Jakarta Post\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJakarta[0.045]; newspapers[0.045]; newspapers[0.045]; stories[0.044]; stories[0.044]; Indonesia[0.044]; Indonesia[0.044]; Kompas[0.044]; Indonesian[0.043]; Indonesian[0.043]; Indonesian[0.043]; Media[0.041]; Philippines[0.040]; Digest[0.040]; effort[0.040]; Reuters[0.040]; Reuters[0.040]; business[0.040]; Press[0.039]; boost[0.039]; businessmen[0.039]; Philippine[0.039]; institutions[0.039]; direct[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kompas\u001b[39m ==> ENTITY: \u001b[32mKompas\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnewspapers[0.048]; newspapers[0.048]; Jakarta[0.047]; Jakarta[0.047]; summary[0.044]; Indonesian[0.044]; Indonesian[0.044]; Indonesian[0.044]; stories[0.043]; stories[0.043]; Philippines[0.043]; Liberation[0.042]; Friday[0.042]; major[0.042]; President[0.042]; business[0.042]; government[0.042]; Suharto[0.042]; political[0.042]; Fax[0.041]; Post[0.041]; exports[0.041]; Representatives[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tanjung Priok\u001b[39m ==> ENTITY: \u001b[32mTanjung Priok\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.129:0.129[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nport[0.045]; Indonesia[0.044]; Indonesia[0.044]; Indonesia[0.044]; southern[0.042]; Jakarta[0.042]; Jakarta[0.042]; South[0.042]; Indonesian[0.042]; state[0.041]; region[0.041]; Philippines[0.041]; government[0.041]; government[0.041]; government[0.041]; Sukarnoputri[0.041]; Timor[0.041]; Timor[0.041]; Central[0.041]; congress[0.040]; Putra[0.040]; rebel[0.040]; Philippine[0.040]; Pt[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Megawati Sukarnoputri\u001b[39m ==> ENTITY: \u001b[32mMegawati Sukarnoputri\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.047]; Megawati[0.044]; Indonesia[0.044]; Indonesia[0.044]; Indonesia[0.044]; leader[0.043]; party[0.043]; Party[0.043]; government[0.041]; government[0.041]; Indonesian[0.041]; congress[0.041]; Jakarta[0.041]; Jakarta[0.041]; Putra[0.041]; ousted[0.041]; ousted[0.041]; Central[0.041]; rebel[0.041]; Pt[0.041]; Republika[0.040]; Priok[0.040]; Timor[0.040]; Timor[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.250:0.250[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.050]; Indonesia[0.050]; Indonesian[0.045]; Philippines[0.045]; country[0.044]; Korea[0.043]; Jakarta[0.042]; Jakarta[0.042]; Jakarta[0.042]; region[0.042]; southern[0.041]; Philippine[0.041]; national[0.041]; Timor[0.041]; Timor[0.041]; Putra[0.040]; Sukarnoputri[0.040]; South[0.040]; government[0.039]; government[0.039]; efforts[0.039]; start[0.038]; Central[0.038]; Tanjung[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Philippines ==> ENTITY: \u001b[32mPhilippines\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.180:-0.180[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhilippine[0.048]; country[0.045]; national[0.045]; Indonesia[0.043]; Indonesia[0.043]; Indonesia[0.043]; Representatives[0.043]; Korea[0.042]; southern[0.042]; National[0.042]; President[0.042]; President[0.042]; region[0.042]; government[0.042]; serving[0.041]; Thursday[0.041]; Moro[0.040]; start[0.040]; Friday[0.040]; Telephone[0.039]; Putra[0.039]; Liberation[0.039]; businessmen[0.039]; peace[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesian ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.176:0.176[0]; log p(e|m)= -0.742:-0.742[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesian[0.048]; Indonesian[0.048]; Philippines[0.048]; country[0.048]; Jakarta[0.045]; Jakarta[0.045]; southern[0.045]; Philippine[0.044]; businessmen[0.044]; experiences[0.043]; Kompas[0.043]; stories[0.042]; stories[0.042]; government[0.042]; business[0.041]; newspapers[0.041]; newspapers[0.041]; National[0.041]; institutions[0.041]; Reuters[0.041]; Reuters[0.041]; Moro[0.041]; delegates[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Republika ==> ENTITY: \u001b[32mRepublika (Indonesian newspaper)\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.198:0.198[0]; log p(e|m)= -0.837:-0.837[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.043]; Indonesia[0.043]; Indonesia[0.043]; June[0.043]; Media[0.043]; Democratic[0.043]; Nasional[0.042]; national[0.042]; Indonesian[0.042]; Liberation[0.042]; Jakarta[0.042]; Jakarta[0.042]; Thursday[0.042]; government[0.042]; government[0.042]; government[0.042]; Korea[0.041]; Sukarnoputri[0.041]; started[0.040]; National[0.040]; President[0.040]; Friday[0.040]; Party[0.040]; South[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.098:0.098[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReuters[0.044]; business[0.044]; newspapers[0.043]; newspapers[0.043]; Press[0.043]; major[0.043]; political[0.043]; southern[0.043]; leading[0.042]; country[0.042]; institutions[0.042]; government[0.041]; social[0.041]; direct[0.041]; Philippines[0.041]; Digest[0.041]; efforts[0.040]; Post[0.040]; stories[0.040]; stories[0.040]; share[0.040]; August[0.040]; exports[0.040]; Kompas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kia\u001b[39m ==> ENTITY: \u001b[32mKia Motors\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKia[0.048]; car[0.045]; cars[0.044]; Motor[0.043]; vehicles[0.043]; Korea[0.043]; South[0.042]; marketed[0.042]; Thursday[0.042]; Media[0.042]; plans[0.041]; Friday[0.041]; efforts[0.041]; direct[0.041]; Corp[0.041]; congress[0.040]; assembling[0.040]; Putra[0.040]; Central[0.040]; rebel[0.040]; Pt[0.040]; June[0.040]; national[0.040]; started[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Moro National Liberation Front\u001b[39m ==> ENTITY: \u001b[32mMoro National Liberation Front\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhilippines[0.048]; Philippine[0.047]; peace[0.046]; national[0.044]; government[0.044]; region[0.042]; country[0.042]; conflict[0.042]; efforts[0.041]; President[0.041]; President[0.041]; effort[0.041]; Indonesia[0.040]; Indonesia[0.040]; Indonesia[0.040]; Representatives[0.040]; bring[0.040]; agreement[0.040]; southern[0.040]; Putra[0.040]; Timor[0.040]; Timor[0.040]; Central[0.040]; negotiations[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Suharto\u001b[39m ==> ENTITY: \u001b[32mSuharto\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.045]; government[0.045]; institutions[0.043]; country[0.043]; Indonesian[0.043]; Indonesian[0.043]; Indonesian[0.043]; Jakarta[0.042]; Jakarta[0.042]; political[0.042]; Philippines[0.042]; Reuters[0.042]; Reuters[0.042]; conflict[0.042]; serving[0.041]; leading[0.041]; Kompas[0.041]; businessmen[0.041]; southern[0.040]; major[0.040]; August[0.040]; region[0.040]; peace[0.040]; Moro[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Megawati\u001b[39m ==> ENTITY: \u001b[32mMegawati Sukarnoputri\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSukarnoputri[0.047]; President[0.047]; Megawati[0.044]; Indonesia[0.044]; leader[0.043]; party[0.043]; Party[0.043]; government[0.041]; government[0.041]; Indonesian[0.041]; congress[0.041]; Jakarta[0.041]; Jakarta[0.041]; Putra[0.041]; ousted[0.041]; ousted[0.041]; Central[0.041]; rebel[0.041]; Pt[0.041]; Republika[0.040]; Priok[0.040]; Timor[0.040]; parties[0.040]; Kia[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Korea ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnational[0.045]; National[0.044]; Central[0.043]; start[0.043]; serving[0.042]; Philippines[0.042]; Kia[0.042]; Kia[0.042]; Indonesia[0.042]; Indonesia[0.042]; Indonesia[0.042]; jointly[0.041]; ousted[0.041]; Friday[0.041]; southern[0.041]; settlement[0.041]; Post[0.041]; President[0.041]; failed[0.041]; assembling[0.041]; Timor[0.041]; Timor[0.041]; high[0.041]; Thursday[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Indonesian Democratic Party\u001b[39m ==> ENTITY: \u001b[32mIndonesian Democratic Party\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSukarnoputri[0.046]; leader[0.044]; President[0.044]; government[0.044]; government[0.044]; party[0.043]; parties[0.042]; Megawati[0.042]; Megawati[0.042]; congress[0.041]; Putra[0.041]; failed[0.041]; Indonesia[0.041]; Indonesia[0.041]; Indonesia[0.041]; Central[0.041]; rebel[0.041]; Pt[0.041]; Nasional[0.041]; ousted[0.040]; ousted[0.040]; Jakarta[0.040]; Jakarta[0.040]; Republika[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Media Indonesia\u001b[39m ==> ENTITY: \u001b[32mMedia Indonesia\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.046]; Indonesia[0.046]; Jakarta[0.044]; Jakarta[0.044]; Jakarta[0.044]; Indonesian[0.044]; Sukarnoputri[0.044]; Republika[0.044]; Nasional[0.042]; country[0.041]; Corp[0.041]; Priok[0.041]; Putra[0.041]; Timor[0.041]; Timor[0.041]; Central[0.040]; Pt[0.040]; Philippines[0.040]; Tanjung[0.040]; Megawati[0.040]; Philippine[0.039]; Kia[0.039]; Kia[0.039]; Korea[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesian ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.146:0.146[0]; log p(e|m)= -0.742:-0.742[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesian[0.047]; Indonesian[0.047]; country[0.047]; Jakarta[0.044]; Jakarta[0.044]; businessmen[0.043]; newspapers[0.043]; newspapers[0.043]; experiences[0.042]; Kompas[0.042]; stories[0.041]; stories[0.041]; business[0.040]; Reuters[0.040]; Reuters[0.040]; state[0.040]; social[0.040]; coming[0.040]; major[0.040]; Representatives[0.039]; asked[0.039]; Digest[0.039]; exports[0.039]; leading[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jakarta\u001b[39m ==> ENTITY: \u001b[32mJakarta\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJakarta[0.049]; Philippines[0.047]; country[0.046]; Indonesian[0.046]; Indonesian[0.046]; Indonesian[0.046]; southern[0.043]; government[0.043]; Philippine[0.041]; Kompas[0.041]; Friday[0.041]; President[0.040]; major[0.040]; political[0.040]; state[0.040]; efforts[0.040]; Representatives[0.039]; summary[0.039]; newspapers[0.039]; newspapers[0.039]; institutions[0.039]; August[0.039]; businessmen[0.039]; business[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Suharto\u001b[39m ==> ENTITY: \u001b[32mSuharto\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.044]; Indonesia[0.044]; Indonesia[0.044]; President[0.044]; government[0.043]; government[0.043]; government[0.043]; Sukarnoputri[0.042]; leader[0.042]; Indonesian[0.041]; Jakarta[0.041]; Jakarta[0.041]; national[0.041]; Priok[0.041]; ousted[0.041]; ousted[0.041]; court[0.041]; son[0.041]; Philippines[0.041]; conflict[0.040]; Megawati[0.040]; Megawati[0.040]; sued[0.040]; Thursday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jakarta\u001b[39m ==> ENTITY: \u001b[32mJakarta\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.049]; Indonesia[0.049]; Indonesia[0.049]; Jakarta[0.046]; Philippines[0.044]; Indonesian[0.043]; region[0.041]; national[0.041]; southern[0.041]; Timor[0.041]; Timor[0.041]; government[0.041]; government[0.041]; government[0.041]; Thursday[0.040]; Priok[0.040]; South[0.040]; Philippine[0.039]; Putra[0.039]; Friday[0.039]; Tanjung[0.039]; Central[0.039]; Korea[0.039]; President[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Philippine ==> ENTITY: \u001b[32mPhilippines\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.175:0.175[0]; log p(e|m)= -0.567:-0.567[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhilippines[0.056]; country[0.047]; national[0.046]; Indonesia[0.045]; Indonesia[0.045]; Indonesia[0.045]; Indonesian[0.044]; southern[0.044]; region[0.044]; Moro[0.044]; institutions[0.043]; vehicles[0.043]; Korea[0.043]; state[0.042]; jointly[0.042]; government[0.042]; Thursday[0.041]; Corp[0.041]; National[0.041]; social[0.041]; Republika[0.041]; Timor[0.040]; Timor[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = PDI ==> ENTITY: \u001b[32mIndonesian Democratic Party\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.151:0.151[0]; log p(e|m)= -1.178:-1.178[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSukarnoputri[0.045]; Media[0.044]; leader[0.043]; President[0.043]; government[0.043]; government[0.043]; Central[0.042]; party[0.042]; national[0.042]; parties[0.042]; Megawati[0.042]; Megawati[0.042]; region[0.042]; Party[0.041]; District[0.041]; Corp[0.041]; started[0.041]; sued[0.040]; court[0.040]; Republika[0.040]; failed[0.040]; Indonesia[0.040]; Indonesia[0.040]; Indonesia[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.098:0.098[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReuters[0.044]; business[0.044]; newspapers[0.043]; newspapers[0.043]; Press[0.043]; major[0.043]; political[0.043]; southern[0.043]; leading[0.042]; country[0.042]; institutions[0.042]; government[0.041]; social[0.041]; direct[0.041]; Philippines[0.041]; Digest[0.041]; efforts[0.040]; Post[0.040]; stories[0.040]; stories[0.040]; share[0.040]; August[0.040]; exports[0.040]; Kompas[0.040]; \t\n\r [===============>..............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s229ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1704/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1019testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Berlin ==> ENTITY: \u001b[32mBerlin\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.154:-0.154[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBerlin[0.049]; studied[0.048]; German[0.045]; German[0.045]; Friday[0.043]; Monday[0.043]; near[0.043]; running[0.042]; male[0.041]; Velten[0.041]; prime[0.041]; old[0.041]; child[0.040]; child[0.040]; child[0.040]; Netherlands[0.040]; couple[0.040]; took[0.040]; said[0.040]; said[0.040]; Public[0.040]; police[0.040]; police[0.040]; sex[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Netherlands ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.081:0.081[0]; log p(e|m)= -0.250:-0.250[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnames[0.045]; sex[0.045]; list[0.043]; list[0.043]; pair[0.043]; couple[0.043]; involved[0.041]; prostitution[0.041]; prostitution[0.041]; Berlin[0.041]; Berlin[0.041]; case[0.041]; case[0.041]; links[0.041]; Monday[0.041]; Reuters[0.041]; German[0.041]; German[0.041]; took[0.041]; missing[0.041]; previous[0.041]; Friday[0.041]; possible[0.040]; Public[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.096:0.096[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntold[0.045]; links[0.044]; prime[0.043]; said[0.043]; said[0.043]; Monday[0.043]; involved[0.042]; spokesman[0.042]; running[0.042]; case[0.042]; Friday[0.042]; prosecutors[0.042]; prosecutors[0.042]; list[0.041]; list[0.041]; possible[0.041]; previous[0.041]; Velten[0.041]; seen[0.041]; abducted[0.041]; names[0.040]; child[0.040]; child[0.040]; Public[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.133:0.133[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGerman[0.049]; Berlin[0.047]; Berlin[0.047]; children[0.043]; studied[0.043]; Netherlands[0.043]; took[0.042]; child[0.042]; child[0.042]; child[0.042]; girl[0.041]; aunt[0.041]; old[0.041]; couple[0.041]; near[0.040]; names[0.040]; prosecutors[0.040]; involved[0.040]; Monday[0.040]; said[0.039]; said[0.039]; missing[0.039]; Nicole[0.039]; ages[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Berlin ==> ENTITY: \u001b[32mBerlin\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.154:-0.154[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBerlin[0.049]; studied[0.048]; German[0.045]; German[0.045]; Friday[0.043]; Monday[0.043]; near[0.043]; prime[0.042]; running[0.042]; Velten[0.041]; Public[0.041]; couple[0.041]; seen[0.040]; Netherlands[0.040]; took[0.040]; old[0.040]; said[0.040]; said[0.040]; ages[0.040]; police[0.040]; police[0.040]; Police[0.040]; previous[0.040]; involved[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Velten\u001b[39m ==> ENTITY: \u001b[32mVelten\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnear[0.044]; school[0.044]; Friday[0.044]; Monday[0.044]; Berlin[0.044]; Berlin[0.044]; Netherlands[0.043]; seen[0.042]; running[0.042]; case[0.042]; case[0.042]; old[0.041]; said[0.041]; said[0.041]; took[0.041]; couple[0.041]; prosecutors[0.041]; prosecutors[0.041]; German[0.040]; German[0.040]; possible[0.040]; studied[0.040]; link[0.040]; links[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.126:0.126[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGerman[0.049]; Berlin[0.047]; Berlin[0.047]; children[0.043]; Netherlands[0.043]; took[0.042]; child[0.042]; child[0.042]; girl[0.041]; aunt[0.041]; old[0.041]; couple[0.041]; said[0.040]; said[0.040]; near[0.040]; prosecutors[0.040]; police[0.040]; police[0.040]; names[0.040]; seen[0.040]; involved[0.040]; case[0.039]; case[0.039]; missing[0.039]; \t\n\r [===============>..............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s239ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1711/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 992testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chechens ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.968:-0.968[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.052]; Chechen[0.048]; military[0.047]; Russian[0.046]; Russian[0.046]; Russia[0.045]; political[0.045]; war[0.045]; Maskhadov[0.043]; Maskhadov[0.043]; statement[0.043]; said[0.042]; signed[0.042]; signed[0.042]; Federation[0.042]; relations[0.042]; rebel[0.041]; leader[0.041]; independent[0.041]; Republic[0.041]; gave[0.041]; peacemaker[0.041]; Alexander[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.050]; Chechnya[0.050]; Chechnya[0.050]; Chechnya[0.050]; troops[0.044]; troops[0.044]; Russia[0.043]; President[0.043]; political[0.043]; war[0.043]; backed[0.043]; Moscow[0.043]; peace[0.042]; Maskhadov[0.042]; border[0.042]; independence[0.042]; unspecified[0.042]; eastern[0.041]; failed[0.041]; thousands[0.041]; said[0.041]; said[0.041]; said[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lebed\u001b[39m ==> ENTITY: \u001b[32mAlexander Lebed\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebed[0.043]; Lebed[0.043]; Lebed[0.043]; Lebed[0.043]; Russian[0.043]; Russian[0.043]; military[0.042]; President[0.042]; Russia[0.042]; Russia[0.042]; Federation[0.041]; Chechnya[0.041]; Chechnya[0.041]; Chechnya[0.041]; Chechnya[0.041]; Chechnya[0.041]; Maskhadov[0.041]; Maskhadov[0.041]; Maskhadov[0.041]; Moscow[0.041]; political[0.040]; political[0.040]; Chechen[0.040]; Khasavyurt[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Russian Federation\u001b[39m ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.049]; Russian[0.048]; military[0.046]; Chechnya[0.045]; Chechnya[0.045]; Chechnya[0.045]; Republic[0.044]; signed[0.044]; signed[0.044]; ceremony[0.043]; leader[0.043]; Chechens[0.043]; Chechen[0.043]; eastern[0.043]; Alexander[0.043]; relations[0.042]; relations[0.042]; Lebed[0.042]; Lebed[0.042]; Lebed[0.042]; Lebed[0.042]; Lebed[0.042]; Lebed[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alexander Lebed\u001b[39m ==> ENTITY: \u001b[32mAlexander Lebed\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebed[0.044]; Lebed[0.044]; Lebed[0.044]; Lebed[0.044]; Russian[0.043]; Russian[0.043]; military[0.043]; Russia[0.042]; Federation[0.042]; Chechnya[0.042]; Chechnya[0.042]; Maskhadov[0.042]; Maskhadov[0.042]; Maskhadov[0.042]; political[0.041]; political[0.041]; Chechen[0.041]; Khasavyurt[0.040]; Chechens[0.040]; leader[0.040]; signed[0.040]; signed[0.040]; war[0.039]; signing[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.046]; win[0.044]; Boris[0.043]; Viktor[0.042]; outside[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; war[0.042]; war[0.042]; Friday[0.042]; took[0.041]; unspecified[0.041]; troops[0.041]; troops[0.041]; Republic[0.041]; Chechnya[0.041]; Chechnya[0.041]; Chechnya[0.041]; Chechnya[0.041]; Chechnya[0.041]; suffered[0.041]; late[0.041]; late[0.041]; powers[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Viktor Chernomyrdin\u001b[39m ==> ENTITY: \u001b[32mViktor Chernomyrdin\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.046]; Russia[0.046]; Moscow[0.045]; Yeltsin[0.044]; Yeltsin[0.044]; Boris[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Prime[0.042]; Minister[0.042]; independence[0.041]; ceremony[0.041]; talks[0.040]; late[0.040]; late[0.040]; said[0.040]; said[0.040]; Lebed[0.040]; Lebed[0.040]; Lebed[0.040]; backed[0.040]; failed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.047]; Chechnya[0.047]; Chechnya[0.047]; Chechnya[0.047]; Chechnya[0.047]; Chechen[0.043]; military[0.042]; Russian[0.041]; Russian[0.041]; Republic[0.041]; troops[0.041]; troops[0.041]; Russia[0.040]; President[0.040]; political[0.040]; war[0.040]; war[0.040]; Moscow[0.040]; peace[0.040]; Maskhadov[0.039]; Maskhadov[0.039]; Maskhadov[0.039]; border[0.039]; independence[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lebed\u001b[39m ==> ENTITY: \u001b[32mAlexander Lebed\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebed[0.045]; Lebed[0.045]; Lebed[0.045]; Lebed[0.045]; Lebed[0.045]; Russian[0.045]; Russian[0.045]; military[0.044]; Russia[0.043]; Federation[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Maskhadov[0.043]; Maskhadov[0.043]; Maskhadov[0.043]; Moscow[0.043]; political[0.042]; political[0.042]; Chechen[0.042]; Khasavyurt[0.042]; Chechens[0.042]; leader[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Aslan Maskhadov\u001b[39m ==> ENTITY: \u001b[32mAslan Maskhadov\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.046]; Chechnya[0.046]; military[0.046]; leader[0.046]; Maskhadov[0.046]; Maskhadov[0.046]; Chechen[0.045]; Chechens[0.044]; relations[0.043]; Khasavyurt[0.043]; signed[0.043]; signed[0.043]; political[0.043]; political[0.043]; December[0.042]; war[0.042]; agreed[0.042]; agreed[0.042]; signing[0.042]; said[0.042]; said[0.042]; Federation[0.042]; peacemaker[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Boris Yeltsin\u001b[39m ==> ENTITY: \u001b[32mBoris Yeltsin\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.047]; Russia[0.045]; Moscow[0.045]; Yeltsin[0.045]; Chernomyrdin[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Viktor[0.042]; unspecified[0.041]; backed[0.041]; Maskhadov[0.041]; sent[0.040]; Lebed[0.040]; Lebed[0.040]; Lebed[0.040]; agreed[0.040]; gave[0.040]; signing[0.040]; relations[0.040]; late[0.040]; late[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Khasavyurt\u001b[39m ==> ENTITY: \u001b[32mKhasavyurt\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechens[0.054]; Chechnya[0.052]; Russian[0.052]; Russian[0.052]; Russia[0.052]; Chechen[0.051]; Republic[0.051]; framework[0.050]; signed[0.050]; signed[0.050]; relations[0.049]; Federation[0.049]; military[0.049]; rebel[0.049]; Maskhadov[0.048]; Maskhadov[0.048]; Maskhadov[0.048]; December[0.048]; independent[0.048]; documents[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.052]; Chechnya[0.052]; Chechnya[0.052]; Chechnya[0.052]; troops[0.045]; troops[0.045]; Russia[0.045]; President[0.045]; political[0.045]; war[0.045]; war[0.045]; backed[0.044]; Moscow[0.044]; peace[0.044]; Maskhadov[0.044]; Maskhadov[0.044]; border[0.044]; independence[0.044]; unspecified[0.044]; eastern[0.043]; failed[0.042]; thousands[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lebed\u001b[39m ==> ENTITY: \u001b[32mAlexander Lebed\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebed[0.043]; Lebed[0.043]; Lebed[0.043]; Lebed[0.043]; Lebed[0.043]; Russian[0.043]; Russian[0.043]; military[0.042]; Russia[0.041]; Russia[0.041]; Federation[0.041]; Chechnya[0.041]; Chechnya[0.041]; Chechnya[0.041]; Chechnya[0.041]; Chechnya[0.041]; Maskhadov[0.041]; Maskhadov[0.041]; Maskhadov[0.041]; Moscow[0.041]; political[0.040]; political[0.040]; Chechen[0.040]; Khasavyurt[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.051]; Chechen[0.047]; Chechens[0.047]; military[0.046]; Russian[0.045]; Russian[0.045]; Republic[0.044]; Russia[0.044]; political[0.044]; political[0.044]; war[0.043]; Maskhadov[0.043]; Maskhadov[0.043]; Maskhadov[0.043]; Khasavyurt[0.043]; statement[0.042]; December[0.042]; said[0.041]; said[0.041]; said[0.041]; signed[0.041]; signed[0.041]; Federation[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yeltsin\u001b[39m ==> ENTITY: \u001b[32mBoris Yeltsin\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.047]; Russia[0.045]; Moscow[0.045]; Yeltsin[0.045]; Chernomyrdin[0.042]; Boris[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Viktor[0.042]; unspecified[0.041]; backed[0.041]; sent[0.040]; Lebed[0.040]; Lebed[0.040]; gave[0.040]; signing[0.040]; late[0.040]; ceremony[0.040]; independence[0.040]; took[0.040]; ordered[0.040]; failed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lebed\u001b[39m ==> ENTITY: \u001b[32mAlexander Lebed\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebed[0.045]; Lebed[0.045]; Lebed[0.045]; Lebed[0.045]; Russian[0.045]; Yeltsin[0.044]; President[0.044]; Russia[0.044]; Federation[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Maskhadov[0.043]; Maskhadov[0.043]; Moscow[0.043]; political[0.042]; Chechen[0.042]; Boris[0.042]; troops[0.041]; troops[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lebed\u001b[39m ==> ENTITY: \u001b[32mAlexander Lebed\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebed[0.045]; Lebed[0.045]; Yeltsin[0.043]; Yeltsin[0.043]; President[0.043]; Russia[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chernomyrdin[0.043]; Moscow[0.042]; political[0.041]; Boris[0.041]; troops[0.041]; troops[0.041]; war[0.040]; signing[0.040]; backed[0.040]; took[0.040]; quell[0.040]; Viktor[0.040]; relations[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Maskhadov\u001b[39m ==> ENTITY: \u001b[32mAslan Maskhadov\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.045]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; military[0.043]; leader[0.043]; Maskhadov[0.043]; Maskhadov[0.043]; Chechen[0.042]; troops[0.041]; troops[0.041]; relations[0.041]; relations[0.041]; Khasavyurt[0.040]; ceremony[0.040]; signed[0.040]; signed[0.040]; political[0.040]; political[0.040]; December[0.040]; war[0.040]; war[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechen Republic\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.049]; Chechnya[0.049]; Chechnya[0.049]; Chechens[0.044]; military[0.043]; political[0.042]; political[0.042]; Russia[0.042]; war[0.041]; war[0.041]; Maskhadov[0.040]; Maskhadov[0.040]; Maskhadov[0.040]; statement[0.040]; border[0.040]; Moscow[0.040]; ceremony[0.040]; eastern[0.040]; Khasavyurt[0.040]; Russian[0.040]; Russian[0.040]; said[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lebed\u001b[39m ==> ENTITY: \u001b[32mAlexander Lebed\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebed[0.045]; Lebed[0.045]; Lebed[0.045]; Russian[0.044]; Russian[0.044]; military[0.043]; Russia[0.043]; Federation[0.043]; Chechnya[0.043]; Maskhadov[0.043]; Maskhadov[0.043]; political[0.041]; Chechen[0.041]; Khasavyurt[0.041]; Chechens[0.041]; leader[0.040]; signed[0.040]; signed[0.040]; war[0.040]; signing[0.040]; December[0.040]; rebel[0.039]; independent[0.039]; relations[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFederation[0.046]; Russian[0.046]; Russian[0.046]; Chechnya[0.044]; Republic[0.043]; signed[0.042]; signed[0.042]; Chechens[0.042]; Alexander[0.041]; Chechen[0.041]; agreed[0.041]; agreed[0.041]; statement[0.041]; military[0.040]; said[0.040]; said[0.040]; war[0.040]; relations[0.040]; Lebed[0.040]; Lebed[0.040]; Lebed[0.040]; Lebed[0.040]; Lebed[0.040]; decision[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.119:0.119[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.045]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; President[0.043]; defeats[0.042]; win[0.042]; Boris[0.042]; eastern[0.042]; Viktor[0.041]; failed[0.041]; ceremony[0.041]; late[0.041]; late[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; agreed[0.041]; bid[0.040]; proposals[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lebed\u001b[39m ==> ENTITY: \u001b[32mAlexander Lebed\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebed[0.045]; Yeltsin[0.044]; Yeltsin[0.044]; President[0.043]; Russia[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chernomyrdin[0.043]; Moscow[0.043]; Boris[0.041]; troops[0.041]; troops[0.041]; war[0.040]; backed[0.040]; took[0.040]; quell[0.040]; Viktor[0.040]; sent[0.040]; independence[0.040]; said[0.039]; people[0.039]; unspecified[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.047]; Chechnya[0.047]; Chechnya[0.047]; Chechnya[0.047]; Chechen[0.044]; Russian[0.042]; Republic[0.041]; troops[0.041]; troops[0.041]; Russia[0.041]; President[0.041]; political[0.041]; war[0.041]; war[0.041]; backed[0.040]; Moscow[0.040]; peace[0.040]; Maskhadov[0.040]; Maskhadov[0.040]; border[0.040]; independence[0.040]; unspecified[0.040]; statement[0.039]; eastern[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Maskhadov\u001b[39m ==> ENTITY: \u001b[32mAslan Maskhadov\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; military[0.045]; leader[0.045]; Maskhadov[0.045]; Maskhadov[0.045]; Chechen[0.044]; troops[0.043]; Chechens[0.043]; relations[0.043]; relations[0.043]; Khasavyurt[0.042]; ceremony[0.042]; signed[0.042]; signed[0.042]; political[0.042]; political[0.042]; December[0.042]; war[0.042]; war[0.042]; sent[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.048]; Chechnya[0.048]; Chechnya[0.048]; Chechnya[0.048]; troops[0.042]; troops[0.042]; Russia[0.042]; President[0.042]; political[0.042]; war[0.041]; backed[0.041]; Moscow[0.041]; peace[0.041]; border[0.041]; independence[0.041]; unspecified[0.040]; eastern[0.039]; failed[0.039]; thousands[0.039]; said[0.039]; said[0.039]; Yeltsin[0.039]; Yeltsin[0.039]; relations[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.146:0.146[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.053]; Russia[0.052]; Federation[0.050]; Chechnya[0.047]; Chechnya[0.047]; Chechens[0.047]; Republic[0.046]; Alexander[0.046]; Chechen[0.045]; Lebed[0.044]; Lebed[0.044]; Lebed[0.044]; Lebed[0.044]; Lebed[0.044]; military[0.044]; war[0.044]; relations[0.044]; gave[0.043]; signed[0.043]; signed[0.043]; December[0.043]; leader[0.043]; \t\n\r [===============>..............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s138ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1739/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1151testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mRochester\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRochester, New York\u001b[39m <---> \u001b[32mRochester, New Hampshire\u001b[39m\t\nSCORES: global= 0.264:0.234[\u001b[31m0.030\u001b[39m]; local()= 0.105:0.132[\u001b[32m0.026\u001b[39m]; log p(e|m)= -1.094:-2.865[\u001b[31m1.771\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRochester[0.046]; Rochester[0.046]; Hampshire[0.045]; Boston[0.043]; motel[0.043]; Department[0.042]; people[0.042]; people[0.042]; television[0.041]; Don[0.041]; Friday[0.041]; saw[0.041]; far[0.041]; gasoline[0.041]; gasoline[0.041]; truck[0.041]; officials[0.041]; officials[0.041]; trucks[0.041]; stations[0.041]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mRochester\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRochester, New York\u001b[39m <---> \u001b[32mRochester, New Hampshire\u001b[39m\t\nSCORES: global= 0.264:0.234[\u001b[31m0.030\u001b[39m]; local()= 0.105:0.132[\u001b[32m0.026\u001b[39m]; log p(e|m)= -1.094:-2.865[\u001b[31m1.771\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRochester[0.046]; Rochester[0.046]; Hampshire[0.045]; Boston[0.043]; motel[0.043]; Department[0.042]; people[0.042]; people[0.042]; television[0.041]; Don[0.041]; Friday[0.041]; saw[0.041]; far[0.041]; gasoline[0.041]; gasoline[0.041]; truck[0.041]; officials[0.041]; officials[0.041]; trucks[0.041]; stations[0.041]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Boston ==> ENTITY: \u001b[32mBoston\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.073:0.073[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHampshire[0.045]; officials[0.043]; officials[0.043]; officials[0.043]; shortly[0.043]; Rochester[0.043]; Rochester[0.043]; Rochester[0.043]; Officials[0.043]; Friday[0.042]; stations[0.042]; Don[0.042]; television[0.041]; got[0.041]; involved[0.041]; people[0.041]; people[0.041]; Department[0.040]; immediately[0.040]; Lilac[0.040]; Lilac[0.040]; details[0.040]; parked[0.040]; Local[0.040]; \t\n\r [===============>..............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s181ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1743/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1144testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jerusalem ==> ENTITY: \u001b[32mJerusalem\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.174:0.174[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.047]; Israel[0.047]; Israel[0.047]; Lebanon[0.045]; Lebanon[0.045]; Lebanon[0.045]; Lebanon[0.045]; Israeli[0.044]; Israeli[0.044]; Islamic[0.042]; Lebanese[0.042]; Syria[0.041]; following[0.040]; Sunday[0.039]; Hizbollah[0.039]; Hizbollah[0.039]; France[0.039]; south[0.039]; south[0.039]; residents[0.039]; army[0.038]; discuss[0.038]; agreement[0.038]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.243:0.243[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.049]; Israel[0.049]; Israel[0.049]; Israeli[0.045]; Israeli[0.045]; Israeli[0.045]; Lebanon[0.043]; Lebanon[0.043]; Lebanon[0.043]; Lebanon[0.043]; Syria[0.043]; Jewish[0.042]; God[0.041]; army[0.040]; army[0.040]; nations[0.040]; Lebanese[0.039]; Hizbollah[0.039]; Hizbollah[0.039]; following[0.037]; war[0.037]; allies[0.037]; States[0.037]; civilian[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = GMT\u001b[39m ==> ENTITY: \u001b[32mGreenwich Mean Time\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.108:0.108[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.043]; Sunday[0.043]; States[0.043]; days[0.042]; following[0.042]; Lebanon[0.042]; Lebanon[0.042]; Lebanon[0.042]; Lebanon[0.042]; April[0.042]; making[0.042]; set[0.042]; Saturday[0.042]; United[0.041]; August[0.041]; Israel[0.041]; Israel[0.041]; Israel[0.041]; Israel[0.041]; Representatives[0.041]; monitoring[0.040]; monitoring[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.242:0.242[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israel[0.050]; Israel[0.050]; Israeli[0.045]; Israeli[0.045]; Israeli[0.045]; Lebanon[0.044]; Lebanon[0.044]; Lebanon[0.044]; Lebanon[0.044]; Syria[0.043]; Jerusalem[0.041]; nations[0.040]; Lebanese[0.040]; Hizbollah[0.039]; Hizbollah[0.039]; agreement[0.038]; discuss[0.038]; following[0.038]; States[0.037]; France[0.037]; Saturday[0.037]; Sunday[0.037]; Representatives[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Party of God\u001b[39m ==> ENTITY: \u001b[32mHezbollah\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.048]; Israel[0.048]; Lebanon[0.048]; Lebanon[0.048]; Hizbollah[0.047]; Israeli[0.046]; Israeli[0.046]; Israeli[0.046]; militia[0.040]; attacks[0.040]; attacks[0.040]; guerrilla[0.040]; guerrilla[0.040]; army[0.039]; Jewish[0.039]; allies[0.039]; guerrillas[0.039]; war[0.039]; troops[0.039]; troops[0.039]; civilian[0.038]; occupation[0.038]; gunmen[0.038]; northern[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.232:0.232[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israel[0.050]; Israel[0.050]; Israeli[0.045]; Israeli[0.045]; Lebanon[0.045]; Lebanon[0.045]; Lebanon[0.045]; Lebanon[0.045]; Jerusalem[0.042]; Syria[0.042]; army[0.041]; army[0.041]; Lebanese[0.040]; Hizbollah[0.039]; Hizbollah[0.039]; following[0.038]; south[0.037]; south[0.037]; said[0.037]; said[0.037]; ceasefire[0.037]; ceasefire[0.037]; Islamic[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanese ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.289:0.289[0]; log p(e|m)= -0.559:-0.559[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.049]; Lebanon[0.049]; Lebanon[0.049]; Lebanon[0.049]; Lebanon[0.049]; Lebanon[0.049]; Syria[0.046]; Israel[0.045]; Israel[0.045]; Israel[0.045]; Israel[0.045]; Jewish[0.042]; Israeli[0.041]; Israeli[0.041]; Israeli[0.041]; Israeli[0.041]; Israeli[0.041]; Jerusalem[0.039]; Hizbollah[0.039]; Hizbollah[0.039]; occupation[0.039]; Islamic[0.039]; local[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanon ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.206:0.206[0]; log p(e|m)= -0.409:-0.409[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.051]; Lebanese[0.046]; Israel[0.046]; Israel[0.046]; Israel[0.046]; Hizbollah[0.044]; Hizbollah[0.044]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; northern[0.041]; occupation[0.040]; troops[0.040]; troops[0.040]; ceasefire[0.040]; allies[0.039]; war[0.039]; militia[0.039]; Jewish[0.039]; April[0.039]; army[0.039]; following[0.039]; area[0.039]; August[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N. Interim Force in Lebanon\u001b[39m ==> ENTITY: \u001b[32mUnited Nations Interim Force in Lebanon\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.044]; Lebanon[0.044]; Lebanon[0.044]; Lebanon[0.044]; Lebanon[0.044]; Lebanese[0.042]; Israel[0.042]; Israel[0.042]; Israel[0.042]; Israel[0.042]; Israeli[0.042]; Israeli[0.042]; Israeli[0.042]; Israeli[0.042]; Israeli[0.042]; civilian[0.042]; troops[0.041]; troops[0.041]; Hizbollah[0.040]; Hizbollah[0.040]; Syria[0.040]; patrol[0.039]; army[0.039]; army[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syria ==> ENTITY: \u001b[32mSyria\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.225:0.225[0]; log p(e|m)= -0.261:-0.261[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.046]; Israel[0.046]; Israel[0.046]; Israel[0.046]; Lebanon[0.046]; Lebanon[0.046]; Lebanon[0.046]; Lebanon[0.046]; Israeli[0.042]; Israeli[0.042]; Israeli[0.042]; Lebanese[0.041]; army[0.040]; army[0.040]; Islamic[0.040]; Hizbollah[0.040]; Hizbollah[0.040]; civilian[0.039]; ceasefire[0.038]; ceasefire[0.038]; agreement[0.038]; Jerusalem[0.038]; south[0.037]; south[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Islamic ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.092:0.092[0]; log p(e|m)= -0.212:-0.212[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.044]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Lebanon[0.043]; Lebanon[0.043]; Lebanon[0.043]; Lebanon[0.043]; Lebanese[0.043]; Hizbollah[0.042]; Hizbollah[0.042]; Jerusalem[0.042]; nations[0.041]; days[0.041]; Israeli[0.041]; Israeli[0.041]; Israeli[0.041]; States[0.041]; group[0.040]; said[0.040]; said[0.040]; monitoring[0.039]; monitoring[0.039]; guerrillas[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.210:0.210[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.051]; Israel[0.051]; Israel[0.051]; Israeli[0.046]; Israeli[0.046]; Jewish[0.046]; Lebanon[0.045]; Lebanon[0.045]; Lebanon[0.045]; army[0.042]; Lebanese[0.040]; area[0.039]; Hizbollah[0.039]; Hizbollah[0.039]; northern[0.038]; following[0.038]; attacks[0.038]; attacks[0.038]; occupation[0.038]; troops[0.038]; troops[0.038]; militia[0.037]; guerrilla[0.037]; guerrilla[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.233:0.233[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.052]; Israel[0.052]; Israeli[0.048]; Israeli[0.048]; Israeli[0.048]; Lebanon[0.046]; Lebanon[0.046]; Lebanon[0.046]; Lebanon[0.046]; Syria[0.045]; Jerusalem[0.043]; nations[0.042]; Lebanese[0.042]; Hizbollah[0.041]; Hizbollah[0.041]; agreement[0.040]; discuss[0.040]; following[0.040]; States[0.039]; France[0.039]; Saturday[0.039]; Sunday[0.039]; Representatives[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.169:0.169[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.053]; Israeli[0.048]; Israeli[0.048]; Lebanon[0.047]; Lebanon[0.047]; Jewish[0.046]; God[0.044]; Hizbollah[0.042]; state[0.041]; civilian[0.040]; troops[0.040]; troops[0.040]; occupation[0.040]; northern[0.040]; forbid[0.039]; war[0.038]; guerrilla[0.038]; guerrilla[0.038]; understandings[0.038]; rule[0.038]; ended[0.038]; April[0.038]; militia[0.038]; area[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hizbollah\u001b[39m ==> ENTITY: \u001b[32mHezbollah\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.245:0.245[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.045]; Israel[0.045]; Israel[0.045]; Israel[0.045]; Lebanon[0.043]; Lebanon[0.043]; Lebanon[0.043]; Lebanon[0.043]; Lebanon[0.043]; Lebanon[0.043]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; Hizbollah[0.042]; Hizbollah[0.042]; Syria[0.042]; Lebanese[0.042]; Islamic[0.037]; attacks[0.037]; attacks[0.037]; militia[0.036]; group[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.082:0.082[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnations[0.045]; group[0.044]; said[0.043]; said[0.043]; Islamic[0.042]; Sunday[0.042]; Saturday[0.042]; France[0.042]; Lebanon[0.042]; Lebanon[0.042]; Lebanon[0.042]; Lebanon[0.042]; August[0.041]; Israel[0.041]; Israel[0.041]; Israel[0.041]; Israel[0.041]; following[0.041]; days[0.041]; April[0.041]; ended[0.040]; Jerusalem[0.040]; Lebanese[0.040]; Representatives[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.232:0.232[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israel[0.050]; Israel[0.050]; Israeli[0.045]; Israeli[0.045]; Lebanon[0.045]; Lebanon[0.045]; Lebanon[0.045]; Lebanon[0.045]; Jerusalem[0.042]; Syria[0.042]; army[0.041]; army[0.041]; Lebanese[0.040]; Hizbollah[0.039]; Hizbollah[0.039]; following[0.038]; south[0.037]; south[0.037]; said[0.037]; said[0.037]; ceasefire[0.037]; ceasefire[0.037]; Islamic[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jewish ==> ENTITY: \u001b[32mJews\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.805:-0.805[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.045]; Israel[0.045]; Israel[0.045]; Israeli[0.044]; Israeli[0.044]; Israeli[0.044]; Lebanon[0.043]; Lebanon[0.043]; Lebanese[0.042]; God[0.042]; area[0.042]; occupation[0.041]; war[0.041]; northern[0.041]; rule[0.040]; ended[0.040]; local[0.040]; forbid[0.040]; said[0.039]; residents[0.039]; south[0.039]; south[0.039]; state[0.039]; village[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = UNIFIL\u001b[39m ==> ENTITY: \u001b[32mUnited Nations Interim Force in Lebanon\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.044]; Lebanon[0.044]; Lebanon[0.044]; Lebanon[0.044]; Lebanon[0.044]; Lebanese[0.042]; Israel[0.042]; Israel[0.042]; Israel[0.042]; Israel[0.042]; Israeli[0.042]; Israeli[0.042]; Israeli[0.042]; Israeli[0.042]; Israeli[0.042]; civilian[0.042]; troops[0.041]; troops[0.041]; Hizbollah[0.040]; Hizbollah[0.040]; Syria[0.040]; patrol[0.039]; army[0.039]; army[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanon ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.220:0.220[0]; log p(e|m)= -0.409:-0.409[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.050]; Lebanon[0.050]; Lebanese[0.046]; Israel[0.045]; Israel[0.045]; Israel[0.045]; Hizbollah[0.043]; Hizbollah[0.043]; Israeli[0.042]; Israeli[0.042]; Israeli[0.042]; northern[0.040]; occupation[0.040]; troops[0.040]; troops[0.040]; ceasefire[0.039]; allies[0.039]; war[0.039]; militia[0.038]; Jewish[0.038]; April[0.038]; army[0.038]; following[0.038]; area[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Naqoura\u001b[39m ==> ENTITY: \u001b[32mNaqoura\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.047]; Lebanon[0.047]; Lebanon[0.047]; Lebanon[0.047]; Lebanese[0.043]; Hizbollah[0.042]; Hizbollah[0.042]; Israel[0.042]; Israel[0.042]; Israel[0.042]; Israel[0.042]; coastal[0.041]; Israeli[0.041]; Israeli[0.041]; Israeli[0.041]; Israeli[0.041]; Syria[0.040]; troops[0.040]; April[0.040]; south[0.039]; south[0.039]; Force[0.039]; August[0.039]; United[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanon ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.250:0.250[0]; log p(e|m)= -0.409:-0.409[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.049]; Lebanon[0.049]; Lebanon[0.049]; Syria[0.045]; Lebanese[0.044]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Israel[0.044]; Hizbollah[0.042]; Hizbollah[0.042]; Israeli[0.041]; Israeli[0.041]; Israeli[0.041]; Islamic[0.039]; ceasefire[0.038]; ceasefire[0.038]; Jerusalem[0.038]; France[0.038]; April[0.037]; army[0.037]; army[0.037]; following[0.037]; nations[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanon ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.236:0.236[0]; log p(e|m)= -0.409:-0.409[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.050]; Lebanon[0.050]; Lebanon[0.050]; Syria[0.046]; Lebanese[0.045]; Israel[0.045]; Israel[0.045]; Israel[0.045]; Hizbollah[0.043]; Hizbollah[0.043]; Israeli[0.042]; Israeli[0.042]; Islamic[0.039]; ceasefire[0.039]; Jerusalem[0.039]; France[0.038]; army[0.038]; following[0.038]; nations[0.038]; Sunday[0.038]; Naqoura[0.038]; residents[0.037]; headquarters[0.037]; States[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hizbollah\u001b[39m ==> ENTITY: \u001b[32mHezbollah\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.047]; Israel[0.047]; Israel[0.047]; Lebanon[0.045]; Lebanon[0.045]; Lebanon[0.045]; Lebanon[0.045]; Israeli[0.045]; Israeli[0.045]; Israeli[0.045]; Hizbollah[0.045]; Syria[0.045]; Lebanese[0.044]; Islamic[0.040]; group[0.038]; Jerusalem[0.038]; army[0.037]; army[0.037]; ceasefire[0.037]; guerrillas[0.037]; following[0.037]; Ceasefire[0.036]; said[0.036]; said[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.264:0.264[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.048]; Israel[0.048]; Israel[0.048]; Israel[0.048]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; Jewish[0.043]; Lebanon[0.043]; Lebanon[0.043]; Lebanon[0.043]; Lebanon[0.043]; Lebanon[0.043]; army[0.040]; army[0.040]; Syria[0.040]; Lebanese[0.038]; Hizbollah[0.038]; Hizbollah[0.038]; Hizbollah[0.038]; civilian[0.036]; troops[0.036]; troops[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hizbollah\u001b[39m ==> ENTITY: \u001b[32mHezbollah\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.049]; Israel[0.049]; Lebanon[0.047]; Lebanon[0.047]; Israeli[0.047]; Israeli[0.047]; Israeli[0.047]; Hizbollah[0.046]; attacks[0.041]; attacks[0.041]; militia[0.040]; Jewish[0.039]; troops[0.039]; troops[0.039]; guerrilla[0.039]; guerrilla[0.039]; civilian[0.039]; army[0.039]; ceasefire[0.039]; war[0.038]; allies[0.038]; guerrillas[0.038]; said[0.037]; occupation[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.277:0.277[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.048]; Israel[0.048]; Israel[0.048]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; Lebanon[0.042]; Lebanon[0.042]; Lebanon[0.042]; Lebanon[0.042]; Lebanon[0.042]; Lebanon[0.042]; Syria[0.041]; Jewish[0.041]; Jerusalem[0.039]; army[0.039]; army[0.039]; nations[0.038]; Lebanese[0.038]; Hizbollah[0.038]; Hizbollah[0.038]; agreement[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.106:0.106[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.044]; Lebanon[0.044]; Lebanon[0.044]; Lebanon[0.044]; army[0.043]; army[0.043]; August[0.043]; States[0.042]; April[0.042]; nations[0.041]; following[0.041]; civilian[0.041]; Saturday[0.041]; south[0.041]; south[0.041]; said[0.041]; said[0.041]; headquarters[0.041]; fighting[0.041]; group[0.040]; days[0.040]; deal[0.040]; coastal[0.040]; Sunday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.210:0.210[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.051]; Israel[0.051]; Israel[0.051]; Israeli[0.046]; Israeli[0.046]; Jewish[0.046]; Lebanon[0.045]; Lebanon[0.045]; Lebanon[0.045]; army[0.042]; Lebanese[0.040]; area[0.039]; Hizbollah[0.039]; Hizbollah[0.039]; northern[0.038]; following[0.038]; attacks[0.038]; attacks[0.038]; occupation[0.038]; troops[0.038]; troops[0.038]; militia[0.037]; guerrilla[0.037]; guerrilla[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lebanon ==> ENTITY: \u001b[32mLebanon\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.242:0.242[0]; log p(e|m)= -0.409:-0.409[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebanon[0.049]; Lebanon[0.049]; Lebanon[0.049]; Syria[0.046]; Lebanese[0.045]; Israel[0.045]; Israel[0.045]; Israel[0.045]; Hizbollah[0.042]; Hizbollah[0.042]; Israeli[0.042]; Israeli[0.042]; Israeli[0.042]; Islamic[0.039]; ceasefire[0.039]; Jerusalem[0.039]; France[0.038]; army[0.038]; army[0.038]; following[0.038]; nations[0.037]; Sunday[0.037]; Naqoura[0.037]; August[0.037]; \t\n\r [================>.............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s152ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1773/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1022testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Helsinki ==> ENTITY: \u001b[32mHelsinki\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.015:0.015[0]; log p(e|m)= -0.015:-0.015[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFim[0.056]; sales[0.054]; Martela[0.054]; June[0.053]; months[0.052]; million[0.051]; Earnings[0.050]; Profit[0.050]; Profit[0.050]; share[0.050]; vs[0.050]; vs[0.050]; vs[0.050]; stated[0.049]; Net[0.048]; H1[0.048]; Vs[0.048]; unless[0.047]; taxes[0.047]; appropriations[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Martela\u001b[39m ==> ENTITY: \u001b[32mMartela\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.045:0.045[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmillion[0.058]; sales[0.058]; months[0.057]; Net[0.055]; stated[0.055]; H1[0.055]; Helsinki[0.053]; appropriations[0.052]; vs[0.052]; vs[0.052]; vs[0.052]; Earnings[0.052]; Fim[0.051]; Profit[0.050]; Profit[0.050]; Vs[0.050]; June[0.050]; share[0.049]; taxes[0.049]; \t\n\r [================>.............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s208ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1775/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 984testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sudanese ==> ENTITY: \u001b[32mSudan\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.198:-0.198[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.047]; Sudanese[0.046]; people[0.045]; people[0.045]; asylum[0.044]; asylum[0.044]; Iraqis[0.044]; Iraqis[0.044]; Iraqi[0.042]; immediate[0.041]; London[0.041]; London[0.041]; seven[0.041]; Hussein[0.040]; forced[0.040]; Saddam[0.040]; regime[0.040]; claimed[0.040]; officials[0.039]; appear[0.039]; said[0.039]; ordinary[0.039]; hostages[0.039]; offered[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mStansted\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mLondon Stansted Airport\u001b[39m <---> \u001b[32mStansted Mountfitchet\u001b[39m\t\nSCORES: global= 0.260:0.249[\u001b[31m0.011\u001b[39m]; local()= 0.156:0.108[\u001b[31m0.048\u001b[39m]; log p(e|m)= -0.267:-1.546[\u001b[31m1.280\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.048]; London[0.048]; Britain[0.046]; airport[0.044]; seven[0.043]; seven[0.043]; seven[0.043]; Monday[0.041]; people[0.041]; people[0.041]; English[0.041]; Saturday[0.041]; Iraq[0.041]; Friday[0.040]; Tuesday[0.040]; jet[0.040]; fly[0.040]; Amman[0.040]; plane[0.040]; plane[0.040]; landed[0.040]; maximum[0.040]; receive[0.040]; considered[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Khartoum ==> ENTITY: \u001b[32mKhartoum\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.040:-0.040[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHussein[0.044]; Amman[0.044]; Iraq[0.043]; aircraft[0.043]; unless[0.043]; regime[0.043]; Monday[0.042]; said[0.042]; explosives[0.042]; plane[0.042]; surrender[0.042]; London[0.042]; shortly[0.042]; sent[0.041]; people[0.041]; following[0.041]; search[0.041]; Iraqis[0.041]; widespread[0.040]; officials[0.040]; life[0.040]; Saddam[0.040]; seven[0.040]; seven[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqis ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.812:-0.812[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.050]; Iraqi[0.047]; Iraqis[0.045]; Britain[0.044]; Hussein[0.042]; Sudanese[0.042]; Sudanese[0.042]; London[0.042]; London[0.042]; Amman[0.042]; asylum[0.042]; asylum[0.042]; asylum[0.042]; Saddam[0.041]; people[0.041]; people[0.041]; Khartoum[0.041]; regime[0.040]; said[0.040]; said[0.040]; persecuted[0.039]; seven[0.039]; seven[0.039]; seven[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.051]; airport[0.049]; Britain[0.048]; plane[0.045]; Stansted[0.045]; jet[0.045]; fly[0.044]; said[0.043]; said[0.043]; Friday[0.043]; Tuesday[0.043]; airliner[0.042]; landed[0.042]; people[0.042]; people[0.042]; English[0.042]; claimed[0.042]; Saturday[0.042]; seven[0.041]; seven[0.041]; ending[0.041]; appear[0.041]; officials[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cyprus ==> ENTITY: \u001b[32mCyprus\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.375:-0.375[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlaw[0.045]; Iraq[0.044]; following[0.043]; ministry[0.043]; widespread[0.042]; sent[0.042]; Monday[0.042]; Amman[0.042]; considered[0.042]; took[0.042]; London[0.042]; allowed[0.041]; Hussein[0.041]; shortly[0.041]; seven[0.041]; seven[0.041]; fake[0.041]; unless[0.041]; asylum[0.041]; English[0.041]; officials[0.041]; said[0.041]; receive[0.040]; surrender[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saddam\u001b[39m ==> ENTITY: \u001b[32mSaddam Hussein\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.049]; Iraqi[0.047]; Hussein[0.046]; regime[0.044]; Iraqis[0.043]; Iraqis[0.043]; Monday[0.042]; officials[0.041]; said[0.041]; said[0.041]; Tuesday[0.041]; life[0.041]; took[0.041]; shortly[0.040]; Britain[0.040]; Amman[0.040]; Friday[0.040]; sent[0.040]; widespread[0.040]; court[0.040]; Khartoum[0.040]; immediate[0.040]; Sudanese[0.040]; Sudanese[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraqis[0.048]; Iraqis[0.048]; Britain[0.046]; Hussein[0.044]; Sudanese[0.044]; Sudanese[0.044]; Saddam[0.043]; people[0.043]; people[0.043]; asylum[0.042]; asylum[0.042]; said[0.042]; jet[0.041]; regime[0.040]; claimed[0.039]; London[0.039]; London[0.039]; airliner[0.039]; plane[0.039]; forced[0.039]; offered[0.039]; fly[0.039]; persecuted[0.039]; speculation[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqis ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.812:-0.812[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraqi[0.048]; Iraqis[0.047]; Britain[0.045]; Hussein[0.043]; Sudanese[0.043]; Sudanese[0.043]; London[0.043]; London[0.043]; asylum[0.043]; asylum[0.043]; Saddam[0.043]; people[0.042]; people[0.042]; regime[0.041]; said[0.041]; said[0.041]; persecuted[0.040]; seven[0.040]; seven[0.040]; officials[0.039]; claimed[0.039]; likely[0.038]; freed[0.038]; hostages[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.061:0.061[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \naircraft[0.049]; plane[0.045]; following[0.044]; Monday[0.043]; shortly[0.043]; said[0.043]; Amman[0.042]; English[0.042]; began[0.041]; seven[0.041]; seven[0.041]; sent[0.041]; stop[0.041]; Iraq[0.041]; officials[0.041]; hijack[0.040]; hijack[0.040]; hijack[0.040]; widespread[0.040]; taken[0.040]; taken[0.040]; explosives[0.040]; considered[0.040]; took[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hussein ==> ENTITY: \u001b[32mSaddam Hussein\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.143:0.143[0]; log p(e|m)= -1.858:-1.858[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.048]; Saddam[0.047]; Iraqi[0.046]; Amman[0.044]; said[0.043]; said[0.043]; regime[0.043]; Iraqis[0.042]; Iraqis[0.042]; people[0.042]; people[0.042]; considered[0.041]; taken[0.041]; taken[0.041]; immediate[0.040]; asylum[0.040]; asylum[0.040]; asylum[0.040]; Britain[0.039]; hostages[0.039]; officials[0.039]; Monday[0.039]; Tuesday[0.039]; Cyprus[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.100:0.100[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.045]; London[0.045]; freed[0.044]; Tuesday[0.043]; considered[0.043]; seized[0.042]; Friday[0.042]; English[0.042]; Saturday[0.042]; seven[0.041]; seven[0.041]; seven[0.041]; appealed[0.041]; said[0.041]; said[0.041]; plane[0.041]; fly[0.041]; law[0.041]; landed[0.041]; aboard[0.041]; claimed[0.041]; jet[0.041]; people[0.040]; people[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.049]; airport[0.047]; Britain[0.046]; plane[0.043]; Stansted[0.043]; jet[0.043]; fly[0.042]; said[0.042]; said[0.042]; Friday[0.041]; Tuesday[0.041]; airliner[0.041]; landed[0.041]; people[0.041]; people[0.041]; claimed[0.040]; Saturday[0.040]; seven[0.040]; seven[0.040]; ending[0.040]; appear[0.039]; officials[0.039]; hijack[0.039]; hijack[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sudanese ==> ENTITY: \u001b[32mSudan\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.110:0.110[0]; log p(e|m)= -0.198:-0.198[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.049]; Sudanese[0.048]; people[0.046]; people[0.046]; asylum[0.045]; asylum[0.045]; asylum[0.045]; Iraqis[0.045]; Iraqis[0.045]; Iraqi[0.044]; immediate[0.042]; London[0.042]; London[0.042]; likely[0.042]; seven[0.042]; seven[0.042]; Hussein[0.042]; forced[0.041]; Saddam[0.041]; regime[0.041]; claimed[0.041]; officials[0.041]; appear[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.129:0.129[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaddam[0.049]; Iraqis[0.047]; Hussein[0.047]; regime[0.045]; Amman[0.044]; Khartoum[0.044]; Cyprus[0.043]; aircraft[0.042]; seven[0.042]; seven[0.042]; following[0.041]; Monday[0.040]; sent[0.040]; shortly[0.040]; officials[0.040]; Tuesday[0.040]; London[0.040]; receive[0.039]; men[0.039]; said[0.039]; Saturday[0.039]; began[0.039]; plane[0.039]; plane[0.039]; \t\n\r [================>.............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 13s164ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1791/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1153testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mKurdish\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mKurdish people\u001b[39m <---> \u001b[32mKurdistan\u001b[39m\t\nSCORES: global= 0.250:0.238[\u001b[31m0.011\u001b[39m]; local()= 0.214:0.241[\u001b[32m0.027\u001b[39m]; log p(e|m)= -0.713:-2.797[\u001b[31m2.084\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdish[0.048]; region[0.046]; northern[0.045]; northern[0.045]; northern[0.045]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; southern[0.044]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iran[0.042]; Iran[0.042]; Arbil[0.042]; troops[0.042]; troops[0.042]; Baghdad[0.042]; forces[0.042]; forces[0.042]; military[0.041]; military[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baghdad\u001b[39m ==> ENTITY: \u001b[32mBaghdad\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.228:0.228[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.046]; Iraq[0.046]; Iraq[0.046]; Iraq[0.046]; city[0.044]; southern[0.043]; Iraqi[0.042]; Iraqi[0.042]; Iraqi[0.042]; northern[0.042]; northern[0.042]; northern[0.042]; region[0.041]; troops[0.040]; troops[0.040]; Kurdish[0.040]; Kurdish[0.040]; military[0.040]; military[0.040]; Arbil[0.039]; forces[0.039]; forces[0.039]; Iran[0.038]; Iran[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mediterranean ==> ENTITY: \u001b[32mMediterranean Sea\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnorthern[0.049]; northern[0.049]; region[0.047]; eastern[0.045]; Kurdish[0.045]; ships[0.044]; ships[0.044]; military[0.043]; Iraq[0.042]; Iraq[0.042]; force[0.040]; force[0.040]; forces[0.040]; forces[0.040]; Washington[0.040]; response[0.040]; Defense[0.040]; expeditionary[0.039]; Gulf[0.039]; Gulf[0.039]; Gulf[0.039]; action[0.039]; Pentagon[0.039]; deploy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Air Force ==> ENTITY: \u001b[32mUnited States Air Force\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.837:-0.837[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.049]; military[0.049]; Iraqi[0.048]; aircraft[0.048]; planes[0.047]; planes[0.047]; forces[0.046]; bases[0.046]; Iraq[0.046]; Iraq[0.046]; personnel[0.046]; troops[0.044]; troops[0.044]; Saudi[0.044]; Air[0.044]; Marines[0.044]; Marines[0.044]; servicemen[0.044]; total[0.043]; force[0.043]; force[0.043]; Readiness[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.029:0.029[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.045]; military[0.045]; Marines[0.044]; Marines[0.044]; nearly[0.043]; personnel[0.043]; aircraft[0.042]; servicemen[0.042]; according[0.041]; ships[0.041]; ships[0.041]; ships[0.041]; States[0.041]; Readiness[0.041]; Air[0.041]; Air[0.041]; Amphibious[0.041]; total[0.041]; seven[0.041]; sailors[0.040]; planes[0.040]; planes[0.040]; troops[0.040]; troops[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bill Clinton\u001b[39m ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.100:0.100[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.045]; Clinton[0.045]; Iraq[0.045]; military[0.044]; Washington[0.043]; Defense[0.043]; Iraqi[0.043]; Pentagon[0.042]; president[0.041]; officials[0.041]; direct[0.041]; Department[0.041]; forces[0.041]; Jim[0.041]; Adams[0.041]; force[0.041]; force[0.041]; use[0.041]; case[0.040]; deploy[0.040]; States[0.040]; said[0.040]; said[0.040]; planning[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.038:0.038[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.047]; military[0.047]; Defense[0.046]; Pentagon[0.045]; aircraft[0.045]; aircraft[0.045]; deploy[0.044]; Iraq[0.044]; Iraq[0.044]; States[0.044]; Department[0.043]; enforce[0.043]; planes[0.043]; time[0.043]; ordered[0.042]; ordered[0.042]; ordered[0.042]; doubled[0.042]; Tennessee[0.042]; southern[0.042]; Yesterday[0.041]; carrier[0.041]; carrier[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tennessee ==> ENTITY: \u001b[32mTennessee\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.067:0.067[0]; log p(e|m)= -0.280:-0.280[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDefense[0.046]; military[0.045]; military[0.045]; Pentagon[0.043]; Troy[0.043]; Department[0.042]; deploy[0.042]; Saturday[0.042]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; held[0.041]; States[0.041]; plans[0.041]; southern[0.041]; flights[0.041]; enforce[0.040]; planning[0.040]; planning[0.040]; Clinton[0.040]; Clinton[0.040]; Clinton[0.040]; added[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCarl Vinson\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mCarl Vinson\u001b[39m <---> \u001b[32mUSS Carl Vinson (CVN-70)\u001b[39m\t\nSCORES: global= 0.291:0.278[\u001b[31m0.012\u001b[39m]; local()= 0.075:0.105[\u001b[32m0.030\u001b[39m]; log p(e|m)= 0.000:-0.286[\u001b[31m0.286\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBill[0.044]; Gulf[0.043]; Gulf[0.043]; Gulf[0.043]; Jim[0.043]; carrier[0.043]; carrier[0.043]; President[0.042]; Pentagon[0.042]; ships[0.042]; ships[0.042]; Department[0.042]; Washington[0.042]; expeditionary[0.041]; military[0.041]; Adams[0.041]; aircraft[0.041]; aircraft[0.041]; States[0.041]; Defense[0.041]; Iraq[0.040]; deploy[0.040]; Doug[0.040]; officials[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iran ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.171:0.171[0]; log p(e|m)= -0.138:-0.138[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.047]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; military[0.044]; military[0.044]; Iraqi[0.042]; Iraqi[0.042]; Kurdish[0.042]; Kurdish[0.042]; troops[0.041]; troops[0.041]; northern[0.040]; northern[0.040]; northern[0.040]; Baghdad[0.040]; region[0.040]; southern[0.040]; forces[0.040]; forces[0.040]; attack[0.039]; attacks[0.039]; developments[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iran ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.163:0.163[0]; log p(e|m)= -0.138:-0.138[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.049]; Iraq[0.047]; Iraq[0.047]; Iraq[0.047]; Iraq[0.047]; military[0.046]; military[0.046]; Iraqi[0.044]; Iraqi[0.044]; Kurdish[0.044]; troops[0.042]; troops[0.042]; northern[0.042]; northern[0.042]; Baghdad[0.041]; region[0.041]; southern[0.041]; forces[0.041]; forces[0.041]; attack[0.041]; attacks[0.041]; developments[0.041]; according[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gulf ==> ENTITY: \u001b[32mPersian Gulf\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.192:0.192[0]; log p(e|m)= -2.163:-2.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGulf[0.048]; region[0.046]; military[0.046]; military[0.046]; Marines[0.044]; Marines[0.044]; Arabia[0.044]; ships[0.044]; ships[0.044]; ships[0.044]; troops[0.044]; troops[0.044]; personnel[0.043]; nearly[0.043]; Saudi[0.043]; States[0.043]; area[0.043]; people[0.043]; servicemen[0.042]; sailors[0.042]; bases[0.041]; aircraft[0.041]; support[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMarines\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMarines\u001b[39m <---> \u001b[32mUnited States Marine Corps\u001b[39m\t\nSCORES: global= 0.258:0.256[\u001b[31m0.002\u001b[39m]; local()= 0.192:0.139[\u001b[31m0.053\u001b[39m]; log p(e|m)= -1.363:-0.454[\u001b[32m0.908\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMarines[0.046]; personnel[0.045]; military[0.044]; military[0.044]; Amphibious[0.043]; ships[0.043]; ships[0.043]; ships[0.043]; sailors[0.042]; servicemen[0.041]; force[0.041]; force[0.041]; area[0.041]; bases[0.040]; troops[0.040]; troops[0.040]; expeditionary[0.040]; expeditionary[0.040]; nearly[0.040]; aircraft[0.040]; exercises[0.040]; planes[0.040]; planes[0.040]; total[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.051]; military[0.046]; warplanes[0.044]; warplanes[0.044]; aircraft[0.044]; aircraft[0.044]; forces[0.043]; planes[0.043]; northern[0.042]; Pentagon[0.042]; eastern[0.041]; Defense[0.041]; region[0.041]; deploy[0.040]; Department[0.040]; force[0.040]; force[0.040]; ships[0.039]; ships[0.039]; officials[0.039]; Gulf[0.039]; Gulf[0.039]; fighter[0.039]; air[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.066:0.066[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.046]; military[0.046]; Pentagon[0.044]; deploy[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; according[0.043]; reports[0.042]; enforce[0.041]; time[0.041]; troops[0.041]; troops[0.041]; ordered[0.041]; doubled[0.041]; Tennessee[0.041]; southern[0.040]; fly[0.040]; want[0.040]; northern[0.040]; northern[0.040]; northern[0.040]; speculate[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.222:0.222[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.050]; Iraq[0.050]; Iraqi[0.046]; Iraqi[0.046]; military[0.045]; military[0.045]; Baghdad[0.044]; aircraft[0.044]; troops[0.043]; troops[0.043]; forces[0.043]; forces[0.043]; Kurdish[0.043]; planes[0.043]; planes[0.043]; bases[0.042]; northern[0.042]; northern[0.042]; southern[0.042]; Arbil[0.041]; Iran[0.040]; Iran[0.040]; region[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.218:0.218[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.048]; Iraq[0.048]; Iraqi[0.044]; Iraqi[0.044]; military[0.044]; military[0.044]; Baghdad[0.042]; aircraft[0.042]; troops[0.042]; troops[0.042]; forces[0.042]; forces[0.042]; Kurdish[0.041]; planes[0.041]; planes[0.041]; personnel[0.040]; bases[0.040]; northern[0.040]; northern[0.040]; southern[0.040]; Arbil[0.039]; Iran[0.039]; Iran[0.039]; region[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWashington[0.044]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; Saturday[0.042]; Saturday[0.042]; Adams[0.042]; action[0.042]; added[0.042]; Tennessee[0.042]; Troy[0.041]; Jim[0.041]; Department[0.041]; Iraq[0.041]; Iraq[0.041]; Carl[0.041]; order[0.041]; Doug[0.041]; northern[0.041]; northern[0.041]; officials[0.041]; defense[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.028:0.028[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.045]; military[0.045]; Marines[0.044]; Marines[0.044]; nearly[0.043]; personnel[0.043]; aircraft[0.042]; servicemen[0.042]; according[0.041]; ships[0.041]; ships[0.041]; ships[0.041]; States[0.041]; Readiness[0.041]; Air[0.041]; Air[0.041]; Amphibious[0.041]; total[0.041]; seven[0.041]; sailors[0.040]; planes[0.040]; planes[0.040]; troops[0.040]; bases[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Saudi Arabia ==> ENTITY: \u001b[32mSaudi Arabia\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.255:-0.255[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.050]; military[0.050]; personnel[0.047]; troops[0.045]; Marines[0.045]; Marines[0.045]; Gulf[0.044]; Gulf[0.044]; aircraft[0.044]; people[0.043]; servicemen[0.043]; force[0.043]; force[0.043]; support[0.042]; bases[0.042]; Group[0.042]; group[0.042]; ships[0.042]; ships[0.042]; ships[0.042]; Readiness[0.041]; seven[0.041]; primarily[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.216:0.216[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.049]; Iraq[0.049]; Iraqi[0.046]; Kurdish[0.045]; Kurdish[0.045]; Baghdad[0.043]; military[0.042]; military[0.042]; northern[0.041]; northern[0.041]; southern[0.041]; Pentagon[0.041]; eastern[0.040]; Arbil[0.040]; region[0.040]; forces[0.040]; forces[0.040]; forces[0.040]; said[0.039]; said[0.039]; said[0.039]; said[0.039]; said[0.039]; Defense[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.244:0.244[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.047]; Iraq[0.047]; Iraq[0.047]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Baghdad[0.042]; aircraft[0.041]; military[0.041]; military[0.041]; troops[0.041]; troops[0.041]; forces[0.041]; forces[0.041]; forces[0.041]; Kurdish[0.041]; Kurdish[0.041]; planes[0.041]; northern[0.039]; northern[0.039]; northern[0.039]; southern[0.039]; Arbil[0.039]; region[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gulf ==> ENTITY: \u001b[32mPersian Gulf\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.201:0.201[0]; log p(e|m)= -2.163:-2.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.049]; Gulf[0.048]; Gulf[0.048]; military[0.046]; region[0.044]; Mediterranean[0.044]; ships[0.044]; ships[0.044]; Iraqi[0.044]; Iraqi[0.044]; Pentagon[0.044]; States[0.043]; eastern[0.043]; northern[0.042]; deploy[0.042]; Department[0.042]; aircraft[0.042]; aircraft[0.042]; forces[0.042]; forces[0.042]; Defense[0.041]; force[0.041]; force[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.045:0.045[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.045]; military[0.045]; nearly[0.044]; personnel[0.044]; aircraft[0.043]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; according[0.042]; States[0.042]; Air[0.041]; Air[0.041]; reports[0.041]; enforce[0.041]; planes[0.041]; planes[0.041]; time[0.041]; troops[0.041]; troops[0.041]; doubled[0.041]; bases[0.040]; Force[0.040]; Force[0.040]; Tennessee[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.046:0.046[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.045]; military[0.045]; nearly[0.044]; personnel[0.044]; aircraft[0.043]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; according[0.042]; States[0.042]; Air[0.041]; Air[0.041]; reports[0.041]; enforce[0.041]; planes[0.041]; planes[0.041]; time[0.041]; troops[0.041]; troops[0.041]; doubled[0.041]; bases[0.040]; Force[0.040]; Force[0.040]; southern[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.041:0.041[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.046]; order[0.045]; Defense[0.044]; aircraft[0.043]; aircraft[0.043]; Washington[0.043]; Iraq[0.042]; ships[0.042]; ships[0.042]; States[0.042]; case[0.042]; Department[0.042]; planes[0.041]; ordered[0.041]; ordered[0.041]; ordered[0.041]; Yesterday[0.040]; carrier[0.040]; carrier[0.040]; fly[0.040]; Carl[0.040]; fighter[0.040]; northern[0.040]; Adams[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.027:0.027[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.045]; military[0.045]; Marines[0.044]; Marines[0.044]; nearly[0.043]; personnel[0.043]; aircraft[0.042]; servicemen[0.042]; ships[0.041]; ships[0.041]; ships[0.041]; States[0.041]; Readiness[0.041]; Air[0.041]; Air[0.041]; Amphibious[0.041]; total[0.041]; seven[0.041]; sailors[0.040]; planes[0.040]; planes[0.040]; troops[0.040]; bases[0.040]; based[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMarines\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMarines\u001b[39m <---> \u001b[32mUnited States Marine Corps\u001b[39m\t\nSCORES: global= 0.257:0.256[\u001b[31m0.001\u001b[39m]; local()= 0.188:0.142[\u001b[31m0.045\u001b[39m]; log p(e|m)= -1.363:-0.454[\u001b[32m0.908\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMarines[0.046]; personnel[0.046]; military[0.044]; military[0.044]; Amphibious[0.043]; ships[0.043]; ships[0.043]; ships[0.043]; sailors[0.042]; servicemen[0.041]; force[0.041]; force[0.041]; area[0.041]; bases[0.041]; troops[0.041]; expeditionary[0.041]; expeditionary[0.041]; nearly[0.040]; exercises[0.040]; planes[0.040]; total[0.040]; Tarawa[0.039]; support[0.039]; Readiness[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gulf ==> ENTITY: \u001b[32mPersian Gulf\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.160:0.160[0]; log p(e|m)= -2.163:-2.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.047]; Gulf[0.047]; military[0.045]; eastern[0.044]; Mediterranean[0.044]; northern[0.043]; ships[0.043]; ships[0.043]; Iraqi[0.043]; States[0.042]; aircraft[0.041]; aircraft[0.041]; force[0.040]; force[0.040]; Clinton[0.040]; Clinton[0.040]; warplanes[0.040]; warplanes[0.040]; air[0.040]; planes[0.040]; carrier[0.040]; carrier[0.040]; expeditionary[0.039]; Washington[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraqi[0.047]; warplanes[0.045]; warplanes[0.045]; aircraft[0.045]; aircraft[0.045]; military[0.045]; planes[0.044]; forces[0.043]; northern[0.042]; Pentagon[0.042]; region[0.041]; deploy[0.041]; Defense[0.040]; force[0.040]; force[0.040]; eastern[0.040]; ships[0.040]; ships[0.040]; Washington[0.040]; Department[0.040]; Gulf[0.040]; Gulf[0.040]; fighter[0.039]; air[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Clinton\u001b[39m ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.089:0.089[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinton[0.045]; Clinton[0.045]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; military[0.045]; military[0.045]; enforce[0.044]; campaign[0.044]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iran[0.043]; Iran[0.043]; response[0.042]; response[0.042]; response[0.042]; officials[0.042]; officials[0.042]; according[0.042]; troops[0.042]; troops[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.222:0.222[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.047]; Iraq[0.047]; Iraq[0.047]; Iraqi[0.044]; Kurdish[0.043]; military[0.043]; military[0.043]; military[0.043]; Baghdad[0.042]; aircraft[0.041]; troops[0.041]; troops[0.041]; forces[0.041]; forces[0.041]; planes[0.041]; planes[0.041]; personnel[0.040]; bases[0.039]; northern[0.039]; northern[0.039]; southern[0.039]; Iran[0.039]; Iran[0.039]; Arbil[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.046:0.046[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.045]; military[0.045]; military[0.045]; nearly[0.044]; personnel[0.044]; aircraft[0.043]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; according[0.042]; States[0.041]; Air[0.041]; Air[0.041]; reports[0.041]; enforce[0.041]; planes[0.041]; planes[0.041]; troops[0.040]; troops[0.040]; doubled[0.040]; bases[0.040]; Force[0.040]; Force[0.040]; southern[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Department of Defense\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Defense\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPentagon[0.046]; military[0.045]; Enterprise[0.043]; Iraq[0.042]; Iraq[0.042]; States[0.042]; deploy[0.042]; President[0.042]; force[0.041]; force[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; region[0.041]; ordered[0.041]; ordered[0.041]; ordered[0.041]; ordered[0.041]; officials[0.041]; planning[0.041]; planning[0.041]; United[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCarl Vinson\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mCarl Vinson\u001b[39m <---> \u001b[32mUSS Carl Vinson (CVN-70)\u001b[39m\t\nSCORES: global= 0.290:0.279[\u001b[31m0.011\u001b[39m]; local()= 0.070:0.123[\u001b[32m0.053\u001b[39m]; log p(e|m)= 0.000:-0.286[\u001b[31m0.286\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGulf[0.046]; Gulf[0.046]; Marines[0.045]; Marines[0.045]; nearly[0.044]; personnel[0.044]; sailors[0.044]; ships[0.044]; ships[0.044]; ships[0.044]; exercises[0.044]; seven[0.044]; expeditionary[0.044]; military[0.043]; military[0.043]; States[0.043]; servicemen[0.042]; Tarawa[0.042]; support[0.042]; Amphibious[0.042]; crews[0.042]; Force[0.041]; fly[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.031:0.031[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.045]; military[0.045]; Marines[0.045]; Marines[0.045]; nearly[0.044]; personnel[0.044]; servicemen[0.042]; ships[0.041]; ships[0.041]; ships[0.041]; States[0.041]; Readiness[0.041]; Air[0.041]; Amphibious[0.041]; total[0.041]; seven[0.041]; sailors[0.041]; troops[0.040]; bases[0.040]; based[0.040]; Force[0.040]; Tarawa[0.040]; Carl[0.039]; expeditionary[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.033:0.033[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.045]; order[0.045]; Defense[0.044]; Pentagon[0.044]; aircraft[0.043]; aircraft[0.043]; Washington[0.043]; deploy[0.042]; Iraq[0.042]; ships[0.042]; ships[0.042]; States[0.042]; case[0.042]; Department[0.041]; planes[0.041]; ordered[0.041]; ordered[0.041]; ordered[0.041]; Yesterday[0.040]; carrier[0.040]; carrier[0.040]; fly[0.040]; Carl[0.040]; fighter[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.226:0.226[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.049]; Iraqi[0.045]; Iraqi[0.045]; Baghdad[0.043]; forces[0.043]; forces[0.043]; forces[0.043]; Kurdish[0.043]; Kurdish[0.043]; military[0.042]; military[0.042]; northern[0.041]; northern[0.041]; southern[0.041]; Pentagon[0.041]; Arbil[0.040]; region[0.040]; Iran[0.040]; Iran[0.040]; Gulf[0.039]; Defense[0.039]; force[0.039]; Department[0.039]; officials[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Clinton\u001b[39m ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.091:0.091[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.046]; Clinton[0.045]; Iraq[0.045]; military[0.044]; Washington[0.044]; Iraqi[0.043]; Bill[0.043]; Department[0.043]; president[0.043]; officials[0.042]; States[0.041]; Jim[0.041]; Adams[0.041]; force[0.041]; force[0.041]; use[0.041]; case[0.041]; said[0.040]; said[0.040]; warplanes[0.040]; warplanes[0.040]; planes[0.039]; action[0.039]; order[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.251:0.251[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.047]; Iraq[0.047]; Iraq[0.047]; Iraq[0.047]; Iraqi[0.044]; Iraqi[0.044]; Kurdish[0.043]; Kurdish[0.043]; military[0.043]; military[0.043]; Baghdad[0.041]; aircraft[0.041]; troops[0.041]; troops[0.041]; forces[0.040]; forces[0.040]; forces[0.040]; planes[0.040]; northern[0.039]; northern[0.039]; northern[0.039]; southern[0.039]; Iran[0.038]; Iran[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Enterprise ==> ENTITY: \u001b[32mUSS Enterprise (CVN-65)\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.160:0.160[0]; log p(e|m)= -2.659:-2.659[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nships[0.047]; ships[0.047]; military[0.046]; carrier[0.045]; carrier[0.045]; aircraft[0.045]; aircraft[0.045]; Pentagon[0.045]; planes[0.044]; fly[0.043]; planning[0.043]; planning[0.043]; Defense[0.042]; Iraq[0.042]; deploy[0.042]; Gulf[0.042]; Gulf[0.042]; Gulf[0.042]; Washington[0.042]; expeditionary[0.042]; air[0.042]; Adams[0.041]; high[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Enterprise ==> ENTITY: \u001b[32mUSS Enterprise (CVN-65)\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.167:0.167[0]; log p(e|m)= -2.659:-2.659[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.044]; military[0.044]; military[0.044]; aircraft[0.043]; sailors[0.043]; personnel[0.042]; Marines[0.042]; Marines[0.042]; planes[0.042]; planes[0.042]; area[0.042]; carriers[0.042]; fly[0.041]; according[0.041]; bases[0.041]; nearly[0.041]; Air[0.041]; Air[0.041]; Iraq[0.040]; Iraq[0.040]; Iraq[0.040]; exercises[0.040]; Gulf[0.040]; addition[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Air Force ==> ENTITY: \u001b[32mUnited States Air Force\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.093:0.093[0]; log p(e|m)= -0.837:-0.837[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.044]; military[0.044]; military[0.044]; Iraqi[0.044]; Iraqi[0.044]; aircraft[0.044]; planes[0.043]; planes[0.043]; forces[0.042]; forces[0.042]; bases[0.042]; Iraq[0.041]; Iraq[0.041]; Iraq[0.041]; personnel[0.041]; Baghdad[0.041]; Iran[0.041]; Iran[0.041]; troops[0.040]; troops[0.040]; troops[0.040]; Saudi[0.040]; force[0.040]; force[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pentagon ==> ENTITY: \u001b[32mThe Pentagon\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.160:0.160[0]; log p(e|m)= -0.146:-0.146[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDefense[0.047]; military[0.045]; military[0.045]; Department[0.044]; Iraq[0.043]; Iraq[0.043]; officials[0.042]; officials[0.042]; States[0.041]; planes[0.041]; deploy[0.041]; aircraft[0.041]; aircraft[0.041]; stop[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; planning[0.040]; planning[0.040]; eastern[0.040]; attacks[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gulf ==> ENTITY: \u001b[32mPersian Gulf\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.183:0.183[0]; log p(e|m)= -2.163:-2.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGulf[0.047]; military[0.044]; military[0.044]; Marines[0.042]; Marines[0.042]; Arabia[0.042]; ships[0.042]; ships[0.042]; ships[0.042]; troops[0.042]; personnel[0.042]; nearly[0.042]; Saudi[0.042]; States[0.042]; area[0.042]; people[0.041]; servicemen[0.041]; sailors[0.041]; support[0.040]; force[0.040]; seven[0.040]; total[0.040]; includes[0.039]; Air[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Clinton\u001b[39m ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.100:0.100[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinton[0.044]; Clinton[0.044]; Iraq[0.044]; Iraq[0.044]; Iraq[0.044]; military[0.043]; military[0.043]; enforce[0.042]; campaign[0.042]; Iraqi[0.042]; Iraqi[0.042]; Iran[0.041]; Iran[0.041]; Pentagon[0.041]; response[0.041]; response[0.041]; response[0.041]; president[0.041]; officials[0.040]; direct[0.040]; Department[0.040]; troops[0.040]; attacks[0.040]; sent[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.033:0.033[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.045]; order[0.045]; Defense[0.044]; Pentagon[0.044]; aircraft[0.043]; aircraft[0.043]; Washington[0.043]; deploy[0.042]; Iraq[0.042]; ships[0.042]; ships[0.042]; States[0.042]; case[0.042]; Department[0.041]; planes[0.041]; ordered[0.041]; ordered[0.041]; ordered[0.041]; Yesterday[0.040]; carrier[0.040]; carrier[0.040]; fly[0.040]; Carl[0.040]; fighter[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arbil\u001b[39m ==> ENTITY: \u001b[32mErbil\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.240:0.240[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdish[0.046]; Kurdish[0.046]; region[0.044]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; Iraqi[0.042]; Iraqi[0.042]; Iraqi[0.042]; troops[0.042]; troops[0.042]; Baghdad[0.042]; northern[0.042]; northern[0.042]; northern[0.042]; southern[0.041]; military[0.040]; military[0.040]; city[0.039]; forces[0.039]; forces[0.039]; officials[0.038]; officials[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.054:0.054[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.046]; order[0.046]; aircraft[0.044]; aircraft[0.044]; Washington[0.044]; Iraq[0.043]; States[0.043]; case[0.043]; planes[0.042]; ordered[0.042]; use[0.042]; carrier[0.041]; carrier[0.041]; fly[0.041]; Carl[0.041]; fighter[0.040]; northern[0.040]; Adams[0.040]; Jim[0.040]; expeditionary[0.040]; Bill[0.040]; eastern[0.040]; force[0.040]; force[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.072:0.072[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBill[0.044]; President[0.044]; Department[0.043]; Adams[0.043]; States[0.043]; said[0.042]; said[0.042]; president[0.042]; military[0.042]; Clinton[0.042]; Clinton[0.042]; Saturday[0.041]; aircraft[0.041]; aircraft[0.041]; eastern[0.041]; order[0.041]; northern[0.041]; Jim[0.041]; Defense[0.041]; Iraq[0.041]; United[0.040]; Carl[0.040]; planes[0.040]; air[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.032:0.032[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.045]; military[0.045]; Marines[0.044]; Marines[0.044]; nearly[0.043]; personnel[0.043]; aircraft[0.042]; servicemen[0.042]; Iraq[0.041]; Iraq[0.041]; according[0.041]; ships[0.041]; ships[0.041]; ships[0.041]; States[0.041]; Readiness[0.041]; Air[0.041]; Air[0.041]; Amphibious[0.041]; reports[0.041]; seven[0.040]; sailors[0.040]; planes[0.040]; planes[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.084:0.084[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnearly[0.046]; group[0.044]; according[0.044]; said[0.043]; said[0.043]; people[0.042]; seven[0.042]; heavily[0.042]; based[0.042]; total[0.042]; personnel[0.042]; ground[0.042]; Carl[0.041]; Iraq[0.041]; Iraq[0.041]; primarily[0.041]; northern[0.041]; addition[0.041]; defense[0.041]; military[0.040]; military[0.040]; region[0.040]; exercises[0.040]; bases[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Clinton\u001b[39m ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.099:0.099[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinton[0.044]; Clinton[0.044]; Iraq[0.044]; Iraq[0.044]; Iraq[0.044]; military[0.043]; campaign[0.043]; enforce[0.042]; Iraqi[0.042]; Iraqi[0.042]; Department[0.042]; president[0.042]; Pentagon[0.041]; Defense[0.041]; response[0.041]; response[0.041]; officials[0.041]; officials[0.041]; direct[0.041]; forces[0.040]; forces[0.040]; States[0.040]; deploy[0.040]; attacks[0.039]; \t\n\r [================>.............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s801ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1847/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1106testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Madrid ==> ENTITY: \u001b[32mMadrid\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMadrid[0.081]; Soccer[0.078]; Spanish[0.075]; Spanish[0.075]; Coruna[0.074]; game[0.073]; played[0.073]; Deportivo[0.072]; Real[0.069]; division[0.066]; Saturday[0.066]; Division[0.066]; Result[0.066]; Result[0.066]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spanish ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.108:0.108[0]; log p(e|m)= -1.241:-1.241[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpanish[0.089]; Soccer[0.077]; Madrid[0.074]; Madrid[0.074]; game[0.073]; Division[0.073]; played[0.072]; division[0.071]; Real[0.070]; Deportivo[0.069]; Coruna[0.069]; Saturday[0.065]; Result[0.062]; Result[0.062]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Madrid\u001b[39m ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.087]; Madrid[0.083]; Deportivo[0.080]; game[0.080]; Coruna[0.079]; Division[0.078]; played[0.077]; Spanish[0.076]; Spanish[0.076]; division[0.076]; Saturday[0.072]; Result[0.068]; Result[0.068]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spanish ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.108:0.108[0]; log p(e|m)= -1.241:-1.241[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpanish[0.089]; Soccer[0.077]; Madrid[0.074]; Madrid[0.074]; game[0.073]; Division[0.073]; played[0.072]; division[0.071]; Real[0.070]; Deportivo[0.069]; Coruna[0.069]; Saturday[0.065]; Result[0.062]; Result[0.062]; \t\n\r [================>.............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s820ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1851/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 962testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cunningham\u001b[39m ==> ENTITY: \u001b[32mRandall Cunningham\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nquarterback[0.047]; Cunningham[0.046]; Cunningham[0.046]; Cunningham[0.046]; Cunningham[0.046]; Eagles[0.044]; Eagles[0.044]; yards[0.044]; rusher[0.043]; players[0.043]; Randall[0.043]; Randall[0.043]; Bowl[0.043]; Philadelphia[0.042]; Philadelphia[0.042]; Philadelphia[0.042]; played[0.042]; athletic[0.042]; rushed[0.042]; Football[0.042]; Football[0.042]; forward[0.042]; Retires[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pro Bowl\u001b[39m ==> ENTITY: \u001b[32mPro Bowl\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFootball[0.052]; Football[0.052]; quarterback[0.051]; selection[0.051]; players[0.051]; played[0.050]; Eagles[0.050]; Eagles[0.050]; Eagles[0.050]; Eagles[0.050]; franchise[0.050]; Philadelphia[0.050]; Philadelphia[0.050]; Philadelphia[0.050]; Philadelphia[0.050]; rusher[0.049]; career[0.049]; career[0.049]; Thursday[0.049]; quarterbacks[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cunningham\u001b[39m ==> ENTITY: \u001b[32mRandall Cunningham\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCunningham[0.045]; Cunningham[0.045]; touchdowns[0.044]; quarterbacks[0.044]; Eagles[0.043]; Eagles[0.043]; Eagles[0.043]; yards[0.043]; second[0.042]; Randall[0.042]; Randall[0.042]; Philadelphia[0.041]; Philadelphia[0.041]; Philadelphia[0.041]; forward[0.041]; career[0.041]; passes[0.040]; fans[0.040]; fans[0.040]; great[0.040]; franchise[0.040]; said[0.039]; said[0.039]; spending[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPhiladelphia\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPhiladelphia Eagles\u001b[39m <---> \u001b[32mPhiladelphia\u001b[39m\t\nSCORES: global= 0.252:0.240[\u001b[31m0.012\u001b[39m]; local()= 0.199:0.123[\u001b[31m0.076\u001b[39m]; log p(e|m)= -4.510:-0.211[\u001b[32m4.299\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.046]; Philadelphia[0.046]; Philadelphia[0.046]; franchise[0.045]; fans[0.043]; fans[0.043]; second[0.043]; played[0.042]; Eagles[0.041]; Eagles[0.041]; Eagles[0.041]; Eagles[0.041]; career[0.041]; career[0.041]; Jeffrey[0.041]; history[0.041]; Pro[0.040]; Thursday[0.040]; round[0.040]; broadcaster[0.040]; Bowl[0.039]; time[0.039]; time[0.039]; cornerstone[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NFL ==> ENTITY: \u001b[32mNational Football League\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.194:0.194[0]; log p(e|m)= -0.019:-0.019[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbroadcaster[0.044]; players[0.043]; Philadelphia[0.043]; Philadelphia[0.043]; Philadelphia[0.043]; Philadelphia[0.043]; franchise[0.043]; second[0.042]; history[0.042]; played[0.042]; Bowl[0.042]; Eagles[0.041]; Eagles[0.041]; Eagles[0.041]; Eagles[0.041]; fans[0.041]; fans[0.041]; Thursday[0.041]; quarterbacks[0.041]; career[0.040]; career[0.040]; athletic[0.040]; signed[0.040]; round[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cunningham\u001b[39m ==> ENTITY: \u001b[32mRandall Cunningham\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nquarterback[0.045]; Cunningham[0.044]; Cunningham[0.044]; Cunningham[0.044]; Cunningham[0.044]; quarterbacks[0.042]; Eagles[0.042]; Eagles[0.042]; Eagles[0.042]; Eagles[0.042]; yards[0.042]; rusher[0.041]; players[0.041]; Randall[0.041]; Randall[0.041]; Randall[0.041]; Randall[0.041]; Bowl[0.040]; Philadelphia[0.040]; Philadelphia[0.040]; Philadelphia[0.040]; Philadelphia[0.040]; Philadelphia[0.040]; played[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cunningham\u001b[39m ==> ENTITY: \u001b[32mRandall Cunningham\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nquarterback[0.045]; Cunningham[0.044]; Cunningham[0.044]; Cunningham[0.044]; Cunningham[0.044]; quarterbacks[0.043]; Eagles[0.042]; Eagles[0.042]; Eagles[0.042]; yards[0.042]; rusher[0.041]; players[0.041]; Randall[0.041]; Randall[0.041]; Randall[0.041]; Randall[0.041]; Bowl[0.041]; Philadelphia[0.040]; Philadelphia[0.040]; Philadelphia[0.040]; Philadelphia[0.040]; played[0.040]; athletic[0.040]; rushed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Eagles ==> ENTITY: \u001b[32mPhiladelphia Eagles\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.193:0.193[0]; log p(e|m)= -2.477:-2.477[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.044]; Philadelphia[0.044]; Philadelphia[0.044]; Philadelphia[0.044]; Eagles[0.043]; Eagles[0.043]; Eagles[0.043]; yards[0.043]; yards[0.043]; second[0.042]; franchise[0.042]; touchdowns[0.041]; Jeffrey[0.041]; rushed[0.041]; career[0.041]; career[0.041]; Pro[0.041]; Randall[0.040]; Randall[0.040]; fans[0.040]; fans[0.040]; round[0.040]; Bowl[0.040]; carries[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Eagles ==> ENTITY: \u001b[32mPhiladelphia Eagles\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.196:0.196[0]; log p(e|m)= -2.477:-2.477[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.044]; Philadelphia[0.044]; Philadelphia[0.044]; Philadelphia[0.044]; Eagles[0.043]; Eagles[0.043]; Eagles[0.043]; yards[0.043]; yards[0.043]; second[0.042]; players[0.042]; franchise[0.042]; touchdowns[0.041]; Jeffrey[0.041]; rushed[0.041]; career[0.041]; career[0.041]; Pro[0.041]; played[0.040]; Randall[0.040]; Randall[0.040]; fans[0.040]; fans[0.040]; round[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Eagles ==> ENTITY: \u001b[32mPhiladelphia Eagles\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.213:0.213[0]; log p(e|m)= -2.477:-2.477[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.043]; Philadelphia[0.043]; Philadelphia[0.043]; Philadelphia[0.043]; Philadelphia[0.043]; Bowl[0.043]; Eagles[0.042]; Eagles[0.042]; Eagles[0.042]; quarterback[0.042]; Football[0.042]; Football[0.042]; second[0.041]; players[0.041]; franchise[0.041]; League[0.041]; yards[0.041]; athletic[0.041]; quarterbacks[0.041]; career[0.040]; career[0.040]; Jeffrey[0.040]; Pro[0.040]; played[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NFL ==> ENTITY: \u001b[32mNational Football League\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.019:-0.019[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.047]; Philadelphia[0.043]; Philadelphia[0.043]; Philadelphia[0.043]; Football[0.043]; Football[0.043]; players[0.042]; selection[0.042]; played[0.042]; quarterback[0.042]; Eagles[0.042]; Eagles[0.042]; National[0.041]; career[0.041]; career[0.041]; Thursday[0.041]; Bowl[0.041]; Pro[0.041]; time[0.040]; time[0.040]; time[0.040]; rusher[0.040]; American[0.040]; athletic[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Randall Cunningham\u001b[39m ==> ENTITY: \u001b[32mRandall Cunningham\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nquarterback[0.045]; Cunningham[0.044]; Cunningham[0.044]; Cunningham[0.044]; Cunningham[0.044]; Eagles[0.043]; Eagles[0.043]; yards[0.042]; rusher[0.042]; players[0.041]; Randall[0.041]; Bowl[0.041]; Philadelphia[0.041]; Philadelphia[0.041]; Philadelphia[0.041]; played[0.041]; athletic[0.040]; rushed[0.040]; Football[0.040]; Football[0.040]; forward[0.040]; Retires[0.040]; career[0.040]; career[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jeffrey Lurie\u001b[39m ==> ENTITY: \u001b[32mJeffrey Lurie\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfranchise[0.046]; career[0.044]; career[0.044]; Philadelphia[0.043]; Philadelphia[0.043]; Philadelphia[0.043]; Philadelphia[0.043]; touchdowns[0.042]; played[0.042]; Eagles[0.042]; Eagles[0.042]; Eagles[0.042]; Eagles[0.042]; quarterbacks[0.041]; yards[0.041]; yards[0.041]; broadcaster[0.041]; Thursday[0.040]; fans[0.040]; fans[0.040]; history[0.040]; exciting[0.040]; passes[0.040]; second[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPhiladelphia\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPhiladelphia Eagles\u001b[39m <---> \u001b[32mPhiladelphia\u001b[39m\t\nSCORES: global= 0.254:0.239[\u001b[31m0.015\u001b[39m]; local()= 0.216:0.113[\u001b[31m0.103\u001b[39m]; log p(e|m)= -4.510:-0.211[\u001b[32m4.299\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.045]; Philadelphia[0.045]; Philadelphia[0.045]; Philadelphia[0.045]; franchise[0.044]; players[0.043]; fans[0.042]; fans[0.042]; second[0.042]; League[0.042]; played[0.041]; Football[0.041]; Football[0.041]; Eagles[0.041]; Eagles[0.041]; Eagles[0.041]; Eagles[0.041]; career[0.041]; career[0.041]; quarterback[0.040]; Jeffrey[0.040]; history[0.040]; Pro[0.039]; Thursday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Philadelphia Eagles ==> ENTITY: \u001b[32mPhiladelphia Eagles\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.203:0.203[0]; log p(e|m)= -0.114:-0.114[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.044]; Philadelphia[0.044]; Philadelphia[0.044]; yards[0.043]; quarterback[0.042]; Eagles[0.042]; Eagles[0.042]; Eagles[0.042]; Football[0.042]; Football[0.042]; Randall[0.042]; Randall[0.042]; Randall[0.042]; Randall[0.042]; Thursday[0.041]; players[0.041]; franchise[0.041]; League[0.041]; rushed[0.041]; career[0.040]; career[0.040]; Pro[0.040]; played[0.040]; rusher[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NFL ==> ENTITY: \u001b[32mNational Football League\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.158:0.158[0]; log p(e|m)= -0.019:-0.019[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbroadcaster[0.044]; Philadelphia[0.044]; Philadelphia[0.044]; Philadelphia[0.044]; franchise[0.044]; history[0.043]; Eagles[0.042]; Eagles[0.042]; Eagles[0.042]; fans[0.042]; fans[0.042]; second[0.042]; quarterbacks[0.041]; career[0.041]; signed[0.041]; round[0.041]; time[0.040]; cornerstone[0.040]; Randall[0.040]; Randall[0.040]; touchdowns[0.040]; Cunningham[0.040]; Cunningham[0.040]; Cunningham[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Randall\u001b[39m ==> ENTITY: \u001b[32mRandall Cunningham\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.208:0.208[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCunningham[0.046]; Cunningham[0.046]; Cunningham[0.046]; Cunningham[0.046]; Cunningham[0.046]; touchdowns[0.044]; quarterbacks[0.044]; Eagles[0.044]; Eagles[0.044]; Eagles[0.044]; Eagles[0.044]; yards[0.044]; yards[0.044]; second[0.043]; Randall[0.043]; Bowl[0.042]; Philadelphia[0.042]; Philadelphia[0.042]; Philadelphia[0.042]; Philadelphia[0.042]; played[0.042]; rushed[0.042]; forward[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPhiladelphia\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPhiladelphia Eagles\u001b[39m <---> \u001b[32mPhiladelphia\u001b[39m\t\nSCORES: global= 0.254:0.239[\u001b[31m0.015\u001b[39m]; local()= 0.195:0.086[\u001b[31m0.109\u001b[39m]; log p(e|m)= -4.510:-0.211[\u001b[32m4.299\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.046]; Philadelphia[0.046]; players[0.044]; fans[0.044]; League[0.043]; played[0.043]; Football[0.042]; Football[0.042]; Eagles[0.042]; Eagles[0.042]; career[0.042]; career[0.042]; quarterback[0.041]; Pro[0.041]; Thursday[0.041]; American[0.041]; National[0.040]; Bowl[0.040]; time[0.040]; time[0.040]; time[0.040]; leading[0.040]; forward[0.039]; rusher[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cunningham\u001b[39m ==> ENTITY: \u001b[32mRandall Cunningham\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nquarterback[0.048]; Cunningham[0.047]; Cunningham[0.047]; Cunningham[0.047]; Cunningham[0.047]; Cunningham[0.047]; touchdowns[0.046]; quarterbacks[0.046]; Eagles[0.046]; Eagles[0.046]; Eagles[0.046]; Eagles[0.046]; yards[0.045]; yards[0.045]; rusher[0.044]; players[0.044]; second[0.044]; Randall[0.044]; Randall[0.044]; Randall[0.044]; Randall[0.044]; Bowl[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cunningham\u001b[39m ==> ENTITY: \u001b[32mRandall Cunningham\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCunningham[0.046]; touchdowns[0.044]; quarterbacks[0.044]; Eagles[0.044]; Eagles[0.044]; yards[0.044]; second[0.043]; Randall[0.043]; Randall[0.043]; Philadelphia[0.042]; Philadelphia[0.042]; forward[0.042]; passes[0.041]; fans[0.041]; great[0.041]; franchise[0.041]; said[0.040]; said[0.040]; spending[0.040]; Jeffrey[0.040]; time[0.040]; round[0.040]; exciting[0.039]; completed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = National Football League\u001b[39m ==> ENTITY: \u001b[32mNational Football League\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFootball[0.044]; Philadelphia[0.043]; Philadelphia[0.043]; Philadelphia[0.043]; quarterback[0.043]; players[0.043]; played[0.043]; Eagles[0.042]; Eagles[0.042]; career[0.042]; career[0.042]; Thursday[0.041]; rusher[0.041]; Bowl[0.041]; selection[0.041]; American[0.041]; athletic[0.041]; leading[0.040]; Pro[0.040]; time[0.040]; time[0.040]; time[0.040]; Randall[0.040]; Randall[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPhiladelphia\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPhiladelphia Eagles\u001b[39m <---> \u001b[32mPhiladelphia\u001b[39m\t\nSCORES: global= 0.249:0.236[\u001b[31m0.013\u001b[39m]; local()= 0.172:0.077[\u001b[31m0.095\u001b[39m]; log p(e|m)= -4.510:-0.211[\u001b[32m4.299\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.050]; Philadelphia[0.050]; franchise[0.049]; fans[0.047]; fans[0.047]; Eagles[0.047]; Eagles[0.047]; Eagles[0.047]; second[0.047]; career[0.045]; Jeffrey[0.045]; history[0.045]; yards[0.044]; yards[0.044]; touchdowns[0.044]; round[0.044]; broadcaster[0.043]; time[0.043]; cornerstone[0.043]; forward[0.043]; rushed[0.042]; signed[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Randall\u001b[39m ==> ENTITY: \u001b[32mRandall Cunningham\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nquarterback[0.045]; Cunningham[0.043]; Cunningham[0.043]; Cunningham[0.043]; Cunningham[0.043]; Cunningham[0.043]; touchdowns[0.042]; quarterbacks[0.042]; Eagles[0.042]; Eagles[0.042]; Eagles[0.042]; Eagles[0.042]; yards[0.042]; yards[0.042]; rusher[0.041]; players[0.041]; second[0.041]; Randall[0.041]; Bowl[0.040]; Philadelphia[0.040]; Philadelphia[0.040]; Philadelphia[0.040]; Philadelphia[0.040]; played[0.040]; \t\n\r [================>.............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s721ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1874/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1131testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Havana ==> ENTITY: \u001b[32mHavana\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.160:0.160[0]; log p(e|m)= -0.071:-0.071[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCuba[0.046]; Cuba[0.046]; Soler[0.046]; Soler[0.046]; Cuban[0.045]; Cuban[0.045]; Santiago[0.043]; city[0.043]; Jose[0.042]; Jose[0.042]; Castro[0.042]; Castro[0.042]; dies[0.041]; Saturday[0.040]; native[0.040]; age[0.040]; Granma[0.040]; Batista[0.040]; Fidel[0.040]; January[0.039]; Sleeping[0.039]; life[0.039]; titles[0.039]; official[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mGranma\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mGranma (yacht)\u001b[39m <---> \u001b[32mGranma (newspaper)\u001b[39m\t\nSCORES: global= 0.262:0.257[\u001b[31m0.005\u001b[39m]; local()= 0.119:0.125[\u001b[32m0.006\u001b[39m]; log p(e|m)= -0.270:-0.924[\u001b[31m0.653\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGranma[0.047]; Cuba[0.045]; Cuba[0.045]; Cuban[0.044]; Cuban[0.044]; Havana[0.044]; published[0.043]; newspaper[0.043]; city[0.042]; Saturday[0.042]; official[0.041]; Castro[0.041]; Castro[0.041]; government[0.041]; Santiago[0.040]; January[0.040]; Fidel[0.040]; revolution[0.040]; reported[0.040]; Fulgencio[0.039]; Batista[0.039]; Jose[0.039]; Jose[0.039]; day[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Prensa Latina\u001b[39m ==> ENTITY: \u001b[32mPrensa Latina\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.092:0.092[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGranma[0.048]; interview[0.044]; interview[0.044]; America[0.044]; January[0.044]; published[0.044]; stories[0.043]; shortly[0.043]; recently[0.042]; Puig[0.042]; Puig[0.042]; given[0.041]; experiences[0.041]; said[0.041]; said[0.041]; said[0.041]; Collapse[0.040]; World[0.040]; followed[0.040]; Saturday[0.040]; realism[0.039]; ideas[0.039]; modern[0.039]; people[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fulgencio Batista\u001b[39m ==> ENTITY: \u001b[32mFulgencio Batista\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCastro[0.047]; Castro[0.047]; Cuba[0.047]; Cuba[0.047]; Jose[0.046]; Jose[0.046]; Havana[0.045]; government[0.045]; Fidel[0.045]; Cuban[0.044]; Cuban[0.044]; Santiago[0.044]; revolution[0.043]; President[0.042]; shortly[0.042]; followed[0.041]; Soler[0.041]; Soler[0.041]; America[0.041]; published[0.040]; brought[0.040]; recently[0.040]; January[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Santiago de Cuba\u001b[39m ==> ENTITY: \u001b[32mSantiago de Cuba\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCuba[0.046]; Cuban[0.046]; Cuban[0.046]; Havana[0.044]; Castro[0.042]; Castro[0.042]; January[0.042]; Soler[0.042]; Soler[0.042]; Fidel[0.041]; Granma[0.041]; Granma[0.041]; native[0.041]; Jose[0.041]; Jose[0.041]; city[0.041]; day[0.041]; Saturday[0.040]; Batista[0.040]; America[0.040]; titles[0.040]; shortly[0.040]; Fulgencio[0.039]; government[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mGranma\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mGranma (yacht)\u001b[39m <---> \u001b[32mGranma (newspaper)\u001b[39m\t\nSCORES: global= 0.263:0.259[\u001b[31m0.003\u001b[39m]; local()= 0.114:0.135[\u001b[32m0.021\u001b[39m]; log p(e|m)= -0.270:-0.924[\u001b[31m0.653\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGranma[0.048]; Cuba[0.045]; Cuban[0.045]; published[0.044]; published[0.044]; newspaper[0.043]; city[0.043]; Saturday[0.042]; Saturday[0.042]; Castro[0.042]; Castro[0.042]; stories[0.041]; government[0.041]; Santiago[0.041]; January[0.040]; interview[0.040]; interview[0.040]; Fidel[0.040]; revolution[0.040]; reported[0.040]; Fulgencio[0.040]; shortly[0.040]; Batista[0.039]; day[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cuba ==> ENTITY: \u001b[32mCuba\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.159:0.159[0]; log p(e|m)= -0.240:-0.240[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCuba[0.053]; Cuban[0.046]; Cuban[0.046]; Castro[0.045]; Castro[0.045]; Jose[0.045]; Jose[0.045]; Santiago[0.044]; government[0.044]; Havana[0.044]; Fidel[0.043]; Soler[0.042]; Soler[0.042]; Batista[0.042]; titles[0.042]; President[0.042]; published[0.042]; Granma[0.042]; January[0.041]; official[0.041]; native[0.041]; brought[0.041]; revolution[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fidel Castro\u001b[39m ==> ENTITY: \u001b[32mFidel Castro\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCuba[0.047]; Cuba[0.047]; Castro[0.046]; Cuban[0.045]; Cuban[0.045]; Santiago[0.043]; Havana[0.043]; Jose[0.042]; Jose[0.042]; published[0.041]; President[0.041]; Batista[0.041]; America[0.041]; revolution[0.041]; government[0.041]; interview[0.040]; recently[0.040]; Saturday[0.040]; day[0.040]; Granma[0.039]; Granma[0.039]; shortly[0.039]; January[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cuban ==> ENTITY: \u001b[32mCuba\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.162:0.162[0]; log p(e|m)= -0.375:-0.375[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCuba[0.050]; Cuba[0.050]; Cuban[0.044]; Jose[0.043]; Jose[0.043]; Havana[0.043]; native[0.043]; Castro[0.042]; Castro[0.042]; recently[0.041]; revolution[0.041]; government[0.041]; acclaimed[0.040]; America[0.040]; Santiago[0.040]; Fidel[0.040]; Soler[0.040]; Soler[0.040]; Batista[0.040]; novelist[0.040]; modern[0.039]; Soul[0.039]; Granma[0.039]; Granma[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAmerica\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAmericas\u001b[39m <---> \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.263:0.262[\u001b[31m0.001\u001b[39m]; local()= 0.090:0.076[\u001b[31m0.014\u001b[39m]; log p(e|m)= -1.666:-0.785[\u001b[32m0.881\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbook[0.044]; life[0.043]; life[0.043]; Soul[0.043]; novel[0.043]; native[0.043]; novels[0.042]; languages[0.042]; stories[0.042]; interview[0.042]; interview[0.042]; Woman[0.041]; recently[0.041]; author[0.041]; published[0.041]; books[0.041]; experiences[0.041]; Cuba[0.041]; gift[0.041]; deals[0.040]; called[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cuban ==> ENTITY: \u001b[32mCuba\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.152:0.152[0]; log p(e|m)= -0.375:-0.375[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCuba[0.050]; Cuba[0.050]; Cuban[0.044]; Jose[0.044]; Jose[0.044]; Havana[0.043]; native[0.043]; Castro[0.043]; Castro[0.043]; revolution[0.041]; government[0.041]; acclaimed[0.041]; Santiago[0.040]; Fidel[0.040]; Soler[0.040]; Soler[0.040]; Batista[0.040]; novelist[0.040]; Granma[0.039]; brought[0.039]; deals[0.039]; titles[0.038]; President[0.038]; translate[0.038]; \t\n\r [=================>............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s730ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1885/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1123testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.081:0.081[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.046]; Bbc[0.044]; Saturday[0.043]; radio[0.043]; radio[0.043]; saying[0.043]; Friday[0.043]; near[0.042]; late[0.042]; security[0.042]; services[0.042]; election[0.042]; people[0.041]; people[0.041]; people[0.041]; authorities[0.041]; blast[0.040]; blast[0.040]; reported[0.040]; report[0.040]; Broadcasting[0.040]; Islamic[0.040]; quoted[0.040]; expected[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Algerian ==> ENTITY: \u001b[32mAlgeria\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.209:-0.209[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlgeria[0.048]; Algeria[0.048]; violence[0.044]; Algiers[0.043]; political[0.043]; security[0.043]; army[0.043]; win[0.042]; Islamic[0.041]; services[0.041]; town[0.041]; people[0.041]; people[0.041]; people[0.041]; killed[0.040]; killed[0.040]; backed[0.040]; Corporation[0.040]; election[0.040]; Friday[0.040]; saying[0.039]; British[0.039]; expected[0.039]; general[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Islamic ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.113:0.113[0]; log p(e|m)= -0.212:-0.212[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfundamentalists[0.044]; security[0.044]; people[0.044]; people[0.044]; people[0.044]; political[0.043]; services[0.042]; British[0.042]; saying[0.042]; Algeria[0.042]; Algeria[0.042]; Algerian[0.042]; authorities[0.041]; estimated[0.041]; backed[0.041]; expected[0.041]; general[0.041]; Friday[0.040]; late[0.040]; violence[0.040]; army[0.040]; die[0.040]; quoted[0.040]; near[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Algeria ==> ENTITY: \u001b[32mAlgeria\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.337:-0.337[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlgeria[0.049]; army[0.047]; security[0.045]; Algerian[0.044]; political[0.044]; Algiers[0.043]; win[0.042]; town[0.042]; violence[0.041]; Islamic[0.041]; Friday[0.041]; people[0.041]; people[0.041]; people[0.041]; British[0.041]; late[0.040]; general[0.040]; backed[0.040]; services[0.040]; near[0.040]; Saturday[0.039]; killed[0.039]; killed[0.039]; expected[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Algiers ==> ENTITY: \u001b[32mAlgiers\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.129:0.129[0]; log p(e|m)= -0.104:-0.104[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntown[0.050]; Algeria[0.049]; Algeria[0.049]; Algerian[0.049]; estimated[0.045]; near[0.044]; army[0.044]; violence[0.043]; general[0.043]; Saturday[0.043]; Friday[0.043]; late[0.043]; security[0.042]; expected[0.042]; Islamic[0.042]; authorities[0.042]; people[0.042]; people[0.042]; people[0.042]; killed[0.041]; killed[0.041]; exploded[0.041]; British[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = BBC ==> ENTITY: \u001b[32mBBC\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.100:0.100[0]; log p(e|m)= -0.014:-0.014[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.047]; Broadcasting[0.045]; London[0.045]; quoted[0.044]; radio[0.044]; radio[0.044]; reported[0.044]; report[0.044]; saying[0.044]; services[0.042]; Corporation[0.042]; Friday[0.040]; people[0.040]; people[0.040]; people[0.040]; late[0.040]; general[0.039]; backed[0.039]; Saturday[0.039]; security[0.039]; election[0.039]; Islamic[0.039]; expected[0.039]; political[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = British Broadcasting Corporation\u001b[39m ==> ENTITY: \u001b[32mBBC\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.073:0.073[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBbc[0.051]; London[0.046]; radio[0.046]; radio[0.046]; Saturday[0.045]; Friday[0.043]; late[0.043]; reported[0.042]; quoted[0.042]; report[0.041]; army[0.041]; services[0.041]; grenade[0.041]; blast[0.040]; blast[0.040]; saying[0.040]; injured[0.040]; backed[0.039]; Algiers[0.039]; win[0.039]; expected[0.039]; general[0.039]; security[0.038]; estimated[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Algeria ==> ENTITY: \u001b[32mAlgeria\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.337:-0.337[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlgeria[0.049]; army[0.047]; security[0.045]; Algerian[0.044]; political[0.044]; Algiers[0.043]; win[0.042]; town[0.042]; violence[0.041]; Islamic[0.041]; Friday[0.041]; people[0.041]; people[0.041]; people[0.041]; British[0.041]; late[0.040]; general[0.040]; backed[0.040]; services[0.040]; near[0.040]; Saturday[0.039]; killed[0.039]; killed[0.039]; expected[0.039]; \t\n\r [=================>............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s756ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1893/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 952testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lithuania ==> ENTITY: \u001b[32mLithuania national football team\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.151:0.151[0]; log p(e|m)= -2.129:-2.129[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLithuania[0.055]; Lithuania[0.055]; soccer[0.053]; Romania[0.052]; Romania[0.052]; Romania[0.052]; Soccer[0.052]; match[0.052]; European[0.049]; Match[0.048]; beat[0.046]; Bucharest[0.046]; Scorers[0.045]; halftime[0.044]; Beat[0.044]; Mihai[0.044]; Friday[0.044]; Cosmin[0.043]; Contra[0.041]; Attendance[0.041]; Danius[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cosmin Contra\u001b[39m ==> ENTITY: \u001b[32mCosmin Contra\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.112:0.112[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.058]; soccer[0.055]; match[0.054]; Romania[0.053]; Romania[0.053]; Romania[0.053]; Match[0.052]; European[0.051]; Friday[0.050]; halftime[0.050]; Bucharest[0.050]; Scorers[0.049]; Attendance[0.049]; Lithuania[0.047]; Lithuania[0.047]; Lithuania[0.047]; Mihai[0.046]; Danius[0.046]; beat[0.045]; Beat[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mihai Tararache\u001b[39m ==> ENTITY: \u001b[32mMihai Tararache\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.089:0.089[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.052]; Cosmin[0.051]; Soccer[0.051]; Romania[0.050]; Romania[0.050]; Romania[0.050]; match[0.049]; halftime[0.048]; Bucharest[0.048]; Scorers[0.048]; Attendance[0.048]; Lithuania[0.046]; Lithuania[0.046]; Lithuania[0.046]; beat[0.046]; Match[0.046]; European[0.046]; Beat[0.045]; Danius[0.045]; Friday[0.045]; Contra[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bucharest ==> ENTITY: \u001b[32mBucharest\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.018:-0.018[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRomania[0.055]; Romania[0.055]; Romania[0.055]; match[0.055]; soccer[0.053]; Soccer[0.053]; Cosmin[0.052]; European[0.051]; Mihai[0.051]; Match[0.050]; Friday[0.050]; Lithuania[0.049]; Lithuania[0.049]; Lithuania[0.049]; beat[0.048]; Attendance[0.046]; Beat[0.046]; halftime[0.045]; Scorers[0.045]; Contra[0.044]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEuropean\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUEFA\u001b[39m <---> \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.241:0.238[\u001b[31m0.003\u001b[39m]; local()= 0.092:0.023[\u001b[31m0.069\u001b[39m]; log p(e|m)= -3.576:-0.256[\u001b[32m3.319\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.057]; Soccer[0.056]; match[0.055]; Match[0.053]; Romania[0.053]; Romania[0.053]; Romania[0.053]; Lithuania[0.051]; Lithuania[0.051]; Lithuania[0.051]; Bucharest[0.051]; beat[0.049]; Cosmin[0.048]; Attendance[0.047]; Friday[0.047]; Mihai[0.046]; Beat[0.045]; halftime[0.045]; Contra[0.045]; Scorers[0.044]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mLithuania\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mLithuania national football team\u001b[39m <---> \u001b[32mLithuania national under-21 football team\u001b[39m\t\nSCORES: global= 0.265:0.217[\u001b[31m0.049\u001b[39m]; local()= 0.151:0.129[\u001b[31m0.023\u001b[39m]; log p(e|m)= -2.129:-3.912[\u001b[31m1.783\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLithuania[0.055]; Lithuania[0.055]; soccer[0.053]; Romania[0.052]; Romania[0.052]; Romania[0.052]; Soccer[0.052]; match[0.052]; European[0.049]; Match[0.048]; beat[0.046]; Bucharest[0.046]; Scorers[0.045]; halftime[0.044]; Beat[0.044]; Mihai[0.044]; Friday[0.044]; Cosmin[0.043]; Contra[0.041]; Attendance[0.041]; Danius[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mLithuania\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mLithuania national football team\u001b[39m <---> \u001b[32mLithuania national under-21 football team\u001b[39m\t\nSCORES: global= 0.265:0.217[\u001b[31m0.049\u001b[39m]; local()= 0.151:0.129[\u001b[31m0.023\u001b[39m]; log p(e|m)= -2.129:-3.912[\u001b[31m1.783\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLithuania[0.055]; Lithuania[0.055]; soccer[0.053]; Romania[0.052]; Romania[0.052]; Romania[0.052]; Soccer[0.052]; match[0.052]; European[0.049]; Match[0.048]; beat[0.046]; Bucharest[0.046]; Scorers[0.045]; halftime[0.044]; Beat[0.044]; Mihai[0.044]; Friday[0.044]; Cosmin[0.043]; Contra[0.041]; Attendance[0.041]; Danius[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mRomania\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRomania\u001b[39m <---> \u001b[32mRomania national football team\u001b[39m\t\nSCORES: global= 0.249:0.248[\u001b[31m0.001\u001b[39m]; local()= 0.124:0.163[\u001b[32m0.039\u001b[39m]; log p(e|m)= -0.212:-3.058[\u001b[31m2.846\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRomania[0.057]; Romania[0.057]; Lithuania[0.054]; Lithuania[0.054]; Lithuania[0.054]; soccer[0.052]; match[0.052]; Bucharest[0.051]; European[0.050]; Soccer[0.050]; halftime[0.049]; Beat[0.049]; beat[0.049]; Match[0.048]; Mihai[0.048]; Cosmin[0.047]; Scorers[0.047]; Contra[0.045]; Friday[0.044]; Attendance[0.042]; \t\n\r [=================>............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s743ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1903/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1078testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gascoigne\u001b[39m ==> ENTITY: \u001b[32mPaul Gascoigne\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.049]; England[0.049]; Gascoigne[0.045]; Gascoigne[0.045]; Hoddle[0.044]; player[0.043]; team[0.043]; Paul[0.043]; Paul[0.043]; players[0.042]; Ince[0.041]; Ince[0.041]; Ince[0.041]; Glenn[0.041]; joking[0.040]; Milan[0.040]; midfielder[0.039]; stadium[0.039]; Saturday[0.039]; Soccer[0.039]; Inter[0.039]; mate[0.039]; Moldova[0.038]; Moldova[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMoldova\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMoldova national football team\u001b[39m <---> \u001b[32mMoldova\u001b[39m\t\nSCORES: global= 0.260:0.254[\u001b[31m0.007\u001b[39m]; local()= 0.194:0.096[\u001b[31m0.098\u001b[39m]; log p(e|m)= -2.419:-0.473[\u001b[32m1.946\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoldova[0.048]; Moldova[0.048]; England[0.047]; England[0.047]; England[0.047]; team[0.046]; Soccer[0.045]; players[0.044]; player[0.044]; qualifier[0.044]; stadium[0.043]; stadium[0.043]; Chisinau[0.043]; Chisinau[0.043]; Cup[0.042]; midfielder[0.042]; Sunday[0.041]; Saturday[0.041]; group[0.040]; coaching[0.040]; clash[0.040]; press[0.040]; World[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moldova ==> ENTITY: \u001b[32mMoldova national football team\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.168:0.168[0]; log p(e|m)= -2.419:-2.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; England[0.047]; Moldova[0.046]; team[0.046]; players[0.045]; player[0.045]; qualifier[0.044]; stadium[0.043]; stadium[0.043]; Cup[0.042]; Sunday[0.041]; Saturday[0.041]; coaching[0.040]; group[0.040]; clash[0.040]; press[0.040]; World[0.040]; Chisinau[0.040]; Hoddle[0.039]; Hoddle[0.039]; Inter[0.039]; crowd[0.039]; time[0.039]; Glenn[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Inter Milan\u001b[39m ==> ENTITY: \u001b[32mInter Milan\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.047]; player[0.046]; players[0.045]; team[0.044]; Cup[0.043]; coaching[0.042]; England[0.042]; England[0.042]; England[0.042]; group[0.042]; midfielder[0.042]; qualifier[0.042]; stadium[0.042]; stadium[0.042]; time[0.041]; past[0.041]; crowd[0.041]; Sunday[0.039]; clash[0.039]; World[0.039]; tried[0.039]; press[0.039]; landed[0.039]; old[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ince\u001b[39m ==> ENTITY: \u001b[32mPaul Ince\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInce[0.045]; Ince[0.045]; England[0.045]; England[0.045]; England[0.045]; team[0.044]; Glenn[0.044]; Hoddle[0.043]; Hoddle[0.043]; players[0.042]; Paul[0.042]; Paul[0.042]; player[0.041]; coaching[0.040]; qualifier[0.040]; midfielder[0.040]; Soccer[0.040]; time[0.040]; Gascoigne[0.040]; Gascoigne[0.040]; Gascoigne[0.040]; trouble[0.039]; joking[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paul Gascoigne\u001b[39m ==> ENTITY: \u001b[32mPaul Gascoigne\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.048]; England[0.048]; England[0.048]; Gascoigne[0.044]; Gascoigne[0.044]; Hoddle[0.043]; Hoddle[0.043]; player[0.042]; team[0.042]; Paul[0.042]; players[0.041]; coaching[0.040]; Ince[0.040]; Ince[0.040]; Ince[0.040]; Glenn[0.040]; joking[0.040]; Milan[0.040]; time[0.039]; midfielder[0.039]; stadium[0.039]; Saturday[0.039]; Soccer[0.039]; Inter[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hoddle\u001b[39m ==> ENTITY: \u001b[32mGlenn Hoddle\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.048]; England[0.048]; England[0.048]; Hoddle[0.045]; Ince[0.043]; Ince[0.043]; player[0.043]; players[0.042]; team[0.042]; Gascoigne[0.041]; Gascoigne[0.041]; Glenn[0.041]; Paul[0.041]; Paul[0.041]; coaching[0.040]; qualifier[0.040]; Cup[0.040]; time[0.040]; midfielder[0.040]; declined[0.040]; Inter[0.040]; past[0.039]; Milan[0.038]; World[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chisinau\u001b[39m ==> ENTITY: \u001b[32mChișinău\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.091:0.091[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoldova[0.045]; Moldova[0.045]; Moldova[0.045]; coaching[0.044]; qualifier[0.043]; Chisinau[0.042]; Saturday[0.042]; past[0.042]; midfielder[0.041]; time[0.041]; Sunday[0.041]; old[0.041]; press[0.041]; stadium[0.041]; stadium[0.041]; Soccer[0.041]; team[0.041]; declined[0.041]; World[0.041]; Paul[0.040]; Paul[0.040]; landed[0.040]; Inter[0.040]; practical[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ince\u001b[39m ==> ENTITY: \u001b[32mPaul Ince\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInce[0.045]; Ince[0.045]; England[0.045]; England[0.045]; team[0.045]; Glenn[0.045]; Hoddle[0.044]; players[0.043]; Paul[0.043]; Paul[0.043]; player[0.042]; midfielder[0.041]; Soccer[0.040]; Gascoigne[0.040]; Gascoigne[0.040]; Gascoigne[0.040]; trouble[0.040]; joking[0.040]; Saturday[0.040]; mate[0.039]; stadium[0.039]; Inter[0.039]; Milan[0.039]; past[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMoldova\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMoldova national football team\u001b[39m <---> \u001b[32mMoldova\u001b[39m\t\nSCORES: global= 0.259:0.253[\u001b[31m0.007\u001b[39m]; local()= 0.159:0.058[\u001b[31m0.101\u001b[39m]; log p(e|m)= -2.419:-0.473[\u001b[32m1.946\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoldova[0.047]; England[0.047]; England[0.047]; team[0.046]; Soccer[0.045]; players[0.044]; player[0.044]; stadium[0.043]; Chisinau[0.043]; Chisinau[0.043]; midfielder[0.041]; Saturday[0.041]; group[0.040]; coaching[0.040]; clash[0.040]; press[0.040]; Hoddle[0.039]; Hoddle[0.039]; Inter[0.039]; crowd[0.039]; Latest[0.039]; Glenn[0.039]; Milan[0.038]; past[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland national football team\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.153:0.153[0]; log p(e|m)= -2.781:-2.781[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.050]; team[0.045]; players[0.044]; player[0.044]; Cup[0.043]; qualifier[0.042]; coaching[0.042]; Hoddle[0.042]; Hoddle[0.042]; stadium[0.041]; stadium[0.041]; time[0.041]; Saturday[0.041]; Ince[0.041]; Ince[0.041]; Inter[0.040]; Gascoigne[0.040]; Gascoigne[0.040]; Glenn[0.040]; Paul[0.040]; Paul[0.040]; Milan[0.040]; Sunday[0.040]; World[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Glenn Hoddle\u001b[39m ==> ENTITY: \u001b[32mGlenn Hoddle\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.048]; England[0.048]; England[0.048]; Hoddle[0.045]; Ince[0.042]; Ince[0.042]; Ince[0.042]; player[0.042]; players[0.042]; team[0.042]; Gascoigne[0.041]; Gascoigne[0.041]; Gascoigne[0.041]; Soccer[0.040]; Paul[0.040]; Paul[0.040]; coaching[0.040]; qualifier[0.040]; Cup[0.040]; time[0.039]; midfielder[0.039]; declined[0.039]; Inter[0.039]; past[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paul Ince\u001b[39m ==> ENTITY: \u001b[32mPaul Ince\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInce[0.045]; Ince[0.045]; England[0.045]; England[0.045]; England[0.045]; team[0.044]; Glenn[0.044]; Hoddle[0.043]; Hoddle[0.043]; players[0.042]; Paul[0.042]; player[0.041]; coaching[0.040]; qualifier[0.040]; midfielder[0.040]; Soccer[0.040]; time[0.040]; Gascoigne[0.040]; Gascoigne[0.040]; Gascoigne[0.040]; trouble[0.039]; joking[0.039]; Saturday[0.039]; Cup[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chisinau\u001b[39m ==> ENTITY: \u001b[32mChișinău\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.065:0.065[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoldova[0.046]; Moldova[0.046]; coaching[0.045]; Chisinau[0.043]; Saturday[0.042]; past[0.042]; midfielder[0.042]; old[0.042]; press[0.042]; stadium[0.042]; Soccer[0.042]; team[0.042]; Paul[0.041]; Paul[0.041]; landed[0.041]; Inter[0.041]; practical[0.040]; player[0.040]; Latest[0.040]; Glenn[0.040]; England[0.040]; England[0.040]; tried[0.040]; players[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland national football team\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.157:0.157[0]; log p(e|m)= -2.781:-2.781[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.049]; England[0.049]; team[0.044]; players[0.043]; Soccer[0.043]; player[0.043]; Cup[0.042]; qualifier[0.042]; coaching[0.041]; Hoddle[0.041]; Hoddle[0.041]; midfielder[0.041]; stadium[0.041]; stadium[0.041]; time[0.040]; Saturday[0.040]; Ince[0.040]; Ince[0.040]; Ince[0.040]; Inter[0.040]; Gascoigne[0.039]; Gascoigne[0.039]; Gascoigne[0.039]; Glenn[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEngland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEngland national football team\u001b[39m <---> \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.249:0.240[\u001b[31m0.009\u001b[39m]; local()= 0.144:0.056[\u001b[31m0.089\u001b[39m]; log p(e|m)= -2.781:-0.190[\u001b[32m2.591\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.052]; team[0.047]; players[0.046]; Soccer[0.046]; player[0.045]; coaching[0.044]; Hoddle[0.044]; Hoddle[0.044]; midfielder[0.043]; stadium[0.043]; time[0.043]; Saturday[0.042]; Ince[0.042]; Ince[0.042]; Ince[0.042]; Inter[0.042]; Gascoigne[0.042]; Gascoigne[0.042]; Gascoigne[0.042]; Glenn[0.042]; Paul[0.042]; Paul[0.042]; Milan[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gascoigne\u001b[39m ==> ENTITY: \u001b[32mPaul Gascoigne\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.050]; England[0.050]; England[0.050]; Gascoigne[0.045]; Gascoigne[0.045]; Hoddle[0.045]; Hoddle[0.045]; player[0.044]; team[0.044]; Paul[0.043]; Paul[0.043]; players[0.043]; coaching[0.042]; Ince[0.042]; Ince[0.042]; Ince[0.042]; qualifier[0.042]; Glenn[0.042]; joking[0.041]; Milan[0.041]; Cup[0.041]; time[0.040]; midfielder[0.040]; \t\n\r [=================>............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s659ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1921/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1092testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.021:0.021[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.045]; Tennis[0.045]; Tennis[0.045]; Saturday[0.044]; Saturday[0.044]; Championships[0.043]; National[0.043]; round[0.043]; round[0.043]; Women[0.041]; Judith[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; Zvereva[0.041]; Natasha[0.041]; Open[0.041]; Open[0.041]; number[0.040]; Men[0.040]; Japan[0.040]; Naoko[0.039]; Republic[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Martina Hingis\u001b[39m ==> ENTITY: \u001b[32mMartina Hingis\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.232:0.232[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.044]; Open[0.044]; Tennis[0.044]; Tennis[0.044]; Tennis[0.044]; Zvereva[0.043]; singles[0.043]; singles[0.043]; round[0.042]; round[0.042]; Graf[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; Steffi[0.041]; Enqvist[0.041]; Switzerland[0.041]; Czech[0.041]; Pablo[0.040]; number[0.039]; Natasha[0.039]; Austria[0.038]; Championships[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEcuador\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEcuador national football team\u001b[39m <---> \u001b[32mEcuador\u001b[39m\t\nSCORES: global= 0.247:0.244[\u001b[31m0.003\u001b[39m]; local()= 0.119:0.052[\u001b[31m0.067\u001b[39m]; log p(e|m)= -2.397:-0.214[\u001b[32m2.182\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.046]; Tennis[0.046]; Tennis[0.046]; Men[0.045]; Switzerland[0.044]; Women[0.043]; Championships[0.042]; Pablo[0.042]; Germany[0.042]; Sweden[0.042]; Austria[0.042]; Japan[0.041]; Belarus[0.041]; Open[0.040]; Open[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; Saturday[0.040]; Saturday[0.040]; singles[0.040]; singles[0.040]; denotes[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech Republic ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.044]; Sweden[0.044]; Championships[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; Tennis[0.042]; Tennis[0.042]; Tennis[0.042]; Austria[0.042]; Germany[0.042]; Open[0.042]; Open[0.042]; Belarus[0.042]; singles[0.041]; singles[0.041]; round[0.041]; round[0.041]; Martina[0.041]; Graf[0.041]; National[0.040]; number[0.040]; Men[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Switzerland ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.100:0.100[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.045]; Austria[0.044]; round[0.044]; round[0.044]; Germany[0.043]; Belarus[0.043]; Tennis[0.042]; Tennis[0.042]; Tennis[0.042]; Ecuador[0.042]; Men[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; Championships[0.042]; Republic[0.040]; Martina[0.040]; singles[0.040]; singles[0.040]; Saturday[0.040]; Saturday[0.040]; Japan[0.040]; Czech[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Natasha Zvereva\u001b[39m ==> ENTITY: \u001b[32mNatasha Zvereva\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMartina[0.046]; Graf[0.045]; Tennis[0.044]; Tennis[0.044]; Tennis[0.044]; round[0.043]; round[0.043]; Steffi[0.043]; Enqvist[0.043]; Open[0.042]; Open[0.042]; singles[0.042]; singles[0.042]; number[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; Czech[0.040]; Judith[0.040]; Hingis[0.040]; Championships[0.039]; Thomas[0.039]; Pablo[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Thomas Enqvist\u001b[39m ==> ENTITY: \u001b[32mThomas Enqvist\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.044]; Tennis[0.044]; Tennis[0.044]; round[0.043]; round[0.043]; Open[0.043]; Open[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; Pablo[0.042]; singles[0.042]; singles[0.042]; Martina[0.041]; Championships[0.041]; Graf[0.040]; Zvereva[0.040]; Steffi[0.040]; Czech[0.040]; Men[0.039]; Sweden[0.039]; number[0.039]; Switzerland[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japan ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.075:0.075[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChampionships[0.046]; Switzerland[0.044]; Men[0.043]; Germany[0.043]; Austria[0.043]; Belarus[0.043]; Women[0.042]; Ecuador[0.042]; Republic[0.042]; round[0.042]; round[0.042]; Sweden[0.042]; National[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; Saturday[0.041]; Saturday[0.041]; denotes[0.040]; Tennis[0.040]; Tennis[0.040]; Tennis[0.040]; number[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U S Open ==> ENTITY: \u001b[32mUS Open (tennis)\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.223:0.223[0]; log p(e|m)= -1.252:-1.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.048]; Tennis[0.048]; Tennis[0.048]; Championships[0.045]; Open[0.043]; Open[0.043]; singles[0.042]; singles[0.042]; round[0.042]; round[0.042]; Martina[0.042]; Enqvist[0.041]; Graf[0.041]; National[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; Pablo[0.039]; Men[0.039]; Steffi[0.039]; Saturday[0.039]; Saturday[0.039]; Zvereva[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Steffi Graf\u001b[39m ==> ENTITY: \u001b[32mSteffi Graf\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.206:0.206[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMartina[0.047]; Tennis[0.046]; Tennis[0.046]; Tennis[0.046]; Open[0.043]; Open[0.043]; round[0.042]; round[0.042]; Switzerland[0.042]; Enqvist[0.042]; Zvereva[0.041]; singles[0.041]; singles[0.041]; Czech[0.041]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; Championships[0.040]; Austria[0.040]; Hingis[0.040]; Thomas[0.039]; Pablo[0.039]; Sweden[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.102:0.102[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.046]; Austria[0.044]; Germany[0.044]; Championships[0.043]; Tennis[0.042]; Tennis[0.042]; Tennis[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; Open[0.042]; Open[0.042]; Belarus[0.042]; Women[0.041]; Ecuador[0.041]; round[0.041]; round[0.041]; singles[0.041]; singles[0.041]; National[0.041]; Japan[0.040]; Czech[0.040]; Martina[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Judith Wiesner\u001b[39m ==> ENTITY: \u001b[32mJudith Wiesner\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEnqvist[0.044]; Tennis[0.044]; Tennis[0.044]; Tennis[0.044]; Open[0.044]; Open[0.044]; Graf[0.042]; singles[0.042]; singles[0.042]; round[0.042]; round[0.042]; Championships[0.041]; Martina[0.041]; Switzerland[0.041]; Hingis[0.041]; Austria[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; Zvereva[0.040]; Steffi[0.040]; number[0.039]; Sweden[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belarus ==> ENTITY: \u001b[32mBelarus\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.422:-0.422[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMen[0.049]; Austria[0.044]; Sweden[0.043]; Championships[0.043]; Germany[0.043]; Switzerland[0.043]; Czech[0.042]; Women[0.042]; Tennis[0.042]; Tennis[0.042]; Tennis[0.042]; Republic[0.042]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; Ecuador[0.040]; round[0.040]; round[0.040]; singles[0.040]; singles[0.040]; National[0.040]; number[0.039]; Japan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.S. Open Tennis Championships\u001b[39m ==> ENTITY: \u001b[32mUS Open (tennis)\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.208:0.208[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.050]; Tennis[0.050]; Open[0.045]; singles[0.043]; singles[0.043]; round[0.043]; round[0.043]; Martina[0.043]; Enqvist[0.042]; Graf[0.042]; Championships[0.041]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; Pablo[0.040]; Men[0.040]; Steffi[0.040]; Zvereva[0.040]; number[0.039]; Saturday[0.039]; Saturday[0.039]; Czech[0.039]; Women[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Austria ==> ENTITY: \u001b[32mAustria\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.111:0.111[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.046]; Germany[0.045]; singles[0.044]; singles[0.044]; Tennis[0.043]; Tennis[0.043]; Tennis[0.043]; Open[0.042]; Open[0.042]; Belarus[0.042]; Sweden[0.042]; round[0.041]; round[0.041]; Ecuador[0.041]; Republic[0.041]; Women[0.041]; Graf[0.041]; Enqvist[0.040]; National[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; Czech[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.073:0.073[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustria[0.046]; Switzerland[0.046]; Sweden[0.044]; Championships[0.044]; Republic[0.043]; Belarus[0.043]; Tennis[0.042]; Tennis[0.042]; Tennis[0.042]; round[0.042]; round[0.042]; Czech[0.042]; seeding[0.041]; Women[0.041]; Japan[0.040]; Ecuador[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; Men[0.040]; Open[0.040]; Open[0.040]; Graf[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Petra Langrova\u001b[39m ==> ENTITY: \u001b[32mPetra Langrová\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMartina[0.047]; Republic[0.045]; Czech[0.045]; Enqvist[0.044]; Tennis[0.043]; Tennis[0.043]; Tennis[0.043]; Austria[0.042]; Switzerland[0.042]; Graf[0.042]; Germany[0.042]; Judith[0.041]; Steffi[0.041]; Championships[0.041]; Zvereva[0.040]; singles[0.040]; singles[0.040]; Pablo[0.040]; Belarus[0.040]; round[0.040]; round[0.040]; Sweden[0.040]; Hingis[0.040]; Ecuador[0.040]; \t\n\r [=================>............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s609ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1938/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1032testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chinese ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.129:0.129[0]; log p(e|m)= -1.419:-1.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChinese[0.051]; China[0.050]; Shanghai[0.045]; local[0.044]; Ling[0.043]; Tong[0.043]; province[0.043]; Hong[0.042]; Kong[0.041]; Jiangsu[0.041]; set[0.041]; Wuxi[0.041]; Wuxi[0.041]; group[0.041]; based[0.041]; growing[0.041]; Pacific[0.040]; million[0.040]; existing[0.039]; tap[0.039]; telecommunications[0.039]; linking[0.039]; venture[0.038]; venture[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.059:0.059[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmillion[0.046]; major[0.045]; major[0.045]; billion[0.043]; Europe[0.042]; percent[0.042]; line[0.042]; China[0.042]; China[0.042]; distribution[0.042]; distribution[0.042]; coming[0.042]; listed[0.041]; sales[0.041]; systems[0.041]; systems[0.041]; profits[0.041]; group[0.041]; establish[0.041]; company[0.041]; industrial[0.041]; global[0.040]; starting[0.040]; value[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pirelli\u001b[39m ==> ENTITY: \u001b[32mPirelli\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPirelli[0.049]; Pirelli[0.049]; Pirelli[0.049]; Pirelli[0.049]; Pirelli[0.049]; industrial[0.042]; industrial[0.042]; Italian[0.042]; Paula[0.041]; companies[0.041]; technical[0.041]; Milan[0.040]; news[0.040]; news[0.040]; second[0.039]; networks[0.039]; example[0.039]; central[0.039]; million[0.039]; million[0.039]; marks[0.039]; late[0.039]; compared[0.038]; September[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Europe ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.106:0.106[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsales[0.044]; government[0.044]; Asia[0.043]; profits[0.043]; company[0.043]; million[0.042]; distribution[0.042]; distribution[0.042]; trading[0.042]; investments[0.042]; global[0.042]; percent[0.041]; industrial[0.041]; billion[0.041]; focusing[0.041]; group[0.041]; China[0.041]; China[0.041]; major[0.041]; major[0.041]; large[0.041]; carriers[0.041]; supplier[0.041]; strategic[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pirelli\u001b[39m ==> ENTITY: \u001b[32mPirelli\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPirelli[0.048]; Pirelli[0.048]; Pirelli[0.048]; Pirelli[0.048]; Pirelli[0.048]; industrial[0.042]; industrial[0.042]; Shanghai[0.042]; Italian[0.042]; Paula[0.041]; companies[0.041]; technical[0.040]; company[0.040]; technology[0.040]; factory[0.040]; telecommunications[0.040]; news[0.039]; partnership[0.039]; example[0.039]; partners[0.039]; million[0.038]; million[0.038]; marks[0.038]; late[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = CITIC Pacific\u001b[39m ==> ENTITY: \u001b[32mCITIC Pacific\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncompany[0.047]; China[0.045]; Shanghai[0.044]; million[0.043]; million[0.043]; Eastern[0.042]; Chinese[0.042]; Chinese[0.042]; Company[0.042]; venture[0.041]; venture[0.041]; division[0.041]; annual[0.041]; invest[0.041]; telecommunications[0.041]; announced[0.041]; market[0.041]; developing[0.040]; production[0.040]; growth[0.040]; partnership[0.040]; joint[0.040]; giant[0.040]; markets[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.125:0.125[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.050]; Chinese[0.047]; representing[0.045]; Asia[0.044]; Europe[0.043]; central[0.043]; establish[0.043]; second[0.042]; Hong[0.041]; government[0.041]; group[0.041]; chairman[0.041]; starting[0.041]; base[0.040]; global[0.040]; involvement[0.040]; executive[0.040]; major[0.040]; major[0.040]; said[0.040]; trading[0.040]; chief[0.040]; billion[0.039]; East[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChinese[0.049]; Chinese[0.049]; Shanghai[0.045]; Wuxi[0.043]; Wuxi[0.043]; Wuxi[0.043]; Kong[0.043]; Hong[0.042]; Ling[0.042]; Jiangsu[0.042]; Tong[0.041]; production[0.041]; announced[0.041]; partnership[0.041]; partners[0.041]; province[0.040]; Friday[0.040]; joint[0.040]; based[0.039]; include[0.039]; set[0.039]; venture[0.039]; venture[0.039]; Pacific[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italian ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.096:0.096[0]; log p(e|m)= -0.759:-0.759[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMilan[0.048]; lire[0.046]; late[0.045]; growing[0.044]; companies[0.044]; September[0.044]; present[0.044]; exporting[0.044]; said[0.043]; Chinese[0.043]; Chinese[0.043]; showed[0.043]; Eastern[0.043]; widely[0.043]; province[0.043]; million[0.043]; million[0.043]; know[0.043]; group[0.043]; reach[0.042]; control[0.042]; percent[0.042]; expand[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pirelli\u001b[39m ==> ENTITY: \u001b[32mPirelli\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPirelli[0.049]; Pirelli[0.049]; Pirelli[0.049]; Pirelli[0.049]; industrial[0.042]; industrial[0.042]; Italian[0.042]; Paula[0.042]; companies[0.041]; technical[0.041]; Milan[0.041]; news[0.040]; line[0.040]; starting[0.040]; second[0.040]; networks[0.039]; networks[0.039]; example[0.039]; representing[0.039]; central[0.039]; million[0.039]; million[0.039]; late[0.039]; compared[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = India ==> ENTITY: \u001b[32mIndia\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.058:0.058[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMalaysia[0.046]; technology[0.044]; companies[0.043]; Indonesia[0.043]; Eastern[0.042]; million[0.042]; million[0.042]; Chinese[0.042]; reach[0.042]; time[0.042]; percent[0.042]; company[0.041]; existing[0.041]; existing[0.041]; September[0.041]; half[0.041]; province[0.041]; partnership[0.041]; widely[0.041]; markets[0.041]; markets[0.041]; cable[0.041]; compared[0.040]; partners[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jiangsu\u001b[39m ==> ENTITY: \u001b[32mJiangsu\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChinese[0.045]; Chinese[0.045]; Wuxi[0.044]; Wuxi[0.044]; Wuxi[0.044]; China[0.043]; Shanghai[0.043]; expected[0.043]; province[0.043]; Tong[0.042]; joint[0.041]; announced[0.041]; September[0.041]; Hong[0.041]; late[0.040]; expand[0.040]; expect[0.040]; set[0.040]; division[0.040]; industrial[0.040]; industrial[0.040]; Ling[0.040]; Eastern[0.040]; India[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = CITIC Pacific\u001b[39m ==> ENTITY: \u001b[32mCITIC Pacific\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncompany[0.049]; China[0.047]; China[0.047]; investments[0.046]; million[0.045]; profits[0.044]; billion[0.044]; Asia[0.044]; global[0.044]; strategic[0.043]; supplier[0.043]; sales[0.043]; trading[0.043]; executive[0.043]; ventures[0.042]; production[0.042]; value[0.042]; major[0.042]; major[0.042]; joint[0.042]; property[0.041]; trunk[0.041]; chairman[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chinese ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.136:0.136[0]; log p(e|m)= -1.419:-1.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChinese[0.051]; China[0.049]; Shanghai[0.045]; local[0.043]; Kong[0.043]; Ling[0.043]; Tong[0.043]; Hong[0.043]; province[0.043]; production[0.042]; Jiangsu[0.041]; Wuxi[0.041]; Wuxi[0.041]; Wuxi[0.041]; group[0.040]; growing[0.040]; venture[0.039]; venture[0.039]; Pacific[0.039]; million[0.039]; million[0.039]; marks[0.039]; expected[0.039]; tap[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pirelli\u001b[39m ==> ENTITY: \u001b[32mPirelli\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPirelli[0.049]; Pirelli[0.049]; Pirelli[0.049]; Pirelli[0.049]; Pirelli[0.049]; industrial[0.042]; industrial[0.042]; Italian[0.042]; Paula[0.041]; companies[0.041]; technical[0.041]; Milan[0.040]; news[0.040]; news[0.040]; second[0.039]; networks[0.039]; example[0.039]; central[0.039]; million[0.039]; million[0.039]; marks[0.039]; late[0.039]; compared[0.038]; September[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.095:0.095[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAsia[0.046]; Kong[0.045]; Hong[0.045]; Europe[0.044]; chairman[0.043]; joint[0.042]; representing[0.042]; global[0.042]; executive[0.042]; major[0.041]; major[0.041]; said[0.041]; trading[0.041]; chief[0.041]; billion[0.041]; East[0.041]; Pacific[0.041]; focusing[0.041]; industrial[0.040]; listed[0.040]; distribution[0.040]; supplier[0.040]; investments[0.040]; coming[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Milan ==> ENTITY: \u001b[32mMilan\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.066:0.066[0]; log p(e|m)= -0.301:-0.301[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItalian[0.048]; operate[0.044]; million[0.043]; based[0.042]; long[0.042]; plant[0.042]; telecommunications[0.042]; announced[0.042]; industrial[0.042]; linking[0.042]; existing[0.042]; existing[0.042]; cable[0.041]; growing[0.041]; Friday[0.041]; market[0.041]; Shanghai[0.041]; partnership[0.041]; local[0.041]; set[0.041]; company[0.040]; awaited[0.040]; factory[0.040]; province[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Asia ==> ENTITY: \u001b[32mAsia\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.113:0.113[0]; log p(e|m)= -0.150:-0.150[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurope[0.048]; China[0.044]; China[0.044]; China[0.044]; global[0.044]; chief[0.043]; executive[0.043]; government[0.042]; East[0.042]; Chinese[0.042]; chairman[0.041]; presence[0.041]; officer[0.041]; million[0.040]; million[0.040]; market[0.040]; market[0.040]; group[0.040]; second[0.040]; central[0.040]; Pacific[0.040]; demand[0.040]; demand[0.040]; representing[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pirelli\u001b[39m ==> ENTITY: \u001b[32mPirelli\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPirelli[0.049]; Pirelli[0.049]; Pirelli[0.049]; Pirelli[0.049]; Pirelli[0.049]; industrial[0.042]; industrial[0.042]; Italian[0.042]; Paula[0.041]; companies[0.041]; technical[0.040]; Milan[0.040]; technology[0.040]; news[0.040]; news[0.040]; networks[0.039]; example[0.039]; million[0.039]; million[0.039]; marks[0.039]; late[0.038]; compared[0.038]; September[0.038]; grow[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wuxi\u001b[39m ==> ENTITY: \u001b[32mWuxi\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWuxi[0.050]; Wuxi[0.050]; Chinese[0.047]; Shanghai[0.047]; China[0.047]; Jiangsu[0.047]; province[0.045]; long[0.044]; fast[0.043]; expected[0.042]; Eastern[0.042]; India[0.042]; Hong[0.042]; local[0.042]; Tong[0.042]; Kong[0.041]; half[0.041]; Citic[0.041]; joint[0.041]; factory[0.041]; set[0.041]; Ling[0.041]; expect[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pirelli\u001b[39m ==> ENTITY: \u001b[32mPirelli\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPirelli[0.050]; Pirelli[0.050]; Pirelli[0.050]; industrial[0.043]; Shanghai[0.043]; Italian[0.043]; David[0.042]; announced[0.042]; Milan[0.041]; company[0.041]; technology[0.041]; factory[0.041]; telecommunications[0.041]; giant[0.040]; partnership[0.040]; partners[0.040]; million[0.040]; million[0.040]; marks[0.039]; long[0.039]; growing[0.039]; growing[0.039]; include[0.039]; group[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chinese ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.143:0.143[0]; log p(e|m)= -1.419:-1.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChinese[0.051]; China[0.050]; China[0.050]; Asia[0.046]; local[0.044]; widely[0.042]; exporting[0.041]; group[0.041]; government[0.041]; companies[0.041]; central[0.040]; representing[0.040]; said[0.040]; said[0.040]; largest[0.040]; time[0.040]; million[0.039]; million[0.039]; expected[0.039]; expected[0.039]; second[0.039]; lines[0.039]; lines[0.039]; easier[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Malaysia ==> ENTITY: \u001b[32mMalaysia\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.068:0.068[0]; log p(e|m)= -0.202:-0.202[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.049]; India[0.048]; late[0.044]; fast[0.042]; Shanghai[0.042]; expected[0.041]; expected[0.041]; time[0.041]; Pirelli[0.041]; Pirelli[0.041]; Pirelli[0.041]; Pirelli[0.041]; Pirelli[0.041]; Italian[0.041]; division[0.041]; province[0.041]; expect[0.041]; Chinese[0.041]; technical[0.041]; exporting[0.040]; half[0.040]; developing[0.040]; developing[0.040]; compared[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = GMT\u001b[39m ==> ENTITY: \u001b[32mGreenwich Mean Time\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.100:0.100[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntime[0.044]; million[0.044]; million[0.044]; largest[0.043]; September[0.043]; news[0.042]; news[0.042]; late[0.042]; starting[0.042]; representing[0.042]; percent[0.042]; developing[0.041]; easier[0.041]; expected[0.041]; expected[0.041]; systems[0.041]; central[0.041]; example[0.041]; second[0.041]; base[0.041]; companies[0.041]; half[0.040]; group[0.040]; compared[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.118:0.118[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMalaysia[0.051]; India[0.048]; Chinese[0.045]; companies[0.045]; company[0.044]; time[0.044]; million[0.044]; million[0.044]; group[0.043]; widely[0.043]; percent[0.043]; local[0.043]; partnership[0.043]; marks[0.043]; operate[0.042]; shares[0.042]; shares[0.042]; province[0.042]; division[0.042]; showed[0.042]; management[0.042]; technical[0.041]; September[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wuxi\u001b[39m ==> ENTITY: \u001b[32mWuxi\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWuxi[0.047]; Wuxi[0.047]; Chinese[0.045]; Chinese[0.045]; Shanghai[0.045]; China[0.045]; Jiangsu[0.045]; province[0.043]; long[0.042]; fast[0.041]; expected[0.041]; tap[0.040]; Eastern[0.040]; India[0.040]; Hong[0.040]; local[0.040]; Tong[0.040]; Kong[0.040]; half[0.040]; Citic[0.039]; joint[0.039]; factory[0.039]; set[0.039]; Ling[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pirelli Cables\u001b[39m ==> ENTITY: \u001b[32mPirelli\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPirelli[0.052]; supplier[0.044]; strategic[0.043]; SpA[0.042]; major[0.042]; major[0.042]; company[0.042]; line[0.042]; second[0.042]; Europe[0.042]; optical[0.041]; Marco[0.041]; Asia[0.041]; joint[0.041]; Pacific[0.041]; trunk[0.041]; starting[0.041]; industrial[0.041]; optic[0.040]; distribution[0.040]; distribution[0.040]; global[0.040]; cables[0.040]; cables[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pirelli\u001b[39m ==> ENTITY: \u001b[32mPirelli\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPirelli[0.051]; Pirelli[0.051]; industrial[0.044]; Italian[0.044]; David[0.043]; announced[0.043]; Milan[0.042]; company[0.042]; factory[0.042]; telecommunications[0.042]; giant[0.041]; partnership[0.041]; long[0.040]; growing[0.040]; group[0.040]; joint[0.040]; called[0.040]; existing[0.040]; look[0.039]; Wuxi[0.039]; Wuxi[0.039]; set[0.039]; market[0.039]; Jones[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italian ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.069:0.069[0]; log p(e|m)= -0.759:-0.759[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMilan[0.047]; Chinese[0.045]; Chinese[0.045]; local[0.043]; growing[0.043]; telecommunications[0.043]; group[0.043]; based[0.043]; David[0.042]; linking[0.041]; called[0.041]; China[0.041]; province[0.041]; million[0.041]; Friday[0.041]; announced[0.041]; existing[0.041]; existing[0.041]; joint[0.040]; growth[0.040]; operate[0.040]; partnership[0.040]; set[0.040]; partners[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chinese ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.144:0.144[0]; log p(e|m)= -1.419:-1.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChinese[0.051]; China[0.050]; Indonesia[0.048]; Malaysia[0.047]; India[0.043]; widely[0.041]; exporting[0.041]; group[0.041]; Eastern[0.041]; companies[0.041]; growing[0.041]; said[0.040]; largest[0.040]; time[0.040]; million[0.040]; million[0.040]; expected[0.040]; expected[0.040]; expected[0.040]; second[0.039]; expect[0.039]; half[0.039]; showed[0.039]; percent[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.050]; Chinese[0.048]; Chinese[0.048]; Asia[0.044]; central[0.043]; establish[0.043]; second[0.042]; government[0.041]; chairman[0.041]; local[0.041]; starting[0.041]; percent[0.040]; base[0.040]; global[0.040]; involvement[0.040]; executive[0.040]; exporting[0.040]; large[0.040]; said[0.040]; chief[0.040]; billion[0.040]; showed[0.039]; technical[0.039]; largest[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pirelli\u001b[39m ==> ENTITY: \u001b[32mPirelli\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPirelli[0.049]; Pirelli[0.049]; Pirelli[0.049]; Pirelli[0.049]; industrial[0.042]; industrial[0.042]; Shanghai[0.042]; Italian[0.042]; Paula[0.042]; companies[0.041]; technical[0.041]; company[0.040]; technology[0.040]; factory[0.040]; news[0.040]; news[0.040]; partnership[0.039]; example[0.039]; partners[0.039]; million[0.039]; million[0.039]; marks[0.039]; late[0.039]; compared[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Shanghai ==> ENTITY: \u001b[32mShanghai\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.120:0.120[0]; log p(e|m)= -0.101:-0.101[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.045]; Chinese[0.045]; Chinese[0.045]; companies[0.042]; Malaysia[0.042]; Wuxi[0.042]; Wuxi[0.042]; Wuxi[0.042]; Pirelli[0.042]; Pirelli[0.042]; Pirelli[0.042]; Pirelli[0.042]; Kong[0.041]; September[0.041]; India[0.041]; Indonesia[0.041]; announced[0.041]; late[0.040]; Citic[0.040]; group[0.040]; group[0.040]; province[0.040]; expected[0.040]; Hong[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Milan ==> ENTITY: \u001b[32mMilan\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.301:-0.301[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItalian[0.047]; September[0.043]; starting[0.043]; second[0.043]; percent[0.042]; million[0.042]; million[0.042]; late[0.042]; central[0.042]; operator[0.042]; control[0.041]; expected[0.041]; expected[0.041]; establish[0.041]; necessitate[0.041]; lire[0.041]; industrial[0.041]; industrial[0.041]; line[0.041]; systems[0.041]; government[0.041]; distribution[0.041]; Paula[0.040]; management[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pirelli\u001b[39m ==> ENTITY: \u001b[32mPirelli\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPirelli[0.050]; Pirelli[0.050]; industrial[0.044]; Shanghai[0.043]; Italian[0.043]; David[0.042]; announced[0.042]; Milan[0.042]; company[0.042]; technology[0.042]; factory[0.041]; telecommunications[0.041]; giant[0.041]; partnership[0.041]; partners[0.040]; million[0.040]; long[0.040]; growing[0.040]; include[0.039]; group[0.039]; joint[0.039]; called[0.039]; existing[0.039]; existing[0.039]; \t\n\r [=================>............................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s536ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1973/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1119testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Stirling ==> ENTITY: \u001b[32mStirling\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.243:-0.243[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.048]; Scotland[0.048]; Scottish[0.046]; Scottish[0.046]; Scottish[0.046]; Scots[0.045]; election[0.041]; British[0.041]; narrowly[0.041]; earlier[0.040]; Labour[0.040]; Labour[0.040]; Labour[0.040]; won[0.040]; won[0.040]; parliament[0.040]; parliament[0.040]; parliament[0.040]; plans[0.040]; held[0.040]; referendum[0.039]; referendum[0.039]; referendum[0.039]; general[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBlair\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPrime Minister of the United Kingdom\u001b[39m <---> \u001b[32mTony Blair\u001b[39m\t\nSCORES: global= 0.266:0.264[\u001b[31m0.001\u001b[39m]; local()= 0.144:0.144[\u001b[31m0.000\u001b[39m]; log p(e|m)= -0.693:0.000[\u001b[32m0.693\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.045]; Blair[0.045]; Scotland[0.042]; Scotland[0.042]; John[0.042]; party[0.042]; party[0.042]; Labour[0.042]; Labour[0.042]; Labour[0.042]; narrowly[0.042]; Tony[0.041]; Party[0.041]; Party[0.041]; Scottish[0.041]; Scottish[0.041]; Scottish[0.041]; Scottish[0.041]; Prime[0.041]; election[0.041]; election[0.041]; won[0.040]; won[0.040]; asking[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Labour ==> ENTITY: \u001b[32mLabour Party (UK)\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.176:0.176[0]; log p(e|m)= -0.341:-0.341[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConservative[0.046]; Scotland[0.045]; Scottish[0.044]; Scottish[0.044]; Scottish[0.044]; Conservatives[0.044]; Party[0.043]; Party[0.043]; John[0.042]; Minister[0.041]; election[0.041]; election[0.041]; National[0.041]; parliament[0.040]; parliament[0.040]; parliament[0.040]; general[0.040]; Scots[0.040]; party[0.040]; referendum[0.040]; seats[0.040]; support[0.040]; said[0.039]; consistently[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Conservative Party ==> ENTITY: \u001b[32mConservative Party (UK)\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.176:0.176[0]; log p(e|m)= -0.403:-0.403[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConservatives[0.045]; Scottish[0.044]; Scottish[0.044]; Scottish[0.044]; Scotland[0.043]; Labour[0.043]; Labour[0.043]; party[0.043]; Party[0.043]; John[0.041]; won[0.041]; election[0.041]; election[0.041]; election[0.041]; run[0.041]; seats[0.041]; Minister[0.041]; support[0.041]; National[0.040]; Scots[0.040]; parliament[0.040]; parliament[0.040]; parliament[0.040]; seeking[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scottish ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.160:0.160[0]; log p(e|m)= -0.480:-0.480[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.050]; Scotland[0.050]; British[0.046]; Scottish[0.046]; Scottish[0.046]; Scottish[0.046]; Scots[0.046]; John[0.042]; Stirling[0.042]; parliaments[0.040]; general[0.040]; general[0.040]; English[0.039]; separate[0.039]; separate[0.039]; separate[0.039]; powers[0.039]; Labour[0.039]; Labour[0.039]; Labour[0.039]; parliament[0.039]; parliament[0.039]; parliament[0.039]; parliament[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scottish ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.480:-0.480[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.053]; Scotland[0.053]; British[0.049]; Scottish[0.049]; Scottish[0.049]; Scots[0.049]; Stirling[0.044]; general[0.042]; general[0.042]; separate[0.041]; separate[0.041]; separate[0.041]; powers[0.041]; Labour[0.041]; Labour[0.041]; Labour[0.041]; parliament[0.041]; parliament[0.041]; parliament[0.041]; parliament[0.041]; Tony[0.041]; set[0.040]; narrowly[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scottish ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.124:0.124[0]; log p(e|m)= -0.480:-0.480[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.052]; Scottish[0.047]; Scottish[0.047]; Scots[0.047]; John[0.043]; Conservative[0.042]; Conservatives[0.042]; English[0.041]; parliaments[0.041]; independence[0.041]; general[0.041]; says[0.040]; separate[0.040]; separate[0.040]; powers[0.040]; Labour[0.040]; parliament[0.040]; parliament[0.040]; parliament[0.040]; seats[0.040]; National[0.039]; election[0.039]; election[0.039]; old[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Labour ==> ENTITY: \u001b[32mLabour Party (UK)\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.182:0.182[0]; log p(e|m)= -0.341:-0.341[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConservative[0.045]; Scotland[0.044]; Scotland[0.044]; Scottish[0.043]; Scottish[0.043]; Scottish[0.043]; Scottish[0.043]; Labour[0.043]; Labour[0.043]; leader[0.042]; British[0.042]; Tony[0.041]; John[0.041]; Stirling[0.041]; party[0.040]; party[0.040]; Blair[0.040]; Blair[0.040]; Party[0.040]; Party[0.040]; Minister[0.040]; election[0.040]; election[0.040]; narrowly[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Scottish Labour Party\u001b[39m ==> ENTITY: \u001b[32mScottish Labour Party\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.045]; Scotland[0.045]; Scottish[0.045]; Scottish[0.045]; leader[0.043]; Labour[0.042]; Labour[0.042]; party[0.042]; Stirling[0.042]; Party[0.041]; Blair[0.041]; Blair[0.041]; Scots[0.041]; election[0.041]; Tony[0.041]; executive[0.040]; narrowly[0.040]; British[0.040]; referendum[0.040]; referendum[0.040]; referendum[0.040]; general[0.040]; won[0.040]; won[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = British Labour Party\u001b[39m ==> ENTITY: \u001b[32mLabour Party (UK)\u001b[39m\t\nSCORES: global= 0.311:0.311[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLabour[0.045]; Labour[0.045]; Scotland[0.043]; Scotland[0.043]; leader[0.042]; Scottish[0.042]; Scottish[0.042]; Scottish[0.042]; narrowly[0.042]; party[0.042]; party[0.042]; election[0.042]; Tony[0.042]; Blair[0.042]; Blair[0.042]; Party[0.041]; won[0.041]; won[0.041]; general[0.040]; general[0.040]; parliament[0.040]; parliament[0.040]; parliament[0.040]; Stirling[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scotland ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.229:-0.229[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScottish[0.046]; Scottish[0.046]; Scottish[0.046]; Scots[0.045]; Conservative[0.044]; Conservatives[0.044]; win[0.042]; Labour[0.041]; National[0.041]; referendum[0.041]; referendum[0.041]; party[0.041]; seats[0.041]; election[0.040]; election[0.040]; parliament[0.040]; parliament[0.040]; parliament[0.040]; Party[0.040]; Party[0.040]; members[0.040]; parliaments[0.040]; John[0.040]; Minister[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = John Major\u001b[39m ==> ENTITY: \u001b[32mJohn Major\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nConservative[0.047]; Conservatives[0.045]; Labour[0.043]; Labour[0.043]; Scottish[0.042]; Scottish[0.042]; Scottish[0.042]; Scottish[0.042]; Scotland[0.041]; Scotland[0.041]; general[0.041]; general[0.041]; Blair[0.041]; Scots[0.041]; parliament[0.041]; parliament[0.041]; parliament[0.041]; parliament[0.041]; parliament[0.041]; election[0.041]; election[0.041]; election[0.041]; won[0.040]; Prime[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scotland ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.142:0.142[0]; log p(e|m)= -0.229:-0.229[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.050]; Scottish[0.046]; Scottish[0.046]; Scottish[0.046]; Scots[0.044]; Stirling[0.043]; British[0.042]; win[0.041]; Labour[0.041]; Labour[0.041]; Labour[0.041]; won[0.040]; won[0.040]; Saturday[0.040]; Tony[0.040]; victory[0.040]; referendum[0.040]; referendum[0.040]; referendum[0.040]; referendum[0.040]; party[0.040]; party[0.040]; voted[0.040]; held[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Scottish National Party\u001b[39m ==> ENTITY: \u001b[32mScottish National Party\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.049]; Scottish[0.048]; Scottish[0.048]; Conservative[0.047]; Conservatives[0.046]; Labour[0.044]; party[0.044]; John[0.044]; election[0.044]; election[0.044]; win[0.043]; support[0.043]; polls[0.042]; Party[0.042]; seats[0.042]; consistently[0.041]; Minister[0.041]; seeking[0.041]; members[0.041]; referendum[0.041]; wanted[0.041]; run[0.041]; general[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scotland ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.149:0.149[0]; log p(e|m)= -0.229:-0.229[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.054]; Scottish[0.050]; Scottish[0.050]; Scottish[0.050]; Scots[0.048]; Stirling[0.047]; British[0.046]; Labour[0.044]; Labour[0.044]; Labour[0.044]; won[0.044]; won[0.044]; Saturday[0.044]; Tony[0.044]; victory[0.044]; referendum[0.044]; referendum[0.044]; referendum[0.044]; party[0.044]; voted[0.043]; held[0.043]; election[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scottish ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.480:-0.480[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.051]; Scottish[0.047]; Scottish[0.047]; Scottish[0.047]; Scots[0.047]; John[0.043]; Conservative[0.042]; Conservatives[0.041]; English[0.041]; parliaments[0.041]; independence[0.041]; general[0.040]; general[0.040]; says[0.040]; separate[0.040]; separate[0.040]; powers[0.040]; Labour[0.039]; Labour[0.039]; parliament[0.039]; parliament[0.039]; parliament[0.039]; parliament[0.039]; seats[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mTony Blair\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPrime Minister of the United Kingdom\u001b[39m <---> \u001b[32mTony Blair\u001b[39m\t\nSCORES: global= 0.266:0.264[\u001b[31m0.001\u001b[39m]; local()= 0.148:0.148[\u001b[31m0.000\u001b[39m]; log p(e|m)= -0.693:0.000[\u001b[32m0.693\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.046]; Blair[0.045]; Scotland[0.042]; Scotland[0.042]; party[0.042]; party[0.042]; Labour[0.042]; Labour[0.042]; Labour[0.042]; narrowly[0.042]; Party[0.042]; Party[0.042]; Scottish[0.041]; Scottish[0.041]; Scottish[0.041]; election[0.041]; election[0.041]; won[0.041]; won[0.041]; asking[0.041]; leader[0.040]; parliament[0.040]; parliament[0.040]; parliament[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = English ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.071:0.071[0]; log p(e|m)= -1.355:-1.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.046]; John[0.046]; independence[0.044]; National[0.043]; Scots[0.043]; Scottish[0.042]; Scottish[0.042]; Scottish[0.042]; Scottish[0.042]; popular[0.042]; parliaments[0.042]; powers[0.041]; Conservative[0.041]; seeking[0.041]; parliament[0.041]; parliament[0.041]; parliament[0.041]; parliament[0.041]; says[0.040]; general[0.040]; general[0.040]; set[0.040]; separate[0.040]; separate[0.040]; \t\n\r [==================>...........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s461ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1991/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1141testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hong Kong ==> ENTITY: \u001b[32mHong Kong\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.157:0.157[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTaiwan[0.047]; Chinese[0.046]; mainland[0.046]; China[0.046]; Kong[0.045]; Kong[0.045]; Kong[0.045]; Hong[0.043]; Hong[0.043]; Hong[0.043]; Chen[0.042]; Chen[0.042]; Chun[0.041]; Saturday[0.040]; admitted[0.039]; Court[0.039]; army[0.039]; army[0.039]; collect[0.039]; High[0.039]; Friday[0.039]; said[0.038]; said[0.038]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = High Court ==> ENTITY: \u001b[32mHigh Court (Hong Kong)\u001b[39m\t\nSCORES: global= 0.240:0.240[0]; local()= 0.146:0.146[0]; log p(e|m)= -2.659:-2.659[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncourt[0.047]; hearing[0.045]; guilty[0.042]; Kong[0.042]; Kong[0.042]; Kong[0.042]; Kong[0.042]; likely[0.042]; sentencing[0.041]; admitted[0.041]; admitted[0.041]; ex[0.041]; Chen[0.041]; Chen[0.041]; secretary[0.041]; said[0.041]; said[0.041]; said[0.041]; Hong[0.041]; Hong[0.041]; Hong[0.041]; Hong[0.041]; told[0.041]; Chun[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Taiwan ==> ENTITY: \u001b[32mTaiwan\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.078:-0.078[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.049]; mainland[0.048]; Chinese[0.046]; Chen[0.045]; Chen[0.045]; Nationalist[0.044]; Hong[0.041]; Hong[0.041]; army[0.041]; army[0.041]; Kong[0.041]; Kong[0.041]; Kong[0.041]; Chun[0.041]; said[0.040]; said[0.040]; said[0.040]; Friday[0.040]; regime[0.040]; High[0.039]; ex[0.039]; admitted[0.039]; admitted[0.039]; likely[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hong Kong ==> ENTITY: \u001b[32mHong Kong\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.151:0.151[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTaiwan[0.051]; Chinese[0.050]; mainland[0.049]; China[0.049]; Kong[0.048]; Kong[0.048]; Kong[0.048]; Hong[0.047]; Hong[0.047]; Hong[0.047]; Chen[0.046]; Chen[0.046]; Chun[0.044]; pounds[0.044]; Saturday[0.043]; admitted[0.042]; Court[0.042]; army[0.042]; army[0.042]; collect[0.042]; High[0.042]; Friday[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTaiwan[0.052]; Chinese[0.048]; mainland[0.048]; Chen[0.043]; Chen[0.043]; army[0.042]; army[0.042]; Hong[0.042]; Hong[0.042]; Hong[0.042]; Kong[0.041]; Kong[0.041]; Kong[0.041]; admitted[0.041]; admitted[0.041]; Nationalist[0.040]; ex[0.039]; Chun[0.039]; regime[0.039]; said[0.039]; said[0.039]; said[0.039]; Friday[0.039]; ancestors[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chinese ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.126:0.126[0]; log p(e|m)= -1.419:-1.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.050]; mainland[0.046]; Taiwan[0.045]; ancestors[0.044]; Chen[0.043]; Chen[0.043]; Hong[0.042]; Hong[0.042]; Hong[0.042]; Kong[0.041]; Kong[0.041]; Kong[0.041]; Kong[0.041]; sons[0.041]; Chun[0.040]; said[0.040]; said[0.040]; said[0.040]; generations[0.040]; Nationalist[0.039]; army[0.039]; army[0.039]; ex[0.039]; likely[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hong Kong ==> ENTITY: \u001b[32mHong Kong\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.126:0.126[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChinese[0.047]; Kong[0.046]; Kong[0.046]; Kong[0.046]; Hong[0.044]; Hong[0.044]; Hong[0.044]; Chen[0.043]; Chen[0.043]; Chun[0.042]; Saturday[0.041]; admitted[0.040]; Court[0.040]; army[0.040]; army[0.040]; collect[0.040]; High[0.040]; Friday[0.039]; said[0.039]; said[0.039]; said[0.039]; trafficking[0.039]; sentencing[0.039]; jails[0.039]; \t\n\r [==================>...........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s468ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 1998/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1115testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lithuania ==> ENTITY: \u001b[32mLithuania national football team\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.155:0.155[0]; log p(e|m)= -2.129:-2.129[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLithuania[0.057]; Lithuania[0.057]; soccer[0.056]; Romania[0.054]; Romania[0.054]; Romania[0.054]; Soccer[0.054]; match[0.054]; European[0.051]; Qualifier[0.050]; beat[0.048]; Bucharest[0.048]; Scorers[0.047]; halftime[0.046]; Beat[0.046]; Mihai[0.046]; Friday[0.046]; Cosmin[0.045]; Contra[0.043]; Danius[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mRomania\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRomania\u001b[39m <---> \u001b[32mRomania national football team\u001b[39m\t\nSCORES: global= 0.253:0.246[\u001b[31m0.007\u001b[39m]; local()= 0.131:0.165[\u001b[32m0.034\u001b[39m]; log p(e|m)= -0.212:-3.058[\u001b[31m2.846\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRomania[0.060]; Romania[0.060]; Lithuania[0.056]; Lithuania[0.056]; Lithuania[0.056]; soccer[0.054]; match[0.054]; Bucharest[0.054]; European[0.052]; Soccer[0.052]; halftime[0.051]; Beat[0.051]; beat[0.051]; Mihai[0.050]; Qualifier[0.049]; Cosmin[0.049]; Scorers[0.049]; Contra[0.047]; Friday[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mihai Tararache\u001b[39m ==> ENTITY: \u001b[32mMihai Tararache\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.092:0.092[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.054]; Cosmin[0.054]; Soccer[0.053]; Romania[0.052]; Romania[0.052]; Romania[0.052]; Qualifier[0.052]; match[0.051]; halftime[0.051]; Bucharest[0.050]; Scorers[0.050]; Lithuania[0.048]; Lithuania[0.048]; Lithuania[0.048]; beat[0.048]; European[0.048]; Beat[0.047]; Danius[0.047]; Friday[0.047]; Contra[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cosmin Contra\u001b[39m ==> ENTITY: \u001b[32mCosmin Contra\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.113:0.113[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.061]; soccer[0.058]; match[0.057]; Romania[0.056]; Romania[0.056]; Romania[0.056]; Qualifier[0.055]; European[0.054]; Friday[0.053]; halftime[0.052]; Bucharest[0.052]; Scorers[0.052]; Lithuania[0.049]; Lithuania[0.049]; Lithuania[0.049]; Mihai[0.049]; Danius[0.048]; beat[0.047]; Beat[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lithuania ==> ENTITY: \u001b[32mLithuania national football team\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.155:0.155[0]; log p(e|m)= -2.129:-2.129[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLithuania[0.057]; Lithuania[0.057]; soccer[0.056]; Romania[0.054]; Romania[0.054]; Romania[0.054]; Soccer[0.054]; match[0.054]; European[0.051]; Qualifier[0.050]; beat[0.048]; Bucharest[0.048]; Scorers[0.047]; halftime[0.046]; Beat[0.046]; Mihai[0.046]; Friday[0.046]; Cosmin[0.045]; Contra[0.043]; Danius[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bucharest ==> ENTITY: \u001b[32mBucharest\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.018:-0.018[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nQualifier[0.057]; Romania[0.057]; Romania[0.057]; Romania[0.057]; match[0.057]; soccer[0.056]; Soccer[0.055]; Cosmin[0.054]; European[0.053]; Mihai[0.053]; Friday[0.052]; Lithuania[0.051]; Lithuania[0.051]; Lithuania[0.051]; beat[0.050]; Beat[0.048]; halftime[0.047]; Scorers[0.047]; Contra[0.046]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mRomania\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRomania\u001b[39m <---> \u001b[32mRomania national football team\u001b[39m\t\nSCORES: global= 0.253:0.246[\u001b[31m0.007\u001b[39m]; local()= 0.131:0.165[\u001b[32m0.034\u001b[39m]; log p(e|m)= -0.212:-3.058[\u001b[31m2.846\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRomania[0.060]; Romania[0.060]; Lithuania[0.056]; Lithuania[0.056]; Lithuania[0.056]; soccer[0.054]; match[0.054]; Bucharest[0.054]; European[0.052]; Soccer[0.052]; halftime[0.051]; Beat[0.051]; beat[0.051]; Mihai[0.050]; Qualifier[0.049]; Cosmin[0.049]; Scorers[0.049]; Contra[0.047]; Friday[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lithuania ==> ENTITY: \u001b[32mLithuania national football team\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.155:0.155[0]; log p(e|m)= -2.129:-2.129[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLithuania[0.057]; Lithuania[0.057]; soccer[0.056]; Romania[0.054]; Romania[0.054]; Romania[0.054]; Soccer[0.054]; match[0.054]; European[0.051]; Qualifier[0.050]; beat[0.048]; Bucharest[0.048]; Scorers[0.047]; halftime[0.046]; Beat[0.046]; Mihai[0.046]; Friday[0.046]; Cosmin[0.045]; Contra[0.043]; Danius[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mRomania\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRomania\u001b[39m <---> \u001b[32mRomania national football team\u001b[39m\t\nSCORES: global= 0.253:0.246[\u001b[31m0.007\u001b[39m]; local()= 0.131:0.165[\u001b[32m0.034\u001b[39m]; log p(e|m)= -0.212:-3.058[\u001b[31m2.846\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRomania[0.060]; Romania[0.060]; Lithuania[0.056]; Lithuania[0.056]; Lithuania[0.056]; soccer[0.054]; match[0.054]; Bucharest[0.054]; European[0.052]; Soccer[0.052]; halftime[0.051]; Beat[0.051]; beat[0.051]; Mihai[0.050]; Qualifier[0.049]; Cosmin[0.049]; Scorers[0.049]; Contra[0.047]; Friday[0.046]; \t\n\r [==================>...........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s452ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2007/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1069testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bury ==> ENTITY: \u001b[32mBury F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.210:0.210[0]; log p(e|m)= -0.722:-0.722[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.043]; Bradford[0.043]; Barnet[0.043]; Rochdale[0.043]; League[0.042]; Colchester[0.042]; Doncaster[0.042]; Peterborough[0.042]; Walsall[0.042]; Stockport[0.042]; Wolverhampton[0.042]; Charlton[0.041]; Swindon[0.041]; Grimsby[0.041]; Barnsley[0.041]; Burnley[0.041]; Watford[0.041]; Wycombe[0.041]; Ipswich[0.041]; Scunthorpe[0.041]; Hartlepool[0.041]; Carlisle[0.041]; Chesterfield[0.041]; Rotherham[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bristol City\u001b[39m ==> ENTITY: \u001b[32mBristol City F.C.\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.218:0.218[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBristol[0.044]; Hull[0.043]; League[0.043]; Manchester[0.043]; Watford[0.042]; Cardiff[0.042]; matches[0.042]; Charlton[0.042]; City[0.042]; Walsall[0.042]; Norwich[0.042]; Portsmouth[0.042]; Darlington[0.041]; Reading[0.041]; Gillingham[0.041]; Chesterfield[0.041]; Stockport[0.041]; Ipswich[0.041]; Millwall[0.041]; Colchester[0.041]; Southend[0.041]; Shrewsbury[0.041]; Barnsley[0.041]; Soccer[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Port Vale\u001b[39m ==> ENTITY: \u001b[32mPort Vale F.C.\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.043]; Chesterfield[0.043]; Walsall[0.042]; Bury[0.042]; Bristol[0.042]; Bristol[0.042]; Southend[0.042]; Charlton[0.042]; Peterborough[0.042]; Stockport[0.042]; Watford[0.042]; Barnsley[0.042]; Portsmouth[0.041]; City[0.041]; City[0.041]; Norwich[0.041]; Ipswich[0.041]; Scunthorpe[0.041]; Rotherham[0.041]; Swindon[0.041]; Oldham[0.041]; Wycombe[0.041]; Gillingham[0.041]; Burnley[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wrexham ==> ENTITY: \u001b[32mWrexham F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.202:0.202[0]; log p(e|m)= -0.882:-0.882[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCardiff[0.045]; Hull[0.044]; Stockport[0.042]; Walsall[0.042]; Rotherham[0.042]; Chesterfield[0.042]; Bristol[0.042]; Bristol[0.042]; Shrewsbury[0.042]; Exeter[0.042]; Peterborough[0.041]; Watford[0.041]; Park[0.041]; Charlton[0.041]; Chester[0.041]; Bournemouth[0.041]; Northampton[0.041]; Birmingham[0.041]; Gillingham[0.041]; Bury[0.041]; Oldham[0.041]; Scunthorpe[0.041]; Luton[0.041]; Rochdale[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brighton ==> ENTITY: \u001b[32mBrighton & Hove Albion F.C.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.206:0.206[0]; log p(e|m)= -3.124:-3.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYork[0.048]; County[0.044]; Bristol[0.043]; Bristol[0.043]; Northampton[0.042]; Charlton[0.042]; Watford[0.042]; Hull[0.042]; Cardiff[0.042]; Chesterfield[0.041]; Carlisle[0.041]; Exeter[0.041]; Walsall[0.041]; Millwall[0.041]; Birmingham[0.041]; Bournemouth[0.041]; Manchester[0.041]; Southend[0.041]; Stockport[0.041]; Blackpool[0.041]; Burnley[0.041]; Shrewsbury[0.041]; Barnsley[0.041]; Luton[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Grimsby ==> ENTITY: \u001b[32mGrimsby Town F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.228:0.228[0]; log p(e|m)= -2.604:-2.604[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBristol[0.043]; Bristol[0.043]; Ipswich[0.042]; Southend[0.042]; League[0.042]; Rotherham[0.042]; Swindon[0.042]; Norwich[0.042]; Burnley[0.042]; Portsmouth[0.042]; Peterborough[0.042]; Bournemouth[0.042]; Gillingham[0.042]; Charlton[0.042]; Chesterfield[0.042]; Huddersfield[0.041]; Shrewsbury[0.041]; Stockport[0.041]; Watford[0.041]; Barnsley[0.041]; Bury[0.041]; Oldham[0.041]; Manchester[0.041]; Bradford[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swindon ==> ENTITY: \u001b[32mSwindon Town F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.236:0.236[0]; log p(e|m)= -2.489:-2.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; Bristol[0.043]; Bristol[0.043]; Charlton[0.043]; Stockport[0.042]; Millwall[0.042]; Walsall[0.042]; Colchester[0.042]; Darlington[0.042]; Reading[0.042]; Burnley[0.041]; Watford[0.041]; Southend[0.041]; Scunthorpe[0.041]; Chesterfield[0.041]; Gillingham[0.041]; matches[0.041]; Peterborough[0.041]; Portsmouth[0.041]; Oldham[0.041]; Barnsley[0.041]; Wolverhampton[0.041]; Rotherham[0.041]; Ipswich[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hartlepool ==> ENTITY: \u001b[32mHartlepool United F.C.\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.194:0.194[0]; log p(e|m)= -1.754:-1.754[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.044]; Stockport[0.043]; Cardiff[0.043]; Charlton[0.042]; Bristol[0.042]; Bristol[0.042]; Park[0.042]; Carlisle[0.042]; Hereford[0.042]; Walsall[0.042]; Rotherham[0.042]; Bury[0.042]; City[0.041]; City[0.041]; Peterborough[0.041]; Rochdale[0.041]; York[0.041]; Wycombe[0.041]; Chesterfield[0.041]; Bournemouth[0.041]; Burnley[0.041]; Darlington[0.041]; Brighton[0.041]; Scunthorpe[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mansfield ==> ENTITY: \u001b[32mMansfield Town F.C.\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.242:0.242[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.044]; Stockport[0.043]; Carlisle[0.043]; Walsall[0.042]; Shrewsbury[0.042]; Bury[0.042]; Colchester[0.042]; Watford[0.042]; Charlton[0.042]; Peterborough[0.042]; Chesterfield[0.042]; Rochdale[0.041]; Northampton[0.041]; Bristol[0.041]; Bristol[0.041]; Wycombe[0.041]; Hereford[0.041]; Cardiff[0.041]; Bournemouth[0.041]; Burnley[0.041]; Rotherham[0.041]; Exeter[0.041]; Scunthorpe[0.041]; Gillingham[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Blackpool ==> ENTITY: \u001b[32mBlackpool F.C.\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.205:0.205[0]; log p(e|m)= -0.853:-0.853[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; Hull[0.043]; Southend[0.043]; Bury[0.043]; Wycombe[0.042]; Rotherham[0.042]; Rochdale[0.042]; Rovers[0.042]; Millwall[0.042]; Walsall[0.042]; Doncaster[0.042]; Watford[0.042]; Huddersfield[0.042]; Wrexham[0.041]; Swindon[0.041]; Stockport[0.041]; Colchester[0.041]; Charlton[0.041]; Scunthorpe[0.041]; Gillingham[0.041]; Barnsley[0.041]; Peterborough[0.041]; Bradford[0.041]; Cardiff[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wolverhampton ==> ENTITY: \u001b[32mWolverhampton Wanderers F.C.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.191:0.191[0]; log p(e|m)= -2.273:-2.273[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; Charlton[0.043]; matches[0.043]; Watford[0.043]; Birmingham[0.042]; Bristol[0.042]; Bristol[0.042]; Bradford[0.042]; Oxford[0.042]; Manchester[0.042]; Burnley[0.042]; Walsall[0.042]; Bolton[0.042]; Oldham[0.041]; Millwall[0.041]; Huddersfield[0.041]; Reading[0.041]; Barnsley[0.041]; Stockport[0.041]; Swindon[0.041]; Blackpool[0.041]; City[0.041]; City[0.041]; Bury[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Crewe ==> ENTITY: \u001b[32mCrewe Alexandra F.C.\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.205:0.205[0]; log p(e|m)= -2.830:-2.830[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWolverhampton[0.043]; Hull[0.043]; League[0.043]; Watford[0.043]; Bristol[0.043]; Bristol[0.043]; Norwich[0.042]; Charlton[0.042]; Shrewsbury[0.042]; Portsmouth[0.042]; Chesterfield[0.042]; Walsall[0.042]; Birmingham[0.042]; Carlisle[0.041]; Manchester[0.041]; Colchester[0.041]; Ipswich[0.041]; Burnley[0.041]; Swindon[0.041]; matches[0.041]; Barnsley[0.041]; Stockport[0.041]; Wycombe[0.041]; Peterborough[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Watford ==> ENTITY: \u001b[32mWatford F.C.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.202:0.202[0]; log p(e|m)= -0.960:-0.960[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.043]; Hull[0.043]; matches[0.043]; Birmingham[0.043]; Cardiff[0.042]; Burnley[0.042]; Portsmouth[0.042]; Walsall[0.042]; Norwich[0.042]; Millwall[0.042]; Manchester[0.042]; Bristol[0.042]; Bristol[0.042]; Stockport[0.041]; Reading[0.041]; Colchester[0.041]; Shrewsbury[0.041]; Bury[0.041]; Chesterfield[0.041]; Swindon[0.041]; Luton[0.041]; Carlisle[0.041]; Blackpool[0.041]; Scunthorpe[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scunthorpe ==> ENTITY: \u001b[32mScunthorpe United F.C.\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.228:0.228[0]; log p(e|m)= -1.510:-1.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.045]; Charlton[0.043]; Stockport[0.043]; Birmingham[0.042]; Cardiff[0.042]; Walsall[0.042]; Burnley[0.042]; Rotherham[0.042]; Watford[0.042]; Bristol[0.042]; Bristol[0.042]; Rochdale[0.041]; Colchester[0.041]; Swindon[0.041]; Barnsley[0.041]; Chesterfield[0.041]; Oldham[0.041]; Reading[0.041]; Carlisle[0.041]; Millwall[0.041]; Peterborough[0.041]; Hartlepool[0.041]; Darlington[0.041]; Bury[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cambridge United\u001b[39m ==> ENTITY: \u001b[32mCambridge United F.C.\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.249:0.249[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.045]; Charlton[0.043]; Manchester[0.043]; Colchester[0.042]; Watford[0.042]; Bristol[0.042]; Bristol[0.042]; Peterborough[0.042]; Cardiff[0.042]; Burnley[0.042]; Birmingham[0.041]; Northampton[0.041]; Division[0.041]; Division[0.041]; Carlisle[0.041]; Scunthorpe[0.041]; Walsall[0.041]; Millwall[0.041]; Southend[0.041]; Luton[0.041]; Gillingham[0.041]; Darlington[0.041]; Stockport[0.041]; Rochdale[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Birmingham ==> ENTITY: \u001b[32mBirmingham City F.C.\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.176:0.176[0]; log p(e|m)= -3.101:-3.101[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.043]; Wolverhampton[0.043]; Hull[0.043]; League[0.043]; Bristol[0.043]; Bristol[0.043]; Watford[0.043]; Luton[0.042]; Manchester[0.042]; matches[0.042]; Burnley[0.041]; Carlisle[0.041]; Walsall[0.041]; Stockport[0.041]; City[0.041]; City[0.041]; London[0.041]; Shrewsbury[0.041]; Oxford[0.041]; Peterborough[0.041]; Millwall[0.041]; Norwich[0.041]; Blackpool[0.041]; Southend[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rochdale ==> ENTITY: \u001b[32mRochdale A.F.C.\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.223:0.223[0]; log p(e|m)= -0.952:-0.952[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChesterfield[0.043]; Hull[0.043]; Stockport[0.043]; Walsall[0.042]; Rotherham[0.042]; Carlisle[0.042]; Peterborough[0.042]; Bury[0.042]; Colchester[0.042]; Shrewsbury[0.042]; Scunthorpe[0.042]; Darlington[0.042]; Rovers[0.042]; Bristol[0.042]; Bristol[0.042]; Queens[0.041]; Burnley[0.041]; Bournemouth[0.041]; Millwall[0.041]; Watford[0.041]; Gillingham[0.041]; Chester[0.041]; Wycombe[0.041]; Luton[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scarborough ==> ENTITY: \u001b[32mScarborough F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.217:0.217[0]; log p(e|m)= -1.814:-1.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPeterborough[0.043]; Bournemouth[0.043]; Rochdale[0.043]; Walsall[0.043]; Bristol[0.043]; Bristol[0.043]; Rotherham[0.043]; Carlisle[0.042]; Stockport[0.042]; Hull[0.042]; Darlington[0.041]; Hartlepool[0.041]; Cardiff[0.041]; Watford[0.041]; Colchester[0.041]; Doncaster[0.041]; Wrexham[0.041]; Luton[0.041]; Blackpool[0.041]; Exeter[0.041]; Bury[0.041]; Burnley[0.041]; Wycombe[0.041]; Torquay[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Huddersfield ==> ENTITY: \u001b[32mHuddersfield Town F.C.\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.210:0.210[0]; log p(e|m)= -2.283:-2.283[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.044]; League[0.043]; Bradford[0.043]; Wolverhampton[0.043]; Charlton[0.042]; Rotherham[0.042]; Bristol[0.042]; Bristol[0.042]; Chesterfield[0.042]; Manchester[0.042]; Stockport[0.042]; Walsall[0.041]; Birmingham[0.041]; Burnley[0.041]; Park[0.041]; Watford[0.041]; Bury[0.041]; Millwall[0.041]; Oldham[0.041]; Barnsley[0.041]; Luton[0.041]; Portsmouth[0.041]; City[0.041]; City[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wycombe ==> ENTITY: \u001b[32mWycombe Wanderers F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.231:0.231[0]; log p(e|m)= -1.146:-1.146[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.045]; matches[0.043]; Hull[0.042]; Reading[0.042]; Watford[0.042]; Charlton[0.042]; Walsall[0.042]; Stockport[0.042]; Carlisle[0.042]; Shrewsbury[0.042]; Chesterfield[0.042]; Swindon[0.041]; Millwall[0.041]; Colchester[0.041]; Oxford[0.041]; Barnsley[0.041]; Norwich[0.041]; Rochdale[0.041]; Oldham[0.041]; Bristol[0.041]; Bristol[0.041]; Cardiff[0.041]; Northampton[0.041]; Bury[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gillingham ==> ENTITY: \u001b[32mGillingham F.C.\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.226:0.226[0]; log p(e|m)= -0.257:-0.257[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.043]; Watford[0.043]; Charlton[0.043]; Colchester[0.042]; Southend[0.042]; Walsall[0.042]; Stockport[0.042]; Norwich[0.042]; matches[0.042]; Peterborough[0.042]; City[0.042]; City[0.042]; Millwall[0.042]; Chesterfield[0.042]; Tranmere[0.042]; Rochdale[0.042]; Birmingham[0.041]; Scunthorpe[0.041]; Oldham[0.041]; Burnley[0.041]; Portsmouth[0.041]; Swindon[0.041]; Rotherham[0.041]; Bury[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chesterfield ==> ENTITY: \u001b[32mChesterfield F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.200:0.200[0]; log p(e|m)= -1.269:-1.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.043]; Birmingham[0.043]; Carlisle[0.043]; Stockport[0.042]; Oxford[0.042]; Bradford[0.042]; Bristol[0.042]; Bristol[0.042]; Wolverhampton[0.042]; Northampton[0.042]; Colchester[0.042]; Walsall[0.042]; Oldham[0.042]; Rotherham[0.041]; Shrewsbury[0.041]; matches[0.041]; Norwich[0.041]; York[0.041]; Rochdale[0.041]; Portsmouth[0.041]; Watford[0.041]; Burnley[0.041]; Manchester[0.041]; Bury[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oxford ==> ENTITY: \u001b[32mOxford United F.C.\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.212:0.212[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCambridge[0.044]; League[0.044]; Charlton[0.042]; Wolverhampton[0.042]; Peterborough[0.042]; Shrewsbury[0.042]; Birmingham[0.042]; Watford[0.042]; Walsall[0.042]; Stockport[0.042]; Millwall[0.042]; City[0.042]; City[0.042]; Gillingham[0.041]; Burnley[0.041]; Bury[0.041]; Reading[0.041]; Bristol[0.041]; Bristol[0.041]; Cardiff[0.041]; Bournemouth[0.041]; matches[0.041]; Norwich[0.041]; Swindon[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cardiff ==> ENTITY: \u001b[32mCardiff City F.C.\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.215:0.215[0]; log p(e|m)= -3.474:-3.474[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.044]; Bristol[0.043]; Bristol[0.043]; Carlisle[0.042]; Charlton[0.042]; Watford[0.042]; Rotherham[0.042]; Southend[0.042]; Millwall[0.042]; Exeter[0.042]; Walsall[0.041]; Shrewsbury[0.041]; Stockport[0.041]; Burnley[0.041]; Park[0.041]; Chesterfield[0.041]; Birmingham[0.041]; Bournemouth[0.041]; Northampton[0.041]; Colchester[0.041]; Wrexham[0.041]; Rochdale[0.041]; Swindon[0.041]; Bury[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colchester ==> ENTITY: \u001b[32mColchester United F.C.\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.202:0.202[0]; log p(e|m)= -2.865:-2.865[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.044]; Charlton[0.044]; Shrewsbury[0.043]; Watford[0.042]; Stockport[0.042]; Manchester[0.042]; Gillingham[0.042]; Burnley[0.042]; Walsall[0.042]; Peterborough[0.042]; Exeter[0.042]; Bristol[0.042]; Bristol[0.042]; Chesterfield[0.041]; Birmingham[0.041]; Oxford[0.041]; Carlisle[0.041]; Reading[0.041]; Barnsley[0.041]; Southend[0.041]; Millwall[0.041]; Cambridge[0.041]; Northampton[0.041]; Swindon[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barnsley ==> ENTITY: \u001b[32mBarnsley F.C.\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.205:0.205[0]; log p(e|m)= -0.713:-0.713[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWrexham[0.045]; Walsall[0.043]; Shrewsbury[0.043]; Swindon[0.043]; Wycombe[0.042]; Stoke[0.042]; Burnley[0.042]; Luton[0.042]; Colchester[0.042]; Scunthorpe[0.042]; Ipswich[0.042]; Rotherham[0.042]; Bury[0.042]; Huddersfield[0.042]; Southend[0.041]; Charlton[0.041]; Stockport[0.041]; Wolverhampton[0.041]; Cardiff[0.041]; Watford[0.041]; Blackpool[0.041]; Hull[0.041]; Millwall[0.040]; Crewe[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Northampton ==> ENTITY: \u001b[32mNorthampton Town F.C.\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.230:0.230[0]; log p(e|m)= -3.244:-3.244[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.044]; Cardiff[0.043]; Stockport[0.043]; Colchester[0.042]; Wycombe[0.042]; Walsall[0.042]; Watford[0.042]; Millwall[0.042]; Gillingham[0.042]; Shrewsbury[0.042]; Rochdale[0.042]; Chesterfield[0.042]; Scunthorpe[0.041]; Burnley[0.041]; Peterborough[0.041]; York[0.041]; Carlisle[0.041]; Bury[0.041]; Rotherham[0.041]; Bristol[0.041]; Bristol[0.041]; Bournemouth[0.041]; Luton[0.041]; Wrexham[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Exeter ==> ENTITY: \u001b[32mExeter City F.C.\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.231:0.231[0]; log p(e|m)= -3.650:-3.650[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.043]; Bristol[0.043]; Bristol[0.043]; Watford[0.042]; Carlisle[0.042]; Cardiff[0.042]; Stockport[0.042]; Shrewsbury[0.042]; Colchester[0.042]; Walsall[0.042]; Gillingham[0.042]; Bury[0.042]; Northampton[0.042]; Rochdale[0.041]; Burnley[0.041]; Chester[0.041]; Peterborough[0.041]; Millwall[0.041]; Chesterfield[0.041]; Bournemouth[0.041]; Rotherham[0.041]; Scunthorpe[0.041]; Cambridge[0.041]; Hereford[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bournemouth ==> ENTITY: \u001b[32mA.F.C. Bournemouth\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.195:0.195[0]; log p(e|m)= -1.058:-1.058[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTorquay[0.044]; Southend[0.043]; Blackpool[0.043]; Swindon[0.043]; Stockport[0.042]; Scunthorpe[0.042]; Doncaster[0.042]; Walsall[0.042]; Shrewsbury[0.042]; Wycombe[0.042]; Wrexham[0.042]; Notts[0.041]; Gillingham[0.041]; Hartlepool[0.041]; Wolverhampton[0.041]; League[0.041]; Grimsby[0.041]; Huddersfield[0.041]; Burnley[0.041]; Rovers[0.041]; Hull[0.041]; Rochdale[0.041]; Colchester[0.041]; Rotherham[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oldham ==> ENTITY: \u001b[32mOldham Athletic A.F.C.\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.219:0.219[0]; log p(e|m)= -2.865:-2.865[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.045]; Bradford[0.044]; Huddersfield[0.043]; Watford[0.042]; Charlton[0.042]; Bristol[0.042]; Bristol[0.042]; Walsall[0.042]; Chesterfield[0.042]; Peterborough[0.042]; Stockport[0.042]; Bury[0.042]; Rotherham[0.042]; Millwall[0.041]; matches[0.041]; Burnley[0.041]; Gillingham[0.041]; Southend[0.041]; Barnsley[0.041]; Scunthorpe[0.041]; Wycombe[0.041]; Tranmere[0.040]; Wolverhampton[0.040]; Rovers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Luton ==> ENTITY: \u001b[32mLuton Town F.C.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.220:0.220[0]; log p(e|m)= -2.551:-2.551[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.044]; Charlton[0.043]; Walsall[0.042]; Watford[0.042]; Stockport[0.042]; Chesterfield[0.042]; Gillingham[0.042]; Shrewsbury[0.042]; matches[0.042]; Burnley[0.042]; Scunthorpe[0.042]; Manchester[0.042]; Bristol[0.042]; Bristol[0.042]; Birmingham[0.041]; Rotherham[0.041]; Millwall[0.041]; Norwich[0.041]; Wolverhampton[0.041]; Bournemouth[0.041]; Colchester[0.041]; Darlington[0.041]; Portsmouth[0.041]; Swindon[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rotherham ==> ENTITY: \u001b[32mRotherham United F.C.\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.235:0.235[0]; log p(e|m)= -2.645:-2.645[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.043]; Charlton[0.043]; Watford[0.042]; Stockport[0.042]; Bristol[0.042]; Bristol[0.042]; Walsall[0.042]; Colchester[0.042]; Oldham[0.042]; matches[0.042]; Huddersfield[0.042]; Bradford[0.042]; Burnley[0.042]; Cardiff[0.041]; Gillingham[0.041]; Carlisle[0.041]; Barnsley[0.041]; Northampton[0.041]; Scunthorpe[0.041]; Bury[0.041]; Millwall[0.041]; Darlington[0.041]; Southend[0.041]; Swindon[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tranmere ==> ENTITY: \u001b[32mTranmere Rovers F.C.\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.234:0.234[0]; log p(e|m)= -0.600:-0.600[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.044]; Watford[0.043]; League[0.043]; matches[0.043]; Burnley[0.043]; Stockport[0.042]; Rotherham[0.042]; Portsmouth[0.042]; Gillingham[0.042]; Southend[0.042]; Millwall[0.042]; Barnsley[0.042]; Manchester[0.041]; Birmingham[0.041]; Blackpool[0.041]; Bournemouth[0.041]; Ipswich[0.041]; Oldham[0.041]; Rovers[0.041]; Wycombe[0.041]; Swindon[0.041]; Norwich[0.041]; Chesterfield[0.041]; Bury[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hereford ==> ENTITY: \u001b[32mHereford United F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.221:0.221[0]; log p(e|m)= -2.765:-2.765[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCounty[0.045]; Chesterfield[0.043]; Hull[0.043]; Stockport[0.042]; Walsall[0.042]; Southend[0.042]; Shrewsbury[0.042]; Colchester[0.042]; Manchester[0.042]; Watford[0.042]; Rochdale[0.042]; Rotherham[0.042]; Swindon[0.041]; Charlton[0.041]; Gillingham[0.041]; Peterborough[0.041]; Scunthorpe[0.041]; Northampton[0.041]; Carlisle[0.041]; Millwall[0.041]; Burnley[0.041]; Barnsley[0.041]; Bournemouth[0.041]; Wycombe[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Doncaster ==> ENTITY: \u001b[32mDoncaster Rovers F.C.\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.217:0.217[0]; log p(e|m)= -2.957:-2.957[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.044]; Rotherham[0.043]; Charlton[0.042]; Walsall[0.042]; Rochdale[0.042]; Carlisle[0.042]; Southend[0.042]; Stockport[0.042]; Watford[0.042]; Wrexham[0.042]; Cardiff[0.041]; Birmingham[0.041]; Bristol[0.041]; Bristol[0.041]; Burnley[0.041]; Chesterfield[0.041]; Hereford[0.041]; Scunthorpe[0.041]; Millwall[0.041]; Peterborough[0.041]; Blackpool[0.041]; Rovers[0.041]; Luton[0.041]; Swindon[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.088:0.088[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; Manchester[0.044]; League[0.044]; Charlton[0.043]; Birmingham[0.043]; Bristol[0.043]; Bristol[0.043]; Watford[0.042]; Soccer[0.042]; soccer[0.042]; City[0.041]; City[0.041]; Portsmouth[0.041]; Wolverhampton[0.041]; Oxford[0.041]; Blackpool[0.041]; Norwich[0.041]; Park[0.041]; Stockport[0.041]; Barnsley[0.040]; Southend[0.040]; Stoke[0.040]; Reading[0.040]; Bolton[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Manchester City\u001b[39m ==> ENTITY: \u001b[32mManchester City F.C.\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; Charlton[0.044]; Hull[0.044]; Portsmouth[0.043]; Watford[0.043]; Birmingham[0.042]; matches[0.042]; Division[0.042]; Division[0.042]; Division[0.042]; Burnley[0.041]; Millwall[0.041]; Bolton[0.041]; Cardiff[0.041]; Soccer[0.041]; Colchester[0.041]; Bristol[0.041]; Bristol[0.041]; Norwich[0.041]; City[0.041]; Gillingham[0.041]; Ipswich[0.041]; soccer[0.041]; Southend[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chester ==> ENTITY: \u001b[32mChester City F.C.\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.216:0.216[0]; log p(e|m)= -3.219:-3.219[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.045]; Stockport[0.043]; Watford[0.042]; Walsall[0.042]; Carlisle[0.042]; Bristol[0.042]; Bristol[0.042]; Bury[0.042]; Gillingham[0.042]; Peterborough[0.042]; Burnley[0.042]; Colchester[0.041]; Chesterfield[0.041]; Darlington[0.041]; Rotherham[0.041]; Bournemouth[0.041]; Scunthorpe[0.041]; Rochdale[0.041]; City[0.041]; Northampton[0.041]; Millwall[0.041]; Cardiff[0.041]; Wycombe[0.041]; Shrewsbury[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Crystal Palace ==> ENTITY: \u001b[32mCrystal Palace F.C.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.192:0.192[0]; log p(e|m)= -0.511:-0.511[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; Charlton[0.043]; Watford[0.043]; Birmingham[0.042]; Manchester[0.042]; Bolton[0.042]; Portsmouth[0.042]; Walsall[0.042]; Ipswich[0.042]; matches[0.042]; Millwall[0.042]; soccer[0.041]; Stockport[0.041]; Burnley[0.041]; Park[0.041]; Luton[0.041]; Southend[0.041]; Norwich[0.041]; Bristol[0.041]; Bristol[0.041]; Barnsley[0.041]; Shrewsbury[0.041]; Chesterfield[0.041]; Reading[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Peterborough ==> ENTITY: \u001b[32mPeterborough United F.C.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.221:0.221[0]; log p(e|m)= -3.170:-3.170[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.045]; League[0.043]; Bristol[0.043]; Bristol[0.043]; Charlton[0.042]; Shrewsbury[0.042]; Portsmouth[0.042]; Walsall[0.042]; Ipswich[0.042]; Birmingham[0.042]; Wolverhampton[0.042]; Swindon[0.041]; Manchester[0.041]; Stockport[0.041]; Colchester[0.041]; Rochdale[0.041]; Norwich[0.041]; Carlisle[0.041]; Watford[0.041]; Bury[0.041]; Southend[0.041]; Bradford[0.041]; Rotherham[0.041]; Exeter[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Millwall\u001b[39m ==> ENTITY: \u001b[32mMillwall F.C.\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.228:0.228[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.044]; Watford[0.043]; Charlton[0.043]; Cardiff[0.042]; Birmingham[0.042]; Tranmere[0.042]; Brighton[0.042]; Stockport[0.042]; Portsmouth[0.042]; Manchester[0.042]; Walsall[0.042]; Colchester[0.041]; Shrewsbury[0.041]; Reading[0.041]; Carlisle[0.041]; Chesterfield[0.041]; Burnley[0.041]; Southend[0.041]; matches[0.041]; Gillingham[0.041]; Luton[0.041]; City[0.041]; City[0.041]; Norwich[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burnley ==> ENTITY: \u001b[32mBurnley F.C.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.198:0.198[0]; log p(e|m)= -0.992:-0.992[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChesterfield[0.047]; County[0.046]; Hull[0.045]; Charlton[0.044]; Watford[0.044]; Stockport[0.044]; Walsall[0.044]; Rotherham[0.044]; Colchester[0.044]; Oldham[0.044]; Bradford[0.043]; Gillingham[0.043]; Carlisle[0.043]; Barnsley[0.043]; Scunthorpe[0.043]; Bury[0.043]; Millwall[0.043]; Darlington[0.043]; Southend[0.043]; Swindon[0.043]; Northampton[0.042]; Portsmouth[0.042]; Rochdale[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Reading ==> ENTITY: \u001b[32mReading F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.179:0.179[0]; log p(e|m)= -1.201:-1.201[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.045]; Bristol[0.043]; Bristol[0.043]; Charlton[0.043]; Wolverhampton[0.042]; Shrewsbury[0.042]; Watford[0.042]; Oxford[0.042]; English[0.042]; English[0.042]; Peterborough[0.041]; Bolton[0.041]; Swindon[0.041]; matches[0.041]; Millwall[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Birmingham[0.041]; Manchester[0.041]; Burnley[0.041]; Cardiff[0.041]; Norwich[0.041]; Walsall[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Darlington ==> ENTITY: \u001b[32mDarlington F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.213:0.213[0]; log p(e|m)= -1.871:-1.871[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.043]; Bristol[0.043]; Bristol[0.043]; Chesterfield[0.043]; Charlton[0.042]; Northampton[0.042]; Stockport[0.042]; Exeter[0.042]; Colchester[0.042]; Peterborough[0.042]; Rotherham[0.041]; Barnsley[0.041]; Scunthorpe[0.041]; York[0.041]; Gillingham[0.041]; Carlisle[0.041]; Swindon[0.041]; Watford[0.041]; Walsall[0.041]; Hartlepool[0.041]; County[0.041]; Doncaster[0.041]; Cardiff[0.041]; Burnley[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Fulham ==> ENTITY: \u001b[32mFulham F.C.\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.184:0.184[0]; log p(e|m)= -0.322:-0.322[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.044]; Cardiff[0.044]; Watford[0.043]; Bolton[0.043]; Rotherham[0.043]; Manchester[0.043]; Hull[0.043]; Bristol[0.042]; Bristol[0.042]; Wigan[0.042]; Stockport[0.042]; Birmingham[0.042]; Northampton[0.041]; Colchester[0.041]; Millwall[0.041]; Burnley[0.041]; Southend[0.041]; Peterborough[0.041]; Chesterfield[0.041]; Rochdale[0.041]; Doncaster[0.041]; Blackpool[0.041]; United[0.040]; Luton[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portsmouth ==> ENTITY: \u001b[32mPortsmouth F.C.\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.220:0.220[0]; log p(e|m)= -1.231:-1.231[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.045]; League[0.045]; Watford[0.043]; Bristol[0.042]; Bristol[0.042]; Millwall[0.042]; Burnley[0.042]; Manchester[0.042]; Birmingham[0.042]; Palace[0.042]; Norwich[0.042]; matches[0.041]; Gillingham[0.041]; Ipswich[0.041]; Bolton[0.041]; Southend[0.041]; Barnsley[0.041]; Shrewsbury[0.041]; City[0.041]; City[0.041]; Reading[0.041]; Chesterfield[0.041]; Blackpool[0.040]; Soccer[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Charlton ==> ENTITY: \u001b[32mCharlton Athletic F.C.\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.217:0.217[0]; log p(e|m)= -2.104:-2.104[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCardiff[0.043]; League[0.043]; Watford[0.043]; Hull[0.043]; Colchester[0.042]; Norwich[0.042]; Manchester[0.042]; Portsmouth[0.042]; Bolton[0.042]; Bristol[0.042]; Bristol[0.042]; Shrewsbury[0.042]; Millwall[0.041]; Burnley[0.041]; Walsall[0.041]; Reading[0.041]; Southend[0.041]; matches[0.041]; Gillingham[0.041]; Blackpool[0.041]; City[0.041]; City[0.041]; Stockport[0.041]; Birmingham[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = English ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.133:0.133[0]; log p(e|m)= -1.355:-1.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEnglish[0.046]; League[0.045]; matches[0.043]; Division[0.042]; Division[0.042]; Oxford[0.042]; Peterborough[0.042]; Watford[0.042]; Birmingham[0.042]; Bristol[0.042]; Bristol[0.042]; Wycombe[0.042]; Swindon[0.041]; Huddersfield[0.041]; Barnsley[0.041]; Southend[0.041]; Bournemouth[0.041]; Oldham[0.041]; Bury[0.041]; Chesterfield[0.041]; Norwich[0.041]; Luton[0.041]; Stockport[0.041]; Portsmouth[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bristol Rovers\u001b[39m ==> ENTITY: \u001b[32mBristol Rovers F.C.\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.222:0.222[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBristol[0.049]; Hull[0.044]; Scunthorpe[0.043]; Shrewsbury[0.042]; Portsmouth[0.042]; Manchester[0.042]; Watford[0.042]; Ipswich[0.042]; Southend[0.041]; Exeter[0.041]; Stoke[0.041]; Colchester[0.041]; Tranmere[0.041]; Burnley[0.041]; Darlington[0.041]; Rotherham[0.041]; Northampton[0.041]; Cardiff[0.041]; Charlton[0.041]; Stockport[0.041]; Norwich[0.041]; Blackpool[0.040]; Wigan[0.040]; Barnsley[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ipswich ==> ENTITY: \u001b[32mIpswich Town F.C.\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.214:0.214[0]; log p(e|m)= -2.688:-2.688[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; Charlton[0.044]; Watford[0.043]; Bristol[0.043]; Bristol[0.043]; Manchester[0.042]; Bolton[0.042]; Portsmouth[0.042]; Birmingham[0.042]; Norwich[0.041]; Burnley[0.041]; Scunthorpe[0.041]; Millwall[0.041]; Stockport[0.041]; Shrewsbury[0.041]; Walsall[0.041]; matches[0.041]; Barnsley[0.041]; Reading[0.041]; Southend[0.041]; Crewe[0.041]; Gillingham[0.041]; Chesterfield[0.041]; Wycombe[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Notts County\u001b[39m ==> ENTITY: \u001b[32mNotts County F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.231:0.231[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.043]; Charlton[0.042]; Walsall[0.042]; Chesterfield[0.042]; Division[0.042]; Division[0.042]; Division[0.042]; Burnley[0.042]; Manchester[0.042]; Colchester[0.042]; Watford[0.042]; Bury[0.042]; Gillingham[0.041]; Scunthorpe[0.041]; Millwall[0.041]; Norwich[0.041]; Barnsley[0.041]; Stockport[0.041]; Portsmouth[0.041]; Ipswich[0.041]; Southend[0.041]; Peterborough[0.041]; Rotherham[0.041]; Luton[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Stoke ==> ENTITY: \u001b[32mStoke City F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.202:0.202[0]; log p(e|m)= -0.341:-0.341[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.043]; Charlton[0.043]; Wolverhampton[0.042]; Bristol[0.042]; Bristol[0.042]; Manchester[0.042]; Shrewsbury[0.042]; Portsmouth[0.042]; Norwich[0.042]; Watford[0.042]; City[0.042]; City[0.042]; Birmingham[0.042]; Southend[0.041]; Stockport[0.041]; Ipswich[0.041]; Cardiff[0.041]; Burnley[0.041]; matches[0.041]; Wrexham[0.041]; Blackpool[0.041]; Colchester[0.041]; Peterborough[0.041]; Doncaster[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Carlisle ==> ENTITY: \u001b[32mCarlisle United F.C.\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.194:0.194[0]; log p(e|m)= -3.612:-3.612[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBristol[0.043]; Bristol[0.043]; Hull[0.043]; Rochdale[0.043]; Peterborough[0.042]; Manchester[0.042]; Stockport[0.042]; Cardiff[0.042]; Colchester[0.042]; Walsall[0.042]; York[0.041]; Exeter[0.041]; Charlton[0.041]; Bury[0.041]; Luton[0.041]; Watford[0.041]; City[0.041]; City[0.041]; Bournemouth[0.041]; Park[0.041]; Burnley[0.041]; Swindon[0.041]; Chesterfield[0.041]; Wycombe[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hull ==> ENTITY: \u001b[32mHull City A.F.C.\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.223:0.223[0]; log p(e|m)= -3.772:-3.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWalsall[0.043]; Cardiff[0.043]; Charlton[0.042]; Rochdale[0.042]; Manchester[0.042]; Watford[0.042]; Burnley[0.042]; Bristol[0.042]; Bristol[0.042]; Colchester[0.042]; Millwall[0.042]; Stockport[0.042]; Scunthorpe[0.042]; Carlisle[0.041]; Chesterfield[0.041]; Rotherham[0.041]; Birmingham[0.041]; Peterborough[0.041]; Hereford[0.041]; City[0.041]; City[0.041]; Bury[0.041]; Barnsley[0.041]; Shrewsbury[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Queens Park Rangers\u001b[39m ==> ENTITY: \u001b[32mQueens Park Rangers F.C.\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.047]; Charlton[0.045]; Hull[0.045]; Watford[0.044]; Millwall[0.044]; Walsall[0.044]; matches[0.044]; Burnley[0.044]; Manchester[0.043]; Rochdale[0.043]; Chesterfield[0.043]; Luton[0.043]; Blackpool[0.043]; Southend[0.043]; Cardiff[0.043]; Barnsley[0.043]; Stockport[0.043]; Birmingham[0.043]; Colchester[0.043]; Scunthorpe[0.043]; Gillingham[0.043]; Rotherham[0.043]; Portsmouth[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norwich ==> ENTITY: \u001b[32mNorwich City F.C.\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.193:0.193[0]; log p(e|m)= -2.733:-2.733[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.044]; League[0.043]; Wolverhampton[0.043]; Watford[0.043]; Birmingham[0.042]; matches[0.042]; Portsmouth[0.042]; Manchester[0.042]; Burnley[0.042]; Bristol[0.042]; Bristol[0.042]; Chesterfield[0.041]; Peterborough[0.041]; Millwall[0.041]; Barnsley[0.041]; Walsall[0.041]; Ipswich[0.041]; Bolton[0.041]; Southend[0.041]; Shrewsbury[0.041]; Luton[0.041]; Blackpool[0.041]; Stockport[0.041]; Swindon[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Stockport ==> ENTITY: \u001b[32mStockport County F.C.\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.225:0.225[0]; log p(e|m)= -2.526:-2.526[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.044]; Cardiff[0.043]; York[0.043]; Bradford[0.042]; Rotherham[0.042]; Walsall[0.042]; Bury[0.042]; Northampton[0.042]; Chesterfield[0.042]; Chester[0.042]; Southend[0.042]; Rochdale[0.041]; League[0.041]; Norwich[0.041]; Watford[0.041]; Scunthorpe[0.041]; Swindon[0.041]; Bristol[0.041]; Bristol[0.041]; Bournemouth[0.041]; Huddersfield[0.041]; Wolverhampton[0.041]; Colchester[0.041]; Burnley[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = English ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.132:0.132[0]; log p(e|m)= -1.355:-1.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEnglish[0.046]; League[0.046]; matches[0.043]; Division[0.042]; Division[0.042]; Oxford[0.042]; Peterborough[0.042]; Birmingham[0.042]; Bristol[0.042]; Bristol[0.042]; Wycombe[0.042]; Swindon[0.041]; Huddersfield[0.041]; Barnsley[0.041]; Southend[0.041]; Bournemouth[0.041]; Oldham[0.041]; Bury[0.041]; Norwich[0.041]; Stockport[0.041]; Portsmouth[0.041]; Bradford[0.041]; Reading[0.040]; Ipswich[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bradford ==> ENTITY: \u001b[32mBradford City A.F.C.\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.243:0.243[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.045]; Charlton[0.043]; matches[0.042]; Burnley[0.042]; Bristol[0.042]; Bristol[0.042]; Watford[0.042]; Huddersfield[0.042]; Stockport[0.042]; Manchester[0.042]; Rotherham[0.042]; Millwall[0.042]; Reading[0.041]; Chesterfield[0.041]; City[0.041]; City[0.041]; Barnsley[0.041]; Oldham[0.041]; Swindon[0.041]; Tranmere[0.041]; Ipswich[0.041]; Bury[0.041]; Peterborough[0.041]; Wycombe[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Torquay ==> ENTITY: \u001b[32mTorquay United F.C.\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.206:0.206[0]; log p(e|m)= -2.017:-2.017[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.044]; Carlisle[0.043]; Park[0.043]; Shrewsbury[0.042]; Stockport[0.042]; Walsall[0.042]; Colchester[0.042]; Northampton[0.042]; Bristol[0.042]; Bristol[0.042]; Chesterfield[0.042]; Rochdale[0.042]; Watford[0.041]; Gillingham[0.041]; City[0.041]; Scunthorpe[0.041]; Exeter[0.041]; Cardiff[0.041]; Chester[0.041]; Bury[0.041]; Millwall[0.041]; Rotherham[0.041]; Hereford[0.041]; Peterborough[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Shrewsbury ==> ENTITY: \u001b[32mShrewsbury Town F.C.\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.228:0.228[0]; log p(e|m)= -3.244:-3.244[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.044]; Cambridge[0.043]; Oxford[0.043]; Bristol[0.042]; Bristol[0.042]; Norwich[0.042]; Watford[0.042]; Colchester[0.042]; Manchester[0.042]; Carlisle[0.042]; Chesterfield[0.042]; Portsmouth[0.042]; Charlton[0.042]; Stockport[0.042]; Ipswich[0.041]; Wolverhampton[0.041]; Exeter[0.041]; Walsall[0.041]; City[0.041]; City[0.041]; Northampton[0.041]; Cardiff[0.041]; Rochdale[0.041]; Peterborough[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Walsall ==> ENTITY: \u001b[32mWalsall F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.192:0.192[0]; log p(e|m)= -0.620:-0.620[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWolverhampton[0.045]; Hull[0.043]; Birmingham[0.043]; Rochdale[0.042]; Stockport[0.042]; Bristol[0.042]; Bristol[0.042]; Southend[0.042]; Charlton[0.042]; Luton[0.042]; Swindon[0.042]; City[0.041]; City[0.041]; Norwich[0.041]; Peterborough[0.041]; Hereford[0.041]; Millwall[0.041]; Burnley[0.041]; Watford[0.041]; Bury[0.041]; Scunthorpe[0.041]; Rotherham[0.041]; Darlington[0.041]; Shrewsbury[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Southend ==> ENTITY: \u001b[32mSouthend United F.C.\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.244:0.244[0]; log p(e|m)= -1.772:-1.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.043]; League[0.043]; Bristol[0.042]; Bristol[0.042]; Chesterfield[0.042]; Millwall[0.042]; Walsall[0.042]; Colchester[0.042]; Gillingham[0.042]; Stockport[0.042]; Oldham[0.042]; Watford[0.042]; Reading[0.042]; Shrewsbury[0.041]; Norwich[0.041]; Burnley[0.041]; Scunthorpe[0.041]; Portsmouth[0.041]; matches[0.041]; Manchester[0.041]; Barnsley[0.041]; Bury[0.041]; City[0.041]; City[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barnet ==> ENTITY: \u001b[32mBarnet F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.218:0.218[0]; log p(e|m)= -0.679:-0.679[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBournemouth[0.044]; Charlton[0.043]; Wycombe[0.043]; Watford[0.042]; Walsall[0.042]; Colchester[0.042]; Hull[0.042]; Shrewsbury[0.042]; Bury[0.042]; Rochdale[0.042]; Stockport[0.042]; Gillingham[0.042]; Burnley[0.042]; Rotherham[0.041]; Millwall[0.041]; Luton[0.041]; Carlisle[0.041]; Barnsley[0.041]; Blackpool[0.041]; Peterborough[0.041]; Hartlepool[0.041]; Chesterfield[0.041]; Northampton[0.041]; Cambridge[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Leyton Orient\u001b[39m ==> ENTITY: \u001b[32mLeyton Orient F.C.\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.223:0.223[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.043]; Hull[0.043]; Colchester[0.042]; Cardiff[0.042]; Stockport[0.042]; Walsall[0.042]; Chesterfield[0.042]; Watford[0.042]; Millwall[0.042]; Burnley[0.042]; Manchester[0.042]; Bournemouth[0.042]; Rochdale[0.042]; Northampton[0.041]; Rotherham[0.041]; Gillingham[0.041]; Bristol[0.041]; Bristol[0.041]; Shrewsbury[0.041]; Carlisle[0.041]; Barnsley[0.041]; Blackpool[0.041]; Scunthorpe[0.041]; Bolton[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wigan ==> ENTITY: \u001b[32mWigan Athletic F.C.\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.170:0.170[0]; log p(e|m)= -2.323:-2.323[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBristol[0.044]; Bristol[0.044]; Cardiff[0.043]; Hull[0.043]; Northampton[0.043]; Colchester[0.042]; Exeter[0.042]; Shrewsbury[0.042]; Rochdale[0.042]; Watford[0.042]; Blackpool[0.042]; Bolton[0.041]; City[0.041]; United[0.041]; Carlisle[0.041]; Rotherham[0.041]; Luton[0.041]; Peterborough[0.041]; Stockport[0.041]; Bury[0.041]; Burnley[0.041]; Walsall[0.041]; Hartlepool[0.040]; Millwall[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bolton ==> ENTITY: \u001b[32mBolton Wanderers F.C.\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.177:0.177[0]; log p(e|m)= -2.283:-2.283[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStockport[0.044]; Charlton[0.043]; League[0.043]; Hull[0.042]; Manchester[0.042]; Mansfield[0.042]; Bradford[0.042]; Watford[0.042]; York[0.042]; Northampton[0.042]; Oldham[0.042]; Luton[0.041]; Burnley[0.041]; matches[0.041]; Millwall[0.041]; Bristol[0.041]; Bristol[0.041]; Walsall[0.041]; Norwich[0.041]; Rochdale[0.041]; Tranmere[0.041]; Exeter[0.041]; Shrewsbury[0.041]; Carlisle[0.041]; \t\n\r [==================>...........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 12s41ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2075/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 993testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sarajevo\u001b[39m ==> ENTITY: \u001b[32mSarajevo\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.114:0.114[0]; log p(e|m)= -0.002:-0.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBosnia[0.051]; Bosnian[0.046]; Bosnian[0.046]; national[0.043]; specify[0.042]; guarantees[0.042]; nationalist[0.042]; Izetbegovic[0.041]; said[0.041]; said[0.041]; refugee[0.041]; President[0.041]; officials[0.041]; government[0.041]; government[0.041]; visit[0.041]; citing[0.040]; called[0.040]; provided[0.040]; provided[0.040]; Friday[0.040]; Friday[0.040]; Alija[0.040]; reassured[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = John Kornblum\u001b[39m ==> ENTITY: \u001b[32mJohn C. Kornblum\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.116:0.116[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSecretary[0.046]; Bosnia[0.046]; Bosnia[0.046]; President[0.045]; Sarajevo[0.045]; Security[0.044]; Alija[0.044]; national[0.044]; State[0.044]; provided[0.043]; provided[0.043]; Cooperation[0.043]; September[0.043]; Izetbegovic[0.043]; Bosnian[0.043]; Bosnian[0.043]; Organisation[0.043]; guarantees[0.042]; August[0.042]; Wednesday[0.042]; Wednesday[0.042]; Friday[0.042]; Friday[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = SDA ==> ENTITY: \u001b[32mParty of Democratic Action\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.142:0.142[0]; log p(e|m)= -0.395:-0.395[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBosnia[0.047]; Party[0.046]; elections[0.043]; elections[0.043]; elections[0.043]; Democratic[0.043]; higher[0.042]; offices[0.042]; parties[0.042]; radio[0.041]; joined[0.041]; Election[0.041]; decided[0.041]; Action[0.041]; prevent[0.041]; balloting[0.041]; voting[0.041]; voting[0.041]; voting[0.041]; followers[0.040]; abroad[0.040]; provided[0.040]; week[0.040]; living[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Party of Democratic Action\u001b[39m ==> ENTITY: \u001b[32mParty of Democratic Action\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBosnia[0.047]; parties[0.044]; Bosnian[0.043]; Bosnian[0.043]; elections[0.043]; elections[0.043]; elections[0.043]; elections[0.043]; voting[0.042]; voting[0.042]; voting[0.042]; ethnic[0.042]; joined[0.041]; Serb[0.041]; Alija[0.041]; September[0.041]; leader[0.040]; State[0.040]; national[0.040]; August[0.040]; balloting[0.040]; boycott[0.040]; President[0.040]; visit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = SDA ==> ENTITY: \u001b[32mParty of Democratic Action\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.161:0.161[0]; log p(e|m)= -0.395:-0.395[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBosnia[0.045]; Bosnia[0.045]; Party[0.043]; party[0.043]; party[0.043]; nationalist[0.043]; government[0.043]; government[0.043]; Bosnian[0.042]; Bosnian[0.042]; Action[0.042]; national[0.041]; elections[0.041]; elections[0.041]; Democratic[0.041]; leader[0.041]; Serb[0.041]; John[0.040]; vote[0.040]; radio[0.040]; radio[0.040]; radio[0.040]; Election[0.040]; voting[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bosnian ==> ENTITY: \u001b[32mBosnia and Herzegovina\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.165:0.165[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBosnia[0.052]; Bosnia[0.052]; Sarajevo[0.046]; Bosnian[0.045]; Alija[0.042]; refugee[0.042]; Moslem[0.042]; Moslem[0.042]; national[0.042]; nationalist[0.041]; living[0.041]; voters[0.040]; abroad[0.040]; said[0.040]; said[0.040]; Izetbegovic[0.040]; August[0.040]; boycott[0.039]; boycott[0.039]; boycott[0.039]; government[0.039]; government[0.039]; envoy[0.039]; envoy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = SDA ==> ENTITY: \u001b[32mParty of Democratic Action\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.395:-0.395[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBosnia[0.046]; Party[0.045]; government[0.043]; government[0.043]; officials[0.043]; Bosnian[0.043]; Bosnian[0.043]; elections[0.042]; elections[0.042]; elections[0.042]; Democratic[0.042]; leader[0.041]; higher[0.041]; offices[0.041]; radio[0.041]; radio[0.041]; radio[0.041]; Wednesday[0.040]; Wednesday[0.040]; joined[0.040]; Election[0.040]; decided[0.040]; Action[0.040]; parties[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bosnian ==> ENTITY: \u001b[32mBosnia and Herzegovina\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.176:0.176[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBosnia[0.053]; Bosnia[0.053]; Sarajevo[0.047]; Bosnian[0.047]; Serb[0.045]; Alija[0.044]; refugee[0.043]; refugees[0.043]; Moslem[0.043]; Moslem[0.043]; national[0.043]; Europe[0.043]; nationalist[0.043]; living[0.042]; voters[0.041]; abroad[0.041]; said[0.041]; said[0.041]; Izetbegovic[0.041]; August[0.041]; boycott[0.040]; boycott[0.040]; boycott[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moslem ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.103:0.103[0]; log p(e|m)= -1.214:-1.214[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoslem[0.049]; Bosnia[0.047]; Bosnian[0.045]; Bosnian[0.045]; Sarajevo[0.043]; refugee[0.043]; Alija[0.042]; nationalist[0.042]; citing[0.041]; said[0.041]; said[0.041]; national[0.041]; voters[0.041]; visit[0.041]; Izetbegovic[0.040]; met[0.040]; Secretary[0.040]; government[0.040]; government[0.040]; leader[0.040]; boycott[0.040]; boycott[0.040]; officials[0.040]; ruling[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moslem ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.103:0.103[0]; log p(e|m)= -1.214:-1.214[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoslem[0.049]; Bosnia[0.047]; Bosnian[0.045]; Bosnian[0.045]; Sarajevo[0.043]; refugee[0.043]; Alija[0.042]; nationalist[0.042]; citing[0.041]; said[0.041]; said[0.041]; national[0.041]; voters[0.041]; visit[0.041]; Izetbegovic[0.040]; met[0.040]; Secretary[0.040]; government[0.040]; government[0.040]; leader[0.040]; boycott[0.040]; boycott[0.040]; officials[0.040]; ruling[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSerb\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSerbs\u001b[39m <---> \u001b[32mSerbia\u001b[39m\t\nSCORES: global= 0.255:0.248[\u001b[31m0.007\u001b[39m]; local()= 0.122:0.108[\u001b[31m0.014\u001b[39m]; log p(e|m)= -0.027:-2.172[\u001b[31m2.144\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBosnia[0.049]; ethnic[0.047]; Bosnian[0.047]; Alija[0.046]; war[0.046]; living[0.045]; refugees[0.044]; refugees[0.044]; Europe[0.043]; Izetbegovic[0.043]; August[0.043]; municipal[0.043]; Wednesday[0.042]; Wednesday[0.042]; Friday[0.042]; President[0.042]; boycott[0.042]; Democratic[0.042]; Secretary[0.042]; officials[0.042]; Party[0.042]; leader[0.042]; September[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alija Izetbegovic\u001b[39m ==> ENTITY: \u001b[32mAlija Izetbegović\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBosnia[0.047]; Bosnia[0.047]; President[0.043]; Bosnian[0.043]; Bosnian[0.043]; nationalist[0.042]; envoy[0.042]; envoy[0.042]; Secretary[0.042]; elections[0.042]; elections[0.042]; elections[0.042]; Sarajevo[0.041]; government[0.041]; government[0.041]; Serb[0.041]; national[0.040]; leader[0.040]; State[0.040]; August[0.040]; boycott[0.040]; boycott[0.040]; boycott[0.040]; refugee[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnational[0.046]; John[0.044]; week[0.043]; Democratic[0.043]; presumably[0.042]; especially[0.042]; officials[0.042]; report[0.042]; abroad[0.041]; day[0.041]; radio[0.041]; radio[0.041]; radio[0.041]; September[0.041]; August[0.041]; citing[0.041]; broadcast[0.041]; began[0.041]; State[0.041]; specify[0.041]; President[0.040]; Secretary[0.040]; Wednesday[0.040]; Wednesday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bosnia ==> ENTITY: \u001b[32mBosnia and Herzegovina\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.111:0.111[0]; log p(e|m)= -0.345:-0.345[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSarajevo[0.046]; Bosnian[0.046]; Bosnian[0.046]; John[0.046]; national[0.044]; refugee[0.043]; government[0.043]; government[0.043]; report[0.042]; citing[0.041]; ruling[0.041]; boycott[0.041]; boycott[0.041]; Ruling[0.040]; Friday[0.040]; Friday[0.040]; Secretary[0.040]; nationalist[0.040]; Izetbegovic[0.040]; broadcast[0.040]; officials[0.040]; State[0.040]; envoy[0.040]; envoy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnational[0.045]; John[0.044]; abroad[0.044]; week[0.043]; Democratic[0.043]; Europe[0.043]; presumably[0.042]; report[0.041]; day[0.041]; radio[0.041]; radio[0.041]; radio[0.041]; September[0.041]; August[0.041]; citing[0.041]; broadcast[0.041]; began[0.041]; State[0.041]; Wednesday[0.041]; Wednesday[0.041]; specify[0.041]; President[0.040]; Secretary[0.040]; ruling[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.091:0.091[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnational[0.046]; John[0.044]; week[0.043]; Democratic[0.043]; presumably[0.043]; officials[0.042]; report[0.042]; day[0.042]; radio[0.042]; radio[0.042]; radio[0.042]; September[0.041]; August[0.041]; citing[0.041]; broadcast[0.041]; began[0.041]; State[0.041]; specify[0.041]; President[0.041]; Secretary[0.041]; Wednesday[0.040]; Wednesday[0.040]; ruling[0.040]; vote[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bosnia ==> ENTITY: \u001b[32mBosnia and Herzegovina\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.140:0.140[0]; log p(e|m)= -0.345:-0.345[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBosnian[0.045]; Bosnian[0.045]; war[0.045]; John[0.045]; Serb[0.043]; national[0.043]; refugees[0.042]; refugees[0.042]; government[0.042]; government[0.042]; Europe[0.041]; report[0.041]; ethnic[0.041]; August[0.041]; citing[0.041]; September[0.040]; week[0.040]; living[0.040]; boycott[0.040]; Wednesday[0.040]; Wednesday[0.040]; Friday[0.040]; Secretary[0.040]; Izetbegovic[0.039]; \t\n\r [===================>..........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s979ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2092/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 998testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Naina ==> ENTITY: \u001b[32mNaina Yeltsina\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.198:0.198[0]; log p(e|m)= -1.609:-1.609[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNaina[0.045]; Naina[0.045]; Naina[0.045]; Moscow[0.043]; Moscow[0.043]; visits[0.042]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; visited[0.041]; visited[0.041]; Russian[0.041]; Russian[0.041]; Russian[0.041]; Boris[0.041]; wife[0.041]; wife[0.041]; Sergei[0.041]; getting[0.040]; hospital[0.039]; hospital[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Interfax\u001b[39m ==> ENTITY: \u001b[32mInterfax\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.243:0.243[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInterfax[0.046]; news[0.045]; news[0.045]; Moscow[0.045]; Moscow[0.045]; Moscow[0.045]; Russian[0.044]; Russian[0.044]; Russian[0.044]; reports[0.044]; Clinical[0.042]; agency[0.041]; Sergei[0.041]; quoted[0.040]; quoted[0.040]; Yastrzhembsky[0.040]; Yastrzhembsky[0.040]; hospital[0.039]; hospital[0.039]; spokesman[0.039]; saying[0.039]; saying[0.039]; Hospital[0.038]; Friday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Barvikha\u001b[39m ==> ENTITY: \u001b[32mBarvikha\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinical[0.044]; sanatorium[0.044]; Moscow[0.043]; Moscow[0.043]; Moscow[0.043]; Moscow[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; hospital[0.042]; Hospital[0.041]; president[0.040]; Yastrzhembsky[0.040]; Yastrzhembsky[0.040]; treated[0.040]; denied[0.040]; vacation[0.040]; outside[0.040]; outside[0.040]; Central[0.040]; office[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.175:0.175[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.047]; Moscow[0.047]; Moscow[0.047]; Russian[0.044]; Russian[0.044]; Barvikha[0.043]; Sergei[0.042]; outside[0.041]; outside[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; second[0.040]; Interfax[0.040]; Friday[0.040]; taken[0.040]; Saturday[0.040]; Saturday[0.040]; attacks[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.189:0.189[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.048]; Moscow[0.048]; Moscow[0.048]; Sergei[0.046]; Russian[0.045]; Russian[0.045]; Russian[0.045]; Boris[0.045]; Barvikha[0.043]; outside[0.042]; outside[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; second[0.042]; Friday[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sergei Yastrzhembsky\u001b[39m ==> ENTITY: \u001b[32mSergey Yastrzhembsky\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npresident[0.050]; Clinical[0.048]; President[0.048]; Moscow[0.046]; Moscow[0.046]; Moscow[0.046]; said[0.046]; said[0.046]; spokesman[0.046]; hospital[0.046]; hospital[0.046]; Yastrzhembsky[0.045]; saying[0.045]; saying[0.045]; Hospital[0.045]; Central[0.044]; telephone[0.044]; quoted[0.044]; quoted[0.044]; Interfax[0.044]; Interfax[0.044]; agency[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Boris Yeltsin\u001b[39m ==> ENTITY: \u001b[32mBoris Yeltsin\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.045]; Russian[0.043]; Russian[0.043]; Russian[0.043]; Moscow[0.043]; Moscow[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Sergei[0.042]; Interfax[0.041]; Interfax[0.041]; hospital[0.040]; hospital[0.040]; Yastrzhembsky[0.039]; Yastrzhembsky[0.039]; visited[0.039]; visited[0.039]; news[0.039]; news[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yeltsin\u001b[39m ==> ENTITY: \u001b[32mBoris Yeltsin\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.043]; Russian[0.043]; Moscow[0.043]; Moscow[0.043]; Moscow[0.043]; Moscow[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; president[0.042]; Clinical[0.042]; Sergei[0.042]; elected[0.041]; hospital[0.040]; office[0.039]; Yastrzhembsky[0.039]; Yastrzhembsky[0.039]; Yastrzhembsky[0.039]; Hospital[0.039]; denied[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yastrzhembsky\u001b[39m ==> ENTITY: \u001b[32mSergey Yastrzhembsky\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npresident[0.045]; Clinical[0.044]; President[0.044]; Moscow[0.042]; Moscow[0.042]; Moscow[0.042]; Moscow[0.042]; said[0.042]; said[0.042]; said[0.042]; spokesman[0.042]; office[0.042]; hospital[0.042]; hospital[0.042]; Yastrzhembsky[0.041]; Yastrzhembsky[0.041]; saying[0.041]; saying[0.041]; Hospital[0.041]; denied[0.040]; Central[0.040]; telephone[0.040]; quoted[0.040]; quoted[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.049]; Moscow[0.049]; outside[0.047]; office[0.042]; Barvikha[0.042]; second[0.042]; Saturday[0.042]; night[0.041]; attacks[0.041]; miles[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Central[0.041]; twice[0.041]; July[0.041]; spend[0.041]; sanatorium[0.040]; taken[0.040]; return[0.039]; Hospital[0.039]; holiday[0.039]; hospital[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.170:0.170[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.048]; Russian[0.048]; Moscow[0.045]; Moscow[0.045]; Moscow[0.045]; Boris[0.043]; Sergei[0.042]; Interfax[0.042]; Interfax[0.042]; spoken[0.041]; Central[0.041]; Yastrzhembsky[0.040]; Yastrzhembsky[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; president[0.040]; President[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yeltsin\u001b[39m ==> ENTITY: \u001b[32mBoris Yeltsin\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.045]; Russian[0.044]; Russian[0.044]; Russian[0.044]; Moscow[0.044]; Moscow[0.044]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Yeltsin[0.043]; Sergei[0.042]; Interfax[0.041]; Interfax[0.041]; Boris[0.041]; hospital[0.040]; hospital[0.040]; Yastrzhembsky[0.040]; visited[0.039]; visited[0.039]; news[0.039]; news[0.039]; saying[0.039]; visits[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.170:0.170[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.048]; Russian[0.045]; Russian[0.045]; Russian[0.045]; Sergei[0.043]; Boris[0.043]; outside[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Interfax[0.041]; Interfax[0.041]; Friday[0.041]; Saturday[0.040]; Yastrzhembsky[0.040]; Yastrzhembsky[0.040]; visited[0.040]; visited[0.040]; operation[0.039]; hospital[0.039]; hospital[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yeltsin\u001b[39m ==> ENTITY: \u001b[32mBoris Yeltsin\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.044]; Russian[0.042]; Russian[0.042]; Russian[0.042]; Moscow[0.042]; Moscow[0.042]; Moscow[0.042]; Moscow[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; president[0.042]; Clinical[0.042]; Sergei[0.041]; elected[0.041]; Interfax[0.040]; Boris[0.040]; hospital[0.039]; hospital[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yeltsin\u001b[39m ==> ENTITY: \u001b[32mBoris Yeltsin\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.044]; Russian[0.042]; Russian[0.042]; Russian[0.042]; Moscow[0.042]; Moscow[0.042]; Moscow[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; president[0.042]; Clinical[0.042]; Sergei[0.041]; elected[0.041]; Interfax[0.040]; Interfax[0.040]; Boris[0.040]; hospital[0.039]; hospital[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Naina ==> ENTITY: \u001b[32mNaina Yeltsina\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.218:0.218[0]; log p(e|m)= -1.609:-1.609[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNaina[0.045]; Naina[0.045]; Naina[0.045]; Naina[0.045]; Moscow[0.042]; Moscow[0.042]; Moscow[0.042]; visits[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; Yeltsin[0.041]; visited[0.041]; visited[0.041]; Russian[0.041]; Russian[0.041]; Russian[0.041]; Boris[0.041]; wife[0.040]; wife[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yeltsin\u001b[39m ==> ENTITY: \u001b[32mBoris Yeltsin\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.046]; Moscow[0.045]; Moscow[0.045]; Moscow[0.045]; Moscow[0.045]; Yeltsin[0.045]; Yeltsin[0.045]; Yeltsin[0.045]; Yeltsin[0.045]; Yeltsin[0.045]; president[0.045]; Clinical[0.045]; elected[0.044]; hospital[0.042]; office[0.042]; Yastrzhembsky[0.042]; Yastrzhembsky[0.042]; Hospital[0.041]; denied[0.041]; visited[0.041]; Barvikha[0.041]; saying[0.040]; Central[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Central Clinical Hospital\u001b[39m ==> ENTITY: \u001b[32mCentral Clinical Hospital\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.045]; Russian[0.045]; Moscow[0.044]; Moscow[0.044]; Moscow[0.044]; Moscow[0.044]; Sergei[0.042]; Barvikha[0.042]; treated[0.041]; hospital[0.041]; hospital[0.041]; taken[0.041]; seen[0.041]; denied[0.041]; visited[0.041]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Interfax[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Interfax\u001b[39m ==> ENTITY: \u001b[32mInterfax\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.248:0.248[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInterfax[0.049]; news[0.048]; news[0.048]; Moscow[0.047]; Moscow[0.047]; Russian[0.046]; Russian[0.046]; Russian[0.046]; reports[0.046]; agency[0.043]; Sergei[0.043]; quoted[0.043]; Yastrzhembsky[0.042]; Yastrzhembsky[0.042]; hospital[0.041]; hospital[0.041]; spokesman[0.041]; saying[0.041]; Friday[0.040]; President[0.040]; Boris[0.040]; telephone[0.039]; Earlier[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.176:0.176[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.048]; Russian[0.048]; Moscow[0.045]; Moscow[0.045]; Moscow[0.045]; Boris[0.043]; Sergei[0.042]; Interfax[0.042]; Interfax[0.042]; spoken[0.041]; Central[0.041]; Yastrzhembsky[0.040]; Yastrzhembsky[0.040]; Yastrzhembsky[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.048]; Moscow[0.048]; Moscow[0.048]; Central[0.042]; outside[0.042]; outside[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; second[0.041]; Barvikha[0.041]; office[0.041]; Saturday[0.040]; attacks[0.040]; twice[0.040]; July[0.040]; miles[0.040]; spend[0.040]; Yastrzhembsky[0.040]; Yastrzhembsky[0.040]; sanatorium[0.039]; taken[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.173:0.173[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.050]; Russian[0.050]; Moscow[0.047]; Moscow[0.047]; Boris[0.045]; Sergei[0.045]; Interfax[0.044]; Interfax[0.044]; spoken[0.043]; Yastrzhembsky[0.042]; Yastrzhembsky[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; Yeltsin[0.042]; President[0.042]; visited[0.041]; visited[0.041]; outside[0.040]; children[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yastrzhembsky\u001b[39m ==> ENTITY: \u001b[32mSergey Yastrzhembsky\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npresident[0.046]; Clinical[0.045]; Moscow[0.043]; Moscow[0.043]; Moscow[0.043]; Moscow[0.043]; said[0.043]; said[0.043]; office[0.042]; hospital[0.042]; Yastrzhembsky[0.042]; saying[0.042]; Hospital[0.041]; denied[0.041]; Central[0.041]; telephone[0.041]; quoted[0.041]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; Yeltsin[0.040]; aides[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yeltsin\u001b[39m ==> ENTITY: \u001b[32mBoris Yeltsin\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.046]; Russian[0.044]; Russian[0.044]; Russian[0.044]; Moscow[0.044]; Moscow[0.044]; Moscow[0.044]; Yeltsin[0.044]; Yeltsin[0.044]; Yeltsin[0.044]; Yeltsin[0.044]; Yeltsin[0.044]; Yeltsin[0.044]; Yeltsin[0.044]; Yeltsin[0.044]; Yeltsin[0.044]; president[0.043]; Clinical[0.043]; Sergei[0.042]; elected[0.042]; Interfax[0.042]; Interfax[0.042]; Boris[0.041]; \t\n\r [===================>..........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s878ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2116/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1087testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moldova ==> ENTITY: \u001b[32mMoldova national football team\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.140:0.140[0]; log p(e|m)= -2.419:-2.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoldova[0.056]; England[0.052]; England[0.052]; soccer[0.050]; European[0.050]; Soccer[0.049]; qualifier[0.049]; Qualifier[0.047]; Chisinau[0.047]; Saturday[0.045]; group[0.044]; championship[0.044]; beat[0.044]; minute[0.043]; Bruce[0.043]; Beat[0.042]; Darren[0.042]; halftime[0.042]; Attendance[0.042]; Scorers[0.041]; Dyer[0.039]; Eadie[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland national football team\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.099:0.099[0]; log p(e|m)= -2.781:-2.781[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.056]; soccer[0.052]; Soccer[0.050]; qualifier[0.050]; Qualifier[0.048]; Bruce[0.047]; Darren[0.047]; European[0.046]; beat[0.046]; Saturday[0.046]; championship[0.045]; Dyer[0.045]; Moldova[0.044]; Moldova[0.044]; group[0.044]; Scorers[0.043]; minute[0.043]; halftime[0.042]; Attendance[0.042]; Beat[0.041]; Eadie[0.040]; Chisinau[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Darren Eadie\u001b[39m ==> ENTITY: \u001b[32mDarren Eadie\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.052:0.052[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaturday[0.052]; Soccer[0.052]; qualifier[0.050]; Qualifier[0.050]; soccer[0.050]; England[0.049]; England[0.049]; Bruce[0.049]; halftime[0.049]; Scorers[0.048]; Attendance[0.048]; championship[0.048]; European[0.047]; group[0.046]; minute[0.046]; beat[0.046]; Chisinau[0.045]; Dyer[0.044]; Moldova[0.044]; Moldova[0.044]; Beat[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bruce Dyer\u001b[39m ==> ENTITY: \u001b[32mBruce Dyer\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.070:0.070[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.055]; England[0.055]; Darren[0.053]; Saturday[0.050]; qualifier[0.049]; Soccer[0.049]; Qualifier[0.049]; minute[0.049]; soccer[0.048]; halftime[0.048]; Scorers[0.048]; Attendance[0.048]; championship[0.047]; European[0.045]; beat[0.045]; group[0.044]; Chisinau[0.044]; Moldova[0.044]; Moldova[0.044]; Eadie[0.043]; Beat[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chisinau\u001b[39m ==> ENTITY: \u001b[32mChișinău\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.039:0.039[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoldova[0.052]; Moldova[0.052]; qualifier[0.051]; Qualifier[0.051]; halftime[0.049]; Scorers[0.049]; Attendance[0.049]; Saturday[0.049]; European[0.048]; Soccer[0.048]; soccer[0.047]; minute[0.047]; championship[0.047]; England[0.046]; England[0.046]; Darren[0.045]; group[0.045]; beat[0.045]; Bruce[0.045]; Beat[0.045]; Eadie[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = European ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.241:0.241[0]; local()= 0.007:0.007[0]; log p(e|m)= -0.256:-0.256[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.055]; Soccer[0.054]; England[0.052]; England[0.052]; qualifier[0.052]; Qualifier[0.051]; group[0.049]; championship[0.049]; Moldova[0.048]; Moldova[0.048]; beat[0.047]; Saturday[0.047]; minute[0.046]; Chisinau[0.045]; Attendance[0.045]; Bruce[0.044]; Beat[0.043]; Darren[0.043]; halftime[0.043]; Scorers[0.042]; Dyer[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moldova ==> ENTITY: \u001b[32mMoldova national football team\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.140:0.140[0]; log p(e|m)= -2.419:-2.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoldova[0.056]; England[0.052]; England[0.052]; soccer[0.050]; European[0.050]; Soccer[0.049]; qualifier[0.049]; Qualifier[0.047]; Chisinau[0.047]; Saturday[0.045]; group[0.044]; championship[0.044]; beat[0.044]; minute[0.043]; Bruce[0.043]; Beat[0.042]; Darren[0.042]; halftime[0.042]; Attendance[0.042]; Scorers[0.041]; Dyer[0.039]; Eadie[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland national football team\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.099:0.099[0]; log p(e|m)= -2.781:-2.781[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.056]; soccer[0.052]; Soccer[0.050]; qualifier[0.050]; Qualifier[0.048]; Bruce[0.047]; Darren[0.047]; European[0.046]; beat[0.046]; Saturday[0.046]; championship[0.045]; Dyer[0.045]; Moldova[0.044]; Moldova[0.044]; group[0.044]; Scorers[0.043]; minute[0.043]; halftime[0.042]; Attendance[0.042]; Beat[0.041]; Eadie[0.040]; Chisinau[0.039]; \t\n\r [===================>..........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s874ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2124/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1162testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dhaka ==> ENTITY: \u001b[32mDhaka\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.021:-0.021[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDhaka[0.049]; Dhaka[0.049]; Rizvi[0.046]; Khurshid[0.043]; Shakil[0.043]; Alam[0.043]; brokers[0.043]; said[0.041]; said[0.041]; said[0.041]; investors[0.041]; investors[0.041]; Friday[0.041]; Exchange[0.041]; close[0.040]; mainly[0.040]; Stock[0.040]; Moslem[0.040]; sidelines[0.040]; weekend[0.040]; following[0.040]; selling[0.039]; overseas[0.039]; stocks[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dhaka Stock Exchange\u001b[39m ==> ENTITY: \u001b[32mDhaka Stock Exchange\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDhaka[0.048]; Dhaka[0.048]; brokers[0.047]; investors[0.045]; investors[0.045]; Friday[0.044]; close[0.044]; Rizvi[0.044]; Brokers[0.044]; selling[0.043]; stocks[0.043]; Khurshid[0.043]; buying[0.043]; market[0.043]; market[0.043]; market[0.043]; market[0.043]; overseas[0.042]; mainly[0.042]; players[0.041]; broker[0.041]; uptrend[0.041]; sidelines[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMoslem\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMuslim\u001b[39m <---> \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.254:0.251[\u001b[31m0.003\u001b[39m]; local()= 0.106:0.091[\u001b[31m0.015\u001b[39m]; log p(e|m)= -0.308:-1.214[\u001b[31m0.906\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBangladesh[0.044]; Bangladesh[0.044]; continued[0.043]; mainly[0.043]; Rizvi[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; engineering[0.042]; expected[0.042]; expected[0.042]; expected[0.042]; following[0.042]; Atlas[0.041]; like[0.041]; Shakil[0.041]; Dhaka[0.040]; Dhaka[0.040]; Dhaka[0.040]; Alam[0.040]; strong[0.040]; despite[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mDhaka\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mDhaka\u001b[39m <---> \u001b[32mDhaka Stock Exchange\u001b[39m\t\nSCORES: global= 0.268:0.245[\u001b[31m0.023\u001b[39m]; local()= 0.100:0.137[\u001b[32m0.037\u001b[39m]; log p(e|m)= -0.021:-2.397[\u001b[31m2.376\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDhaka[0.051]; Dhaka[0.051]; Rizvi[0.048]; Khurshid[0.045]; Shakil[0.045]; Alam[0.045]; brokers[0.045]; said[0.043]; said[0.043]; said[0.043]; investors[0.043]; investors[0.043]; Friday[0.043]; Exchange[0.042]; mainly[0.042]; Stock[0.042]; Moslem[0.042]; sidelines[0.041]; weekend[0.041]; selling[0.041]; overseas[0.041]; stocks[0.041]; buying[0.041]; \t\n\r [===================>..........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s907ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2128/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 995testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Minsk ==> ENTITY: \u001b[32mMinsk\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.071:-0.071[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.045]; Belarus[0.045]; Belarus[0.045]; Soviet[0.045]; republic[0.045]; Sayenko[0.044]; national[0.043]; state[0.042]; November[0.042]; million[0.042]; Semyon[0.042]; Friday[0.041]; community[0.041]; European[0.041]; Larisa[0.040]; Alexander[0.040]; people[0.040]; parliament[0.040]; world[0.040]; fact[0.039]; attempting[0.039]; Lukashenko[0.039]; Lukashenko[0.039]; Reuters[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lukashenko\u001b[39m ==> ENTITY: \u001b[32mAlexander Lukashenko\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.217:0.217[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.044]; President[0.044]; president[0.044]; president[0.044]; Belarus[0.043]; Belarus[0.043]; Belarus[0.043]; Lukashenko[0.043]; Lukashenko[0.043]; Lukashenko[0.043]; parliament[0.043]; parliament[0.043]; elected[0.042]; Soviet[0.042]; Minsk[0.041]; political[0.040]; constitution[0.040]; constitution[0.040]; constitution[0.040]; referendum[0.039]; accused[0.039]; republic[0.039]; European[0.039]; senior[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.042:0.042[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnearly[0.046]; Russia[0.046]; groups[0.044]; leaders[0.043]; said[0.043]; believes[0.043]; week[0.042]; taking[0.042]; people[0.042]; federation[0.042]; Moscow[0.042]; Moscow[0.042]; strong[0.041]; economic[0.041]; April[0.041]; Sergei[0.040]; backing[0.040]; political[0.040]; political[0.040]; signed[0.040]; granted[0.040]; responded[0.040]; protested[0.040]; renewed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.049]; Russia[0.046]; Russia[0.046]; Soviet[0.045]; won[0.045]; Sergei[0.044]; Union[0.042]; battled[0.042]; power[0.042]; nearly[0.041]; April[0.041]; office[0.041]; period[0.041]; domination[0.040]; taking[0.040]; federation[0.040]; responded[0.040]; order[0.040]; week[0.040]; fight[0.039]; parliament[0.039]; parliament[0.039]; signed[0.039]; restore[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoviet[0.046]; Moscow[0.046]; won[0.045]; president[0.043]; Sergei[0.043]; parliament[0.043]; parliament[0.043]; parliament[0.043]; signed[0.043]; economic[0.042]; federation[0.041]; representatives[0.041]; said[0.041]; said[0.041]; Union[0.041]; represented[0.040]; period[0.040]; Lukashenko[0.040]; Lukashenko[0.040]; Lukashenko[0.040]; Lukashenko[0.040]; battled[0.040]; constitution[0.040]; constitution[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = European ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.083:0.083[0]; log p(e|m)= -0.256:-0.256[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnational[0.046]; elected[0.045]; majority[0.045]; represented[0.045]; world[0.044]; said[0.044]; community[0.044]; seat[0.044]; won[0.044]; people[0.043]; people[0.043]; charge[0.043]; million[0.043]; Belarus[0.043]; Belarus[0.043]; Belarus[0.043]; asked[0.043]; representatives[0.043]; say[0.043]; regionally[0.043]; accused[0.043]; referendum[0.042]; November[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belarus ==> ENTITY: \u001b[32mBelarus\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.422:-0.422[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.046]; Belarus[0.046]; Soviet[0.044]; national[0.044]; European[0.043]; parliament[0.043]; parliament[0.043]; Sayenko[0.042]; referendum[0.042]; majority[0.042]; president[0.041]; president[0.041]; Minsk[0.041]; republic[0.041]; representatives[0.041]; President[0.041]; President[0.041]; accused[0.040]; elected[0.040]; November[0.040]; constitution[0.040]; constitution[0.040]; constitution[0.040]; world[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.046]; Moscow[0.046]; Sergei[0.043]; parliament[0.043]; parliament[0.043]; signed[0.043]; federation[0.042]; nationalist[0.042]; nationalist[0.042]; nationalist[0.042]; economic[0.042]; United[0.041]; granted[0.041]; nearly[0.041]; said[0.041]; said[0.041]; States[0.041]; April[0.040]; period[0.040]; Lukashenko[0.040]; Lukashenko[0.040]; Lukashenko[0.040]; battled[0.040]; renewed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Soviet Union ==> ENTITY: \u001b[32mSoviet Union\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.114:0.114[0]; log p(e|m)= -0.076:-0.076[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.048]; represented[0.046]; Moscow[0.045]; Moscow[0.045]; federation[0.045]; won[0.044]; Sergei[0.042]; trade[0.041]; nationalist[0.041]; nationalist[0.041]; groups[0.041]; people[0.041]; political[0.041]; political[0.041]; collapse[0.040]; named[0.040]; opponents[0.040]; quasi[0.040]; period[0.040]; April[0.040]; pact[0.040]; union[0.040]; members[0.040]; hands[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lukashenko\u001b[39m ==> ENTITY: \u001b[32mAlexander Lukashenko\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.224:0.224[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.044]; president[0.044]; Belarus[0.043]; presidential[0.043]; Lukashenko[0.043]; Lukashenko[0.043]; Lukashenko[0.043]; parliament[0.043]; parliament[0.043]; parliament[0.043]; parliament[0.043]; elected[0.042]; Russia[0.042]; Soviet[0.042]; trade[0.041]; won[0.041]; political[0.040]; polls[0.040]; office[0.040]; constitution[0.040]; constitution[0.040]; constitution[0.040]; referendum[0.039]; Union[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belarus ==> ENTITY: \u001b[32mBelarus\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.422:-0.422[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.047]; Belarus[0.047]; Soviet[0.045]; European[0.043]; parliament[0.043]; Sayenko[0.043]; referendum[0.042]; president[0.042]; Minsk[0.041]; republic[0.041]; national[0.041]; President[0.041]; President[0.041]; accused[0.041]; November[0.041]; constitution[0.041]; constitution[0.041]; world[0.041]; Lukashenko[0.040]; Lukashenko[0.040]; community[0.040]; said[0.039]; powers[0.039]; Alexander[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alexander Lukashenko\u001b[39m ==> ENTITY: \u001b[32mAlexander Lukashenko\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.045]; President[0.045]; president[0.044]; Belarus[0.044]; Belarus[0.044]; Belarus[0.044]; Lukashenko[0.044]; parliament[0.044]; parliament[0.044]; elected[0.043]; Soviet[0.042]; Minsk[0.042]; constitution[0.040]; constitution[0.040]; constitution[0.040]; referendum[0.040]; accused[0.040]; republic[0.040]; European[0.040]; senior[0.040]; Sayenko[0.039]; November[0.039]; national[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lukashenko\u001b[39m ==> ENTITY: \u001b[32mAlexander Lukashenko\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStates[0.043]; presidential[0.043]; Lukashenko[0.043]; Lukashenko[0.043]; Lukashenko[0.043]; parliament[0.043]; parliament[0.043]; opposition[0.042]; opposition[0.042]; Russia[0.042]; Russia[0.042]; economic[0.042]; Soviet[0.042]; groups[0.041]; democratic[0.041]; trade[0.041]; Moscow[0.041]; Moscow[0.041]; United[0.041]; won[0.041]; political[0.040]; political[0.040]; polls[0.040]; granted[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lukashenko\u001b[39m ==> ENTITY: \u001b[32mAlexander Lukashenko\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.209:0.209[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.045]; president[0.044]; presidential[0.043]; Lukashenko[0.043]; Lukashenko[0.043]; Lukashenko[0.043]; parliament[0.043]; parliament[0.043]; parliament[0.043]; elected[0.042]; Russia[0.042]; Soviet[0.042]; trade[0.042]; won[0.041]; political[0.041]; polls[0.041]; office[0.040]; constitution[0.040]; constitution[0.040]; constitution[0.040]; Union[0.040]; European[0.040]; Sergei[0.039]; current[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.060:0.060[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntold[0.045]; million[0.044]; European[0.043]; charge[0.043]; said[0.043]; world[0.042]; regionally[0.042]; parliament[0.042]; parliament[0.042]; senate[0.042]; national[0.042]; Alexander[0.041]; people[0.041]; Friday[0.041]; aides[0.041]; establish[0.041]; dictatorship[0.041]; dictatorship[0.041]; state[0.041]; fact[0.041]; November[0.041]; majority[0.040]; republic[0.040]; increase[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lukashenko\u001b[39m ==> ENTITY: \u001b[32mAlexander Lukashenko\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStates[0.044]; Lukashenko[0.044]; Lukashenko[0.044]; parliament[0.044]; parliament[0.044]; opposition[0.043]; opposition[0.043]; Russia[0.043]; economic[0.042]; groups[0.042]; democratic[0.042]; trade[0.042]; Moscow[0.042]; Moscow[0.042]; United[0.041]; political[0.041]; political[0.041]; granted[0.040]; constitution[0.040]; referendum[0.040]; protested[0.040]; federation[0.040]; signed[0.039]; Sergei[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.126:0.126[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.053]; Russia[0.053]; Soviet[0.050]; Sergei[0.048]; Union[0.046]; battled[0.046]; federation[0.045]; nearly[0.045]; April[0.045]; States[0.044]; office[0.044]; period[0.044]; taking[0.044]; responded[0.044]; economic[0.043]; week[0.043]; nationalist[0.043]; nationalist[0.043]; nationalist[0.043]; parliament[0.043]; parliament[0.043]; domination[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belarus ==> ENTITY: \u001b[32mBelarus\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.422:-0.422[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.046]; Belarus[0.046]; Soviet[0.044]; national[0.044]; European[0.043]; parliament[0.043]; parliament[0.043]; Sayenko[0.042]; referendum[0.042]; majority[0.042]; president[0.041]; Minsk[0.041]; republic[0.041]; President[0.041]; President[0.041]; accused[0.041]; November[0.040]; constitution[0.040]; constitution[0.040]; constitution[0.040]; world[0.040]; Lukashenko[0.040]; Lukashenko[0.040]; community[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.089:0.089[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntold[0.044]; trade[0.044]; political[0.044]; links[0.043]; main[0.042]; European[0.042]; said[0.042]; said[0.042]; said[0.042]; world[0.042]; regionally[0.041]; parliament[0.041]; parliament[0.041]; parliament[0.041]; current[0.041]; senate[0.041]; period[0.041]; fact[0.041]; million[0.041]; say[0.041]; polls[0.041]; adviser[0.040]; people[0.040]; people[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Soviet ==> ENTITY: \u001b[32mSoviet Union\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.102:0.102[0]; log p(e|m)= -0.098:-0.098[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.048]; Belarus[0.048]; Belarus[0.048]; republic[0.047]; President[0.047]; President[0.047]; elected[0.047]; Sayenko[0.046]; Minsk[0.046]; president[0.045]; Semyon[0.045]; parliament[0.045]; parliament[0.045]; representatives[0.045]; represented[0.045]; November[0.044]; national[0.044]; members[0.044]; speaker[0.044]; named[0.043]; Alexander[0.043]; Larisa[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lukashenko\u001b[39m ==> ENTITY: \u001b[32mAlexander Lukashenko\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStates[0.044]; Lukashenko[0.044]; Lukashenko[0.044]; parliament[0.044]; parliament[0.044]; opposition[0.043]; opposition[0.043]; Russia[0.043]; economic[0.042]; groups[0.042]; democratic[0.042]; trade[0.042]; Moscow[0.042]; Moscow[0.042]; United[0.041]; political[0.041]; political[0.041]; granted[0.040]; constitution[0.040]; referendum[0.040]; protested[0.040]; federation[0.040]; signed[0.039]; Sergei[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lukashenko\u001b[39m ==> ENTITY: \u001b[32mAlexander Lukashenko\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npresident[0.044]; presidential[0.044]; Lukashenko[0.043]; Lukashenko[0.043]; Lukashenko[0.043]; parliament[0.043]; parliament[0.043]; parliament[0.043]; elected[0.042]; Russia[0.042]; Soviet[0.042]; democratic[0.042]; trade[0.042]; won[0.041]; political[0.041]; polls[0.041]; office[0.040]; constitution[0.040]; constitution[0.040]; constitution[0.040]; referendum[0.040]; Union[0.040]; European[0.040]; signed[0.039]; \t\n\r [===================>..........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s811ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2150/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1006testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Southeast Asian\u001b[39m ==> ENTITY: \u001b[32mSoutheast Asia\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.111:0.111[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.050]; Singapore[0.050]; East[0.046]; East[0.046]; Indonesian[0.046]; South[0.045]; Korean[0.044]; origin[0.044]; seasonal[0.043]; Australian[0.043]; range[0.043]; Saran[0.043]; said[0.043]; Kong[0.042]; London[0.042]; seen[0.042]; Hong[0.041]; market[0.041]; market[0.041]; market[0.041]; fourth[0.041]; fit[0.041]; bullion[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = South Korean\u001b[39m ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.107:0.107[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKorean[0.048]; Koreans[0.048]; Seoul[0.046]; Singapore[0.042]; Singapore[0.042]; price[0.042]; presidential[0.042]; recent[0.042]; Hong[0.041]; government[0.041]; steady[0.041]; fourth[0.041]; months[0.041]; week[0.041]; election[0.041]; demand[0.040]; prices[0.040]; prices[0.040]; prices[0.040]; previous[0.040]; South[0.040]; selling[0.040]; start[0.040]; ahead[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = International Monetary Fund\u001b[39m ==> ENTITY: \u001b[32mInternational Monetary Fund\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.100:0.100[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrade[0.046]; market[0.045]; countries[0.044]; prices[0.044]; demand[0.043]; price[0.042]; report[0.042]; ahead[0.042]; September[0.042]; sell[0.041]; October[0.041]; October[0.041]; basis[0.041]; debts[0.041]; versus[0.041]; remained[0.041]; developing[0.041]; retracement[0.040]; arbitrage[0.040]; Thursday[0.040]; mid[0.040]; selling[0.040]; contract[0.040]; eased[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Seoul ==> ENTITY: \u001b[32mSeoul\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.038:-0.038[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfourth[0.045]; Korean[0.045]; Korean[0.045]; Koreans[0.043]; government[0.043]; South[0.043]; South[0.043]; Singapore[0.042]; Singapore[0.042]; trade[0.042]; start[0.042]; finish[0.042]; Thursday[0.042]; closed[0.041]; Friday[0.041]; Hong[0.041]; recent[0.041]; months[0.041]; demand[0.041]; quarter[0.040]; selling[0.040]; revival[0.040]; marked[0.039]; following[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = South Korean\u001b[39m ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.105:0.105[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSeoul[0.051]; Korean[0.049]; Asian[0.049]; Koreans[0.048]; Singapore[0.047]; Singapore[0.047]; price[0.043]; South[0.043]; government[0.042]; fourth[0.042]; gold[0.042]; gold[0.042]; gold[0.042]; London[0.041]; week[0.041]; week[0.041]; week[0.041]; Indonesian[0.041]; Indonesian[0.041]; demand[0.041]; prices[0.041]; prices[0.041]; prices[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.084:0.084[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYork[0.046]; closed[0.043]; September[0.043]; September[0.043]; International[0.043]; added[0.043]; October[0.042]; October[0.042]; Thursday[0.042]; week[0.042]; mid[0.042]; market[0.042]; Monetary[0.042]; said[0.041]; said[0.041]; helped[0.041]; support[0.041]; expected[0.040]; demand[0.040]; sell[0.040]; meet[0.040]; Silver[0.040]; report[0.040]; contract[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hong Kong ==> ENTITY: \u001b[32mHong Kong\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.048]; fourth[0.045]; Asian[0.045]; week[0.044]; week[0.044]; week[0.044]; spot[0.043]; Australian[0.043]; previous[0.042]; East[0.041]; East[0.041]; Friday[0.041]; continuing[0.041]; range[0.041]; better[0.041]; position[0.041]; quarter[0.040]; seen[0.040]; Moribund[0.039]; thumped[0.039]; fit[0.039]; awaiting[0.039]; unchanged[0.039]; bars[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australian ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.491:-0.491[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseasonal[0.045]; market[0.044]; market[0.044]; market[0.044]; South[0.044]; South[0.044]; fourth[0.042]; following[0.042]; Singapore[0.042]; Singapore[0.042]; trade[0.041]; origin[0.041]; range[0.041]; better[0.041]; week[0.041]; week[0.041]; week[0.041]; Asian[0.041]; Friday[0.040]; sell[0.040]; bars[0.040]; sales[0.040]; spot[0.040]; fit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.068:0.068[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclosed[0.045]; September[0.044]; September[0.044]; added[0.044]; York[0.044]; October[0.043]; October[0.043]; Thursday[0.043]; market[0.043]; saw[0.042]; said[0.042]; said[0.042]; expected[0.041]; demand[0.041]; think[0.041]; meet[0.041]; start[0.040]; contract[0.040]; support[0.040]; range[0.040]; basis[0.039]; Silver[0.039]; Comex[0.039]; remained[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesian ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.146:0.146[0]; log p(e|m)= -0.742:-0.742[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesian[0.047]; countries[0.045]; Singapore[0.045]; recent[0.043]; International[0.043]; origin[0.042]; remained[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; Korean[0.041]; Korean[0.041]; Koreans[0.041]; mid[0.041]; South[0.041]; South[0.041]; government[0.040]; report[0.040]; week[0.040]; crackdown[0.040]; ahead[0.040]; Hong[0.040]; following[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.053]; Asian[0.050]; start[0.045]; Australian[0.044]; London[0.044]; fourth[0.043]; Seoul[0.043]; Indonesian[0.043]; Indonesian[0.043]; Koreans[0.043]; following[0.043]; ahead[0.043]; gold[0.043]; gold[0.043]; spot[0.042]; Korean[0.042]; Korean[0.042]; week[0.042]; week[0.042]; week[0.042]; coming[0.042]; market[0.042]; trade[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.118:0.118[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.051]; Asian[0.048]; Australian[0.042]; London[0.042]; fourth[0.042]; Seoul[0.042]; Indonesian[0.042]; following[0.041]; gold[0.041]; gold[0.041]; gold[0.041]; spot[0.041]; Korean[0.041]; Korean[0.041]; week[0.041]; week[0.041]; week[0.041]; coming[0.040]; market[0.040]; market[0.040]; market[0.040]; trade[0.040]; South[0.040]; South[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hong Kong ==> ENTITY: \u001b[32mHong Kong\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.095:0.095[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nselling[0.044]; following[0.043]; week[0.043]; Seoul[0.043]; start[0.043]; mid[0.042]; months[0.042]; spot[0.042]; Korean[0.042]; Koreans[0.042]; finish[0.041]; October[0.041]; October[0.041]; International[0.041]; ahead[0.041]; countries[0.041]; versus[0.041]; coming[0.041]; sell[0.041]; sell[0.041]; volume[0.041]; price[0.041]; revival[0.041]; South[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.069:0.069[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEast[0.043]; selling[0.043]; spot[0.043]; market[0.042]; market[0.042]; week[0.042]; week[0.042]; week[0.042]; fourth[0.042]; sell[0.042]; South[0.042]; South[0.042]; trade[0.042]; Friday[0.042]; marked[0.041]; previous[0.041]; following[0.041]; volume[0.041]; recent[0.041]; thumped[0.041]; start[0.041]; sales[0.041]; Asian[0.040]; months[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.092:0.092[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclosed[0.044]; International[0.043]; helped[0.043]; recent[0.043]; week[0.042]; months[0.042]; October[0.042]; October[0.042]; following[0.042]; Thursday[0.042]; mid[0.042]; Monetary[0.042]; trade[0.042]; said[0.041]; said[0.041]; said[0.041]; South[0.041]; Continued[0.041]; selling[0.041]; ahead[0.041]; demand[0.041]; Friday[0.040]; sell[0.040]; start[0.040]; \t\n\r [===================>..........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s853ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2166/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1159testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brussels ==> ENTITY: \u001b[32mBrussels\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.113:0.113[0]; log p(e|m)= -0.025:-0.025[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.052]; Belgium[0.052]; Belgium[0.052]; second[0.046]; Belgian[0.046]; town[0.045]; Tuesday[0.044]; week[0.044]; southern[0.043]; southern[0.043]; car[0.042]; said[0.042]; police[0.042]; Chatelet[0.042]; Reuters[0.041]; told[0.041]; Saturday[0.041]; window[0.041]; automatic[0.040]; automatic[0.040]; automatic[0.040]; range[0.040]; night[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgium ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.242:-0.242[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.059]; Belgium[0.059]; Belgian[0.051]; Brussels[0.049]; week[0.047]; second[0.046]; said[0.046]; southern[0.045]; southern[0.045]; Tuesday[0.045]; Reuters[0.044]; Saturday[0.044]; told[0.044]; weapons[0.043]; weapons[0.043]; weapons[0.043]; range[0.042]; town[0.042]; car[0.041]; window[0.041]; stolen[0.041]; stolen[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kalashnikov ==> ENTITY: \u001b[32mAK-47\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.162:0.162[0]; log p(e|m)= -0.131:-0.131[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKalashnikov[0.048]; rifles[0.048]; weapons[0.044]; weapons[0.044]; weapons[0.044]; Uzi[0.044]; policeman[0.044]; automatic[0.043]; automatic[0.043]; automatic[0.043]; shooting[0.043]; range[0.041]; car[0.040]; arms[0.040]; arms[0.040]; arms[0.040]; stolen[0.039]; stolen[0.039]; stole[0.039]; southern[0.039]; southern[0.039]; police[0.039]; Tuesday[0.039]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgium ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.242:-0.242[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.059]; Belgium[0.059]; Belgian[0.051]; Brussels[0.049]; week[0.047]; second[0.046]; said[0.046]; southern[0.045]; southern[0.045]; Tuesday[0.045]; Reuters[0.044]; Saturday[0.044]; told[0.044]; weapons[0.043]; weapons[0.043]; weapons[0.043]; range[0.042]; town[0.042]; car[0.041]; window[0.041]; stolen[0.041]; stolen[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Uzi\u001b[39m ==> ENTITY: \u001b[32mUzi\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrifles[0.045]; weapons[0.044]; weapons[0.044]; weapons[0.044]; policeman[0.043]; arms[0.042]; arms[0.042]; arms[0.042]; range[0.042]; second[0.042]; automatic[0.042]; automatic[0.042]; automatic[0.042]; Kalashnikov[0.041]; Kalashnikov[0.041]; told[0.041]; shooting[0.041]; Saturday[0.041]; Reuters[0.041]; week[0.040]; Tuesday[0.040]; car[0.040]; stolen[0.039]; stolen[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kalashnikov ==> ENTITY: \u001b[32mAK-47\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.162:0.162[0]; log p(e|m)= -0.131:-0.131[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKalashnikov[0.048]; rifles[0.048]; weapons[0.044]; weapons[0.044]; weapons[0.044]; Uzi[0.044]; policeman[0.044]; automatic[0.043]; automatic[0.043]; automatic[0.043]; shooting[0.043]; range[0.041]; car[0.040]; arms[0.040]; arms[0.040]; arms[0.040]; stolen[0.039]; stolen[0.039]; stole[0.039]; southern[0.039]; southern[0.039]; police[0.039]; Tuesday[0.039]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.024:0.024[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntold[0.046]; southern[0.045]; southern[0.045]; week[0.044]; Chatelet[0.043]; said[0.043]; Tuesday[0.042]; automatic[0.042]; automatic[0.042]; automatic[0.042]; second[0.042]; range[0.041]; weapons[0.041]; weapons[0.041]; weapons[0.041]; Belgian[0.040]; Belgium[0.040]; Belgium[0.040]; Belgium[0.040]; Brussels[0.040]; Kalashnikov[0.040]; Kalashnikov[0.040]; store[0.040]; store[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgian ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.124:-0.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.054]; Belgium[0.054]; Belgium[0.054]; southern[0.043]; southern[0.043]; said[0.042]; Brussels[0.041]; town[0.041]; week[0.041]; weapons[0.040]; weapons[0.040]; weapons[0.040]; Tuesday[0.040]; told[0.040]; Fal[0.040]; Reuters[0.039]; car[0.039]; arms[0.039]; arms[0.039]; arms[0.039]; stolen[0.039]; stolen[0.039]; range[0.039]; second[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgium ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.242:-0.242[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.059]; Belgium[0.059]; Belgian[0.051]; Brussels[0.049]; week[0.047]; second[0.046]; said[0.046]; southern[0.045]; southern[0.045]; Tuesday[0.045]; Reuters[0.044]; Saturday[0.044]; told[0.044]; weapons[0.043]; weapons[0.043]; weapons[0.043]; range[0.042]; town[0.042]; car[0.041]; window[0.041]; stolen[0.041]; stolen[0.041]; \t\n\r [===================>..........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s818ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2175/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1158testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brussels ==> ENTITY: \u001b[32mBrussels\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.051:0.051[0]; log p(e|m)= -0.025:-0.025[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSaturday[0.046]; Saturday[0.046]; official[0.043]; Treaty[0.043]; Organisation[0.043]; northern[0.042]; northern[0.042]; city[0.042]; Kurd[0.042]; said[0.041]; declined[0.041]; North[0.041]; Arbil[0.041]; Earlier[0.041]; reports[0.041]; Nato[0.040]; Nato[0.040]; Iraq[0.040]; Iraq[0.040]; Iraq[0.040]; Reuters[0.040]; fighting[0.040]; spokesman[0.040]; Kurdistan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.186:0.186[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.053]; Iraq[0.053]; Iraqi[0.049]; Iraqi[0.049]; Kurdistan[0.048]; troops[0.046]; forces[0.046]; northern[0.044]; northern[0.044]; Arbil[0.043]; Kurd[0.043]; said[0.042]; official[0.042]; Organisation[0.041]; conflict[0.040]; Reuters[0.040]; declined[0.040]; reports[0.040]; fighting[0.040]; armed[0.039]; Nato[0.039]; Nato[0.039]; North[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.184:0.184[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.051]; Iraq[0.051]; Iraqi[0.047]; Iraqi[0.047]; Kurdistan[0.046]; troops[0.045]; forces[0.044]; northern[0.043]; northern[0.043]; Arbil[0.042]; Kurd[0.041]; said[0.040]; official[0.040]; Organisation[0.039]; conflict[0.039]; Reuters[0.039]; reports[0.038]; fighting[0.038]; Nato[0.038]; Nato[0.038]; North[0.038]; declined[0.038]; Saturday[0.037]; Saturday[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kurdistan Democratic Party\u001b[39m ==> ENTITY: \u001b[32mKurdistan Democratic Party\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArbil[0.045]; Iraqi[0.045]; Iraqi[0.045]; Kurd[0.045]; Organisation[0.044]; Iraq[0.044]; Iraq[0.044]; Iraq[0.044]; forces[0.044]; leader[0.043]; official[0.042]; troops[0.042]; northern[0.041]; northern[0.041]; said[0.040]; armed[0.039]; Nato[0.039]; Nato[0.039]; Earlier[0.039]; spokesman[0.039]; conflict[0.039]; Reuters[0.038]; told[0.038]; Brussels[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.193:0.193[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.050]; Iraq[0.050]; Iraq[0.050]; Kurdistan[0.048]; Iraqi[0.047]; northern[0.045]; northern[0.045]; Kurd[0.044]; troops[0.044]; forces[0.043]; Arbil[0.042]; said[0.040]; reports[0.040]; official[0.039]; conflict[0.038]; declined[0.038]; leader[0.037]; city[0.037]; fighting[0.037]; Reuters[0.037]; situation[0.037]; armed[0.037]; North[0.037]; Organisation[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = KDP\u001b[39m ==> ENTITY: \u001b[32mKurdistan Democratic Party\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.049]; Arbil[0.045]; Iraqi[0.044]; Iraqi[0.044]; Kurd[0.044]; Iraq[0.044]; Iraq[0.044]; Iraq[0.044]; forces[0.044]; leader[0.043]; Party[0.042]; troops[0.042]; northern[0.041]; northern[0.041]; official[0.040]; Democratic[0.040]; fighting[0.039]; armed[0.039]; said[0.039]; Nato[0.039]; Nato[0.039]; Earlier[0.039]; spokesman[0.039]; conflict[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = North Atlantic Treaty Organisation\u001b[39m ==> ENTITY: \u001b[32mNATO\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.098:0.098[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNato[0.047]; Nato[0.047]; Iraq[0.044]; Iraq[0.044]; Iraq[0.044]; troops[0.044]; forces[0.044]; official[0.042]; spokesman[0.042]; Arbil[0.041]; Brussels[0.041]; conflict[0.041]; said[0.040]; armed[0.040]; Iraqi[0.040]; Iraqi[0.040]; fighting[0.040]; reports[0.040]; situation[0.040]; Kurd[0.040]; northern[0.040]; northern[0.040]; Kurdistan[0.039]; Reuters[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.082:0.082[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nreports[0.047]; told[0.044]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; attacking[0.043]; northern[0.043]; northern[0.043]; official[0.043]; troops[0.042]; said[0.042]; Iraqi[0.041]; Iraqi[0.041]; forces[0.041]; fighting[0.041]; spokesman[0.041]; Democratic[0.041]; closely[0.040]; conflict[0.040]; city[0.040]; situation[0.040]; Atlantic[0.040]; leader[0.040]; Earlier[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NATO\u001b[39m ==> ENTITY: \u001b[32mNATO\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.120:0.120[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNato[0.046]; troops[0.045]; Iraq[0.044]; Iraq[0.044]; Iraq[0.044]; forces[0.044]; Iraqi[0.042]; Iraqi[0.042]; said[0.042]; Brussels[0.042]; official[0.041]; Treaty[0.041]; conflict[0.041]; reports[0.041]; Atlantic[0.041]; spokesman[0.041]; fighting[0.040]; northern[0.040]; northern[0.040]; told[0.040]; declined[0.040]; Organisation[0.040]; North[0.039]; Earlier[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arbil\u001b[39m ==> ENTITY: \u001b[32mErbil\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.049]; Iraq[0.046]; Iraq[0.046]; Iraq[0.046]; Kurd[0.045]; Iraqi[0.045]; Iraqi[0.045]; troops[0.045]; northern[0.044]; northern[0.044]; city[0.042]; forces[0.042]; Saturday[0.040]; Saturday[0.040]; official[0.040]; leader[0.039]; Democratic[0.038]; said[0.038]; Nato[0.038]; Nato[0.038]; conflict[0.037]; fighting[0.037]; Brussels[0.037]; Party[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.184:0.184[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.051]; Iraq[0.051]; Iraqi[0.047]; Iraqi[0.047]; Kurdistan[0.046]; troops[0.045]; forces[0.045]; northern[0.043]; northern[0.043]; Arbil[0.042]; Kurd[0.042]; said[0.040]; official[0.040]; Organisation[0.039]; conflict[0.039]; Reuters[0.039]; reports[0.038]; fighting[0.038]; declined[0.038]; Nato[0.038]; Nato[0.038]; North[0.037]; Saturday[0.037]; Saturday[0.037]; \t\n\r [===================>..........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s811ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2186/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 956testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Seoul ==> ENTITY: \u001b[32mSeoul\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.176:0.176[0]; log p(e|m)= -0.038:-0.038[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChonbuk[0.050]; Games[0.047]; Soccer[0.045]; Soccer[0.045]; won[0.044]; Pohang[0.044]; Pohang[0.044]; soccer[0.044]; Korean[0.044]; Korean[0.044]; games[0.044]; games[0.044]; played[0.043]; played[0.043]; Puchon[0.042]; Puchon[0.042]; Ulsan[0.042]; Ulsan[0.042]; South[0.042]; South[0.042]; Anyang[0.041]; Thursday[0.041]; Thursday[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pohang ==> ENTITY: \u001b[32mPohang Steelers\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.250:0.250[0]; log p(e|m)= -1.058:-1.058[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPohang[0.055]; Chonbuk[0.048]; Chonbuk[0.048]; Ulsan[0.044]; Ulsan[0.044]; Seoul[0.043]; Soccer[0.042]; Soccer[0.042]; soccer[0.041]; Pusan[0.041]; Korean[0.040]; Korean[0.040]; Chonan[0.040]; Anyang[0.040]; Thursday[0.040]; Thursday[0.040]; goals[0.040]; goals[0.040]; games[0.040]; games[0.040]; won[0.039]; Puchon[0.039]; Puchon[0.039]; Standings[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Anyang ==> ENTITY: \u001b[32mFC Seoul\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.243:0.243[0]; log p(e|m)= -1.382:-1.382[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChonbuk[0.050]; Chonbuk[0.050]; soccer[0.046]; Seoul[0.046]; Soccer[0.046]; Soccer[0.046]; won[0.045]; Pohang[0.045]; Pohang[0.045]; games[0.044]; games[0.044]; played[0.043]; played[0.043]; Ulsan[0.042]; Ulsan[0.042]; Korean[0.042]; Korean[0.042]; goals[0.041]; goals[0.041]; Pusan[0.040]; Puchon[0.040]; Puchon[0.040]; Games[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ulsan ==> ENTITY: \u001b[32mUlsan Hyundai FC\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.220:0.220[0]; log p(e|m)= -1.732:-1.732[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChonbuk[0.047]; Chonbuk[0.047]; Ulsan[0.047]; Seoul[0.045]; Pohang[0.044]; Pohang[0.044]; Pusan[0.042]; Korean[0.042]; Korean[0.042]; soccer[0.042]; won[0.041]; Anyang[0.041]; games[0.041]; games[0.041]; Soccer[0.041]; Soccer[0.041]; Games[0.040]; Puchon[0.039]; Puchon[0.039]; Chonan[0.039]; South[0.039]; South[0.039]; goals[0.039]; goals[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = South Korean\u001b[39m ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChonbuk[0.047]; Korean[0.045]; Soccer[0.045]; Soccer[0.045]; soccer[0.044]; games[0.044]; games[0.044]; Seoul[0.043]; Pohang[0.043]; Pohang[0.043]; Ulsan[0.043]; Ulsan[0.043]; Games[0.042]; Pusan[0.041]; played[0.040]; played[0.040]; won[0.040]; South[0.040]; Puchon[0.039]; Puchon[0.039]; Anyang[0.038]; goals[0.038]; goals[0.038]; Standings[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pohang ==> ENTITY: \u001b[32mPohang Steelers\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.250:0.250[0]; log p(e|m)= -1.058:-1.058[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPohang[0.055]; Chonbuk[0.048]; Chonbuk[0.048]; Ulsan[0.044]; Ulsan[0.044]; Seoul[0.043]; Soccer[0.042]; Soccer[0.042]; soccer[0.041]; Pusan[0.041]; Korean[0.040]; Korean[0.040]; Chonan[0.040]; Anyang[0.040]; Thursday[0.040]; Thursday[0.040]; goals[0.040]; goals[0.040]; games[0.040]; games[0.040]; won[0.039]; Puchon[0.039]; Puchon[0.039]; Standings[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ulsan ==> ENTITY: \u001b[32mUlsan Hyundai FC\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.220:0.220[0]; log p(e|m)= -1.732:-1.732[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChonbuk[0.047]; Chonbuk[0.047]; Ulsan[0.047]; Seoul[0.045]; Pohang[0.044]; Pohang[0.044]; Pusan[0.042]; Korean[0.042]; Korean[0.042]; soccer[0.042]; won[0.041]; Anyang[0.041]; games[0.041]; games[0.041]; Soccer[0.041]; Soccer[0.041]; Games[0.040]; Puchon[0.039]; Puchon[0.039]; Chonan[0.039]; South[0.039]; South[0.039]; goals[0.039]; goals[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = South Korean\u001b[39m ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChonbuk[0.047]; Korean[0.046]; Soccer[0.045]; Soccer[0.045]; soccer[0.044]; games[0.044]; games[0.044]; Seoul[0.044]; Pohang[0.043]; Pohang[0.043]; Ulsan[0.043]; Ulsan[0.043]; Games[0.042]; played[0.041]; played[0.041]; won[0.040]; South[0.040]; Puchon[0.040]; Puchon[0.040]; Anyang[0.039]; goals[0.038]; goals[0.038]; Standings[0.038]; points[0.037]; \t\n\r [====================>.........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s783ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2194/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1039testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Auckland ==> ENTITY: \u001b[32mAuckland\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.297:-0.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAuckland[0.050]; Zealand[0.044]; mid[0.044]; South[0.044]; Donald[0.043]; coast[0.043]; Friday[0.042]; transport[0.042]; said[0.041]; said[0.041]; success[0.041]; affairs[0.041]; running[0.041]; infrastructure[0.041]; infrastructure[0.041]; infrastructure[0.041]; place[0.041]; place[0.041]; territory[0.040]; world[0.040]; China[0.040]; visit[0.040]; largest[0.040]; future[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Zealand ==> ENTITY: \u001b[32mNew Zealand\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.107:0.107[0]; log p(e|m)= -0.238:-0.238[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAuckland[0.045]; Auckland[0.045]; Britain[0.045]; world[0.043]; South[0.042]; China[0.042]; China[0.042]; mid[0.042]; Donald[0.042]; telecommunications[0.042]; acknowledged[0.041]; largest[0.041]; place[0.041]; place[0.041]; territory[0.041]; visit[0.041]; practice[0.041]; speech[0.041]; success[0.041]; issuing[0.040]; managing[0.040]; Friday[0.040]; means[0.040]; home[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wellington ==> ENTITY: \u001b[32mWellington\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.050:0.050[0]; log p(e|m)= -0.926:-0.926[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.047]; home[0.046]; months[0.042]; infrastructure[0.042]; acknowledged[0.042]; place[0.042]; said[0.042]; financial[0.041]; reserves[0.041]; great[0.041]; practice[0.041]; enormous[0.041]; future[0.041]; future[0.041]; future[0.041]; running[0.041]; people[0.041]; raising[0.041]; constitutional[0.041]; expenditure[0.041]; managing[0.041]; handover[0.041]; seek[0.040]; currency[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hong Kong ==> ENTITY: \u001b[32mHong Kong\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.045]; Kong[0.044]; Kong[0.044]; Kong[0.044]; Kong[0.044]; Britain[0.043]; Hong[0.043]; Hong[0.043]; Hong[0.043]; Hong[0.043]; currency[0.041]; place[0.041]; Tsang[0.041]; Tsang[0.041]; home[0.041]; months[0.041]; majority[0.040]; world[0.040]; constitutional[0.040]; Auckland[0.040]; future[0.039]; future[0.039]; future[0.039]; managing[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hong Kong ==> ENTITY: \u001b[32mHong Kong\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.168:0.168[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChinese[0.044]; China[0.044]; China[0.044]; port[0.043]; Kong[0.043]; Kong[0.043]; Kong[0.043]; Kong[0.043]; Kong[0.043]; Britain[0.042]; Hong[0.042]; Hong[0.042]; Hong[0.042]; Hong[0.042]; Hong[0.042]; transport[0.041]; currency[0.040]; place[0.040]; place[0.040]; home[0.040]; mid[0.040]; months[0.040]; coast[0.039]; world[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tsang\u001b[39m ==> ENTITY: \u001b[32mDonald Tsang\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSecretary[0.048]; Tsang[0.048]; Tsang[0.048]; Tsang[0.048]; speech[0.047]; China[0.046]; China[0.046]; Hk[0.045]; visit[0.045]; said[0.045]; said[0.045]; said[0.045]; handover[0.045]; months[0.045]; acknowledged[0.045]; post[0.044]; financial[0.044]; Chinese[0.044]; Kong[0.044]; Kong[0.044]; Kong[0.044]; Kong[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hong Kong ==> ENTITY: \u001b[32mHong Kong\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.118:0.118[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.045]; Kong[0.044]; Kong[0.044]; Kong[0.044]; Kong[0.044]; Britain[0.043]; Hong[0.043]; Hong[0.043]; Hong[0.043]; Hong[0.043]; currency[0.042]; place[0.041]; Tsang[0.041]; home[0.041]; months[0.041]; majority[0.040]; constitutional[0.040]; Auckland[0.040]; future[0.040]; future[0.040]; future[0.040]; managing[0.040]; issuing[0.039]; raising[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.048:0.048[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nterritory[0.053]; China[0.052]; China[0.052]; sovereignty[0.051]; Chinese[0.051]; speech[0.050]; world[0.050]; Secretary[0.050]; months[0.050]; autonomy[0.050]; South[0.050]; mid[0.049]; Kong[0.049]; Kong[0.049]; Kong[0.049]; Kong[0.049]; Kong[0.049]; Kong[0.049]; Zealand[0.049]; people[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tsang\u001b[39m ==> ENTITY: \u001b[32mDonald Tsang\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.121:0.121[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSecretary[0.045]; Tsang[0.045]; Tsang[0.045]; speech[0.044]; China[0.044]; Hk[0.043]; visit[0.043]; said[0.043]; said[0.043]; post[0.042]; Chinese[0.042]; Kong[0.041]; Kong[0.041]; Hong[0.041]; Hong[0.041]; affairs[0.041]; Financial[0.040]; Friday[0.040]; Auckland[0.040]; Auckland[0.040]; running[0.039]; success[0.039]; Zealand[0.039]; transport[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Donald Tsang\u001b[39m ==> ENTITY: \u001b[32mDonald Tsang\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSecretary[0.045]; Tsang[0.044]; Tsang[0.044]; speech[0.044]; China[0.043]; China[0.043]; Hk[0.042]; visit[0.042]; said[0.042]; said[0.042]; handover[0.042]; post[0.041]; Chinese[0.041]; Kong[0.041]; Kong[0.041]; Hong[0.041]; Hong[0.041]; affairs[0.041]; Financial[0.040]; Britain[0.040]; Friday[0.040]; Auckland[0.039]; Auckland[0.039]; running[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hong Kong ==> ENTITY: \u001b[32mHong Kong\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.045]; Kong[0.044]; Kong[0.044]; Kong[0.044]; Kong[0.044]; Britain[0.043]; Hong[0.043]; Hong[0.043]; Hong[0.043]; Hong[0.043]; currency[0.042]; place[0.041]; Tsang[0.041]; home[0.041]; months[0.041]; majority[0.040]; constitutional[0.040]; future[0.040]; future[0.040]; future[0.040]; managing[0.040]; issuing[0.039]; raising[0.039]; Wellington[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.049]; Chinese[0.047]; currency[0.043]; territory[0.043]; Britain[0.042]; world[0.042]; Financial[0.042]; months[0.041]; issuing[0.041]; Hong[0.041]; Hong[0.041]; Hong[0.041]; Hong[0.041]; Hong[0.041]; financial[0.041]; mid[0.041]; affairs[0.041]; Kong[0.040]; Kong[0.040]; Kong[0.040]; Kong[0.040]; Kong[0.040]; acknowledged[0.040]; speech[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Auckland ==> ENTITY: \u001b[32mAuckland\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.104:0.104[0]; log p(e|m)= -0.297:-0.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAuckland[0.049]; Britain[0.044]; Zealand[0.044]; mid[0.043]; Donald[0.043]; South[0.043]; expenditure[0.042]; financial[0.042]; Friday[0.042]; months[0.042]; visit[0.042]; transport[0.041]; practice[0.041]; success[0.041]; affairs[0.041]; seek[0.040]; running[0.040]; infrastructure[0.040]; infrastructure[0.040]; infrastructure[0.040]; place[0.040]; place[0.040]; future[0.040]; future[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hong Kong ==> ENTITY: \u001b[32mHong Kong\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChinese[0.046]; China[0.046]; port[0.045]; Kong[0.045]; Britain[0.044]; Hong[0.043]; transport[0.043]; place[0.042]; place[0.042]; mid[0.041]; coast[0.041]; world[0.041]; infrastructure[0.040]; infrastructure[0.040]; infrastructure[0.040]; Auckland[0.040]; Auckland[0.040]; Tsang[0.040]; Tsang[0.040]; Tsang[0.040]; territory[0.040]; running[0.040]; South[0.040]; necessary[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hong Kong ==> ENTITY: \u001b[32mHong Kong\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.159:0.159[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChinese[0.047]; China[0.046]; China[0.046]; Kong[0.045]; Kong[0.045]; Kong[0.045]; Kong[0.045]; Britain[0.044]; Hong[0.044]; Hong[0.044]; Hong[0.044]; Hong[0.044]; currency[0.043]; place[0.042]; Tsang[0.042]; Tsang[0.042]; home[0.042]; mid[0.042]; months[0.042]; majority[0.041]; world[0.041]; constitutional[0.041]; Auckland[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chinese ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.142:0.142[0]; log p(e|m)= -1.419:-1.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.050]; China[0.050]; world[0.044]; territory[0.043]; Britain[0.043]; Hong[0.042]; Hong[0.042]; Hong[0.042]; Kong[0.041]; Kong[0.041]; Kong[0.041]; speech[0.041]; said[0.041]; said[0.041]; South[0.041]; means[0.040]; sovereignty[0.040]; mid[0.040]; Tsang[0.040]; Tsang[0.040]; Tsang[0.040]; Donald[0.040]; largest[0.040]; autonomy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tsang\u001b[39m ==> ENTITY: \u001b[32mDonald Tsang\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTsang[0.044]; speech[0.044]; China[0.043]; China[0.043]; visit[0.042]; said[0.042]; said[0.042]; handover[0.042]; months[0.042]; acknowledged[0.042]; financial[0.041]; Kong[0.041]; Kong[0.041]; Kong[0.041]; Kong[0.041]; Kong[0.041]; Hong[0.041]; Hong[0.041]; Hong[0.041]; Hong[0.041]; Hong[0.041]; affairs[0.041]; priorities[0.041]; Wellington[0.040]; \t\n\r [====================>.........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s699ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2211/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1112testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mThe Curragh\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mCurragh\u001b[39m <---> \u001b[32mCurragh Racecourse\u001b[39m\t\nSCORES: global= 0.281:0.276[\u001b[31m0.005\u001b[39m]; local()= 0.155:0.220[\u001b[32m0.065\u001b[39m]; log p(e|m)= 0.000:-0.717[\u001b[31m0.717\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nridden[0.047]; Stakes[0.046]; Stakes[0.046]; Paddy[0.046]; Tattersalls[0.045]; Tattersalls[0.045]; race[0.045]; run[0.044]; Horse[0.044]; Dublin[0.044]; furlongs[0.044]; Saturday[0.044]; Lad[0.043]; Racing[0.043]; Hannon[0.043]; trained[0.042]; olds[0.042]; Britain[0.042]; lengths[0.042]; lengths[0.042]; winning[0.041]; owned[0.041]; John[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.068:0.068[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrace[0.048]; David[0.046]; winning[0.046]; Richard[0.046]; Racing[0.045]; Peter[0.045]; John[0.045]; run[0.044]; Harrison[0.043]; Dublin[0.042]; Winner[0.040]; Warren[0.040]; Miss[0.040]; ridden[0.040]; Horse[0.040]; Paddy[0.040]; half[0.039]; Saturday[0.039]; lengths[0.039]; lengths[0.039]; joint[0.039]; Result[0.038]; Result[0.038]; trained[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Richard Hannon\u001b[39m ==> ENTITY: \u001b[32mRichard Hannon\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHorse[0.046]; Tattersalls[0.044]; Tattersalls[0.044]; John[0.043]; Lad[0.043]; David[0.042]; Stakes[0.042]; Stakes[0.042]; Breeders[0.042]; Breeders[0.042]; Curragh[0.042]; Britain[0.042]; furlongs[0.041]; lengths[0.041]; lengths[0.041]; Warren[0.041]; trained[0.041]; Winner[0.041]; olds[0.041]; Miss[0.040]; Paddy[0.040]; ridden[0.040]; Peter[0.040]; Racing[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dublin ==> ENTITY: \u001b[32mDublin\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPaddy[0.050]; Connor[0.046]; John[0.045]; Britain[0.044]; Saturday[0.044]; Hannon[0.044]; Warren[0.043]; Harrison[0.043]; winning[0.042]; Peter[0.042]; David[0.042]; Richard[0.041]; Stakes[0.040]; Stakes[0.040]; owned[0.040]; Bloomfield[0.040]; half[0.040]; Curragh[0.040]; Tattersalls[0.040]; Tattersalls[0.040]; ridden[0.039]; Pelham[0.039]; run[0.039]; Miss[0.039]; \t\n\r [====================>.........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s725ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2215/4791 \r\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = KDP\u001b[39m ==> ENTITY: \u001b[32mKurdistan Democratic Party\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBarzani[0.045]; Barzani[0.045]; Arbil[0.045]; Iraqi[0.045]; Iraqi[0.045]; Hussein[0.044]; Talabani[0.044]; Talabani[0.044]; forces[0.044]; Saddam[0.042]; Saddam[0.042]; troops[0.042]; President[0.041]; militarily[0.040]; August[0.040]; Iran[0.040]; foreign[0.039]; armed[0.039]; said[0.039]; said[0.039]; said[0.039]; tanks[0.039]; attacks[0.038]; controlled[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saddam Hussein\u001b[39m ==> ENTITY: \u001b[32mSaddam Hussein\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.047]; Saddam[0.046]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; military[0.042]; military[0.042]; Kurdistan[0.041]; Baghdad[0.041]; Iran[0.041]; Kurds[0.041]; attacks[0.041]; Talabani[0.041]; Talabani[0.041]; Talabani[0.041]; Aziz[0.040]; Aziz[0.040]; Arbil[0.040]; forces[0.040]; forces[0.040]; aid[0.040]; troops[0.040]; capability[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.098:0.098[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.046]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Iran[0.044]; Saddam[0.042]; Saddam[0.042]; response[0.042]; Baghdad[0.042]; military[0.042]; President[0.042]; officials[0.042]; order[0.041]; August[0.041]; sent[0.041]; message[0.041]; foreign[0.040]; Hussein[0.040]; said[0.040]; said[0.040]; said[0.040]; forces[0.040]; troops[0.040]; Arbil[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.223:0.223[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.052]; Iraq[0.052]; Iraqi[0.047]; Iraqi[0.047]; Kurdistan[0.046]; Baghdad[0.045]; Baghdad[0.045]; troops[0.045]; forces[0.045]; Kurdish[0.044]; military[0.044]; northern[0.043]; support[0.042]; Talabani[0.041]; said[0.040]; said[0.040]; official[0.040]; Aziz[0.040]; Aziz[0.040]; Aziz[0.040]; Aziz[0.040]; Barzani[0.040]; Barzani[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iraqi Kurds\u001b[39m ==> ENTITY: \u001b[32mIraqi Kurdistan\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.240:0.240[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.047]; Kurdish[0.047]; northern[0.044]; Iraq[0.044]; Iraq[0.044]; Hussein[0.043]; Iraqi[0.042]; Iraqi[0.042]; Iraqi[0.042]; Arbil[0.041]; Talabani[0.041]; Talabani[0.041]; Talabani[0.041]; Barzani[0.041]; Barzani[0.041]; Barzani[0.041]; Barzani[0.041]; troops[0.041]; troops[0.041]; Baghdad[0.041]; forces[0.040]; forces[0.040]; Saddam[0.040]; Saddam[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saddam\u001b[39m ==> ENTITY: \u001b[32mSaddam Hussein\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.050]; Saddam[0.050]; Iraqi[0.048]; Iraqi[0.048]; Iraqi[0.048]; Iraqi[0.048]; Hussein[0.047]; military[0.045]; Kurdistan[0.045]; Baghdad[0.045]; Iran[0.045]; Kurds[0.044]; attacks[0.044]; Talabani[0.044]; Talabani[0.044]; Talabani[0.044]; Aziz[0.044]; Aziz[0.044]; Arbil[0.043]; forces[0.043]; troops[0.043]; capability[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Massoud Barzani\u001b[39m ==> ENTITY: \u001b[32mMasoud Barzani\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.242:0.242[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.046]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; Barzani[0.044]; Barzani[0.044]; Barzani[0.044]; Kurdish[0.044]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Talabani[0.044]; Talabani[0.044]; Hussein[0.043]; Kurds[0.043]; President[0.043]; Baghdad[0.042]; Baghdad[0.042]; Baghdad[0.042]; troops[0.042]; leader[0.042]; Massoud[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.240:0.240[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.050]; Iraq[0.050]; Iraq[0.050]; Iraq[0.050]; Kurdistan[0.047]; Iraqi[0.046]; Kurdish[0.046]; northern[0.044]; Baghdad[0.043]; Baghdad[0.043]; troops[0.043]; military[0.042]; military[0.042]; Talabani[0.042]; Barzani[0.041]; Barzani[0.041]; forces[0.041]; aid[0.040]; aid[0.040]; Aziz[0.040]; Aziz[0.040]; Aziz[0.040]; Aziz[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iran ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.138:-0.138[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.046]; military[0.045]; military[0.045]; Kurdistan[0.044]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Kurds[0.042]; official[0.042]; troops[0.042]; Agency[0.041]; Baghdad[0.041]; forces[0.041]; statement[0.041]; Saddam[0.040]; Saddam[0.040]; attacks[0.040]; Hussein[0.040]; threat[0.040]; Massoud[0.040]; support[0.040]; officials[0.040]; allies[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baghdad\u001b[39m ==> ENTITY: \u001b[32mBaghdad\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.049]; Iraq[0.049]; Iraq[0.049]; Iraqi[0.045]; Iraqi[0.045]; Baghdad[0.045]; northern[0.045]; Kurdistan[0.044]; troops[0.043]; Kurdish[0.042]; military[0.042]; Talabani[0.040]; Aziz[0.039]; Aziz[0.039]; Aziz[0.039]; forces[0.039]; Barzani[0.039]; Barzani[0.039]; Jalal[0.039]; Massoud[0.039]; Massoud[0.039]; official[0.038]; aid[0.037]; aid[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arbil\u001b[39m ==> ENTITY: \u001b[32mErbil\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurds[0.050]; Iraqi[0.047]; Iraqi[0.047]; Iraqi[0.047]; troops[0.046]; Baghdad[0.046]; Barzani[0.045]; Barzani[0.045]; Talabani[0.044]; Talabani[0.044]; Hussein[0.044]; people[0.043]; forces[0.043]; Saddam[0.043]; Saddam[0.043]; areas[0.042]; officials[0.042]; Iran[0.041]; sent[0.040]; foreign[0.040]; President[0.040]; August[0.040]; militarily[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baghdad\u001b[39m ==> ENTITY: \u001b[32mBaghdad\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.227:0.227[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.048]; Iraq[0.048]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Kurdistan[0.043]; troops[0.042]; Kurdish[0.041]; military[0.041]; military[0.041]; Arbil[0.041]; forces[0.041]; forces[0.041]; Hussein[0.041]; Kurds[0.040]; northern[0.040]; Saddam[0.040]; Saddam[0.040]; official[0.040]; Talabani[0.040]; Talabani[0.040]; Talabani[0.040]; Iran[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jalal Talabani\u001b[39m ==> ENTITY: \u001b[32mJalal Talabani\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.045]; Talabani[0.045]; Kurdish[0.045]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Iraq[0.044]; Iraq[0.044]; Iraq[0.044]; Iraq[0.044]; Kurds[0.043]; Barzani[0.043]; Barzani[0.043]; Barzani[0.043]; Barzani[0.043]; Hussein[0.042]; Aziz[0.042]; Aziz[0.042]; Aziz[0.042]; Aziz[0.042]; Saddam[0.042]; Saddam[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.237:0.237[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.048]; Iraq[0.048]; Iraq[0.048]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Kurdistan[0.043]; Baghdad[0.042]; Baghdad[0.042]; Baghdad[0.042]; troops[0.042]; forces[0.042]; Kurdish[0.041]; military[0.041]; military[0.041]; Kurds[0.040]; northern[0.040]; support[0.039]; Talabani[0.039]; allies[0.038]; said[0.038]; said[0.038]; said[0.038]; official[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.218:0.218[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.049]; Iraqi[0.046]; Iraqi[0.046]; Iraqi[0.046]; Kurdistan[0.044]; Kurds[0.043]; Baghdad[0.043]; military[0.042]; troops[0.042]; Hussein[0.041]; Talabani[0.041]; Talabani[0.041]; Talabani[0.041]; forces[0.041]; Saddam[0.040]; Saddam[0.040]; capability[0.040]; Iran[0.040]; Arbil[0.040]; people[0.039]; tanks[0.039]; Barzani[0.039]; Barzani[0.039]; foreign[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.237:0.237[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.048]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Saddam[0.044]; Saddam[0.044]; Kurdistan[0.043]; Baghdad[0.042]; troops[0.042]; forces[0.042]; forces[0.042]; Hussein[0.041]; Kurdish[0.041]; military[0.041]; military[0.041]; Kurds[0.040]; northern[0.040]; support[0.039]; Iran[0.039]; foreign[0.039]; Talabani[0.039]; Talabani[0.039]; Talabani[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Talabani\u001b[39m ==> ENTITY: \u001b[32mJalal Talabani\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.044]; Talabani[0.043]; Talabani[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraq[0.043]; Kurds[0.042]; Barzani[0.042]; Barzani[0.042]; Barzani[0.042]; Hussein[0.041]; Aziz[0.041]; Aziz[0.041]; Saddam[0.041]; Saddam[0.041]; Jalal[0.041]; Arbil[0.040]; Baghdad[0.040]; forces[0.040]; troops[0.040]; President[0.039]; official[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iraqi News Agency\u001b[39m ==> ENTITY: \u001b[32mNational Iraqi News Agency\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.047]; Iraq[0.047]; Iraq[0.047]; Baghdad[0.046]; Baghdad[0.046]; Baghdad[0.046]; Kurdistan[0.046]; Iraqi[0.046]; Iraqi[0.046]; Iraqi[0.046]; Hussein[0.045]; Saddam[0.043]; Saddam[0.043]; Kurdish[0.042]; official[0.041]; Talabani[0.041]; Talabani[0.041]; Talabani[0.041]; leadership[0.040]; militarily[0.040]; military[0.040]; military[0.040]; rebel[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Massoud Barzani\u001b[39m ==> ENTITY: \u001b[32mMasoud Barzani\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.228:0.228[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.046]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; Barzani[0.044]; Barzani[0.044]; Kurdish[0.044]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Talabani[0.044]; Hussein[0.043]; Kurds[0.043]; President[0.043]; Baghdad[0.042]; Baghdad[0.042]; Baghdad[0.042]; troops[0.042]; leader[0.042]; Massoud[0.041]; military[0.041]; military[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.183:0.183[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraqi[0.047]; Iraqi[0.047]; Kurds[0.044]; Baghdad[0.044]; troops[0.044]; forces[0.043]; Talabani[0.042]; Talabani[0.042]; Hussein[0.042]; Saddam[0.042]; Saddam[0.042]; capability[0.041]; Iran[0.041]; Arbil[0.041]; people[0.041]; tanks[0.040]; Barzani[0.040]; Barzani[0.040]; foreign[0.040]; said[0.040]; said[0.040]; said[0.040]; message[0.039]; officials[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.216:0.216[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.050]; Iraq[0.050]; Iraqi[0.046]; Kurdistan[0.045]; Baghdad[0.044]; Baghdad[0.044]; troops[0.044]; Kurdish[0.043]; military[0.043]; northern[0.042]; forces[0.041]; support[0.041]; Talabani[0.040]; said[0.039]; said[0.039]; Aziz[0.039]; Aziz[0.039]; Aziz[0.039]; Barzani[0.039]; Barzani[0.039]; aid[0.039]; aid[0.039]; Massoud[0.038]; Massoud[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = KDP\u001b[39m ==> ENTITY: \u001b[32mKurdistan Democratic Party\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.047]; Kurdish[0.046]; Barzani[0.043]; Barzani[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; Talabani[0.042]; forces[0.042]; leader[0.042]; Baghdad[0.042]; Baghdad[0.042]; leadership[0.041]; support[0.040]; troops[0.040]; Aziz[0.040]; Aziz[0.040]; Aziz[0.040]; northern[0.039]; Union[0.039]; Jalal[0.039]; military[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Barzani\u001b[39m ==> ENTITY: \u001b[32mMasoud Barzani\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.229:0.229[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.044]; Iraq[0.043]; Barzani[0.042]; Barzani[0.042]; Barzani[0.042]; Kurdish[0.042]; Iraqi[0.042]; Iraqi[0.042]; Iraqi[0.042]; Iraqi[0.042]; Talabani[0.042]; Talabani[0.042]; Talabani[0.042]; Hussein[0.042]; Kurds[0.042]; Arbil[0.041]; President[0.041]; Baghdad[0.041]; troops[0.040]; leader[0.040]; Massoud[0.040]; Massoud[0.040]; military[0.040]; military[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Talabani\u001b[39m ==> ENTITY: \u001b[32mJalal Talabani\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTalabani[0.051]; Iraqi[0.050]; Iraqi[0.050]; Iraqi[0.050]; Iraq[0.050]; Kurds[0.049]; Barzani[0.048]; Barzani[0.048]; Hussein[0.048]; Aziz[0.048]; Saddam[0.048]; Saddam[0.048]; Arbil[0.047]; Baghdad[0.047]; forces[0.046]; troops[0.046]; President[0.046]; Iran[0.045]; allies[0.045]; officials[0.045]; military[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.044:0.044[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.047]; military[0.047]; order[0.046]; Iraq[0.044]; Iraq[0.044]; foreign[0.042]; troops[0.042]; statement[0.042]; August[0.042]; aware[0.041]; northern[0.041]; Union[0.041]; News[0.041]; sent[0.040]; message[0.040]; chief[0.040]; capability[0.040]; officials[0.040]; Agency[0.040]; Iraqi[0.040]; Iraqi[0.040]; Iraqi[0.040]; Iraqi[0.040]; scale[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Patriotic Union of Kurdistan\u001b[39m ==> ENTITY: \u001b[32mPatriotic Union of Kurdistan\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdish[0.044]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; Kurds[0.043]; Iraqi[0.042]; Iraqi[0.042]; Iraqi[0.042]; Iraqi[0.042]; leadership[0.042]; Talabani[0.041]; Talabani[0.041]; forces[0.041]; Barzani[0.041]; Barzani[0.041]; Barzani[0.041]; Barzani[0.041]; Baghdad[0.040]; Baghdad[0.040]; Baghdad[0.040]; leader[0.040]; support[0.040]; official[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Barzani\u001b[39m ==> ENTITY: \u001b[32mMasoud Barzani\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.226:0.226[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.045]; Iraq[0.044]; Barzani[0.043]; Barzani[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Talabani[0.042]; Talabani[0.042]; Talabani[0.042]; Hussein[0.042]; Kurds[0.042]; Arbil[0.042]; President[0.041]; Baghdad[0.041]; troops[0.041]; Massoud[0.040]; military[0.040]; military[0.040]; Aziz[0.040]; Aziz[0.040]; Saddam[0.040]; Saddam[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baghdad\u001b[39m ==> ENTITY: \u001b[32mBaghdad\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.049]; Iraq[0.049]; Iraq[0.049]; Iraqi[0.045]; Iraqi[0.045]; Baghdad[0.045]; northern[0.045]; Kurdistan[0.044]; troops[0.043]; Kurdish[0.042]; military[0.042]; Talabani[0.040]; Aziz[0.039]; Aziz[0.039]; Aziz[0.039]; forces[0.039]; Barzani[0.039]; Barzani[0.039]; Jalal[0.039]; Massoud[0.039]; Massoud[0.039]; official[0.038]; aid[0.037]; aid[0.037]; \t\n\r [====================>.........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s530ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2247/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1126testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nairobi\u001b[39m ==> ENTITY: \u001b[32mNairobi\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.112:0.112[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngroup[0.045]; group[0.045]; group[0.045]; Rwanda[0.044]; Democracy[0.043]; operation[0.042]; Sunday[0.042]; Saturday[0.042]; aims[0.042]; calls[0.041]; camps[0.041]; demands[0.041]; called[0.041]; Rally[0.041]; Rwandan[0.041]; Rwandan[0.041]; High[0.041]; refugee[0.040]; refugee[0.040]; refugee[0.040]; workers[0.040]; statement[0.040]; urged[0.040]; eastern[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zairean\u001b[39m ==> ENTITY: \u001b[32mDemocratic Republic of the Congo\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaire[0.046]; Zaire[0.046]; Rwanda[0.044]; Rwanda[0.044]; Rwanda[0.044]; Tanzania[0.044]; Zairean[0.043]; Zairean[0.043]; Zairean[0.043]; Rwandan[0.042]; Rwandan[0.042]; Rwandan[0.042]; border[0.041]; camps[0.041]; camps[0.041]; refugees[0.040]; refugees[0.040]; refugees[0.040]; refugees[0.040]; refugee[0.040]; troops[0.040]; people[0.039]; governments[0.038]; born[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zairean\u001b[39m ==> ENTITY: \u001b[32mDemocratic Republic of the Congo\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.218:0.218[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaire[0.045]; Zaire[0.045]; Rwanda[0.044]; Rwanda[0.044]; Rwanda[0.044]; Tanzania[0.043]; eastern[0.043]; Zairean[0.043]; Zairean[0.043]; Zairean[0.043]; Rwandan[0.041]; Rwandan[0.041]; Rwandan[0.041]; Rwandan[0.041]; border[0.040]; camps[0.040]; camps[0.040]; camps[0.040]; refugees[0.040]; refugees[0.040]; refugees[0.040]; refugees[0.040]; refugees[0.040]; refugees[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rwanda ==> ENTITY: \u001b[32mRwanda\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.220:0.220[0]; log p(e|m)= -0.429:-0.429[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRwanda[0.048]; Rwanda[0.048]; Rwanda[0.048]; Rwandan[0.045]; Rwandan[0.045]; Rwandan[0.045]; Zaire[0.045]; refugees[0.043]; refugees[0.043]; refugees[0.043]; refugees[0.043]; refugees[0.043]; refugees[0.043]; troops[0.043]; Zairean[0.042]; Zairean[0.042]; Zairean[0.042]; government[0.042]; camps[0.042]; camps[0.042]; border[0.042]; eastern[0.041]; repatriation[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.105:0.105[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstatement[0.047]; government[0.044]; officials[0.043]; operation[0.043]; policing[0.042]; said[0.042]; said[0.042]; said[0.042]; troops[0.042]; calls[0.042]; agents[0.042]; group[0.041]; group[0.041]; group[0.041]; urged[0.041]; called[0.041]; Rwanda[0.041]; Rwanda[0.041]; refugee[0.041]; refugee[0.041]; refugee[0.041]; demands[0.040]; lobby[0.040]; explaining[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rwandan\u001b[39m ==> ENTITY: \u001b[32mRwanda\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.235:0.235[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRwanda[0.046]; Rwanda[0.046]; Rwanda[0.046]; Rwanda[0.046]; Tanzania[0.043]; Rwandan[0.043]; Rwandan[0.043]; Zaire[0.042]; Zaire[0.042]; Hutus[0.042]; genocide[0.041]; refugees[0.041]; refugees[0.041]; refugees[0.041]; refugees[0.041]; refugees[0.041]; Hutu[0.040]; census[0.040]; refugee[0.040]; Zairean[0.040]; Zairean[0.040]; Zairean[0.040]; Zairean[0.040]; people[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tanzania ==> ENTITY: \u001b[32mTanzania\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.166:0.166[0]; log p(e|m)= -0.257:-0.257[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAugust[0.046]; Rwandan[0.043]; Rwandan[0.043]; Rwandan[0.043]; Rwanda[0.043]; Rwanda[0.043]; Zaire[0.043]; Zaire[0.043]; camps[0.042]; camps[0.042]; refugee[0.042]; home[0.042]; week[0.041]; refugees[0.041]; refugees[0.041]; refugees[0.041]; Zairean[0.041]; Zairean[0.041]; Zairean[0.041]; people[0.040]; visit[0.040]; nearly[0.040]; wa[0.040]; start[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rwanda ==> ENTITY: \u001b[32mRwanda\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.236:0.236[0]; log p(e|m)= -0.429:-0.429[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRwanda[0.046]; Rwanda[0.046]; Rwanda[0.046]; Tanzania[0.043]; Rwandan[0.043]; Rwandan[0.043]; Rwandan[0.043]; Rwandan[0.043]; Zaire[0.042]; Zaire[0.042]; genocide[0.041]; refugees[0.041]; refugees[0.041]; refugees[0.041]; refugees[0.041]; refugees[0.041]; troops[0.040]; troops[0.040]; Zairean[0.040]; Zairean[0.040]; Zairean[0.040]; Zairean[0.040]; government[0.040]; refugee[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zairean\u001b[39m ==> ENTITY: \u001b[32mDemocratic Republic of the Congo\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaire[0.046]; Rwanda[0.044]; Rwanda[0.044]; Rwanda[0.044]; eastern[0.044]; Zairean[0.043]; Zairean[0.043]; Zairean[0.043]; Rwandan[0.042]; Rwandan[0.042]; Rwandan[0.042]; border[0.041]; camps[0.041]; camps[0.041]; camps[0.041]; refugees[0.040]; refugees[0.040]; refugees[0.040]; refugees[0.040]; refugees[0.040]; refugee[0.040]; troops[0.040]; government[0.040]; governments[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = RDR ==> ENTITY: \u001b[32mRepublican Rally for Democracy in Rwanda\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.166:0.166[0]; log p(e|m)= -1.423:-1.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRwanda[0.046]; Rwanda[0.046]; government[0.044]; Rwandan[0.044]; Rwandan[0.044]; Rwandan[0.044]; refugees[0.044]; refugees[0.044]; refugees[0.044]; refugees[0.044]; refugees[0.044]; Democracy[0.044]; group[0.043]; group[0.043]; group[0.043]; camps[0.043]; camps[0.043]; Rally[0.043]; officials[0.042]; refugee[0.042]; refugee[0.042]; refugee[0.042]; urged[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rwandan\u001b[39m ==> ENTITY: \u001b[32mRwanda\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRwanda[0.048]; Rwanda[0.048]; Rwanda[0.048]; Tanzania[0.045]; Rwandan[0.045]; Rwandan[0.045]; Zaire[0.044]; Zaire[0.044]; Hutus[0.044]; genocide[0.043]; refugees[0.043]; refugees[0.043]; refugees[0.043]; Hutu[0.042]; refugee[0.042]; Zairean[0.042]; Zairean[0.042]; Zairean[0.042]; people[0.042]; camps[0.042]; camps[0.042]; troops[0.041]; wa[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rwandan\u001b[39m ==> ENTITY: \u001b[32mRwanda\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRwanda[0.048]; Rwandan[0.045]; Zaire[0.044]; refugees[0.043]; refugees[0.043]; refugees[0.043]; census[0.042]; census[0.042]; census[0.042]; census[0.042]; refugee[0.042]; refugee[0.042]; refugee[0.042]; camps[0.041]; Nairobi[0.041]; eastern[0.041]; Refugees[0.040]; Refugees[0.040]; use[0.039]; statement[0.039]; group[0.039]; group[0.039]; group[0.039]; aid[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zaire ==> ENTITY: \u001b[32mDemocratic Republic of the Congo\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.219:0.219[0]; log p(e|m)= -2.002:-2.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaire[0.045]; Rwanda[0.044]; Rwanda[0.044]; Rwanda[0.044]; Rwanda[0.044]; Tanzania[0.043]; eastern[0.043]; Zairean[0.043]; Zairean[0.043]; Zairean[0.043]; Zairean[0.043]; Rwandan[0.041]; Rwandan[0.041]; Rwandan[0.041]; border[0.040]; troops[0.040]; camps[0.040]; camps[0.040]; camps[0.040]; refugees[0.040]; refugees[0.040]; refugees[0.040]; refugees[0.040]; refugees[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rwanda ==> ENTITY: \u001b[32mRwanda\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.219:0.219[0]; log p(e|m)= -0.429:-0.429[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRwanda[0.046]; Rwanda[0.046]; Rwandan[0.043]; Rwandan[0.043]; Rwandan[0.043]; Rwandan[0.043]; Zaire[0.043]; refugees[0.041]; refugees[0.041]; refugees[0.041]; refugees[0.041]; refugees[0.041]; refugees[0.041]; troops[0.041]; troops[0.041]; Zairean[0.041]; Zairean[0.041]; Zairean[0.041]; Zairean[0.041]; government[0.041]; refugee[0.040]; camps[0.040]; camps[0.040]; camps[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rwandan\u001b[39m ==> ENTITY: \u001b[32mRwanda\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.213:0.213[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRwanda[0.046]; Rwanda[0.046]; Rwanda[0.046]; Rwanda[0.046]; Rwandan[0.043]; Zaire[0.043]; refugees[0.041]; refugees[0.041]; refugees[0.041]; refugees[0.041]; refugees[0.041]; census[0.041]; census[0.041]; census[0.041]; census[0.041]; Zairean[0.040]; Zairean[0.040]; Zairean[0.040]; camps[0.040]; camps[0.040]; government[0.040]; troops[0.039]; eastern[0.039]; eastern[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rwanda ==> ENTITY: \u001b[32mRwanda\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.208:0.208[0]; log p(e|m)= -0.429:-0.429[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRwanda[0.047]; Tanzania[0.044]; Rwandan[0.044]; Rwandan[0.044]; Rwandan[0.044]; Zaire[0.044]; Zaire[0.044]; genocide[0.043]; refugees[0.042]; refugees[0.042]; refugees[0.042]; refugee[0.041]; Zairean[0.041]; Zairean[0.041]; camps[0.041]; saying[0.040]; troops[0.040]; Hutu[0.040]; governments[0.039]; expulsions[0.039]; Hutus[0.039]; week[0.039]; repatriation[0.039]; repatriation[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zairean\u001b[39m ==> ENTITY: \u001b[32mDemocratic Republic of the Congo\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.223:0.223[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaire[0.045]; Zaire[0.045]; Rwanda[0.044]; Rwanda[0.044]; Rwanda[0.044]; Rwanda[0.044]; Tanzania[0.043]; eastern[0.043]; Zairean[0.043]; Zairean[0.043]; Zairean[0.043]; Rwandan[0.041]; Rwandan[0.041]; Rwandan[0.041]; Rwandan[0.041]; border[0.040]; camps[0.040]; camps[0.040]; camps[0.040]; refugees[0.040]; refugees[0.040]; refugees[0.040]; refugees[0.040]; refugees[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hutu\u001b[39m ==> ENTITY: \u001b[32mHutu\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRwanda[0.045]; Rwanda[0.045]; Rwanda[0.045]; Rwandan[0.044]; Rwandan[0.044]; Rwandan[0.044]; Hutus[0.044]; genocide[0.042]; Zairean[0.042]; Zairean[0.042]; Zairean[0.042]; refugees[0.041]; refugees[0.041]; refugees[0.041]; Zaire[0.041]; Zaire[0.041]; refugee[0.041]; people[0.040]; Tanzania[0.040]; nearly[0.040]; expulsions[0.039]; troops[0.039]; camps[0.039]; camps[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kengo wa Dondo\u001b[39m ==> ENTITY: \u001b[32mLéon Kengo\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaire[0.050]; Zaire[0.050]; government[0.050]; Minister[0.048]; Zairean[0.048]; Zairean[0.048]; Zairean[0.048]; Zairean[0.048]; starting[0.048]; Rwanda[0.047]; Rwanda[0.047]; Rwanda[0.047]; Rwandan[0.047]; Rwandan[0.047]; Rwandan[0.047]; Rwandan[0.047]; governments[0.046]; Hutu[0.046]; born[0.046]; Tuesday[0.046]; nearly[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.118:0.118[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstatement[0.048]; week[0.045]; week[0.045]; government[0.045]; governments[0.045]; officials[0.044]; operation[0.044]; agreed[0.044]; policing[0.043]; weekly[0.043]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; troops[0.043]; Tuesday[0.043]; agents[0.043]; group[0.043]; repatriation[0.042]; border[0.042]; urged[0.042]; million[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rwandan\u001b[39m ==> ENTITY: \u001b[32mRwanda\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRwanda[0.049]; Rwandan[0.045]; Zaire[0.045]; refugees[0.043]; refugees[0.043]; census[0.043]; census[0.043]; census[0.043]; refugee[0.043]; refugee[0.043]; refugee[0.043]; camps[0.042]; Nairobi[0.042]; eastern[0.041]; Refugees[0.041]; Refugees[0.041]; group[0.040]; group[0.040]; aid[0.039]; Sunday[0.039]; explaining[0.039]; called[0.039]; policing[0.038]; workers[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zaire ==> ENTITY: \u001b[32mDemocratic Republic of the Congo\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.157:0.157[0]; log p(e|m)= -2.002:-2.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRwanda[0.046]; eastern[0.046]; Rwandan[0.044]; Rwandan[0.044]; Rwandan[0.044]; troops[0.043]; camps[0.042]; refugees[0.042]; refugees[0.042]; refugees[0.042]; government[0.042]; refugee[0.042]; refugee[0.042]; refugee[0.042]; group[0.041]; group[0.041]; group[0.041]; Nairobi[0.040]; aid[0.040]; Saturday[0.040]; said[0.039]; said[0.039]; mark[0.039]; ink[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = RDR ==> ENTITY: \u001b[32mRepublican Rally for Democracy in Rwanda\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.146:0.146[0]; log p(e|m)= -1.423:-1.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRwanda[0.044]; Rwanda[0.044]; Rwanda[0.044]; genocide[0.043]; Rwandan[0.042]; Rwandan[0.042]; Rwandan[0.042]; refugees[0.042]; refugees[0.042]; refugees[0.042]; refugees[0.042]; refugees[0.042]; camps[0.041]; camps[0.041]; camps[0.041]; closure[0.041]; Hutus[0.041]; home[0.041]; refugee[0.041]; saying[0.041]; eastern[0.041]; governments[0.040]; Goma[0.040]; nearly[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rwandan\u001b[39m ==> ENTITY: \u001b[32mRwanda\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.239:0.239[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRwanda[0.045]; Rwanda[0.045]; Rwanda[0.045]; Rwanda[0.045]; Tanzania[0.043]; Rwandan[0.042]; Rwandan[0.042]; Rwandan[0.042]; Zaire[0.042]; Zaire[0.042]; Hutus[0.042]; genocide[0.041]; refugees[0.041]; refugees[0.041]; refugees[0.041]; refugees[0.041]; refugees[0.041]; Hutu[0.040]; census[0.040]; refugee[0.040]; Zairean[0.040]; Zairean[0.040]; Zairean[0.040]; Zairean[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Goma\u001b[39m ==> ENTITY: \u001b[32mGoma\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDondo[0.044]; Rwanda[0.043]; Rwanda[0.043]; Rwanda[0.043]; Rwandan[0.043]; Rwandan[0.043]; Rwandan[0.043]; Zairean[0.042]; Zairean[0.042]; Zairean[0.042]; Zairean[0.042]; refugees[0.041]; refugees[0.041]; refugees[0.041]; refugees[0.041]; refugees[0.041]; refugees[0.041]; troops[0.041]; troops[0.041]; border[0.040]; eastern[0.040]; refugee[0.040]; start[0.040]; town[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zaire ==> ENTITY: \u001b[32mDemocratic Republic of the Congo\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.201:0.201[0]; log p(e|m)= -2.002:-2.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZaire[0.047]; Rwanda[0.045]; Rwanda[0.045]; Tanzania[0.044]; Zairean[0.044]; Zairean[0.044]; Zairean[0.044]; Rwandan[0.042]; Rwandan[0.042]; Rwandan[0.042]; troops[0.041]; camps[0.041]; camps[0.041]; refugees[0.041]; refugees[0.041]; refugees[0.041]; refugee[0.040]; nearly[0.040]; people[0.039]; million[0.039]; million[0.039]; million[0.039]; governments[0.039]; born[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = RDR ==> ENTITY: \u001b[32mRepublican Rally for Democracy in Rwanda\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.154:0.154[0]; log p(e|m)= -1.423:-1.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRwanda[0.046]; Rwanda[0.046]; government[0.044]; Rwandan[0.044]; Rwandan[0.044]; Rwandan[0.044]; refugees[0.044]; refugees[0.044]; refugees[0.044]; Democracy[0.044]; group[0.043]; group[0.043]; group[0.043]; camps[0.043]; Rally[0.043]; officials[0.043]; refugee[0.043]; refugee[0.043]; refugee[0.043]; urged[0.042]; eastern[0.042]; policing[0.042]; Zaire[0.042]; \t\n\r [====================>.........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s401ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2274/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1043testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Beijing ==> ENTITY: \u001b[32mBeijing\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.042:-0.042[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.053]; China[0.053]; Friday[0.046]; June[0.045]; July[0.045]; southwestern[0.045]; southwest[0.045]; Yunnan[0.045]; Xinhua[0.045]; province[0.044]; total[0.043]; late[0.043]; authorities[0.042]; local[0.041]; said[0.041]; said[0.041]; said[0.041]; making[0.041]; group[0.041]; people[0.040]; people[0.040]; people[0.040]; left[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Xinhua\u001b[39m ==> ENTITY: \u001b[32mXinhua News Agency\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBeijing[0.045]; province[0.045]; China[0.045]; China[0.045]; area[0.045]; residents[0.045]; southwest[0.044]; southwestern[0.043]; Yunnan[0.043]; news[0.042]; reports[0.042]; county[0.042]; June[0.041]; July[0.041]; agency[0.040]; agency[0.040]; local[0.040]; authorities[0.039]; Local[0.039]; people[0.039]; people[0.039]; people[0.039]; toxic[0.038]; toxic[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Xinhua\u001b[39m ==> ENTITY: \u001b[32mXinhua News Agency\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.118:0.118[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nprovince[0.047]; area[0.047]; residents[0.046]; southwestern[0.045]; Yunnan[0.045]; reports[0.044]; county[0.043]; June[0.042]; July[0.042]; agency[0.042]; local[0.042]; authorities[0.041]; Local[0.040]; people[0.040]; toxic[0.040]; toxic[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; poisoning[0.039]; launched[0.039]; methanol[0.039]; methanol[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.055]; Beijing[0.051]; southwest[0.048]; residents[0.047]; southwestern[0.047]; Yunnan[0.046]; launched[0.046]; July[0.046]; Xinhua[0.046]; province[0.046]; June[0.046]; local[0.045]; county[0.045]; late[0.044]; total[0.044]; authorities[0.044]; received[0.043]; Friday[0.043]; said[0.043]; said[0.043]; said[0.043]; making[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yunnan\u001b[39m ==> ENTITY: \u001b[32mYunnan\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsouthwestern[0.046]; province[0.044]; China[0.044]; China[0.044]; southwest[0.043]; Beijing[0.043]; Xinhua[0.043]; Xinhua[0.043]; caused[0.043]; local[0.042]; area[0.042]; killed[0.042]; people[0.040]; people[0.040]; people[0.040]; left[0.040]; late[0.040]; county[0.040]; selling[0.040]; selling[0.040]; police[0.040]; July[0.040]; authorities[0.039]; total[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Huize\u001b[39m ==> ENTITY: \u001b[32mHuize County\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYunnan[0.051]; province[0.050]; China[0.050]; China[0.050]; area[0.049]; county[0.047]; Xinhua[0.047]; Beijing[0.047]; southwestern[0.047]; southwest[0.045]; Bootleg[0.044]; residents[0.044]; authorities[0.044]; examinations[0.043]; liquor[0.043]; liquor[0.043]; liquor[0.043]; showed[0.043]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.051]; Beijing[0.047]; southwest[0.045]; residents[0.043]; southwestern[0.043]; Yunnan[0.043]; July[0.042]; June[0.042]; Xinhua[0.042]; province[0.042]; local[0.042]; county[0.041]; late[0.041]; total[0.041]; making[0.040]; Friday[0.040]; said[0.040]; said[0.040]; police[0.039]; Police[0.039]; people[0.039]; people[0.039]; people[0.039]; group[0.039]; \t\n\r [====================>.........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s403ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2281/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1063testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Marino ==> ENTITY: \u001b[32mSan Marino national football team\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.144:0.144[0]; log p(e|m)= -2.216:-2.216[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSan[0.047]; qualifier[0.045]; Marino[0.045]; soccer[0.045]; European[0.044]; Cardiff[0.043]; Wales[0.043]; Wales[0.043]; Qualifier[0.042]; minute[0.042]; Soccer[0.041]; Cup[0.041]; Cup[0.041]; Saturday[0.041]; John[0.041]; Robinson[0.040]; World[0.040]; World[0.040]; beat[0.040]; halftime[0.040]; Mark[0.039]; group[0.039]; Andy[0.039]; Dean[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEuropean\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUEFA\u001b[39m <---> \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.240:0.239[\u001b[31m0.001\u001b[39m]; local()= 0.104:0.039[\u001b[31m0.065\u001b[39m]; log p(e|m)= -3.576:-0.256[\u001b[32m3.319\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.048]; Soccer[0.048]; World[0.046]; World[0.046]; Cup[0.046]; Cup[0.046]; qualifier[0.046]; Qualifier[0.045]; Cardiff[0.044]; group[0.044]; Wales[0.043]; Wales[0.043]; Saturday[0.042]; John[0.042]; Mark[0.042]; beat[0.042]; Andy[0.041]; Dean[0.041]; minute[0.041]; Hughes[0.041]; Robinson[0.040]; Attendance[0.040]; 2nd[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dean Saunders\u001b[39m ==> ENTITY: \u001b[32mDean Saunders\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.112:0.112[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHughes[0.048]; Cardiff[0.046]; soccer[0.044]; Soccer[0.044]; Mark[0.043]; Andy[0.043]; Robinson[0.042]; European[0.042]; qualifier[0.041]; John[0.041]; Qualifier[0.041]; Wales[0.041]; Wales[0.041]; Beat[0.040]; Cup[0.040]; Cup[0.040]; halftime[0.040]; beat[0.040]; World[0.040]; World[0.040]; Scorers[0.040]; Attendance[0.040]; Saturday[0.040]; minute[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Marino ==> ENTITY: \u001b[32mSan Marino national football team\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.144:0.144[0]; log p(e|m)= -2.216:-2.216[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSan[0.047]; qualifier[0.045]; Marino[0.045]; soccer[0.045]; European[0.044]; Cardiff[0.043]; Wales[0.043]; Wales[0.043]; Qualifier[0.042]; minute[0.042]; Soccer[0.041]; Cup[0.041]; Cup[0.041]; Saturday[0.041]; John[0.041]; Robinson[0.040]; World[0.040]; World[0.040]; beat[0.040]; halftime[0.040]; Mark[0.039]; group[0.039]; Andy[0.039]; Dean[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mark Hughes ==> ENTITY: \u001b[32mMark Hughes\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.087:0.087[0]; log p(e|m)= -0.105:-0.105[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCardiff[0.047]; Saunders[0.045]; Cup[0.044]; Cup[0.044]; Qualifier[0.044]; Andy[0.043]; Wales[0.043]; Wales[0.043]; Saturday[0.043]; Soccer[0.043]; soccer[0.042]; Robinson[0.042]; qualifier[0.042]; European[0.041]; Scorers[0.041]; John[0.040]; group[0.039]; Dean[0.039]; beat[0.039]; World[0.039]; World[0.039]; 2nd[0.039]; halftime[0.038]; Melville[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wales ==> ENTITY: \u001b[32mWales national football team\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.147:0.147[0]; log p(e|m)= -2.453:-2.453[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWales[0.054]; Cardiff[0.048]; Cup[0.045]; Cup[0.045]; Qualifier[0.044]; soccer[0.044]; Soccer[0.043]; qualifier[0.043]; Hughes[0.043]; Robinson[0.042]; European[0.042]; Andy[0.042]; Mark[0.041]; John[0.041]; World[0.040]; World[0.040]; Saturday[0.039]; Dean[0.039]; beat[0.038]; halftime[0.038]; Saunders[0.038]; Melville[0.037]; Scorers[0.037]; minute[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cardiff ==> ENTITY: \u001b[32mCardiff\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.088:0.088[0]; log p(e|m)= -0.205:-0.205[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWales[0.047]; Wales[0.047]; Cup[0.045]; Cup[0.045]; Hughes[0.044]; Soccer[0.043]; Mark[0.043]; soccer[0.043]; Saturday[0.043]; Andy[0.043]; Robinson[0.042]; Dean[0.041]; John[0.041]; qualifier[0.040]; Qualifier[0.040]; beat[0.040]; halftime[0.040]; minute[0.039]; World[0.039]; World[0.039]; Saunders[0.039]; European[0.039]; group[0.038]; Melville[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andy Melville\u001b[39m ==> ENTITY: \u001b[32mAndy Melville\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.094:0.094[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCardiff[0.048]; Wales[0.045]; Wales[0.045]; Hughes[0.044]; Mark[0.043]; Soccer[0.043]; soccer[0.043]; qualifier[0.042]; Qualifier[0.042]; Dean[0.042]; John[0.042]; minute[0.041]; Robinson[0.041]; halftime[0.041]; Scorers[0.041]; Attendance[0.041]; Saunders[0.041]; Saturday[0.040]; 2nd[0.040]; Cup[0.039]; Cup[0.039]; San[0.039]; San[0.039]; beat[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Cup ==> ENTITY: \u001b[32mFIFA World Cup\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.095:0.095[0]; log p(e|m)= -1.002:-1.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.048]; Soccer[0.046]; Wales[0.046]; Wales[0.046]; Cardiff[0.046]; Cup[0.045]; qualifier[0.045]; World[0.044]; European[0.044]; Qualifier[0.043]; Saturday[0.041]; Mark[0.040]; beat[0.040]; minute[0.040]; Andy[0.040]; halftime[0.040]; Robinson[0.040]; group[0.039]; John[0.039]; Beat[0.038]; 2nd[0.038]; Scorers[0.038]; Dean[0.037]; Hughes[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = John Robinson ==> ENTITY: \u001b[32mJohn Robinson (footballer)\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.086:0.086[0]; log p(e|m)= -2.976:-2.976[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMark[0.049]; Dean[0.048]; Hughes[0.048]; Andy[0.046]; Saunders[0.045]; Melville[0.043]; Wales[0.043]; Wales[0.043]; Cardiff[0.042]; Soccer[0.042]; Saturday[0.041]; group[0.041]; soccer[0.041]; European[0.040]; Marino[0.040]; Marino[0.040]; Qualifier[0.040]; Attendance[0.039]; halftime[0.039]; Cup[0.038]; Cup[0.038]; qualifier[0.038]; 2nd[0.038]; beat[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wales ==> ENTITY: \u001b[32mWales national football team\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.147:0.147[0]; log p(e|m)= -2.453:-2.453[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWales[0.054]; Cardiff[0.048]; Cup[0.045]; Cup[0.045]; Qualifier[0.044]; soccer[0.044]; Soccer[0.043]; qualifier[0.043]; Hughes[0.043]; Robinson[0.042]; European[0.042]; Andy[0.042]; Mark[0.041]; John[0.041]; World[0.040]; World[0.040]; Saturday[0.039]; Dean[0.039]; beat[0.038]; halftime[0.038]; Saunders[0.038]; Melville[0.037]; Scorers[0.037]; minute[0.037]; \t\n\r [=====================>........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s360ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2293/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 968testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Geneva ==> ENTITY: \u001b[32mGeneva\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.103:0.103[0]; log p(e|m)= -0.247:-0.247[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseven[0.045]; Germany[0.044]; conduct[0.044]; Swiss[0.044]; Liege[0.043]; Liege[0.043]; team[0.043]; Friday[0.043]; Belgian[0.042]; Belgian[0.042]; Didier[0.041]; final[0.041]; came[0.041]; club[0.041]; club[0.041]; sent[0.041]; heavily[0.041]; games[0.041]; Guy[0.041]; Uefa[0.040]; verbal[0.040]; matches[0.040]; Bisconti[0.040]; Roberto[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mGermany\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mGermany national football team\u001b[39m <---> \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.241:0.240[\u001b[31m0.001\u001b[39m]; local()= 0.176:0.072[\u001b[31m0.104\u001b[39m]; log p(e|m)= -3.352:-0.155[\u001b[32m3.197\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.051]; Soccer[0.049]; match[0.049]; match[0.049]; team[0.048]; games[0.048]; Euro[0.046]; club[0.046]; club[0.046]; Uefa[0.046]; captain[0.045]; won[0.045]; goals[0.044]; goal[0.044]; final[0.044]; August[0.043]; Intertoto[0.043]; banned[0.043]; banned[0.043]; seven[0.043]; tie[0.043]; took[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bisconti\u001b[39m ==> ENTITY: \u001b[32mRoberto Bisconti\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.044]; club[0.044]; Uefa[0.043]; matches[0.043]; games[0.043]; Karlsruhe[0.043]; Karlsruhe[0.043]; Karlsruhe[0.043]; Belgian[0.042]; Belgian[0.042]; Didier[0.042]; team[0.042]; Bisconti[0.042]; Liege[0.041]; Roberto[0.041]; Friday[0.041]; August[0.041]; match[0.041]; match[0.041]; Germany[0.040]; Swiss[0.040]; Intertoto[0.040]; Ernst[0.039]; came[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgian ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.124:-0.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgian[0.049]; matches[0.047]; match[0.045]; match[0.045]; Germany[0.045]; club[0.045]; club[0.045]; Swiss[0.045]; Soccer[0.044]; games[0.044]; Uefa[0.044]; won[0.043]; Euro[0.043]; Liege[0.042]; Liege[0.042]; team[0.042]; Intertoto[0.042]; pushing[0.041]; goals[0.041]; sidelined[0.041]; qualified[0.041]; Didier[0.041]; seven[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Intertoto\u001b[39m ==> ENTITY: \u001b[32mUEFA Intertoto Cup\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.046]; club[0.046]; matches[0.045]; won[0.044]; Soccer[0.043]; Uefa[0.043]; games[0.043]; team[0.042]; final[0.042]; match[0.042]; match[0.042]; Euro[0.042]; Liege[0.041]; Liege[0.041]; sidelined[0.041]; goals[0.041]; Karlsruhe[0.040]; Karlsruhe[0.040]; Karlsruhe[0.040]; Germany[0.040]; Roberto[0.040]; August[0.039]; Swiss[0.039]; late[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Karlsruhe ==> ENTITY: \u001b[32mKarlsruher SC\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.208:0.208[0]; log p(e|m)= -2.226:-2.226[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.046]; club[0.046]; matches[0.045]; Germany[0.044]; team[0.044]; match[0.043]; match[0.043]; games[0.043]; won[0.042]; Cup[0.042]; Karlsruhe[0.041]; Karlsruhe[0.041]; tie[0.041]; Intertoto[0.041]; final[0.041]; Ernst[0.041]; goals[0.041]; took[0.040]; goal[0.040]; attack[0.039]; August[0.039]; Uefa[0.039]; late[0.039]; Swiss[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swiss ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.049]; match[0.047]; match[0.047]; games[0.046]; team[0.045]; Germany[0.045]; won[0.044]; Soccer[0.044]; final[0.044]; Uefa[0.043]; Uefa[0.043]; Euro[0.043]; tie[0.043]; Friday[0.043]; Cup[0.042]; goal[0.042]; seven[0.042]; club[0.042]; club[0.042]; August[0.041]; goals[0.041]; referee[0.041]; referee[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Standard Liege\u001b[39m ==> ENTITY: \u001b[32mStandard Liège\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.047]; club[0.047]; Soccer[0.045]; Liege[0.045]; games[0.044]; match[0.043]; Intertoto[0.043]; team[0.043]; Didier[0.042]; Karlsruhe[0.042]; Karlsruhe[0.042]; seven[0.042]; Uefa[0.041]; Belgian[0.041]; Belgian[0.041]; Germany[0.041]; goal[0.040]; came[0.040]; Club[0.040]; final[0.040]; Euro[0.039]; Friday[0.039]; francs[0.038]; Guy[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgian ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.107:0.107[0]; log p(e|m)= -0.124:-0.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgian[0.047]; matches[0.046]; match[0.044]; Germany[0.044]; club[0.044]; club[0.044]; Swiss[0.043]; Soccer[0.043]; games[0.043]; Uefa[0.042]; Euro[0.042]; Liege[0.041]; Liege[0.041]; team[0.041]; Intertoto[0.041]; pushing[0.040]; sidelined[0.040]; Didier[0.040]; seven[0.039]; referee[0.039]; referee[0.039]; final[0.039]; goal[0.039]; Club[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = UEFA Cup ==> ENTITY: \u001b[32mUEFA Europa League\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.164:0.164[0]; log p(e|m)= -0.261:-0.261[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.047]; won[0.047]; match[0.045]; games[0.045]; aggregate[0.045]; team[0.044]; tie[0.043]; goals[0.042]; Karlsruhe[0.042]; Karlsruhe[0.042]; goal[0.042]; sidelined[0.042]; Euro[0.041]; qualified[0.041]; seven[0.040]; Roberto[0.040]; fined[0.040]; took[0.039]; late[0.039]; Didier[0.039]; August[0.039]; ties[0.039]; thanks[0.039]; Swiss[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Guy Hellers\u001b[39m ==> ENTITY: \u001b[32mGuy Hellers\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.048]; Soccer[0.047]; match[0.046]; match[0.046]; Cup[0.044]; team[0.044]; Karlsruhe[0.044]; Karlsruhe[0.044]; Karlsruhe[0.044]; games[0.044]; club[0.043]; club[0.043]; Ernst[0.043]; dismissed[0.043]; Germany[0.042]; won[0.042]; Euro[0.042]; Uefa[0.042]; Uefa[0.042]; goals[0.042]; Liege[0.042]; Liege[0.042]; qualified[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Karlsruhe ==> ENTITY: \u001b[32mKarlsruher SC\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.219:0.219[0]; log p(e|m)= -2.226:-2.226[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.047]; club[0.046]; club[0.046]; matches[0.045]; Germany[0.044]; team[0.043]; games[0.043]; Club[0.043]; match[0.043]; match[0.043]; won[0.042]; Karlsruhe[0.041]; Karlsruhe[0.041]; tie[0.041]; Intertoto[0.040]; final[0.040]; Ernst[0.040]; goals[0.040]; took[0.039]; goal[0.039]; Players[0.039]; came[0.039]; Uefa[0.039]; late[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = UEFA\u001b[39m ==> ENTITY: \u001b[32mUEFA\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.047]; match[0.045]; club[0.044]; club[0.044]; games[0.044]; team[0.043]; Soccer[0.043]; Euro[0.043]; Germany[0.043]; final[0.042]; seven[0.042]; banned[0.042]; banned[0.042]; Intertoto[0.041]; goal[0.041]; Roberto[0.040]; Swiss[0.040]; Players[0.040]; Didier[0.039]; Club[0.039]; fined[0.039]; Liege[0.039]; Liege[0.039]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Roberto Bisconti\u001b[39m ==> ENTITY: \u001b[32mRoberto Bisconti\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.043]; club[0.043]; Uefa[0.043]; Uefa[0.043]; matches[0.043]; games[0.042]; Karlsruhe[0.042]; Karlsruhe[0.042]; Karlsruhe[0.042]; Belgian[0.042]; Belgian[0.042]; Soccer[0.042]; Didier[0.042]; team[0.042]; Bisconti[0.042]; Liege[0.041]; Liege[0.041]; Friday[0.041]; August[0.041]; match[0.040]; match[0.040]; Germany[0.040]; Swiss[0.040]; Intertoto[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Standard Liege\u001b[39m ==> ENTITY: \u001b[32mStandard Liège\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.046]; club[0.046]; matches[0.046]; Soccer[0.044]; Liege[0.044]; games[0.043]; match[0.042]; match[0.042]; Intertoto[0.042]; team[0.042]; won[0.041]; Didier[0.041]; Karlsruhe[0.041]; Karlsruhe[0.041]; Karlsruhe[0.041]; seven[0.041]; Uefa[0.040]; Belgian[0.040]; Belgian[0.040]; Germany[0.040]; goal[0.039]; came[0.039]; Club[0.039]; August[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Karlsruhe ==> ENTITY: \u001b[32mKarlsruher SC\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.221:0.221[0]; log p(e|m)= -2.226:-2.226[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.047]; club[0.046]; club[0.046]; matches[0.045]; Germany[0.044]; team[0.043]; Club[0.043]; match[0.043]; match[0.043]; games[0.042]; won[0.042]; Cup[0.041]; Karlsruhe[0.041]; Karlsruhe[0.041]; tie[0.041]; Intertoto[0.040]; final[0.040]; Ernst[0.040]; goals[0.040]; took[0.039]; goal[0.039]; Players[0.039]; attack[0.039]; came[0.039]; \t\n\r [=====================>........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 11s309ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2309/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1073testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Walter Little\u001b[39m ==> ENTITY: \u001b[32mWalter Little (rugby player)\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHenry[0.046]; second[0.046]; Marshall[0.046]; Wayne[0.046]; scrum[0.045]; Gary[0.045]; Mehrtens[0.044]; Stransky[0.044]; Mark[0.043]; Springboks[0.043]; Springboks[0.043]; Joel[0.043]; South[0.043]; ruck[0.043]; captain[0.043]; half[0.042]; half[0.042]; late[0.042]; Honiball[0.042]; Justin[0.042]; Justin[0.042]; Fritz[0.042]; Black[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mehrtens\u001b[39m ==> ENTITY: \u001b[32mAndrew Mehrtens\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nscrum[0.045]; Springboks[0.044]; Springboks[0.044]; second[0.043]; time[0.042]; time[0.042]; Wayne[0.042]; ruck[0.042]; Honiball[0.042]; score[0.042]; score[0.042]; Blacks[0.041]; comeback[0.041]; try[0.041]; try[0.041]; injury[0.041]; half[0.041]; half[0.041]; raced[0.041]; Venter[0.040]; Venter[0.040]; Gary[0.040]; captain[0.040]; Marshall[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andre Joubert\u001b[39m ==> ENTITY: \u001b[32mAndré Joubert\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpringbok[0.053]; Springboks[0.052]; Honiball[0.051]; scrum[0.051]; scrum[0.051]; Fullback[0.051]; flanker[0.050]; African[0.050]; Mehrtens[0.050]; Venter[0.050]; final[0.050]; Black[0.049]; Westhuizen[0.049]; Blacks[0.049]; victories[0.049]; minutes[0.049]; minutes[0.049]; minutes[0.049]; second[0.049]; South[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = James Dalton ==> ENTITY: \u001b[32mJames Dalton (rugby player)\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.480:-0.480[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJoel[0.045]; Michael[0.045]; Mark[0.045]; Wayne[0.044]; Craig[0.044]; Glenn[0.044]; Andrew[0.044]; Jeff[0.044]; Zinzan[0.044]; South[0.044]; Honiball[0.044]; Ian[0.043]; Sean[0.043]; Garry[0.043]; Gary[0.043]; Wilson[0.043]; Glen[0.043]; captain[0.043]; captain[0.043]; Robin[0.043]; Zealand[0.042]; Jones[0.042]; Jones[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Garry Pagel\u001b[39m ==> ENTITY: \u001b[32mGarry Pagel\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.045]; Mehrtens[0.045]; Honiball[0.044]; Zinzan[0.042]; captain[0.042]; captain[0.042]; Henry[0.042]; Fitzpatrick[0.042]; Stransky[0.042]; time[0.041]; Venter[0.041]; Joubert[0.041]; Danie[0.041]; Swart[0.041]; Jones[0.041]; Jones[0.041]; Dawie[0.041]; Fritz[0.041]; Africa[0.041]; Walter[0.041]; Theron[0.041]; ruck[0.041]; Kobus[0.040]; Glenn[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = All Blacks\u001b[39m ==> ENTITY: \u001b[32mNew Zealand national rugby union team\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.048]; Springboks[0.045]; scrum[0.045]; Springbok[0.044]; Zealand[0.042]; Zealand[0.042]; series[0.042]; series[0.042]; season[0.042]; world[0.042]; South[0.041]; South[0.041]; Ellis[0.041]; fifth[0.041]; Johannesburg[0.041]; African[0.041]; Saturday[0.041]; arrived[0.040]; Africa[0.040]; Park[0.040]; opening[0.040]; try[0.040]; half[0.040]; late[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Henry Honiball\u001b[39m ==> ENTITY: \u001b[32mHenry Honiball\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nscrum[0.046]; Springboks[0.046]; Springboks[0.046]; Andrew[0.045]; Andrew[0.045]; Westhuizen[0.044]; Joubert[0.044]; Joubert[0.044]; Joubert[0.044]; Pieter[0.044]; Fullback[0.043]; Mehrtens[0.043]; Mehrtens[0.043]; Mehrtens[0.043]; second[0.043]; Marshall[0.042]; winger[0.042]; half[0.042]; half[0.042]; half[0.042]; half[0.042]; half[0.042]; knock[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Robin Brooke\u001b[39m ==> ENTITY: \u001b[32mRobin Brooke\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZinzan[0.046]; Mehrtens[0.044]; Ieremia[0.044]; Craig[0.044]; Brooke[0.043]; Wayne[0.043]; James[0.042]; Cullen[0.042]; Jones[0.042]; Jones[0.042]; Westhuizen[0.041]; captain[0.041]; captain[0.041]; Bunce[0.041]; Ian[0.041]; Sean[0.041]; Justin[0.041]; Glen[0.041]; Michael[0.040]; Gary[0.040]; Venter[0.040]; Kronfeld[0.040]; Heerden[0.040]; Kobus[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ian Jones\u001b[39m ==> ENTITY: \u001b[32mIan Jones (rugby union)\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.135:0.135[0]; log p(e|m)= -0.002:-0.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGlenn[0.046]; Zinzan[0.044]; Garry[0.044]; Michael[0.044]; Craig[0.044]; James[0.043]; Mark[0.043]; Jones[0.042]; Wayne[0.042]; Mehrtens[0.042]; Sean[0.042]; Andrew[0.041]; Glen[0.041]; Walter[0.041]; Robin[0.041]; captain[0.041]; captain[0.041]; Ieremia[0.041]; Dalton[0.040]; Frank[0.040]; Bunce[0.040]; Marshall[0.040]; Fitzpatrick[0.040]; Venter[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = All Black\u001b[39m ==> ENTITY: \u001b[32mNew Zealand national rugby union team\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpringboks[0.046]; Springboks[0.046]; Springboks[0.046]; hooker[0.044]; Welsh[0.044]; Mehrtens[0.043]; Mehrtens[0.043]; Honiball[0.042]; second[0.042]; winger[0.041]; time[0.041]; Blacks[0.041]; half[0.041]; half[0.041]; comeback[0.040]; referee[0.040]; try[0.040]; try[0.040]; try[0.040]; score[0.040]; score[0.040]; kicked[0.040]; kicked[0.040]; kicked[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South African ==> ENTITY: \u001b[32mSouth Africa national rugby union team\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.195:0.195[0]; log p(e|m)= -3.270:-3.270[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpringbok[0.045]; South[0.044]; series[0.043]; series[0.043]; scrum[0.043]; scrum[0.043]; Africa[0.043]; season[0.042]; flanker[0.042]; Zealand[0.042]; Zealand[0.042]; second[0.042]; Saturday[0.041]; fifth[0.041]; world[0.041]; final[0.041]; champions[0.041]; Park[0.041]; late[0.041]; Ellis[0.041]; opening[0.040]; successive[0.040]; successive[0.040]; Venter[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Justin Marshall\u001b[39m ==> ENTITY: \u001b[32mJustin Marshall\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nscrum[0.047]; Mehrtens[0.047]; Wayne[0.047]; Blacks[0.045]; Gary[0.045]; James[0.044]; second[0.044]; Springboks[0.044]; Springboks[0.044]; captain[0.044]; Mark[0.043]; Justin[0.043]; Henry[0.043]; Walter[0.043]; going[0.043]; half[0.042]; ruck[0.042]; Joel[0.042]; Honiball[0.042]; followed[0.042]; try[0.041]; try[0.041]; Hurter[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Africa ==> ENTITY: \u001b[32mSouth Africa national rugby union team\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.162:0.162[0]; log p(e|m)= -3.612:-3.612[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpringboks[0.047]; Springboks[0.047]; scrum[0.044]; Honiball[0.043]; Zealand[0.043]; second[0.042]; captain[0.042]; Mark[0.042]; penalty[0.042]; penalty[0.042]; Mehrtens[0.041]; Venter[0.041]; Venter[0.041]; injury[0.041]; Stransky[0.041]; James[0.040]; score[0.040]; score[0.040]; Justin[0.040]; Justin[0.040]; Joubert[0.040]; Joubert[0.040]; time[0.040]; Wayne[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Christian Cullen\u001b[39m ==> ENTITY: \u001b[32mChristian Cullen\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.128:0.128[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZinzan[0.045]; Zealand[0.044]; Mehrtens[0.043]; Honiball[0.043]; James[0.042]; time[0.042]; captain[0.042]; captain[0.042]; Andrew[0.042]; Jones[0.042]; Jones[0.042]; Michael[0.042]; Wayne[0.041]; injury[0.041]; Ieremia[0.041]; Glen[0.041]; South[0.041]; Henry[0.041]; Craig[0.041]; Pieter[0.040]; Glenn[0.040]; der[0.040]; Heerden[0.040]; Kobus[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Springboks\u001b[39m ==> ENTITY: \u001b[32mSouth Africa national rugby union team\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpringboks[0.047]; Welsh[0.046]; scrum[0.045]; day[0.043]; Honiball[0.043]; Bevan[0.043]; second[0.042]; Fullback[0.041]; Mehrtens[0.041]; Mehrtens[0.041]; Mehrtens[0.041]; winger[0.041]; final[0.041]; Westhuizen[0.041]; referee[0.040]; Pieter[0.040]; inside[0.040]; Little[0.040]; Joubert[0.040]; Joubert[0.040]; Joubert[0.040]; penalties[0.040]; penalties[0.040]; kicked[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Olo Brown\u001b[39m ==> ENTITY: \u001b[32mOlo Brown\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZinzan[0.045]; captain[0.044]; Zealand[0.044]; Ieremia[0.043]; Andrew[0.042]; Venter[0.042]; Glen[0.042]; James[0.042]; Ian[0.042]; Jones[0.042]; Jones[0.042]; Craig[0.041]; Little[0.041]; Walter[0.041]; Garry[0.041]; Brooke[0.041]; Brooke[0.041]; Mehrtens[0.041]; Wilson[0.041]; Andre[0.041]; Fitzpatrick[0.040]; Heerden[0.040]; Kobus[0.040]; Marius[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sean Fitzpatrick\u001b[39m ==> ENTITY: \u001b[32mSean Fitzpatrick\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJames[0.044]; Mark[0.044]; Michael[0.044]; Wayne[0.044]; Glenn[0.044]; captain[0.044]; Garry[0.043]; Craig[0.043]; Ian[0.042]; Andrew[0.042]; Zinzan[0.042]; Zealand[0.042]; Glen[0.041]; Wilson[0.041]; Mehrtens[0.040]; Cullen[0.040]; Jones[0.040]; Jones[0.040]; Ieremia[0.040]; Justin[0.040]; Dalton[0.040]; Marshall[0.040]; Taylor[0.040]; Jeff[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Zealand ==> ENTITY: \u001b[32mNew Zealand national rugby union team\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.184:0.184[0]; log p(e|m)= -3.817:-3.817[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.048]; recording[0.046]; Springboks[0.044]; scrum[0.043]; Springbok[0.043]; series[0.043]; series[0.043]; flanker[0.042]; Zealand[0.042]; world[0.042]; fifth[0.041]; Johannesburg[0.041]; African[0.041]; highest[0.041]; South[0.040]; South[0.040]; Ellis[0.040]; season[0.040]; Africa[0.040]; opening[0.040]; successive[0.040]; successive[0.040]; Park[0.040]; Saturday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sean Fitzpatrick\u001b[39m ==> ENTITY: \u001b[32mSean Fitzpatrick\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.045]; Springboks[0.043]; Springboks[0.043]; Springboks[0.043]; Welsh[0.043]; Blacks[0.043]; missed[0.042]; Honiball[0.042]; comeback[0.042]; time[0.042]; Mehrtens[0.042]; Mehrtens[0.042]; going[0.042]; half[0.041]; half[0.041]; Derek[0.041]; long[0.041]; late[0.040]; ahead[0.040]; kicked[0.040]; kicked[0.040]; kicked[0.040]; hooker[0.040]; scored[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andre Joubert\u001b[39m ==> ENTITY: \u001b[32mAndré Joubert\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJoubert[0.045]; Stransky[0.044]; Springboks[0.043]; Springboks[0.043]; Swart[0.043]; Honiball[0.042]; scrum[0.042]; Hurter[0.042]; Dawie[0.042]; Danie[0.042]; Japie[0.042]; Kobus[0.042]; Mehrtens[0.041]; Venter[0.041]; Venter[0.041]; Black[0.041]; Westhuizen[0.041]; Westhuizen[0.041]; Blacks[0.041]; minutes[0.040]; second[0.040]; Heerden[0.040]; Schalkwyk[0.040]; ruck[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Japie Mulder\u001b[39m ==> ENTITY: \u001b[32mJapie Mulder\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.045]; Springboks[0.045]; Zealand[0.043]; Westhuizen[0.043]; Westhuizen[0.043]; Venter[0.043]; Venter[0.043]; Mehrtens[0.043]; Africa[0.042]; Kobus[0.042]; Danie[0.041]; kicked[0.041]; Swart[0.041]; Dawie[0.041]; Joubert[0.041]; Joubert[0.041]; close[0.041]; Blacks[0.041]; Pieter[0.040]; der[0.040]; der[0.040]; Heerden[0.040]; Marius[0.040]; Theron[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = All Blacks\u001b[39m ==> ENTITY: \u001b[32mNew Zealand national rugby union team\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpringboks[0.049]; scrum[0.049]; scrum[0.049]; Springbok[0.048]; final[0.047]; flanker[0.046]; Zealand[0.046]; series[0.046]; series[0.046]; second[0.046]; Honiball[0.045]; South[0.045]; African[0.044]; Henry[0.044]; Mehrtens[0.044]; Mehrtens[0.044]; Mehrtens[0.044]; arrived[0.044]; missed[0.044]; missed[0.044]; opening[0.043]; try[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alama Ieremia\u001b[39m ==> ENTITY: \u001b[32mAlama Ieremia\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZinzan[0.047]; Little[0.045]; Mehrtens[0.045]; captain[0.044]; captain[0.044]; Honiball[0.044]; Zealand[0.043]; Venter[0.042]; Bunce[0.042]; Heerden[0.042]; Stransky[0.041]; Fyvie[0.041]; Swart[0.041]; Glen[0.041]; Kronfeld[0.040]; Hendriks[0.040]; Joubert[0.040]; Pieter[0.040]; der[0.040]; Kobus[0.040]; Jones[0.040]; Jones[0.040]; Marius[0.040]; Japie[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jeff Wilson\u001b[39m ==> ENTITY: \u001b[32mJeff Wilson (sportsman)\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGlenn[0.044]; Ian[0.043]; Zinzan[0.043]; Wayne[0.043]; Gary[0.043]; Garry[0.043]; Craig[0.042]; James[0.042]; Andrew[0.042]; Mehrtens[0.042]; Henry[0.042]; Mark[0.042]; Michael[0.041]; Joel[0.041]; Taylor[0.041]; Jones[0.041]; Jones[0.041]; captain[0.041]; captain[0.041]; Justin[0.040]; Justin[0.040]; Sean[0.040]; Ieremia[0.040]; Glen[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andrew Mehrtens\u001b[39m ==> ENTITY: \u001b[32mAndrew Mehrtens\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.229:0.229[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nscrum[0.044]; Springboks[0.044]; Springboks[0.044]; Mehrtens[0.043]; Mehrtens[0.043]; second[0.043]; hooker[0.042]; final[0.042]; points[0.041]; Honiball[0.041]; Honiball[0.041]; score[0.041]; Blacks[0.041]; missed[0.041]; missed[0.041]; half[0.041]; half[0.041]; half[0.041]; half[0.041]; half[0.041]; Welsh[0.041]; ahead[0.041]; point[0.040]; scoring[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Craig Dowd\u001b[39m ==> ENTITY: \u001b[32mCraig Dowd\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.089:0.089[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncaptain[0.045]; Zealand[0.044]; Zinzan[0.043]; Mehrtens[0.043]; Hurter[0.043]; Garry[0.042]; Heerden[0.042]; Kobus[0.042]; Marius[0.042]; Kronfeld[0.042]; Little[0.042]; Fitzpatrick[0.042]; Kruge[0.042]; Ian[0.042]; Sean[0.041]; Dawie[0.041]; Ieremia[0.041]; Alama[0.041]; Andrew[0.041]; Glenn[0.040]; Marshall[0.040]; Pagel[0.040]; Josh[0.040]; James[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Springboks\u001b[39m ==> ENTITY: \u001b[32mSouth Africa national rugby union team\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpringboks[0.056]; Welsh[0.054]; scrum[0.053]; Honiball[0.051]; South[0.051]; Bevan[0.051]; ruck[0.050]; second[0.050]; hooker[0.050]; Mehrtens[0.049]; winger[0.049]; Africa[0.048]; Venter[0.048]; comeback[0.048]; injury[0.048]; Westhuizen[0.048]; Westhuizen[0.048]; referee[0.048]; Pieter[0.048]; Little[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Springbok ==> ENTITY: \u001b[32mSouth Africa national rugby union team\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.186:0.186[0]; log p(e|m)= -0.957:-0.957[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseries[0.044]; series[0.044]; African[0.044]; scrum[0.043]; scrum[0.043]; season[0.043]; flanker[0.042]; Zealand[0.042]; Zealand[0.042]; South[0.042]; historic[0.042]; Saturday[0.042]; fifth[0.042]; second[0.042]; recording[0.041]; Africa[0.041]; Park[0.041]; Ellis[0.041]; final[0.040]; Venter[0.040]; champions[0.040]; Westhuizen[0.040]; late[0.040]; highest[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Joel Stransky\u001b[39m ==> ENTITY: \u001b[32mJoel Stransky\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhalf[0.045]; Wayne[0.044]; South[0.043]; scrum[0.043]; Mark[0.043]; Springboks[0.042]; captain[0.042]; Andrew[0.042]; Garry[0.042]; James[0.042]; quarter[0.042]; Wilson[0.042]; Marshall[0.042]; score[0.041]; Gary[0.041]; Hurter[0.041]; Honiball[0.041]; time[0.041]; Blacks[0.041]; penalty[0.041]; penalty[0.041]; kicked[0.040]; try[0.040]; Theron[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Henry Honiball\u001b[39m ==> ENTITY: \u001b[32mHenry Honiball\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWayne[0.046]; scrum[0.046]; Springboks[0.045]; Zinzan[0.045]; Zealand[0.044]; Andrew[0.044]; ruck[0.044]; Westhuizen[0.044]; Joubert[0.043]; Joubert[0.043]; captain[0.043]; Pieter[0.043]; South[0.043]; Hurter[0.043]; Mark[0.043]; Heerden[0.043]; try[0.043]; Dawie[0.042]; Mehrtens[0.042]; Mehrtens[0.042]; Swart[0.042]; Stransky[0.042]; Joost[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Africa ==> ENTITY: \u001b[32mSouth Africa national rugby union team\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.196:0.196[0]; log p(e|m)= -3.612:-3.612[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.050]; Springboks[0.047]; Springbok[0.046]; South[0.046]; series[0.045]; series[0.045]; African[0.045]; season[0.044]; Zealand[0.043]; Zealand[0.043]; Johannesburg[0.043]; Union[0.043]; Saturday[0.043]; fifth[0.043]; world[0.042]; champions[0.042]; victory[0.042]; Park[0.042]; late[0.042]; Ellis[0.042]; victories[0.042]; successive[0.041]; successive[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Walter Little\u001b[39m ==> ENTITY: \u001b[32mWalter Little (rugby player)\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBunce[0.045]; Frank[0.044]; Glen[0.043]; Henry[0.043]; Dalton[0.042]; Marshall[0.042]; Wayne[0.042]; James[0.042]; Garry[0.042]; Glenn[0.042]; Andrew[0.042]; Gary[0.042]; Wilson[0.041]; Jones[0.041]; Jones[0.041]; Zinzan[0.041]; Craig[0.041]; Brown[0.041]; Mehrtens[0.041]; Osborne[0.041]; Taylor[0.041]; Stransky[0.041]; Ian[0.040]; Mark[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Frank Bunce\u001b[39m ==> ENTITY: \u001b[32mFrank Bunce\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZinzan[0.046]; Andrew[0.045]; Mehrtens[0.045]; Zealand[0.045]; James[0.044]; Walter[0.044]; Michael[0.044]; Gary[0.043]; Ieremia[0.043]; Glenn[0.043]; Ian[0.043]; Henry[0.043]; Mark[0.043]; Jeff[0.043]; Honiball[0.043]; Garry[0.043]; Justin[0.043]; Hurter[0.043]; Joel[0.043]; Wayne[0.043]; Glen[0.043]; Dalton[0.042]; Craig[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Springboks\u001b[39m ==> ENTITY: \u001b[32mSouth Africa national rugby union team\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.055]; series[0.044]; season[0.043]; South[0.043]; Zealand[0.043]; Zealand[0.043]; Saturday[0.042]; fifth[0.042]; Johannesburg[0.042]; Park[0.042]; Ellis[0.042]; Union[0.042]; Africa[0.041]; champions[0.041]; highest[0.040]; successive[0.040]; successive[0.040]; world[0.040]; victory[0.040]; victories[0.039]; run[0.039]; total[0.039]; tries[0.039]; Blacks[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mehrtens\u001b[39m ==> ENTITY: \u001b[32mAndrew Mehrtens\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.228:0.228[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nscrum[0.044]; Springboks[0.044]; Springboks[0.044]; Mehrtens[0.043]; Mehrtens[0.043]; time[0.042]; hooker[0.042]; final[0.042]; points[0.041]; Honiball[0.041]; Honiball[0.041]; score[0.041]; Blacks[0.041]; missed[0.041]; missed[0.041]; try[0.041]; half[0.041]; half[0.041]; half[0.041]; half[0.041]; half[0.041]; Welsh[0.041]; ahead[0.041]; point[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Zealand ==> ENTITY: \u001b[32mNew Zealand national rugby union team\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.182:0.182[0]; log p(e|m)= -3.817:-3.817[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.048]; recording[0.046]; Springboks[0.044]; scrum[0.043]; Springbok[0.043]; series[0.042]; series[0.042]; flanker[0.042]; Zealand[0.042]; second[0.042]; world[0.042]; fifth[0.041]; Johannesburg[0.041]; African[0.041]; highest[0.041]; South[0.040]; South[0.040]; Ellis[0.040]; season[0.040]; Africa[0.040]; opening[0.040]; successive[0.040]; successive[0.040]; Park[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = All Black\u001b[39m ==> ENTITY: \u001b[32mNew Zealand national rugby union team\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpringboks[0.050]; Springboks[0.050]; scrum[0.050]; hooker[0.048]; Welsh[0.047]; Mehrtens[0.046]; South[0.046]; Honiball[0.046]; second[0.045]; time[0.044]; time[0.044]; ruck[0.044]; Little[0.044]; Blacks[0.044]; half[0.044]; half[0.044]; comeback[0.044]; Africa[0.044]; referee[0.044]; try[0.043]; try[0.043]; try[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Joost van der Westhuizen\u001b[39m ==> ENTITY: \u001b[32mJoost van der Westhuizen\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpringboks[0.045]; injury[0.044]; captain[0.043]; scrum[0.043]; Venter[0.043]; Honiball[0.043]; Mehrtens[0.042]; Mehrtens[0.042]; Hurter[0.042]; South[0.042]; Danie[0.041]; Dawie[0.041]; Zinzan[0.041]; minutes[0.041]; Swart[0.041]; kicked[0.041]; Joubert[0.041]; Joubert[0.041]; Kobus[0.041]; Africa[0.041]; Ieremia[0.041]; Pieter[0.040]; Heerden[0.040]; Marius[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gary Teichmann\u001b[39m ==> ENTITY: \u001b[32mGary Teichmann\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncaptain[0.045]; Honiball[0.044]; South[0.044]; Swart[0.043]; Wayne[0.043]; Springboks[0.043]; Zinzan[0.043]; Venter[0.042]; Stransky[0.042]; Westhuizen[0.042]; scrum[0.042]; Japie[0.042]; Hurter[0.041]; Andrew[0.041]; Kobus[0.041]; Mark[0.041]; Glenn[0.041]; Mehrtens[0.040]; Mehrtens[0.040]; Michael[0.040]; Andre[0.040]; Andre[0.040]; Taylor[0.040]; Dawie[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Walter Little\u001b[39m ==> ENTITY: \u001b[32mWalter Little (rugby player)\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nflanker[0.046]; Henry[0.044]; second[0.044]; Marshall[0.043]; scrum[0.043]; scrum[0.043]; Andrew[0.043]; Andrew[0.043]; Mehrtens[0.042]; Mehrtens[0.042]; Mehrtens[0.042]; Fullback[0.041]; Springboks[0.041]; South[0.041]; Springbok[0.041]; half[0.040]; half[0.040]; half[0.040]; half[0.040]; half[0.040]; half[0.040]; arrived[0.040]; late[0.040]; Honiball[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Glen Osborne ==> ENTITY: \u001b[32mGlen Osborne\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.277:-0.277[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZinzan[0.044]; Stransky[0.043]; Little[0.042]; Ieremia[0.042]; Craig[0.042]; Dalton[0.042]; Marshall[0.042]; Mehrtens[0.042]; Zealand[0.042]; James[0.042]; Hurter[0.042]; Jones[0.041]; Jones[0.041]; Taylor[0.041]; Ian[0.041]; Andrew[0.041]; Cullen[0.041]; Garry[0.041]; Wilson[0.041]; Bunce[0.041]; Michael[0.041]; Walter[0.041]; Joel[0.041]; Wayne[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andrew Mehrtens\u001b[39m ==> ENTITY: \u001b[32mAndrew Mehrtens\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZinzan[0.046]; Garry[0.043]; Wayne[0.043]; Honiball[0.043]; Zealand[0.043]; Ieremia[0.042]; Jones[0.042]; Jones[0.042]; Ian[0.042]; Hurter[0.041]; Venter[0.041]; Glenn[0.041]; Gary[0.041]; captain[0.041]; captain[0.041]; Cullen[0.041]; Marshall[0.041]; Stransky[0.041]; Justin[0.041]; Mark[0.041]; Craig[0.041]; Westhuizen[0.040]; Little[0.040]; Danie[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ellis Park\u001b[39m ==> ENTITY: \u001b[32mEllis Park Stadium\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.048]; South[0.045]; South[0.045]; Johannesburg[0.044]; Springboks[0.043]; Saturday[0.042]; Zealand[0.042]; Zealand[0.042]; historic[0.042]; Springbok[0.041]; opening[0.041]; making[0.041]; world[0.041]; scrum[0.041]; African[0.040]; total[0.040]; series[0.040]; series[0.040]; run[0.040]; season[0.040]; flanker[0.040]; Union[0.040]; late[0.040]; champions[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Springboks\u001b[39m ==> ENTITY: \u001b[32mSouth Africa national rugby union team\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpringboks[0.049]; Welsh[0.048]; day[0.045]; Honiball[0.045]; Honiball[0.045]; Bevan[0.045]; hooker[0.044]; Mehrtens[0.043]; Mehrtens[0.043]; Mehrtens[0.043]; winger[0.043]; final[0.043]; comeback[0.043]; Westhuizen[0.042]; referee[0.042]; Pieter[0.042]; inside[0.042]; Joubert[0.042]; Joubert[0.042]; Joubert[0.042]; Joubert[0.042]; time[0.042]; penalties[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = All Black\u001b[39m ==> ENTITY: \u001b[32mNew Zealand national rugby union team\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.051]; Springboks[0.046]; Zealand[0.045]; Zealand[0.045]; recording[0.044]; South[0.042]; Ellis[0.042]; season[0.042]; Saturday[0.042]; Park[0.041]; champions[0.041]; series[0.041]; tries[0.040]; Blacks[0.040]; world[0.040]; Johannesburg[0.040]; successive[0.040]; successive[0.040]; Africa[0.040]; highest[0.040]; Union[0.039]; total[0.039]; late[0.039]; pride[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Springboks\u001b[39m ==> ENTITY: \u001b[32mSouth Africa national rugby union team\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpringboks[0.047]; scrum[0.045]; Honiball[0.043]; Honiball[0.043]; South[0.043]; ruck[0.043]; second[0.042]; hooker[0.042]; Mehrtens[0.041]; Danie[0.041]; Africa[0.041]; Venter[0.041]; comeback[0.041]; injury[0.041]; Westhuizen[0.041]; Westhuizen[0.041]; Westhuizen[0.041]; Stransky[0.041]; Pieter[0.041]; Little[0.040]; Joubert[0.040]; Joubert[0.040]; Joubert[0.040]; Joubert[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Joost van der Westhuizen\u001b[39m ==> ENTITY: \u001b[32mJoost van der Westhuizen\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.135:0.135[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpringbok[0.045]; Fullback[0.044]; scrum[0.043]; scrum[0.043]; Venter[0.043]; South[0.042]; flanker[0.042]; pride[0.042]; pride[0.042]; African[0.041]; minutes[0.041]; minutes[0.041]; try[0.041]; second[0.041]; Joubert[0.041]; emphatic[0.041]; victory[0.041]; triumph[0.041]; season[0.041]; final[0.040]; Saturday[0.040]; Zealand[0.040]; Zealand[0.040]; champions[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mark Andrews ==> ENTITY: \u001b[32mMark Andrews (rugby player)\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.178:0.178[0]; log p(e|m)= -1.143:-1.143[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncaptain[0.045]; captain[0.045]; Andrew[0.045]; Honiball[0.045]; Michael[0.044]; Robin[0.044]; Stransky[0.044]; Sean[0.044]; Zinzan[0.044]; scrum[0.043]; Marius[0.043]; South[0.043]; Venter[0.043]; Justin[0.043]; Justin[0.043]; Justin[0.043]; Africa[0.043]; Dawie[0.043]; Ian[0.043]; Joel[0.042]; Swart[0.042]; Kobus[0.042]; Japie[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fritz van Heerden\u001b[39m ==> ENTITY: \u001b[32mFritz van Heerden\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nscrum[0.045]; Stransky[0.044]; Joel[0.043]; Henry[0.043]; South[0.042]; Swart[0.042]; Wilson[0.042]; Zinzan[0.042]; Jeff[0.042]; Danie[0.042]; Dawie[0.042]; Hurter[0.041]; Mark[0.041]; Honiball[0.041]; Mehrtens[0.041]; Mehrtens[0.041]; Taylor[0.041]; Craig[0.041]; Glen[0.041]; Jones[0.041]; Jones[0.041]; Justin[0.041]; Justin[0.041]; Justin[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Justin Marshall\u001b[39m ==> ENTITY: \u001b[32mJustin Marshall\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.221:0.221[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nscrum[0.044]; scrum[0.044]; Mehrtens[0.044]; Mehrtens[0.044]; Mehrtens[0.044]; flanker[0.043]; Blacks[0.043]; Springbok[0.042]; second[0.042]; Springboks[0.042]; Springboks[0.042]; final[0.041]; Fullback[0.041]; Andrew[0.041]; Andrew[0.041]; ahead[0.041]; Henry[0.041]; Walter[0.040]; half[0.040]; half[0.040]; half[0.040]; half[0.040]; half[0.040]; half[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Justin Marshall\u001b[39m ==> ENTITY: \u001b[32mJustin Marshall\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMehrtens[0.044]; Wayne[0.044]; Ian[0.043]; Zinzan[0.043]; Craig[0.042]; Gary[0.042]; Glenn[0.042]; James[0.042]; Jones[0.042]; Jones[0.042]; captain[0.042]; captain[0.042]; Garry[0.042]; Zealand[0.042]; Sean[0.041]; Brown[0.041]; Andrew[0.041]; Mark[0.041]; Henry[0.041]; Walter[0.040]; Taylor[0.040]; Michael[0.040]; Wilson[0.040]; Honiball[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zinzan Brooke\u001b[39m ==> ENTITY: \u001b[32mZinzan Brooke\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.137:0.137[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCraig[0.044]; Wayne[0.043]; Andrew[0.043]; Brooke[0.043]; Ian[0.043]; Gary[0.043]; Mark[0.042]; James[0.042]; Ieremia[0.042]; Sean[0.042]; Garry[0.042]; Michael[0.042]; Glenn[0.042]; captain[0.041]; captain[0.041]; Robin[0.041]; Henry[0.041]; Venter[0.041]; Glen[0.041]; Zealand[0.041]; Mehrtens[0.040]; Frank[0.040]; Jeff[0.040]; Honiball[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Welsh ==> ENTITY: \u001b[32mWales\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.061:0.061[0]; log p(e|m)= -1.079:-1.079[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpringboks[0.043]; Springboks[0.043]; Springboks[0.043]; second[0.043]; hooker[0.042]; missed[0.042]; Andrew[0.042]; Sean[0.042]; minute[0.041]; Honiball[0.041]; Honiball[0.041]; goal[0.041]; half[0.041]; half[0.041]; half[0.041]; time[0.041]; minutes[0.041]; minutes[0.041]; scored[0.041]; ahead[0.041]; comeback[0.041]; winger[0.041]; Henry[0.041]; appeared[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = All Black\u001b[39m ==> ENTITY: \u001b[32mNew Zealand national rugby union team\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nscrum[0.047]; scrum[0.047]; Springbok[0.046]; flanker[0.046]; Zealand[0.046]; Zealand[0.046]; recording[0.045]; African[0.044]; South[0.043]; South[0.043]; Ellis[0.043]; season[0.043]; Fullback[0.043]; Saturday[0.043]; Park[0.042]; second[0.042]; final[0.042]; champions[0.042]; series[0.042]; series[0.042]; tries[0.041]; tries[0.041]; Little[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = All Blacks\u001b[39m ==> ENTITY: \u001b[32mNew Zealand national rugby union team\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpringboks[0.046]; Springboks[0.046]; scrum[0.046]; second[0.043]; Honiball[0.042]; hooker[0.042]; South[0.042]; captain[0.041]; comeback[0.041]; outside[0.041]; Henry[0.041]; Mehrtens[0.041]; Africa[0.041]; time[0.041]; time[0.041]; break[0.041]; score[0.041]; score[0.041]; try[0.040]; try[0.040]; try[0.040]; Wayne[0.040]; half[0.040]; half[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rugby Union\u001b[39m ==> ENTITY: \u001b[32mRugby union\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpringboks[0.047]; Zealand[0.044]; Zealand[0.044]; South[0.044]; Ellis[0.043]; season[0.043]; Park[0.042]; champions[0.042]; Saturday[0.042]; victories[0.041]; highest[0.041]; tries[0.041]; victory[0.041]; Andy[0.041]; Johannesburg[0.041]; fifth[0.041]; successive[0.041]; successive[0.041]; Blacks[0.040]; series[0.040]; scored[0.040]; defeat[0.040]; world[0.040]; pride[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Johannesburg ==> ENTITY: \u001b[32mJohannesburg\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.095:-0.095[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.047]; Park[0.046]; South[0.045]; Saturday[0.043]; fifth[0.043]; series[0.042]; series[0.042]; Springboks[0.042]; Ellis[0.042]; Zealand[0.042]; Zealand[0.042]; season[0.041]; world[0.041]; historic[0.041]; scored[0.041]; victory[0.041]; Blacks[0.040]; Africa[0.040]; highest[0.040]; triumph[0.040]; Union[0.040]; arrived[0.039]; victories[0.039]; crowd[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kobus Wiese\u001b[39m ==> ENTITY: \u001b[32mKobus Wiese\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHoniball[0.045]; Venter[0.044]; scrum[0.044]; Hurter[0.044]; South[0.043]; Africa[0.043]; Japie[0.042]; Joubert[0.042]; Dawie[0.042]; captain[0.042]; captain[0.042]; Mehrtens[0.041]; Danie[0.041]; Swart[0.041]; Stransky[0.041]; Zinzan[0.041]; Kronfeld[0.040]; Theron[0.040]; ruck[0.040]; line[0.040]; Fritz[0.040]; Heerden[0.040]; Westhuizen[0.040]; Pieter[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marius Hurter\u001b[39m ==> ENTITY: \u001b[32mMarius Hurter\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.045]; Honiball[0.045]; scrum[0.044]; Japie[0.043]; Dawie[0.043]; Danie[0.043]; Africa[0.043]; Kobus[0.043]; Venter[0.042]; captain[0.041]; captain[0.041]; Heerden[0.041]; Swart[0.041]; Westhuizen[0.041]; Mehrtens[0.041]; Joubert[0.041]; ruck[0.040]; Pieter[0.040]; Stransky[0.040]; Andre[0.040]; Andre[0.040]; Zinzan[0.040]; Teichmann[0.040]; Zealand[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andrew Mehrtens\u001b[39m ==> ENTITY: \u001b[32mAndrew Mehrtens\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.230:0.230[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nscrum[0.044]; Springboks[0.044]; Springboks[0.044]; Mehrtens[0.043]; Mehrtens[0.043]; flanker[0.043]; second[0.042]; final[0.042]; Fullback[0.041]; points[0.041]; Honiball[0.041]; Blacks[0.041]; missed[0.041]; missed[0.041]; try[0.041]; half[0.041]; half[0.041]; half[0.041]; half[0.041]; half[0.041]; ahead[0.041]; point[0.040]; scoring[0.040]; Andrew[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Josh Kronfeld\u001b[39m ==> ENTITY: \u001b[32mJosh Kronfeld\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.121:0.121[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIeremia[0.046]; Zinzan[0.045]; Zealand[0.044]; Mehrtens[0.043]; Hurter[0.042]; Ian[0.042]; Venter[0.042]; Michael[0.042]; Westhuizen[0.042]; Gary[0.041]; Andrew[0.041]; Sean[0.041]; Wayne[0.041]; Mark[0.041]; Glenn[0.041]; captain[0.041]; captain[0.041]; Honiball[0.041]; Little[0.041]; Joost[0.041]; Justin[0.040]; Craig[0.040]; der[0.040]; Heerden[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Michael Jones ==> ENTITY: \u001b[32mMichael Jones (rugby union)\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.127:0.127[0]; log p(e|m)= -1.002:-1.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJames[0.043]; Zinzan[0.043]; Wayne[0.043]; Jones[0.043]; Ieremia[0.043]; Glenn[0.042]; Zealand[0.042]; captain[0.042]; captain[0.042]; Sean[0.042]; Cullen[0.042]; Garry[0.042]; Glen[0.041]; Fitzpatrick[0.041]; Mark[0.041]; Wilson[0.041]; Taylor[0.041]; Marshall[0.041]; Gary[0.041]; Craig[0.041]; Andrews[0.041]; Hurter[0.041]; Ian[0.040]; Christian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Honiball\u001b[39m ==> ENTITY: \u001b[32mHenry Honiball\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpringboks[0.044]; Springboks[0.044]; Springboks[0.044]; hooker[0.043]; Andrew[0.042]; Westhuizen[0.042]; Westhuizen[0.042]; Joubert[0.042]; Joubert[0.042]; Joubert[0.042]; Joubert[0.042]; Joubert[0.042]; Pieter[0.042]; quick[0.041]; try[0.041]; try[0.041]; try[0.041]; Mehrtens[0.041]; Mehrtens[0.041]; second[0.040]; converted[0.040]; winger[0.040]; half[0.040]; half[0.040]; \t\n\r [=====================>........................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 10s935ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2373/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 958testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jonas Bjorkman\u001b[39m ==> ENTITY: \u001b[32mJonas Björkman\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.277:0.277[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPioline[0.046]; Sjeng[0.044]; singles[0.043]; singles[0.043]; singles[0.043]; singles[0.043]; Ivanisevic[0.043]; Goran[0.042]; Rikl[0.042]; Mantilla[0.042]; Corretja[0.041]; round[0.041]; round[0.041]; round[0.041]; round[0.041]; Philippoussis[0.041]; Switzerland[0.040]; second[0.040]; second[0.040]; second[0.040]; Czech[0.040]; Czech[0.040]; Sweden[0.040]; Sweden[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.048]; Switzerland[0.045]; Croatia[0.043]; Romania[0.042]; Spain[0.042]; Spain[0.042]; Spain[0.042]; Spain[0.042]; Netherlands[0.042]; second[0.041]; second[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; Goran[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.241:0.241[0]; local()= 0.083:0.083[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArgentina[0.045]; Pioline[0.045]; round[0.044]; round[0.044]; round[0.044]; round[0.044]; round[0.044]; round[0.044]; Slovakia[0.044]; second[0.044]; second[0.044]; Switzerland[0.043]; bat[0.043]; Haarhuis[0.043]; singles[0.043]; singles[0.043]; singles[0.043]; singles[0.043]; singles[0.043]; Germany[0.043]; Germany[0.043]; Spain[0.043]; Spain[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Doug Flach\u001b[39m ==> ENTITY: \u001b[32mDoug Flach\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIvanisevic[0.043]; Davenport[0.043]; singles[0.042]; singles[0.042]; singles[0.042]; singles[0.042]; round[0.042]; round[0.042]; round[0.042]; round[0.042]; Tennis[0.042]; Tennis[0.042]; Tennis[0.042]; Rikl[0.042]; Mark[0.042]; David[0.042]; David[0.042]; Bjorkman[0.041]; Philippoussis[0.041]; Mantilla[0.041]; Tim[0.041]; Open[0.041]; Open[0.041]; Scott[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Romania ==> ENTITY: \u001b[32mRomania\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.154:0.154[0]; log p(e|m)= -0.212:-0.212[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArgentina[0.052]; Men[0.052]; Men[0.052]; Men[0.052]; France[0.051]; France[0.051]; France[0.051]; Switzerland[0.050]; Germany[0.050]; Spain[0.049]; Spain[0.049]; Spain[0.049]; Spain[0.049]; Republic[0.049]; Republic[0.049]; Republic[0.049]; Women[0.049]; Women[0.049]; Pioline[0.048]; Russia[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cedric Pioline\u001b[39m ==> ENTITY: \u001b[32mCédric Pioline\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.228:0.228[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHaarhuis[0.071]; round[0.067]; round[0.067]; round[0.067]; round[0.067]; round[0.067]; singles[0.067]; singles[0.067]; singles[0.067]; singles[0.067]; singles[0.067]; Corretja[0.066]; Mantilla[0.065]; Davenport[0.065]; Jiri[0.064]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ines Gorrochategui\u001b[39m ==> ENTITY: \u001b[32mInés Gorrochategui\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArgentina[0.045]; Croatia[0.044]; Rikl[0.043]; Ivanisevic[0.043]; singles[0.042]; singles[0.042]; singles[0.042]; Tennis[0.042]; Tennis[0.042]; Tennis[0.042]; Spain[0.042]; Championships[0.041]; Sjeng[0.041]; Schalken[0.041]; Netherlands[0.041]; Goran[0.041]; David[0.041]; Australia[0.041]; Australia[0.041]; Tillstrom[0.041]; Sweden[0.040]; Testud[0.040]; Mark[0.040]; number[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.239:0.239[0]; local()= 0.070:0.070[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScott[0.049]; Russia[0.048]; David[0.048]; David[0.048]; Mark[0.047]; Croatia[0.047]; Tennis[0.046]; Tennis[0.046]; Tennis[0.046]; second[0.046]; second[0.046]; Draper[0.045]; Doug[0.044]; round[0.044]; round[0.044]; round[0.044]; round[0.044]; Championships[0.044]; Sweden[0.044]; Sweden[0.044]; Alexander[0.044]; Andrei[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gabriela Sabatini\u001b[39m ==> ENTITY: \u001b[32mGabriela Sabatini\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.218:0.218[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHaarhuis[0.044]; Pioline[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; singles[0.041]; singles[0.041]; singles[0.041]; singles[0.041]; singles[0.041]; Jiri[0.041]; round[0.041]; round[0.041]; round[0.041]; round[0.041]; round[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Argentina ==> ENTITY: \u001b[32mArgentina\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.160:0.160[0]; log p(e|m)= -0.300:-0.300[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.051]; Spain[0.051]; Spain[0.051]; Switzerland[0.049]; Australia[0.048]; Roberto[0.048]; France[0.047]; France[0.047]; Slovakia[0.047]; Germany[0.047]; Germany[0.047]; second[0.047]; second[0.047]; second[0.047]; Romania[0.047]; Netherlands[0.047]; Women[0.046]; Women[0.046]; Sweden[0.046]; Sweden[0.046]; Pioline[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.085:0.085[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArgentina[0.056]; Switzerland[0.055]; Spain[0.055]; Slovakia[0.054]; Sweden[0.053]; Sweden[0.053]; singles[0.052]; singles[0.052]; singles[0.052]; singles[0.052]; singles[0.052]; singles[0.052]; Italy[0.052]; Germany[0.052]; Germany[0.052]; second[0.052]; second[0.052]; second[0.052]; Netherlands[0.051]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Felix Mantilla\u001b[39m ==> ENTITY: \u001b[32mFélix Mantilla Botella\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.248:0.248[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPioline[0.053]; round[0.052]; round[0.052]; round[0.052]; round[0.052]; round[0.052]; Ivanisevic[0.051]; Sjeng[0.051]; Mark[0.050]; Philippoussis[0.050]; Spain[0.050]; Spain[0.050]; David[0.049]; David[0.049]; Rikl[0.049]; Ines[0.049]; second[0.048]; second[0.048]; second[0.048]; Goran[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bohdan Ulihrach\u001b[39m ==> ENTITY: \u001b[32mBohdan Ulihrach\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHaarhuis[0.044]; round[0.043]; round[0.043]; round[0.043]; round[0.043]; round[0.043]; Sampras[0.042]; Czech[0.041]; Czech[0.041]; Czech[0.041]; Bernd[0.041]; Jiri[0.041]; second[0.041]; second[0.041]; second[0.041]; Michael[0.041]; singles[0.041]; singles[0.041]; singles[0.041]; singles[0.041]; singles[0.041]; Randriantefy[0.041]; Edberg[0.041]; leg[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Monica Seles\u001b[39m ==> ENTITY: \u001b[32mMonica Seles\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.248:0.248[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHaarhuis[0.045]; second[0.044]; second[0.044]; Randriantefy[0.043]; singles[0.042]; singles[0.042]; singles[0.042]; singles[0.042]; round[0.042]; round[0.042]; round[0.042]; round[0.042]; Ulihrach[0.042]; Edberg[0.041]; Johansson[0.041]; Andrea[0.041]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; Lisa[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.049]; Argentina[0.043]; Tennis[0.042]; Tennis[0.042]; Tennis[0.042]; Croatia[0.042]; Britain[0.042]; second[0.042]; second[0.042]; Open[0.042]; Open[0.042]; South[0.042]; Mark[0.041]; singles[0.041]; singles[0.041]; singles[0.041]; National[0.041]; Spain[0.041]; Sweden[0.041]; Sweden[0.041]; David[0.040]; David[0.040]; Scott[0.040]; Philippoussis[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Africa ==> ENTITY: \u001b[32mSouth Africa\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.263:-0.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.045]; Australia[0.045]; Africa[0.043]; Men[0.043]; Men[0.043]; South[0.043]; Switzerland[0.043]; Sweden[0.042]; Sweden[0.042]; Spain[0.042]; Spain[0.042]; Spain[0.042]; Spain[0.042]; France[0.041]; France[0.041]; France[0.041]; Romania[0.041]; Women[0.041]; Women[0.041]; second[0.040]; second[0.040]; Netherlands[0.040]; Coetzer[0.040]; Britain[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lindsay Davenport\u001b[39m ==> ENTITY: \u001b[32mLindsay Davenport\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.256:0.256[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPioline[0.065]; Rikl[0.064]; Mantilla[0.063]; round[0.063]; round[0.063]; round[0.063]; round[0.063]; second[0.063]; second[0.063]; Corretja[0.062]; Sjeng[0.062]; singles[0.062]; singles[0.062]; singles[0.062]; singles[0.062]; Philippoussis[0.061]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.082:0.082[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArgentina[0.059]; Switzerland[0.058]; Slovakia[0.057]; Sweden[0.056]; Sweden[0.056]; singles[0.055]; singles[0.055]; singles[0.055]; singles[0.055]; singles[0.055]; singles[0.055]; Italy[0.055]; Germany[0.055]; Germany[0.055]; second[0.055]; second[0.055]; second[0.055]; Netherlands[0.053]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.174:0.174[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.053]; Germany[0.053]; Argentina[0.050]; Slovakia[0.049]; Italy[0.048]; Republic[0.047]; Republic[0.047]; Republic[0.047]; Sweden[0.047]; Sweden[0.047]; Sweden[0.047]; Netherlands[0.046]; round[0.046]; round[0.046]; round[0.046]; round[0.046]; round[0.046]; round[0.046]; second[0.046]; second[0.046]; second[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.049]; Argentina[0.047]; Switzerland[0.046]; Pioline[0.044]; Haarhuis[0.044]; singles[0.043]; singles[0.043]; singles[0.043]; singles[0.043]; singles[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Germany[0.043]; Australia[0.043]; Republic[0.042]; Republic[0.042]; Romania[0.042]; second[0.042]; second[0.042]; second[0.042]; Netherlands[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Roberto Carretero\u001b[39m ==> ENTITY: \u001b[32mRoberto Carretero\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nround[0.069]; round[0.069]; round[0.069]; round[0.069]; round[0.069]; Pioline[0.068]; Haarhuis[0.068]; Jiri[0.067]; Corretja[0.066]; Bjorkman[0.065]; Spain[0.065]; Spain[0.065]; Spain[0.065]; Paul[0.064]; Alexander[0.064]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sjeng Schalken\u001b[39m ==> ENTITY: \u001b[32mSjeng Schalken\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.260:0.260[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsingles[0.057]; singles[0.057]; singles[0.057]; singles[0.057]; second[0.056]; second[0.056]; Rikl[0.056]; Tennis[0.056]; Tennis[0.056]; Ivanisevic[0.055]; Open[0.055]; Open[0.055]; Philippoussis[0.054]; round[0.054]; round[0.054]; round[0.054]; round[0.054]; Testud[0.053]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pete Sampras\u001b[39m ==> ENTITY: \u001b[32mPete Sampras\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.251:0.251[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHaarhuis[0.051]; Pioline[0.050]; Corretja[0.047]; Michael[0.046]; Ulihrach[0.045]; singles[0.045]; singles[0.045]; singles[0.045]; singles[0.045]; singles[0.045]; beat[0.045]; beat[0.045]; beat[0.045]; beat[0.045]; beat[0.045]; beat[0.045]; beat[0.045]; beat[0.045]; beat[0.045]; beat[0.045]; Paul[0.045]; Alex[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.087:0.087[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArgentina[0.056]; Switzerland[0.055]; Spain[0.055]; Spain[0.055]; Slovakia[0.054]; Sweden[0.053]; Sweden[0.053]; singles[0.052]; singles[0.052]; singles[0.052]; singles[0.052]; singles[0.052]; Germany[0.052]; Germany[0.052]; second[0.052]; second[0.052]; second[0.052]; Romania[0.051]; Netherlands[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Goran Ivanisevic\u001b[39m ==> ENTITY: \u001b[32mGoran Ivanišević\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.050]; Tennis[0.050]; Tennis[0.050]; singles[0.049]; singles[0.049]; singles[0.049]; Mark[0.048]; second[0.048]; second[0.048]; Rikl[0.048]; round[0.048]; round[0.048]; round[0.048]; Open[0.047]; Open[0.047]; Philippoussis[0.047]; David[0.046]; David[0.046]; Championships[0.046]; Bjorkman[0.045]; Mantilla[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.151:0.151[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.050]; Argentina[0.045]; Russia[0.043]; Russia[0.043]; Australia[0.043]; Australia[0.043]; Romania[0.042]; second[0.042]; second[0.042]; second[0.042]; Pioline[0.041]; Britain[0.041]; Croatia[0.041]; Mantilla[0.041]; David[0.041]; David[0.041]; Rikl[0.040]; France[0.040]; France[0.040]; France[0.040]; France[0.040]; Sweden[0.040]; Sweden[0.040]; Ivanisevic[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dally Randriantefy\u001b[39m ==> ENTITY: \u001b[32mDally Randriantefy\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHaarhuis[0.045]; round[0.045]; round[0.045]; round[0.045]; round[0.045]; singles[0.043]; singles[0.043]; singles[0.043]; singles[0.043]; second[0.043]; second[0.043]; Raymond[0.041]; Ulihrach[0.041]; Bernd[0.041]; Korda[0.040]; leg[0.040]; Johansson[0.040]; Stefan[0.040]; retired[0.039]; Edberg[0.039]; Australia[0.039]; Seles[0.038]; Petr[0.038]; Hendrik[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Todd Martin\u001b[39m ==> ENTITY: \u001b[32mTodd Martin\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.227:0.227[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nround[0.046]; round[0.046]; round[0.046]; round[0.046]; Edberg[0.044]; singles[0.044]; singles[0.044]; singles[0.044]; singles[0.044]; second[0.044]; second[0.044]; Bernd[0.044]; Randriantefy[0.043]; Ulihrach[0.043]; leg[0.043]; Stefan[0.043]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; Seles[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.114:0.114[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.050]; Slovakia[0.044]; Sweden[0.044]; Sweden[0.044]; Italy[0.043]; Republic[0.043]; Republic[0.043]; Czech[0.043]; Czech[0.043]; women[0.041]; second[0.041]; second[0.041]; Johansson[0.041]; injury[0.041]; men[0.041]; leg[0.040]; round[0.040]; round[0.040]; round[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.052]; Spain[0.052]; Spain[0.052]; Spain[0.052]; Spain[0.052]; Britain[0.050]; Sweden[0.050]; Sweden[0.050]; Sweden[0.050]; Russia[0.050]; Russia[0.050]; singles[0.049]; singles[0.049]; singles[0.049]; singles[0.049]; Germany[0.049]; second[0.049]; second[0.049]; Romania[0.049]; Netherlands[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Anne-Gaelle Sidot\u001b[39m ==> ENTITY: \u001b[32mAnne-Gaëlle Sidot\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsingles[0.058]; singles[0.058]; singles[0.058]; singles[0.058]; Rikl[0.056]; round[0.055]; round[0.055]; round[0.055]; round[0.055]; Pioline[0.055]; France[0.055]; France[0.055]; France[0.055]; Bjorkman[0.055]; Women[0.055]; Women[0.055]; Corretja[0.054]; Schalken[0.053]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.154:0.154[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.048]; Sweden[0.048]; Slovakia[0.044]; Argentina[0.043]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Italy[0.042]; Netherlands[0.041]; second[0.041]; second[0.041]; second[0.041]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; leg[0.040]; Edberg[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.086:0.086[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArgentina[0.045]; Open[0.044]; Open[0.044]; Spain[0.043]; Championships[0.043]; number[0.042]; Britain[0.042]; Sweden[0.042]; Sweden[0.042]; singles[0.041]; singles[0.041]; singles[0.041]; singles[0.041]; second[0.041]; second[0.041]; round[0.041]; round[0.041]; round[0.041]; round[0.041]; National[0.041]; Australia[0.040]; Australia[0.040]; Netherlands[0.040]; Croatia[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.185:0.185[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.056]; Spain[0.056]; Argentina[0.051]; second[0.049]; second[0.049]; second[0.049]; Switzerland[0.048]; Romania[0.047]; Haarhuis[0.047]; Pioline[0.047]; Germany[0.046]; Martinez[0.046]; Roberto[0.046]; David[0.046]; Netherlands[0.045]; Corretja[0.045]; France[0.045]; France[0.045]; Sweden[0.045]; Sweden[0.045]; Sweden[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alex Corretja\u001b[39m ==> ENTITY: \u001b[32mÀlex Corretja\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.244:0.244[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPioline[0.050]; Haarhuis[0.050]; singles[0.049]; singles[0.049]; singles[0.049]; singles[0.049]; singles[0.049]; round[0.048]; round[0.048]; round[0.048]; round[0.048]; round[0.048]; second[0.047]; second[0.047]; Raymond[0.047]; Bjorkman[0.046]; Jiri[0.046]; Spain[0.046]; Spain[0.046]; Spain[0.046]; Sampras[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Netherlands ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.111:0.111[0]; log p(e|m)= -0.250:-0.250[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.044]; Tennis[0.044]; Australia[0.043]; Australia[0.043]; Spain[0.043]; Spain[0.043]; Argentina[0.043]; Women[0.043]; Women[0.043]; Sweden[0.042]; Sweden[0.042]; Britain[0.041]; Croatia[0.041]; Men[0.041]; Men[0.041]; France[0.041]; France[0.041]; France[0.041]; Championships[0.040]; round[0.040]; round[0.040]; round[0.040]; round[0.040]; Sjeng[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jiri Novak\u001b[39m ==> ENTITY: \u001b[32mJiří Novák\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.264:0.264[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHaarhuis[0.045]; Pioline[0.044]; singles[0.044]; singles[0.044]; singles[0.044]; singles[0.044]; singles[0.044]; round[0.042]; round[0.042]; round[0.042]; round[0.042]; round[0.042]; Corretja[0.041]; second[0.041]; second[0.041]; second[0.041]; Ulihrach[0.041]; Raymond[0.040]; Sampras[0.040]; Roberto[0.040]; Czech[0.039]; Czech[0.039]; Czech[0.039]; Switzerland[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paul Haarhuis\u001b[39m ==> ENTITY: \u001b[32mPaul Haarhuis\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPioline[0.046]; Michael[0.043]; Sampras[0.043]; Pete[0.042]; Jiri[0.042]; Raymond[0.042]; Seles[0.042]; singles[0.041]; singles[0.041]; singles[0.041]; singles[0.041]; singles[0.041]; singles[0.041]; Corretja[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ukraine ==> ENTITY: \u001b[32mUkraine\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.354:-0.354[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSlovakia[0.052]; Argentina[0.048]; Australia[0.047]; Sweden[0.046]; Sweden[0.046]; Sweden[0.046]; Germany[0.046]; Germany[0.046]; women[0.046]; men[0.046]; men[0.046]; Czech[0.045]; Czech[0.045]; Czech[0.045]; Italy[0.045]; Martin[0.044]; Andrei[0.044]; Republic[0.044]; Republic[0.044]; Republic[0.044]; Bohdan[0.044]; Netherlands[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Alexander Volkov ==> ENTITY: \u001b[32mAlexander Vladimirovich Volkov\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.279:0.279[0]; log p(e|m)= -0.650:-0.650[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRikl[0.043]; Andrei[0.043]; round[0.042]; round[0.042]; round[0.042]; round[0.042]; round[0.042]; Mark[0.042]; Russia[0.042]; Russia[0.042]; Pioline[0.041]; Philippoussis[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andrea Gaudenzi\u001b[39m ==> ENTITY: \u001b[32mAndrea Gaudenzi\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.225:0.225[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nround[0.048]; round[0.048]; round[0.048]; round[0.048]; Seles[0.046]; singles[0.046]; singles[0.046]; singles[0.046]; Johansson[0.046]; Edberg[0.045]; Randriantefy[0.045]; Bernd[0.045]; Ulihrach[0.045]; Korda[0.044]; second[0.044]; second[0.044]; leg[0.044]; Italy[0.044]; Stefan[0.044]; Raymond[0.044]; Lisa[0.044]; Martin[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Stefan Edberg\u001b[39m ==> ENTITY: \u001b[32mStefan Edberg\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMartin[0.045]; round[0.045]; round[0.045]; round[0.045]; Randriantefy[0.045]; Johansson[0.045]; Ulihrach[0.045]; Seles[0.044]; Czech[0.043]; Czech[0.043]; singles[0.043]; singles[0.043]; singles[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; Korda[0.043]; Bernd[0.043]; Thomas[0.042]; leg[0.042]; Bohdan[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tim Henman\u001b[39m ==> ENTITY: \u001b[32mTim Henman\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.216:0.216[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.045]; Open[0.045]; Tennis[0.043]; Tennis[0.043]; Tennis[0.043]; round[0.042]; round[0.042]; round[0.042]; round[0.042]; David[0.042]; David[0.042]; Mark[0.042]; Philippoussis[0.042]; number[0.042]; Sjeng[0.042]; Ivanisevic[0.042]; Rikl[0.040]; singles[0.040]; singles[0.040]; singles[0.040]; singles[0.040]; Bjorkman[0.040]; Mantilla[0.040]; Spain[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.168:0.168[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.051]; France[0.051]; Argentina[0.049]; Switzerland[0.048]; Pioline[0.046]; singles[0.045]; singles[0.045]; singles[0.045]; singles[0.045]; Spain[0.045]; Spain[0.045]; Spain[0.045]; Spain[0.045]; Rikl[0.044]; Germany[0.044]; Australia[0.044]; Republic[0.044]; Republic[0.044]; Romania[0.044]; second[0.044]; second[0.044]; Netherlands[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Conchita Martinez\u001b[39m ==> ENTITY: \u001b[32mConchita Martínez\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.237:0.237[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsingles[0.047]; singles[0.047]; singles[0.047]; singles[0.047]; Rikl[0.047]; second[0.047]; second[0.047]; Pioline[0.046]; Corretja[0.046]; Mantilla[0.045]; beat[0.044]; beat[0.044]; beat[0.044]; beat[0.044]; beat[0.044]; beat[0.044]; beat[0.044]; beat[0.044]; beat[0.044]; beat[0.044]; beat[0.044]; beat[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.085:0.085[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.052]; Argentina[0.050]; singles[0.046]; singles[0.046]; singles[0.046]; Spain[0.046]; Rikl[0.045]; Australia[0.045]; Australia[0.045]; Open[0.045]; Open[0.045]; Republic[0.045]; second[0.045]; second[0.045]; Scott[0.045]; Netherlands[0.044]; Croatia[0.044]; Testud[0.044]; Tennis[0.044]; Tennis[0.044]; Tennis[0.044]; Britain[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Slovakia ==> ENTITY: \u001b[32mSlovakia\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.367:-0.367[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArgentina[0.048]; Italy[0.048]; Germany[0.046]; Germany[0.046]; Sweden[0.046]; Sweden[0.046]; second[0.046]; second[0.046]; second[0.046]; Czech[0.046]; Czech[0.046]; Czech[0.046]; Men[0.045]; beat[0.044]; beat[0.044]; beat[0.044]; beat[0.044]; beat[0.044]; beat[0.044]; beat[0.044]; beat[0.044]; beat[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.185:0.185[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.056]; Spain[0.056]; Argentina[0.051]; Switzerland[0.051]; France[0.049]; France[0.049]; second[0.048]; second[0.048]; Germany[0.047]; Romania[0.047]; Haarhuis[0.047]; Pioline[0.046]; Martinez[0.046]; Roberto[0.046]; David[0.046]; Netherlands[0.045]; Corretja[0.045]; Australia[0.045]; Sweden[0.045]; Sweden[0.045]; Sweden[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Switzerland ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.166:0.166[0]; log p(e|m)= -0.186:-0.186[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.047]; Sweden[0.047]; Argentina[0.047]; round[0.046]; round[0.046]; round[0.046]; round[0.046]; round[0.046]; second[0.046]; second[0.046]; Pioline[0.046]; Germany[0.046]; Romania[0.046]; Spain[0.045]; Spain[0.045]; Spain[0.045]; France[0.044]; France[0.044]; Netherlands[0.044]; Men[0.044]; Men[0.044]; Haarhuis[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.051:0.051[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.044]; Tennis[0.044]; Tennis[0.044]; Friday[0.044]; Friday[0.044]; David[0.044]; second[0.043]; Championships[0.043]; National[0.042]; round[0.042]; round[0.042]; Mark[0.042]; Britain[0.041]; Women[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; Open[0.040]; Open[0.040]; Scott[0.040]; number[0.040]; Argentina[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = David Rikl\u001b[39m ==> ENTITY: \u001b[32mDavid Rikl\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.044]; Open[0.044]; singles[0.043]; singles[0.043]; singles[0.043]; singles[0.043]; second[0.042]; second[0.042]; round[0.042]; round[0.042]; round[0.042]; round[0.042]; Championships[0.042]; Tennis[0.041]; Tennis[0.041]; Sjeng[0.041]; Ivanisevic[0.041]; Bjorkman[0.041]; David[0.041]; Scott[0.040]; Testud[0.040]; Mantilla[0.040]; Mark[0.039]; Henman[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Michael Tebbutt\u001b[39m ==> ENTITY: \u001b[32mMichael Tebbutt\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.058]; Haarhuis[0.052]; Jiri[0.052]; Paul[0.051]; Jan[0.051]; Pioline[0.051]; Thomas[0.050]; second[0.050]; second[0.050]; second[0.050]; Sampras[0.050]; Argentina[0.049]; Randriantefy[0.049]; Seles[0.048]; Petr[0.048]; Switzerland[0.048]; Hendrik[0.048]; Corretja[0.048]; Slovakia[0.048]; Ulihrach[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lisa Raymond\u001b[39m ==> ENTITY: \u001b[32mLisa Raymond\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.278:0.278[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHaarhuis[0.047]; singles[0.047]; singles[0.047]; singles[0.047]; singles[0.047]; singles[0.047]; singles[0.047]; round[0.047]; round[0.047]; round[0.047]; round[0.047]; round[0.047]; round[0.047]; Randriantefy[0.044]; second[0.044]; second[0.044]; second[0.044]; Andrea[0.044]; Seles[0.043]; Jiri[0.043]; Johansson[0.043]; Martin[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Croatia ==> ENTITY: \u001b[32mCroatia\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.350:-0.350[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.043]; Tennis[0.043]; Tennis[0.043]; Argentina[0.043]; Australia[0.043]; Australia[0.043]; Rikl[0.043]; singles[0.042]; singles[0.042]; singles[0.042]; Republic[0.042]; Championships[0.042]; Ivanisevic[0.041]; Spain[0.041]; Open[0.041]; Open[0.041]; France[0.041]; France[0.041]; round[0.041]; round[0.041]; round[0.041]; Men[0.040]; Men[0.040]; Netherlands[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.240:0.240[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScott[0.045]; Russia[0.045]; David[0.044]; David[0.044]; Mark[0.043]; Croatia[0.043]; Romania[0.042]; second[0.042]; second[0.042]; second[0.042]; Draper[0.041]; Doug[0.041]; round[0.041]; round[0.041]; round[0.041]; round[0.041]; Sweden[0.041]; Sweden[0.041]; Alexander[0.041]; Andrei[0.040]; singles[0.040]; singles[0.040]; singles[0.040]; singles[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bernd Karbacher\u001b[39m ==> ENTITY: \u001b[32mBernd Karbacher\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsingles[0.046]; singles[0.046]; singles[0.046]; Korda[0.045]; Czech[0.045]; Czech[0.045]; Ulihrach[0.044]; round[0.044]; round[0.044]; round[0.044]; Edberg[0.044]; Stefan[0.044]; Randriantefy[0.044]; Seles[0.043]; Petr[0.043]; leg[0.043]; Jan[0.043]; Germany[0.041]; Germany[0.041]; retired[0.041]; second[0.041]; second[0.041]; Hendrik[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech Republic ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.162:0.162[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSlovakia[0.046]; Italy[0.044]; Sweden[0.044]; Sweden[0.044]; Czech[0.044]; Czech[0.044]; second[0.044]; second[0.044]; second[0.044]; Martin[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; Germany[0.043]; Germany[0.043]; Republic[0.043]; Republic[0.043]; injury[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.049]; Spain[0.049]; Spain[0.049]; Argentina[0.045]; second[0.042]; second[0.042]; Switzerland[0.042]; Romania[0.041]; Pioline[0.041]; Germany[0.041]; Martinez[0.041]; Roberto[0.040]; Mantilla[0.040]; David[0.040]; David[0.040]; Rikl[0.040]; Netherlands[0.040]; Corretja[0.040]; France[0.040]; France[0.040]; France[0.040]; Sweden[0.039]; Sweden[0.039]; Sweden[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Helena Sukova\u001b[39m ==> ENTITY: \u001b[32mHelena Suková\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRikl[0.043]; Pioline[0.043]; singles[0.043]; singles[0.043]; singles[0.043]; singles[0.043]; singles[0.043]; Czech[0.043]; Czech[0.043]; Corretja[0.042]; Irina[0.042]; Bjorkman[0.041]; second[0.041]; second[0.041]; second[0.041]; Linda[0.041]; Mantilla[0.041]; Andrei[0.041]; Barbara[0.041]; Anne[0.040]; David[0.040]; David[0.040]; Sjeng[0.040]; Gabriela[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech Republic ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.162:0.162[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSlovakia[0.046]; Italy[0.044]; Sweden[0.044]; Sweden[0.044]; Czech[0.044]; Czech[0.044]; second[0.044]; second[0.044]; second[0.044]; Martin[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; Germany[0.043]; Germany[0.043]; Republic[0.043]; Republic[0.043]; injury[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mark Philippoussis\u001b[39m ==> ENTITY: \u001b[32mMark Philippoussis\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.222:0.222[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.048]; second[0.048]; Open[0.047]; Open[0.047]; round[0.046]; round[0.046]; round[0.046]; round[0.046]; Tennis[0.046]; Tennis[0.046]; Tennis[0.046]; David[0.045]; David[0.045]; Rikl[0.045]; Ivanisevic[0.044]; singles[0.044]; singles[0.044]; singles[0.044]; singles[0.044]; Testud[0.044]; Bjorkman[0.044]; Mantilla[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Linda Wild\u001b[39m ==> ENTITY: \u001b[32mLinda Wild\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHaarhuis[0.063]; round[0.063]; round[0.063]; round[0.063]; round[0.063]; round[0.063]; round[0.063]; singles[0.063]; singles[0.063]; singles[0.063]; singles[0.063]; singles[0.063]; singles[0.063]; Jiri[0.062]; Pioline[0.062]; Sampras[0.062]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArgentina[0.056]; Switzerland[0.055]; Spain[0.054]; Spain[0.054]; Spain[0.054]; Sweden[0.052]; Sweden[0.052]; singles[0.052]; singles[0.052]; singles[0.052]; singles[0.052]; singles[0.052]; singles[0.052]; Germany[0.052]; Germany[0.052]; second[0.052]; second[0.052]; second[0.052]; Romania[0.051]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Argentina ==> ENTITY: \u001b[32mArgentina\u001b[39m\t\nSCORES: global= 0.241:0.241[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.300:-0.300[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.054]; Men[0.053]; Men[0.053]; Championships[0.051]; Australia[0.051]; Australia[0.051]; France[0.050]; France[0.050]; second[0.050]; second[0.050]; Netherlands[0.049]; Tennis[0.049]; Tennis[0.049]; Tennis[0.049]; Women[0.049]; Sweden[0.049]; Croatia[0.049]; round[0.048]; round[0.048]; round[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech Republic ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRomania[0.043]; Sweden[0.043]; Sweden[0.043]; Argentina[0.043]; Czech[0.042]; second[0.042]; second[0.042]; Croatia[0.042]; Championships[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; Tennis[0.041]; Tennis[0.041]; Republic[0.041]; Open[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kimberly Po\u001b[39m ==> ENTITY: \u001b[32mKimberly Po\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHaarhuis[0.044]; singles[0.043]; singles[0.043]; singles[0.043]; singles[0.043]; singles[0.043]; singles[0.043]; Lisa[0.043]; Monica[0.042]; Andrea[0.042]; Randriantefy[0.042]; Jiri[0.042]; women[0.041]; round[0.041]; round[0.041]; round[0.041]; round[0.041]; round[0.041]; round[0.041]; Linda[0.041]; Stefan[0.040]; Raymond[0.040]; Todd[0.040]; Michael[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hendrik Dreekmann\u001b[39m ==> ENTITY: \u001b[32mHendrik Dreekmann\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHaarhuis[0.044]; round[0.043]; round[0.043]; round[0.043]; round[0.043]; round[0.043]; round[0.043]; singles[0.042]; singles[0.042]; singles[0.042]; singles[0.042]; singles[0.042]; singles[0.042]; Seles[0.042]; Edberg[0.041]; Bernd[0.041]; Jiri[0.041]; Sampras[0.041]; Stefan[0.040]; Ulihrach[0.040]; second[0.040]; second[0.040]; second[0.040]; Netherlands[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.S. Open Tennis Championships\u001b[39m ==> ENTITY: \u001b[32mUS Open (tennis)\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.049]; Tennis[0.049]; Open[0.044]; singles[0.043]; singles[0.043]; round[0.043]; round[0.043]; second[0.041]; Scott[0.041]; Rikl[0.041]; Championships[0.041]; Philippoussis[0.041]; Sjeng[0.041]; Mark[0.041]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; Ivanisevic[0.040]; David[0.040]; Men[0.040]; Henman[0.039]; Guy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.080:0.080[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSlovakia[0.044]; Sweden[0.043]; Sweden[0.043]; singles[0.042]; singles[0.042]; singles[0.042]; singles[0.042]; Italy[0.042]; Germany[0.042]; Germany[0.042]; second[0.042]; second[0.042]; round[0.042]; round[0.042]; round[0.042]; round[0.042]; Australia[0.041]; Netherlands[0.041]; Thomas[0.041]; Martin[0.041]; Women[0.040]; Michael[0.040]; Republic[0.040]; Republic[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mikael Tillstrom\u001b[39m ==> ENTITY: \u001b[32mMikael Tillström\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMark[0.054]; David[0.054]; David[0.054]; Sjeng[0.053]; Alex[0.053]; singles[0.053]; singles[0.053]; singles[0.053]; singles[0.053]; round[0.053]; round[0.053]; round[0.053]; round[0.053]; Rikl[0.053]; Ivanisevic[0.052]; Goran[0.052]; Pioline[0.052]; Bjorkman[0.052]; Tim[0.051]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.044]; Australia[0.044]; second[0.043]; second[0.043]; Scott[0.043]; Republic[0.043]; France[0.043]; France[0.043]; Mark[0.043]; David[0.042]; David[0.042]; Sweden[0.042]; Sweden[0.042]; York[0.042]; Spain[0.042]; number[0.041]; South[0.041]; Alexander[0.041]; Netherlands[0.040]; Guy[0.040]; Russia[0.040]; Russia[0.040]; Friday[0.040]; Friday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech Republic ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.162:0.162[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRomania[0.045]; Switzerland[0.045]; Sweden[0.044]; Sweden[0.044]; Sweden[0.044]; Argentina[0.044]; Czech[0.044]; second[0.043]; second[0.043]; second[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; Germany[0.043]; Republic[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Amanda Coetzer\u001b[39m ==> ENTITY: \u001b[32mAmanda Coetzer\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.245:0.245[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nround[0.065]; round[0.065]; round[0.065]; round[0.065]; round[0.065]; Pioline[0.064]; Sjeng[0.062]; Sampras[0.061]; Jiri[0.061]; Switzerland[0.061]; singles[0.061]; singles[0.061]; singles[0.061]; singles[0.061]; singles[0.061]; Rikl[0.061]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Madagascar ==> ENTITY: \u001b[32mMadagascar\u001b[39m\t\nSCORES: global= 0.239:0.239[0]; local()= 0.076:0.076[0]; log p(e|m)= -0.307:-0.307[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMichael[0.043]; round[0.043]; round[0.043]; round[0.043]; round[0.043]; Italy[0.043]; men[0.043]; men[0.043]; Australia[0.042]; Slovakia[0.042]; second[0.042]; second[0.042]; leg[0.041]; Martin[0.041]; Netherlands[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; Germany[0.041]; Germany[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.049]; Argentina[0.043]; Tennis[0.042]; Tennis[0.042]; Tennis[0.042]; Croatia[0.042]; Britain[0.042]; second[0.042]; second[0.042]; Open[0.042]; Open[0.042]; South[0.041]; Mark[0.041]; singles[0.041]; singles[0.041]; singles[0.041]; singles[0.041]; National[0.041]; Spain[0.041]; Spain[0.041]; Sweden[0.041]; Sweden[0.041]; David[0.040]; David[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Barbara Rittner\u001b[39m ==> ENTITY: \u001b[32mBarbara Rittner\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.225:0.225[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHaarhuis[0.046]; second[0.045]; second[0.045]; second[0.045]; singles[0.044]; singles[0.044]; singles[0.044]; singles[0.044]; singles[0.044]; singles[0.044]; round[0.044]; round[0.044]; round[0.044]; round[0.044]; round[0.044]; round[0.044]; Pioline[0.043]; Davenport[0.043]; Corretja[0.043]; Jiri[0.043]; Sampras[0.042]; Lisa[0.042]; Novak[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.217:0.217[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.080]; Switzerland[0.074]; Argentina[0.073]; Spain[0.073]; Spain[0.073]; Spain[0.073]; Romania[0.072]; Sweden[0.071]; Sweden[0.071]; Republic[0.069]; Republic[0.069]; France[0.068]; France[0.068]; Netherlands[0.068]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Guy Forget\u001b[39m ==> ENTITY: \u001b[32mGuy Forget\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsingles[0.058]; singles[0.058]; singles[0.058]; singles[0.058]; singles[0.058]; Rikl[0.056]; Ivanisevic[0.055]; Davenport[0.055]; France[0.055]; France[0.055]; France[0.055]; round[0.055]; round[0.055]; round[0.055]; round[0.055]; Bjorkman[0.055]; Sjeng[0.054]; number[0.054]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Thomas Johansson\u001b[39m ==> ENTITY: \u001b[32mThomas Johansson\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.257:0.257[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nround[0.044]; round[0.044]; round[0.044]; round[0.044]; round[0.044]; round[0.044]; Haarhuis[0.043]; Randriantefy[0.042]; singles[0.042]; singles[0.042]; singles[0.042]; singles[0.042]; singles[0.042]; Martin[0.041]; Edberg[0.040]; Sampras[0.040]; second[0.040]; second[0.040]; second[0.040]; Bernd[0.040]; Jiri[0.040]; Slovakia[0.040]; Andrei[0.040]; Seles[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andrei Medvedev\u001b[39m ==> ENTITY: \u001b[32mAndriy Medvedev\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHaarhuis[0.062]; round[0.056]; round[0.056]; round[0.056]; round[0.056]; round[0.056]; Sampras[0.056]; Seles[0.055]; singles[0.055]; singles[0.055]; singles[0.055]; singles[0.055]; singles[0.055]; Bernd[0.055]; Jiri[0.055]; second[0.054]; second[0.054]; second[0.054]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Scott Draper\u001b[39m ==> ENTITY: \u001b[32mScott Draper\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.206:0.206[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.062]; Open[0.062]; Tennis[0.061]; Tennis[0.061]; Tennis[0.061]; Philippoussis[0.059]; Testud[0.058]; Rikl[0.058]; Mark[0.058]; singles[0.058]; singles[0.058]; singles[0.058]; Ivanisevic[0.058]; round[0.057]; round[0.057]; round[0.057]; Bjorkman[0.057]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.048]; Argentina[0.043]; Croatia[0.043]; Romania[0.042]; Spain[0.042]; Spain[0.042]; Spain[0.042]; Spain[0.042]; Netherlands[0.042]; second[0.041]; second[0.041]; second[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U S Open ==> ENTITY: \u001b[32mUS Open (tennis)\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.207:0.207[0]; log p(e|m)= -1.252:-1.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.048]; Tennis[0.048]; Tennis[0.048]; Championships[0.045]; Open[0.043]; Open[0.043]; singles[0.042]; singles[0.042]; round[0.042]; round[0.042]; Scott[0.040]; Australia[0.040]; Australia[0.040]; second[0.040]; Philippoussis[0.040]; Sjeng[0.040]; Mark[0.040]; National[0.040]; beat[0.039]; beat[0.039]; beat[0.039]; beat[0.039]; beat[0.039]; Ivanisevic[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sandrine Testud\u001b[39m ==> ENTITY: \u001b[32mSandrine Testud\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.208:0.208[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsingles[0.045]; singles[0.045]; singles[0.045]; Open[0.045]; Open[0.045]; Tennis[0.045]; Tennis[0.045]; Tennis[0.045]; round[0.045]; round[0.045]; round[0.045]; Sjeng[0.044]; Philippoussis[0.043]; Rikl[0.043]; Ivanisevic[0.043]; France[0.042]; France[0.042]; second[0.042]; second[0.042]; Ines[0.041]; Mantilla[0.041]; Championships[0.041]; Schalken[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech Republic ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.175:0.175[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSlovakia[0.046]; Romania[0.045]; Switzerland[0.045]; Sweden[0.044]; Sweden[0.044]; Argentina[0.044]; Czech[0.044]; Czech[0.044]; second[0.043]; second[0.043]; second[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; beat[0.043]; Germany[0.043]; Germany[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Africa ==> ENTITY: \u001b[32mSouth Africa\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.263:-0.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArgentina[0.050]; Men[0.049]; Men[0.049]; Men[0.049]; South[0.049]; Switzerland[0.049]; Africa[0.048]; Spain[0.048]; Spain[0.048]; Spain[0.048]; Spain[0.048]; France[0.047]; France[0.047]; France[0.047]; Germany[0.047]; Women[0.046]; Women[0.046]; second[0.046]; second[0.046]; second[0.046]; Romania[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Irina Spirlea\u001b[39m ==> ENTITY: \u001b[32mIrina Spîrlea\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPioline[0.045]; Rikl[0.045]; Mantilla[0.042]; Jiri[0.042]; singles[0.042]; singles[0.042]; singles[0.042]; singles[0.042]; singles[0.042]; round[0.042]; round[0.042]; round[0.042]; round[0.042]; round[0.042]; Czech[0.041]; Czech[0.041]; Czech[0.041]; Bjorkman[0.041]; Corretja[0.041]; second[0.041]; second[0.041]; second[0.041]; Davenport[0.041]; Rittner[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.091:0.091[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSlovakia[0.044]; Germany[0.043]; Germany[0.043]; second[0.043]; second[0.043]; men[0.043]; men[0.043]; women[0.042]; Republic[0.042]; Republic[0.042]; Andrea[0.041]; round[0.041]; round[0.041]; round[0.041]; Sweden[0.041]; Sweden[0.041]; Randriantefy[0.041]; singles[0.041]; singles[0.041]; singles[0.041]; Monica[0.040]; injury[0.040]; Martin[0.040]; Ukraine[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andrei Olhovskiy\u001b[39m ==> ENTITY: \u001b[32mAndrei Olhovskiy\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsingles[0.045]; singles[0.045]; singles[0.045]; singles[0.045]; Rikl[0.044]; Mark[0.044]; Tennis[0.044]; Tennis[0.044]; Tennis[0.044]; David[0.044]; David[0.044]; Mantilla[0.043]; round[0.043]; round[0.043]; round[0.043]; round[0.043]; Open[0.043]; Open[0.043]; Scott[0.043]; Ivanisevic[0.043]; Doug[0.042]; Henman[0.042]; Bjorkman[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Asa Carlsson\u001b[39m ==> ENTITY: \u001b[32mÅsa Svensson\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHaarhuis[0.047]; singles[0.045]; singles[0.045]; singles[0.045]; singles[0.045]; singles[0.045]; Sweden[0.045]; Sweden[0.045]; Pioline[0.044]; round[0.044]; round[0.044]; round[0.044]; round[0.044]; round[0.044]; Sampras[0.043]; Johansson[0.043]; Jiri[0.042]; Corretja[0.042]; Jan[0.042]; Switzerland[0.041]; second[0.041]; second[0.041]; second[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.190:0.190[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.087]; Switzerland[0.081]; Argentina[0.078]; Germany[0.077]; Germany[0.077]; Romania[0.076]; Spain[0.076]; Spain[0.076]; Spain[0.076]; Netherlands[0.075]; second[0.074]; second[0.074]; second[0.074]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Netherlands ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.250:-0.250[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwitzerland[0.044]; Australia[0.043]; Germany[0.043]; Germany[0.043]; Spain[0.043]; Spain[0.043]; Argentina[0.043]; Women[0.043]; Women[0.043]; Sweden[0.042]; Sweden[0.042]; Slovakia[0.042]; women[0.042]; Men[0.041]; Men[0.041]; France[0.041]; men[0.041]; round[0.040]; round[0.040]; round[0.040]; round[0.040]; round[0.040]; round[0.040]; Pioline[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.047]; France[0.047]; Argentina[0.046]; singles[0.042]; singles[0.042]; singles[0.042]; singles[0.042]; singles[0.042]; Spain[0.041]; Spain[0.041]; Rikl[0.041]; Australia[0.041]; Australia[0.041]; Republic[0.041]; Republic[0.041]; Romania[0.041]; second[0.041]; second[0.041]; second[0.041]; Scott[0.040]; Netherlands[0.040]; Croatia[0.040]; Testud[0.040]; Tennis[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jan Kroslak\u001b[39m ==> ENTITY: \u001b[32mJán Krošlák\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMartin[0.048]; round[0.046]; round[0.046]; round[0.046]; round[0.046]; round[0.046]; Czech[0.046]; Czech[0.046]; Czech[0.046]; Slovakia[0.046]; singles[0.046]; singles[0.046]; singles[0.046]; singles[0.046]; singles[0.046]; Haarhuis[0.045]; Bernd[0.045]; Jiri[0.045]; Raymond[0.045]; leg[0.044]; Stefan[0.044]; Paul[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Petr Korda\u001b[39m ==> ENTITY: \u001b[32mPetr Korda\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMartin[0.052]; Ulihrach[0.052]; Haarhuis[0.051]; Czech[0.050]; Czech[0.050]; Czech[0.050]; Bernd[0.050]; Jiri[0.050]; round[0.050]; round[0.050]; round[0.050]; round[0.050]; round[0.050]; Michael[0.050]; Randriantefy[0.050]; leg[0.049]; men[0.049]; men[0.049]; Stefan[0.049]; Sampras[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.070:0.070[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSlovakia[0.048]; Sweden[0.047]; Sweden[0.047]; singles[0.046]; singles[0.046]; singles[0.046]; singles[0.046]; Italy[0.046]; Germany[0.046]; Germany[0.046]; second[0.045]; second[0.045]; round[0.045]; round[0.045]; round[0.045]; round[0.045]; Thomas[0.044]; Martin[0.044]; Republic[0.044]; Republic[0.044]; Czech[0.044]; Czech[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.049]; Slovakia[0.045]; Germany[0.043]; Germany[0.043]; Italy[0.043]; second[0.042]; second[0.042]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; leg[0.041]; Edberg[0.041]; round[0.040]; round[0.040]; round[0.040]; singles[0.040]; singles[0.040]; singles[0.040]; Johansson[0.040]; injury[0.040]; women[0.040]; Petr[0.040]; \t\n\r [======================>.......................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 10s428ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2470/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1080testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Benetton ==> ENTITY: \u001b[32mTreviso Basket\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.114:0.114[0]; log p(e|m)= -2.180:-2.180[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.053]; basketball[0.051]; tournament[0.050]; Beograd[0.048]; Beograd[0.048]; Tournament[0.048]; international[0.047]; Yugoslavia[0.046]; Dinamo[0.045]; halftime[0.045]; Saturday[0.045]; place[0.045]; place[0.045]; Alba[0.044]; Belgrade[0.044]; Germany[0.044]; Red[0.043]; Russia[0.043]; Star[0.042]; Fifth[0.041]; Results[0.041]; Results[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.042:0.042[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.051]; Yugoslavia[0.049]; Belgrade[0.049]; international[0.049]; Germany[0.049]; basketball[0.048]; Italy[0.048]; Dinamo[0.048]; Tournament[0.047]; Saturday[0.046]; place[0.046]; place[0.046]; Beograd[0.044]; Beograd[0.044]; halftime[0.044]; Star[0.043]; Benetton[0.043]; Red[0.043]; Results[0.042]; Results[0.042]; Fifth[0.041]; Alba[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Yugoslavia ==> ENTITY: \u001b[32mYugoslavia\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.085:0.085[0]; log p(e|m)= -0.654:-0.654[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.051]; tournament[0.051]; Belgrade[0.050]; international[0.050]; Germany[0.049]; Russia[0.047]; basketball[0.047]; Beograd[0.046]; Beograd[0.046]; place[0.046]; place[0.046]; Fifth[0.045]; Tournament[0.044]; Saturday[0.044]; Dinamo[0.044]; Red[0.043]; halftime[0.043]; Star[0.042]; Results[0.042]; Results[0.042]; Benetton[0.041]; Alba[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.068:0.068[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.052]; international[0.051]; Yugoslavia[0.051]; Tournament[0.050]; Germany[0.049]; place[0.047]; place[0.047]; Saturday[0.047]; Russia[0.046]; basketball[0.046]; Dinamo[0.046]; Belgrade[0.044]; Benetton[0.044]; halftime[0.044]; Beograd[0.043]; Beograd[0.043]; Star[0.042]; Results[0.042]; Results[0.042]; Red[0.042]; Alba[0.041]; Fifth[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mRed Star\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRed Star Belgrade\u001b[39m <---> \u001b[32mKK Crvena zvezda\u001b[39m\t\nSCORES: global= 0.266:0.247[\u001b[31m0.019\u001b[39m]; local()= 0.111:0.133[\u001b[32m0.022\u001b[39m]; log p(e|m)= -1.036:-2.465[\u001b[31m1.429\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgrade[0.054]; tournament[0.053]; basketball[0.053]; Beograd[0.053]; Beograd[0.053]; Tournament[0.050]; Dinamo[0.050]; Yugoslavia[0.050]; international[0.049]; Italy[0.048]; Saturday[0.048]; place[0.046]; place[0.046]; Russia[0.046]; Germany[0.045]; halftime[0.045]; Benetton[0.044]; Alba[0.043]; Fifth[0.042]; Results[0.041]; Results[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.050:0.050[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.054]; international[0.052]; Italy[0.051]; Yugoslavia[0.050]; basketball[0.049]; Tournament[0.048]; Russia[0.047]; place[0.047]; place[0.047]; Belgrade[0.046]; Beograd[0.044]; Beograd[0.044]; Dinamo[0.044]; halftime[0.043]; Saturday[0.043]; Red[0.043]; Results[0.042]; Results[0.042]; Star[0.041]; Benetton[0.041]; Fifth[0.041]; Alba[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgrade ==> ENTITY: \u001b[32mBelgrade\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.091:0.091[0]; log p(e|m)= -0.052:-0.052[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYugoslavia[0.052]; tournament[0.052]; Beograd[0.049]; Beograd[0.049]; Germany[0.048]; basketball[0.047]; Saturday[0.047]; Tournament[0.047]; place[0.047]; place[0.047]; Italy[0.047]; international[0.046]; Red[0.045]; halftime[0.044]; Russia[0.044]; Dinamo[0.043]; Star[0.043]; Fifth[0.043]; Benetton[0.042]; Alba[0.040]; Results[0.040]; Results[0.040]; \t\n\r [======================>.......................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 10s424ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2477/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1099testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dublin ==> ENTITY: \u001b[32mDublin\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.013:0.013[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.061]; Saturday[0.061]; retained[0.058]; title[0.058]; round[0.058]; Mexico[0.057]; retired[0.056]; Featherweight[0.056]; Prince[0.055]; Manuel[0.055]; Title[0.055]; Boxing[0.054]; corner[0.054]; featherweight[0.053]; Medina[0.053]; Hamed[0.053]; Naseem[0.052]; Naseem[0.052]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Naseem Hamed\u001b[39m ==> ENTITY: \u001b[32mNaseem Hamed\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFeatherweight[0.066]; Boxing[0.065]; featherweight[0.064]; Britain[0.062]; title[0.061]; Mexico[0.060]; Naseem[0.060]; Title[0.059]; Medina[0.059]; Manuel[0.058]; round[0.058]; retained[0.056]; Saturday[0.056]; Dublin[0.055]; Prince[0.055]; retired[0.055]; corner[0.053]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Manuel Medina ==> ENTITY: \u001b[32mManuel Medina (boxer)\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.186:0.186[0]; log p(e|m)= -0.300:-0.300[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfeatherweight[0.065]; title[0.065]; Boxing[0.064]; retained[0.064]; Mexico[0.063]; round[0.062]; Featherweight[0.062]; Saturday[0.059]; Title[0.058]; Hamed[0.057]; corner[0.057]; retired[0.056]; Britain[0.056]; Dublin[0.054]; Naseem[0.054]; Naseem[0.054]; Prince[0.051]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.013:0.013[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDublin[0.062]; title[0.059]; retained[0.059]; Prince[0.058]; Title[0.058]; Saturday[0.058]; corner[0.057]; Mexico[0.057]; retired[0.057]; round[0.056]; Boxing[0.056]; Hamed[0.054]; featherweight[0.052]; Naseem[0.052]; Naseem[0.052]; Featherweight[0.052]; Medina[0.051]; Manuel[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexico ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.044:0.044[0]; log p(e|m)= -0.183:-0.183[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nManuel[0.062]; Boxing[0.059]; Saturday[0.059]; retained[0.058]; round[0.058]; retired[0.057]; Britain[0.057]; Medina[0.057]; title[0.056]; Featherweight[0.056]; featherweight[0.055]; Title[0.053]; Prince[0.053]; Naseem[0.052]; Naseem[0.052]; Hamed[0.052]; Dublin[0.051]; corner[0.051]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = WBO\u001b[39m ==> ENTITY: \u001b[32mWorld Boxing Organization\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBoxing[0.059]; featherweight[0.058]; Featherweight[0.057]; retained[0.057]; title[0.056]; Mexico[0.054]; retired[0.054]; Title[0.054]; Hamed[0.054]; round[0.053]; Britain[0.052]; Saturday[0.051]; Medina[0.050]; Naseem[0.050]; Naseem[0.050]; Manuel[0.049]; Dublin[0.048]; Prince[0.047]; corner[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Naseem\u001b[39m ==> ENTITY: \u001b[32mNaseem Hamed\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHamed[0.062]; Featherweight[0.061]; Boxing[0.061]; featherweight[0.060]; Britain[0.058]; title[0.057]; Mexico[0.056]; Naseem[0.056]; Title[0.055]; Medina[0.055]; Manuel[0.054]; round[0.054]; retained[0.053]; Saturday[0.052]; Dublin[0.052]; Prince[0.051]; retired[0.051]; corner[0.049]; \t\n\r [======================>.......................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 10s406ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2484/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1077testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgrade ==> ENTITY: \u001b[32mBelgrade\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.114:0.114[0]; log p(e|m)= -0.052:-0.052[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; League[0.047]; Vojvodina[0.045]; Vojvodina[0.045]; soccer[0.045]; Division[0.044]; matches[0.044]; Yugoslav[0.044]; Yugoslav[0.044]; Mladost[0.043]; Mladost[0.043]; Soccer[0.043]; Borac[0.043]; Zemun[0.043]; points[0.043]; played[0.043]; Partizan[0.043]; Partizan[0.043]; Buducnost[0.043]; Saturday[0.042]; won[0.041]; goals[0.040]; goals[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = OFK Kikinda\u001b[39m ==> ENTITY: \u001b[32mFK Kikinda\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRadnicki[0.046]; Radnicki[0.046]; Beograd[0.044]; Beograd[0.044]; Kikinda[0.042]; Rudar[0.042]; Buducnost[0.042]; Buducnost[0.042]; Buducnost[0.042]; Obilic[0.042]; Obilic[0.042]; Vojvodina[0.042]; Zeleznik[0.042]; Zeleznik[0.042]; played[0.042]; Borac[0.041]; Loznica[0.040]; Loznica[0.040]; Rad[0.040]; Mladost[0.040]; Mladost[0.040]; Standings[0.040]; Division[0.039]; Bj[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vojvodina ==> ENTITY: \u001b[32mFK Vojvodina\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.219:0.219[0]; log p(e|m)= -1.211:-1.211[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; league[0.043]; Vojvodina[0.043]; played[0.043]; Partizan[0.042]; Partizan[0.042]; matches[0.042]; Buducnost[0.042]; Buducnost[0.042]; Buducnost[0.042]; Borac[0.042]; Borac[0.042]; Belgrade[0.042]; soccer[0.042]; Rudar[0.041]; Obilic[0.041]; drawn[0.041]; points[0.041]; won[0.041]; Beograd[0.041]; Yugoslav[0.040]; Yugoslav[0.040]; Division[0.040]; Division[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBuducnost\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mKK Budućnost Podgorica\u001b[39m <---> \u001b[32mFK Budućnost Podgorica\u001b[39m\t\nSCORES: global= 0.282:0.281[\u001b[31m0.001\u001b[39m]; local()= 0.180:0.194[\u001b[32m0.014\u001b[39m]; log p(e|m)= -1.100:-0.693[\u001b[32m0.406\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.045]; soccer[0.044]; Buducnost[0.044]; Belgrade[0.044]; Borac[0.043]; matches[0.043]; League[0.043]; Partizan[0.042]; Partizan[0.042]; Soccer[0.042]; Vojvodina[0.042]; Vojvodina[0.042]; Zemun[0.041]; Zemun[0.041]; played[0.041]; Mladost[0.041]; Mladost[0.041]; won[0.041]; Division[0.040]; points[0.040]; Yugoslav[0.040]; Yugoslav[0.040]; Hajduk[0.039]; Hajduk[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spartak ==> ENTITY: \u001b[32mFK Spartak Zlatibor Voda\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.184:0.184[0]; log p(e|m)= -1.839:-1.839[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpartak[0.047]; Buducnost[0.047]; Buducnost[0.047]; Radnicki[0.046]; Radnicki[0.046]; Obilic[0.045]; Obilic[0.045]; Rudar[0.044]; Rudar[0.044]; played[0.044]; Standings[0.044]; Beograd[0.044]; Beograd[0.044]; Division[0.043]; Mladost[0.042]; Mladost[0.042]; Loznica[0.042]; Loznica[0.042]; Zeleznik[0.041]; Zeleznik[0.041]; Kikinda[0.041]; Kikinda[0.041]; Sunday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rad ==> ENTITY: \u001b[32mFK Rad\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.186:0.186[0]; log p(e|m)= -1.211:-1.211[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.047]; Rad[0.047]; Soccer[0.045]; League[0.045]; league[0.043]; matches[0.043]; Buducnost[0.042]; Belgrade[0.042]; Borac[0.041]; Vojvodina[0.041]; Vojvodina[0.041]; played[0.041]; Partizan[0.041]; Partizan[0.041]; Yugoslav[0.041]; Yugoslav[0.041]; goals[0.040]; goals[0.040]; Hajduk[0.040]; Hajduk[0.040]; Zemun[0.040]; Zemun[0.040]; Mladost[0.039]; Mladost[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sloboda ==> ENTITY: \u001b[32mFK Sloboda Tuzla\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.169:0.169[0]; log p(e|m)= -0.423:-0.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSloboda[0.047]; Buducnost[0.044]; Buducnost[0.044]; Buducnost[0.044]; Radnicki[0.044]; Obilic[0.042]; Obilic[0.042]; Zemun[0.042]; Beograd[0.042]; Beograd[0.042]; Rudar[0.042]; played[0.042]; Partizan[0.042]; Vojvodina[0.042]; Zeleznik[0.041]; Borac[0.041]; Mladost[0.041]; Mladost[0.041]; Standings[0.040]; Kikinda[0.040]; Kikinda[0.040]; Spartak[0.040]; Spartak[0.040]; Hajduk[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sutjeska ==> ENTITY: \u001b[32mFK Sutjeska Nikšić\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.186:0.186[0]; log p(e|m)= -0.518:-0.518[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBeograd[0.043]; Beograd[0.043]; Buducnost[0.043]; Buducnost[0.043]; Buducnost[0.043]; Radnicki[0.043]; Radnicki[0.043]; Borac[0.042]; Obilic[0.042]; Obilic[0.042]; played[0.041]; Sloboda[0.041]; Sloboda[0.041]; Mladost[0.041]; Mladost[0.041]; Kikinda[0.041]; Kikinda[0.041]; Rudar[0.041]; Rudar[0.041]; Zemun[0.041]; Loznica[0.041]; Loznica[0.041]; Zeleznik[0.040]; Zeleznik[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mZemun\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mZemun\u001b[39m <---> \u001b[32mFK Zemun\u001b[39m\t\nSCORES: global= 0.285:0.281[\u001b[31m0.004\u001b[39m]; local()= 0.155:0.188[\u001b[32m0.033\u001b[39m]; log p(e|m)= 0.000:-0.851[\u001b[31m0.851\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRadnicki[0.046]; Buducnost[0.043]; Buducnost[0.043]; Buducnost[0.043]; Kikinda[0.043]; Kikinda[0.043]; Obilic[0.043]; Obilic[0.043]; won[0.042]; Loznica[0.042]; Beograd[0.042]; Vojvodina[0.041]; Vojvodina[0.041]; played[0.041]; Zemun[0.041]; Partizan[0.041]; Partizan[0.041]; Borac[0.040]; Borac[0.040]; Mladost[0.040]; Mladost[0.040]; Mladost[0.040]; Sloboda[0.040]; Saturday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Partizan ==> ENTITY: \u001b[32mFK Partizan\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.182:0.182[0]; log p(e|m)= -0.194:-0.194[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.046]; league[0.044]; League[0.044]; Soccer[0.044]; matches[0.044]; Partizan[0.043]; Buducnost[0.043]; Buducnost[0.043]; played[0.043]; Yugoslav[0.042]; Yugoslav[0.042]; Borac[0.042]; won[0.042]; Belgrade[0.042]; Vojvodina[0.040]; Vojvodina[0.040]; lost[0.040]; Zemun[0.039]; Zemun[0.039]; Division[0.039]; Mladost[0.039]; Mladost[0.039]; Standings[0.039]; Standings[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBorac\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFK Borac Banja Luka\u001b[39m <---> \u001b[32mFK Borac Čačak\u001b[39m\t\nSCORES: global= 0.275:0.268[\u001b[31m0.008\u001b[39m]; local()= 0.203:0.193[\u001b[31m0.010\u001b[39m]; log p(e|m)= -0.136:-1.000[\u001b[31m0.864\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; League[0.044]; soccer[0.044]; matches[0.044]; Buducnost[0.044]; Soccer[0.043]; played[0.043]; Partizan[0.042]; Partizan[0.042]; Belgrade[0.042]; Vojvodina[0.041]; Vojvodina[0.041]; Yugoslav[0.041]; Yugoslav[0.041]; Mladost[0.041]; Mladost[0.041]; Zemun[0.041]; Zemun[0.041]; Rad[0.040]; Rad[0.040]; Division[0.040]; won[0.040]; Standings[0.039]; Standings[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sutjeska ==> ENTITY: \u001b[32mFK Sutjeska Nikšić\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.185:0.185[0]; log p(e|m)= -0.518:-0.518[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVojvodina[0.045]; Beograd[0.043]; Beograd[0.043]; Buducnost[0.043]; Buducnost[0.043]; Buducnost[0.043]; Radnicki[0.043]; Radnicki[0.043]; Borac[0.042]; Obilic[0.042]; Obilic[0.042]; played[0.041]; Sloboda[0.041]; Sloboda[0.041]; Mladost[0.041]; Kikinda[0.041]; Kikinda[0.041]; Rudar[0.041]; Rudar[0.041]; Zemun[0.041]; Loznica[0.041]; Loznica[0.041]; Zeleznik[0.040]; Zeleznik[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mYugoslav\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSocialist Federal Republic of Yugoslavia\u001b[39m <---> \u001b[32mYugoslavia\u001b[39m\t\nSCORES: global= 0.244:0.238[\u001b[31m0.006\u001b[39m]; local()= 0.112:0.094[\u001b[31m0.018\u001b[39m]; log p(e|m)= -0.826:-0.759[\u001b[32m0.066\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.048]; matches[0.047]; Belgrade[0.046]; League[0.046]; soccer[0.045]; Yugoslav[0.045]; Buducnost[0.045]; played[0.044]; won[0.044]; Vojvodina[0.043]; Vojvodina[0.043]; Partizan[0.043]; Partizan[0.043]; Borac[0.043]; Soccer[0.043]; Hajduk[0.042]; Hajduk[0.042]; Mladost[0.042]; Mladost[0.042]; Division[0.041]; Zemun[0.041]; goals[0.040]; goals[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = OFK Beograd\u001b[39m ==> ENTITY: \u001b[32mOFK Beograd\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.231:0.231[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRadnicki[0.045]; Radnicki[0.045]; Buducnost[0.045]; Buducnost[0.045]; Buducnost[0.045]; Rudar[0.043]; Borac[0.043]; Beograd[0.043]; played[0.042]; Obilic[0.042]; Obilic[0.042]; Vojvodina[0.042]; Zeleznik[0.041]; Loznica[0.040]; Loznica[0.040]; Mladost[0.040]; Mladost[0.040]; Zemun[0.040]; Kikinda[0.040]; Kikinda[0.040]; Hajduk[0.039]; Bj[0.039]; Division[0.039]; Rad[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Partizan ==> ENTITY: \u001b[32mFK Partizan\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.194:-0.194[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.046]; league[0.044]; League[0.044]; Soccer[0.044]; matches[0.043]; Partizan[0.043]; Buducnost[0.043]; Buducnost[0.043]; Buducnost[0.043]; played[0.042]; Yugoslav[0.042]; Yugoslav[0.042]; Belgrade[0.042]; Borac[0.042]; Borac[0.042]; won[0.041]; Vojvodina[0.040]; Vojvodina[0.040]; lost[0.040]; Mladost[0.039]; Mladost[0.039]; Mladost[0.039]; Zemun[0.039]; Zemun[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Vojvodina ==> ENTITY: \u001b[32mFK Vojvodina\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.199:0.199[0]; log p(e|m)= -1.211:-1.211[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; league[0.044]; Vojvodina[0.043]; played[0.043]; Partizan[0.043]; Partizan[0.043]; matches[0.043]; Buducnost[0.042]; Buducnost[0.042]; Borac[0.042]; Belgrade[0.042]; soccer[0.042]; drawn[0.041]; points[0.041]; won[0.041]; Yugoslav[0.041]; Yugoslav[0.041]; Division[0.041]; Soccer[0.041]; lost[0.040]; goals[0.040]; goals[0.040]; Zemun[0.040]; Zemun[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = OFK Beograd\u001b[39m ==> ENTITY: \u001b[32mOFK Beograd\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.237:0.237[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRadnicki[0.047]; Radnicki[0.047]; Buducnost[0.047]; Buducnost[0.047]; Buducnost[0.047]; Rudar[0.045]; Rudar[0.045]; Borac[0.044]; Beograd[0.044]; played[0.044]; Obilic[0.044]; Obilic[0.044]; Zeleznik[0.042]; Zeleznik[0.042]; Loznica[0.042]; Loznica[0.042]; Mladost[0.042]; Mladost[0.042]; Kikinda[0.041]; Kikinda[0.041]; Bj[0.041]; Bj[0.041]; Division[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBuducnost\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mKK Budućnost Podgorica\u001b[39m <---> \u001b[32mFK Budućnost Podgorica\u001b[39m\t\nSCORES: global= 0.282:0.281[\u001b[31m0.001\u001b[39m]; local()= 0.197:0.205[\u001b[32m0.008\u001b[39m]; log p(e|m)= -1.100:-0.693[\u001b[32m0.406\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRadnicki[0.044]; Buducnost[0.044]; Buducnost[0.044]; Borac[0.044]; Partizan[0.042]; Partizan[0.042]; Beograd[0.042]; Beograd[0.042]; Vojvodina[0.042]; Vojvodina[0.042]; Rudar[0.042]; Zemun[0.041]; played[0.041]; Obilic[0.041]; Obilic[0.041]; Mladost[0.041]; Mladost[0.041]; won[0.041]; Division[0.040]; Loznica[0.040]; Loznica[0.040]; points[0.040]; Hajduk[0.040]; Spartak[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spartak ==> ENTITY: \u001b[32mFK Spartak Zlatibor Voda\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.191:0.191[0]; log p(e|m)= -1.839:-1.839[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpartak[0.044]; Buducnost[0.044]; Buducnost[0.044]; Buducnost[0.044]; Radnicki[0.043]; Radnicki[0.043]; Obilic[0.043]; Obilic[0.043]; Borac[0.043]; Rudar[0.042]; Rudar[0.042]; played[0.042]; Standings[0.042]; Beograd[0.041]; Beograd[0.041]; Division[0.041]; Zemun[0.040]; Mladost[0.040]; Mladost[0.040]; Hajduk[0.040]; Loznica[0.040]; Loznica[0.040]; Zeleznik[0.039]; Zeleznik[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mHajduk\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHNK Hajduk Split\u001b[39m <---> \u001b[32mFK Hajduk Kula\u001b[39m\t\nSCORES: global= 0.278:0.271[\u001b[31m0.007\u001b[39m]; local()= 0.247:0.232[\u001b[31m0.014\u001b[39m]; log p(e|m)= -0.904:-1.820[\u001b[31m0.916\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.046]; Buducnost[0.046]; Buducnost[0.046]; Buducnost[0.046]; league[0.045]; Radnicki[0.045]; Borac[0.044]; Borac[0.044]; matches[0.044]; Rudar[0.044]; Belgrade[0.043]; Beograd[0.043]; League[0.043]; Obilic[0.043]; Obilic[0.043]; Vojvodina[0.043]; Vojvodina[0.043]; Partizan[0.043]; Partizan[0.043]; Hajduk[0.041]; Zemun[0.041]; Zemun[0.041]; Sloboda[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Red Star ==> ENTITY: \u001b[32mRed Star Belgrade\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.180:0.180[0]; log p(e|m)= -1.036:-1.036[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; League[0.044]; Partizan[0.044]; Partizan[0.044]; Belgrade[0.044]; soccer[0.044]; Soccer[0.043]; Buducnost[0.043]; Buducnost[0.043]; matches[0.043]; Borac[0.042]; won[0.041]; played[0.041]; Vojvodina[0.041]; Vojvodina[0.041]; Mladost[0.040]; Mladost[0.040]; Zemun[0.040]; Zemun[0.040]; Star[0.040]; Yugoslav[0.040]; Yugoslav[0.040]; goals[0.039]; goals[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mYugoslav\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSocialist Federal Republic of Yugoslavia\u001b[39m <---> \u001b[32mYugoslavia\u001b[39m\t\nSCORES: global= 0.244:0.238[\u001b[31m0.006\u001b[39m]; local()= 0.112:0.094[\u001b[31m0.018\u001b[39m]; log p(e|m)= -0.826:-0.759[\u001b[32m0.066\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.048]; matches[0.047]; Belgrade[0.046]; League[0.046]; soccer[0.045]; Yugoslav[0.045]; Buducnost[0.045]; played[0.044]; won[0.044]; Vojvodina[0.043]; Vojvodina[0.043]; Partizan[0.043]; Partizan[0.043]; Borac[0.043]; Soccer[0.043]; Hajduk[0.042]; Hajduk[0.042]; Mladost[0.042]; Mladost[0.042]; Division[0.041]; Zemun[0.041]; goals[0.040]; goals[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mHajduk\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHNK Hajduk Split\u001b[39m <---> \u001b[32mFK Hajduk Kula\u001b[39m\t\nSCORES: global= 0.276:0.271[\u001b[31m0.006\u001b[39m]; local()= 0.203:0.209[\u001b[32m0.006\u001b[39m]; log p(e|m)= -0.904:-1.820[\u001b[31m0.916\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.045]; Buducnost[0.045]; league[0.045]; Borac[0.044]; matches[0.044]; Belgrade[0.043]; League[0.043]; Soccer[0.042]; Vojvodina[0.042]; Vojvodina[0.042]; Partizan[0.042]; Partizan[0.042]; Hajduk[0.041]; Zemun[0.040]; Zemun[0.040]; Mladost[0.040]; Mladost[0.040]; Yugoslav[0.040]; Yugoslav[0.040]; played[0.040]; Standings[0.040]; Standings[0.040]; won[0.040]; lost[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sloboda ==> ENTITY: \u001b[32mFK Sloboda Tuzla\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.167:0.167[0]; log p(e|m)= -0.423:-0.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSloboda[0.048]; Radnicki[0.044]; Radnicki[0.044]; played[0.043]; Buducnost[0.043]; Buducnost[0.043]; Spartak[0.043]; Spartak[0.043]; Beograd[0.042]; Beograd[0.042]; Rudar[0.042]; Rudar[0.042]; Zeleznik[0.041]; Zeleznik[0.041]; Borac[0.041]; Mladost[0.041]; Mladost[0.041]; Obilic[0.041]; Obilic[0.041]; Kikinda[0.041]; Kikinda[0.041]; Standings[0.040]; Loznica[0.039]; Loznica[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBorac\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFK Borac Banja Luka\u001b[39m <---> \u001b[32mFK Borac Čačak\u001b[39m\t\nSCORES: global= 0.276:0.266[\u001b[31m0.010\u001b[39m]; local()= 0.230:0.193[\u001b[31m0.037\u001b[39m]; log p(e|m)= -0.136:-1.000[\u001b[31m0.864\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRadnicki[0.045]; Buducnost[0.044]; Buducnost[0.044]; Buducnost[0.044]; played[0.043]; Partizan[0.043]; Rudar[0.042]; Beograd[0.042]; Beograd[0.042]; Obilic[0.042]; Obilic[0.042]; Vojvodina[0.041]; Loznica[0.041]; Loznica[0.041]; Mladost[0.041]; Mladost[0.041]; Zemun[0.041]; Rad[0.041]; Division[0.040]; won[0.040]; Standings[0.040]; Kikinda[0.040]; Kikinda[0.040]; Zeleznik[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Red Star ==> ENTITY: \u001b[32mRed Star Belgrade\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.184:0.184[0]; log p(e|m)= -1.036:-1.036[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.048]; League[0.046]; Partizan[0.045]; Partizan[0.045]; Belgrade[0.045]; soccer[0.045]; Soccer[0.045]; Buducnost[0.045]; Buducnost[0.045]; matches[0.044]; Borac[0.043]; Borac[0.043]; won[0.043]; played[0.042]; Vojvodina[0.042]; Vojvodina[0.042]; Mladost[0.042]; Mladost[0.042]; Zemun[0.042]; Zemun[0.042]; Star[0.041]; Yugoslav[0.041]; Yugoslav[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = OFK Kikinda\u001b[39m ==> ENTITY: \u001b[32mFK Kikinda\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRadnicki[0.048]; Radnicki[0.048]; Beograd[0.045]; Beograd[0.045]; Kikinda[0.044]; Rudar[0.044]; Rudar[0.044]; Buducnost[0.044]; Buducnost[0.044]; Buducnost[0.044]; Obilic[0.043]; Obilic[0.043]; Zeleznik[0.043]; Zeleznik[0.043]; played[0.043]; Borac[0.043]; Loznica[0.042]; Loznica[0.042]; Rad[0.042]; Mladost[0.041]; Mladost[0.041]; Standings[0.041]; Division[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mZemun\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mZemun\u001b[39m <---> \u001b[32mFK Zemun\u001b[39m\t\nSCORES: global= 0.285:0.281[\u001b[31m0.003\u001b[39m]; local()= 0.129:0.174[\u001b[32m0.045\u001b[39m]; log p(e|m)= 0.000:-0.851[\u001b[31m0.851\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.049]; Buducnost[0.046]; League[0.045]; Belgrade[0.045]; Yugoslav[0.044]; Yugoslav[0.044]; Soccer[0.044]; won[0.044]; soccer[0.044]; matches[0.044]; Vojvodina[0.043]; Vojvodina[0.043]; played[0.043]; Zemun[0.043]; Partizan[0.043]; Partizan[0.043]; Borac[0.042]; Mladost[0.042]; Mladost[0.042]; Saturday[0.042]; lost[0.041]; Rad[0.041]; Rad[0.041]; \t\n\r [=======================>......................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 10s252ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2514/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 990testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nagoum Yamassoum\u001b[39m ==> ENTITY: \u001b[32mNagoum Yamassoum\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nuniversity[0.044]; university[0.044]; university[0.044]; minister[0.044]; government[0.043]; government[0.043]; University[0.043]; office[0.042]; September[0.042]; Minister[0.042]; students[0.042]; students[0.042]; students[0.042]; students[0.042]; Education[0.041]; Chad[0.040]; Chad[0.040]; held[0.040]; held[0.040]; said[0.040]; said[0.040]; said[0.040]; state[0.040]; examinations[0.040]; \t\n\r [=======================>......................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 10s281ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2515/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 987testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Liechtenstein ==> ENTITY: \u001b[32mLiechtenstein national football team\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.186:0.186[0]; log p(e|m)= -1.995:-1.995[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLiechtenstein[0.047]; soccer[0.046]; match[0.045]; Soccer[0.043]; play[0.043]; play[0.043]; qualifying[0.043]; team[0.043]; team[0.043]; team[0.043]; Cup[0.043]; international[0.042]; Dublin[0.041]; competitive[0.041]; Saturday[0.040]; ahead[0.040]; Irish[0.040]; makes[0.040]; McCarthy[0.040]; McCarthy[0.040]; debut[0.040]; World[0.039]; joins[0.038]; Friday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Birmingham ==> ENTITY: \u001b[32mBirmingham City F.C.\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.151:0.151[0]; log p(e|m)= -3.101:-3.101[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.045]; City[0.044]; Norwich[0.044]; Soccer[0.043]; team[0.042]; team[0.042]; team[0.042]; manager[0.042]; Dublin[0.042]; Dublin[0.042]; soccer[0.042]; ahead[0.041]; play[0.041]; play[0.041]; joins[0.041]; Gary[0.041]; debut[0.041]; Cup[0.041]; Neill[0.041]; Neill[0.041]; Saturday[0.041]; qualifying[0.040]; Ian[0.040]; Keith[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Irwin ==> ENTITY: \u001b[32mDenis Irwin\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.159:0.159[0]; log p(e|m)= -3.170:-3.170[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmanager[0.044]; Birmingham[0.044]; Gary[0.043]; match[0.043]; Soccer[0.043]; team[0.042]; team[0.042]; team[0.042]; Cup[0.042]; Ian[0.042]; Townsend[0.042]; Keith[0.041]; Phil[0.041]; soccer[0.041]; City[0.041]; Neill[0.041]; Neill[0.041]; debut[0.041]; Norwich[0.041]; McCarthy[0.041]; McCarthy[0.041]; Saturday[0.041]; qualifying[0.041]; announced[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Given ==> ENTITY: \u001b[32mShay Given\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.175:0.175[0]; log p(e|m)= -0.284:-0.284[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.045]; Birmingham[0.044]; Soccer[0.043]; play[0.042]; play[0.042]; Staunton[0.042]; City[0.042]; Keith[0.042]; team[0.042]; team[0.042]; team[0.042]; Gary[0.042]; debut[0.041]; Neill[0.041]; Neill[0.041]; Dublin[0.041]; Dublin[0.041]; soccer[0.041]; Cup[0.041]; international[0.041]; Phil[0.041]; Ian[0.041]; Irish[0.041]; Niall[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gary Breen\u001b[39m ==> ENTITY: \u001b[32mGary Breen\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBirmingham[0.044]; Soccer[0.044]; manager[0.043]; team[0.043]; team[0.043]; team[0.043]; Cup[0.042]; match[0.042]; Breen[0.042]; international[0.042]; City[0.042]; Phil[0.042]; McCarthy[0.041]; McCarthy[0.041]; debut[0.041]; Neill[0.041]; Neill[0.041]; Norwich[0.041]; Ian[0.041]; play[0.040]; play[0.040]; soccer[0.040]; Keith[0.040]; Staunton[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Liechtenstein ==> ENTITY: \u001b[32mLiechtenstein national football team\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.187:0.187[0]; log p(e|m)= -1.995:-1.995[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLiechtenstein[0.047]; soccer[0.045]; match[0.045]; Soccer[0.043]; play[0.043]; play[0.043]; qualifying[0.043]; team[0.043]; team[0.043]; team[0.043]; Cup[0.043]; international[0.042]; Dublin[0.041]; Dublin[0.041]; competitive[0.041]; Saturday[0.040]; ahead[0.040]; Irish[0.040]; makes[0.040]; McCarthy[0.039]; McCarthy[0.039]; debut[0.039]; World[0.039]; joins[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Harte\u001b[39m ==> ENTITY: \u001b[32mIan Harte\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHarte[0.045]; team[0.043]; team[0.043]; team[0.043]; City[0.043]; play[0.042]; play[0.042]; match[0.042]; Keith[0.042]; manager[0.042]; ahead[0.041]; announced[0.041]; Gary[0.041]; Birmingham[0.041]; debut[0.041]; McCarthy[0.041]; McCarthy[0.041]; international[0.041]; Neill[0.041]; Neill[0.041]; Saturday[0.041]; Irish[0.041]; Quinn[0.041]; Quinn[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Breen\u001b[39m ==> ENTITY: \u001b[32mGary Breen\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBirmingham[0.044]; Soccer[0.043]; manager[0.043]; team[0.043]; team[0.043]; team[0.043]; Cup[0.042]; match[0.042]; Gary[0.042]; Breen[0.042]; international[0.042]; City[0.042]; Phil[0.041]; McCarthy[0.041]; McCarthy[0.041]; debut[0.041]; Neill[0.041]; Neill[0.041]; Norwich[0.041]; Ian[0.041]; play[0.040]; play[0.040]; soccer[0.040]; Keith[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Phil Babb\u001b[39m ==> ENTITY: \u001b[32mPhil Babb\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.045]; Birmingham[0.044]; match[0.044]; City[0.042]; manager[0.042]; McCarthy[0.042]; McCarthy[0.042]; international[0.042]; Staunton[0.042]; Irish[0.042]; team[0.041]; team[0.041]; team[0.041]; Norwich[0.041]; Keith[0.041]; Gary[0.041]; Dublin[0.041]; Dublin[0.041]; Ian[0.041]; soccer[0.041]; Neill[0.041]; Neill[0.041]; debut[0.041]; Cup[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = McCarthy\u001b[39m ==> ENTITY: \u001b[32mMick McCarthy\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDublin[0.045]; manager[0.044]; Birmingham[0.044]; Irish[0.043]; McCarthy[0.043]; team[0.043]; team[0.043]; team[0.043]; Niall[0.042]; Gary[0.042]; Keith[0.042]; Staunton[0.041]; Mick[0.041]; Neill[0.041]; debut[0.041]; Harte[0.041]; Harte[0.041]; Ian[0.041]; Quinn[0.040]; Norwich[0.040]; joins[0.040]; Cup[0.040]; play[0.040]; play[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIrish\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIreland\u001b[39m <---> \u001b[32mRepublic of Ireland\u001b[39m\t\nSCORES: global= 0.236:0.233[\u001b[31m0.002\u001b[39m]; local()= 0.117:0.131[\u001b[32m0.013\u001b[39m]; log p(e|m)= -1.457:-2.254[\u001b[31m0.797\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDublin[0.048]; Dublin[0.048]; team[0.047]; team[0.047]; team[0.047]; Cup[0.044]; match[0.044]; play[0.043]; play[0.043]; soccer[0.043]; Niall[0.043]; international[0.043]; Staunton[0.043]; Soccer[0.042]; Norwich[0.042]; McCarthy[0.042]; McCarthy[0.042]; Birmingham[0.042]; Neill[0.042]; Neill[0.042]; Quinn[0.042]; Quinn[0.042]; City[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Staunton ==> ENTITY: \u001b[32mSteve Staunton\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.173:0.173[0]; log p(e|m)= -2.343:-2.343[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIrish[0.048]; Dublin[0.045]; Dublin[0.045]; match[0.045]; manager[0.044]; team[0.044]; team[0.044]; team[0.044]; Birmingham[0.043]; international[0.043]; play[0.043]; play[0.043]; Keith[0.043]; announced[0.043]; Neill[0.043]; Neill[0.043]; debut[0.042]; Gary[0.042]; Ian[0.042]; joins[0.042]; Soccer[0.042]; City[0.042]; Niall[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ian Harte\u001b[39m ==> ENTITY: \u001b[32mIan Harte\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHarte[0.045]; team[0.043]; team[0.043]; team[0.043]; City[0.043]; play[0.042]; play[0.042]; match[0.042]; Keith[0.042]; manager[0.042]; ahead[0.041]; announced[0.041]; Gary[0.041]; Birmingham[0.041]; debut[0.041]; McCarthy[0.041]; McCarthy[0.041]; international[0.041]; Neill[0.041]; Neill[0.041]; Saturday[0.041]; Irish[0.041]; Quinn[0.041]; Quinn[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Houghton ==> ENTITY: \u001b[32mRay Houghton\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.192:0.192[0]; log p(e|m)= -3.101:-3.101[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGary[0.045]; Birmingham[0.044]; team[0.043]; team[0.043]; team[0.043]; Keith[0.042]; Neill[0.042]; Neill[0.042]; Quinn[0.042]; Quinn[0.042]; match[0.041]; play[0.041]; play[0.041]; Phil[0.041]; Ian[0.041]; debut[0.041]; City[0.041]; McCarthy[0.041]; Norwich[0.041]; manager[0.041]; Staunton[0.040]; international[0.040]; Irish[0.040]; Mick[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = McAteer\u001b[39m ==> ENTITY: \u001b[32mJason McAteer\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnames[0.047]; team[0.043]; team[0.043]; team[0.043]; match[0.043]; Irish[0.043]; Cup[0.042]; Dublin[0.042]; Dublin[0.042]; Birmingham[0.042]; Gary[0.041]; soccer[0.041]; Norwich[0.041]; Soccer[0.041]; Phil[0.041]; Niall[0.041]; McLoughlin[0.041]; manager[0.041]; competitive[0.041]; City[0.040]; Keith[0.040]; play[0.040]; play[0.040]; Ian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dublin ==> ENTITY: \u001b[32mDublin\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.158:0.158[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDublin[0.048]; Irish[0.045]; Birmingham[0.042]; Breen[0.042]; Breen[0.042]; match[0.042]; Niall[0.042]; Mick[0.042]; play[0.042]; play[0.042]; McCarthy[0.042]; McCarthy[0.042]; Gary[0.041]; Quinn[0.041]; Quinn[0.041]; Ian[0.041]; team[0.041]; team[0.041]; team[0.041]; McLoughlin[0.041]; Harte[0.041]; Harte[0.041]; City[0.040]; Norwich[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Norwich City\u001b[39m ==> ENTITY: \u001b[32mNorwich City F.C.\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBirmingham[0.046]; team[0.046]; team[0.046]; team[0.046]; qualifying[0.046]; match[0.045]; Keith[0.044]; ahead[0.044]; Ian[0.043]; manager[0.043]; Phil[0.043]; Cup[0.043]; play[0.043]; play[0.043]; Soccer[0.043]; Neill[0.043]; Neill[0.043]; Gary[0.043]; debut[0.042]; Saturday[0.042]; soccer[0.042]; joins[0.041]; McAteer[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Quinn\u001b[39m ==> ENTITY: \u001b[32mNiall Quinn\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBirmingham[0.044]; Quinn[0.043]; team[0.043]; team[0.043]; Neill[0.042]; Neill[0.042]; Ian[0.042]; Keith[0.042]; Dublin[0.042]; Dublin[0.042]; Irish[0.042]; McCarthy[0.042]; Gary[0.042]; qualifying[0.041]; debut[0.041]; match[0.041]; play[0.041]; play[0.041]; Norwich[0.041]; manager[0.041]; City[0.041]; World[0.040]; Saturday[0.040]; Staunton[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mick McCarthy\u001b[39m ==> ENTITY: \u001b[32mMick McCarthy\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDublin[0.044]; Dublin[0.044]; manager[0.043]; Birmingham[0.043]; Irish[0.043]; McCarthy[0.043]; team[0.042]; team[0.042]; team[0.042]; Niall[0.042]; Gary[0.042]; Keith[0.042]; Staunton[0.041]; Neill[0.041]; Neill[0.041]; debut[0.041]; Harte[0.041]; Harte[0.041]; Ian[0.040]; Quinn[0.040]; Quinn[0.040]; Norwich[0.040]; joins[0.040]; Cup[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = McLoughlin ==> ENTITY: \u001b[32mAlan McLoughlin\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.130:0.130[0]; log p(e|m)= -2.364:-2.364[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIan[0.050]; Keith[0.047]; Gary[0.044]; Mick[0.044]; Irwin[0.043]; Phil[0.043]; Neill[0.042]; Neill[0.042]; play[0.042]; play[0.042]; Irish[0.041]; Niall[0.041]; soccer[0.041]; Townsend[0.041]; Dublin[0.040]; Dublin[0.040]; Cup[0.040]; Norwich[0.040]; team[0.040]; team[0.040]; team[0.040]; manager[0.039]; match[0.039]; Houghton[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Niall Quinn\u001b[39m ==> ENTITY: \u001b[32mNiall Quinn\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBirmingham[0.044]; Quinn[0.043]; team[0.043]; team[0.043]; team[0.043]; Neill[0.042]; Neill[0.042]; Ian[0.042]; Keith[0.042]; Dublin[0.042]; Dublin[0.042]; Irish[0.042]; McCarthy[0.042]; McCarthy[0.042]; Gary[0.041]; qualifying[0.041]; debut[0.041]; match[0.041]; play[0.041]; play[0.041]; Norwich[0.041]; manager[0.041]; City[0.040]; World[0.040]; \t\n\r [=======================>......................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 10s153ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2538/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 999testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mChechens\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mChechnya\u001b[39m <---> \u001b[32mChechen people\u001b[39m\t\nSCORES: global= 0.256:0.247[\u001b[31m0.010\u001b[39m]; local()= 0.162:0.175[\u001b[32m0.014\u001b[39m]; log p(e|m)= -0.968:0.000[\u001b[32m0.968\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.049]; Chechen[0.046]; military[0.045]; region[0.045]; Russian[0.043]; Russia[0.042]; people[0.042]; separatist[0.042]; political[0.042]; conflict[0.042]; tens[0.042]; peace[0.042]; peace[0.042]; month[0.041]; Maskhadov[0.041]; Maskhadov[0.041]; statement[0.040]; said[0.040]; thousands[0.039]; rebel[0.039]; settlement[0.039]; settlement[0.039]; leader[0.039]; hoped[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Aslan Maskhadov\u001b[39m ==> ENTITY: \u001b[32mAslan Maskhadov\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.142:0.142[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.045]; military[0.045]; leader[0.045]; Maskhadov[0.044]; Chechen[0.043]; Chechens[0.043]; month[0.042]; conflict[0.042]; peace[0.042]; peace[0.042]; separatist[0.042]; Khasavyurt[0.042]; political[0.041]; said[0.041]; peacemaker[0.041]; Lebed[0.040]; Lebed[0.040]; Lebed[0.040]; Lebed[0.040]; rebel[0.040]; outside[0.040]; Friday[0.040]; Russian[0.040]; region[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lebed\u001b[39m ==> ENTITY: \u001b[32mAlexander Lebed\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebed[0.045]; Lebed[0.045]; Lebed[0.045]; Russian[0.044]; military[0.044]; Russia[0.043]; Chechnya[0.043]; Maskhadov[0.043]; Maskhadov[0.043]; separatist[0.042]; month[0.042]; political[0.042]; Chechen[0.042]; Khasavyurt[0.041]; Chechens[0.041]; leader[0.041]; rebel[0.040]; conflict[0.040]; Alexander[0.040]; said[0.039]; people[0.039]; region[0.039]; died[0.039]; peacemaker[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alexander Lebed\u001b[39m ==> ENTITY: \u001b[32mAlexander Lebed\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebed[0.045]; Lebed[0.045]; Lebed[0.045]; Russian[0.045]; military[0.044]; Russia[0.043]; Chechnya[0.043]; Maskhadov[0.043]; Maskhadov[0.043]; separatist[0.042]; month[0.042]; political[0.042]; Chechen[0.042]; Khasavyurt[0.041]; Chechens[0.041]; leader[0.041]; rebel[0.040]; conflict[0.040]; said[0.039]; people[0.039]; region[0.039]; died[0.039]; peacemaker[0.039]; sign[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechens[0.047]; Chechen[0.046]; military[0.045]; region[0.044]; Russian[0.043]; Russia[0.042]; people[0.042]; separatist[0.042]; political[0.042]; conflict[0.042]; tens[0.042]; peace[0.042]; peace[0.042]; Khasavyurt[0.041]; month[0.041]; Maskhadov[0.041]; Maskhadov[0.041]; statement[0.040]; thousands[0.040]; said[0.040]; rebel[0.039]; leader[0.039]; peacemaker[0.038]; day[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lebed\u001b[39m ==> ENTITY: \u001b[32mAlexander Lebed\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebed[0.045]; Lebed[0.045]; Lebed[0.045]; Russian[0.044]; military[0.044]; Russia[0.043]; Chechnya[0.043]; Maskhadov[0.043]; Maskhadov[0.043]; separatist[0.042]; month[0.042]; political[0.042]; Chechen[0.042]; Khasavyurt[0.041]; Chechens[0.041]; leader[0.041]; rebel[0.040]; conflict[0.040]; Alexander[0.040]; said[0.039]; people[0.039]; region[0.039]; died[0.039]; peacemaker[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.047]; Chechnya[0.045]; region[0.044]; Chechens[0.043]; month[0.043]; Alexander[0.042]; round[0.042]; Chechen[0.042]; military[0.041]; Lebed[0.041]; Lebed[0.041]; Lebed[0.041]; Lebed[0.041]; separatist[0.041]; peace[0.041]; peace[0.041]; tens[0.041]; start[0.040]; leader[0.040]; sign[0.040]; conflict[0.040]; Maskhadov[0.040]; Maskhadov[0.040]; outside[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lebed\u001b[39m ==> ENTITY: \u001b[32mAlexander Lebed\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLebed[0.045]; Lebed[0.045]; Lebed[0.045]; Russian[0.044]; military[0.044]; Russia[0.043]; Chechnya[0.043]; Maskhadov[0.043]; Maskhadov[0.043]; separatist[0.042]; month[0.042]; political[0.042]; Chechen[0.042]; Khasavyurt[0.041]; Chechens[0.041]; leader[0.041]; rebel[0.040]; conflict[0.040]; Alexander[0.040]; said[0.039]; people[0.039]; region[0.039]; died[0.039]; peacemaker[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chechen ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.159:0.159[0]; log p(e|m)= -0.863:-0.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.049]; Chechens[0.047]; military[0.044]; region[0.044]; Russian[0.043]; Russia[0.042]; people[0.042]; conflict[0.042]; separatist[0.042]; political[0.042]; tens[0.042]; peace[0.041]; peace[0.041]; Khasavyurt[0.041]; month[0.041]; Maskhadov[0.041]; Maskhadov[0.041]; statement[0.040]; said[0.040]; thousands[0.039]; rebel[0.039]; settlement[0.039]; settlement[0.039]; leader[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Khasavyurt\u001b[39m ==> ENTITY: \u001b[32mKhasavyurt\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechens[0.046]; Chechnya[0.045]; Russian[0.044]; Russia[0.044]; Chechen[0.044]; framework[0.043]; region[0.042]; separatist[0.042]; military[0.042]; rebel[0.042]; Maskhadov[0.041]; Maskhadov[0.041]; conflict[0.041]; Friday[0.041]; people[0.041]; month[0.041]; day[0.041]; outside[0.040]; Lebed[0.040]; Lebed[0.040]; Lebed[0.040]; Lebed[0.040]; said[0.040]; peacemaker[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Maskhadov\u001b[39m ==> ENTITY: \u001b[32mAslan Maskhadov\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.142:0.142[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.045]; military[0.045]; leader[0.045]; Maskhadov[0.044]; Chechen[0.043]; Chechens[0.043]; month[0.042]; conflict[0.042]; peace[0.042]; peace[0.042]; separatist[0.042]; Khasavyurt[0.042]; political[0.041]; said[0.041]; peacemaker[0.041]; Lebed[0.040]; Lebed[0.040]; Lebed[0.040]; Lebed[0.040]; rebel[0.040]; outside[0.040]; Friday[0.040]; Russian[0.040]; region[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.119:0.119[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.049]; Chechnya[0.045]; region[0.044]; Chechens[0.044]; Alexander[0.043]; Chechen[0.043]; month[0.043]; Lebed[0.041]; Lebed[0.041]; Lebed[0.041]; Lebed[0.041]; military[0.041]; separatist[0.041]; peace[0.041]; peace[0.041]; tens[0.041]; leader[0.040]; conflict[0.040]; Maskhadov[0.040]; Maskhadov[0.040]; people[0.040]; started[0.040]; outside[0.040]; died[0.040]; \t\n\r [=======================>......................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 10s105ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2550/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1037testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.110:0.110[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncommunity[0.044]; land[0.043]; town[0.043]; town[0.043]; island[0.043]; tiny[0.043]; beneath[0.042]; Arctic[0.042]; Arctic[0.042]; came[0.042]; people[0.042]; chilly[0.042]; rest[0.041]; Rolf[0.041]; close[0.041]; Thursday[0.041]; work[0.041]; knit[0.041]; handful[0.041]; flags[0.041]; remember[0.040]; half[0.040]; said[0.040]; bringing[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ukrainians ==> ENTITY: \u001b[32mUkraine\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.099:0.099[0]; log p(e|m)= -1.852:-1.852[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussians[0.048]; Russian[0.047]; communities[0.044]; people[0.042]; people[0.042]; people[0.042]; lived[0.042]; said[0.042]; said[0.042]; soccer[0.042]; local[0.041]; live[0.041]; Norwegian[0.041]; Norwegian[0.041]; work[0.041]; towns[0.041]; survivors[0.040]; protecting[0.040]; Imagine[0.040]; Thursday[0.040]; Thursday[0.040]; settlement[0.040]; bears[0.040]; officials[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.141:0.141[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.050]; international[0.045]; Barentsburg[0.044]; fjord[0.043]; northern[0.043]; summer[0.042]; round[0.042]; wooden[0.042]; mountainous[0.042]; Spitzbergen[0.041]; town[0.041]; island[0.041]; nations[0.041]; church[0.041]; Russia[0.041]; Russia[0.041]; time[0.040]; lies[0.040]; place[0.040]; months[0.040]; breakfast[0.040]; miles[0.040]; people[0.039]; treaty[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.133:0.133[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.050]; Russia[0.049]; Ukrainian[0.047]; Moscow[0.047]; church[0.042]; sent[0.041]; Barentsburg[0.041]; Barentsburg[0.041]; Norwegian[0.041]; Norwegian[0.041]; soccer[0.041]; said[0.040]; said[0.040]; given[0.040]; wooden[0.040]; left[0.040]; place[0.040]; time[0.040]; time[0.040]; houses[0.040]; home[0.040]; having[0.040]; winter[0.040]; asked[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.140:0.140[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.050]; Russians[0.049]; Ukrainian[0.047]; Ukrainians[0.045]; work[0.041]; Barentsburg[0.041]; Barentsburg[0.041]; Barentsburg[0.041]; words[0.041]; village[0.041]; lived[0.041]; people[0.041]; people[0.041]; bears[0.041]; Norwegian[0.041]; Norwegian[0.041]; Norwegian[0.041]; soccer[0.040]; said[0.040]; said[0.040]; said[0.040]; settlement[0.040]; asked[0.040]; towns[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Barentsburg\u001b[39m ==> ENTITY: \u001b[32mBarentsburg\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpitzbergen[0.052]; Arctic[0.051]; Barentsburg[0.049]; polar[0.048]; Norwegian[0.048]; Norwegian[0.048]; island[0.046]; island[0.046]; settlement[0.046]; communities[0.045]; east[0.044]; mountain[0.044]; Johan[0.044]; officials[0.044]; bears[0.044]; site[0.044]; winter[0.043]; summer[0.043]; Russians[0.043]; towns[0.043]; work[0.043]; mining[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.077:0.077[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntold[0.044]; reporters[0.044]; parts[0.044]; local[0.043]; luggage[0.043]; towns[0.042]; said[0.042]; said[0.042]; site[0.042]; Thursday[0.042]; Thursday[0.042]; east[0.041]; trouble[0.041]; mining[0.041]; work[0.041]; bears[0.041]; remember[0.041]; people[0.041]; people[0.041]; people[0.041]; people[0.041]; finding[0.041]; Ukrainians[0.041]; lot[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.075:0.075[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.046]; Russia[0.046]; home[0.044]; place[0.044]; miles[0.044]; Barentsburg[0.043]; Barentsburg[0.043]; Ukrainian[0.043]; people[0.042]; town[0.042]; muddy[0.041]; waters[0.041]; summer[0.041]; km[0.040]; total[0.040]; reporters[0.040]; left[0.040]; wooden[0.040]; village[0.040]; fjord[0.040]; houses[0.040]; lies[0.040]; northern[0.040]; church[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norwegian ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.607:-0.607[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorwegian[0.049]; ate[0.045]; communities[0.044]; asked[0.042]; town[0.042]; lived[0.042]; summer[0.042]; words[0.042]; told[0.041]; island[0.041]; village[0.041]; home[0.041]; live[0.041]; visit[0.041]; ride[0.041]; having[0.041]; winter[0.041]; given[0.041]; know[0.040]; people[0.040]; people[0.040]; finding[0.040]; church[0.040]; feels[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.104:0.104[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.049]; international[0.048]; Norway[0.044]; Norway[0.044]; northern[0.043]; nations[0.042]; sees[0.042]; round[0.042]; rules[0.041]; total[0.041]; group[0.041]; gave[0.041]; summer[0.041]; months[0.041]; treaty[0.041]; Inhabited[0.040]; establish[0.040]; plunged[0.040]; endures[0.040]; chosen[0.040]; winter[0.040]; lies[0.039]; island[0.039]; terms[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pyramiden\u001b[39m ==> ENTITY: \u001b[32mPyramiden\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBarentsburg[0.047]; Barentsburg[0.047]; Norwegian[0.043]; Norwegian[0.043]; sealed[0.043]; work[0.042]; site[0.042]; Russian[0.042]; Spitzbergen[0.042]; Johan[0.042]; east[0.041]; island[0.041]; island[0.041]; Russians[0.041]; said[0.041]; said[0.041]; Ukrainians[0.041]; settlement[0.041]; polar[0.040]; lot[0.040]; bears[0.040]; mining[0.040]; towns[0.040]; wrecked[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Arctic ==> ENTITY: \u001b[32mArctic\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.107:0.107[0]; log p(e|m)= -0.102:-0.102[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArctic[0.053]; Norway[0.046]; land[0.043]; island[0.043]; chilly[0.042]; tiny[0.042]; beneath[0.042]; town[0.042]; town[0.042]; life[0.042]; coal[0.041]; sky[0.041]; close[0.041]; half[0.041]; sight[0.040]; plane[0.040]; Russian[0.040]; windblown[0.040]; people[0.040]; Thursday[0.040]; rest[0.040]; flying[0.040]; said[0.040]; bringing[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.087:0.087[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.044]; land[0.043]; work[0.043]; Arctic[0.043]; Arctic[0.043]; Arctic[0.043]; beneath[0.042]; lot[0.042]; came[0.042]; people[0.042]; town[0.041]; town[0.041]; mourning[0.041]; tiny[0.041]; life[0.041]; community[0.041]; east[0.041]; half[0.041]; said[0.041]; island[0.041]; island[0.041]; sight[0.041]; bringing[0.041]; close[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Barentsburg\u001b[39m ==> ENTITY: \u001b[32mBarentsburg\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBarentsburg[0.047]; Norwegian[0.046]; Norwegian[0.046]; island[0.044]; communities[0.043]; Johan[0.042]; town[0.042]; workers[0.042]; winter[0.041]; overnight[0.041]; summer[0.041]; mining[0.041]; plane[0.041]; plane[0.041]; plane[0.041]; plane[0.041]; helicopter[0.040]; hours[0.040]; Russian[0.040]; Russian[0.040]; people[0.040]; sent[0.040]; village[0.040]; airport[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norwegian ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.607:-0.607[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorwegian[0.048]; Arctic[0.046]; Spitzbergen[0.045]; east[0.043]; communities[0.043]; polar[0.043]; island[0.043]; island[0.043]; visit[0.041]; lived[0.041]; towns[0.041]; summer[0.041]; Barentsburg[0.041]; Barentsburg[0.041]; words[0.041]; slope[0.040]; live[0.040]; mountain[0.040]; Russians[0.040]; winter[0.040]; people[0.040]; people[0.040]; people[0.040]; caretaker[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russians ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.128:0.128[0]; log p(e|m)= -1.877:-1.877[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.049]; Ukrainians[0.044]; communities[0.043]; Barentsburg[0.043]; Barentsburg[0.043]; officials[0.042]; towns[0.042]; Arctic[0.042]; live[0.041]; east[0.041]; people[0.041]; people[0.041]; people[0.041]; island[0.041]; island[0.041]; settlement[0.041]; local[0.041]; lived[0.040]; Spitzbergen[0.040]; said[0.040]; said[0.040]; soccer[0.040]; Norwegian[0.040]; Norwegian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.048]; Russian[0.046]; Moscow[0.046]; Ukrainian[0.045]; Norway[0.043]; Norway[0.043]; international[0.043]; Barentsburg[0.042]; northern[0.042]; island[0.041]; Winter[0.041]; round[0.041]; treaty[0.041]; nations[0.040]; total[0.040]; group[0.040]; run[0.040]; summer[0.040]; place[0.040]; months[0.040]; town[0.040]; home[0.040]; day[0.040]; time[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ukrainian ==> ENTITY: \u001b[32mUkraine\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.129:0.129[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.047]; Russian[0.047]; Russia[0.046]; Moscow[0.044]; time[0.042]; time[0.042]; soccer[0.042]; Norwegian[0.042]; Norwegian[0.042]; workers[0.042]; church[0.041]; communities[0.041]; home[0.041]; town[0.041]; said[0.040]; said[0.040]; Barentsburg[0.040]; Barentsburg[0.040]; village[0.040]; place[0.040]; given[0.040]; sent[0.040]; summer[0.039]; asked[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Arctic ==> ENTITY: \u001b[32mArctic\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.102:-0.102[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArctic[0.053]; Norway[0.046]; land[0.043]; island[0.043]; chilly[0.042]; tiny[0.042]; beneath[0.042]; town[0.042]; town[0.042]; life[0.042]; coal[0.041]; sky[0.041]; helicopter[0.041]; close[0.041]; half[0.041]; sight[0.040]; plane[0.040]; Russian[0.040]; windblown[0.040]; people[0.040]; Thursday[0.040]; rest[0.040]; flying[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Barentsburg\u001b[39m ==> ENTITY: \u001b[32mBarentsburg\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBarentsburg[0.046]; Barentsburg[0.046]; polar[0.045]; Norwegian[0.045]; Norwegian[0.045]; Norwegian[0.045]; island[0.043]; settlement[0.043]; communities[0.042]; Johan[0.041]; officials[0.041]; bears[0.041]; site[0.041]; workers[0.040]; winter[0.040]; summer[0.040]; Russians[0.040]; towns[0.040]; work[0.040]; mining[0.040]; mining[0.040]; plane[0.040]; plane[0.040]; hours[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norway ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.227:-0.227[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorway[0.050]; international[0.045]; place[0.044]; fjord[0.043]; nations[0.043]; northern[0.043]; summer[0.042]; wooden[0.042]; Spitzbergen[0.042]; fewer[0.041]; round[0.041]; island[0.041]; day[0.041]; total[0.041]; lies[0.040]; months[0.040]; rules[0.040]; Barentsburg[0.040]; people[0.040]; Russia[0.040]; Russia[0.040]; terms[0.040]; group[0.040]; miles[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Arctic ==> ENTITY: \u001b[32mArctic\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.102:-0.102[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpitzbergen[0.048]; polar[0.047]; bears[0.044]; Barentsburg[0.043]; icy[0.043]; mountain[0.042]; east[0.042]; island[0.042]; Norwegian[0.042]; snow[0.042]; expanses[0.041]; town[0.041]; land[0.041]; life[0.041]; survivors[0.041]; parts[0.040]; Russians[0.040]; helicopter[0.040]; close[0.040]; sight[0.040]; lot[0.040]; Ukrainians[0.040]; plane[0.040]; plane[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Barentsburg\u001b[39m ==> ENTITY: \u001b[32mBarentsburg\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpitzbergen[0.048]; Norway[0.046]; Norway[0.046]; waters[0.044]; fjord[0.043]; miles[0.043]; island[0.043]; northern[0.043]; international[0.042]; treaty[0.042]; mountainous[0.041]; km[0.041]; lies[0.041]; town[0.041]; winter[0.040]; overnight[0.040]; summer[0.040]; place[0.040]; months[0.040]; nations[0.040]; total[0.040]; plane[0.039]; plane[0.039]; Winter[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norwegian ==> ENTITY: \u001b[32mNorway\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.119:0.119[0]; log p(e|m)= -0.607:-0.607[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorwegian[0.051]; Norwegian[0.051]; soccer[0.047]; communities[0.046]; lived[0.044]; towns[0.043]; summer[0.043]; words[0.043]; island[0.043]; village[0.043]; live[0.043]; Russians[0.042]; competing[0.042]; having[0.042]; winter[0.042]; people[0.042]; people[0.042]; loss[0.042]; finding[0.042]; work[0.042]; feels[0.042]; minutes[0.042]; polar[0.042]; \t\n\r [=======================>......................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s978ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2574/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 964testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Milwaukee ==> ENTITY: \u001b[32mMilwaukee\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.100:0.100[0]; log p(e|m)= -0.282:-0.282[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMilwaukee[0.044]; Milwaukee[0.044]; Tuesday[0.044]; Thursday[0.044]; winning[0.044]; Championship[0.043]; professional[0.043]; ahead[0.043]; Wisconsin[0.042]; round[0.042]; lead[0.042]; debut[0.042]; opening[0.041]; million[0.041]; pro[0.041]; Billy[0.041]; Amateur[0.041]; Takes[0.040]; hole[0.040]; unprecedented[0.040]; Bob[0.040]; Greater[0.040]; Estes[0.040]; old[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = PGA Tour\u001b[39m ==> ENTITY: \u001b[32mPGA Tour\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.044]; Golf[0.044]; Golf[0.044]; list[0.043]; fourth[0.043]; Open[0.043]; Open[0.043]; second[0.043]; second[0.043]; week[0.042]; week[0.042]; week[0.042]; players[0.041]; hole[0.041]; Woods[0.041]; strokes[0.041]; skipped[0.040]; World[0.040]; World[0.040]; money[0.040]; place[0.040]; place[0.040]; card[0.040]; eagled[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Estes\u001b[39m ==> ENTITY: \u001b[32mBob Estes\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTexas[0.043]; round[0.043]; round[0.043]; finish[0.043]; record[0.043]; course[0.042]; Golf[0.042]; Tour[0.042]; Scott[0.042]; finished[0.042]; place[0.042]; place[0.042]; Open[0.042]; Open[0.042]; Open[0.042]; Classic[0.041]; tie[0.041]; par[0.041]; Billy[0.041]; career[0.041]; short[0.041]; pro[0.041]; win[0.040]; good[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = World Series of Golf\u001b[39m ==> ENTITY: \u001b[32mWorld Series of Golf\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.048]; Open[0.046]; Woods[0.046]; Golf[0.045]; Tour[0.044]; winner[0.044]; second[0.044]; second[0.044]; fourth[0.044]; winning[0.043]; week[0.043]; week[0.043]; week[0.043]; hole[0.043]; place[0.042]; took[0.042]; players[0.042]; skipped[0.042]; champ[0.042]; Scott[0.042]; Mark[0.042]; Mark[0.042]; Tom[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Woods\u001b[39m ==> ENTITY: \u001b[32mTiger Woods\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.243:0.243[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngolf[0.050]; Golf[0.048]; Woods[0.046]; Woods[0.046]; Woods[0.046]; Amateur[0.043]; Open[0.043]; Tiger[0.042]; round[0.041]; round[0.041]; round[0.041]; professional[0.041]; Championship[0.041]; play[0.040]; final[0.040]; pro[0.040]; Parnevik[0.040]; Today[0.039]; winning[0.039]; strokes[0.038]; million[0.038]; par[0.038]; Thursday[0.038]; lead[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Henke\u001b[39m ==> ENTITY: \u001b[32mNolan Henke\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngolf[0.046]; Amateur[0.045]; Golf[0.045]; Open[0.044]; play[0.043]; round[0.042]; round[0.042]; Championship[0.042]; professional[0.042]; Estes[0.042]; Parnevik[0.042]; Henke[0.041]; Henke[0.041]; pro[0.041]; par[0.041]; eagled[0.041]; lead[0.040]; Woods[0.040]; Woods[0.040]; Woods[0.040]; Woods[0.040]; start[0.040]; start[0.040]; Bob[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Payne Stewart\u001b[39m ==> ENTITY: \u001b[32mPayne Stewart\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScott[0.044]; Golf[0.044]; Golf[0.044]; hole[0.043]; strokes[0.043]; tournament[0.042]; Woods[0.042]; tie[0.042]; second[0.041]; second[0.041]; Mark[0.041]; Mark[0.041]; winning[0.041]; Tour[0.041]; win[0.041]; bogeys[0.041]; Mickelson[0.041]; par[0.041]; fourth[0.041]; victories[0.041]; week[0.041]; week[0.041]; week[0.041]; Phil[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nolan Henke\u001b[39m ==> ENTITY: \u001b[32mNolan Henke\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmateur[0.046]; Golf[0.045]; Open[0.045]; round[0.043]; round[0.043]; Championship[0.043]; professional[0.043]; Estes[0.042]; Parnevik[0.042]; Henke[0.042]; Henke[0.042]; pro[0.042]; par[0.041]; eagled[0.041]; lead[0.041]; Woods[0.040]; Woods[0.040]; Woods[0.040]; Bob[0.040]; winning[0.040]; Waldorf[0.040]; Debut[0.040]; strokes[0.040]; Pro[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andrew Magee\u001b[39m ==> ENTITY: \u001b[32mAndrew Magee\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.044]; Open[0.044]; Tour[0.044]; round[0.044]; round[0.044]; pro[0.043]; course[0.043]; Woods[0.042]; Classic[0.042]; place[0.041]; place[0.041]; hole[0.041]; holes[0.041]; career[0.041]; playing[0.041]; second[0.041]; second[0.041]; week[0.041]; tie[0.041]; win[0.041]; eagled[0.041]; fourth[0.041]; finish[0.040]; Mayfair[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Billy Mayfair\u001b[39m ==> ENTITY: \u001b[32mBilly Mayfair\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.046]; Golf[0.046]; Tour[0.045]; Tour[0.045]; tournament[0.045]; Scott[0.044]; Tom[0.044]; Phil[0.044]; Estes[0.044]; Mark[0.043]; Mark[0.043]; Open[0.043]; Open[0.043]; Open[0.043]; career[0.043]; Woods[0.043]; pro[0.043]; win[0.042]; players[0.042]; victories[0.042]; tie[0.042]; place[0.042]; place[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Woods\u001b[39m ==> ENTITY: \u001b[32mTiger Woods\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.233:0.233[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngolf[0.051]; Woods[0.046]; Woods[0.046]; Amateur[0.044]; Open[0.043]; Tiger[0.042]; game[0.042]; game[0.042]; playing[0.041]; round[0.041]; round[0.041]; round[0.041]; round[0.041]; professional[0.041]; Championship[0.041]; play[0.041]; final[0.040]; pro[0.040]; Parnevik[0.040]; Today[0.040]; winning[0.040]; strokes[0.039]; million[0.039]; Thursday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Billy Andrade\u001b[39m ==> ENTITY: \u001b[32mBilly Andrade\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngolf[0.046]; Golf[0.046]; Championship[0.045]; Amateur[0.044]; round[0.043]; round[0.043]; round[0.043]; play[0.042]; professional[0.042]; Estes[0.042]; Open[0.041]; Bob[0.041]; Woods[0.041]; Woods[0.041]; Woods[0.041]; Woods[0.041]; Tiger[0.041]; pro[0.040]; Parnevik[0.040]; Waldorf[0.039]; Debut[0.039]; position[0.039]; lead[0.039]; winning[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mark Brooks ==> ENTITY: \u001b[32mMark Brooks (golfer)\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.178:-0.178[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMark[0.047]; Scott[0.045]; Tom[0.043]; Open[0.042]; Billy[0.042]; Golf[0.042]; Golf[0.042]; Mayfair[0.042]; Defending[0.042]; Phil[0.041]; tournament[0.041]; Tour[0.041]; second[0.041]; second[0.041]; winner[0.041]; skipped[0.041]; fourth[0.041]; Payne[0.041]; World[0.041]; World[0.041]; took[0.040]; hole[0.040]; winning[0.040]; Stewart[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Greater Milwaukee Open\u001b[39m ==> ENTITY: \u001b[32mU.S. Bank Championship in Milwaukee\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngolf[0.045]; Golf[0.045]; Championship[0.045]; round[0.044]; round[0.044]; play[0.042]; Amateur[0.042]; Woods[0.042]; Woods[0.042]; Woods[0.042]; Woods[0.042]; strokes[0.041]; lead[0.041]; opening[0.041]; opening[0.041]; pro[0.041]; ahead[0.041]; winning[0.041]; professional[0.040]; Estes[0.040]; Thursday[0.040]; debut[0.040]; Parnevik[0.040]; Tuesday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Texas Open\u001b[39m ==> ENTITY: \u001b[32mValero Texas Open\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.237:0.237[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.045]; round[0.044]; Woods[0.043]; second[0.043]; second[0.043]; Open[0.043]; Open[0.043]; Tour[0.042]; fourth[0.042]; players[0.042]; shot[0.041]; shot[0.041]; final[0.041]; Mayfair[0.041]; pro[0.041]; place[0.041]; place[0.041]; Classic[0.041]; playing[0.040]; week[0.040]; win[0.040]; victories[0.040]; Estes[0.040]; course[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tom Lehman\u001b[39m ==> ENTITY: \u001b[32mTom Lehman\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.045]; second[0.044]; second[0.044]; Open[0.044]; Tour[0.043]; strokes[0.043]; Golf[0.043]; Golf[0.043]; Scott[0.042]; tied[0.042]; Mark[0.041]; Mark[0.041]; Woods[0.041]; list[0.041]; players[0.041]; place[0.041]; skipped[0.041]; Hoch[0.041]; champion[0.040]; champion[0.040]; took[0.040]; Phil[0.040]; winning[0.040]; Mickelson[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Woods\u001b[39m ==> ENTITY: \u001b[32mTiger Woods\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.050]; Woods[0.048]; Woods[0.048]; Amateur[0.045]; Open[0.045]; Tiger[0.044]; round[0.043]; professional[0.043]; Championship[0.042]; pro[0.041]; Parnevik[0.041]; winning[0.041]; strokes[0.040]; million[0.040]; par[0.040]; Thursday[0.040]; lead[0.039]; Tuesday[0.039]; ahead[0.039]; Bob[0.039]; hole[0.039]; opening[0.038]; birdie[0.038]; Billy[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Scott Hoch\u001b[39m ==> ENTITY: \u001b[32mScott Hoch\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.246:0.246[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.044]; Golf[0.043]; Golf[0.043]; Mickelson[0.043]; hole[0.043]; Woods[0.043]; players[0.043]; Open[0.042]; Open[0.042]; Open[0.042]; tie[0.042]; par[0.042]; strokes[0.042]; place[0.041]; place[0.041]; tied[0.041]; second[0.041]; second[0.041]; Tour[0.041]; Estes[0.040]; bogeys[0.040]; finish[0.040]; win[0.040]; Phil[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = PGA Tour\u001b[39m ==> ENTITY: \u001b[32mPGA Tour\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.227:0.227[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.044]; fourth[0.043]; Open[0.043]; Open[0.043]; Open[0.043]; second[0.043]; second[0.043]; final[0.042]; round[0.042]; round[0.042]; week[0.042]; players[0.042]; playing[0.041]; victories[0.041]; Classic[0.041]; hole[0.041]; Woods[0.041]; Wednesday[0.041]; win[0.041]; pro[0.040]; World[0.040]; place[0.040]; place[0.040]; finish[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jesper Parnevik\u001b[39m ==> ENTITY: \u001b[32mJesper Parnevik\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngolf[0.046]; Golf[0.045]; Open[0.045]; professional[0.044]; play[0.043]; round[0.042]; round[0.042]; round[0.042]; Amateur[0.042]; Championship[0.042]; Woods[0.042]; Woods[0.042]; Woods[0.042]; Woods[0.042]; strokes[0.042]; pro[0.042]; winning[0.041]; Bob[0.040]; final[0.040]; Tiger[0.039]; debut[0.039]; Pro[0.039]; Billy[0.039]; surprise[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tiger Woods\u001b[39m ==> ENTITY: \u001b[32mTiger Woods\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.235:0.235[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngolf[0.051]; Golf[0.048]; Woods[0.046]; Woods[0.046]; Woods[0.046]; Amateur[0.044]; Open[0.043]; round[0.042]; round[0.042]; professional[0.042]; Championship[0.041]; play[0.041]; pro[0.040]; Parnevik[0.040]; Today[0.040]; winning[0.040]; strokes[0.039]; million[0.039]; par[0.039]; Thursday[0.039]; lead[0.038]; Tuesday[0.038]; day[0.038]; ahead[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWorld Series of Golf\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mWGC-Bridgestone Invitational\u001b[39m <---> \u001b[32mWorld Series of Golf\u001b[39m\t\nSCORES: global= 0.285:0.284[\u001b[31m0.001\u001b[39m]; local()= 0.262:0.176[\u001b[31m0.086\u001b[39m]; log p(e|m)= -0.711:0.000[\u001b[32m0.711\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.045]; Open[0.044]; Open[0.044]; Open[0.044]; Woods[0.043]; Golf[0.043]; Classic[0.042]; Tour[0.042]; Tour[0.042]; winner[0.042]; second[0.042]; second[0.042]; fourth[0.041]; winning[0.041]; win[0.041]; week[0.041]; week[0.041]; week[0.041]; hole[0.040]; place[0.040]; place[0.040]; took[0.040]; players[0.040]; skipped[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Henke\u001b[39m ==> ENTITY: \u001b[32mNolan Henke\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngolf[0.047]; Amateur[0.045]; play[0.043]; Tour[0.043]; round[0.042]; round[0.042]; round[0.042]; Championship[0.042]; playing[0.042]; Estes[0.042]; Parnevik[0.042]; victories[0.042]; Henke[0.041]; pro[0.041]; pro[0.041]; course[0.041]; eagled[0.041]; career[0.041]; Woods[0.040]; Woods[0.040]; start[0.040]; start[0.040]; Bob[0.040]; finished[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nortel Open\u001b[39m ==> ENTITY: \u001b[32mChrysler Classic of Tucson\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.239:0.239[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.045]; Golf[0.045]; Classic[0.044]; Open[0.044]; Open[0.044]; round[0.043]; Tour[0.042]; Tour[0.042]; Woods[0.042]; skipped[0.042]; Estes[0.042]; course[0.042]; winner[0.041]; week[0.041]; place[0.041]; place[0.041]; tie[0.040]; second[0.040]; second[0.040]; winning[0.040]; win[0.040]; finish[0.040]; victories[0.040]; fourth[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.S. Amateur Championship\u001b[39m ==> ENTITY: \u001b[32mUnited States Amateur Championship (golf)\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.224:0.224[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngolf[0.046]; Woods[0.044]; Woods[0.044]; Woods[0.044]; Woods[0.044]; Golf[0.044]; Open[0.044]; round[0.043]; round[0.043]; round[0.043]; professional[0.042]; time[0.041]; Championship[0.041]; Bob[0.041]; Estes[0.040]; Billy[0.040]; final[0.040]; play[0.040]; winning[0.040]; Thursday[0.039]; pro[0.039]; Tiger[0.039]; Tuesday[0.039]; day[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bob Estes\u001b[39m ==> ENTITY: \u001b[32mBob Estes\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngolf[0.049]; round[0.048]; round[0.048]; round[0.048]; Amateur[0.048]; Golf[0.047]; professional[0.046]; strokes[0.046]; Championship[0.046]; Open[0.046]; par[0.045]; Billy[0.045]; Makes[0.045]; pro[0.045]; good[0.044]; long[0.044]; Parnevik[0.044]; play[0.044]; winning[0.043]; ahead[0.043]; Waldorf[0.043]; pace[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Phil Mickelson\u001b[39m ==> ENTITY: \u001b[32mPhil Mickelson\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.264:0.264[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.045]; Golf[0.044]; Golf[0.044]; Open[0.044]; Open[0.044]; Open[0.044]; Woods[0.043]; second[0.042]; second[0.042]; fourth[0.042]; Tour[0.042]; Scott[0.042]; players[0.041]; strokes[0.041]; hole[0.040]; week[0.040]; week[0.040]; week[0.040]; par[0.040]; tie[0.040]; champion[0.040]; champion[0.040]; finish[0.040]; shot[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Henke\u001b[39m ==> ENTITY: \u001b[32mNolan Henke\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmateur[0.046]; Golf[0.046]; Open[0.045]; round[0.043]; Championship[0.043]; professional[0.043]; Estes[0.042]; Parnevik[0.042]; Henke[0.042]; Henke[0.042]; pro[0.042]; par[0.041]; lead[0.041]; Woods[0.041]; Woods[0.041]; Woods[0.041]; Bob[0.040]; winning[0.040]; Nolan[0.040]; Waldorf[0.040]; Debut[0.040]; strokes[0.040]; Pro[0.040]; opening[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Milwaukee ==> ENTITY: \u001b[32mMilwaukee\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.100:0.100[0]; log p(e|m)= -0.282:-0.282[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMilwaukee[0.044]; Milwaukee[0.044]; Tuesday[0.044]; Thursday[0.044]; winning[0.043]; Championship[0.043]; professional[0.042]; ahead[0.042]; Wisconsin[0.042]; round[0.042]; round[0.042]; lead[0.041]; debut[0.041]; opening[0.041]; opening[0.041]; million[0.041]; pro[0.041]; Billy[0.040]; Amateur[0.040]; Takes[0.040]; hole[0.040]; unprecedented[0.040]; Bob[0.040]; left[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = BellSouth Classic\u001b[39m ==> ENTITY: \u001b[32mAT&T Classic\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.230:0.230[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nround[0.045]; round[0.045]; second[0.043]; second[0.043]; Golf[0.043]; Tour[0.043]; fourth[0.042]; Wednesday[0.042]; Woods[0.041]; Open[0.041]; Open[0.041]; Open[0.041]; final[0.041]; finish[0.041]; champion[0.041]; champion[0.041]; playing[0.041]; win[0.041]; week[0.041]; champ[0.041]; victories[0.040]; place[0.040]; place[0.040]; got[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. Open ==> ENTITY: \u001b[32mU.S. Open (golf)\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.211:0.211[0]; log p(e|m)= -0.307:-0.307[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.046]; Golf[0.045]; Golf[0.045]; Woods[0.045]; par[0.044]; strokes[0.044]; hole[0.044]; Open[0.044]; Open[0.044]; Open[0.044]; Mickelson[0.044]; second[0.043]; second[0.043]; fourth[0.043]; shot[0.043]; champion[0.043]; champion[0.043]; Scott[0.042]; bogeys[0.042]; eagled[0.042]; tie[0.042]; Payne[0.042]; Classic[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Henke\u001b[39m ==> ENTITY: \u001b[32mNolan Henke\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngolf[0.046]; Open[0.044]; Open[0.044]; Classic[0.044]; play[0.043]; win[0.043]; Tour[0.043]; round[0.042]; round[0.042]; playing[0.042]; Estes[0.042]; victories[0.041]; tie[0.041]; Henke[0.041]; pro[0.041]; course[0.041]; par[0.041]; eagled[0.040]; career[0.040]; players[0.040]; Woods[0.040]; start[0.040]; start[0.040]; finished[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Duffy Waldorf\u001b[39m ==> ENTITY: \u001b[32mDuffy Waldorf\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngolf[0.045]; Golf[0.044]; professional[0.044]; play[0.044]; round[0.044]; round[0.044]; round[0.044]; Championship[0.043]; Open[0.043]; Amateur[0.042]; Woods[0.042]; Woods[0.042]; Woods[0.042]; Woods[0.042]; strokes[0.042]; pro[0.040]; Debut[0.040]; Parnevik[0.039]; start[0.039]; start[0.039]; finished[0.039]; final[0.039]; Estes[0.039]; Pro[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Woods\u001b[39m ==> ENTITY: \u001b[32mTiger Woods\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.234:0.234[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.048]; Golf[0.048]; tournament[0.044]; Open[0.043]; Open[0.043]; Open[0.043]; Tour[0.042]; Tour[0.042]; fourth[0.042]; playing[0.042]; players[0.041]; second[0.041]; second[0.041]; course[0.041]; Mickelson[0.041]; win[0.040]; pro[0.040]; winning[0.040]; Scott[0.040]; champion[0.040]; champion[0.040]; Phil[0.039]; Classic[0.039]; tie[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wisconsin ==> ENTITY: \u001b[32mWisconsin\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.067:0.067[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChampionship[0.048]; winning[0.044]; Tuesday[0.043]; round[0.043]; round[0.043]; Thursday[0.043]; Amateur[0.042]; Milwaukee[0.042]; Milwaukee[0.042]; Milwaukee[0.042]; debut[0.042]; ahead[0.041]; professional[0.041]; Billy[0.041]; Bob[0.041]; Golf[0.041]; opening[0.041]; opening[0.041]; Nolan[0.040]; left[0.040]; unprecedented[0.040]; shots[0.040]; pro[0.040]; settled[0.040]; \t\n\r [========================>.....................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s814ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2609/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1120testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.165:0.165[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraqi[0.047]; Iraqi[0.047]; Iraqi[0.047]; Iraqi[0.047]; Kurdish[0.044]; Kurdish[0.044]; Arbil[0.042]; Arbil[0.042]; Arbil[0.042]; Britain[0.041]; Britain[0.041]; involvement[0.041]; involvement[0.041]; involvement[0.041]; involvement[0.041]; allies[0.041]; officials[0.040]; said[0.040]; said[0.040]; said[0.040]; Foreign[0.040]; half[0.039]; helicopters[0.039]; backed[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arbil\u001b[39m ==> ENTITY: \u001b[32mErbil\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdish[0.049]; Kurdish[0.049]; Kurds[0.049]; Arbil[0.048]; Arbil[0.048]; Iraq[0.047]; Iraqi[0.046]; Iraqi[0.046]; Iraqi[0.046]; Iraqi[0.046]; Iraqi[0.046]; city[0.043]; city[0.043]; officials[0.042]; Saturday[0.041]; faction[0.040]; lies[0.039]; close[0.039]; Gulf[0.039]; backed[0.039]; rebel[0.039]; half[0.039]; safe[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = British ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.047:0.047[0]; log p(e|m)= -0.489:-0.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.050]; military[0.047]; John[0.044]; helicopters[0.043]; Major[0.043]; Office[0.042]; Office[0.042]; War[0.042]; planes[0.042]; Foreign[0.041]; Foreign[0.041]; artillery[0.041]; area[0.041]; spokesman[0.040]; spokesman[0.040]; tanks[0.040]; French[0.040]; officials[0.040]; Iraqi[0.040]; Iraqi[0.040]; Iraqi[0.040]; Iraqi[0.040]; patrolled[0.040]; attack[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = John Major\u001b[39m ==> ENTITY: \u001b[32mJohn Major\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.135:0.135[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.048]; British[0.048]; Office[0.044]; Office[0.044]; Prime[0.043]; Foreign[0.043]; Foreign[0.043]; Minister[0.042]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; French[0.041]; military[0.041]; declined[0.040]; spokesman[0.040]; spokesman[0.040]; seen[0.040]; close[0.040]; close[0.040]; backed[0.040]; set[0.040]; taken[0.040]; faction[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arbil\u001b[39m ==> ENTITY: \u001b[32mErbil\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.218:0.218[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdish[0.047]; Kurdish[0.047]; Kurds[0.046]; Arbil[0.045]; Arbil[0.045]; Iraq[0.045]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; area[0.043]; military[0.041]; city[0.040]; city[0.040]; officials[0.039]; Saturday[0.039]; faction[0.038]; lies[0.037]; close[0.037]; close[0.037]; Gulf[0.037]; backed[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.206:0.206[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.048]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Kurdish[0.045]; Kurdish[0.045]; Kurds[0.044]; military[0.044]; British[0.042]; Britain[0.041]; Britain[0.041]; Arbil[0.040]; Arbil[0.040]; Arbil[0.040]; artillery[0.039]; said[0.038]; said[0.038]; said[0.038]; helicopters[0.038]; officials[0.037]; Gulf[0.037]; tanks[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.185:0.185[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.050]; Iraqi[0.047]; Iraqi[0.047]; Iraqi[0.047]; Kurdish[0.047]; Kurdish[0.047]; Britain[0.044]; Britain[0.044]; British[0.044]; Arbil[0.042]; Arbil[0.042]; Arbil[0.042]; said[0.040]; said[0.040]; said[0.040]; Gulf[0.039]; Foreign[0.039]; War[0.038]; London[0.038]; allies[0.038]; officials[0.038]; close[0.037]; helicopters[0.037]; seen[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.164:0.164[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraqi[0.047]; Iraqi[0.047]; Iraqi[0.047]; Iraqi[0.047]; Kurdish[0.045]; military[0.045]; helicopters[0.044]; Kurds[0.044]; British[0.044]; British[0.044]; planes[0.043]; Arbil[0.040]; artillery[0.040]; tanks[0.040]; said[0.039]; said[0.039]; said[0.039]; officials[0.039]; Gulf[0.039]; backed[0.038]; Foreign[0.038]; Foreign[0.038]; attack[0.038]; War[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gulf War\u001b[39m ==> ENTITY: \u001b[32mGulf War\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.104:0.104[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.049]; Iraqi[0.049]; Iraqi[0.049]; Iraqi[0.049]; Iraqi[0.049]; Iraqi[0.049]; Iraqi[0.049]; helicopters[0.046]; Britain[0.046]; Kurdish[0.046]; Kurdish[0.046]; British[0.046]; British[0.046]; Kurds[0.045]; planes[0.044]; Arbil[0.043]; Arbil[0.043]; John[0.042]; allies[0.042]; allies[0.042]; tanks[0.042]; artillery[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.048]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Kurdish[0.043]; Kurdish[0.043]; military[0.043]; helicopters[0.042]; Kurds[0.042]; British[0.042]; British[0.042]; planes[0.041]; Britain[0.041]; Britain[0.041]; Arbil[0.039]; Arbil[0.039]; Arbil[0.039]; artillery[0.039]; tanks[0.039]; said[0.038]; said[0.038]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.056:0.056[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.055]; London[0.045]; condemned[0.044]; Iraq[0.044]; Office[0.043]; helicopters[0.042]; Foreign[0.042]; condemn[0.042]; artillery[0.041]; Saturday[0.041]; officials[0.041]; Iraqi[0.040]; Iraqi[0.040]; Iraqi[0.040]; Iraqi[0.040]; attack[0.040]; attack[0.040]; involvement[0.040]; involvement[0.040]; involvement[0.040]; involvement[0.040]; heavy[0.040]; heavy[0.040]; allies[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.073:0.073[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.048]; Britain[0.048]; said[0.043]; said[0.043]; said[0.043]; condemned[0.043]; set[0.042]; Saturday[0.041]; condemn[0.041]; attack[0.041]; attack[0.041]; Iraq[0.041]; officials[0.041]; Foreign[0.041]; Office[0.041]; city[0.040]; city[0.040]; Gulf[0.040]; seen[0.040]; involvement[0.040]; involvement[0.040]; involvement[0.040]; involvement[0.040]; called[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = British ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.061:0.061[0]; log p(e|m)= -0.489:-0.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.063]; military[0.059]; John[0.055]; helicopters[0.053]; Major[0.053]; Office[0.053]; Office[0.053]; War[0.053]; planes[0.052]; Foreign[0.052]; Foreign[0.052]; artillery[0.051]; area[0.051]; Saturday[0.051]; spokesman[0.050]; spokesman[0.050]; tanks[0.050]; French[0.050]; officials[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.122:0.122[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncondemns[0.049]; Iraq[0.049]; condemn[0.048]; Iraqi[0.047]; Iraqi[0.047]; Iraqi[0.047]; Iraqi[0.047]; Iraqi[0.047]; Iraqi[0.047]; condemned[0.046]; Britain[0.045]; Britain[0.045]; British[0.044]; British[0.044]; military[0.044]; officials[0.044]; Foreign[0.044]; Foreign[0.044]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.070:0.070[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJohn[0.045]; British[0.045]; British[0.045]; military[0.044]; War[0.044]; allies[0.043]; control[0.042]; half[0.042]; said[0.042]; said[0.042]; said[0.042]; fighting[0.042]; declined[0.041]; city[0.041]; rebel[0.041]; called[0.041]; Foreign[0.041]; Foreign[0.041]; area[0.040]; seen[0.040]; protect[0.040]; Gulf[0.040]; Major[0.040]; Minister[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iraqi Kurds\u001b[39m ==> ENTITY: \u001b[32mIraqi Kurdistan\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdish[0.050]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Arbil[0.044]; Arbil[0.044]; area[0.043]; military[0.042]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; city[0.040]; city[0.040]; seen[0.040]; British[0.039]; British[0.039]; faction[0.039]; artillery[0.039]; allies[0.039]; allies[0.039]; rebel[0.039]; condemn[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arbil\u001b[39m ==> ENTITY: \u001b[32mErbil\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdish[0.050]; Kurdish[0.050]; Arbil[0.048]; Arbil[0.048]; Iraq[0.048]; Iraqi[0.047]; Iraqi[0.047]; Iraqi[0.047]; Iraqi[0.047]; city[0.043]; city[0.043]; officials[0.042]; Saturday[0.041]; faction[0.041]; lies[0.040]; close[0.040]; backed[0.040]; rebel[0.040]; half[0.039]; safe[0.039]; said[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.059:0.059[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.057]; London[0.046]; condemned[0.046]; Iraq[0.045]; Office[0.045]; helicopters[0.044]; War[0.044]; Foreign[0.044]; condemn[0.043]; artillery[0.043]; Saturday[0.042]; officials[0.042]; called[0.042]; Iraqi[0.042]; Iraqi[0.042]; Iraqi[0.042]; Iraqi[0.042]; attack[0.042]; attack[0.042]; involvement[0.042]; involvement[0.042]; involvement[0.042]; involvement[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.056:0.056[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.057]; War[0.052]; British[0.051]; British[0.051]; Office[0.051]; Office[0.051]; John[0.051]; planes[0.051]; Major[0.051]; helicopters[0.051]; declined[0.050]; set[0.049]; called[0.048]; Foreign[0.048]; Foreign[0.048]; involvement[0.048]; involvement[0.048]; officials[0.048]; area[0.048]; seen[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.211:0.211[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.054]; Iraqi[0.051]; Iraqi[0.051]; Iraqi[0.051]; Iraqi[0.051]; Iraqi[0.051]; Kurdish[0.049]; Kurdish[0.049]; military[0.049]; helicopters[0.047]; Kurds[0.047]; British[0.047]; British[0.047]; planes[0.046]; Britain[0.046]; Britain[0.046]; Arbil[0.044]; Arbil[0.044]; Arbil[0.044]; artillery[0.044]; tanks[0.044]; \t\n\r [========================>.....................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s758ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2631/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1033testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Amsterdam ==> ENTITY: \u001b[32mAmsterdam\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.064:-0.064[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDutch[0.046]; Dutch[0.046]; today[0.043]; needed[0.042]; delayed[0.042]; delayed[0.042]; Friday[0.042]; European[0.042]; time[0.042]; time[0.042]; contract[0.042]; wise[0.041]; begin[0.041]; said[0.041]; said[0.041]; expected[0.041]; chance[0.041]; spokesman[0.041]; spokesman[0.041]; start[0.041]; start[0.041]; rarely[0.040]; activity[0.040]; familiar[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.075:0.075[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nExchange[0.047]; trading[0.044]; trading[0.044]; futures[0.044]; futures[0.044]; traded[0.043]; Investors[0.043]; told[0.042]; today[0.042]; brokers[0.042]; use[0.042]; German[0.042]; European[0.041]; newsroom[0.040]; electronic[0.040]; said[0.040]; said[0.040]; said[0.040]; Options[0.040]; price[0.040]; accustomed[0.040]; needed[0.040]; date[0.040]; portfolios[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Amsterdam ==> ENTITY: \u001b[32mAmsterdam\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.084:0.084[0]; log p(e|m)= -0.064:-0.064[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDutch[0.047]; currently[0.044]; newsroom[0.043]; time[0.043]; van[0.043]; told[0.043]; wise[0.042]; date[0.042]; speak[0.042]; participants[0.042]; ago[0.041]; German[0.041]; weeks[0.041]; commitment[0.041]; reinvigorate[0.041]; start[0.041]; start[0.041]; start[0.041]; said[0.041]; fixed[0.041]; ready[0.040]; failed[0.040]; limited[0.040]; agreed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dutch ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.794:-0.794[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDutch[0.051]; European[0.047]; Amsterdam[0.045]; needed[0.042]; trading[0.041]; trading[0.041]; today[0.041]; time[0.041]; time[0.041]; van[0.041]; said[0.041]; said[0.041]; said[0.041]; expected[0.041]; initiative[0.041]; accustomed[0.040]; Reuters[0.040]; familiar[0.040]; activity[0.040]; changes[0.040]; start[0.040]; start[0.040]; Friday[0.040]; traded[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.048:0.048[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nspeak[0.044]; Dutch[0.043]; today[0.043]; start[0.042]; start[0.042]; start[0.042]; reinvigorate[0.042]; initiative[0.042]; electronic[0.042]; involved[0.042]; limited[0.042]; ready[0.042]; needed[0.042]; weeks[0.042]; expected[0.041]; failed[0.041]; participants[0.041]; use[0.041]; banks[0.041]; trading[0.041]; Reuters[0.040]; currently[0.040]; date[0.040]; ago[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dutch ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.067:0.067[0]; log p(e|m)= -0.794:-0.794[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGerman[0.045]; Amsterdam[0.044]; speak[0.044]; attempt[0.043]; Lex[0.042]; time[0.042]; time[0.042]; failed[0.042]; making[0.042]; actual[0.042]; said[0.041]; said[0.041]; today[0.041]; changes[0.041]; start[0.041]; start[0.041]; start[0.041]; trading[0.041]; van[0.041]; participants[0.041]; fixed[0.041]; limited[0.040]; needed[0.040]; ready[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dutch ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.076:0.076[0]; log p(e|m)= -0.794:-0.794[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDutch[0.051]; European[0.046]; Amsterdam[0.044]; attempt[0.043]; chance[0.042]; time[0.042]; time[0.042]; Friday[0.042]; making[0.042]; rarely[0.042]; start[0.041]; trading[0.041]; trading[0.041]; said[0.041]; said[0.041]; needed[0.040]; ready[0.040]; today[0.040]; initiative[0.040]; accustomed[0.040]; expected[0.040]; familiar[0.040]; activity[0.040]; revival[0.039]; \t\n\r [========================>.....................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s770ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2638/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1118testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.048]; British[0.046]; British[0.046]; member[0.044]; headlines[0.044]; Thursday[0.043]; public[0.043]; said[0.043]; said[0.043]; day[0.042]; launched[0.041]; Saturday[0.041]; days[0.041]; visiting[0.041]; trip[0.041]; trip[0.041]; arrested[0.039]; southern[0.039]; twice[0.039]; spokesman[0.039]; nationwide[0.039]; Trevor[0.039]; Holland[0.039]; Holland[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Worthing ==> ENTITY: \u001b[32mWorthing\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.125:0.125[0]; log p(e|m)= -0.109:-0.109[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.044]; earlier[0.044]; member[0.043]; town[0.043]; London[0.043]; London[0.043]; Britain[0.042]; public[0.042]; park[0.042]; visiting[0.042]; moved[0.042]; day[0.041]; Saturday[0.041]; toilet[0.041]; twice[0.041]; British[0.041]; British[0.041]; coast[0.041]; similar[0.041]; children[0.041]; south[0.041]; captured[0.040]; girls[0.040]; Thursday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Holland ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.152:0.152[0]; log p(e|m)= -2.781:-2.781[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHolland[0.049]; Holland[0.049]; Belgium[0.048]; England[0.047]; Britain[0.046]; twice[0.044]; town[0.041]; southern[0.041]; moved[0.041]; south[0.040]; trip[0.040]; trip[0.040]; visiting[0.040]; headlines[0.040]; Thursday[0.040]; public[0.040]; park[0.039]; girls[0.039]; girls[0.039]; London[0.039]; day[0.039]; old[0.039]; old[0.039]; similar[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.101:0.101[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.048]; Britain[0.047]; member[0.044]; headlines[0.043]; Thursday[0.043]; public[0.043]; said[0.042]; said[0.042]; day[0.042]; moved[0.042]; launched[0.041]; earlier[0.041]; Saturday[0.041]; days[0.041]; south[0.041]; visiting[0.040]; outrage[0.040]; trip[0.040]; trip[0.040]; attention[0.040]; similar[0.040]; incident[0.039]; southern[0.039]; twice[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.093:0.093[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.050]; coast[0.045]; southern[0.045]; London[0.045]; south[0.042]; Belgium[0.042]; headlines[0.042]; trip[0.042]; Holland[0.041]; Holland[0.041]; similar[0.041]; public[0.041]; visiting[0.041]; twice[0.041]; launched[0.041]; nationwide[0.041]; popular[0.040]; member[0.040]; search[0.040]; theme[0.040]; town[0.040]; secure[0.039]; reading[0.039]; earlier[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = British ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.489:-0.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.053]; England[0.052]; London[0.047]; southern[0.044]; children[0.044]; visiting[0.044]; headlines[0.043]; popular[0.043]; days[0.043]; twice[0.043]; member[0.043]; old[0.043]; launched[0.042]; public[0.042]; Holland[0.042]; Holland[0.042]; coast[0.042]; Saturday[0.042]; Trevor[0.042]; day[0.042]; Thursday[0.041]; said[0.041]; said[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Holland ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.154:0.154[0]; log p(e|m)= -2.781:-2.781[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHolland[0.049]; Holland[0.049]; Belgium[0.048]; England[0.047]; Britain[0.046]; twice[0.044]; British[0.042]; British[0.042]; town[0.041]; southern[0.041]; moved[0.040]; south[0.040]; trip[0.040]; trip[0.040]; visiting[0.040]; headlines[0.040]; Thursday[0.040]; public[0.039]; park[0.039]; girls[0.039]; girls[0.039]; London[0.039]; London[0.039]; day[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.126:0.126[0]; log p(e|m)= -0.190:-0.190[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.048]; British[0.045]; British[0.045]; Holland[0.044]; Holland[0.044]; Holland[0.044]; day[0.043]; twice[0.043]; headlines[0.043]; London[0.042]; London[0.042]; Saturday[0.042]; Thursday[0.040]; south[0.040]; town[0.040]; Belgium[0.040]; visiting[0.040]; similar[0.040]; southern[0.040]; nationwide[0.040]; earlier[0.039]; Worthing[0.039]; public[0.039]; secure[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgium ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.114:0.114[0]; log p(e|m)= -0.242:-0.242[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.047]; Holland[0.047]; Holland[0.047]; England[0.046]; sex[0.043]; captured[0.043]; visiting[0.042]; member[0.042]; southern[0.041]; child[0.041]; scandal[0.041]; launched[0.041]; trip[0.041]; Thursday[0.041]; London[0.040]; children[0.040]; earlier[0.040]; secure[0.040]; headlines[0.040]; public[0.039]; gross[0.039]; south[0.039]; girls[0.039]; girls[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = British ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.489:-0.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.052]; England[0.050]; London[0.046]; southern[0.042]; children[0.042]; headlines[0.042]; days[0.042]; twice[0.042]; member[0.041]; old[0.041]; launched[0.041]; public[0.041]; Holland[0.040]; Holland[0.040]; coast[0.040]; Saturday[0.040]; Trevor[0.040]; day[0.040]; Thursday[0.040]; said[0.040]; said[0.040]; trip[0.039]; trip[0.039]; newsagent[0.039]; \t\n\r [========================>.....................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s787ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2648/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 974testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sri Lanka ==> ENTITY: \u001b[32mSri Lanka national cricket team\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.200:0.200[0]; log p(e|m)= -2.263:-2.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCricket[0.055]; Australia[0.054]; Australia[0.054]; Colombo[0.054]; Lanka[0.054]; match[0.053]; overs[0.053]; overs[0.053]; tournament[0.052]; wickets[0.051]; day[0.051]; World[0.048]; Sri[0.048]; Friday[0.048]; scored[0.047]; Overs[0.047]; night[0.046]; limited[0.045]; Series[0.045]; Singer[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.171:0.171[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.059]; match[0.054]; Cricket[0.054]; tournament[0.052]; Colombo[0.051]; day[0.050]; overs[0.049]; overs[0.049]; wickets[0.048]; World[0.047]; Lanka[0.047]; Lanka[0.047]; scored[0.046]; Friday[0.046]; Overs[0.045]; night[0.044]; Series[0.044]; limited[0.043]; Sri[0.042]; Sri[0.042]; Singer[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colombo ==> ENTITY: \u001b[32mColombo\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.194:-0.194[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLanka[0.053]; Lanka[0.053]; Cricket[0.053]; match[0.050]; Australia[0.050]; Australia[0.050]; tournament[0.049]; Sri[0.049]; Sri[0.049]; day[0.047]; World[0.047]; overs[0.047]; overs[0.047]; Series[0.046]; limited[0.046]; wickets[0.046]; Overs[0.046]; Friday[0.046]; scored[0.044]; night[0.043]; Singer[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Singer World Series\u001b[39m ==> ENTITY: \u001b[32mSinger World Series\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.060]; Australia[0.060]; Colombo[0.058]; Cricket[0.058]; Lanka[0.056]; Lanka[0.056]; match[0.053]; Sri[0.053]; Sri[0.053]; overs[0.052]; overs[0.052]; wickets[0.051]; tournament[0.051]; Friday[0.051]; Overs[0.049]; day[0.048]; scored[0.048]; limited[0.048]; night[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sri Lanka ==> ENTITY: \u001b[32mSri Lanka national cricket team\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.200:0.200[0]; log p(e|m)= -2.263:-2.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCricket[0.055]; Australia[0.054]; Australia[0.054]; Colombo[0.054]; Lanka[0.054]; match[0.053]; overs[0.053]; overs[0.053]; tournament[0.052]; wickets[0.051]; day[0.051]; World[0.048]; Sri[0.048]; Friday[0.048]; scored[0.047]; Overs[0.047]; night[0.046]; limited[0.045]; Series[0.045]; Singer[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.171:0.171[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.059]; match[0.054]; Cricket[0.054]; tournament[0.052]; Colombo[0.051]; day[0.050]; overs[0.049]; overs[0.049]; wickets[0.048]; World[0.047]; Lanka[0.047]; Lanka[0.047]; scored[0.046]; Friday[0.046]; Overs[0.045]; night[0.044]; Series[0.044]; limited[0.043]; Sri[0.042]; Sri[0.042]; Singer[0.040]; \t\n\r [========================>.....................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s766ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2654/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1161testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Venice ==> ENTITY: \u001b[32mVenice\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.048]; film[0.045]; film[0.045]; film[0.045]; Venice[0.045]; Venice[0.045]; screened[0.044]; Film[0.044]; starring[0.043]; premiere[0.043]; Michael[0.042]; Michael[0.042]; festival[0.040]; born[0.040]; Britain[0.039]; Neil[0.039]; Festival[0.039]; Saturday[0.039]; Ira[0.039]; Ira[0.039]; life[0.039]; early[0.039]; release[0.039]; Jordan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Aidan Quinn\u001b[39m ==> ENTITY: \u001b[32mAidan Quinn\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nactor[0.048]; actor[0.048]; film[0.044]; film[0.044]; film[0.044]; plays[0.043]; portrays[0.043]; nominated[0.042]; Harry[0.042]; role[0.041]; Oscar[0.041]; Kitty[0.041]; played[0.041]; Ira[0.040]; story[0.040]; Collins[0.040]; Collins[0.040]; Collins[0.040]; Collins[0.040]; Northern[0.040]; portrayal[0.040]; love[0.039]; born[0.039]; wanted[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIRA\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mProvisional Irish Republican Army\u001b[39m <---> \u001b[32mIrish Republican Army\u001b[39m\t\nSCORES: global= 0.261:0.259[\u001b[31m0.002\u001b[39m]; local()= 0.093:0.088[\u001b[31m0.006\u001b[39m]; log p(e|m)= -0.970:-1.053[\u001b[31m0.082\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIrish[0.047]; Ireland[0.046]; Ireland[0.046]; British[0.043]; British[0.043]; wrote[0.042]; Harry[0.042]; Warner[0.042]; screenplay[0.042]; Collins[0.041]; Collins[0.041]; guerilla[0.041]; soldier[0.041]; List[0.040]; Aidan[0.040]; political[0.040]; film[0.040]; film[0.040]; film[0.040]; film[0.040]; Oscar[0.040]; nominated[0.040]; plays[0.040]; sectarian[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Harry Boland\u001b[39m ==> ENTITY: \u001b[32mHarry Boland\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.049]; Dublin[0.048]; born[0.043]; Aidan[0.043]; played[0.042]; statesman[0.042]; actor[0.042]; actor[0.042]; soldier[0.042]; Northern[0.041]; nominated[0.041]; Quinn[0.041]; Kiernan[0.041]; time[0.041]; wanted[0.040]; Kitty[0.040]; Collins[0.040]; Collins[0.040]; Collins[0.040]; Collins[0.040]; developed[0.040]; Noting[0.040]; Leeson[0.040]; friend[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Northern Ireland ==> ENTITY: \u001b[32mNorthern Ireland\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.251:-0.251[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArmy[0.045]; Irish[0.045]; Irish[0.045]; Irish[0.045]; Dublin[0.044]; British[0.042]; British[0.042]; sectarian[0.042]; Britain[0.042]; Michael[0.041]; Michael[0.041]; Neil[0.041]; Collins[0.041]; Collins[0.041]; Collins[0.041]; Italy[0.041]; Jordan[0.041]; Jordan[0.041]; Jordan[0.041]; Jordan[0.041]; history[0.039]; starring[0.039]; Saturday[0.039]; time[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Julia Roberts\u001b[39m ==> ENTITY: \u001b[32mJulia Roberts\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstarring[0.046]; Neil[0.045]; film[0.044]; film[0.044]; film[0.044]; film[0.044]; film[0.044]; Michael[0.043]; Michael[0.043]; premiere[0.043]; screenplay[0.042]; says[0.041]; Neeson[0.040]; Ira[0.040]; Ira[0.040]; defends[0.040]; Film[0.040]; Collins[0.040]; Collins[0.040]; release[0.040]; wrote[0.040]; British[0.039]; British[0.039]; Controversial[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCollins\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMichael Collins (film)\u001b[39m <---> \u001b[32mMichael Collins (Irish leader)\u001b[39m\t\nSCORES: global= 0.275:0.260[\u001b[31m0.015\u001b[39m]; local()= 0.189:0.096[\u001b[31m0.092\u001b[39m]; log p(e|m)= -1.671:-0.734[\u001b[32m0.937\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDublin[0.044]; Ireland[0.044]; Irish[0.043]; film[0.043]; film[0.043]; film[0.043]; film[0.043]; Collins[0.043]; Collins[0.043]; nominated[0.043]; actor[0.042]; actor[0.042]; played[0.042]; time[0.041]; time[0.041]; copied[0.041]; portrays[0.041]; released[0.040]; British[0.040]; extras[0.040]; today[0.040]; Quinn[0.040]; plays[0.040]; performance[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIRA\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mProvisional Irish Republican Army\u001b[39m <---> \u001b[32mIrish Republican Army\u001b[39m\t\nSCORES: global= 0.256:0.255[\u001b[31m0.002\u001b[39m]; local()= 0.105:0.100[\u001b[31m0.005\u001b[39m]; log p(e|m)= -0.970:-1.053[\u001b[31m0.082\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIrish[0.046]; Irish[0.046]; Ireland[0.045]; Ira[0.045]; Dublin[0.044]; Army[0.044]; British[0.043]; starring[0.042]; Neil[0.041]; wrote[0.041]; Britain[0.041]; screenplay[0.041]; Michael[0.041]; Michael[0.041]; Collins[0.041]; Collins[0.041]; film[0.040]; film[0.040]; film[0.040]; film[0.040]; film[0.040]; Liam[0.039]; Julia[0.039]; recounts[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Warner Bros ==> ENTITY: \u001b[32mWarner Bros.\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.126:0.126[0]; log p(e|m)= -0.023:-0.023[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nactor[0.046]; release[0.046]; film[0.045]; film[0.045]; film[0.045]; released[0.044]; Ira[0.043]; portrayal[0.042]; screenplay[0.042]; Michael[0.042]; Britain[0.042]; depiction[0.041]; wrote[0.041]; came[0.040]; early[0.040]; British[0.040]; British[0.040]; Crying[0.040]; today[0.039]; Collins[0.039]; Collins[0.039]; Collins[0.039]; time[0.039]; time[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jordan\u001b[39m ==> ENTITY: \u001b[32mNeil Jordan\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfilm[0.049]; film[0.049]; film[0.049]; film[0.049]; actor[0.047]; actor[0.047]; played[0.047]; Ireland[0.046]; plays[0.046]; Aidan[0.045]; nominated[0.045]; Jordan[0.045]; Jordan[0.045]; story[0.044]; role[0.044]; Quinn[0.044]; Dublin[0.044]; portrays[0.044]; Kiernan[0.043]; Britain[0.043]; Northern[0.043]; List[0.043]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCollins\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMichael Collins (film)\u001b[39m <---> \u001b[32mMichael Collins (Irish leader)\u001b[39m\t\nSCORES: global= 0.271:0.260[\u001b[31m0.011\u001b[39m]; local()= 0.146:0.096[\u001b[31m0.051\u001b[39m]; log p(e|m)= -1.671:-0.734[\u001b[32m0.937\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.046]; film[0.045]; film[0.045]; film[0.045]; Collins[0.045]; Collins[0.045]; Michael[0.044]; scenes[0.042]; Britain[0.042]; extras[0.042]; today[0.041]; reported[0.041]; surrounding[0.041]; story[0.041]; comrades[0.040]; Noting[0.040]; Jordan[0.040]; Jordan[0.040]; historical[0.040]; wanted[0.040]; shooting[0.039]; said[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yitzak Shamir\u001b[39m ==> ENTITY: \u001b[32mYitzhak Shamir\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.142:0.142[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Jordan[0.045]; Jordan[0.045]; Jordan[0.045]; Jordan[0.045]; Jordan[0.045]; statement[0.042]; statement[0.042]; Bros[0.042]; peace[0.042]; said[0.041]; said[0.041]; said[0.041]; Northern[0.041]; saying[0.041]; nominated[0.040]; released[0.040]; political[0.039]; contemporary[0.039]; achieved[0.039]; statesman[0.039]; rival[0.039]; history[0.039]; played[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIrish\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIreland\u001b[39m <---> \u001b[32mRepublic of Ireland\u001b[39m\t\nSCORES: global= 0.251:0.237[\u001b[31m0.014\u001b[39m]; local()= 0.188:0.165[\u001b[31m0.023\u001b[39m]; log p(e|m)= -1.457:-2.254[\u001b[31m0.797\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.060]; Irish[0.053]; Irish[0.053]; Britain[0.050]; British[0.047]; British[0.047]; Michael[0.046]; history[0.046]; contemporary[0.045]; Liam[0.044]; wrote[0.044]; time[0.043]; today[0.043]; Collins[0.043]; Collins[0.043]; Collins[0.043]; early[0.042]; Army[0.042]; fought[0.042]; world[0.042]; Northern[0.041]; Republican[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Irish Republican Army\u001b[39m ==> ENTITY: \u001b[32mIrish Republican Army\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIrish[0.048]; Irish[0.048]; Ireland[0.047]; Dublin[0.046]; British[0.044]; British[0.044]; Britain[0.042]; Liam[0.041]; Collins[0.041]; Collins[0.041]; Michael[0.041]; Michael[0.041]; political[0.041]; Bros[0.040]; Neil[0.040]; sectarian[0.040]; time[0.040]; Northern[0.040]; released[0.040]; politicians[0.040]; release[0.039]; Intelligence[0.039]; controversial[0.039]; festival[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCollins\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMichael Collins (film)\u001b[39m <---> \u001b[32mMichael Collins (Irish leader)\u001b[39m\t\nSCORES: global= 0.276:0.260[\u001b[31m0.015\u001b[39m]; local()= 0.193:0.096[\u001b[31m0.097\u001b[39m]; log p(e|m)= -1.671:-0.734[\u001b[32m0.937\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCollins[0.044]; Collins[0.044]; Collins[0.044]; Dublin[0.044]; Ireland[0.044]; plays[0.043]; film[0.043]; film[0.043]; film[0.043]; nominated[0.042]; born[0.042]; actor[0.042]; actor[0.042]; Quinn[0.041]; played[0.041]; time[0.041]; Boland[0.040]; portrays[0.040]; scenes[0.040]; Aidan[0.040]; extras[0.039]; today[0.039]; performance[0.039]; Oscar[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Venice Film Festival\u001b[39m ==> ENTITY: \u001b[32mVenice Film Festival\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVenice[0.050]; Venice[0.050]; festival[0.044]; film[0.044]; film[0.044]; film[0.044]; film[0.044]; film[0.044]; screened[0.043]; Italy[0.043]; premiere[0.042]; starring[0.041]; screenplay[0.040]; British[0.039]; British[0.039]; Michael[0.039]; Michael[0.039]; Saturday[0.039]; director[0.039]; release[0.039]; depiction[0.039]; Julia[0.038]; Britain[0.038]; controversial[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCollins\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMichael Collins (film)\u001b[39m <---> \u001b[32mMichael Collins (Irish leader)\u001b[39m\t\nSCORES: global= 0.275:0.260[\u001b[31m0.014\u001b[39m]; local()= 0.189:0.099[\u001b[31m0.090\u001b[39m]; log p(e|m)= -1.671:-0.734[\u001b[32m0.937\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.044]; Ireland[0.044]; Irish[0.043]; film[0.043]; film[0.043]; film[0.043]; film[0.043]; Collins[0.043]; Collins[0.043]; nominated[0.043]; actor[0.042]; actor[0.042]; time[0.041]; time[0.041]; copied[0.041]; wrote[0.041]; portrays[0.041]; released[0.040]; British[0.040]; British[0.040]; today[0.040]; Quinn[0.040]; plays[0.040]; performance[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kitty Kiernan\u001b[39m ==> ENTITY: \u001b[32mKitty Kiernan\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.125:0.125[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.046]; Dublin[0.046]; Collins[0.044]; Collins[0.044]; Collins[0.044]; today[0.044]; Quinn[0.042]; Aidan[0.042]; letters[0.042]; story[0.042]; Northern[0.042]; set[0.041]; certain[0.040]; world[0.040]; Boland[0.040]; historical[0.040]; unpaid[0.040]; mysterious[0.040]; time[0.040]; love[0.040]; contemporary[0.040]; born[0.040]; Harry[0.040]; Leeson[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.047]; Army[0.044]; Britain[0.043]; British[0.043]; Venice[0.043]; Venice[0.043]; Venice[0.043]; Irish[0.043]; Irish[0.043]; Michael[0.042]; Michael[0.042]; early[0.042]; born[0.041]; Saturday[0.041]; starring[0.041]; Dublin[0.041]; fan[0.040]; independence[0.040]; Collins[0.040]; Collins[0.040]; Republican[0.040]; film[0.040]; film[0.040]; film[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mOscar\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAcademy Awards\u001b[39m <---> \u001b[32mAcademy Award for Best Actor\u001b[39m\t\nSCORES: global= 0.254:0.231[\u001b[31m0.023\u001b[39m]; local()= 0.156:0.182[\u001b[32m0.026\u001b[39m]; log p(e|m)= -0.265:-4.510[\u001b[31m4.245\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nactor[0.047]; actor[0.047]; film[0.046]; film[0.046]; film[0.046]; film[0.046]; nominated[0.045]; plays[0.042]; portrayal[0.042]; certain[0.041]; played[0.040]; Ira[0.040]; portrays[0.040]; role[0.040]; scenes[0.040]; Harry[0.040]; performance[0.039]; born[0.039]; extras[0.039]; Collins[0.039]; Collins[0.039]; Collins[0.039]; Collins[0.039]; List[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jordan\u001b[39m ==> ENTITY: \u001b[32mNeil Jordan\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfilm[0.045]; film[0.045]; film[0.045]; Irish[0.044]; Irish[0.044]; actor[0.043]; actor[0.043]; Ireland[0.042]; Ireland[0.042]; nominated[0.041]; Jordan[0.041]; Jordan[0.041]; Jordan[0.041]; British[0.041]; British[0.041]; screenplay[0.041]; wrote[0.041]; Bros[0.040]; Warner[0.040]; portrayal[0.040]; Britain[0.039]; early[0.039]; Northern[0.039]; Northern[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.049]; British[0.049]; Neil[0.046]; Ireland[0.046]; Michael[0.043]; Michael[0.043]; history[0.042]; Irish[0.041]; Irish[0.041]; Irish[0.041]; released[0.041]; release[0.040]; controversial[0.040]; Italy[0.040]; Northern[0.040]; ruled[0.040]; early[0.040]; saying[0.040]; born[0.040]; feared[0.040]; Jordan[0.040]; Jordan[0.040]; Jordan[0.040]; Jordan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.104:0.104[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.048]; events[0.043]; events[0.043]; sought[0.043]; based[0.043]; today[0.043]; historical[0.042]; Michael[0.042]; speeches[0.041]; maintained[0.041]; Jordan[0.041]; Jordan[0.041]; important[0.041]; possible[0.041]; film[0.041]; film[0.041]; film[0.041]; said[0.041]; said[0.041]; said[0.041]; reported[0.041]; certain[0.041]; wanted[0.041]; south[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Liam Neeson\u001b[39m ==> ENTITY: \u001b[32mLiam Neeson\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstarring[0.050]; film[0.048]; film[0.048]; film[0.048]; film[0.048]; film[0.048]; Neil[0.048]; Michael[0.047]; Michael[0.047]; premiere[0.046]; Collins[0.045]; Collins[0.045]; says[0.044]; Julia[0.044]; release[0.044]; Army[0.044]; screenplay[0.043]; Roberts[0.043]; screened[0.042]; Northern[0.042]; recounts[0.042]; Film[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJordan[0.046]; Jordan[0.046]; Jordan[0.046]; Jordan[0.046]; Jordan[0.046]; Ireland[0.043]; Shamir[0.043]; played[0.042]; terrorism[0.042]; history[0.042]; British[0.041]; statement[0.041]; statement[0.041]; said[0.041]; said[0.041]; said[0.041]; China[0.041]; peace[0.040]; perceived[0.040]; political[0.040]; time[0.039]; time[0.039]; achieved[0.039]; depiction[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Michael Collins ==> ENTITY: \u001b[32mMichael Collins (film)\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.112:0.112[0]; log p(e|m)= -1.671:-1.671[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.047]; film[0.046]; film[0.046]; film[0.046]; Collins[0.046]; Britain[0.043]; story[0.043]; today[0.042]; surrounding[0.041]; wanted[0.041]; Jordan[0.041]; reported[0.041]; comrades[0.041]; life[0.041]; shooting[0.040]; said[0.040]; said[0.040]; sought[0.040]; negotiated[0.040]; dramatically[0.039]; speeches[0.039]; possible[0.039]; true[0.039]; based[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = British ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.141:0.141[0]; log p(e|m)= -0.489:-0.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.050]; British[0.046]; film[0.045]; film[0.045]; film[0.045]; film[0.045]; starring[0.044]; Ireland[0.044]; Irish[0.042]; Irish[0.042]; Irish[0.042]; Michael[0.041]; China[0.040]; history[0.039]; independence[0.039]; independence[0.039]; independence[0.039]; Film[0.039]; screenplay[0.039]; Warner[0.039]; Jordan[0.039]; Jordan[0.039]; Jordan[0.039]; Jordan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Roberts\u001b[39m ==> ENTITY: \u001b[32mJulia Roberts\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nactor[0.047]; actor[0.047]; film[0.044]; film[0.044]; film[0.044]; film[0.044]; Oscar[0.043]; story[0.042]; Harry[0.042]; nominated[0.042]; plays[0.042]; portrays[0.041]; scenes[0.041]; List[0.041]; Kitty[0.041]; role[0.040]; extras[0.040]; love[0.040]; Collins[0.040]; Collins[0.040]; Collins[0.040]; played[0.039]; make[0.039]; think[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dublin ==> ENTITY: \u001b[32mDublin\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.053]; played[0.046]; Aidan[0.045]; Quinn[0.044]; location[0.043]; Boland[0.042]; today[0.042]; Collins[0.042]; Collins[0.042]; Collins[0.042]; Kiernan[0.042]; born[0.041]; based[0.041]; time[0.041]; plays[0.041]; Jordan[0.040]; Jordan[0.040]; Jordan[0.040]; Jordan[0.040]; historical[0.040]; Kitty[0.040]; set[0.039]; Noting[0.039]; citizens[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mao Tse-Tung\u001b[39m ==> ENTITY: \u001b[32mMao Zedong\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.115:0.115[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.048]; revolutionary[0.045]; Tung[0.044]; Bros[0.043]; history[0.042]; saying[0.042]; Northern[0.042]; guerilla[0.042]; independence[0.042]; independence[0.042]; wrote[0.041]; political[0.041]; British[0.041]; copied[0.041]; perceived[0.041]; world[0.040]; Jordan[0.040]; Jordan[0.040]; Jordan[0.040]; Jordan[0.040]; contemporary[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jordan\u001b[39m ==> ENTITY: \u001b[32mNeil Jordan\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.247:0.247[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstarring[0.044]; Michael[0.044]; Michael[0.044]; film[0.044]; film[0.044]; film[0.044]; film[0.044]; film[0.044]; Neil[0.044]; Irish[0.042]; Irish[0.042]; Irish[0.042]; premiere[0.042]; Ireland[0.041]; Julia[0.040]; Film[0.040]; Jordan[0.040]; Jordan[0.040]; Jordan[0.040]; British[0.040]; British[0.040]; screenplay[0.039]; wrote[0.039]; Liam[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMichael Collins\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMichael Collins (film)\u001b[39m <---> \u001b[32mMichael Collins (Irish leader)\u001b[39m\t\nSCORES: global= 0.277:0.262[\u001b[31m0.015\u001b[39m]; local()= 0.220:0.125[\u001b[31m0.094\u001b[39m]; log p(e|m)= -1.671:-0.734[\u001b[32m0.937\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMichael[0.046]; Collins[0.045]; Dublin[0.045]; Ireland[0.045]; Irish[0.044]; Irish[0.044]; Irish[0.044]; film[0.044]; film[0.044]; film[0.044]; film[0.044]; film[0.044]; starring[0.044]; Neil[0.044]; born[0.043]; release[0.042]; Liam[0.042]; time[0.042]; wrote[0.042]; premiere[0.042]; Neeson[0.042]; Army[0.042]; Britain[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jordan\u001b[39m ==> ENTITY: \u001b[32mNeil Jordan\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfilm[0.052]; film[0.052]; film[0.052]; actor[0.049]; actor[0.049]; played[0.049]; Ireland[0.048]; plays[0.048]; Aidan[0.047]; nominated[0.047]; Jordan[0.047]; Jordan[0.047]; Jordan[0.047]; role[0.046]; Quinn[0.046]; Dublin[0.046]; portrays[0.046]; Kiernan[0.045]; portrayal[0.045]; Northern[0.045]; List[0.045]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIRA\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mProvisional Irish Republican Army\u001b[39m <---> \u001b[32mIrish Republican Army\u001b[39m\t\nSCORES: global= 0.257:0.256[\u001b[31m0.002\u001b[39m]; local()= 0.106:0.099[\u001b[31m0.007\u001b[39m]; log p(e|m)= -0.970:-1.053[\u001b[31m0.082\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIrish[0.047]; Irish[0.047]; Ira[0.046]; Dublin[0.045]; Army[0.045]; starring[0.043]; Neil[0.042]; Britain[0.042]; Michael[0.042]; Michael[0.042]; Collins[0.041]; Collins[0.041]; film[0.041]; film[0.041]; film[0.041]; Jordan[0.040]; Liam[0.040]; Julia[0.040]; premiere[0.040]; Film[0.040]; release[0.039]; early[0.039]; festival[0.039]; Intelligence[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Neil Jordan\u001b[39m ==> ENTITY: \u001b[32mNeil Jordan\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.217:0.217[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstarring[0.045]; Michael[0.045]; Michael[0.045]; film[0.045]; film[0.045]; film[0.045]; film[0.045]; Irish[0.043]; Irish[0.043]; premiere[0.042]; Ireland[0.041]; screened[0.041]; Julia[0.041]; Film[0.040]; Jordan[0.040]; British[0.040]; Liam[0.040]; director[0.040]; Army[0.040]; Dublin[0.040]; Neeson[0.039]; Britain[0.039]; early[0.039]; Northern[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = The Crying Game\u001b[39m ==> ENTITY: \u001b[32mThe Crying Game\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.050]; British[0.050]; film[0.049]; film[0.049]; film[0.049]; Britain[0.047]; plays[0.046]; Irish[0.045]; screenplay[0.045]; nominated[0.045]; actor[0.045]; actor[0.045]; release[0.045]; Ireland[0.044]; Ireland[0.044]; released[0.044]; Bros[0.043]; portrayal[0.043]; wrote[0.043]; sympathetic[0.043]; Oscar[0.043]; political[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jordan\u001b[39m ==> ENTITY: \u001b[32mNeil Jordan\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.110:0.110[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMichael[0.050]; film[0.049]; film[0.049]; Ireland[0.046]; story[0.044]; Britain[0.043]; Collins[0.042]; Collins[0.042]; sought[0.041]; wanted[0.041]; said[0.040]; said[0.040]; events[0.040]; today[0.040]; ll[0.040]; comrades[0.040]; speeches[0.040]; dramatically[0.039]; life[0.039]; extensive[0.039]; broken[0.039]; assumptions[0.039]; think[0.039]; important[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jordan\u001b[39m ==> ENTITY: \u001b[32mNeil Jordan\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfilm[0.045]; film[0.045]; film[0.045]; film[0.045]; Irish[0.043]; actor[0.043]; actor[0.043]; played[0.043]; Ireland[0.042]; plays[0.042]; Aidan[0.041]; nominated[0.041]; Jordan[0.041]; Jordan[0.041]; Jordan[0.041]; Jordan[0.041]; British[0.041]; role[0.040]; Quinn[0.040]; Dublin[0.040]; Bros[0.040]; portrays[0.040]; Warner[0.039]; Kiernan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ireland ==> ENTITY: \u001b[32mIreland\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.093:0.093[0]; log p(e|m)= -0.566:-0.566[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.051]; Michael[0.046]; today[0.043]; Collins[0.043]; Collins[0.043]; north[0.043]; south[0.043]; lost[0.043]; surrounding[0.042]; important[0.041]; Jordan[0.041]; making[0.041]; sought[0.041]; make[0.040]; possible[0.040]; comrades[0.040]; film[0.040]; film[0.040]; said[0.040]; life[0.040]; events[0.040]; death[0.039]; think[0.039]; settlement[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jordan\u001b[39m ==> ENTITY: \u001b[32mNeil Jordan\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMichael[0.046]; film[0.045]; film[0.045]; film[0.045]; Irish[0.044]; Irish[0.044]; Irish[0.044]; Ireland[0.042]; Julia[0.041]; Jordan[0.041]; Jordan[0.041]; Jordan[0.041]; British[0.041]; British[0.041]; screenplay[0.041]; wrote[0.041]; Army[0.040]; Bros[0.040]; Neeson[0.040]; Warner[0.040]; portrayal[0.040]; Britain[0.039]; early[0.039]; Northern[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIrish\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIreland\u001b[39m <---> \u001b[32mRepublic of Ireland\u001b[39m\t\nSCORES: global= 0.249:0.236[\u001b[31m0.014\u001b[39m]; local()= 0.207:0.187[\u001b[31m0.020\u001b[39m]; log p(e|m)= -1.457:-2.254[\u001b[31m0.797\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.054]; Irish[0.048]; Irish[0.048]; Dublin[0.046]; Britain[0.046]; British[0.043]; British[0.043]; Michael[0.042]; Michael[0.042]; history[0.041]; born[0.041]; Neil[0.041]; Liam[0.040]; Italy[0.040]; wrote[0.040]; time[0.039]; Collins[0.039]; Collins[0.039]; early[0.039]; Army[0.038]; fought[0.038]; Republican[0.038]; said[0.038]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jordan\u001b[39m ==> ENTITY: \u001b[32mNeil Jordan\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfilm[0.049]; film[0.049]; film[0.049]; film[0.049]; actor[0.047]; actor[0.047]; played[0.047]; Ireland[0.046]; plays[0.046]; Aidan[0.045]; nominated[0.045]; Jordan[0.045]; Jordan[0.045]; Jordan[0.045]; story[0.044]; role[0.044]; Quinn[0.044]; Dublin[0.044]; portrays[0.044]; Kiernan[0.043]; Northern[0.043]; List[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCollins\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMichael Collins (film)\u001b[39m <---> \u001b[32mMichael Collins (Irish leader)\u001b[39m\t\nSCORES: global= 0.275:0.259[\u001b[31m0.016\u001b[39m]; local()= 0.181:0.080[\u001b[31m0.102\u001b[39m]; log p(e|m)= -1.671:-0.734[\u001b[32m0.937\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDublin[0.045]; Ireland[0.044]; film[0.043]; film[0.043]; film[0.043]; film[0.043]; Collins[0.043]; Collins[0.043]; nominated[0.043]; actor[0.042]; actor[0.042]; played[0.042]; time[0.041]; portrays[0.041]; scenes[0.041]; Britain[0.041]; story[0.041]; extras[0.040]; today[0.040]; Quinn[0.040]; plays[0.040]; performance[0.040]; mysterious[0.040]; Aidan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfilm[0.046]; film[0.046]; film[0.046]; film[0.046]; Mao[0.043]; Tse[0.043]; history[0.042]; Ireland[0.042]; warfare[0.042]; Tung[0.042]; played[0.041]; guerilla[0.041]; British[0.041]; Jordan[0.040]; Jordan[0.040]; Jordan[0.040]; Jordan[0.040]; Jordan[0.040]; actor[0.040]; actor[0.040]; Northern[0.040]; world[0.040]; nominated[0.040]; contemporary[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Michael Collins ==> ENTITY: \u001b[32mMichael Collins (film)\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.208:0.208[0]; log p(e|m)= -1.671:-1.671[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDublin[0.045]; Ireland[0.044]; Irish[0.043]; Irish[0.043]; film[0.043]; film[0.043]; film[0.043]; film[0.043]; starring[0.043]; Collins[0.043]; Neil[0.043]; Michael[0.042]; release[0.041]; wrote[0.041]; premiere[0.041]; Neeson[0.041]; Army[0.041]; Britain[0.040]; British[0.040]; Liam[0.040]; defends[0.039]; Intelligence[0.039]; screenplay[0.039]; Film[0.039]; \t\n\r [========================>.....................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s504ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2699/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1013testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Libya ==> ENTITY: \u001b[32mLibya\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.125:0.125[0]; log p(e|m)= -0.305:-0.305[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLibya[0.053]; Libya[0.053]; Libya[0.053]; government[0.049]; rights[0.047]; said[0.046]; human[0.046]; Friday[0.046]; noted[0.045]; leader[0.044]; passport[0.044]; Department[0.044]; Department[0.044]; spokesman[0.043]; view[0.043]; restriction[0.043]; accepting[0.043]; travel[0.043]; refused[0.043]; prize[0.042]; Washington[0.042]; awarded[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.109:0.109[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstates[0.046]; pledged[0.046]; Department[0.046]; Department[0.046]; Department[0.046]; rally[0.045]; said[0.044]; said[0.044]; government[0.044]; Wednesday[0.043]; international[0.043]; January[0.043]; October[0.043]; Nation[0.043]; flight[0.043]; travel[0.042]; March[0.042]; award[0.042]; black[0.042]; receive[0.042]; led[0.042]; view[0.041]; denied[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.S. Treasury Department\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of the Treasury\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDepartment[0.046]; Department[0.046]; Department[0.046]; Department[0.046]; Treasury[0.044]; Washington[0.044]; Washington[0.044]; January[0.041]; said[0.041]; said[0.041]; noted[0.041]; noted[0.041]; Wednesday[0.041]; government[0.040]; October[0.040]; pledged[0.040]; State[0.040]; State[0.040]; billion[0.040]; accepting[0.040]; awarded[0.040]; receive[0.039]; international[0.039]; March[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = State Department\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of State\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \naward[0.047]; award[0.047]; Department[0.046]; Department[0.046]; Washington[0.046]; Washington[0.046]; awarded[0.044]; receive[0.043]; State[0.041]; Libya[0.040]; Libya[0.040]; Libya[0.040]; Libya[0.040]; view[0.040]; noted[0.040]; government[0.040]; January[0.040]; Wednesday[0.040]; passport[0.039]; Nation[0.039]; said[0.039]; application[0.039]; Treasury[0.039]; October[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scotland ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.054:0.054[0]; log p(e|m)= -0.229:-0.229[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninternational[0.045]; men[0.044]; Washington[0.043]; leader[0.043]; January[0.043]; denied[0.042]; group[0.042]; Wednesday[0.042]; list[0.042]; October[0.042]; award[0.041]; Department[0.041]; pledged[0.041]; March[0.041]; receive[0.041]; black[0.041]; Nation[0.041]; Libya[0.041]; Libya[0.041]; states[0.041]; Libyan[0.040]; Libyan[0.040]; sponsor[0.040]; Lockerbie[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Louis Farrakhan\u001b[39m ==> ENTITY: \u001b[32mLouis Farrakhan\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFarrakhan[0.050]; Farrakhan[0.050]; Farrakhan[0.050]; denied[0.045]; black[0.045]; Washington[0.044]; leader[0.044]; leader[0.044]; State[0.044]; State[0.044]; noted[0.043]; talked[0.043]; said[0.042]; Department[0.042]; Department[0.042]; Department[0.042]; ve[0.041]; government[0.041]; refused[0.041]; tight[0.041]; award[0.041]; award[0.041]; going[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nation of Islam\u001b[39m ==> ENTITY: \u001b[32mNation of Islam\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.132:0.132[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nblack[0.045]; Farrakhan[0.044]; Farrakhan[0.044]; Farrakhan[0.044]; Farrakhan[0.044]; talked[0.043]; states[0.042]; Washington[0.042]; pledged[0.042]; noted[0.042]; noted[0.042]; said[0.042]; said[0.042]; human[0.042]; denied[0.041]; group[0.041]; award[0.040]; rally[0.040]; terrorism[0.040]; leader[0.040]; rights[0.040]; government[0.040]; men[0.040]; thousands[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Libyan ==> ENTITY: \u001b[32mLibya\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.163:0.163[0]; log p(e|m)= -0.273:-0.273[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLibya[0.047]; Libya[0.047]; Libya[0.047]; Libyan[0.046]; Tripoli[0.045]; Muammar[0.044]; Gaddafi[0.043]; international[0.042]; leader[0.041]; said[0.040]; thousands[0.040]; brought[0.040]; noted[0.040]; states[0.040]; bombing[0.040]; sanctions[0.040]; Islam[0.040]; terrorism[0.040]; travel[0.040]; Farrakhan[0.040]; Farrakhan[0.040]; Farrakhan[0.040]; Washington[0.039]; refused[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Libyan ==> ENTITY: \u001b[32mLibya\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.196:0.196[0]; log p(e|m)= -0.273:-0.273[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLibya[0.046]; Libya[0.046]; Libya[0.046]; Libya[0.046]; Libya[0.046]; Libyan[0.045]; Tripoli[0.044]; Muammar[0.043]; Gaddafi[0.042]; government[0.042]; international[0.041]; leader[0.040]; leader[0.040]; said[0.040]; said[0.040]; thousands[0.039]; brought[0.039]; noted[0.039]; noted[0.039]; states[0.039]; bombing[0.039]; sanctions[0.039]; Islam[0.039]; terrorism[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Farrakhan\u001b[39m ==> ENTITY: \u001b[32mLouis Farrakhan\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFarrakhan[0.047]; Farrakhan[0.047]; Farrakhan[0.047]; Farrakhan[0.047]; Farrakhan[0.047]; Nation[0.042]; denied[0.042]; black[0.041]; black[0.041]; Washington[0.041]; Washington[0.041]; leader[0.041]; leader[0.041]; State[0.040]; State[0.040]; group[0.040]; noted[0.040]; talked[0.040]; Islam[0.040]; rally[0.040]; Louis[0.039]; said[0.039]; said[0.039]; met[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.068:0.068[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWashington[0.045]; case[0.044]; case[0.044]; award[0.043]; award[0.043]; State[0.042]; State[0.042]; Department[0.042]; Department[0.042]; Department[0.042]; noted[0.042]; Wednesday[0.042]; travel[0.041]; speculate[0.041]; speculate[0.041]; elaborate[0.041]; refused[0.041]; Louis[0.040]; black[0.040]; application[0.040]; receive[0.040]; Friday[0.040]; awarded[0.040]; accepting[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Treasury Department\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of the Treasury\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDepartment[0.048]; Treasury[0.046]; Washington[0.046]; Wednesday[0.044]; January[0.044]; government[0.044]; March[0.042]; billion[0.042]; October[0.042]; pledged[0.042]; leader[0.041]; receive[0.041]; imposition[0.040]; application[0.040]; met[0.040]; terrorism[0.040]; noted[0.040]; noted[0.040]; sanctions[0.040]; list[0.040]; said[0.040]; said[0.040]; accepting[0.040]; passport[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Farrakhan\u001b[39m ==> ENTITY: \u001b[32mLouis Farrakhan\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.113:0.113[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFarrakhan[0.049]; Farrakhan[0.049]; black[0.044]; Washington[0.043]; leader[0.043]; State[0.043]; State[0.043]; noted[0.043]; talked[0.042]; Louis[0.042]; said[0.042]; Department[0.041]; Department[0.041]; ve[0.040]; government[0.040]; refused[0.040]; tight[0.040]; award[0.040]; going[0.040]; happen[0.040]; happen[0.040]; Glyn[0.039]; Friday[0.039]; travel[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.118:0.118[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninternational[0.046]; Libya[0.043]; Libya[0.043]; states[0.043]; leader[0.043]; terrorism[0.042]; Islam[0.042]; October[0.042]; March[0.042]; noted[0.042]; January[0.042]; Nation[0.041]; Washington[0.041]; imposition[0.041]; refusal[0.041]; Department[0.041]; said[0.041]; Libyan[0.041]; Libyan[0.041]; bombing[0.041]; group[0.040]; Man[0.040]; sanctions[0.040]; receive[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Farrakhan\u001b[39m ==> ENTITY: \u001b[32mLouis Farrakhan\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFarrakhan[0.048]; Farrakhan[0.048]; Farrakhan[0.048]; Nation[0.043]; denied[0.043]; black[0.042]; Washington[0.042]; leader[0.042]; State[0.041]; group[0.041]; noted[0.041]; noted[0.041]; talked[0.041]; Islam[0.041]; rally[0.041]; said[0.040]; said[0.040]; sanctions[0.040]; met[0.040]; list[0.040]; Department[0.039]; Department[0.039]; Department[0.039]; pledged[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Libya ==> ENTITY: \u001b[32mLibya\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.190:0.190[0]; log p(e|m)= -0.305:-0.305[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLibya[0.046]; Libya[0.046]; Libya[0.046]; Libya[0.046]; Libyan[0.045]; Libyan[0.045]; Tripoli[0.043]; government[0.042]; international[0.042]; Muammar[0.042]; Gaddafi[0.041]; rights[0.041]; terrorism[0.040]; Wednesday[0.040]; January[0.040]; said[0.040]; said[0.040]; October[0.040]; human[0.040]; Friday[0.039]; denied[0.039]; thousands[0.039]; March[0.039]; group[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Libya ==> ENTITY: \u001b[32mLibya\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.160:0.160[0]; log p(e|m)= -0.305:-0.305[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLibya[0.048]; Libya[0.048]; Libya[0.048]; Libyan[0.046]; government[0.044]; Muammar[0.043]; Gaddafi[0.042]; rights[0.042]; Wednesday[0.041]; January[0.041]; said[0.041]; human[0.041]; Friday[0.041]; denied[0.041]; group[0.040]; noted[0.040]; leader[0.040]; leader[0.040]; passport[0.040]; pledged[0.039]; Department[0.039]; Department[0.039]; Department[0.039]; spokesman[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pan Am flight 103\u001b[39m ==> ENTITY: \u001b[32mPan Am Flight 103\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLockerbie[0.049]; bombing[0.045]; Libya[0.044]; Libya[0.044]; Libyan[0.044]; Libyan[0.044]; Scotland[0.044]; international[0.043]; Tripoli[0.042]; October[0.042]; denied[0.042]; terrorism[0.041]; January[0.041]; Wednesday[0.041]; March[0.041]; suspects[0.040]; said[0.040]; group[0.040]; Gaddafi[0.040]; Muammar[0.040]; list[0.039]; sanctions[0.039]; Department[0.038]; Department[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Muammar Gaddafi\u001b[39m ==> ENTITY: \u001b[32mMuammar Gaddafi\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLockerbie[0.047]; Libyan[0.044]; Libyan[0.044]; Libya[0.044]; Libya[0.044]; Libya[0.044]; Libya[0.044]; Libya[0.044]; flight[0.043]; Tripoli[0.043]; bombing[0.042]; Scotland[0.041]; government[0.041]; international[0.041]; leader[0.040]; leader[0.040]; case[0.040]; travel[0.040]; October[0.040]; denied[0.040]; passport[0.039]; terrorism[0.039]; January[0.039]; group[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Libya ==> ENTITY: \u001b[32mLibya\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.183:0.183[0]; log p(e|m)= -0.305:-0.305[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLibya[0.047]; Libya[0.047]; Libya[0.047]; Libyan[0.045]; Libyan[0.045]; Tripoli[0.044]; government[0.043]; international[0.043]; Muammar[0.042]; Gaddafi[0.041]; terrorism[0.041]; Wednesday[0.041]; January[0.041]; said[0.040]; October[0.040]; sanctions[0.040]; denied[0.040]; thousands[0.039]; March[0.039]; group[0.039]; noted[0.039]; noted[0.039]; leader[0.039]; peaceful[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Farrakhan\u001b[39m ==> ENTITY: \u001b[32mLouis Farrakhan\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFarrakhan[0.049]; Farrakhan[0.049]; Farrakhan[0.049]; Farrakhan[0.049]; Nation[0.044]; denied[0.044]; black[0.043]; black[0.043]; Washington[0.043]; leader[0.043]; leader[0.043]; State[0.043]; group[0.043]; noted[0.042]; noted[0.042]; talked[0.042]; Islam[0.042]; rally[0.042]; Louis[0.041]; said[0.041]; said[0.041]; met[0.041]; list[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.088:0.088[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWashington[0.044]; Washington[0.044]; billion[0.043]; case[0.043]; case[0.043]; October[0.042]; January[0.042]; award[0.042]; award[0.042]; Department[0.041]; Department[0.041]; Department[0.041]; Department[0.041]; awarded[0.041]; noted[0.041]; receive[0.041]; pledged[0.041]; prize[0.041]; Nation[0.041]; Wednesday[0.041]; March[0.041]; travel[0.041]; speculate[0.041]; speculate[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Libya ==> ENTITY: \u001b[32mLibya\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.159:0.159[0]; log p(e|m)= -0.305:-0.305[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLibya[0.048]; Libyan[0.046]; Libyan[0.046]; Tripoli[0.045]; international[0.044]; Muammar[0.043]; Gaddafi[0.042]; terrorism[0.042]; January[0.042]; said[0.041]; October[0.041]; sanctions[0.041]; thousands[0.040]; March[0.040]; group[0.040]; noted[0.040]; leader[0.040]; peaceful[0.040]; pledged[0.040]; men[0.040]; states[0.040]; imposition[0.039]; brought[0.039]; Department[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.050:0.050[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWashington[0.048]; case[0.046]; case[0.046]; awarded[0.046]; award[0.045]; Department[0.044]; Department[0.044]; noted[0.044]; travel[0.044]; speculate[0.043]; speculate[0.043]; Friday[0.043]; refused[0.043]; black[0.043]; accepting[0.042]; prize[0.042]; going[0.042]; said[0.042]; Davies[0.042]; Davies[0.042]; Libya[0.042]; Libya[0.042]; Libya[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.076:0.076[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDepartment[0.045]; Department[0.045]; said[0.044]; State[0.044]; State[0.044]; government[0.044]; travel[0.043]; Friday[0.042]; award[0.042]; Louis[0.042]; tight[0.042]; awarded[0.041]; view[0.041]; noted[0.041]; accepting[0.041]; elaborate[0.040]; talked[0.040]; refused[0.040]; leader[0.040]; going[0.040]; rights[0.040]; case[0.039]; case[0.039]; restriction[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Million Man March\u001b[39m ==> ENTITY: \u001b[32mMillion Man March\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrally[0.044]; Farrakhan[0.044]; Farrakhan[0.044]; Farrakhan[0.044]; Farrakhan[0.044]; pledged[0.043]; group[0.043]; denied[0.043]; Nation[0.042]; black[0.042]; men[0.042]; Wednesday[0.041]; State[0.041]; said[0.041]; said[0.041]; government[0.041]; Washington[0.041]; talked[0.041]; brought[0.041]; noted[0.040]; noted[0.040]; terrorism[0.040]; leader[0.040]; sanctions[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mLockerbie\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPan Am Flight 103\u001b[39m <---> \u001b[32mLockerbie\u001b[39m\t\nSCORES: global= 0.281:0.279[\u001b[31m0.002\u001b[39m]; local()= 0.158:0.108[\u001b[31m0.051\u001b[39m]; log p(e|m)= -1.070:0.000[\u001b[32m1.070\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nflight[0.045]; Scotland[0.044]; bombing[0.044]; Libya[0.044]; Libya[0.044]; Libyan[0.044]; Libyan[0.044]; international[0.043]; Tripoli[0.042]; October[0.042]; denied[0.042]; terrorism[0.041]; January[0.041]; said[0.041]; Wednesday[0.040]; March[0.040]; suspects[0.040]; Washington[0.040]; black[0.040]; group[0.040]; sponsor[0.040]; Gaddafi[0.039]; Muammar[0.039]; men[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Farrakhan\u001b[39m ==> ENTITY: \u001b[32mLouis Farrakhan\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFarrakhan[0.048]; Farrakhan[0.048]; Farrakhan[0.048]; Nation[0.043]; denied[0.043]; black[0.042]; Washington[0.042]; leader[0.042]; State[0.041]; group[0.041]; noted[0.041]; noted[0.041]; talked[0.041]; Islam[0.041]; rally[0.041]; said[0.040]; said[0.040]; sanctions[0.040]; met[0.040]; list[0.040]; Department[0.039]; Department[0.039]; Department[0.039]; pledged[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Libya ==> ENTITY: \u001b[32mLibya\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.176:0.176[0]; log p(e|m)= -0.305:-0.305[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLibya[0.047]; Libya[0.047]; Libya[0.047]; Libya[0.047]; Libyan[0.045]; government[0.043]; international[0.043]; Muammar[0.042]; Gaddafi[0.042]; rights[0.041]; Wednesday[0.041]; January[0.041]; said[0.040]; said[0.040]; October[0.040]; human[0.040]; Friday[0.040]; denied[0.040]; thousands[0.039]; March[0.039]; group[0.039]; noted[0.039]; leader[0.039]; leader[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.075:0.075[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScotland[0.046]; Washington[0.044]; international[0.044]; Pan[0.042]; men[0.042]; October[0.042]; January[0.042]; award[0.042]; states[0.042]; Department[0.042]; Department[0.042]; noted[0.042]; noted[0.042]; Wednesday[0.041]; group[0.041]; list[0.041]; Nation[0.041]; travel[0.041]; March[0.041]; imposition[0.040]; denied[0.040]; Davies[0.040]; Million[0.040]; met[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = State Department\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of State\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \naward[0.048]; Department[0.047]; Department[0.047]; Washington[0.047]; awarded[0.046]; State[0.042]; Libya[0.041]; Libya[0.041]; Libya[0.041]; Libya[0.041]; view[0.041]; noted[0.041]; government[0.041]; Wednesday[0.041]; passport[0.040]; said[0.040]; Treasury[0.040]; travel[0.039]; accepting[0.039]; spokesman[0.039]; Friday[0.039]; prize[0.039]; denied[0.039]; rights[0.039]; \t\n\r [=========================>....................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s328ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2730/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1042testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tokyo ==> ENTITY: \u001b[32mTokyo\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.020:0.020[0]; log p(e|m)= -0.093:-0.093[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nyen[0.047]; September[0.047]; Corp[0.045]; million[0.044]; million[0.044]; million[0.044]; parent[0.043]; manufacturer[0.042]; leading[0.042]; Previous[0.042]; months[0.042]; semiconductor[0.042]; Sales[0.041]; unless[0.041]; Yamada[0.040]; Yamada[0.040]; Forecast[0.040]; Forecast[0.040]; Latest[0.040]; Current[0.040]; billions[0.040]; Parent[0.039]; specified[0.039]; Net[0.039]; \t\n\r [=========================>....................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s724ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2731/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1137testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.138:0.138[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.050]; China[0.050]; Beijing[0.047]; Tianjin[0.044]; Tianjin[0.044]; Tianjin[0.044]; northern[0.043]; city[0.043]; annual[0.042]; Daily[0.040]; facilities[0.040]; companies[0.040]; recent[0.040]; United[0.040]; States[0.040]; production[0.039]; Saturday[0.039]; boasts[0.039]; plant[0.039]; plant[0.039]; said[0.039]; said[0.039]; attracting[0.039]; multinational[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tianjin\u001b[39m ==> ENTITY: \u001b[32mTianjin\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.045]; China[0.045]; China[0.045]; China[0.045]; Beijing[0.045]; Tianjin[0.045]; Tianjin[0.045]; Tianjin[0.045]; facilities[0.044]; infrastructure[0.041]; plans[0.041]; build[0.041]; build[0.041]; city[0.041]; blender[0.040]; multinational[0.040]; tonnes[0.040]; Daily[0.039]; Corp[0.039]; recent[0.039]; northern[0.039]; attracting[0.039]; Saturday[0.039]; billion[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.142:0.142[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.050]; China[0.050]; Tianjin[0.044]; Tianjin[0.044]; Tianjin[0.044]; Tianjin[0.044]; northern[0.043]; city[0.043]; annual[0.042]; Daily[0.041]; based[0.040]; facilities[0.040]; rising[0.040]; companies[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; recent[0.040]; Corp[0.040]; attracted[0.039]; firms[0.039]; plans[0.039]; production[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tianjin\u001b[39m ==> ENTITY: \u001b[32mTianjin\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.045]; China[0.045]; China[0.045]; China[0.045]; Beijing[0.044]; facilities[0.044]; Tianjin[0.044]; Tianjin[0.044]; Tianjin[0.044]; infrastructure[0.041]; companies[0.041]; annual[0.041]; build[0.041]; build[0.041]; Daily[0.040]; city[0.040]; attracted[0.040]; attracting[0.040]; blender[0.040]; multinational[0.040]; plans[0.039]; tonnes[0.039]; recent[0.039]; firms[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.154:0.154[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.049]; China[0.049]; China[0.049]; Beijing[0.046]; Tianjin[0.043]; Tianjin[0.043]; Tianjin[0.043]; Tianjin[0.043]; northern[0.043]; city[0.042]; annual[0.041]; Daily[0.040]; facilities[0.039]; rising[0.039]; companies[0.039]; recent[0.039]; United[0.039]; States[0.039]; attracted[0.039]; production[0.039]; Saturday[0.039]; boasts[0.039]; plant[0.038]; plant[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mobil\u001b[39m ==> ENTITY: \u001b[32mMobil\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nShell[0.048]; Petroleum[0.044]; oil[0.044]; oil[0.044]; companies[0.043]; lubricants[0.043]; Caltex[0.042]; Caltex[0.042]; multinational[0.042]; based[0.042]; facilities[0.042]; Chemical[0.041]; Multinational[0.041]; production[0.041]; plant[0.041]; plant[0.041]; products[0.040]; plans[0.040]; Corp[0.040]; billion[0.040]; demand[0.040]; firms[0.040]; annual[0.040]; recent[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dow Chemical\u001b[39m ==> ENTITY: \u001b[32mDow Chemical Company\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.126:0.126[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncompanies[0.047]; Chemical[0.045]; Corp[0.044]; Dow[0.043]; multinational[0.042]; China[0.042]; China[0.042]; China[0.042]; plant[0.042]; plant[0.042]; facilities[0.041]; build[0.041]; build[0.041]; plans[0.041]; invest[0.041]; invest[0.041]; recent[0.041]; lubricants[0.040]; ethylene[0.040]; ethylene[0.040]; boasts[0.040]; Beijing[0.040]; billion[0.040]; Petroleum[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.050:0.050[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBeijing[0.049]; China[0.046]; China[0.046]; China[0.046]; details[0.045]; Saturday[0.045]; recent[0.044]; northern[0.043]; gave[0.043]; Tianjin[0.043]; Tianjin[0.043]; Tianjin[0.043]; said[0.043]; said[0.043]; said[0.043]; facilities[0.042]; invest[0.042]; invest[0.042]; boasts[0.042]; Daily[0.042]; plans[0.042]; annual[0.042]; range[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mShell\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRoyal Dutch Shell\u001b[39m <---> \u001b[32mShell Oil Company\u001b[39m\t\nSCORES: global= 0.264:0.243[\u001b[31m0.021\u001b[39m]; local()= 0.174:0.142[\u001b[31m0.032\u001b[39m]; log p(e|m)= -0.654:-2.364[\u001b[31m1.711\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncompanies[0.044]; Petroleum[0.044]; facilities[0.043]; plans[0.043]; Mobil[0.042]; oil[0.042]; oil[0.042]; multinational[0.042]; plant[0.042]; plant[0.042]; firms[0.042]; build[0.041]; build[0.041]; range[0.041]; Caltex[0.041]; Caltex[0.041]; lubricants[0.041]; Corp[0.041]; products[0.041]; infrastructure[0.041]; Multinational[0.040]; gave[0.040]; gave[0.040]; production[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = China Daily\u001b[39m ==> ENTITY: \u001b[32mChina Daily\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnewspaper[0.048]; newspaper[0.048]; newspaper[0.048]; China[0.048]; China[0.048]; China[0.048]; Beijing[0.047]; plans[0.043]; Tianjin[0.042]; Tianjin[0.042]; Tianjin[0.042]; Tianjin[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; attracting[0.041]; companies[0.041]; invest[0.041]; invest[0.041]; ethylene[0.040]; ethylene[0.040]; firms[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Beijing ==> ENTITY: \u001b[32mBeijing\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.042:-0.042[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.049]; China[0.049]; China[0.049]; Tianjin[0.045]; Tianjin[0.045]; Tianjin[0.045]; northern[0.043]; facilities[0.042]; city[0.041]; plans[0.041]; build[0.041]; build[0.041]; Saturday[0.041]; Daily[0.041]; gave[0.039]; boasts[0.039]; infrastructure[0.039]; recent[0.039]; venture[0.039]; venture[0.039]; billion[0.038]; said[0.038]; said[0.038]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Caltex\u001b[39m ==> ENTITY: \u001b[32mCaltex\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCaltex[0.048]; Shell[0.048]; Mobil[0.047]; companies[0.046]; Petroleum[0.045]; multinational[0.045]; oil[0.044]; oil[0.044]; Multinational[0.043]; Corp[0.043]; lubricants[0.043]; China[0.043]; China[0.043]; China[0.043]; firms[0.043]; facilities[0.042]; plans[0.042]; billion[0.042]; venture[0.041]; lube[0.041]; States[0.041]; ethylene[0.041]; production[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tianjin\u001b[39m ==> ENTITY: \u001b[32mTianjin\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.045]; China[0.045]; China[0.045]; China[0.045]; Beijing[0.044]; facilities[0.044]; Tianjin[0.044]; Tianjin[0.044]; Tianjin[0.044]; infrastructure[0.041]; companies[0.041]; annual[0.041]; build[0.041]; build[0.041]; Daily[0.040]; city[0.040]; attracted[0.040]; attracting[0.040]; blender[0.040]; multinational[0.040]; plans[0.039]; tonnes[0.039]; recent[0.039]; firms[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tianjin\u001b[39m ==> ENTITY: \u001b[32mTianjin\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.046]; China[0.046]; China[0.046]; facilities[0.045]; Tianjin[0.044]; Tianjin[0.044]; Tianjin[0.044]; infrastructure[0.041]; companies[0.041]; annual[0.041]; build[0.041]; build[0.041]; Daily[0.041]; city[0.041]; attracted[0.041]; attracting[0.041]; blender[0.040]; multinational[0.040]; plans[0.040]; tonnes[0.040]; recent[0.040]; firms[0.039]; Corp[0.039]; northern[0.039]; \t\n\r [=========================>....................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s678ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2745/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 976testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bucharest ==> ENTITY: \u001b[32mBucharest\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.018:-0.018[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBucharest[0.048]; Romania[0.046]; Romanian[0.044]; town[0.043]; month[0.042]; month[0.042]; rate[0.042]; rising[0.042]; exchange[0.042]; doubled[0.042]; Bistrita[0.041]; percent[0.041]; company[0.041]; company[0.041]; showed[0.041]; listed[0.041]; equipment[0.041]; state[0.041]; central[0.040]; complained[0.040]; high[0.040]; stock[0.040]; Transylvanian[0.040]; leu[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bucharest ==> ENTITY: \u001b[32mBucharest\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.120:0.120[0]; log p(e|m)= -0.018:-0.018[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBucharest[0.048]; Romania[0.045]; June[0.044]; Romanian[0.044]; town[0.042]; January[0.042]; month[0.042]; month[0.042]; rate[0.041]; hit[0.041]; rising[0.041]; rising[0.041]; months[0.041]; exchange[0.041]; doubled[0.041]; costs[0.041]; Bistrita[0.041]; percent[0.040]; percent[0.040]; percent[0.040]; company[0.040]; company[0.040]; company[0.040]; showed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Transylvanian\u001b[39m ==> ENTITY: \u001b[32mTransylvania\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRomanian[0.046]; central[0.045]; Bistrita[0.043]; town[0.043]; complained[0.043]; Bucharest[0.043]; Bucharest[0.043]; month[0.042]; month[0.042]; equivalent[0.041]; said[0.041]; said[0.041]; said[0.041]; estimated[0.041]; water[0.041]; showed[0.041]; lei[0.041]; lei[0.041]; state[0.041]; concentrated[0.041]; exported[0.040]; listed[0.040]; rising[0.040]; rising[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bistrita\u001b[39m ==> ENTITY: \u001b[32mBistrița\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.121:0.121[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntown[0.047]; listed[0.043]; Transylvanian[0.043]; leu[0.043]; central[0.043]; hit[0.043]; water[0.043]; Romanian[0.043]; exported[0.042]; state[0.042]; rate[0.041]; Bucharest[0.041]; Bucharest[0.041]; tonnes[0.041]; tonnes[0.041]; volume[0.041]; lei[0.041]; lei[0.041]; estimated[0.041]; said[0.040]; said[0.040]; said[0.040]; concentrated[0.040]; equivalent[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Romania ==> ENTITY: \u001b[32mRomania\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.100:0.100[0]; log p(e|m)= -0.212:-0.212[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBucharest[0.047]; Bucharest[0.047]; Romanian[0.047]; month[0.045]; month[0.045]; lei[0.042]; lei[0.042]; state[0.042]; doubled[0.042]; exchange[0.042]; said[0.041]; period[0.041]; company[0.040]; company[0.040]; equivalent[0.040]; complained[0.040]; rate[0.040]; showed[0.040]; rose[0.040]; percent[0.040]; listed[0.040]; high[0.040]; based[0.039]; cut[0.039]; \t\n\r [=========================>....................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s728ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2750/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1044testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Thai ==> ENTITY: \u001b[32mThailand\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.139:0.139[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThai[0.052]; Singapore[0.047]; Singapore[0.047]; Singapore[0.047]; Singapore[0.047]; people[0.044]; Friday[0.042]; Central[0.042]; said[0.040]; said[0.040]; half[0.040]; nearly[0.040]; death[0.040]; heroin[0.040]; trafficking[0.040]; trafficking[0.040]; farmer[0.040]; grams[0.039]; grams[0.039]; grams[0.039]; Bureau[0.039]; drug[0.039]; drug[0.039]; trafficker[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.051]; Singapore[0.051]; Singapore[0.051]; Singapore[0.051]; Thai[0.043]; Thai[0.043]; mandatory[0.042]; sentence[0.042]; crimes[0.041]; nearly[0.040]; Changi[0.040]; Friday[0.039]; Prison[0.039]; Central[0.039]; marijuana[0.039]; drug[0.039]; drug[0.039]; cannabis[0.039]; cannabis[0.039]; trafficking[0.039]; trafficking[0.039]; said[0.039]; said[0.039]; half[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.051]; Singapore[0.051]; Singapore[0.051]; Singapore[0.051]; Thai[0.042]; Thai[0.042]; Changi[0.042]; mandatory[0.042]; sentence[0.041]; various[0.041]; crimes[0.040]; related[0.040]; nearly[0.040]; half[0.039]; half[0.039]; charges[0.039]; Friday[0.039]; kg[0.039]; Prison[0.039]; Central[0.039]; marijuana[0.038]; drug[0.038]; drug[0.038]; drug[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Central Narcotics Bureau\u001b[39m ==> ENTITY: \u001b[32mCentral Narcotics Bureau\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndrug[0.046]; drug[0.046]; drug[0.046]; trafficking[0.043]; trafficking[0.043]; heroin[0.043]; cannabis[0.042]; cannabis[0.042]; Singapore[0.041]; Singapore[0.041]; Singapore[0.041]; Singapore[0.041]; Singapore[0.041]; oz[0.041]; Changi[0.041]; charges[0.041]; mandatory[0.041]; related[0.040]; morphine[0.040]; crimes[0.040]; Prison[0.040]; kg[0.040]; marijuana[0.039]; various[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.114:0.114[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.052]; Singapore[0.052]; Thai[0.044]; mandatory[0.043]; sentence[0.043]; charges[0.042]; crimes[0.042]; related[0.041]; Changi[0.041]; nearly[0.041]; Central[0.041]; Friday[0.040]; Prison[0.040]; various[0.040]; marijuana[0.040]; drug[0.040]; drug[0.040]; cannabis[0.040]; cannabis[0.040]; trafficking[0.040]; trafficking[0.040]; said[0.040]; said[0.040]; Bureau[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.052]; Singapore[0.052]; Singapore[0.052]; Thai[0.043]; Thai[0.043]; mandatory[0.043]; sentence[0.042]; crimes[0.041]; nearly[0.041]; Changi[0.041]; Friday[0.040]; Prison[0.040]; Central[0.040]; marijuana[0.039]; drug[0.039]; drug[0.039]; cannabis[0.039]; cannabis[0.039]; trafficking[0.039]; trafficking[0.039]; said[0.039]; said[0.039]; half[0.039]; Bureau[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Changi Prison\u001b[39m ==> ENTITY: \u001b[32mChangi Prison\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.045]; Singapore[0.045]; Singapore[0.045]; Singapore[0.045]; Singapore[0.045]; Friday[0.043]; charges[0.042]; hanged[0.042]; hanged[0.042]; oz[0.042]; sentence[0.042]; crimes[0.041]; Thai[0.041]; Thai[0.041]; nearly[0.041]; morphine[0.041]; Central[0.040]; related[0.040]; kg[0.040]; guilty[0.040]; people[0.040]; arrested[0.040]; mandatory[0.039]; death[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Thai ==> ENTITY: \u001b[32mThailand\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.150:0.150[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThai[0.052]; Singapore[0.046]; Singapore[0.046]; Singapore[0.046]; Singapore[0.046]; Singapore[0.046]; people[0.043]; Friday[0.041]; Central[0.041]; said[0.040]; said[0.040]; half[0.040]; half[0.040]; nearly[0.040]; death[0.040]; various[0.040]; heroin[0.040]; trafficking[0.039]; trafficking[0.039]; farmer[0.039]; grams[0.039]; grams[0.039]; grams[0.039]; Bureau[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Singapore ==> ENTITY: \u001b[32mSingapore\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.192:-0.192[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSingapore[0.052]; Singapore[0.052]; Singapore[0.052]; Thai[0.044]; Thai[0.044]; mandatory[0.043]; sentence[0.043]; nearly[0.041]; Changi[0.041]; Friday[0.040]; Prison[0.040]; Central[0.040]; marijuana[0.039]; drug[0.039]; drug[0.039]; cannabis[0.039]; cannabis[0.039]; trafficking[0.039]; trafficking[0.039]; said[0.039]; said[0.039]; half[0.039]; Bureau[0.039]; hanged[0.039]; \t\n\r [=========================>....................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s710ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2759/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 960testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew York\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew York Yankees\u001b[39m <---> \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.238:0.234[\u001b[31m0.005\u001b[39m]; local()= 0.195:0.107[\u001b[31m0.088\u001b[39m]; log p(e|m)= -5.116:-1.790[\u001b[32m3.326\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.045]; games[0.045]; Baseball[0.044]; Baseball[0.044]; League[0.043]; League[0.043]; League[0.043]; League[0.043]; City[0.042]; Baltimore[0.042]; Pittsburgh[0.042]; Cincinnati[0.042]; Houston[0.041]; Thursday[0.041]; Thursday[0.041]; played[0.041]; Detroit[0.041]; York[0.040]; York[0.040]; American[0.040]; Atlanta[0.040]; Montreal[0.039]; National[0.039]; Chicago[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMilwaukee\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAtlanta Braves\u001b[39m <---> \u001b[32mMilwaukee Brewers\u001b[39m\t\nSCORES: global= 0.252:0.250[\u001b[31m0.002\u001b[39m]; local()= 0.222:0.194[\u001b[31m0.028\u001b[39m]; log p(e|m)= -3.912:-3.411[\u001b[32m0.501\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; Baseball[0.044]; Baseball[0.044]; Cincinnati[0.044]; team[0.043]; Pittsburgh[0.043]; League[0.043]; League[0.043]; League[0.043]; League[0.043]; Baltimore[0.042]; Houston[0.042]; Detroit[0.042]; Atlanta[0.041]; played[0.041]; Thursday[0.041]; Thursday[0.041]; York[0.040]; York[0.040]; York[0.040]; Chicago[0.040]; City[0.040]; Montreal[0.039]; Minnesota[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Los Angeles ==> ENTITY: \u001b[32mLos Angeles Dodgers\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.189:0.189[0]; log p(e|m)= -4.711:-4.711[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; team[0.044]; California[0.043]; League[0.043]; League[0.043]; League[0.043]; League[0.043]; Montreal[0.042]; Baseball[0.042]; Baseball[0.042]; Cincinnati[0.042]; Houston[0.042]; Caps[0.042]; Pittsburgh[0.041]; Baltimore[0.041]; Detroit[0.041]; City[0.040]; played[0.040]; Colorado[0.040]; Thursday[0.040]; Thursday[0.040]; Atlanta[0.040]; Florida[0.040]; Chicago[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Houston ==> ENTITY: \u001b[32mHouston Astros\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.179:0.179[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.046]; games[0.045]; League[0.044]; League[0.044]; League[0.044]; League[0.044]; Baltimore[0.043]; Cincinnati[0.043]; Baseball[0.042]; Baseball[0.042]; Pittsburgh[0.042]; Thursday[0.041]; Thursday[0.041]; City[0.041]; played[0.041]; home[0.040]; Atlanta[0.040]; National[0.040]; American[0.040]; Detroit[0.040]; Colorado[0.040]; Florida[0.039]; Chicago[0.039]; California[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSan Diego\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSan Diego Padres\u001b[39m <---> \u001b[32mSan Diego\u001b[39m\t\nSCORES: global= 0.258:0.244[\u001b[31m0.014\u001b[39m]; local()= 0.209:0.126[\u001b[31m0.084\u001b[39m]; log p(e|m)= -3.912:-0.188[\u001b[32m3.724\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; League[0.044]; League[0.044]; League[0.044]; League[0.044]; team[0.043]; Houston[0.043]; California[0.043]; Cincinnati[0.043]; Baseball[0.042]; Baseball[0.042]; Baltimore[0.041]; Thursday[0.041]; Thursday[0.041]; Pittsburgh[0.041]; played[0.040]; Detroit[0.040]; Florida[0.040]; National[0.040]; City[0.040]; Atlanta[0.040]; Colorado[0.039]; American[0.039]; Seattle[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kansas City ==> ENTITY: \u001b[32mKansas City Royals\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.207:0.207[0]; log p(e|m)= -3.576:-3.576[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.049]; team[0.049]; Baseball[0.047]; Baseball[0.047]; Houston[0.047]; League[0.047]; League[0.047]; League[0.047]; League[0.047]; Baltimore[0.046]; Cincinnati[0.045]; Pittsburgh[0.045]; Detroit[0.044]; played[0.044]; Thursday[0.044]; Thursday[0.044]; Montreal[0.044]; Louis[0.044]; Milwaukee[0.043]; National[0.043]; home[0.043]; Atlanta[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colorado ==> ENTITY: \u001b[32mColorado Rockies\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.175:0.175[0]; log p(e|m)= -4.510:-4.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; League[0.044]; League[0.044]; League[0.044]; League[0.044]; team[0.044]; Montreal[0.044]; Caps[0.042]; Baseball[0.042]; Baseball[0.042]; Cincinnati[0.042]; played[0.041]; American[0.041]; Pittsburgh[0.041]; Houston[0.041]; Florida[0.041]; Atlanta[0.040]; Kansas[0.040]; Louis[0.040]; Detroit[0.040]; National[0.040]; California[0.039]; Baltimore[0.039]; Chicago[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew York\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew York Yankees\u001b[39m <---> \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.238:0.234[\u001b[31m0.005\u001b[39m]; local()= 0.195:0.107[\u001b[31m0.088\u001b[39m]; log p(e|m)= -5.116:-1.790[\u001b[32m3.326\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.045]; games[0.045]; Baseball[0.044]; Baseball[0.044]; League[0.043]; League[0.043]; League[0.043]; League[0.043]; City[0.042]; Baltimore[0.042]; Pittsburgh[0.042]; Cincinnati[0.042]; Houston[0.041]; Thursday[0.041]; Thursday[0.041]; played[0.041]; Detroit[0.041]; York[0.040]; York[0.040]; American[0.040]; Atlanta[0.040]; Montreal[0.039]; National[0.039]; Chicago[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Florida ==> ENTITY: \u001b[32mMiami Marlins\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.200:0.200[0]; log p(e|m)= -4.828:-4.828[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.048]; team[0.046]; Caps[0.045]; Cincinnati[0.045]; Montreal[0.044]; Baseball[0.044]; Baseball[0.044]; League[0.044]; League[0.044]; League[0.044]; League[0.044]; Pittsburgh[0.044]; Houston[0.043]; played[0.043]; California[0.043]; Milwaukee[0.043]; Baltimore[0.042]; Atlanta[0.042]; Colorado[0.042]; Detroit[0.042]; Chicago[0.042]; Kansas[0.041]; American[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York Yankees\u001b[39m\t\nSCORES: global= 0.238:0.238[0]; local()= 0.195:0.195[0]; log p(e|m)= -5.116:-5.116[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.045]; games[0.045]; Baseball[0.044]; Baseball[0.044]; League[0.043]; League[0.043]; League[0.043]; League[0.043]; City[0.042]; Baltimore[0.042]; Pittsburgh[0.042]; Cincinnati[0.042]; Houston[0.041]; Thursday[0.041]; Thursday[0.041]; played[0.041]; Detroit[0.041]; York[0.040]; York[0.040]; American[0.040]; Atlanta[0.040]; Montreal[0.039]; National[0.039]; Chicago[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Major League ==> ENTITY: \u001b[32mMajor League Baseball\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.196:0.196[0]; log p(e|m)= -0.478:-0.478[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; League[0.046]; League[0.046]; team[0.045]; games[0.044]; Baseball[0.044]; Baseball[0.044]; Cincinnati[0.043]; Houston[0.042]; Major[0.042]; Pittsburgh[0.042]; played[0.042]; Baltimore[0.041]; National[0.040]; Detroit[0.040]; American[0.040]; York[0.040]; York[0.040]; York[0.040]; Atlanta[0.039]; Florida[0.039]; Colorado[0.039]; City[0.039]; Chicago[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = American League\u001b[39m ==> ENTITY: \u001b[32mAmerican League\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.045]; League[0.045]; League[0.045]; team[0.044]; Cincinnati[0.044]; Baseball[0.043]; Baseball[0.043]; games[0.043]; Pittsburgh[0.042]; Detroit[0.042]; Baltimore[0.042]; Houston[0.041]; Chicago[0.041]; Thursday[0.040]; Thursday[0.040]; played[0.040]; National[0.040]; California[0.040]; Atlanta[0.040]; Colorado[0.040]; Minnesota[0.040]; Milwaukee[0.040]; Montreal[0.039]; home[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Seattle ==> ENTITY: \u001b[32mSeattle Mariners\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.193:0.193[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.047]; games[0.045]; Baseball[0.044]; Baseball[0.044]; League[0.043]; League[0.043]; League[0.043]; League[0.043]; Houston[0.043]; Cincinnati[0.042]; Baltimore[0.042]; Pittsburgh[0.042]; played[0.042]; Thursday[0.040]; Thursday[0.040]; Detroit[0.040]; Atlanta[0.040]; City[0.040]; Milwaukee[0.040]; National[0.040]; Florida[0.040]; California[0.039]; Colorado[0.039]; American[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Minnesota ==> ENTITY: \u001b[32mMinnesota Twins\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.205:0.205[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; team[0.045]; Baseball[0.044]; Baseball[0.044]; League[0.043]; League[0.043]; League[0.043]; League[0.043]; Montreal[0.042]; played[0.042]; Detroit[0.042]; Caps[0.042]; Baltimore[0.042]; Cincinnati[0.042]; Houston[0.041]; Thursday[0.041]; Thursday[0.041]; Pittsburgh[0.041]; Colorado[0.040]; Atlanta[0.039]; Kansas[0.039]; Louis[0.039]; Florida[0.039]; Milwaukee[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pittsburgh ==> ENTITY: \u001b[32mPittsburgh Pirates\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.194:0.194[0]; log p(e|m)= -4.017:-4.017[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.049]; Cincinnati[0.047]; Baseball[0.047]; Baseball[0.047]; team[0.047]; League[0.047]; League[0.047]; League[0.047]; League[0.047]; Montreal[0.046]; Baltimore[0.045]; played[0.045]; Florida[0.045]; Houston[0.045]; Detroit[0.045]; Caps[0.044]; Louis[0.044]; Atlanta[0.044]; Colorado[0.043]; National[0.043]; City[0.043]; American[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cincinnati ==> ENTITY: \u001b[32mCincinnati Reds\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.193:0.193[0]; log p(e|m)= -3.772:-3.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; team[0.046]; Baseball[0.043]; Baseball[0.043]; League[0.043]; League[0.043]; League[0.043]; League[0.043]; Houston[0.043]; Pittsburgh[0.042]; Baltimore[0.042]; Thursday[0.041]; Thursday[0.041]; played[0.041]; Louis[0.041]; Detroit[0.041]; American[0.040]; Milwaukee[0.040]; National[0.040]; Minnesota[0.040]; Atlanta[0.040]; Montreal[0.040]; Kansas[0.039]; City[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Major League ==> ENTITY: \u001b[32mMajor League Baseball\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.196:0.196[0]; log p(e|m)= -0.478:-0.478[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; League[0.046]; League[0.046]; team[0.045]; games[0.044]; Baseball[0.044]; Baseball[0.044]; Cincinnati[0.043]; Houston[0.042]; Major[0.042]; Pittsburgh[0.042]; played[0.042]; Baltimore[0.041]; National[0.040]; Detroit[0.040]; American[0.040]; York[0.040]; York[0.040]; York[0.040]; Atlanta[0.039]; Florida[0.039]; Colorado[0.039]; City[0.039]; Chicago[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Baltimore ==> ENTITY: \u001b[32mBaltimore Orioles\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.201:0.201[0]; log p(e|m)= -4.075:-4.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; Houston[0.044]; Cincinnati[0.044]; Pittsburgh[0.044]; League[0.043]; League[0.043]; League[0.043]; League[0.043]; team[0.043]; Baseball[0.043]; Baseball[0.043]; played[0.042]; National[0.042]; Detroit[0.042]; Florida[0.041]; Atlanta[0.040]; York[0.040]; York[0.040]; York[0.040]; Thursday[0.040]; Thursday[0.040]; Chicago[0.039]; Montreal[0.039]; City[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMontreal\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMontreal Expos\u001b[39m <---> \u001b[32mMontreal\u001b[39m\t\nSCORES: global= 0.248:0.237[\u001b[31m0.011\u001b[39m]; local()= 0.201:0.065[\u001b[31m0.136\u001b[39m]; log p(e|m)= -4.510:-0.110[\u001b[32m4.400\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; team[0.044]; Baltimore[0.043]; Baseball[0.043]; Baseball[0.043]; Caps[0.043]; League[0.042]; League[0.042]; League[0.042]; League[0.042]; Houston[0.042]; Cincinnati[0.042]; Louis[0.042]; Detroit[0.041]; Pittsburgh[0.041]; Thursday[0.041]; Thursday[0.041]; Atlanta[0.041]; played[0.040]; York[0.040]; York[0.040]; York[0.040]; Florida[0.039]; Chicago[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Detroit ==> ENTITY: \u001b[32mDetroit Tigers\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.202:0.202[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; team[0.045]; Baseball[0.044]; Baseball[0.044]; Caps[0.043]; Cincinnati[0.043]; Montreal[0.043]; League[0.042]; League[0.042]; League[0.042]; League[0.042]; Pittsburgh[0.042]; played[0.041]; Houston[0.041]; Thursday[0.040]; Thursday[0.040]; Baltimore[0.040]; Atlanta[0.040]; City[0.040]; Chicago[0.040]; American[0.040]; Colorado[0.039]; Louis[0.039]; National[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlanta ==> ENTITY: \u001b[32mAtlanta Braves\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.214:0.214[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.045]; games[0.045]; Baseball[0.043]; Baseball[0.043]; League[0.043]; League[0.043]; League[0.043]; League[0.043]; Cincinnati[0.043]; Houston[0.043]; Pittsburgh[0.043]; Diego[0.042]; Baltimore[0.042]; Detroit[0.041]; Los[0.041]; played[0.040]; Thursday[0.040]; Thursday[0.040]; Chicago[0.040]; Milwaukee[0.040]; Florida[0.039]; York[0.039]; York[0.039]; York[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = National League ==> ENTITY: \u001b[32mNational League\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.194:0.194[0]; log p(e|m)= -0.075:-0.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; League[0.046]; League[0.046]; games[0.046]; Baseball[0.044]; Baseball[0.044]; Cincinnati[0.042]; team[0.042]; Detroit[0.042]; Baltimore[0.041]; played[0.041]; Houston[0.041]; Pittsburgh[0.041]; York[0.040]; York[0.040]; York[0.040]; Atlanta[0.040]; City[0.040]; Milwaukee[0.040]; Thursday[0.040]; Thursday[0.040]; Seattle[0.039]; Montreal[0.039]; Colorado[0.039]; \t\n\r [=========================>....................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s618ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2783/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1135testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = OIC\u001b[39m ==> ENTITY: \u001b[32mOrganisation of Islamic Cooperation\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntext[0.048]; Organisation[0.046]; Islamic[0.046]; Moslem[0.045]; Conference[0.044]; Indonesia[0.042]; formal[0.042]; Indonesian[0.042]; support[0.042]; support[0.042]; peace[0.042]; friendship[0.041]; Jakarta[0.041]; agreement[0.040]; hails[0.040]; supporting[0.040]; deal[0.040]; Suharto[0.040]; Suharto[0.040]; told[0.040]; talks[0.040]; reporters[0.040]; served[0.040]; Philippine[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = OIC\u001b[39m ==> ENTITY: \u001b[32mOrganisation of Islamic Cooperation\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntext[0.047]; Organisation[0.046]; Islamic[0.045]; Moslem[0.045]; Conference[0.044]; Indonesia[0.042]; formal[0.042]; Indonesian[0.042]; Monday[0.041]; support[0.041]; support[0.041]; peace[0.041]; peace[0.041]; peace[0.041]; friendship[0.041]; Jakarta[0.041]; scheduled[0.040]; said[0.040]; agreement[0.040]; agreement[0.040]; agreement[0.040]; hails[0.040]; supporting[0.040]; deal[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Philippine ==> ENTITY: \u001b[32mPhilippines\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.567:-0.567[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nManila[0.049]; Mindanao[0.047]; history[0.044]; text[0.044]; local[0.044]; southern[0.044]; island[0.043]; Ramos[0.042]; Ramos[0.042]; formal[0.041]; said[0.041]; Monday[0.041]; era[0.041]; supporting[0.040]; letter[0.040]; letter[0.040]; Friday[0.040]; Jakarta[0.040]; peace[0.040]; peace[0.040]; war[0.040]; bring[0.040]; host[0.040]; served[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moslem ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.108:0.108[0]; log p(e|m)= -1.214:-1.214[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIslamic[0.051]; southern[0.045]; history[0.045]; signing[0.044]; war[0.043]; said[0.042]; Indonesia[0.042]; hails[0.042]; peace[0.041]; peace[0.041]; peace[0.041]; lives[0.041]; scheduled[0.041]; local[0.041]; support[0.040]; support[0.040]; agreement[0.040]; agreement[0.040]; agreement[0.040]; Monday[0.040]; strife[0.040]; gratitude[0.040]; gratitude[0.040]; told[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Suharto\u001b[39m ==> ENTITY: \u001b[32mSuharto\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.046]; president[0.044]; president[0.044]; support[0.043]; support[0.043]; Indonesian[0.043]; Suharto[0.043]; Jakarta[0.043]; reporters[0.042]; conflict[0.042]; Monday[0.041]; said[0.041]; Islamic[0.041]; signing[0.041]; served[0.040]; released[0.040]; culminated[0.040]; scheduled[0.040]; hails[0.040]; peace[0.040]; peace[0.040]; peace[0.040]; rebels[0.040]; supporting[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mindanao\u001b[39m ==> ENTITY: \u001b[32mMindanao\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhilippine[0.050]; Manila[0.049]; southern[0.048]; island[0.046]; peace[0.046]; peace[0.046]; Monday[0.044]; rebels[0.044]; agreement[0.044]; agreement[0.044]; agreement[0.044]; Friday[0.043]; war[0.043]; history[0.043]; formal[0.042]; culminated[0.041]; negotiations[0.041]; Saturday[0.041]; support[0.041]; signing[0.041]; said[0.040]; scheduled[0.040]; century[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ramos\u001b[39m ==> ENTITY: \u001b[32mFidel V. Ramos\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhilippine[0.048]; Philippine[0.048]; Mindanao[0.046]; Manila[0.045]; Manila[0.045]; Manila[0.045]; president[0.042]; president[0.042]; Ramos[0.041]; Ramos[0.041]; signing[0.041]; agreement[0.040]; agreement[0.040]; agreement[0.040]; reporters[0.040]; peace[0.040]; peace[0.040]; peace[0.040]; support[0.039]; support[0.039]; released[0.039]; Monday[0.039]; rebels[0.039]; Conference[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesia ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.124:0.124[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesian[0.048]; Jakarta[0.045]; peace[0.044]; Philippine[0.043]; text[0.042]; conflict[0.042]; local[0.042]; host[0.042]; supporting[0.041]; support[0.041]; support[0.041]; Suharto[0.041]; Suharto[0.041]; talks[0.041]; series[0.041]; negotiations[0.041]; Islamic[0.041]; Manila[0.041]; Manila[0.041]; president[0.041]; president[0.041]; agreement[0.040]; Conference[0.040]; rebels[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fidel Ramos\u001b[39m ==> ENTITY: \u001b[32mFidel V. Ramos\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhilippine[0.050]; Manila[0.046]; Manila[0.046]; Manila[0.046]; president[0.043]; president[0.043]; Ramos[0.042]; Ramos[0.042]; signing[0.042]; agreement[0.041]; agreement[0.041]; reporters[0.041]; peace[0.040]; peace[0.040]; support[0.040]; support[0.040]; released[0.040]; Monday[0.040]; rebels[0.040]; Conference[0.040]; served[0.039]; deal[0.039]; said[0.039]; supporting[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Suharto\u001b[39m ==> ENTITY: \u001b[32mSuharto\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.046]; president[0.044]; president[0.044]; support[0.043]; support[0.043]; Indonesian[0.043]; Suharto[0.043]; Jakarta[0.043]; reporters[0.042]; conflict[0.042]; island[0.042]; Monday[0.041]; said[0.041]; Islamic[0.041]; signing[0.041]; history[0.041]; southern[0.041]; lives[0.040]; war[0.040]; served[0.040]; released[0.040]; culminated[0.040]; scheduled[0.040]; hails[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Organisation of Islamic Conference\u001b[39m ==> ENTITY: \u001b[32mOrganisation of Islamic Cooperation\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.044]; Monday[0.043]; support[0.043]; support[0.043]; peace[0.042]; peace[0.042]; peace[0.042]; said[0.042]; agreement[0.042]; agreement[0.042]; agreement[0.042]; supporting[0.041]; rebels[0.041]; told[0.041]; formal[0.041]; talks[0.041]; reporters[0.041]; Philippine[0.041]; conflict[0.041]; scheduled[0.041]; Moslem[0.041]; negotiations[0.041]; host[0.041]; Friday[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Philippine ==> ENTITY: \u001b[32mPhilippines\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.114:0.114[0]; log p(e|m)= -0.567:-0.567[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nManila[0.051]; Manila[0.051]; Indonesia[0.047]; Indonesian[0.045]; text[0.045]; Ramos[0.044]; Ramos[0.044]; Monday[0.044]; local[0.043]; supporting[0.042]; formal[0.042]; letter[0.042]; letter[0.042]; Friday[0.042]; president[0.042]; president[0.042]; released[0.042]; peace[0.042]; peace[0.042]; host[0.042]; expressed[0.041]; served[0.041]; series[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jakarta\u001b[39m ==> ENTITY: \u001b[32mJakarta\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.049]; southern[0.046]; Manila[0.046]; Manila[0.046]; Manila[0.046]; Indonesian[0.046]; history[0.045]; island[0.045]; Monday[0.045]; Moslem[0.043]; Friday[0.043]; scheduled[0.043]; Philippine[0.042]; Philippine[0.042]; released[0.042]; local[0.042]; Islamic[0.042]; reporters[0.042]; extend[0.042]; served[0.041]; text[0.041]; series[0.041]; century[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Indonesian ==> ENTITY: \u001b[32mIndonesia\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.135:0.135[0]; log p(e|m)= -0.742:-0.742[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIndonesia[0.052]; Manila[0.047]; Manila[0.047]; Manila[0.047]; Philippine[0.046]; Jakarta[0.044]; text[0.042]; local[0.042]; released[0.041]; said[0.041]; Islamic[0.041]; Moslem[0.041]; Conference[0.040]; formal[0.040]; reporters[0.039]; host[0.039]; scheduled[0.039]; series[0.039]; deal[0.039]; signing[0.039]; agreement[0.039]; agreement[0.039]; agreement[0.039]; expressed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ramos\u001b[39m ==> ENTITY: \u001b[32mFidel V. Ramos\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhilippine[0.050]; Mindanao[0.047]; Manila[0.047]; president[0.044]; Ramos[0.042]; Ramos[0.042]; signing[0.042]; agreement[0.041]; agreement[0.041]; agreement[0.041]; reporters[0.041]; peace[0.041]; peace[0.041]; support[0.041]; released[0.041]; Monday[0.040]; rebels[0.040]; Conference[0.040]; served[0.040]; said[0.040]; supporting[0.039]; quarter[0.039]; excellency[0.039]; claimed[0.039]; \t\n\r [=========================>....................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s569ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2798/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 982testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Magna ==> ENTITY: \u001b[32mMagna International\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.193:0.193[0]; log p(e|m)= -1.262:-1.262[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMagna[0.048]; Magna[0.048]; business[0.043]; car[0.043]; purchased[0.043]; acquired[0.043]; Ontario[0.043]; American[0.042]; announcement[0.041]; North[0.041]; contracts[0.041]; maker[0.041]; Corp[0.041]; agreed[0.041]; vehicle[0.041]; vehicle[0.041]; acquisition[0.040]; makers[0.040]; seat[0.040]; Lear[0.040]; James[0.040]; shares[0.040]; shares[0.040]; shares[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Magna ==> ENTITY: \u001b[32mMagna International\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.189:0.189[0]; log p(e|m)= -1.262:-1.262[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFord[0.052]; model[0.051]; executives[0.050]; auto[0.049]; business[0.049]; car[0.049]; company[0.049]; company[0.049]; company[0.049]; plans[0.048]; American[0.048]; financial[0.047]; North[0.047]; contracts[0.046]; deal[0.046]; Corp[0.046]; vehicle[0.046]; makers[0.045]; changeovers[0.045]; Lear[0.045]; James[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ford ==> ENTITY: \u001b[32mFord Motor Company\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.340:-0.340[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncompany[0.045]; company[0.045]; model[0.045]; John[0.044]; executives[0.043]; competitive[0.043]; automakers[0.043]; production[0.042]; sales[0.042]; sales[0.042]; Johnson[0.042]; interiors[0.041]; Douglas[0.041]; Douglas[0.041]; suppliers[0.041]; market[0.041]; added[0.041]; working[0.041]; million[0.041]; million[0.041]; million[0.041]; million[0.041]; seats[0.040]; months[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = North American ==> ENTITY: \u001b[32mNorth America\u001b[39m\t\nSCORES: global= 0.241:0.241[0]; local()= 0.064:0.064[0]; log p(e|m)= -0.136:-0.136[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncompany[0.046]; company[0.046]; business[0.044]; International[0.043]; Thursday[0.042]; based[0.042]; Corp[0.042]; announcement[0.042]; James[0.042]; million[0.041]; million[0.041]; million[0.041]; million[0.041]; truck[0.041]; acquired[0.041]; executives[0.041]; acquisition[0.041]; competitive[0.041]; contracts[0.040]; Ontario[0.040]; maker[0.040]; financial[0.040]; Friday[0.040]; revenue[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Magna ==> ENTITY: \u001b[32mMagna International\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.193:0.193[0]; log p(e|m)= -1.262:-1.262[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMagna[0.048]; Magna[0.048]; Ford[0.046]; model[0.045]; company[0.043]; plans[0.043]; automakers[0.043]; deal[0.041]; interiors[0.041]; interior[0.041]; changeovers[0.041]; Lear[0.040]; million[0.040]; million[0.040]; million[0.040]; million[0.040]; seating[0.040]; sales[0.040]; sales[0.040]; components[0.040]; build[0.040]; suppliers[0.040]; production[0.040]; deals[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Farmington Hills\u001b[39m ==> ENTITY: \u001b[32mFarmington Hills, Michigan\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.126:0.126[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThursday[0.045]; North[0.045]; Friday[0.044]; based[0.044]; American[0.044]; million[0.044]; million[0.044]; acquired[0.042]; Mich[0.042]; outstanding[0.041]; beef[0.041]; York[0.041]; Markham[0.041]; afternoon[0.041]; International[0.041]; Douglas[0.040]; Douglas[0.040]; Douglas[0.040]; vehicle[0.040]; share[0.040]; merger[0.040]; car[0.040]; seat[0.040]; Ontario[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Markham ==> ENTITY: \u001b[32mMarkham, Ontario\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.947:-0.947[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOntario[0.047]; York[0.044]; North[0.043]; American[0.043]; business[0.042]; James[0.042]; Douglas[0.042]; Douglas[0.042]; Douglas[0.042]; Douglas[0.042]; Douglas[0.042]; acquired[0.042]; based[0.041]; seat[0.041]; Hills[0.041]; million[0.041]; million[0.041]; million[0.041]; million[0.041]; financial[0.041]; Thursday[0.040]; Corp[0.040]; Farmington[0.040]; Johnson[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Magna ==> ENTITY: \u001b[32mMagna International\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.189:0.189[0]; log p(e|m)= -1.262:-1.262[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMagna[0.049]; Magna[0.049]; Ford[0.046]; company[0.044]; plans[0.043]; automakers[0.043]; deal[0.042]; interiors[0.041]; interior[0.041]; Lear[0.040]; million[0.040]; million[0.040]; million[0.040]; million[0.040]; seating[0.040]; sales[0.040]; sales[0.040]; components[0.040]; build[0.040]; suppliers[0.040]; production[0.040]; deals[0.040]; earned[0.040]; Wertheim[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Johnson Controls Inc\u001b[39m ==> ENTITY: \u001b[32mJohnson Controls\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncompany[0.047]; company[0.047]; business[0.043]; auto[0.043]; maker[0.043]; Lear[0.043]; American[0.042]; executives[0.042]; vehicle[0.041]; vehicle[0.041]; contracts[0.041]; Corp[0.041]; International[0.041]; car[0.041]; Magna[0.041]; Magna[0.041]; Magna[0.041]; Magna[0.041]; financial[0.041]; makers[0.040]; industry[0.040]; Stock[0.040]; based[0.040]; purchased[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = New York Stock Exchange\u001b[39m ==> ENTITY: \u001b[32mNew York Stock Exchange\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nshares[0.044]; shares[0.044]; shares[0.044]; financial[0.044]; Thursday[0.044]; Corp[0.044]; Friday[0.043]; Shares[0.043]; acquired[0.042]; James[0.042]; business[0.041]; International[0.041]; announcement[0.041]; afternoon[0.041]; said[0.041]; agreed[0.040]; American[0.040]; share[0.040]; car[0.040]; million[0.040]; million[0.040]; million[0.040]; Family[0.040]; price[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Magna ==> ENTITY: \u001b[32mMagna International\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.184:0.184[0]; log p(e|m)= -1.262:-1.262[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMagna[0.049]; Magna[0.049]; Ford[0.046]; company[0.044]; plans[0.043]; automakers[0.043]; deal[0.042]; interiors[0.041]; interior[0.041]; Lear[0.040]; million[0.040]; million[0.040]; million[0.040]; million[0.040]; seating[0.040]; sales[0.040]; sales[0.040]; components[0.040]; build[0.040]; suppliers[0.040]; deals[0.040]; earned[0.040]; Wertheim[0.040]; half[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Magna ==> ENTITY: \u001b[32mMagna International\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.212:0.212[0]; log p(e|m)= -1.262:-1.262[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMagna[0.047]; Magna[0.047]; executives[0.043]; auto[0.043]; business[0.043]; car[0.042]; company[0.042]; company[0.042]; purchased[0.042]; acquired[0.042]; Ontario[0.042]; American[0.042]; financial[0.041]; announcement[0.041]; North[0.041]; contracts[0.040]; maker[0.040]; Corp[0.040]; agreed[0.040]; vehicle[0.040]; vehicle[0.040]; acquisition[0.040]; makers[0.040]; seat[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Johnson Controls\u001b[39m ==> ENTITY: \u001b[32mJohnson Controls\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFord[0.053]; components[0.053]; Lear[0.050]; interiors[0.049]; automakers[0.049]; suppliers[0.049]; sales[0.049]; sales[0.049]; interior[0.048]; Magna[0.047]; Magna[0.047]; Magna[0.047]; plans[0.046]; months[0.046]; seating[0.046]; build[0.046]; panels[0.045]; panels[0.045]; panels[0.045]; panels[0.045]; really[0.045]; \t\n\r [=========================>....................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s497ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2811/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1107testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Turkey ==> ENTITY: \u001b[32mTurkey national football team\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.162:0.162[0]; log p(e|m)= -3.037:-3.037[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.053]; Turkey[0.053]; soccer[0.044]; qualifier[0.044]; Belgium[0.044]; Belgium[0.044]; Belgium[0.044]; Sergen[0.043]; Soccer[0.043]; Cup[0.042]; Cup[0.042]; Yalcin[0.042]; Qualifier[0.041]; seven[0.041]; group[0.040]; beat[0.039]; World[0.039]; World[0.039]; Saturday[0.038]; halftime[0.038]; Attendance[0.038]; Marc[0.037]; Beat[0.037]; Brussels[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brussels ==> ENTITY: \u001b[32mBrussels\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.077:0.077[0]; log p(e|m)= -0.025:-0.025[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.048]; Belgium[0.048]; Belgium[0.048]; soccer[0.046]; Soccer[0.045]; Qualifier[0.044]; Cup[0.044]; Cup[0.044]; Saturday[0.043]; qualifier[0.042]; Degryse[0.042]; World[0.041]; World[0.041]; Marc[0.039]; Attendance[0.039]; seven[0.039]; beat[0.039]; Beat[0.038]; Turkey[0.038]; Turkey[0.038]; Turkey[0.038]; Luis[0.038]; group[0.038]; Scorers[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Turkey ==> ENTITY: \u001b[32mTurkey national football team\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.162:0.162[0]; log p(e|m)= -3.037:-3.037[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.053]; Turkey[0.053]; soccer[0.044]; qualifier[0.044]; Belgium[0.044]; Belgium[0.044]; Belgium[0.044]; Sergen[0.043]; Soccer[0.043]; Cup[0.042]; Cup[0.042]; Yalcin[0.042]; Qualifier[0.041]; seven[0.041]; group[0.040]; beat[0.039]; World[0.039]; World[0.039]; Saturday[0.038]; halftime[0.038]; Attendance[0.038]; Marc[0.037]; Beat[0.037]; Brussels[0.037]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBelgium\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBelgium\u001b[39m <---> \u001b[32mBelgium national football team\u001b[39m\t\nSCORES: global= 0.253:0.252[\u001b[31m0.002\u001b[39m]; local()= 0.127:0.146[\u001b[32m0.019\u001b[39m]; log p(e|m)= -0.242:-2.882[\u001b[31m2.640\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.052]; Belgium[0.052]; soccer[0.048]; Soccer[0.047]; qualifier[0.043]; Cup[0.043]; Cup[0.043]; Marc[0.042]; beat[0.041]; Qualifier[0.041]; Turkey[0.041]; Turkey[0.041]; Turkey[0.041]; seven[0.041]; Degryse[0.039]; Brussels[0.039]; halftime[0.039]; group[0.039]; World[0.039]; World[0.039]; Saturday[0.038]; Sergen[0.038]; Luis[0.038]; Beat[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Cup ==> ENTITY: \u001b[32mFIFA World Cup\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.123:0.123[0]; log p(e|m)= -1.002:-1.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.048]; Soccer[0.048]; qualifier[0.045]; Cup[0.045]; World[0.045]; Qualifier[0.044]; Belgium[0.043]; Belgium[0.043]; Belgium[0.043]; seven[0.042]; Turkey[0.041]; Turkey[0.041]; Turkey[0.041]; Saturday[0.040]; group[0.040]; Sergen[0.040]; Marc[0.040]; Luis[0.040]; Beat[0.039]; Oliveira[0.039]; beat[0.039]; Scorers[0.038]; halftime[0.038]; Yalcin[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marc Degryse\u001b[39m ==> ENTITY: \u001b[32mMarc Degryse\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.101:0.101[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.050]; Belgium[0.050]; Belgium[0.050]; soccer[0.049]; Cup[0.047]; Cup[0.047]; qualifier[0.046]; Qualifier[0.046]; Soccer[0.046]; World[0.046]; World[0.046]; halftime[0.045]; seven[0.045]; Scorers[0.045]; Attendance[0.045]; Sergen[0.043]; Brussels[0.043]; Luis[0.043]; Saturday[0.043]; Yalcin[0.042]; group[0.042]; beat[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Turkey ==> ENTITY: \u001b[32mTurkey national football team\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.162:0.162[0]; log p(e|m)= -3.037:-3.037[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.053]; Turkey[0.053]; soccer[0.044]; qualifier[0.044]; Belgium[0.044]; Belgium[0.044]; Belgium[0.044]; Sergen[0.043]; Soccer[0.043]; Cup[0.042]; Cup[0.042]; Yalcin[0.042]; Qualifier[0.041]; seven[0.041]; group[0.040]; beat[0.039]; World[0.039]; World[0.039]; Saturday[0.038]; halftime[0.038]; Attendance[0.038]; Marc[0.037]; Beat[0.037]; Brussels[0.037]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBelgium\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBelgium\u001b[39m <---> \u001b[32mBelgium national football team\u001b[39m\t\nSCORES: global= 0.253:0.252[\u001b[31m0.002\u001b[39m]; local()= 0.127:0.146[\u001b[32m0.019\u001b[39m]; log p(e|m)= -0.242:-2.882[\u001b[31m2.640\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.052]; Belgium[0.052]; soccer[0.048]; Soccer[0.047]; qualifier[0.043]; Cup[0.043]; Cup[0.043]; Marc[0.042]; beat[0.041]; Qualifier[0.041]; Turkey[0.041]; Turkey[0.041]; Turkey[0.041]; seven[0.041]; Degryse[0.039]; Brussels[0.039]; halftime[0.039]; group[0.039]; World[0.039]; World[0.039]; Saturday[0.038]; Sergen[0.038]; Luis[0.038]; Beat[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Luis Oliveira\u001b[39m ==> ENTITY: \u001b[32mLuís Oliveira\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.047]; Soccer[0.045]; Belgium[0.044]; Belgium[0.044]; Belgium[0.044]; seven[0.044]; Degryse[0.043]; qualifier[0.042]; Qualifier[0.042]; halftime[0.041]; Cup[0.041]; Cup[0.041]; Scorers[0.041]; Saturday[0.041]; Attendance[0.041]; group[0.041]; Sergen[0.040]; Brussels[0.040]; World[0.039]; World[0.039]; Turkey[0.039]; Turkey[0.039]; Turkey[0.039]; Marc[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBelgium\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBelgium\u001b[39m <---> \u001b[32mBelgium national football team\u001b[39m\t\nSCORES: global= 0.253:0.252[\u001b[31m0.002\u001b[39m]; local()= 0.127:0.146[\u001b[32m0.019\u001b[39m]; log p(e|m)= -0.242:-2.882[\u001b[31m2.640\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.052]; Belgium[0.052]; soccer[0.048]; Soccer[0.047]; qualifier[0.043]; Cup[0.043]; Cup[0.043]; Marc[0.042]; beat[0.041]; Qualifier[0.041]; Turkey[0.041]; Turkey[0.041]; Turkey[0.041]; seven[0.041]; Degryse[0.039]; Brussels[0.039]; halftime[0.039]; group[0.039]; World[0.039]; World[0.039]; Saturday[0.038]; Sergen[0.038]; Luis[0.038]; Beat[0.037]; \t\n\r [==========================>...................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s469ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2822/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1005testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Phnom Penh\u001b[39m ==> ENTITY: \u001b[32mPhnom Penh\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.118:0.118[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCambodia[0.049]; Cambodian[0.047]; Norodom[0.046]; Khmer[0.046]; August[0.043]; Sihanouk[0.043]; Sihanouk[0.043]; royal[0.041]; Friday[0.041]; Friday[0.041]; upholding[0.041]; colonial[0.041]; gates[0.041]; Nation[0.040]; following[0.040]; T3[0.040]; era[0.040]; late[0.040]; sent[0.040]; level[0.040]; Party[0.039]; corruption[0.039]; Court[0.039]; member[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Norodom Sihanouk\u001b[39m ==> ENTITY: \u001b[32mNorodom Sihanouk\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSihanouk[0.046]; Cambodia[0.045]; Cambodian[0.044]; Phnom[0.043]; Khmer[0.043]; Penh[0.043]; King[0.043]; King[0.043]; royal[0.043]; Supreme[0.042]; pardon[0.041]; French[0.041]; Party[0.041]; August[0.041]; pardoned[0.041]; sent[0.040]; late[0.040]; era[0.040]; colonial[0.040]; level[0.040]; Friday[0.040]; Friday[0.040]; Nation[0.040]; member[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Norodom Ranariddh\u001b[39m ==> ENTITY: \u001b[32mNorodom Ranariddh\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.120:0.120[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSihanouk[0.046]; king[0.045]; Party[0.043]; Sen[0.043]; member[0.043]; member[0.043]; Khmer[0.042]; serving[0.042]; Prince[0.042]; committee[0.042]; agreed[0.042]; opposition[0.041]; prosecuted[0.041]; sent[0.041]; Hun[0.041]; judge[0.041]; amnesty[0.041]; amnesty[0.041]; Rattana[0.041]; decision[0.040]; overturned[0.040]; following[0.040]; Chan[0.040]; courts[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sihanouk\u001b[39m ==> ENTITY: \u001b[32mNorodom Sihanouk\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorodom[0.047]; king[0.046]; Sihanouk[0.046]; Khmer[0.043]; King[0.043]; Ranariddh[0.042]; Supreme[0.042]; serving[0.042]; promised[0.042]; Chan[0.042]; term[0.041]; French[0.041]; Rattana[0.041]; Party[0.041]; week[0.041]; week[0.041]; Prince[0.041]; June[0.040]; August[0.040]; sent[0.040]; late[0.040]; era[0.040]; Premiers[0.040]; colonial[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hun Sen\u001b[39m ==> ENTITY: \u001b[32mHun Sen\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.116:0.116[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorodom[0.047]; Ranariddh[0.043]; Khmer[0.043]; Rattana[0.043]; Sihanouk[0.043]; Party[0.043]; alleging[0.042]; week[0.042]; week[0.042]; following[0.042]; Chan[0.042]; term[0.041]; serving[0.041]; judge[0.041]; riels[0.041]; decision[0.041]; promised[0.041]; amnesty[0.040]; amnesty[0.040]; king[0.040]; fellow[0.040]; released[0.040]; earlier[0.040]; long[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cambodia ==> ENTITY: \u001b[32mCambodia\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.129:0.129[0]; log p(e|m)= -0.273:-0.273[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCambodian[0.046]; Sihanouk[0.045]; Sihanouk[0.045]; Norodom[0.045]; Khmer[0.044]; colonial[0.044]; Penh[0.043]; member[0.042]; Phnom[0.042]; courts[0.042]; pardon[0.041]; pardoned[0.041]; royal[0.040]; French[0.040]; Party[0.040]; level[0.040]; Friday[0.040]; Friday[0.040]; following[0.040]; Nation[0.040]; August[0.040]; late[0.039]; committee[0.039]; era[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sihanouk\u001b[39m ==> ENTITY: \u001b[32mNorodom Sihanouk\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.142:0.142[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorodom[0.046]; Sihanouk[0.045]; Sihanouk[0.045]; Cambodia[0.044]; Cambodian[0.043]; Phnom[0.042]; Khmer[0.042]; Penh[0.042]; King[0.042]; King[0.042]; royal[0.042]; Supreme[0.041]; serving[0.041]; promised[0.041]; Chan[0.041]; pardon[0.040]; term[0.040]; French[0.040]; Rattana[0.040]; Party[0.040]; week[0.040]; August[0.040]; pardoned[0.040]; sent[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.241:0.241[0]; local()= 0.033:0.033[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncolonial[0.051]; Cambodia[0.047]; upholding[0.045]; article[0.045]; Khmer[0.044]; rulings[0.044]; pardon[0.044]; appeal[0.044]; era[0.043]; prosecuted[0.043]; Cambodian[0.043]; late[0.043]; overturned[0.042]; Friday[0.042]; Friday[0.042]; courts[0.042]; August[0.042]; Supreme[0.042]; imprisonment[0.042]; Party[0.042]; Nation[0.042]; royal[0.042]; Penh[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cambodian ==> ENTITY: \u001b[32mCambodia\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.331:-0.331[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKhmer[0.052]; Cambodia[0.051]; Sihanouk[0.045]; Sihanouk[0.045]; Norodom[0.045]; colonial[0.044]; Penh[0.043]; Phnom[0.042]; pardon[0.041]; royal[0.041]; pardoned[0.041]; French[0.040]; Party[0.040]; Nation[0.040]; late[0.039]; King[0.039]; King[0.039]; era[0.039]; level[0.039]; sent[0.039]; alleged[0.039]; gave[0.038]; Hen[0.038]; August[0.038]; \t\n\r [==========================>...................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s550ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2831/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1062testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = English ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.117:0.117[0]; log p(e|m)= -1.355:-1.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.047]; English[0.047]; English[0.047]; rugby[0.047]; League[0.046]; Cardiff[0.045]; league[0.045]; Welsh[0.044]; Welsh[0.044]; Welsh[0.044]; matches[0.043]; Swansea[0.043]; Bath[0.043]; Bristol[0.043]; Harlequins[0.043]; Northampton[0.042]; Leicester[0.042]; Saracens[0.041]; Gloucester[0.041]; division[0.041]; division[0.041]; Irish[0.041]; Wasps[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scottish ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.480:-0.480[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScottish[0.047]; Scottish[0.047]; Cardiff[0.046]; League[0.046]; Welsh[0.045]; Welsh[0.045]; Welsh[0.045]; matches[0.045]; league[0.045]; Irish[0.045]; division[0.043]; division[0.043]; Rugby[0.043]; National[0.043]; Stirling[0.042]; Swansea[0.042]; Bristol[0.042]; Leicester[0.041]; Currie[0.041]; premier[0.041]; London[0.040]; London[0.040]; rugby[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bridgend ==> ENTITY: \u001b[32mBridgend Ravens\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.224:0.224[0]; log p(e|m)= -2.040:-2.040[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.045]; Harlequins[0.044]; Rugby[0.044]; Cardiff[0.043]; Saracens[0.043]; league[0.042]; Swansea[0.042]; Northampton[0.042]; Welsh[0.042]; Welsh[0.042]; Welsh[0.042]; Wasps[0.042]; League[0.042]; Caerphilly[0.042]; Bristol[0.041]; Bath[0.041]; Llanelli[0.041]; matches[0.041]; Gloucester[0.041]; Treorchy[0.040]; Ebbw[0.040]; Leicester[0.040]; Irish[0.039]; Heriot[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Boroughmuir\u001b[39m ==> ENTITY: \u001b[32mBoroughmuir RFC\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.046]; Rugby[0.046]; league[0.045]; Stirling[0.043]; matches[0.043]; Heriot[0.042]; Harlequins[0.042]; League[0.042]; Welsh[0.042]; Welsh[0.042]; Welsh[0.042]; Cardiff[0.042]; Watsonians[0.041]; Wasps[0.041]; Scottish[0.041]; Scottish[0.041]; Scottish[0.041]; Hawick[0.040]; Orrell[0.040]; Currie[0.040]; Bath[0.040]; Saracens[0.040]; Leicester[0.040]; Llanelli[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Hartlepool ==> ENTITY: \u001b[32mWest Hartlepool R.F.C.\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.182:0.182[0]; log p(e|m)= -0.865:-0.865[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.051]; league[0.050]; Rugby[0.049]; rugby[0.048]; matches[0.047]; Bristol[0.046]; Cardiff[0.046]; Swansea[0.046]; division[0.046]; division[0.046]; Orrell[0.045]; Newport[0.045]; Bridgend[0.044]; Harlequins[0.044]; Caerphilly[0.044]; Vale[0.044]; Northampton[0.044]; Boroughmuir[0.044]; Ebbw[0.043]; Stirling[0.043]; National[0.043]; Saturday[0.043]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mHarlequins\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHarlequin F.C.\u001b[39m <---> \u001b[32mLondon Broncos\u001b[39m\t\nSCORES: global= 0.287:0.270[\u001b[31m0.017\u001b[39m]; local()= 0.234:0.199[\u001b[31m0.035\u001b[39m]; log p(e|m)= 0.000:-1.501[\u001b[31m1.501\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.047]; rugby[0.046]; Leicester[0.044]; matches[0.044]; League[0.044]; league[0.044]; Bath[0.043]; Cardiff[0.043]; Saracens[0.043]; Wasps[0.042]; Bristol[0.042]; Northampton[0.042]; Gloucester[0.041]; Swansea[0.040]; Welsh[0.040]; Welsh[0.040]; Welsh[0.040]; Boroughmuir[0.040]; Stirling[0.039]; London[0.039]; London[0.039]; division[0.039]; division[0.039]; Orrell[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Caerphilly ==> ENTITY: \u001b[32mCaerphilly RFC\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.201:0.201[0]; log p(e|m)= -1.645:-1.645[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.046]; rugby[0.046]; League[0.046]; league[0.046]; Welsh[0.046]; Welsh[0.046]; Welsh[0.046]; Swansea[0.044]; Bridgend[0.044]; Cardiff[0.044]; County[0.043]; Harlequins[0.043]; Saracens[0.043]; Llanelli[0.043]; matches[0.042]; Wasps[0.042]; Treorchy[0.042]; Bristol[0.042]; Newport[0.042]; Ebbw[0.042]; National[0.042]; Orrell[0.041]; Hartlepool[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = English ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.117:0.117[0]; log p(e|m)= -1.355:-1.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.047]; English[0.047]; English[0.047]; rugby[0.047]; League[0.046]; Cardiff[0.045]; league[0.045]; Welsh[0.044]; Welsh[0.044]; Welsh[0.044]; matches[0.043]; Swansea[0.043]; Bath[0.043]; Bristol[0.043]; Harlequins[0.043]; Northampton[0.042]; Leicester[0.042]; Saracens[0.041]; Gloucester[0.041]; division[0.041]; division[0.041]; Irish[0.041]; Wasps[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Saracens ==> ENTITY: \u001b[32mSaracens F.C.\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.217:0.217[0]; log p(e|m)= -0.567:-0.567[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.050]; rugby[0.050]; Harlequins[0.047]; Bath[0.047]; Leicester[0.046]; league[0.045]; Wasps[0.045]; League[0.045]; Northampton[0.044]; Bristol[0.044]; Gloucester[0.043]; Cardiff[0.043]; matches[0.042]; Welsh[0.042]; Welsh[0.042]; Welsh[0.042]; Saturday[0.041]; London[0.041]; London[0.041]; West[0.041]; Irish[0.040]; Llanelli[0.040]; Newbridge[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mTreorchy\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mTreorchy\u001b[39m <---> \u001b[32mTreorchy RFC\u001b[39m\t\nSCORES: global= 0.288:0.284[\u001b[31m0.004\u001b[39m]; local()= 0.126:0.203[\u001b[32m0.076\u001b[39m]; log p(e|m)= 0.000:-0.942[\u001b[31m0.942\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.045]; league[0.045]; Rugby[0.045]; Welsh[0.044]; Welsh[0.044]; Welsh[0.044]; Bridgend[0.043]; Caerphilly[0.042]; Cardiff[0.042]; Harlequins[0.042]; Saracens[0.042]; Swansea[0.041]; Wasps[0.041]; League[0.041]; Northampton[0.041]; Bristol[0.041]; Llanelli[0.041]; Leicester[0.040]; Ebbw[0.040]; Vale[0.040]; West[0.039]; Boroughmuir[0.039]; Gloucester[0.039]; matches[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bath ==> ENTITY: \u001b[32mBath Rugby\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.206:0.206[0]; log p(e|m)= -2.604:-2.604[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBristol[0.048]; Gloucester[0.048]; Northampton[0.047]; Cardiff[0.045]; Swansea[0.044]; Leicester[0.043]; Llanelli[0.043]; Saracens[0.042]; Rugby[0.042]; Stirling[0.041]; rugby[0.041]; Harlequins[0.041]; Dunvant[0.041]; Wasps[0.040]; Hartlepool[0.040]; Welsh[0.040]; Welsh[0.040]; Welsh[0.040]; league[0.039]; Bridgend[0.039]; Scottish[0.039]; Scottish[0.039]; Scottish[0.039]; Caerphilly[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = London Irish\u001b[39m ==> ENTITY: \u001b[32mLondon Irish\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.240:0.240[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.047]; rugby[0.046]; Harlequins[0.046]; Wasps[0.044]; Saracens[0.043]; Cardiff[0.043]; league[0.043]; Leicester[0.043]; League[0.042]; Bath[0.042]; Northampton[0.041]; Swansea[0.041]; Welsh[0.041]; Welsh[0.041]; Welsh[0.041]; matches[0.041]; Gloucester[0.041]; Bristol[0.040]; Llanelli[0.040]; Bridgend[0.039]; Newbridge[0.039]; Boroughmuir[0.038]; Caerphilly[0.038]; Ebbw[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hawick ==> ENTITY: \u001b[32mHawick RFC\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.160:0.160[0]; log p(e|m)= -1.682:-1.682[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.046]; league[0.045]; rugby[0.045]; League[0.044]; matches[0.043]; Scottish[0.042]; Scottish[0.042]; Scottish[0.042]; Stirling[0.042]; County[0.041]; Watsonians[0.041]; Saturday[0.041]; Cardiff[0.041]; Boroughmuir[0.041]; Welsh[0.041]; Welsh[0.041]; Welsh[0.041]; Heriot[0.040]; Ebbw[0.040]; Caerphilly[0.040]; Bridgend[0.040]; National[0.040]; Currie[0.040]; premier[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scottish ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.480:-0.480[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScottish[0.047]; Scottish[0.047]; Cardiff[0.046]; League[0.046]; Welsh[0.045]; Welsh[0.045]; Welsh[0.045]; Irish[0.045]; matches[0.045]; league[0.044]; division[0.043]; division[0.043]; Rugby[0.043]; National[0.043]; Stirling[0.042]; Swansea[0.042]; Bristol[0.042]; Leicester[0.041]; Currie[0.041]; premier[0.041]; London[0.040]; London[0.040]; rugby[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Llanelli ==> ENTITY: \u001b[32mLlanelli RFC\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.210:0.210[0]; log p(e|m)= -1.911:-1.911[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.045]; rugby[0.045]; Rugby[0.044]; Welsh[0.043]; Welsh[0.043]; Welsh[0.043]; Cardiff[0.043]; Harlequins[0.043]; Bath[0.043]; Swansea[0.042]; League[0.042]; matches[0.042]; Caerphilly[0.042]; Saracens[0.041]; Bridgend[0.041]; Bristol[0.040]; Wasps[0.040]; Northampton[0.040]; Ebbw[0.040]; Gloucester[0.040]; Newport[0.039]; Leicester[0.039]; Stirling[0.039]; Irish[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCurrie\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mCurrie\u001b[39m <---> \u001b[32mCurrie RFC\u001b[39m\t\nSCORES: global= 0.251:0.234[\u001b[31m0.017\u001b[39m]; local()= 0.072:0.173[\u001b[32m0.101\u001b[39m]; log p(e|m)= -0.176:-4.017[\u001b[31m3.842\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.047]; league[0.046]; rugby[0.045]; Leicester[0.045]; Swansea[0.045]; Cardiff[0.045]; Scottish[0.044]; Scottish[0.044]; Scottish[0.044]; Hartlepool[0.044]; Boroughmuir[0.044]; Bristol[0.044]; matches[0.043]; Harlequins[0.043]; Stirling[0.043]; Northampton[0.042]; Forest[0.042]; County[0.042]; Watsonians[0.042]; Heriot[0.042]; Union[0.042]; Saturday[0.042]; Llanelli[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scottish ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.126:0.126[0]; log p(e|m)= -0.480:-0.480[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScottish[0.049]; Scottish[0.049]; Cardiff[0.048]; League[0.048]; Welsh[0.047]; Welsh[0.047]; Welsh[0.047]; Irish[0.047]; matches[0.047]; league[0.046]; division[0.045]; division[0.045]; Rugby[0.045]; National[0.045]; Swansea[0.044]; Bristol[0.044]; Leicester[0.043]; Currie[0.043]; premier[0.043]; London[0.042]; London[0.042]; rugby[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swansea ==> ENTITY: \u001b[32mSwansea RFC\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.233:0.233[0]; log p(e|m)= -2.375:-2.375[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.046]; Rugby[0.045]; Cardiff[0.044]; Welsh[0.044]; Welsh[0.044]; Welsh[0.044]; Llanelli[0.043]; Caerphilly[0.042]; matches[0.042]; League[0.042]; Saracens[0.041]; Harlequins[0.041]; Bridgend[0.041]; league[0.041]; Bristol[0.041]; Northampton[0.041]; Ebbw[0.040]; Leicester[0.040]; Gloucester[0.040]; Newbridge[0.040]; Bath[0.040]; Saturday[0.039]; Wasps[0.039]; Treorchy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leicester ==> ENTITY: \u001b[32mLeicester Tigers\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.193:0.193[0]; log p(e|m)= -3.194:-3.194[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.048]; rugby[0.047]; League[0.045]; Bristol[0.044]; league[0.043]; Harlequins[0.043]; Bath[0.043]; Cardiff[0.042]; matches[0.042]; Swansea[0.042]; Saracens[0.041]; Gloucester[0.041]; London[0.041]; London[0.041]; Wasps[0.041]; Welsh[0.041]; Welsh[0.041]; Welsh[0.041]; Northampton[0.041]; Saturday[0.039]; Hartlepool[0.039]; West[0.039]; National[0.039]; Newport[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Northampton ==> ENTITY: \u001b[32mNorthampton Saints\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.239:0.239[0]; log p(e|m)= -3.147:-3.147[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.049]; rugby[0.048]; League[0.046]; Harlequins[0.046]; league[0.045]; Saracens[0.045]; Cardiff[0.045]; matches[0.044]; Swansea[0.044]; Bath[0.044]; Leicester[0.044]; Wasps[0.043]; Gloucester[0.043]; Welsh[0.042]; Welsh[0.042]; Welsh[0.042]; Bristol[0.042]; National[0.042]; Llanelli[0.042]; Hartlepool[0.041]; Caerphilly[0.041]; Bridgend[0.040]; Irish[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Welsh ==> ENTITY: \u001b[32mWales\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.149:0.149[0]; log p(e|m)= -1.079:-1.079[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.046]; rugby[0.046]; Welsh[0.045]; Welsh[0.045]; Cardiff[0.045]; matches[0.042]; English[0.042]; English[0.042]; English[0.042]; League[0.041]; league[0.041]; Scottish[0.041]; Scottish[0.041]; Scottish[0.041]; Bridgend[0.041]; Bath[0.041]; Llanelli[0.041]; Harlequins[0.040]; Bristol[0.040]; Swansea[0.040]; union[0.040]; Saracens[0.040]; Leicester[0.039]; Treorchy[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNewbridge\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNewbridge, Caerphilly\u001b[39m <---> \u001b[32mNewbridge RFC\u001b[39m\t\nSCORES: global= 0.254:0.243[\u001b[31m0.011\u001b[39m]; local()= 0.194:0.132[\u001b[31m0.062\u001b[39m]; log p(e|m)= -1.833:-2.386[\u001b[31m0.553\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.044]; Caerphilly[0.043]; Bridgend[0.043]; Welsh[0.043]; Welsh[0.043]; Welsh[0.043]; Rugby[0.042]; Irish[0.042]; Bristol[0.042]; Llanelli[0.042]; Swansea[0.042]; Treorchy[0.041]; Hawick[0.041]; league[0.041]; Dunvant[0.041]; Cardiff[0.041]; Scottish[0.041]; Scottish[0.041]; Scottish[0.041]; Hartlepool[0.041]; West[0.041]; Ebbw[0.041]; Stirling[0.040]; Bath[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sale ==> ENTITY: \u001b[32mSale Sharks\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.193:0.193[0]; log p(e|m)= -2.040:-2.040[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.046]; Cardiff[0.044]; rugby[0.044]; Bath[0.044]; Leicester[0.044]; Bristol[0.043]; Harlequins[0.043]; league[0.043]; Saracens[0.043]; Northampton[0.042]; Gloucester[0.042]; matches[0.042]; Wasps[0.042]; League[0.041]; Swansea[0.041]; Welsh[0.041]; Welsh[0.041]; Welsh[0.041]; Caerphilly[0.039]; Bridgend[0.039]; Llanelli[0.039]; Dunvant[0.039]; Newbridge[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jed-Forest\u001b[39m ==> ENTITY: \u001b[32mJed-Forest RFC\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; League[0.045]; rugby[0.044]; National[0.043]; Stirling[0.043]; Heriot[0.042]; Hawick[0.042]; Boroughmuir[0.042]; Watsonians[0.042]; Vale[0.042]; Harlequins[0.042]; division[0.041]; division[0.041]; Scottish[0.041]; Scottish[0.041]; Scottish[0.041]; Cardiff[0.041]; Forest[0.041]; premier[0.040]; Melrose[0.040]; West[0.040]; Saracens[0.040]; Bath[0.040]; Bristol[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Watsonians\u001b[39m ==> ENTITY: \u001b[32mWatsonians RFC\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.047]; league[0.044]; League[0.044]; Stirling[0.043]; matches[0.042]; Scottish[0.042]; Scottish[0.042]; Boroughmuir[0.042]; Welsh[0.042]; Welsh[0.042]; Cardiff[0.042]; Heriot[0.042]; Gloucester[0.041]; Leicester[0.041]; Currie[0.041]; Harlequins[0.041]; Swansea[0.041]; National[0.040]; Bath[0.040]; Wasps[0.040]; Saracens[0.040]; Llanelli[0.040]; Northampton[0.040]; Caerphilly[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dunvant ==> ENTITY: \u001b[32mDunvant RFC\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.263:0.263[0]; log p(e|m)= -1.082:-1.082[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.049]; Rugby[0.048]; Harlequins[0.043]; Saracens[0.043]; Welsh[0.042]; Welsh[0.042]; Welsh[0.042]; Bridgend[0.042]; Wasps[0.042]; union[0.042]; Bath[0.042]; Llanelli[0.042]; Cardiff[0.041]; Caerphilly[0.041]; Swansea[0.041]; Union[0.041]; Orrell[0.040]; Ebbw[0.040]; West[0.040]; league[0.039]; Treorchy[0.039]; Leicester[0.039]; County[0.039]; Bristol[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Orrell ==> ENTITY: \u001b[32mOrrell R.U.F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.185:0.185[0]; log p(e|m)= -0.931:-0.931[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorthampton[0.045]; Rugby[0.044]; Cardiff[0.043]; Leicester[0.043]; Bristol[0.043]; rugby[0.043]; Gloucester[0.043]; Swansea[0.043]; Bath[0.043]; Harlequins[0.042]; League[0.042]; league[0.042]; Bridgend[0.041]; Welsh[0.041]; Welsh[0.041]; Welsh[0.041]; Ebbw[0.041]; matches[0.041]; Saracens[0.040]; Boroughmuir[0.040]; Caerphilly[0.039]; English[0.039]; English[0.039]; English[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gloucester ==> ENTITY: \u001b[32mGloucester Rugby\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.235:0.235[0]; log p(e|m)= -3.058:-3.058[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.047]; rugby[0.046]; Cardiff[0.044]; Bristol[0.044]; Harlequins[0.043]; Bath[0.043]; Leicester[0.042]; Saracens[0.042]; Welsh[0.042]; Welsh[0.042]; Welsh[0.042]; Northampton[0.042]; Wasps[0.042]; matches[0.041]; Newport[0.041]; Swansea[0.041]; league[0.041]; Llanelli[0.040]; County[0.040]; Bridgend[0.040]; Caerphilly[0.039]; West[0.039]; League[0.039]; Treorchy[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCardiff\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mCardiff\u001b[39m <---> \u001b[32mCardiff RFC\u001b[39m\t\nSCORES: global= 0.245:0.243[\u001b[31m0.003\u001b[39m]; local()= 0.166:0.205[\u001b[32m0.039\u001b[39m]; log p(e|m)= -0.205:-2.781[\u001b[31m2.576\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.045]; rugby[0.045]; Swansea[0.043]; Harlequins[0.043]; Welsh[0.043]; Welsh[0.043]; Welsh[0.043]; league[0.043]; matches[0.043]; Bristol[0.042]; Leicester[0.042]; Saracens[0.042]; Wasps[0.042]; League[0.041]; Northampton[0.041]; Bath[0.041]; Llanelli[0.040]; Newport[0.040]; Gloucester[0.040]; Bridgend[0.040]; London[0.040]; London[0.040]; Stirling[0.040]; Caerphilly[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rugby Union\u001b[39m ==> ENTITY: \u001b[32mRugby union\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.225:0.225[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.053]; Cardiff[0.044]; Welsh[0.043]; Welsh[0.043]; Welsh[0.043]; Bath[0.043]; league[0.043]; matches[0.042]; Harlequins[0.042]; Bristol[0.042]; League[0.042]; Gloucester[0.041]; Saracens[0.041]; Leicester[0.041]; Swansea[0.041]; Bridgend[0.041]; Llanelli[0.040]; Caerphilly[0.040]; Treorchy[0.040]; Wasps[0.040]; Northampton[0.039]; National[0.039]; Newport[0.039]; Irish[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Welsh ==> ENTITY: \u001b[32mWales\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.149:0.149[0]; log p(e|m)= -1.079:-1.079[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.046]; rugby[0.046]; Welsh[0.045]; Welsh[0.045]; Cardiff[0.045]; matches[0.042]; English[0.042]; English[0.042]; English[0.042]; League[0.041]; league[0.041]; Scottish[0.041]; Scottish[0.041]; Scottish[0.041]; Bridgend[0.041]; Bath[0.041]; Llanelli[0.041]; Harlequins[0.040]; Bristol[0.040]; Swansea[0.040]; union[0.040]; Saracens[0.040]; Leicester[0.039]; Treorchy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Welsh ==> ENTITY: \u001b[32mWales\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.149:0.149[0]; log p(e|m)= -1.079:-1.079[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.046]; rugby[0.046]; Welsh[0.045]; Welsh[0.045]; Cardiff[0.045]; matches[0.042]; English[0.042]; English[0.042]; English[0.042]; League[0.041]; league[0.041]; Scottish[0.041]; Scottish[0.041]; Scottish[0.041]; Bridgend[0.041]; Bath[0.041]; Llanelli[0.041]; Harlequins[0.040]; Bristol[0.040]; Swansea[0.040]; union[0.040]; Saracens[0.040]; Leicester[0.039]; Treorchy[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNewport\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNewport\u001b[39m <---> \u001b[32mNewport RFC\u001b[39m\t\nSCORES: global= 0.256:0.250[\u001b[31m0.006\u001b[39m]; local()= 0.142:0.208[\u001b[32m0.066\u001b[39m]; log p(e|m)= -0.697:-3.381[\u001b[31m2.684\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCardiff[0.045]; rugby[0.045]; Welsh[0.044]; Welsh[0.044]; Welsh[0.044]; Rugby[0.044]; Bristol[0.043]; Swansea[0.042]; Northampton[0.042]; league[0.042]; Gloucester[0.042]; Bath[0.042]; League[0.041]; Caerphilly[0.041]; Bridgend[0.041]; matches[0.041]; Llanelli[0.040]; Harlequins[0.040]; Ebbw[0.040]; Saturday[0.040]; West[0.040]; Leicester[0.040]; Union[0.039]; Dunvant[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wasps ==> ENTITY: \u001b[32mLondon Wasps\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.230:0.230[0]; log p(e|m)= -0.070:-0.070[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.046]; rugby[0.045]; Leicester[0.044]; Northampton[0.044]; Harlequins[0.043]; Cardiff[0.043]; Saracens[0.043]; Gloucester[0.043]; Bristol[0.043]; league[0.042]; Bath[0.042]; League[0.041]; division[0.041]; division[0.041]; matches[0.041]; Welsh[0.041]; Welsh[0.041]; Welsh[0.041]; Swansea[0.040]; Llanelli[0.039]; London[0.039]; London[0.039]; Bridgend[0.039]; Caerphilly[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bristol ==> ENTITY: \u001b[32mBristol Rugby\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.236:0.236[0]; log p(e|m)= -3.772:-3.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.045]; rugby[0.045]; Cardiff[0.044]; Harlequins[0.044]; Leicester[0.044]; league[0.044]; Northampton[0.043]; Swansea[0.043]; Bath[0.043]; Saracens[0.043]; Wasps[0.042]; Gloucester[0.042]; League[0.041]; Welsh[0.041]; Welsh[0.041]; Welsh[0.041]; Newport[0.040]; matches[0.040]; Llanelli[0.040]; London[0.039]; London[0.039]; Saturday[0.039]; Irish[0.039]; County[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Melrose ==> ENTITY: \u001b[32mMelrose RFC\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.129:0.129[0]; log p(e|m)= -3.079:-3.079[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.046]; league[0.046]; League[0.044]; Bristol[0.043]; Stirling[0.043]; matches[0.042]; London[0.042]; London[0.042]; Saturday[0.042]; Welsh[0.042]; Welsh[0.042]; Scottish[0.042]; Scottish[0.042]; National[0.041]; Currie[0.041]; Hawick[0.041]; West[0.041]; Gloucester[0.041]; Heriot[0.040]; premier[0.040]; County[0.040]; Newport[0.040]; Boroughmuir[0.040]; Bath[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEbbw Vale\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEbbw Vale\u001b[39m <---> \u001b[32mEbbw Vale RFC\u001b[39m\t\nSCORES: global= 0.267:0.263[\u001b[31m0.005\u001b[39m]; local()= 0.160:0.217[\u001b[32m0.056\u001b[39m]; log p(e|m)= 0.000:-1.864[\u001b[31m1.864\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.044]; rugby[0.044]; Cardiff[0.043]; league[0.043]; Swansea[0.043]; Llanelli[0.042]; League[0.042]; matches[0.042]; Caerphilly[0.042]; Bridgend[0.042]; Gloucester[0.042]; Saracens[0.042]; Harlequins[0.042]; Bristol[0.042]; Welsh[0.042]; Welsh[0.042]; Welsh[0.042]; Northampton[0.041]; Newport[0.041]; Bath[0.041]; Leicester[0.040]; Wasps[0.040]; Newbridge[0.039]; Orrell[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.096:0.096[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.048]; league[0.045]; League[0.045]; Irish[0.044]; Bristol[0.044]; matches[0.044]; Northampton[0.042]; English[0.042]; English[0.042]; English[0.042]; Cardiff[0.042]; Bath[0.041]; Rugby[0.041]; Gloucester[0.041]; Newbridge[0.041]; Bridgend[0.041]; Leicester[0.040]; Saturday[0.040]; West[0.040]; Swansea[0.040]; division[0.039]; division[0.039]; National[0.039]; Orrell[0.039]; \t\n\r [==========================>...................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s290ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2869/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 953testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Volgograd ==> ENTITY: \u001b[32mFC Volgograd\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nVolgograd[0.059]; Togliatti[0.046]; league[0.044]; league[0.044]; Rotor[0.043]; Rotor[0.043]; Rotor[0.043]; Rotor[0.043]; Rotor[0.043]; club[0.043]; Russian[0.040]; Dynamo[0.040]; Anatoly[0.040]; games[0.040]; Soccer[0.040]; Moscow[0.040]; Moscow[0.040]; September[0.039]; committee[0.039]; Lada[0.038]; Uefa[0.038]; stands[0.038]; players[0.037]; game[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rotor ==> ENTITY: \u001b[32mFC Rotor Volgograd\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.182:0.182[0]; log p(e|m)= -1.440:-1.440[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; league[0.046]; players[0.045]; club[0.044]; club[0.044]; championship[0.043]; games[0.042]; Rotor[0.042]; Rotor[0.042]; Rotor[0.042]; Rotor[0.042]; Soccer[0.042]; closed[0.041]; doors[0.040]; Cup[0.040]; game[0.040]; game[0.040]; play[0.040]; play[0.040]; play[0.040]; fans[0.040]; Dynamo[0.039]; Dynamo[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dynamo Moscow ==> ENTITY: \u001b[32mFC Dynamo Moscow\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.222:0.222[0]; log p(e|m)= -0.453:-0.453[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; league[0.047]; club[0.047]; club[0.047]; championship[0.047]; Dynamo[0.046]; games[0.045]; Moscow[0.045]; Moscow[0.045]; Russian[0.045]; players[0.044]; United[0.043]; Cup[0.043]; game[0.042]; game[0.042]; Uefa[0.041]; fans[0.041]; play[0.041]; play[0.041]; Alania[0.041]; September[0.041]; Manchester[0.041]; Togliatti[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mLada Togliatti\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHC Lada Togliatti\u001b[39m <---> \u001b[32mFC Lada-Togliatti Togliatti\u001b[39m\t\nSCORES: global= 0.279:0.264[\u001b[31m0.015\u001b[39m]; local()= 0.214:0.183[\u001b[31m0.032\u001b[39m]; log p(e|m)= 0.000:-0.293[\u001b[31m0.293\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; league[0.047]; games[0.044]; championship[0.044]; Cup[0.044]; Dynamo[0.043]; Dynamo[0.043]; club[0.042]; club[0.042]; Russian[0.042]; players[0.041]; Moscow[0.041]; Moscow[0.041]; Moscow[0.041]; Moscow[0.041]; Soccer[0.040]; game[0.040]; game[0.040]; Uefa[0.040]; September[0.040]; placed[0.039]; Friday[0.039]; Anatoly[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rotor ==> ENTITY: \u001b[32mFC Rotor Volgograd\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.182:0.182[0]; log p(e|m)= -1.440:-1.440[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; league[0.046]; players[0.045]; club[0.044]; club[0.044]; championship[0.043]; games[0.042]; Rotor[0.042]; Rotor[0.042]; Rotor[0.042]; Rotor[0.042]; Soccer[0.042]; closed[0.041]; doors[0.040]; Cup[0.040]; game[0.040]; game[0.040]; play[0.040]; play[0.040]; play[0.040]; fans[0.040]; Dynamo[0.039]; Dynamo[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Volgograd ==> ENTITY: \u001b[32mFC Volgograd\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.180:0.180[0]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTogliatti[0.047]; league[0.045]; league[0.045]; Rotor[0.044]; Rotor[0.044]; Rotor[0.044]; club[0.044]; club[0.044]; Russian[0.041]; Dynamo[0.041]; Dynamo[0.041]; Anatoly[0.041]; games[0.041]; Moscow[0.041]; Moscow[0.041]; Moscow[0.041]; September[0.040]; Alania[0.040]; hand[0.040]; committee[0.040]; Lada[0.039]; Uefa[0.039]; stands[0.039]; players[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Manchester United\u001b[39m ==> ENTITY: \u001b[32mManchester United F.C.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.050]; league[0.050]; club[0.044]; club[0.044]; games[0.043]; Soccer[0.042]; players[0.042]; game[0.042]; game[0.042]; championship[0.041]; defeat[0.041]; defeat[0.041]; fans[0.041]; play[0.040]; play[0.040]; play[0.040]; Uefa[0.040]; points[0.040]; Cup[0.040]; Dynamo[0.040]; Dynamo[0.040]; home[0.039]; home[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rotor ==> ENTITY: \u001b[32mFC Rotor Volgograd\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.163:0.163[0]; log p(e|m)= -1.440:-1.440[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; league[0.047]; players[0.046]; club[0.045]; Rotor[0.043]; Rotor[0.043]; Rotor[0.043]; Soccer[0.043]; closed[0.042]; doors[0.041]; game[0.041]; play[0.041]; play[0.041]; Dynamo[0.040]; Saturday[0.040]; Lada[0.040]; Russian[0.040]; leaders[0.040]; Togliatti[0.040]; Moscow[0.040]; Moscow[0.040]; Cup[0.040]; disciplinary[0.040]; stands[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rotor Volgograd\u001b[39m ==> ENTITY: \u001b[32mFC Rotor Volgograd\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.047]; league[0.047]; players[0.046]; club[0.045]; games[0.043]; Soccer[0.043]; championship[0.042]; Rotor[0.042]; Rotor[0.042]; Rotor[0.042]; Rotor[0.042]; game[0.041]; play[0.041]; play[0.041]; play[0.041]; Dynamo[0.040]; Russian[0.040]; leaders[0.040]; Togliatti[0.040]; Moscow[0.040]; Moscow[0.040]; Cup[0.040]; disciplinary[0.039]; stands[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.118:0.118[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.047]; league[0.045]; league[0.045]; Russian[0.044]; club[0.043]; club[0.043]; games[0.043]; Dynamo[0.043]; home[0.042]; Cup[0.042]; Anatoly[0.041]; championship[0.041]; game[0.041]; Alania[0.041]; Uefa[0.040]; play[0.040]; play[0.040]; Friday[0.040]; Saturday[0.040]; stands[0.040]; ended[0.040]; Lada[0.039]; September[0.039]; placed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = UEFA Cup ==> ENTITY: \u001b[32mUEFA Europa League\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.192:0.192[0]; log p(e|m)= -0.261:-0.261[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.045]; league[0.045]; club[0.044]; club[0.044]; games[0.044]; Soccer[0.044]; game[0.042]; game[0.042]; Dynamo[0.041]; Dynamo[0.041]; players[0.041]; placed[0.041]; championship[0.041]; Manchester[0.041]; play[0.041]; play[0.041]; play[0.041]; Alania[0.040]; ended[0.040]; disciplinary[0.040]; spell[0.040]; fans[0.040]; points[0.040]; United[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.106:0.106[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; game[0.043]; game[0.043]; players[0.043]; Moscow[0.043]; Moscow[0.043]; Moscow[0.043]; Moscow[0.043]; club[0.042]; club[0.042]; league[0.042]; league[0.042]; Uefa[0.042]; Soccer[0.041]; play[0.041]; play[0.041]; play[0.041]; Dynamo[0.041]; Dynamo[0.041]; fans[0.040]; defeat[0.040]; defeat[0.040]; closed[0.040]; Manchester[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.106:0.106[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.047]; league[0.045]; league[0.045]; Russian[0.044]; fans[0.044]; club[0.043]; games[0.043]; Dynamo[0.042]; home[0.042]; home[0.042]; players[0.042]; Cup[0.041]; Soccer[0.041]; Anatoly[0.041]; championship[0.041]; game[0.041]; Uefa[0.040]; play[0.040]; play[0.040]; play[0.040]; Friday[0.040]; Saturday[0.040]; stands[0.039]; Fans[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rotor ==> ENTITY: \u001b[32mFC Rotor Volgograd\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.195:0.195[0]; log p(e|m)= -1.440:-1.440[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.051]; league[0.051]; players[0.049]; club[0.048]; club[0.048]; games[0.046]; Rotor[0.046]; Rotor[0.046]; Rotor[0.046]; championship[0.046]; game[0.044]; game[0.044]; play[0.044]; play[0.044]; play[0.044]; Dynamo[0.043]; Dynamo[0.043]; Lada[0.043]; Russian[0.043]; leaders[0.043]; leaders[0.043]; Togliatti[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dynamo Moscow ==> ENTITY: \u001b[32mFC Dynamo Moscow\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.212:0.212[0]; log p(e|m)= -0.453:-0.453[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.045]; league[0.045]; club[0.045]; championship[0.045]; Dynamo[0.045]; games[0.043]; Moscow[0.043]; Moscow[0.043]; Russian[0.043]; players[0.043]; Soccer[0.041]; United[0.041]; Cup[0.041]; game[0.040]; game[0.040]; Togliatti[0.040]; Uefa[0.040]; fans[0.040]; play[0.039]; play[0.039]; play[0.039]; Alania[0.039]; September[0.039]; Manchester[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Alania ==> ENTITY: \u001b[32mFC Alania Vladikavkaz\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.207:0.207[0]; log p(e|m)= -0.749:-0.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.045]; league[0.045]; club[0.045]; club[0.045]; Moscow[0.044]; Moscow[0.044]; Moscow[0.044]; games[0.043]; Russian[0.043]; Dynamo[0.043]; Dynamo[0.043]; players[0.042]; September[0.041]; said[0.040]; Uefa[0.040]; defeat[0.040]; defeat[0.040]; championship[0.039]; Lada[0.039]; United[0.039]; game[0.039]; game[0.039]; home[0.039]; home[0.039]; \t\n\r [==========================>...................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 9s246ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2885/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1057testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.106:0.106[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.051]; Canada[0.047]; Britain[0.046]; Zealand[0.046]; final[0.046]; final[0.046]; championship[0.046]; championship[0.046]; Italy[0.046]; Italy[0.046]; Lithuania[0.045]; record[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; Germany[0.045]; medal[0.045]; Women[0.045]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.196:0.196[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.047]; France[0.047]; France[0.047]; matches[0.043]; Italy[0.043]; Italy[0.043]; championship[0.042]; championship[0.042]; finals[0.042]; team[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Australia[0.041]; final[0.041]; final[0.041]; Canada[0.040]; semifinalist[0.040]; Zealand[0.040]; Britain[0.040]; record[0.039]; Philippe[0.039]; Philippe[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.046]; Latvia[0.045]; matches[0.045]; Lithuania[0.044]; finals[0.044]; finals[0.044]; Marty[0.043]; championship[0.043]; championship[0.043]; Pavel[0.042]; Czech[0.041]; Britain[0.041]; Ainars[0.041]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; Republic[0.040]; Italy[0.040]; Italy[0.040]; Svetlana[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lithuania ==> ENTITY: \u001b[32mLithuania\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.397:-0.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.045]; Russia[0.045]; Russia[0.045]; Italy[0.043]; Italy[0.043]; team[0.043]; time[0.043]; England[0.042]; Ukraine[0.042]; qualifying[0.042]; Track[0.042]; finals[0.041]; individual[0.041]; championships[0.041]; Britain[0.041]; medal[0.041]; Australia[0.040]; Women[0.040]; Women[0.040]; track[0.040]; Cycling[0.040]; Championship[0.040]; Germany[0.040]; Germany[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lithuania ==> ENTITY: \u001b[32mLithuania\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.129:0.129[0]; log p(e|m)= -0.397:-0.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nchampionship[0.044]; championship[0.044]; matches[0.044]; Latvia[0.043]; Russia[0.043]; Russia[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; team[0.043]; time[0.042]; World[0.041]; World[0.041]; finals[0.041]; Finland[0.041]; Canada[0.041]; Britain[0.040]; medal[0.040]; Ainars[0.040]; Women[0.040]; Germany[0.039]; Germany[0.039]; Germany[0.039]; losing[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.211:0.211[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.051]; Italy[0.051]; team[0.043]; matches[0.043]; Lithuania[0.041]; finals[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; losing[0.041]; Germany[0.041]; Germany[0.041]; final[0.041]; final[0.041]; Australia[0.041]; Australia[0.041]; Britain[0.040]; championship[0.040]; championship[0.040]; world[0.040]; world[0.040]; Zealand[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.146:0.146[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.049]; England[0.045]; team[0.045]; Italy[0.044]; Italy[0.044]; Championship[0.043]; qualifying[0.043]; Australia[0.042]; Britain[0.041]; championships[0.041]; round[0.041]; Women[0.041]; finals[0.040]; Philippe[0.040]; world[0.040]; world[0.040]; world[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; semifinals[0.040]; Cycling[0.040]; Michel[0.040]; cycling[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Anton Chantyr\u001b[39m ==> ENTITY: \u001b[32mAnton Shantyr\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.118:0.118[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNikolai[0.045]; Thorsten[0.043]; Heiko[0.043]; Natalia[0.043]; semifinals[0.042]; Olga[0.042]; Philippe[0.042]; semifinalist[0.042]; Lithuania[0.042]; Svetlana[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Tatian[0.041]; finals[0.041]; team[0.041]; Kathrin[0.041]; Cyril[0.041]; record[0.041]; record[0.041]; Trentino[0.041]; Rund[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Frederic Magne\u001b[39m ==> ENTITY: \u001b[32mFrédéric Magné\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nworld[0.045]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; Philippe[0.043]; sprint[0.043]; Darryn[0.042]; team[0.042]; Buran[0.042]; finals[0.041]; finals[0.041]; Michel[0.041]; Florian[0.041]; Pavel[0.041]; Cyril[0.041]; Australia[0.041]; Australia[0.041]; Trentino[0.041]; pursuit[0.040]; pursuit[0.040]; Tyler[0.040]; Ainars[0.040]; Arrue[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.046]; France[0.046]; France[0.046]; team[0.043]; matches[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; finals[0.042]; finals[0.042]; championship[0.041]; championship[0.041]; championship[0.041]; Lithuania[0.041]; Lithuania[0.041]; Australia[0.040]; Australia[0.040]; Republic[0.040]; final[0.040]; Latvia[0.040]; Britain[0.039]; Philippe[0.038]; world[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.092:0.092[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.051]; Australia[0.051]; Republic[0.050]; France[0.050]; France[0.050]; France[0.050]; Italy[0.048]; Italy[0.048]; championship[0.048]; Darryn[0.047]; world[0.047]; World[0.047]; Hill[0.047]; Lithuania[0.047]; Russia[0.046]; Russia[0.046]; Latvia[0.046]; Florian[0.046]; eliminated[0.046]; caught[0.045]; pursuit[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Danilo Hondo\u001b[39m ==> ENTITY: \u001b[32mDanilo Hondo\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.046:0.046[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThorsten[0.044]; Heiko[0.044]; semifinals[0.043]; semifinalist[0.043]; final[0.042]; Tatian[0.042]; Kathrin[0.042]; trial[0.042]; Trentino[0.042]; Rund[0.042]; Kouznetsov[0.042]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Anton[0.041]; fastest[0.041]; Tyler[0.041]; Wynd[0.041]; Freitag[0.041]; team[0.041]; Bos[0.040]; Australia[0.040]; Australia[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.047]; Russia[0.047]; Russia[0.047]; England[0.045]; Lithuania[0.045]; Ukraine[0.045]; medal[0.045]; championships[0.044]; finals[0.044]; losing[0.043]; round[0.043]; semifinalist[0.043]; qualifying[0.043]; record[0.042]; record[0.042]; Championship[0.042]; semifinals[0.042]; Germany[0.042]; Germany[0.042]; Britain[0.042]; Natalia[0.041]; Manchester[0.041]; Nikolai[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Olga Slioussareva\u001b[39m ==> ENTITY: \u001b[32mOlga Slyusareva\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmedal[0.051]; Russia[0.045]; Russia[0.045]; Russia[0.045]; Nikolai[0.045]; bronze[0.044]; Lithuania[0.044]; championship[0.044]; championship[0.044]; team[0.044]; Thorsten[0.044]; Heiko[0.043]; Tanya[0.043]; Anton[0.043]; Oksana[0.042]; semifinalist[0.042]; finals[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; sprint[0.042]; Kathrin[0.041]; Florian[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.138:0.138[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.049]; Russia[0.049]; Russia[0.049]; matches[0.047]; Lithuania[0.047]; Canada[0.047]; medal[0.046]; finals[0.046]; championship[0.045]; championship[0.045]; losing[0.045]; semifinalist[0.045]; Finland[0.045]; Germany[0.044]; Germany[0.044]; Germany[0.044]; Germany[0.044]; record[0.044]; final[0.043]; final[0.043]; Britain[0.043]; Nikolai[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.103:0.103[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.053]; finals[0.047]; Britain[0.045]; championship[0.045]; Italy[0.045]; Italy[0.045]; Lithuania[0.044]; Women[0.044]; Latvia[0.043]; France[0.043]; France[0.043]; France[0.043]; Russia[0.043]; Russia[0.043]; World[0.043]; Darryn[0.042]; world[0.042]; quarter[0.041]; eliminated[0.041]; Czech[0.041]; Republic[0.041]; Jane[0.041]; Marty[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Antonella Bellutti\u001b[39m ==> ENTITY: \u001b[32mAntonella Bellutti\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npursuit[0.046]; world[0.044]; World[0.044]; Natalia[0.044]; Italy[0.043]; Italy[0.043]; championship[0.043]; Buran[0.042]; finals[0.042]; Women[0.042]; Nothstein[0.042]; Pavel[0.041]; Australia[0.041]; Australia[0.041]; Russia[0.041]; Russia[0.041]; Tyler[0.041]; Hill[0.040]; Darryn[0.040]; Arrue[0.040]; France[0.040]; France[0.040]; quarter[0.040]; Britain[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andrea Collinelli\u001b[39m ==> ENTITY: \u001b[32mAndrea Collinelli\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.120:0.120[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npursuit[0.050]; pursuit[0.050]; Italy[0.047]; Italy[0.047]; Italy[0.047]; sprint[0.047]; individual[0.047]; Buran[0.045]; finals[0.045]; team[0.045]; Kathrin[0.045]; Florian[0.044]; Pavel[0.044]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; Czech[0.044]; time[0.044]; Trentino[0.044]; Canada[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.208:0.208[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.046]; France[0.046]; France[0.046]; France[0.046]; France[0.046]; team[0.042]; matches[0.042]; Italy[0.041]; Italy[0.041]; Italy[0.041]; finals[0.041]; finals[0.041]; championship[0.041]; championship[0.041]; championship[0.041]; Lithuania[0.040]; Germany[0.040]; Germany[0.040]; Australia[0.040]; Australia[0.040]; Republic[0.040]; final[0.040]; Latvia[0.039]; Canada[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Edouard Gritsoun\u001b[39m ==> ENTITY: \u001b[32mEduard Gritsun\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.129:0.129[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNikolai[0.047]; Olga[0.045]; Anton[0.044]; Natalia[0.044]; Svetlana[0.043]; Philippe[0.043]; Thorsten[0.043]; Heiko[0.042]; team[0.042]; semifinals[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; semifinalist[0.041]; Tatian[0.041]; finals[0.041]; Kathrin[0.040]; Jane[0.040]; Trentino[0.040]; Rund[0.040]; Kouznetsov[0.040]; Magali[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marty Nothstein\u001b[39m ==> ENTITY: \u001b[32mMarty Nothstein\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nchampionship[0.044]; championship[0.044]; championship[0.044]; sprint[0.043]; finals[0.043]; finals[0.043]; team[0.043]; Marion[0.043]; matches[0.042]; Buran[0.042]; final[0.042]; Australia[0.042]; Australia[0.042]; pursuit[0.041]; pursuit[0.041]; Florian[0.041]; Pavel[0.041]; Trentino[0.041]; Record[0.040]; Tyler[0.040]; Ainars[0.040]; Arrue[0.040]; eliminated[0.040]; world[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.054:0.054[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.053]; finals[0.053]; finals[0.053]; matches[0.052]; final[0.052]; championship[0.051]; championship[0.051]; championship[0.051]; Britain[0.050]; Latvia[0.050]; eliminated[0.049]; Record[0.049]; Russia[0.049]; Russia[0.049]; Italy[0.049]; Italy[0.049]; Italy[0.049]; World[0.048]; World[0.048]; World[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.111:0.111[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.048]; team[0.048]; championships[0.046]; finals[0.045]; round[0.044]; record[0.044]; record[0.044]; Britain[0.044]; Italy[0.043]; Italy[0.043]; Manchester[0.043]; qualifying[0.043]; Championship[0.043]; Lithuania[0.043]; Germany[0.043]; Women[0.042]; Track[0.042]; France[0.042]; France[0.042]; world[0.042]; world[0.042]; world[0.042]; semifinals[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.179:0.179[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.046]; Germany[0.046]; Germany[0.046]; Italy[0.043]; Italy[0.043]; France[0.043]; France[0.043]; France[0.043]; team[0.042]; team[0.042]; Lithuania[0.042]; Lithuania[0.042]; Australia[0.041]; Britain[0.041]; Britain[0.041]; final[0.041]; Canada[0.040]; beat[0.040]; beat[0.040]; semifinals[0.040]; Finland[0.040]; Zealand[0.039]; World[0.039]; World[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.047]; Russia[0.047]; Russia[0.047]; matches[0.046]; Lithuania[0.045]; Canada[0.045]; medal[0.045]; finals[0.044]; championship[0.043]; championship[0.043]; losing[0.043]; semifinalist[0.043]; Finland[0.043]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; record[0.042]; final[0.042]; final[0.042]; Britain[0.042]; Ainars[0.041]; Nikolai[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lithuania ==> ENTITY: \u001b[32mLithuania\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.120:0.120[0]; log p(e|m)= -0.397:-0.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.047]; Russia[0.047]; Italy[0.045]; Italy[0.045]; finals[0.043]; Republic[0.042]; Australia[0.042]; Britain[0.042]; Czech[0.041]; World[0.041]; Marion[0.041]; championship[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; quarter[0.040]; eliminated[0.040]; France[0.040]; France[0.040]; world[0.040]; Pavel[0.039]; Svetlana[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ukraine ==> ENTITY: \u001b[32mUkraine\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.140:0.140[0]; log p(e|m)= -0.354:-0.354[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.050]; Lithuania[0.046]; Russia[0.046]; Russia[0.046]; Russia[0.046]; championships[0.043]; England[0.041]; Australia[0.041]; qualifying[0.041]; Germany[0.040]; Germany[0.040]; medal[0.040]; cycling[0.040]; France[0.040]; France[0.040]; France[0.040]; Championship[0.040]; Svetlana[0.040]; Italy[0.040]; Italy[0.040]; finals[0.040]; Cycling[0.040]; semifinalist[0.039]; Britain[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Philippe Ermenault\u001b[39m ==> ENTITY: \u001b[32mPhilippe Ermenault\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.243:0.243[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npursuit[0.048]; pursuit[0.048]; team[0.047]; France[0.045]; France[0.045]; France[0.045]; France[0.045]; France[0.045]; France[0.045]; world[0.044]; individual[0.043]; Frederic[0.043]; sprint[0.042]; Canada[0.042]; Michel[0.042]; championship[0.042]; championship[0.042]; championship[0.042]; Magali[0.041]; Moreau[0.041]; Jean[0.041]; Buran[0.040]; Magne[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Francis Moreau\u001b[39m ==> ENTITY: \u001b[32mFrancis Moreau\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npursuit[0.045]; pursuit[0.045]; championship[0.043]; championship[0.043]; championship[0.043]; World[0.042]; World[0.042]; World[0.042]; team[0.042]; Buran[0.042]; finals[0.041]; finals[0.041]; Record[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; Kathrin[0.041]; world[0.041]; sprint[0.041]; Florian[0.041]; Pavel[0.041]; Trentino[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.191:0.191[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.047]; France[0.047]; France[0.047]; team[0.043]; Italy[0.043]; Italy[0.043]; finals[0.043]; Lithuania[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Australia[0.041]; Australia[0.041]; semifinals[0.041]; final[0.041]; Canada[0.040]; semifinalist[0.040]; Britain[0.040]; medal[0.040]; record[0.039]; record[0.039]; Natalia[0.039]; Philippe[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pavel Buran\u001b[39m ==> ENTITY: \u001b[32mPavel Buran\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.118:0.118[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCzech[0.047]; Republic[0.045]; sprint[0.045]; world[0.045]; Natalia[0.044]; individual[0.044]; team[0.044]; finals[0.044]; finals[0.044]; Florian[0.043]; Nothstein[0.043]; Ainars[0.043]; Trentino[0.043]; final[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Britain[0.043]; Tyler[0.042]; World[0.042]; World[0.042]; World[0.042]; Latvia[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marion Clignet\u001b[39m ==> ENTITY: \u001b[32mMarion Clignet\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.086:0.086[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBuran[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; finals[0.043]; finals[0.043]; Florian[0.042]; Pavel[0.042]; sprint[0.042]; Trentino[0.042]; championship[0.042]; championship[0.042]; Tyler[0.041]; Ainars[0.041]; Arrue[0.041]; matches[0.041]; quarter[0.041]; quarter[0.041]; Bos[0.040]; pursuit[0.040]; Magne[0.040]; Australia[0.040]; Australia[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Felicia Ballanger\u001b[39m ==> ENTITY: \u001b[32mFélicia Ballanger\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.128:0.128[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThorsten[0.044]; Heiko[0.043]; semifinals[0.043]; trial[0.042]; France[0.042]; France[0.042]; France[0.042]; France[0.042]; semifinalist[0.042]; Mauro[0.042]; Mauro[0.042]; Tatian[0.042]; medal[0.042]; world[0.042]; world[0.042]; Kathrin[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Trentino[0.041]; Trentino[0.041]; Rund[0.041]; Kouznetsov[0.041]; record[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.047]; Russia[0.047]; Russia[0.047]; England[0.045]; Lithuania[0.045]; Ukraine[0.045]; medal[0.045]; championships[0.044]; finals[0.044]; losing[0.043]; round[0.043]; semifinalist[0.043]; qualifying[0.043]; record[0.042]; record[0.042]; Championship[0.042]; semifinals[0.042]; Germany[0.042]; Germany[0.042]; Britain[0.042]; Natalia[0.041]; Manchester[0.041]; Nikolai[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tanya Dubnicoff\u001b[39m ==> ENTITY: \u001b[32mTanya Dubnicoff\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.100:0.100[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.049]; medal[0.048]; team[0.043]; bronze[0.043]; Thorsten[0.043]; Heiko[0.042]; World[0.042]; World[0.042]; championship[0.042]; championship[0.042]; Australia[0.041]; semifinalist[0.041]; sprint[0.041]; finals[0.041]; Kathrin[0.040]; world[0.040]; Women[0.040]; Florian[0.040]; Donna[0.040]; Trentino[0.040]; Rund[0.040]; time[0.040]; Kouznetsov[0.040]; Wendy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.047]; matches[0.046]; Australia[0.045]; medal[0.044]; championship[0.042]; championship[0.042]; Latvia[0.042]; finals[0.042]; Finland[0.042]; Women[0.042]; final[0.041]; final[0.041]; Britain[0.041]; France[0.040]; France[0.040]; France[0.040]; France[0.040]; France[0.040]; World[0.040]; World[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; semifinalist[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marion Clignet\u001b[39m ==> ENTITY: \u001b[32mMarion Clignet\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.137:0.137[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncycling[0.048]; track[0.047]; Cycling[0.045]; championships[0.045]; Track[0.043]; semifinals[0.042]; France[0.042]; France[0.042]; Tatian[0.042]; finals[0.042]; qualifying[0.041]; Trentino[0.041]; Kouznetsov[0.041]; record[0.041]; record[0.041]; Saturday[0.041]; team[0.040]; Anton[0.040]; Tyler[0.040]; Cristiano[0.040]; quarter[0.039]; Bos[0.039]; pursuit[0.039]; pursuit[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Florian Rousseau\u001b[39m ==> ENTITY: \u001b[32mFlorian Rousseau\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfinals[0.044]; finals[0.044]; championship[0.044]; championship[0.044]; championship[0.044]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; team[0.042]; world[0.041]; Australia[0.041]; Australia[0.041]; Buran[0.041]; sprint[0.041]; matches[0.041]; Kathrin[0.040]; Pavel[0.040]; Germany[0.040]; Trentino[0.040]; Natalia[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.176:0.176[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.046]; France[0.046]; France[0.046]; team[0.043]; matches[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; finals[0.042]; finals[0.042]; championship[0.041]; championship[0.041]; championship[0.041]; Lithuania[0.041]; Germany[0.041]; Australia[0.040]; Australia[0.040]; Republic[0.040]; final[0.040]; Latvia[0.040]; Canada[0.040]; Zealand[0.039]; Britain[0.039]; Britain[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.058:0.058[0]; log p(e|m)= -0.190:-0.190[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.049]; Manchester[0.048]; team[0.047]; Championship[0.046]; Britain[0.046]; championships[0.045]; qualifying[0.045]; finals[0.044]; Lithuania[0.044]; France[0.043]; Italy[0.043]; World[0.042]; World[0.042]; Saturday[0.042]; Russia[0.042]; Russia[0.042]; Women[0.042]; semifinals[0.042]; Track[0.041]; quarter[0.041]; record[0.041]; round[0.041]; cycling[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Philippe Ermenault\u001b[39m ==> ENTITY: \u001b[32mPhilippe Ermenault\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.217:0.217[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npursuit[0.047]; team[0.046]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; world[0.043]; world[0.043]; world[0.043]; Canada[0.041]; medal[0.041]; Edouard[0.041]; Michel[0.041]; Thorsten[0.041]; Heiko[0.040]; time[0.040]; Magali[0.040]; Moreau[0.040]; record[0.040]; record[0.040]; Jean[0.040]; semifinals[0.040]; semifinalist[0.039]; Tatian[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = USA ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.059:0.059[0]; log p(e|m)= -0.130:-0.130[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.046]; Australia[0.045]; Australia[0.045]; Lithuania[0.044]; finals[0.044]; finals[0.044]; championship[0.043]; championship[0.043]; Britain[0.042]; Italy[0.042]; Italy[0.042]; Women[0.041]; World[0.041]; World[0.041]; Latvia[0.041]; France[0.040]; France[0.040]; France[0.040]; Nothstein[0.040]; Roberto[0.040]; quarter[0.039]; quarter[0.039]; sprint[0.039]; Republic[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yvonne McGregor\u001b[39m ==> ENTITY: \u001b[32mYvonne McGregor\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.068:0.068[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJane[0.046]; Darryn[0.044]; Lucy[0.043]; Buran[0.043]; Britain[0.043]; finals[0.043]; finals[0.043]; Women[0.042]; Florian[0.042]; Pavel[0.042]; world[0.042]; France[0.041]; France[0.041]; France[0.041]; Tyler[0.041]; Ainars[0.041]; Arrue[0.041]; quarter[0.040]; quarter[0.040]; Sharman[0.040]; matches[0.040]; Australia[0.040]; Australia[0.040]; Magne[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yvonne McGregor\u001b[39m ==> ENTITY: \u001b[32mYvonne McGregor\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.092:0.092[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncycling[0.046]; Jane[0.045]; Thorsten[0.043]; Heiko[0.043]; Jean[0.042]; Lucy[0.042]; Cycling[0.042]; semifinals[0.042]; Britain[0.042]; Cyril[0.042]; Tatian[0.041]; finals[0.041]; Women[0.041]; team[0.041]; Trentino[0.041]; Rund[0.041]; world[0.041]; world[0.041]; world[0.041]; Kouznetsov[0.041]; France[0.040]; France[0.040]; Anton[0.040]; Saturday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.191:0.191[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.047]; France[0.047]; France[0.047]; team[0.043]; team[0.043]; Italy[0.042]; Italy[0.042]; Italy[0.042]; championship[0.042]; Lithuania[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Australia[0.041]; semifinals[0.041]; final[0.040]; final[0.040]; Canada[0.040]; semifinalist[0.040]; Zealand[0.040]; Britain[0.040]; medal[0.040]; record[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.183:0.183[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.046]; Germany[0.046]; Germany[0.046]; matches[0.046]; team[0.043]; Italy[0.043]; Italy[0.043]; France[0.042]; France[0.042]; France[0.042]; France[0.042]; France[0.042]; Australia[0.041]; Australia[0.041]; Britain[0.041]; Latvia[0.040]; Lithuania[0.040]; finals[0.040]; Canada[0.040]; championship[0.039]; championship[0.039]; Finland[0.039]; Zealand[0.039]; time[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Zealand ==> ENTITY: \u001b[32mNew Zealand\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.238:-0.238[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.047]; Australia[0.047]; team[0.045]; matches[0.043]; Canada[0.042]; final[0.042]; final[0.042]; Lithuania[0.042]; finals[0.042]; Britain[0.042]; World[0.041]; World[0.041]; world[0.041]; Italy[0.041]; Italy[0.041]; Darryn[0.040]; championship[0.040]; championship[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Latvia[0.040]; time[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.175:0.175[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.052]; team[0.044]; Lithuania[0.042]; finals[0.042]; France[0.042]; France[0.042]; France[0.042]; qualifying[0.042]; losing[0.042]; Germany[0.041]; Germany[0.041]; Germany[0.041]; final[0.041]; Australia[0.041]; Britain[0.041]; championships[0.041]; semifinals[0.040]; medal[0.040]; Trentino[0.040]; record[0.040]; record[0.040]; Andrea[0.040]; round[0.040]; semifinalist[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Latvia ==> ENTITY: \u001b[32mLatvia\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.161:0.161[0]; log p(e|m)= -0.351:-0.351[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.045]; Lithuania[0.044]; Canada[0.043]; matches[0.043]; finals[0.043]; finals[0.043]; Finland[0.042]; Women[0.042]; championship[0.042]; championship[0.042]; championship[0.042]; Russia[0.042]; Russia[0.042]; World[0.042]; World[0.042]; World[0.042]; Britain[0.040]; Britain[0.040]; Germany[0.040]; Ainars[0.040]; Czech[0.040]; Italy[0.040]; Italy[0.040]; Italy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Darryn Hill\u001b[39m ==> ENTITY: \u001b[32mDarryn Hill\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.083:0.083[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; Australia[0.048]; sprint[0.047]; pursuit[0.042]; pursuit[0.042]; championship[0.041]; championship[0.041]; championship[0.041]; Buran[0.041]; Britain[0.041]; Britain[0.041]; team[0.041]; finals[0.041]; finals[0.041]; world[0.041]; Kathrin[0.041]; Florian[0.040]; Pavel[0.040]; Record[0.040]; Trentino[0.040]; Zealand[0.040]; Tyler[0.040]; Wynd[0.040]; Freitag[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lucy Tyler-Sharman\u001b[39m ==> ENTITY: \u001b[32mLucy Tyler-Sharman\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.046]; championships[0.045]; cycling[0.044]; track[0.043]; qualifying[0.043]; pursuit[0.043]; pursuit[0.043]; Track[0.043]; Cycling[0.042]; Thorsten[0.042]; Britain[0.041]; world[0.041]; world[0.041]; world[0.041]; World[0.041]; World[0.041]; semifinals[0.041]; record[0.040]; record[0.040]; fastest[0.040]; Tatian[0.040]; Championship[0.040]; finals[0.040]; Trentino[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Francis Moreau\u001b[39m ==> ENTITY: \u001b[32mFrancis Moreau\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.132:0.132[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmedal[0.045]; pursuit[0.044]; record[0.044]; record[0.044]; Thorsten[0.043]; Heiko[0.043]; semifinals[0.042]; World[0.042]; team[0.042]; semifinalist[0.041]; Tatian[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; time[0.041]; Kathrin[0.041]; world[0.041]; world[0.041]; world[0.041]; Trentino[0.040]; Rund[0.040]; trial[0.040]; Kouznetsov[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lucy Tyler-Sharman\u001b[39m ==> ENTITY: \u001b[32mLucy Tyler-Sharman\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.081:0.081[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.048]; Australia[0.048]; pursuit[0.045]; championship[0.044]; Britain[0.043]; world[0.043]; World[0.043]; Darryn[0.042]; Buran[0.042]; finals[0.042]; finals[0.042]; Florian[0.041]; Pavel[0.041]; individual[0.041]; Sharman[0.040]; matches[0.040]; Ainars[0.040]; Arrue[0.040]; Nothstein[0.040]; quarter[0.040]; Magne[0.039]; Italy[0.039]; Italy[0.039]; Marty[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.047]; Latvia[0.046]; Lithuania[0.045]; Marty[0.044]; championship[0.043]; finals[0.043]; Pavel[0.043]; Czech[0.042]; Britain[0.041]; Ainars[0.041]; Natalia[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; beat[0.041]; Republic[0.041]; Italy[0.040]; Italy[0.040]; Svetlana[0.040]; eliminated[0.040]; World[0.040]; Tyler[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Antonella Bellutti\u001b[39m ==> ENTITY: \u001b[32mAntonella Bellutti\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncycling[0.045]; Track[0.045]; championships[0.044]; Cycling[0.044]; pursuit[0.043]; pursuit[0.043]; world[0.042]; world[0.042]; world[0.042]; track[0.042]; World[0.042]; World[0.042]; Natalia[0.041]; Italy[0.041]; Italy[0.041]; semifinals[0.041]; record[0.040]; record[0.040]; team[0.040]; qualifying[0.040]; Tatian[0.040]; finals[0.040]; Women[0.040]; Trentino[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Annett Neumann\u001b[39m ==> ENTITY: \u001b[32mAnnett Neumann\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.121:0.121[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmedal[0.047]; Thorsten[0.043]; Heiko[0.043]; record[0.043]; championship[0.042]; semifinals[0.042]; bronze[0.042]; semifinalist[0.042]; Adler[0.041]; Adler[0.041]; fastest[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Kathrin[0.041]; World[0.041]; World[0.041]; pursuit[0.041]; pursuit[0.041]; Trentino[0.041]; Trentino[0.041]; Rund[0.041]; Kouznetsov[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.056:0.056[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nchampionships[0.045]; team[0.045]; record[0.044]; record[0.044]; finals[0.044]; qualifying[0.043]; medal[0.042]; Britain[0.042]; track[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Track[0.041]; Championship[0.041]; England[0.041]; semifinals[0.041]; semifinalist[0.041]; losing[0.040]; World[0.040]; World[0.040]; beat[0.040]; beat[0.040]; round[0.040]; Australia[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Finland ==> ENTITY: \u001b[32mFinland\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.161:0.161[0]; log p(e|m)= -0.245:-0.245[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; team[0.045]; Latvia[0.044]; Lithuania[0.043]; Canada[0.043]; Germany[0.042]; Germany[0.042]; Germany[0.042]; Germany[0.042]; finals[0.042]; championship[0.041]; championship[0.041]; final[0.041]; final[0.041]; medal[0.041]; world[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Britain[0.041]; Russia[0.040]; Russia[0.040]; losing[0.040]; Ainars[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wendy Everson\u001b[39m ==> ENTITY: \u001b[32mWendy Everson\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmedal[0.046]; Australia[0.046]; Australia[0.046]; sprint[0.045]; Thorsten[0.044]; Heiko[0.044]; world[0.044]; final[0.044]; final[0.044]; World[0.043]; World[0.043]; matches[0.043]; time[0.043]; Women[0.043]; trial[0.043]; semifinalist[0.043]; fastest[0.043]; finals[0.043]; team[0.042]; Darryn[0.042]; Kathrin[0.042]; Britain[0.042]; Florian[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andrea Collinelli\u001b[39m ==> ENTITY: \u001b[32mAndrea Collinelli\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrack[0.046]; cycling[0.045]; pursuit[0.044]; pursuit[0.044]; medal[0.043]; championships[0.043]; Italy[0.042]; Italy[0.042]; Thorsten[0.042]; Heiko[0.042]; individual[0.042]; semifinals[0.041]; Marion[0.041]; semifinalist[0.041]; record[0.041]; record[0.041]; Tatian[0.040]; finals[0.040]; qualifying[0.040]; team[0.040]; France[0.040]; France[0.040]; France[0.040]; France[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.126:0.126[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.056]; matches[0.051]; team[0.050]; finals[0.049]; finals[0.049]; Britain[0.047]; Britain[0.047]; Zealand[0.047]; final[0.047]; championship[0.047]; championship[0.047]; championship[0.047]; Italy[0.047]; Italy[0.047]; Italy[0.047]; Italy[0.047]; Lithuania[0.046]; Germany[0.046]; Darryn[0.046]; world[0.046]; Latvia[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.064:0.064[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.046]; finals[0.046]; finals[0.046]; matches[0.045]; final[0.045]; championship[0.044]; championship[0.044]; championship[0.044]; Britain[0.043]; Britain[0.043]; Latvia[0.043]; eliminated[0.043]; Record[0.043]; Russia[0.043]; Russia[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Finland[0.042]; World[0.042]; World[0.042]; World[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.174:0.174[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.048]; Germany[0.048]; Germany[0.048]; Italy[0.045]; Italy[0.045]; France[0.045]; France[0.045]; France[0.045]; team[0.044]; Lithuania[0.044]; Australia[0.043]; Australia[0.043]; Britain[0.043]; Britain[0.043]; final[0.042]; Canada[0.042]; beat[0.042]; beat[0.042]; semifinals[0.042]; Finland[0.041]; Zealand[0.041]; World[0.041]; losing[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.054]; Lithuania[0.043]; finals[0.043]; France[0.043]; France[0.043]; Australia[0.042]; Australia[0.042]; Britain[0.042]; eliminated[0.042]; championship[0.042]; Republic[0.041]; Roberto[0.041]; quarter[0.040]; Marty[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; world[0.040]; Natalia[0.040]; World[0.040]; Antonella[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.125:0.125[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.053]; qualifying[0.048]; Australia[0.047]; track[0.047]; France[0.047]; France[0.047]; Championship[0.045]; Italy[0.045]; Italy[0.045]; Germany[0.045]; Germany[0.045]; Manchester[0.045]; world[0.044]; world[0.044]; world[0.044]; World[0.044]; World[0.044]; Lithuania[0.044]; cycling[0.044]; Saturday[0.044]; fastest[0.044]; team[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.171:0.171[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.046]; Germany[0.046]; Germany[0.046]; Italy[0.043]; Italy[0.043]; Italy[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; team[0.042]; Lithuania[0.042]; Australia[0.041]; Britain[0.041]; championship[0.041]; final[0.040]; final[0.040]; Canada[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; semifinals[0.040]; Finland[0.039]; Zealand[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Manchester ==> ENTITY: \u001b[32mManchester\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.085:0.085[0]; log p(e|m)= -0.322:-0.322[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.051]; Championship[0.049]; Britain[0.048]; championships[0.046]; team[0.046]; Marion[0.045]; Australia[0.045]; finals[0.044]; Saturday[0.044]; record[0.044]; World[0.043]; World[0.043]; round[0.043]; quarter[0.042]; Women[0.042]; fastest[0.042]; cycling[0.041]; Cycling[0.041]; qualifying[0.041]; track[0.041]; Track[0.040]; Jane[0.040]; individual[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.057]; England[0.049]; team[0.048]; Championship[0.047]; France[0.046]; France[0.046]; Australia[0.046]; qualifying[0.046]; finals[0.046]; Britain[0.045]; semifinals[0.044]; Trentino[0.044]; record[0.044]; record[0.044]; Manchester[0.044]; Andrea[0.044]; round[0.044]; championships[0.044]; Track[0.043]; cycling[0.043]; quarter[0.043]; Lithuania[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Czech Republic ==> ENTITY: \u001b[32mCzech Republic\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.124:0.124[0]; log p(e|m)= -0.278:-0.278[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; Italy[0.043]; Italy[0.043]; Italy[0.043]; finals[0.042]; finals[0.042]; Lithuania[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; Latvia[0.042]; World[0.041]; World[0.041]; final[0.041]; Pavel[0.041]; Marty[0.040]; championship[0.040]; championship[0.040]; championship[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.045]; Australia[0.043]; Australia[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; sprint[0.042]; final[0.041]; final[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; championship[0.041]; championship[0.041]; Darryn[0.041]; world[0.041]; World[0.041]; World[0.041]; Hill[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.223:0.223[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.051]; Italy[0.051]; team[0.043]; matches[0.043]; Lithuania[0.041]; finals[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; losing[0.041]; Germany[0.040]; Germany[0.040]; Germany[0.040]; Germany[0.040]; final[0.040]; final[0.040]; Australia[0.040]; Australia[0.040]; Britain[0.040]; championship[0.040]; championship[0.040]; world[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.161:0.161[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.047]; Russia[0.047]; Russia[0.047]; Russia[0.047]; Russia[0.047]; Lithuania[0.045]; Ukraine[0.045]; Canada[0.044]; medal[0.044]; finals[0.043]; losing[0.043]; round[0.043]; semifinalist[0.043]; qualifying[0.042]; record[0.042]; record[0.042]; semifinals[0.042]; Britain[0.042]; final[0.041]; Germany[0.041]; Germany[0.041]; Germany[0.041]; Nikolai[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Roberto Chiappa\u001b[39m ==> ENTITY: \u001b[32mRoberto Chiappa\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGiovanna[0.045]; team[0.043]; World[0.043]; World[0.043]; World[0.043]; Record[0.043]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Natalia[0.042]; world[0.042]; Trentino[0.042]; Andrea[0.042]; Buran[0.041]; finals[0.041]; finals[0.041]; championship[0.040]; championship[0.040]; championship[0.040]; Florian[0.040]; Pavel[0.040]; Mauro[0.040]; Jane[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.211:0.211[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.051]; Italy[0.051]; Italy[0.051]; team[0.043]; matches[0.043]; Lithuania[0.041]; Lithuania[0.041]; finals[0.041]; finals[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; final[0.040]; Australia[0.040]; Australia[0.040]; Britain[0.040]; Britain[0.040]; eliminated[0.039]; championship[0.039]; championship[0.039]; championship[0.039]; Republic[0.039]; Trentino[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.182:0.182[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.059]; France[0.059]; France[0.059]; matches[0.054]; Italy[0.054]; Italy[0.054]; championship[0.053]; championship[0.053]; finals[0.053]; finals[0.053]; Australia[0.052]; Australia[0.052]; Republic[0.051]; Britain[0.050]; Natalia[0.049]; Philippe[0.049]; world[0.049]; Russia[0.049]; Russia[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Guido Fulst\u001b[39m ==> ENTITY: \u001b[32mGuido Fulst\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.119:0.119[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThorsten[0.044]; Heiko[0.043]; record[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; Germany[0.043]; semifinals[0.043]; semifinalist[0.042]; Tatian[0.042]; Kathrin[0.041]; team[0.041]; Giovanna[0.041]; fastest[0.041]; Natalia[0.041]; Trentino[0.041]; Rund[0.041]; Kouznetsov[0.041]; Anton[0.041]; Philippe[0.041]; final[0.041]; Tyler[0.041]; Hondo[0.040]; Wynd[0.040]; \t\n\r [===========================>..................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s784ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2962/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 957testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Seoul ==> ENTITY: \u001b[32mSeoul\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.186:0.186[0]; log p(e|m)= -0.038:-0.038[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGames[0.045]; Hyundai[0.043]; Hyundai[0.043]; Hyundai[0.043]; Hyundai[0.043]; won[0.043]; Korean[0.042]; Korean[0.042]; games[0.042]; games[0.042]; games[0.042]; games[0.042]; Samsung[0.042]; Samsung[0.042]; Samsung[0.042]; Samsung[0.042]; played[0.041]; played[0.041]; played[0.041]; South[0.040]; Haitai[0.040]; Haitai[0.040]; Haitai[0.040]; Haitai[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = South Korean\u001b[39m ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbaseball[0.049]; Korean[0.048]; Baseball[0.048]; Hyundai[0.047]; Hyundai[0.047]; Hyundai[0.047]; Hyundai[0.047]; games[0.046]; games[0.046]; games[0.046]; games[0.046]; Seoul[0.046]; Games[0.044]; Haitai[0.044]; Haitai[0.044]; Haitai[0.044]; Haitai[0.044]; Samsung[0.043]; Samsung[0.043]; Samsung[0.043]; Samsung[0.043]; place[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hanwha ==> ENTITY: \u001b[32mHanwha Eagles\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.205:0.205[0]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbaseball[0.046]; Baseball[0.044]; Haitai[0.043]; Haitai[0.043]; Haitai[0.043]; Haitai[0.043]; games[0.043]; games[0.043]; games[0.043]; games[0.043]; professional[0.041]; Seoul[0.041]; won[0.040]; Hanwha[0.040]; Samsung[0.040]; Samsung[0.040]; Samsung[0.040]; Samsung[0.040]; Korean[0.040]; Korean[0.040]; Lotte[0.040]; Lotte[0.040]; Lotte[0.040]; Lotte[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hanwha ==> ENTITY: \u001b[32mHanwha Eagles\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.205:0.205[0]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbaseball[0.046]; Baseball[0.044]; Haitai[0.043]; Haitai[0.043]; Haitai[0.043]; Haitai[0.043]; games[0.043]; games[0.043]; games[0.043]; games[0.043]; professional[0.041]; Seoul[0.041]; won[0.040]; Hanwha[0.040]; Samsung[0.040]; Samsung[0.040]; Samsung[0.040]; Samsung[0.040]; Korean[0.040]; Korean[0.040]; Lotte[0.040]; Lotte[0.040]; Lotte[0.040]; Lotte[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = S. KOREAN\u001b[39m ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSeoul[0.047]; Korean[0.046]; Games[0.045]; Hyundai[0.044]; Hyundai[0.044]; Hyundai[0.044]; Hyundai[0.044]; Samsung[0.041]; Samsung[0.041]; Samsung[0.041]; Samsung[0.041]; games[0.041]; games[0.041]; games[0.041]; games[0.041]; played[0.040]; played[0.040]; played[0.040]; Haitai[0.040]; Haitai[0.040]; Haitai[0.040]; Haitai[0.040]; professional[0.039]; baseball[0.039]; \t\n\r [===========================>..................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s789ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2967/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1139testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Beijing ==> ENTITY: \u001b[32mBeijing\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.154:0.154[0]; log p(e|m)= -0.042:-0.042[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChinese[0.050]; Chinese[0.050]; China[0.047]; Hangzhou[0.044]; Donghai[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; held[0.042]; euphemism[0.042]; Friday[0.041]; capital[0.040]; dissident[0.040]; dissident[0.040]; class[0.040]; meet[0.040]; Zhejiang[0.040]; Police[0.039]; Police[0.039]; Police[0.039]; gave[0.039]; hold[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hangzhou\u001b[39m ==> ENTITY: \u001b[32mHangzhou\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHangzhou[0.048]; China[0.048]; Zhejiang[0.045]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; class[0.041]; class[0.041]; reached[0.041]; Friday[0.041]; month[0.041]; province[0.040]; returning[0.040]; Saturday[0.040]; maximum[0.039]; demanded[0.039]; days[0.039]; eastern[0.039]; capital[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.135:0.135[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChinese[0.048]; Beijing[0.046]; military[0.045]; Hangzhou[0.043]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; month[0.042]; Revolution[0.042]; Revolution[0.042]; June[0.041]; Democracy[0.041]; Tiananmen[0.040]; demonstrations[0.040]; authorities[0.040]; loss[0.040]; serve[0.039]; family[0.039]; family[0.039]; administrative[0.039]; dissident[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hangzhou\u001b[39m ==> ENTITY: \u001b[32mHangzhou\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHangzhou[0.047]; China[0.047]; Chinese[0.046]; Chinese[0.046]; Beijing[0.045]; Donghai[0.045]; Zhejiang[0.044]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; class[0.040]; class[0.040]; reached[0.040]; Friday[0.040]; province[0.040]; Saturday[0.039]; eastern[0.039]; capital[0.038]; family[0.038]; family[0.038]; family[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Democracy Wall\u001b[39m ==> ENTITY: \u001b[32mDemocracy Wall\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBeijing[0.047]; Tiananmen[0.045]; China[0.044]; Chinese[0.044]; Hangzhou[0.044]; month[0.043]; Revolution[0.043]; Revolution[0.043]; Square[0.043]; protests[0.041]; June[0.041]; dissidents[0.041]; Wang[0.041]; Wang[0.041]; Wang[0.041]; Wang[0.041]; Wang[0.041]; Wang[0.041]; democracy[0.040]; movement[0.040]; sent[0.040]; days[0.040]; dissident[0.040]; demonstrations[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chinese ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.165:0.165[0]; log p(e|m)= -1.419:-1.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChinese[0.051]; China[0.049]; Beijing[0.045]; Hangzhou[0.043]; province[0.043]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Donghai[0.042]; eastern[0.042]; family[0.041]; family[0.041]; group[0.040]; group[0.040]; Zhejiang[0.040]; said[0.039]; said[0.039]; class[0.039]; say[0.039]; capital[0.039]; dissident[0.039]; dissident[0.039]; euphemism[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York-based ==> ENTITY: \u001b[32mNew York\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.017:-0.017[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngroup[0.044]; group[0.044]; group[0.044]; Revolution[0.044]; based[0.044]; Human[0.043]; Friday[0.042]; Saturday[0.042]; Cultural[0.041]; class[0.041]; class[0.041]; said[0.041]; said[0.041]; said[0.041]; held[0.041]; family[0.041]; family[0.041]; province[0.041]; sent[0.041]; sent[0.041]; gave[0.040]; ideological[0.040]; eastern[0.040]; hold[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Beijing ==> ENTITY: \u001b[32mBeijing\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.042:-0.042[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.047]; Chinese[0.046]; Tiananmen[0.043]; military[0.043]; Square[0.043]; month[0.043]; Revolution[0.043]; Wang[0.042]; Wang[0.042]; Wang[0.042]; serve[0.042]; Democracy[0.042]; dissidents[0.041]; street[0.041]; appeared[0.041]; dissident[0.041]; June[0.041]; crushed[0.040]; demonstrations[0.040]; loss[0.040]; returning[0.040]; movement[0.040]; pro[0.040]; ordered[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zhejiang\u001b[39m ==> ENTITY: \u001b[32mZhejiang\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.236:0.236[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChinese[0.045]; Chinese[0.045]; province[0.045]; Hangzhou[0.044]; Hangzhou[0.044]; China[0.044]; eastern[0.043]; Beijing[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Donghai[0.042]; capital[0.039]; Saturday[0.039]; family[0.039]; family[0.039]; family[0.039]; coercive[0.039]; Revolution[0.038]; held[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chinese ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.170:0.170[0]; log p(e|m)= -1.419:-1.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChinese[0.050]; China[0.049]; Beijing[0.045]; Hangzhou[0.043]; province[0.043]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Donghai[0.042]; eastern[0.042]; family[0.041]; family[0.041]; group[0.040]; group[0.040]; group[0.040]; Zhejiang[0.039]; said[0.039]; said[0.039]; said[0.039]; class[0.039]; say[0.039]; say[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mTiananmen Square\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mTiananmen Square protests of 1989\u001b[39m <---> \u001b[32mTiananmen Square\u001b[39m\t\nSCORES: global= 0.278:0.271[\u001b[31m0.007\u001b[39m]; local()= 0.187:0.148[\u001b[31m0.038\u001b[39m]; log p(e|m)= 0.000:-0.018[\u001b[31m0.018\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.047]; Chinese[0.047]; Beijing[0.047]; protests[0.044]; demonstrations[0.043]; Revolution[0.042]; dissidents[0.042]; Wall[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Democracy[0.042]; month[0.042]; movement[0.041]; dissident[0.040]; June[0.040]; said[0.040]; said[0.040]; students[0.040]; street[0.040]; democracy[0.039]; authorities[0.039]; jailed[0.039]; imposed[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mHuman Rights in China\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHuman rights in China\u001b[39m <---> \u001b[32mHuman Rights in China (organization)\u001b[39m\t\nSCORES: global= 0.296:0.287[\u001b[31m0.009\u001b[39m]; local()= 0.172:0.168[\u001b[31m0.004\u001b[39m]; log p(e|m)= 0.000:-0.068[\u001b[31m0.068\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBeijing[0.046]; Chinese[0.046]; Chinese[0.046]; York[0.043]; dissident[0.042]; dissident[0.042]; group[0.042]; group[0.042]; group[0.042]; group[0.042]; Wang[0.041]; Wang[0.041]; Wang[0.041]; Wang[0.041]; Wang[0.041]; Wang[0.041]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; Donghai[0.040]; held[0.040]; sent[0.040]; sent[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chinese ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.119:0.119[0]; log p(e|m)= -1.419:-1.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.050]; Beijing[0.046]; military[0.046]; Wang[0.043]; Wang[0.043]; protests[0.042]; Square[0.042]; Democracy[0.042]; Revolution[0.042]; month[0.042]; June[0.041]; dissidents[0.041]; Western[0.041]; demonstrations[0.041]; Tiananmen[0.041]; said[0.040]; poor[0.040]; serve[0.040]; Cultural[0.040]; administrative[0.040]; administrative[0.040]; movement[0.040]; dissident[0.039]; authorities[0.039]; \t\n\r [===========================>..................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s732ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 2980/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 973testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sri Lanka ==> ENTITY: \u001b[32mSri Lanka national cricket team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.200:0.200[0]; log p(e|m)= -2.263:-2.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.045]; Kaluwitharana[0.044]; Kaluwitharana[0.044]; Australia[0.043]; Australia[0.043]; Australia[0.043]; Colombo[0.043]; Lanka[0.043]; Muralitharan[0.043]; Muralitharan[0.043]; Cricket[0.043]; match[0.042]; wickets[0.041]; Jayasuriya[0.041]; Jayasuriya[0.041]; Ponting[0.041]; World[0.040]; Vaas[0.040]; McGrath[0.040]; Dharmasena[0.040]; Chandana[0.040]; Chandana[0.040]; Chandana[0.040]; Chandana[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Vaas\u001b[39m ==> ENTITY: \u001b[32mChaminda Vaas\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKaluwitharana[0.045]; Mahanama[0.043]; overs[0.043]; Muralitharan[0.042]; Muralitharan[0.042]; bat[0.042]; Aravinda[0.042]; India[0.042]; Jayasuriya[0.042]; Jayasuriya[0.042]; McGrath[0.042]; McGrath[0.042]; wickets[0.042]; wickets[0.042]; Chandana[0.041]; Chandana[0.041]; Dharmasena[0.041]; Waugh[0.041]; Waugh[0.041]; Waugh[0.041]; Lanka[0.041]; lbw[0.041]; lbw[0.041]; Gillespie[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dharmasena\u001b[39m ==> ENTITY: \u001b[32mKumar Dharmasena\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRanatunga[0.044]; Ranatunga[0.044]; Muralitharan[0.043]; Muralitharan[0.043]; Muralitharan[0.043]; Mahanama[0.043]; Dharmasena[0.043]; Lanka[0.041]; Jayasuriya[0.041]; Jayasuriya[0.041]; Jayasuriya[0.041]; Kaluwitharana[0.041]; Kaluwitharana[0.041]; Kaluwitharana[0.041]; Bowling[0.041]; Waugh[0.041]; Waugh[0.041]; Sri[0.041]; Chandana[0.041]; Chandana[0.041]; Chandana[0.041]; Chandana[0.041]; Chandana[0.041]; Chandana[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Muralitharan\u001b[39m ==> ENTITY: \u001b[32mMuttiah Muralitharan\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.229:0.229[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKaluwitharana[0.046]; Muralitharan[0.045]; Mahanama[0.044]; McGrath[0.043]; McGrath[0.043]; India[0.043]; overs[0.042]; Aravinda[0.042]; wickets[0.041]; wickets[0.041]; bat[0.041]; Jayasuriya[0.041]; Jayasuriya[0.041]; Chandana[0.041]; Chandana[0.041]; Lanka[0.041]; Waugh[0.041]; Waugh[0.041]; Waugh[0.041]; match[0.041]; Dharmasena[0.041]; lbw[0.040]; lbw[0.040]; Ranatunga[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kaluwitharana\u001b[39m ==> ENTITY: \u001b[32mRomesh Kaluwitharana\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKaluwitharana[0.047]; cricket[0.045]; match[0.044]; Muralitharan[0.044]; Muralitharan[0.044]; Muralitharan[0.044]; Australia[0.044]; Australia[0.044]; Australia[0.044]; Dharmasena[0.044]; Dharmasena[0.044]; Vaas[0.044]; Chandana[0.043]; Chandana[0.043]; Chandana[0.043]; Chandana[0.043]; Chandana[0.043]; Colombo[0.043]; overs[0.043]; Jayasuriya[0.042]; Jayasuriya[0.042]; Jayasuriya[0.042]; Jayasuriya[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dharmasena\u001b[39m ==> ENTITY: \u001b[32mKumar Dharmasena\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.051]; Cricket[0.049]; Ranatunga[0.047]; Colombo[0.047]; Muralitharan[0.046]; Muralitharan[0.046]; Dharmasena[0.046]; Australia[0.045]; Australia[0.045]; Australia[0.045]; match[0.045]; Lanka[0.045]; Lanka[0.045]; Jayasuriya[0.044]; Jayasuriya[0.044]; Kaluwitharana[0.044]; Kaluwitharana[0.044]; Bowling[0.044]; Waugh[0.044]; Waugh[0.044]; Sri[0.044]; Sri[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = McGrath ==> ENTITY: \u001b[32mGlenn McGrath\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.194:0.194[0]; log p(e|m)= -2.189:-2.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMcGrath[0.050]; Kaluwitharana[0.049]; Waugh[0.048]; Waugh[0.048]; Waugh[0.048]; Gillespie[0.048]; lbw[0.048]; lbw[0.048]; overs[0.048]; overs[0.048]; match[0.048]; bat[0.047]; wickets[0.047]; wickets[0.047]; wickets[0.047]; wickets[0.047]; Bevan[0.047]; Healy[0.047]; Muralitharan[0.046]; Muralitharan[0.046]; Vaas[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sri Lanka ==> ENTITY: \u001b[32mSri Lanka national cricket team\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.203:0.203[0]; log p(e|m)= -2.263:-2.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.049]; Kaluwitharana[0.047]; Kaluwitharana[0.047]; Australia[0.047]; Australia[0.047]; Australia[0.047]; Colombo[0.046]; Lanka[0.046]; Muralitharan[0.046]; Muralitharan[0.046]; Cricket[0.046]; match[0.046]; overs[0.046]; wickets[0.044]; wickets[0.044]; Jayasuriya[0.044]; Jayasuriya[0.044]; Ponting[0.044]; World[0.044]; Vaas[0.044]; McGrath[0.043]; Dharmasena[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Aravinda de Silva\u001b[39m ==> ENTITY: \u001b[32mAravinda de Silva\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMuralitharan[0.045]; India[0.044]; Mahanama[0.044]; McGrath[0.044]; McGrath[0.044]; Waugh[0.043]; Waugh[0.043]; bat[0.043]; match[0.043]; Bevan[0.042]; wickets[0.042]; wickets[0.042]; Dharmasena[0.042]; overs[0.042]; lbw[0.042]; Chandana[0.040]; Vaas[0.040]; Fleming[0.040]; Fleming[0.040]; Zimbabwe[0.040]; Gillespie[0.039]; Match[0.039]; lb[0.039]; Bowling[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = India ==> ENTITY: \u001b[32mIndia national cricket team\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.146:0.146[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWaugh[0.044]; Waugh[0.044]; match[0.044]; Mahanama[0.043]; overs[0.043]; wickets[0.043]; wickets[0.043]; bat[0.043]; Match[0.043]; Muralitharan[0.042]; Zimbabwe[0.042]; Gillespie[0.042]; Bevan[0.042]; Bowling[0.041]; Aravinda[0.041]; McGrath[0.041]; McGrath[0.041]; Fleming[0.041]; Dharmasena[0.040]; September[0.040]; Chandana[0.040]; Man[0.039]; Vaas[0.039]; Series[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sri Lanka ==> ENTITY: \u001b[32mSri Lanka national cricket team\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.181:0.181[0]; log p(e|m)= -2.263:-2.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKaluwitharana[0.048]; Kaluwitharana[0.048]; Kaluwitharana[0.048]; Muralitharan[0.047]; Muralitharan[0.047]; Muralitharan[0.047]; overs[0.046]; overs[0.046]; Mahanama[0.046]; wickets[0.045]; wickets[0.045]; wickets[0.045]; wickets[0.045]; Jayasuriya[0.045]; Jayasuriya[0.045]; Jayasuriya[0.045]; Vaas[0.044]; McGrath[0.044]; McGrath[0.044]; McGrath[0.044]; Dharmasena[0.044]; Dharmasena[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ranatunga\u001b[39m ==> ENTITY: \u001b[32mArjuna Ranatunga\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nColombo[0.047]; cricket[0.044]; Cricket[0.043]; Kaluwitharana[0.042]; Kaluwitharana[0.042]; Kaluwitharana[0.042]; Muralitharan[0.042]; Muralitharan[0.042]; Muralitharan[0.042]; Lanka[0.042]; Lanka[0.042]; Lanka[0.042]; Dharmasena[0.042]; Dharmasena[0.042]; match[0.042]; Australia[0.041]; Australia[0.041]; Australia[0.041]; wickets[0.040]; wickets[0.040]; Jayasuriya[0.040]; Jayasuriya[0.040]; Jayasuriya[0.040]; Jayasuriya[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = S. Waugh\u001b[39m ==> ENTITY: \u001b[32mSteve Waugh\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWaugh[0.046]; Waugh[0.046]; McGrath[0.046]; McGrath[0.046]; Gillespie[0.043]; Bevan[0.043]; Fleming[0.043]; Fleming[0.043]; Fleming[0.043]; Fleming[0.043]; match[0.042]; India[0.042]; Healy[0.042]; Kaluwitharana[0.041]; Mahanama[0.040]; bat[0.039]; wickets[0.039]; wickets[0.039]; Muralitharan[0.039]; Muralitharan[0.039]; overs[0.039]; Bowling[0.039]; Match[0.039]; Zimbabwe[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.190:0.190[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.049]; Australia[0.049]; cricket[0.047]; match[0.044]; Cricket[0.044]; Waugh[0.043]; Waugh[0.043]; Colombo[0.042]; Muralitharan[0.041]; Muralitharan[0.041]; McGrath[0.041]; overs[0.040]; Kaluwitharana[0.040]; Kaluwitharana[0.040]; Gillespie[0.040]; wickets[0.040]; wickets[0.040]; Bevan[0.040]; Ponting[0.040]; Ranatunga[0.040]; Slater[0.039]; Scoreboard[0.039]; Scoreboard[0.039]; Fleming[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Muralitharan\u001b[39m ==> ENTITY: \u001b[32mMuttiah Muralitharan\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.250:0.250[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKaluwitharana[0.054]; Kaluwitharana[0.054]; Kaluwitharana[0.054]; Muralitharan[0.053]; Muralitharan[0.053]; Muralitharan[0.053]; Mahanama[0.051]; McGrath[0.050]; McGrath[0.050]; overs[0.049]; overs[0.049]; wickets[0.048]; wickets[0.048]; wickets[0.048]; wickets[0.048]; bat[0.048]; Ponting[0.048]; Jayasuriya[0.048]; Jayasuriya[0.048]; Jayasuriya[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colombo ==> ENTITY: \u001b[32mColombo\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.192:0.192[0]; log p(e|m)= -0.194:-0.194[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.047]; Lanka[0.046]; Lanka[0.046]; Cricket[0.046]; match[0.044]; Chandana[0.044]; Chandana[0.044]; Chandana[0.044]; Chandana[0.044]; Australia[0.044]; Australia[0.044]; Australia[0.044]; Muralitharan[0.043]; Muralitharan[0.043]; Scoreboard[0.043]; Scoreboard[0.043]; Sri[0.043]; Sri[0.043]; World[0.041]; Ranatunga[0.041]; Jayasuriya[0.041]; Jayasuriya[0.041]; overs[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U. Chandana\u001b[39m ==> ENTITY: \u001b[32mU. Chandana\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChandana[0.048]; bat[0.046]; Muralitharan[0.046]; Muralitharan[0.046]; Kaluwitharana[0.045]; Jayasuriya[0.044]; Jayasuriya[0.044]; Vaas[0.044]; match[0.044]; Ranatunga[0.043]; overs[0.043]; overs[0.043]; Sri[0.043]; Lanka[0.043]; Dharmasena[0.042]; Dharmasena[0.042]; run[0.042]; Mahanama[0.042]; lb[0.042]; Bowling[0.042]; Bowling[0.042]; lbw[0.041]; lbw[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Muralitharan\u001b[39m ==> ENTITY: \u001b[32mMuttiah Muralitharan\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.242:0.242[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKaluwitharana[0.047]; Kaluwitharana[0.047]; Muralitharan[0.046]; Muralitharan[0.046]; cricket[0.045]; McGrath[0.044]; Australia[0.044]; Australia[0.044]; Australia[0.044]; Colombo[0.043]; overs[0.043]; Cricket[0.043]; wickets[0.042]; wickets[0.042]; Ponting[0.042]; Jayasuriya[0.042]; Jayasuriya[0.042]; Jayasuriya[0.042]; Chandana[0.042]; Chandana[0.042]; Chandana[0.042]; Chandana[0.042]; Chandana[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Zimbabwe ==> ENTITY: \u001b[32mZimbabwe national cricket team\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.153:0.153[0]; log p(e|m)= -2.273:-2.273[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.046]; India[0.045]; bat[0.043]; overs[0.043]; Muralitharan[0.043]; McGrath[0.043]; McGrath[0.043]; Match[0.042]; wickets[0.042]; wickets[0.042]; Gillespie[0.042]; Waugh[0.042]; Waugh[0.042]; Series[0.042]; Fleming[0.041]; Aravinda[0.041]; Bevan[0.041]; Bowling[0.041]; September[0.041]; Dharmasena[0.040]; Chandana[0.040]; Vaas[0.040]; Man[0.039]; Law[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jayasuriya\u001b[39m ==> ENTITY: \u001b[32mSanath Jayasuriya\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.209:0.209[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.046]; Kaluwitharana[0.046]; Kaluwitharana[0.046]; Colombo[0.045]; Muralitharan[0.044]; Muralitharan[0.044]; Cricket[0.044]; Australia[0.044]; Australia[0.044]; Australia[0.044]; Jayasuriya[0.043]; Chandana[0.043]; Chandana[0.043]; Chandana[0.043]; Chandana[0.043]; overs[0.043]; Lanka[0.043]; Lanka[0.043]; match[0.043]; McGrath[0.042]; Sri[0.042]; Sri[0.042]; Vaas[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Healy ==> ENTITY: \u001b[32mIan Healy\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.149:0.149[0]; log p(e|m)= -3.650:-3.650[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHealy[0.044]; McGrath[0.043]; McGrath[0.043]; McGrath[0.043]; Waugh[0.043]; Waugh[0.043]; overs[0.043]; overs[0.043]; bat[0.042]; Gillespie[0.042]; Fleming[0.041]; Fleming[0.041]; Fleming[0.041]; Fleming[0.041]; Fleming[0.041]; run[0.041]; Kaluwitharana[0.041]; Kaluwitharana[0.041]; Kaluwitharana[0.041]; wickets[0.040]; wickets[0.040]; wickets[0.040]; wickets[0.040]; Mahanama[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Muralitharan\u001b[39m ==> ENTITY: \u001b[32mMuttiah Muralitharan\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.246:0.246[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKaluwitharana[0.054]; Kaluwitharana[0.054]; Kaluwitharana[0.054]; Muralitharan[0.053]; Muralitharan[0.053]; cricket[0.051]; McGrath[0.050]; Australia[0.050]; Australia[0.050]; Australia[0.050]; Colombo[0.049]; overs[0.049]; Cricket[0.049]; wickets[0.048]; wickets[0.048]; Ponting[0.048]; Jayasuriya[0.048]; Jayasuriya[0.048]; Jayasuriya[0.048]; Jayasuriya[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = McGrath ==> ENTITY: \u001b[32mGlenn McGrath\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.179:0.179[0]; log p(e|m)= -2.189:-2.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMcGrath[0.044]; Kaluwitharana[0.044]; Waugh[0.043]; Waugh[0.043]; Waugh[0.043]; Gillespie[0.043]; lbw[0.042]; lbw[0.042]; overs[0.042]; match[0.042]; bat[0.042]; wickets[0.042]; wickets[0.042]; Bevan[0.042]; Healy[0.042]; Muralitharan[0.041]; Vaas[0.041]; Fleming[0.040]; Fleming[0.040]; Fleming[0.040]; Fleming[0.040]; India[0.040]; Jayasuriya[0.040]; run[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gillespie ==> ENTITY: \u001b[32mJason Gillespie\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.170:0.170[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbat[0.044]; Kaluwitharana[0.043]; McGrath[0.043]; McGrath[0.043]; Waugh[0.042]; Waugh[0.042]; Waugh[0.042]; lbw[0.042]; lbw[0.042]; wickets[0.042]; wickets[0.042]; overs[0.042]; match[0.042]; Fleming[0.041]; Fleming[0.041]; Fleming[0.041]; Fleming[0.041]; Muralitharan[0.041]; Match[0.041]; Bevan[0.041]; Vaas[0.040]; Zimbabwe[0.040]; India[0.040]; September[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jayasuriya\u001b[39m ==> ENTITY: \u001b[32mSanath Jayasuriya\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.216:0.216[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKaluwitharana[0.044]; Kaluwitharana[0.044]; Kaluwitharana[0.044]; Muralitharan[0.043]; Muralitharan[0.043]; Muralitharan[0.043]; Muralitharan[0.043]; Jayasuriya[0.042]; Jayasuriya[0.042]; Chandana[0.041]; Chandana[0.041]; Chandana[0.041]; Chandana[0.041]; Chandana[0.041]; Chandana[0.041]; overs[0.041]; overs[0.041]; Lanka[0.041]; Mahanama[0.041]; McGrath[0.041]; McGrath[0.041]; Sri[0.040]; Vaas[0.040]; bat[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Vaas\u001b[39m ==> ENTITY: \u001b[32mChaminda Vaas\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKaluwitharana[0.047]; Kaluwitharana[0.047]; cricket[0.045]; overs[0.044]; Muralitharan[0.044]; Muralitharan[0.044]; Muralitharan[0.044]; Colombo[0.043]; Jayasuriya[0.043]; Jayasuriya[0.043]; McGrath[0.043]; wickets[0.043]; wickets[0.043]; Cricket[0.043]; Chandana[0.043]; Chandana[0.043]; Chandana[0.043]; Chandana[0.043]; Dharmasena[0.043]; Dharmasena[0.043]; Australia[0.042]; Australia[0.042]; Australia[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dharmasena\u001b[39m ==> ENTITY: \u001b[32mKumar Dharmasena\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRanatunga[0.044]; Muralitharan[0.044]; Muralitharan[0.044]; Mahanama[0.044]; India[0.043]; match[0.042]; Lanka[0.042]; Jayasuriya[0.042]; Jayasuriya[0.042]; Kaluwitharana[0.042]; Aravinda[0.042]; Bowling[0.042]; Waugh[0.042]; Waugh[0.042]; Waugh[0.042]; Sri[0.042]; Chandana[0.041]; Chandana[0.041]; September[0.040]; bat[0.040]; Match[0.040]; lb[0.039]; Zimbabwe[0.039]; Vaas[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.191:0.191[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.049]; Australia[0.049]; cricket[0.047]; match[0.044]; Cricket[0.044]; Waugh[0.043]; Waugh[0.043]; Colombo[0.042]; Muralitharan[0.041]; Muralitharan[0.041]; McGrath[0.041]; Kaluwitharana[0.041]; Kaluwitharana[0.041]; Gillespie[0.040]; Bevan[0.040]; Ponting[0.040]; Ranatunga[0.040]; Slater[0.040]; Scoreboard[0.039]; Scoreboard[0.039]; Fleming[0.039]; World[0.039]; Lanka[0.039]; Lanka[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Singer World Series\u001b[39m ==> ENTITY: \u001b[32mSinger World Series\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.224:0.224[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.044]; Australia[0.044]; Australia[0.044]; cricket[0.043]; Colombo[0.043]; Cricket[0.043]; Muralitharan[0.042]; Muralitharan[0.042]; Gillespie[0.042]; McGrath[0.042]; Lanka[0.042]; Lanka[0.042]; Waugh[0.042]; Waugh[0.042]; Bevan[0.041]; Kaluwitharana[0.041]; Kaluwitharana[0.041]; Chandana[0.040]; Chandana[0.040]; Chandana[0.040]; Chandana[0.040]; Ponting[0.040]; Fleming[0.040]; match[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jayasuriya\u001b[39m ==> ENTITY: \u001b[32mSanath Jayasuriya\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.222:0.222[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.044]; Kaluwitharana[0.043]; Kaluwitharana[0.043]; Kaluwitharana[0.043]; Colombo[0.042]; Muralitharan[0.042]; Muralitharan[0.042]; Muralitharan[0.042]; Cricket[0.042]; Australia[0.042]; Australia[0.042]; Australia[0.042]; Jayasuriya[0.041]; Jayasuriya[0.041]; Jayasuriya[0.041]; Chandana[0.041]; Chandana[0.041]; Chandana[0.041]; Chandana[0.041]; Chandana[0.041]; overs[0.041]; Lanka[0.041]; Lanka[0.041]; Lanka[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = S. Waugh\u001b[39m ==> ENTITY: \u001b[32mSteve Waugh\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWaugh[0.046]; Waugh[0.046]; McGrath[0.046]; McGrath[0.046]; McGrath[0.046]; Gillespie[0.043]; Gillespie[0.043]; Fleming[0.042]; Fleming[0.042]; Fleming[0.042]; Fleming[0.042]; Fleming[0.042]; Healy[0.041]; Kaluwitharana[0.040]; Kaluwitharana[0.040]; Mahanama[0.040]; bat[0.039]; wickets[0.039]; wickets[0.039]; wickets[0.039]; wickets[0.039]; Muralitharan[0.039]; Muralitharan[0.039]; Muralitharan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kaluwitharana\u001b[39m ==> ENTITY: \u001b[32mRomesh Kaluwitharana\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKaluwitharana[0.045]; Kaluwitharana[0.045]; cricket[0.043]; match[0.042]; Muralitharan[0.042]; Muralitharan[0.042]; Muralitharan[0.042]; Australia[0.042]; Australia[0.042]; Australia[0.042]; Dharmasena[0.042]; Dharmasena[0.042]; Vaas[0.042]; Chandana[0.041]; Chandana[0.041]; Chandana[0.041]; Chandana[0.041]; Chandana[0.041]; Colombo[0.041]; overs[0.041]; Jayasuriya[0.040]; Jayasuriya[0.040]; Jayasuriya[0.040]; Jayasuriya[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = S. Waugh\u001b[39m ==> ENTITY: \u001b[32mSteve Waugh\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWaugh[0.048]; Australia[0.048]; Australia[0.048]; Australia[0.048]; McGrath[0.048]; cricket[0.046]; Gillespie[0.045]; Bevan[0.045]; Cricket[0.045]; Fleming[0.044]; match[0.044]; Healy[0.043]; Kaluwitharana[0.042]; Kaluwitharana[0.042]; Ponting[0.042]; wickets[0.041]; wickets[0.041]; Slater[0.041]; Muralitharan[0.040]; Muralitharan[0.040]; Muralitharan[0.040]; overs[0.040]; Bowling[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lehmann ==> ENTITY: \u001b[32mDarren Lehmann\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.130:0.130[0]; log p(e|m)= -2.813:-2.813[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.045]; Waugh[0.043]; Waugh[0.043]; Bevan[0.043]; India[0.043]; McGrath[0.042]; McGrath[0.042]; wickets[0.042]; wickets[0.042]; bat[0.042]; overs[0.042]; Match[0.042]; September[0.042]; Gillespie[0.042]; Fleming[0.041]; Fleming[0.041]; Fleming[0.041]; Muralitharan[0.041]; Law[0.040]; Vaas[0.040]; Mahanama[0.040]; Bowling[0.040]; Ranatunga[0.039]; Aravinda[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bevan ==> ENTITY: \u001b[32mMichael Bevan\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.137:0.137[0]; log p(e|m)= -2.976:-2.976[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.046]; bat[0.045]; overs[0.044]; September[0.043]; McGrath[0.043]; McGrath[0.043]; Waugh[0.043]; Waugh[0.043]; Match[0.042]; Muralitharan[0.042]; wickets[0.042]; wickets[0.042]; India[0.041]; Mahanama[0.041]; Gillespie[0.041]; Fleming[0.041]; Fleming[0.041]; Fleming[0.041]; lbw[0.040]; lbw[0.040]; Bowling[0.040]; Vaas[0.039]; Aravinda[0.039]; Ranatunga[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.190:0.190[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.049]; Australia[0.049]; cricket[0.047]; match[0.044]; Cricket[0.044]; Waugh[0.043]; Waugh[0.043]; Colombo[0.042]; Muralitharan[0.041]; Muralitharan[0.041]; McGrath[0.041]; overs[0.040]; Kaluwitharana[0.040]; Kaluwitharana[0.040]; Gillespie[0.040]; wickets[0.040]; wickets[0.040]; Bevan[0.040]; Ponting[0.040]; Ranatunga[0.040]; Slater[0.039]; Scoreboard[0.039]; Scoreboard[0.039]; Fleming[0.039]; \t\n\r [============================>.................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s565ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3022/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1111testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dublin ==> ENTITY: \u001b[32mDublin\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.025:0.025[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJames[0.058]; Saturday[0.055]; American[0.053]; Heath[0.053]; title[0.052]; round[0.052]; seventh[0.051]; defended[0.050]; bout[0.049]; knocked[0.049]; successfully[0.048]; Boxing[0.048]; Title[0.048]; Nate[0.048]; Miller[0.048]; Miller[0.048]; Cruiserweight[0.048]; compatriot[0.047]; cruiserweight[0.047]; Defends[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = American ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.015:0.015[0]; log p(e|m)= -0.130:-0.130[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJames[0.058]; seventh[0.057]; Saturday[0.052]; Heath[0.051]; Miller[0.051]; Miller[0.051]; title[0.051]; Boxing[0.051]; round[0.050]; Title[0.050]; Dublin[0.050]; Nate[0.049]; defended[0.049]; bout[0.048]; cruiserweight[0.048]; Cruiserweight[0.047]; knocked[0.047]; compatriot[0.047]; successfully[0.047]; Defends[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nate Miller\u001b[39m ==> ENTITY: \u001b[32mNate Miller\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCruiserweight[0.060]; cruiserweight[0.059]; bout[0.056]; title[0.056]; Title[0.055]; Miller[0.055]; seventh[0.055]; Boxing[0.054]; knocked[0.054]; round[0.053]; defended[0.052]; James[0.052]; Heath[0.052]; Saturday[0.050]; compatriot[0.048]; American[0.048]; Dublin[0.047]; Defends[0.046]; successfully[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = James Heath ==> ENTITY: \u001b[32mJames Heath (boxer)\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.146:0.146[0]; log p(e|m)= -1.890:-1.890[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCruiserweight[0.058]; cruiserweight[0.057]; Title[0.056]; bout[0.056]; Miller[0.055]; Miller[0.055]; Boxing[0.054]; title[0.054]; seventh[0.053]; round[0.052]; Nate[0.052]; compatriot[0.052]; American[0.052]; knocked[0.051]; successfully[0.050]; Dublin[0.050]; Saturday[0.049]; defended[0.047]; Defends[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Miller\u001b[39m ==> ENTITY: \u001b[32mNate Miller\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.142:0.142[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCruiserweight[0.056]; cruiserweight[0.056]; bout[0.053]; title[0.053]; Title[0.053]; Miller[0.053]; seventh[0.052]; Boxing[0.051]; Nate[0.051]; knocked[0.051]; round[0.050]; defended[0.050]; James[0.050]; Heath[0.049]; Saturday[0.047]; compatriot[0.046]; American[0.046]; Dublin[0.045]; Defends[0.044]; successfully[0.044]; \t\n\r [============================>.................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s625ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3027/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1143testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iranian ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.571:-0.571[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIranian[0.046]; Iranian[0.046]; Kurdistan[0.045]; Kurdish[0.044]; Kurdish[0.044]; region[0.044]; Iraq[0.044]; Iraq[0.044]; Iraq[0.044]; Kurds[0.043]; northern[0.042]; northern[0.042]; Iraqi[0.042]; Iraqi[0.042]; Baghdad[0.041]; Ankara[0.040]; Istanbul[0.040]; Faik[0.039]; said[0.039]; said[0.039]; group[0.039]; say[0.039]; north[0.038]; occupied[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iran ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.172:0.172[0]; log p(e|m)= -0.138:-0.138[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.047]; Turkey[0.045]; Kurdistan[0.044]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Kurdish[0.043]; Kurds[0.042]; official[0.042]; northern[0.042]; Baghdad[0.041]; Baghdad[0.041]; Baghdad[0.041]; region[0.041]; casualties[0.040]; attacks[0.040]; said[0.040]; said[0.040]; said[0.040]; Arbil[0.040]; Arbil[0.040]; Gulf[0.040]; Gulf[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.240:0.240[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.047]; Iraq[0.047]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Kurdish[0.043]; Kurdistan[0.042]; Kurdistan[0.042]; helicopters[0.042]; Kurds[0.042]; Baghdad[0.042]; Baghdad[0.042]; Baghdad[0.042]; Iranian[0.041]; Iranian[0.041]; Turkey[0.041]; planes[0.041]; fighters[0.039]; northern[0.039]; northern[0.039]; allied[0.039]; Iran[0.039]; Arbil[0.039]; Arbil[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = PUK\u001b[39m ==> ENTITY: \u001b[32mPatriotic Union of Kurdistan\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.047]; Kurdish[0.045]; Iraq[0.044]; Kurds[0.044]; Arbil[0.043]; Arbil[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; allied[0.042]; Baghdad[0.041]; Baghdad[0.041]; Baghdad[0.041]; Saturday[0.041]; official[0.040]; northern[0.040]; region[0.040]; backed[0.040]; led[0.040]; main[0.039]; repositioning[0.039]; casualties[0.039]; Northern[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iranian ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.228:0.228[0]; log p(e|m)= -0.571:-0.571[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIranian[0.045]; Iranian[0.045]; Iranian[0.045]; Kurdistan[0.044]; Kurdish[0.043]; Kurdish[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; region[0.042]; region[0.042]; Kurds[0.042]; northern[0.041]; northern[0.041]; northern[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; Baghdad[0.040]; Baghdad[0.040]; Arbil[0.039]; Ankara[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.104:0.104[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReuters[0.045]; told[0.043]; told[0.043]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; region[0.042]; northern[0.042]; northern[0.042]; wake[0.042]; official[0.042]; troops[0.041]; troops[0.041]; border[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; headquarters[0.041]; Iraqi[0.041]; Iraqi[0.041]; established[0.041]; Istanbul[0.040]; miles[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = KDP\u001b[39m ==> ENTITY: \u001b[32mKurdistan Democratic Party\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.047]; Kurdish[0.046]; Kurds[0.044]; Arbil[0.043]; Arbil[0.043]; allied[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraq[0.042]; Iraq[0.042]; Baghdad[0.041]; Baghdad[0.041]; Baghdad[0.041]; backed[0.040]; backing[0.040]; region[0.040]; casualties[0.039]; Turkey[0.039]; Puk[0.039]; northern[0.039]; northern[0.039]; Union[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.227:0.227[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.049]; Iraq[0.049]; Iraqi[0.045]; Iraqi[0.045]; Kurdistan[0.044]; Baghdad[0.043]; troops[0.043]; troops[0.043]; Kurdish[0.042]; Kurdish[0.042]; Kurds[0.042]; region[0.042]; northern[0.041]; northern[0.041]; Iranian[0.039]; Iranian[0.039]; Iranian[0.039]; Iranian[0.039]; force[0.039]; occupied[0.039]; fighters[0.038]; border[0.038]; withdrawn[0.038]; established[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baghdad\u001b[39m ==> ENTITY: \u001b[32mBaghdad\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.216:0.216[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.047]; Iraq[0.047]; Baghdad[0.047]; Iraqi[0.047]; Iraqi[0.047]; Iraqi[0.047]; Iraqi[0.047]; Kurdistan[0.045]; Kurdistan[0.045]; Kurdish[0.044]; Arbil[0.043]; Arbil[0.043]; city[0.042]; region[0.042]; fighters[0.042]; Iranian[0.042]; Iranian[0.042]; northern[0.041]; northern[0.041]; tanks[0.041]; helicopters[0.041]; fighting[0.041]; fighting[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baghdad\u001b[39m ==> ENTITY: \u001b[32mBaghdad\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.050]; Iraqi[0.046]; Iraqi[0.046]; Iraqi[0.046]; Baghdad[0.046]; Kurdistan[0.045]; Kurdish[0.043]; Turkey[0.043]; Kurds[0.042]; Arbil[0.042]; Arbil[0.042]; northern[0.042]; city[0.041]; region[0.041]; War[0.041]; Iran[0.041]; Gulf[0.039]; Gulf[0.039]; main[0.038]; Northern[0.038]; allied[0.038]; based[0.037]; accuses[0.037]; led[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.066:0.066[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.044]; Iraq[0.044]; War[0.044]; planes[0.043]; helicopters[0.043]; intended[0.043]; based[0.043]; size[0.043]; December[0.042]; northern[0.042]; northern[0.042]; Union[0.041]; Bill[0.041]; backing[0.041]; far[0.041]; force[0.041]; area[0.040]; region[0.040]; know[0.040]; know[0.040]; Iraqi[0.040]; Iraqi[0.040]; Iraqi[0.040]; Iraqi[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = KDP\u001b[39m ==> ENTITY: \u001b[32mKurdistan Democratic Party\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.051]; Kurdish[0.050]; Kurdish[0.050]; Arbil[0.047]; Iraqi[0.046]; Iraqi[0.046]; Iraqi[0.046]; Iraqi[0.046]; Iraq[0.046]; Iraq[0.046]; Iraq[0.046]; Baghdad[0.045]; Baghdad[0.045]; Faik[0.044]; Party[0.044]; backed[0.044]; backed[0.044]; troops[0.043]; troops[0.043]; region[0.043]; region[0.043]; casualties[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = KDP\u001b[39m ==> ENTITY: \u001b[32mKurdistan Democratic Party\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.051]; Kurdish[0.049]; Kurdish[0.049]; Kurds[0.048]; Arbil[0.047]; Iraqi[0.046]; Iraqi[0.046]; Iraqi[0.046]; Iraqi[0.046]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; Baghdad[0.045]; Baghdad[0.045]; Faik[0.044]; Party[0.043]; backed[0.043]; backed[0.043]; troops[0.043]; troops[0.043]; region[0.043]; casualties[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = KDP\u001b[39m ==> ENTITY: \u001b[32mKurdistan Democratic Party\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.047]; Kurdish[0.046]; Kurds[0.044]; Arbil[0.043]; Arbil[0.043]; allied[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraq[0.042]; Baghdad[0.042]; Baghdad[0.042]; Baghdad[0.042]; backed[0.040]; backing[0.040]; region[0.040]; casualties[0.040]; Turkey[0.040]; Puk[0.039]; northern[0.039]; Union[0.039]; President[0.039]; official[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baghdad\u001b[39m ==> ENTITY: \u001b[32mBaghdad\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.049]; Iraq[0.049]; Iraq[0.049]; Iraqi[0.045]; Iraqi[0.045]; Kurdistan[0.044]; troops[0.042]; troops[0.042]; Kurdish[0.042]; Kurdish[0.042]; Kurds[0.041]; northern[0.041]; northern[0.041]; Iranian[0.040]; Iranian[0.040]; Iranian[0.040]; Iranian[0.040]; region[0.039]; Istanbul[0.039]; fighters[0.038]; border[0.038]; north[0.038]; telephone[0.038]; Party[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arbil\u001b[39m ==> ENTITY: \u001b[32mErbil\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.242:0.242[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.045]; Kurdistan[0.045]; Kurdish[0.045]; Kurds[0.044]; Arbil[0.043]; region[0.043]; Iraq[0.043]; Iraq[0.043]; Iraqi[0.042]; Iraqi[0.042]; Iraqi[0.042]; Iraqi[0.042]; area[0.041]; Baghdad[0.041]; Baghdad[0.041]; Baghdad[0.041]; northern[0.041]; northern[0.041]; Ankara[0.040]; Turkey[0.040]; border[0.039]; city[0.039]; Faik[0.038]; west[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBaghdad\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBaghdad\u001b[39m <---> \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.280:0.268[\u001b[31m0.012\u001b[39m]; local()= 0.207:0.215[\u001b[32m0.007\u001b[39m]; log p(e|m)= 0.000:-2.564[\u001b[31m2.564\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.049]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Baghdad[0.045]; Baghdad[0.045]; Kurdistan[0.044]; Kurdish[0.042]; Turkey[0.042]; Kurds[0.041]; Arbil[0.041]; Arbil[0.041]; northern[0.041]; city[0.040]; Iran[0.040]; region[0.039]; tanks[0.039]; helicopters[0.039]; fighting[0.039]; fighting[0.039]; artillery[0.038]; casualties[0.038]; official[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.226:0.226[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.050]; Iraq[0.050]; Iraqi[0.045]; Iraqi[0.045]; Kurdistan[0.044]; Baghdad[0.044]; troops[0.043]; troops[0.043]; Kurdish[0.043]; Kurdish[0.043]; Kurds[0.042]; northern[0.041]; northern[0.041]; region[0.040]; Iranian[0.040]; Iranian[0.040]; Iranian[0.040]; fighters[0.039]; force[0.039]; occupied[0.039]; established[0.038]; backed[0.038]; Istanbul[0.037]; Ankara[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Patriotic Union of Kurdistan\u001b[39m ==> ENTITY: \u001b[32mPatriotic Union of Kurdistan\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdish[0.045]; Iraq[0.044]; Kurds[0.044]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; allied[0.043]; Arbil[0.043]; Arbil[0.043]; Baghdad[0.041]; Baghdad[0.041]; Baghdad[0.041]; official[0.041]; northern[0.041]; region[0.041]; Puk[0.041]; backed[0.040]; led[0.040]; main[0.040]; repositioning[0.040]; casualties[0.040]; Northern[0.040]; backing[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = KDP\u001b[39m ==> ENTITY: \u001b[32mKurdistan Democratic Party\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.047]; Kurdistan[0.047]; Kurdish[0.046]; Arbil[0.043]; Arbil[0.043]; allied[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraq[0.042]; Iraq[0.042]; Baghdad[0.041]; Baghdad[0.041]; Faik[0.041]; Party[0.040]; backed[0.040]; backing[0.040]; region[0.040]; casualties[0.039]; Puk[0.039]; northern[0.039]; northern[0.039]; Union[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ankara\u001b[39m ==> ENTITY: \u001b[32mAnkara\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nregion[0.046]; area[0.045]; area[0.045]; north[0.043]; northern[0.043]; northern[0.043]; northern[0.043]; Istanbul[0.043]; city[0.043]; Faik[0.042]; Kurdish[0.042]; Kurdish[0.042]; Kurdistan[0.041]; fighting[0.041]; Arbil[0.041]; Kurds[0.040]; west[0.040]; Iraq[0.040]; Iraq[0.040]; Iraq[0.040]; km[0.040]; official[0.040]; established[0.040]; miles[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iraqi Kurdish\u001b[39m ==> ENTITY: \u001b[32mIraqi Kurdistan\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.230:0.230[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.049]; Kurdish[0.049]; Kurds[0.048]; northern[0.047]; northern[0.047]; Iraq[0.046]; Iraq[0.046]; Iraq[0.046]; region[0.046]; Iraqi[0.044]; troops[0.043]; troops[0.043]; Baghdad[0.043]; border[0.043]; north[0.042]; Faik[0.041]; Iranian[0.040]; Iranian[0.040]; Iranian[0.040]; Iranian[0.040]; occupied[0.040]; Ankara[0.040]; fighters[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gulf ==> ENTITY: \u001b[32mPersian Gulf\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.173:0.173[0]; log p(e|m)= -2.163:-2.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.046]; Iraq[0.046]; Gulf[0.046]; region[0.044]; city[0.043]; Turkey[0.043]; northern[0.042]; northern[0.042]; west[0.042]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; area[0.041]; War[0.040]; Baghdad[0.040]; Baghdad[0.040]; Baghdad[0.040]; Kurdistan[0.040]; Iranian[0.040]; Iranian[0.040]; helicopters[0.040]; Iran[0.040]; border[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iranian ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.214:0.214[0]; log p(e|m)= -0.571:-0.571[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIranian[0.045]; Iranian[0.045]; Kurdistan[0.044]; Kurdistan[0.044]; Kurdish[0.043]; Kurdish[0.043]; Iraq[0.043]; Iraq[0.043]; region[0.043]; region[0.043]; northern[0.041]; northern[0.041]; northern[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; Baghdad[0.040]; Baghdad[0.040]; Arbil[0.040]; Arbil[0.040]; Ankara[0.039]; Faik[0.039]; occupied[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.214:0.214[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Kurdistan[0.044]; helicopters[0.043]; Baghdad[0.043]; Baghdad[0.043]; Baghdad[0.043]; Kurdish[0.042]; Turkey[0.042]; planes[0.042]; Kurds[0.041]; northern[0.041]; allied[0.040]; Arbil[0.040]; Arbil[0.040]; region[0.040]; Iran[0.040]; tanks[0.040]; artillery[0.039]; Gulf[0.039]; Gulf[0.039]; War[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iraqi Kurdish\u001b[39m ==> ENTITY: \u001b[32mIraqi Kurdistan\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.049]; Kurds[0.047]; northern[0.046]; Iraq[0.045]; region[0.045]; Turkey[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Arbil[0.042]; Arbil[0.042]; Baghdad[0.042]; Baghdad[0.042]; Baghdad[0.042]; allied[0.040]; Iran[0.040]; Puk[0.039]; official[0.039]; main[0.038]; Northern[0.038]; led[0.038]; repositioning[0.038]; city[0.038]; artillery[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.066:0.066[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.044]; Iraq[0.044]; War[0.044]; planes[0.043]; helicopters[0.043]; intended[0.043]; based[0.043]; size[0.043]; December[0.042]; northern[0.042]; northern[0.042]; Union[0.041]; Bill[0.041]; backing[0.041]; far[0.041]; force[0.041]; area[0.040]; region[0.040]; know[0.040]; know[0.040]; Iraqi[0.040]; Iraqi[0.040]; Iraqi[0.040]; Iraqi[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.237:0.237[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.050]; Iraq[0.050]; Iraqi[0.046]; Iraqi[0.046]; Iraqi[0.046]; Iraqi[0.046]; Kurdistan[0.045]; helicopters[0.044]; Baghdad[0.044]; Baghdad[0.044]; troops[0.043]; troops[0.043]; Kurdish[0.043]; Kurdish[0.043]; fighters[0.042]; northern[0.041]; northern[0.041]; northern[0.041]; Arbil[0.041]; region[0.041]; region[0.041]; artillery[0.040]; tanks[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kurdistan Democratic Party\u001b[39m ==> ENTITY: \u001b[32mKurdistan Democratic Party\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdish[0.046]; Kurdish[0.046]; Kurds[0.044]; Arbil[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; Baghdad[0.041]; Baghdad[0.041]; Faik[0.041]; Iranian[0.041]; Iranian[0.041]; Iranian[0.041]; Iranian[0.041]; backed[0.040]; backed[0.040]; official[0.040]; troops[0.040]; troops[0.040]; region[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arbil\u001b[39m ==> ENTITY: \u001b[32mErbil\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.217:0.217[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.046]; Kurdish[0.046]; Kurds[0.045]; Arbil[0.044]; region[0.044]; Iraq[0.044]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; area[0.042]; Baghdad[0.042]; Baghdad[0.042]; Baghdad[0.042]; northern[0.042]; Turkey[0.041]; border[0.040]; city[0.040]; west[0.039]; Saturday[0.038]; official[0.038]; Puk[0.038]; Iranian[0.037]; Iran[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iraqi Kurds\u001b[39m ==> ENTITY: \u001b[32mIraqi Kurdistan\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.215:0.215[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.047]; Kurdish[0.047]; Kurdish[0.047]; northern[0.045]; northern[0.045]; Iraq[0.044]; Iraq[0.044]; region[0.043]; Iraqi[0.042]; Iranian[0.041]; Iranian[0.041]; Iranian[0.041]; troops[0.041]; troops[0.041]; Baghdad[0.041]; north[0.040]; area[0.040]; Faik[0.040]; said[0.039]; Ankara[0.039]; occupied[0.038]; force[0.037]; Istanbul[0.037]; miles[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Istanbul ==> ENTITY: \u001b[32mIstanbul\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.025:-0.025[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAnkara[0.044]; region[0.043]; Faik[0.043]; Kurds[0.042]; Kurdish[0.042]; Kurdish[0.042]; northern[0.042]; northern[0.042]; Iranian[0.042]; Iranian[0.042]; Iranian[0.042]; Kurdistan[0.042]; occupied[0.042]; Iraq[0.041]; Iraq[0.041]; Iraq[0.041]; Iraqi[0.041]; Iraqi[0.041]; group[0.041]; troops[0.041]; troops[0.041]; fighters[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.225:0.225[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.048]; Iraq[0.048]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Kurdish[0.044]; Kurdistan[0.043]; Kurdistan[0.043]; helicopters[0.042]; Baghdad[0.042]; Baghdad[0.042]; Iranian[0.041]; Iranian[0.041]; planes[0.041]; fighters[0.040]; northern[0.040]; northern[0.040]; allied[0.039]; Iran[0.039]; Arbil[0.039]; Arbil[0.039]; region[0.039]; artillery[0.039]; tanks[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.238:0.238[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.048]; Iraq[0.048]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Kurdish[0.043]; Kurdistan[0.042]; helicopters[0.042]; Kurds[0.042]; Baghdad[0.042]; Baghdad[0.042]; Baghdad[0.042]; Iranian[0.041]; Iranian[0.041]; Turkey[0.041]; planes[0.041]; fighters[0.040]; northern[0.040]; northern[0.040]; allied[0.039]; Iran[0.039]; Arbil[0.039]; Arbil[0.039]; region[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.105:0.105[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.045]; Iraq[0.045]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iran[0.043]; response[0.042]; Iranian[0.042]; Iranian[0.042]; Baghdad[0.041]; Baghdad[0.041]; President[0.041]; Kurdistan[0.041]; Kurdistan[0.041]; occupied[0.041]; December[0.041]; official[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; backed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nReuters[0.049]; told[0.047]; told[0.047]; Iraq[0.046]; Iraq[0.046]; Iraq[0.046]; region[0.046]; northern[0.045]; northern[0.045]; northern[0.045]; main[0.045]; wake[0.045]; official[0.045]; troops[0.045]; troops[0.045]; border[0.045]; said[0.044]; said[0.044]; said[0.044]; said[0.044]; said[0.044]; headquarters[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gulf War\u001b[39m ==> ENTITY: \u001b[32mGulf War\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.048]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Iran[0.043]; helicopters[0.043]; Kurdistan[0.042]; Kurdish[0.042]; Baghdad[0.041]; Baghdad[0.041]; Baghdad[0.041]; Turkey[0.041]; Gulf[0.041]; Kurds[0.041]; planes[0.041]; region[0.040]; firepower[0.040]; Arbil[0.040]; Arbil[0.040]; allied[0.040]; northern[0.039]; December[0.039]; tanks[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bill Clinton\u001b[39m ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.075:0.075[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.054]; Iraq[0.053]; Iraq[0.053]; Iraqi[0.051]; Iraqi[0.051]; Iraqi[0.051]; Iraqi[0.051]; accuses[0.050]; Iran[0.050]; response[0.049]; close[0.049]; official[0.049]; backing[0.049]; told[0.049]; Iranian[0.049]; Iranian[0.049]; attacks[0.049]; Turkey[0.048]; Union[0.048]; intended[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iranian ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.217:0.217[0]; log p(e|m)= -0.571:-0.571[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIranian[0.047]; Iranian[0.047]; Iranian[0.047]; Kurdistan[0.046]; Kurdish[0.045]; Kurdish[0.045]; region[0.045]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; Kurds[0.045]; northern[0.043]; northern[0.043]; Iraqi[0.043]; Iraqi[0.043]; Baghdad[0.042]; Ankara[0.041]; Istanbul[0.041]; Faik[0.041]; said[0.040]; said[0.040]; say[0.040]; north[0.040]; \t\n\r [============================>.................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s426ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3067/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1148testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.176:0.176[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.051]; Kurdistan[0.048]; Iraqi[0.047]; Iraqi[0.047]; Kurdish[0.047]; military[0.046]; northern[0.045]; Baghdad[0.044]; troops[0.044]; Arbil[0.042]; Arbil[0.042]; said[0.040]; reports[0.040]; officials[0.039]; official[0.039]; tanks[0.039]; factions[0.038]; city[0.038]; city[0.038]; city[0.038]; fighting[0.037]; Reuters[0.037]; seen[0.037]; tank[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.064:0.064[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nreports[0.047]; told[0.045]; Iraq[0.044]; northern[0.043]; official[0.043]; troops[0.043]; said[0.042]; denies[0.042]; Iraqi[0.042]; Iraqi[0.042]; Iraqi[0.042]; military[0.041]; fighting[0.041]; Democratic[0.041]; Baghdad[0.041]; officials[0.040]; city[0.040]; city[0.040]; city[0.040]; think[0.040]; rival[0.040]; advance[0.040]; relief[0.040]; relief[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.176:0.176[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.051]; Kurdistan[0.048]; Iraqi[0.047]; Iraqi[0.047]; Kurdish[0.047]; military[0.046]; northern[0.045]; Baghdad[0.044]; troops[0.044]; Arbil[0.042]; Arbil[0.042]; said[0.040]; reports[0.040]; officials[0.039]; official[0.039]; tanks[0.039]; factions[0.038]; city[0.038]; city[0.038]; city[0.038]; fighting[0.037]; Reuters[0.037]; seen[0.037]; tank[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arbil\u001b[39m ==> ENTITY: \u001b[32mErbil\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.048]; Kurdish[0.048]; Arbil[0.046]; Iraq[0.046]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; troops[0.044]; Baghdad[0.044]; northern[0.044]; military[0.042]; city[0.041]; city[0.041]; city[0.041]; officials[0.040]; advance[0.040]; Saturday[0.039]; official[0.039]; succeed[0.038]; think[0.038]; Democratic[0.038]; factions[0.037]; said[0.037]; seen[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = KDP\u001b[39m ==> ENTITY: \u001b[32mKurdistan Democratic Party\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.049]; Kurdish[0.047]; Arbil[0.045]; Arbil[0.045]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Iraq[0.044]; Baghdad[0.043]; factions[0.042]; Party[0.042]; troops[0.041]; northern[0.041]; military[0.040]; official[0.040]; advance[0.040]; Democratic[0.039]; fighting[0.039]; succeed[0.039]; said[0.039]; rival[0.038]; tanks[0.038]; officials[0.038]; told[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.074:0.074[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.047]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; reports[0.044]; Baghdad[0.042]; military[0.042]; officials[0.042]; Kurdistan[0.042]; official[0.041]; said[0.041]; troops[0.041]; denies[0.041]; Party[0.041]; Arbil[0.041]; Arbil[0.041]; assault[0.040]; United[0.040]; told[0.040]; Nations[0.040]; advance[0.040]; aware[0.040]; Kurdish[0.040]; Reuters[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baghdad\u001b[39m ==> ENTITY: \u001b[32mBaghdad\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.050]; city[0.047]; city[0.047]; city[0.047]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; northern[0.045]; Kurdistan[0.044]; troops[0.043]; Kurdish[0.043]; military[0.042]; Arbil[0.042]; Arbil[0.042]; official[0.041]; officials[0.039]; advance[0.037]; Nations[0.037]; Reuters[0.037]; reports[0.037]; tanks[0.036]; Saturday[0.036]; factions[0.036]; assault[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kurdistan Democratic Party\u001b[39m ==> ENTITY: \u001b[32mKurdistan Democratic Party\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdish[0.048]; Arbil[0.045]; Arbil[0.045]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Iraq[0.044]; military[0.044]; Baghdad[0.044]; factions[0.043]; official[0.042]; troops[0.042]; northern[0.041]; said[0.040]; advance[0.039]; succeed[0.039]; rival[0.039]; tanks[0.039]; officials[0.039]; seen[0.038]; Reuters[0.038]; told[0.038]; United[0.038]; fighting[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.166:0.166[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraqi[0.047]; Iraqi[0.047]; Iraqi[0.047]; military[0.047]; Kurdistan[0.046]; Baghdad[0.045]; troops[0.045]; Kurdish[0.044]; northern[0.043]; Arbil[0.042]; Arbil[0.042]; officials[0.040]; said[0.040]; official[0.040]; tanks[0.039]; advance[0.039]; Reuters[0.039]; reports[0.039]; fighting[0.038]; seen[0.038]; tank[0.038]; factions[0.038]; Saturday[0.037]; assault[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arbil\u001b[39m ==> ENTITY: \u001b[32mErbil\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdistan[0.048]; Kurdish[0.048]; Arbil[0.046]; Iraq[0.046]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; troops[0.044]; Baghdad[0.044]; northern[0.044]; military[0.042]; city[0.041]; city[0.041]; city[0.041]; officials[0.040]; advance[0.040]; Saturday[0.039]; official[0.039]; succeed[0.038]; think[0.038]; Democratic[0.038]; factions[0.037]; said[0.037]; seen[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.176:0.176[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.051]; Kurdistan[0.048]; Iraqi[0.047]; Iraqi[0.047]; Kurdish[0.047]; military[0.046]; northern[0.045]; Baghdad[0.044]; troops[0.044]; Arbil[0.042]; Arbil[0.042]; said[0.040]; reports[0.040]; officials[0.039]; official[0.039]; tanks[0.039]; factions[0.038]; city[0.038]; city[0.038]; city[0.038]; fighting[0.037]; Reuters[0.037]; seen[0.037]; tank[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = United Nations\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.074:0.074[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.050]; Iraqi[0.047]; Iraqi[0.047]; Iraqi[0.047]; reports[0.046]; official[0.045]; Baghdad[0.044]; military[0.044]; officials[0.044]; Kurdistan[0.044]; Reuters[0.043]; troops[0.043]; Arbil[0.043]; Arbil[0.043]; Kurdish[0.042]; assault[0.042]; said[0.042]; advance[0.041]; told[0.041]; denies[0.041]; fighting[0.040]; Saturday[0.040]; northern[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.074:0.074[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.047]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; reports[0.044]; Baghdad[0.042]; military[0.042]; officials[0.042]; Kurdistan[0.042]; official[0.041]; said[0.041]; troops[0.041]; denies[0.041]; Party[0.041]; Arbil[0.041]; Arbil[0.041]; assault[0.040]; United[0.040]; told[0.040]; Nations[0.040]; advance[0.040]; aware[0.040]; Kurdish[0.040]; Reuters[0.039]; \t\n\r [============================>.................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s409ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3081/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1023testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.065:0.065[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnumber[0.046]; Canada[0.045]; opened[0.044]; Canadian[0.043]; Canadian[0.043]; Canadian[0.043]; Canadian[0.043]; narrower[0.042]; numbers[0.041]; Friday[0.041]; open[0.041]; fell[0.041]; fell[0.041]; benchmark[0.041]; Jim[0.041]; account[0.041]; current[0.041]; data[0.041]; lower[0.040]; pulled[0.040]; market[0.040]; Toronto[0.040]; morning[0.040]; bond[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canadian ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.163:0.163[0]; log p(e|m)= -0.412:-0.412[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.052]; Canada[0.052]; Canada[0.052]; Canadian[0.048]; Canadian[0.048]; Canadian[0.048]; Toronto[0.043]; particularly[0.040]; economic[0.040]; Jim[0.039]; billion[0.039]; income[0.039]; account[0.039]; account[0.039]; Wednesday[0.039]; close[0.039]; supportive[0.038]; number[0.038]; current[0.038]; current[0.038]; said[0.038]; said[0.038]; expected[0.038]; moved[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Labour Day\u001b[39m ==> ENTITY: \u001b[32mLabour Day\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nholiday[0.045]; economic[0.045]; weekends[0.043]; Friday[0.043]; Monday[0.043]; early[0.042]; early[0.042]; Thursday[0.042]; July[0.042]; month[0.042]; month[0.042]; Toronto[0.042]; data[0.041]; closes[0.041]; August[0.041]; number[0.041]; Canadian[0.041]; Canadian[0.041]; Canadian[0.041]; Canadian[0.041]; release[0.040]; typically[0.040]; financial[0.040]; closed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnumber[0.051]; Canada[0.049]; Canada[0.049]; Canada[0.049]; release[0.049]; fourth[0.048]; billion[0.047]; billion[0.047]; reported[0.047]; reported[0.047]; second[0.047]; second[0.047]; early[0.047]; early[0.047]; news[0.047]; August[0.047]; July[0.047]; moved[0.047]; report[0.047]; expected[0.046]; expected[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canadian ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.182:0.182[0]; log p(e|m)= -0.412:-0.412[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.051]; Canada[0.051]; Canada[0.051]; Canadian[0.047]; Canadian[0.047]; Canadian[0.047]; Canadian[0.047]; Toronto[0.042]; support[0.040]; economic[0.039]; rate[0.039]; news[0.039]; strong[0.039]; strong[0.039]; billion[0.039]; billion[0.039]; domestic[0.039]; Monday[0.038]; account[0.038]; Wednesday[0.038]; showing[0.038]; financial[0.038]; close[0.038]; release[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.151:0.151[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.051]; Canada[0.051]; Canadian[0.047]; Canadian[0.047]; Canadian[0.047]; Canadian[0.047]; Toronto[0.042]; Friday[0.042]; Friday[0.042]; reported[0.041]; Wednesday[0.041]; number[0.040]; second[0.040]; expected[0.040]; fourth[0.039]; basis[0.038]; basis[0.038]; open[0.038]; numbers[0.038]; particularly[0.038]; morning[0.038]; quarter[0.038]; quarter[0.038]; quarter[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.067:0.067[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnumber[0.045]; Canada[0.044]; Canada[0.044]; current[0.044]; percent[0.042]; percent[0.042]; percent[0.042]; data[0.042]; Canadian[0.042]; Canadian[0.042]; Canadian[0.042]; Canadian[0.042]; basis[0.041]; basis[0.041]; narrower[0.041]; numbers[0.041]; Jim[0.041]; fell[0.041]; fell[0.041]; particularly[0.041]; Wednesday[0.040]; account[0.040]; benchmark[0.040]; benchmark[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = GMT\u001b[39m ==> ENTITY: \u001b[32mGreenwich Mean Time\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.142:0.142[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMonday[0.046]; Thursday[0.046]; Friday[0.045]; Managers[0.045]; recommending[0.045]; month[0.044]; month[0.044]; news[0.044]; early[0.044]; early[0.044]; July[0.043]; percent[0.043]; percent[0.043]; percent[0.043]; percent[0.043]; percent[0.043]; percent[0.043]; provided[0.043]; August[0.043]; expected[0.042]; managers[0.042]; second[0.042]; holiday[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canadian ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.182:0.182[0]; log p(e|m)= -0.412:-0.412[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.051]; Canada[0.051]; Canada[0.051]; Canadian[0.047]; Canadian[0.047]; Canadian[0.047]; Canadian[0.047]; Toronto[0.042]; support[0.040]; economic[0.039]; rate[0.039]; news[0.039]; strong[0.039]; strong[0.039]; billion[0.039]; billion[0.039]; domestic[0.039]; Monday[0.038]; account[0.038]; Wednesday[0.038]; showing[0.038]; financial[0.038]; close[0.038]; release[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mU.S.\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUnited States dollar\u001b[39m <---> \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.255:0.254[\u001b[31m0.001\u001b[39m]; local()= 0.141:0.078[\u001b[31m0.063\u001b[39m]; log p(e|m)= -4.075:-0.140[\u001b[32m3.934\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmonth[0.045]; month[0.045]; cash[0.043]; typically[0.043]; early[0.042]; early[0.042]; percent[0.042]; percent[0.042]; percent[0.042]; percent[0.042]; percent[0.042]; Canadian[0.042]; Canadian[0.042]; basis[0.041]; basis[0.041]; Jeffrey[0.041]; financial[0.041]; Day[0.041]; fell[0.041]; fell[0.041]; closed[0.040]; benchmark[0.040]; Thursday[0.040]; Monday[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mU.S.\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUnited States dollar\u001b[39m <---> \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.255:0.254[\u001b[31m0.001\u001b[39m]; local()= 0.161:0.096[\u001b[31m0.064\u001b[39m]; log p(e|m)= -4.075:-0.140[\u001b[32m3.934\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnumber[0.050]; Canada[0.049]; Canada[0.049]; Canada[0.049]; Canada[0.049]; current[0.048]; current[0.048]; fourth[0.047]; billion[0.047]; billion[0.047]; reported[0.047]; reported[0.047]; second[0.047]; second[0.047]; domestic[0.047]; percent[0.047]; percent[0.047]; percent[0.047]; percent[0.047]; expected[0.046]; expected[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canadian ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.111:0.111[0]; log p(e|m)= -0.412:-0.412[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.055]; Canadian[0.050]; Canadian[0.050]; Canadian[0.050]; Toronto[0.045]; particularly[0.042]; economic[0.042]; income[0.041]; account[0.041]; Jim[0.040]; number[0.040]; current[0.040]; think[0.040]; lower[0.039]; director[0.039]; fixed[0.039]; numbers[0.039]; softer[0.039]; softer[0.039]; Friday[0.038]; open[0.038]; positive[0.038]; said[0.038]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canadian ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.412:-0.412[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.057]; Canadian[0.052]; Canadian[0.052]; Canadian[0.052]; Toronto[0.047]; particularly[0.044]; economic[0.043]; income[0.042]; account[0.042]; Jim[0.042]; number[0.042]; current[0.042]; think[0.041]; lower[0.041]; director[0.040]; fixed[0.040]; numbers[0.040]; softer[0.040]; softer[0.040]; Friday[0.040]; open[0.040]; narrowed[0.040]; positive[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnumber[0.051]; Canada[0.049]; Canada[0.049]; Canada[0.049]; release[0.049]; fourth[0.048]; billion[0.047]; billion[0.047]; reported[0.047]; reported[0.047]; second[0.047]; second[0.047]; early[0.047]; early[0.047]; news[0.047]; August[0.047]; July[0.047]; moved[0.047]; report[0.047]; expected[0.046]; expected[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Statistics Canada\u001b[39m ==> ENTITY: \u001b[32mStatistics Canada\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanadian[0.044]; Canadian[0.044]; Canadian[0.044]; Canadian[0.044]; Canada[0.044]; Canada[0.044]; Canada[0.044]; Td[0.044]; income[0.042]; numbers[0.041]; data[0.041]; data[0.041]; data[0.041]; rate[0.041]; current[0.041]; current[0.041]; expected[0.041]; expected[0.041]; economic[0.041]; economic[0.041]; number[0.040]; report[0.040]; reported[0.039]; reported[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canadian ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.159:0.159[0]; log p(e|m)= -0.412:-0.412[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.055]; Canada[0.055]; Canada[0.055]; Canadian[0.050]; Canadian[0.050]; Canadian[0.050]; Toronto[0.045]; particularly[0.042]; economic[0.042]; income[0.041]; account[0.041]; close[0.041]; Jim[0.040]; number[0.040]; current[0.040]; Wednesday[0.040]; Statistics[0.040]; think[0.040]; lower[0.039]; reported[0.039]; director[0.039]; fixed[0.039]; numbers[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canadian ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.412:-0.412[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanadian[0.049]; Canadian[0.049]; Canadian[0.049]; Canadian[0.049]; Toronto[0.044]; month[0.042]; month[0.042]; support[0.042]; cash[0.041]; economic[0.040]; rate[0.040]; strong[0.040]; strong[0.040]; Monday[0.040]; Thursday[0.040]; showing[0.040]; financial[0.040]; close[0.040]; release[0.039]; early[0.039]; early[0.039]; Association[0.039]; Managers[0.039]; number[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canadian ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.412:-0.412[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanadian[0.049]; Canadian[0.049]; Canadian[0.049]; Canadian[0.049]; Toronto[0.044]; month[0.042]; support[0.042]; cash[0.042]; economic[0.041]; rate[0.040]; strong[0.040]; strong[0.040]; domestic[0.040]; Monday[0.040]; Thursday[0.040]; showing[0.040]; financial[0.040]; close[0.040]; release[0.039]; early[0.039]; early[0.039]; Association[0.039]; Managers[0.039]; number[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Toronto ==> ENTITY: \u001b[32mToronto\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.123:-0.123[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.049]; Canadian[0.046]; Canadian[0.046]; Canadian[0.046]; Canadian[0.046]; opened[0.043]; Friday[0.042]; economic[0.041]; particularly[0.041]; Jim[0.041]; open[0.040]; lower[0.040]; Webber[0.040]; said[0.040]; said[0.040]; number[0.040]; research[0.040]; current[0.040]; spread[0.040]; income[0.040]; numbers[0.040]; spreads[0.039]; spreads[0.039]; director[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.174:0.174[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.051]; Canada[0.051]; Canada[0.051]; Canadian[0.047]; Canadian[0.047]; Canadian[0.047]; Canadian[0.047]; reported[0.041]; reported[0.041]; number[0.040]; release[0.040]; second[0.040]; second[0.040]; showing[0.039]; fourth[0.039]; Wednesday[0.038]; gross[0.038]; numbers[0.038]; basis[0.038]; basis[0.038]; report[0.038]; expected[0.038]; expected[0.038]; provided[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.092:0.092[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmonth[0.046]; month[0.046]; number[0.045]; cash[0.043]; Canadian[0.042]; Canadian[0.042]; Canadian[0.042]; Jeffrey[0.042]; early[0.042]; early[0.042]; news[0.042]; report[0.041]; Day[0.041]; basis[0.041]; basis[0.041]; August[0.041]; fell[0.041]; fell[0.041]; financial[0.041]; July[0.040]; Monday[0.040]; rose[0.040]; Friday[0.040]; recommending[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mU.S.\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUnited States dollar\u001b[39m <---> \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.253:0.253[\u001b[31m0.000\u001b[39m]; local()= 0.109:0.058[\u001b[31m0.050\u001b[39m]; log p(e|m)= -4.075:-0.140[\u001b[32m3.934\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnumber[0.046]; Canada[0.045]; current[0.044]; percent[0.042]; percent[0.042]; percent[0.042]; data[0.042]; Canadian[0.042]; Canadian[0.042]; Canadian[0.042]; Canadian[0.042]; basis[0.042]; basis[0.042]; narrower[0.041]; numbers[0.041]; Jim[0.041]; fell[0.041]; fell[0.041]; particularly[0.041]; Wednesday[0.040]; account[0.040]; benchmark[0.040]; benchmark[0.040]; lower[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canadian ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.412:-0.412[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanadian[0.049]; Canadian[0.049]; Canadian[0.049]; Canadian[0.049]; Toronto[0.044]; month[0.042]; month[0.042]; support[0.042]; cash[0.041]; economic[0.040]; rate[0.040]; strong[0.040]; strong[0.040]; Monday[0.040]; Thursday[0.040]; showing[0.040]; financial[0.040]; close[0.040]; release[0.039]; early[0.039]; early[0.039]; Association[0.039]; Managers[0.039]; number[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.149:0.149[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.052]; Canada[0.052]; Canadian[0.047]; Canadian[0.047]; Toronto[0.043]; Friday[0.042]; reported[0.041]; reported[0.041]; Wednesday[0.041]; number[0.041]; news[0.040]; second[0.040]; second[0.040]; expected[0.040]; expected[0.040]; showing[0.040]; July[0.040]; fourth[0.040]; August[0.039]; basis[0.039]; basis[0.039]; report[0.038]; provided[0.038]; release[0.038]; \t\n\r [============================>.................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s256ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3105/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1133testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Michoacan\u001b[39m ==> ENTITY: \u001b[32mMichoacán\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.045]; Gonzalo[0.045]; Michoacan[0.045]; Michoacan[0.045]; state[0.044]; state[0.044]; state[0.044]; Mexican[0.044]; Mexican[0.044]; western[0.043]; south[0.043]; Morelia[0.042]; local[0.041]; town[0.041]; miles[0.041]; capital[0.040]; Ak[0.040]; said[0.040]; City[0.039]; military[0.039]; commander[0.039]; Friday[0.038]; km[0.038]; army[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexican ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.397:-0.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexican[0.053]; Mexico[0.050]; style[0.044]; group[0.042]; group[0.042]; Michoacan[0.042]; Michoacan[0.042]; Michoacan[0.042]; City[0.042]; miles[0.041]; Gonzalo[0.041]; local[0.041]; Montoya[0.040]; Montoya[0.040]; south[0.040]; western[0.040]; state[0.040]; state[0.040]; state[0.040]; town[0.040]; military[0.040]; military[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = AK-47s\u001b[39m ==> ENTITY: \u001b[32mAK-47\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrifles[0.051]; armed[0.047]; armed[0.047]; armed[0.047]; attackers[0.045]; attack[0.044]; launched[0.044]; Army[0.044]; fatigues[0.043]; military[0.043]; military[0.043]; military[0.043]; assault[0.043]; wounding[0.042]; wounding[0.042]; attacked[0.042]; attacked[0.042]; attacks[0.042]; soldier[0.042]; rebels[0.042]; men[0.041]; men[0.041]; commander[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Morelia ==> ENTITY: \u001b[32mMorelia\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.123:-0.123[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntown[0.049]; state[0.045]; state[0.045]; state[0.045]; Michoacan[0.045]; Michoacan[0.045]; Michoacan[0.045]; Mexico[0.044]; Gonzalo[0.043]; Mexican[0.043]; Mexican[0.043]; area[0.042]; City[0.041]; local[0.040]; miles[0.040]; south[0.039]; Popular[0.039]; capital[0.039]; Montoya[0.039]; Montoya[0.039]; western[0.038]; police[0.038]; police[0.038]; army[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Popular Revolutionary Army\u001b[39m ==> ENTITY: \u001b[32mPopular Revolutionary Army\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.135:0.135[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexican[0.044]; military[0.044]; military[0.044]; attacks[0.043]; killing[0.043]; armed[0.042]; armed[0.042]; wounding[0.042]; linked[0.042]; fatigues[0.042]; Ak[0.042]; states[0.042]; convoy[0.041]; kidnapping[0.041]; Morelia[0.041]; rebels[0.041]; state[0.041]; saying[0.041]; attack[0.041]; rifles[0.041]; coordinated[0.040]; south[0.040]; commander[0.040]; miles[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexico City ==> ENTITY: \u001b[32mMexico City\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.028:-0.028[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexican[0.044]; Mexican[0.044]; Montoya[0.044]; western[0.044]; state[0.043]; state[0.043]; state[0.043]; south[0.042]; Michoacan[0.042]; Michoacan[0.042]; Michoacan[0.042]; local[0.042]; saying[0.041]; Friday[0.041]; military[0.041]; Morelia[0.041]; miles[0.041]; high[0.040]; Gonzalo[0.040]; town[0.040]; capital[0.040]; army[0.040]; group[0.040]; group[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Michoacan\u001b[39m ==> ENTITY: \u001b[32mMichoacán\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.046]; Gonzalo[0.046]; Michoacan[0.046]; Michoacan[0.046]; state[0.045]; state[0.045]; state[0.045]; Mexican[0.045]; Mexican[0.045]; western[0.045]; area[0.044]; south[0.044]; Morelia[0.043]; local[0.043]; town[0.042]; miles[0.042]; capital[0.042]; Ak[0.041]; said[0.041]; said[0.041]; said[0.041]; people[0.041]; City[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Michoacan\u001b[39m ==> ENTITY: \u001b[32mMichoacán\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.045]; Gonzalo[0.044]; Michoacan[0.044]; Michoacan[0.044]; state[0.044]; state[0.044]; state[0.044]; Mexican[0.044]; Mexican[0.044]; western[0.043]; south[0.043]; Morelia[0.042]; local[0.041]; town[0.041]; miles[0.041]; capital[0.040]; Ak[0.040]; said[0.040]; said[0.040]; City[0.039]; military[0.039]; military[0.039]; commander[0.038]; Friday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexican ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.088:0.088[0]; log p(e|m)= -0.397:-0.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstyle[0.045]; style[0.045]; Gonzalo[0.044]; linked[0.044]; group[0.043]; area[0.043]; people[0.043]; people[0.043]; states[0.042]; Montoya[0.042]; Montoya[0.042]; Popular[0.041]; Army[0.041]; said[0.041]; said[0.041]; day[0.041]; high[0.041]; thought[0.040]; men[0.040]; styled[0.040]; launched[0.040]; clothing[0.040]; told[0.040]; comes[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexican ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.157:0.157[0]; log p(e|m)= -0.397:-0.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexican[0.053]; Mexico[0.051]; group[0.042]; Michoacan[0.042]; Michoacan[0.042]; Michoacan[0.042]; miles[0.042]; Gonzalo[0.041]; western[0.041]; local[0.041]; Montoya[0.041]; south[0.041]; state[0.041]; state[0.041]; state[0.041]; town[0.040]; military[0.040]; said[0.040]; heavily[0.040]; Morelia[0.040]; saying[0.040]; City[0.040]; Friday[0.039]; told[0.039]; \t\n\r [============================>.................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s221ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3115/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1041testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tokyo ==> ENTITY: \u001b[32mTokyo\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.023:0.023[0]; log p(e|m)= -0.093:-0.093[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nyen[0.046]; company[0.046]; August[0.045]; Corp[0.044]; Corp[0.044]; million[0.043]; million[0.043]; million[0.043]; million[0.043]; million[0.043]; million[0.043]; parent[0.042]; months[0.042]; Previous[0.041]; unless[0.040]; Sales[0.040]; Latest[0.040]; Current[0.039]; billions[0.039]; Parent[0.039]; specified[0.039]; Net[0.038]; Note[0.038]; forecast[0.038]; \t\n\r [============================>.................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s223ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3116/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1000testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.049]; Moscow[0.049]; city[0.046]; Russian[0.046]; Olga[0.043]; streets[0.043]; Valery[0.043]; street[0.042]; Friday[0.041]; Interfax[0.041]; Interfax[0.041]; Interfax[0.041]; Thursday[0.041]; seven[0.041]; seven[0.041]; agency[0.040]; dollars[0.040]; dollars[0.040]; confiscation[0.039]; appeared[0.039]; court[0.039]; news[0.039]; morning[0.039]; died[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Interfax\u001b[39m ==> ENTITY: \u001b[32mInterfax\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.221:0.221[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInterfax[0.053]; Interfax[0.053]; news[0.051]; Moscow[0.051]; Moscow[0.051]; Moscow[0.051]; Russian[0.050]; agency[0.047]; quoted[0.046]; Thursday[0.045]; hospital[0.045]; Valery[0.044]; saying[0.044]; Friday[0.044]; said[0.041]; said[0.041]; confiscation[0.041]; Judge[0.041]; morning[0.041]; Olga[0.040]; Ivankov[0.040]; Ivankov[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Interfax\u001b[39m ==> ENTITY: \u001b[32mInterfax\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInterfax[0.049]; Interfax[0.049]; news[0.048]; Moscow[0.048]; Moscow[0.048]; Moscow[0.048]; agency[0.044]; quoted[0.043]; Thursday[0.042]; hospital[0.042]; Valery[0.042]; saying[0.041]; Friday[0.041]; said[0.039]; said[0.039]; confiscation[0.038]; Judge[0.038]; morning[0.038]; Olga[0.038]; Ivankov[0.037]; Ivankov[0.037]; stabbed[0.037]; stabbed[0.037]; appeared[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.049]; Moscow[0.049]; Russian[0.046]; Olga[0.044]; Valery[0.043]; Friday[0.042]; Interfax[0.042]; Interfax[0.042]; Interfax[0.042]; Thursday[0.041]; seven[0.041]; seven[0.041]; ordered[0.041]; agency[0.041]; city[0.041]; dollars[0.040]; dollars[0.040]; confiscation[0.040]; appeared[0.040]; court[0.039]; news[0.039]; said[0.039]; said[0.039]; morning[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.049]; Moscow[0.049]; Russian[0.046]; Olga[0.044]; Valery[0.043]; Friday[0.042]; Interfax[0.042]; Interfax[0.042]; Interfax[0.042]; attacks[0.041]; Thursday[0.041]; seven[0.041]; seven[0.041]; agency[0.041]; city[0.041]; dollars[0.040]; dollars[0.040]; confiscation[0.040]; appeared[0.040]; court[0.039]; saying[0.039]; hospital[0.039]; news[0.039]; died[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Interfax\u001b[39m ==> ENTITY: \u001b[32mInterfax\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.221:0.221[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInterfax[0.053]; Interfax[0.053]; news[0.051]; Moscow[0.051]; Moscow[0.051]; Moscow[0.051]; Russian[0.050]; agency[0.047]; quoted[0.046]; Thursday[0.045]; hospital[0.045]; Valery[0.044]; saying[0.044]; Friday[0.044]; said[0.041]; said[0.041]; confiscation[0.041]; Judge[0.041]; morning[0.041]; Olga[0.040]; Ivankov[0.040]; Ivankov[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.131:0.131[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.047]; Moscow[0.047]; Moscow[0.047]; Olga[0.045]; Ivankov[0.044]; Ivankov[0.044]; Valery[0.044]; illegally[0.043]; illegally[0.043]; Interfax[0.041]; Interfax[0.041]; said[0.040]; said[0.040]; appeared[0.040]; agency[0.040]; Friday[0.040]; seven[0.039]; seven[0.039]; woman[0.039]; dollars[0.039]; dollars[0.039]; news[0.039]; ordered[0.039]; Thursday[0.039]; \t\n\r [============================>.................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s221ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3123/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1109testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dublin ==> ENTITY: \u001b[32mDublin\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.080:0.080[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJames[0.048]; win[0.047]; Saturday[0.045]; wins[0.045]; record[0.044]; Knock[0.044]; score[0.044]; American[0.044]; champion[0.044]; Heath[0.044]; Heath[0.044]; failed[0.043]; title[0.043]; contest[0.043]; round[0.043]; round[0.043]; defending[0.042]; opponent[0.042]; knock[0.042]; knock[0.042]; followed[0.042]; landed[0.042]; seventh[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = American ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.043:0.043[0]; log p(e|m)= -0.130:-0.130[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfifth[0.046]; James[0.046]; seventh[0.045]; record[0.044]; wins[0.044]; fellow[0.043]; win[0.043]; start[0.042]; straight[0.041]; score[0.041]; landed[0.041]; started[0.041]; champion[0.041]; Miller[0.040]; Miller[0.040]; Miller[0.040]; Miller[0.040]; Miller[0.040]; title[0.040]; took[0.040]; round[0.040]; round[0.040]; round[0.040]; failed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Miller\u001b[39m ==> ENTITY: \u001b[32mNate Miller\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncruiserweight[0.045]; fights[0.043]; bout[0.043]; title[0.043]; wins[0.043]; Miller[0.042]; Miller[0.042]; Miller[0.042]; win[0.042]; champion[0.042]; defending[0.042]; seventh[0.042]; fifth[0.042]; Nate[0.042]; landed[0.041]; lefts[0.041]; fight[0.041]; score[0.041]; round[0.040]; round[0.040]; round[0.040]; seconds[0.040]; hook[0.040]; James[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Miller\u001b[39m ==> ENTITY: \u001b[32mNate Miller\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncruiserweight[0.045]; fights[0.043]; bout[0.043]; title[0.043]; wins[0.042]; Title[0.042]; Miller[0.042]; Miller[0.042]; Miller[0.042]; Miller[0.042]; win[0.042]; champion[0.042]; defending[0.042]; seventh[0.042]; fifth[0.041]; Nate[0.041]; landed[0.041]; lefts[0.041]; fight[0.040]; score[0.040]; round[0.040]; round[0.040]; round[0.040]; seconds[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nate Miller\u001b[39m ==> ENTITY: \u001b[32mNate Miller\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncruiserweight[0.045]; fights[0.043]; bout[0.043]; title[0.043]; wins[0.042]; Title[0.042]; Miller[0.042]; Miller[0.042]; Miller[0.042]; Miller[0.042]; win[0.042]; champion[0.042]; defending[0.042]; seventh[0.042]; fifth[0.041]; Boxing[0.041]; landed[0.041]; lefts[0.041]; score[0.040]; round[0.040]; round[0.040]; seconds[0.040]; hook[0.040]; hook[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Miller\u001b[39m ==> ENTITY: \u001b[32mNate Miller\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncruiserweight[0.045]; fights[0.043]; bout[0.043]; title[0.043]; wins[0.042]; Title[0.042]; Miller[0.042]; Miller[0.042]; Miller[0.042]; Miller[0.042]; win[0.042]; champion[0.042]; defending[0.042]; seventh[0.042]; fifth[0.041]; Boxing[0.041]; Nate[0.041]; landed[0.041]; lefts[0.040]; fight[0.040]; score[0.040]; round[0.040]; round[0.040]; round[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = James Heath ==> ENTITY: \u001b[32mJames Heath (boxer)\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.186:0.186[0]; log p(e|m)= -1.890:-1.890[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncruiserweight[0.044]; fights[0.044]; Heath[0.044]; Title[0.043]; bout[0.043]; Miller[0.042]; Miller[0.042]; Miller[0.042]; Miller[0.042]; Miller[0.042]; Boxing[0.042]; title[0.042]; win[0.041]; fifth[0.041]; went[0.041]; seventh[0.041]; round[0.040]; round[0.040]; round[0.040]; wins[0.040]; gave[0.040]; charge[0.040]; Nate[0.040]; champion[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Miller\u001b[39m ==> ENTITY: \u001b[32mNate Miller\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncruiserweight[0.045]; fights[0.043]; bout[0.043]; title[0.043]; wins[0.042]; Title[0.042]; Miller[0.042]; Miller[0.042]; Miller[0.042]; win[0.042]; champion[0.042]; defending[0.042]; seventh[0.042]; Boxing[0.042]; Nate[0.041]; landed[0.041]; lefts[0.041]; score[0.040]; round[0.040]; seconds[0.040]; hook[0.040]; hook[0.040]; James[0.040]; record[0.040]; \t\n\r [=============================>................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s189ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3131/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1074testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Africa ==> ENTITY: \u001b[32mSouth Africa national rugby union team\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.191:0.191[0]; log p(e|m)= -3.612:-3.612[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.048]; South[0.044]; South[0.044]; match[0.043]; series[0.043]; series[0.043]; Africa[0.042]; Africa[0.042]; Honiball[0.042]; Honiball[0.042]; Zealand[0.042]; Zealand[0.042]; Zealand[0.042]; Johannesburg[0.041]; Union[0.041]; Saturday[0.041]; final[0.041]; Park[0.040]; Ellis[0.040]; Mehrtens[0.040]; Mehrtens[0.040]; win[0.040]; Andrew[0.039]; Justin[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Africa ==> ENTITY: \u001b[32mSouth Africa national rugby union team\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.191:0.191[0]; log p(e|m)= -3.612:-3.612[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.048]; South[0.044]; South[0.044]; match[0.043]; series[0.043]; series[0.043]; Africa[0.042]; Africa[0.042]; Honiball[0.042]; Honiball[0.042]; Zealand[0.042]; Zealand[0.042]; Zealand[0.042]; Johannesburg[0.041]; Union[0.041]; Saturday[0.041]; final[0.041]; Park[0.040]; Ellis[0.040]; Mehrtens[0.040]; Mehrtens[0.040]; win[0.040]; Andrew[0.039]; Justin[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andrew Mehrtens\u001b[39m ==> ENTITY: \u001b[32mAndrew Mehrtens\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.049]; Mehrtens[0.045]; Ellis[0.043]; final[0.043]; match[0.043]; Honiball[0.042]; Honiball[0.042]; Zealand[0.042]; Zealand[0.042]; Zealand[0.042]; Blacks[0.042]; Conversions[0.041]; Marshall[0.041]; win[0.041]; Justin[0.040]; Johannesburg[0.040]; Scorers[0.040]; Joubert[0.040]; Joubert[0.040]; Westhuizen[0.040]; Little[0.040]; Sean[0.040]; der[0.040]; Henry[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sean Fitzpatrick\u001b[39m ==> ENTITY: \u001b[32mSean Fitzpatrick\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.045]; match[0.043]; Henry[0.043]; Ellis[0.042]; Andrew[0.042]; series[0.042]; series[0.042]; final[0.042]; Zealand[0.042]; Zealand[0.042]; Zealand[0.042]; South[0.042]; South[0.042]; South[0.042]; Blacks[0.042]; win[0.041]; Honiball[0.041]; Honiball[0.041]; Mehrtens[0.041]; Mehrtens[0.041]; Africa[0.041]; Africa[0.041]; Africa[0.041]; Johannesburg[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Joost van der Westhuizen\u001b[39m ==> ENTITY: \u001b[32mJoost van der Westhuizen\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.128:0.128[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.050]; match[0.047]; Honiball[0.047]; Honiball[0.047]; Mehrtens[0.047]; Mehrtens[0.047]; South[0.046]; South[0.046]; South[0.046]; Joubert[0.045]; Joubert[0.045]; Africa[0.045]; Africa[0.045]; Africa[0.045]; Scorers[0.044]; Johannesburg[0.044]; final[0.044]; Saturday[0.044]; win[0.044]; Zealand[0.044]; Zealand[0.044]; Zealand[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Zealand ==> ENTITY: \u001b[32mNew Zealand national rugby union team\u001b[39m\t\nSCORES: global= 0.241:0.241[0]; local()= 0.180:0.180[0]; log p(e|m)= -3.817:-3.817[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.049]; match[0.044]; final[0.044]; series[0.043]; series[0.043]; Zealand[0.043]; Zealand[0.043]; Johannesburg[0.042]; South[0.041]; South[0.041]; South[0.041]; Mehrtens[0.041]; Mehrtens[0.041]; Ellis[0.041]; Africa[0.041]; Africa[0.041]; Africa[0.041]; Park[0.040]; Honiball[0.040]; Honiball[0.040]; Saturday[0.040]; Andrew[0.040]; Union[0.040]; win[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Zealand ==> ENTITY: \u001b[32mNew Zealand national rugby union team\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.179:0.179[0]; log p(e|m)= -3.817:-3.817[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.049]; match[0.044]; final[0.044]; series[0.043]; series[0.043]; Zealand[0.043]; Zealand[0.043]; Johannesburg[0.042]; South[0.041]; South[0.041]; South[0.041]; Mehrtens[0.041]; Mehrtens[0.041]; Ellis[0.041]; Africa[0.041]; Africa[0.041]; Africa[0.041]; Park[0.040]; Honiball[0.040]; Honiball[0.040]; Saturday[0.040]; Andrew[0.040]; win[0.040]; beat[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Walter Little\u001b[39m ==> ENTITY: \u001b[32mWalter Little (rugby player)\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.139:0.139[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.047]; Ellis[0.046]; Henry[0.045]; Marshall[0.044]; Andrew[0.043]; Mehrtens[0.043]; Mehrtens[0.043]; South[0.041]; South[0.041]; South[0.041]; Sean[0.041]; Fitzpatrick[0.041]; Saturday[0.041]; Honiball[0.041]; Honiball[0.041]; Tries[0.040]; Tries[0.040]; Zealand[0.040]; Zealand[0.040]; Zealand[0.040]; Justin[0.040]; Park[0.040]; match[0.040]; win[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andre Joubert\u001b[39m ==> ENTITY: \u001b[32mAndré Joubert\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.046]; Joubert[0.045]; Johannesburg[0.043]; Honiball[0.043]; Honiball[0.043]; match[0.042]; Mehrtens[0.042]; Mehrtens[0.042]; final[0.042]; Ellis[0.042]; Conversions[0.042]; Westhuizen[0.041]; Blacks[0.041]; win[0.041]; Tries[0.041]; Tries[0.041]; South[0.041]; South[0.041]; South[0.041]; Conversion[0.040]; Saturday[0.040]; Marshall[0.039]; Little[0.039]; Scorers[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rugby Union\u001b[39m ==> ENTITY: \u001b[32mRugby union\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.046]; Zealand[0.045]; Zealand[0.045]; Zealand[0.045]; Mehrtens[0.045]; Mehrtens[0.045]; South[0.045]; South[0.045]; South[0.045]; Ellis[0.044]; Park[0.044]; final[0.043]; Honiball[0.043]; Honiball[0.043]; Johannesburg[0.043]; Saturday[0.043]; win[0.042]; Conversions[0.042]; Scorers[0.042]; series[0.042]; series[0.042]; Blacks[0.041]; Little[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Johannesburg ==> ENTITY: \u001b[32mJohannesburg\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.095:-0.095[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.050]; Park[0.049]; South[0.048]; South[0.048]; South[0.048]; match[0.047]; Saturday[0.045]; Africa[0.045]; Africa[0.045]; Africa[0.045]; series[0.045]; series[0.045]; Honiball[0.045]; Honiball[0.045]; Ellis[0.044]; Zealand[0.044]; Zealand[0.044]; Zealand[0.044]; final[0.044]; win[0.044]; Blacks[0.043]; Union[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Justin Marshall\u001b[39m ==> ENTITY: \u001b[32mJustin Marshall\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.047]; Mehrtens[0.045]; Mehrtens[0.045]; Blacks[0.044]; Zealand[0.043]; Zealand[0.043]; Zealand[0.043]; final[0.042]; Ellis[0.042]; Sean[0.042]; match[0.042]; Andrew[0.042]; Henry[0.042]; Walter[0.041]; Honiball[0.041]; Honiball[0.041]; Saturday[0.040]; Andre[0.040]; Westhuizen[0.039]; South[0.039]; South[0.039]; South[0.039]; Scorers[0.039]; win[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Henry Honiball\u001b[39m ==> ENTITY: \u001b[32mHenry Honiball\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHoniball[0.044]; Zealand[0.043]; Zealand[0.043]; Zealand[0.043]; Rugby[0.043]; match[0.043]; Andrew[0.043]; Westhuizen[0.042]; Joubert[0.042]; Joubert[0.042]; Johannesburg[0.042]; South[0.042]; South[0.042]; South[0.042]; Mehrtens[0.041]; Mehrtens[0.041]; Conversions[0.041]; Ellis[0.041]; Joost[0.041]; Marshall[0.041]; Tries[0.040]; Tries[0.040]; win[0.040]; Little[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ellis Park\u001b[39m ==> ENTITY: \u001b[32mEllis Park Stadium\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.048]; match[0.045]; South[0.045]; South[0.045]; South[0.045]; Johannesburg[0.044]; Saturday[0.042]; Zealand[0.042]; Zealand[0.042]; Zealand[0.042]; final[0.042]; Honiball[0.041]; Honiball[0.041]; win[0.040]; series[0.040]; series[0.040]; Union[0.040]; Henry[0.039]; Africa[0.039]; Africa[0.039]; Africa[0.039]; Scorers[0.039]; Blacks[0.039]; der[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mehrtens\u001b[39m ==> ENTITY: \u001b[32mAndrew Mehrtens\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.049]; Mehrtens[0.045]; Ellis[0.043]; final[0.043]; match[0.043]; Honiball[0.042]; Honiball[0.042]; Zealand[0.042]; Zealand[0.042]; Zealand[0.042]; Blacks[0.042]; Conversions[0.041]; Andrew[0.041]; Marshall[0.041]; win[0.041]; Justin[0.040]; Johannesburg[0.040]; Scorers[0.040]; Joubert[0.040]; Joubert[0.040]; Westhuizen[0.040]; Little[0.040]; Sean[0.040]; der[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = All Blacks\u001b[39m ==> ENTITY: \u001b[32mNew Zealand national rugby union team\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.048]; match[0.044]; final[0.043]; Zealand[0.042]; Zealand[0.042]; Zealand[0.042]; series[0.042]; series[0.042]; Honiball[0.042]; Honiball[0.042]; South[0.041]; South[0.041]; South[0.041]; Ellis[0.041]; Johannesburg[0.041]; Saturday[0.041]; Henry[0.041]; Mehrtens[0.040]; Mehrtens[0.040]; Africa[0.040]; Africa[0.040]; Africa[0.040]; Park[0.040]; win[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Honiball\u001b[39m ==> ENTITY: \u001b[32mHenry Honiball\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHoniball[0.044]; Zealand[0.043]; Zealand[0.043]; Zealand[0.043]; Rugby[0.043]; match[0.043]; Henry[0.043]; Andrew[0.042]; Westhuizen[0.042]; Joubert[0.042]; Joubert[0.042]; Johannesburg[0.042]; South[0.042]; South[0.042]; South[0.042]; Mehrtens[0.041]; Mehrtens[0.041]; Conversions[0.041]; Ellis[0.041]; Joost[0.041]; Marshall[0.040]; Tries[0.040]; Tries[0.040]; win[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Africa ==> ENTITY: \u001b[32mSouth Africa national rugby union team\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.191:0.191[0]; log p(e|m)= -3.612:-3.612[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.048]; South[0.044]; South[0.044]; match[0.043]; series[0.043]; series[0.043]; Africa[0.042]; Africa[0.042]; Honiball[0.042]; Honiball[0.042]; Zealand[0.042]; Zealand[0.042]; Zealand[0.042]; Johannesburg[0.041]; Union[0.041]; Saturday[0.041]; final[0.041]; Park[0.040]; Ellis[0.040]; Mehrtens[0.040]; Mehrtens[0.040]; win[0.040]; Andrew[0.039]; Justin[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New Zealand ==> ENTITY: \u001b[32mNew Zealand national rugby union team\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.179:0.179[0]; log p(e|m)= -3.817:-3.817[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.049]; match[0.044]; final[0.044]; series[0.043]; series[0.043]; Zealand[0.043]; Zealand[0.043]; Johannesburg[0.042]; South[0.041]; South[0.041]; South[0.041]; Mehrtens[0.041]; Mehrtens[0.041]; Ellis[0.041]; Africa[0.041]; Africa[0.041]; Africa[0.041]; Park[0.040]; Honiball[0.040]; Honiball[0.040]; Saturday[0.040]; Andrew[0.040]; win[0.040]; beat[0.040]; \t\n\r [=============================>................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s38ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3150/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1048testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wigan ==> ENTITY: \u001b[32mWigan Warriors\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.195:0.195[0]; log p(e|m)= -1.802:-1.802[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.062]; rugby[0.062]; league[0.061]; Rugby[0.060]; England[0.060]; Bradford[0.059]; Bradford[0.059]; played[0.059]; Wigan[0.056]; Wigan[0.056]; Saturday[0.055]; premiership[0.055]; semifinal[0.051]; English[0.051]; Bulls[0.051]; Semifinal[0.050]; Beat[0.047]; Result[0.046]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWigan\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mWigan Warriors\u001b[39m <---> \u001b[32mWigan\u001b[39m\t\nSCORES: global= 0.269:0.238[\u001b[31m0.030\u001b[39m]; local()= 0.195:0.127[\u001b[31m0.068\u001b[39m]; log p(e|m)= -1.802:-0.302[\u001b[32m1.499\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.062]; rugby[0.062]; league[0.061]; Rugby[0.060]; England[0.060]; Bradford[0.059]; Bradford[0.059]; played[0.059]; Wigan[0.056]; Wigan[0.056]; Saturday[0.055]; premiership[0.055]; semifinal[0.051]; English[0.051]; Bulls[0.051]; Semifinal[0.050]; Beat[0.047]; Result[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bradford Bulls\u001b[39m ==> ENTITY: \u001b[32mBradford Bulls\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.067]; rugby[0.065]; Bradford[0.064]; England[0.064]; League[0.063]; league[0.063]; played[0.060]; Wigan[0.059]; Wigan[0.059]; Wigan[0.059]; premiership[0.058]; Semifinal[0.057]; semifinal[0.056]; Saturday[0.055]; Result[0.051]; English[0.051]; Beat[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wigan ==> ENTITY: \u001b[32mWigan Warriors\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.195:0.195[0]; log p(e|m)= -1.802:-1.802[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.062]; rugby[0.062]; league[0.061]; Rugby[0.060]; England[0.060]; Bradford[0.059]; Bradford[0.059]; played[0.059]; Wigan[0.056]; Wigan[0.056]; Saturday[0.055]; premiership[0.055]; semifinal[0.051]; English[0.051]; Bulls[0.051]; Semifinal[0.050]; Beat[0.047]; Result[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bradford ==> ENTITY: \u001b[32mBradford Bulls\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.189:0.189[0]; log p(e|m)= -3.817:-3.817[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.064]; league[0.063]; Rugby[0.062]; Bradford[0.061]; rugby[0.061]; England[0.060]; played[0.058]; Saturday[0.055]; Wigan[0.055]; Wigan[0.055]; Wigan[0.055]; premiership[0.054]; English[0.052]; Bulls[0.051]; semifinal[0.050]; Semifinal[0.049]; Beat[0.048]; Result[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = English ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.067:0.067[0]; log p(e|m)= -1.355:-1.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.065]; Rugby[0.065]; rugby[0.065]; League[0.064]; league[0.063]; played[0.057]; Bradford[0.057]; Bradford[0.057]; Wigan[0.054]; Wigan[0.054]; Wigan[0.054]; Saturday[0.053]; premiership[0.051]; semifinal[0.050]; Bulls[0.049]; Semifinal[0.048]; Beat[0.047]; Result[0.046]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEngland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEngland national rugby league team\u001b[39m <---> \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.241:0.235[\u001b[31m0.006\u001b[39m]; local()= 0.190:0.061[\u001b[31m0.129\u001b[39m]; log p(e|m)= -4.269:-0.190[\u001b[32m4.079\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRugby[0.065]; rugby[0.064]; League[0.064]; league[0.062]; Bradford[0.058]; Bradford[0.058]; played[0.057]; Wigan[0.056]; Wigan[0.056]; Wigan[0.056]; English[0.054]; Saturday[0.053]; premiership[0.053]; Bulls[0.052]; semifinal[0.050]; Semifinal[0.048]; Beat[0.047]; Result[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rugby League\u001b[39m ==> ENTITY: \u001b[32mRugby league\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrugby[0.075]; England[0.067]; league[0.064]; Bradford[0.062]; Bradford[0.062]; played[0.061]; premiership[0.061]; Wigan[0.059]; Wigan[0.059]; Wigan[0.059]; Saturday[0.057]; Bulls[0.055]; English[0.053]; semifinal[0.053]; Semifinal[0.053]; Beat[0.052]; Result[0.049]; \t\n\r [=============================>................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 8s27ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3158/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1010testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.S\u001b[39m ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.078:0.078[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncontinental[0.046]; season[0.045]; Washington[0.045]; States[0.044]; Department[0.043]; time[0.042]; Paul[0.042]; senior[0.041]; probably[0.041]; probably[0.041]; expected[0.041]; Mexican[0.041]; Mexican[0.041]; Mexican[0.041]; Secretary[0.041]; November[0.041]; lift[0.041]; today[0.041]; entering[0.041]; entering[0.041]; Agriculture[0.040]; Agriculture[0.040]; Friday[0.040]; restricted[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Michoacan\u001b[39m ==> ENTITY: \u001b[32mMichoacán\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.045]; Mexico[0.045]; Mexican[0.044]; Mexican[0.044]; Mexican[0.044]; Mexican[0.044]; area[0.043]; country[0.043]; country[0.043]; agriculture[0.042]; avocados[0.042]; avocados[0.042]; orchards[0.042]; department[0.041]; California[0.041]; Agriculture[0.041]; said[0.040]; said[0.040]; agencies[0.040]; avocado[0.039]; say[0.039]; shows[0.039]; officials[0.039]; officials[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = California ==> ENTITY: \u001b[32mCalifornia\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCalifornia[0.048]; Department[0.046]; country[0.045]; season[0.044]; Mexican[0.043]; Mexican[0.043]; Mexican[0.043]; Mexican[0.043]; states[0.042]; industry[0.042]; senior[0.041]; American[0.041]; growers[0.041]; growers[0.041]; growers[0.041]; Northern[0.040]; probably[0.040]; international[0.040]; officials[0.040]; produce[0.040]; avocados[0.040]; avocados[0.040]; avocados[0.040]; say[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dan Glickman\u001b[39m ==> ENTITY: \u001b[32mDan Glickman\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNortheastern[0.045]; Secretary[0.043]; Department[0.043]; Department[0.043]; November[0.043]; November[0.043]; Agriculture[0.042]; Agriculture[0.042]; Agriculture[0.042]; senior[0.042]; Washington[0.042]; said[0.041]; said[0.041]; said[0.041]; Northern[0.041]; February[0.041]; officials[0.041]; administration[0.041]; entering[0.041]; entering[0.041]; season[0.040]; controversial[0.040]; Paul[0.040]; possible[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.069:0.069[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCalifornia[0.044]; California[0.044]; Mexico[0.044]; Commission[0.043]; officials[0.043]; officials[0.043]; shows[0.042]; rules[0.042]; Department[0.042]; Department[0.042]; probably[0.041]; want[0.041]; expect[0.041]; issue[0.041]; issue[0.041]; issue[0.041]; significant[0.041]; data[0.041]; data[0.041]; established[0.041]; group[0.041]; group[0.041]; reconsider[0.041]; lifting[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.044]; Washington[0.044]; continental[0.043]; time[0.043]; probably[0.042]; probably[0.042]; Paul[0.042]; said[0.042]; said[0.042]; said[0.042]; trade[0.041]; Friday[0.041]; Mexican[0.041]; Mexican[0.041]; Mexican[0.041]; Mexican[0.041]; today[0.041]; November[0.041]; restricted[0.041]; Department[0.040]; Department[0.040]; likely[0.040]; Dan[0.040]; tomorrow[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = World Trade Organisation\u001b[39m ==> ENTITY: \u001b[32mWorld Trade Organization\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninternational[0.047]; export[0.046]; Northeastern[0.045]; Commission[0.045]; imports[0.042]; administration[0.042]; safeguards[0.042]; restrictions[0.041]; agriculture[0.041]; trading[0.041]; rules[0.041]; rules[0.041]; Northern[0.041]; American[0.041]; proposal[0.041]; proposal[0.041]; states[0.041]; agencies[0.040]; country[0.040]; country[0.040]; data[0.040]; established[0.040]; approved[0.040]; officials[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.074:0.074[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.046]; American[0.044]; international[0.043]; World[0.043]; California[0.043]; Mexico[0.043]; Mexico[0.043]; Commission[0.042]; officials[0.042]; officials[0.042]; data[0.041]; data[0.041]; basis[0.041]; contend[0.041]; shows[0.041]; rules[0.041]; rules[0.041]; Department[0.041]; lifting[0.040]; growing[0.040]; issue[0.040]; significant[0.040]; significant[0.040]; producers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.047]; country[0.047]; American[0.046]; international[0.045]; World[0.044]; California[0.044]; California[0.044]; Mexico[0.044]; billion[0.044]; officials[0.043]; states[0.043]; season[0.043]; basis[0.043]; basis[0.043]; contend[0.043]; ago[0.042]; rules[0.042]; Department[0.042]; February[0.042]; limited[0.042]; largely[0.042]; probably[0.042]; growing[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexican ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.397:-0.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexican[0.050]; Mexican[0.050]; Mexican[0.050]; Mexican[0.050]; country[0.043]; avocado[0.042]; continental[0.041]; largely[0.041]; avocados[0.040]; avocados[0.040]; avocados[0.040]; avocados[0.040]; international[0.040]; trade[0.040]; States[0.040]; markets[0.040]; trading[0.040]; California[0.040]; California[0.040]; administration[0.040]; states[0.039]; likely[0.039]; restrictions[0.038]; billion[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.092:0.092[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.045]; Washington[0.044]; time[0.044]; States[0.043]; continental[0.043]; officials[0.042]; expected[0.042]; Mexican[0.042]; Mexican[0.042]; Mexican[0.042]; Mexican[0.042]; Mexican[0.042]; United[0.042]; ago[0.041]; Department[0.041]; Department[0.041]; probably[0.041]; probably[0.041]; November[0.041]; Paul[0.041]; come[0.040]; mean[0.040]; possible[0.040]; restricted[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexico ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.183:-0.183[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.051]; Mexican[0.047]; Mexican[0.047]; Mexican[0.047]; Mexican[0.047]; country[0.047]; country[0.047]; international[0.044]; built[0.043]; area[0.043]; California[0.043]; American[0.043]; Michoacan[0.042]; restrictions[0.042]; opposition[0.041]; strict[0.041]; established[0.041]; allowed[0.041]; World[0.041]; growing[0.041]; agencies[0.041]; limited[0.041]; experts[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.100:0.100[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStates[0.045]; Department[0.044]; Department[0.044]; proposed[0.043]; officials[0.043]; continental[0.043]; November[0.042]; United[0.042]; Paul[0.042]; Friday[0.041]; Secretary[0.041]; likely[0.041]; today[0.041]; senior[0.041]; said[0.041]; said[0.041]; said[0.041]; time[0.041]; probably[0.041]; probably[0.041]; season[0.041]; restricted[0.040]; entering[0.040]; entering[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = California ==> ENTITY: \u001b[32mCalifornia\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.039:0.039[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nreconsider[0.045]; studying[0.045]; Mexico[0.044]; industry[0.044]; science[0.044]; delayed[0.044]; based[0.043]; probably[0.042]; Department[0.042]; administration[0.041]; administration[0.041]; ultimately[0.041]; Tom[0.041]; expect[0.041]; position[0.041]; study[0.040]; want[0.040]; issue[0.040]; issue[0.040]; issue[0.040]; change[0.040]; Administration[0.040]; officials[0.040]; officials[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.045]; country[0.045]; American[0.044]; international[0.043]; World[0.042]; California[0.042]; California[0.042]; California[0.042]; Mexico[0.042]; billion[0.042]; Commission[0.041]; officials[0.041]; data[0.041]; basis[0.041]; basis[0.041]; contend[0.041]; shows[0.041]; states[0.041]; rules[0.040]; rules[0.040]; Department[0.040]; February[0.040]; limited[0.040]; largely[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.094:0.094[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.046]; American[0.044]; international[0.043]; World[0.043]; California[0.043]; California[0.043]; billion[0.043]; officials[0.042]; states[0.042]; season[0.042]; basis[0.041]; basis[0.041]; contend[0.041]; ago[0.041]; rules[0.041]; Department[0.041]; February[0.041]; limited[0.041]; largely[0.041]; probably[0.041]; probably[0.041]; Paul[0.040]; November[0.040]; November[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexican ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.397:-0.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexican[0.051]; Mexican[0.051]; Mexican[0.051]; Mexican[0.051]; continental[0.042]; avocados[0.041]; avocados[0.041]; avocados[0.041]; avocados[0.041]; trade[0.041]; States[0.041]; markets[0.041]; administration[0.040]; Washington[0.040]; states[0.040]; likely[0.039]; Northeastern[0.039]; probably[0.039]; probably[0.039]; expected[0.038]; restricted[0.038]; February[0.038]; plan[0.038]; borders[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Agriculture Department\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Agriculture\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCommission[0.044]; pest[0.043]; agency[0.043]; agency[0.043]; agriculture[0.043]; growers[0.043]; growers[0.043]; administration[0.042]; administration[0.042]; department[0.042]; department[0.042]; approved[0.042]; study[0.042]; avocados[0.041]; established[0.041]; American[0.041]; agencies[0.041]; export[0.041]; Mexico[0.040]; Mexico[0.040]; data[0.040]; data[0.040]; officials[0.040]; eliminate[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexican ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.162:0.162[0]; log p(e|m)= -0.397:-0.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexican[0.049]; Mexican[0.049]; Mexican[0.049]; Mexican[0.049]; Mexico[0.047]; country[0.043]; country[0.043]; American[0.042]; avocado[0.041]; avocado[0.041]; largely[0.040]; avocados[0.040]; avocados[0.040]; avocados[0.040]; international[0.040]; say[0.039]; area[0.039]; trading[0.039]; California[0.039]; California[0.039]; administration[0.039]; Michoacan[0.039]; states[0.039]; restrictions[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexican ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.154:0.154[0]; log p(e|m)= -0.397:-0.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexican[0.049]; Mexican[0.049]; Mexican[0.049]; Mexico[0.047]; Mexico[0.047]; country[0.043]; country[0.043]; American[0.042]; avocado[0.041]; avocado[0.041]; avocados[0.040]; avocados[0.040]; Avocado[0.040]; international[0.040]; say[0.039]; area[0.039]; trading[0.039]; California[0.039]; group[0.039]; administration[0.039]; Michoacan[0.039]; growing[0.038]; rules[0.038]; rules[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = California ==> ENTITY: \u001b[32mCalifornia\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCalifornia[0.048]; Department[0.046]; country[0.045]; season[0.044]; Mexican[0.043]; Mexican[0.043]; Mexican[0.043]; Mexican[0.043]; states[0.042]; industry[0.042]; senior[0.041]; American[0.041]; growers[0.041]; growers[0.041]; growers[0.041]; Northern[0.040]; growing[0.040]; probably[0.040]; international[0.040]; officials[0.040]; produce[0.040]; avocados[0.040]; avocados[0.040]; avocados[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Agriculture Department\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Agriculture\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAgriculture[0.048]; Agriculture[0.048]; Department[0.046]; Northeastern[0.045]; growers[0.042]; growers[0.042]; administration[0.042]; Washington[0.042]; Secretary[0.041]; senior[0.041]; avocados[0.041]; avocados[0.041]; avocados[0.041]; avocados[0.041]; Northern[0.041]; States[0.040]; officials[0.040]; officials[0.040]; proposed[0.040]; country[0.040]; import[0.040]; season[0.040]; billion[0.040]; states[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Agriculture Department\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Agriculture\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAdministration[0.047]; agency[0.046]; administration[0.045]; department[0.044]; officials[0.043]; California[0.042]; avocado[0.042]; avocado[0.042]; available[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; investigation[0.041]; reconsider[0.041]; industry[0.040]; possible[0.040]; based[0.040]; ll[0.040]; position[0.040]; reopen[0.040]; significant[0.040]; science[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexican ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.125:0.125[0]; log p(e|m)= -0.397:-0.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexican[0.053]; Mexican[0.053]; continental[0.044]; avocados[0.043]; avocados[0.043]; avocados[0.043]; trade[0.042]; States[0.042]; markets[0.042]; Washington[0.042]; probably[0.041]; probably[0.041]; expected[0.040]; restricted[0.040]; today[0.040]; said[0.040]; said[0.040]; said[0.040]; Department[0.039]; officials[0.039]; likely[0.039]; entering[0.039]; entering[0.039]; United[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.S. Agriculture Department\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Agriculture\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAgriculture[0.049]; Agriculture[0.049]; Department[0.047]; Department[0.047]; Washington[0.043]; Secretary[0.042]; senior[0.042]; avocados[0.042]; avocados[0.042]; avocados[0.042]; States[0.041]; officials[0.041]; proposed[0.041]; import[0.041]; season[0.040]; markets[0.040]; said[0.039]; said[0.039]; said[0.039]; Mexican[0.039]; Mexican[0.039]; Mexican[0.039]; continental[0.039]; restricted[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = American ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.118:0.118[0]; log p(e|m)= -0.130:-0.130[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexican[0.044]; Mexican[0.044]; Mexican[0.044]; Mexican[0.044]; country[0.044]; country[0.044]; Mexico[0.044]; Mexico[0.044]; shows[0.042]; international[0.042]; largely[0.042]; California[0.041]; California[0.041]; California[0.041]; limited[0.040]; producers[0.040]; growing[0.040]; said[0.040]; World[0.040]; contend[0.040]; industry[0.039]; Department[0.039]; established[0.039]; scientific[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexico ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.111:0.111[0]; log p(e|m)= -0.183:-0.183[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.050]; Mexican[0.046]; Mexican[0.046]; country[0.046]; built[0.042]; area[0.042]; California[0.042]; California[0.042]; American[0.042]; Michoacan[0.041]; opposition[0.041]; strict[0.041]; ahead[0.040]; established[0.040]; allowed[0.040]; growing[0.040]; agencies[0.040]; experts[0.040]; position[0.040]; reconsider[0.040]; officials[0.040]; officials[0.040]; data[0.040]; data[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexican ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.173:0.173[0]; log p(e|m)= -0.397:-0.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexican[0.048]; Mexican[0.048]; Mexican[0.048]; Mexican[0.048]; Mexican[0.048]; Mexico[0.046]; country[0.042]; country[0.042]; American[0.041]; avocado[0.040]; avocado[0.040]; largely[0.040]; avocados[0.039]; avocados[0.039]; avocados[0.039]; international[0.039]; say[0.039]; area[0.039]; trading[0.039]; California[0.039]; California[0.039]; administration[0.038]; Michoacan[0.038]; states[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexican ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.141:0.141[0]; log p(e|m)= -0.397:-0.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexican[0.050]; Mexican[0.050]; Mexican[0.050]; Mexican[0.050]; country[0.044]; avocado[0.042]; largely[0.041]; avocados[0.041]; avocados[0.041]; avocados[0.041]; international[0.040]; trade[0.040]; markets[0.040]; trading[0.040]; California[0.040]; California[0.040]; administration[0.040]; states[0.039]; likely[0.039]; restrictions[0.039]; billion[0.039]; Northeastern[0.039]; probably[0.039]; probably[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexican ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.397:-0.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexican[0.050]; Mexican[0.050]; Mexico[0.048]; Mexico[0.048]; country[0.044]; American[0.043]; avocado[0.042]; avocado[0.042]; avocados[0.041]; avocados[0.041]; Avocado[0.041]; international[0.041]; say[0.040]; area[0.040]; trading[0.040]; California[0.040]; group[0.040]; administration[0.040]; Michoacan[0.040]; growing[0.039]; rules[0.038]; rules[0.038]; shows[0.038]; lifting[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexican ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.397:-0.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexican[0.055]; Mexican[0.055]; continental[0.046]; avocados[0.045]; avocados[0.045]; avocados[0.045]; trade[0.044]; States[0.044]; markets[0.044]; Washington[0.043]; likely[0.043]; probably[0.042]; probably[0.042]; expected[0.042]; restricted[0.041]; said[0.041]; said[0.041]; said[0.041]; November[0.041]; today[0.041]; Department[0.041]; Department[0.041]; officials[0.041]; \t\n\r [=============================>................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 7s844ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3189/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 965testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = FIFA\u001b[39m ==> ENTITY: \u001b[32mFIFA\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.046]; Soccer[0.046]; Fifa[0.045]; game[0.044]; game[0.044]; game[0.044]; Brazil[0.043]; national[0.043]; match[0.042]; match[0.042]; team[0.042]; Real[0.042]; Madrid[0.040]; Madrid[0.040]; European[0.040]; Brazilian[0.040]; Spanish[0.040]; Netherlands[0.040]; division[0.039]; Deportivo[0.039]; suspension[0.039]; suspended[0.039]; tour[0.039]; world[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazilian ==> ENTITY: \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.161:0.161[0]; log p(e|m)= -0.273:-0.273[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpanish[0.046]; Brazil[0.046]; soccer[0.045]; Soccer[0.044]; team[0.044]; match[0.043]; match[0.043]; Fifa[0.041]; Fifa[0.041]; national[0.041]; game[0.041]; game[0.041]; game[0.041]; Mauro[0.041]; division[0.041]; Deportivo[0.041]; Silva[0.041]; Silva[0.041]; Silva[0.041]; Real[0.040]; European[0.040]; Netherlands[0.039]; Coruna[0.039]; world[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEuropean\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUEFA\u001b[39m <---> \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.245:0.239[\u001b[31m0.006\u001b[39m]; local()= 0.162:0.081[\u001b[31m0.081\u001b[39m]; log p(e|m)= -3.576:-0.256[\u001b[32m3.319\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.045]; Soccer[0.044]; Fifa[0.043]; Fifa[0.043]; match[0.043]; match[0.043]; team[0.043]; national[0.042]; game[0.042]; game[0.042]; game[0.042]; tour[0.041]; Madrid[0.041]; Madrid[0.041]; world[0.041]; Spanish[0.041]; suspended[0.041]; suspension[0.040]; division[0.040]; Coruna[0.040]; Brazil[0.040]; Netherlands[0.040]; Real[0.040]; Deportivo[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Real Madrid\u001b[39m ==> ENTITY: \u001b[32mReal Madrid C.F.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.047]; soccer[0.047]; Madrid[0.044]; team[0.044]; Deportivo[0.043]; game[0.043]; game[0.043]; game[0.043]; match[0.043]; match[0.043]; Coruna[0.042]; European[0.041]; Spanish[0.041]; division[0.041]; midfielder[0.041]; Brazil[0.040]; national[0.040]; Brazilian[0.040]; Fifa[0.040]; Fifa[0.040]; Wednesday[0.039]; Mauro[0.039]; lost[0.038]; tour[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Silva\u001b[39m ==> ENTITY: \u001b[32mMauro Silva\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDeportivo[0.044]; Soccer[0.044]; Silva[0.043]; Silva[0.043]; Mauro[0.043]; team[0.043]; Brazilian[0.043]; Brazil[0.042]; soccer[0.042]; midfielder[0.042]; Madrid[0.041]; Madrid[0.041]; absence[0.041]; Real[0.041]; suspension[0.041]; Fifa[0.041]; Fifa[0.041]; Coruna[0.041]; match[0.041]; match[0.041]; game[0.041]; game[0.041]; game[0.041]; division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spanish ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.127:0.127[0]; log p(e|m)= -1.241:-1.241[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.045]; Soccer[0.044]; match[0.044]; match[0.044]; Brazil[0.043]; Madrid[0.042]; Madrid[0.042]; national[0.042]; game[0.042]; game[0.042]; game[0.042]; European[0.042]; team[0.042]; Fifa[0.041]; Fifa[0.041]; Netherlands[0.041]; division[0.041]; world[0.040]; Real[0.040]; Brazilian[0.040]; Deportivo[0.040]; tour[0.039]; Coruna[0.039]; midfielder[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBrazil\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBrazil national football team\u001b[39m <---> \u001b[32mBrazil\u001b[39m\t\nSCORES: global= 0.256:0.249[\u001b[31m0.007\u001b[39m]; local()= 0.188:0.162[\u001b[31m0.026\u001b[39m]; log p(e|m)= -2.590:-0.213[\u001b[32m2.377\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.045]; Soccer[0.045]; team[0.044]; match[0.043]; match[0.043]; Brazilian[0.043]; Fifa[0.042]; Fifa[0.042]; game[0.042]; game[0.042]; game[0.042]; national[0.042]; Silva[0.042]; Silva[0.042]; Silva[0.042]; Spanish[0.041]; Madrid[0.041]; Madrid[0.041]; European[0.041]; Netherlands[0.040]; world[0.039]; Real[0.039]; division[0.039]; Mauro[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = FIFA\u001b[39m ==> ENTITY: \u001b[32mFIFA\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.046]; Soccer[0.046]; Fifa[0.045]; game[0.044]; game[0.044]; game[0.044]; Brazil[0.043]; national[0.043]; match[0.042]; match[0.042]; team[0.042]; Real[0.042]; Madrid[0.040]; Madrid[0.040]; European[0.040]; Brazilian[0.040]; Spanish[0.040]; Netherlands[0.040]; division[0.039]; Deportivo[0.039]; suspension[0.039]; suspended[0.039]; tour[0.039]; world[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Madrid ==> ENTITY: \u001b[32mMadrid\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMadrid[0.046]; soccer[0.046]; Brazil[0.044]; Soccer[0.044]; national[0.043]; Spanish[0.043]; midfielder[0.042]; Coruna[0.042]; game[0.041]; game[0.041]; game[0.041]; team[0.041]; match[0.041]; match[0.041]; saying[0.041]; European[0.041]; Deportivo[0.041]; Wednesday[0.041]; Fifa[0.040]; Fifa[0.040]; Brazilian[0.040]; tour[0.040]; governing[0.039]; key[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mRussia\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRussia national football team\u001b[39m <---> \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.250:0.237[\u001b[31m0.013\u001b[39m]; local()= 0.173:0.067[\u001b[31m0.106\u001b[39m]; log p(e|m)= -3.540:-0.203[\u001b[32m3.337\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.046]; team[0.046]; soccer[0.046]; match[0.045]; match[0.045]; game[0.045]; game[0.045]; game[0.045]; European[0.042]; national[0.042]; Netherlands[0.041]; Fifa[0.040]; Fifa[0.040]; governing[0.040]; key[0.040]; midfielder[0.040]; world[0.040]; Brazil[0.039]; Saturday[0.039]; Saturday[0.039]; Wednesday[0.039]; division[0.039]; lost[0.039]; absence[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mauro Silva\u001b[39m ==> ENTITY: \u001b[32mMauro Silva\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDeportivo[0.044]; Soccer[0.044]; Silva[0.043]; Silva[0.043]; team[0.043]; Brazilian[0.043]; Brazil[0.042]; soccer[0.042]; midfielder[0.042]; Madrid[0.042]; Madrid[0.042]; absence[0.041]; Real[0.041]; suspension[0.041]; Fifa[0.041]; Fifa[0.041]; Coruna[0.041]; match[0.041]; match[0.041]; game[0.041]; game[0.041]; game[0.041]; division[0.039]; Spanish[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Netherlands ==> ENTITY: \u001b[32mNetherlands national football team\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.142:0.142[0]; log p(e|m)= -2.659:-2.659[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.046]; match[0.046]; match[0.046]; team[0.046]; soccer[0.046]; national[0.044]; Fifa[0.043]; Fifa[0.043]; game[0.043]; game[0.043]; game[0.043]; European[0.042]; Brazil[0.041]; governing[0.041]; midfielder[0.040]; tour[0.040]; saying[0.039]; world[0.039]; Madrid[0.039]; Madrid[0.039]; Wednesday[0.039]; suspension[0.039]; Russia[0.038]; lost[0.038]; \t\n\r [=============================>................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 7s744ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3202/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1121testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.077:0.077[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.049]; airport[0.047]; Britain[0.046]; Flight[0.043]; plane[0.043]; Stansted[0.043]; flying[0.043]; week[0.042]; following[0.042]; Tuesday[0.041]; Amman[0.041]; airliner[0.041]; landed[0.041]; people[0.041]; Jordan[0.040]; air[0.040]; Saturday[0.040]; seven[0.040]; appeared[0.040]; north[0.040]; hijack[0.039]; taking[0.039]; hijacking[0.039]; entered[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Amman\u001b[39m ==> ENTITY: \u001b[32mAmman\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJordan[0.049]; north[0.045]; Iraq[0.045]; Sudan[0.044]; airport[0.044]; Sudanese[0.044]; Khartoum[0.044]; London[0.044]; London[0.044]; Iraqi[0.044]; people[0.044]; life[0.043]; Britain[0.043]; Britain[0.043]; Britain[0.043]; week[0.043]; seven[0.042]; seven[0.042]; Iraqis[0.042]; means[0.042]; appeared[0.042]; following[0.042]; following[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.113:0.113[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.055]; London[0.044]; Iraq[0.044]; freed[0.043]; held[0.042]; Tuesday[0.042]; aircraft[0.042]; following[0.042]; week[0.042]; law[0.041]; English[0.041]; considered[0.041]; Jordan[0.041]; receive[0.041]; seven[0.040]; seven[0.040]; entered[0.040]; claimed[0.040]; Sudan[0.040]; asylum[0.040]; asylum[0.040]; widespread[0.040]; life[0.040]; men[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jordan ==> ENTITY: \u001b[32mJordan\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.175:0.175[0]; log p(e|m)= -0.536:-0.536[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmman[0.046]; Iraq[0.045]; Sudan[0.045]; Britain[0.044]; Britain[0.044]; Britain[0.044]; Iraqi[0.043]; week[0.042]; Sudanese[0.042]; Tuesday[0.041]; appearance[0.041]; following[0.041]; following[0.041]; Khartoum[0.041]; considered[0.041]; Iraqis[0.040]; widespread[0.040]; seven[0.040]; seven[0.040]; saying[0.040]; political[0.040]; appeared[0.040]; north[0.040]; entered[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Khartoum ==> ENTITY: \u001b[32mKhartoum\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.146:0.146[0]; log p(e|m)= -0.040:-0.040[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSudanese[0.046]; Sudan[0.046]; north[0.045]; Jordan[0.044]; Amman[0.043]; Iraq[0.042]; airport[0.042]; following[0.042]; following[0.042]; Britain[0.042]; Britain[0.042]; Britain[0.042]; Iraqi[0.041]; freed[0.041]; Tuesday[0.041]; London[0.041]; London[0.041]; saying[0.040]; plane[0.040]; people[0.040]; week[0.040]; appearance[0.040]; aircraft[0.040]; Iraqis[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Stansted ==> ENTITY: \u001b[32mLondon Stansted Airport\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.171:0.171[0]; log p(e|m)= -0.267:-0.267[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnorth[0.054]; London[0.048]; London[0.048]; Britain[0.046]; Britain[0.046]; Britain[0.046]; airport[0.045]; seven[0.043]; seven[0.043]; aircraft[0.043]; week[0.043]; people[0.042]; English[0.042]; police[0.042]; Saturday[0.041]; Iraq[0.041]; Tuesday[0.041]; Amman[0.041]; air[0.041]; Flight[0.041]; plane[0.041]; landed[0.040]; maximum[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqis ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.126:0.126[0]; log p(e|m)= -0.812:-0.812[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraqi[0.050]; Britain[0.046]; Jordan[0.046]; Sudan[0.045]; Sudanese[0.045]; London[0.045]; Amman[0.045]; people[0.043]; Khartoum[0.043]; seven[0.041]; week[0.041]; appeared[0.040]; following[0.040]; held[0.040]; freed[0.040]; hostages[0.039]; businessmen[0.039]; plane[0.039]; men[0.039]; Tuesday[0.039]; Saturday[0.039]; preliminary[0.038]; landed[0.038]; taking[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.106:0.106[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.054]; London[0.044]; London[0.044]; Iraq[0.043]; freed[0.043]; flying[0.042]; held[0.042]; Tuesday[0.042]; following[0.042]; week[0.041]; piracy[0.041]; Saturday[0.041]; seven[0.041]; Flight[0.041]; Jordan[0.040]; plane[0.040]; Iraqi[0.040]; landed[0.040]; aboard[0.040]; claimed[0.040]; Sudan[0.040]; people[0.040]; asylum[0.040]; air[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.105:0.105[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.047]; airport[0.046]; aircraft[0.046]; Britain[0.045]; Britain[0.045]; Britain[0.045]; Flight[0.042]; plane[0.042]; Stansted[0.042]; flying[0.042]; week[0.041]; following[0.041]; following[0.041]; saying[0.041]; Tuesday[0.040]; Amman[0.040]; airliner[0.040]; landed[0.040]; people[0.039]; English[0.039]; Jordan[0.039]; claimed[0.039]; air[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sudan ==> ENTITY: \u001b[32mSudan\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.285:-0.285[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSudanese[0.060]; Khartoum[0.057]; Iraq[0.049]; Iraqi[0.046]; Iraqis[0.045]; Amman[0.044]; Britain[0.043]; Britain[0.043]; Britain[0.043]; Jordan[0.040]; imprisonment[0.038]; imprisonment[0.038]; asylum[0.038]; asylum[0.038]; political[0.038]; north[0.038]; widespread[0.038]; piracy[0.038]; businessmen[0.038]; freed[0.038]; hijacking[0.037]; hostages[0.037]; London[0.037]; London[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.118:0.118[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraqis[0.048]; Britain[0.045]; Jordan[0.045]; Sudan[0.045]; Sudanese[0.044]; Amman[0.044]; freed[0.043]; people[0.043]; Khartoum[0.043]; held[0.042]; appeared[0.042]; week[0.041]; following[0.040]; political[0.040]; claimed[0.040]; aboard[0.040]; ages[0.040]; plane[0.039]; accused[0.039]; north[0.039]; charged[0.039]; charged[0.039]; landed[0.039]; Tuesday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Britain ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.138:0.138[0]; log p(e|m)= -0.611:-0.611[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritain[0.054]; Britain[0.054]; London[0.044]; freed[0.043]; held[0.042]; Tuesday[0.041]; following[0.041]; following[0.041]; week[0.041]; piracy[0.041]; Iraq[0.041]; law[0.041]; English[0.041]; considered[0.040]; Sudan[0.040]; Saturday[0.040]; Jordan[0.040]; seven[0.040]; seven[0.040]; entered[0.039]; claimed[0.039]; people[0.039]; asylum[0.039]; asylum[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sudanese ==> ENTITY: \u001b[32mSudan\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.198:-0.198[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSudan[0.048]; Britain[0.046]; Britain[0.046]; Khartoum[0.045]; people[0.044]; asylum[0.043]; Iraqis[0.043]; following[0.042]; Jordan[0.042]; political[0.042]; Iraq[0.042]; north[0.042]; Iraqi[0.042]; Amman[0.040]; businessmen[0.040]; London[0.040]; London[0.040]; seven[0.040]; appeared[0.039]; claimed[0.039]; means[0.039]; ranged[0.038]; hostages[0.038]; saying[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJordan[0.046]; Sudan[0.045]; Sudanese[0.044]; Amman[0.044]; Khartoum[0.044]; Britain[0.043]; Britain[0.043]; Britain[0.043]; seven[0.042]; seven[0.042]; saying[0.041]; following[0.041]; following[0.041]; accused[0.041]; widespread[0.041]; trial[0.041]; aircraft[0.040]; north[0.040]; men[0.040]; week[0.040]; ordered[0.040]; appearance[0.040]; political[0.039]; Tuesday[0.039]; \t\n\r [=============================>................] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 7s697ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3217/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1052testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Northampton ==> ENTITY: \u001b[32mNorthampton\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.511:-0.511[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.049]; Martin[0.047]; tournament[0.045]; play[0.044]; won[0.043]; Australia[0.043]; final[0.042]; British[0.042]; British[0.042]; British[0.042]; scores[0.042]; Saturday[0.042]; Robert[0.041]; Robert[0.041]; Colin[0.041]; round[0.041]; Ian[0.040]; Final[0.040]; Argentina[0.040]; Italy[0.039]; Montgomerie[0.038]; Scores[0.038]; Woosnam[0.038]; golf[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.050]; Ireland[0.045]; Mark[0.044]; Argentina[0.044]; Stephen[0.043]; David[0.042]; Colin[0.042]; Phillip[0.042]; Ian[0.042]; Italy[0.042]; Spain[0.041]; Spain[0.041]; Sweden[0.041]; Sweden[0.041]; Sweden[0.041]; Peter[0.041]; Peter[0.041]; Peter[0.041]; France[0.040]; Adam[0.040]; Robert[0.040]; Mitchell[0.040]; Mike[0.040]; Philip[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Robert Coles ==> ENTITY: \u001b[32mRobert Coles (golfer)\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.207:0.207[0]; log p(e|m)= -0.614:-0.614[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMasters[0.049]; Stephen[0.047]; David[0.047]; Robert[0.047]; Montgomerie[0.047]; Martin[0.046]; tournament[0.046]; Hedblom[0.046]; British[0.046]; British[0.046]; Philip[0.046]; round[0.045]; golf[0.045]; England[0.045]; Lebouc[0.044]; Phillip[0.044]; Mark[0.044]; France[0.044]; Australia[0.044]; Australia[0.044]; Australia[0.044]; Ireland[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mark Roe\u001b[39m ==> ENTITY: \u001b[32mMark Roe\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.053:0.053[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKlas[0.044]; David[0.043]; Spain[0.042]; Spain[0.042]; Montgomerie[0.042]; Joakim[0.042]; Ireland[0.042]; France[0.042]; Argentina[0.042]; Sweden[0.042]; Sweden[0.042]; Sweden[0.042]; Lebouc[0.041]; Australia[0.041]; Australia[0.041]; Colin[0.041]; Stephen[0.041]; Hunter[0.041]; Clayton[0.041]; Robert[0.041]; Hedblom[0.041]; Linhart[0.041]; Woosnam[0.041]; Mike[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.192:0.192[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.052]; Spain[0.052]; Argentina[0.047]; tournament[0.046]; won[0.045]; Italy[0.044]; Jose[0.043]; England[0.043]; Ireland[0.043]; Miguel[0.043]; Miguel[0.043]; final[0.043]; Pedro[0.042]; France[0.042]; Eriksson[0.042]; play[0.042]; Martin[0.042]; Jimenez[0.042]; Golf[0.041]; Sweden[0.041]; Sweden[0.041]; round[0.041]; Mark[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.192:0.192[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.049]; Spain[0.049]; Argentina[0.045]; England[0.045]; tournament[0.044]; won[0.043]; Italy[0.043]; Jose[0.042]; Ireland[0.041]; Miguel[0.041]; Miguel[0.041]; final[0.041]; Pedro[0.041]; France[0.040]; Eriksson[0.040]; play[0.040]; Martin[0.040]; Jimenez[0.040]; Australia[0.040]; Golf[0.040]; Sweden[0.040]; Sweden[0.040]; round[0.040]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.184:0.184[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.050]; Spain[0.050]; Argentina[0.046]; won[0.044]; Italy[0.043]; Jose[0.042]; Ireland[0.042]; Miguel[0.041]; Miguel[0.041]; Pedro[0.041]; David[0.041]; France[0.040]; Eriksson[0.040]; play[0.040]; Martin[0.040]; Jimenez[0.040]; Australia[0.040]; Australia[0.040]; Sweden[0.040]; Sweden[0.040]; Sweden[0.040]; Mark[0.039]; Robert[0.039]; Stephen[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Robert Allenby\u001b[39m ==> ENTITY: \u001b[32mRobert Allenby\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.047]; round[0.046]; won[0.044]; golf[0.044]; play[0.043]; Masters[0.043]; Masters[0.043]; Golf[0.043]; final[0.042]; Allenby[0.041]; Colin[0.041]; Martin[0.041]; Montgomerie[0.041]; Australia[0.041]; Ian[0.041]; Klas[0.041]; scores[0.040]; hole[0.040]; Robert[0.040]; Lebouc[0.040]; Philip[0.040]; British[0.039]; British[0.039]; British[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.132:0.132[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.050]; tournament[0.045]; England[0.045]; Argentina[0.044]; won[0.044]; Ireland[0.043]; Italy[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; play[0.042]; Golf[0.042]; Eriksson[0.042]; round[0.041]; France[0.040]; Australia[0.040]; Australia[0.040]; Australia[0.040]; final[0.040]; golf[0.039]; Final[0.039]; Martin[0.038]; scores[0.038]; Klas[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Klas Eriksson\u001b[39m ==> ENTITY: \u001b[32mKlas Eriksson\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.046]; Sweden[0.046]; Sweden[0.046]; Hedblom[0.044]; play[0.043]; won[0.043]; Montgomerie[0.043]; Pyman[0.042]; British[0.041]; Lebouc[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Woosnam[0.041]; Joakim[0.040]; Argentina[0.040]; Ireland[0.040]; France[0.040]; Australia[0.040]; Australia[0.040]; Australia[0.040]; Allenby[0.040]; Allenby[0.040]; hole[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jose Coceres\u001b[39m ==> ENTITY: \u001b[32mJosé Cóceres\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.045]; golf[0.045]; Masters[0.045]; Masters[0.045]; Spain[0.044]; Spain[0.044]; Spain[0.044]; Argentina[0.044]; tournament[0.043]; Sweden[0.042]; Sweden[0.042]; round[0.041]; Italy[0.041]; Klas[0.040]; England[0.040]; Miguel[0.040]; Miguel[0.040]; Ireland[0.040]; won[0.040]; play[0.040]; France[0.039]; Pedro[0.039]; Jimenez[0.039]; scores[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Costantino Rocca\u001b[39m ==> ENTITY: \u001b[32mCostantino Rocca\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngolf[0.046]; tournament[0.046]; Golf[0.044]; round[0.043]; Masters[0.043]; Masters[0.043]; final[0.043]; Montgomerie[0.043]; Woosnam[0.042]; scores[0.042]; Lebouc[0.042]; hole[0.041]; play[0.041]; Ireland[0.041]; British[0.040]; British[0.040]; British[0.040]; Klas[0.040]; Allenby[0.040]; Allenby[0.040]; won[0.040]; England[0.039]; Italy[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.125:0.125[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.050]; Ireland[0.045]; Mark[0.044]; Argentina[0.044]; Stephen[0.043]; David[0.043]; Colin[0.042]; Phillip[0.042]; Ian[0.042]; Spain[0.041]; Sweden[0.041]; Sweden[0.041]; Sweden[0.041]; Peter[0.041]; Peter[0.041]; Peter[0.041]; France[0.040]; Adam[0.040]; Robert[0.040]; Mitchell[0.040]; Mike[0.040]; Philip[0.040]; Jose[0.039]; Hunter[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.125:0.125[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.049]; Ireland[0.047]; tournament[0.046]; Argentina[0.046]; round[0.044]; final[0.044]; won[0.044]; Italy[0.044]; Martin[0.043]; Spain[0.043]; Spain[0.043]; Sweden[0.043]; Sweden[0.043]; Masters[0.043]; Masters[0.043]; Ian[0.043]; play[0.043]; Peter[0.043]; France[0.042]; Robert[0.042]; Robert[0.042]; Colin[0.041]; Saturday[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Philip Walton\u001b[39m ==> ENTITY: \u001b[32mPhilip Walton\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMasters[0.048]; round[0.046]; Ireland[0.045]; golf[0.045]; Sweden[0.045]; Sweden[0.045]; Sweden[0.045]; tournament[0.045]; final[0.044]; play[0.044]; Colin[0.043]; British[0.042]; British[0.042]; Hunter[0.042]; Spain[0.042]; Spain[0.042]; Spain[0.042]; Ian[0.042]; David[0.042]; Stephen[0.042]; won[0.042]; Hedblom[0.042]; Lebouc[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iain Pyman\u001b[39m ==> ENTITY: \u001b[32mIain Pyman\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.127:0.127[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.046]; Sweden[0.046]; Sweden[0.046]; Hedblom[0.045]; Lebouc[0.045]; Colin[0.045]; Montgomerie[0.045]; Klas[0.044]; France[0.044]; Mitchell[0.043]; Walton[0.043]; Peter[0.043]; Peter[0.043]; Peter[0.043]; Spain[0.043]; Ireland[0.043]; Roe[0.042]; Hunter[0.042]; Joakim[0.042]; David[0.042]; Antoine[0.042]; Philip[0.042]; Mark[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.190:-0.190[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.047]; play[0.045]; tournament[0.044]; Northampton[0.043]; Argentina[0.043]; British[0.042]; British[0.042]; British[0.042]; final[0.042]; Martin[0.042]; Spain[0.042]; Spain[0.042]; Colin[0.041]; won[0.041]; France[0.041]; Ian[0.041]; Italy[0.041]; Robert[0.041]; Robert[0.041]; Sweden[0.040]; Saturday[0.040]; scores[0.040]; Final[0.039]; round[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Phillip Price\u001b[39m ==> ENTITY: \u001b[32mPhillip Price\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.116:0.116[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplay[0.044]; Montgomerie[0.043]; Woosnam[0.042]; Stephen[0.042]; David[0.042]; Mitchell[0.042]; Sweden[0.042]; Sweden[0.042]; Sweden[0.042]; Lebouc[0.042]; Robert[0.042]; Mark[0.042]; Hedblom[0.042]; Ireland[0.042]; won[0.041]; Klas[0.041]; Philip[0.041]; Colin[0.041]; Mike[0.041]; Clayton[0.041]; Australia[0.041]; Australia[0.041]; Coles[0.041]; Jose[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.146:0.146[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArgentina[0.047]; Ireland[0.045]; England[0.045]; Italy[0.044]; Spain[0.042]; Spain[0.042]; Spain[0.042]; tournament[0.042]; Australia[0.042]; Australia[0.042]; Australia[0.042]; final[0.042]; Sweden[0.041]; Sweden[0.041]; Sweden[0.041]; round[0.041]; Mike[0.040]; play[0.040]; won[0.040]; Martin[0.040]; Mark[0.040]; British[0.040]; British[0.040]; British[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Miguel Angel Jimenez\u001b[39m ==> ENTITY: \u001b[32mMiguel Ángel Jiménez\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.218:0.218[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngolf[0.046]; tournament[0.044]; Masters[0.044]; Masters[0.044]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Golf[0.043]; round[0.042]; play[0.042]; Sweden[0.042]; Sweden[0.042]; hole[0.041]; final[0.041]; Miguel[0.040]; Jose[0.040]; Martin[0.040]; Montgomerie[0.040]; Woosnam[0.040]; Argentina[0.040]; won[0.040]; British[0.039]; British[0.039]; British[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Miguel Angel Martin\u001b[39m ==> ENTITY: \u001b[32mMiguel Ángel Martín (golfer)\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMasters[0.045]; Masters[0.045]; tournament[0.044]; golf[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; round[0.043]; Argentina[0.042]; Jose[0.042]; Miguel[0.042]; play[0.042]; Golf[0.041]; Sweden[0.041]; Sweden[0.041]; final[0.041]; Klas[0.041]; Woosnam[0.040]; Pedro[0.040]; hole[0.040]; England[0.040]; Lebouc[0.040]; scores[0.040]; Montgomerie[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.051]; Sweden[0.051]; Argentina[0.045]; won[0.045]; Ireland[0.044]; Italy[0.044]; Spain[0.044]; Spain[0.044]; Spain[0.044]; play[0.043]; Eriksson[0.042]; France[0.040]; Australia[0.040]; Australia[0.040]; Australia[0.040]; Martin[0.039]; Klas[0.039]; Joakim[0.039]; David[0.039]; Colin[0.038]; Pedro[0.037]; Woosnam[0.037]; Adam[0.037]; Stephen[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Argentina ==> ENTITY: \u001b[32mArgentina\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.178:0.178[0]; log p(e|m)= -0.300:-0.300[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.045]; Spain[0.045]; Spain[0.045]; Italy[0.044]; tournament[0.044]; England[0.044]; Ireland[0.044]; Australia[0.042]; Australia[0.042]; Australia[0.042]; France[0.042]; Jose[0.041]; Miguel[0.041]; Miguel[0.041]; won[0.041]; Sweden[0.040]; Sweden[0.040]; Pedro[0.040]; final[0.040]; round[0.040]; play[0.040]; Eriksson[0.039]; Saturday[0.039]; Martin[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = David Gilford\u001b[39m ==> ENTITY: \u001b[32mDavid Gilford\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.119:0.119[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.044]; Sweden[0.044]; Sweden[0.044]; Adam[0.044]; Mark[0.043]; Klas[0.043]; Hedblom[0.043]; Colin[0.043]; Montgomerie[0.042]; Peter[0.041]; Peter[0.041]; Peter[0.041]; Stephen[0.041]; Robert[0.041]; Joakim[0.041]; Lebouc[0.041]; France[0.041]; Philip[0.040]; Ireland[0.040]; Australia[0.040]; Australia[0.040]; Mike[0.040]; Spain[0.040]; Phillip[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Peter Hedblom\u001b[39m ==> ENTITY: \u001b[32mPeter Hedblom\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.102:0.102[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.045]; Sweden[0.045]; Sweden[0.045]; Lebouc[0.045]; Montgomerie[0.043]; Klas[0.043]; Colin[0.043]; Antoine[0.042]; Peter[0.042]; Peter[0.042]; Robert[0.042]; Philip[0.041]; Roe[0.041]; Spain[0.041]; Australia[0.040]; Australia[0.040]; Pyman[0.040]; Stephen[0.040]; Linhart[0.040]; Phillip[0.040]; Iain[0.040]; Pedro[0.040]; Mike[0.040]; Adam[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIreland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIreland\u001b[39m <---> \u001b[32mRepublic of Ireland\u001b[39m\t\nSCORES: global= 0.254:0.251[\u001b[31m0.003\u001b[39m]; local()= 0.158:0.130[\u001b[31m0.028\u001b[39m]; log p(e|m)= -0.566:-1.505[\u001b[31m0.939\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.044]; Australia[0.044]; Australia[0.044]; Argentina[0.043]; British[0.043]; British[0.043]; Italy[0.042]; Martin[0.042]; play[0.042]; tournament[0.042]; Colin[0.041]; Stephen[0.041]; won[0.041]; final[0.041]; Spain[0.041]; Spain[0.041]; Spain[0.041]; Ian[0.041]; France[0.041]; McAllister[0.040]; Sweden[0.040]; Sweden[0.040]; Sweden[0.040]; David[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ian Woosnam\u001b[39m ==> ENTITY: \u001b[32mIan Woosnam\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngolf[0.047]; tournament[0.047]; Golf[0.046]; round[0.044]; Montgomerie[0.044]; won[0.043]; Masters[0.043]; Masters[0.043]; play[0.042]; England[0.042]; final[0.042]; Allenby[0.041]; Allenby[0.041]; Klas[0.040]; Martin[0.040]; British[0.040]; British[0.040]; British[0.040]; Mark[0.040]; Colin[0.040]; Lebouc[0.039]; Sweden[0.039]; Sweden[0.039]; hole[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Allenby\u001b[39m ==> ENTITY: \u001b[32mRobert Allenby\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.049]; round[0.048]; won[0.046]; golf[0.045]; play[0.045]; Masters[0.045]; Masters[0.045]; Golf[0.045]; final[0.044]; Allenby[0.043]; Colin[0.043]; Martin[0.043]; Montgomerie[0.043]; Australia[0.043]; Ian[0.042]; Klas[0.042]; scores[0.042]; hole[0.042]; Robert[0.042]; Robert[0.042]; Lebouc[0.041]; Phillip[0.041]; Philip[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Adam Hunter ==> ENTITY: \u001b[32mAdam Hunter (golfer)\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.116:0.116[0]; log p(e|m)= -1.162:-1.162[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIain[0.046]; Ian[0.045]; Colin[0.045]; David[0.045]; Mark[0.044]; play[0.044]; Stephen[0.043]; Robert[0.043]; Phillip[0.042]; Philip[0.042]; Peter[0.042]; Peter[0.042]; Peter[0.042]; Australia[0.042]; Australia[0.042]; Rocca[0.040]; Mike[0.040]; Klas[0.039]; won[0.039]; Ireland[0.039]; Spain[0.039]; Spain[0.039]; Woosnam[0.039]; Hedblom[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = British Masters\u001b[39m ==> ENTITY: \u001b[32mBritish Masters\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.048]; golf[0.048]; Masters[0.047]; tournament[0.046]; won[0.045]; round[0.045]; British[0.045]; British[0.045]; Robert[0.044]; Martin[0.043]; Colin[0.043]; Ian[0.043]; England[0.043]; Montgomerie[0.042]; Woosnam[0.042]; Jimenez[0.042]; play[0.042]; final[0.042]; Saturday[0.041]; Spain[0.041]; Spain[0.041]; Sweden[0.041]; Lebouc[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = British Masters\u001b[39m ==> ENTITY: \u001b[32mBritish Masters\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.045]; golf[0.045]; Ireland[0.045]; Masters[0.044]; tournament[0.044]; won[0.043]; round[0.043]; British[0.042]; British[0.042]; Robert[0.042]; Robert[0.042]; Peter[0.041]; Mitchell[0.041]; Philip[0.041]; Martin[0.041]; Colin[0.041]; Ian[0.040]; England[0.040]; Montgomerie[0.040]; Woosnam[0.040]; Jimenez[0.040]; play[0.040]; final[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Colin Montgomerie\u001b[39m ==> ENTITY: \u001b[32mColin Montgomerie\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngolf[0.049]; tournament[0.046]; Golf[0.045]; Masters[0.042]; Masters[0.042]; won[0.042]; Hedblom[0.042]; Woosnam[0.042]; round[0.042]; play[0.042]; British[0.042]; British[0.042]; British[0.042]; final[0.041]; David[0.041]; hole[0.040]; Ian[0.040]; Ireland[0.040]; Klas[0.040]; scores[0.040]; Spain[0.040]; Spain[0.040]; Spain[0.040]; Mark[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pedro Linhart\u001b[39m ==> ENTITY: \u001b[32mPedro Linhart\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.104:0.104[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.044]; Spain[0.044]; Spain[0.044]; Klas[0.043]; play[0.043]; Jose[0.043]; won[0.042]; Sweden[0.042]; Sweden[0.042]; Sweden[0.042]; Miguel[0.042]; Miguel[0.042]; Argentina[0.042]; Hedblom[0.041]; Joakim[0.041]; Lebouc[0.041]; Jimenez[0.041]; Australia[0.041]; Australia[0.041]; Australia[0.041]; hole[0.040]; Ian[0.040]; Rocca[0.040]; Mike[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Joakim Haeggman\u001b[39m ==> ENTITY: \u001b[32mJoakim Haeggman\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.047]; Sweden[0.047]; golf[0.045]; play[0.044]; Masters[0.044]; Masters[0.044]; tournament[0.044]; Golf[0.044]; round[0.042]; Montgomerie[0.042]; won[0.042]; England[0.041]; Klas[0.041]; Lebouc[0.040]; Woosnam[0.040]; final[0.040]; Eriksson[0.039]; Roe[0.039]; Ireland[0.039]; Final[0.039]; Spain[0.039]; Spain[0.039]; Spain[0.039]; France[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.045]; Ireland[0.045]; Australia[0.045]; Australia[0.045]; England[0.045]; Argentina[0.044]; Spain[0.043]; Spain[0.043]; Spain[0.043]; won[0.043]; France[0.042]; final[0.042]; round[0.042]; play[0.041]; Sweden[0.040]; Sweden[0.040]; Golf[0.040]; Eriksson[0.040]; Northampton[0.039]; Final[0.039]; British[0.039]; British[0.039]; British[0.039]; Rocca[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.075:0.075[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.053]; Ireland[0.046]; Spain[0.046]; Eriksson[0.044]; Colin[0.042]; France[0.042]; Australia[0.042]; Australia[0.042]; Peter[0.042]; Peter[0.042]; Peter[0.042]; Klas[0.041]; Mark[0.041]; David[0.041]; Mitchell[0.040]; Mike[0.040]; Stephen[0.040]; Iain[0.040]; Pedro[0.039]; Philip[0.039]; Adam[0.039]; Robert[0.039]; Lebouc[0.039]; McAllister[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Stephen McAllister\u001b[39m ==> ENTITY: \u001b[32mStephen McAllister\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.122:0.122[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSweden[0.045]; Sweden[0.045]; Ireland[0.044]; Colin[0.044]; Hedblom[0.043]; Iain[0.043]; Hunter[0.043]; Klas[0.043]; Lebouc[0.042]; Montgomerie[0.041]; Mark[0.041]; Peter[0.041]; Peter[0.041]; Peter[0.041]; Spain[0.041]; David[0.041]; Robert[0.041]; Roe[0.041]; Philip[0.041]; France[0.040]; Australia[0.040]; Australia[0.040]; Pyman[0.040]; Walton[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Peter Mitchell ==> ENTITY: \u001b[32mPeter Mitchell (golfer)\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.157:0.157[0]; log p(e|m)= -1.204:-1.204[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngolf[0.046]; Masters[0.044]; tournament[0.043]; British[0.043]; British[0.043]; Hedblom[0.042]; Mark[0.042]; Peter[0.042]; Peter[0.042]; Ian[0.042]; Martin[0.041]; Australia[0.041]; Australia[0.041]; Australia[0.041]; Lebouc[0.041]; Sweden[0.041]; Sweden[0.041]; Sweden[0.041]; Adam[0.041]; David[0.041]; Colin[0.041]; Iain[0.040]; play[0.040]; Philip[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = British ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.489:-0.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.048]; British[0.048]; England[0.047]; Ireland[0.045]; Martin[0.044]; Philip[0.043]; Ian[0.043]; Peter[0.042]; Robert[0.042]; Robert[0.042]; Australia[0.042]; Colin[0.042]; won[0.041]; final[0.041]; Sweden[0.040]; France[0.040]; Italy[0.040]; Spain[0.040]; Spain[0.040]; play[0.039]; round[0.038]; Argentina[0.038]; Mitchell[0.038]; Saturday[0.038]; \t\n\r [==============================>...............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 7s449ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3256/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1026testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.204:0.204[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.053]; Israeli[0.048]; Arab[0.047]; Syria[0.046]; Syria[0.046]; Palestinian[0.045]; Netanyahu[0.043]; Arabs[0.042]; policy[0.041]; Damascus[0.041]; Damascus[0.041]; government[0.040]; government[0.040]; government[0.040]; war[0.040]; war[0.040]; peace[0.039]; settlement[0.039]; settlement[0.039]; said[0.039]; said[0.039]; said[0.039]; Radio[0.039]; settlements[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.218:0.218[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.053]; Israel[0.053]; Israel[0.053]; Israeli[0.048]; Arab[0.047]; Syria[0.046]; Syria[0.046]; Jerusalem[0.045]; Netanyahu[0.043]; international[0.043]; international[0.043]; Arabs[0.043]; policy[0.042]; talks[0.041]; government[0.040]; policies[0.040]; trade[0.040]; peace[0.039]; peace[0.039]; peace[0.039]; settlements[0.039]; support[0.039]; settlement[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.275:0.275[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.049]; Israel[0.049]; Israel[0.049]; Israel[0.049]; Israel[0.049]; Israeli[0.044]; Israeli[0.044]; Jerusalem[0.041]; Arab[0.041]; Arab[0.041]; Arab[0.041]; Syria[0.040]; Syria[0.040]; Arabs[0.040]; Arabs[0.040]; Palestinians[0.040]; Gaza[0.039]; West[0.039]; West[0.039]; international[0.038]; international[0.038]; Bank[0.038]; Bank[0.038]; Netanyahu[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.242:0.242[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.054]; Israel[0.054]; Israeli[0.049]; Israeli[0.049]; Israeli[0.049]; Arab[0.048]; Arab[0.048]; Syria[0.046]; Palestinian[0.046]; Gaza[0.045]; Palestinians[0.044]; international[0.044]; international[0.044]; Arabs[0.044]; Arabs[0.044]; West[0.043]; West[0.043]; policy[0.042]; Bank[0.042]; Bank[0.042]; Damascus[0.041]; Damascus[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yasser Arafat\u001b[39m ==> ENTITY: \u001b[32mYasser Arafat\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.049]; Israel[0.049]; Israel[0.049]; Israeli[0.047]; Israeli[0.047]; Israeli[0.047]; Palestinian[0.046]; Gaza[0.045]; Palestinians[0.044]; Syria[0.043]; Arab[0.043]; Arab[0.043]; Arab[0.043]; Damascus[0.043]; Damascus[0.043]; Damascus[0.043]; Netanyahu[0.042]; settlements[0.040]; Arabs[0.039]; Arabs[0.039]; Bank[0.039]; Bank[0.039]; President[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Benjamin Netanyahu\u001b[39m ==> ENTITY: \u001b[32mBenjamin Netanyahu\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.052]; Israel[0.052]; Israeli[0.047]; Palestinian[0.045]; Arab[0.042]; Palestinians[0.042]; Syria[0.042]; Syria[0.042]; Prime[0.042]; President[0.041]; Arabs[0.041]; Bank[0.040]; said[0.040]; said[0.040]; said[0.040]; policy[0.040]; Damascus[0.040]; Damascus[0.040]; West[0.039]; Minister[0.039]; week[0.039]; settlements[0.039]; settlements[0.039]; Arafat[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syria ==> ENTITY: \u001b[32mSyria\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.239:0.239[0]; log p(e|m)= -0.261:-0.261[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.047]; Israel[0.047]; Israel[0.047]; Israel[0.047]; Israel[0.047]; Israel[0.047]; Arab[0.044]; Arab[0.044]; Arab[0.044]; Arabs[0.043]; Israeli[0.042]; Damascus[0.042]; international[0.040]; international[0.040]; government[0.039]; government[0.039]; policy[0.038]; Jerusalem[0.038]; Netanyahu[0.038]; peace[0.037]; peace[0.037]; peace[0.037]; force[0.037]; occupied[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Damascus ==> ENTITY: \u001b[32mDamascus\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.204:0.204[0]; log p(e|m)= -0.120:-0.120[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.047]; Arab[0.046]; Arab[0.046]; Arab[0.046]; Israel[0.045]; Israel[0.045]; Israel[0.045]; Israeli[0.044]; Israeli[0.044]; Israeli[0.044]; Jerusalem[0.044]; Arabs[0.044]; Arabs[0.044]; Palestinian[0.043]; government[0.043]; government[0.043]; Gaza[0.043]; declaration[0.043]; Palestinians[0.042]; Yasser[0.042]; Arafat[0.040]; support[0.040]; week[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.201:0.201[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.052]; Israel[0.052]; Israeli[0.047]; Arab[0.046]; Syria[0.045]; Syria[0.045]; Jerusalem[0.044]; Netanyahu[0.043]; international[0.042]; Arabs[0.042]; policy[0.041]; talks[0.040]; government[0.039]; policies[0.039]; trade[0.039]; peace[0.038]; peace[0.038]; peace[0.038]; settlements[0.038]; support[0.038]; settlement[0.038]; said[0.038]; efforts[0.038]; refusal[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syria ==> ENTITY: \u001b[32mSyria\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.213:0.213[0]; log p(e|m)= -0.261:-0.261[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.051]; Israel[0.051]; Israel[0.051]; Arab[0.048]; Arabs[0.047]; Israeli[0.046]; Damascus[0.045]; Damascus[0.045]; government[0.042]; government[0.042]; government[0.042]; policy[0.042]; state[0.042]; settlements[0.041]; Netanyahu[0.041]; peace[0.041]; war[0.041]; war[0.041]; settlement[0.040]; settlement[0.040]; said[0.040]; said[0.040]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syria ==> ENTITY: \u001b[32mSyria\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.219:0.219[0]; log p(e|m)= -0.261:-0.261[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.051]; Israel[0.050]; Israel[0.050]; Arab[0.048]; Arabs[0.047]; Israeli[0.046]; Palestinian[0.045]; Damascus[0.045]; Damascus[0.045]; government[0.042]; government[0.042]; government[0.042]; policy[0.041]; state[0.041]; settlements[0.041]; Netanyahu[0.041]; Arafat[0.041]; peace[0.041]; war[0.040]; war[0.040]; settlement[0.040]; settlement[0.040]; Yasser[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.257:0.257[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israel[0.050]; Israel[0.050]; Israeli[0.046]; Israeli[0.046]; Israeli[0.046]; Arab[0.045]; Arab[0.045]; Arab[0.045]; Syria[0.044]; Syria[0.044]; Jerusalem[0.043]; Gaza[0.042]; Palestinians[0.042]; Netanyahu[0.041]; international[0.041]; international[0.041]; Arabs[0.041]; Arabs[0.041]; West[0.040]; West[0.040]; policy[0.039]; talks[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mGaza\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mGaza Strip\u001b[39m <---> \u001b[32mGaza\u001b[39m\t\nSCORES: global= 0.262:0.258[\u001b[31m0.004\u001b[39m]; local()= 0.244:0.242[\u001b[31m0.002\u001b[39m]; log p(e|m)= -1.374:-0.304[\u001b[32m1.071\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.046]; Israel[0.046]; Palestinian[0.045]; Israeli[0.044]; Israeli[0.044]; Israeli[0.044]; Israeli[0.044]; Palestinians[0.043]; Arab[0.041]; Arab[0.041]; Arab[0.041]; Bank[0.041]; Bank[0.041]; West[0.041]; West[0.041]; Arabs[0.041]; Arabs[0.041]; settlements[0.040]; Damascus[0.039]; Damascus[0.039]; settlement[0.039]; settlement[0.039]; force[0.039]; week[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.213:0.213[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.051]; Israel[0.051]; Palestinian[0.044]; settlements[0.043]; settlements[0.043]; Palestinians[0.043]; Arab[0.043]; Gaza[0.042]; Syria[0.042]; Syria[0.042]; Arabs[0.042]; West[0.042]; West[0.042]; Netanyahu[0.041]; Bank[0.041]; Bank[0.041]; settlement[0.040]; settlement[0.040]; Damascus[0.038]; Damascus[0.038]; Yasser[0.038]; Radio[0.038]; peace[0.038]; policy[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jerusalem ==> ENTITY: \u001b[32mJerusalem\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.180:0.180[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.048]; Israel[0.048]; Israel[0.048]; Israel[0.048]; Israeli[0.044]; occupied[0.044]; Arabs[0.044]; Arab[0.043]; settlements[0.042]; Damascus[0.042]; Syria[0.041]; Syria[0.041]; settlement[0.040]; Netanyahu[0.040]; held[0.039]; called[0.039]; refusal[0.039]; peace[0.039]; peace[0.039]; peace[0.039]; force[0.039]; said[0.038]; said[0.038]; solidarity[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.261:0.261[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israel[0.050]; Israel[0.050]; Israel[0.050]; Israeli[0.046]; Israeli[0.046]; Israeli[0.046]; Arab[0.045]; Arab[0.045]; Arab[0.045]; Syria[0.043]; Syria[0.043]; Jerusalem[0.043]; Netanyahu[0.041]; international[0.041]; international[0.041]; Arabs[0.041]; Arabs[0.041]; policy[0.039]; talks[0.039]; Damascus[0.039]; government[0.038]; government[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Damascus ==> ENTITY: \u001b[32mDamascus\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.120:-0.120[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDamascus[0.046]; Syria[0.046]; Syria[0.046]; Arab[0.045]; Israel[0.044]; Israel[0.044]; Israeli[0.043]; Arabs[0.043]; Palestinian[0.043]; government[0.042]; government[0.042]; government[0.042]; Yasser[0.041]; settlements[0.040]; policy[0.040]; condemned[0.040]; state[0.039]; said[0.039]; said[0.039]; said[0.039]; war[0.039]; war[0.039]; confirm[0.039]; Netanyahu[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.244:0.244[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israel[0.050]; Israel[0.050]; Israeli[0.045]; Israeli[0.045]; Palestinian[0.043]; settlements[0.042]; Palestinians[0.042]; Arab[0.042]; Arab[0.042]; Arab[0.042]; Gaza[0.041]; Arabs[0.041]; West[0.041]; West[0.041]; Bank[0.040]; Bank[0.040]; international[0.039]; settlement[0.039]; settlement[0.039]; Damascus[0.037]; Damascus[0.037]; Yasser[0.037]; Radio[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Syria ==> ENTITY: \u001b[32mSyria\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.207:0.207[0]; log p(e|m)= -0.261:-0.261[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSyria[0.049]; Israel[0.048]; Israel[0.048]; Israel[0.048]; Israel[0.048]; Arab[0.046]; Arabs[0.045]; Israeli[0.044]; Jerusalem[0.042]; international[0.041]; international[0.041]; government[0.040]; policy[0.039]; occupied[0.039]; settlements[0.039]; Netanyahu[0.039]; peace[0.039]; peace[0.039]; peace[0.039]; force[0.038]; settlement[0.038]; said[0.038]; said[0.038]; refusal[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.199:0.199[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.057]; Israeli[0.052]; Arab[0.051]; Syria[0.049]; Syria[0.049]; Palestinian[0.049]; Netanyahu[0.047]; Arabs[0.046]; policy[0.045]; Damascus[0.044]; Damascus[0.044]; government[0.043]; government[0.043]; government[0.043]; week[0.043]; war[0.043]; war[0.043]; peace[0.042]; President[0.042]; Arafat[0.042]; settlement[0.042]; settlement[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Netanyahu\u001b[39m ==> ENTITY: \u001b[32mBenjamin Netanyahu\u001b[39m\t\nSCORES: global= 0.312:0.312[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israel[0.050]; Israel[0.050]; Israel[0.050]; Israeli[0.046]; Jerusalem[0.044]; Arab[0.042]; Arab[0.042]; Arab[0.042]; Syria[0.041]; Syria[0.041]; Arabs[0.040]; said[0.039]; said[0.039]; policy[0.039]; talks[0.039]; Damascus[0.039]; international[0.038]; international[0.038]; criticism[0.038]; criticism[0.038]; support[0.038]; settlements[0.038]; government[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.242:0.242[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.052]; Israel[0.052]; Israel[0.052]; Israeli[0.047]; Israeli[0.047]; Palestinian[0.045]; settlements[0.044]; Palestinians[0.044]; Arab[0.043]; Arab[0.043]; Gaza[0.043]; Arabs[0.043]; West[0.043]; West[0.043]; Bank[0.042]; Bank[0.042]; international[0.041]; international[0.041]; settlement[0.040]; Damascus[0.039]; Damascus[0.039]; Yasser[0.039]; Radio[0.039]; \t\n\r [==============================>...............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 7s365ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3280/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1067testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Millwall\u001b[39m ==> ENTITY: \u001b[32mMillwall F.C.\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.217:0.217[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.046]; Sheffield[0.044]; Watford[0.044]; Charlton[0.043]; Brentford[0.042]; Manchester[0.042]; Shrewsbury[0.042]; Chesterfield[0.042]; Burnley[0.042]; Southend[0.042]; Gillingham[0.042]; City[0.041]; Crewe[0.041]; Bristol[0.041]; Plymouth[0.041]; Bournemouth[0.041]; Oldham[0.041]; Bury[0.041]; Rovers[0.041]; Blackpool[0.041]; Peterborough[0.040]; Division[0.040]; Grimsby[0.040]; Notts[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Shrewsbury ==> ENTITY: \u001b[32mShrewsbury Town F.C.\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.218:0.218[0]; log p(e|m)= -3.244:-3.244[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Bristol[0.043]; Bristol[0.043]; Watford[0.043]; Chesterfield[0.043]; Charlton[0.043]; Sheffield[0.042]; Plymouth[0.042]; City[0.042]; City[0.042]; Peterborough[0.042]; Southend[0.042]; Gillingham[0.042]; Millwall[0.041]; Crewe[0.041]; Burnley[0.041]; Brentford[0.041]; Bournemouth[0.041]; Oldham[0.041]; Grimsby[0.041]; Bury[0.040]; York[0.040]; Blackpool[0.040]; West[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Leyton Orient\u001b[39m ==> ENTITY: \u001b[32mLeyton Orient F.C.\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwansea[0.056]; Colchester[0.056]; Cardiff[0.056]; Rochdale[0.055]; Northampton[0.055]; Carlisle[0.054]; Scunthorpe[0.054]; Barnet[0.053]; Brighton[0.053]; Mansfield[0.052]; Exeter[0.052]; Darlington[0.051]; Chester[0.051]; United[0.051]; Doncaster[0.051]; Hereford[0.050]; Scarborough[0.050]; Cambridge[0.050]; Lincoln[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ipswich ==> ENTITY: \u001b[32mIpswich Town F.C.\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.200:0.200[0]; log p(e|m)= -2.688:-2.688[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; Sheffield[0.046]; Manchester[0.045]; Bolton[0.045]; Portsmouth[0.045]; Birmingham[0.045]; Norwich[0.045]; matches[0.044]; Barnsley[0.044]; Reading[0.044]; Tranmere[0.043]; Oldham[0.043]; Swindon[0.043]; City[0.043]; played[0.043]; Stoke[0.042]; Park[0.042]; Wolverhampton[0.042]; Huddersfield[0.042]; Bradford[0.042]; Oxford[0.042]; United[0.042]; Grimsby[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Luton ==> ENTITY: \u001b[32mLuton Town F.C.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.212:0.212[0]; log p(e|m)= -2.551:-2.551[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.046]; Hull[0.045]; Walsall[0.043]; Watford[0.043]; Stockport[0.042]; Gillingham[0.042]; Burnley[0.042]; Scunthorpe[0.042]; Bristol[0.042]; Bristol[0.042]; Rotherham[0.042]; Bournemouth[0.042]; Crewe[0.042]; Cardiff[0.041]; Blackpool[0.041]; Peterborough[0.041]; Wrexham[0.041]; Wycombe[0.041]; City[0.040]; Rovers[0.040]; Torquay[0.040]; Notts[0.040]; Hartlepool[0.040]; Division[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wycombe ==> ENTITY: \u001b[32mWycombe Wanderers F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.213:0.213[0]; log p(e|m)= -1.146:-1.146[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Hull[0.044]; Watford[0.043]; Walsall[0.043]; Stockport[0.043]; Carlisle[0.043]; Bristol[0.042]; Bristol[0.042]; Cardiff[0.042]; Burnley[0.042]; Rotherham[0.042]; Peterborough[0.042]; Luton[0.042]; Scunthorpe[0.042]; Gillingham[0.042]; Crewe[0.041]; Wrexham[0.041]; City[0.041]; Bournemouth[0.041]; Notts[0.041]; Rovers[0.040]; Torquay[0.040]; Division[0.039]; Hartlepool[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Norwich ==> ENTITY: \u001b[32mNorwich City F.C.\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.183:0.183[0]; log p(e|m)= -2.733:-2.733[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; league[0.046]; Birmingham[0.046]; matches[0.045]; Portsmouth[0.045]; Barnsley[0.044]; Bolton[0.044]; Ipswich[0.044]; Division[0.044]; Reading[0.044]; Wolverhampton[0.043]; Standings[0.043]; Park[0.043]; Swindon[0.043]; Tranmere[0.043]; Stoke[0.042]; Soccer[0.042]; played[0.042]; London[0.042]; goals[0.042]; goals[0.042]; Queens[0.042]; Bradford[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Huddersfield ==> ENTITY: \u001b[32mHuddersfield Town F.C.\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.198:0.198[0]; log p(e|m)= -2.283:-2.283[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBradford[0.044]; Sheffield[0.044]; Wolverhampton[0.044]; Charlton[0.043]; Manchester[0.043]; Birmingham[0.042]; Park[0.042]; Bury[0.042]; Oldham[0.042]; Portsmouth[0.042]; City[0.042]; Brentford[0.042]; Swindon[0.042]; Southend[0.041]; Plymouth[0.041]; Reading[0.041]; Tranmere[0.041]; Ipswich[0.041]; West[0.040]; Grimsby[0.040]; Rangers[0.040]; Bolton[0.040]; Oxford[0.040]; United[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scarborough ==> ENTITY: \u001b[32mScarborough F.C.\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.206:0.206[0]; log p(e|m)= -1.814:-1.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWalsall[0.044]; Rotherham[0.044]; Carlisle[0.043]; Stockport[0.043]; Hull[0.043]; Swansea[0.043]; Darlington[0.043]; Hartlepool[0.043]; Cardiff[0.042]; Doncaster[0.042]; Wrexham[0.042]; Exeter[0.042]; Torquay[0.042]; Cambridge[0.042]; Scunthorpe[0.041]; Chester[0.041]; Division[0.041]; Hereford[0.041]; Brighton[0.041]; Barnet[0.041]; Northampton[0.040]; United[0.040]; Wigan[0.039]; Lincoln[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Northampton ==> ENTITY: \u001b[32mNorthampton Town F.C.\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.218:0.218[0]; log p(e|m)= -3.244:-3.244[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.045]; Cardiff[0.044]; Stockport[0.044]; Colchester[0.044]; Swansea[0.044]; Walsall[0.044]; Scunthorpe[0.043]; Carlisle[0.042]; Rotherham[0.042]; Chester[0.042]; Cambridge[0.041]; Barnet[0.041]; Exeter[0.041]; Torquay[0.041]; Hereford[0.041]; Darlington[0.041]; Hartlepool[0.041]; Doncaster[0.040]; Brighton[0.040]; Division[0.040]; United[0.040]; Wigan[0.040]; Scarborough[0.039]; Orient[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hereford ==> ENTITY: \u001b[32mHereford United F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.193:0.193[0]; log p(e|m)= -2.765:-2.765[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nColchester[0.060]; Rochdale[0.059]; Swansea[0.059]; Northampton[0.058]; Carlisle[0.058]; United[0.057]; Exeter[0.057]; Chester[0.057]; Darlington[0.056]; Barnet[0.056]; Doncaster[0.055]; Brighton[0.055]; Scarborough[0.054]; Mansfield[0.054]; Cambridge[0.053]; Lincoln[0.051]; Orient[0.050]; Leyton[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Southend ==> ENTITY: \u001b[32mSouthend United F.C.\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.228:0.228[0]; log p(e|m)= -1.772:-1.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.045]; Sheffield[0.044]; Chesterfield[0.043]; Millwall[0.043]; Oldham[0.043]; Watford[0.043]; Reading[0.043]; Brentford[0.043]; Shrewsbury[0.042]; Burnley[0.042]; Manchester[0.042]; Bury[0.042]; City[0.042]; Huddersfield[0.042]; Plymouth[0.042]; Bournemouth[0.042]; Birmingham[0.041]; Grimsby[0.041]; Blackpool[0.040]; United[0.040]; York[0.039]; Oxford[0.039]; Division[0.039]; Palace[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swindon ==> ENTITY: \u001b[32mSwindon Town F.C.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.202:0.202[0]; log p(e|m)= -2.489:-2.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; league[0.044]; Reading[0.043]; matches[0.043]; Portsmouth[0.042]; Oldham[0.042]; Barnsley[0.042]; Wolverhampton[0.042]; Ipswich[0.042]; Manchester[0.042]; Norwich[0.042]; Tranmere[0.042]; Birmingham[0.042]; City[0.041]; Soccer[0.041]; Oxford[0.041]; Huddersfield[0.041]; Bradford[0.041]; Bolton[0.041]; Division[0.040]; Park[0.040]; played[0.040]; London[0.040]; soccer[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Blackpool ==> ENTITY: \u001b[32mBlackpool F.C.\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.201:0.201[0]; log p(e|m)= -0.853:-0.853[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthend[0.044]; Bury[0.044]; Preston[0.044]; Rovers[0.043]; Millwall[0.043]; Watford[0.043]; Charlton[0.042]; Sheffield[0.042]; Gillingham[0.042]; Brentford[0.042]; Peterborough[0.042]; Oldham[0.042]; Bournemouth[0.041]; Shrewsbury[0.041]; Notts[0.041]; Luton[0.041]; Burnley[0.041]; Grimsby[0.041]; Bristol[0.041]; Bristol[0.041]; Crewe[0.040]; Plymouth[0.040]; Chesterfield[0.040]; City[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colchester ==> ENTITY: \u001b[32mColchester United F.C.\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.177:0.177[0]; log p(e|m)= -2.865:-2.865[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwansea[0.066]; Exeter[0.066]; Cambridge[0.065]; Northampton[0.065]; Rochdale[0.064]; United[0.064]; Darlington[0.064]; Brighton[0.063]; Barnet[0.063]; Hereford[0.063]; Mansfield[0.062]; Chester[0.062]; Doncaster[0.061]; Leyton[0.059]; Lincoln[0.058]; Orient[0.058]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mansfield ==> ENTITY: \u001b[32mMansfield Town F.C.\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.217:0.217[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwansea[0.076]; Colchester[0.076]; Rochdale[0.075]; Hereford[0.074]; Exeter[0.074]; Chester[0.073]; Darlington[0.073]; Doncaster[0.071]; Brighton[0.071]; Barnet[0.071]; United[0.069]; Cambridge[0.069]; Orient[0.065]; Leyton[0.063]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wrexham ==> ENTITY: \u001b[32mWrexham F.C.\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.206:0.206[0]; log p(e|m)= -0.882:-0.882[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.047]; Cardiff[0.045]; Hull[0.044]; Stockport[0.043]; Walsall[0.043]; Rotherham[0.043]; Bristol[0.042]; Bristol[0.042]; Peterborough[0.042]; Watford[0.042]; Bournemouth[0.042]; Gillingham[0.042]; Scunthorpe[0.041]; Luton[0.041]; Carlisle[0.041]; Burnley[0.041]; Crewe[0.041]; City[0.040]; Hartlepool[0.040]; Scarborough[0.040]; Wycombe[0.040]; Rovers[0.040]; Torquay[0.039]; Division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Stockport ==> ENTITY: \u001b[32mStockport County F.C.\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.218:0.218[0]; log p(e|m)= -2.526:-2.526[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.045]; Preston[0.044]; Cardiff[0.044]; Rotherham[0.043]; Walsall[0.043]; Northampton[0.043]; Watford[0.042]; Scunthorpe[0.042]; Bristol[0.042]; Bristol[0.042]; Bournemouth[0.042]; Gillingham[0.042]; County[0.042]; Wrexham[0.041]; Carlisle[0.041]; Luton[0.041]; Crewe[0.041]; City[0.041]; Peterborough[0.041]; Wycombe[0.041]; Torquay[0.040]; Hartlepool[0.040]; Scarborough[0.039]; Division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oxford ==> ENTITY: \u001b[32mOxford United F.C.\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.210:0.210[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSheffield[0.044]; Charlton[0.044]; Wolverhampton[0.043]; Brentford[0.043]; Birmingham[0.043]; City[0.043]; Bury[0.042]; Reading[0.042]; Swindon[0.042]; Southend[0.042]; Portsmouth[0.042]; Chesterfield[0.042]; Ipswich[0.042]; Bradford[0.041]; Manchester[0.041]; Park[0.041]; Oldham[0.041]; Huddersfield[0.041]; Plymouth[0.041]; Grimsby[0.041]; Bolton[0.041]; United[0.040]; Queens[0.040]; Division[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barnet ==> ENTITY: \u001b[32mBarnet F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.196:0.196[0]; log p(e|m)= -0.679:-0.679[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWalsall[0.044]; Colchester[0.044]; Hull[0.044]; Rochdale[0.043]; Swansea[0.043]; Carlisle[0.043]; Hartlepool[0.042]; Northampton[0.042]; Cambridge[0.042]; Scunthorpe[0.042]; Cardiff[0.042]; Darlington[0.042]; Exeter[0.041]; Hereford[0.041]; Brighton[0.041]; Mansfield[0.041]; Fulham[0.041]; Chester[0.041]; Torquay[0.041]; Doncaster[0.040]; United[0.040]; Scarborough[0.040]; Wigan[0.040]; Division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bolton ==> ENTITY: \u001b[32mBolton Wanderers F.C.\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.167:0.167[0]; log p(e|m)= -2.283:-2.283[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; league[0.046]; Bradford[0.045]; Portsmouth[0.045]; Soccer[0.044]; matches[0.044]; Norwich[0.044]; Tranmere[0.044]; soccer[0.044]; Birmingham[0.044]; Reading[0.044]; Ipswich[0.044]; Barnsley[0.043]; Wolverhampton[0.043]; Park[0.043]; Huddersfield[0.043]; Swindon[0.043]; Queens[0.042]; Oxford[0.042]; Stoke[0.042]; Saturday[0.042]; Division[0.042]; played[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Charlton ==> ENTITY: \u001b[32mCharlton Athletic F.C.\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.200:0.200[0]; log p(e|m)= -2.104:-2.104[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWatford[0.044]; Sheffield[0.044]; Manchester[0.043]; Bristol[0.043]; Shrewsbury[0.043]; Millwall[0.043]; Burnley[0.043]; Reading[0.042]; Southend[0.042]; Blackpool[0.042]; City[0.042]; Birmingham[0.042]; Brentford[0.042]; Bury[0.041]; Chesterfield[0.041]; Oldham[0.041]; Oxford[0.041]; United[0.041]; Plymouth[0.041]; Bournemouth[0.041]; Huddersfield[0.041]; Division[0.040]; Rovers[0.040]; Grimsby[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burnley ==> ENTITY: \u001b[32mBurnley F.C.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.187:0.187[0]; log p(e|m)= -0.992:-0.992[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.047]; Chesterfield[0.046]; County[0.045]; Sheffield[0.043]; Charlton[0.043]; Watford[0.043]; Gillingham[0.042]; Brentford[0.042]; Bury[0.041]; Millwall[0.041]; Southend[0.041]; Bristol[0.041]; Bristol[0.041]; Shrewsbury[0.041]; Peterborough[0.041]; Wycombe[0.041]; Wrexham[0.041]; City[0.040]; Plymouth[0.040]; Luton[0.040]; Bournemouth[0.040]; Grimsby[0.040]; Crewe[0.040]; Blackpool[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bournemouth ==> ENTITY: \u001b[32mA.F.C. Bournemouth\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.201:0.201[0]; log p(e|m)= -1.058:-1.058[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouthend[0.044]; Blackpool[0.044]; Stockport[0.043]; Shrewsbury[0.043]; Brentford[0.043]; Wycombe[0.043]; Wrexham[0.043]; Preston[0.042]; Notts[0.042]; Gillingham[0.042]; Grimsby[0.042]; Burnley[0.042]; Rovers[0.041]; Charlton[0.041]; Peterborough[0.041]; Crewe[0.041]; Watford[0.041]; Bristol[0.041]; Bristol[0.041]; City[0.040]; Luton[0.040]; Millwall[0.040]; Bury[0.040]; Plymouth[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Crystal Palace ==> ENTITY: \u001b[32mCrystal Palace F.C.\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.190:0.190[0]; log p(e|m)= -0.511:-0.511[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.046]; Sheffield[0.046]; Birmingham[0.045]; Manchester[0.045]; Bolton[0.045]; Portsmouth[0.044]; Ipswich[0.044]; Park[0.044]; Southend[0.044]; Norwich[0.044]; Barnsley[0.044]; Reading[0.043]; City[0.043]; Tranmere[0.043]; Oldham[0.043]; Swindon[0.043]; Wolverhampton[0.043]; Stoke[0.042]; Huddersfield[0.042]; Bradford[0.042]; United[0.042]; Grimsby[0.042]; Queens[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rotherham ==> ENTITY: \u001b[32mRotherham United F.C.\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.216:0.216[0]; log p(e|m)= -2.645:-2.645[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.047]; Hull[0.044]; Watford[0.043]; Stockport[0.043]; Bristol[0.043]; Bristol[0.043]; Walsall[0.043]; Cardiff[0.042]; Gillingham[0.042]; Carlisle[0.042]; Northampton[0.042]; Scunthorpe[0.042]; Luton[0.042]; Rovers[0.041]; Wycombe[0.041]; Peterborough[0.041]; Wrexham[0.041]; City[0.041]; Crewe[0.040]; Division[0.040]; Hartlepool[0.040]; Torquay[0.040]; Notts[0.039]; Wigan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Walsall ==> ENTITY: \u001b[32mWalsall F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.183:0.183[0]; log p(e|m)= -0.620:-0.620[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Hull[0.044]; Stockport[0.043]; Bristol[0.043]; Bristol[0.043]; Luton[0.042]; City[0.042]; Peterborough[0.042]; Scunthorpe[0.042]; Rotherham[0.042]; Cardiff[0.041]; Gillingham[0.041]; Crewe[0.041]; Wycombe[0.041]; Wrexham[0.041]; Carlisle[0.041]; Division[0.041]; Notts[0.041]; Hartlepool[0.041]; Barnet[0.040]; Scarborough[0.040]; Northampton[0.040]; Torquay[0.040]; Rovers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Manchester City\u001b[39m ==> ENTITY: \u001b[32mManchester City F.C.\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.045]; Portsmouth[0.045]; Sheffield[0.044]; Birmingham[0.044]; Division[0.043]; Millwall[0.043]; Ipswich[0.042]; Brentford[0.042]; Southend[0.042]; Shrewsbury[0.042]; United[0.042]; Reading[0.042]; Chesterfield[0.041]; Park[0.041]; Plymouth[0.041]; Swindon[0.041]; Bury[0.041]; Wolverhampton[0.041]; Grimsby[0.040]; Oldham[0.040]; Huddersfield[0.040]; Rangers[0.040]; Bradford[0.040]; Queens[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Exeter ==> ENTITY: \u001b[32mExeter City F.C.\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.217:0.217[0]; log p(e|m)= -3.650:-3.650[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.045]; Carlisle[0.044]; Cardiff[0.044]; Colchester[0.043]; Swansea[0.043]; Northampton[0.043]; Rochdale[0.043]; Chester[0.043]; Scunthorpe[0.042]; Cambridge[0.042]; Hereford[0.042]; Brighton[0.042]; Darlington[0.042]; United[0.041]; Mansfield[0.041]; Barnet[0.041]; Torquay[0.041]; Division[0.041]; Wigan[0.040]; Scarborough[0.040]; Hartlepool[0.040]; Doncaster[0.040]; Fulham[0.039]; Orient[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Darlington ==> ENTITY: \u001b[32mDarlington F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.198:0.198[0]; log p(e|m)= -1.871:-1.871[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.047]; Swansea[0.046]; Northampton[0.046]; Exeter[0.045]; Colchester[0.045]; Scunthorpe[0.045]; Carlisle[0.045]; Hartlepool[0.045]; Doncaster[0.045]; Cardiff[0.045]; Mansfield[0.044]; Rochdale[0.044]; Scarborough[0.044]; Torquay[0.043]; Hereford[0.043]; United[0.043]; Chester[0.042]; Cambridge[0.042]; Barnet[0.042]; Brighton[0.042]; Lincoln[0.041]; Orient[0.039]; Leyton[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Plymouth ==> ENTITY: \u001b[32mPlymouth Argyle F.C.\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.219:0.219[0]; log p(e|m)= -4.075:-4.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSheffield[0.045]; Watford[0.044]; Birmingham[0.044]; Charlton[0.043]; Burnley[0.043]; Reading[0.042]; Southend[0.042]; Millwall[0.042]; Manchester[0.042]; Oldham[0.042]; Bristol[0.042]; Brentford[0.042]; Chesterfield[0.042]; Bury[0.041]; City[0.041]; Peterborough[0.041]; Shrewsbury[0.041]; Blackpool[0.041]; Bournemouth[0.041]; Huddersfield[0.040]; Grimsby[0.040]; Division[0.040]; Rovers[0.040]; Oxford[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bradford ==> ENTITY: \u001b[32mBradford City A.F.C.\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.220:0.220[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.045]; league[0.045]; Sheffield[0.044]; Huddersfield[0.043]; matches[0.043]; Manchester[0.042]; Reading[0.042]; City[0.042]; Barnsley[0.042]; Oldham[0.042]; Swindon[0.042]; Tranmere[0.042]; Ipswich[0.041]; played[0.041]; Norwich[0.041]; Portsmouth[0.041]; Division[0.041]; Birmingham[0.041]; Wolverhampton[0.040]; Soccer[0.040]; soccer[0.040]; Park[0.040]; Rangers[0.040]; Bolton[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Watford ==> ENTITY: \u001b[32mWatford F.C.\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.197:0.197[0]; log p(e|m)= -0.960:-0.960[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Charlton[0.044]; Brentford[0.043]; Burnley[0.042]; Millwall[0.042]; Bristol[0.042]; Bristol[0.042]; Stockport[0.042]; Shrewsbury[0.042]; Bury[0.042]; Chesterfield[0.042]; Luton[0.042]; Blackpool[0.042]; Southend[0.041]; Bournemouth[0.041]; Gillingham[0.041]; Wycombe[0.041]; Crewe[0.041]; Peterborough[0.041]; City[0.041]; Rotherham[0.041]; Wrexham[0.040]; Plymouth[0.040]; Rovers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bristol Rovers\u001b[39m ==> ENTITY: \u001b[32mBristol Rovers F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.216:0.216[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBristol[0.051]; Shrewsbury[0.044]; Plymouth[0.044]; Watford[0.043]; Burnley[0.043]; Rotherham[0.043]; Charlton[0.043]; Stockport[0.042]; Blackpool[0.042]; Bournemouth[0.042]; Millwall[0.042]; Peterborough[0.042]; Gillingham[0.041]; Wrexham[0.041]; Walsall[0.041]; Crewe[0.041]; Wycombe[0.041]; Brentford[0.041]; Luton[0.040]; Chesterfield[0.040]; Bury[0.039]; Preston[0.039]; York[0.038]; Notts[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Barnsley ==> ENTITY: \u001b[32mBarnsley F.C.\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.180:0.180[0]; log p(e|m)= -0.713:-0.713[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBradford[0.044]; league[0.044]; League[0.044]; Norwich[0.044]; Birmingham[0.043]; Tranmere[0.043]; Ipswich[0.043]; Swindon[0.043]; Wolverhampton[0.043]; Bolton[0.042]; matches[0.042]; Portsmouth[0.042]; Park[0.042]; Stoke[0.042]; Soccer[0.041]; Vale[0.041]; English[0.040]; English[0.040]; played[0.040]; Rangers[0.040]; Saturday[0.040]; won[0.039]; goals[0.039]; goals[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wigan ==> ENTITY: \u001b[32mWigan Athletic F.C.\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.153:0.153[0]; log p(e|m)= -2.323:-2.323[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBristol[0.045]; Cardiff[0.044]; Hull[0.044]; Preston[0.044]; Northampton[0.044]; Exeter[0.043]; City[0.042]; Carlisle[0.042]; Rotherham[0.042]; Luton[0.042]; Peterborough[0.042]; Stockport[0.042]; Walsall[0.041]; Hartlepool[0.041]; Crewe[0.041]; Wycombe[0.041]; Fulham[0.040]; Scunthorpe[0.040]; Wrexham[0.040]; Gillingham[0.040]; Scarborough[0.040]; Barnet[0.039]; Torquay[0.039]; Notts[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Fulham ==> ENTITY: \u001b[32mFulham F.C.\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.163:0.163[0]; log p(e|m)= -0.322:-0.322[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.046]; Cardiff[0.045]; Rotherham[0.044]; Hull[0.044]; Bristol[0.044]; Wigan[0.043]; Stockport[0.043]; Northampton[0.042]; United[0.042]; Luton[0.041]; Walsall[0.041]; City[0.041]; Barnet[0.041]; Wrexham[0.041]; Wycombe[0.041]; Exeter[0.041]; Carlisle[0.041]; Scunthorpe[0.040]; Division[0.040]; Gillingham[0.040]; Torquay[0.040]; Crewe[0.040]; Cambridge[0.040]; Notts[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.053:0.053[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.048]; league[0.047]; League[0.047]; Soccer[0.045]; soccer[0.045]; played[0.045]; Portsmouth[0.044]; Wolverhampton[0.044]; Norwich[0.044]; Park[0.043]; Barnsley[0.043]; Stoke[0.043]; Bolton[0.043]; Swindon[0.043]; Division[0.043]; won[0.043]; Rangers[0.042]; Tranmere[0.042]; Bradford[0.042]; Saturday[0.042]; goals[0.042]; goals[0.042]; Queens[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Bromwich ==> ENTITY: \u001b[32mWest Bromwich Albion F.C.\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.201:0.201[0]; log p(e|m)= -1.501:-1.501[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.044]; Birmingham[0.044]; Wolverhampton[0.043]; Sheffield[0.043]; City[0.043]; Bradford[0.043]; Millwall[0.042]; Oxford[0.042]; Manchester[0.042]; Portsmouth[0.042]; Swindon[0.042]; Bury[0.041]; Oldham[0.041]; Brentford[0.041]; Reading[0.041]; Chesterfield[0.041]; Ipswich[0.041]; Blackpool[0.041]; Southend[0.041]; Shrewsbury[0.040]; United[0.040]; Division[0.040]; Plymouth[0.040]; Grimsby[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Doncaster ==> ENTITY: \u001b[32mDoncaster Rovers F.C.\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.192:0.192[0]; log p(e|m)= -2.957:-2.957[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSwansea[0.051]; Rochdale[0.051]; Carlisle[0.051]; Cardiff[0.050]; Hereford[0.050]; Scunthorpe[0.050]; Exeter[0.049]; Colchester[0.049]; Scarborough[0.048]; Darlington[0.048]; Northampton[0.048]; Torquay[0.048]; Chester[0.047]; Brighton[0.047]; Barnet[0.047]; Mansfield[0.047]; Cambridge[0.046]; United[0.045]; Lincoln[0.044]; Orient[0.044]; Leyton[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brighton ==> ENTITY: \u001b[32mBrighton & Hove Albion F.C.\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.186:0.186[0]; log p(e|m)= -3.124:-3.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorthampton[0.057]; Swansea[0.057]; Carlisle[0.056]; Exeter[0.055]; Colchester[0.054]; Chester[0.054]; Scunthorpe[0.054]; Rochdale[0.054]; Cambridge[0.053]; Hereford[0.052]; Scarborough[0.052]; Barnet[0.052]; Darlington[0.051]; Mansfield[0.051]; United[0.051]; Doncaster[0.051]; Leyton[0.049]; Orient[0.048]; Lincoln[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Queens Park Rangers\u001b[39m ==> ENTITY: \u001b[32mQueens Park Rangers F.C.\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; league[0.044]; matches[0.043]; Manchester[0.043]; standings[0.042]; Barnsley[0.042]; Birmingham[0.042]; Portsmouth[0.042]; City[0.042]; Reading[0.042]; Soccer[0.042]; Bolton[0.041]; Stoke[0.041]; Ipswich[0.041]; Norwich[0.041]; Standings[0.041]; played[0.041]; Tranmere[0.041]; Swindon[0.041]; Division[0.041]; Wolverhampton[0.041]; won[0.041]; Bradford[0.041]; soccer[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brentford ==> ENTITY: \u001b[32mBrentford F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.213:0.213[0]; log p(e|m)= -0.361:-0.361[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.046]; Charlton[0.045]; Watford[0.044]; Sheffield[0.043]; Millwall[0.043]; Burnley[0.042]; Southend[0.042]; Reading[0.041]; Bournemouth[0.041]; Blackpool[0.041]; Bury[0.041]; Manchester[0.041]; Shrewsbury[0.041]; Oldham[0.041]; Bristol[0.041]; Peterborough[0.041]; Chesterfield[0.041]; Huddersfield[0.041]; City[0.041]; Grimsby[0.041]; West[0.040]; Plymouth[0.040]; Rovers[0.040]; United[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Peterborough ==> ENTITY: \u001b[32mPeterborough United F.C.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.210:0.210[0]; log p(e|m)= -3.170:-3.170[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Bristol[0.043]; Bristol[0.043]; Shrewsbury[0.043]; Walsall[0.042]; Stockport[0.042]; Watford[0.042]; Bury[0.042]; Rotherham[0.042]; Brentford[0.042]; Luton[0.042]; Plymouth[0.042]; Chesterfield[0.042]; Gillingham[0.041]; Millwall[0.041]; Wycombe[0.041]; Burnley[0.041]; Bournemouth[0.041]; Division[0.041]; Division[0.041]; Wrexham[0.041]; York[0.041]; Blackpool[0.041]; City[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Preston ==> ENTITY: \u001b[32mPreston North End F.C.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.170:0.170[0]; log p(e|m)= -3.194:-3.194[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWatford[0.044]; Walsall[0.043]; Bristol[0.043]; Bristol[0.043]; Burnley[0.043]; Stockport[0.043]; Chesterfield[0.042]; City[0.042]; Brentford[0.042]; Gillingham[0.042]; Millwall[0.042]; Bury[0.042]; Peterborough[0.042]; Shrewsbury[0.042]; Blackpool[0.041]; Rotherham[0.041]; Bournemouth[0.041]; Wycombe[0.041]; York[0.041]; Luton[0.041]; Crewe[0.041]; Wrexham[0.040]; Rovers[0.040]; County[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Port Vale\u001b[39m ==> ENTITY: \u001b[32mPort Vale F.C.\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSheffield[0.043]; Southend[0.043]; Charlton[0.043]; Barnsley[0.043]; Portsmouth[0.043]; City[0.043]; Norwich[0.043]; Ipswich[0.042]; Swindon[0.042]; Oldham[0.042]; Park[0.042]; Manchester[0.042]; Reading[0.042]; Grimsby[0.041]; Birmingham[0.041]; United[0.041]; Tranmere[0.041]; Bradford[0.041]; Huddersfield[0.041]; Division[0.040]; Wolverhampton[0.040]; Stoke[0.040]; Rangers[0.040]; Bolton[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hull ==> ENTITY: \u001b[32mHull City A.F.C.\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.200:0.200[0]; log p(e|m)= -3.772:-3.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWalsall[0.044]; Cardiff[0.044]; Bristol[0.043]; Stockport[0.043]; Scunthorpe[0.043]; Carlisle[0.043]; Rotherham[0.043]; City[0.042]; Scarborough[0.042]; Gillingham[0.042]; Wycombe[0.042]; Crewe[0.042]; Luton[0.041]; Wrexham[0.041]; Hartlepool[0.041]; Northampton[0.041]; Fulham[0.041]; Exeter[0.041]; Darlington[0.041]; Wigan[0.041]; Barnet[0.040]; Cambridge[0.040]; United[0.040]; Torquay[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hartlepool ==> ENTITY: \u001b[32mHartlepool United F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.176:0.176[0]; log p(e|m)= -1.754:-1.754[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.045]; Stockport[0.044]; Cardiff[0.043]; Bristol[0.043]; Carlisle[0.043]; Walsall[0.043]; Rotherham[0.043]; City[0.042]; Wycombe[0.042]; Darlington[0.042]; Scunthorpe[0.042]; Wrexham[0.041]; United[0.041]; Chester[0.041]; Exeter[0.041]; Scarborough[0.041]; Northampton[0.041]; Barnet[0.041]; Luton[0.041]; Torquay[0.041]; Notts[0.041]; Gillingham[0.040]; Cambridge[0.040]; Wigan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tranmere ==> ENTITY: \u001b[32mTranmere Rovers F.C.\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.204:0.204[0]; log p(e|m)= -0.600:-0.600[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; matches[0.044]; league[0.043]; Portsmouth[0.043]; Barnsley[0.043]; Birmingham[0.043]; Ipswich[0.042]; played[0.042]; Swindon[0.042]; Norwich[0.042]; Huddersfield[0.042]; Reading[0.042]; Park[0.042]; Soccer[0.042]; Bolton[0.042]; soccer[0.041]; Wolverhampton[0.041]; Stoke[0.041]; Bradford[0.040]; won[0.040]; Port[0.040]; goals[0.040]; goals[0.040]; Division[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oldham ==> ENTITY: \u001b[32mOldham Athletic A.F.C.\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.204:0.204[0]; log p(e|m)= -2.865:-2.865[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBradford[0.045]; Huddersfield[0.045]; Charlton[0.043]; Chesterfield[0.043]; Bury[0.043]; Millwall[0.043]; Plymouth[0.043]; Sheffield[0.042]; Southend[0.042]; Brentford[0.042]; Shrewsbury[0.041]; Swindon[0.041]; Ipswich[0.041]; City[0.041]; Manchester[0.041]; Reading[0.041]; Birmingham[0.041]; Portsmouth[0.041]; United[0.040]; Division[0.040]; Grimsby[0.040]; Oxford[0.040]; Blackpool[0.040]; York[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swansea ==> ENTITY: \u001b[32mSwansea City A.F.C.\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.160:0.160[0]; log p(e|m)= -2.604:-2.604[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNorthampton[0.062]; Colchester[0.062]; Rochdale[0.062]; Exeter[0.060]; Chester[0.060]; Mansfield[0.060]; Barnet[0.060]; Brighton[0.060]; Hereford[0.059]; United[0.059]; Darlington[0.058]; Scarborough[0.058]; Doncaster[0.057]; Orient[0.057]; Cambridge[0.056]; Lincoln[0.055]; Leyton[0.053]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rochdale ==> ENTITY: \u001b[32mRochdale A.F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.952:-0.952[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nColchester[0.070]; Swansea[0.070]; Darlington[0.070]; Chester[0.069]; Exeter[0.068]; Mansfield[0.068]; Doncaster[0.068]; Hereford[0.067]; Barnet[0.067]; Brighton[0.066]; United[0.066]; Cambridge[0.065]; Lincoln[0.063]; Orient[0.061]; Leyton[0.060]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = English ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.102:0.102[0]; log p(e|m)= -1.355:-1.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEnglish[0.048]; League[0.047]; league[0.046]; matches[0.044]; Division[0.044]; Swindon[0.043]; Barnsley[0.043]; played[0.042]; Norwich[0.042]; Portsmouth[0.042]; Bradford[0.042]; London[0.041]; Soccer[0.041]; Wolverhampton[0.040]; Stoke[0.040]; Tranmere[0.040]; soccer[0.040]; won[0.040]; Park[0.040]; Queens[0.040]; Saturday[0.039]; standings[0.039]; Bolton[0.038]; Standings[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bury ==> ENTITY: \u001b[32mBury F.C.\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.196:0.196[0]; log p(e|m)= -0.722:-0.722[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.044]; Peterborough[0.044]; Sheffield[0.043]; Charlton[0.043]; Grimsby[0.043]; Burnley[0.043]; Watford[0.042]; Chesterfield[0.042]; Southend[0.042]; Bournemouth[0.042]; Oldham[0.042]; City[0.042]; Brentford[0.042]; Blackpool[0.042]; Huddersfield[0.041]; Millwall[0.041]; Manchester[0.041]; Shrewsbury[0.041]; Crewe[0.041]; Bristol[0.040]; Oxford[0.040]; Plymouth[0.040]; United[0.040]; Division[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portsmouth ==> ENTITY: \u001b[32mPortsmouth F.C.\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.207:0.207[0]; log p(e|m)= -1.231:-1.231[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; league[0.043]; Manchester[0.043]; Birmingham[0.043]; Palace[0.043]; Norwich[0.043]; Sheffield[0.043]; matches[0.043]; Ipswich[0.042]; Bolton[0.042]; Barnsley[0.042]; City[0.042]; Reading[0.042]; Tranmere[0.041]; Park[0.041]; Stoke[0.041]; Swindon[0.041]; United[0.041]; Oldham[0.040]; English[0.040]; Wolverhampton[0.040]; played[0.040]; Division[0.040]; Huddersfield[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Crewe ==> ENTITY: \u001b[32mCrewe Alexandra F.C.\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.191:0.191[0]; log p(e|m)= -2.830:-2.830[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Hull[0.044]; Watford[0.043]; Bristol[0.043]; Bristol[0.043]; Shrewsbury[0.043]; Chesterfield[0.042]; Walsall[0.042]; Burnley[0.042]; Stockport[0.042]; Wycombe[0.042]; Peterborough[0.042]; Gillingham[0.041]; Bury[0.041]; Bournemouth[0.041]; Millwall[0.041]; Luton[0.041]; York[0.041]; Blackpool[0.041]; Rotherham[0.041]; Wrexham[0.041]; City[0.040]; Rovers[0.040]; Notts[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gillingham ==> ENTITY: \u001b[32mGillingham F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.202:0.202[0]; log p(e|m)= -0.257:-0.257[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Hull[0.044]; Watford[0.043]; Walsall[0.043]; Stockport[0.043]; Peterborough[0.042]; City[0.042]; Millwall[0.042]; Chesterfield[0.042]; Burnley[0.042]; Notts[0.042]; Rotherham[0.041]; Wycombe[0.041]; Bristol[0.041]; Bristol[0.041]; Blackpool[0.041]; Shrewsbury[0.041]; Bournemouth[0.041]; Luton[0.041]; Crewe[0.041]; Wrexham[0.041]; Hartlepool[0.040]; County[0.040]; Rovers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Birmingham ==> ENTITY: \u001b[32mBirmingham City F.C.\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.166:0.166[0]; log p(e|m)= -3.101:-3.101[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.045]; Wolverhampton[0.045]; Sheffield[0.044]; Manchester[0.043]; City[0.043]; Oxford[0.042]; Norwich[0.042]; Southend[0.042]; Oldham[0.042]; Barnsley[0.042]; Swindon[0.042]; Portsmouth[0.042]; Bradford[0.042]; Reading[0.042]; Ipswich[0.041]; Tranmere[0.041]; Plymouth[0.041]; Bolton[0.041]; Park[0.040]; Huddersfield[0.040]; West[0.040]; United[0.040]; Grimsby[0.040]; Queens[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Torquay ==> ENTITY: \u001b[32mTorquay United F.C.\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.191:0.191[0]; log p(e|m)= -2.017:-2.017[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.045]; Carlisle[0.044]; Stockport[0.043]; Walsall[0.043]; Northampton[0.043]; Bristol[0.043]; City[0.042]; Scunthorpe[0.042]; Exeter[0.042]; Cardiff[0.042]; Chester[0.042]; Rotherham[0.042]; Darlington[0.042]; Wycombe[0.042]; Wrexham[0.041]; Division[0.041]; Luton[0.041]; Hartlepool[0.041]; United[0.040]; Doncaster[0.040]; Barnet[0.040]; Scarborough[0.040]; Notts[0.039]; Wigan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cardiff ==> ENTITY: \u001b[32mCardiff City F.C.\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.194:0.194[0]; log p(e|m)= -3.474:-3.474[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.045]; Bristol[0.044]; Carlisle[0.044]; Rotherham[0.043]; Exeter[0.043]; Walsall[0.043]; Stockport[0.043]; Northampton[0.042]; Wrexham[0.042]; Scunthorpe[0.042]; Luton[0.042]; Doncaster[0.042]; Wycombe[0.042]; Darlington[0.041]; City[0.041]; Wigan[0.041]; Hartlepool[0.041]; Scarborough[0.041]; Chester[0.040]; Barnet[0.040]; Fulham[0.040]; Torquay[0.039]; Cambridge[0.039]; United[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sheffield United\u001b[39m ==> ENTITY: \u001b[32mSheffield United F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCharlton[0.044]; Millwall[0.044]; Birmingham[0.044]; Burnley[0.043]; Southend[0.043]; Manchester[0.043]; Bury[0.043]; Brentford[0.042]; Oldham[0.042]; City[0.042]; Blackpool[0.042]; Ipswich[0.042]; Reading[0.042]; Portsmouth[0.042]; Shrewsbury[0.042]; Chesterfield[0.042]; Huddersfield[0.041]; Grimsby[0.041]; Plymouth[0.040]; Bradford[0.040]; Division[0.039]; Palace[0.039]; Vale[0.039]; York[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wolverhampton ==> ENTITY: \u001b[32mWolverhampton Wanderers F.C.\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.175:0.175[0]; log p(e|m)= -2.273:-2.273[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; matches[0.044]; league[0.043]; Birmingham[0.043]; Bradford[0.043]; Oxford[0.043]; Manchester[0.042]; Bolton[0.042]; Huddersfield[0.042]; Reading[0.042]; Barnsley[0.042]; Swindon[0.042]; City[0.041]; Soccer[0.041]; Portsmouth[0.041]; Tranmere[0.041]; Norwich[0.041]; Ipswich[0.041]; Stoke[0.041]; Park[0.041]; played[0.040]; won[0.040]; soccer[0.040]; London[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chesterfield ==> ENTITY: \u001b[32mChesterfield F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.197:0.197[0]; log p(e|m)= -1.269:-1.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Oxford[0.043]; Bristol[0.043]; Oldham[0.042]; Shrewsbury[0.042]; Sheffield[0.042]; York[0.042]; Watford[0.042]; Burnley[0.042]; Manchester[0.042]; Bury[0.042]; Charlton[0.042]; Millwall[0.042]; Southend[0.042]; Plymouth[0.041]; Bournemouth[0.041]; Brentford[0.041]; Peterborough[0.041]; Gillingham[0.041]; City[0.041]; Crewe[0.041]; Blackpool[0.040]; Grimsby[0.040]; United[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = English ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.098:0.098[0]; log p(e|m)= -1.355:-1.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEnglish[0.048]; League[0.047]; league[0.046]; matches[0.045]; Division[0.044]; Swindon[0.043]; Barnsley[0.043]; played[0.042]; Norwich[0.042]; Bradford[0.042]; London[0.041]; Soccer[0.041]; Wolverhampton[0.041]; Stoke[0.040]; Tranmere[0.040]; soccer[0.040]; won[0.040]; Park[0.040]; Queens[0.040]; Saturday[0.040]; standings[0.039]; Bolton[0.038]; Standings[0.038]; Rangers[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Stoke ==> ENTITY: \u001b[32mStoke City F.C.\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.175:0.175[0]; log p(e|m)= -0.341:-0.341[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.047]; league[0.045]; League[0.044]; Portsmouth[0.044]; Norwich[0.043]; Ipswich[0.042]; Standings[0.042]; Barnsley[0.042]; Bradford[0.042]; played[0.042]; Wolverhampton[0.042]; Swindon[0.042]; Tranmere[0.041]; Soccer[0.041]; soccer[0.041]; standings[0.041]; Bolton[0.041]; Park[0.041]; points[0.040]; Saturday[0.039]; won[0.039]; goals[0.039]; goals[0.039]; London[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cambridge United\u001b[39m ==> ENTITY: \u001b[32mCambridge United F.C.\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.049]; Swansea[0.046]; Colchester[0.046]; Cardiff[0.046]; Northampton[0.045]; Carlisle[0.045]; Scunthorpe[0.045]; Darlington[0.045]; Rochdale[0.045]; Exeter[0.044]; Barnet[0.043]; Mansfield[0.043]; Scarborough[0.043]; Hereford[0.043]; Brighton[0.043]; Torquay[0.043]; Hartlepool[0.043]; Doncaster[0.042]; Chester[0.042]; Fulham[0.042]; Lincoln[0.040]; Leyton[0.039]; Orient[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Notts County\u001b[39m ==> ENTITY: \u001b[32mNotts County F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.217:0.217[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPreston[0.045]; Hull[0.044]; Walsall[0.043]; Division[0.043]; Burnley[0.043]; Watford[0.042]; Gillingham[0.042]; Millwall[0.042]; Stockport[0.042]; Peterborough[0.042]; Rotherham[0.042]; Luton[0.042]; Shrewsbury[0.041]; Bournemouth[0.041]; Wycombe[0.041]; Wrexham[0.041]; Crewe[0.041]; Bristol[0.041]; Bristol[0.041]; City[0.040]; Rovers[0.040]; Blackpool[0.040]; Hartlepool[0.040]; Torquay[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bristol City\u001b[39m ==> ENTITY: \u001b[32mBristol City F.C.\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBristol[0.046]; Preston[0.044]; Hull[0.044]; Watford[0.043]; Cardiff[0.043]; Walsall[0.043]; Gillingham[0.042]; Stockport[0.042]; Shrewsbury[0.042]; Burnley[0.042]; Bournemouth[0.042]; Peterborough[0.041]; Crewe[0.041]; Wycombe[0.041]; Blackpool[0.041]; Rotherham[0.041]; Luton[0.041]; Rovers[0.041]; Division[0.040]; Wrexham[0.040]; Torquay[0.040]; Hartlepool[0.040]; Notts[0.040]; Wigan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Reading ==> ENTITY: \u001b[32mReading F.C.\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.179:0.179[0]; log p(e|m)= -1.201:-1.201[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOxford[0.046]; Sheffield[0.044]; Charlton[0.044]; Wolverhampton[0.043]; Bolton[0.042]; Swindon[0.042]; Birmingham[0.042]; Manchester[0.042]; Norwich[0.042]; Brentford[0.042]; Portsmouth[0.042]; City[0.042]; Ipswich[0.041]; West[0.041]; Tranmere[0.041]; Bradford[0.041]; Park[0.041]; Southend[0.041]; Oldham[0.041]; Plymouth[0.040]; Huddersfield[0.040]; Palace[0.040]; Rangers[0.040]; Grimsby[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scunthorpe ==> ENTITY: \u001b[32mScunthorpe United F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.199:0.199[0]; log p(e|m)= -1.510:-1.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.047]; Stockport[0.044]; Cardiff[0.044]; Walsall[0.044]; Rotherham[0.043]; Carlisle[0.043]; Hartlepool[0.043]; Darlington[0.043]; Luton[0.042]; Northampton[0.042]; Wrexham[0.042]; Wycombe[0.042]; Doncaster[0.041]; United[0.041]; Exeter[0.041]; Chester[0.041]; Brighton[0.041]; Torquay[0.040]; Wigan[0.040]; Scarborough[0.040]; Barnet[0.040]; Division[0.039]; Fulham[0.039]; Cambridge[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Carlisle ==> ENTITY: \u001b[32mCarlisle United F.C.\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.177:0.177[0]; log p(e|m)= -3.612:-3.612[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHull[0.045]; Stockport[0.043]; Cardiff[0.043]; Rotherham[0.043]; Walsall[0.043]; Exeter[0.043]; Wycombe[0.042]; Northampton[0.042]; United[0.042]; Hartlepool[0.042]; Hereford[0.042]; Scarborough[0.042]; Scunthorpe[0.042]; Wrexham[0.042]; Division[0.042]; Doncaster[0.042]; Darlington[0.041]; Chester[0.041]; Brighton[0.040]; Cambridge[0.040]; Torquay[0.040]; Barnet[0.040]; Wigan[0.039]; Fulham[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chester ==> ENTITY: \u001b[32mChester City F.C.\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.191:0.191[0]; log p(e|m)= -3.219:-3.219[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCarlisle[0.048]; Swansea[0.048]; Colchester[0.047]; Darlington[0.047]; Scunthorpe[0.047]; Rochdale[0.047]; Northampton[0.047]; Cardiff[0.047]; United[0.046]; Exeter[0.046]; Torquay[0.045]; Barnet[0.045]; Hartlepool[0.045]; Hereford[0.045]; Doncaster[0.045]; Brighton[0.045]; Scarborough[0.045]; Cambridge[0.044]; Mansfield[0.044]; Orient[0.043]; Lincoln[0.042]; Leyton[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Grimsby ==> ENTITY: \u001b[32mGrimsby Town F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.212:0.212[0]; log p(e|m)= -2.604:-2.604[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSheffield[0.044]; Ipswich[0.043]; Southend[0.043]; Plymouth[0.043]; Burnley[0.043]; Bournemouth[0.043]; Charlton[0.043]; Chesterfield[0.043]; Brentford[0.043]; Huddersfield[0.042]; Shrewsbury[0.042]; Bury[0.042]; Oldham[0.042]; Manchester[0.042]; Birmingham[0.042]; Millwall[0.042]; Blackpool[0.042]; City[0.041]; Reading[0.041]; Oxford[0.039]; York[0.039]; Division[0.039]; United[0.039]; West[0.038]; \t\n\r [===============================>..............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 6s901ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3355/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1089testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jansher Khan\u001b[39m ==> ENTITY: \u001b[32mJansher Khan\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.047]; Open[0.045]; Open[0.045]; Squash[0.045]; number[0.045]; Semifinal[0.044]; Semifinal[0.044]; Australia[0.043]; Australia[0.043]; beat[0.043]; beat[0.043]; Saturday[0.042]; Scotland[0.041]; Nicol[0.041]; Eyles[0.040]; Hong[0.039]; Hong[0.039]; Hong[0.039]; Kong[0.039]; Kong[0.039]; Kong[0.039]; Results[0.039]; seeding[0.039]; Rodney[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.061:0.061[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.053]; Pakistan[0.046]; number[0.045]; Scotland[0.044]; Open[0.044]; Open[0.044]; Peter[0.043]; Squash[0.042]; Jansher[0.042]; Nicol[0.042]; Saturday[0.041]; Khan[0.041]; Rodney[0.041]; Anthony[0.041]; beat[0.040]; beat[0.040]; Eyles[0.040]; Hill[0.040]; seeding[0.039]; Semifinal[0.039]; Semifinal[0.039]; Hong[0.038]; Hong[0.038]; Hong[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scotland ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.241:0.241[0]; local()= 0.022:0.022[0]; log p(e|m)= -0.229:-0.229[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.048]; Australia[0.047]; Australia[0.047]; number[0.045]; Squash[0.045]; Nicol[0.045]; Saturday[0.043]; Jansher[0.043]; Khan[0.042]; Peter[0.042]; Rodney[0.042]; beat[0.041]; beat[0.041]; Hill[0.041]; Anthony[0.040]; seeding[0.040]; Semifinal[0.040]; Semifinal[0.040]; Open[0.040]; Open[0.040]; Eyles[0.039]; results[0.038]; prefix[0.037]; denotes[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Peter Nicol\u001b[39m ==> ENTITY: \u001b[32mPeter Nicol\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.122:0.122[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJansher[0.051]; Squash[0.051]; Open[0.049]; Open[0.049]; Scotland[0.048]; Semifinal[0.048]; Semifinal[0.048]; Australia[0.047]; Australia[0.047]; number[0.047]; Khan[0.046]; results[0.044]; beat[0.044]; beat[0.044]; seeding[0.044]; Pakistan[0.044]; Saturday[0.043]; Rodney[0.042]; Results[0.042]; prefix[0.041]; Eyles[0.041]; Hill[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPakistan\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPakistan national cricket team\u001b[39m <---> \u001b[32mPakistan\u001b[39m\t\nSCORES: global= 0.244:0.241[\u001b[31m0.003\u001b[39m]; local()= 0.071:0.016[\u001b[31m0.055\u001b[39m]; log p(e|m)= -2.749:-0.219[\u001b[32m2.529\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSquash[0.050]; Australia[0.046]; Australia[0.046]; Jansher[0.046]; Khan[0.046]; Semifinal[0.045]; Semifinal[0.045]; Open[0.044]; Open[0.044]; Saturday[0.043]; number[0.040]; beat[0.040]; beat[0.040]; Results[0.040]; seeding[0.040]; Scotland[0.039]; Eyles[0.039]; Nicol[0.039]; Hong[0.039]; Hong[0.039]; Hong[0.039]; Hill[0.038]; results[0.038]; Rodney[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.061:0.061[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.053]; Pakistan[0.046]; number[0.045]; Scotland[0.045]; Open[0.044]; Open[0.044]; Peter[0.043]; Squash[0.043]; Jansher[0.042]; Nicol[0.042]; Saturday[0.042]; Khan[0.041]; Rodney[0.041]; Anthony[0.041]; beat[0.040]; beat[0.040]; Eyles[0.040]; Hill[0.040]; seeding[0.039]; Semifinal[0.038]; Semifinal[0.038]; Hong[0.038]; Hong[0.038]; Hong[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rodney Eyles\u001b[39m ==> ENTITY: \u001b[32mRodney Eyles\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJansher[0.047]; Australia[0.046]; Australia[0.046]; Squash[0.045]; Open[0.044]; Open[0.044]; Semifinal[0.043]; Semifinal[0.043]; Nicol[0.043]; Peter[0.041]; Pakistan[0.041]; Anthony[0.041]; Kong[0.040]; Kong[0.040]; Kong[0.040]; Scotland[0.040]; Hong[0.040]; Hong[0.040]; Hong[0.040]; seeding[0.039]; Khan[0.039]; Hill[0.038]; beat[0.038]; beat[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hong Kong ==> ENTITY: \u001b[32mHong Kong\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.061:0.061[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHong[0.047]; Hong[0.047]; Kong[0.047]; Kong[0.047]; Pakistan[0.045]; Australia[0.045]; Australia[0.045]; Squash[0.043]; Open[0.043]; Open[0.043]; Results[0.043]; Jansher[0.043]; Scotland[0.041]; number[0.040]; Hill[0.039]; Khan[0.039]; Saturday[0.039]; Nicol[0.038]; beat[0.038]; beat[0.038]; Semifinal[0.038]; Semifinal[0.038]; Eyles[0.037]; seeding[0.037]; \t\n\r [===============================>..............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 6s857ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3363/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1110testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazil ==> ENTITY: \u001b[32mBrazil national football team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.140:0.140[0]; log p(e|m)= -2.590:-2.590[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.049]; Brazil[0.049]; soccer[0.048]; Soccer[0.048]; Netherlands[0.046]; Netherlands[0.046]; International[0.042]; Draw[0.042]; time[0.041]; drew[0.041]; friendly[0.041]; minute[0.040]; Friendly[0.040]; Dutch[0.040]; half[0.040]; Saturday[0.040]; Goncalves[0.039]; Marcello[0.039]; Ronald[0.039]; Giovanni[0.038]; Scorers[0.038]; Van[0.038]; Amsterdam[0.038]; Gastel[0.036]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mThe Netherlands\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNetherlands\u001b[39m <---> \u001b[32mNetherlands national football team\u001b[39m\t\nSCORES: global= 0.254:0.251[\u001b[31m0.003\u001b[39m]; local()= 0.107:0.140[\u001b[32m0.033\u001b[39m]; log p(e|m)= -0.108:-2.976[\u001b[31m2.868\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.051]; Soccer[0.049]; soccer[0.048]; Dutch[0.047]; Brazil[0.044]; Brazil[0.044]; Brazil[0.044]; Van[0.044]; Friendly[0.043]; Amsterdam[0.042]; drew[0.042]; time[0.041]; minute[0.041]; friendly[0.041]; International[0.041]; Saturday[0.040]; half[0.040]; Draw[0.039]; Scorers[0.038]; Boer[0.038]; Ronald[0.037]; Giovanni[0.036]; Gastel[0.035]; Goncalves[0.035]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Van Gastel\u001b[39m ==> ENTITY: \u001b[32mJean-Paul van Gastel\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.129:0.129[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.047]; Netherlands[0.047]; Netherlands[0.047]; Giovanni[0.045]; Soccer[0.045]; Dutch[0.044]; Amsterdam[0.044]; Scorers[0.042]; Saturday[0.041]; Marcello[0.041]; International[0.041]; drew[0.041]; Ronald[0.041]; Friendly[0.041]; Goncalves[0.041]; minute[0.040]; Brazil[0.040]; Brazil[0.040]; Brazil[0.040]; time[0.039]; Draw[0.039]; Boer[0.039]; friendly[0.039]; half[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ronald de Boer\u001b[39m ==> ENTITY: \u001b[32mRonald de Boer\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.102:0.102[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.048]; Soccer[0.046]; Netherlands[0.045]; Netherlands[0.045]; Dutch[0.044]; Brazil[0.043]; Brazil[0.043]; Brazil[0.043]; time[0.043]; Scorers[0.042]; Amsterdam[0.041]; minute[0.041]; friendly[0.041]; Van[0.041]; half[0.041]; Gastel[0.040]; Goncalves[0.040]; International[0.040]; Saturday[0.039]; Giovanni[0.039]; pen[0.039]; Marcello[0.039]; Friendly[0.039]; Draw[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazil ==> ENTITY: \u001b[32mBrazil national football team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.140:0.140[0]; log p(e|m)= -2.590:-2.590[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.049]; Brazil[0.049]; soccer[0.048]; Soccer[0.048]; Netherlands[0.046]; Netherlands[0.046]; International[0.042]; Draw[0.042]; time[0.041]; drew[0.041]; friendly[0.041]; minute[0.040]; Friendly[0.040]; Dutch[0.040]; half[0.040]; Saturday[0.040]; Goncalves[0.039]; Marcello[0.039]; Ronald[0.039]; Giovanni[0.038]; Scorers[0.038]; Van[0.038]; Amsterdam[0.038]; Gastel[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Amsterdam ==> ENTITY: \u001b[32mAmsterdam\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.096:0.096[0]; log p(e|m)= -0.064:-0.064[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.050]; Netherlands[0.050]; Dutch[0.045]; time[0.044]; Van[0.044]; Ronald[0.044]; Soccer[0.044]; soccer[0.043]; drew[0.042]; International[0.041]; Brazil[0.041]; Brazil[0.041]; Brazil[0.041]; friendly[0.041]; minute[0.041]; Saturday[0.041]; half[0.040]; Boer[0.040]; Friendly[0.040]; pen[0.038]; Draw[0.038]; Giovanni[0.037]; Gastel[0.036]; Goncalves[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Netherlands ==> ENTITY: \u001b[32mNetherlands national football team\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.140:0.140[0]; log p(e|m)= -2.659:-2.659[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.050]; Soccer[0.049]; soccer[0.048]; Dutch[0.046]; Brazil[0.046]; Brazil[0.046]; Brazil[0.046]; Van[0.043]; Amsterdam[0.042]; drew[0.041]; time[0.041]; minute[0.041]; International[0.041]; friendly[0.040]; Saturday[0.040]; half[0.040]; Friendly[0.039]; Ronald[0.039]; Draw[0.038]; Goncalves[0.038]; Scorers[0.038]; Boer[0.037]; Gastel[0.036]; Giovanni[0.035]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brazil ==> ENTITY: \u001b[32mBrazil national football team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.140:0.140[0]; log p(e|m)= -2.590:-2.590[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.049]; Brazil[0.049]; soccer[0.048]; Soccer[0.048]; Netherlands[0.046]; Netherlands[0.046]; International[0.042]; Draw[0.042]; time[0.041]; drew[0.041]; friendly[0.041]; minute[0.040]; Friendly[0.040]; Dutch[0.040]; half[0.040]; Saturday[0.040]; Goncalves[0.039]; Marcello[0.039]; Ronald[0.039]; Giovanni[0.038]; Scorers[0.038]; Van[0.038]; Amsterdam[0.038]; Gastel[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Giovanni ==> ENTITY: \u001b[32mGiovanni Silva de Oliveira\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.124:0.124[0]; log p(e|m)= -1.309:-1.309[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.047]; Brazil[0.047]; Brazil[0.047]; Marcello[0.045]; Soccer[0.044]; soccer[0.044]; Amsterdam[0.043]; Dutch[0.043]; Goncalves[0.043]; Netherlands[0.042]; Netherlands[0.042]; drew[0.042]; minute[0.041]; time[0.040]; friendly[0.040]; Saturday[0.040]; Ronald[0.039]; Friendly[0.039]; Gastel[0.039]; Scorers[0.039]; International[0.039]; Draw[0.039]; half[0.039]; pen[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mDutch\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNetherlands\u001b[39m <---> \u001b[32mNetherlands national football team\u001b[39m\t\nSCORES: global= 0.254:0.238[\u001b[31m0.016\u001b[39m]; local()= 0.109:0.142[\u001b[32m0.033\u001b[39m]; log p(e|m)= -0.794:-4.711[\u001b[31m3.916\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.051]; Netherlands[0.051]; soccer[0.045]; Brazil[0.045]; Brazil[0.045]; Brazil[0.045]; Van[0.044]; Soccer[0.044]; Amsterdam[0.043]; Boer[0.042]; drew[0.042]; friendly[0.041]; minute[0.041]; time[0.041]; International[0.040]; Saturday[0.040]; Friendly[0.040]; half[0.039]; Ronald[0.039]; Goncalves[0.038]; Draw[0.038]; Giovanni[0.037]; Scorers[0.037]; Gastel[0.037]; \t\n\r [===============================>..............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 6s788ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3373/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1030testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Commerce Department\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Commerce\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDepartment[0.051]; January[0.045]; officials[0.045]; months[0.044]; Consumer[0.044]; Consumer[0.044]; month[0.044]; adjusted[0.043]; Washington[0.043]; July[0.043]; July[0.043]; July[0.043]; July[0.043]; June[0.043]; June[0.043]; spending[0.043]; spending[0.043]; said[0.043]; said[0.043]; Spending[0.042]; growth[0.042]; Glenn[0.042]; revised[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.075:0.075[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndollar[0.044]; dollar[0.044]; generally[0.044]; orders[0.043]; orders[0.043]; previously[0.043]; Orders[0.042]; money[0.041]; billion[0.041]; billion[0.041]; highest[0.041]; October[0.041]; followed[0.041]; earned[0.041]; earned[0.041]; climbed[0.041]; July[0.041]; July[0.041]; July[0.041]; July[0.041]; July[0.041]; July[0.041]; weekly[0.041]; sales[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.110:0.110[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDepartment[0.045]; Department[0.045]; Glenn[0.044]; said[0.044]; said[0.044]; July[0.044]; July[0.044]; July[0.044]; June[0.044]; June[0.044]; month[0.043]; Somerville[0.043]; rate[0.043]; rate[0.043]; annual[0.043]; spending[0.043]; spending[0.043]; revised[0.043]; January[0.043]; income[0.043]; Commerce[0.042]; officials[0.042]; Friday[0.042]; \t\n\r [===============================>..............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 6s771ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3378/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1004testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paramaribo\u001b[39m ==> ENTITY: \u001b[32mParamaribo\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nParamaribo[0.047]; Surinamese[0.046]; Surinamese[0.046]; town[0.044]; Surinam[0.043]; Surinam[0.043]; Netherlands[0.043]; born[0.043]; east[0.042]; reached[0.042]; miles[0.042]; Friday[0.041]; Freddy[0.041]; rebel[0.041]; Ronny[0.041]; days[0.040]; km[0.040]; lost[0.040]; said[0.039]; police[0.039]; police[0.039]; dropped[0.039]; dropped[0.039]; mining[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Surinamese\u001b[39m ==> ENTITY: \u001b[32mSuriname\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSurinamese[0.049]; Surinam[0.048]; Surinam[0.048]; Netherlands[0.047]; Paramaribo[0.043]; Paramaribo[0.043]; second[0.041]; east[0.041]; born[0.041]; days[0.041]; lost[0.041]; time[0.040]; Brunswijk[0.040]; Brunswijk[0.040]; Brunswijk[0.040]; Brunswijk[0.040]; Brunswijk[0.040]; town[0.040]; said[0.040]; rebel[0.039]; dropped[0.039]; dropped[0.039]; custody[0.039]; guerrilla[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Desi Bouterse\u001b[39m ==> ENTITY: \u001b[32mDési Bouterse\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.128:0.128[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.045]; military[0.044]; charge[0.044]; accused[0.043]; Brunswijk[0.043]; Brunswijk[0.043]; Brunswijk[0.043]; Brunswijk[0.043]; Surinamese[0.043]; denied[0.042]; rebel[0.042]; months[0.042]; custody[0.042]; served[0.041]; charged[0.041]; regime[0.041]; Freddy[0.040]; days[0.040]; murder[0.040]; Paramaribo[0.040]; led[0.040]; time[0.040]; group[0.040]; second[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paramaribo\u001b[39m ==> ENTITY: \u001b[32mParamaribo\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nParamaribo[0.046]; Surinamese[0.045]; Surinamese[0.045]; served[0.044]; town[0.043]; Surinam[0.043]; Surinam[0.043]; Netherlands[0.043]; born[0.042]; east[0.042]; reached[0.042]; miles[0.041]; Friday[0.041]; Freddy[0.041]; military[0.040]; rebel[0.040]; rebel[0.040]; Ronny[0.040]; businessman[0.040]; days[0.040]; regime[0.040]; months[0.039]; km[0.039]; Bouterse[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Surinam\u001b[39m ==> ENTITY: \u001b[32mSuriname\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSurinam[0.048]; Netherlands[0.047]; Surinamese[0.047]; Surinamese[0.047]; Paramaribo[0.043]; Paramaribo[0.043]; lost[0.042]; Ronny[0.042]; second[0.041]; east[0.041]; dropped[0.041]; dropped[0.041]; born[0.041]; Friday[0.041]; rebel[0.040]; days[0.040]; town[0.040]; Freddy[0.040]; Flamboyant[0.040]; miles[0.039]; free[0.039]; agreement[0.039]; said[0.039]; attempted[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Moengo\u001b[39m ==> ENTITY: \u001b[32mMoengo\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.110:0.110[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntown[0.046]; Surinam[0.044]; Surinam[0.044]; Paramaribo[0.043]; Paramaribo[0.043]; Surinamese[0.043]; Surinamese[0.043]; born[0.043]; mining[0.042]; mining[0.042]; bar[0.042]; Freddy[0.042]; east[0.042]; military[0.041]; rebel[0.041]; rebel[0.041]; visitor[0.040]; businessman[0.040]; regime[0.040]; room[0.040]; days[0.040]; Ronny[0.039]; said[0.039]; custody[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Netherlands ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.250:-0.250[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSurinam[0.044]; Surinam[0.044]; second[0.043]; Surinamese[0.043]; Surinamese[0.043]; military[0.043]; dropped[0.043]; dropped[0.043]; reached[0.042]; town[0.042]; months[0.042]; time[0.042]; lost[0.042]; days[0.041]; late[0.041]; group[0.041]; charge[0.040]; guerrilla[0.040]; Friday[0.040]; free[0.040]; successful[0.040]; east[0.040]; agreement[0.040]; served[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Surinam\u001b[39m ==> ENTITY: \u001b[32mSuriname\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSurinam[0.048]; Netherlands[0.048]; Surinamese[0.047]; Surinamese[0.047]; Paramaribo[0.043]; Paramaribo[0.043]; Ronny[0.042]; east[0.041]; dropped[0.041]; dropped[0.041]; born[0.041]; Friday[0.041]; rebel[0.041]; days[0.041]; town[0.040]; Freddy[0.040]; Flamboyant[0.040]; miles[0.040]; free[0.039]; agreement[0.039]; said[0.039]; attempted[0.039]; reached[0.039]; guerrilla[0.039]; \t\n\r [===============================>..............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 6s750ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3386/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 981testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.081:0.081[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChicago[0.047]; Chicago[0.047]; Chicago[0.047]; said[0.043]; stronger[0.042]; December[0.042]; expected[0.042]; February[0.042]; August[0.042]; August[0.042]; area[0.041]; market[0.041]; National[0.041]; markets[0.041]; shortened[0.041]; monthly[0.041]; sharply[0.040]; pre[0.040]; analysts[0.040]; sold[0.040]; session[0.040]; futures[0.040]; futures[0.040]; futures[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.066:0.066[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmajor[0.045]; calendar[0.043]; television[0.042]; Federal[0.042]; Federal[0.042]; Federal[0.042]; news[0.042]; decision[0.042]; day[0.042]; speaking[0.042]; reach[0.042]; reflecting[0.042]; Lawrence[0.042]; appears[0.041]; Alan[0.041]; September[0.041]; price[0.041]; price[0.041]; piece[0.041]; July[0.041]; July[0.041]; slightly[0.041]; little[0.041]; Jackson[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Federal Reserve\u001b[39m ==> ENTITY: \u001b[32mFederal Reserve System\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfunds[0.043]; September[0.043]; September[0.043]; notes[0.043]; notes[0.043]; notes[0.043]; fund[0.043]; helped[0.042]; December[0.042]; December[0.042]; December[0.042]; swaps[0.041]; Eurodollar[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; firm[0.041]; firm[0.041]; July[0.041]; monthly[0.041]; data[0.041]; futures[0.041]; April[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japanese ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.073:0.073[0]; log p(e|m)= -0.992:-0.992[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsaid[0.044]; said[0.044]; said[0.044]; levels[0.044]; short[0.042]; area[0.042]; shortened[0.042]; mid[0.042]; level[0.042]; sharply[0.041]; September[0.041]; September[0.041]; helped[0.041]; highest[0.041]; expected[0.041]; traders[0.041]; lower[0.041]; lower[0.041]; index[0.041]; index[0.041]; rose[0.041]; August[0.041]; August[0.041]; average[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Eurodollars\u001b[39m ==> ENTITY: \u001b[32mEurodollar\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurodollars[0.049]; trade[0.045]; notes[0.045]; price[0.044]; example[0.043]; bonds[0.043]; data[0.043]; Traders[0.042]; economic[0.042]; bps[0.041]; bps[0.041]; bps[0.041]; stance[0.041]; bills[0.041]; reach[0.041]; major[0.040]; stability[0.040]; really[0.040]; said[0.040]; said[0.040]; Fed[0.040]; Fed[0.040]; adopt[0.040]; attracted[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChicago[0.046]; Chicago[0.046]; Chicago[0.046]; area[0.045]; said[0.042]; said[0.042]; September[0.042]; stronger[0.042]; December[0.042]; largest[0.041]; February[0.041]; August[0.041]; August[0.041]; helped[0.041]; market[0.041]; July[0.041]; National[0.040]; close[0.040]; firm[0.040]; markets[0.040]; shortened[0.040]; monthly[0.040]; expected[0.040]; sharply[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lawrence Lindsey\u001b[39m ==> ENTITY: \u001b[32mLawrence B. Lindsey\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCommittee[0.044]; Greenspan[0.043]; September[0.042]; economic[0.042]; Federal[0.042]; Federal[0.042]; Federal[0.042]; economy[0.042]; chairman[0.042]; Alan[0.041]; growth[0.041]; July[0.041]; July[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; December[0.041]; December[0.041]; April[0.041]; symposium[0.041]; rates[0.041]; rates[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = FOMC\u001b[39m ==> ENTITY: \u001b[32mFederal Open Market Committee\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmeeting[0.049]; economic[0.048]; Federal[0.047]; adopt[0.047]; provided[0.047]; decision[0.046]; stability[0.046]; inflation[0.046]; Greenspan[0.045]; chairman[0.045]; Fed[0.045]; Fed[0.045]; Fed[0.045]; growth[0.045]; trade[0.045]; economy[0.045]; revisions[0.045]; tightening[0.044]; central[0.044]; bank[0.044]; data[0.044]; stance[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Federal Reserve\u001b[39m ==> ENTITY: \u001b[32mFederal Reserve System\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbank[0.044]; Federal[0.044]; Federal[0.044]; chairman[0.043]; Committee[0.043]; central[0.043]; September[0.042]; notes[0.042]; Greenspan[0.042]; Reserve[0.042]; provided[0.042]; economy[0.041]; rates[0.041]; rates[0.041]; December[0.041]; December[0.041]; December[0.041]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; curb[0.040]; July[0.040]; monthly[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Eurodollars\u001b[39m ==> ENTITY: \u001b[32mEurodollar\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.227:0.227[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nswaps[0.046]; Eurodollar[0.045]; Eurodollar[0.045]; futures[0.044]; futures[0.044]; market[0.043]; notes[0.042]; notes[0.042]; notes[0.042]; funds[0.042]; cash[0.042]; traders[0.041]; price[0.041]; sold[0.040]; data[0.040]; spreads[0.040]; bought[0.040]; bought[0.040]; fund[0.040]; buying[0.040]; buying[0.040]; buying[0.040]; buying[0.040]; unwinding[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.119:0.119[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChicago[0.046]; Chicago[0.046]; Chicago[0.046]; area[0.045]; said[0.042]; said[0.042]; September[0.042]; stronger[0.042]; December[0.041]; largest[0.041]; February[0.041]; August[0.041]; August[0.041]; helped[0.041]; market[0.041]; July[0.041]; National[0.040]; mid[0.040]; close[0.040]; firm[0.040]; markets[0.040]; shortened[0.040]; monthly[0.040]; expected[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fed\u001b[39m ==> ENTITY: \u001b[32mFederal Reserve System\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCommittee[0.045]; meeting[0.045]; meeting[0.045]; Federal[0.045]; Federal[0.045]; Federal[0.045]; economic[0.044]; chairman[0.044]; rates[0.044]; rates[0.044]; adopt[0.044]; bank[0.043]; central[0.043]; provided[0.043]; decision[0.043]; September[0.043]; stability[0.043]; Greenspan[0.043]; inflation[0.042]; inflation[0.042]; Alan[0.042]; economy[0.042]; December[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.100:0.100[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChicago[0.051]; Chicago[0.051]; Chicago[0.051]; said[0.046]; July[0.046]; stronger[0.046]; December[0.046]; expected[0.046]; February[0.045]; August[0.045]; August[0.045]; area[0.045]; market[0.045]; National[0.045]; markets[0.044]; shortened[0.044]; monthly[0.044]; sharply[0.044]; pre[0.044]; analysts[0.044]; sold[0.043]; session[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Eurodollars\u001b[39m ==> ENTITY: \u001b[32mEurodollar\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurodollars[0.050]; trade[0.046]; notes[0.046]; example[0.044]; bonds[0.043]; data[0.043]; economic[0.042]; bps[0.042]; bps[0.042]; bps[0.042]; stance[0.041]; bills[0.041]; really[0.040]; said[0.040]; said[0.040]; Fed[0.040]; Fed[0.040]; adopt[0.040]; attracted[0.040]; lot[0.039]; March[0.039]; March[0.039]; December[0.039]; December[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.064:0.064[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncalendar[0.043]; television[0.043]; Federal[0.042]; Federal[0.042]; Federal[0.042]; news[0.042]; monthly[0.042]; decision[0.042]; day[0.042]; speaking[0.042]; reach[0.042]; reflecting[0.042]; Lawrence[0.042]; appears[0.042]; major[0.041]; Alan[0.041]; September[0.041]; price[0.041]; price[0.041]; piece[0.041]; July[0.041]; July[0.041]; slightly[0.041]; little[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.085:0.085[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmajor[0.043]; mid[0.043]; September[0.042]; September[0.042]; strike[0.042]; April[0.042]; firm[0.042]; firm[0.042]; February[0.042]; level[0.042]; level[0.042]; December[0.042]; December[0.042]; close[0.042]; close[0.042]; percent[0.041]; good[0.041]; July[0.041]; Japanese[0.041]; levels[0.041]; long[0.041]; short[0.041]; said[0.041]; said[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Eurodollar\u001b[39m ==> ENTITY: \u001b[32mEurodollar\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.224:0.224[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurodollars[0.046]; swaps[0.046]; Eurodollar[0.045]; futures[0.044]; futures[0.044]; futures[0.044]; futures[0.044]; market[0.043]; traders[0.041]; price[0.041]; markets[0.041]; debt[0.041]; debt[0.041]; sold[0.040]; data[0.040]; spreads[0.040]; bought[0.040]; bought[0.040]; buying[0.040]; unwinding[0.040]; short[0.040]; Traders[0.040]; investors[0.039]; rose[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.072:0.072[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntelevision[0.043]; news[0.043]; notes[0.043]; decision[0.042]; day[0.042]; speaking[0.042]; reach[0.042]; Lawrence[0.042]; appears[0.042]; example[0.042]; major[0.042]; Alan[0.042]; price[0.041]; piece[0.041]; July[0.041]; Jackson[0.041]; December[0.041]; December[0.041]; annual[0.041]; bit[0.041]; March[0.041]; March[0.041]; data[0.041]; bank[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Federal Open Market Committee\u001b[39m ==> ENTITY: \u001b[32mFederal Open Market Committee\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmeeting[0.048]; rates[0.047]; rates[0.047]; Federal[0.047]; Federal[0.047]; adopt[0.047]; provided[0.046]; decision[0.046]; stability[0.046]; inflation[0.045]; inflation[0.045]; monthly[0.045]; Greenspan[0.045]; chairman[0.045]; Fed[0.044]; Fed[0.044]; growth[0.044]; economy[0.044]; calendar[0.044]; tightening[0.044]; central[0.044]; bank[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alan Greenspan\u001b[39m ==> ENTITY: \u001b[32mAlan Greenspan\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \neconomic[0.048]; Federal[0.046]; Federal[0.046]; Federal[0.046]; economy[0.044]; Committee[0.044]; growth[0.044]; rates[0.043]; rates[0.043]; focus[0.043]; stability[0.043]; chairman[0.043]; trade[0.043]; inflation[0.043]; inflation[0.043]; Reserve[0.043]; Reserve[0.043]; adopt[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; tightening[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.062:0.062[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmajor[0.044]; cash[0.043]; cash[0.043]; level[0.042]; highest[0.042]; calendar[0.042]; television[0.042]; Federal[0.042]; Federal[0.042]; Federal[0.042]; notes[0.042]; monthly[0.042]; day[0.041]; speaking[0.041]; reflecting[0.041]; Lawrence[0.041]; appears[0.041]; Alan[0.041]; September[0.041]; price[0.041]; July[0.041]; slightly[0.041]; little[0.040]; fund[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Eurodollar\u001b[39m ==> ENTITY: \u001b[32mEurodollar\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.221:0.221[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurodollars[0.046]; swaps[0.046]; Eurodollar[0.045]; futures[0.044]; futures[0.044]; market[0.043]; notes[0.043]; notes[0.043]; funds[0.042]; traders[0.041]; price[0.041]; markets[0.041]; sold[0.040]; data[0.040]; spreads[0.040]; bought[0.040]; bought[0.040]; buying[0.040]; buying[0.040]; buying[0.040]; unwinding[0.040]; short[0.040]; Traders[0.040]; sellers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.062:0.062[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsold[0.044]; National[0.044]; largest[0.043]; highest[0.043]; Association[0.043]; finished[0.043]; monthly[0.043]; expected[0.042]; February[0.042]; August[0.042]; August[0.042]; shortened[0.042]; December[0.041]; points[0.041]; rose[0.041]; sharply[0.041]; session[0.041]; pre[0.041]; level[0.041]; lower[0.041]; lower[0.041]; market[0.040]; debt[0.040]; debt[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.067:0.067[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsold[0.045]; National[0.044]; level[0.044]; highest[0.044]; monthly[0.043]; expected[0.043]; February[0.042]; August[0.042]; August[0.042]; shortened[0.042]; July[0.042]; Association[0.042]; December[0.041]; rose[0.041]; sharply[0.041]; session[0.041]; pre[0.041]; lower[0.041]; lower[0.041]; market[0.041]; debt[0.041]; debt[0.041]; bond[0.040]; area[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = CNBC\u001b[39m ==> ENTITY: \u001b[32mCNBC\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.135:0.135[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntelevision[0.049]; news[0.047]; network[0.045]; growth[0.043]; September[0.043]; day[0.043]; cable[0.041]; focus[0.041]; Earlier[0.041]; December[0.041]; December[0.041]; provided[0.041]; July[0.041]; July[0.041]; Alan[0.041]; analyst[0.041]; reach[0.041]; economy[0.040]; calendar[0.040]; adding[0.040]; higher[0.040]; Market[0.040]; major[0.040]; reflecting[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fed\u001b[39m ==> ENTITY: \u001b[32mFederal Reserve System\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmeeting[0.046]; economic[0.045]; chairman[0.045]; adopt[0.045]; bank[0.044]; central[0.044]; provided[0.044]; decision[0.044]; stability[0.044]; Greenspan[0.044]; inflation[0.043]; Alan[0.043]; economy[0.043]; December[0.043]; December[0.043]; Fed[0.043]; Fed[0.043]; growth[0.042]; trade[0.042]; revisions[0.042]; March[0.042]; March[0.042]; network[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Federal Reserve\u001b[39m ==> ENTITY: \u001b[32mFederal Reserve System\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbank[0.044]; Federal[0.044]; Federal[0.044]; chairman[0.044]; Committee[0.043]; central[0.043]; September[0.042]; Greenspan[0.042]; Reserve[0.042]; provided[0.042]; economy[0.041]; decision[0.041]; rates[0.041]; rates[0.041]; December[0.041]; December[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; curb[0.040]; July[0.040]; July[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.081:0.081[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsold[0.043]; cash[0.043]; monthly[0.042]; highest[0.042]; reportedly[0.042]; Federal[0.042]; February[0.042]; August[0.042]; September[0.042]; September[0.042]; July[0.042]; notes[0.042]; notes[0.042]; notes[0.042]; circulated[0.042]; short[0.041]; December[0.041]; December[0.041]; rose[0.041]; mid[0.041]; price[0.041]; sessions[0.041]; helped[0.041]; fund[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fed\u001b[39m ==> ENTITY: \u001b[32mFederal Reserve System\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCommittee[0.044]; meeting[0.043]; meeting[0.043]; Federal[0.043]; Federal[0.043]; economic[0.043]; chairman[0.043]; rates[0.043]; adopt[0.042]; bank[0.042]; central[0.042]; provided[0.042]; decision[0.042]; September[0.041]; stability[0.041]; Greenspan[0.041]; inflation[0.041]; Alan[0.041]; economy[0.040]; December[0.040]; December[0.040]; Fed[0.040]; Fed[0.040]; Market[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = National Association of Purchasing Management\u001b[39m ==> ENTITY: \u001b[32mInstitute for Supply Management\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmonthly[0.044]; Chicago[0.043]; Chicago[0.043]; Chicago[0.043]; Chicago[0.043]; index[0.043]; index[0.043]; largest[0.043]; market[0.042]; highest[0.042]; stronger[0.042]; firm[0.042]; markets[0.042]; helped[0.042]; analysts[0.042]; December[0.041]; futures[0.041]; futures[0.041]; futures[0.041]; futures[0.041]; February[0.040]; session[0.040]; level[0.040]; expected[0.040]; \t\n\r [===============================>..............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 6s576ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3416/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 994testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgrade ==> ENTITY: \u001b[32mBelgrade\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.052:-0.052[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairport[0.045]; airport[0.045]; airport[0.045]; airport[0.045]; airport[0.045]; Belgrade[0.043]; Belgrade[0.043]; Belgrade[0.043]; Yugoslav[0.043]; flights[0.042]; Batajnica[0.042]; Batajnica[0.042]; Aerodrom[0.040]; passengers[0.040]; traffic[0.040]; Beograd[0.040]; September[0.040]; Friday[0.040]; buses[0.040]; Sept[0.039]; closed[0.039]; secured[0.039]; runway[0.039]; runway[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgrade ==> ENTITY: \u001b[32mBelgrade\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.052:-0.052[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairport[0.044]; airport[0.044]; airport[0.044]; airport[0.044]; airport[0.044]; airport[0.044]; airport[0.044]; Belgrade[0.042]; Belgrade[0.042]; Belgrade[0.042]; Belgrade[0.042]; Belgrade[0.042]; Yugoslav[0.042]; flights[0.041]; Batajnica[0.041]; Batajnica[0.041]; Batajnica[0.041]; Aerodrom[0.039]; passengers[0.039]; traffic[0.039]; traffic[0.039]; Beograd[0.039]; September[0.039]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgrade ==> ENTITY: \u001b[32mBelgrade\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.052:-0.052[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairport[0.046]; airport[0.046]; airport[0.046]; airport[0.046]; Belgrade[0.044]; Belgrade[0.044]; Batajnica[0.043]; Batajnica[0.043]; Batajnica[0.043]; passengers[0.041]; traffic[0.041]; buses[0.041]; Air[0.041]; air[0.040]; secured[0.040]; secured[0.040]; maintenance[0.040]; transport[0.040]; Traffic[0.040]; saying[0.039]; Ljubomir[0.039]; said[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSurcin\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBelgrade Nikola Tesla Airport\u001b[39m <---> \u001b[32mSurčin\u001b[39m\t\nSCORES: global= 0.295:0.278[\u001b[31m0.017\u001b[39m]; local()= 0.233:0.138[\u001b[31m0.095\u001b[39m]; log p(e|m)= -1.575:0.000[\u001b[32m1.575\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairport[0.056]; airport[0.056]; airport[0.056]; airport[0.056]; airport[0.056]; airport[0.056]; airport[0.056]; flights[0.052]; Aerodrom[0.052]; Batajnica[0.051]; Batajnica[0.051]; Batajnica[0.051]; Beograd[0.050]; passengers[0.050]; traffic[0.050]; traffic[0.050]; Tanjug[0.050]; Tanjug[0.050]; Tanjug[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Batajnica\u001b[39m ==> ENTITY: \u001b[32mBatajnica\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBatajnica[0.044]; Batajnica[0.044]; Air[0.044]; Beograd[0.042]; Belgrade[0.042]; Belgrade[0.042]; Belgrade[0.042]; Belgrade[0.042]; Belgrade[0.042]; Belgrade[0.042]; Branko[0.042]; air[0.041]; airport[0.041]; airport[0.041]; airport[0.041]; airport[0.041]; airport[0.041]; airport[0.041]; airport[0.041]; Aerodrom[0.041]; nearby[0.040]; Amra[0.040]; buses[0.040]; Yugoslav[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Batajnica\u001b[39m ==> ENTITY: \u001b[32mBatajnica\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBatajnica[0.044]; Batajnica[0.044]; Air[0.044]; Beograd[0.042]; Belgrade[0.042]; Belgrade[0.042]; Belgrade[0.042]; Belgrade[0.042]; Belgrade[0.042]; Belgrade[0.042]; Branko[0.042]; air[0.041]; airport[0.041]; airport[0.041]; airport[0.041]; airport[0.041]; airport[0.041]; airport[0.041]; airport[0.041]; Aerodrom[0.041]; nearby[0.040]; Amra[0.040]; buses[0.040]; Yugoslav[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgrade ==> ENTITY: \u001b[32mBelgrade\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.052:-0.052[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairport[0.045]; airport[0.045]; airport[0.045]; airport[0.045]; airport[0.045]; Belgrade[0.043]; Belgrade[0.043]; Belgrade[0.043]; Yugoslav[0.043]; flights[0.042]; Batajnica[0.042]; Batajnica[0.042]; Aerodrom[0.040]; passengers[0.040]; traffic[0.040]; Beograd[0.040]; September[0.040]; Friday[0.040]; buses[0.040]; Sept[0.039]; closed[0.039]; secured[0.039]; runway[0.039]; runway[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Yugoslav ==> ENTITY: \u001b[32mYugoslavia\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.126:0.126[0]; log p(e|m)= -0.759:-0.759[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgrade[0.054]; Belgrade[0.054]; Belgrade[0.054]; Belgrade[0.054]; Belgrade[0.054]; Batajnica[0.051]; Batajnica[0.051]; Batajnica[0.051]; Aerodrom[0.050]; transport[0.049]; Ljubomir[0.048]; September[0.048]; air[0.048]; said[0.048]; said[0.048]; Tanjug[0.048]; Tanjug[0.048]; Tanjug[0.048]; modernization[0.048]; saying[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgrade ==> ENTITY: \u001b[32mBelgrade\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.103:0.103[0]; log p(e|m)= -0.052:-0.052[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairport[0.045]; airport[0.045]; airport[0.045]; airport[0.045]; airport[0.045]; Belgrade[0.043]; Belgrade[0.043]; Yugoslav[0.043]; flights[0.042]; Batajnica[0.042]; Batajnica[0.042]; Batajnica[0.042]; passengers[0.040]; traffic[0.040]; traffic[0.040]; September[0.040]; Friday[0.040]; buses[0.040]; Air[0.040]; air[0.039]; closed[0.039]; secured[0.039]; secured[0.039]; runway[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tanjug\u001b[39m ==> ENTITY: \u001b[32mTanjug\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.135:0.135[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnews[0.043]; agency[0.043]; agency[0.043]; Tanjug[0.043]; Tanjug[0.043]; Beograd[0.043]; Belgrade[0.043]; Belgrade[0.043]; Belgrade[0.043]; Belgrade[0.043]; Belgrade[0.043]; Belgrade[0.043]; Branko[0.042]; September[0.041]; Ljubomir[0.040]; Batajnica[0.040]; Batajnica[0.040]; Batajnica[0.040]; schedules[0.040]; Director[0.040]; Director[0.040]; Friday[0.040]; Yugoslav[0.040]; modernization[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgrade ==> ENTITY: \u001b[32mBelgrade\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.052:-0.052[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairport[0.045]; airport[0.045]; airport[0.045]; airport[0.045]; airport[0.045]; Belgrade[0.043]; Belgrade[0.043]; Belgrade[0.043]; Yugoslav[0.043]; flights[0.042]; Batajnica[0.042]; Batajnica[0.042]; Aerodrom[0.040]; passengers[0.040]; traffic[0.040]; Beograd[0.040]; September[0.040]; Friday[0.040]; buses[0.040]; Sept[0.039]; closed[0.039]; secured[0.039]; runway[0.039]; runway[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tanjug\u001b[39m ==> ENTITY: \u001b[32mTanjug\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.119:0.119[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnews[0.044]; agency[0.044]; Tanjug[0.044]; Tanjug[0.044]; Belgrade[0.044]; Belgrade[0.044]; Belgrade[0.044]; Branko[0.043]; September[0.042]; Ljubomir[0.041]; Batajnica[0.041]; Batajnica[0.041]; Batajnica[0.041]; schedules[0.041]; Director[0.041]; Director[0.041]; Friday[0.041]; Yugoslav[0.040]; modernization[0.040]; saying[0.040]; air[0.040]; million[0.040]; General[0.039]; Administration[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tanjug\u001b[39m ==> ENTITY: \u001b[32mTanjug\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnews[0.044]; agency[0.044]; agency[0.044]; Tanjug[0.044]; Tanjug[0.044]; Beograd[0.043]; Belgrade[0.043]; Belgrade[0.043]; Belgrade[0.043]; Belgrade[0.043]; Belgrade[0.043]; September[0.041]; Ljubomir[0.041]; Batajnica[0.041]; Batajnica[0.041]; Batajnica[0.041]; schedules[0.041]; Director[0.040]; Friday[0.040]; Yugoslav[0.040]; modernization[0.040]; saying[0.039]; air[0.039]; million[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Batajnica\u001b[39m ==> ENTITY: \u001b[32mBatajnica\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBatajnica[0.045]; Batajnica[0.045]; Air[0.044]; Belgrade[0.043]; Belgrade[0.043]; Belgrade[0.043]; Branko[0.042]; air[0.042]; airport[0.042]; airport[0.042]; airport[0.042]; airport[0.042]; airport[0.042]; nearby[0.041]; Amra[0.041]; buses[0.041]; Yugoslav[0.041]; Ljubomir[0.040]; schedules[0.040]; traffic[0.040]; modernization[0.040]; Tanjug[0.040]; Tanjug[0.040]; Tanjug[0.040]; \t\n\r [===============================>..............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 6s486ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3430/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1116testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.085:0.085[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.047]; Conservative[0.046]; office[0.044]; saying[0.043]; said[0.043]; said[0.043]; bombing[0.042]; Bernard[0.042]; chart[0.042]; Sunday[0.042]; prime[0.042]; Telegraph[0.041]; Irish[0.041]; narrowly[0.041]; conference[0.041]; secretary[0.041]; Party[0.041]; says[0.040]; consulted[0.040]; ago[0.040]; threats[0.040]; Paper[0.040]; Margaret[0.040]; quoted[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = My Turn ==> ENTITY: \u001b[32mMy Turn (memoir)\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.106:0.106[0]; log p(e|m)= -1.339:-1.339[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nuniversity[0.043]; Nancy[0.043]; consulted[0.043]; consulted[0.043]; Ronald[0.043]; wife[0.042]; husband[0.042]; Reagan[0.042]; biography[0.042]; stars[0.042]; watch[0.042]; said[0.042]; said[0.042]; telling[0.042]; thought[0.042]; president[0.041]; quoted[0.041]; personality[0.041]; forceful[0.041]; astrologer[0.041]; admitted[0.041]; plan[0.040]; dubbed[0.040]; schedule[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.035:0.035[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nregularly[0.045]; asking[0.044]; newspaper[0.043]; schedule[0.043]; heard[0.042]; recall[0.042]; indicated[0.042]; degree[0.042]; Turn[0.042]; watch[0.042]; biography[0.041]; Orr[0.041]; Orr[0.041]; Orr[0.041]; added[0.041]; telephone[0.041]; said[0.041]; said[0.041]; admitted[0.041]; Ronald[0.041]; thought[0.041]; Thatcher[0.041]; Thatcher[0.041]; Reagan[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Conservative Party ==> ENTITY: \u001b[32mConservative Party (UK)\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.403:-0.403[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.045]; Bernard[0.045]; Bernard[0.045]; secretary[0.044]; minister[0.043]; narrowly[0.043]; Orr[0.042]; Orr[0.042]; Orr[0.042]; Orr[0.042]; Orr[0.042]; London[0.041]; Irish[0.041]; office[0.041]; Thatcher[0.041]; Thatcher[0.041]; Thatcher[0.041]; Thatcher[0.041]; conference[0.040]; political[0.040]; says[0.040]; Telegraph[0.040]; Ingram[0.039]; Ingram[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = The Sunday Telegraph\u001b[39m ==> ENTITY: \u001b[32mThe Sunday Telegraph\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.049]; newspaper[0.048]; London[0.045]; quoted[0.044]; told[0.043]; saying[0.042]; Conservative[0.041]; Thatcher[0.041]; Thatcher[0.041]; Thatcher[0.041]; Thatcher[0.041]; says[0.041]; Ingram[0.041]; Irish[0.041]; press[0.041]; added[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; Paper[0.040]; decade[0.039]; asked[0.039]; asked[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ronald Reagan\u001b[39m ==> ENTITY: \u001b[32mRonald Reagan\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.081:0.081[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntelling[0.046]; president[0.044]; security[0.044]; let[0.043]; help[0.043]; asking[0.043]; dubbed[0.042]; said[0.042]; said[0.042]; political[0.042]; plan[0.042]; know[0.041]; personality[0.041]; biography[0.041]; recall[0.041]; stars[0.041]; load[0.041]; watch[0.041]; thought[0.041]; Nancy[0.040]; Thatcher[0.040]; Thatcher[0.040]; telephone[0.040]; admitted[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Thatcher\u001b[39m ==> ENTITY: \u001b[32mMargaret Thatcher\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.046]; Conservative[0.044]; Thatcher[0.043]; Thatcher[0.043]; Thatcher[0.043]; Thatcher[0.043]; Thatcher[0.043]; narrowly[0.042]; life[0.042]; London[0.041]; office[0.041]; minister[0.041]; conference[0.041]; future[0.041]; future[0.041]; asked[0.041]; asked[0.041]; political[0.041]; secretary[0.040]; Irish[0.040]; security[0.040]; prime[0.040]; Party[0.040]; Iron[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Thatcher\u001b[39m ==> ENTITY: \u001b[32mMargaret Thatcher\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.046]; Conservative[0.044]; Thatcher[0.043]; Thatcher[0.043]; Thatcher[0.043]; narrowly[0.043]; London[0.041]; prime[0.041]; office[0.041]; minister[0.041]; decade[0.041]; conference[0.041]; asked[0.041]; asked[0.041]; said[0.041]; said[0.041]; life[0.041]; Party[0.041]; saying[0.041]; secretary[0.041]; says[0.041]; Irish[0.041]; future[0.040]; attempt[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Margaret Thatcher\u001b[39m ==> ENTITY: \u001b[32mMargaret Thatcher\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.159:0.159[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.046]; Conservative[0.044]; Thatcher[0.043]; Thatcher[0.043]; Thatcher[0.043]; narrowly[0.043]; London[0.041]; prime[0.041]; office[0.041]; minister[0.041]; decade[0.041]; conference[0.041]; asked[0.041]; asked[0.041]; said[0.041]; said[0.041]; life[0.041]; saying[0.041]; secretary[0.041]; says[0.041]; Irish[0.041]; Party[0.040]; future[0.040]; future[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nancy ==> ENTITY: \u001b[32mNancy Reagan\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.121:0.121[0]; log p(e|m)= -2.718:-2.718[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbiography[0.046]; Reagan[0.045]; husband[0.043]; wife[0.043]; Thatcher[0.043]; Thatcher[0.043]; Ronald[0.042]; admitted[0.042]; president[0.042]; personality[0.042]; Lady[0.042]; dubbed[0.041]; stars[0.041]; know[0.041]; added[0.041]; telling[0.041]; thought[0.041]; help[0.041]; let[0.041]; moves[0.040]; said[0.040]; said[0.040]; watch[0.040]; recall[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bernard Ingham\u001b[39m ==> ENTITY: \u001b[32mBernard Ingham\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.216:0.216[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecretary[0.048]; Telegraph[0.048]; press[0.048]; Conservative[0.048]; newspaper[0.047]; newspaper[0.047]; saying[0.047]; political[0.046]; minister[0.045]; said[0.045]; said[0.045]; said[0.045]; said[0.045]; told[0.045]; conference[0.044]; quoted[0.044]; quoted[0.044]; forceful[0.044]; asking[0.044]; asked[0.044]; asked[0.044]; prime[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Thatcher\u001b[39m ==> ENTITY: \u001b[32mMargaret Thatcher\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.050]; Conservative[0.048]; Thatcher[0.047]; Thatcher[0.047]; Thatcher[0.047]; narrowly[0.046]; life[0.046]; London[0.045]; office[0.045]; minister[0.045]; conference[0.045]; future[0.045]; future[0.045]; asked[0.045]; asked[0.045]; political[0.044]; secretary[0.044]; Irish[0.044]; security[0.044]; prime[0.044]; Party[0.044]; attempt[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = British ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.082:0.082[0]; log p(e|m)= -0.489:-0.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIrish[0.047]; London[0.045]; Thatcher[0.044]; Thatcher[0.044]; Thatcher[0.044]; Thatcher[0.044]; Margaret[0.043]; Conservative[0.043]; Bernard[0.043]; decade[0.042]; says[0.041]; Ingram[0.041]; Ira[0.041]; prime[0.041]; astrologer[0.040]; astrologer[0.040]; life[0.040]; Telegraph[0.040]; chart[0.040]; said[0.039]; said[0.039]; future[0.039]; future[0.039]; office[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Thatcher\u001b[39m ==> ENTITY: \u001b[32mMargaret Thatcher\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.128:0.128[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThatcher[0.044]; Thatcher[0.044]; watch[0.043]; telling[0.042]; asked[0.042]; Reagan[0.042]; stars[0.042]; wife[0.042]; admitted[0.042]; secretary[0.041]; consulted[0.041]; consulted[0.041]; regularly[0.041]; security[0.041]; Iron[0.041]; told[0.041]; said[0.041]; said[0.041]; said[0.041]; biography[0.041]; telephone[0.041]; telephone[0.041]; asking[0.041]; thought[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = IRA ==> ENTITY: \u001b[32mProvisional Irish Republican Army\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.970:-0.970[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIrish[0.050]; British[0.046]; guerrilla[0.046]; bombing[0.044]; London[0.042]; Party[0.042]; office[0.041]; consulted[0.041]; decade[0.041]; attempt[0.041]; indicated[0.041]; conference[0.041]; Sunday[0.041]; threats[0.041]; Conservative[0.041]; ago[0.040]; indications[0.040]; prime[0.040]; saying[0.040]; press[0.040]; asked[0.040]; asked[0.040]; Margaret[0.040]; narrowly[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Thatcher\u001b[39m ==> ENTITY: \u001b[32mMargaret Thatcher\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.128:0.128[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThatcher[0.044]; Thatcher[0.044]; watch[0.043]; telling[0.042]; asked[0.042]; Reagan[0.042]; stars[0.042]; wife[0.042]; admitted[0.042]; secretary[0.041]; consulted[0.041]; consulted[0.041]; regularly[0.041]; security[0.041]; Iron[0.041]; told[0.041]; said[0.041]; said[0.041]; said[0.041]; biography[0.041]; telephone[0.041]; telephone[0.041]; asking[0.041]; thought[0.041]; \t\n\r [================================>.............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 6s362ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3447/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1140testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Beijing ==> ENTITY: \u001b[32mBeijing\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.164:0.164[0]; log p(e|m)= -0.042:-0.042[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChinese[0.049]; China[0.046]; China[0.046]; China[0.046]; reached[0.042]; month[0.042]; Donghai[0.041]; Donghai[0.041]; Wang[0.041]; Wang[0.041]; Wang[0.041]; Wang[0.041]; Wang[0.041]; Wang[0.041]; held[0.041]; euphemism[0.041]; Democracy[0.041]; Friday[0.040]; dissident[0.040]; dissident[0.040]; dissident[0.040]; class[0.039]; meet[0.039]; Police[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York-based ==> ENTITY: \u001b[32mNew York\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.092:0.092[0]; log p(e|m)= -0.017:-0.017[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nreleased[0.044]; group[0.044]; group[0.044]; group[0.044]; based[0.043]; month[0.043]; Human[0.043]; administrative[0.043]; movement[0.042]; Friday[0.042]; education[0.041]; education[0.041]; health[0.041]; Saturday[0.041]; class[0.041]; said[0.041]; said[0.041]; said[0.041]; serve[0.040]; held[0.040]; family[0.040]; family[0.040]; Democracy[0.040]; sent[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Beijing ==> ENTITY: \u001b[32mBeijing\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.042:-0.042[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.048]; Tiananmen[0.044]; reached[0.044]; military[0.043]; Square[0.043]; month[0.043]; Wang[0.042]; Wang[0.042]; Wang[0.042]; serve[0.042]; Democracy[0.042]; street[0.041]; dissident[0.041]; June[0.041]; crushed[0.041]; demonstrations[0.040]; loss[0.040]; movement[0.040]; pro[0.040]; ordered[0.040]; veteran[0.040]; protests[0.040]; immediately[0.040]; Wall[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Democracy Wall\u001b[39m ==> ENTITY: \u001b[32mDemocracy Wall\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBeijing[0.051]; Beijing[0.051]; Tiananmen[0.048]; China[0.048]; China[0.048]; Chinese[0.047]; month[0.046]; Square[0.046]; Donghai[0.045]; Donghai[0.045]; protests[0.044]; June[0.044]; Wang[0.044]; Wang[0.044]; Wang[0.044]; Wang[0.044]; Wang[0.044]; Wang[0.044]; Wang[0.044]; democracy[0.044]; movement[0.044]; sent[0.043]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mTiananmen Square\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mTiananmen Square protests of 1989\u001b[39m <---> \u001b[32mTiananmen Square\u001b[39m\t\nSCORES: global= 0.277:0.271[\u001b[31m0.006\u001b[39m]; local()= 0.151:0.128[\u001b[31m0.023\u001b[39m]; log p(e|m)= 0.000:-0.018[\u001b[31m0.018\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.048]; Beijing[0.047]; protests[0.045]; demonstrations[0.044]; Wall[0.043]; Wang[0.043]; Wang[0.043]; Wang[0.043]; Democracy[0.042]; month[0.042]; movement[0.042]; dissident[0.041]; June[0.041]; students[0.041]; street[0.040]; democracy[0.040]; jailed[0.040]; imposed[0.040]; military[0.040]; administrative[0.040]; immediately[0.039]; pro[0.039]; euphemism[0.039]; class[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chinese ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.172:0.172[0]; log p(e|m)= -1.419:-1.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.049]; China[0.049]; China[0.049]; Beijing[0.045]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Donghai[0.041]; Donghai[0.041]; Democracy[0.041]; family[0.040]; family[0.040]; month[0.040]; group[0.040]; group[0.040]; group[0.040]; said[0.039]; said[0.039]; said[0.039]; class[0.039]; say[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.167:0.167[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.049]; China[0.049]; Chinese[0.047]; Beijing[0.045]; Beijing[0.045]; military[0.044]; Wang[0.041]; Wang[0.041]; Wang[0.041]; Wang[0.041]; Wang[0.041]; Wang[0.041]; Wang[0.041]; month[0.041]; Donghai[0.040]; Donghai[0.040]; June[0.040]; Democracy[0.040]; Tiananmen[0.039]; demonstrations[0.039]; serve[0.039]; family[0.038]; family[0.038]; administrative[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mHuman Rights in China\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHuman rights in China\u001b[39m <---> \u001b[32mHuman Rights in China (organization)\u001b[39m\t\nSCORES: global= 0.295:0.288[\u001b[31m0.007\u001b[39m]; local()= 0.158:0.190[\u001b[32m0.031\u001b[39m]; log p(e|m)= 0.000:-0.068[\u001b[31m0.068\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.051]; China[0.051]; Beijing[0.049]; Chinese[0.049]; York[0.045]; Democracy[0.045]; dissident[0.045]; dissident[0.045]; dissident[0.045]; group[0.044]; group[0.044]; group[0.044]; administrative[0.044]; Wang[0.044]; Wang[0.044]; Wang[0.044]; Wang[0.044]; Wang[0.044]; Wang[0.044]; education[0.044]; education[0.044]; movement[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.184:0.184[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.053]; China[0.053]; Chinese[0.053]; Beijing[0.049]; Wang[0.047]; Wang[0.047]; Wang[0.047]; Wang[0.047]; Wang[0.047]; Wang[0.047]; Donghai[0.046]; Donghai[0.046]; month[0.045]; class[0.042]; dissident[0.042]; dissident[0.042]; sent[0.042]; held[0.041]; said[0.041]; said[0.041]; said[0.041]; Friday[0.041]; \t\n\r [================================>.............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 6s346ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3456/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1054testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pakistan ==> ENTITY: \u001b[32mPakistan national cricket team\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.160:0.160[0]; log p(e|m)= -2.749:-2.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.052]; Pakistan[0.052]; series[0.048]; England[0.048]; England[0.048]; England[0.048]; England[0.048]; Ijaz[0.043]; Cricket[0.043]; Edgbaston[0.042]; Ahmed[0.041]; Hollioake[0.041]; day[0.040]; second[0.040]; Saturday[0.039]; innings[0.039]; international[0.039]; runs[0.038]; beat[0.037]; Birmingham[0.037]; Scores[0.035]; Runs[0.035]; Beat[0.034]; closed[0.034]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEngland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEngland cricket team\u001b[39m <---> \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.257:0.236[\u001b[31m0.021\u001b[39m]; local()= 0.152:0.070[\u001b[31m0.082\u001b[39m]; log p(e|m)= -3.297:-0.190[\u001b[32m3.107\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.049]; England[0.049]; England[0.049]; Birmingham[0.046]; Cricket[0.044]; Edgbaston[0.043]; Hollioake[0.043]; international[0.043]; Pakistan[0.042]; Pakistan[0.042]; Pakistan[0.042]; series[0.042]; day[0.042]; second[0.042]; Saturday[0.040]; Ijaz[0.040]; innings[0.040]; runs[0.040]; beat[0.038]; Scores[0.038]; Ahmed[0.037]; Runs[0.037]; Knight[0.036]; Beat[0.035]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland cricket team\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.152:0.152[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.049]; England[0.049]; England[0.049]; Birmingham[0.046]; Cricket[0.044]; Edgbaston[0.043]; Hollioake[0.043]; international[0.043]; Pakistan[0.042]; Pakistan[0.042]; Pakistan[0.042]; series[0.042]; day[0.042]; second[0.042]; Saturday[0.040]; Ijaz[0.040]; innings[0.040]; runs[0.040]; beat[0.038]; Scores[0.038]; Ahmed[0.037]; Runs[0.037]; Knight[0.036]; Beat[0.035]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Edgbaston ==> ENTITY: \u001b[32mEdgbaston Cricket Ground\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.451:-0.451[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.046]; England[0.046]; England[0.046]; England[0.046]; second[0.044]; innings[0.044]; series[0.044]; Cricket[0.043]; Hollioake[0.043]; Birmingham[0.043]; day[0.042]; runs[0.042]; beat[0.041]; Scores[0.041]; Pakistan[0.041]; Pakistan[0.041]; Pakistan[0.041]; Saturday[0.040]; international[0.039]; Runs[0.038]; Ijaz[0.038]; Beat[0.038]; Second[0.037]; Ahmed[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pakistan ==> ENTITY: \u001b[32mPakistan national cricket team\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.160:0.160[0]; log p(e|m)= -2.749:-2.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.052]; Pakistan[0.052]; series[0.048]; England[0.048]; England[0.048]; England[0.048]; England[0.048]; Ijaz[0.043]; Cricket[0.043]; Edgbaston[0.042]; Ahmed[0.041]; Hollioake[0.041]; day[0.040]; second[0.040]; Saturday[0.039]; innings[0.039]; international[0.039]; runs[0.038]; beat[0.037]; Birmingham[0.037]; Scores[0.035]; Runs[0.035]; Beat[0.034]; closed[0.034]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Birmingham ==> ENTITY: \u001b[32mBirmingham\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.083:0.083[0]; log p(e|m)= -0.402:-0.402[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; England[0.047]; England[0.047]; England[0.047]; Edgbaston[0.045]; second[0.045]; innings[0.044]; Cricket[0.044]; Hollioake[0.044]; runs[0.042]; day[0.041]; series[0.041]; Saturday[0.040]; Pakistan[0.040]; Pakistan[0.040]; Pakistan[0.040]; beat[0.040]; international[0.039]; closed[0.039]; Ijaz[0.038]; Ahmed[0.038]; Runs[0.037]; Second[0.037]; Scores[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pakistan ==> ENTITY: \u001b[32mPakistan national cricket team\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.160:0.160[0]; log p(e|m)= -2.749:-2.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.052]; Pakistan[0.052]; series[0.048]; England[0.048]; England[0.048]; England[0.048]; England[0.048]; Ijaz[0.043]; Cricket[0.043]; Edgbaston[0.042]; Ahmed[0.041]; Hollioake[0.041]; day[0.040]; second[0.040]; Saturday[0.039]; innings[0.039]; international[0.039]; runs[0.038]; beat[0.037]; Birmingham[0.037]; Scores[0.035]; Runs[0.035]; Beat[0.034]; closed[0.034]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland cricket team\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.152:0.152[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.049]; England[0.049]; England[0.049]; Birmingham[0.046]; Cricket[0.044]; Edgbaston[0.043]; Hollioake[0.043]; international[0.043]; Pakistan[0.042]; Pakistan[0.042]; Pakistan[0.042]; series[0.042]; day[0.042]; second[0.042]; Saturday[0.040]; Ijaz[0.040]; innings[0.040]; runs[0.040]; beat[0.038]; Scores[0.038]; Ahmed[0.037]; Runs[0.037]; Knight[0.036]; Beat[0.035]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ijaz Ahmed\u001b[39m ==> ENTITY: \u001b[32mIjaz Ahmed (cricketer)\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.049]; Pakistan[0.049]; Pakistan[0.049]; Edgbaston[0.046]; England[0.043]; England[0.043]; England[0.043]; England[0.043]; series[0.043]; Cricket[0.043]; day[0.043]; Hollioake[0.042]; innings[0.040]; international[0.040]; runs[0.040]; second[0.040]; Saturday[0.040]; Birmingham[0.039]; Runs[0.038]; Scores[0.038]; beat[0.038]; Dayer[0.037]; Beat[0.036]; closed[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland cricket team\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.152:0.152[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.049]; England[0.049]; England[0.049]; Birmingham[0.046]; Cricket[0.044]; Edgbaston[0.043]; Hollioake[0.043]; international[0.043]; Pakistan[0.042]; Pakistan[0.042]; Pakistan[0.042]; series[0.042]; day[0.042]; second[0.042]; Saturday[0.040]; Ijaz[0.040]; innings[0.040]; runs[0.040]; beat[0.038]; Scores[0.038]; Ahmed[0.037]; Runs[0.037]; Knight[0.036]; Beat[0.035]; \t\n\r [================================>.............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 6s425ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3466/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1055testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Netherlands ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.189:0.189[0]; log p(e|m)= -0.250:-0.250[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.048]; Netherlands[0.048]; Netherlands[0.048]; Netherlands[0.048]; Denmark[0.045]; Germany[0.044]; Spain[0.044]; Italy[0.042]; Italy[0.042]; Roermond[0.042]; Landgraaf[0.042]; Landgraaf[0.042]; final[0.041]; Standings[0.040]; Cycling[0.040]; Tour[0.040]; Tour[0.040]; Final[0.039]; sixth[0.039]; time[0.038]; time[0.038]; Jesper[0.038]; Saturday[0.038]; Ludwig[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rabobank\u001b[39m ==> ENTITY: \u001b[32mBelkin Pro Cycling Team\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBreukink[0.052]; Breukink[0.052]; Polti[0.050]; Rabobank[0.050]; Rabobank[0.050]; Rabobank[0.050]; stages[0.048]; overall[0.048]; Armstrong[0.048]; Spain[0.048]; Belgium[0.047]; Marco[0.047]; Ekimov[0.046]; Ekimov[0.046]; Lance[0.046]; Netherlands[0.045]; Netherlands[0.045]; Netherlands[0.045]; Netherlands[0.045]; Denmark[0.045]; Denmark[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rabobank\u001b[39m ==> ENTITY: \u001b[32mBelkin Pro Cycling Team\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBreukink[0.045]; Breukink[0.045]; Cycling[0.045]; Polti[0.043]; Rabobank[0.043]; Rabobank[0.043]; Rabobank[0.043]; Tour[0.043]; Tour[0.043]; Landgraaf[0.041]; Landgraaf[0.041]; stages[0.041]; overall[0.041]; Armstrong[0.041]; Spain[0.041]; Belgium[0.041]; sixth[0.040]; Marco[0.040]; Ekimov[0.040]; Ekimov[0.040]; Roermond[0.040]; stage[0.040]; Lance[0.040]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lance Armstrong\u001b[39m ==> ENTITY: \u001b[32mLance Armstrong\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEkimov[0.044]; Ekimov[0.044]; Michael[0.044]; Breukink[0.043]; Breukink[0.043]; Polti[0.043]; Skibby[0.042]; Skibby[0.042]; Telekom[0.041]; Telekom[0.041]; stages[0.041]; Rolf[0.041]; seconds[0.041]; seconds[0.041]; seconds[0.041]; seconds[0.041]; Johan[0.041]; Marco[0.041]; time[0.041]; time[0.041]; time[0.041]; time[0.041]; Erik[0.041]; Erik[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Erik Breukink\u001b[39m ==> ENTITY: \u001b[32mErik Breukink\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBreukink[0.053]; Polti[0.051]; Cycling[0.050]; Landgraaf[0.050]; Landgraaf[0.050]; stage[0.047]; Marco[0.047]; Tour[0.047]; Tour[0.047]; Roermond[0.047]; overall[0.047]; Skibby[0.047]; Ekimov[0.047]; Ekimov[0.047]; Telekom[0.046]; Telekom[0.046]; Rolf[0.046]; Johan[0.046]; stages[0.046]; Spain[0.046]; Belgium[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = TVM ==> ENTITY: \u001b[32mTVM (cycling team)\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.088:0.088[0]; log p(e|m)= -0.844:-0.844[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTelekom[0.043]; Rabobank[0.043]; Rabobank[0.043]; Rabobank[0.043]; Rolf[0.043]; Johan[0.043]; Erik[0.042]; Breukink[0.042]; Belgium[0.042]; time[0.042]; time[0.042]; time[0.042]; Italy[0.041]; Italy[0.041]; Mx[0.041]; seconds[0.041]; Pavanello[0.041]; Netherlands[0.041]; Netherlands[0.041]; Spain[0.041]; Maarten[0.041]; Anguita[0.041]; Technogym[0.040]; stages[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.205:0.205[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.053]; Italy[0.053]; Spain[0.044]; Germany[0.042]; final[0.042]; overall[0.042]; sixth[0.042]; Denmark[0.042]; Denmark[0.042]; Marco[0.041]; Giovanni[0.041]; Belgium[0.041]; Sweden[0.040]; time[0.040]; time[0.040]; time[0.040]; Tour[0.040]; Tour[0.040]; Netherlands[0.040]; Netherlands[0.040]; Netherlands[0.040]; Netherlands[0.040]; Netherlands[0.040]; Luca[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Collstrop\u001b[39m ==> ENTITY: \u001b[32mCycle Collstrop\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.113:0.113[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPolti[0.046]; Landgraaf[0.045]; Tour[0.043]; Armstrong[0.043]; Breukink[0.043]; Breukink[0.043]; Telekom[0.042]; Telekom[0.042]; Rolf[0.042]; Johan[0.042]; Roermond[0.042]; Erik[0.041]; Erik[0.041]; Rabobank[0.041]; Rabobank[0.041]; Rabobank[0.041]; Rabobank[0.041]; Olaf[0.041]; Skibby[0.040]; Skibby[0.040]; Italy[0.040]; Italy[0.040]; Italy[0.040]; mins[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Olaf Ludwig\u001b[39m ==> ENTITY: \u001b[32mOlaf Ludwig\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCycling[0.045]; Landgraaf[0.044]; Landgraaf[0.044]; Breukink[0.044]; Polti[0.044]; Johan[0.043]; Tour[0.043]; Tour[0.043]; Ekimov[0.043]; Telekom[0.041]; Telekom[0.041]; Rabobank[0.041]; Rabobank[0.041]; Germany[0.041]; Rolf[0.041]; Roermond[0.041]; Erik[0.041]; Belgium[0.041]; overall[0.040]; mins[0.039]; Jesper[0.039]; Pavanello[0.039]; Skibby[0.039]; Anguita[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = TVM ==> ENTITY: \u001b[32mTVM (cycling team)\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.844:-0.844[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLandgraaf[0.051]; Landgraaf[0.051]; Cycling[0.049]; Polti[0.048]; hours[0.048]; Telekom[0.048]; Telekom[0.048]; Rabobank[0.048]; Rabobank[0.048]; Rabobank[0.048]; Rabobank[0.048]; Rolf[0.047]; Johan[0.047]; Roermond[0.047]; Erik[0.047]; Erik[0.047]; Breukink[0.047]; Breukink[0.047]; Saturday[0.046]; Olaf[0.046]; Belgium[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Andersson ==> ENTITY: \u001b[32mMichael Andersson (cyclist)\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.152:0.152[0]; log p(e|m)= -0.926:-0.926[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAndersson[0.050]; Sweden[0.049]; Michael[0.047]; time[0.047]; time[0.047]; time[0.047]; Denmark[0.046]; Denmark[0.046]; Erik[0.046]; Johan[0.046]; Telekom[0.045]; Rolf[0.045]; Italy[0.045]; Italy[0.045]; Breukink[0.044]; Spain[0.044]; Sorensen[0.044]; Pavanello[0.044]; Mx[0.044]; Belgium[0.043]; Ekimov[0.043]; Ekimov[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Netherlands ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.171:0.171[0]; log p(e|m)= -0.250:-0.250[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.048]; Netherlands[0.048]; Netherlands[0.048]; Belgium[0.046]; Denmark[0.046]; Denmark[0.046]; Spain[0.045]; Sweden[0.044]; Italy[0.043]; Italy[0.043]; Final[0.040]; Bakker[0.040]; Johan[0.039]; Erik[0.039]; Erik[0.039]; Sorensen[0.039]; overall[0.038]; time[0.038]; time[0.038]; time[0.038]; time[0.038]; Maarten[0.038]; Jesper[0.038]; Ludwig[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rabobank\u001b[39m ==> ENTITY: \u001b[32mBelkin Pro Cycling Team\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBreukink[0.046]; Breukink[0.046]; Rabobank[0.044]; Rabobank[0.044]; Rabobank[0.044]; stages[0.042]; overall[0.042]; Armstrong[0.042]; Spain[0.042]; Belgium[0.042]; Marco[0.041]; Ekimov[0.041]; Ekimov[0.041]; Lance[0.041]; Netherlands[0.040]; Netherlands[0.040]; Netherlands[0.040]; Denmark[0.040]; Denmark[0.040]; seconds[0.040]; seconds[0.040]; Skibby[0.040]; Skibby[0.040]; Telekom[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.174:0.174[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.048]; Denmark[0.048]; Germany[0.043]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Spain[0.042]; Netherlands[0.042]; Netherlands[0.042]; Netherlands[0.042]; Netherlands[0.042]; Netherlands[0.042]; overall[0.042]; Belgium[0.041]; final[0.041]; Andersson[0.041]; Andersson[0.041]; sixth[0.040]; Jesper[0.040]; Erik[0.039]; Erik[0.039]; Russia[0.039]; Tour[0.039]; Landgraaf[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rabobank\u001b[39m ==> ENTITY: \u001b[32mBelkin Pro Cycling Team\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBreukink[0.045]; Breukink[0.045]; Cycling[0.045]; Polti[0.043]; Rabobank[0.043]; Rabobank[0.043]; Rabobank[0.043]; Tour[0.043]; Tour[0.043]; Landgraaf[0.041]; Landgraaf[0.041]; stages[0.041]; overall[0.041]; Armstrong[0.041]; Spain[0.041]; Belgium[0.041]; sixth[0.040]; Marco[0.040]; Ekimov[0.040]; Ekimov[0.040]; Roermond[0.040]; stage[0.040]; Lance[0.040]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tristan Hoffman\u001b[39m ==> ENTITY: \u001b[32mTristan Hoffman\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRabobank[0.045]; Rabobank[0.045]; Rabobank[0.045]; Cycling[0.044]; Johan[0.044]; Landgraaf[0.043]; Landgraaf[0.043]; Polti[0.043]; Armstrong[0.043]; Breukink[0.042]; Olaf[0.041]; Belgium[0.041]; Dekker[0.041]; Jesper[0.041]; Telekom[0.040]; Telekom[0.040]; Standings[0.040]; Rolf[0.040]; Roermond[0.040]; Tour[0.040]; Tour[0.040]; Ludwig[0.040]; Erik[0.040]; Erik[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.064:0.064[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.045]; Denmark[0.044]; Denmark[0.044]; Sweden[0.043]; overall[0.043]; Russia[0.043]; Italy[0.042]; Italy[0.042]; Italy[0.042]; Final[0.042]; time[0.042]; time[0.042]; time[0.042]; time[0.042]; Netherlands[0.041]; Netherlands[0.041]; Netherlands[0.041]; Netherlands[0.041]; Lombardi[0.041]; Michael[0.040]; Belgium[0.040]; Andersson[0.039]; Andersson[0.039]; Sorensen[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.046]; Denmark[0.046]; Spain[0.044]; Italy[0.044]; Italy[0.044]; Belgium[0.044]; Netherlands[0.043]; Netherlands[0.043]; Netherlands[0.043]; Netherlands[0.043]; Netherlands[0.043]; Sweden[0.042]; final[0.041]; Andersson[0.040]; Erik[0.040]; overall[0.039]; time[0.039]; time[0.039]; time[0.039]; Russia[0.039]; Landgraaf[0.039]; Landgraaf[0.039]; Tour[0.039]; Tour[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.206:0.206[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.069]; Italy[0.069]; Spain[0.057]; Standings[0.055]; Germany[0.055]; final[0.055]; overall[0.055]; sixth[0.054]; Denmark[0.054]; Denmark[0.054]; Belgium[0.053]; Sweden[0.053]; Giovanni[0.053]; time[0.053]; time[0.053]; time[0.053]; time[0.053]; Marco[0.052]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Motorola ==> ENTITY: \u001b[32m7-Eleven (cycling team)\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.132:0.132[0]; log p(e|m)= -1.766:-1.766[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArmstrong[0.045]; Lance[0.045]; Breukink[0.045]; Breukink[0.045]; Ekimov[0.045]; Ekimov[0.045]; Erik[0.045]; Erik[0.045]; Michael[0.044]; Telekom[0.044]; overall[0.044]; Rolf[0.044]; Marco[0.044]; Johan[0.043]; Skibby[0.043]; Skibby[0.043]; Polti[0.042]; Giovanni[0.042]; Spain[0.042]; mins[0.042]; Denmark[0.042]; Denmark[0.042]; Pavanello[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tour of the Netherlands\u001b[39m ==> ENTITY: \u001b[32mRonde van Nederland\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLandgraaf[0.046]; Landgraaf[0.046]; Tour[0.045]; Polti[0.044]; Belgium[0.044]; Roermond[0.043]; Cycling[0.043]; Breukink[0.043]; Netherlands[0.042]; Netherlands[0.042]; Netherlands[0.042]; Netherlands[0.042]; Ekimov[0.041]; Denmark[0.040]; Rabobank[0.040]; Rabobank[0.040]; Standings[0.040]; Telekom[0.040]; Telekom[0.040]; stage[0.040]; Johan[0.039]; final[0.039]; Erik[0.039]; Germany[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Skibby\u001b[39m ==> ENTITY: \u001b[32mJesper Skibby\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.045]; Netherlands[0.044]; Netherlands[0.044]; Belgium[0.043]; Breukink[0.043]; Ekimov[0.042]; Ekimov[0.042]; Italy[0.042]; Italy[0.042]; Rabobank[0.042]; Rabobank[0.042]; Rabobank[0.042]; Armstrong[0.042]; Telekom[0.041]; stages[0.041]; Sweden[0.041]; Rolf[0.041]; Johan[0.041]; Erik[0.041]; Lance[0.040]; Spain[0.040]; Andersson[0.039]; Andersson[0.039]; Technogym[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rolf Sorensen\u001b[39m ==> ENTITY: \u001b[32mRolf Sørensen\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPolti[0.052]; Breukink[0.049]; Breukink[0.049]; Armstrong[0.049]; Belgium[0.049]; Italy[0.049]; Italy[0.049]; Italy[0.049]; Ekimov[0.048]; Ekimov[0.048]; Lance[0.048]; Skibby[0.047]; Skibby[0.047]; Denmark[0.047]; Denmark[0.047]; Telekom[0.046]; Telekom[0.046]; stages[0.046]; Johan[0.046]; Erik[0.045]; Erik[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Vyacheslav Ekimov\u001b[39m ==> ENTITY: \u001b[32mViatcheslav Ekimov\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEkimov[0.045]; Tour[0.044]; Tour[0.044]; Landgraaf[0.044]; Landgraaf[0.044]; Cycling[0.043]; sixth[0.043]; Armstrong[0.042]; Breukink[0.042]; Breukink[0.042]; Telekom[0.041]; Telekom[0.041]; Rolf[0.041]; Johan[0.041]; Roermond[0.041]; Lance[0.041]; Erik[0.041]; Erik[0.041]; Polti[0.040]; Olaf[0.040]; seconds[0.040]; seconds[0.040]; seconds[0.040]; seconds[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = TVM ==> ENTITY: \u001b[32mTVM (cycling team)\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.129:0.129[0]; log p(e|m)= -0.844:-0.844[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLandgraaf[0.045]; Landgraaf[0.045]; Cycling[0.043]; Polti[0.042]; hours[0.042]; Telekom[0.042]; Telekom[0.042]; Rabobank[0.042]; Rabobank[0.042]; Rabobank[0.042]; Rolf[0.042]; Johan[0.041]; Roermond[0.041]; Erik[0.041]; Erik[0.041]; Breukink[0.041]; Saturday[0.041]; Olaf[0.041]; Belgium[0.041]; time[0.041]; time[0.041]; time[0.041]; Tour[0.041]; Tour[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ludwig\u001b[39m ==> ENTITY: \u001b[32mOlaf Ludwig\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBreukink[0.045]; Breukink[0.045]; Johan[0.044]; Ekimov[0.044]; Ekimov[0.044]; Armstrong[0.042]; Telekom[0.042]; Rabobank[0.042]; Rabobank[0.042]; Rabobank[0.042]; Rabobank[0.042]; Rolf[0.042]; Erik[0.041]; Erik[0.041]; Belgium[0.041]; overall[0.041]; Marco[0.041]; Jesper[0.040]; Pavanello[0.040]; Skibby[0.040]; Skibby[0.040]; Maarten[0.040]; Anguita[0.040]; Mx[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Giovanni Lombardi\u001b[39m ==> ENTITY: \u001b[32mGiovanni Lombardi\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBreukink[0.044]; Polti[0.044]; Landgraaf[0.043]; Landgraaf[0.043]; Tour[0.043]; Tour[0.043]; Italy[0.042]; Italy[0.042]; Marco[0.042]; Cycling[0.042]; Skibby[0.041]; time[0.041]; time[0.041]; time[0.041]; Ekimov[0.041]; Ekimov[0.041]; Armstrong[0.041]; Spain[0.041]; Erik[0.040]; Telekom[0.040]; Telekom[0.040]; Rolf[0.040]; Denmark[0.040]; Denmark[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Johan Capiot\u001b[39m ==> ENTITY: \u001b[32mJohan Capiot\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.224:0.224[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.046]; Breukink[0.044]; Breukink[0.044]; Landgraaf[0.044]; Polti[0.043]; Jesper[0.043]; Sweden[0.042]; Maarten[0.042]; Tour[0.042]; Denmark[0.042]; Denmark[0.042]; Erik[0.041]; Erik[0.041]; Olaf[0.041]; Andersson[0.041]; Andersson[0.041]; Saturday[0.040]; Rabobank[0.040]; Rabobank[0.040]; Rabobank[0.040]; Rabobank[0.040]; Lance[0.040]; Rolf[0.040]; Armstrong[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Breukink\u001b[39m ==> ENTITY: \u001b[32mErik Breukink\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBreukink[0.048]; Marco[0.043]; overall[0.042]; Skibby[0.042]; Skibby[0.042]; Ekimov[0.042]; Ekimov[0.042]; Telekom[0.042]; Rolf[0.042]; Johan[0.041]; stages[0.041]; Spain[0.041]; Belgium[0.041]; Rabobank[0.041]; Rabobank[0.041]; Rabobank[0.041]; Rabobank[0.041]; Erik[0.041]; time[0.041]; time[0.041]; time[0.041]; Netherlands[0.041]; Netherlands[0.041]; Netherlands[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jesper Skibby\u001b[39m ==> ENTITY: \u001b[32mJesper Skibby\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCycling[0.043]; Polti[0.043]; Denmark[0.043]; Denmark[0.043]; Landgraaf[0.043]; Landgraaf[0.043]; Netherlands[0.042]; Netherlands[0.042]; Netherlands[0.042]; Netherlands[0.042]; Netherlands[0.042]; Netherlands[0.042]; Netherlands[0.042]; Tour[0.042]; Tour[0.042]; Belgium[0.041]; Breukink[0.041]; Breukink[0.041]; Roermond[0.040]; Ekimov[0.040]; Ekimov[0.040]; Italy[0.040]; Italy[0.040]; Italy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ekimov\u001b[39m ==> ENTITY: \u001b[32mViatcheslav Ekimov\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEkimov[0.050]; Armstrong[0.048]; Breukink[0.047]; Breukink[0.047]; Telekom[0.046]; Rolf[0.046]; Johan[0.046]; Lance[0.046]; Erik[0.046]; Erik[0.046]; Polti[0.045]; seconds[0.045]; seconds[0.045]; seconds[0.045]; Russia[0.045]; overall[0.045]; Skibby[0.044]; Skibby[0.044]; Belgium[0.044]; Vyacheslav[0.044]; stages[0.044]; Pavanello[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marco Lietti\u001b[39m ==> ENTITY: \u001b[32mMarco Lietti\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.047]; Italy[0.047]; Italy[0.047]; Luca[0.043]; Polti[0.042]; Spain[0.042]; Belgium[0.042]; Ekimov[0.041]; Ekimov[0.041]; time[0.041]; time[0.041]; time[0.041]; time[0.041]; Telekom[0.041]; Sweden[0.041]; Rolf[0.040]; Michael[0.040]; Johan[0.040]; Netherlands[0.040]; Netherlands[0.040]; Netherlands[0.040]; Netherlands[0.040]; Erik[0.040]; Erik[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.151:0.151[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.044]; Italy[0.044]; Italy[0.044]; Denmark[0.043]; Denmark[0.043]; Germany[0.043]; final[0.043]; overall[0.042]; Belgium[0.042]; Netherlands[0.042]; Netherlands[0.042]; Netherlands[0.042]; Netherlands[0.042]; Netherlands[0.042]; Netherlands[0.042]; sixth[0.041]; Sweden[0.041]; time[0.040]; time[0.040]; time[0.040]; time[0.040]; Russia[0.040]; Anguita[0.040]; Tour[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Netherlands ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.229:0.229[0]; log p(e|m)= -0.250:-0.250[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.048]; Netherlands[0.048]; Netherlands[0.048]; Netherlands[0.048]; Netherlands[0.048]; Belgium[0.046]; Denmark[0.045]; Denmark[0.045]; Germany[0.044]; Spain[0.044]; Sweden[0.044]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Roermond[0.042]; Landgraaf[0.042]; Landgraaf[0.042]; Russia[0.042]; final[0.041]; sixth[0.039]; Erik[0.039]; Erik[0.039]; Cycling[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Polti\u001b[39m ==> ENTITY: \u001b[32mTeam Polti\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCycling[0.045]; Tour[0.044]; Tour[0.044]; Italy[0.044]; Italy[0.044]; Italy[0.044]; Armstrong[0.043]; Landgraaf[0.043]; Landgraaf[0.043]; Breukink[0.043]; time[0.042]; time[0.042]; time[0.042]; Lance[0.041]; Telekom[0.040]; Telekom[0.040]; Rolf[0.040]; Marco[0.040]; Ekimov[0.040]; Ekimov[0.040]; Giovanni[0.040]; Johan[0.040]; Roermond[0.040]; Luca[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Denmark ==> ENTITY: \u001b[32mDenmark\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.177:0.177[0]; log p(e|m)= -0.268:-0.268[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.051]; Sweden[0.047]; Netherlands[0.043]; Netherlands[0.043]; Netherlands[0.043]; Netherlands[0.043]; Germany[0.043]; Belgium[0.042]; Andersson[0.041]; Andersson[0.041]; Spain[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; Jesper[0.041]; Erik[0.041]; Erik[0.041]; overall[0.040]; Final[0.040]; time[0.039]; time[0.039]; time[0.039]; time[0.039]; Sorensen[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Landgraaf\u001b[39m ==> ENTITY: \u001b[32mLandgraaf\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.099:0.099[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRoermond[0.046]; Landgraaf[0.046]; Netherlands[0.044]; Netherlands[0.044]; Netherlands[0.044]; Netherlands[0.044]; Netherlands[0.044]; km[0.042]; Denmark[0.042]; Rabobank[0.041]; Rabobank[0.041]; Germany[0.041]; Telekom[0.041]; Spain[0.041]; Erik[0.040]; Breukink[0.040]; Italy[0.040]; Italy[0.040]; Olaf[0.040]; Tour[0.040]; Tour[0.040]; Saturday[0.040]; Onda[0.039]; mins[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Roermond\u001b[39m ==> ENTITY: \u001b[32mRoermond\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.046]; Netherlands[0.045]; Netherlands[0.045]; Netherlands[0.045]; Netherlands[0.045]; Netherlands[0.045]; Landgraaf[0.044]; Landgraaf[0.044]; Denmark[0.041]; Telekom[0.041]; Telekom[0.041]; Johan[0.041]; Erik[0.040]; km[0.040]; Germany[0.040]; Spain[0.040]; Breukink[0.040]; Rabobank[0.040]; Rabobank[0.040]; Olaf[0.040]; Sweden[0.039]; Italy[0.039]; Italy[0.039]; mins[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.083:0.083[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsixth[0.045]; overall[0.044]; Denmark[0.044]; Denmark[0.044]; Sweden[0.043]; final[0.042]; Michael[0.042]; Germany[0.042]; Spain[0.041]; Final[0.041]; Final[0.041]; Erik[0.041]; Erik[0.041]; Belgium[0.041]; Netherlands[0.041]; Netherlands[0.041]; Netherlands[0.041]; Netherlands[0.041]; Netherlands[0.041]; Netherlands[0.041]; Netherlands[0.041]; Italy[0.041]; Italy[0.041]; Italy[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Maarten den Bakker\u001b[39m ==> ENTITY: \u001b[32mMaarten den Bakker\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.132:0.132[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.048]; Netherlands[0.046]; Netherlands[0.046]; Denmark[0.044]; Denmark[0.044]; Telekom[0.042]; Rolf[0.042]; Sweden[0.042]; Johan[0.042]; Erik[0.042]; Breukink[0.041]; Ekimov[0.041]; Ekimov[0.041]; Rabobank[0.041]; Rabobank[0.041]; Rabobank[0.041]; Rabobank[0.041]; Dekker[0.040]; Skibby[0.040]; Skibby[0.040]; Pavanello[0.040]; Italy[0.040]; Italy[0.040]; Anguita[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgium ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.207:0.207[0]; log p(e|m)= -0.242:-0.242[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.044]; Denmark[0.044]; Netherlands[0.044]; Netherlands[0.044]; Netherlands[0.044]; Netherlands[0.044]; Netherlands[0.044]; Germany[0.043]; Spain[0.043]; Sweden[0.042]; Italy[0.042]; Italy[0.042]; Italy[0.042]; final[0.041]; Roermond[0.041]; overall[0.040]; Tour[0.040]; Landgraaf[0.040]; Johan[0.040]; Polti[0.040]; Final[0.040]; Erik[0.039]; Erik[0.039]; Maarten[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Netherlands ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.250:-0.250[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.051]; Belgium[0.049]; Denmark[0.049]; Denmark[0.049]; Spain[0.047]; Sweden[0.047]; Italy[0.046]; Italy[0.046]; Final[0.043]; Bakker[0.042]; Johan[0.042]; Erik[0.042]; Sorensen[0.041]; overall[0.041]; time[0.041]; time[0.041]; time[0.041]; Maarten[0.041]; Ludwig[0.040]; Breukink[0.040]; Marco[0.040]; Andersson[0.040]; Andersson[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Denmark ==> ENTITY: \u001b[32mDenmark\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.189:0.189[0]; log p(e|m)= -0.268:-0.268[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDenmark[0.051]; Sweden[0.046]; final[0.043]; Netherlands[0.042]; Netherlands[0.042]; Netherlands[0.042]; Netherlands[0.042]; Netherlands[0.042]; Netherlands[0.042]; Netherlands[0.042]; Germany[0.042]; sixth[0.042]; Belgium[0.042]; Spain[0.041]; Italy[0.040]; Italy[0.040]; Italy[0.040]; Russia[0.040]; Andersson[0.039]; Andersson[0.039]; Final[0.039]; Final[0.039]; Jesper[0.039]; overall[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Netherlands ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.226:0.226[0]; log p(e|m)= -0.250:-0.250[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.048]; Netherlands[0.048]; Netherlands[0.048]; Netherlands[0.048]; Netherlands[0.048]; Belgium[0.045]; Denmark[0.045]; Denmark[0.045]; Germany[0.044]; Spain[0.044]; Sweden[0.043]; Italy[0.043]; Italy[0.043]; Italy[0.043]; Roermond[0.042]; Landgraaf[0.042]; Landgraaf[0.042]; Russia[0.042]; final[0.041]; Final[0.040]; Final[0.040]; Cycling[0.040]; sixth[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.056]; Italy[0.056]; Spain[0.046]; overall[0.044]; Denmark[0.044]; Denmark[0.044]; Marco[0.043]; Belgium[0.043]; Sweden[0.042]; time[0.042]; time[0.042]; time[0.042]; time[0.042]; Netherlands[0.042]; Netherlands[0.042]; Netherlands[0.042]; Netherlands[0.042]; Luca[0.042]; Ekimov[0.041]; Ekimov[0.041]; Russia[0.041]; Final[0.041]; Technogym[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Erik Dekker\u001b[39m ==> ENTITY: \u001b[32mErik Dekker\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBreukink[0.046]; Breukink[0.046]; Belgium[0.045]; Ekimov[0.045]; Ekimov[0.045]; Rabobank[0.044]; Rabobank[0.044]; Rabobank[0.044]; Rabobank[0.044]; Marco[0.044]; Erik[0.044]; stages[0.044]; Lance[0.044]; Armstrong[0.044]; Telekom[0.043]; Rolf[0.043]; Johan[0.043]; Maarten[0.042]; seconds[0.042]; seconds[0.042]; Skibby[0.042]; Skibby[0.042]; Spain[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Michael Andersson ==> ENTITY: \u001b[32mMichael Andersson (cyclist)\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.160:0.160[0]; log p(e|m)= -0.926:-0.926[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAndersson[0.046]; Tour[0.044]; Sweden[0.044]; Landgraaf[0.044]; time[0.042]; time[0.042]; time[0.042]; time[0.042]; Jesper[0.042]; Denmark[0.042]; Denmark[0.042]; Erik[0.041]; Erik[0.041]; Johan[0.041]; Telekom[0.041]; Telekom[0.041]; Rolf[0.041]; Roermond[0.040]; Italy[0.040]; Italy[0.040]; Italy[0.040]; Breukink[0.040]; Breukink[0.040]; Spain[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Landgraaf\u001b[39m ==> ENTITY: \u001b[32mLandgraaf\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.112:0.112[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRoermond[0.047]; Landgraaf[0.047]; Netherlands[0.046]; Netherlands[0.046]; Netherlands[0.046]; Netherlands[0.046]; Netherlands[0.046]; Belgium[0.045]; Sweden[0.044]; km[0.043]; Denmark[0.043]; Rabobank[0.043]; Rabobank[0.043]; Germany[0.042]; Telekom[0.042]; Telekom[0.042]; Johan[0.042]; Spain[0.042]; Erik[0.042]; Breukink[0.041]; Italy[0.041]; Italy[0.041]; Olaf[0.041]; \t\n\r [================================>.............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 6s263ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3514/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 983testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.077:0.077[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.053]; stg[0.046]; stg[0.046]; following[0.046]; said[0.045]; market[0.045]; Thursday[0.044]; south[0.044]; trade[0.044]; deal[0.044]; Friday[0.044]; marked[0.043]; Jim[0.043]; reported[0.043]; Trade[0.042]; follow[0.041]; think[0.041]; think[0.041]; high[0.041]; high[0.041]; Uk[0.041]; Uk[0.041]; gains[0.041]; \t\n\r [================================>.............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 6s324ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3517/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1157testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dini\u001b[39m ==> ENTITY: \u001b[32mLamberto Dini\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.087:0.087[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.047]; Italian[0.045]; Italian[0.045]; president[0.043]; Dini[0.042]; Dini[0.042]; Community[0.042]; Lamberto[0.042]; Egidio[0.041]; Minister[0.041]; Minister[0.041]; Rome[0.041]; Rome[0.041]; held[0.041]; ministry[0.041]; negotiator[0.041]; negotiator[0.041]; international[0.041]; Foreign[0.040]; said[0.040]; Nyerere[0.040]; Nyerere[0.040]; Nyerere[0.040]; Nyerere[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burundi ==> ENTITY: \u001b[32mBurundi\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.374:-0.374[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurundi[0.057]; country[0.045]; Buyoya[0.045]; Buyoya[0.045]; Buyoya[0.045]; Hutu[0.043]; Hutu[0.043]; Tutsi[0.043]; Nyerere[0.042]; army[0.041]; people[0.041]; time[0.040]; civilians[0.040]; factions[0.040]; president[0.039]; elected[0.039]; International[0.039]; coup[0.039]; coup[0.039]; Friday[0.039]; major[0.039]; Monday[0.039]; Sunday[0.039]; killed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burundi ==> ENTITY: \u001b[32mBurundi\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.162:0.162[0]; log p(e|m)= -0.374:-0.374[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurundi[0.054]; Burundi[0.054]; Burundi[0.054]; Tanzanian[0.044]; Buyoya[0.043]; Buyoya[0.043]; Buyoya[0.043]; Community[0.041]; Tutsi[0.041]; region[0.040]; Nyerere[0.040]; Nyerere[0.040]; Nyerere[0.040]; Nyerere[0.040]; Catholic[0.040]; private[0.039]; Foreign[0.039]; international[0.039]; time[0.038]; July[0.038]; ministry[0.038]; ministry[0.038]; military[0.038]; held[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hutu\u001b[39m ==> ENTITY: \u001b[32mHutu\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTutsi[0.046]; Burundi[0.046]; Burundi[0.046]; Hutu[0.045]; Buyoya[0.045]; Buyoya[0.045]; Buyoya[0.045]; country[0.043]; army[0.042]; rebels[0.042]; people[0.042]; civilians[0.041]; factions[0.041]; Peace[0.040]; president[0.040]; coup[0.040]; coup[0.040]; parties[0.040]; killed[0.039]; condemned[0.039]; major[0.039]; Council[0.039]; said[0.039]; ruled[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burundi ==> ENTITY: \u001b[32mBurundi\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.160:0.160[0]; log p(e|m)= -0.374:-0.374[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurundi[0.054]; Burundi[0.054]; Burundi[0.054]; Burundi[0.054]; Tanzanian[0.044]; Community[0.041]; region[0.040]; Nyerere[0.040]; Nyerere[0.040]; Nyerere[0.040]; Nyerere[0.040]; Nyerere[0.040]; Catholic[0.040]; private[0.039]; Foreign[0.039]; international[0.039]; July[0.038]; ministry[0.038]; ministry[0.038]; military[0.038]; held[0.038]; Italy[0.038]; envoy[0.038]; president[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Julius Nyerere\u001b[39m ==> ENTITY: \u001b[32mJulius Nyerere\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNyerere[0.044]; Nyerere[0.044]; Nyerere[0.044]; region[0.044]; Tanzanian[0.043]; president[0.043]; Burundi[0.042]; Burundi[0.042]; Burundi[0.042]; Burundi[0.042]; Burundi[0.042]; coup[0.042]; held[0.041]; said[0.041]; said[0.041]; private[0.041]; Lakes[0.041]; visit[0.041]; international[0.041]; military[0.040]; following[0.040]; Howard[0.040]; Community[0.040]; Catholic[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Buyoya\u001b[39m ==> ENTITY: \u001b[32mPierre Buyoya\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBuyoya[0.047]; Buyoya[0.047]; coup[0.043]; coup[0.043]; coup[0.043]; elected[0.043]; Burundi[0.043]; Burundi[0.043]; Burundi[0.043]; military[0.042]; Hutu[0.042]; Hutu[0.042]; president[0.041]; etat[0.041]; army[0.041]; condemned[0.041]; Peace[0.040]; factions[0.040]; democratically[0.040]; Tutsi[0.039]; rebels[0.039]; country[0.039]; ruled[0.039]; embargo[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italian ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.141:0.141[0]; log p(e|m)= -0.759:-0.759[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.055]; Italian[0.052]; Rome[0.046]; Rome[0.046]; Roman[0.045]; region[0.044]; international[0.043]; Catholic[0.043]; Lamberto[0.042]; particular[0.040]; Egidio[0.040]; following[0.040]; developments[0.040]; special[0.039]; said[0.039]; private[0.039]; week[0.039]; Wolpe[0.039]; Howard[0.039]; visit[0.038]; Community[0.038]; latest[0.038]; met[0.038]; held[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nyerere\u001b[39m ==> ENTITY: \u001b[32mJulius Nyerere\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNyerere[0.043]; Nyerere[0.043]; Nyerere[0.043]; Nyerere[0.043]; region[0.043]; Tanzanian[0.042]; president[0.042]; Buyoya[0.042]; Burundi[0.042]; Burundi[0.042]; Burundi[0.042]; Burundi[0.042]; Burundi[0.042]; coup[0.041]; coup[0.041]; major[0.041]; held[0.041]; said[0.040]; said[0.040]; said[0.040]; private[0.040]; Lakes[0.040]; visit[0.040]; international[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tutsi\u001b[39m ==> ENTITY: \u001b[32mTutsi\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurundi[0.045]; Burundi[0.045]; Burundi[0.045]; Burundi[0.045]; Hutu[0.044]; Hutu[0.044]; Buyoya[0.043]; Buyoya[0.043]; Buyoya[0.043]; country[0.042]; army[0.042]; region[0.042]; rebels[0.041]; people[0.041]; particular[0.040]; military[0.040]; factions[0.040]; coup[0.039]; coup[0.039]; coup[0.039]; parties[0.039]; civilians[0.039]; Peace[0.039]; major[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italian ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.114:0.114[0]; log p(e|m)= -0.759:-0.759[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.053]; Italian[0.050]; Rome[0.044]; Rome[0.044]; Rome[0.044]; Roman[0.042]; region[0.042]; Burundi[0.042]; Burundi[0.042]; Burundi[0.042]; Burundi[0.042]; Burundi[0.042]; major[0.042]; international[0.041]; Catholic[0.041]; Lamberto[0.040]; Tanzanian[0.039]; coup[0.039]; coup[0.039]; July[0.039]; military[0.039]; etat[0.039]; private[0.039]; condemned[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nyerere\u001b[39m ==> ENTITY: \u001b[32mJulius Nyerere\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNyerere[0.044]; Nyerere[0.044]; Nyerere[0.044]; Nyerere[0.044]; region[0.044]; Tanzanian[0.043]; president[0.042]; Burundi[0.042]; Burundi[0.042]; Burundi[0.042]; Burundi[0.042]; Burundi[0.042]; coup[0.042]; held[0.041]; said[0.041]; said[0.041]; private[0.041]; Lakes[0.041]; visit[0.040]; international[0.040]; military[0.040]; following[0.040]; Howard[0.040]; Community[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lamberto Dini\u001b[39m ==> ENTITY: \u001b[32mLamberto Dini\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.103:0.103[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.046]; Italian[0.045]; Italian[0.045]; president[0.043]; Lakes[0.042]; Dini[0.042]; Dini[0.042]; military[0.042]; Community[0.042]; following[0.041]; Egidio[0.041]; Minister[0.041]; Minister[0.041]; Rome[0.041]; Rome[0.041]; held[0.041]; ministry[0.041]; negotiator[0.041]; negotiator[0.041]; July[0.041]; international[0.040]; Foreign[0.040]; said[0.040]; region[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nyerere\u001b[39m ==> ENTITY: \u001b[32mJulius Nyerere\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.046]; Nyerere[0.043]; Nyerere[0.043]; region[0.043]; Buyoya[0.042]; Buyoya[0.042]; Buyoya[0.042]; Burundi[0.042]; Burundi[0.042]; Burundi[0.042]; Burundi[0.042]; Burundi[0.042]; Burundi[0.042]; coup[0.041]; coup[0.041]; major[0.041]; held[0.041]; said[0.040]; said[0.040]; said[0.040]; private[0.040]; Lakes[0.040]; visit[0.040]; international[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burundi ==> ENTITY: \u001b[32mBurundi\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.149:0.149[0]; log p(e|m)= -0.374:-0.374[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurundi[0.056]; Burundi[0.056]; Burundi[0.056]; Tanzanian[0.045]; Community[0.042]; region[0.042]; Nyerere[0.041]; Nyerere[0.041]; Nyerere[0.041]; Nyerere[0.041]; private[0.040]; Catholic[0.040]; Foreign[0.040]; international[0.040]; ministry[0.039]; envoy[0.039]; Italy[0.039]; president[0.039]; week[0.038]; said[0.038]; Julius[0.037]; held[0.037]; Saturday[0.037]; special[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burundi ==> ENTITY: \u001b[32mBurundi\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.152:0.152[0]; log p(e|m)= -0.374:-0.374[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurundi[0.056]; country[0.045]; Buyoya[0.045]; Buyoya[0.045]; Buyoya[0.045]; Hutu[0.043]; Hutu[0.043]; Tutsi[0.043]; Nyerere[0.042]; army[0.041]; people[0.040]; time[0.040]; July[0.040]; civilians[0.040]; ministry[0.040]; factions[0.039]; military[0.039]; president[0.039]; elected[0.039]; International[0.039]; coup[0.039]; coup[0.039]; coup[0.039]; statement[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dini\u001b[39m ==> ENTITY: \u001b[32mLamberto Dini\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.112:0.112[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.046]; Italian[0.045]; Italian[0.045]; president[0.043]; Lakes[0.042]; Dini[0.042]; Dini[0.042]; military[0.042]; Community[0.042]; Lamberto[0.041]; following[0.041]; Egidio[0.041]; presented[0.041]; Minister[0.041]; Minister[0.041]; Rome[0.041]; Rome[0.041]; Rome[0.041]; held[0.041]; ministry[0.041]; ministry[0.041]; negotiator[0.040]; negotiator[0.040]; July[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.085:0.085[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninternational[0.049]; Italy[0.047]; week[0.046]; latest[0.046]; International[0.044]; following[0.044]; July[0.044]; statement[0.044]; particular[0.044]; Howard[0.043]; held[0.043]; Foreign[0.043]; Saturday[0.043]; respect[0.043]; military[0.042]; Monday[0.042]; Sunday[0.042]; private[0.042]; met[0.042]; region[0.042]; prize[0.041]; gave[0.041]; Peace[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tanzanian\u001b[39m ==> ENTITY: \u001b[32mTanzania\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfollowing[0.047]; Catholic[0.045]; region[0.045]; Burundi[0.043]; Burundi[0.043]; Burundi[0.043]; Burundi[0.043]; Burundi[0.043]; latest[0.043]; meets[0.042]; Foreign[0.042]; international[0.042]; Community[0.041]; ministry[0.040]; ministry[0.040]; held[0.040]; particular[0.040]; private[0.040]; week[0.040]; visit[0.040]; developments[0.040]; said[0.039]; said[0.039]; Italy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pierre Buyoya\u001b[39m ==> ENTITY: \u001b[32mPierre Buyoya\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBuyoya[0.047]; Buyoya[0.047]; coup[0.043]; coup[0.043]; coup[0.043]; elected[0.043]; Burundi[0.042]; Burundi[0.042]; Burundi[0.042]; Burundi[0.042]; military[0.042]; Hutu[0.042]; Hutu[0.042]; president[0.041]; etat[0.041]; army[0.041]; condemned[0.041]; Peace[0.040]; factions[0.040]; democratically[0.040]; Tutsi[0.039]; rebels[0.039]; country[0.039]; ruled[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Roman Catholic\u001b[39m ==> ENTITY: \u001b[32mCatholic Church\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.125:0.125[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCommunity[0.045]; Rome[0.044]; Rome[0.044]; Rome[0.044]; Italy[0.044]; statement[0.044]; private[0.043]; region[0.042]; particular[0.042]; respect[0.041]; Italian[0.041]; Italian[0.041]; ministry[0.041]; ministry[0.041]; said[0.041]; said[0.041]; week[0.040]; Pierre[0.040]; Egidio[0.040]; international[0.040]; major[0.040]; special[0.040]; July[0.040]; Volunteers[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rome ==> ENTITY: \u001b[32mRome\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRoman[0.050]; Catholic[0.046]; region[0.044]; Italian[0.044]; country[0.042]; latest[0.042]; military[0.042]; statement[0.041]; people[0.041]; army[0.041]; Egidio[0.041]; held[0.041]; following[0.041]; elected[0.041]; Sunday[0.041]; condemned[0.040]; Howard[0.040]; July[0.040]; special[0.040]; particular[0.040]; ministry[0.040]; said[0.040]; said[0.040]; respect[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Howard Wolpe\u001b[39m ==> ENTITY: \u001b[32mHoward Wolpe\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.090:0.090[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCouncil[0.043]; Foreign[0.043]; president[0.043]; visit[0.043]; International[0.042]; Burundi[0.042]; Burundi[0.042]; Burundi[0.042]; Burundi[0.042]; Burundi[0.042]; Lakes[0.042]; Security[0.042]; envoy[0.042]; military[0.042]; Community[0.042]; Monday[0.041]; July[0.041]; international[0.041]; Minister[0.040]; Minister[0.040]; ministry[0.040]; ministry[0.040]; negotiator[0.040]; negotiator[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burundi ==> ENTITY: \u001b[32mBurundi\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.159:0.159[0]; log p(e|m)= -0.374:-0.374[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurundi[0.054]; Burundi[0.054]; Burundi[0.054]; Burundi[0.054]; Tanzanian[0.044]; Community[0.041]; region[0.040]; Nyerere[0.040]; Nyerere[0.040]; Nyerere[0.040]; Nyerere[0.040]; Nyerere[0.040]; private[0.039]; Catholic[0.039]; Foreign[0.039]; international[0.039]; July[0.038]; envoy[0.038]; visit[0.038]; ministry[0.038]; ministry[0.038]; president[0.038]; military[0.038]; Italy[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hutu\u001b[39m ==> ENTITY: \u001b[32mHutu\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTutsi[0.046]; Burundi[0.046]; Burundi[0.046]; Hutu[0.045]; Buyoya[0.045]; Buyoya[0.045]; Buyoya[0.045]; country[0.043]; army[0.042]; rebels[0.042]; people[0.042]; civilians[0.041]; factions[0.041]; Peace[0.040]; president[0.040]; coup[0.040]; coup[0.040]; parties[0.040]; killed[0.039]; condemned[0.039]; major[0.039]; Council[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.129:0.129[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItalian[0.048]; Italian[0.048]; Rome[0.046]; Rome[0.046]; international[0.044]; week[0.043]; Lamberto[0.043]; Roman[0.043]; Egidio[0.041]; informed[0.041]; held[0.041]; Saturday[0.040]; said[0.040]; Catholic[0.040]; closely[0.040]; special[0.040]; president[0.040]; latest[0.040]; met[0.039]; visit[0.039]; private[0.039]; arrived[0.039]; talks[0.039]; Foreign[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nyerere\u001b[39m ==> ENTITY: \u001b[32mJulius Nyerere\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNyerere[0.044]; Nyerere[0.044]; Nyerere[0.044]; region[0.044]; Tanzanian[0.043]; president[0.043]; Burundi[0.043]; Burundi[0.043]; Burundi[0.043]; Burundi[0.043]; held[0.042]; said[0.041]; private[0.041]; Lakes[0.041]; visit[0.041]; international[0.041]; Howard[0.041]; Community[0.040]; Julius[0.040]; Catholic[0.040]; developments[0.040]; Minister[0.040]; Minister[0.040]; ministry[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mGreat Lakes\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAfrican Great Lakes\u001b[39m <---> \u001b[32mGreat Lakes\u001b[39m\t\nSCORES: global= 0.260:0.250[\u001b[31m0.010\u001b[39m]; local()= 0.155:0.032[\u001b[31m0.123\u001b[39m]; log p(e|m)= -2.477:-0.100[\u001b[32m2.377\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nregion[0.048]; major[0.045]; following[0.043]; international[0.043]; Friday[0.042]; Burundi[0.042]; Burundi[0.042]; Burundi[0.042]; Burundi[0.042]; Burundi[0.042]; Monday[0.041]; July[0.041]; held[0.041]; Tanzanian[0.041]; said[0.041]; said[0.041]; said[0.041]; parties[0.041]; prize[0.041]; particular[0.041]; Community[0.041]; latest[0.041]; week[0.040]; visit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Buyoya\u001b[39m ==> ENTITY: \u001b[32mPierre Buyoya\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBuyoya[0.047]; Buyoya[0.047]; coup[0.043]; coup[0.043]; coup[0.043]; elected[0.043]; Burundi[0.043]; Burundi[0.043]; Burundi[0.043]; military[0.042]; Hutu[0.042]; Hutu[0.042]; president[0.041]; etat[0.041]; army[0.041]; condemned[0.041]; Peace[0.040]; factions[0.040]; democratically[0.040]; Tutsi[0.039]; rebels[0.039]; country[0.039]; ruled[0.039]; embargo[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Burundi ==> ENTITY: \u001b[32mBurundi\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.164:0.164[0]; log p(e|m)= -0.374:-0.374[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBurundi[0.054]; Burundi[0.054]; Burundi[0.054]; Burundi[0.054]; Tanzanian[0.043]; Buyoya[0.042]; Community[0.040]; Tutsi[0.040]; region[0.040]; Nyerere[0.040]; Nyerere[0.040]; Nyerere[0.040]; Nyerere[0.040]; Nyerere[0.040]; Catholic[0.039]; private[0.039]; Foreign[0.038]; international[0.038]; time[0.038]; July[0.038]; ministry[0.037]; ministry[0.037]; military[0.037]; held[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rome ==> ENTITY: \u001b[32mRome\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.162:0.162[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRome[0.050]; Roman[0.049]; Catholic[0.049]; Italy[0.048]; region[0.043]; Italian[0.043]; Italian[0.043]; latest[0.041]; week[0.041]; military[0.041]; international[0.040]; statement[0.040]; private[0.040]; Egidio[0.040]; following[0.040]; visit[0.040]; Howard[0.039]; July[0.039]; particular[0.039]; ministry[0.039]; ministry[0.039]; said[0.039]; said[0.039]; respect[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N. Security Council\u001b[39m ==> ENTITY: \u001b[32mUnited Nations Security Council\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCouncil[0.045]; resolution[0.044]; condemned[0.044]; statement[0.043]; military[0.042]; flatly[0.042]; country[0.042]; embargo[0.042]; following[0.042]; envoy[0.042]; negotiations[0.041]; Monday[0.041]; July[0.041]; Organisation[0.041]; Buyoya[0.041]; Buyoya[0.041]; Buyoya[0.041]; president[0.041]; talks[0.041]; talks[0.041]; International[0.041]; latest[0.041]; Peace[0.041]; monitoring[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Rome ==> ENTITY: \u001b[32mRome\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.151:0.151[0]; log p(e|m)= -0.252:-0.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRome[0.051]; Roman[0.050]; Italy[0.049]; Catholic[0.045]; region[0.044]; Italian[0.043]; Italian[0.043]; week[0.042]; latest[0.041]; international[0.041]; Egidio[0.040]; visit[0.040]; Howard[0.040]; particular[0.040]; ministry[0.040]; said[0.040]; respect[0.040]; special[0.039]; Foreign[0.039]; met[0.039]; held[0.039]; private[0.039]; Saturday[0.039]; Great[0.038]; \t\n\r [=================================>............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 6s108ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3550/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1149testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tunis ==> ENTITY: \u001b[32mTunis\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.033:-0.033[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTunisia[0.046]; Tunisia[0.046]; Tunisia[0.046]; Tunisia[0.046]; Tunisian[0.044]; Tunisian[0.044]; Tunisian[0.044]; France[0.044]; France[0.044]; Tripoli[0.044]; al[0.042]; Abidine[0.041]; parliament[0.040]; Habib[0.040]; Aug[0.040]; immigrants[0.039]; Saturday[0.039]; exports[0.039]; Palestinian[0.039]; represent[0.039]; dinars[0.039]; Zine[0.039]; Digest[0.038]; legal[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tunisian ==> ENTITY: \u001b[32mTunisia\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.187:0.187[0]; log p(e|m)= -0.131:-0.131[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTunisia[0.045]; Tunisia[0.045]; Tunisia[0.045]; Tunisia[0.045]; Tunisian[0.045]; Tunisian[0.045]; Tunis[0.044]; al[0.043]; France[0.043]; France[0.043]; Tripoli[0.043]; Libyan[0.042]; Abidine[0.041]; Ali[0.041]; parliament[0.040]; revolution[0.040]; Palestinian[0.040]; Habib[0.040]; immigrants[0.039]; Ben[0.039]; parts[0.039]; legal[0.038]; dinars[0.038]; Reuters[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Libyan ==> ENTITY: \u001b[32mLibya\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.140:0.140[0]; log p(e|m)= -0.273:-0.273[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTripoli[0.045]; Tunisia[0.044]; Tunisia[0.044]; Tunisia[0.044]; immigrants[0.044]; Tunisian[0.043]; Tunisian[0.043]; Tunisian[0.043]; al[0.043]; revolution[0.042]; dinars[0.042]; France[0.041]; France[0.041]; Palestinian[0.041]; Ali[0.041]; parliament[0.040]; parts[0.040]; Habib[0.040]; residents[0.040]; respect[0.040]; Reuters[0.039]; dignity[0.039]; Trade[0.039]; Authority[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Palestinian Authority\u001b[39m ==> ENTITY: \u001b[32mPalestinian National Authority\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.111:0.111[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nal[0.045]; Tunis[0.045]; Habib[0.044]; Tunisian[0.043]; Tunisian[0.043]; Tunisian[0.043]; Tunisia[0.043]; Tunisia[0.043]; Tunisia[0.043]; Tunisia[0.043]; Tripoli[0.043]; Libyan[0.042]; parliament[0.042]; President[0.041]; talks[0.041]; Ali[0.040]; parts[0.040]; Reuters[0.040]; legal[0.040]; dinars[0.039]; Press[0.039]; amounted[0.039]; dinar[0.039]; million[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.094:0.094[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.051]; Tunisia[0.044]; Tunisia[0.044]; Tunisia[0.044]; Tunisia[0.044]; revolution[0.043]; parliament[0.042]; amounted[0.041]; Presse[0.041]; Tunisian[0.041]; Tunisian[0.041]; Tunisian[0.041]; Le[0.041]; represent[0.041]; million[0.040]; Saturday[0.040]; parts[0.040]; respect[0.040]; immigrants[0.040]; welcomes[0.040]; President[0.040]; press[0.040]; Tunis[0.040]; called[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tunisia ==> ENTITY: \u001b[32mTunisia\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.180:0.180[0]; log p(e|m)= -0.393:-0.393[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTunisia[0.046]; Tunisia[0.046]; Tunisia[0.046]; Tunisian[0.044]; Tunisian[0.044]; Tunisian[0.044]; France[0.043]; France[0.043]; Palestinian[0.043]; Tunis[0.043]; Libyan[0.042]; Tripoli[0.041]; Abidine[0.041]; al[0.041]; Ali[0.040]; Habib[0.040]; parliament[0.040]; Authority[0.040]; revolution[0.040]; talks[0.039]; represent[0.039]; respect[0.039]; Ben[0.038]; leading[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Habib Boulares\u001b[39m ==> ENTITY: \u001b[32mHabib Boularès\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.043]; Ali[0.043]; Ben[0.043]; press[0.043]; Tunisia[0.043]; Tunisia[0.043]; Tunisia[0.043]; Tunisia[0.043]; parliament[0.043]; Tunisian[0.042]; Tunisian[0.042]; Tunisian[0.042]; leading[0.041]; immigrants[0.041]; Libyan[0.041]; Press[0.041]; Abidine[0.041]; Speaker[0.041]; exports[0.041]; Tunis[0.040]; al[0.040]; Trade[0.040]; Presse[0.040]; dinars[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tunisian ==> ENTITY: \u001b[32mTunisia\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.190:0.190[0]; log p(e|m)= -0.131:-0.131[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTunisia[0.045]; Tunisia[0.045]; Tunisia[0.045]; Tunisia[0.045]; Tunisian[0.044]; Tunisian[0.044]; Tunis[0.044]; al[0.043]; France[0.043]; France[0.043]; Tripoli[0.043]; Abidine[0.041]; Ali[0.041]; immigrants[0.040]; Libyan[0.040]; parliament[0.040]; revolution[0.040]; Habib[0.040]; Palestinian[0.039]; Ben[0.039]; parts[0.039]; million[0.039]; legal[0.038]; dinars[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tunisia ==> ENTITY: \u001b[32mTunisia\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.172:0.172[0]; log p(e|m)= -0.393:-0.393[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTunisia[0.046]; Tunisia[0.046]; Tunisia[0.046]; Tunisian[0.045]; Tunisian[0.045]; Tunisian[0.045]; France[0.044]; France[0.044]; Palestinian[0.044]; Tunis[0.043]; Tripoli[0.042]; Habib[0.041]; parliament[0.040]; Authority[0.040]; talks[0.040]; represent[0.039]; respect[0.039]; parts[0.039]; leading[0.039]; dinars[0.039]; million[0.039]; Zine[0.038]; Trade[0.038]; Saturday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.081:0.081[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npress[0.046]; immigrants[0.045]; Press[0.044]; parts[0.044]; leading[0.043]; Libyan[0.042]; Trade[0.042]; parliament[0.042]; celebrations[0.042]; Digest[0.041]; Tripoli[0.041]; says[0.041]; stories[0.041]; stories[0.041]; Habib[0.041]; million[0.041]; Palestinian[0.041]; exports[0.041]; anniversary[0.040]; verified[0.040]; amounted[0.040]; legal[0.040]; does[0.040]; revolution[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zine al-Abidine Ben Ali\u001b[39m ==> ENTITY: \u001b[32mZine El Abidine Ben Ali\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTunisia[0.046]; Tunisia[0.046]; Tunisia[0.046]; Tunisia[0.046]; Tunisian[0.044]; Tunisian[0.044]; Tunisian[0.044]; Tunis[0.042]; Libyan[0.042]; President[0.042]; parliament[0.042]; legal[0.041]; Habib[0.041]; revolution[0.041]; Ali[0.040]; Tripoli[0.040]; Reuters[0.040]; France[0.040]; France[0.040]; dignity[0.039]; leading[0.039]; called[0.039]; Palestinian[0.039]; press[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tunisia ==> ENTITY: \u001b[32mTunisia\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.169:0.169[0]; log p(e|m)= -0.393:-0.393[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTunisia[0.045]; Tunisia[0.045]; Tunisia[0.045]; Tunisian[0.044]; Tunisian[0.044]; Tunisian[0.044]; France[0.043]; France[0.043]; Palestinian[0.043]; Tunis[0.042]; Libyan[0.042]; Tripoli[0.041]; Abidine[0.041]; al[0.041]; Ali[0.040]; Habib[0.040]; legal[0.040]; parliament[0.040]; Authority[0.040]; revolution[0.040]; talks[0.039]; President[0.039]; welcomes[0.039]; respect[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.094:0.094[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.051]; Tunisia[0.044]; Tunisia[0.044]; Tunisia[0.044]; Tunisia[0.044]; revolution[0.043]; parliament[0.042]; amounted[0.041]; Presse[0.041]; Tunisian[0.041]; Tunisian[0.041]; Tunisian[0.041]; Le[0.041]; represent[0.041]; million[0.040]; Saturday[0.040]; parts[0.040]; respect[0.040]; immigrants[0.040]; welcomes[0.040]; President[0.040]; press[0.040]; Tunis[0.040]; called[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tunisian ==> ENTITY: \u001b[32mTunisia\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.183:0.183[0]; log p(e|m)= -0.131:-0.131[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTunisia[0.045]; Tunisia[0.045]; Tunisia[0.045]; Tunisia[0.045]; Tunisian[0.045]; Tunisian[0.045]; Tunis[0.044]; al[0.043]; France[0.043]; France[0.043]; Tripoli[0.043]; Libyan[0.042]; Abidine[0.041]; Ali[0.041]; parliament[0.040]; Palestinian[0.040]; Habib[0.040]; immigrants[0.039]; Ben[0.039]; parts[0.039]; legal[0.038]; dinars[0.038]; Reuters[0.038]; Zine[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Le Temps\u001b[39m ==> ENTITY: \u001b[32mLe Temps\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresse[0.046]; France[0.046]; France[0.046]; Press[0.043]; La[0.043]; press[0.043]; leading[0.042]; Tunisian[0.042]; Tunisian[0.042]; Tunisian[0.042]; million[0.042]; Reuters[0.041]; Tunis[0.041]; Digest[0.041]; spare[0.041]; stories[0.040]; stories[0.040]; Saturday[0.040]; Aug[0.040]; Tunisia[0.040]; Tunisia[0.040]; Tunisia[0.040]; Tunisia[0.040]; exports[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mLa Presse\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mLa Presse (Canadian newspaper)\u001b[39m <---> \u001b[32mLa Presse (French newspaper)\u001b[39m\t\nSCORES: global= 0.290:0.285[\u001b[31m0.005\u001b[39m]; local()= 0.081:0.099[\u001b[32m0.018\u001b[39m]; log p(e|m)= 0.000:-0.043[\u001b[31m0.043\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPress[0.047]; Le[0.046]; Digest[0.046]; France[0.045]; France[0.045]; press[0.044]; million[0.042]; says[0.042]; welcomes[0.042]; President[0.041]; amounted[0.041]; stories[0.041]; stories[0.041]; called[0.041]; exports[0.041]; Zine[0.040]; does[0.040]; anniversary[0.040]; Saturday[0.040]; leading[0.040]; parliament[0.040]; dinars[0.040]; Speaker[0.040]; dinar[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tunisia ==> ENTITY: \u001b[32mTunisia\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.169:0.169[0]; log p(e|m)= -0.393:-0.393[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTunisia[0.045]; Tunisia[0.045]; Tunisia[0.045]; Tunisian[0.044]; Tunisian[0.044]; Tunisian[0.044]; France[0.043]; France[0.043]; Palestinian[0.043]; Tunis[0.042]; Libyan[0.042]; Tripoli[0.041]; Abidine[0.041]; al[0.041]; Ali[0.040]; Habib[0.040]; legal[0.040]; parliament[0.040]; Authority[0.040]; revolution[0.040]; talks[0.039]; President[0.039]; welcomes[0.039]; respect[0.039]; \t\n\r [=================================>............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 6s49ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3567/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1008testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.039:0.039[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncontest[0.047]; Ross[0.046]; President[0.043]; economic[0.043]; list[0.043]; candidates[0.043]; partying[0.042]; race[0.041]; Bob[0.041]; season[0.041]; promises[0.041]; presidential[0.041]; Clinton[0.041]; puts[0.041]; includes[0.041]; add[0.041]; say[0.041]; people[0.041]; raise[0.041]; believe[0.041]; Federal[0.040]; dollars[0.040]; Street[0.040]; Street[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nasdaq\u001b[39m ==> ENTITY: \u001b[32mNASDAQ\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstock[0.052]; market[0.048]; closed[0.048]; closed[0.048]; Exchange[0.047]; Stock[0.045]; American[0.044]; Friday[0.043]; Friday[0.043]; stocks[0.043]; stocks[0.043]; week[0.043]; week[0.043]; week[0.043]; week[0.043]; November[0.042]; hold[0.041]; said[0.041]; said[0.041]; higher[0.041]; economic[0.040]; Standard[0.040]; lower[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = George Bush ==> ENTITY: \u001b[32mGeorge H. W. Bush\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.810:-0.810[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncampaign[0.046]; American[0.044]; Reagan[0.044]; Clinton[0.044]; John[0.044]; presidential[0.043]; president[0.043]; Washington[0.042]; Democrat[0.041]; candidates[0.041]; candidate[0.041]; won[0.041]; won[0.041]; won[0.041]; election[0.041]; Johnson[0.041]; budget[0.041]; North[0.040]; surge[0.040]; taxes[0.040]; firm[0.040]; Mondale[0.040]; Ronald[0.040]; week[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Standard & Poor\u001b[39m ==> ENTITY: \u001b[32mS&P 500\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstock[0.045]; index[0.044]; index[0.044]; index[0.044]; index[0.044]; market[0.044]; lower[0.042]; Exchange[0.042]; higher[0.042]; stocks[0.041]; stocks[0.041]; Stock[0.041]; Nasdaq[0.041]; week[0.041]; week[0.041]; week[0.041]; week[0.041]; composite[0.041]; American[0.041]; Dow[0.040]; closed[0.040]; closed[0.040]; risk[0.040]; November[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dole\u001b[39m ==> ENTITY: \u001b[32mBob Dole\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJohn[0.049]; votes[0.045]; Clinton[0.044]; Clinton[0.044]; election[0.044]; American[0.044]; president[0.043]; Washington[0.042]; political[0.042]; promise[0.041]; home[0.041]; win[0.041]; House[0.041]; promises[0.040]; promises[0.040]; week[0.040]; Dole[0.040]; said[0.040]; said[0.040]; said[0.040]; agenda[0.040]; spare[0.040]; gains[0.040]; carries[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dole\u001b[39m ==> ENTITY: \u001b[32mBob Dole\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.096:0.096[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinton[0.046]; candidates[0.046]; Johnson[0.045]; Perot[0.044]; political[0.043]; economic[0.043]; promises[0.042]; said[0.041]; said[0.041]; Hugh[0.041]; financial[0.041]; financial[0.041]; raise[0.041]; believe[0.041]; list[0.041]; legislation[0.041]; touches[0.041]; shots[0.041]; dollars[0.040]; dynamic[0.040]; economy[0.040]; drop[0.040]; includes[0.040]; socks[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Whitewater ==> ENTITY: \u001b[32mWhitewater controversy\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.082:0.082[0]; log p(e|m)= -1.073:-1.073[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinton[0.050]; Clintons[0.049]; November[0.049]; edge[0.048]; pattern[0.048]; looks[0.048]; looks[0.048]; election[0.048]; election[0.048]; election[0.048]; high[0.048]; summer[0.047]; candidates[0.047]; presidency[0.047]; average[0.047]; make[0.047]; turn[0.047]; record[0.046]; scandal[0.046]; industrial[0.046]; probably[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ross Perot\u001b[39m ==> ENTITY: \u001b[32mRoss Perot\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npresidential[0.047]; Clinton[0.046]; Clinton[0.046]; candidates[0.044]; President[0.043]; Bob[0.043]; economic[0.042]; economy[0.042]; political[0.042]; political[0.042]; dollars[0.041]; Dole[0.041]; Dole[0.041]; invested[0.041]; contest[0.040]; race[0.040]; scaring[0.040]; thing[0.040]; impact[0.040]; Reserve[0.040]; financial[0.040]; financial[0.040]; raise[0.040]; includes[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = White House\u001b[39m ==> ENTITY: \u001b[32mWhite House\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.006:-0.006[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGeorge[0.046]; Washington[0.045]; Clinton[0.044]; John[0.044]; campaign[0.043]; Reagan[0.043]; American[0.043]; Bush[0.043]; presidential[0.043]; president[0.042]; election[0.041]; rate[0.041]; Services[0.041]; Johnson[0.040]; North[0.040]; second[0.040]; week[0.040]; Ronald[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; Mondale[0.040]; Democrat[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mDow Jones\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mDow Jones & Company\u001b[39m <---> \u001b[32mDow Jones Industrial Average\u001b[39m\t\nSCORES: global= 0.268:0.264[\u001b[31m0.004\u001b[39m]; local()= 0.093:0.212[\u001b[32m0.119\u001b[39m]; log p(e|m)= 0.000:-0.423[\u001b[31m0.423\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsurge[0.043]; sharply[0.043]; decade[0.042]; highs[0.042]; market[0.042]; market[0.042]; market[0.042]; low[0.042]; corporate[0.042]; stocks[0.042]; boosted[0.042]; Stocks[0.042]; term[0.042]; term[0.042]; rate[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; high[0.041]; average[0.041]; narrowed[0.040]; jumped[0.040]; pattern[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dole\u001b[39m ==> ENTITY: \u001b[32mBob Dole\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.115:0.115[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinton[0.045]; Clinton[0.045]; Johnson[0.045]; election[0.045]; president[0.044]; home[0.042]; House[0.042]; week[0.041]; Dole[0.041]; said[0.041]; said[0.041]; said[0.041]; agenda[0.041]; spare[0.041]; Hugh[0.041]; gains[0.041]; carries[0.041]; federal[0.040]; touches[0.040]; board[0.040]; White[0.040]; dynamic[0.040]; hire[0.040]; drop[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = American Stock Exchange\u001b[39m ==> ENTITY: \u001b[32mAmerican Stock Exchange\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstock[0.047]; market[0.045]; Nasdaq[0.045]; closed[0.044]; closed[0.044]; stocks[0.043]; stocks[0.043]; Friday[0.042]; Friday[0.042]; said[0.042]; November[0.042]; week[0.041]; week[0.041]; week[0.041]; week[0.041]; prices[0.040]; index[0.040]; index[0.040]; index[0.040]; index[0.040]; lower[0.040]; higher[0.040]; Standard[0.039]; Dow[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Federal Reserve\u001b[39m ==> ENTITY: \u001b[32mFederal Reserve System\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfinancial[0.046]; financial[0.046]; President[0.045]; legislation[0.045]; economic[0.042]; dollars[0.042]; investment[0.042]; economy[0.042]; rates[0.042]; includes[0.042]; chief[0.041]; said[0.041]; said[0.041]; Johnson[0.041]; Ross[0.041]; York[0.040]; market[0.040]; market[0.040]; market[0.040]; billions[0.040]; worry[0.040]; presidential[0.040]; trade[0.040]; add[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clinton ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.123:0.123[0]; log p(e|m)= -1.917:-1.917[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncampaign[0.046]; George[0.046]; Bush[0.045]; rate[0.045]; John[0.045]; presidential[0.045]; rates[0.045]; American[0.045]; taxes[0.044]; Washington[0.044]; Reagan[0.044]; tax[0.043]; tax[0.043]; tax[0.043]; president[0.043]; Democrat[0.043]; North[0.042]; week[0.042]; candidates[0.042]; candidate[0.042]; Johnson[0.042]; election[0.041]; second[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Democrat ==> ENTITY: \u001b[32mDemocratic Party (United States)\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.139:0.139[0]; log p(e|m)= -0.062:-0.062[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.045]; votes[0.044]; Reagan[0.044]; Washington[0.044]; George[0.044]; John[0.043]; election[0.042]; candidate[0.042]; campaign[0.042]; candidates[0.042]; Clinton[0.041]; jumped[0.041]; presidential[0.041]; elections[0.041]; president[0.041]; Johnson[0.041]; second[0.041]; political[0.041]; Mondale[0.040]; won[0.040]; won[0.040]; won[0.040]; House[0.040]; win[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clinton ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.095:0.095[0]; log p(e|m)= -1.917:-1.917[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinton[0.045]; North[0.044]; tax[0.044]; tax[0.044]; taxes[0.043]; taxes[0.043]; taxes[0.043]; American[0.043]; John[0.043]; Washington[0.042]; federal[0.042]; president[0.042]; income[0.042]; budget[0.041]; people[0.041]; include[0.041]; Street[0.040]; White[0.040]; House[0.040]; election[0.040]; increase[0.040]; lower[0.040]; board[0.039]; percent[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clinton ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.046:0.046[0]; log p(e|m)= -1.917:-1.917[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlist[0.045]; Albany[0.044]; people[0.043]; Johnson[0.043]; economic[0.043]; include[0.043]; Street[0.042]; Ross[0.042]; Wall[0.041]; dollars[0.041]; legislation[0.041]; road[0.041]; drop[0.041]; Federal[0.041]; political[0.041]; includes[0.041]; middle[0.041]; economy[0.041]; individuals[0.041]; rates[0.041]; rates[0.041]; rates[0.041]; raise[0.040]; add[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dole\u001b[39m ==> ENTITY: \u001b[32mBob Dole\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinton[0.045]; candidates[0.045]; polls[0.045]; election[0.045]; election[0.045]; presidency[0.043]; political[0.042]; political[0.042]; happens[0.042]; economic[0.042]; hold[0.042]; does[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; surge[0.041]; probably[0.040]; Jones[0.040]; horizontal[0.040]; record[0.040]; holding[0.039]; edge[0.039]; narrowed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.089:0.089[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.049]; George[0.048]; campaign[0.048]; election[0.047]; John[0.047]; American[0.046]; Democrat[0.046]; presidential[0.046]; votes[0.045]; Clinton[0.045]; Clinton[0.045]; promised[0.045]; said[0.045]; said[0.045]; said[0.045]; president[0.045]; president[0.045]; second[0.045]; elections[0.045]; people[0.044]; incentives[0.044]; North[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = White House\u001b[39m ==> ENTITY: \u001b[32mWhite House\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.006:-0.006[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.046]; Washington[0.046]; John[0.044]; federal[0.044]; Clinton[0.044]; Clinton[0.044]; presidential[0.043]; mutual[0.043]; president[0.043]; election[0.041]; agenda[0.041]; claims[0.041]; Services[0.041]; North[0.041]; said[0.041]; said[0.041]; said[0.041]; people[0.040]; week[0.040]; income[0.040]; percent[0.040]; increase[0.040]; hire[0.040]; home[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mDow Jones\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mDow Jones & Company\u001b[39m <---> \u001b[32mDow Jones Industrial Average\u001b[39m\t\nSCORES: global= 0.269:0.263[\u001b[31m0.005\u001b[39m]; local()= 0.141:0.242[\u001b[32m0.101\u001b[39m]; log p(e|m)= 0.000:-0.423[\u001b[31m0.423\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstock[0.048]; index[0.047]; index[0.047]; index[0.047]; index[0.047]; market[0.046]; lower[0.046]; stocks[0.046]; stocks[0.046]; Stock[0.046]; higher[0.045]; composite[0.045]; American[0.045]; closed[0.045]; closed[0.045]; Nasdaq[0.045]; Exchange[0.044]; economic[0.044]; week[0.044]; week[0.044]; week[0.044]; week[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clinton ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.078:0.078[0]; log p(e|m)= -1.917:-1.917[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.047]; list[0.044]; presidential[0.044]; York[0.043]; people[0.043]; economic[0.042]; includes[0.042]; race[0.042]; Street[0.042]; Street[0.042]; Ross[0.041]; zone[0.041]; zone[0.041]; Wall[0.041]; Wall[0.041]; dollars[0.041]; legislation[0.041]; road[0.041]; Bob[0.041]; Federal[0.041]; political[0.041]; political[0.041]; candidates[0.040]; economy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Walter Mondale\u001b[39m ==> ENTITY: \u001b[32mWalter Mondale\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJohn[0.046]; George[0.044]; campaign[0.044]; presidential[0.043]; election[0.043]; candidate[0.043]; Reagan[0.042]; Clinton[0.042]; Democrat[0.042]; candidates[0.042]; votes[0.042]; won[0.041]; won[0.041]; won[0.041]; elections[0.041]; American[0.041]; president[0.041]; Ronald[0.040]; Johnson[0.040]; term[0.040]; term[0.040]; term[0.040]; Washington[0.040]; second[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clinton ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.078:0.078[0]; log p(e|m)= -1.917:-1.917[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nelection[0.046]; election[0.046]; people[0.044]; polls[0.044]; low[0.044]; presidency[0.044]; set[0.044]; Whitewater[0.044]; average[0.044]; candidates[0.044]; edge[0.044]; Clintons[0.043]; hold[0.043]; reflects[0.043]; like[0.043]; like[0.043]; said[0.043]; said[0.043]; said[0.043]; said[0.043]; probably[0.043]; holding[0.043]; high[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ronald Reagan\u001b[39m ==> ENTITY: \u001b[32mRonald Reagan\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.045]; John[0.045]; George[0.045]; decade[0.044]; presidential[0.044]; Washington[0.044]; Clinton[0.044]; term[0.044]; term[0.044]; Democrat[0.044]; Bush[0.043]; tax[0.043]; tax[0.043]; tax[0.043]; tax[0.043]; promise[0.043]; promise[0.043]; campaign[0.043]; candidates[0.043]; candidate[0.043]; budget[0.043]; taxes[0.043]; taxes[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clinton ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.072:0.072[0]; log p(e|m)= -1.917:-1.917[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinton[0.045]; tax[0.044]; tax[0.044]; rates[0.044]; rates[0.044]; taxes[0.044]; taxes[0.044]; taxes[0.044]; federal[0.043]; president[0.042]; income[0.042]; people[0.042]; include[0.041]; White[0.041]; Street[0.040]; House[0.040]; election[0.040]; increase[0.040]; lower[0.040]; drop[0.040]; board[0.040]; percent[0.040]; cost[0.040]; individuals[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Treasury ==> ENTITY: \u001b[32mUnited States Department of the Treasury\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.108:0.108[0]; log p(e|m)= -1.013:-1.013[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJohn[0.044]; rate[0.043]; Services[0.043]; George[0.043]; Ronald[0.042]; week[0.042]; tax[0.042]; tax[0.042]; tax[0.042]; term[0.042]; term[0.042]; term[0.042]; bond[0.042]; bond[0.042]; market[0.041]; consulting[0.041]; American[0.041]; corporate[0.041]; rates[0.041]; rates[0.041]; low[0.041]; higher[0.041]; Reagan[0.041]; surge[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bob Dole\u001b[39m ==> ENTITY: \u001b[32mBob Dole\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.107:0.107[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npresidential[0.047]; President[0.046]; Clinton[0.045]; candidates[0.045]; Ross[0.044]; Perot[0.044]; contest[0.043]; political[0.042]; political[0.042]; race[0.042]; economic[0.042]; promises[0.041]; financial[0.040]; financial[0.040]; raise[0.040]; believe[0.040]; Twilight[0.040]; list[0.040]; legislation[0.040]; season[0.040]; dollars[0.040]; economy[0.039]; includes[0.039]; socks[0.039]; \t\n\r [=================================>............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 5s832ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3601/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1014testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = European ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.097:0.097[0]; log p(e|m)= -0.256:-0.256[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMediterranean[0.046]; fleet[0.045]; fleet[0.045]; fleet[0.045]; week[0.044]; week[0.044]; week[0.044]; total[0.044]; total[0.044]; total[0.044]; remained[0.044]; number[0.044]; worldwide[0.044]; contract[0.043]; contract[0.043]; percent[0.043]; percent[0.043]; percent[0.043]; Gulf[0.042]; Gulf[0.042]; Gulf[0.042]; August[0.041]; unchanged[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mediterranean ==> ENTITY: \u001b[32mMediterranean Sea\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuropean[0.048]; fleet[0.047]; fleet[0.047]; fleet[0.047]; August[0.047]; number[0.046]; total[0.046]; total[0.046]; total[0.046]; Gulf[0.046]; Gulf[0.046]; Gulf[0.046]; worldwide[0.045]; York[0.045]; York[0.045]; Services[0.044]; week[0.044]; week[0.044]; week[0.044]; prior[0.044]; working[0.044]; working[0.044]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew York\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew York\u001b[39m <---> \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.243:0.242[\u001b[31m0.001\u001b[39m]; local()= 0.031:0.029[\u001b[31m0.002\u001b[39m]; log p(e|m)= -0.289:-1.790[\u001b[31m1.501\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfleet[0.044]; fleet[0.044]; fleet[0.044]; week[0.043]; week[0.043]; week[0.043]; August[0.042]; Offshore[0.042]; Mediterranean[0.042]; said[0.042]; prior[0.041]; based[0.041]; working[0.041]; working[0.041]; Gulf[0.041]; Gulf[0.041]; Gulf[0.041]; rigs[0.041]; rigs[0.041]; rigs[0.041]; rigs[0.041]; total[0.041]; total[0.041]; total[0.041]; \t\n\r [=================================>............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 5s846ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3604/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1101testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = American ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.009:0.009[0]; log p(e|m)= -0.130:-0.130[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRamon[0.057]; earned[0.054]; Johnson[0.054]; Johnson[0.054]; Tom[0.053]; Saturday[0.051]; points[0.050]; title[0.050]; Boxing[0.050]; Verdict[0.049]; Wins[0.049]; unanimous[0.049]; Dublin[0.049]; Venezuela[0.049]; decision[0.049]; Guzman[0.048]; featherweight[0.048]; defended[0.046]; successfully[0.046]; Unanimous[0.045]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mTom Johnson\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mTom Johnson (journalist)\u001b[39m <---> \u001b[32mTom Johnson (boxer)\u001b[39m\t\nSCORES: global= 0.245:0.236[\u001b[31m0.009\u001b[39m]; local()= 0.048:0.107[\u001b[32m0.059\u001b[39m]; log p(e|m)= -1.382:-2.397[\u001b[31m1.015\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJohnson[0.060]; unanimous[0.056]; American[0.055]; title[0.055]; Ramon[0.055]; earned[0.055]; featherweight[0.055]; decision[0.054]; successfully[0.054]; Saturday[0.053]; Guzman[0.052]; Wins[0.052]; defended[0.051]; Boxing[0.051]; points[0.049]; Venezuela[0.049]; Dublin[0.049]; Unanimous[0.048]; Verdict[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Venezuela ==> ENTITY: \u001b[32mVenezuela\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.033:0.033[0]; log p(e|m)= -0.212:-0.212[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRamon[0.055]; earned[0.054]; defended[0.054]; American[0.053]; Boxing[0.053]; title[0.052]; Guzman[0.052]; Johnson[0.052]; Johnson[0.052]; Unanimous[0.050]; Wins[0.049]; Saturday[0.049]; decision[0.049]; points[0.048]; unanimous[0.048]; featherweight[0.048]; Tom[0.047]; successfully[0.047]; Verdict[0.046]; Dublin[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dublin ==> ENTITY: \u001b[32mDublin\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.023:0.023[0]; log p(e|m)= -0.232:-0.232[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTom[0.060]; Saturday[0.057]; American[0.055]; Johnson[0.055]; Johnson[0.055]; earned[0.054]; points[0.054]; title[0.054]; decision[0.052]; defended[0.052]; Venezuela[0.051]; Ramon[0.051]; Verdict[0.050]; successfully[0.050]; Boxing[0.050]; Unanimous[0.050]; Wins[0.050]; unanimous[0.050]; featherweight[0.049]; \t\n\r [=================================>............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 5s904ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3609/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1018testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = African ==> ENTITY: \u001b[32mAfrica\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.911:-0.911[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrican[0.049]; immigrants[0.045]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; home[0.041]; apparently[0.041]; said[0.041]; fly[0.041]; ago[0.040]; Air[0.040]; Air[0.040]; Air[0.040]; Air[0.040]; man[0.040]; man[0.040]; flight[0.040]; Bernard[0.040]; week[0.040]; Friday[0.039]; raid[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Air France\u001b[39m ==> ENTITY: \u001b[32mAir France\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.233:0.233[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAir[0.046]; Air[0.046]; Air[0.046]; flight[0.045]; Paris[0.044]; Paris[0.044]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; released[0.041]; fly[0.041]; charters[0.040]; spokesman[0.040]; Friday[0.040]; scheduled[0.039]; said[0.039]; said[0.039]; brief[0.039]; week[0.039]; day[0.039]; workers[0.038]; apparently[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alain Juppe\u001b[39m ==> ENTITY: \u001b[32mAlain Juppé\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.044]; France[0.044]; Minister[0.044]; government[0.043]; government[0.043]; Prime[0.043]; Separately[0.043]; Francois[0.043]; Juppe[0.042]; Juppe[0.042]; said[0.042]; said[0.042]; parliament[0.042]; office[0.042]; chairman[0.041]; spokesman[0.041]; Paris[0.041]; military[0.041]; conservative[0.040]; confirmed[0.040]; personnel[0.040]; commitment[0.040]; air[0.040]; told[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Charles de Gaulle ==> ENTITY: \u001b[32mCharles de Gaulle Airport\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.224:0.224[0]; log p(e|m)= -1.537:-1.537[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nParis[0.046]; airport[0.045]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; flight[0.043]; Air[0.042]; Air[0.042]; Air[0.042]; Air[0.042]; Bernard[0.041]; military[0.041]; jets[0.040]; jets[0.040]; jets[0.040]; civilian[0.040]; civilian[0.040]; Socialist[0.040]; headquarters[0.040]; air[0.040]; flying[0.040]; Francois[0.040]; north[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = CFDT\u001b[39m ==> ENTITY: \u001b[32mFrench Democratic Confederation of Labour\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npro[0.047]; union[0.046]; Socialist[0.046]; unions[0.046]; unions[0.046]; France[0.046]; France[0.046]; France[0.046]; France[0.046]; Francois[0.044]; personnel[0.042]; air[0.042]; Paris[0.042]; Saint[0.042]; Minister[0.042]; staff[0.042]; Prime[0.042]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; civilian[0.041]; formal[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Air France\u001b[39m ==> ENTITY: \u001b[32mAir France\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.275:0.275[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAir[0.046]; Air[0.046]; Air[0.046]; Air[0.046]; flight[0.046]; Paris[0.045]; Paris[0.045]; Paris[0.045]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; headquarters[0.042]; jets[0.042]; airport[0.041]; released[0.041]; fly[0.041]; charters[0.041]; spokesman[0.040]; Friday[0.040]; scheduled[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = African ==> ENTITY: \u001b[32mAfrica\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.135:0.135[0]; log p(e|m)= -0.911:-0.911[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrican[0.049]; African[0.049]; immigrants[0.046]; people[0.044]; said[0.042]; home[0.041]; apparently[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; France[0.041]; fly[0.041]; ago[0.040]; church[0.040]; church[0.040]; church[0.040]; man[0.040]; man[0.040]; flight[0.040]; Bernard[0.040]; week[0.040]; day[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Air France\u001b[39m ==> ENTITY: \u001b[32mAir France\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAir[0.046]; Air[0.046]; plane[0.045]; Paris[0.044]; France[0.044]; France[0.044]; headquarters[0.042]; flying[0.041]; jets[0.041]; jets[0.041]; jets[0.041]; airport[0.041]; Separately[0.041]; released[0.041]; charters[0.040]; charters[0.040]; air[0.040]; spokesman[0.040]; spokesman[0.040]; personnel[0.040]; Alain[0.040]; transform[0.040]; office[0.040]; companies[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = CGT ==> ENTITY: \u001b[32mGeneral Confederation of Labour (France)\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.190:0.190[0]; log p(e|m)= -0.770:-0.770[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nunion[0.046]; unions[0.045]; unions[0.045]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; Socialist[0.043]; Paris[0.042]; obtain[0.042]; personnel[0.042]; scheduled[0.041]; Francois[0.041]; pro[0.041]; people[0.040]; people[0.040]; companies[0.040]; Alain[0.040]; week[0.040]; Bernard[0.040]; released[0.040]; failed[0.040]; formal[0.040]; use[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Conakry\u001b[39m ==> ENTITY: \u001b[32mConakry\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.178:0.178[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGuinean[0.044]; north[0.043]; civilian[0.043]; headquarters[0.043]; airport[0.042]; church[0.042]; church[0.042]; church[0.042]; scheduled[0.042]; Saint[0.042]; France[0.042]; France[0.042]; France[0.042]; France[0.042]; France[0.042]; France[0.042]; France[0.042]; Friday[0.041]; said[0.041]; said[0.041]; Paris[0.041]; Paris[0.041]; Paris[0.041]; week[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.210:0.210[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.049]; France[0.049]; France[0.049]; France[0.049]; France[0.049]; Socialist[0.042]; Paris[0.042]; Paris[0.042]; unions[0.040]; week[0.040]; African[0.040]; African[0.040]; union[0.040]; denounced[0.040]; pro[0.039]; protest[0.039]; Bernard[0.039]; Air[0.039]; Air[0.039]; Air[0.039]; Air[0.039]; Saint[0.038]; failed[0.038]; immigrants[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Paris ==> ENTITY: \u001b[32mParis\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.187:0.187[0]; log p(e|m)= -0.115:-0.115[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nParis[0.047]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; flight[0.044]; Air[0.043]; Air[0.043]; Air[0.043]; Air[0.043]; apparently[0.042]; fly[0.040]; Saint[0.040]; home[0.040]; scheduled[0.040]; man[0.039]; man[0.039]; week[0.039]; workers[0.039]; Friday[0.038]; Bernard[0.038]; people[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Air France\u001b[39m ==> ENTITY: \u001b[32mAir France\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.223:0.223[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAir[0.045]; Air[0.045]; Air[0.045]; flight[0.045]; plane[0.044]; Paris[0.044]; France[0.043]; France[0.043]; France[0.043]; headquarters[0.042]; flying[0.041]; jets[0.041]; jets[0.041]; airport[0.041]; released[0.040]; charters[0.040]; charters[0.040]; air[0.040]; spokesman[0.039]; spokesman[0.039]; personnel[0.039]; Alain[0.039]; transform[0.039]; scheduled[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Air France\u001b[39m ==> ENTITY: \u001b[32mAir France\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.285:0.285[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAir[0.044]; Air[0.044]; Air[0.044]; Air[0.044]; Air[0.044]; flight[0.044]; Paris[0.042]; Paris[0.042]; Paris[0.042]; France[0.042]; France[0.042]; France[0.042]; France[0.042]; France[0.042]; France[0.042]; France[0.042]; headquarters[0.040]; jets[0.040]; airport[0.039]; released[0.039]; fly[0.039]; charters[0.039]; spokesman[0.038]; Friday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Paris ==> ENTITY: \u001b[32mParis\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.188:0.188[0]; log p(e|m)= -0.115:-0.115[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nParis[0.047]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; flight[0.044]; Air[0.043]; Air[0.043]; Air[0.043]; Air[0.043]; apparently[0.042]; fly[0.040]; Saint[0.040]; home[0.040]; scheduled[0.039]; man[0.039]; man[0.039]; week[0.039]; residence[0.039]; workers[0.038]; Friday[0.038]; Bernard[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = CFDT\u001b[39m ==> ENTITY: \u001b[32mFrench Democratic Confederation of Labour\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.135:0.135[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nunions[0.046]; unions[0.046]; France[0.046]; France[0.046]; conservative[0.044]; government[0.044]; government[0.044]; Francois[0.044]; Juppe[0.043]; Juppe[0.043]; Juppe[0.043]; military[0.043]; Alain[0.043]; notably[0.043]; personnel[0.042]; air[0.042]; Paris[0.042]; Minister[0.042]; staff[0.042]; Prime[0.042]; said[0.042]; said[0.042]; said[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Paris ==> ENTITY: \u001b[32mParis\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.115:-0.115[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nairport[0.046]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; flight[0.044]; Air[0.043]; Air[0.043]; Air[0.043]; Air[0.043]; apparently[0.042]; Alain[0.042]; headquarters[0.041]; Charles[0.041]; air[0.040]; Francois[0.040]; released[0.040]; plane[0.040]; home[0.040]; scheduled[0.040]; office[0.039]; flying[0.039]; Blanc[0.039]; north[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Air France\u001b[39m ==> ENTITY: \u001b[32mAir France\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.228:0.228[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAir[0.046]; Air[0.046]; Air[0.046]; flight[0.046]; Paris[0.044]; Paris[0.044]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; France[0.044]; fly[0.041]; charters[0.040]; spokesman[0.040]; Friday[0.040]; scheduled[0.040]; said[0.039]; week[0.039]; workers[0.038]; apparently[0.038]; Saint[0.038]; Bernard[0.038]; expelled[0.038]; unions[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = CFDT\u001b[39m ==> ENTITY: \u001b[32mFrench Democratic Confederation of Labour\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npro[0.045]; union[0.043]; Socialist[0.043]; unions[0.043]; unions[0.043]; unions[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; Paris[0.040]; Paris[0.040]; Paris[0.040]; Saint[0.039]; staff[0.039]; expelled[0.039]; said[0.039]; said[0.039]; civilian[0.039]; formal[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.208:0.208[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.049]; France[0.049]; France[0.049]; France[0.049]; France[0.049]; Socialist[0.042]; Paris[0.042]; Paris[0.042]; unions[0.040]; week[0.040]; African[0.040]; African[0.040]; African[0.040]; union[0.040]; denounced[0.039]; pro[0.039]; protest[0.039]; Bernard[0.039]; Air[0.039]; Air[0.039]; Air[0.039]; Air[0.039]; said[0.038]; Saint[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Christian Blanc\u001b[39m ==> ENTITY: \u001b[32mChristian Blanc\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrancois[0.046]; Alain[0.044]; France[0.044]; France[0.044]; office[0.043]; Paris[0.043]; pro[0.042]; government[0.042]; government[0.042]; conservative[0.042]; confirmed[0.041]; military[0.041]; formal[0.041]; parliament[0.041]; Separately[0.041]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; personnel[0.040]; air[0.040]; north[0.040]; Socialist[0.040]; headquarters[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = African ==> ENTITY: \u001b[32mAfrica\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.911:-0.911[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrican[0.048]; immigrants[0.044]; immigrants[0.044]; personnel[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; France[0.043]; Charles[0.042]; chairman[0.042]; people[0.041]; people[0.041]; formal[0.040]; said[0.040]; said[0.040]; north[0.040]; civilian[0.040]; home[0.040]; headquarters[0.040]; scheduled[0.039]; Christian[0.039]; \t\n\r [=================================>............] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 5s793ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3631/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1020testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dutch ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.154:0.154[0]; log p(e|m)= -0.794:-0.794[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.051]; Dutch[0.049]; Dutch[0.049]; Dutch[0.049]; German[0.044]; German[0.044]; northern[0.042]; Groningen[0.041]; Groningen[0.041]; Groningen[0.041]; Friday[0.041]; service[0.041]; Nicole[0.040]; Nicole[0.040]; public[0.040]; city[0.040]; sexual[0.039]; said[0.039]; said[0.039]; companion[0.039]; companion[0.039]; showed[0.039]; old[0.038]; know[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Groningen ==> ENTITY: \u001b[32mGroningen\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.207:0.207[0]; log p(e|m)= -0.533:-0.533[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.049]; Groningen[0.048]; Groningen[0.048]; Dutch[0.046]; Dutch[0.046]; Dutch[0.046]; Dutch[0.046]; northern[0.046]; city[0.043]; service[0.042]; near[0.042]; list[0.042]; German[0.042]; German[0.042]; German[0.042]; thanks[0.042]; Friday[0.042]; Monday[0.041]; early[0.041]; Public[0.041]; Berlin[0.041]; school[0.041]; public[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.124:0.124[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBerlin[0.046]; German[0.045]; early[0.043]; probably[0.043]; possible[0.043]; added[0.043]; previous[0.043]; Monday[0.042]; studied[0.042]; Velten[0.042]; Dutch[0.041]; Dutch[0.041]; involved[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; signs[0.041]; led[0.040]; names[0.040]; requested[0.040]; child[0.040]; child[0.040]; showed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dutch ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.169:0.169[0]; log p(e|m)= -0.794:-0.794[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.050]; Dutch[0.048]; Dutch[0.048]; Dutch[0.048]; German[0.043]; German[0.043]; German[0.043]; couple[0.043]; northern[0.041]; early[0.041]; Groningen[0.040]; Groningen[0.040]; Groningen[0.040]; Monday[0.040]; Friday[0.040]; service[0.040]; Nicole[0.039]; Nicole[0.039]; children[0.039]; public[0.039]; Berlin[0.039]; city[0.039]; sexual[0.039]; list[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.141:0.141[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGerman[0.050]; Berlin[0.047]; northern[0.046]; Dutch[0.045]; Dutch[0.045]; Dutch[0.045]; Dutch[0.045]; Netherlands[0.044]; girl[0.043]; girl[0.043]; girl[0.043]; aunt[0.043]; aunt[0.043]; old[0.042]; early[0.042]; said[0.042]; said[0.042]; said[0.042]; city[0.042]; showed[0.042]; service[0.042]; near[0.041]; missing[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dutch ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.794:-0.794[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDutch[0.050]; Germany[0.047]; German[0.045]; probably[0.043]; car[0.042]; couple[0.042]; couple[0.042]; couple[0.042]; early[0.042]; Monday[0.042]; studied[0.041]; Berlin[0.041]; extradited[0.040]; released[0.040]; sexual[0.040]; list[0.040]; list[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; companion[0.040]; child[0.040]; child[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dutch ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.172:0.172[0]; log p(e|m)= -0.794:-0.794[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.050]; Dutch[0.048]; Dutch[0.048]; Dutch[0.048]; German[0.043]; German[0.043]; German[0.043]; couple[0.043]; northern[0.041]; early[0.041]; Groningen[0.040]; Groningen[0.040]; Groningen[0.040]; Monday[0.040]; Friday[0.040]; service[0.040]; studied[0.040]; Nicole[0.039]; Nicole[0.039]; children[0.039]; public[0.039]; Berlin[0.039]; city[0.039]; sexual[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.112:0.112[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGerman[0.049]; northern[0.045]; Dutch[0.044]; Dutch[0.044]; Dutch[0.044]; Dutch[0.044]; Netherlands[0.043]; girl[0.042]; girl[0.042]; aunt[0.041]; aunt[0.041]; old[0.041]; said[0.041]; said[0.041]; city[0.041]; showed[0.041]; service[0.041]; missing[0.040]; Nicole[0.040]; Nicole[0.040]; Friday[0.040]; public[0.039]; alive[0.039]; alive[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Netherlands ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.159:0.159[0]; log p(e|m)= -0.250:-0.250[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDutch[0.048]; Dutch[0.048]; Dutch[0.048]; Dutch[0.048]; public[0.044]; Groningen[0.043]; Groningen[0.043]; Groningen[0.043]; German[0.043]; German[0.043]; northern[0.041]; service[0.041]; early[0.041]; city[0.040]; sexual[0.040]; Friday[0.040]; said[0.039]; said[0.039]; thanks[0.039]; showed[0.038]; missing[0.038]; Missing[0.038]; abuse[0.037]; know[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dutch ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.152:0.152[0]; log p(e|m)= -0.794:-0.794[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDutch[0.052]; Germany[0.049]; German[0.047]; couple[0.046]; couple[0.046]; couple[0.046]; probably[0.044]; studied[0.043]; children[0.042]; Berlin[0.042]; extradited[0.042]; released[0.042]; Monday[0.042]; said[0.042]; said[0.042]; said[0.042]; car[0.042]; child[0.042]; child[0.042]; media[0.042]; led[0.041]; names[0.041]; added[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Velten\u001b[39m ==> ENTITY: \u001b[32mVelten\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnorthern[0.050]; Germany[0.047]; Groningen[0.046]; Groningen[0.046]; near[0.045]; school[0.045]; Monday[0.045]; Berlin[0.045]; city[0.044]; probably[0.043]; running[0.043]; old[0.043]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; couple[0.042]; couple[0.042]; couple[0.042]; prosecutors[0.042]; prosecutors[0.042]; early[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Groningen ==> ENTITY: \u001b[32mGroningen\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.212:0.212[0]; log p(e|m)= -0.533:-0.533[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.047]; Groningen[0.046]; Groningen[0.046]; Dutch[0.045]; Dutch[0.045]; Dutch[0.045]; Dutch[0.045]; northern[0.044]; city[0.041]; service[0.041]; near[0.041]; German[0.040]; German[0.040]; German[0.040]; thanks[0.040]; Friday[0.040]; Monday[0.040]; early[0.040]; Berlin[0.039]; school[0.039]; public[0.039]; said[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Groningen ==> ENTITY: \u001b[32mGroningen\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.184:0.184[0]; log p(e|m)= -0.533:-0.533[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGroningen[0.049]; Groningen[0.049]; city[0.047]; Netherlands[0.047]; Dutch[0.045]; Dutch[0.045]; Dutch[0.045]; Dutch[0.045]; northern[0.044]; service[0.041]; Friday[0.040]; early[0.040]; public[0.039]; said[0.039]; said[0.039]; German[0.039]; German[0.039]; police[0.039]; campsite[0.038]; campsite[0.038]; campsite[0.038]; showed[0.038]; missing[0.038]; signs[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Berlin ==> ENTITY: \u001b[32mBerlin\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.154:-0.154[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstudied[0.054]; Germany[0.053]; German[0.051]; German[0.051]; city[0.049]; Monday[0.048]; near[0.048]; northern[0.048]; released[0.047]; running[0.047]; early[0.047]; service[0.047]; Velten[0.046]; Public[0.046]; suspected[0.046]; couple[0.046]; couple[0.046]; couple[0.046]; public[0.046]; old[0.045]; probably[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dutch ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.173:0.173[0]; log p(e|m)= -0.794:-0.794[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNetherlands[0.051]; Dutch[0.049]; Dutch[0.049]; Dutch[0.049]; German[0.044]; German[0.044]; northern[0.042]; early[0.041]; Groningen[0.041]; Groningen[0.041]; Groningen[0.041]; service[0.041]; public[0.040]; city[0.040]; sexual[0.039]; said[0.039]; said[0.039]; said[0.039]; old[0.039]; showed[0.038]; Friday[0.038]; know[0.038]; aunt[0.038]; aunt[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.150:0.150[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.050]; Berlin[0.045]; children[0.043]; northern[0.042]; studied[0.042]; Dutch[0.042]; Dutch[0.042]; Dutch[0.042]; Dutch[0.042]; child[0.041]; child[0.041]; early[0.041]; probably[0.041]; girl[0.041]; aunt[0.041]; aunt[0.041]; aunt[0.041]; old[0.041]; couple[0.040]; couple[0.040]; couple[0.040]; city[0.040]; service[0.040]; near[0.040]; \t\n\r [==================================>...........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 5s734ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3647/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 970testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nice ==> ENTITY: \u001b[32mOGC Nice\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.233:0.233[0]; log p(e|m)= -2.137:-2.137[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; league[0.046]; club[0.044]; Paris[0.044]; France[0.044]; Guingamp[0.043]; matches[0.043]; played[0.043]; played[0.043]; season[0.042]; French[0.042]; Albert[0.041]; win[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Soccer[0.040]; division[0.039]; international[0.039]; Thursday[0.038]; defeat[0.038]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baratelli\u001b[39m ==> ENTITY: \u001b[32mDominique Baratelli\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; league[0.046]; matches[0.045]; played[0.043]; played[0.043]; Soccer[0.043]; France[0.042]; Coach[0.042]; Guingamp[0.041]; season[0.041]; coach[0.041]; club[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Germain[0.041]; French[0.040]; Paris[0.040]; international[0.039]; goalkeeper[0.039]; division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nice ==> ENTITY: \u001b[32mOGC Nice\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.233:0.233[0]; log p(e|m)= -2.137:-2.137[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; league[0.046]; club[0.044]; Paris[0.044]; France[0.044]; Guingamp[0.043]; matches[0.043]; played[0.043]; played[0.043]; season[0.042]; French[0.042]; Albert[0.041]; win[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Soccer[0.040]; division[0.039]; international[0.039]; Thursday[0.038]; defeat[0.038]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nice ==> ENTITY: \u001b[32mOGC Nice\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.233:0.233[0]; log p(e|m)= -2.137:-2.137[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; league[0.046]; club[0.044]; Paris[0.044]; France[0.044]; Guingamp[0.043]; matches[0.043]; played[0.043]; played[0.043]; season[0.042]; French[0.042]; Albert[0.041]; win[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Soccer[0.040]; division[0.039]; international[0.039]; Thursday[0.038]; defeat[0.038]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.154:0.154[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.050]; matches[0.045]; Soccer[0.044]; international[0.043]; league[0.043]; league[0.043]; Paris[0.043]; Guingamp[0.043]; club[0.042]; played[0.042]; played[0.042]; Albert[0.040]; Dominique[0.040]; season[0.040]; Nice[0.040]; Nice[0.040]; Nice[0.040]; Nice[0.040]; Nice[0.040]; Nice[0.040]; win[0.040]; home[0.039]; defeat[0.039]; Thursday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paris St Germain\u001b[39m ==> ENTITY: \u001b[32mParis Saint-Germain F.C.\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.228:0.228[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.046]; Guingamp[0.045]; league[0.045]; league[0.045]; matches[0.044]; Soccer[0.043]; season[0.043]; France[0.042]; win[0.042]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; defeat[0.041]; Coach[0.041]; division[0.041]; French[0.040]; played[0.040]; played[0.040]; international[0.039]; struggling[0.039]; Friday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baratelli\u001b[39m ==> ENTITY: \u001b[32mDominique Baratelli\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; league[0.046]; matches[0.045]; played[0.043]; played[0.043]; Soccer[0.043]; France[0.042]; Coach[0.042]; Guingamp[0.041]; season[0.041]; coach[0.041]; club[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Germain[0.041]; French[0.040]; Paris[0.040]; international[0.039]; goalkeeper[0.039]; division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.049]; league[0.049]; club[0.046]; Soccer[0.045]; Guingamp[0.044]; matches[0.044]; season[0.044]; French[0.043]; international[0.042]; Paris[0.042]; played[0.042]; played[0.042]; division[0.040]; coach[0.039]; win[0.039]; Nice[0.039]; Nice[0.039]; Nice[0.039]; Nice[0.039]; Nice[0.039]; Nice[0.039]; Germain[0.039]; defeat[0.038]; Thursday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Albert Emon\u001b[39m ==> ENTITY: \u001b[32mAlbert Emon\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.045]; league[0.045]; France[0.044]; played[0.043]; played[0.043]; coach[0.043]; Coach[0.043]; Guingamp[0.043]; Soccer[0.043]; club[0.043]; matches[0.042]; season[0.042]; Dominique[0.041]; Paris[0.041]; international[0.041]; French[0.041]; win[0.041]; Nice[0.040]; Nice[0.040]; Nice[0.040]; Nice[0.040]; Nice[0.040]; Nice[0.040]; division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nice ==> ENTITY: \u001b[32mOGC Nice\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.233:0.233[0]; log p(e|m)= -2.137:-2.137[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; league[0.046]; club[0.044]; Paris[0.044]; France[0.044]; Guingamp[0.043]; matches[0.043]; played[0.043]; played[0.043]; season[0.042]; French[0.042]; Albert[0.041]; win[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Soccer[0.040]; division[0.039]; international[0.039]; Thursday[0.038]; defeat[0.038]; said[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNice\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mOGC Nice\u001b[39m <---> \u001b[32mNice\u001b[39m\t\nSCORES: global= 0.279:0.251[\u001b[31m0.028\u001b[39m]; local()= 0.233:0.180[\u001b[31m0.052\u001b[39m]; log p(e|m)= -2.137:-0.260[\u001b[32m1.877\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; league[0.046]; club[0.044]; Paris[0.044]; France[0.044]; Guingamp[0.043]; matches[0.043]; played[0.043]; played[0.043]; season[0.042]; French[0.042]; Albert[0.041]; win[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Soccer[0.040]; division[0.039]; international[0.039]; Thursday[0.038]; defeat[0.038]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nice ==> ENTITY: \u001b[32mOGC Nice\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.233:0.233[0]; log p(e|m)= -2.137:-2.137[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; league[0.046]; club[0.044]; Paris[0.044]; France[0.044]; Guingamp[0.043]; matches[0.043]; played[0.043]; played[0.043]; season[0.042]; French[0.042]; Albert[0.041]; win[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Soccer[0.040]; division[0.039]; international[0.039]; Thursday[0.038]; defeat[0.038]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dominique Baratelli\u001b[39m ==> ENTITY: \u001b[32mDominique Baratelli\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.046]; league[0.046]; matches[0.045]; played[0.043]; played[0.043]; Soccer[0.043]; France[0.042]; Coach[0.042]; Guingamp[0.041]; season[0.041]; coach[0.041]; club[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Germain[0.041]; French[0.040]; Paris[0.040]; international[0.039]; goalkeeper[0.039]; division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Guingamp\u001b[39m ==> ENTITY: \u001b[32mEn Avant de Guingamp\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nleague[0.045]; league[0.045]; club[0.044]; France[0.044]; played[0.043]; played[0.043]; matches[0.043]; season[0.042]; Soccer[0.042]; division[0.042]; Coach[0.041]; French[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Nice[0.041]; Paris[0.041]; Dominique[0.040]; coach[0.040]; win[0.040]; international[0.039]; home[0.039]; \t\n\r [==================================>...........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 5s655ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3661/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 954testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Roberto Duran\u001b[39m ==> ENTITY: \u001b[32mRoberto Durán\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.230:0.230[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nboxing[0.048]; bout[0.046]; middleweight[0.046]; Duran[0.045]; Duran[0.045]; Duran[0.045]; Duran[0.045]; Roberto[0.045]; Boxing[0.044]; career[0.043]; fight[0.043]; Cruz[0.043]; title[0.043]; title[0.043]; Mexican[0.043]; fifth[0.042]; Panamanian[0.042]; age[0.042]; Panama[0.042]; Panama[0.042]; Panama[0.042]; ring[0.042]; prime[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Panamanian\u001b[39m ==> ENTITY: \u001b[32mPanama\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.091:0.091[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPanamanian[0.045]; match[0.044]; won[0.044]; fifth[0.044]; points[0.043]; win[0.043]; Ramon[0.042]; Camacho[0.042]; Camacho[0.042]; Hector[0.042]; took[0.042]; championships[0.041]; Puerto[0.041]; record[0.041]; close[0.041]; close[0.041]; earn[0.041]; title[0.041]; title[0.041]; Rico[0.040]; career[0.040]; career[0.040]; position[0.040]; time[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Puerto Rico ==> ENTITY: \u001b[32mPuerto Rico\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.114:0.114[0]; log p(e|m)= -0.297:-0.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRamon[0.043]; Camacho[0.043]; Camacho[0.043]; Panamanian[0.043]; Panamanian[0.043]; championships[0.042]; President[0.042]; earn[0.042]; boxers[0.042]; Hector[0.042]; round[0.042]; bout[0.042]; won[0.042]; Duran[0.041]; Duran[0.041]; record[0.041]; June[0.041]; boxer[0.041]; points[0.041]; took[0.041]; win[0.041]; time[0.040]; Saturday[0.040]; title[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Panama City\u001b[39m ==> ENTITY: \u001b[32mPanama City\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.119:0.119[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPanama[0.045]; Panama[0.045]; boxing[0.043]; Roberto[0.043]; Roberto[0.043]; Cruz[0.043]; Sands[0.043]; City[0.043]; Mexican[0.042]; Panamanian[0.042]; championships[0.041]; match[0.041]; Duran[0.041]; Duran[0.041]; Duran[0.041]; Duran[0.041]; Duran[0.041]; Duran[0.041]; Saturday[0.040]; Boxing[0.040]; prime[0.040]; close[0.040]; close[0.040]; different[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCamacho\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mJosé Antonio Camacho\u001b[39m <---> \u001b[32mHéctor Camacho\u001b[39m\t\nSCORES: global= 0.243:0.242[\u001b[31m0.001\u001b[39m]; local()= 0.107:0.208[\u001b[32m0.101\u001b[39m]; log p(e|m)= -1.091:-3.101[\u001b[31m2.010\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nboxer[0.046]; boxers[0.044]; Camacho[0.044]; bout[0.043]; Hector[0.043]; Ramon[0.043]; Boxing[0.042]; fights[0.042]; middleweight[0.042]; middleweight[0.042]; welterweight[0.041]; career[0.041]; career[0.041]; rematch[0.041]; Duran[0.041]; fight[0.041]; Rico[0.041]; Puerto[0.041]; record[0.041]; Panamanian[0.041]; Panamanian[0.041]; Manzanares[0.040]; championships[0.039]; decision[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Panama ==> ENTITY: \u001b[32mPanama\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.113:0.113[0]; log p(e|m)= -0.326:-0.326[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nboxing[0.052]; Panama[0.051]; Panama[0.051]; Roberto[0.049]; Roberto[0.049]; middleweight[0.049]; Panamanian[0.048]; Mexican[0.048]; bout[0.048]; match[0.048]; career[0.048]; Boxing[0.047]; Cruz[0.047]; legend[0.046]; City[0.046]; City[0.046]; title[0.046]; title[0.046]; fifth[0.046]; fight[0.045]; home[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Panamanian\u001b[39m ==> ENTITY: \u001b[32mPanama\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPanama[0.054]; Panama[0.054]; Panama[0.054]; Roberto[0.053]; Roberto[0.053]; match[0.052]; fifth[0.052]; Mexican[0.052]; Cruz[0.050]; close[0.048]; title[0.048]; title[0.048]; legend[0.048]; home[0.048]; career[0.048]; boxing[0.048]; City[0.048]; City[0.048]; Ariel[0.047]; long[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Panamanian\u001b[39m ==> ENTITY: \u001b[32mPanama\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.097:0.097[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPanamanian[0.046]; points[0.044]; win[0.043]; Ramon[0.043]; Camacho[0.043]; Camacho[0.043]; Hector[0.042]; took[0.042]; championships[0.042]; Puerto[0.042]; record[0.042]; close[0.041]; earn[0.041]; title[0.041]; Rico[0.041]; career[0.041]; career[0.041]; position[0.041]; time[0.041]; City[0.041]; different[0.040]; Duran[0.040]; round[0.040]; June[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.056:0.056[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \njunior[0.045]; told[0.045]; Stone[0.044]; record[0.044]; prime[0.043]; said[0.042]; world[0.042]; spans[0.042]; soil[0.042]; classes[0.041]; Legend[0.041]; legend[0.041]; career[0.041]; career[0.041]; career[0.041]; fight[0.041]; defying[0.041]; time[0.041]; home[0.041]; position[0.041]; frontier[0.041]; City[0.040]; City[0.040]; decades[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Panama City\u001b[39m ==> ENTITY: \u001b[32mPanama City\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.113:0.113[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPanama[0.045]; Panama[0.045]; boxing[0.044]; Roberto[0.043]; Roberto[0.043]; Cruz[0.043]; Sands[0.043]; City[0.043]; Mexican[0.042]; Panamanian[0.042]; match[0.042]; Duran[0.041]; Duran[0.041]; Duran[0.041]; Duran[0.041]; Duran[0.041]; Duran[0.041]; Saturday[0.040]; Boxing[0.040]; prime[0.040]; Time[0.040]; billed[0.040]; long[0.039]; career[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlantic City ==> ENTITY: \u001b[32mAtlantic City, New Jersey\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.078:-0.078[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nround[0.044]; time[0.044]; fight[0.044]; fights[0.043]; win[0.043]; bout[0.043]; career[0.042]; career[0.042]; took[0.042]; loses[0.042]; controversial[0.042]; close[0.042]; record[0.041]; Saturday[0.041]; title[0.041]; decision[0.041]; Boxing[0.041]; rematch[0.041]; Association[0.041]; June[0.041]; spans[0.041]; hopes[0.040]; points[0.040]; frontier[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexican ==> ENTITY: \u001b[32mMexico\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.397:-0.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRoberto[0.044]; Roberto[0.044]; match[0.044]; Panama[0.043]; Panama[0.043]; Panama[0.043]; boxer[0.043]; boxing[0.042]; career[0.042]; career[0.042]; Cruz[0.041]; different[0.041]; City[0.041]; City[0.041]; won[0.041]; Panamanian[0.041]; fifth[0.040]; Duran[0.040]; Duran[0.040]; Duran[0.040]; Duran[0.040]; Duran[0.040]; Duran[0.040]; prime[0.040]; \t\n\r [==================================>...........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 5s609ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3673/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1003testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bogota\u001b[39m ==> ENTITY: \u001b[32mBogotá\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nColombia[0.051]; Colombia[0.051]; Bogota[0.048]; Colombian[0.045]; located[0.044]; area[0.043]; Narino[0.043]; Casa[0.041]; Friday[0.041]; historic[0.041]; said[0.040]; presidential[0.040]; presidential[0.040]; Municipal[0.040]; office[0.040]; landmines[0.040]; told[0.039]; downtown[0.039]; added[0.039]; near[0.039]; real[0.039]; yards[0.039]; meters[0.038]; presidency[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colombia ==> ENTITY: \u001b[32mColombia\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.223:-0.223[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nColombia[0.052]; Colombian[0.048]; Narino[0.047]; Bogota[0.046]; Bogota[0.046]; area[0.044]; area[0.044]; backed[0.041]; downtown[0.040]; historic[0.040]; presidency[0.040]; Casa[0.040]; police[0.040]; police[0.040]; police[0.040]; police[0.040]; near[0.040]; Municipal[0.039]; landmines[0.039]; adding[0.039]; said[0.039]; said[0.039]; said[0.039]; told[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bogota\u001b[39m ==> ENTITY: \u001b[32mBogotá\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBogota[0.047]; Colombia[0.046]; Colombia[0.046]; historic[0.045]; located[0.043]; women[0.043]; area[0.043]; area[0.043]; downtown[0.043]; Narino[0.042]; Colombian[0.042]; near[0.042]; Casa[0.041]; Friday[0.041]; say[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; presidential[0.039]; presidential[0.039]; Municipal[0.039]; landmines[0.039]; landmines[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colombia ==> ENTITY: \u001b[32mColombia\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.223:-0.223[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nColombia[0.053]; Colombian[0.049]; Narino[0.048]; Bogota[0.047]; Bogota[0.047]; area[0.044]; downtown[0.041]; historic[0.040]; presidency[0.040]; Casa[0.040]; landmines[0.040]; police[0.040]; police[0.040]; near[0.040]; added[0.040]; Municipal[0.040]; said[0.039]; said[0.039]; adding[0.039]; told[0.039]; palace[0.039]; palace[0.039]; mines[0.039]; telephone[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Casa de Narino\u001b[39m ==> ENTITY: \u001b[32mCasa de Nariño\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.092:0.092[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \noffice[0.048]; Bogota[0.048]; Bogota[0.048]; palace[0.047]; palace[0.047]; Colombia[0.047]; Colombia[0.047]; ornate[0.047]; yards[0.046]; Colombian[0.046]; presidency[0.045]; located[0.045]; claymore[0.045]; claymore[0.045]; presidential[0.044]; presidential[0.044]; historic[0.044]; near[0.044]; meters[0.043]; downtown[0.043]; plastic[0.043]; plastic[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.103:0.103[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntold[0.045]; real[0.043]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; mines[0.042]; mines[0.042]; mines[0.042]; added[0.042]; experts[0.042]; backed[0.041]; spokesman[0.041]; spokesman[0.041]; search[0.041]; office[0.041]; Friday[0.041]; Bogota[0.041]; Bogota[0.041]; possible[0.041]; possible[0.041]; area[0.041]; area[0.041]; historic[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colombian ==> ENTITY: \u001b[32mColombia\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.113:0.113[0]; log p(e|m)= -0.320:-0.320[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nColombia[0.052]; Colombia[0.052]; Bogota[0.045]; Bogota[0.045]; Narino[0.043]; landmines[0.042]; presidency[0.041]; Casa[0.041]; area[0.041]; weapons[0.041]; police[0.041]; police[0.041]; office[0.040]; cause[0.040]; real[0.040]; palace[0.040]; palace[0.040]; Police[0.040]; historic[0.040]; spokesman[0.040]; located[0.039]; said[0.039]; powerful[0.039]; powerful[0.039]; \t\n\r [==================================>...........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 5s575ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3680/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 959testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oakland ==> ENTITY: \u001b[32mOakland Athletics\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.210:0.210[0]; log p(e|m)= -3.912:-3.912[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPittsburgh[0.044]; Oakland[0.044]; League[0.043]; Cleveland[0.043]; Cleveland[0.043]; Houston[0.042]; California[0.042]; California[0.042]; Cincinnati[0.042]; Cincinnati[0.042]; Philadelphia[0.042]; Baltimore[0.041]; Boston[0.041]; Boston[0.041]; Detroit[0.041]; Detroit[0.041]; Central[0.041]; Central[0.041]; Toronto[0.040]; Toronto[0.040]; Texas[0.040]; Texas[0.040]; Florida[0.040]; Florida[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Baltimore ==> ENTITY: \u001b[32mBaltimore Orioles\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.218:0.218[0]; log p(e|m)= -4.075:-4.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.044]; Cleveland[0.044]; Cleveland[0.044]; Houston[0.043]; Cincinnati[0.043]; Cincinnati[0.043]; Pittsburgh[0.043]; Oakland[0.042]; Oakland[0.042]; Boston[0.042]; Boston[0.042]; League[0.042]; National[0.041]; Detroit[0.041]; Detroit[0.041]; Toronto[0.041]; Toronto[0.041]; City[0.041]; City[0.041]; Montreal[0.040]; Florida[0.040]; Florida[0.040]; Seattle[0.040]; Seattle[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMilwaukee\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAtlanta Braves\u001b[39m <---> \u001b[32mMilwaukee Brewers\u001b[39m\t\nSCORES: global= 0.253:0.251[\u001b[31m0.002\u001b[39m]; local()= 0.243:0.213[\u001b[31m0.030\u001b[39m]; log p(e|m)= -3.912:-3.411[\u001b[32m0.501\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.044]; games[0.043]; games[0.043]; Cleveland[0.043]; Cleveland[0.043]; Baseball[0.043]; Baseball[0.043]; League[0.042]; League[0.042]; League[0.042]; League[0.042]; Boston[0.042]; Boston[0.042]; Baltimore[0.041]; Baltimore[0.041]; Houston[0.041]; Detroit[0.041]; Detroit[0.041]; Oakland[0.041]; Oakland[0.041]; Atlanta[0.040]; Central[0.040]; Central[0.040]; played[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Western Division ==> ENTITY: \u001b[32mNational League West\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.177:0.177[0]; log p(e|m)= -1.295:-1.295[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; Division[0.043]; Division[0.043]; Houston[0.043]; Houston[0.043]; Oakland[0.043]; Texas[0.043]; Cleveland[0.042]; Cincinnati[0.042]; Cincinnati[0.042]; Pittsburgh[0.042]; Pittsburgh[0.042]; Philadelphia[0.041]; Philadelphia[0.041]; Toronto[0.041]; Eastern[0.041]; Colorado[0.041]; Colorado[0.041]; Detroit[0.040]; York[0.040]; York[0.040]; York[0.040]; National[0.040]; Baltimore[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Diego ==> ENTITY: \u001b[32mSan Diego Padres\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.225:0.225[0]; log p(e|m)= -3.912:-3.912[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; Houston[0.043]; Houston[0.043]; Oakland[0.043]; California[0.043]; California[0.043]; Cincinnati[0.043]; Cincinnati[0.043]; Cleveland[0.042]; Philadelphia[0.042]; Philadelphia[0.042]; August[0.042]; August[0.042]; Baltimore[0.041]; Pittsburgh[0.041]; Pittsburgh[0.041]; Boston[0.041]; Toronto[0.040]; Detroit[0.040]; Texas[0.040]; Friday[0.040]; Friday[0.040]; Florida[0.040]; Florida[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Baltimore ==> ENTITY: \u001b[32mBaltimore Orioles\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.214:0.214[0]; log p(e|m)= -4.075:-4.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.048]; games[0.048]; Cleveland[0.048]; Cleveland[0.048]; League[0.047]; League[0.047]; League[0.047]; won[0.046]; Baseball[0.046]; Baseball[0.046]; Boston[0.046]; played[0.045]; Oakland[0.045]; Detroit[0.045]; Detroit[0.045]; Toronto[0.045]; Toronto[0.045]; winning[0.044]; lost[0.043]; York[0.043]; York[0.043]; York[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oakland ==> ENTITY: \u001b[32mOakland Athletics\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.220:0.220[0]; log p(e|m)= -3.912:-3.912[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPittsburgh[0.045]; Oakland[0.045]; games[0.045]; games[0.045]; League[0.044]; League[0.044]; League[0.044]; Cleveland[0.044]; Cleveland[0.044]; Baseball[0.044]; Houston[0.044]; California[0.044]; California[0.044]; Cincinnati[0.043]; Philadelphia[0.043]; Baltimore[0.043]; Baltimore[0.043]; Boston[0.043]; Boston[0.043]; Detroit[0.042]; Detroit[0.042]; Central[0.042]; Central[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pittsburgh ==> ENTITY: \u001b[32mPittsburgh Pirates\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.195:0.195[0]; log p(e|m)= -4.017:-4.017[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPittsburgh[0.045]; Philadelphia[0.045]; Philadelphia[0.045]; Cincinnati[0.044]; Cincinnati[0.044]; Montreal[0.044]; Montreal[0.044]; Florida[0.042]; Florida[0.042]; Houston[0.042]; Houston[0.042]; York[0.041]; York[0.041]; Colorado[0.041]; Colorado[0.041]; Chicago[0.040]; Chicago[0.040]; Atlanta[0.040]; Atlanta[0.040]; Louis[0.040]; Louis[0.040]; Central[0.039]; Friday[0.039]; Western[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Francisco ==> ENTITY: \u001b[32mSan Francisco Giants\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.161:0.161[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrancisco[0.051]; San[0.049]; San[0.049]; San[0.049]; Diego[0.045]; Diego[0.045]; Los[0.045]; Los[0.045]; Houston[0.044]; Houston[0.044]; Philadelphia[0.043]; Philadelphia[0.043]; Cincinnati[0.042]; Cincinnati[0.042]; Pittsburgh[0.042]; Pittsburgh[0.042]; Florida[0.040]; Florida[0.040]; Atlanta[0.040]; Atlanta[0.040]; York[0.040]; York[0.040]; August[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMilwaukee\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAtlanta Braves\u001b[39m <---> \u001b[32mMilwaukee Brewers\u001b[39m\t\nSCORES: global= 0.254:0.251[\u001b[31m0.003\u001b[39m]; local()= 0.243:0.205[\u001b[31m0.038\u001b[39m]; log p(e|m)= -3.912:-3.411[\u001b[32m0.501\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.044]; Cleveland[0.044]; Cleveland[0.044]; Cincinnati[0.043]; Pittsburgh[0.043]; League[0.042]; Boston[0.042]; Boston[0.042]; Baltimore[0.042]; Baltimore[0.042]; Houston[0.042]; Detroit[0.042]; Detroit[0.042]; Oakland[0.041]; Oakland[0.041]; Atlanta[0.041]; Central[0.041]; Central[0.041]; Milwaukee[0.041]; Toronto[0.040]; Toronto[0.040]; York[0.040]; York[0.040]; York[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cleveland ==> ENTITY: \u001b[32mCleveland Indians\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.235:0.235[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.046]; games[0.044]; games[0.044]; Baseball[0.042]; Baseball[0.042]; League[0.042]; League[0.042]; League[0.042]; League[0.042]; Baltimore[0.042]; Baltimore[0.042]; Boston[0.041]; Boston[0.041]; Oakland[0.041]; Oakland[0.041]; Detroit[0.041]; Detroit[0.041]; American[0.041]; Thursday[0.040]; Thursday[0.040]; Toronto[0.040]; Toronto[0.040]; played[0.040]; Atlanta[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Seattle ==> ENTITY: \u001b[32mSeattle Mariners\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.191:0.191[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSeattle[0.044]; Cleveland[0.043]; Cleveland[0.043]; League[0.043]; Houston[0.042]; Cincinnati[0.042]; Cincinnati[0.042]; Oakland[0.042]; Oakland[0.042]; Baltimore[0.042]; Pittsburgh[0.042]; Philadelphia[0.042]; Western[0.042]; Western[0.042]; Boston[0.041]; Toronto[0.041]; Toronto[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Detroit[0.040]; Detroit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = National League Eastern Division\u001b[39m ==> ENTITY: \u001b[32mNational League East\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.044]; Cleveland[0.044]; Philadelphia[0.044]; Philadelphia[0.044]; Cincinnati[0.043]; Cincinnati[0.043]; Pittsburgh[0.043]; Houston[0.042]; Oakland[0.042]; Oakland[0.042]; Atlanta[0.041]; Atlanta[0.041]; Baltimore[0.041]; Boston[0.041]; Detroit[0.041]; Detroit[0.041]; Toronto[0.041]; Toronto[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Montreal[0.040]; Montreal[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colorado ==> ENTITY: \u001b[32mColorado Rockies\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.198:0.198[0]; log p(e|m)= -4.510:-4.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nColorado[0.055]; Montreal[0.045]; Montreal[0.045]; Toronto[0.044]; California[0.043]; Philadelphia[0.043]; Philadelphia[0.043]; Kansas[0.043]; Cincinnati[0.043]; Cincinnati[0.043]; Pittsburgh[0.043]; Pittsburgh[0.043]; Houston[0.043]; Houston[0.043]; Florida[0.043]; Florida[0.043]; Boston[0.043]; Cleveland[0.042]; Texas[0.042]; League[0.042]; Atlanta[0.042]; Atlanta[0.042]; Oakland[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York Yankees\u001b[39m\t\nSCORES: global= 0.238:0.238[0]; local()= 0.174:0.174[0]; log p(e|m)= -5.116:-5.116[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.047]; Philadelphia[0.047]; Pittsburgh[0.045]; Pittsburgh[0.045]; Cincinnati[0.045]; Cincinnati[0.045]; Houston[0.045]; Houston[0.045]; York[0.044]; Central[0.043]; Atlanta[0.043]; Friday[0.043]; Division[0.043]; Division[0.043]; Montreal[0.043]; Montreal[0.043]; Chicago[0.042]; Chicago[0.042]; Florida[0.042]; Florida[0.042]; Colorado[0.041]; Colorado[0.041]; Western[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colorado ==> ENTITY: \u001b[32mColorado Rockies\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.164:0.164[0]; log p(e|m)= -4.510:-4.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nColorado[0.054]; Montreal[0.044]; Montreal[0.044]; Philadelphia[0.043]; Philadelphia[0.043]; Cincinnati[0.042]; Cincinnati[0.042]; Pittsburgh[0.042]; Pittsburgh[0.042]; Houston[0.042]; Houston[0.042]; Florida[0.042]; Florida[0.042]; Atlanta[0.041]; Louis[0.041]; Louis[0.041]; Central[0.040]; Chicago[0.040]; Chicago[0.040]; Western[0.039]; Division[0.039]; Division[0.039]; Friday[0.039]; August[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Toronto ==> ENTITY: \u001b[32mToronto Blue Jays\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.204:0.204[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nToronto[0.045]; games[0.044]; games[0.044]; League[0.043]; League[0.043]; League[0.043]; League[0.043]; Baseball[0.043]; Baseball[0.043]; played[0.042]; Boston[0.042]; Boston[0.042]; Cleveland[0.041]; Cleveland[0.041]; Detroit[0.040]; Detroit[0.040]; won[0.040]; Baltimore[0.040]; Baltimore[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; winning[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cleveland ==> ENTITY: \u001b[32mCleveland Indians\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.219:0.219[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.047]; Philadelphia[0.043]; Cincinnati[0.043]; League[0.043]; Houston[0.043]; Baltimore[0.042]; Baltimore[0.042]; Boston[0.042]; Boston[0.042]; Pittsburgh[0.042]; Detroit[0.041]; Detroit[0.041]; Oakland[0.041]; Oakland[0.041]; Toronto[0.041]; Toronto[0.041]; Atlanta[0.041]; York[0.041]; York[0.041]; York[0.041]; City[0.040]; City[0.040]; Texas[0.040]; Texas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlanta ==> ENTITY: \u001b[32mAtlanta Braves\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.235:0.235[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.045]; Philadelphia[0.045]; Cleveland[0.045]; Cleveland[0.045]; League[0.044]; Atlanta[0.044]; Cincinnati[0.044]; Cincinnati[0.044]; Houston[0.044]; Houston[0.044]; Pittsburgh[0.044]; Pittsburgh[0.044]; Diego[0.043]; Diego[0.043]; Boston[0.043]; Baltimore[0.043]; Oakland[0.043]; Oakland[0.043]; Detroit[0.042]; Detroit[0.042]; Los[0.042]; Los[0.042]; Toronto[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Montreal ==> ENTITY: \u001b[32mMontreal Expos\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.246:0.246[0]; log p(e|m)= -4.510:-4.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMontreal[0.044]; Toronto[0.043]; Baltimore[0.043]; League[0.043]; Houston[0.042]; Houston[0.042]; Cincinnati[0.042]; Cincinnati[0.042]; Cleveland[0.042]; Cleveland[0.042]; Philadelphia[0.042]; Philadelphia[0.042]; Boston[0.042]; Louis[0.042]; Detroit[0.041]; Detroit[0.041]; Pittsburgh[0.041]; Pittsburgh[0.041]; Oakland[0.041]; Oakland[0.041]; Atlanta[0.041]; Atlanta[0.041]; Friday[0.040]; Friday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kansas City ==> ENTITY: \u001b[32mKansas City Royals\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.225:0.225[0]; log p(e|m)= -3.576:-3.576[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; games[0.044]; Baseball[0.043]; Baseball[0.043]; Houston[0.042]; League[0.042]; League[0.042]; League[0.042]; League[0.042]; Cleveland[0.042]; Cleveland[0.042]; Toronto[0.042]; Toronto[0.042]; Baltimore[0.041]; Baltimore[0.041]; Philadelphia[0.041]; Oakland[0.041]; Oakland[0.041]; City[0.041]; won[0.040]; Detroit[0.040]; Detroit[0.040]; played[0.040]; winning[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Francisco ==> ENTITY: \u001b[32mSan Francisco Giants\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.179:0.179[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; Philadelphia[0.044]; Philadelphia[0.044]; Cleveland[0.043]; Houston[0.043]; Houston[0.043]; California[0.043]; Oakland[0.042]; Cincinnati[0.042]; Cincinnati[0.042]; Diego[0.041]; Diego[0.041]; Francisco[0.041]; Baltimore[0.041]; San[0.041]; San[0.041]; San[0.041]; Pittsburgh[0.041]; Pittsburgh[0.041]; Boston[0.041]; Texas[0.041]; Seattle[0.040]; Los[0.040]; Los[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlanta ==> ENTITY: \u001b[32mAtlanta Braves\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.207:0.207[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.048]; Philadelphia[0.048]; League[0.047]; Atlanta[0.047]; Cincinnati[0.047]; Cincinnati[0.047]; Houston[0.046]; Houston[0.046]; Pittsburgh[0.046]; Pittsburgh[0.046]; Diego[0.046]; Diego[0.046]; Boston[0.046]; Baltimore[0.045]; Oakland[0.045]; Los[0.044]; Los[0.044]; Western[0.043]; Francisco[0.043]; Francisco[0.043]; Chicago[0.043]; Chicago[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kansas City ==> ENTITY: \u001b[32mKansas City Royals\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.229:0.229[0]; log p(e|m)= -3.576:-3.576[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; Houston[0.043]; League[0.043]; League[0.043]; Cleveland[0.043]; Cleveland[0.043]; Toronto[0.042]; Toronto[0.042]; Baltimore[0.042]; Baltimore[0.042]; Philadelphia[0.042]; Oakland[0.042]; Oakland[0.042]; City[0.042]; Cincinnati[0.041]; Pittsburgh[0.041]; Detroit[0.041]; Detroit[0.041]; Boston[0.040]; Boston[0.040]; Montreal[0.040]; Louis[0.040]; Texas[0.040]; Texas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Texas ==> ENTITY: \u001b[32mTexas Rangers (baseball)\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.209:0.209[0]; log p(e|m)= -4.962:-4.962[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTexas[0.047]; Cleveland[0.045]; Cleveland[0.045]; Houston[0.045]; League[0.044]; Baltimore[0.044]; Baltimore[0.044]; Cincinnati[0.044]; Division[0.043]; Division[0.043]; Division[0.043]; Division[0.043]; Division[0.043]; Oakland[0.043]; Oakland[0.043]; Philadelphia[0.043]; Detroit[0.042]; Detroit[0.042]; Friday[0.042]; Friday[0.042]; Pittsburgh[0.042]; August[0.042]; August[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Florida ==> ENTITY: \u001b[32mMiami Marlins\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.224:0.224[0]; log p(e|m)= -4.828:-4.828[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFlorida[0.046]; Cincinnati[0.042]; Cincinnati[0.042]; Philadelphia[0.042]; Philadelphia[0.042]; Montreal[0.042]; Montreal[0.042]; League[0.042]; Pittsburgh[0.042]; Pittsburgh[0.042]; Toronto[0.042]; Texas[0.042]; Texas[0.042]; Cleveland[0.042]; Cleveland[0.042]; Houston[0.041]; Houston[0.041]; California[0.041]; California[0.041]; Milwaukee[0.041]; Milwaukee[0.041]; Oakland[0.040]; Oakland[0.040]; Baltimore[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York Yankees\u001b[39m\t\nSCORES: global= 0.239:0.239[0]; local()= 0.204:0.204[0]; log p(e|m)= -5.116:-5.116[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.046]; Philadelphia[0.046]; Cleveland[0.045]; Boston[0.045]; League[0.045]; City[0.044]; City[0.044]; Baltimore[0.044]; Pittsburgh[0.044]; Pittsburgh[0.044]; Cincinnati[0.044]; Cincinnati[0.044]; Houston[0.043]; Houston[0.043]; Toronto[0.043]; Oakland[0.043]; Oakland[0.043]; Detroit[0.043]; York[0.042]; York[0.042]; Central[0.041]; Atlanta[0.041]; Atlanta[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York Yankees\u001b[39m\t\nSCORES: global= 0.239:0.239[0]; local()= 0.208:0.208[0]; log p(e|m)= -5.116:-5.116[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; games[0.045]; Baseball[0.044]; Baseball[0.044]; Cleveland[0.043]; Cleveland[0.043]; Boston[0.042]; League[0.042]; League[0.042]; League[0.042]; City[0.042]; City[0.042]; Baltimore[0.042]; Toronto[0.041]; Toronto[0.041]; Oakland[0.041]; Thursday[0.041]; Thursday[0.041]; played[0.041]; Detroit[0.041]; Detroit[0.041]; York[0.040]; American[0.040]; won[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Florida ==> ENTITY: \u001b[32mMiami Marlins\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.221:0.221[0]; log p(e|m)= -4.828:-4.828[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFlorida[0.051]; Cincinnati[0.047]; Cincinnati[0.047]; Philadelphia[0.047]; Philadelphia[0.047]; Montreal[0.047]; Montreal[0.047]; League[0.046]; Pittsburgh[0.046]; Pittsburgh[0.046]; Houston[0.045]; Houston[0.045]; California[0.045]; Oakland[0.044]; Baltimore[0.044]; Boston[0.044]; Atlanta[0.044]; Atlanta[0.044]; Colorado[0.044]; Colorado[0.044]; Chicago[0.044]; Chicago[0.044]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew York\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew York Yankees\u001b[39m <---> \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.238:0.232[\u001b[31m0.006\u001b[39m]; local()= 0.186:0.098[\u001b[31m0.088\u001b[39m]; log p(e|m)= -5.116:-1.790[\u001b[32m3.326\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; games[0.045]; Baseball[0.044]; Baseball[0.044]; Cleveland[0.043]; Boston[0.043]; League[0.043]; League[0.043]; League[0.043]; City[0.042]; Baltimore[0.042]; Toronto[0.041]; Thursday[0.041]; Thursday[0.041]; played[0.041]; Detroit[0.041]; York[0.040]; American[0.040]; won[0.040]; Central[0.040]; Division[0.039]; Division[0.039]; Division[0.039]; winning[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Detroit ==> ENTITY: \u001b[32mDetroit Tigers\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.227:0.227[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDetroit[0.044]; Cleveland[0.043]; Cleveland[0.043]; Cincinnati[0.043]; Montreal[0.043]; Philadelphia[0.043]; League[0.042]; League[0.042]; Toronto[0.042]; Toronto[0.042]; Oakland[0.042]; Oakland[0.042]; Pittsburgh[0.041]; Boston[0.041]; Boston[0.041]; Houston[0.041]; Baltimore[0.041]; Baltimore[0.041]; Division[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Detroit ==> ENTITY: \u001b[32mDetroit Tigers\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.246:0.246[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.053]; games[0.053]; Detroit[0.049]; Baseball[0.049]; Baseball[0.049]; Cleveland[0.048]; Cleveland[0.048]; Montreal[0.048]; League[0.047]; League[0.047]; League[0.047]; League[0.047]; Toronto[0.047]; Toronto[0.047]; Thursday[0.046]; Thursday[0.046]; Boston[0.046]; Boston[0.046]; played[0.046]; Oakland[0.046]; Oakland[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York Yankees\u001b[39m\t\nSCORES: global= 0.238:0.238[0]; local()= 0.203:0.203[0]; log p(e|m)= -5.116:-5.116[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.044]; Cleveland[0.043]; Cleveland[0.043]; Boston[0.043]; Boston[0.043]; League[0.043]; City[0.042]; City[0.042]; Baltimore[0.042]; Baltimore[0.042]; Pittsburgh[0.042]; Cincinnati[0.042]; Houston[0.041]; Toronto[0.041]; Toronto[0.041]; Oakland[0.041]; Oakland[0.041]; Detroit[0.041]; Detroit[0.041]; York[0.040]; Central[0.040]; Central[0.040]; Atlanta[0.040]; Atlanta[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Philadelphia ==> ENTITY: \u001b[32mPhiladelphia Phillies\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.220:0.220[0]; log p(e|m)= -4.510:-4.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.046]; Cleveland[0.045]; Houston[0.043]; Houston[0.043]; League[0.043]; Cincinnati[0.043]; Cincinnati[0.043]; Baltimore[0.043]; Pittsburgh[0.042]; Pittsburgh[0.042]; Boston[0.042]; Oakland[0.042]; Oakland[0.042]; Detroit[0.041]; Atlanta[0.040]; Atlanta[0.040]; Florida[0.040]; Florida[0.040]; York[0.040]; York[0.040]; York[0.040]; Toronto[0.040]; Milwaukee[0.040]; Milwaukee[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Central Division ==> ENTITY: \u001b[32mNational League Central\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.184:0.184[0]; log p(e|m)= -0.989:-0.989[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPittsburgh[0.044]; Pittsburgh[0.044]; League[0.044]; Division[0.044]; Division[0.044]; Division[0.044]; Cincinnati[0.043]; Cincinnati[0.043]; Cleveland[0.043]; Philadelphia[0.042]; Philadelphia[0.042]; Boston[0.042]; Detroit[0.041]; Baltimore[0.041]; Western[0.041]; Western[0.041]; Houston[0.040]; Houston[0.040]; Toronto[0.040]; Oakland[0.040]; Oakland[0.040]; Eastern[0.039]; Atlanta[0.039]; Atlanta[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Western Division ==> ENTITY: \u001b[32mNational League West\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.212:0.212[0]; log p(e|m)= -1.295:-1.295[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; League[0.044]; League[0.044]; League[0.044]; Boston[0.042]; Boston[0.042]; games[0.042]; games[0.042]; Philadelphia[0.041]; Cleveland[0.041]; Cleveland[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; winning[0.041]; won[0.041]; Detroit[0.041]; Detroit[0.041]; Texas[0.041]; Texas[0.041]; standings[0.041]; Oakland[0.041]; Oakland[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cincinnati ==> ENTITY: \u001b[32mCincinnati Reds\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.193:0.193[0]; log p(e|m)= -3.772:-3.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCincinnati[0.050]; League[0.048]; Philadelphia[0.047]; Philadelphia[0.047]; Oakland[0.047]; Houston[0.047]; Houston[0.047]; Pittsburgh[0.047]; Pittsburgh[0.047]; Baltimore[0.046]; Louis[0.045]; Louis[0.045]; National[0.044]; Atlanta[0.044]; Atlanta[0.044]; Montreal[0.044]; Montreal[0.044]; Friday[0.044]; Florida[0.043]; Florida[0.043]; Chicago[0.043]; Chicago[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Diego ==> ENTITY: \u001b[32mSan Diego Padres\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.189:0.189[0]; log p(e|m)= -3.912:-3.912[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDiego[0.046]; San[0.045]; San[0.045]; San[0.045]; League[0.044]; Oakland[0.043]; Cincinnati[0.042]; Cincinnati[0.042]; Houston[0.042]; Houston[0.042]; Philadelphia[0.041]; Philadelphia[0.041]; August[0.041]; Baltimore[0.041]; Los[0.041]; Los[0.041]; Pittsburgh[0.040]; Pittsburgh[0.040]; Francisco[0.040]; Francisco[0.040]; Louis[0.039]; Louis[0.039]; Friday[0.039]; National[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Boston ==> ENTITY: \u001b[32mBoston Red Sox\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.226:0.226[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBoston[0.046]; games[0.046]; games[0.046]; Cleveland[0.043]; Cleveland[0.043]; Baseball[0.042]; Baseball[0.042]; League[0.042]; League[0.042]; League[0.042]; Toronto[0.042]; Toronto[0.042]; Detroit[0.041]; Detroit[0.041]; Baltimore[0.041]; Baltimore[0.041]; played[0.041]; won[0.040]; Oakland[0.040]; Oakland[0.040]; Central[0.040]; Texas[0.040]; Texas[0.040]; winning[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Major League Baseball\u001b[39m ==> ENTITY: \u001b[32mMajor League Baseball\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBaseball[0.046]; games[0.044]; games[0.044]; League[0.044]; League[0.044]; won[0.044]; winning[0.043]; Cleveland[0.043]; played[0.042]; Oakland[0.042]; Detroit[0.042]; Baltimore[0.041]; standings[0.041]; Standings[0.041]; Boston[0.041]; York[0.040]; York[0.040]; Thursday[0.040]; Thursday[0.040]; Toronto[0.040]; percentage[0.040]; Texas[0.040]; Milwaukee[0.039]; Chicago[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Minnesota ==> ENTITY: \u001b[32mMinnesota Twins\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.237:0.237[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; games[0.044]; Baseball[0.043]; Baseball[0.043]; Cleveland[0.042]; Cleveland[0.042]; Minnesota[0.042]; League[0.042]; League[0.042]; League[0.042]; League[0.042]; Oakland[0.041]; Oakland[0.041]; played[0.041]; Detroit[0.041]; Detroit[0.041]; Baltimore[0.041]; Baltimore[0.041]; Philadelphia[0.041]; Thursday[0.041]; Thursday[0.041]; National[0.040]; Boston[0.040]; Boston[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Houston ==> ENTITY: \u001b[32mHouston Astros\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.195:0.195[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.048]; League[0.044]; Texas[0.043]; Texas[0.043]; Cleveland[0.043]; Baltimore[0.042]; Cincinnati[0.042]; Cincinnati[0.042]; Philadelphia[0.042]; Philadelphia[0.042]; Oakland[0.042]; Oakland[0.042]; Pittsburgh[0.041]; Pittsburgh[0.041]; Boston[0.041]; City[0.041]; City[0.041]; Atlanta[0.040]; Atlanta[0.040]; National[0.040]; Toronto[0.040]; Friday[0.040]; Friday[0.040]; Detroit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cincinnati ==> ENTITY: \u001b[32mCincinnati Reds\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.201:0.201[0]; log p(e|m)= -3.772:-3.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCincinnati[0.045]; Cleveland[0.044]; League[0.043]; Philadelphia[0.043]; Philadelphia[0.043]; Oakland[0.043]; Oakland[0.043]; Houston[0.043]; Houston[0.043]; Pittsburgh[0.042]; Pittsburgh[0.042]; Baltimore[0.042]; Louis[0.041]; Louis[0.041]; Detroit[0.041]; Boston[0.041]; Toronto[0.041]; Milwaukee[0.040]; National[0.040]; Minnesota[0.040]; Texas[0.040]; Texas[0.040]; Atlanta[0.040]; Atlanta[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Texas ==> ENTITY: \u001b[32mTexas Rangers (baseball)\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.194:0.194[0]; log p(e|m)= -4.962:-4.962[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTexas[0.045]; Baseball[0.043]; Cleveland[0.042]; Cleveland[0.042]; Houston[0.042]; games[0.042]; games[0.042]; League[0.042]; League[0.042]; League[0.042]; League[0.042]; Baltimore[0.042]; Baltimore[0.042]; Cincinnati[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Standings[0.041]; Oakland[0.041]; Oakland[0.041]; Philadelphia[0.041]; standings[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Montreal ==> ENTITY: \u001b[32mMontreal Expos\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.200:0.200[0]; log p(e|m)= -4.510:-4.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMontreal[0.045]; League[0.044]; Houston[0.043]; Houston[0.043]; Cincinnati[0.043]; Cincinnati[0.043]; Philadelphia[0.043]; Philadelphia[0.043]; Louis[0.042]; Louis[0.042]; Pittsburgh[0.042]; Pittsburgh[0.042]; Atlanta[0.042]; Atlanta[0.042]; Friday[0.041]; York[0.040]; York[0.040]; Florida[0.040]; Florida[0.040]; Chicago[0.040]; Chicago[0.040]; Division[0.040]; Division[0.040]; Division[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Toronto ==> ENTITY: \u001b[32mToronto Blue Jays\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.206:0.206[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nToronto[0.048]; Montreal[0.046]; League[0.045]; Boston[0.044]; Boston[0.044]; Cleveland[0.044]; Cleveland[0.044]; Cincinnati[0.044]; Pittsburgh[0.044]; Philadelphia[0.043]; Detroit[0.043]; Detroit[0.043]; Houston[0.043]; Baltimore[0.043]; Baltimore[0.043]; Division[0.042]; Division[0.042]; Division[0.042]; Division[0.042]; Division[0.042]; Division[0.042]; Atlanta[0.042]; Louis[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Central Division ==> ENTITY: \u001b[32mNational League Central\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.185:0.185[0]; log p(e|m)= -0.989:-0.989[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; League[0.044]; League[0.044]; League[0.044]; games[0.043]; games[0.043]; Division[0.043]; Division[0.043]; Division[0.043]; Cleveland[0.042]; Cleveland[0.042]; Boston[0.041]; Boston[0.041]; winning[0.041]; won[0.041]; Detroit[0.041]; Detroit[0.041]; Western[0.040]; Baltimore[0.040]; Baltimore[0.040]; Baseball[0.040]; Baseball[0.040]; Toronto[0.040]; Toronto[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Boston ==> ENTITY: \u001b[32mBoston Red Sox\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.212:0.212[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBoston[0.047]; York[0.045]; York[0.045]; Montreal[0.043]; Cleveland[0.043]; Cleveland[0.043]; Philadelphia[0.043]; Pittsburgh[0.042]; League[0.042]; City[0.042]; City[0.042]; Toronto[0.042]; Toronto[0.042]; Cincinnati[0.041]; Detroit[0.041]; Detroit[0.041]; Baltimore[0.041]; Houston[0.041]; Oakland[0.040]; Oakland[0.040]; Louis[0.040]; Colorado[0.039]; Western[0.039]; Western[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Major League ==> ENTITY: \u001b[32mMajor League Baseball\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.177:0.177[0]; log p(e|m)= -0.478:-0.478[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.048]; League[0.048]; games[0.046]; games[0.046]; Baseball[0.046]; Baseball[0.046]; won[0.045]; Cleveland[0.044]; winning[0.044]; Major[0.044]; played[0.043]; Baltimore[0.043]; Detroit[0.042]; Boston[0.042]; Division[0.042]; Division[0.042]; Division[0.042]; American[0.041]; Texas[0.041]; Toronto[0.041]; Chicago[0.041]; Standings[0.041]; Milwaukee[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = American League Eastern Division\u001b[39m ==> ENTITY: \u001b[32mAmerican League East\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.230:0.230[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.046]; League[0.046]; Boston[0.046]; Cleveland[0.045]; winning[0.045]; won[0.045]; Detroit[0.045]; Detroit[0.045]; Baltimore[0.044]; games[0.044]; games[0.044]; Toronto[0.044]; Toronto[0.044]; Oakland[0.043]; Baseball[0.042]; Baseball[0.042]; standings[0.042]; Division[0.042]; Division[0.042]; lost[0.042]; played[0.041]; Chicago[0.041]; Chicago[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pittsburgh ==> ENTITY: \u001b[32mPittsburgh Pirates\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.230:0.230[0]; log p(e|m)= -4.017:-4.017[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.046]; Philadelphia[0.046]; Pittsburgh[0.045]; Cleveland[0.045]; Cincinnati[0.045]; Cincinnati[0.045]; League[0.044]; Montreal[0.044]; Montreal[0.044]; Baltimore[0.043]; Boston[0.043]; Toronto[0.043]; Oakland[0.043]; Oakland[0.043]; Florida[0.043]; Florida[0.043]; Houston[0.043]; Houston[0.043]; Detroit[0.043]; Louis[0.042]; Louis[0.042]; Atlanta[0.042]; Atlanta[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Seattle ==> ENTITY: \u001b[32mSeattle Mariners\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.220:0.220[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.048]; games[0.048]; Seattle[0.048]; Baseball[0.047]; Cleveland[0.046]; Cleveland[0.046]; League[0.046]; League[0.046]; League[0.046]; Houston[0.045]; Cincinnati[0.045]; Oakland[0.045]; Oakland[0.045]; Baltimore[0.045]; Baltimore[0.045]; Philadelphia[0.045]; played[0.045]; Western[0.044]; Boston[0.044]; Boston[0.044]; Toronto[0.044]; Toronto[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Philadelphia ==> ENTITY: \u001b[32mPhiladelphia Phillies\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.193:0.193[0]; log p(e|m)= -4.510:-4.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.049]; Cincinnati[0.045]; Cincinnati[0.045]; Pittsburgh[0.045]; Pittsburgh[0.045]; Houston[0.044]; Houston[0.044]; Montreal[0.044]; Montreal[0.044]; York[0.044]; York[0.044]; Chicago[0.043]; Chicago[0.043]; Atlanta[0.043]; Atlanta[0.043]; Division[0.042]; Division[0.042]; Division[0.042]; Friday[0.042]; Louis[0.042]; Louis[0.042]; Florida[0.042]; Florida[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Houston ==> ENTITY: \u001b[32mHouston Astros\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.165:0.165[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHouston[0.051]; Cincinnati[0.046]; Cincinnati[0.046]; Philadelphia[0.045]; Philadelphia[0.045]; Pittsburgh[0.045]; Pittsburgh[0.045]; Atlanta[0.043]; Atlanta[0.043]; Friday[0.043]; Western[0.043]; Colorado[0.043]; Colorado[0.043]; Florida[0.042]; Florida[0.042]; Chicago[0.042]; Chicago[0.042]; August[0.042]; Central[0.042]; Division[0.042]; Division[0.042]; York[0.041]; York[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Minnesota ==> ENTITY: \u001b[32mMinnesota Twins\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.218:0.218[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCleveland[0.044]; Cleveland[0.044]; Minnesota[0.044]; League[0.043]; Oakland[0.043]; Oakland[0.043]; Detroit[0.042]; Detroit[0.042]; Baltimore[0.042]; Baltimore[0.042]; Philadelphia[0.042]; Cincinnati[0.042]; Houston[0.042]; Pittsburgh[0.041]; National[0.041]; Boston[0.041]; Boston[0.041]; Texas[0.040]; Texas[0.040]; Central[0.040]; Central[0.040]; Friday[0.040]; Toronto[0.040]; Toronto[0.040]; \t\n\r [==================================>...........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 5s211ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3743/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1038testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.053]; won[0.048]; wins[0.046]; Guangdong[0.045]; Guangdong[0.045]; Administration[0.045]; province[0.044]; group[0.044]; statement[0.044]; included[0.043]; training[0.043]; said[0.042]; said[0.042]; broad[0.042]; billion[0.042]; fixed[0.042]; Friday[0.042]; crowns[0.042]; services[0.042]; media[0.041]; Telecommunications[0.041]; public[0.041]; says[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swedish ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.096:0.096[0]; log p(e|m)= -0.631:-0.631[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStockholm[0.046]; public[0.044]; crowns[0.044]; band[0.044]; province[0.044]; included[0.043]; won[0.042]; statement[0.042]; wins[0.042]; Ab[0.042]; services[0.041]; said[0.041]; said[0.041]; multi[0.041]; Ericsson[0.041]; Ericsson[0.041]; Ericsson[0.041]; Ericsson[0.041]; Friday[0.040]; Skr[0.040]; media[0.040]; China[0.040]; China[0.040]; communication[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Stockholm ==> ENTITY: \u001b[32mStockholm\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.077:0.077[0]; log p(e|m)= -0.104:-0.104[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwon[0.047]; band[0.045]; company[0.045]; crowns[0.044]; services[0.044]; multi[0.043]; Friday[0.042]; worth[0.042]; Telecommunications[0.042]; Ericsson[0.041]; Ericsson[0.041]; Ericsson[0.041]; network[0.041]; network[0.041]; statement[0.041]; equipment[0.041]; equipment[0.041]; included[0.040]; public[0.040]; said[0.040]; said[0.040]; said[0.040]; billion[0.040]; communication[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Intelligent Network\u001b[39m ==> ENTITY: \u001b[32mIntelligent Network\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nservices[0.054]; network[0.051]; network[0.051]; communication[0.050]; Telecommunications[0.050]; multi[0.049]; switching[0.048]; broad[0.048]; telecoms[0.048]; telecoms[0.048]; fixed[0.047]; Skr[0.046]; crowns[0.046]; products[0.046]; products[0.046]; training[0.045]; Reuters[0.045]; Stockholm[0.045]; Stockholm[0.045]; equipment[0.045]; equipment[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ericsson ==> ENTITY: \u001b[32mEricsson\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.199:0.199[0]; log p(e|m)= -0.040:-0.040[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEricsson[0.049]; Ericsson[0.049]; Ericsson[0.049]; company[0.048]; equipment[0.044]; equipment[0.044]; telecoms[0.043]; telecoms[0.043]; won[0.043]; Swedish[0.043]; Stockholm[0.043]; Stockholm[0.043]; services[0.043]; network[0.042]; network[0.042]; crowns[0.042]; wins[0.042]; Telecommunications[0.042]; products[0.042]; products[0.042]; communication[0.041]; switching[0.041]; says[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ericsson ==> ENTITY: \u001b[32mEricsson\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.040:-0.040[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEricsson[0.048]; Ericsson[0.048]; company[0.047]; equipment[0.043]; equipment[0.043]; telecoms[0.042]; telecoms[0.042]; Swedish[0.042]; services[0.042]; network[0.041]; network[0.041]; crowns[0.041]; Stockholm[0.041]; Stockholm[0.041]; communication[0.041]; Telecommunications[0.041]; products[0.041]; products[0.041]; won[0.040]; switching[0.040]; said[0.039]; said[0.039]; said[0.039]; billion[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Stockholm ==> ENTITY: \u001b[32mStockholm\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.104:-0.104[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwon[0.046]; Swedish[0.046]; crowns[0.043]; wins[0.043]; services[0.043]; Ericsson[0.043]; Ericsson[0.043]; Ericsson[0.043]; Ericsson[0.043]; public[0.043]; Friday[0.042]; group[0.042]; Telecommunications[0.041]; included[0.040]; band[0.040]; statement[0.040]; says[0.040]; Lm[0.040]; communication[0.040]; equipment[0.040]; equipment[0.040]; fixed[0.040]; network[0.039]; network[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.051]; won[0.045]; Guangdong[0.043]; Guangdong[0.043]; Administration[0.043]; wins[0.042]; province[0.042]; statement[0.042]; Network[0.041]; included[0.041]; training[0.041]; said[0.041]; said[0.041]; said[0.041]; broad[0.041]; group[0.040]; billion[0.040]; fixed[0.040]; equipment[0.040]; equipment[0.040]; Reuters[0.040]; media[0.040]; network[0.040]; network[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Guangdong\u001b[39m ==> ENTITY: \u001b[32mGuangdong\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.117:0.117[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.048]; China[0.048]; province[0.048]; Guangdong[0.047]; wins[0.043]; won[0.043]; Administration[0.042]; media[0.042]; Skr[0.041]; Friday[0.041]; crowns[0.041]; spokesman[0.041]; worth[0.040]; Stockholm[0.040]; Stockholm[0.040]; statement[0.040]; order[0.040]; order[0.040]; order[0.040]; order[0.040]; said[0.039]; said[0.039]; said[0.039]; communication[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.118:0.118[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmedia[0.044]; services[0.044]; company[0.044]; told[0.044]; billion[0.044]; network[0.043]; network[0.043]; statement[0.042]; products[0.042]; products[0.042]; newsroom[0.042]; said[0.041]; said[0.041]; said[0.041]; Network[0.041]; Post[0.041]; equipment[0.041]; equipment[0.041]; public[0.040]; fixed[0.040]; spokesman[0.040]; Friday[0.040]; Swedish[0.040]; multi[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ericsson ==> ENTITY: \u001b[32mEricsson\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.174:0.174[0]; log p(e|m)= -0.040:-0.040[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEricsson[0.049]; Ericsson[0.049]; equipment[0.043]; equipment[0.043]; telecoms[0.043]; telecoms[0.043]; Swedish[0.042]; network[0.042]; network[0.042]; crowns[0.042]; Stockholm[0.042]; Telecommunications[0.041]; products[0.041]; communication[0.041]; wins[0.040]; won[0.040]; switching[0.040]; says[0.040]; Skr[0.040]; said[0.039]; said[0.039]; billion[0.039]; included[0.039]; fixed[0.039]; \t\n\r [===================================>..........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 5s142ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3754/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1113testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = 1998 World Cup\u001b[39m ==> ENTITY: \u001b[32m1998 FIFA World Cup\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWorld[0.048]; Cup[0.048]; Soccer[0.046]; England[0.045]; international[0.045]; summer[0.045]; national[0.044]; sport[0.044]; European[0.044]; team[0.044]; Germany[0.044]; players[0.044]; France[0.044]; France[0.044]; won[0.044]; championship[0.042]; win[0.042]; Saturday[0.041]; Klinsmann[0.041]; Klinsmann[0.041]; Klinsmann[0.041]; Retire[0.040]; striker[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sueddeutsche Zeitung\u001b[39m ==> ENTITY: \u001b[32mSüddeutsche Zeitung\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.125:0.125[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; Germany[0.047]; quoted[0.046]; German[0.046]; Coach[0.044]; saying[0.043]; Juergen[0.042]; Bonn[0.042]; France[0.041]; France[0.041]; international[0.041]; week[0.041]; said[0.041]; said[0.041]; said[0.041]; daily[0.041]; European[0.040]; national[0.040]; Poland[0.040]; think[0.039]; think[0.039]; Retire[0.039]; Soccer[0.039]; ve[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Euro 96\u001b[39m ==> ENTITY: \u001b[32mUEFA Euro 1996\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.044]; won[0.044]; squad[0.044]; match[0.044]; European[0.043]; team[0.043]; Poland[0.043]; Germany[0.043]; Germany[0.043]; players[0.043]; players[0.043]; national[0.043]; win[0.042]; Coach[0.041]; sport[0.041]; championship[0.041]; Berti[0.041]; Klinsmann[0.040]; France[0.040]; week[0.040]; summer[0.039]; Vogts[0.039]; captain[0.039]; aiming[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Berti Vogts\u001b[39m ==> ENTITY: \u001b[32mBerti Vogts\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.157:0.157[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.045]; Germany[0.045]; players[0.044]; players[0.044]; England[0.044]; match[0.044]; squad[0.043]; Klinsmann[0.043]; team[0.043]; Poland[0.042]; Coach[0.042]; Euro[0.041]; national[0.041]; European[0.041]; friendly[0.041]; captain[0.040]; win[0.040]; week[0.040]; summer[0.040]; won[0.040]; sport[0.039]; France[0.039]; France[0.039]; Zeitung[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Klinsmann\u001b[39m ==> ENTITY: \u001b[32mJürgen Klinsmann\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.044]; Klinsmann[0.044]; Klinsmann[0.044]; team[0.044]; match[0.043]; players[0.043]; players[0.043]; Germany[0.043]; Germany[0.043]; squad[0.042]; England[0.042]; Coach[0.041]; European[0.041]; international[0.041]; won[0.040]; Cup[0.040]; Cup[0.040]; Euro[0.040]; national[0.040]; summer[0.040]; Juergen[0.040]; Poland[0.040]; win[0.039]; Bonn[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Cup ==> ENTITY: \u001b[32mFIFA World Cup\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.164:0.164[0]; log p(e|m)= -1.002:-1.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.045]; England[0.044]; national[0.044]; Cup[0.043]; team[0.043]; Germany[0.043]; sport[0.043]; international[0.043]; World[0.042]; players[0.042]; players[0.042]; European[0.042]; France[0.042]; France[0.042]; won[0.041]; Klinsmann[0.041]; Klinsmann[0.041]; Klinsmann[0.041]; win[0.040]; championship[0.040]; striker[0.039]; summer[0.039]; captain[0.039]; aiming[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland national football team\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.177:0.177[0]; log p(e|m)= -3.324:-3.324[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.044]; players[0.044]; players[0.044]; England[0.044]; Germany[0.043]; Germany[0.043]; European[0.043]; squad[0.043]; Euro[0.042]; team[0.042]; won[0.042]; national[0.042]; week[0.042]; Berti[0.042]; win[0.041]; summer[0.041]; friendly[0.041]; experience[0.041]; sport[0.041]; aiming[0.040]; championship[0.040]; Saturday[0.040]; France[0.039]; old[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany national football team\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.181:0.181[0]; log p(e|m)= -3.352:-3.352[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; Soccer[0.044]; match[0.043]; England[0.043]; team[0.043]; Poland[0.043]; players[0.043]; players[0.043]; international[0.042]; squad[0.042]; Klinsmann[0.042]; Klinsmann[0.042]; Klinsmann[0.042]; Euro[0.041]; national[0.041]; German[0.041]; European[0.040]; week[0.040]; Juergen[0.040]; win[0.040]; France[0.040]; France[0.040]; captain[0.040]; won[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Klinsmann\u001b[39m ==> ENTITY: \u001b[32mJürgen Klinsmann\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.045]; Klinsmann[0.045]; Klinsmann[0.045]; team[0.045]; players[0.044]; Germany[0.044]; England[0.043]; European[0.042]; international[0.042]; won[0.041]; Cup[0.041]; Cup[0.041]; national[0.041]; summer[0.041]; Juergen[0.041]; win[0.040]; Bonn[0.040]; striker[0.040]; German[0.040]; captain[0.039]; championship[0.039]; World[0.039]; World[0.039]; Zeitung[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.138:0.138[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.050]; England[0.045]; players[0.044]; Klinsmann[0.043]; Klinsmann[0.043]; Klinsmann[0.043]; Soccer[0.043]; European[0.043]; team[0.042]; France[0.042]; France[0.042]; Cup[0.041]; Cup[0.041]; national[0.041]; international[0.041]; win[0.041]; won[0.040]; summer[0.040]; Juergen[0.040]; striker[0.040]; Sueddeutsche[0.040]; World[0.039]; World[0.039]; Zeitung[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany national football team\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.172:0.172[0]; log p(e|m)= -3.352:-3.352[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; match[0.044]; England[0.043]; team[0.043]; Poland[0.043]; players[0.043]; players[0.043]; national[0.043]; squad[0.043]; Klinsmann[0.043]; France[0.043]; European[0.042]; Euro[0.042]; week[0.041]; win[0.040]; captain[0.040]; sport[0.040]; won[0.040]; championship[0.040]; Saturday[0.040]; summer[0.039]; Berti[0.039]; Vogts[0.039]; aiming[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mFrance\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFrance national football team\u001b[39m <---> \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.246:0.242[\u001b[31m0.004\u001b[39m]; local()= 0.180:0.120[\u001b[31m0.060\u001b[39m]; log p(e|m)= -3.381:-0.163[\u001b[32m3.219\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.048]; England[0.044]; team[0.044]; squad[0.043]; players[0.042]; players[0.042]; Soccer[0.042]; Euro[0.042]; Saturday[0.042]; Germany[0.042]; Germany[0.042]; match[0.041]; international[0.041]; Cup[0.041]; Cup[0.041]; national[0.041]; championship[0.041]; European[0.041]; week[0.041]; win[0.040]; sport[0.040]; captain[0.040]; won[0.040]; Poland[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mFrance\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFrance national football team\u001b[39m <---> \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.245:0.241[\u001b[31m0.004\u001b[39m]; local()= 0.171:0.102[\u001b[31m0.069\u001b[39m]; log p(e|m)= -3.381:-0.163[\u001b[32m3.219\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.049]; England[0.045]; team[0.045]; Soccer[0.044]; players[0.043]; players[0.043]; Germany[0.043]; Cup[0.042]; Cup[0.042]; international[0.042]; national[0.042]; championship[0.042]; European[0.042]; win[0.041]; sport[0.041]; won[0.041]; captain[0.040]; summer[0.040]; Klinsmann[0.039]; Klinsmann[0.039]; Klinsmann[0.039]; Saturday[0.039]; World[0.039]; World[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bonn\u001b[39m ==> ENTITY: \u001b[32mBonn\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.108:0.108[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.046]; players[0.044]; Soccer[0.044]; team[0.043]; German[0.043]; international[0.042]; Saturday[0.042]; France[0.042]; France[0.042]; summer[0.042]; European[0.042]; Juergen[0.042]; nucleus[0.041]; planned[0.041]; Cup[0.041]; Cup[0.041]; won[0.041]; championship[0.041]; Sueddeutsche[0.041]; national[0.040]; England[0.040]; win[0.040]; Zeitung[0.040]; old[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Juergen Klinsmann\u001b[39m ==> ENTITY: \u001b[32mJürgen Klinsmann\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.045]; Klinsmann[0.045]; Klinsmann[0.045]; team[0.045]; players[0.044]; Germany[0.044]; England[0.043]; European[0.042]; international[0.042]; won[0.042]; Cup[0.041]; Cup[0.041]; national[0.041]; summer[0.041]; win[0.041]; Bonn[0.040]; striker[0.040]; sport[0.040]; German[0.040]; captain[0.039]; championship[0.039]; World[0.039]; World[0.039]; Zeitung[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEngland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEngland national football team\u001b[39m <---> \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.247:0.234[\u001b[31m0.013\u001b[39m]; local()= 0.154:0.066[\u001b[31m0.088\u001b[39m]; log p(e|m)= -2.781:-0.190[\u001b[32m2.591\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsquad[0.045]; team[0.045]; players[0.045]; players[0.045]; match[0.044]; Soccer[0.044]; international[0.043]; Cup[0.042]; Cup[0.042]; national[0.042]; captain[0.041]; sport[0.041]; week[0.041]; win[0.041]; Poland[0.041]; France[0.041]; France[0.041]; won[0.040]; European[0.040]; Euro[0.040]; Saturday[0.040]; championship[0.039]; Germany[0.039]; Germany[0.039]; \t\n\r [===================================>..........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 5s51ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3771/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 988testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lagos ==> ENTITY: \u001b[32mLagos\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUyo[0.046]; Nigerian[0.046]; Nigeria[0.046]; Nigeria[0.046]; eastern[0.045]; national[0.044]; town[0.043]; naira[0.043]; naira[0.043]; services[0.041]; gang[0.041]; returned[0.041]; transporting[0.040]; agency[0.040]; corporal[0.040]; corporal[0.040]; corporal[0.040]; Agency[0.039]; carry[0.039]; carry[0.039]; loot[0.039]; loot[0.039]; reported[0.039]; truck[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nigeria ==> ENTITY: \u001b[32mNigeria\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.120:0.120[0]; log p(e|m)= -0.315:-0.315[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNigeria[0.050]; Nigerian[0.046]; national[0.045]; admitted[0.043]; having[0.042]; Lagos[0.042]; returned[0.042]; services[0.042]; town[0.041]; eastern[0.041]; corporal[0.041]; corporal[0.041]; corporal[0.041]; said[0.041]; news[0.040]; Uyo[0.040]; transporting[0.040]; Little[0.040]; naira[0.040]; naira[0.040]; valued[0.040]; rightful[0.040]; gang[0.040]; know[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Uyo\u001b[39m ==> ENTITY: \u001b[32mUyo\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.111:0.111[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLagos[0.046]; Nigerian[0.046]; Nigeria[0.046]; Nigeria[0.046]; Friday[0.045]; national[0.044]; town[0.042]; returned[0.042]; naira[0.041]; naira[0.041]; naira[0.041]; conditioners[0.041]; conditioners[0.041]; rightful[0.040]; air[0.040]; air[0.040]; valued[0.040]; eastern[0.040]; reported[0.040]; said[0.040]; Police[0.039]; paid[0.039]; paid[0.039]; agency[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nigerian ==> ENTITY: \u001b[32mNigeria\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.075:-0.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNigeria[0.049]; Nigeria[0.049]; national[0.044]; paid[0.044]; paid[0.044]; admitted[0.042]; stolen[0.042]; reported[0.042]; Lagos[0.042]; unnamed[0.042]; having[0.041]; naira[0.041]; naira[0.041]; services[0.041]; thieves[0.040]; thieves[0.040]; eastern[0.040]; said[0.040]; know[0.040]; hire[0.040]; gang[0.040]; Agency[0.039]; quoted[0.039]; Little[0.039]; \t\n\r [===================================>..........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 5s98ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3775/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1031testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.052:0.052[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYork[0.044]; latest[0.043]; mid[0.043]; Metals[0.043]; Thursday[0.043]; July[0.043]; stocks[0.042]; opening[0.042]; Stocks[0.041]; Stocks[0.041]; percent[0.041]; Markets[0.041]; Markets[0.041]; Blue[0.041]; boosted[0.041]; close[0.041]; close[0.041]; close[0.041]; World[0.041]; World[0.041]; rates[0.041]; edged[0.041]; Overnight[0.041]; Brent[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Morris\u001b[39m ==> ENTITY: \u001b[32mDick Morris\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncampaign[0.047]; Clinton[0.044]; Clinton[0.044]; Clinton[0.044]; Clinton[0.044]; yearlong[0.043]; President[0.042]; President[0.042]; Democratic[0.042]; election[0.042]; Republican[0.042]; Morris[0.042]; tabloid[0.041]; October[0.040]; political[0.040]; political[0.040]; political[0.040]; September[0.040]; September[0.040]; strategist[0.040]; quit[0.040]; months[0.040]; bid[0.040]; York[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Commerce Department\u001b[39m ==> ENTITY: \u001b[32mUnited States Department of Commerce\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \neconomic[0.045]; employment[0.044]; report[0.044]; report[0.044]; Bros[0.042]; domestic[0.042]; months[0.042]; months[0.042]; month[0.042]; month[0.042]; month[0.042]; reports[0.042]; February[0.041]; reported[0.041]; adjusted[0.041]; chief[0.041]; July[0.041]; July[0.041]; June[0.041]; Sales[0.040]; April[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Treasury ==> ENTITY: \u001b[32mUnited States Department of the Treasury\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.110:0.110[0]; log p(e|m)= -1.013:-1.013[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nExchange[0.045]; dollar[0.044]; shares[0.044]; James[0.042]; term[0.042]; Ralph[0.042]; bond[0.042]; bond[0.042]; bond[0.042]; bond[0.042]; market[0.042]; market[0.042]; monthly[0.041]; million[0.041]; lower[0.041]; lower[0.041]; rates[0.041]; yield[0.041]; employment[0.041]; rose[0.041]; chief[0.041]; drop[0.041]; Wednesday[0.040]; money[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = FTSE 100\u001b[39m ==> ENTITY: \u001b[32mFTSE 100 Index\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.055]; Exchange[0.054]; Exchange[0.054]; December[0.051]; October[0.051]; September[0.051]; September[0.051]; closed[0.050]; closed[0.050]; Overseas[0.050]; percent[0.049]; day[0.049]; index[0.049]; share[0.049]; earlier[0.049]; Mercantile[0.048]; record[0.048]; session[0.048]; rose[0.048]; fall[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Japanese ==> ENTITY: \u001b[32mJapan\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.067:0.067[0]; log p(e|m)= -0.992:-0.992[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \neconomy[0.045]; Nikkei[0.043]; percent[0.042]; percent[0.042]; percent[0.042]; percent[0.042]; mark[0.042]; issues[0.042]; mid[0.042]; slightly[0.042]; strong[0.042]; sank[0.042]; highest[0.041]; lower[0.041]; broader[0.041]; average[0.041]; July[0.041]; July[0.041]; July[0.041]; long[0.041]; rates[0.041]; largest[0.041]; German[0.041]; yen[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Republican ==> ENTITY: \u001b[32mRepublican Party (United States)\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.131:-0.131[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncampaign[0.046]; Democratic[0.043]; President[0.042]; President[0.042]; Clinton[0.042]; Clinton[0.042]; Clinton[0.042]; Clinton[0.042]; Morris[0.042]; Morris[0.042]; York[0.042]; resigning[0.042]; second[0.042]; election[0.042]; late[0.042]; Wednesday[0.041]; bid[0.041]; past[0.041]; September[0.041]; September[0.041]; February[0.040]; political[0.040]; political[0.040]; political[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew York Cotton Exchange\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31m1 Hanover Square\u001b[39m <---> \u001b[32mNew York Cotton Exchange\u001b[39m\t\nSCORES: global= 0.288:0.279[\u001b[31m0.008\u001b[39m]; local()= 0.101:0.159[\u001b[32m0.058\u001b[39m]; log p(e|m)= 0.000:-0.564[\u001b[31m0.564\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nExchange[0.045]; London[0.044]; trading[0.044]; York[0.043]; stocks[0.043]; markets[0.042]; closed[0.042]; closed[0.042]; closed[0.042]; Mercantile[0.042]; September[0.042]; September[0.042]; cotton[0.042]; earlier[0.041]; Wednesday[0.041]; Oil[0.041]; December[0.041]; October[0.041]; session[0.041]; indicating[0.040]; hurricanes[0.040]; climbed[0.040]; despite[0.040]; Overseas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dick Morris\u001b[39m ==> ENTITY: \u001b[32mDick Morris\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncampaign[0.047]; Clinton[0.044]; Clinton[0.044]; Clinton[0.044]; Clinton[0.044]; yearlong[0.043]; President[0.042]; President[0.042]; Democratic[0.042]; election[0.042]; Republican[0.042]; Morris[0.042]; tabloid[0.041]; political[0.041]; political[0.041]; political[0.041]; report[0.040]; February[0.040]; strategist[0.040]; April[0.040]; quit[0.040]; months[0.040]; months[0.040]; months[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dow ==> ENTITY: \u001b[32mDow Jones Industrial Average\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.183:0.183[0]; log p(e|m)= -1.519:-1.519[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstock[0.044]; month[0.043]; market[0.043]; market[0.043]; lower[0.043]; Stock[0.042]; close[0.042]; psychological[0.042]; percent[0.042]; percent[0.042]; percent[0.042]; swings[0.041]; fell[0.041]; nation[0.041]; telling[0.041]; million[0.041]; land[0.041]; reports[0.041]; Department[0.041]; declining[0.041]; shares[0.041]; domestic[0.041]; James[0.041]; reported[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mDow Jones\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mDow Jones & Company\u001b[39m <---> \u001b[32mDow Jones Industrial Average\u001b[39m\t\nSCORES: global= 0.266:0.266[\u001b[31m0.000\u001b[39m]; local()= 0.102:0.232[\u001b[32m0.130\u001b[39m]; log p(e|m)= 0.000:-0.423[\u001b[31m0.423\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndrop[0.045]; market[0.042]; market[0.042]; lower[0.042]; loss[0.042]; loss[0.042]; stocks[0.042]; Stock[0.042]; sank[0.042]; boosted[0.042]; York[0.042]; close[0.042]; close[0.042]; close[0.042]; Stocks[0.041]; rose[0.041]; point[0.041]; term[0.041]; mid[0.041]; biggest[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; percent[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clinton ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.136:0.136[0]; log p(e|m)= -1.917:-1.917[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinton[0.046]; Clinton[0.046]; Clinton[0.046]; President[0.045]; President[0.045]; campaign[0.044]; Democratic[0.042]; late[0.042]; Republican[0.041]; York[0.041]; grown[0.041]; September[0.041]; September[0.041]; confidential[0.040]; October[0.040]; Morris[0.040]; Morris[0.040]; Street[0.040]; December[0.040]; shared[0.040]; Iraq[0.040]; Wednesday[0.040]; low[0.040]; election[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nikkei ==> ENTITY: \u001b[32mNikkei 225\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.169:0.169[0]; log p(e|m)= -0.465:-0.465[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapanese[0.046]; yen[0.043]; stocks[0.043]; close[0.042]; close[0.042]; close[0.042]; Markets[0.042]; Markets[0.042]; mid[0.042]; July[0.042]; London[0.041]; Metals[0.041]; World[0.041]; World[0.041]; Stocks[0.041]; Stocks[0.041]; industrial[0.041]; Dow[0.041]; Dow[0.041]; percent[0.041]; Thursday[0.041]; average[0.041]; ended[0.041]; dollar[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tokyo ==> ENTITY: \u001b[32mTokyo\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.055:0.055[0]; log p(e|m)= -0.093:-0.093[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrecord[0.045]; Nikkei[0.044]; Exchange[0.044]; October[0.043]; closed[0.043]; London[0.043]; earlier[0.043]; December[0.043]; day[0.042]; session[0.042]; fall[0.042]; Overseas[0.041]; finish[0.041]; key[0.041]; key[0.041]; climbed[0.041]; despite[0.040]; share[0.040]; rose[0.040]; afternoon[0.040]; edging[0.040]; percent[0.040]; average[0.040]; indicating[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Democratic ==> ENTITY: \u001b[32mDemocratic Party (United States)\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.141:0.141[0]; log p(e|m)= -0.076:-0.076[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRepublican[0.045]; late[0.044]; election[0.043]; Morris[0.042]; Morris[0.042]; campaign[0.042]; York[0.042]; second[0.042]; Clinton[0.042]; Clinton[0.042]; Clinton[0.042]; Clinton[0.042]; President[0.041]; President[0.041]; Wednesday[0.041]; yearlong[0.041]; September[0.041]; September[0.041]; term[0.041]; October[0.041]; Iraq[0.041]; resigning[0.040]; past[0.040]; bid[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fed\u001b[39m ==> ENTITY: \u001b[32mFederal Reserve System\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrate[0.045]; rate[0.045]; rate[0.045]; Street[0.045]; economic[0.045]; Commerce[0.044]; key[0.044]; month[0.044]; month[0.044]; month[0.044]; report[0.044]; report[0.044]; clients[0.043]; months[0.043]; months[0.043]; monthly[0.042]; James[0.042]; Department[0.042]; David[0.042]; June[0.042]; employment[0.042]; nation[0.042]; expanded[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = New York Stock Exchange\u001b[39m ==> ENTITY: \u001b[32mNew York Stock Exchange\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nshares[0.044]; Wednesday[0.044]; Thursday[0.044]; closed[0.044]; Friday[0.043]; ended[0.042]; mid[0.042]; market[0.042]; market[0.042]; James[0.042]; sank[0.041]; close[0.041]; fell[0.041]; stocks[0.041]; largest[0.041]; said[0.041]; said[0.041]; Raymond[0.041]; July[0.041]; July[0.041]; July[0.041]; Wall[0.040]; equities[0.040]; rose[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.045:0.045[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrecord[0.048]; pound[0.045]; average[0.043]; gallon[0.043]; gallon[0.043]; October[0.042]; percent[0.042]; climbed[0.042]; September[0.042]; September[0.042]; despite[0.042]; day[0.042]; earlier[0.041]; December[0.041]; Overseas[0.041]; share[0.041]; Iraq[0.040]; higher[0.040]; higher[0.040]; Exchange[0.040]; Exchange[0.040]; edging[0.040]; York[0.040]; York[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clinton ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.137:0.137[0]; log p(e|m)= -1.917:-1.917[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinton[0.046]; Clinton[0.046]; Clinton[0.046]; President[0.045]; President[0.045]; campaign[0.044]; Democratic[0.042]; late[0.042]; Republican[0.041]; homes[0.041]; homes[0.041]; month[0.041]; month[0.041]; grown[0.041]; confidential[0.040]; Morris[0.040]; Morris[0.040]; shared[0.040]; July[0.040]; Wednesday[0.040]; election[0.039]; said[0.039]; said[0.039]; Dick[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nkey[0.044]; President[0.044]; second[0.043]; Overseas[0.043]; September[0.042]; September[0.042]; despite[0.042]; campaign[0.042]; October[0.042]; documents[0.042]; December[0.041]; higher[0.041]; higher[0.041]; yearlong[0.041]; low[0.041]; Clinton[0.041]; Clinton[0.041]; London[0.041]; Wednesday[0.041]; sharply[0.041]; reported[0.041]; finished[0.041]; said[0.041]; departure[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clinton ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.127:0.127[0]; log p(e|m)= -1.917:-1.917[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinton[0.046]; Clinton[0.046]; Clinton[0.046]; President[0.045]; President[0.045]; campaign[0.044]; Democratic[0.042]; Republican[0.041]; homes[0.041]; homes[0.041]; June[0.041]; month[0.041]; month[0.041]; July[0.041]; April[0.040]; confidential[0.040]; Morris[0.040]; Morris[0.040]; Street[0.040]; election[0.039]; said[0.039]; said[0.039]; adjusted[0.039]; estimated[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Raymond James\u001b[39m ==> ENTITY: \u001b[32mRaymond James Financial\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclients[0.047]; million[0.047]; David[0.046]; Bros[0.045]; stock[0.044]; equities[0.044]; month[0.044]; Ralph[0.043]; Exchange[0.043]; shares[0.043]; loss[0.043]; York[0.043]; equity[0.043]; market[0.043]; market[0.043]; Stock[0.042]; technical[0.042]; July[0.042]; July[0.042]; July[0.042]; Commerce[0.042]; Wednesday[0.042]; volume[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nikkei ==> ENTITY: \u001b[32mNikkei 225\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.166:0.166[0]; log p(e|m)= -0.465:-0.465[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nindex[0.047]; Exchange[0.045]; Tokyo[0.045]; closed[0.043]; Overseas[0.043]; December[0.043]; session[0.042]; London[0.042]; earlier[0.042]; record[0.041]; percent[0.041]; day[0.041]; average[0.041]; higher[0.041]; rose[0.041]; climbed[0.041]; shed[0.041]; despite[0.041]; key[0.040]; key[0.040]; afternoon[0.040]; growing[0.040]; fall[0.040]; share[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.054:0.054[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJuly[0.044]; London[0.043]; latest[0.043]; Thursday[0.043]; World[0.043]; World[0.043]; mid[0.042]; sank[0.042]; surprisingly[0.042]; term[0.042]; close[0.042]; close[0.042]; close[0.042]; stocks[0.042]; Metals[0.041]; Sept[0.041]; Strong[0.041]; Bond[0.041]; Dow[0.041]; Overnight[0.041]; biggest[0.040]; Stocks[0.040]; Stocks[0.040]; Aug[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = FTSE\u001b[39m ==> ENTITY: \u001b[32mFTSE 100 Index\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.048]; stocks[0.045]; Markets[0.045]; Markets[0.045]; Data[0.045]; closed[0.044]; largest[0.044]; Stocks[0.044]; Stocks[0.044]; Metals[0.044]; Rates[0.043]; rates[0.043]; Thursday[0.043]; mid[0.043]; percent[0.043]; Summary[0.043]; ended[0.042]; July[0.042]; July[0.042]; term[0.042]; biggest[0.042]; Sept[0.042]; Nikkei[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clinton ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.145:0.145[0]; log p(e|m)= -1.917:-1.917[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinton[0.046]; Clinton[0.046]; Clinton[0.046]; President[0.045]; President[0.045]; campaign[0.045]; Democratic[0.042]; late[0.042]; Republican[0.041]; York[0.041]; September[0.041]; September[0.041]; confidential[0.040]; October[0.040]; Morris[0.040]; Morris[0.040]; Street[0.040]; Iraq[0.040]; low[0.040]; Wednesday[0.040]; election[0.040]; said[0.040]; said[0.040]; moderate[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.056:0.056[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrecord[0.047]; trading[0.046]; Exchange[0.044]; Exchange[0.044]; York[0.043]; York[0.043]; earlier[0.042]; Mercantile[0.042]; September[0.041]; September[0.041]; October[0.041]; areas[0.041]; session[0.041]; day[0.041]; December[0.041]; average[0.041]; closed[0.040]; closed[0.040]; percent[0.040]; heating[0.040]; despite[0.040]; Overseas[0.040]; Cotton[0.040]; climbed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wall Street ==> ENTITY: \u001b[32mWall Street\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.040:-0.040[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntrading[0.046]; market[0.044]; rate[0.043]; rate[0.043]; markets[0.043]; late[0.043]; York[0.042]; President[0.042]; President[0.042]; Exchange[0.042]; term[0.042]; dollar[0.042]; dollar[0.042]; yearlong[0.042]; stock[0.041]; months[0.041]; resigning[0.041]; Dick[0.040]; stocks[0.040]; second[0.040]; sharply[0.040]; fortunes[0.040]; campaign[0.040]; quit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wall Street ==> ENTITY: \u001b[32mWall Street\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.119:0.119[0]; log p(e|m)= -0.040:-0.040[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmarket[0.045]; economic[0.044]; David[0.044]; rate[0.044]; James[0.044]; nation[0.044]; York[0.043]; Exchange[0.042]; stock[0.042]; months[0.041]; money[0.041]; key[0.041]; equities[0.041]; employment[0.041]; lower[0.041]; Stock[0.041]; month[0.041]; month[0.041]; equity[0.040]; domestic[0.040]; resurfaced[0.040]; expanded[0.040]; telling[0.040]; report[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = New York Mercantile Exchange\u001b[39m ==> ENTITY: \u001b[32mNew York Mercantile Exchange\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nExchange[0.048]; trading[0.046]; dollar[0.042]; dollar[0.042]; markets[0.042]; London[0.041]; Oil[0.041]; closed[0.041]; closed[0.041]; closed[0.041]; cents[0.041]; cents[0.041]; cents[0.041]; cents[0.041]; cents[0.041]; cents[0.041]; barrel[0.041]; oil[0.041]; stocks[0.041]; record[0.041]; crude[0.041]; index[0.041]; climbed[0.040]; cent[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.091:0.091[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nlatest[0.044]; mid[0.043]; July[0.043]; July[0.043]; July[0.043]; million[0.043]; volume[0.042]; highest[0.042]; opening[0.042]; Japanese[0.042]; ended[0.042]; Thursday[0.041]; Jump[0.041]; advances[0.041]; economy[0.041]; lower[0.041]; strong[0.041]; fell[0.041]; sank[0.040]; broader[0.040]; drop[0.040]; largest[0.040]; biggest[0.040]; Gold[0.040]; \t\n\r [===================================>..........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 4s915ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3806/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1104testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Enzo Scifo\u001b[39m ==> ENTITY: \u001b[32mEnzo Scifo\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.044]; Belgium[0.044]; Belgium[0.044]; France[0.043]; Tayfun[0.042]; Dirk[0.042]; Verheyen[0.042]; Belgian[0.042]; finals[0.042]; started[0.042]; Sukur[0.042]; suffered[0.042]; Degryse[0.041]; Degryse[0.041]; Bertrand[0.041]; Marc[0.041]; came[0.041]; Saffet[0.041]; coach[0.041]; fumbled[0.041]; Wilfried[0.041]; Luc[0.040]; second[0.040]; Claeys[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gert Verheyen\u001b[39m ==> ENTITY: \u001b[32mGert Verheyen\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.127:0.127[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.047]; Belgium[0.047]; Belgium[0.047]; Tayfun[0.042]; Dirk[0.042]; coach[0.042]; second[0.042]; started[0.042]; half[0.042]; Sukur[0.042]; suffered[0.041]; ball[0.041]; ball[0.041]; Saffet[0.041]; goal[0.040]; goal[0.040]; fumbled[0.040]; Wilfried[0.040]; Claeys[0.040]; Korkut[0.040]; time[0.040]; Schepens[0.040]; Belgian[0.040]; Filip[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Luis Oliveira\u001b[39m ==> ENTITY: \u001b[32mLuís Oliveira\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazilian[0.046]; June[0.044]; Belgium[0.044]; seven[0.043]; seven[0.043]; Degryse[0.043]; Dirk[0.043]; home[0.042]; born[0.042]; finals[0.042]; Euro[0.041]; Belgian[0.041]; minutes[0.041]; minutes[0.041]; minutes[0.041]; minutes[0.041]; ball[0.041]; penalty[0.040]; Filip[0.040]; English[0.040]; group[0.040]; earlier[0.040]; Elleray[0.040]; qualifying[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Oguz Cetin\u001b[39m ==> ENTITY: \u001b[32mOğuz Çetin\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCetin[0.044]; Yalcin[0.044]; Saffet[0.043]; Sukur[0.043]; Unsal[0.042]; Kafkas[0.042]; friendlies[0.042]; Turkey[0.042]; Turkey[0.042]; Tayfun[0.042]; coach[0.042]; coach[0.042]; Dirk[0.042]; Hakan[0.042]; Hakan[0.042]; Turkish[0.042]; Verheyen[0.041]; April[0.041]; ball[0.041]; Sergen[0.041]; Orhan[0.040]; Korkut[0.040]; Wilfried[0.040]; Claeys[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arif Erdem\u001b[39m ==> ENTITY: \u001b[32mArif Erdem\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTayfun[0.044]; Turkey[0.043]; Turkey[0.043]; Yalcin[0.043]; Sergen[0.043]; Dirk[0.042]; friendlies[0.042]; started[0.042]; Verheyen[0.042]; Saffet[0.042]; Turkish[0.042]; Sukur[0.042]; Hakan[0.042]; Hakan[0.042]; coach[0.041]; coach[0.041]; goal[0.041]; Cetin[0.041]; Cetin[0.041]; victory[0.041]; Ercan[0.041]; Wilfried[0.041]; Claeys[0.040]; ball[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Turkey ==> ENTITY: \u001b[32mTurkey national football team\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.168:0.168[0]; log p(e|m)= -3.037:-3.037[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.051]; Turkish[0.046]; Turkish[0.046]; Turkish[0.046]; second[0.044]; team[0.043]; match[0.043]; Sergen[0.042]; Belgium[0.041]; Belgium[0.041]; victory[0.041]; Soccer[0.041]; Cup[0.041]; Yalcin[0.040]; Yalcin[0.040]; break[0.040]; trouble[0.040]; substitute[0.040]; upset[0.040]; hopes[0.039]; half[0.039]; half[0.039]; half[0.039]; dozens[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgium ==> ENTITY: \u001b[32mBelgium national football team\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.170:0.170[0]; log p(e|m)= -2.882:-2.882[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.052]; Belgium[0.052]; France[0.047]; finals[0.044]; Van[0.044]; Euro[0.042]; qualifying[0.042]; second[0.042]; Turkey[0.041]; seven[0.041]; seven[0.041]; winners[0.040]; time[0.040]; June[0.040]; minutes[0.040]; minutes[0.040]; minutes[0.040]; Dirk[0.040]; Degryse[0.039]; Filip[0.039]; goal[0.039]; goal[0.039]; group[0.039]; upset[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia national football team\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.124:0.124[0]; log p(e|m)= -3.540:-3.540[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfriendlies[0.047]; victory[0.044]; Turkey[0.044]; Belgium[0.044]; Italy[0.043]; coach[0.043]; coach[0.043]; goal[0.043]; April[0.043]; started[0.042]; suffered[0.042]; Turkish[0.042]; Paul[0.042]; Alpay[0.041]; Sergen[0.041]; Kafkas[0.041]; expected[0.041]; draws[0.040]; strong[0.039]; bit[0.039]; said[0.039]; said[0.039]; succeeded[0.039]; Unsal[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mItaly\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mItaly\u001b[39m <---> \u001b[32mItaly national football team\u001b[39m\t\nSCORES: global= 0.245:0.244[\u001b[31m0.002\u001b[39m]; local()= 0.076:0.099[\u001b[32m0.022\u001b[39m]; log p(e|m)= -0.206:-2.957[\u001b[31m2.751\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfriendlies[0.047]; Belgium[0.044]; Turkey[0.044]; goal[0.043]; coach[0.043]; coach[0.043]; victory[0.043]; suffered[0.043]; Russia[0.042]; April[0.042]; strong[0.042]; expected[0.042]; Sergen[0.042]; Alpay[0.042]; started[0.041]; draws[0.041]; Turkish[0.040]; succeeded[0.040]; Paul[0.040]; Van[0.039]; Van[0.039]; Van[0.039]; Van[0.039]; fast[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bertrand Crasson\u001b[39m ==> ENTITY: \u001b[32mBertrand Crasson\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.048]; Belgium[0.048]; Belgium[0.048]; Belgian[0.045]; France[0.043]; Degryse[0.042]; Degryse[0.042]; minutes[0.041]; minutes[0.041]; Tayfun[0.041]; Dirk[0.041]; Verheyen[0.041]; second[0.041]; coach[0.041]; Sukur[0.040]; Claeys[0.040]; seven[0.040]; finals[0.040]; time[0.040]; Saffet[0.039]; fumbled[0.039]; Wilfried[0.039]; ball[0.039]; ball[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marc Degryse\u001b[39m ==> ENTITY: \u001b[32mMarc Degryse\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.132:0.132[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.046]; minutes[0.043]; minutes[0.043]; minutes[0.043]; minutes[0.043]; minutes[0.043]; Dirk[0.043]; English[0.042]; Belgian[0.042]; hopes[0.041]; seven[0.041]; ball[0.041]; half[0.041]; half[0.041]; half[0.041]; second[0.041]; Filip[0.041]; defender[0.041]; took[0.041]; hero[0.040]; David[0.040]; born[0.040]; Elleray[0.040]; Sergen[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Frederic Peiremans\u001b[39m ==> ENTITY: \u001b[32mFrédéric Peiremans\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.065:0.065[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgian[0.047]; Belgium[0.047]; Belgium[0.047]; Tayfun[0.043]; Dirk[0.043]; Verheyen[0.042]; Luc[0.042]; Sukur[0.042]; Luis[0.041]; Saffet[0.041]; coach[0.041]; fumbled[0.041]; Wilfried[0.041]; Claeys[0.041]; ball[0.041]; ball[0.041]; Gilles[0.041]; Korkut[0.041]; Gert[0.040]; Schepens[0.040]; Filip[0.040]; Pascal[0.040]; bit[0.040]; Paul[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marc Degryse\u001b[39m ==> ENTITY: \u001b[32mMarc Degryse\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.079:0.079[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.045]; Belgium[0.045]; Tayfun[0.043]; minutes[0.043]; minutes[0.043]; Dirk[0.043]; Verheyen[0.042]; came[0.042]; Sukur[0.042]; Belgian[0.041]; Van[0.041]; Van[0.041]; Van[0.041]; Van[0.041]; time[0.041]; Saffet[0.041]; coach[0.041]; coach[0.041]; fumbled[0.041]; Wilfried[0.041]; Claeys[0.041]; ball[0.041]; ball[0.041]; Korkut[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tayfun Korkut\u001b[39m ==> ENTITY: \u001b[32mTayfun Korkut\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUnsal[0.045]; Yalcin[0.044]; Saffet[0.044]; Turkey[0.043]; Turkey[0.043]; coach[0.043]; coach[0.043]; Sergen[0.043]; Turkish[0.042]; Italy[0.042]; Kafkas[0.041]; Cetin[0.041]; Cetin[0.041]; Dirk[0.041]; Alpay[0.041]; Verheyen[0.041]; Belgium[0.041]; Belgium[0.041]; Orhan[0.041]; friendlies[0.040]; Sukur[0.040]; April[0.040]; Hakan[0.040]; Hakan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hakan Sukur\u001b[39m ==> ENTITY: \u001b[32mHakan Şükür\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHakan[0.044]; Yalcin[0.044]; Turkey[0.044]; Turkey[0.044]; Italy[0.042]; friendlies[0.042]; Tayfun[0.042]; Sergen[0.042]; Alpay[0.042]; Dirk[0.042]; Turkish[0.042]; goal[0.041]; Verheyen[0.041]; Belgium[0.041]; Belgium[0.041]; Kafkas[0.041]; Saffet[0.041]; Cetin[0.041]; Cetin[0.041]; Orhan[0.041]; started[0.041]; Unsal[0.040]; Filip[0.040]; suffered[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arif Erdem\u001b[39m ==> ENTITY: \u001b[32mArif Erdem\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.120:0.120[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.044]; seven[0.043]; Yalcin[0.043]; Sergen[0.043]; Dirk[0.043]; time[0.043]; Verheyen[0.043]; Hakan[0.042]; finals[0.042]; second[0.041]; goal[0.041]; goal[0.041]; Euro[0.041]; fumbled[0.041]; Claeys[0.041]; qualifying[0.041]; ball[0.041]; ball[0.041]; Recep[0.041]; Orhan[0.041]; Gert[0.041]; Schepens[0.041]; June[0.041]; came[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Filip De Wilde\u001b[39m ==> ENTITY: \u001b[32mFilip De Wilde\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.048]; match[0.044]; seven[0.043]; minutes[0.042]; minutes[0.042]; minutes[0.042]; minutes[0.042]; minutes[0.042]; sent[0.042]; Marc[0.041]; goalkeeper[0.041]; goalkeeper[0.041]; Dirk[0.041]; Belgian[0.041]; half[0.041]; half[0.041]; half[0.041]; defender[0.041]; team[0.041]; fans[0.041]; referee[0.041]; penalty[0.041]; scoring[0.040]; took[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dirk Medved\u001b[39m ==> ENTITY: \u001b[32mDirk Medved\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.057:0.057[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.044]; Belgium[0.044]; Belgium[0.044]; Tayfun[0.043]; Verheyen[0.043]; finals[0.043]; Sukur[0.042]; Belgian[0.042]; Saffet[0.041]; coach[0.041]; fumbled[0.041]; Wilfried[0.041]; Claeys[0.041]; France[0.041]; ball[0.041]; ball[0.041]; Korkut[0.041]; Gert[0.041]; Schepens[0.041]; Filip[0.041]; winners[0.041]; decisive[0.041]; beaten[0.040]; Sergen[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Luc Nilis\u001b[39m ==> ENTITY: \u001b[32mLuc Nilis\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.122:0.122[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.044]; Belgium[0.044]; started[0.043]; second[0.043]; Van[0.043]; Van[0.043]; Van[0.043]; Van[0.043]; Tayfun[0.042]; Dirk[0.042]; Marc[0.042]; Verheyen[0.042]; Sukur[0.041]; came[0.041]; goal[0.041]; goal[0.041]; suffered[0.041]; ball[0.041]; ball[0.041]; Yalcin[0.040]; Sergen[0.040]; coach[0.040]; coach[0.040]; Saffet[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Luis Oliveira\u001b[39m ==> ENTITY: \u001b[32mLuís Oliveira\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.044]; Belgium[0.044]; Belgium[0.044]; April[0.043]; second[0.043]; Degryse[0.043]; Degryse[0.043]; Tayfun[0.042]; Dirk[0.042]; Verheyen[0.042]; Sukur[0.042]; Teams[0.041]; coach[0.041]; coach[0.041]; started[0.041]; Belgian[0.041]; minutes[0.041]; minutes[0.041]; time[0.041]; Saffet[0.041]; fumbled[0.040]; Wilfried[0.040]; Claeys[0.040]; Enzo[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Turkey ==> ENTITY: \u001b[32mTurkey national football team\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.126:0.126[0]; log p(e|m)= -3.037:-3.037[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nqualifying[0.050]; Euro[0.045]; Belgium[0.045]; Belgium[0.045]; finals[0.044]; seven[0.044]; seven[0.044]; winners[0.044]; left[0.044]; past[0.044]; defender[0.043]; score[0.043]; failing[0.043]; David[0.043]; earlier[0.043]; upset[0.043]; half[0.042]; barely[0.042]; Marc[0.042]; France[0.042]; slipped[0.042]; scoring[0.042]; right[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saffet Sancakli\u001b[39m ==> ENTITY: \u001b[32mSaffet Sancaklı\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfriendlies[0.046]; Turkey[0.045]; Turkey[0.045]; Turkish[0.043]; Tayfun[0.042]; Dirk[0.042]; Verheyen[0.042]; Unsal[0.042]; Cetin[0.042]; Cetin[0.042]; Belgium[0.042]; Belgium[0.042]; Yalcin[0.041]; Sukur[0.041]; Hakan[0.041]; Hakan[0.041]; Alpay[0.041]; Kafkas[0.041]; Italy[0.040]; Orhan[0.040]; Sergen[0.040]; coach[0.040]; coach[0.040]; goal[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Orhan Cikirikci\u001b[39m ==> ENTITY: \u001b[32mOrhan Çıkırıkçı\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.046]; Turkey[0.046]; friendlies[0.045]; Turkish[0.043]; Hakan[0.043]; Hakan[0.043]; Unsal[0.042]; Yalcin[0.041]; Recep[0.041]; Sergen[0.041]; Tayfun[0.041]; Cetin[0.041]; Cetin[0.041]; Kafkas[0.041]; Saffet[0.041]; Alpay[0.041]; April[0.041]; coach[0.040]; coach[0.040]; Dirk[0.040]; Verheyen[0.040]; Belgium[0.040]; Belgium[0.040]; Italy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Orhan Cikirikci\u001b[39m ==> ENTITY: \u001b[32mOrhan Çıkırıkçı\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.149:0.149[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.048]; Hakan[0.045]; Unsal[0.044]; Yalcin[0.043]; Recep[0.043]; Sergen[0.043]; Euro[0.043]; Cetin[0.043]; Dirk[0.042]; June[0.042]; Verheyen[0.042]; Belgium[0.041]; Belgium[0.041]; Belgium[0.041]; finals[0.041]; qualifying[0.040]; Erdem[0.040]; fumbled[0.040]; Claeys[0.040]; Teams[0.040]; ball[0.040]; ball[0.040]; Gert[0.040]; Schepens[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pascal Renier\u001b[39m ==> ENTITY: \u001b[32mPascal Renier\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.120:0.120[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.048]; Belgium[0.048]; Belgium[0.048]; Belgian[0.045]; France[0.044]; Gilles[0.043]; Marc[0.042]; Tayfun[0.041]; Dirk[0.041]; Verheyen[0.041]; Sukur[0.040]; Luc[0.040]; Bertrand[0.040]; finals[0.040]; bar[0.040]; Degryse[0.040]; Degryse[0.040]; Saffet[0.040]; coach[0.039]; fumbled[0.039]; Wilfried[0.039]; Claeys[0.039]; ball[0.039]; ball[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgium ==> ENTITY: \u001b[32mBelgium national football team\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.131:0.131[0]; log p(e|m)= -2.882:-2.882[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgian[0.045]; Euro[0.044]; Turkey[0.043]; seven[0.043]; took[0.042]; given[0.042]; defender[0.042]; second[0.042]; opponent[0.042]; June[0.041]; minutes[0.041]; minutes[0.041]; minutes[0.041]; minutes[0.041]; minutes[0.041]; Degryse[0.041]; referee[0.041]; substitute[0.041]; goalkeeper[0.041]; goalkeeper[0.041]; volley[0.041]; goal[0.041]; order[0.041]; ball[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gilles De Bilde\u001b[39m ==> ENTITY: \u001b[32mGilles De Bilde\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.065:0.065[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsuffered[0.043]; Tayfun[0.043]; Dirk[0.043]; Verheyen[0.043]; Belgium[0.042]; Belgium[0.042]; Belgium[0.042]; coach[0.042]; coach[0.042]; Sukur[0.042]; Belgian[0.042]; Saffet[0.041]; Van[0.041]; Van[0.041]; Van[0.041]; Van[0.041]; Van[0.041]; fumbled[0.041]; Wilfried[0.041]; Claeys[0.041]; ball[0.041]; ball[0.041]; Korkut[0.041]; Gert[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mTurkey\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mTurkey\u001b[39m <---> \u001b[32mTurkey national football team\u001b[39m\t\nSCORES: global= 0.250:0.248[\u001b[31m0.002\u001b[39m]; local()= 0.190:0.182[\u001b[31m0.008\u001b[39m]; log p(e|m)= -0.209:-3.037[\u001b[31m2.827\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.052]; Turkish[0.047]; Belgium[0.043]; Belgium[0.043]; Belgium[0.043]; Sergen[0.042]; second[0.042]; Alpay[0.042]; Kafkas[0.042]; Yalcin[0.041]; Tayfun[0.041]; Orhan[0.041]; Orhan[0.041]; Ercan[0.041]; Saffet[0.040]; Hakan[0.040]; Hakan[0.040]; coach[0.040]; coach[0.040]; Unsal[0.040]; Cetin[0.040]; Cetin[0.040]; Teams[0.040]; Recep[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = 1998 World Cup\u001b[39m ==> ENTITY: \u001b[32m1998 FIFA World Cup\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.127:0.127[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.047]; match[0.047]; team[0.045]; Belgium[0.044]; Belgium[0.044]; second[0.044]; Turkey[0.042]; Turkey[0.042]; campaign[0.042]; crowd[0.041]; marred[0.041]; took[0.041]; fans[0.041]; victory[0.041]; seats[0.040]; substitute[0.040]; half[0.040]; half[0.040]; half[0.040]; upset[0.040]; minutes[0.040]; shortly[0.039]; hopes[0.039]; given[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgian ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.124:-0.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.053]; Belgium[0.053]; friendlies[0.045]; Italy[0.044]; Turkey[0.042]; Turkey[0.042]; said[0.041]; said[0.041]; coach[0.041]; coach[0.041]; Bertrand[0.040]; Claeys[0.040]; Van[0.040]; Van[0.040]; Van[0.040]; Van[0.040]; Van[0.040]; Degryse[0.040]; Gilles[0.040]; expected[0.040]; goal[0.040]; Luis[0.040]; Marc[0.039]; Filip[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wilfried Van Moer\u001b[39m ==> ENTITY: \u001b[32mWilfried Van Moer\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.099:0.099[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.046]; Belgium[0.046]; friendlies[0.044]; coach[0.043]; coach[0.043]; Tayfun[0.042]; Dirk[0.042]; Verheyen[0.042]; Gilles[0.041]; Sukur[0.041]; goal[0.041]; Belgian[0.041]; Van[0.041]; Van[0.041]; Van[0.041]; Van[0.041]; Bertrand[0.041]; victory[0.041]; Luc[0.041]; Saffet[0.040]; Claeys[0.040]; Korkut[0.040]; Filip[0.040]; Gert[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Turkey ==> ENTITY: \u001b[32mTurkey national football team\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.155:0.155[0]; log p(e|m)= -3.037:-3.037[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.052]; Turkish[0.047]; Turkish[0.047]; second[0.044]; team[0.044]; match[0.044]; Belgium[0.042]; Belgium[0.042]; victory[0.042]; Soccer[0.042]; Cup[0.042]; break[0.041]; trouble[0.041]; substitute[0.041]; upset[0.040]; half[0.040]; half[0.040]; dozens[0.039]; campaign[0.039]; shortly[0.039]; fans[0.039]; marred[0.038]; minutes[0.038]; took[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgium ==> ENTITY: \u001b[32mBelgium national football team\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.175:0.175[0]; log p(e|m)= -2.882:-2.882[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.051]; Belgium[0.051]; France[0.047]; finals[0.044]; Van[0.044]; Marc[0.042]; Euro[0.042]; qualifying[0.042]; second[0.042]; Gilles[0.041]; Turkey[0.041]; seven[0.041]; Luc[0.040]; winners[0.040]; time[0.040]; June[0.039]; minutes[0.039]; minutes[0.039]; Dirk[0.039]; Degryse[0.039]; Degryse[0.039]; Filip[0.039]; goal[0.039]; goal[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = English ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.074:0.074[0]; log p(e|m)= -1.355:-1.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.046]; David[0.045]; second[0.044]; Elleray[0.043]; born[0.043]; seven[0.042]; Marc[0.042]; slipped[0.042]; referee[0.041]; half[0.041]; half[0.041]; half[0.041]; half[0.041]; area[0.041]; area[0.041]; earlier[0.041]; sent[0.041]; took[0.041]; turned[0.041]; ripped[0.040]; substitute[0.040]; barely[0.040]; Turkish[0.040]; scoring[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Geoffrey Claeys\u001b[39m ==> ENTITY: \u001b[32mGeoffrey Claeys\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.047]; Belgium[0.047]; Belgium[0.047]; Belgian[0.043]; started[0.042]; Tayfun[0.042]; Dirk[0.041]; Filip[0.041]; Verheyen[0.041]; coach[0.041]; Marc[0.041]; second[0.041]; Nico[0.041]; Sukur[0.041]; Degryse[0.041]; Degryse[0.041]; finals[0.041]; Bertrand[0.041]; minutes[0.040]; minutes[0.040]; ball[0.040]; ball[0.040]; Luc[0.040]; Kafkas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hakan Unsal\u001b[39m ==> ENTITY: \u001b[32mHakan Ünsal\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nstarted[0.045]; Turkey[0.043]; Turkey[0.043]; Hakan[0.043]; Tayfun[0.042]; Dirk[0.042]; Sergen[0.042]; Verheyen[0.042]; Turkish[0.042]; friendlies[0.042]; came[0.042]; ball[0.042]; ball[0.042]; Sukur[0.042]; minutes[0.041]; second[0.041]; Yalcin[0.041]; Saffet[0.041]; said[0.041]; said[0.041]; Alpay[0.041]; coach[0.040]; coach[0.040]; goal[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Recep Cetin\u001b[39m ==> ENTITY: \u001b[32mRecep Çetin\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.045]; Turkey[0.045]; Saffet[0.043]; Turkish[0.043]; Tayfun[0.042]; Dirk[0.042]; Orhan[0.042]; Orhan[0.042]; Verheyen[0.042]; Unsal[0.042]; Sukur[0.041]; Alpay[0.041]; Kafkas[0.041]; started[0.041]; coach[0.041]; coach[0.041]; Hakan[0.041]; Hakan[0.041]; Yalcin[0.041]; Wilfried[0.040]; Claeys[0.040]; ball[0.040]; ball[0.040]; Korkut[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ogun Temizkanoglu\u001b[39m ==> ENTITY: \u001b[32mOgün Temizkanoğlu\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.106:0.106[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.048]; defender[0.045]; Dirk[0.044]; referee[0.043]; born[0.043]; June[0.043]; finals[0.043]; Belgium[0.042]; Belgium[0.042]; Orhan[0.042]; fumbled[0.042]; ball[0.042]; goalkeeper[0.041]; Erdem[0.041]; English[0.040]; incisive[0.040]; Arif[0.040]; group[0.040]; Elleray[0.040]; sent[0.040]; angled[0.040]; penalty[0.040]; Euro[0.040]; strike[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mTurkey\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mTurkey\u001b[39m <---> \u001b[32mTurkey national football team\u001b[39m\t\nSCORES: global= 0.250:0.249[\u001b[31m0.002\u001b[39m]; local()= 0.188:0.186[\u001b[31m0.002\u001b[39m]; log p(e|m)= -0.209:-3.037[\u001b[31m2.827\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.052]; Italy[0.047]; Turkish[0.047]; friendlies[0.044]; Belgium[0.043]; Sergen[0.042]; Alpay[0.042]; Kafkas[0.042]; Yalcin[0.041]; Tayfun[0.041]; Orhan[0.041]; Ercan[0.041]; Saffet[0.040]; Hakan[0.040]; Hakan[0.040]; Unsal[0.040]; Cetin[0.040]; Cetin[0.040]; coach[0.040]; coach[0.040]; Recep[0.040]; Russia[0.039]; April[0.039]; victory[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgian ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.112:0.112[0]; log p(e|m)= -0.124:-0.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.053]; match[0.046]; Cup[0.043]; English[0.043]; team[0.043]; defender[0.042]; hopes[0.042]; second[0.041]; Brazilian[0.041]; Marc[0.041]; victory[0.041]; Turkey[0.041]; took[0.041]; minutes[0.040]; minutes[0.040]; minutes[0.040]; minutes[0.040]; Degryse[0.040]; born[0.040]; referee[0.040]; substitute[0.040]; half[0.040]; half[0.040]; half[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Degryse\u001b[39m ==> ENTITY: \u001b[32mMarc Degryse\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.113:0.113[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.045]; Belgium[0.045]; Belgium[0.045]; France[0.042]; Tayfun[0.042]; minutes[0.042]; minutes[0.042]; Dirk[0.042]; Verheyen[0.042]; Marc[0.042]; came[0.042]; finals[0.041]; Van[0.041]; time[0.041]; seven[0.041]; fumbled[0.041]; Claeys[0.040]; ball[0.040]; ball[0.040]; June[0.040]; Korkut[0.040]; Degryse[0.040]; half[0.040]; qualifying[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brussels ==> ENTITY: \u001b[32mBrussels\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.025:-0.025[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.048]; Belgium[0.048]; match[0.046]; Soccer[0.045]; Cup[0.044]; took[0.044]; second[0.043]; team[0.043]; marred[0.042]; World[0.041]; minutes[0.041]; victory[0.041]; fans[0.041]; kicked[0.040]; campaign[0.040]; order[0.040]; break[0.040]; given[0.039]; shortly[0.039]; Turkish[0.039]; Turkish[0.039]; Turkish[0.039]; Crowd[0.039]; restore[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gunther Schepens\u001b[39m ==> ENTITY: \u001b[32mGunther Schepens\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.121:0.121[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDegryse[0.043]; Degryse[0.043]; Belgium[0.043]; Belgium[0.043]; Belgium[0.043]; Dirk[0.043]; Tayfun[0.042]; Verheyen[0.042]; finals[0.042]; ball[0.042]; ball[0.042]; Belgian[0.042]; started[0.042]; Pascal[0.042]; Marc[0.041]; Sukur[0.041]; goal[0.041]; Oliveira[0.041]; Saffet[0.040]; second[0.040]; coach[0.040]; fumbled[0.040]; Wilfried[0.040]; Claeys[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dirk Medved\u001b[39m ==> ENTITY: \u001b[32mDirk Medved\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.091:0.091[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.045]; defender[0.044]; born[0.043]; Belgian[0.043]; ball[0.042]; Filip[0.042]; goalkeeper[0.042]; goalkeeper[0.042]; Euro[0.042]; Elleray[0.042]; referee[0.041]; English[0.041]; Sergen[0.041]; penalty[0.041]; right[0.041]; goal[0.041]; incisive[0.041]; minutes[0.041]; minutes[0.041]; minutes[0.041]; minutes[0.041]; June[0.041]; past[0.041]; angled[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ogun Temizkanoglu\u001b[39m ==> ENTITY: \u001b[32mOgün Temizkanoğlu\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.049]; Turkey[0.049]; friendlies[0.048]; Unsal[0.047]; Turkish[0.046]; Cetin[0.046]; Cetin[0.046]; Saffet[0.046]; Tayfun[0.046]; Kafkas[0.046]; Hakan[0.045]; Hakan[0.045]; Alpay[0.045]; Dirk[0.044]; Recep[0.044]; Verheyen[0.044]; Yalcin[0.044]; Sergen[0.044]; April[0.044]; Sukur[0.044]; coach[0.043]; coach[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alpay Ozalan\u001b[39m ==> ENTITY: \u001b[32mAlpay Özalan\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.129:0.129[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.046]; Turkey[0.046]; Sergen[0.044]; Turkish[0.044]; second[0.043]; friendlies[0.043]; Yalcin[0.042]; Tayfun[0.042]; Dirk[0.042]; Verheyen[0.042]; Italy[0.041]; Sukur[0.041]; suffered[0.041]; Hakan[0.041]; Hakan[0.041]; half[0.041]; ball[0.041]; ball[0.041]; Unsal[0.040]; Saffet[0.040]; Kafkas[0.040]; coach[0.040]; coach[0.040]; Wilfried[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Euro 96\u001b[39m ==> ENTITY: \u001b[32mUEFA Euro 1996\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nqualifying[0.046]; second[0.045]; finals[0.044]; winners[0.044]; Belgium[0.043]; Belgium[0.043]; Belgium[0.043]; Turkey[0.042]; score[0.042]; seven[0.042]; seven[0.042]; France[0.041]; Dirk[0.041]; penalty[0.041]; group[0.040]; goal[0.040]; goal[0.040]; automatically[0.040]; minutes[0.040]; minutes[0.040]; minutes[0.040]; minutes[0.040]; time[0.040]; defender[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Van Moer\u001b[39m ==> ENTITY: \u001b[32mWilfried Van Moer\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.092:0.092[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.046]; friendlies[0.045]; coach[0.044]; coach[0.044]; Tayfun[0.043]; Verheyen[0.042]; Gilles[0.042]; Sukur[0.042]; goal[0.042]; Belgian[0.042]; Van[0.042]; Van[0.042]; Van[0.042]; victory[0.041]; Luc[0.041]; Saffet[0.041]; Wilfried[0.041]; Korkut[0.041]; Gert[0.040]; Moer[0.040]; Moer[0.040]; started[0.040]; Sergen[0.040]; suffered[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paul Van Himst\u001b[39m ==> ENTITY: \u001b[32mPaul Van Himst\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.129:0.129[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.045]; coach[0.044]; coach[0.044]; Van[0.044]; Van[0.044]; Van[0.044]; victory[0.044]; friendlies[0.043]; Belgian[0.043]; Tayfun[0.042]; Degryse[0.042]; Gilles[0.041]; Marc[0.041]; Sukur[0.041]; Luc[0.041]; Saffet[0.040]; Wilfried[0.040]; started[0.040]; goal[0.040]; Korkut[0.040]; Bilde[0.039]; April[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBelgium\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBelgium\u001b[39m <---> \u001b[32mBelgium national football team\u001b[39m\t\nSCORES: global= 0.250:0.249[\u001b[31m0.001\u001b[39m]; local()= 0.166:0.174[\u001b[32m0.008\u001b[39m]; log p(e|m)= -0.242:-2.882[\u001b[31m2.640\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.051]; Belgium[0.051]; France[0.047]; finals[0.044]; Belgian[0.044]; Van[0.044]; Marc[0.042]; qualifying[0.042]; second[0.042]; Gilles[0.041]; Turkey[0.041]; seven[0.041]; Luc[0.040]; winners[0.040]; coach[0.040]; minutes[0.039]; minutes[0.039]; Dirk[0.039]; Degryse[0.039]; Degryse[0.039]; Filip[0.039]; goal[0.039]; Bertrand[0.039]; group[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = David Elleray\u001b[39m ==> ENTITY: \u001b[32mDavid Elleray\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.086:0.086[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nreferee[0.048]; penalty[0.046]; Dirk[0.045]; given[0.044]; second[0.044]; substitute[0.044]; English[0.044]; goalkeeper[0.044]; goalkeeper[0.044]; Belgium[0.044]; ball[0.043]; defender[0.043]; past[0.043]; seven[0.043]; Filip[0.043]; pride[0.043]; score[0.043]; Sergen[0.042]; half[0.042]; half[0.042]; half[0.042]; incisive[0.042]; took[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nico Van Kerckhoven\u001b[39m ==> ENTITY: \u001b[32mNico Van Kerckhoven\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.106:0.106[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.046]; Belgium[0.046]; Belgium[0.046]; Van[0.045]; Belgian[0.044]; coach[0.042]; Tayfun[0.042]; Dirk[0.042]; Verheyen[0.042]; Luc[0.042]; Marc[0.041]; Sukur[0.041]; finals[0.041]; Gilles[0.040]; France[0.040]; Saffet[0.040]; Pascal[0.040]; fumbled[0.040]; Wilfried[0.040]; Claeys[0.040]; ball[0.040]; ball[0.040]; Korkut[0.040]; Teams[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBelgium\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBelgium\u001b[39m <---> \u001b[32mBelgium national football team\u001b[39m\t\nSCORES: global= 0.249:0.249[\u001b[31m0.000\u001b[39m]; local()= 0.103:0.120[\u001b[32m0.016\u001b[39m]; log p(e|m)= -0.242:-2.882[\u001b[31m2.640\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.053]; team[0.047]; match[0.046]; Soccer[0.045]; Cup[0.044]; victory[0.043]; Turkey[0.042]; Turkey[0.042]; took[0.042]; given[0.042]; minutes[0.041]; Lauwers[0.041]; order[0.040]; Brussels[0.040]; upset[0.040]; World[0.040]; marred[0.039]; Turkish[0.039]; Turkish[0.039]; crowd[0.039]; fans[0.039]; seats[0.039]; campaign[0.039]; Bert[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgium ==> ENTITY: \u001b[32mBelgium national football team\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.124:0.124[0]; log p(e|m)= -2.882:-2.882[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.053]; team[0.047]; match[0.046]; Soccer[0.045]; Cup[0.043]; second[0.043]; victory[0.043]; Turkey[0.042]; Turkey[0.042]; took[0.041]; given[0.041]; minutes[0.040]; Lauwers[0.040]; substitute[0.040]; order[0.040]; Brussels[0.040]; upset[0.040]; World[0.039]; marred[0.039]; Turkish[0.039]; Turkish[0.039]; Turkish[0.039]; crowd[0.039]; fans[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Abdullah Ercan\u001b[39m ==> ENTITY: \u001b[32mAbdullah Ercan\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.209:0.209[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurkey[0.044]; Turkey[0.044]; coach[0.044]; coach[0.044]; friendlies[0.044]; Alpay[0.043]; Hakan[0.042]; Hakan[0.042]; started[0.042]; Cetin[0.041]; Cetin[0.041]; Kafkas[0.041]; Tayfun[0.041]; Saffet[0.041]; Turkish[0.041]; Dirk[0.041]; Orhan[0.041]; Verheyen[0.041]; Erdem[0.041]; Yalcin[0.041]; Sukur[0.040]; Recep[0.040]; Unsal[0.039]; Oguz[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tolunay Kafkas\u001b[39m ==> ENTITY: \u001b[32mTolunay Kafkas\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.163:0.163[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfriendlies[0.047]; coach[0.044]; coach[0.044]; Turkey[0.043]; Turkey[0.043]; Saffet[0.043]; Tayfun[0.042]; Alpay[0.042]; Dirk[0.042]; Verheyen[0.041]; Italy[0.041]; Hakan[0.041]; Hakan[0.041]; Belgium[0.041]; Belgium[0.041]; Belgium[0.041]; Yalcin[0.041]; Sukur[0.041]; Turkish[0.041]; Sergen[0.040]; April[0.040]; started[0.040]; Wilfried[0.040]; Claeys[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.048]; Belgium[0.048]; Belgium[0.048]; finals[0.048]; Euro[0.046]; second[0.045]; qualifying[0.044]; Turkey[0.043]; came[0.043]; half[0.043]; half[0.043]; seven[0.042]; seven[0.042]; time[0.042]; Bertrand[0.042]; upset[0.042]; single[0.042]; afford[0.042]; goal[0.042]; goal[0.042]; Van[0.042]; decisive[0.041]; score[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBelgium\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBelgium\u001b[39m <---> \u001b[32mBelgium national football team\u001b[39m\t\nSCORES: global= 0.247:0.246[\u001b[31m0.001\u001b[39m]; local()= 0.113:0.135[\u001b[32m0.023\u001b[39m]; log p(e|m)= -0.242:-2.882[\u001b[31m2.640\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfriendlies[0.045]; Van[0.045]; Van[0.045]; Van[0.045]; Van[0.045]; Belgian[0.044]; Italy[0.044]; victory[0.043]; Gilles[0.043]; Turkey[0.042]; Turkey[0.042]; coach[0.041]; coach[0.041]; April[0.040]; Paul[0.040]; goal[0.040]; started[0.040]; Wilfried[0.040]; Turkish[0.039]; Russia[0.039]; Sergen[0.039]; suffered[0.039]; draws[0.039]; Luis[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Van Moer\u001b[39m ==> ENTITY: \u001b[32mWilfried Van Moer\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.094:0.094[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.047]; friendlies[0.045]; coach[0.044]; coach[0.044]; Tayfun[0.043]; Gilles[0.042]; Sukur[0.042]; goal[0.042]; Belgian[0.042]; Van[0.042]; Van[0.042]; Van[0.042]; victory[0.041]; Saffet[0.041]; Wilfried[0.041]; Korkut[0.041]; Moer[0.040]; Moer[0.040]; started[0.040]; Sergen[0.040]; suffered[0.040]; April[0.040]; Italy[0.040]; expected[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Filip De Wilde\u001b[39m ==> ENTITY: \u001b[32mFilip De Wilde\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.048]; Belgium[0.048]; Belgium[0.048]; France[0.043]; seven[0.043]; Gilles[0.042]; minutes[0.042]; minutes[0.042]; time[0.041]; Pascal[0.041]; Marc[0.041]; Tayfun[0.041]; Dirk[0.041]; Belgian[0.041]; coach[0.041]; half[0.041]; Luc[0.041]; Verheyen[0.041]; Sukur[0.040]; Bertrand[0.040]; finals[0.040]; Degryse[0.040]; Degryse[0.040]; came[0.039]; \t\n\r [====================================>.........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 4s577ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3867/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 997testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Yugoslavia ==> ENTITY: \u001b[32mYugoslavia\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.171:0.171[0]; log p(e|m)= -0.654:-0.654[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYugoslavia[0.051]; Yugoslavia[0.051]; Yugoslavia[0.051]; Poland[0.048]; country[0.045]; Republic[0.043]; Warsaw[0.043]; countries[0.042]; Poles[0.042]; Polish[0.041]; September[0.041]; clinch[0.040]; trip[0.039]; saying[0.039]; granting[0.039]; revive[0.039]; Friday[0.039]; agency[0.038]; agency[0.038]; officials[0.038]; culture[0.038]; Ministry[0.038]; governments[0.038]; agreement[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rosati\u001b[39m ==> ENTITY: \u001b[32mDariusz Rosati\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.046]; Poland[0.046]; Yugoslavia[0.043]; Yugoslavia[0.043]; Yugoslavia[0.043]; Yugoslavia[0.043]; President[0.043]; Dariusz[0.043]; Warsaw[0.042]; Polish[0.042]; governments[0.041]; September[0.041]; Minister[0.041]; Minister[0.041]; Poles[0.040]; Republic[0.040]; countries[0.040]; Rosati[0.040]; Rosati[0.040]; news[0.040]; Montenegro[0.040]; Yugoslav[0.040]; visas[0.040]; visas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgrade ==> ENTITY: \u001b[32mBelgrade\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.149:0.149[0]; log p(e|m)= -0.052:-0.052[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSerbian[0.045]; Yugoslavia[0.045]; Yugoslavia[0.045]; Yugoslavia[0.045]; Belgrade[0.044]; Serbs[0.044]; Poland[0.043]; Poland[0.043]; Yugoslav[0.043]; Bosnian[0.042]; Slobodan[0.042]; clinch[0.041]; fighting[0.041]; Milosevic[0.041]; country[0.040]; level[0.040]; Montenegro[0.040]; diplomatic[0.040]; officials[0.040]; Republic[0.040]; April[0.040]; Poles[0.039]; wants[0.039]; support[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgrade ==> ENTITY: \u001b[32mBelgrade\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.152:0.152[0]; log p(e|m)= -0.052:-0.052[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSerbian[0.045]; Yugoslavia[0.044]; Yugoslavia[0.044]; Yugoslavia[0.044]; Belgrade[0.044]; Serbs[0.044]; Poland[0.043]; Poland[0.043]; Yugoslav[0.043]; Bosnian[0.042]; Slobodan[0.042]; clinch[0.041]; fighting[0.041]; Milosevic[0.041]; country[0.040]; level[0.040]; Montenegro[0.040]; Friday[0.040]; diplomatic[0.040]; officials[0.040]; Republic[0.040]; April[0.040]; Poles[0.039]; countries[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Serbian ==> ENTITY: \u001b[32mSerbia\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.163:0.163[0]; log p(e|m)= -1.252:-1.252[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYugoslavia[0.046]; Yugoslavia[0.046]; Yugoslavia[0.046]; Yugoslavia[0.046]; Montenegro[0.045]; Republic[0.044]; country[0.043]; Yugoslav[0.042]; Serbs[0.042]; Poland[0.042]; Poland[0.042]; Bosnian[0.041]; Belgrade[0.041]; Belgrade[0.041]; links[0.040]; Slobodan[0.040]; saying[0.040]; culture[0.040]; governments[0.040]; Poles[0.039]; ties[0.039]; wants[0.039]; politicians[0.039]; understanding[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mForeign Ministry\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mForeign minister\u001b[39m <---> \u001b[32mMinistry of Foreign Affairs (Poland)\u001b[39m\t\nSCORES: global= 0.243:0.239[\u001b[31m0.004\u001b[39m]; local()= 0.149:0.181[\u001b[32m0.031\u001b[39m]; log p(e|m)= -0.355:-3.037[\u001b[31m2.681\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nforeign[0.046]; Foreign[0.046]; Foreign[0.046]; Warsaw[0.046]; Poland[0.045]; Poland[0.045]; Poland[0.045]; diplomatic[0.044]; Yugoslavia[0.043]; Yugoslavia[0.043]; Yugoslavia[0.043]; Yugoslavia[0.043]; Yugoslavia[0.043]; countries[0.043]; Polish[0.043]; Federal[0.043]; ambassadorial[0.042]; ties[0.042]; country[0.042]; Yugoslav[0.042]; President[0.042]; Republic[0.042]; officials[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Yugoslavia ==> ENTITY: \u001b[32mYugoslavia\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.179:0.179[0]; log p(e|m)= -0.654:-0.654[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYugoslavia[0.051]; Yugoslavia[0.051]; Yugoslavia[0.051]; Poland[0.047]; country[0.045]; Serbian[0.043]; Republic[0.043]; Warsaw[0.042]; countries[0.042]; Poles[0.041]; Slobodan[0.041]; Polish[0.041]; September[0.040]; trip[0.039]; saying[0.039]; granting[0.039]; foreign[0.038]; revive[0.038]; agency[0.038]; agency[0.038]; Friday[0.038]; officials[0.038]; news[0.038]; culture[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.130:0.130[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.049]; Montenegro[0.046]; country[0.045]; Yugoslavia[0.044]; Yugoslavia[0.044]; Yugoslavia[0.044]; Poles[0.043]; Republic[0.043]; economic[0.042]; despite[0.041]; Belgrade[0.041]; Belgrade[0.041]; Serbian[0.041]; trip[0.040]; President[0.040]; Yugoslav[0.040]; sign[0.040]; protection[0.040]; countries[0.040]; trade[0.040]; Friday[0.040]; currently[0.040]; visiting[0.040]; imposed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Slobodan Milosevic\u001b[39m ==> ENTITY: \u001b[32mSlobodan Milošević\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYugoslavia[0.045]; Yugoslavia[0.045]; Yugoslavia[0.045]; Yugoslavia[0.045]; Yugoslav[0.044]; Bosnian[0.043]; Belgrade[0.042]; Belgrade[0.042]; Serbian[0.042]; Republic[0.041]; Serbs[0.041]; President[0.041]; Montenegro[0.041]; ties[0.041]; country[0.040]; saying[0.040]; news[0.040]; despite[0.040]; support[0.040]; September[0.040]; politicians[0.040]; cooperation[0.040]; Poland[0.040]; Poland[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Federal Republic of Yugoslavia ==> ENTITY: \u001b[32mSocialist Federal Republic of Yugoslavia\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.211:0.211[0]; log p(e|m)= -0.687:-0.687[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYugoslavia[0.048]; Yugoslavia[0.048]; Yugoslavia[0.048]; Yugoslavia[0.048]; Poland[0.045]; Poland[0.045]; Poland[0.045]; Montenegro[0.043]; Yugoslav[0.042]; country[0.042]; Belgrade[0.041]; Belgrade[0.041]; Serbian[0.041]; Bosnian[0.040]; Warsaw[0.040]; Serbs[0.040]; countries[0.039]; Poles[0.039]; despite[0.038]; Slobodan[0.038]; Polish[0.038]; diplomatic[0.038]; ties[0.037]; level[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Polish ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.814:-0.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.049]; Poles[0.046]; Warsaw[0.045]; country[0.045]; Yugoslavia[0.044]; Yugoslavia[0.044]; Yugoslavia[0.044]; Yugoslavia[0.044]; Republic[0.043]; Foreign[0.041]; Foreign[0.041]; Foreign[0.041]; countries[0.041]; September[0.040]; Dariusz[0.040]; currently[0.039]; visit[0.039]; visit[0.039]; trip[0.039]; Friday[0.039]; news[0.039]; reported[0.039]; Ministry[0.039]; governments[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.137:0.137[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPolish[0.046]; Warsaw[0.045]; country[0.045]; Yugoslavia[0.044]; Yugoslavia[0.044]; Yugoslavia[0.044]; Yugoslavia[0.044]; Poles[0.044]; Republic[0.044]; Dariusz[0.042]; granting[0.041]; trip[0.041]; visit[0.040]; visit[0.040]; September[0.040]; countries[0.040]; Friday[0.040]; currently[0.040]; clinch[0.040]; foreign[0.039]; Federal[0.039]; frozen[0.039]; Ministry[0.039]; governments[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Yugoslav ==> ENTITY: \u001b[32mYugoslavia\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.202:0.202[0]; log p(e|m)= -0.759:-0.759[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYugoslavia[0.049]; Yugoslavia[0.049]; Yugoslavia[0.049]; Poland[0.045]; Poland[0.045]; country[0.043]; Poles[0.042]; Belgrade[0.042]; Belgrade[0.042]; Serbs[0.042]; Montenegro[0.042]; Serbian[0.042]; Bosnian[0.041]; Republic[0.041]; culture[0.040]; Slobodan[0.040]; countries[0.040]; revived[0.039]; revive[0.039]; saying[0.039]; economic[0.038]; Milosevic[0.038]; fighting[0.038]; seeking[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Yugoslavia ==> ENTITY: \u001b[32mYugoslavia\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.234:0.234[0]; log p(e|m)= -0.654:-0.654[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYugoslavia[0.048]; Yugoslavia[0.048]; Yugoslavia[0.048]; Yugoslavia[0.048]; Poland[0.044]; Poland[0.044]; Poland[0.044]; Montenegro[0.043]; Yugoslav[0.042]; country[0.042]; Belgrade[0.041]; Belgrade[0.041]; Serbian[0.041]; Republic[0.040]; Bosnian[0.040]; Serbs[0.040]; Warsaw[0.040]; countries[0.039]; Poles[0.039]; Slobodan[0.038]; despite[0.038]; Polish[0.038]; diplomatic[0.037]; clinch[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = PAP ==> ENTITY: \u001b[32mPolish Press Agency\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.146:0.146[0]; log p(e|m)= -2.659:-2.659[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.045]; Poland[0.045]; protocols[0.045]; Warsaw[0.044]; Polish[0.044]; news[0.043]; Yugoslavia[0.042]; Yugoslavia[0.042]; Yugoslavia[0.042]; Yugoslavia[0.042]; protection[0.041]; visit[0.041]; visit[0.041]; country[0.041]; agency[0.041]; agency[0.041]; President[0.041]; Republic[0.041]; Minister[0.041]; Minister[0.041]; September[0.040]; foreign[0.039]; frozen[0.039]; politicians[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.118:0.118[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsaying[0.045]; country[0.043]; granting[0.042]; President[0.042]; officials[0.042]; agency[0.042]; Republic[0.042]; support[0.042]; Yugoslavia[0.042]; Yugoslavia[0.042]; Yugoslavia[0.042]; Foreign[0.042]; ties[0.042]; imposed[0.042]; diplomatic[0.042]; politicians[0.041]; Federal[0.041]; April[0.041]; foreign[0.041]; quoted[0.041]; economic[0.041]; embargo[0.041]; protocols[0.041]; cooperation[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Montenegro ==> ENTITY: \u001b[32mMontenegro\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.165:0.165[0]; log p(e|m)= -0.449:-0.449[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYugoslavia[0.046]; Yugoslavia[0.046]; Yugoslavia[0.046]; country[0.044]; Republic[0.044]; Yugoslav[0.043]; Poland[0.043]; Poland[0.043]; Belgrade[0.042]; Belgrade[0.042]; Serbian[0.042]; Serbs[0.042]; Slobodan[0.041]; President[0.041]; crime[0.041]; economic[0.040]; despite[0.040]; Bosnian[0.040]; protection[0.040]; conditions[0.039]; Milosevic[0.039]; countries[0.039]; culture[0.039]; imposed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Warsaw ==> ENTITY: \u001b[32mWarsaw\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.143:-0.143[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.049]; Polish[0.045]; country[0.043]; Yugoslavia[0.043]; Yugoslavia[0.043]; Yugoslavia[0.043]; Yugoslavia[0.043]; Poles[0.043]; Republic[0.042]; Friday[0.041]; September[0.041]; visit[0.041]; visit[0.041]; Dariusz[0.041]; saying[0.041]; trip[0.041]; reported[0.041]; conditions[0.040]; granting[0.040]; Ministry[0.040]; officials[0.040]; frozen[0.040]; protection[0.040]; currently[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPoles\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPoland\u001b[39m <---> \u001b[32mPoles\u001b[39m\t\nSCORES: global= 0.270:0.253[\u001b[31m0.016\u001b[39m]; local()= 0.168:0.152[\u001b[31m0.016\u001b[39m]; log p(e|m)= -2.273:-0.448[\u001b[32m1.825\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.047]; Poland[0.047]; Poland[0.047]; Polish[0.045]; country[0.043]; Warsaw[0.042]; visas[0.042]; visas[0.042]; foreign[0.041]; Yugoslavia[0.041]; Yugoslavia[0.041]; Yugoslavia[0.041]; Yugoslavia[0.041]; Yugoslavia[0.041]; despite[0.041]; officials[0.040]; countries[0.040]; Serbs[0.040]; granting[0.040]; economic[0.040]; Yugoslav[0.040]; crime[0.039]; Republic[0.039]; visiting[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Poland ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.153:-0.153[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.049]; Montenegro[0.045]; country[0.044]; April[0.044]; Yugoslavia[0.043]; Yugoslavia[0.043]; Yugoslavia[0.043]; Poles[0.043]; Republic[0.043]; visiting[0.041]; economic[0.041]; despite[0.041]; clinch[0.041]; Belgrade[0.041]; Belgrade[0.041]; Serbian[0.041]; support[0.040]; President[0.040]; Yugoslav[0.040]; imposed[0.039]; trade[0.039]; currently[0.039]; wants[0.039]; ties[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dariusz Rosati\u001b[39m ==> ENTITY: \u001b[32mDariusz Rosati\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.047]; Yugoslavia[0.044]; Yugoslavia[0.044]; Yugoslavia[0.044]; Yugoslavia[0.044]; Warsaw[0.043]; Polish[0.043]; governments[0.042]; September[0.042]; Minister[0.042]; Minister[0.042]; Poles[0.041]; Republic[0.041]; countries[0.041]; Rosati[0.041]; Rosati[0.041]; news[0.041]; visas[0.040]; visas[0.040]; protocols[0.040]; conditions[0.039]; Friday[0.039]; ministries[0.039]; Ministry[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Yugoslavia ==> ENTITY: \u001b[32mYugoslavia\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.204:0.204[0]; log p(e|m)= -0.654:-0.654[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYugoslavia[0.050]; Yugoslavia[0.050]; Poland[0.046]; Poland[0.046]; Montenegro[0.044]; Yugoslav[0.043]; country[0.043]; Belgrade[0.043]; Belgrade[0.043]; Serbian[0.042]; Republic[0.042]; Bosnian[0.041]; Serbs[0.041]; countries[0.040]; Poles[0.040]; Slobodan[0.040]; despite[0.040]; ties[0.039]; clinch[0.038]; level[0.038]; Milosevic[0.038]; fighting[0.038]; diplomatic[0.038]; trip[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rosati\u001b[39m ==> ENTITY: \u001b[32mDariusz Rosati\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.048]; Poland[0.048]; economic[0.045]; Yugoslavia[0.045]; Yugoslavia[0.045]; Yugoslavia[0.045]; Yugoslavia[0.045]; President[0.045]; Dariusz[0.044]; trade[0.043]; governments[0.043]; April[0.043]; September[0.043]; Minister[0.043]; Poles[0.042]; cooperation[0.042]; Republic[0.042]; countries[0.042]; Rosati[0.042]; Rosati[0.042]; news[0.042]; Montenegro[0.042]; Yugoslav[0.041]; \t\n\r [====================================>.........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 4s445ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3890/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 971testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lentini\u001b[39m ==> ENTITY: \u001b[32mGianluigi Lentini\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMilan[0.044]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Gianluigi[0.044]; Italian[0.044]; club[0.044]; Torino[0.043]; involved[0.042]; international[0.042]; loaned[0.042]; serie[0.042]; Soccer[0.042]; Atalanta[0.041]; Atalanta[0.041]; Atalanta[0.041]; player[0.041]; Sport[0.040]; joined[0.040]; Lentini[0.040]; Lentini[0.040]; subsequently[0.039]; suffered[0.039]; million[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = serie A ==> ENTITY: \u001b[32mSerie A\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.232:0.232[0]; log p(e|m)= -0.018:-0.018[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTorino[0.045]; club[0.044]; Soccer[0.044]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Italian[0.043]; player[0.042]; Sport[0.042]; international[0.042]; Bergamo[0.041]; Atalanta[0.041]; Atalanta[0.041]; Atalanta[0.041]; Atalanta[0.041]; Turin[0.040]; joined[0.040]; following[0.040]; loaned[0.040]; Gazzetta[0.039]; struggled[0.039]; deal[0.039]; deal[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fabio Capello\u001b[39m ==> ENTITY: \u001b[32mFabio Capello\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMondonico[0.047]; player[0.045]; Sport[0.044]; season[0.044]; coach[0.044]; coach[0.044]; Milan[0.043]; Milan[0.043]; Torino[0.043]; Torino[0.043]; following[0.043]; ex[0.042]; Turin[0.041]; million[0.040]; subsequently[0.040]; injuries[0.040]; joined[0.040]; Atalanta[0.040]; Atalanta[0.040]; far[0.040]; struggled[0.039]; Gazzetta[0.039]; suffered[0.039]; head[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Torino ==> ENTITY: \u001b[32mTorino F.C.\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.237:0.237[0]; log p(e|m)= -0.526:-0.526[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurin[0.047]; Torino[0.045]; club[0.044]; dello[0.043]; season[0.043]; Mondonico[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; car[0.042]; Italian[0.042]; Bergamo[0.040]; Soccer[0.040]; Fabio[0.040]; Sport[0.040]; loaned[0.040]; record[0.039]; Emiliano[0.039]; Atalanta[0.039]; Atalanta[0.039]; Atalanta[0.039]; Atalanta[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atalanta ==> ENTITY: \u001b[32mAtalanta B.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.210:0.210[0]; log p(e|m)= -0.347:-0.347[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTorino[0.045]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Milan[0.044]; club[0.044]; Italian[0.043]; Soccer[0.043]; Turin[0.042]; Atalanta[0.042]; Atalanta[0.042]; Atalanta[0.042]; Sport[0.042]; international[0.042]; loaned[0.041]; player[0.040]; serie[0.040]; hero[0.040]; Bergamo[0.040]; reunites[0.040]; joined[0.040]; struggled[0.039]; financial[0.039]; suffered[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gianluigi Lentini\u001b[39m ==> ENTITY: \u001b[32mGianluigi Lentini\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMilan[0.044]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Italian[0.044]; club[0.044]; Torino[0.043]; involved[0.042]; international[0.042]; loaned[0.042]; serie[0.042]; Soccer[0.042]; following[0.041]; Atalanta[0.041]; Atalanta[0.041]; Atalanta[0.041]; player[0.041]; Sport[0.040]; car[0.040]; struggled[0.040]; joined[0.040]; Lentini[0.040]; Lentini[0.040]; subsequently[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Emiliano Mondonico\u001b[39m ==> ENTITY: \u001b[32mEmiliano Mondonico\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nseason[0.045]; Fabio[0.045]; Torino[0.044]; Torino[0.044]; coach[0.044]; coach[0.044]; following[0.043]; Atalanta[0.043]; Atalanta[0.043]; head[0.043]; player[0.042]; Milan[0.042]; Milan[0.042]; Bergamo[0.041]; joined[0.041]; Turin[0.041]; subsequently[0.040]; regain[0.040]; injuries[0.040]; suffered[0.039]; ex[0.039]; investigated[0.039]; magistrates[0.039]; struggled[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atalanta ==> ENTITY: \u001b[32mAtalanta B.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.203:0.203[0]; log p(e|m)= -0.347:-0.347[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTorino[0.045]; Torino[0.045]; Mondonico[0.045]; Milan[0.044]; Milan[0.044]; club[0.044]; season[0.043]; Turin[0.043]; Atalanta[0.042]; Atalanta[0.042]; Sport[0.042]; Fabio[0.041]; loaned[0.041]; player[0.041]; serie[0.040]; hero[0.040]; Bergamo[0.040]; reunites[0.040]; joined[0.040]; struggled[0.040]; financial[0.039]; suffered[0.039]; million[0.039]; car[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gazzetta dello Sport\u001b[39m ==> ENTITY: \u001b[32mLa Gazzetta dello Sport\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItalian[0.045]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Fabio[0.043]; player[0.042]; Torino[0.042]; Turin[0.042]; club[0.041]; reported[0.041]; Soccer[0.041]; newspapers[0.041]; involved[0.040]; Atalanta[0.040]; Atalanta[0.040]; Atalanta[0.040]; Atalanta[0.040]; loaned[0.040]; international[0.040]; Bergamo[0.040]; alleged[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Milan ==> ENTITY: \u001b[32mA.C. Milan\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.214:0.214[0]; log p(e|m)= -1.959:-1.959[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.046]; Milan[0.046]; Milan[0.046]; Milan[0.046]; Soccer[0.045]; Torino[0.045]; Italian[0.044]; Sport[0.042]; loaned[0.041]; serie[0.041]; player[0.041]; Atalanta[0.041]; Atalanta[0.041]; Atalanta[0.041]; dello[0.041]; joined[0.040]; international[0.040]; million[0.040]; Gazzetta[0.039]; suffered[0.039]; believed[0.039]; deal[0.039]; deal[0.039]; subsequently[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMilan\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mA.C. Milan\u001b[39m <---> \u001b[32mMilan\u001b[39m\t\nSCORES: global= 0.276:0.250[\u001b[31m0.026\u001b[39m]; local()= 0.211:0.140[\u001b[31m0.072\u001b[39m]; log p(e|m)= -1.959:-0.301[\u001b[32m1.658\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.046]; Milan[0.046]; Milan[0.046]; Milan[0.046]; Soccer[0.045]; Torino[0.045]; Italian[0.044]; Sport[0.042]; international[0.041]; loaned[0.041]; serie[0.041]; player[0.041]; Atalanta[0.041]; Atalanta[0.041]; Atalanta[0.041]; dello[0.041]; joined[0.040]; million[0.040]; Gazzetta[0.039]; suffered[0.039]; believed[0.039]; near[0.039]; deal[0.039]; deal[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lentini\u001b[39m ==> ENTITY: \u001b[32mGianluigi Lentini\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMilan[0.046]; Milan[0.046]; club[0.046]; Fabio[0.045]; Torino[0.045]; Torino[0.045]; Mondonico[0.044]; involved[0.044]; serie[0.044]; Turin[0.043]; following[0.043]; season[0.043]; Atalanta[0.043]; Atalanta[0.043]; Atalanta[0.043]; player[0.043]; Bergamo[0.043]; Sport[0.042]; car[0.042]; struggled[0.042]; joined[0.041]; Lentini[0.041]; subsequently[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atalanta ==> ENTITY: \u001b[32mAtalanta B.C.\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.231:0.231[0]; log p(e|m)= -0.347:-0.347[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTorino[0.044]; Torino[0.044]; Mondonico[0.044]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; club[0.043]; Italian[0.042]; Soccer[0.042]; season[0.042]; Turin[0.042]; Atalanta[0.041]; Atalanta[0.041]; Atalanta[0.041]; Sport[0.041]; international[0.041]; Fabio[0.040]; loaned[0.040]; player[0.040]; serie[0.039]; hero[0.039]; Bergamo[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Milan ==> ENTITY: \u001b[32mA.C. Milan\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.210:0.210[0]; log p(e|m)= -1.959:-1.959[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.048]; Milan[0.047]; Milan[0.047]; Milan[0.047]; Soccer[0.047]; Torino[0.046]; Italian[0.046]; Sport[0.043]; international[0.043]; loaned[0.043]; serie[0.043]; player[0.043]; Atalanta[0.042]; Atalanta[0.042]; Atalanta[0.042]; dello[0.042]; following[0.042]; joined[0.042]; million[0.041]; Gazzetta[0.041]; suffered[0.040]; believed[0.040]; near[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Torino ==> ENTITY: \u001b[32mTorino F.C.\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.214:0.214[0]; log p(e|m)= -0.526:-0.526[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTurin[0.049]; Torino[0.047]; Mondonico[0.045]; Milan[0.045]; Milan[0.045]; car[0.044]; season[0.043]; Bergamo[0.042]; Fabio[0.042]; Emiliano[0.041]; Atalanta[0.041]; Atalanta[0.041]; struggled[0.041]; joined[0.041]; player[0.040]; suffered[0.040]; following[0.039]; deal[0.039]; deal[0.039]; coach[0.039]; coach[0.039]; injuries[0.039]; crash[0.039]; far[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atalanta ==> ENTITY: \u001b[32mAtalanta B.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.209:0.209[0]; log p(e|m)= -0.347:-0.347[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTorino[0.045]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Atalanta[0.044]; Atalanta[0.044]; club[0.044]; Italian[0.043]; Soccer[0.043]; Sport[0.042]; international[0.042]; loaned[0.041]; player[0.041]; serie[0.040]; joined[0.040]; Gazzetta[0.040]; financial[0.039]; suffered[0.039]; million[0.039]; record[0.039]; believed[0.039]; Gianluigi[0.039]; reported[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lentini\u001b[39m ==> ENTITY: \u001b[32mGianluigi Lentini\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMilan[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Milan[0.043]; Gianluigi[0.043]; Italian[0.043]; club[0.043]; Fabio[0.042]; Torino[0.042]; Torino[0.042]; Mondonico[0.041]; involved[0.041]; international[0.041]; loaned[0.041]; serie[0.041]; Soccer[0.041]; Turin[0.041]; following[0.040]; season[0.040]; Atalanta[0.040]; Atalanta[0.040]; Atalanta[0.040]; Atalanta[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italian ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.759:-0.759[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTorino[0.045]; club[0.045]; player[0.045]; Soccer[0.044]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Sport[0.043]; international[0.042]; serie[0.041]; Atalanta[0.041]; Atalanta[0.041]; Atalanta[0.041]; joined[0.041]; dello[0.041]; loaned[0.040]; world[0.040]; Gazzetta[0.040]; record[0.039]; suffered[0.039]; deal[0.039]; deal[0.039]; Gianluigi[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Milan ==> ENTITY: \u001b[32mA.C. Milan\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.237:0.237[0]; log p(e|m)= -1.959:-1.959[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclub[0.045]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Milan[0.044]; Mondonico[0.044]; Soccer[0.044]; Torino[0.043]; Torino[0.043]; Italian[0.043]; Turin[0.041]; Fabio[0.041]; season[0.041]; Sport[0.040]; international[0.040]; loaned[0.040]; serie[0.040]; Bergamo[0.040]; player[0.040]; Atalanta[0.040]; Atalanta[0.040]; Atalanta[0.040]; Atalanta[0.040]; following[0.039]; \t\n\r [====================================>.........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 4s332ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3909/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1122testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.242:0.242[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.048]; Iraq[0.048]; Iraq[0.048]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Kurdish[0.043]; Kurdish[0.043]; Iraqis[0.043]; forces[0.041]; forces[0.041]; northern[0.040]; northern[0.040]; Arbil[0.039]; Arbil[0.039]; region[0.038]; tanks[0.037]; artillery[0.037]; artillery[0.037]; reports[0.036]; advance[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iraqi Kurdish\u001b[39m ==> ENTITY: \u001b[32mIraqi Kurdistan\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdish[0.049]; northern[0.047]; northern[0.047]; Iraq[0.046]; Iraq[0.046]; Iraq[0.046]; Iraqi[0.044]; Arbil[0.043]; troops[0.043]; forces[0.042]; military[0.042]; Iran[0.041]; groups[0.041]; troop[0.039]; main[0.039]; city[0.039]; city[0.039]; escalating[0.038]; said[0.038]; said[0.038]; said[0.038]; rebel[0.038]; Thursday[0.038]; deployment[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = London-based\u001b[39m ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.055:0.055[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.053]; based[0.044]; administrative[0.042]; rebel[0.042]; Arbil[0.042]; Arbil[0.042]; Arbil[0.042]; Arbil[0.042]; National[0.041]; artillery[0.041]; artillery[0.041]; time[0.041]; exile[0.041]; began[0.041]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; came[0.040]; group[0.040]; group[0.040]; independent[0.040]; region[0.040]; Saturday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iran ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.158:0.158[0]; log p(e|m)= -0.138:-0.138[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.046]; Iraq[0.046]; Iraq[0.046]; military[0.045]; Tehran[0.044]; Iraqi[0.043]; Iraqi[0.043]; Kurdish[0.043]; Kurdish[0.043]; troops[0.041]; retaliate[0.041]; northern[0.041]; northern[0.041]; forces[0.040]; attack[0.040]; casualties[0.040]; Washington[0.040]; escalating[0.040]; said[0.039]; said[0.039]; said[0.039]; Arbil[0.039]; Arbil[0.039]; report[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arbil\u001b[39m ==> ENTITY: \u001b[32mErbil\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.253:0.253[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdish[0.044]; Kurdish[0.044]; Kurdish[0.044]; Arbil[0.043]; Arbil[0.043]; Arbil[0.043]; region[0.043]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; northern[0.040]; northern[0.040]; northern[0.040]; military[0.039]; town[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.255:0.255[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.051]; Iraq[0.051]; Iraq[0.051]; Iraq[0.051]; Iraq[0.051]; Iraqi[0.047]; Iraqi[0.047]; Kurdish[0.047]; Kurdish[0.047]; military[0.046]; northern[0.045]; northern[0.045]; troops[0.044]; forces[0.043]; forces[0.043]; deployment[0.043]; Arbil[0.042]; Arbil[0.042]; Iran[0.041]; artillery[0.041]; troop[0.041]; groups[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.278:0.278[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.048]; Iraq[0.048]; Iraq[0.048]; Iraq[0.048]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; military[0.044]; forces[0.042]; forces[0.042]; forces[0.042]; Iraqis[0.042]; Kurdish[0.042]; Kurdish[0.042]; Kurdish[0.042]; northern[0.040]; northern[0.040]; northern[0.040]; troop[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.267:0.267[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.048]; Iraq[0.048]; Iraq[0.048]; Iraq[0.048]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Kurdish[0.045]; Kurdish[0.045]; Kurdish[0.045]; Iraqis[0.043]; northern[0.043]; northern[0.043]; forces[0.041]; forces[0.041]; forces[0.041]; region[0.041]; Arbil[0.040]; Arbil[0.040]; Arbil[0.040]; Arbil[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arbil\u001b[39m ==> ENTITY: \u001b[32mErbil\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.247:0.247[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdish[0.045]; Kurdish[0.045]; Arbil[0.043]; Arbil[0.043]; Arbil[0.043]; region[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; north[0.041]; northern[0.041]; northern[0.041]; Iraqis[0.041]; town[0.039]; city[0.039]; forces[0.038]; forces[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bill Clinton\u001b[39m ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.077:0.077[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPresident[0.045]; Iraq[0.044]; Iraq[0.044]; Iraq[0.044]; Iraq[0.044]; Iraq[0.044]; military[0.043]; Washington[0.043]; Iraqi[0.042]; Iraqi[0.042]; Iraqi[0.042]; Iraqi[0.042]; Iran[0.041]; retaliate[0.041]; told[0.040]; troops[0.040]; forces[0.040]; forces[0.040]; accused[0.040]; asked[0.040]; deployment[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLondon[0.050]; said[0.043]; said[0.043]; based[0.043]; advance[0.042]; received[0.042]; National[0.042]; began[0.041]; Saturday[0.041]; reports[0.041]; group[0.041]; group[0.041]; attack[0.041]; north[0.041]; Iraq[0.041]; Iraq[0.041]; Iraq[0.041]; Iraq[0.041]; time[0.041]; says[0.041]; miles[0.041]; city[0.040]; region[0.040]; advanced[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tehran\u001b[39m ==> ENTITY: \u001b[32mTehran\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.049]; northern[0.045]; northern[0.045]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; Arbil[0.042]; city[0.041]; city[0.041]; Kurdish[0.041]; Kurdish[0.041]; Washington[0.041]; forces[0.041]; military[0.040]; escalating[0.040]; President[0.040]; Iraqi[0.040]; Iraqi[0.040]; said[0.040]; said[0.040]; said[0.040]; Friday[0.040]; Thursday[0.040]; report[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.128:0.128[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nreports[0.045]; told[0.043]; London[0.043]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; region[0.042]; Washington[0.042]; northern[0.042]; northern[0.042]; northern[0.042]; advanced[0.041]; report[0.041]; based[0.041]; troops[0.041]; Congress[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; Thursday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arbil\u001b[39m ==> ENTITY: \u001b[32mErbil\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.248:0.248[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdish[0.046]; Kurdish[0.046]; Kurdish[0.046]; Arbil[0.045]; Arbil[0.045]; Arbil[0.045]; region[0.044]; Iraq[0.044]; Iraq[0.044]; Iraq[0.044]; Iraq[0.044]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; north[0.043]; northern[0.042]; northern[0.042]; Iraqis[0.042]; military[0.040]; town[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.247:0.247[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.048]; Iraq[0.048]; Iraq[0.048]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; forces[0.041]; forces[0.041]; Iraqis[0.041]; Kurdish[0.041]; Kurdish[0.041]; northern[0.040]; northern[0.040]; Arbil[0.039]; Arbil[0.039]; Arbil[0.039]; Arbil[0.039]; artillery[0.039]; artillery[0.039]; region[0.038]; tanks[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.229:0.229[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.048]; Iraq[0.048]; Iraq[0.048]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; military[0.044]; deployment[0.042]; troops[0.042]; forces[0.042]; forces[0.042]; Kurdish[0.042]; Kurdish[0.042]; northern[0.040]; northern[0.040]; troop[0.040]; Arbil[0.039]; Arbil[0.039]; Iran[0.039]; artillery[0.039]; Washington[0.038]; casualties[0.037]; report[0.037]; groups[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = GMT\u001b[39m ==> ENTITY: \u001b[32mGreenwich Mean Time\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nThursday[0.053]; Friday[0.053]; time[0.053]; London[0.052]; London[0.052]; region[0.051]; Saturday[0.050]; advanced[0.050]; northern[0.050]; northern[0.050]; northern[0.050]; advance[0.050]; reports[0.049]; told[0.048]; began[0.048]; miles[0.048]; said[0.048]; said[0.048]; said[0.048]; said[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.066:0.066[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmilitary[0.046]; deployment[0.045]; Washington[0.043]; administrative[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; Iraq[0.043]; weeks[0.042]; time[0.041]; report[0.041]; troops[0.041]; ordered[0.041]; asked[0.041]; began[0.041]; troop[0.041]; ago[0.041]; reserved[0.040]; confirmation[0.040]; northern[0.040]; northern[0.040]; northern[0.040]; Bill[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.264:0.264[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.047]; Iraq[0.047]; Iraq[0.047]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; Iraqi[0.043]; military[0.043]; Iraqis[0.041]; Kurdish[0.040]; Kurdish[0.040]; Kurdish[0.040]; forces[0.040]; forces[0.040]; forces[0.040]; northern[0.039]; northern[0.039]; Arbil[0.038]; Arbil[0.038]; Arbil[0.038]; Arbil[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Iraqi National Congress\u001b[39m ==> ENTITY: \u001b[32mIraqi National Congress\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.049]; Iraq[0.049]; Iraq[0.049]; Iraq[0.049]; Iraqi[0.048]; Iraqi[0.048]; Iraqi[0.048]; Iraqi[0.048]; Iraqi[0.048]; Kurdish[0.045]; Kurdish[0.045]; Kurdish[0.045]; Iraqis[0.044]; independent[0.044]; forces[0.043]; forces[0.043]; forces[0.043]; exile[0.043]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.050]; Iraq[0.050]; Iraqi[0.046]; Iraqi[0.046]; military[0.046]; deployment[0.044]; troops[0.043]; forces[0.043]; Kurdish[0.043]; Kurdish[0.043]; northern[0.042]; northern[0.042]; troop[0.041]; Arbil[0.041]; Iran[0.041]; Washington[0.039]; casualties[0.038]; report[0.038]; groups[0.038]; Tehran[0.037]; fighting[0.037]; retaliate[0.037]; escalating[0.037]; President[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqis ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.247:0.247[0]; log p(e|m)= -0.812:-0.812[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.047]; Iraq[0.047]; Iraq[0.047]; Iraq[0.047]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Kurdish[0.042]; Kurdish[0.042]; forces[0.042]; forces[0.042]; exile[0.039]; London[0.039]; London[0.039]; northern[0.039]; northern[0.039]; Arbil[0.038]; Arbil[0.038]; region[0.038]; said[0.037]; said[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.252:0.252[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.047]; Iraq[0.047]; Iraq[0.047]; Iraq[0.047]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Iraqi[0.044]; Kurdish[0.044]; Kurdish[0.044]; Iraqis[0.042]; northern[0.042]; northern[0.042]; region[0.040]; Arbil[0.039]; Arbil[0.039]; Arbil[0.039]; north[0.038]; forces[0.038]; forces[0.038]; said[0.037]; said[0.037]; reports[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arbil\u001b[39m ==> ENTITY: \u001b[32mErbil\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.257:0.257[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKurdish[0.044]; Kurdish[0.044]; Kurdish[0.044]; Arbil[0.042]; Arbil[0.042]; Arbil[0.042]; region[0.042]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; Iraq[0.042]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; Iraqi[0.041]; troops[0.041]; northern[0.040]; northern[0.040]; northern[0.040]; northern[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.218:0.218[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.049]; Iraq[0.049]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; military[0.045]; deployment[0.043]; troops[0.043]; forces[0.042]; forces[0.042]; Kurdish[0.042]; Kurdish[0.042]; northern[0.041]; northern[0.041]; troop[0.040]; Arbil[0.040]; Arbil[0.040]; Iran[0.040]; artillery[0.040]; Washington[0.038]; casualties[0.038]; report[0.037]; groups[0.037]; Tehran[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.290:0.290[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.045]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; Iraq[0.045]; Iraqi[0.042]; Iraqi[0.042]; Iraqi[0.042]; Iraqi[0.042]; Iraqi[0.042]; Iraqi[0.042]; Kurdish[0.042]; Kurdish[0.042]; Kurdish[0.042]; military[0.041]; Iraqis[0.041]; northern[0.040]; northern[0.040]; northern[0.040]; forces[0.038]; forces[0.038]; forces[0.038]; region[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.267:0.267[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.048]; Iraq[0.048]; Iraq[0.048]; Iraq[0.048]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Kurdish[0.045]; Kurdish[0.045]; Kurdish[0.045]; Iraqis[0.043]; northern[0.043]; northern[0.043]; forces[0.041]; forces[0.041]; forces[0.041]; region[0.041]; Arbil[0.040]; Arbil[0.040]; Arbil[0.040]; Arbil[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Washington ==> ENTITY: \u001b[32mWashington, D.C.\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.082:0.082[0]; log p(e|m)= -1.749:-1.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBill[0.044]; President[0.044]; troop[0.043]; troops[0.042]; said[0.042]; said[0.042]; said[0.042]; military[0.042]; members[0.042]; Clinton[0.042]; groups[0.041]; Thursday[0.041]; region[0.041]; deployment[0.041]; came[0.041]; northern[0.041]; northern[0.041]; northern[0.041]; began[0.041]; Iraq[0.041]; Iraq[0.041]; Iraq[0.041]; Iraq[0.041]; Iraq[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraqi ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.258:0.258[0]; log p(e|m)= -0.485:-0.485[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.048]; Iraq[0.048]; Iraq[0.048]; Iraq[0.048]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Iraqi[0.045]; Kurdish[0.045]; Kurdish[0.045]; Iraqis[0.044]; northern[0.043]; northern[0.043]; forces[0.041]; forces[0.041]; region[0.041]; Arbil[0.040]; Arbil[0.040]; Arbil[0.040]; artillery[0.039]; artillery[0.039]; north[0.039]; \t\n\r [====================================>.........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 4s181ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3941/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1156testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.107:0.107[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGerman[0.050]; Der[0.044]; foreign[0.043]; Jens[0.043]; security[0.043]; seized[0.042]; month[0.042]; Bonn[0.042]; Bonn[0.042]; Bonn[0.042]; Spiegel[0.042]; stormed[0.041]; diplomat[0.041]; forces[0.041]; cultural[0.041]; news[0.041]; secret[0.041]; ago[0.040]; home[0.040]; home[0.040]; said[0.040]; said[0.040]; said[0.040]; report[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.049]; Berlin[0.046]; Berlin[0.046]; Berlin[0.046]; German[0.044]; German[0.044]; France[0.044]; world[0.041]; government[0.041]; lose[0.041]; opponent[0.040]; face[0.040]; intelligence[0.040]; received[0.040]; authorities[0.040]; political[0.040]; Spiegel[0.040]; president[0.040]; Der[0.040]; said[0.040]; said[0.040]; relations[0.039]; prosecutors[0.039]; continue[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iranian ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.228:0.228[0]; log p(e|m)= -0.571:-0.571[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIranian[0.046]; Iranian[0.046]; Iran[0.046]; Iran[0.046]; Iran[0.046]; Kurdish[0.044]; Kurdish[0.044]; Tehran[0.042]; Tehran[0.042]; Abolhassan[0.042]; Ali[0.041]; Banisadr[0.041]; Banisadr[0.041]; Banisadr[0.041]; Banisadr[0.041]; added[0.040]; ambassador[0.040]; asylum[0.039]; countries[0.039]; exiled[0.039]; international[0.039]; intelligence[0.039]; Germany[0.039]; Germany[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.137:0.137[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.051]; Germany[0.051]; German[0.048]; Berlin[0.046]; Berlin[0.046]; France[0.043]; Der[0.042]; world[0.041]; government[0.041]; authorities[0.040]; political[0.040]; Spiegel[0.040]; president[0.040]; asylum[0.039]; Iran[0.039]; Iran[0.039]; Iran[0.039]; leaders[0.039]; leaders[0.039]; leaders[0.039]; intelligence[0.039]; exile[0.039]; prosecutors[0.039]; relations[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iranian ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.236:0.236[0]; log p(e|m)= -0.571:-0.571[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIranian[0.064]; Iranian[0.064]; Iranian[0.064]; Iranian[0.064]; Iran[0.064]; Tehran[0.059]; Tehran[0.059]; Abolhassan[0.059]; news[0.058]; Banisadr[0.057]; Banisadr[0.057]; added[0.057]; ambassador[0.056]; embassy[0.055]; countries[0.055]; international[0.054]; exile[0.054]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Der Spiegel\u001b[39m ==> ENTITY: \u001b[32mDer Spiegel\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmagazine[0.052]; German[0.049]; German[0.049]; German[0.049]; report[0.047]; Jens[0.045]; Iranian[0.045]; Iranian[0.045]; Iranian[0.045]; Iranian[0.045]; month[0.045]; principles[0.045]; Iran[0.045]; news[0.044]; Bonn[0.044]; Bonn[0.044]; Bonn[0.044]; Bonn[0.044]; weekly[0.044]; cultural[0.044]; diplomat[0.044]; foreign[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Banisadr\u001b[39m ==> ENTITY: \u001b[32mAbolhassan Banisadr\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.238:0.238[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBanisadr[0.045]; Banisadr[0.045]; Banisadr[0.045]; Banisadr[0.045]; Iranian[0.044]; Iranian[0.044]; Iran[0.044]; Iran[0.044]; Iran[0.044]; Abolhassan[0.043]; Tehran[0.043]; president[0.042]; exile[0.041]; government[0.041]; freed[0.040]; minister[0.040]; detained[0.040]; exiled[0.039]; Ali[0.039]; political[0.039]; extradite[0.039]; extradite[0.039]; accused[0.039]; accused[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.100:0.100[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGerman[0.050]; Der[0.043]; foreign[0.043]; Jens[0.043]; month[0.043]; security[0.042]; state[0.042]; seized[0.042]; Bonn[0.041]; Bonn[0.041]; Bonn[0.041]; Spiegel[0.041]; diplomat[0.041]; stormed[0.041]; forces[0.041]; Iran[0.041]; cultural[0.041]; magazine[0.041]; news[0.040]; secret[0.040]; ago[0.040]; hostile[0.040]; home[0.040]; home[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Abolhassan Banisadr\u001b[39m ==> ENTITY: \u001b[32mAbolhassan Banisadr\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.230:0.230[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBanisadr[0.045]; Banisadr[0.045]; Banisadr[0.045]; Iranian[0.044]; Iranian[0.044]; Iranian[0.044]; Iran[0.044]; Iran[0.044]; Tehran[0.043]; Tehran[0.043]; president[0.042]; exile[0.041]; government[0.041]; government[0.041]; freed[0.041]; detained[0.040]; exiled[0.040]; political[0.039]; ambassador[0.039]; extradite[0.039]; extradite[0.039]; accused[0.039]; Bonn[0.039]; assassination[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.049]; Berlin[0.045]; Berlin[0.045]; Berlin[0.045]; German[0.044]; German[0.044]; France[0.043]; international[0.042]; added[0.042]; world[0.041]; countries[0.041]; received[0.041]; government[0.040]; lose[0.040]; opponent[0.040]; face[0.040]; authorities[0.040]; political[0.040]; Spiegel[0.040]; appeared[0.040]; president[0.039]; currently[0.039]; Der[0.039]; regretted[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iran ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.213:0.213[0]; log p(e|m)= -0.138:-0.138[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.048]; Iranian[0.046]; Iranian[0.046]; Iranian[0.046]; Iranian[0.046]; Tehran[0.043]; Tehran[0.043]; Kurdish[0.043]; international[0.041]; Germany[0.041]; Abolhassan[0.040]; Banisadr[0.040]; Banisadr[0.040]; Banisadr[0.040]; government[0.040]; government[0.040]; ambassador[0.040]; countries[0.039]; said[0.039]; said[0.039]; said[0.039]; said[0.039]; said[0.039]; breach[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iranian ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.198:0.198[0]; log p(e|m)= -0.571:-0.571[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIranian[0.052]; Iranian[0.052]; Iranian[0.052]; Iran[0.051]; Iran[0.051]; cultural[0.048]; Tehran[0.047]; Tehran[0.047]; news[0.046]; diplomat[0.045]; ambassador[0.045]; international[0.044]; embassy[0.043]; month[0.043]; foreign[0.043]; government[0.042]; state[0.042]; report[0.042]; security[0.042]; taken[0.042]; protest[0.042]; protest[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iran ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.177:0.177[0]; log p(e|m)= -0.138:-0.138[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.049]; Iran[0.049]; Iranian[0.048]; Tehran[0.045]; Kurdish[0.044]; Kurdish[0.044]; Germany[0.042]; Germany[0.042]; Banisadr[0.041]; Banisadr[0.041]; Banisadr[0.041]; government[0.041]; intelligence[0.040]; said[0.040]; said[0.040]; allegations[0.040]; told[0.040]; ordering[0.039]; ordering[0.039]; exile[0.039]; Ali[0.039]; continue[0.039]; France[0.039]; world[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tehran\u001b[39m ==> ENTITY: \u001b[32mTehran\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.050]; Iranian[0.047]; Iranian[0.047]; Iranian[0.047]; diplomat[0.043]; security[0.043]; report[0.043]; news[0.041]; forces[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; cultural[0.040]; state[0.040]; foreign[0.040]; month[0.040]; confirm[0.040]; activities[0.039]; hostile[0.039]; spokesman[0.039]; spokesman[0.039]; secret[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bonn\u001b[39m ==> ENTITY: \u001b[32mBonn\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.127:0.127[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBonn[0.048]; Bonn[0.048]; security[0.046]; German[0.046]; German[0.046]; said[0.044]; said[0.044]; said[0.044]; ministry[0.044]; forces[0.044]; month[0.044]; Jens[0.043]; Saturday[0.043]; activities[0.043]; seized[0.042]; Spiegel[0.042]; diplomat[0.042]; Der[0.042]; weekly[0.042]; news[0.041]; cultural[0.041]; ago[0.041]; stormed[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.101:0.101[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.050]; Germany[0.050]; government[0.048]; countries[0.047]; leaders[0.046]; leaders[0.046]; leaders[0.046]; world[0.046]; political[0.046]; president[0.045]; received[0.045]; relations[0.044]; ordering[0.044]; ordering[0.044]; German[0.044]; German[0.044]; Thursday[0.044]; exile[0.044]; minister[0.044]; exiled[0.044]; currently[0.044]; appeared[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Banisadr\u001b[39m ==> ENTITY: \u001b[32mAbolhassan Banisadr\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.233:0.233[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBanisadr[0.045]; Banisadr[0.045]; Banisadr[0.045]; Iranian[0.044]; Iranian[0.044]; Iranian[0.044]; Iran[0.044]; Iran[0.044]; Abolhassan[0.043]; Tehran[0.043]; Tehran[0.043]; president[0.042]; exile[0.041]; government[0.041]; government[0.041]; freed[0.040]; detained[0.040]; exiled[0.039]; political[0.039]; ambassador[0.039]; extradite[0.039]; extradite[0.039]; accused[0.039]; Bonn[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tehran\u001b[39m ==> ENTITY: \u001b[32mTehran\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.054]; Iranian[0.051]; Iranian[0.051]; Iranian[0.051]; Iranian[0.051]; Iranian[0.051]; Tehran[0.049]; protest[0.048]; protest[0.048]; Abolhassan[0.046]; international[0.046]; freed[0.046]; ambassador[0.046]; Banisadr[0.045]; Banisadr[0.045]; Berlin[0.045]; government[0.045]; government[0.045]; embassy[0.045]; news[0.045]; countries[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bonn\u001b[39m ==> ENTITY: \u001b[32mBonn\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBonn[0.045]; Bonn[0.045]; German[0.043]; German[0.043]; German[0.043]; Saturday[0.043]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; ministry[0.042]; ministry[0.042]; weekly[0.041]; month[0.041]; Jens[0.041]; activities[0.041]; forces[0.041]; embassy[0.040]; seized[0.040]; Spiegel[0.040]; diplomat[0.040]; Der[0.039]; news[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Banisadr\u001b[39m ==> ENTITY: \u001b[32mAbolhassan Banisadr\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.246:0.246[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBanisadr[0.046]; Banisadr[0.046]; Banisadr[0.046]; Banisadr[0.046]; Iranian[0.045]; Iranian[0.045]; Iran[0.045]; Iran[0.045]; Iran[0.045]; Iran[0.045]; Abolhassan[0.045]; Tehran[0.044]; president[0.043]; exile[0.042]; government[0.042]; freed[0.042]; minister[0.041]; detained[0.041]; exiled[0.041]; Ali[0.041]; political[0.041]; extradite[0.040]; extradite[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Berlin ==> ENTITY: \u001b[32mBerlin\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.154:-0.154[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBerlin[0.047]; Berlin[0.047]; Germany[0.045]; Germany[0.045]; German[0.044]; Bonn[0.043]; state[0.043]; Spiegel[0.043]; international[0.041]; principles[0.041]; Thursday[0.041]; political[0.041]; received[0.040]; currently[0.040]; ambassador[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; France[0.040]; fleeing[0.040]; magazine[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iran ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.161:0.161[0]; log p(e|m)= -0.138:-0.138[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.050]; Iran[0.050]; Kurdish[0.045]; Kurdish[0.045]; Germany[0.042]; Germany[0.042]; Banisadr[0.042]; Banisadr[0.042]; Banisadr[0.042]; intelligence[0.041]; said[0.041]; said[0.041]; allegations[0.041]; told[0.040]; relations[0.040]; Ali[0.040]; continue[0.040]; France[0.040]; world[0.040]; guarantee[0.039]; urged[0.039]; authorities[0.039]; received[0.039]; Thursday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.142:0.142[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.050]; Germany[0.050]; German[0.048]; Berlin[0.046]; Berlin[0.046]; France[0.043]; Der[0.042]; world[0.041]; authorities[0.041]; political[0.041]; Spiegel[0.040]; leaders[0.040]; leaders[0.040]; Thursday[0.040]; ordering[0.040]; intelligence[0.040]; prosecutors[0.039]; relations[0.039]; said[0.039]; said[0.039]; minister[0.039]; conduct[0.039]; fleeing[0.039]; exiled[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Berlin ==> ENTITY: \u001b[32mBerlin\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.154:-0.154[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBerlin[0.048]; Berlin[0.048]; Germany[0.046]; Germany[0.046]; German[0.044]; German[0.044]; Spiegel[0.043]; Thursday[0.042]; received[0.041]; currently[0.041]; added[0.041]; France[0.040]; fleeing[0.040]; president[0.040]; lives[0.040]; Der[0.040]; world[0.040]; law[0.040]; authorities[0.039]; said[0.039]; said[0.039]; said[0.039]; said[0.039]; appeared[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bonn\u001b[39m ==> ENTITY: \u001b[32mBonn\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBerlin[0.046]; appeared[0.044]; international[0.044]; German[0.043]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; ministry[0.042]; ministry[0.042]; countries[0.041]; government[0.041]; government[0.041]; Jens[0.041]; added[0.041]; activities[0.041]; court[0.041]; currently[0.040]; ordering[0.040]; embassy[0.040]; ambassador[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iranian ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.185:0.185[0]; log p(e|m)= -0.571:-0.571[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIranian[0.051]; Iranian[0.051]; Iranian[0.051]; Iran[0.051]; cultural[0.049]; Tehran[0.048]; Tehran[0.048]; news[0.046]; diplomat[0.045]; ambassador[0.045]; embassy[0.044]; international[0.044]; foreign[0.044]; month[0.043]; government[0.043]; state[0.043]; report[0.042]; security[0.042]; taken[0.042]; protest[0.042]; protest[0.042]; seized[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Der Spiegel\u001b[39m ==> ENTITY: \u001b[32mDer Spiegel\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGermany[0.047]; Germany[0.047]; German[0.045]; German[0.045]; Berlin[0.043]; Berlin[0.043]; Berlin[0.043]; told[0.042]; world[0.042]; Iranian[0.041]; Iranian[0.041]; Iran[0.041]; Iran[0.041]; Iran[0.041]; France[0.041]; allegations[0.040]; intelligence[0.040]; countries[0.040]; said[0.039]; said[0.039]; said[0.039]; accused[0.039]; accused[0.039]; personally[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iran ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.138:-0.138[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIranian[0.049]; Iranian[0.049]; Iranian[0.049]; Tehran[0.046]; security[0.043]; diplomat[0.042]; month[0.041]; said[0.041]; said[0.041]; said[0.041]; report[0.041]; foreign[0.041]; forces[0.041]; German[0.040]; German[0.040]; confirm[0.040]; cultural[0.040]; threatened[0.040]; ago[0.039]; spokesman[0.039]; home[0.039]; home[0.039]; news[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bonn\u001b[39m ==> ENTITY: \u001b[32mBonn\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.127:0.127[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBonn[0.048]; Bonn[0.048]; security[0.046]; German[0.046]; German[0.046]; said[0.044]; said[0.044]; said[0.044]; ministry[0.044]; forces[0.044]; month[0.044]; Jens[0.043]; Saturday[0.043]; activities[0.043]; seized[0.042]; Spiegel[0.042]; diplomat[0.042]; Der[0.042]; weekly[0.042]; news[0.041]; cultural[0.041]; ago[0.041]; stormed[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tehran\u001b[39m ==> ENTITY: \u001b[32mTehran\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.048]; Iran[0.048]; Iran[0.048]; Iranian[0.045]; Iranian[0.045]; Iranian[0.045]; Tehran[0.043]; protest[0.042]; protest[0.042]; Abolhassan[0.040]; international[0.040]; freed[0.040]; ambassador[0.040]; Banisadr[0.040]; Banisadr[0.040]; Banisadr[0.040]; Banisadr[0.040]; Berlin[0.040]; Berlin[0.040]; Berlin[0.040]; government[0.040]; government[0.040]; Ali[0.040]; Kurdish[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Berlin ==> ENTITY: \u001b[32mBerlin\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.160:0.160[0]; log p(e|m)= -0.154:-0.154[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBerlin[0.048]; Berlin[0.048]; Germany[0.046]; Germany[0.046]; German[0.044]; German[0.044]; Spiegel[0.043]; Thursday[0.042]; received[0.041]; currently[0.041]; added[0.041]; France[0.040]; fleeing[0.040]; president[0.040]; lives[0.040]; Der[0.040]; world[0.040]; authorities[0.039]; said[0.039]; said[0.039]; said[0.039]; said[0.039]; said[0.039]; appeared[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iranian ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.220:0.220[0]; log p(e|m)= -0.571:-0.571[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIranian[0.050]; Iranian[0.050]; Iran[0.050]; Iran[0.050]; Kurdish[0.048]; Tehran[0.046]; Tehran[0.046]; Abolhassan[0.046]; Banisadr[0.045]; Banisadr[0.045]; Banisadr[0.045]; Banisadr[0.045]; added[0.045]; ambassador[0.044]; asylum[0.043]; embassy[0.043]; countries[0.043]; exiled[0.043]; international[0.043]; Germany[0.043]; Germany[0.043]; exile[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.106:0.106[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBerlin[0.051]; countries[0.049]; Der[0.047]; government[0.047]; government[0.047]; foreign[0.046]; international[0.046]; Jens[0.046]; state[0.046]; seized[0.045]; month[0.045]; Bonn[0.045]; Bonn[0.045]; Spiegel[0.045]; president[0.045]; ambassador[0.044]; regretted[0.044]; magazine[0.044]; news[0.044]; secret[0.044]; protest[0.043]; protest[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Banisadr\u001b[39m ==> ENTITY: \u001b[32mAbolhassan Banisadr\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBanisadr[0.047]; Banisadr[0.047]; Iran[0.046]; Iran[0.046]; Iran[0.046]; minister[0.042]; Ali[0.042]; political[0.041]; extradite[0.041]; extradite[0.041]; accused[0.041]; intelligence[0.040]; said[0.040]; said[0.040]; leaders[0.040]; leaders[0.040]; asylum[0.040]; allegations[0.040]; Thursday[0.040]; authorities[0.040]; prosecutors[0.040]; Kurdish[0.039]; denies[0.039]; guarantee[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iran ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.237:0.237[0]; log p(e|m)= -0.138:-0.138[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIran[0.057]; Iran[0.057]; Iran[0.057]; Iranian[0.055]; Iranian[0.055]; Tehran[0.051]; Kurdish[0.050]; Kurdish[0.050]; international[0.048]; Germany[0.048]; Germany[0.048]; Abolhassan[0.048]; Banisadr[0.048]; Banisadr[0.048]; Banisadr[0.048]; Banisadr[0.048]; Banisadr[0.048]; government[0.047]; intelligence[0.046]; countries[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iranian ==> ENTITY: \u001b[32mIran\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.571:-0.571[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIranian[0.049]; Iranian[0.049]; Iran[0.048]; cultural[0.046]; Tehran[0.045]; news[0.044]; foreign[0.043]; diplomat[0.043]; month[0.041]; said[0.041]; said[0.041]; said[0.041]; state[0.040]; report[0.040]; security[0.040]; ministry[0.040]; promoting[0.039]; interrogated[0.039]; German[0.039]; German[0.039]; violence[0.039]; writers[0.039]; home[0.039]; home[0.039]; \t\n\r [=====================================>........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 3s986ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3979/4791 \r\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Panama ==> ENTITY: \u001b[32mPanama national football team\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.182:0.182[0]; log p(e|m)= -2.749:-2.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.046]; Panama[0.045]; Panama[0.045]; Soccer[0.045]; Jorge[0.045]; match[0.044]; Canada[0.043]; Canada[0.043]; Canada[0.043]; Concacaf[0.043]; qualifying[0.042]; Valdes[0.042]; beat[0.041]; Qualifier[0.041]; Cup[0.041]; Cup[0.041]; Beat[0.040]; World[0.040]; World[0.040]; semifinal[0.040]; Friday[0.039]; halftime[0.038]; Paul[0.038]; Carlo[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paul Peschisolido\u001b[39m ==> ENTITY: \u001b[32mPaul Peschisolido\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.070:0.070[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.047]; match[0.045]; Cup[0.045]; Cup[0.045]; min[0.044]; Canada[0.043]; Canada[0.043]; Canada[0.043]; Qualifier[0.043]; semifinal[0.042]; halftime[0.042]; Scorers[0.042]; Attendance[0.042]; Concacaf[0.041]; Edmonton[0.041]; qualifying[0.041]; Friday[0.041]; World[0.039]; World[0.039]; beat[0.039]; Panama[0.039]; Panama[0.039]; Panama[0.039]; pen[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Aunger\u001b[39m ==> ENTITY: \u001b[32mGeoff Aunger\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.120:0.120[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.046]; Canada[0.045]; Canada[0.045]; Canada[0.045]; Cup[0.043]; Cup[0.043]; qualifying[0.043]; Peschisolido[0.042]; match[0.042]; Paul[0.042]; min[0.042]; Edmonton[0.042]; Concacaf[0.041]; Qualifier[0.041]; semifinal[0.040]; halftime[0.040]; Scorers[0.040]; Attendance[0.040]; Friday[0.040]; World[0.039]; World[0.039]; Panama[0.039]; Panama[0.039]; Panama[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Panama ==> ENTITY: \u001b[32mPanama national football team\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.182:0.182[0]; log p(e|m)= -2.749:-2.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.046]; Panama[0.045]; Panama[0.045]; Soccer[0.045]; Jorge[0.045]; match[0.044]; Canada[0.043]; Canada[0.043]; Canada[0.043]; Concacaf[0.043]; qualifying[0.042]; Valdes[0.042]; beat[0.041]; Qualifier[0.041]; Cup[0.041]; Cup[0.041]; Beat[0.040]; World[0.040]; World[0.040]; semifinal[0.040]; Friday[0.039]; halftime[0.038]; Paul[0.038]; Carlo[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Edmonton ==> ENTITY: \u001b[32mEdmonton\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.082:0.082[0]; log p(e|m)= -0.270:-0.270[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.050]; Canada[0.050]; Canada[0.050]; Soccer[0.046]; Cup[0.044]; Cup[0.044]; match[0.042]; World[0.042]; World[0.042]; qualifying[0.041]; Friday[0.041]; Qualifier[0.040]; Paul[0.040]; beat[0.040]; Beat[0.040]; Attendance[0.040]; Peschisolido[0.039]; semifinal[0.039]; Concacaf[0.039]; halftime[0.039]; phase[0.038]; Panama[0.038]; Panama[0.038]; Panama[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = World Cup ==> ENTITY: \u001b[32mFIFA World Cup\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.110:0.110[0]; log p(e|m)= -1.002:-1.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.047]; qualifying[0.045]; Concacaf[0.045]; Canada[0.045]; Canada[0.045]; Canada[0.045]; Cup[0.045]; match[0.044]; World[0.044]; Qualifier[0.044]; Panama[0.042]; Panama[0.042]; Panama[0.042]; Luis[0.040]; Jorge[0.040]; beat[0.040]; semifinal[0.039]; Friday[0.039]; Beat[0.039]; Carlo[0.039]; Scorers[0.038]; Edmonton[0.038]; min[0.038]; halftime[0.037]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCONCACAF\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mCONCACAF\u001b[39m <---> \u001b[32mCONCACAF Champions League\u001b[39m\t\nSCORES: global= 0.261:0.253[\u001b[31m0.008\u001b[39m]; local()= 0.153:0.172[\u001b[32m0.019\u001b[39m]; log p(e|m)= 0.000:-0.679[\u001b[31m0.679\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.046]; Cup[0.044]; Cup[0.044]; qualifying[0.044]; Panama[0.044]; Panama[0.044]; Panama[0.044]; Canada[0.043]; Canada[0.043]; Canada[0.043]; Qualifier[0.043]; match[0.042]; World[0.042]; World[0.042]; Luis[0.042]; Jorge[0.042]; beat[0.040]; semifinal[0.039]; Beat[0.039]; Valdes[0.039]; Edmonton[0.038]; Friday[0.038]; halftime[0.037]; Attendance[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Panama ==> ENTITY: \u001b[32mPanama national football team\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.182:0.182[0]; log p(e|m)= -2.749:-2.749[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLuis[0.046]; Panama[0.045]; Panama[0.045]; Soccer[0.045]; Jorge[0.045]; match[0.044]; Canada[0.043]; Canada[0.043]; Canada[0.043]; Concacaf[0.043]; qualifying[0.042]; Valdes[0.042]; beat[0.041]; Qualifier[0.041]; Cup[0.041]; Cup[0.041]; Beat[0.040]; World[0.040]; World[0.040]; semifinal[0.040]; Friday[0.039]; halftime[0.038]; Paul[0.038]; Carlo[0.037]; \t\n\r [=====================================>........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 3s943ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 3988/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1098testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.058:0.058[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.045]; match[0.045]; stadium[0.044]; second[0.043]; fourth[0.043]; night[0.043]; National[0.043]; win[0.042]; win[0.042]; time[0.042]; Richard[0.042]; round[0.042]; champion[0.042]; Friday[0.042]; decided[0.041]; set[0.040]; Open[0.040]; Open[0.040]; Tennis[0.040]; Tennis[0.040]; extended[0.040]; wrestled[0.040]; quit[0.039]; hampered[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Edberg\u001b[39m ==> ENTITY: \u001b[32mStefan Edberg\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHaarhuis[0.046]; Wimbledon[0.043]; match[0.043]; match[0.043]; match[0.043]; Krajicek[0.042]; Krajicek[0.042]; round[0.042]; matches[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; champion[0.041]; Swede[0.041]; Paul[0.041]; twice[0.040]; play[0.040]; countryman[0.040]; serve[0.039]; Richard[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Edberg\u001b[39m ==> ENTITY: \u001b[32mStefan Edberg\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.232:0.232[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.046]; Lendl[0.046]; Wimbledon[0.043]; match[0.043]; match[0.043]; Slam[0.042]; Slam[0.042]; Krajicek[0.042]; round[0.042]; round[0.042]; matches[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; champion[0.041]; champion[0.041]; Swede[0.041]; win[0.040]; win[0.040]; victory[0.040]; Richard[0.039]; retire[0.039]; Ivan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Edberg\u001b[39m ==> ENTITY: \u001b[32mStefan Edberg\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.237:0.237[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.046]; Lendl[0.045]; match[0.043]; match[0.043]; match[0.043]; match[0.043]; Open[0.042]; Open[0.042]; Slam[0.042]; Slam[0.042]; round[0.042]; round[0.042]; Edberg[0.041]; Edberg[0.041]; Edberg[0.041]; Edberg[0.041]; Edberg[0.041]; champion[0.041]; champion[0.041]; win[0.040]; win[0.040]; Stefan[0.040]; victory[0.040]; Bernd[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Paul Haarhuis\u001b[39m ==> ENTITY: \u001b[32mPaul Haarhuis\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEdberg[0.045]; Edberg[0.045]; Edberg[0.045]; Edberg[0.045]; Edberg[0.045]; Edberg[0.045]; tennis[0.045]; match[0.042]; match[0.042]; twice[0.042]; countryman[0.042]; Krajicek[0.042]; champion[0.041]; final[0.041]; game[0.040]; draw[0.040]; play[0.040]; set[0.039]; set[0.039]; set[0.039]; tonight[0.039]; start[0.039]; times[0.038]; realistic[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Edberg\u001b[39m ==> ENTITY: \u001b[32mStefan Edberg\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.232:0.232[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.046]; Tennis[0.046]; Lendl[0.046]; match[0.043]; match[0.043]; Open[0.042]; Open[0.042]; Slam[0.042]; Slam[0.042]; round[0.042]; round[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; champion[0.041]; champion[0.041]; win[0.040]; win[0.040]; Stefan[0.040]; victory[0.040]; Bernd[0.040]; Richard[0.039]; retire[0.039]; Ivan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Edberg\u001b[39m ==> ENTITY: \u001b[32mStefan Edberg\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWimbledon[0.044]; match[0.044]; match[0.044]; match[0.044]; Krajicek[0.043]; round[0.043]; round[0.043]; matches[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; champion[0.042]; Swede[0.042]; twice[0.040]; play[0.040]; serve[0.040]; Richard[0.039]; game[0.039]; second[0.039]; second[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Edberg\u001b[39m ==> ENTITY: \u001b[32mStefan Edberg\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntennis[0.054]; Haarhuis[0.051]; match[0.047]; match[0.047]; Krajicek[0.046]; matches[0.046]; Edberg[0.046]; Edberg[0.046]; Edberg[0.046]; Edberg[0.046]; Edberg[0.046]; champion[0.046]; Paul[0.045]; twice[0.044]; play[0.044]; play[0.044]; countryman[0.043]; final[0.043]; serve[0.043]; game[0.043]; second[0.042]; second[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Edberg\u001b[39m ==> ENTITY: \u001b[32mStefan Edberg\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.233:0.233[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.046]; Lendl[0.046]; Wimbledon[0.043]; match[0.043]; match[0.043]; Slam[0.042]; Slam[0.042]; Krajicek[0.042]; round[0.042]; round[0.042]; matches[0.042]; Edberg[0.041]; Edberg[0.041]; Edberg[0.041]; Edberg[0.041]; Edberg[0.041]; champion[0.041]; champion[0.041]; Swede[0.041]; win[0.040]; win[0.040]; victory[0.040]; play[0.039]; Richard[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Karbacher\u001b[39m ==> ENTITY: \u001b[32mBernd Karbacher\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.215:0.215[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLendl[0.051]; Open[0.050]; Open[0.050]; Tennis[0.049]; victory[0.048]; German[0.047]; round[0.047]; round[0.047]; win[0.047]; win[0.047]; Swede[0.047]; Edberg[0.047]; Edberg[0.047]; Edberg[0.047]; Edberg[0.047]; Edberg[0.047]; Bernd[0.047]; Slam[0.047]; Slam[0.047]; champion[0.046]; champion[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Karbacher\u001b[39m ==> ENTITY: \u001b[32mBernd Karbacher\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.213:0.213[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLendl[0.049]; Open[0.048]; Open[0.048]; Tennis[0.046]; Tennis[0.046]; victory[0.045]; German[0.045]; round[0.045]; round[0.045]; win[0.045]; win[0.045]; Edberg[0.045]; Edberg[0.045]; Edberg[0.045]; Edberg[0.045]; Edberg[0.045]; Bernd[0.045]; Stefan[0.045]; Slam[0.044]; Slam[0.044]; champion[0.044]; champion[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Edberg\u001b[39m ==> ENTITY: \u001b[32mStefan Edberg\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.220:0.220[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHaarhuis[0.046]; Wimbledon[0.043]; match[0.043]; match[0.043]; match[0.043]; Krajicek[0.042]; Krajicek[0.042]; round[0.042]; matches[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; champion[0.041]; champion[0.041]; Swede[0.041]; Paul[0.041]; twice[0.040]; play[0.040]; countryman[0.040]; serve[0.039]; Richard[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Stefan Edberg\u001b[39m ==> ENTITY: \u001b[32mStefan Edberg\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.048]; Tennis[0.048]; match[0.044]; match[0.044]; Open[0.044]; Open[0.044]; Slam[0.043]; round[0.043]; Edberg[0.043]; Edberg[0.043]; Edberg[0.043]; champion[0.042]; win[0.041]; win[0.041]; Bernd[0.041]; Richard[0.040]; second[0.039]; fourth[0.039]; set[0.039]; Grand[0.039]; old[0.038]; event[0.038]; Friday[0.038]; hamstring[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Edberg\u001b[39m ==> ENTITY: \u001b[32mStefan Edberg\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntennis[0.051]; Haarhuis[0.048]; match[0.045]; match[0.045]; Krajicek[0.044]; Edberg[0.043]; Edberg[0.043]; Edberg[0.043]; champion[0.043]; Paul[0.042]; play[0.041]; countryman[0.041]; final[0.041]; game[0.040]; second[0.040]; set[0.040]; set[0.040]; start[0.039]; draw[0.039]; week[0.039]; starting[0.038]; starting[0.038]; little[0.038]; tonight[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wimbledon ==> ENTITY: \u001b[32mThe Championships, Wimbledon\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.209:0.209[0]; log p(e|m)= -0.851:-0.851[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.056]; match[0.055]; match[0.055]; match[0.055]; second[0.053]; second[0.053]; upset[0.053]; round[0.053]; round[0.053]; fifth[0.052]; Krajicek[0.052]; champion[0.052]; Richard[0.052]; set[0.051]; set[0.051]; set[0.051]; set[0.051]; twice[0.051]; play[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Grand Slam ==> ENTITY: \u001b[32mGrand Slam (tennis)\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.247:0.247[0]; log p(e|m)= -0.374:-0.374[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.050]; Open[0.046]; match[0.045]; match[0.045]; match[0.045]; match[0.045]; career[0.044]; Lendl[0.044]; Slam[0.044]; win[0.043]; win[0.043]; event[0.043]; set[0.043]; set[0.043]; round[0.043]; round[0.043]; champion[0.042]; champion[0.042]; fourth[0.042]; second[0.041]; time[0.041]; Grand[0.041]; victory[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.082:0.082[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.048]; match[0.048]; match[0.048]; career[0.043]; win[0.043]; win[0.043]; injury[0.043]; stadium[0.042]; quit[0.042]; second[0.041]; victory[0.041]; time[0.041]; fourth[0.041]; early[0.040]; decided[0.040]; Finn[0.040]; round[0.040]; round[0.040]; Bernd[0.040]; event[0.040]; Stefan[0.040]; ended[0.039]; Richard[0.039]; trailing[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Krajicek\u001b[39m ==> ENTITY: \u001b[32mRichard Krajicek\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.191:0.191[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntennis[0.050]; match[0.046]; match[0.046]; Haarhuis[0.045]; champion[0.042]; twice[0.042]; set[0.041]; set[0.041]; set[0.041]; play[0.041]; final[0.041]; Dutch[0.041]; game[0.041]; start[0.041]; second[0.040]; second[0.040]; draw[0.040]; Edberg[0.040]; Edberg[0.040]; Edberg[0.040]; Edberg[0.040]; Edberg[0.040]; Edberg[0.040]; serve[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Edberg\u001b[39m ==> ENTITY: \u001b[32mStefan Edberg\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.232:0.232[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLendl[0.046]; Wimbledon[0.043]; match[0.043]; match[0.043]; Slam[0.042]; Slam[0.042]; Krajicek[0.042]; round[0.042]; round[0.042]; matches[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; champion[0.041]; champion[0.041]; Swede[0.041]; win[0.040]; win[0.040]; victory[0.040]; play[0.039]; Richard[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Edberg\u001b[39m ==> ENTITY: \u001b[32mStefan Edberg\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.217:0.217[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHaarhuis[0.046]; Wimbledon[0.044]; match[0.043]; match[0.043]; Krajicek[0.042]; Krajicek[0.042]; round[0.042]; matches[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; champion[0.042]; champion[0.042]; Paul[0.041]; twice[0.040]; play[0.040]; countryman[0.040]; final[0.040]; serve[0.039]; Richard[0.039]; game[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Karbacher\u001b[39m ==> ENTITY: \u001b[32mBernd Karbacher\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.208:0.208[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLendl[0.045]; Wimbledon[0.044]; Tennis[0.043]; victory[0.042]; German[0.042]; round[0.042]; round[0.042]; win[0.042]; win[0.042]; Swede[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Slam[0.041]; Slam[0.041]; champion[0.041]; champion[0.041]; match[0.041]; match[0.041]; fifth[0.041]; upset[0.041]; Krajicek[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Grand Slam ==> ENTITY: \u001b[32mGrand Slam (tennis)\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.251:0.251[0]; log p(e|m)= -0.374:-0.374[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.048]; Open[0.044]; Open[0.044]; match[0.044]; match[0.044]; match[0.044]; match[0.044]; career[0.042]; Lendl[0.042]; Slam[0.042]; win[0.041]; win[0.041]; event[0.041]; set[0.041]; set[0.041]; round[0.041]; round[0.041]; champion[0.040]; champion[0.040]; fourth[0.040]; second[0.040]; time[0.040]; Grand[0.039]; victory[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Edberg\u001b[39m ==> ENTITY: \u001b[32mStefan Edberg\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.048]; Tennis[0.048]; match[0.045]; match[0.045]; Open[0.044]; Open[0.044]; round[0.043]; Edberg[0.043]; Edberg[0.043]; champion[0.043]; win[0.042]; win[0.042]; Stefan[0.042]; Bernd[0.041]; Richard[0.040]; second[0.040]; fourth[0.040]; set[0.040]; old[0.039]; Friday[0.039]; hamstring[0.038]; injury[0.038]; wrestled[0.038]; time[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. Open ==> ENTITY: \u001b[32mUS Open (tennis)\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.210:0.210[0]; log p(e|m)= -1.238:-1.238[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.048]; Tennis[0.048]; Open[0.044]; Open[0.044]; match[0.044]; match[0.044]; fourth[0.044]; second[0.043]; round[0.043]; Slam[0.042]; champion[0.041]; win[0.041]; win[0.041]; victory[0.040]; event[0.040]; set[0.039]; Richard[0.039]; National[0.039]; trailing[0.039]; Edberg[0.039]; Edberg[0.039]; Edberg[0.039]; Edberg[0.039]; Grand[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Richard Krajicek\u001b[39m ==> ENTITY: \u001b[32mRichard Krajicek\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.189:0.189[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.046]; match[0.046]; matches[0.044]; Wimbledon[0.043]; champion[0.042]; twice[0.042]; round[0.042]; set[0.042]; set[0.042]; set[0.042]; set[0.042]; play[0.042]; game[0.041]; second[0.041]; second[0.041]; upset[0.041]; Edberg[0.040]; Edberg[0.040]; Edberg[0.040]; Edberg[0.040]; Edberg[0.040]; Edberg[0.040]; Edberg[0.040]; Edberg[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dutch ==> ENTITY: \u001b[32mNetherlands\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.077:0.077[0]; log p(e|m)= -0.794:-0.794[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndoes[0.046]; match[0.046]; match[0.046]; final[0.043]; play[0.043]; longer[0.042]; game[0.042]; second[0.042]; second[0.042]; got[0.041]; plans[0.041]; breaks[0.041]; draw[0.041]; little[0.041]; tonight[0.041]; twice[0.041]; week[0.041]; feel[0.040]; champion[0.040]; chance[0.040]; way[0.040]; Like[0.040]; lost[0.040]; making[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Karbacher\u001b[39m ==> ENTITY: \u001b[32mBernd Karbacher\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLendl[0.045]; Wimbledon[0.044]; victory[0.042]; round[0.042]; round[0.042]; Swede[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Slam[0.041]; champion[0.041]; champion[0.041]; match[0.041]; match[0.041]; fifth[0.041]; upset[0.041]; Krajicek[0.041]; career[0.040]; twice[0.040]; matches[0.040]; Ivan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ivan Lendl\u001b[39m ==> ENTITY: \u001b[32mIvan Lendl\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.256:0.256[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.045]; Open[0.044]; Edberg[0.043]; Edberg[0.043]; Edberg[0.043]; Edberg[0.043]; Edberg[0.043]; match[0.042]; match[0.042]; match[0.042]; match[0.042]; Slam[0.042]; Slam[0.042]; champion[0.041]; champion[0.041]; round[0.041]; round[0.041]; Swede[0.040]; event[0.040]; second[0.039]; win[0.039]; win[0.039]; Bernd[0.039]; fourth[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Edberg\u001b[39m ==> ENTITY: \u001b[32mStefan Edberg\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.225:0.225[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntennis[0.049]; Haarhuis[0.046]; Wimbledon[0.043]; match[0.043]; match[0.043]; Krajicek[0.042]; Krajicek[0.042]; round[0.042]; matches[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; champion[0.041]; champion[0.041]; Paul[0.041]; twice[0.040]; play[0.039]; play[0.039]; countryman[0.039]; final[0.039]; serve[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bernd Karbacher\u001b[39m ==> ENTITY: \u001b[32mBernd Karbacher\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLendl[0.045]; Open[0.044]; Open[0.044]; Tennis[0.043]; Tennis[0.043]; victory[0.042]; German[0.042]; round[0.042]; win[0.042]; win[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Edberg[0.042]; Stefan[0.042]; Slam[0.041]; champion[0.041]; match[0.041]; match[0.041]; Ivan[0.040]; event[0.039]; fourth[0.039]; second[0.039]; hamstring[0.038]; \t\n\r [=====================================>........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 3s763ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4018/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 977testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Warsaw ==> ENTITY: \u001b[32mWarsaw\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.062:0.062[0]; log p(e|m)= -0.143:-0.143[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.071]; Warsaw[0.068]; Polish[0.065]; Friday[0.062]; operation[0.061]; Operation[0.060]; refrained[0.059]; staging[0.059]; National[0.058]; said[0.058]; Bank[0.057]; bank[0.056]; Reverse[0.056]; reverse[0.054]; Newsroom[0.052]; repo[0.052]; Repo[0.051]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = National Bank of Poland\u001b[39m ==> ENTITY: \u001b[32mNational Bank of Poland\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.088:0.088[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarsaw[0.072]; Warsaw[0.072]; Polish[0.069]; bank[0.068]; Newsroom[0.064]; repo[0.062]; Friday[0.061]; said[0.061]; operation[0.061]; Refrains[0.061]; Repo[0.059]; refrained[0.059]; Reverse[0.059]; staging[0.058]; Operation[0.057]; reverse[0.057]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Polish ==> ENTITY: \u001b[32mPoland\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.090:0.090[0]; log p(e|m)= -0.814:-0.814[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPoland[0.083]; Warsaw[0.076]; Warsaw[0.076]; National[0.069]; Operation[0.068]; operation[0.068]; Bank[0.066]; Friday[0.065]; staging[0.065]; said[0.065]; refrained[0.063]; bank[0.063]; Reverse[0.059]; Repo[0.058]; Newsroom[0.057]; \t\n\r [=====================================>........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 3s762ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4021/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1007testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mTripoli\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mTripoli\u001b[39m <---> \u001b[32mTripoli, Lebanon\u001b[39m\t\nSCORES: global= 0.262:0.255[\u001b[31m0.007\u001b[39m]; local()= 0.153:0.096[\u001b[31m0.057\u001b[39m]; log p(e|m)= -0.334:-1.826[\u001b[31m1.492\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTripoli[0.048]; Tripoli[0.048]; Tripoli[0.048]; Libya[0.045]; Gaddafi[0.043]; Libyans[0.043]; Muammar[0.043]; road[0.043]; Sunday[0.043]; capital[0.042]; stadium[0.041]; imposed[0.041]; bombing[0.040]; banners[0.040]; flight[0.040]; airport[0.040]; United[0.039]; coup[0.039]; coup[0.039]; internal[0.039]; serves[0.039]; celebrations[0.039]; celebrations[0.039]; brought[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scotland ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.047:0.047[0]; log p(e|m)= -0.229:-0.229[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKing[0.045]; stadium[0.045]; Nations[0.045]; Sunday[0.043]; Sunday[0.043]; group[0.042]; September[0.042]; African[0.042]; United[0.042]; attend[0.041]; chosen[0.041]; banners[0.041]; Guinea[0.041]; Ghana[0.041]; Mohammed[0.041]; Libya[0.041]; great[0.041]; great[0.041]; achievements[0.041]; flags[0.041]; refusal[0.040]; stages[0.040]; banner[0.040]; Libyans[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Benghazi\u001b[39m ==> ENTITY: \u001b[32mBenghazi\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.135:0.135[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTripoli[0.046]; region[0.046]; mountainous[0.044]; Gaddafi[0.043]; Gaddafi[0.043]; Gaddafi[0.043]; Egypt[0.043]; east[0.042]; militant[0.041]; Friday[0.041]; Libyans[0.041]; based[0.041]; said[0.041]; militants[0.041]; July[0.041]; July[0.041]; Egyptian[0.041]; country[0.041]; Ammar[0.041]; airforce[0.040]; revolution[0.040]; hotbed[0.040]; thousands[0.040]; told[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.090:0.090[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntold[0.045]; region[0.044]; country[0.043]; flights[0.043]; based[0.043]; Ammar[0.042]; Egypt[0.042]; said[0.042]; Libyans[0.042]; celebrations[0.042]; Tripoli[0.041]; improved[0.041]; killed[0.041]; east[0.041]; marking[0.041]; say[0.041]; September[0.041]; diplomats[0.041]; people[0.041]; Friday[0.041]; regularly[0.041]; July[0.040]; July[0.040]; clash[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = African ==> ENTITY: \u001b[32mAfrica\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.911:-0.911[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.046]; leaders[0.045]; diplomats[0.044]; Libya[0.044]; United[0.043]; Niger[0.043]; Pan[0.042]; Nations[0.042]; internal[0.042]; Guinea[0.042]; young[0.041]; army[0.041]; east[0.041]; group[0.041]; King[0.041]; Ghana[0.041]; brought[0.040]; Tripoli[0.040]; Tripoli[0.040]; Libyans[0.040]; revolution[0.040]; revolution[0.040]; revolution[0.040]; lives[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Pan Am\u001b[39m ==> ENTITY: \u001b[32mPan American World Airways\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.073:0.073[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLockerbie[0.048]; flight[0.045]; bombing[0.044]; Libya[0.043]; Scotland[0.043]; Libyans[0.043]; September[0.042]; flights[0.042]; stadium[0.042]; Tripoli[0.042]; Tripoli[0.042]; Tripoli[0.042]; Tripoli[0.042]; trial[0.042]; airport[0.041]; connection[0.041]; expected[0.040]; Sunday[0.040]; Sunday[0.040]; United[0.040]; marking[0.040]; suspects[0.040]; Niger[0.039]; attend[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gaddafi\u001b[39m ==> ENTITY: \u001b[32mMuammar Gaddafi\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGaddafi[0.046]; Gaddafi[0.046]; Tripoli[0.045]; Egypt[0.045]; Egyptian[0.043]; Benghazi[0.042]; country[0.042]; Sudanese[0.042]; militant[0.042]; deported[0.041]; rebel[0.041]; al[0.041]; July[0.041]; July[0.041]; diplomats[0.041]; based[0.041]; people[0.041]; said[0.040]; Akhdar[0.040]; football[0.040]; killed[0.040]; exiled[0.040]; officers[0.040]; bodyguards[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gaddafi\u001b[39m ==> ENTITY: \u001b[32mMuammar Gaddafi\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGaddafi[0.045]; Gaddafi[0.045]; Gaddafi[0.045]; Tripoli[0.044]; Libyans[0.044]; Egypt[0.044]; Ammar[0.042]; Egyptian[0.042]; Benghazi[0.041]; Mohammed[0.041]; September[0.041]; country[0.041]; Sudanese[0.041]; militant[0.041]; revolution[0.041]; revolution[0.041]; deposed[0.040]; deported[0.040]; group[0.040]; rebel[0.040]; al[0.040]; Friday[0.040]; July[0.040]; July[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Niger ==> ENTITY: \u001b[32mNiger\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.126:0.126[0]; log p(e|m)= -0.465:-0.465[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.049]; Ghana[0.045]; east[0.044]; violence[0.043]; Libya[0.043]; African[0.043]; Nations[0.043]; Guinea[0.043]; imposed[0.042]; opponents[0.042]; Scotland[0.041]; based[0.041]; diplomats[0.040]; given[0.040]; Libyans[0.040]; hotbed[0.040]; leaders[0.040]; group[0.040]; internal[0.040]; hand[0.040]; city[0.040]; officers[0.040]; army[0.040]; Mohammed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mona Eltahawy\u001b[39m ==> ENTITY: \u001b[32mMona Eltahawy\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.074:0.074[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLibya[0.044]; sanctions[0.043]; sanctions[0.043]; Tripoli[0.042]; Tripoli[0.042]; Tripoli[0.042]; Tripoli[0.042]; Sunday[0.042]; proclaimed[0.042]; challenge[0.042]; flight[0.042]; Libyans[0.042]; revolution[0.041]; Pan[0.041]; Gaddafi[0.041]; United[0.041]; internal[0.041]; hand[0.041]; promise[0.041]; great[0.041]; defiance[0.041]; banners[0.041]; banner[0.041]; praise[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = United Nations\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.087:0.087[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLibya[0.045]; Niger[0.044]; Square[0.044]; September[0.043]; group[0.043]; bombing[0.042]; Libyans[0.042]; Eltahawy[0.042]; marking[0.042]; refusal[0.042]; sanctions[0.041]; sanctions[0.041]; expected[0.041]; Tripoli[0.041]; Tripoli[0.041]; Tripoli[0.041]; Tripoli[0.041]; flags[0.041]; coup[0.041]; coup[0.041]; flights[0.040]; wanted[0.040]; African[0.040]; Ghana[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Egypt ==> ENTITY: \u001b[32mEgypt\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.153:0.153[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEgyptian[0.049]; al[0.048]; match[0.048]; Ammar[0.047]; Mohammed[0.047]; region[0.047]; country[0.047]; Jebel[0.046]; Akhdar[0.046]; football[0.045]; Samer[0.045]; Tripoli[0.045]; fired[0.044]; deposed[0.044]; dismissed[0.044]; Sudanese[0.044]; July[0.044]; July[0.044]; King[0.044]; militants[0.044]; unrest[0.043]; early[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ghana ==> ENTITY: \u001b[32mGhana\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.392:-0.392[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncountry[0.048]; African[0.045]; Nations[0.044]; Guinea[0.044]; stadium[0.043]; Niger[0.043]; Scotland[0.043]; Mohammed[0.042]; east[0.042]; Ammar[0.041]; brought[0.041]; given[0.041]; officers[0.041]; achievements[0.041]; came[0.041]; United[0.041]; September[0.041]; led[0.040]; Idris[0.040]; great[0.040]; great[0.040]; road[0.040]; marking[0.040]; improved[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lockerbie\u001b[39m ==> ENTITY: \u001b[32mLockerbie\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.117:0.117[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nflight[0.045]; Scotland[0.044]; bombing[0.044]; Libya[0.044]; Libyans[0.043]; flights[0.042]; Tripoli[0.042]; Tripoli[0.042]; Tripoli[0.042]; Tripoli[0.042]; trial[0.042]; September[0.042]; Green[0.041]; Green[0.041]; Square[0.041]; airport[0.041]; marking[0.040]; Sunday[0.040]; Sunday[0.040]; road[0.040]; suspects[0.040]; old[0.040]; city[0.040]; African[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Egyptian ==> ENTITY: \u001b[32mEgypt\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.130:0.130[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEgypt[0.054]; al[0.045]; Akhdar[0.045]; foreigners[0.044]; say[0.044]; Jebel[0.044]; people[0.044]; region[0.044]; Benghazi[0.043]; deported[0.043]; Sudanese[0.043]; fired[0.043]; dismissed[0.043]; workers[0.042]; July[0.042]; July[0.042]; thousands[0.042]; militants[0.042]; exiled[0.042]; officers[0.042]; said[0.042]; unrest[0.042]; arriving[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gaddafi\u001b[39m ==> ENTITY: \u001b[32mMuammar Gaddafi\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGaddafi[0.046]; Gaddafi[0.046]; Tripoli[0.045]; Libyans[0.045]; Egypt[0.045]; Ammar[0.043]; Egyptian[0.043]; Benghazi[0.042]; country[0.041]; Sudanese[0.041]; militant[0.041]; deported[0.041]; rebel[0.041]; al[0.041]; Friday[0.040]; July[0.040]; July[0.040]; diplomats[0.040]; based[0.040]; people[0.040]; told[0.040]; said[0.040]; Akhdar[0.040]; football[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Muammar Gaddafi\u001b[39m ==> ENTITY: \u001b[32mMuammar Gaddafi\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLockerbie[0.048]; Libya[0.045]; flight[0.044]; Tripoli[0.044]; Tripoli[0.044]; Tripoli[0.044]; Tripoli[0.044]; Libyans[0.044]; bombing[0.044]; Scotland[0.043]; flights[0.042]; trial[0.041]; coup[0.041]; coup[0.041]; revolution[0.040]; airport[0.040]; Sunday[0.039]; Sunday[0.039]; suspects[0.039]; anniversary[0.039]; African[0.038]; Eltahawy[0.038]; sanctions[0.038]; sanctions[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Gaddafi\u001b[39m ==> ENTITY: \u001b[32mMuammar Gaddafi\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLockerbie[0.048]; Gaddafi[0.045]; flight[0.044]; Tripoli[0.044]; Tripoli[0.044]; Libyans[0.044]; bombing[0.043]; Scotland[0.042]; Ammar[0.042]; flights[0.041]; trial[0.041]; Mohammed[0.041]; September[0.041]; country[0.041]; militant[0.040]; revolution[0.040]; revolution[0.040]; deposed[0.040]; group[0.040]; airport[0.040]; rebel[0.040]; al[0.040]; leaders[0.040]; Friday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Libya ==> ENTITY: \u001b[32mLibya\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.305:-0.305[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTripoli[0.047]; Tripoli[0.047]; Tripoli[0.047]; Tripoli[0.047]; Libyans[0.046]; Muammar[0.045]; Gaddafi[0.044]; revolution[0.044]; coup[0.044]; coup[0.044]; Niger[0.043]; September[0.043]; sanctions[0.043]; sanctions[0.043]; leaders[0.042]; army[0.042]; Nations[0.042]; stadium[0.042]; group[0.042]; await[0.042]; trial[0.042]; imposed[0.041]; African[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sudanese ==> ENTITY: \u001b[32mSudan\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.198:-0.198[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nregion[0.046]; football[0.045]; Egypt[0.045]; Tripoli[0.044]; people[0.044]; Jebel[0.043]; Benghazi[0.043]; Egyptian[0.042]; foreigners[0.042]; al[0.042]; unrest[0.042]; diplomats[0.041]; mountainous[0.041]; Akhdar[0.041]; arriving[0.041]; militants[0.041]; deported[0.040]; thousands[0.040]; match[0.040]; rebel[0.040]; workers[0.040]; Gaddafi[0.039]; Gaddafi[0.039]; Gaddafi[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mohammed Idris\u001b[39m ==> ENTITY: \u001b[32mIdris of Libya\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmmar[0.045]; Egypt[0.044]; Tripoli[0.043]; Tripoli[0.043]; King[0.043]; deposed[0.043]; region[0.043]; country[0.042]; Gaddafi[0.042]; Gaddafi[0.042]; given[0.042]; al[0.042]; revolution[0.041]; revolution[0.041]; September[0.041]; Akhdar[0.041]; exiled[0.041]; Libyans[0.040]; July[0.040]; Jebel[0.040]; rebel[0.040]; African[0.040]; expected[0.040]; brought[0.040]; \t\n\r [=====================================>........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 3s652ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4042/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1103testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexico ==> ENTITY: \u001b[32mMexico national football team\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.157:0.157[0]; log p(e|m)= -2.865:-2.865[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.065]; Soccer[0.062]; soccer[0.061]; Djorkaeff[0.057]; international[0.056]; France[0.053]; France[0.053]; friendly[0.053]; minute[0.051]; Saturday[0.050]; Youri[0.050]; Nicolas[0.050]; Friendly[0.050]; beat[0.050]; halftime[0.050]; Attendance[0.049]; Beat[0.048]; Scorers[0.047]; Paris[0.045]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mFrance\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFrance\u001b[39m <---> \u001b[32mFrance national football team\u001b[39m\t\nSCORES: global= 0.246:0.244[\u001b[31m0.002\u001b[39m]; local()= 0.082:0.143[\u001b[32m0.061\u001b[39m]; log p(e|m)= -0.163:-3.381[\u001b[31m3.219\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.064]; soccer[0.059]; Soccer[0.058]; Saturday[0.056]; international[0.056]; Paris[0.055]; Djorkaeff[0.055]; Mexico[0.052]; Mexico[0.052]; friendly[0.051]; Nicolas[0.051]; Youri[0.050]; Scorers[0.050]; minute[0.050]; halftime[0.050]; beat[0.049]; Friendly[0.049]; Attendance[0.047]; Beat[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Youri Djorkaeff\u001b[39m ==> ENTITY: \u001b[32mYouri Djorkaeff\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.126:0.126[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.063]; Soccer[0.062]; Nicolas[0.061]; international[0.060]; France[0.060]; France[0.060]; Mexico[0.056]; Mexico[0.056]; halftime[0.055]; Scorers[0.055]; Attendance[0.055]; Saturday[0.053]; Paris[0.053]; minute[0.052]; friendly[0.052]; Friendly[0.051]; beat[0.048]; Beat[0.048]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Paris ==> ENTITY: \u001b[32mParis\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.069:0.069[0]; log p(e|m)= -0.115:-0.115[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.061]; France[0.061]; Nicolas[0.059]; Soccer[0.059]; soccer[0.056]; Saturday[0.056]; beat[0.054]; international[0.052]; Attendance[0.051]; Youri[0.050]; Djorkaeff[0.050]; Mexico[0.050]; Mexico[0.050]; minute[0.050]; Beat[0.050]; friendly[0.050]; Scorers[0.048]; halftime[0.047]; Friendly[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nicolas Ouedec\u001b[39m ==> ENTITY: \u001b[32mNicolas Ouédec\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.097:0.097[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.059]; France[0.058]; France[0.058]; soccer[0.057]; Scorers[0.055]; Youri[0.054]; Paris[0.054]; Djorkaeff[0.053]; halftime[0.053]; Attendance[0.053]; international[0.052]; beat[0.051]; minute[0.051]; Saturday[0.050]; friendly[0.050]; Mexico[0.049]; Mexico[0.049]; Friendly[0.048]; Beat[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mexico ==> ENTITY: \u001b[32mMexico national football team\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.157:0.157[0]; log p(e|m)= -2.865:-2.865[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMexico[0.065]; Soccer[0.062]; soccer[0.061]; Djorkaeff[0.057]; international[0.056]; France[0.053]; France[0.053]; friendly[0.053]; minute[0.051]; Saturday[0.050]; Youri[0.050]; Nicolas[0.050]; Friendly[0.050]; beat[0.050]; halftime[0.050]; Attendance[0.049]; Beat[0.048]; Scorers[0.047]; Paris[0.045]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mFrance\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFrance\u001b[39m <---> \u001b[32mFrance national football team\u001b[39m\t\nSCORES: global= 0.246:0.244[\u001b[31m0.002\u001b[39m]; local()= 0.082:0.143[\u001b[32m0.061\u001b[39m]; log p(e|m)= -0.163:-3.381[\u001b[31m3.219\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.064]; soccer[0.059]; Soccer[0.058]; Saturday[0.056]; international[0.056]; Paris[0.055]; Djorkaeff[0.055]; Mexico[0.052]; Mexico[0.052]; friendly[0.051]; Nicolas[0.051]; Youri[0.050]; Scorers[0.050]; minute[0.050]; halftime[0.050]; beat[0.049]; Friendly[0.049]; Attendance[0.047]; Beat[0.046]; \t\n\r [=====================================>........] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 3s618ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4049/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1070testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.016:0.016[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHampshire[0.047]; Hampshire[0.047]; Hampshire[0.047]; Durham[0.045]; Durham[0.045]; Durham[0.045]; matches[0.044]; Championship[0.044]; Worcestershire[0.043]; Worcestershire[0.043]; Middlesex[0.043]; Middlesex[0.043]; Middlesex[0.043]; Derbyshire[0.043]; Derbyshire[0.043]; Derbyshire[0.043]; Cricket[0.043]; Tufnell[0.042]; Glamorgan[0.042]; Glamorgan[0.042]; Glamorgan[0.042]; championship[0.041]; Chesterfield[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Warwickshire ==> ENTITY: \u001b[32mWarwickshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.197:0.197[0]; log p(e|m)= -1.109:-1.109[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.048]; Warwickshire[0.048]; Essex[0.046]; Lancashire[0.046]; Durham[0.046]; Durham[0.046]; Durham[0.046]; Worcestershire[0.045]; Worcestershire[0.045]; Worcestershire[0.045]; Glamorgan[0.045]; Glamorgan[0.045]; Glamorgan[0.045]; Gloucestershire[0.045]; Sussex[0.045]; Northamptonshire[0.045]; Nottinghamshire[0.045]; Derbyshire[0.044]; Derbyshire[0.044]; Derbyshire[0.044]; Kent[0.044]; Kettleborough[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Derbyshire ==> ENTITY: \u001b[32mDerbyshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.241:0.241[0]; log p(e|m)= -1.483:-1.483[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; Warwickshire[0.044]; Warwickshire[0.044]; Hampshire[0.042]; Hampshire[0.042]; Hampshire[0.042]; Lancashire[0.042]; Essex[0.042]; Worcestershire[0.042]; Worcestershire[0.042]; Worcestershire[0.042]; Sussex[0.041]; Glamorgan[0.041]; Glamorgan[0.041]; Glamorgan[0.041]; Nottinghamshire[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Surrey[0.041]; Surrey[0.041]; Surrey[0.041]; Middlesex[0.041]; Middlesex[0.041]; Middlesex[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mHeadingley\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHeadingley Stadium\u001b[39m <---> \u001b[32mHeadingley\u001b[39m\t\nSCORES: global= 0.270:0.260[\u001b[31m0.011\u001b[39m]; local()= 0.177:0.182[\u001b[32m0.005\u001b[39m]; log p(e|m)= -0.302:-0.571[\u001b[31m0.268\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOval[0.047]; Lancashire[0.047]; Warwickshire[0.047]; Warwickshire[0.047]; Warwickshire[0.047]; Leeds[0.047]; Essex[0.046]; Gloucestershire[0.046]; Glamorgan[0.046]; Glamorgan[0.046]; Worcestershire[0.045]; Worcestershire[0.045]; Worcestershire[0.045]; Nottinghamshire[0.045]; Yorkshire[0.045]; Sussex[0.044]; Derbyshire[0.044]; Derbyshire[0.044]; Derbyshire[0.044]; Bristol[0.044]; Northamptonshire[0.044]; Kent[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Glamorgan ==> ENTITY: \u001b[32mGlamorgan County Cricket Club\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.214:0.214[0]; log p(e|m)= -0.790:-0.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; Warwickshire[0.044]; Warwickshire[0.044]; Glamorgan[0.043]; Glamorgan[0.043]; Hampshire[0.042]; Hampshire[0.042]; Hampshire[0.042]; matches[0.042]; Worcestershire[0.041]; Worcestershire[0.041]; Worcestershire[0.041]; Middlesex[0.041]; Middlesex[0.041]; Middlesex[0.041]; innings[0.041]; wickets[0.041]; Durham[0.041]; Durham[0.041]; Durham[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Oval[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Middlesex ==> ENTITY: \u001b[32mMiddlesex County Cricket Club\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.214:0.214[0]; log p(e|m)= -1.570:-1.570[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMiddlesex[0.043]; Middlesex[0.043]; Hampshire[0.043]; Hampshire[0.043]; Hampshire[0.043]; Warwickshire[0.043]; Warwickshire[0.043]; Surrey[0.043]; Surrey[0.043]; Durham[0.042]; Durham[0.042]; Durham[0.042]; matches[0.041]; Glamorgan[0.041]; Glamorgan[0.041]; Glamorgan[0.041]; Oval[0.041]; Worcestershire[0.041]; Worcestershire[0.041]; Worcestershire[0.041]; Derbyshire[0.040]; Derbyshire[0.040]; Derbyshire[0.040]; Cricket[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Warwickshire ==> ENTITY: \u001b[32mWarwickshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.214:0.214[0]; log p(e|m)= -1.109:-1.109[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.047]; Warwickshire[0.047]; Hampshire[0.047]; Hampshire[0.047]; Hampshire[0.047]; Worcestershire[0.046]; Worcestershire[0.046]; Worcestershire[0.046]; Essex[0.046]; Lancashire[0.045]; Durham[0.045]; Durham[0.045]; Durham[0.045]; Middlesex[0.045]; Middlesex[0.045]; Middlesex[0.045]; Glamorgan[0.045]; Glamorgan[0.045]; Glamorgan[0.045]; Gloucestershire[0.045]; Sussex[0.044]; Nottinghamshire[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hampshire ==> ENTITY: \u001b[32mHampshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.216:0.216[0]; log p(e|m)= -1.924:-1.924[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; Warwickshire[0.044]; Hampshire[0.043]; Hampshire[0.043]; Middlesex[0.042]; Middlesex[0.042]; Middlesex[0.042]; Glamorgan[0.042]; Glamorgan[0.042]; Glamorgan[0.042]; Durham[0.042]; Durham[0.042]; Durham[0.042]; Worcestershire[0.042]; Worcestershire[0.042]; Worcestershire[0.042]; Derbyshire[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Surrey[0.040]; matches[0.040]; Oval[0.040]; County[0.039]; Tufnell[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Derbyshire ==> ENTITY: \u001b[32mDerbyshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.237:0.237[0]; log p(e|m)= -1.483:-1.483[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; Warwickshire[0.044]; Warwickshire[0.044]; Hampshire[0.043]; Hampshire[0.043]; Hampshire[0.043]; Essex[0.042]; Worcestershire[0.042]; Worcestershire[0.042]; Worcestershire[0.042]; Glamorgan[0.041]; Glamorgan[0.041]; Glamorgan[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Surrey[0.041]; Surrey[0.041]; Surrey[0.041]; Middlesex[0.041]; Middlesex[0.041]; Middlesex[0.041]; Durham[0.041]; Durham[0.041]; Durham[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nottinghamshire ==> ENTITY: \u001b[32mNottinghamshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.186:0.186[0]; log p(e|m)= -1.106:-1.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.045]; Warwickshire[0.045]; Warwickshire[0.045]; Essex[0.044]; Lancashire[0.043]; Sussex[0.043]; Northamptonshire[0.043]; Worcestershire[0.043]; Gloucestershire[0.042]; Derbyshire[0.042]; Derbyshire[0.042]; Surrey[0.042]; Surrey[0.042]; Surrey[0.042]; Kent[0.041]; Kettleborough[0.041]; Headingley[0.040]; Yorkshire[0.040]; Oval[0.040]; Bristol[0.039]; innings[0.039]; DeFreitas[0.039]; runs[0.039]; Hove[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.049:0.049[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSurrey[0.046]; Surrey[0.046]; Surrey[0.046]; Hampshire[0.045]; Hampshire[0.045]; Hampshire[0.045]; Essex[0.045]; Sussex[0.044]; Durham[0.044]; Durham[0.044]; Durham[0.044]; Warwickshire[0.043]; Warwickshire[0.043]; Warwickshire[0.043]; matches[0.043]; Lancashire[0.042]; Nottinghamshire[0.042]; Worcestershire[0.042]; Worcestershire[0.042]; Worcestershire[0.042]; Middlesex[0.042]; Middlesex[0.042]; Middlesex[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Glamorgan ==> ENTITY: \u001b[32mGlamorgan County Cricket Club\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.214:0.214[0]; log p(e|m)= -0.790:-0.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; Warwickshire[0.044]; Warwickshire[0.044]; Glamorgan[0.043]; Glamorgan[0.043]; Hampshire[0.042]; Hampshire[0.042]; Hampshire[0.042]; matches[0.042]; Worcestershire[0.041]; Worcestershire[0.041]; Worcestershire[0.041]; Middlesex[0.041]; Middlesex[0.041]; Middlesex[0.041]; innings[0.041]; wickets[0.041]; Durham[0.041]; Durham[0.041]; Durham[0.041]; Kettleborough[0.041]; Derbyshire[0.040]; Derbyshire[0.040]; Derbyshire[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chester-le-Street ==> ENTITY: \u001b[32mRiverside Ground\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.186:0.186[0]; log p(e|m)= -1.452:-1.452[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; Durham[0.043]; Durham[0.043]; Durham[0.043]; Warwickshire[0.043]; Warwickshire[0.043]; Hampshire[0.042]; Hampshire[0.042]; Hampshire[0.042]; Oval[0.042]; county[0.042]; day[0.041]; Worcestershire[0.041]; Worcestershire[0.041]; Worcestershire[0.041]; Middlesex[0.041]; Middlesex[0.041]; Middlesex[0.041]; Street[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Cricket[0.040]; County[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leeds ==> ENTITY: \u001b[32mLeeds\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.166:0.166[0]; log p(e|m)= -0.255:-0.255[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBristol[0.045]; Oval[0.045]; Headingley[0.045]; Lancashire[0.045]; Warwickshire[0.045]; Warwickshire[0.045]; Warwickshire[0.045]; Essex[0.044]; Yorkshire[0.043]; Gloucestershire[0.043]; Glamorgan[0.043]; Glamorgan[0.043]; Worcestershire[0.043]; Worcestershire[0.043]; Worcestershire[0.043]; Nottinghamshire[0.043]; Sussex[0.042]; Derbyshire[0.042]; Derbyshire[0.042]; Derbyshire[0.042]; London[0.042]; Northamptonshire[0.042]; Chesterfield[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Middlesex ==> ENTITY: \u001b[32mMiddlesex County Cricket Club\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.209:0.209[0]; log p(e|m)= -1.570:-1.570[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMiddlesex[0.044]; Middlesex[0.044]; Hampshire[0.043]; Hampshire[0.043]; Hampshire[0.043]; Warwickshire[0.043]; Warwickshire[0.043]; Surrey[0.043]; Durham[0.042]; Durham[0.042]; Durham[0.042]; matches[0.042]; Glamorgan[0.041]; Glamorgan[0.041]; Glamorgan[0.041]; Oval[0.041]; Worcestershire[0.041]; Worcestershire[0.041]; Worcestershire[0.041]; Derbyshire[0.040]; Derbyshire[0.040]; Derbyshire[0.040]; Cricket[0.040]; Tufnell[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Derbyshire ==> ENTITY: \u001b[32mDerbyshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.241:0.241[0]; log p(e|m)= -1.483:-1.483[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; Warwickshire[0.044]; Warwickshire[0.044]; Hampshire[0.042]; Hampshire[0.042]; Hampshire[0.042]; Lancashire[0.042]; Essex[0.042]; Worcestershire[0.042]; Worcestershire[0.042]; Worcestershire[0.042]; Sussex[0.041]; Glamorgan[0.041]; Glamorgan[0.041]; Glamorgan[0.041]; Nottinghamshire[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Surrey[0.041]; Surrey[0.041]; Surrey[0.041]; Middlesex[0.041]; Middlesex[0.041]; Middlesex[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Yorkshire ==> ENTITY: \u001b[32mYorkshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.228:0.228[0]; log p(e|m)= -1.778:-1.778[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; Warwickshire[0.044]; Warwickshire[0.044]; Lancashire[0.043]; Essex[0.043]; Worcestershire[0.042]; Worcestershire[0.042]; Worcestershire[0.042]; Glamorgan[0.042]; Glamorgan[0.042]; Gloucestershire[0.042]; Durham[0.041]; Durham[0.041]; Northamptonshire[0.041]; Nottinghamshire[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Sussex[0.041]; Headingley[0.041]; Surrey[0.040]; Surrey[0.040]; Surrey[0.040]; Bristol[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = English ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.121:0.121[0]; log p(e|m)= -1.355:-1.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEnglish[0.046]; matches[0.043]; Warwickshire[0.043]; Oval[0.042]; Cricket[0.042]; Hampshire[0.042]; Hampshire[0.042]; Hampshire[0.042]; Worcestershire[0.041]; Worcestershire[0.041]; Worcestershire[0.041]; Middlesex[0.041]; Middlesex[0.041]; Middlesex[0.041]; Glamorgan[0.041]; Glamorgan[0.041]; Glamorgan[0.041]; Chesterfield[0.041]; Portsmouth[0.041]; Surrey[0.041]; Championship[0.041]; Durham[0.041]; Durham[0.041]; Durham[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Surrey ==> ENTITY: \u001b[32mSurrey County Cricket Club\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.231:0.231[0]; log p(e|m)= -1.796:-1.796[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHampshire[0.045]; Hampshire[0.045]; Warwickshire[0.045]; Warwickshire[0.045]; Warwickshire[0.045]; Essex[0.045]; Middlesex[0.044]; Surrey[0.044]; Surrey[0.044]; Lancashire[0.044]; Sussex[0.043]; Gloucestershire[0.043]; Worcestershire[0.043]; Worcestershire[0.043]; Worcestershire[0.043]; Durham[0.043]; Durham[0.043]; Durham[0.043]; Northamptonshire[0.042]; Glamorgan[0.042]; Glamorgan[0.042]; Glamorgan[0.042]; Nottinghamshire[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Worcestershire ==> ENTITY: \u001b[32mWorcestershire County Cricket Club\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.213:0.213[0]; log p(e|m)= -0.916:-0.916[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWorcestershire[0.047]; Worcestershire[0.047]; Warwickshire[0.047]; Warwickshire[0.047]; Warwickshire[0.047]; Essex[0.046]; county[0.046]; Hampshire[0.046]; Hampshire[0.046]; Hampshire[0.046]; Middlesex[0.045]; Middlesex[0.045]; Middlesex[0.045]; wickets[0.045]; Glamorgan[0.045]; Glamorgan[0.045]; Glamorgan[0.045]; County[0.045]; Durham[0.044]; Durham[0.044]; Durham[0.044]; innings[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Durham ==> ENTITY: \u001b[32mDurham County Cricket Club\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.241:0.241[0]; log p(e|m)= -2.397:-2.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.047]; Warwickshire[0.047]; Warwickshire[0.047]; Middlesex[0.046]; Middlesex[0.046]; Middlesex[0.046]; Hampshire[0.046]; Hampshire[0.046]; Hampshire[0.046]; Durham[0.046]; Durham[0.046]; Worcestershire[0.045]; Worcestershire[0.045]; Worcestershire[0.045]; Glamorgan[0.045]; Glamorgan[0.045]; Glamorgan[0.045]; matches[0.044]; Derbyshire[0.044]; Derbyshire[0.044]; Derbyshire[0.044]; Cricket[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kent ==> ENTITY: \u001b[32mKent County Cricket Club\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.185:0.185[0]; log p(e|m)= -2.513:-2.513[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEssex[0.045]; Warwickshire[0.045]; Warwickshire[0.045]; Warwickshire[0.045]; Sussex[0.044]; Surrey[0.044]; Surrey[0.044]; Surrey[0.044]; Lancashire[0.043]; Worcestershire[0.043]; Gloucestershire[0.043]; Northamptonshire[0.042]; Derbyshire[0.042]; Kettleborough[0.041]; Nottinghamshire[0.041]; Bristol[0.041]; Hove[0.041]; Yorkshire[0.040]; Headingley[0.039]; Oval[0.039]; runs[0.038]; Wells[0.038]; innings[0.038]; Tunbridge[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hampshire ==> ENTITY: \u001b[32mHampshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.219:0.219[0]; log p(e|m)= -1.924:-1.924[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; Warwickshire[0.044]; Hampshire[0.043]; Hampshire[0.043]; Middlesex[0.042]; Middlesex[0.042]; Middlesex[0.042]; Glamorgan[0.042]; Glamorgan[0.042]; Glamorgan[0.042]; Durham[0.042]; Durham[0.042]; Durham[0.042]; Worcestershire[0.042]; Worcestershire[0.042]; Worcestershire[0.042]; Derbyshire[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Surrey[0.040]; Surrey[0.040]; matches[0.040]; Oval[0.040]; County[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Surrey ==> ENTITY: \u001b[32mSurrey County Cricket Club\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.238:0.238[0]; log p(e|m)= -1.796:-1.796[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHampshire[0.043]; Hampshire[0.043]; Hampshire[0.043]; Warwickshire[0.043]; Warwickshire[0.043]; Warwickshire[0.043]; Essex[0.043]; Middlesex[0.042]; Middlesex[0.042]; Middlesex[0.042]; Surrey[0.042]; Surrey[0.042]; Lancashire[0.042]; Sussex[0.041]; Worcestershire[0.041]; Worcestershire[0.041]; Worcestershire[0.041]; Durham[0.041]; Durham[0.041]; Durham[0.041]; Glamorgan[0.040]; Glamorgan[0.040]; Glamorgan[0.040]; Nottinghamshire[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBristol\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBristol\u001b[39m <---> \u001b[32mCounty Cricket Ground, Bristol\u001b[39m\t\nSCORES: global= 0.260:0.242[\u001b[31m0.018\u001b[39m]; local()= 0.127:0.151[\u001b[32m0.024\u001b[39m]; log p(e|m)= -0.435:-4.343[\u001b[31m3.907\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.047]; Warwickshire[0.047]; Lancashire[0.047]; Gloucestershire[0.045]; Essex[0.045]; Headingley[0.045]; Nottinghamshire[0.045]; Hove[0.044]; Northamptonshire[0.044]; Sussex[0.044]; innings[0.044]; runs[0.044]; Surrey[0.044]; Surrey[0.044]; Kent[0.043]; Yorkshire[0.043]; Leeds[0.043]; Russell[0.042]; Kettleborough[0.042]; Wells[0.041]; Ratcliffe[0.041]; points[0.040]; Lewis[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Durham ==> ENTITY: \u001b[32mDurham County Cricket Club\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.243:0.243[0]; log p(e|m)= -2.397:-2.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.045]; Warwickshire[0.045]; Warwickshire[0.045]; Essex[0.045]; Middlesex[0.044]; Middlesex[0.044]; Middlesex[0.044]; Hampshire[0.044]; Hampshire[0.044]; Hampshire[0.044]; Durham[0.044]; Durham[0.044]; Worcestershire[0.043]; Worcestershire[0.043]; Worcestershire[0.043]; Glamorgan[0.043]; Glamorgan[0.043]; Glamorgan[0.043]; matches[0.042]; Derbyshire[0.042]; Derbyshire[0.042]; Derbyshire[0.042]; Cricket[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Essex ==> ENTITY: \u001b[32mEssex County Cricket Club\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.218:0.218[0]; log p(e|m)= -2.343:-2.343[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; Warwickshire[0.044]; Warwickshire[0.044]; Lancashire[0.043]; Worcestershire[0.043]; Worcestershire[0.043]; Worcestershire[0.043]; Gloucestershire[0.042]; Sussex[0.042]; Kettleborough[0.042]; Surrey[0.042]; Surrey[0.042]; Surrey[0.042]; Northamptonshire[0.041]; Nottinghamshire[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Durham[0.041]; Headingley[0.040]; Kent[0.040]; Oval[0.040]; Yorkshire[0.040]; wickets[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mHove\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHove\u001b[39m <---> \u001b[32mCounty Cricket Ground, Hove\u001b[39m\t\nSCORES: global= 0.270:0.257[\u001b[31m0.012\u001b[39m]; local()= 0.165:0.179[\u001b[32m0.014\u001b[39m]; log p(e|m)= -0.041:-2.703[\u001b[31m2.662\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSussex[0.044]; Essex[0.043]; Oval[0.043]; Warwickshire[0.043]; Warwickshire[0.043]; Warwickshire[0.043]; Nottinghamshire[0.042]; Surrey[0.042]; Surrey[0.042]; Surrey[0.042]; Headingley[0.042]; Lancashire[0.042]; Bristol[0.041]; Worcestershire[0.041]; Worcestershire[0.041]; Worcestershire[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Gloucestershire[0.041]; Kent[0.041]; Northamptonshire[0.040]; London[0.040]; Yorkshire[0.040]; wickets[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Worcestershire ==> ENTITY: \u001b[32mWorcestershire County Cricket Club\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.219:0.219[0]; log p(e|m)= -0.916:-0.916[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWorcestershire[0.043]; Worcestershire[0.043]; Warwickshire[0.043]; Warwickshire[0.043]; Warwickshire[0.043]; Essex[0.042]; county[0.042]; Hampshire[0.042]; Hampshire[0.042]; Hampshire[0.042]; Lancashire[0.042]; Middlesex[0.041]; Middlesex[0.041]; Middlesex[0.041]; wickets[0.041]; Glamorgan[0.041]; Glamorgan[0.041]; Glamorgan[0.041]; Nottinghamshire[0.041]; Sussex[0.041]; Durham[0.041]; Durham[0.041]; Durham[0.041]; innings[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Durham ==> ENTITY: \u001b[32mDurham County Cricket Club\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.241:0.241[0]; log p(e|m)= -2.397:-2.397[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.047]; Warwickshire[0.047]; Warwickshire[0.047]; Middlesex[0.046]; Middlesex[0.046]; Middlesex[0.046]; Hampshire[0.046]; Hampshire[0.046]; Hampshire[0.046]; Durham[0.046]; Durham[0.046]; Worcestershire[0.045]; Worcestershire[0.045]; Worcestershire[0.045]; Glamorgan[0.045]; Glamorgan[0.045]; Glamorgan[0.045]; matches[0.044]; Derbyshire[0.044]; Derbyshire[0.044]; Derbyshire[0.044]; Cricket[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hampshire ==> ENTITY: \u001b[32mHampshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.219:0.219[0]; log p(e|m)= -1.924:-1.924[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; Warwickshire[0.044]; Hampshire[0.043]; Hampshire[0.043]; Middlesex[0.042]; Middlesex[0.042]; Middlesex[0.042]; Glamorgan[0.042]; Glamorgan[0.042]; Glamorgan[0.042]; Durham[0.042]; Durham[0.042]; Durham[0.042]; Worcestershire[0.042]; Worcestershire[0.042]; Worcestershire[0.042]; Derbyshire[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Surrey[0.040]; Surrey[0.040]; matches[0.040]; Oval[0.040]; County[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Surrey ==> ENTITY: \u001b[32mSurrey County Cricket Club\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.222:0.222[0]; log p(e|m)= -1.796:-1.796[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; Warwickshire[0.044]; Warwickshire[0.044]; Essex[0.044]; Surrey[0.043]; Surrey[0.043]; Lancashire[0.042]; Sussex[0.042]; Gloucestershire[0.042]; Worcestershire[0.042]; Worcestershire[0.042]; Worcestershire[0.042]; Durham[0.041]; Durham[0.041]; Durham[0.041]; Northamptonshire[0.041]; Glamorgan[0.041]; Glamorgan[0.041]; Glamorgan[0.041]; Nottinghamshire[0.041]; Derbyshire[0.040]; Derbyshire[0.040]; Derbyshire[0.040]; Kent[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = English County Championship\u001b[39m ==> ENTITY: \u001b[32mCounty Championship\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.225:0.225[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHampshire[0.043]; Hampshire[0.043]; Hampshire[0.043]; matches[0.043]; Worcestershire[0.043]; Worcestershire[0.043]; Middlesex[0.043]; Middlesex[0.043]; Middlesex[0.043]; Glamorgan[0.043]; Glamorgan[0.043]; Glamorgan[0.043]; Durham[0.043]; Durham[0.043]; Durham[0.043]; Derbyshire[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Cricket[0.041]; Tufnell[0.040]; wickets[0.040]; county[0.038]; runs[0.038]; runs[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portsmouth ==> ENTITY: \u001b[32mPortsmouth\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.124:0.124[0]; log p(e|m)= -0.755:-0.755[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.048]; Championship[0.047]; Chesterfield[0.047]; Durham[0.047]; Durham[0.047]; Durham[0.047]; Hampshire[0.046]; Hampshire[0.046]; Hampshire[0.046]; Middlesex[0.046]; Middlesex[0.046]; Middlesex[0.046]; county[0.046]; English[0.045]; English[0.045]; County[0.044]; London[0.044]; London[0.044]; Surrey[0.044]; Street[0.044]; Chester[0.044]; Saturday[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Worcestershire ==> ENTITY: \u001b[32mWorcestershire County Cricket Club\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.215:0.215[0]; log p(e|m)= -0.916:-0.916[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWorcestershire[0.045]; Worcestershire[0.045]; Warwickshire[0.045]; Warwickshire[0.045]; Warwickshire[0.045]; Essex[0.044]; county[0.044]; Hampshire[0.044]; Hampshire[0.044]; Hampshire[0.044]; Middlesex[0.043]; Middlesex[0.043]; Middlesex[0.043]; wickets[0.043]; Glamorgan[0.043]; Glamorgan[0.043]; Glamorgan[0.043]; County[0.043]; Sussex[0.043]; Durham[0.042]; Durham[0.042]; Durham[0.042]; innings[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Gloucestershire ==> ENTITY: \u001b[32mGloucestershire County Cricket Club\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.172:0.172[0]; log p(e|m)= -1.378:-1.378[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.046]; Warwickshire[0.046]; Essex[0.046]; Sussex[0.044]; Lancashire[0.044]; Bristol[0.043]; Northamptonshire[0.043]; Nottinghamshire[0.043]; Surrey[0.043]; Surrey[0.043]; Kent[0.043]; Kettleborough[0.042]; Headingley[0.042]; Leeds[0.040]; innings[0.040]; Yorkshire[0.040]; runs[0.040]; Hove[0.040]; Bicknell[0.039]; Bicknell[0.039]; Wells[0.039]; Lewis[0.039]; Russell[0.038]; points[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Warwickshire ==> ENTITY: \u001b[32mWarwickshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.208:0.208[0]; log p(e|m)= -1.109:-1.109[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.045]; Warwickshire[0.045]; Hampshire[0.045]; Hampshire[0.045]; Hampshire[0.045]; Worcestershire[0.044]; Worcestershire[0.044]; Worcestershire[0.044]; Essex[0.044]; Lancashire[0.043]; county[0.043]; Durham[0.043]; Durham[0.043]; Durham[0.043]; Middlesex[0.043]; Middlesex[0.043]; Middlesex[0.043]; Glamorgan[0.043]; Glamorgan[0.043]; Glamorgan[0.043]; Sussex[0.043]; English[0.042]; Nottinghamshire[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Northamptonshire ==> ENTITY: \u001b[32mNorthamptonshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.149:0.149[0]; log p(e|m)= -1.207:-1.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.048]; Essex[0.046]; Lancashire[0.046]; Gloucestershire[0.045]; Sussex[0.045]; Nottinghamshire[0.045]; Bristol[0.044]; Yorkshire[0.043]; Surrey[0.043]; Surrey[0.043]; Kent[0.043]; Kettleborough[0.043]; Headingley[0.042]; Hove[0.041]; Leeds[0.040]; Wells[0.039]; Bicknell[0.039]; Bicknell[0.039]; Russell[0.039]; Lewis[0.038]; Tunbridge[0.038]; Hamilton[0.038]; Benjamin[0.037]; Evans[0.037]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mChesterfield\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mChesterfield F.C.\u001b[39m <---> \u001b[32mChesterfield\u001b[39m\t\nSCORES: global= 0.255:0.255[\u001b[31m0.000\u001b[39m]; local()= 0.127:0.137[\u001b[32m0.011\u001b[39m]; log p(e|m)= -1.269:-0.939[\u001b[32m0.330\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEssex[0.043]; matches[0.043]; Leeds[0.043]; Portsmouth[0.043]; Oval[0.042]; Hampshire[0.042]; Hampshire[0.042]; Hampshire[0.042]; Chester[0.042]; Championship[0.042]; County[0.042]; Warwickshire[0.042]; Warwickshire[0.042]; Warwickshire[0.042]; Middlesex[0.042]; Middlesex[0.042]; Middlesex[0.042]; Derbyshire[0.042]; Derbyshire[0.042]; Derbyshire[0.042]; county[0.040]; Surrey[0.040]; Surrey[0.040]; Surrey[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Glamorgan ==> ENTITY: \u001b[32mGlamorgan County Cricket Club\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.214:0.214[0]; log p(e|m)= -0.790:-0.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; Warwickshire[0.044]; Warwickshire[0.044]; Glamorgan[0.043]; Glamorgan[0.043]; Hampshire[0.042]; Hampshire[0.042]; Hampshire[0.042]; matches[0.042]; Worcestershire[0.041]; Worcestershire[0.041]; Worcestershire[0.041]; Middlesex[0.041]; Middlesex[0.041]; Middlesex[0.041]; innings[0.041]; wickets[0.041]; Durham[0.041]; Durham[0.041]; Durham[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Oval[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sussex ==> ENTITY: \u001b[32mSussex County Cricket Club\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.181:0.181[0]; log p(e|m)= -1.556:-1.556[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; Warwickshire[0.044]; Warwickshire[0.044]; Essex[0.044]; Surrey[0.043]; Surrey[0.043]; Surrey[0.043]; Lancashire[0.042]; Oval[0.042]; Gloucestershire[0.042]; Worcestershire[0.042]; Worcestershire[0.042]; Kent[0.042]; Northamptonshire[0.041]; Nottinghamshire[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Kettleborough[0.041]; Hove[0.040]; London[0.040]; Bristol[0.040]; Leeds[0.040]; Headingley[0.040]; Yorkshire[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Lancashire ==> ENTITY: \u001b[32mLancashire County Cricket Club\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.182:0.182[0]; log p(e|m)= -1.732:-1.732[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWorcestershire[0.045]; Warwickshire[0.045]; Warwickshire[0.045]; Warwickshire[0.045]; Northamptonshire[0.044]; Essex[0.044]; Gloucestershire[0.043]; Derbyshire[0.042]; Derbyshire[0.042]; Sussex[0.042]; Nottinghamshire[0.042]; Kettleborough[0.042]; Surrey[0.041]; Surrey[0.041]; Surrey[0.041]; Bristol[0.041]; Kent[0.041]; Headingley[0.040]; Leeds[0.040]; Yorkshire[0.040]; Oval[0.039]; Hove[0.039]; innings[0.039]; DeFreitas[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = The Oval\u001b[39m ==> ENTITY: \u001b[32mThe Oval\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.209:0.209[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.047]; Hampshire[0.047]; Hampshire[0.047]; Hampshire[0.047]; Surrey[0.046]; Surrey[0.046]; Surrey[0.046]; Warwickshire[0.046]; Warwickshire[0.046]; Warwickshire[0.046]; Middlesex[0.046]; Middlesex[0.046]; Middlesex[0.046]; Essex[0.046]; Lancashire[0.045]; Sussex[0.045]; Glamorgan[0.044]; Glamorgan[0.044]; Glamorgan[0.044]; Headingley[0.044]; Street[0.044]; Tufnell[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Middlesex ==> ENTITY: \u001b[32mMiddlesex County Cricket Club\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.209:0.209[0]; log p(e|m)= -1.570:-1.570[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMiddlesex[0.044]; Middlesex[0.044]; Hampshire[0.043]; Hampshire[0.043]; Hampshire[0.043]; Warwickshire[0.043]; Warwickshire[0.043]; Surrey[0.043]; Durham[0.042]; Durham[0.042]; Durham[0.042]; matches[0.042]; Glamorgan[0.041]; Glamorgan[0.041]; Glamorgan[0.041]; Oval[0.041]; Worcestershire[0.041]; Worcestershire[0.041]; Worcestershire[0.041]; Derbyshire[0.040]; Derbyshire[0.040]; Derbyshire[0.040]; Cricket[0.040]; Tufnell[0.039]; \t\n\r [======================================>.......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 3s421ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4094/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 955testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jansher Khan\u001b[39m ==> ENTITY: \u001b[32mJansher Khan\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.046]; Open[0.044]; Open[0.044]; Squash[0.044]; number[0.044]; final[0.043]; England[0.043]; England[0.043]; Australia[0.042]; Australia[0.042]; Australia[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; beat[0.042]; Quarter[0.041]; Quarter[0.041]; Friday[0.040]; Scotland[0.040]; Nicol[0.040]; Cairns[0.039]; Final[0.039]; Derek[0.039]; Eyles[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.052]; Australia[0.052]; England[0.049]; England[0.049]; Ireland[0.047]; Pakistan[0.045]; Scotland[0.044]; Open[0.044]; Open[0.044]; Mark[0.043]; final[0.043]; Peter[0.042]; Cairns[0.042]; number[0.042]; Jansher[0.042]; Khan[0.041]; Squash[0.040]; Chris[0.040]; Rodney[0.040]; Derek[0.040]; Friday[0.040]; Anthony[0.040]; Final[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.052]; Australia[0.052]; England[0.049]; England[0.049]; Ireland[0.047]; Pakistan[0.045]; Scotland[0.044]; Open[0.044]; Open[0.044]; Mark[0.043]; final[0.043]; Peter[0.042]; Cairns[0.042]; number[0.042]; Jansher[0.042]; Khan[0.041]; Squash[0.040]; Chris[0.040]; Rodney[0.040]; Derek[0.040]; Friday[0.040]; Anthony[0.040]; Final[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Peter Nicol\u001b[39m ==> ENTITY: \u001b[32mPeter Nicol\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJansher[0.046]; Squash[0.046]; Open[0.044]; Open[0.044]; Scotland[0.044]; Australia[0.042]; Australia[0.042]; Australia[0.042]; England[0.042]; England[0.042]; number[0.042]; final[0.042]; Quarter[0.041]; Quarter[0.041]; Khan[0.041]; Mark[0.041]; Derek[0.041]; Ireland[0.040]; results[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; seeding[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.120:0.120[0]; log p(e|m)= -0.172:-0.172[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.050]; Australia[0.050]; England[0.047]; England[0.047]; Ireland[0.045]; Pakistan[0.043]; Scotland[0.042]; Open[0.042]; Open[0.042]; Mark[0.041]; final[0.041]; Peter[0.041]; Cairns[0.040]; number[0.040]; Jansher[0.040]; Khan[0.039]; Squash[0.039]; Chris[0.039]; Rodney[0.039]; Derek[0.039]; Nicol[0.039]; Friday[0.039]; Anthony[0.039]; Final[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Derek Ryan ==> ENTITY: \u001b[32mDerek Ryan (squash player)\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.145:0.145[0]; log p(e|m)= -1.386:-1.386[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.050]; Scotland[0.045]; England[0.045]; England[0.045]; Jansher[0.044]; Australia[0.044]; Australia[0.044]; Australia[0.044]; number[0.043]; Squash[0.042]; Chris[0.041]; Results[0.041]; Quarter[0.040]; Quarter[0.040]; Dan[0.040]; Pakistan[0.040]; Open[0.040]; Open[0.040]; Mark[0.040]; Anthony[0.039]; Cairns[0.039]; results[0.039]; Peter[0.039]; Nicol[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scotland ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.073:0.073[0]; log p(e|m)= -0.229:-0.229[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.048]; England[0.048]; Ireland[0.047]; Pakistan[0.045]; Australia[0.044]; Australia[0.044]; Australia[0.044]; Mark[0.043]; number[0.042]; Squash[0.042]; Nicol[0.042]; final[0.042]; Derek[0.042]; Walker[0.040]; Jansher[0.040]; Chris[0.040]; Ryan[0.040]; Khan[0.040]; Peter[0.039]; Rodney[0.039]; Final[0.039]; Cairns[0.038]; Dan[0.038]; Friday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hong Kong ==> ENTITY: \u001b[32mHong Kong\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPakistan[0.046]; Australia[0.046]; Australia[0.046]; Australia[0.046]; Squash[0.045]; Kong[0.044]; Kong[0.044]; Open[0.044]; Open[0.044]; Results[0.044]; Jansher[0.044]; Hong[0.044]; Hong[0.044]; Final[0.044]; England[0.043]; England[0.043]; final[0.043]; Ireland[0.042]; Scotland[0.042]; number[0.041]; Hill[0.041]; Khan[0.040]; Nicol[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mark Cairns ==> ENTITY: \u001b[32mMark Cairns (squash player)\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.120:0.120[0]; log p(e|m)= -0.183:-0.183[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSquash[0.047]; Australia[0.046]; Australia[0.046]; Australia[0.046]; Walker[0.045]; Scotland[0.045]; Ireland[0.045]; England[0.045]; England[0.045]; Chris[0.044]; Peter[0.044]; Ryan[0.044]; Jansher[0.043]; Quarter[0.043]; Quarter[0.043]; Nicol[0.042]; Anthony[0.042]; Dan[0.042]; Derek[0.042]; Friday[0.042]; Rodney[0.041]; number[0.040]; final[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIreland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIreland\u001b[39m <---> \u001b[32mRepublic of Ireland\u001b[39m\t\nSCORES: global= 0.245:0.242[\u001b[31m0.003\u001b[39m]; local()= 0.098:0.088[\u001b[31m0.011\u001b[39m]; log p(e|m)= -0.566:-1.505[\u001b[31m0.939\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; England[0.047]; Scotland[0.046]; Australia[0.046]; Australia[0.046]; Australia[0.046]; Pakistan[0.045]; Final[0.042]; final[0.042]; Jansher[0.041]; number[0.040]; Mark[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; Peter[0.040]; Squash[0.040]; Ryan[0.039]; Nicol[0.039]; Chris[0.039]; Derek[0.039]; Friday[0.039]; Cairns[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chris Walker ==> ENTITY: \u001b[32mChris Walker (squash player)\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.153:0.153[0]; log p(e|m)= -1.496:-1.496[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSquash[0.045]; England[0.045]; England[0.045]; Nicol[0.044]; Open[0.043]; Open[0.043]; Mark[0.043]; Jansher[0.042]; Anthony[0.042]; final[0.042]; Ireland[0.042]; Rodney[0.042]; Ryan[0.041]; Cairns[0.041]; Derek[0.041]; Final[0.040]; Australia[0.040]; Australia[0.040]; Australia[0.040]; Scotland[0.040]; Dan[0.040]; Peter[0.039]; number[0.039]; Hill[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Dan Jenson\u001b[39m ==> ENTITY: \u001b[32mDan Jenson\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.047]; Australia[0.047]; Australia[0.047]; Jansher[0.045]; Mark[0.045]; Squash[0.044]; final[0.044]; Open[0.044]; Open[0.044]; Peter[0.044]; Ryan[0.043]; Anthony[0.043]; Quarter[0.043]; Quarter[0.043]; Chris[0.043]; Nicol[0.043]; Cairns[0.043]; Friday[0.042]; England[0.042]; England[0.042]; Derek[0.042]; seeding[0.041]; number[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mPakistan\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mPakistan national cricket team\u001b[39m <---> \u001b[32mPakistan\u001b[39m\t\nSCORES: global= 0.251:0.245[\u001b[31m0.006\u001b[39m]; local()= 0.113:0.057[\u001b[31m0.056\u001b[39m]; log p(e|m)= -2.749:-0.219[\u001b[32m2.529\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.045]; England[0.045]; Australia[0.045]; Australia[0.045]; Australia[0.045]; Jansher[0.045]; Khan[0.045]; Final[0.044]; Squash[0.044]; final[0.041]; Ireland[0.041]; Open[0.040]; Open[0.040]; Cairns[0.040]; Friday[0.040]; number[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; Results[0.040]; Scotland[0.039]; Eyles[0.039]; Nicol[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEngland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEngland cricket team\u001b[39m <---> \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.243:0.240[\u001b[31m0.003\u001b[39m]; local()= 0.136:0.078[\u001b[31m0.058\u001b[39m]; log p(e|m)= -3.297:-0.190[\u001b[32m3.107\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.053]; Australia[0.050]; Australia[0.050]; Australia[0.050]; Ireland[0.049]; Scotland[0.048]; Mark[0.047]; Pakistan[0.046]; final[0.045]; Ryan[0.044]; Chris[0.044]; Peter[0.044]; Jansher[0.044]; Cairns[0.043]; number[0.043]; Anthony[0.043]; Squash[0.043]; Derek[0.043]; Eyles[0.043]; Final[0.042]; Khan[0.042]; Hill[0.042]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEngland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEngland cricket team\u001b[39m <---> \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.243:0.240[\u001b[31m0.003\u001b[39m]; local()= 0.136:0.078[\u001b[31m0.058\u001b[39m]; log p(e|m)= -3.297:-0.190[\u001b[32m3.107\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.053]; Australia[0.050]; Australia[0.050]; Australia[0.050]; Ireland[0.049]; Scotland[0.048]; Mark[0.047]; Pakistan[0.046]; final[0.045]; Ryan[0.044]; Chris[0.044]; Peter[0.044]; Jansher[0.044]; Cairns[0.043]; number[0.043]; Anthony[0.043]; Squash[0.043]; Derek[0.043]; Eyles[0.043]; Final[0.042]; Khan[0.042]; Hill[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rodney Eyles\u001b[39m ==> ENTITY: \u001b[32mRodney Eyles\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJansher[0.048]; Australia[0.047]; Australia[0.047]; Australia[0.047]; Squash[0.046]; Open[0.045]; Open[0.045]; final[0.045]; Nicol[0.043]; Final[0.043]; Peter[0.042]; Pakistan[0.042]; England[0.042]; England[0.042]; Quarter[0.042]; Quarter[0.042]; Mark[0.042]; Cairns[0.042]; Anthony[0.042]; Kong[0.041]; Kong[0.041]; Kong[0.041]; Scotland[0.041]; \t\n\r [======================================>.......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 3s341ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4110/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1011testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bernardin\u001b[39m ==> ENTITY: \u001b[32mJoseph Bernardin\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCatholic[0.048]; archdiocese[0.045]; Cardinal[0.044]; Cardinal[0.044]; Joseph[0.044]; months[0.042]; surgery[0.042]; returned[0.042]; June[0.042]; Chicago[0.042]; Chicago[0.042]; Bernardin[0.041]; Bernardin[0.041]; doctors[0.041]; Wednesday[0.041]; Friday[0.040]; underwent[0.040]; pancreatic[0.040]; said[0.040]; said[0.040]; chemotherapy[0.040]; examinations[0.040]; time[0.040]; told[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bernardin\u001b[39m ==> ENTITY: \u001b[32mJoseph Bernardin\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCatholic[0.047]; archbishop[0.045]; archdiocese[0.044]; church[0.044]; Cardinal[0.043]; Cardinal[0.043]; Joseph[0.043]; months[0.041]; surgery[0.041]; returned[0.041]; June[0.041]; prayerful[0.041]; Chicago[0.041]; Chicago[0.041]; Chicago[0.041]; Bernardin[0.041]; Bernardin[0.041]; Bernardin[0.041]; conference[0.041]; doctors[0.040]; life[0.040]; Wednesday[0.040]; Friday[0.039]; underwent[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.076:0.076[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.050]; month[0.049]; seven[0.047]; series[0.047]; played[0.047]; highest[0.046]; second[0.046]; conferences[0.046]; Previously[0.046]; involving[0.046]; common[0.045]; million[0.045]; voice[0.045]; describing[0.045]; ranks[0.045]; ranking[0.045]; instrumental[0.045]; committee[0.044]; announced[0.043]; deal[0.043]; prominent[0.043]; developed[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bernardin\u001b[39m ==> ENTITY: \u001b[32mJoseph Bernardin\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \narchbishop[0.044]; seminary[0.044]; Catholics[0.044]; bishops[0.044]; bishops[0.044]; archdiocese[0.044]; priests[0.043]; church[0.043]; church[0.043]; church[0.043]; parishioners[0.042]; month[0.042]; month[0.042]; surgery[0.040]; praised[0.040]; policy[0.040]; Chicago[0.040]; Bernardin[0.040]; Bernardin[0.040]; Bernardin[0.040]; Bernardin[0.040]; recanted[0.040]; abuse[0.040]; announced[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Los Angeles ==> ENTITY: \u001b[32mLos Angeles\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.048:0.048[0]; log p(e|m)= -0.101:-0.101[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \narchdiocese[0.045]; priests[0.044]; church[0.044]; church[0.044]; prominent[0.043]; nation[0.043]; parishioners[0.042]; million[0.042]; archbishop[0.042]; scheduled[0.042]; seminary[0.042]; second[0.041]; accused[0.041]; month[0.041]; charges[0.041]; charges[0.041]; highest[0.041]; Italian[0.041]; surrounding[0.041]; took[0.040]; Chicago[0.040]; played[0.040]; official[0.040]; praised[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bernardin\u001b[39m ==> ENTITY: \u001b[32mJoseph Bernardin\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCatholic[0.052]; clergy[0.050]; archbishop[0.050]; archdiocese[0.049]; church[0.048]; months[0.045]; surgery[0.045]; returned[0.045]; undergo[0.045]; June[0.045]; prayerful[0.045]; Chicago[0.045]; Bernardin[0.044]; Bernardin[0.044]; conference[0.044]; doctors[0.044]; doctors[0.044]; life[0.044]; Wednesday[0.044]; Friday[0.043]; underwent[0.043]; pancreatic[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.083:0.083[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.045]; months[0.044]; news[0.044]; Joseph[0.043]; time[0.043]; June[0.043]; having[0.043]; returned[0.043]; followed[0.043]; Wednesday[0.042]; Chicago[0.042]; Chicago[0.042]; Friday[0.041]; indicated[0.041]; conference[0.041]; head[0.041]; extensive[0.041]; largest[0.041]; live[0.040]; Catholic[0.040]; said[0.039]; said[0.039]; says[0.039]; underwent[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = AIDS\u001b[39m ==> ENTITY: \u001b[32mHIV/AIDS\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.075:0.075[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnation[0.045]; soft[0.044]; moderate[0.043]; American[0.042]; ground[0.042]; sexual[0.042]; denied[0.042]; month[0.042]; espousing[0.042]; million[0.042]; spare[0.042]; praised[0.042]; life[0.041]; policy[0.041]; policy[0.041]; policy[0.041]; arms[0.041]; Previously[0.041]; prominent[0.041]; developed[0.041]; Angeles[0.041]; progressive[0.041]; anti[0.040]; seven[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bernardin\u001b[39m ==> ENTITY: \u001b[32mJoseph Bernardin\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \narchbishop[0.045]; seminary[0.045]; archdiocese[0.044]; priests[0.044]; church[0.043]; church[0.043]; parishioners[0.043]; month[0.043]; intestine[0.042]; spine[0.042]; surgery[0.041]; praised[0.041]; undergo[0.041]; scheduled[0.041]; Chicago[0.041]; Bernardin[0.040]; Bernardin[0.040]; Bernardin[0.040]; Bernardin[0.040]; recanted[0.040]; abuse[0.040]; recovery[0.040]; abused[0.040]; duct[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.093:0.093[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nclergy[0.045]; Catholic[0.044]; church[0.044]; Wednesday[0.044]; archdiocese[0.043]; returned[0.042]; Friday[0.042]; said[0.042]; said[0.042]; said[0.042]; said[0.042]; faith[0.041]; told[0.041]; told[0.041]; told[0.041]; June[0.041]; news[0.041]; months[0.041]; time[0.040]; time[0.040]; committed[0.040]; say[0.040]; left[0.040]; archbishop[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.081:0.081[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsecond[0.045]; months[0.044]; Joseph[0.044]; time[0.044]; June[0.043]; having[0.043]; returned[0.043]; followed[0.043]; Wednesday[0.043]; Chicago[0.042]; Chicago[0.042]; Friday[0.042]; indicated[0.042]; head[0.041]; extensive[0.041]; largest[0.041]; live[0.040]; Catholic[0.040]; said[0.040]; said[0.040]; says[0.040]; underwent[0.039]; Cardinal[0.039]; Cardinal[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.106:0.106[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChicago[0.047]; Catholic[0.045]; Wednesday[0.045]; archdiocese[0.044]; largest[0.043]; returned[0.043]; Friday[0.043]; told[0.042]; June[0.042]; months[0.042]; said[0.041]; said[0.041]; time[0.041]; Joseph[0.041]; says[0.041]; having[0.041]; second[0.041]; live[0.040]; underwent[0.040]; Cardinal[0.040]; Cardinal[0.040]; followed[0.039]; indicated[0.039]; need[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bernardin\u001b[39m ==> ENTITY: \u001b[32mJoseph Bernardin\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \narchbishop[0.045]; seminary[0.044]; Catholics[0.044]; bishops[0.044]; archdiocese[0.044]; priests[0.043]; church[0.043]; church[0.043]; church[0.043]; parishioners[0.043]; month[0.042]; month[0.042]; surgery[0.041]; praised[0.041]; policy[0.040]; undergo[0.040]; scheduled[0.040]; Chicago[0.040]; Bernardin[0.040]; Bernardin[0.040]; Bernardin[0.040]; Bernardin[0.040]; recanted[0.040]; abuse[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bernardin\u001b[39m ==> ENTITY: \u001b[32mJoseph Bernardin\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfaith[0.048]; cardinal[0.046]; month[0.046]; intestine[0.045]; spine[0.044]; Pancreatic[0.044]; surgery[0.044]; undergo[0.044]; undergo[0.044]; Surgeons[0.044]; scheduled[0.043]; Chicago[0.043]; doctors[0.043]; recovery[0.043]; early[0.043]; serve[0.043]; duct[0.043]; pain[0.042]; order[0.042]; stomach[0.042]; said[0.042]; said[0.042]; said[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Roman Catholic\u001b[39m ==> ENTITY: \u001b[32mCatholic Church\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.106:0.106[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJoseph[0.046]; archdiocese[0.046]; Cardinal[0.045]; Cardinal[0.045]; June[0.043]; aware[0.043]; Wednesday[0.043]; returned[0.042]; largest[0.042]; Friday[0.041]; indicated[0.041]; extensive[0.041]; said[0.041]; said[0.041]; months[0.041]; Chicago[0.041]; Chicago[0.041]; Chicago[0.041]; life[0.041]; conference[0.040]; need[0.040]; time[0.040]; minister[0.040]; news[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cincinnati ==> ENTITY: \u001b[32mCincinnati\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.090:0.090[0]; log p(e|m)= -0.460:-0.460[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \narchdiocese[0.050]; played[0.047]; archbishop[0.047]; American[0.047]; seven[0.047]; second[0.047]; series[0.046]; student[0.046]; student[0.046]; Previously[0.045]; conferences[0.045]; nation[0.045]; ranking[0.045]; Chicago[0.045]; took[0.045]; seminary[0.044]; priests[0.044]; ranks[0.044]; parishioners[0.044]; instrumental[0.044]; month[0.044]; deal[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bernardin\u001b[39m ==> ENTITY: \u001b[32mJoseph Bernardin\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \narchbishop[0.045]; seminary[0.044]; Catholics[0.044]; bishops[0.044]; bishops[0.044]; archdiocese[0.044]; priests[0.043]; church[0.043]; church[0.043]; church[0.043]; parishioners[0.042]; month[0.042]; praised[0.041]; policy[0.040]; policy[0.040]; policy[0.040]; Chicago[0.040]; Bernardin[0.040]; Bernardin[0.040]; Bernardin[0.040]; Bernardin[0.040]; recanted[0.040]; abuse[0.040]; announced[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.106:0.106[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChicago[0.047]; Catholic[0.045]; Wednesday[0.045]; archdiocese[0.044]; largest[0.043]; returned[0.043]; Friday[0.043]; told[0.042]; June[0.042]; months[0.042]; said[0.041]; said[0.041]; time[0.041]; Joseph[0.041]; says[0.041]; having[0.041]; second[0.041]; live[0.040]; underwent[0.040]; Cardinal[0.040]; Cardinal[0.040]; followed[0.039]; indicated[0.039]; need[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italian ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.088:0.088[0]; log p(e|m)= -0.759:-0.759[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nchurch[0.044]; spoken[0.044]; small[0.044]; early[0.043]; month[0.043]; Chicago[0.043]; praised[0.042]; stonecutter[0.042]; second[0.042]; Cincinnati[0.042]; archbishop[0.041]; priests[0.041]; developed[0.041]; developed[0.041]; seminary[0.041]; nation[0.041]; parts[0.041]; surrounding[0.041]; denied[0.041]; recanted[0.041]; soft[0.041]; parishioners[0.040]; involving[0.040]; prayed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Joseph Bernardin\u001b[39m ==> ENTITY: \u001b[32mJoseph Bernardin\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCatholic[0.048]; archdiocese[0.045]; Cardinal[0.044]; Cardinal[0.044]; months[0.042]; surgery[0.042]; returned[0.042]; June[0.042]; Chicago[0.042]; Chicago[0.042]; Bernardin[0.041]; Bernardin[0.041]; Bernardin[0.041]; doctors[0.041]; life[0.041]; Wednesday[0.041]; Friday[0.040]; underwent[0.040]; pancreatic[0.040]; said[0.040]; said[0.040]; chemotherapy[0.040]; examinations[0.040]; time[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bernardin\u001b[39m ==> ENTITY: \u001b[32mJoseph Bernardin\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.200:0.200[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \narchbishop[0.045]; seminary[0.044]; Catholics[0.044]; bishops[0.044]; bishops[0.044]; archdiocese[0.044]; priests[0.043]; church[0.043]; church[0.043]; church[0.043]; parishioners[0.042]; month[0.042]; praised[0.041]; policy[0.040]; policy[0.040]; policy[0.040]; Chicago[0.040]; Bernardin[0.040]; Bernardin[0.040]; Bernardin[0.040]; Bernardin[0.040]; recanted[0.040]; abuse[0.040]; announced[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chicago ==> ENTITY: \u001b[32mChicago\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.065:0.065[0]; log p(e|m)= -0.207:-0.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nchurch[0.045]; archdiocese[0.044]; nation[0.043]; largest[0.043]; scheduled[0.043]; parishioners[0.043]; prominent[0.043]; played[0.042]; month[0.042]; seminary[0.042]; early[0.042]; million[0.042]; praised[0.041]; Italian[0.041]; Cincinnati[0.041]; involving[0.041]; archbishop[0.041]; spare[0.040]; priests[0.040]; accused[0.040]; second[0.040]; denied[0.040]; resumed[0.040]; shortened[0.040]; \t\n\r [======================================>.......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 3s240ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4132/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1015testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Irvine ==> ENTITY: \u001b[32mIrvine, California\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.060:0.060[0]; log p(e|m)= -0.646:-0.646[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.044]; President[0.044]; campaign[0.043]; campaign[0.043]; campaign[0.043]; campaign[0.043]; debate[0.043]; Calif[0.042]; surrounding[0.042]; week[0.042]; letter[0.042]; Scott[0.041]; meet[0.041]; meet[0.041]; meet[0.041]; party[0.041]; seeking[0.041]; presidential[0.040]; inviting[0.040]; manager[0.040]; manager[0.040]; group[0.040]; Republicans[0.040]; small[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dole ==> ENTITY: \u001b[32mBob Dole\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.185:0.185[0]; log p(e|m)= -1.370:-1.370[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDole[0.044]; Dole[0.044]; Dole[0.044]; Dole[0.044]; President[0.042]; Carolina[0.042]; South[0.042]; campaign[0.042]; campaign[0.042]; campaign[0.042]; campaign[0.042]; campaign[0.042]; Scott[0.042]; Scott[0.042]; Irvine[0.041]; Clinton[0.041]; Clinton[0.041]; Clinton[0.041]; week[0.041]; vice[0.040]; presidential[0.040]; presidential[0.040]; Reed[0.040]; Reed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Texas ==> ENTITY: \u001b[32mTexas\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.061:0.061[0]; log p(e|m)= -0.161:-0.161[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.045]; presidential[0.043]; presidential[0.043]; vice[0.043]; vice[0.043]; Presidential[0.043]; Perot[0.042]; campaign[0.042]; campaign[0.042]; Scott[0.042]; partisan[0.042]; Party[0.042]; South[0.041]; Ross[0.041]; Clinton[0.041]; Carolina[0.041]; candidate[0.041]; League[0.041]; Peter[0.041]; short[0.040]; took[0.040]; mid[0.040]; Jack[0.040]; Reed[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dole ==> ENTITY: \u001b[32mBob Dole\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.171:0.171[0]; log p(e|m)= -1.370:-1.370[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDole[0.044]; Dole[0.044]; Dole[0.044]; Carolina[0.043]; South[0.042]; campaign[0.042]; campaign[0.042]; Scott[0.042]; candidate[0.042]; Clinton[0.042]; week[0.042]; head[0.041]; Presidential[0.041]; Ross[0.041]; September[0.041]; vice[0.041]; vice[0.041]; presidential[0.041]; presidential[0.041]; Jack[0.040]; Reed[0.040]; team[0.040]; campaigns[0.040]; Sept[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dole ==> ENTITY: \u001b[32mBob Dole\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.180:0.180[0]; log p(e|m)= -1.370:-1.370[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDole[0.044]; Dole[0.044]; Dole[0.044]; Dole[0.044]; Dole[0.044]; President[0.042]; Carolina[0.042]; South[0.042]; campaign[0.041]; campaign[0.041]; campaign[0.041]; campaign[0.041]; campaign[0.041]; Scott[0.041]; Scott[0.041]; Irvine[0.041]; candidate[0.041]; Clinton[0.041]; Clinton[0.041]; Clinton[0.041]; week[0.041]; head[0.041]; Ross[0.040]; September[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dole ==> ENTITY: \u001b[32mBob Dole\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.185:0.185[0]; log p(e|m)= -1.370:-1.370[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDole[0.044]; Dole[0.044]; Dole[0.044]; President[0.043]; campaign[0.042]; campaign[0.042]; campaign[0.042]; campaign[0.042]; campaign[0.042]; Scott[0.042]; Scott[0.042]; Irvine[0.042]; Clinton[0.042]; Clinton[0.042]; Clinton[0.042]; week[0.042]; vice[0.041]; presidential[0.041]; presidential[0.041]; Reed[0.040]; Reed[0.040]; team[0.040]; campaigns[0.040]; Sept[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Reform Party ==> ENTITY: \u001b[32mReform Party of the United States of America\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.835:-0.835[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPerot[0.045]; Scott[0.044]; candidate[0.044]; Reed[0.043]; campaign[0.042]; campaign[0.042]; Ross[0.042]; partisan[0.042]; presidential[0.042]; presidential[0.042]; Texas[0.041]; Oct[0.041]; Oct[0.041]; Oct[0.041]; Presidential[0.041]; Clinton[0.041]; Campbell[0.041]; Commission[0.041]; list[0.041]; vice[0.041]; vice[0.041]; Peter[0.040]; Kemp[0.040]; billionaire[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jack Kemp\u001b[39m ==> ENTITY: \u001b[32mJack Kemp\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.043]; campaign[0.043]; campaign[0.043]; Ross[0.043]; candidate[0.043]; Reed[0.043]; vice[0.042]; vice[0.042]; Scott[0.042]; Clinton[0.042]; Texas[0.042]; South[0.042]; Presidential[0.042]; Carolina[0.042]; presidential[0.041]; presidential[0.041]; week[0.041]; League[0.041]; campaigns[0.041]; Gov[0.040]; Perot[0.040]; Campbell[0.040]; September[0.040]; Carroll[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clinton ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.165:0.165[0]; log p(e|m)= -1.917:-1.917[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinton[0.045]; Clinton[0.045]; President[0.044]; campaign[0.043]; campaign[0.043]; campaign[0.043]; campaign[0.043]; campaign[0.043]; South[0.043]; Presidential[0.043]; presidential[0.042]; presidential[0.042]; presidential[0.042]; Carolina[0.042]; list[0.041]; televised[0.041]; aides[0.040]; week[0.040]; surrounding[0.039]; candidate[0.039]; team[0.039]; Scott[0.039]; Scott[0.039]; Texas[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dole ==> ENTITY: \u001b[32mBob Dole\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.173:0.173[0]; log p(e|m)= -1.370:-1.370[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDole[0.046]; Dole[0.046]; President[0.044]; campaign[0.043]; campaign[0.043]; campaign[0.043]; Scott[0.043]; Irvine[0.043]; Clinton[0.043]; Clinton[0.043]; week[0.043]; presidential[0.042]; Reed[0.041]; team[0.041]; campaigns[0.041]; debates[0.040]; debates[0.040]; televised[0.040]; Republicans[0.040]; party[0.040]; Calif[0.039]; debate[0.039]; Friday[0.039]; group[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Republicans ==> ENTITY: \u001b[32mRepublican Party (United States)\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.141:0.141[0]; log p(e|m)= -0.114:-0.114[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncampaign[0.046]; campaign[0.046]; campaign[0.046]; campaign[0.046]; President[0.042]; Clinton[0.042]; Clinton[0.042]; Clinton[0.042]; Scott[0.042]; Scott[0.042]; campaigns[0.042]; Reed[0.042]; Reed[0.042]; party[0.042]; presidential[0.041]; representatives[0.040]; wrote[0.040]; Dole[0.040]; Dole[0.040]; Dole[0.040]; Dole[0.040]; group[0.039]; televised[0.039]; letter[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dole ==> ENTITY: \u001b[32mBob Dole\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.170:0.170[0]; log p(e|m)= -1.370:-1.370[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDole[0.045]; Dole[0.045]; Dole[0.045]; Carolina[0.043]; South[0.043]; campaign[0.042]; campaign[0.042]; Scott[0.042]; candidate[0.042]; Clinton[0.042]; head[0.042]; Presidential[0.042]; Ross[0.041]; September[0.041]; vice[0.041]; vice[0.041]; presidential[0.041]; presidential[0.041]; Jack[0.041]; Reed[0.041]; team[0.040]; campaigns[0.040]; Sept[0.039]; Perot[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSouth Carolina\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSouth Carolina\u001b[39m <---> \u001b[32mGovernor of South Carolina\u001b[39m\t\nSCORES: global= 0.245:0.233[\u001b[31m0.012\u001b[39m]; local()= 0.071:0.137[\u001b[32m0.066\u001b[39m]; log p(e|m)= -0.270:-4.135[\u001b[31m3.865\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.047]; presidential[0.043]; presidential[0.043]; Scott[0.042]; Scott[0.042]; campaign[0.042]; campaign[0.042]; campaign[0.042]; Clinton[0.042]; Clinton[0.042]; candidate[0.042]; campaigns[0.042]; televised[0.042]; Presidential[0.041]; week[0.041]; Texas[0.041]; League[0.041]; Gov[0.041]; Perot[0.040]; vice[0.040]; vice[0.040]; Reed[0.040]; Reed[0.040]; address[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clinton ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.152:0.152[0]; log p(e|m)= -1.917:-1.917[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinton[0.046]; Clinton[0.046]; President[0.045]; campaign[0.044]; campaign[0.044]; campaign[0.044]; campaign[0.044]; campaign[0.044]; presidential[0.043]; presidential[0.043]; South[0.042]; televised[0.041]; aides[0.041]; week[0.040]; team[0.040]; Scott[0.040]; Scott[0.040]; surrounding[0.040]; dates[0.039]; ready[0.039]; debate[0.039]; debate[0.039]; counterparts[0.039]; party[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = League of Women Voters\u001b[39m ==> ENTITY: \u001b[32mLeague of Women Voters\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.092:0.092[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ndebates[0.046]; debates[0.046]; presidential[0.045]; presidential[0.045]; Presidential[0.044]; debate[0.043]; candidate[0.043]; Perot[0.042]; partisan[0.041]; decision[0.041]; vice[0.041]; vice[0.041]; South[0.041]; Texas[0.041]; forums[0.041]; mid[0.040]; Debates[0.040]; Carolina[0.040]; Dole[0.040]; Dole[0.040]; Dole[0.040]; September[0.040]; Party[0.039]; Jack[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ross Perot\u001b[39m ==> ENTITY: \u001b[32mRoss Perot\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \npresidential[0.045]; presidential[0.045]; campaign[0.044]; campaign[0.044]; Clinton[0.044]; billionaire[0.044]; Scott[0.043]; candidate[0.043]; Presidential[0.042]; vice[0.041]; vice[0.041]; Reed[0.041]; Kemp[0.041]; team[0.041]; debates[0.041]; debates[0.041]; debates[0.041]; Carolina[0.040]; Party[0.040]; Texas[0.040]; debate[0.040]; South[0.040]; Jack[0.040]; Campbell[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Calif\u001b[39m ==> ENTITY: \u001b[32mCalifornia\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.051:0.051[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncampaign[0.045]; campaign[0.045]; campaign[0.045]; campaign[0.045]; week[0.044]; Irvine[0.042]; seeking[0.042]; group[0.042]; Friday[0.042]; team[0.042]; manager[0.042]; manager[0.042]; President[0.042]; logistical[0.041]; small[0.041]; representatives[0.041]; released[0.040]; format[0.040]; campaigns[0.040]; pin[0.040]; hammer[0.040]; address[0.040]; participants[0.040]; televised[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Carroll Campbell\u001b[39m ==> ENTITY: \u001b[32mCarroll A. Campbell, Jr.\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncampaign[0.044]; campaign[0.044]; campaign[0.044]; Carolina[0.044]; South[0.043]; presidential[0.042]; presidential[0.042]; vice[0.042]; vice[0.042]; candidate[0.042]; surrounding[0.042]; week[0.041]; Gov[0.041]; Clinton[0.041]; September[0.041]; Scott[0.041]; Scott[0.041]; League[0.041]; Presidential[0.041]; team[0.040]; campaigns[0.040]; Commission[0.040]; Party[0.040]; logistical[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clinton ==> ENTITY: \u001b[32mBill Clinton\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.141:0.141[0]; log p(e|m)= -1.917:-1.917[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nClinton[0.046]; Clinton[0.046]; President[0.045]; campaign[0.044]; campaign[0.044]; campaign[0.044]; campaign[0.044]; campaign[0.044]; presidential[0.043]; televised[0.042]; aides[0.041]; week[0.040]; team[0.040]; Scott[0.040]; Scott[0.040]; surrounding[0.040]; dates[0.040]; ready[0.040]; debate[0.040]; counterparts[0.040]; party[0.039]; details[0.039]; details[0.039]; campaigns[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Commission on Presidential Debates\u001b[39m ==> ENTITY: \u001b[32mCommission on Presidential Debates\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncampaign[0.046]; presidential[0.045]; presidential[0.045]; debates[0.044]; debates[0.044]; Clinton[0.043]; vice[0.042]; vice[0.042]; candidate[0.042]; decision[0.041]; participate[0.041]; debate[0.041]; September[0.041]; sponsored[0.041]; Party[0.041]; partisan[0.040]; list[0.040]; non[0.040]; non[0.040]; allowed[0.040]; billionaire[0.040]; Perot[0.040]; Carolina[0.040]; head[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dole ==> ENTITY: \u001b[32mBob Dole\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.184:0.184[0]; log p(e|m)= -1.370:-1.370[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDole[0.046]; Dole[0.046]; Dole[0.046]; Dole[0.046]; Carolina[0.044]; South[0.044]; campaign[0.044]; campaign[0.044]; campaign[0.044]; Scott[0.044]; Scott[0.044]; candidate[0.043]; Clinton[0.043]; week[0.043]; head[0.043]; Presidential[0.043]; Ross[0.043]; September[0.042]; vice[0.042]; vice[0.042]; presidential[0.042]; presidential[0.042]; Jack[0.042]; \t\n\r [======================================>.......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 3s154ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4153/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1096testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew York\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew York Yankees\u001b[39m <---> \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.239:0.233[\u001b[31m0.006\u001b[39m]; local()= 0.214:0.111[\u001b[31m0.103\u001b[39m]; log p(e|m)= -5.116:-1.790[\u001b[32m3.326\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.045]; games[0.044]; Baseball[0.043]; Baseball[0.043]; game[0.043]; game[0.043]; Philadelphia[0.043]; Cleveland[0.042]; Boston[0.042]; League[0.042]; League[0.042]; League[0.042]; League[0.042]; City[0.041]; Baltimore[0.041]; Pittsburgh[0.041]; Cincinnati[0.041]; Houston[0.041]; Toronto[0.040]; Oakland[0.040]; played[0.040]; Detroit[0.040]; York[0.039]; York[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Francisco ==> ENTITY: \u001b[32mSan Francisco Giants\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.201:0.201[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.044]; games[0.044]; League[0.043]; League[0.043]; League[0.043]; Baseball[0.043]; Philadelphia[0.043]; game[0.043]; game[0.043]; Cleveland[0.042]; Houston[0.042]; California[0.042]; Oakland[0.042]; Cincinnati[0.041]; Diego[0.041]; Baltimore[0.041]; played[0.040]; San[0.040]; Pittsburgh[0.040]; Boston[0.040]; Texas[0.040]; Seattle[0.040]; City[0.040]; American[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlanta ==> ENTITY: \u001b[32mAtlanta Braves\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.236:0.236[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.044]; games[0.044]; Philadelphia[0.043]; Cleveland[0.043]; Baseball[0.042]; Baseball[0.042]; League[0.042]; League[0.042]; League[0.042]; League[0.042]; game[0.042]; game[0.042]; Atlanta[0.042]; Cincinnati[0.042]; Houston[0.042]; Pittsburgh[0.042]; Diego[0.041]; Boston[0.041]; Baltimore[0.041]; Oakland[0.040]; Detroit[0.040]; Los[0.040]; played[0.039]; Toronto[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York Yankees\u001b[39m\t\nSCORES: global= 0.238:0.238[0]; local()= 0.209:0.209[0]; log p(e|m)= -5.116:-5.116[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.047]; games[0.046]; Baseball[0.045]; game[0.045]; game[0.045]; Philadelphia[0.045]; Cleveland[0.044]; Boston[0.044]; League[0.044]; League[0.044]; League[0.044]; City[0.043]; Baltimore[0.043]; Pittsburgh[0.043]; Cincinnati[0.043]; Houston[0.042]; Toronto[0.042]; Oakland[0.042]; played[0.042]; Detroit[0.042]; York[0.041]; York[0.041]; American[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mMilwaukee\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAtlanta Braves\u001b[39m <---> \u001b[32mMilwaukee Brewers\u001b[39m\t\nSCORES: global= 0.253:0.251[\u001b[31m0.002\u001b[39m]; local()= 0.249:0.218[\u001b[31m0.032\u001b[39m]; log p(e|m)= -3.912:-3.411[\u001b[32m0.501\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPhiladelphia[0.045]; games[0.045]; Cleveland[0.045]; Baseball[0.045]; Baseball[0.045]; game[0.044]; game[0.044]; Cincinnati[0.044]; team[0.044]; Pittsburgh[0.044]; League[0.043]; League[0.043]; League[0.043]; League[0.043]; Boston[0.043]; Baltimore[0.043]; Houston[0.043]; Detroit[0.043]; Oakland[0.042]; Atlanta[0.042]; Atlanta[0.042]; played[0.042]; Toronto[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Philadelphia ==> ENTITY: \u001b[32mPhiladelphia Phillies\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.218:0.218[0]; log p(e|m)= -4.510:-4.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; Cleveland[0.046]; team[0.046]; Baseball[0.045]; Baseball[0.045]; game[0.044]; game[0.044]; Houston[0.044]; League[0.044]; League[0.044]; League[0.044]; League[0.044]; Cincinnati[0.044]; Baltimore[0.043]; Pittsburgh[0.043]; Boston[0.043]; Oakland[0.043]; played[0.042]; Detroit[0.042]; Atlanta[0.041]; Atlanta[0.041]; Florida[0.041]; American[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kansas City ==> ENTITY: \u001b[32mKansas City Royals\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.221:0.221[0]; log p(e|m)= -3.576:-3.576[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; team[0.044]; Baseball[0.043]; Baseball[0.043]; Houston[0.043]; League[0.043]; League[0.043]; League[0.043]; League[0.043]; Cleveland[0.042]; Toronto[0.042]; game[0.042]; game[0.042]; Baltimore[0.042]; Philadelphia[0.042]; Oakland[0.041]; Cincinnati[0.041]; Pittsburgh[0.040]; Detroit[0.040]; played[0.040]; Boston[0.040]; Montreal[0.040]; Louis[0.040]; Texas[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = National League ==> ENTITY: \u001b[32mNational League\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.217:0.217[0]; log p(e|m)= -0.075:-0.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.047]; League[0.047]; League[0.047]; games[0.047]; Baseball[0.045]; Baseball[0.045]; Cincinnati[0.044]; team[0.043]; Cleveland[0.043]; game[0.043]; game[0.043]; Philadelphia[0.043]; Detroit[0.043]; Boston[0.043]; Baltimore[0.042]; played[0.042]; Oakland[0.042]; Houston[0.042]; Pittsburgh[0.042]; Toronto[0.042]; York[0.041]; York[0.041]; York[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = San Diego ==> ENTITY: \u001b[32mSan Diego Padres\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.233:0.233[0]; log p(e|m)= -3.912:-3.912[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; League[0.045]; League[0.045]; League[0.045]; League[0.045]; team[0.045]; Houston[0.045]; Oakland[0.044]; California[0.044]; Cincinnati[0.044]; Cleveland[0.044]; Baseball[0.043]; Baseball[0.043]; Philadelphia[0.043]; game[0.043]; game[0.043]; Baltimore[0.043]; Pittsburgh[0.042]; Boston[0.042]; played[0.042]; Toronto[0.041]; Detroit[0.041]; Texas[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Boston ==> ENTITY: \u001b[32mBoston Red Sox\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.221:0.221[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; team[0.045]; game[0.043]; game[0.043]; Montreal[0.043]; Cleveland[0.043]; Caps[0.042]; Philadelphia[0.042]; Baseball[0.042]; Baseball[0.042]; League[0.042]; League[0.042]; League[0.042]; League[0.042]; Toronto[0.041]; Pittsburgh[0.041]; Cincinnati[0.041]; Detroit[0.041]; Baltimore[0.041]; played[0.040]; Houston[0.040]; Oakland[0.040]; Texas[0.039]; Louis[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Texas ==> ENTITY: \u001b[32mTexas Rangers (baseball)\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.223:0.223[0]; log p(e|m)= -4.962:-4.962[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBaseball[0.045]; Baseball[0.045]; team[0.045]; Cleveland[0.045]; game[0.045]; game[0.045]; Houston[0.044]; games[0.044]; League[0.044]; League[0.044]; League[0.044]; League[0.044]; Baltimore[0.044]; Cincinnati[0.043]; Oakland[0.043]; Philadelphia[0.043]; Detroit[0.042]; played[0.042]; Friday[0.042]; Friday[0.042]; Pittsburgh[0.042]; American[0.042]; Boston[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colorado ==> ENTITY: \u001b[32mColorado Rockies\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.193:0.193[0]; log p(e|m)= -4.510:-4.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; League[0.044]; League[0.044]; League[0.044]; team[0.043]; Montreal[0.043]; Toronto[0.042]; game[0.042]; game[0.042]; Caps[0.042]; Philadelphia[0.042]; Baseball[0.041]; Cincinnati[0.041]; played[0.041]; American[0.041]; Pittsburgh[0.041]; Houston[0.041]; Florida[0.041]; Boston[0.041]; Cleveland[0.040]; Texas[0.040]; Atlanta[0.040]; Atlanta[0.040]; Kansas[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Major League ==> ENTITY: \u001b[32mMajor League Baseball\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.220:0.220[0]; log p(e|m)= -0.478:-0.478[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.047]; League[0.047]; League[0.047]; team[0.046]; games[0.045]; Baseball[0.045]; Baseball[0.045]; game[0.045]; game[0.045]; Philadelphia[0.044]; Cincinnati[0.043]; Cleveland[0.043]; Houston[0.043]; Major[0.042]; Pittsburgh[0.042]; played[0.042]; Oakland[0.042]; Baltimore[0.042]; Toronto[0.041]; National[0.041]; Detroit[0.041]; Boston[0.041]; American[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Houston ==> ENTITY: \u001b[32mHouston Astros\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.198:0.198[0]; log p(e|m)= -4.605:-4.605[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.045]; games[0.044]; League[0.044]; League[0.044]; League[0.044]; League[0.044]; game[0.043]; game[0.043]; Texas[0.042]; Cleveland[0.042]; Baltimore[0.042]; Cincinnati[0.042]; Philadelphia[0.042]; Oakland[0.041]; Baseball[0.041]; Pittsburgh[0.041]; Boston[0.040]; City[0.040]; played[0.040]; home[0.040]; Atlanta[0.040]; Atlanta[0.040]; National[0.039]; Toronto[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York Yankees\u001b[39m\t\nSCORES: global= 0.239:0.239[0]; local()= 0.214:0.214[0]; log p(e|m)= -5.116:-5.116[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.045]; games[0.044]; Baseball[0.043]; Baseball[0.043]; game[0.043]; game[0.043]; Philadelphia[0.043]; Cleveland[0.042]; Boston[0.042]; League[0.042]; League[0.042]; League[0.042]; League[0.042]; City[0.041]; Baltimore[0.041]; Pittsburgh[0.041]; Cincinnati[0.041]; Houston[0.041]; Toronto[0.040]; Oakland[0.040]; played[0.040]; Detroit[0.040]; York[0.039]; York[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Toronto ==> ENTITY: \u001b[32mToronto Blue Jays\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.209:0.209[0]; log p(e|m)= -4.423:-4.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; team[0.044]; League[0.043]; League[0.043]; League[0.043]; League[0.043]; Montreal[0.043]; Baseball[0.042]; Baseball[0.042]; played[0.042]; game[0.042]; game[0.042]; Boston[0.042]; Cleveland[0.041]; Cincinnati[0.041]; Pittsburgh[0.041]; Caps[0.041]; Philadelphia[0.041]; Detroit[0.040]; Houston[0.040]; Baltimore[0.040]; Atlanta[0.040]; Atlanta[0.040]; Louis[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Montreal ==> ENTITY: \u001b[32mMontreal Expos\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.221:0.221[0]; log p(e|m)= -4.510:-4.510[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; team[0.044]; Toronto[0.043]; Baltimore[0.042]; Baseball[0.042]; Baseball[0.042]; League[0.042]; League[0.042]; League[0.042]; League[0.042]; Caps[0.042]; game[0.042]; game[0.042]; Houston[0.042]; Cincinnati[0.041]; Cleveland[0.041]; Philadelphia[0.041]; Boston[0.041]; Louis[0.041]; Detroit[0.041]; Pittsburgh[0.041]; Oakland[0.040]; Atlanta[0.040]; Atlanta[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Baltimore ==> ENTITY: \u001b[32mBaltimore Orioles\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.220:0.220[0]; log p(e|m)= -4.075:-4.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; Cleveland[0.045]; Houston[0.044]; Cincinnati[0.044]; Pittsburgh[0.044]; Philadelphia[0.044]; League[0.044]; League[0.044]; League[0.044]; League[0.044]; game[0.044]; game[0.044]; team[0.044]; Oakland[0.044]; Baseball[0.043]; Baseball[0.043]; Boston[0.043]; played[0.043]; National[0.043]; Detroit[0.042]; Toronto[0.042]; Florida[0.042]; Texas[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Florida ==> ENTITY: \u001b[32mMiami Marlins\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.218:0.218[0]; log p(e|m)= -4.828:-4.828[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; team[0.043]; Caps[0.043]; game[0.043]; game[0.043]; Cincinnati[0.042]; Philadelphia[0.042]; Montreal[0.042]; Baseball[0.042]; Baseball[0.042]; League[0.042]; League[0.042]; League[0.042]; League[0.042]; Pittsburgh[0.041]; Toronto[0.041]; Texas[0.041]; Cleveland[0.041]; Houston[0.041]; played[0.041]; California[0.040]; Milwaukee[0.040]; Oakland[0.040]; Baltimore[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oakland ==> ENTITY: \u001b[32mOakland Athletics\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.217:0.217[0]; log p(e|m)= -3.912:-3.912[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPittsburgh[0.044]; games[0.043]; team[0.043]; League[0.043]; League[0.043]; League[0.043]; League[0.043]; Cleveland[0.043]; Baseball[0.042]; Baseball[0.042]; Houston[0.042]; California[0.042]; game[0.042]; game[0.042]; Cincinnati[0.042]; Philadelphia[0.041]; Baltimore[0.041]; Boston[0.041]; Detroit[0.041]; Toronto[0.040]; Texas[0.040]; Florida[0.040]; City[0.040]; played[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Minnesota ==> ENTITY: \u001b[32mMinnesota Twins\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.226:0.226[0]; log p(e|m)= -4.269:-4.269[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.047]; game[0.046]; game[0.046]; team[0.046]; Baseball[0.044]; Baseball[0.044]; Cleveland[0.044]; League[0.043]; League[0.043]; League[0.043]; League[0.043]; Montreal[0.043]; Oakland[0.043]; played[0.043]; Detroit[0.042]; Caps[0.042]; Toronto[0.042]; Baltimore[0.042]; Philadelphia[0.042]; Cincinnati[0.042]; Houston[0.042]; Pittsburgh[0.042]; Boston[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pittsburgh ==> ENTITY: \u001b[32mPittsburgh Pirates\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.220:0.220[0]; log p(e|m)= -4.017:-4.017[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; Philadelphia[0.046]; game[0.045]; game[0.045]; Cleveland[0.044]; Cincinnati[0.044]; Baseball[0.044]; team[0.044]; League[0.044]; League[0.044]; League[0.044]; League[0.044]; Montreal[0.044]; Baltimore[0.043]; Boston[0.043]; Toronto[0.042]; Oakland[0.042]; played[0.042]; Florida[0.042]; Houston[0.042]; Detroit[0.042]; Caps[0.042]; Louis[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Major League ==> ENTITY: \u001b[32mMajor League Baseball\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.220:0.220[0]; log p(e|m)= -0.478:-0.478[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.047]; League[0.047]; League[0.047]; team[0.046]; games[0.045]; Baseball[0.045]; Baseball[0.045]; game[0.045]; game[0.045]; Philadelphia[0.044]; Cincinnati[0.043]; Cleveland[0.043]; Houston[0.043]; Major[0.042]; Pittsburgh[0.042]; played[0.042]; Oakland[0.042]; Baltimore[0.042]; Toronto[0.041]; National[0.041]; Detroit[0.041]; Boston[0.041]; American[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = American League\u001b[39m ==> ENTITY: \u001b[32mAmerican League\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.215:0.215[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.044]; League[0.044]; League[0.044]; team[0.043]; Boston[0.043]; Cincinnati[0.043]; Cleveland[0.043]; Baseball[0.043]; Baseball[0.043]; games[0.042]; game[0.042]; game[0.042]; Philadelphia[0.041]; Pittsburgh[0.041]; Detroit[0.041]; Baltimore[0.041]; Toronto[0.041]; Oakland[0.040]; Houston[0.040]; Chicago[0.040]; Chicago[0.040]; Chicago[0.040]; played[0.039]; Francisco[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlanta ==> ENTITY: \u001b[32mAtlanta Braves\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.236:0.236[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.044]; games[0.044]; Philadelphia[0.043]; Cleveland[0.043]; Baseball[0.042]; Baseball[0.042]; League[0.042]; League[0.042]; League[0.042]; League[0.042]; game[0.042]; game[0.042]; Atlanta[0.042]; Cincinnati[0.042]; Houston[0.042]; Pittsburgh[0.042]; Diego[0.041]; Boston[0.041]; Baltimore[0.041]; Oakland[0.040]; Detroit[0.040]; Los[0.040]; played[0.039]; Toronto[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Los Angeles ==> ENTITY: \u001b[32mLos Angeles Dodgers\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.201:0.201[0]; log p(e|m)= -4.711:-4.711[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.044]; team[0.043]; California[0.043]; League[0.043]; League[0.043]; League[0.043]; League[0.043]; Philadelphia[0.043]; Montreal[0.042]; Cleveland[0.042]; game[0.042]; game[0.042]; Baseball[0.041]; Baseball[0.041]; Toronto[0.041]; Cincinnati[0.041]; Houston[0.041]; Caps[0.041]; Boston[0.041]; Pittsburgh[0.041]; Baltimore[0.040]; Oakland[0.040]; Detroit[0.040]; City[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Seattle ==> ENTITY: \u001b[32mSeattle Mariners\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.210:0.210[0]; log p(e|m)= -3.863:-3.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.046]; games[0.044]; Baseball[0.043]; Baseball[0.043]; game[0.043]; game[0.043]; Cleveland[0.042]; League[0.042]; League[0.042]; League[0.042]; League[0.042]; Houston[0.042]; Cincinnati[0.041]; Oakland[0.041]; Baltimore[0.041]; Pittsburgh[0.041]; Philadelphia[0.041]; played[0.041]; Boston[0.041]; Toronto[0.041]; Detroit[0.040]; Atlanta[0.040]; Atlanta[0.040]; City[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cleveland ==> ENTITY: \u001b[32mCleveland Indians\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.227:0.227[0]; log p(e|m)= -4.200:-4.200[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; team[0.046]; Philadelphia[0.044]; Cincinnati[0.044]; Baseball[0.044]; Baseball[0.044]; game[0.044]; game[0.044]; League[0.044]; League[0.044]; League[0.044]; League[0.044]; Houston[0.044]; Baltimore[0.043]; Boston[0.043]; Oakland[0.043]; Pittsburgh[0.043]; Detroit[0.042]; American[0.042]; Toronto[0.042]; played[0.042]; Atlanta[0.042]; Atlanta[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cincinnati ==> ENTITY: \u001b[32mCincinnati Reds\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.211:0.211[0]; log p(e|m)= -3.772:-3.772[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.045]; team[0.045]; game[0.044]; game[0.044]; Cleveland[0.043]; Baseball[0.043]; Baseball[0.043]; League[0.042]; League[0.042]; League[0.042]; League[0.042]; Philadelphia[0.042]; Oakland[0.042]; Houston[0.042]; Pittsburgh[0.041]; Baltimore[0.041]; played[0.040]; Louis[0.040]; Detroit[0.040]; Boston[0.040]; Toronto[0.040]; American[0.039]; Milwaukee[0.039]; National[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Detroit ==> ENTITY: \u001b[32mDetroit Tigers\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.230:0.230[0]; log p(e|m)= -4.135:-4.135[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngames[0.046]; team[0.044]; Baseball[0.043]; Baseball[0.043]; game[0.043]; game[0.043]; Caps[0.043]; Cleveland[0.042]; Cincinnati[0.042]; Montreal[0.042]; Philadelphia[0.042]; League[0.042]; League[0.042]; League[0.042]; League[0.042]; Toronto[0.041]; Pittsburgh[0.041]; Boston[0.041]; played[0.041]; Oakland[0.040]; Houston[0.040]; Baltimore[0.039]; Atlanta[0.039]; Atlanta[0.039]; \t\n\r [=======================================>......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 2s958ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4187/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1146testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Manama\u001b[39m ==> ENTITY: \u001b[32mManama\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nManama[0.055]; Bahraini[0.048]; Bahraini[0.048]; Daih[0.047]; village[0.044]; village[0.044]; village[0.044]; al[0.043]; Turkish[0.042]; miles[0.041]; residents[0.041]; capital[0.041]; west[0.040]; officials[0.039]; Saturday[0.039]; gutted[0.039]; immediately[0.039]; km[0.039]; said[0.038]; said[0.038]; early[0.038]; hours[0.038]; immediate[0.038]; Government[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bahraini\u001b[39m ==> ENTITY: \u001b[32mBahrain\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.111:0.111[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nal[0.048]; Bahraini[0.047]; Manama[0.045]; Manama[0.045]; restaurant[0.044]; restaurant[0.044]; restaurant[0.044]; Turkish[0.043]; Government[0.042]; Daih[0.042]; Saturday[0.041]; west[0.040]; residents[0.040]; village[0.040]; village[0.040]; village[0.040]; capital[0.040]; said[0.040]; said[0.040]; immediate[0.039]; casualties[0.039]; officials[0.039]; ala[0.039]; miles[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Manama\u001b[39m ==> ENTITY: \u001b[32mManama\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nManama[0.055]; Bahraini[0.048]; Bahraini[0.048]; Daih[0.047]; village[0.044]; village[0.044]; village[0.044]; al[0.043]; Turkish[0.042]; miles[0.041]; residents[0.041]; capital[0.041]; west[0.040]; officials[0.039]; Saturday[0.039]; gutted[0.039]; immediately[0.039]; km[0.039]; said[0.038]; said[0.038]; early[0.038]; hours[0.038]; immediate[0.038]; Government[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bahraini\u001b[39m ==> ENTITY: \u001b[32mBahrain\u001b[39m\t\nSCORES: global= 0.285:0.285[0]; local()= 0.111:0.111[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nal[0.048]; Bahraini[0.047]; Manama[0.045]; Manama[0.045]; restaurant[0.044]; restaurant[0.044]; restaurant[0.044]; Turkish[0.043]; Government[0.042]; Daih[0.042]; Saturday[0.041]; west[0.040]; residents[0.040]; village[0.040]; village[0.040]; village[0.040]; capital[0.040]; said[0.040]; said[0.040]; immediate[0.039]; casualties[0.039]; officials[0.039]; ala[0.039]; miles[0.039]; \t\n\r [=======================================>......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 2s949ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4191/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1082testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portugal ==> ENTITY: \u001b[32mPortugal national football team\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.152:0.152[0]; log p(e|m)= -2.453:-2.453[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPortugal[0.064]; Soccer[0.064]; soccer[0.063]; qualifier[0.059]; Qualifier[0.058]; drew[0.058]; Cup[0.057]; Cup[0.057]; European[0.056]; World[0.055]; World[0.055]; group[0.055]; Saturday[0.053]; Armenia[0.052]; Armenia[0.052]; Draw[0.050]; Attendance[0.047]; Yerevan[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Yerevan\u001b[39m ==> ENTITY: \u001b[32mYerevan\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.094:0.094[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArmenia[0.062]; Armenia[0.062]; Cup[0.059]; Cup[0.059]; soccer[0.059]; European[0.057]; Soccer[0.057]; Attendance[0.056]; Saturday[0.055]; qualifier[0.055]; drew[0.055]; group[0.054]; Qualifier[0.053]; World[0.053]; World[0.053]; Portugal[0.050]; Portugal[0.050]; Draw[0.050]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEuropean\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUEFA\u001b[39m <---> \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.243:0.243[\u001b[31m0.000\u001b[39m]; local()= 0.144:0.068[\u001b[31m0.076\u001b[39m]; log p(e|m)= -3.576:-0.256[\u001b[32m3.319\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.062]; Soccer[0.061]; Cup[0.059]; Cup[0.059]; qualifier[0.059]; Qualifier[0.058]; Portugal[0.057]; Portugal[0.057]; group[0.056]; drew[0.055]; Armenia[0.054]; Armenia[0.054]; Saturday[0.053]; World[0.053]; World[0.053]; Attendance[0.051]; Draw[0.051]; Yerevan[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Portugal ==> ENTITY: \u001b[32mPortugal national football team\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.152:0.152[0]; log p(e|m)= -2.453:-2.453[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nPortugal[0.064]; Soccer[0.064]; soccer[0.063]; qualifier[0.059]; Qualifier[0.058]; drew[0.058]; Cup[0.057]; Cup[0.057]; European[0.056]; World[0.055]; World[0.055]; group[0.055]; Saturday[0.053]; Armenia[0.052]; Armenia[0.052]; Draw[0.050]; Attendance[0.047]; Yerevan[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Armenia ==> ENTITY: \u001b[32mArmenia national football team\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.168:0.168[0]; log p(e|m)= -2.688:-2.688[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArmenia[0.063]; soccer[0.061]; Soccer[0.061]; qualifier[0.059]; Qualifier[0.058]; Yerevan[0.057]; Cup[0.057]; Cup[0.057]; European[0.055]; drew[0.055]; Portugal[0.054]; Portugal[0.054]; Saturday[0.053]; group[0.053]; World[0.051]; World[0.051]; Draw[0.050]; Attendance[0.049]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Armenia ==> ENTITY: \u001b[32mArmenia national football team\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.168:0.168[0]; log p(e|m)= -2.688:-2.688[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArmenia[0.063]; soccer[0.061]; Soccer[0.061]; qualifier[0.059]; Qualifier[0.058]; Yerevan[0.057]; Cup[0.057]; Cup[0.057]; European[0.055]; drew[0.055]; Portugal[0.054]; Portugal[0.054]; Saturday[0.053]; group[0.053]; World[0.051]; World[0.051]; Draw[0.050]; Attendance[0.049]; \t\n\r [=======================================>......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 2s889ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4199/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1034testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bonn\u001b[39m ==> ENTITY: \u001b[32mBonn\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.112:0.112[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurope[0.045]; international[0.045]; Vienna[0.045]; Friday[0.045]; German[0.045]; Security[0.044]; viewed[0.042]; Organisation[0.042]; pro[0.042]; based[0.041]; Moscow[0.041]; chairman[0.041]; limited[0.041]; Cooperation[0.041]; region[0.041]; said[0.040]; record[0.040]; record[0.040]; Swiss[0.040]; told[0.040]; powers[0.040]; head[0.040]; minister[0.040]; leaders[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.223:0.223[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; Chechen[0.043]; military[0.042]; region[0.042]; region[0.042]; Russian[0.041]; Russian[0.041]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; separatist[0.040]; separatist[0.040]; conflict[0.040]; conflict[0.040]; government[0.040]; Zelimkhan[0.040]; Yandarbiyev[0.039]; Moscow[0.039]; declared[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.200:0.200[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.046]; Russia[0.046]; Russia[0.046]; Russia[0.046]; Russia[0.046]; Moscow[0.043]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Europe[0.042]; region[0.041]; region[0.041]; Alexander[0.040]; Chechen[0.040]; international[0.039]; state[0.039]; president[0.039]; German[0.039]; Lebed[0.039]; military[0.039]; sovereignty[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.226:0.226[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.048]; Chechnya[0.048]; Chechnya[0.048]; Chechnya[0.048]; Chechnya[0.048]; Chechen[0.045]; military[0.044]; region[0.044]; region[0.044]; region[0.044]; Russian[0.043]; Russian[0.043]; Russia[0.042]; separatist[0.042]; separatist[0.042]; conflict[0.041]; conflict[0.041]; government[0.041]; Zelimkhan[0.041]; Yandarbiyev[0.041]; Moscow[0.041]; Moscow[0.041]; declared[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.201:0.201[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.047]; Chechnya[0.047]; Chechnya[0.047]; Chechnya[0.047]; Chechnya[0.047]; Chechen[0.043]; military[0.042]; region[0.042]; Russian[0.041]; Russian[0.041]; separatist[0.040]; separatist[0.040]; conflict[0.040]; government[0.040]; Zelimkhan[0.040]; Yandarbiyev[0.040]; Moscow[0.040]; Moscow[0.040]; declared[0.040]; peace[0.040]; peace[0.040]; peace[0.040]; international[0.039]; saying[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.221:0.221[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; Chechen[0.043]; military[0.042]; region[0.042]; region[0.042]; Russian[0.041]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; conflict[0.040]; conflict[0.040]; separatist[0.040]; separatist[0.040]; government[0.040]; Zelimkhan[0.040]; Yandarbiyev[0.039]; Moscow[0.039]; declared[0.039]; peace[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.050]; Chechnya[0.050]; region[0.045]; Russian[0.044]; Russia[0.043]; Russia[0.043]; Russia[0.043]; conflict[0.043]; international[0.042]; borders[0.042]; state[0.041]; said[0.040]; said[0.040]; sovereignty[0.040]; internal[0.040]; small[0.040]; added[0.040]; according[0.040]; restricted[0.039]; member[0.039]; autonomy[0.039]; autonomy[0.039]; accorded[0.039]; called[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chechen ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.231:0.231[0]; log p(e|m)= -0.863:-0.863[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; military[0.042]; region[0.042]; region[0.042]; Russian[0.040]; Russian[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; declared[0.040]; separatist[0.040]; separatist[0.040]; conflict[0.039]; conflict[0.039]; government[0.039]; Zelimkhan[0.039]; according[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swiss ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.113:0.113[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nround[0.046]; Swiss[0.045]; Europe[0.045]; German[0.044]; limited[0.044]; region[0.041]; region[0.041]; region[0.041]; military[0.041]; according[0.041]; conflict[0.041]; conflict[0.041]; viewed[0.041]; Vienna[0.041]; involved[0.041]; pro[0.041]; said[0.041]; said[0.041]; maximum[0.041]; diplomat[0.040]; foreign[0.040]; autonomy[0.040]; powers[0.040]; government[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.152:0.152[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.048]; Moscow[0.045]; Moscow[0.045]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Chechnya[0.043]; Europe[0.043]; region[0.043]; Alexander[0.041]; Chechen[0.041]; international[0.040]; president[0.040]; German[0.040]; Lebed[0.040]; military[0.040]; separatist[0.040]; separatist[0.040]; peace[0.039]; peace[0.039]; peace[0.039]; diplomat[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.186:0.186[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.046]; Russia[0.046]; Russia[0.046]; Russian[0.045]; Russian[0.045]; government[0.043]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; region[0.042]; region[0.042]; president[0.041]; international[0.041]; Alexander[0.040]; round[0.039]; Chechen[0.039]; state[0.039]; military[0.039]; said[0.039]; said[0.039]; said[0.039]; Lebed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.051]; Chechnya[0.051]; Chechnya[0.051]; Chechnya[0.051]; region[0.046]; Russian[0.045]; conflict[0.043]; Moscow[0.043]; Moscow[0.043]; peace[0.043]; peace[0.043]; international[0.043]; pro[0.042]; foreign[0.042]; said[0.041]; internal[0.041]; Security[0.041]; Organisation[0.041]; viewed[0.041]; insisted[0.041]; given[0.040]; involved[0.040]; defended[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.196:0.196[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.050]; Russia[0.050]; Russia[0.050]; Russian[0.048]; Moscow[0.047]; government[0.046]; Chechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; region[0.045]; region[0.045]; president[0.044]; international[0.044]; minister[0.043]; Alexander[0.043]; round[0.043]; Chechen[0.043]; state[0.042]; military[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Zelimkhan Yandarbiyev\u001b[39m ==> ENTITY: \u001b[32mZelimkhan Yandarbiyev\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechen[0.044]; Chechnya[0.044]; Chechnya[0.044]; Chechnya[0.044]; Chechnya[0.044]; Chechnya[0.044]; Chechnya[0.044]; Maskhadov[0.042]; leader[0.042]; military[0.042]; government[0.042]; Russian[0.041]; Russian[0.041]; Moscow[0.041]; member[0.041]; diplomat[0.041]; president[0.041]; said[0.040]; said[0.040]; said[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; Russia[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Swiss ==> ENTITY: \u001b[32mSwitzerland\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.127:0.127[0]; log p(e|m)= -0.139:-0.139[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nround[0.046]; Swiss[0.045]; Europe[0.045]; German[0.044]; limited[0.044]; region[0.042]; region[0.042]; foreign[0.042]; international[0.042]; government[0.041]; military[0.041]; given[0.041]; viewed[0.041]; Vienna[0.041]; face[0.041]; involved[0.041]; record[0.040]; pro[0.040]; said[0.040]; said[0.040]; diplomat[0.040]; powers[0.040]; based[0.040]; defended[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.111:0.111[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.050]; Russia[0.050]; Chechnya[0.044]; Chechnya[0.044]; Chechnya[0.044]; framework[0.042]; international[0.042]; state[0.042]; sovereignty[0.041]; means[0.041]; restricted[0.041]; according[0.041]; autonomy[0.040]; small[0.040]; added[0.040]; maximum[0.040]; principles[0.040]; said[0.040]; said[0.040]; rights[0.040]; conflict[0.040]; allowed[0.040]; policy[0.039]; member[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Aslan Maskhadov\u001b[39m ==> ENTITY: \u001b[32mAslan Maskhadov\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nZelimkhan[0.046]; Chechnya[0.045]; Chechnya[0.045]; Chechnya[0.045]; Chechnya[0.045]; Chechnya[0.045]; Chechnya[0.045]; Chechnya[0.045]; military[0.045]; leader[0.045]; president[0.045]; Yandarbiyev[0.044]; government[0.044]; Chechen[0.043]; given[0.042]; conflict[0.042]; conflict[0.042]; peace[0.042]; separatist[0.042]; separatist[0.042]; minister[0.042]; leaders[0.041]; international[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.047]; Russia[0.047]; Russia[0.047]; Russian[0.045]; Russian[0.045]; government[0.043]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; region[0.042]; region[0.042]; president[0.041]; international[0.041]; Alexander[0.040]; round[0.040]; Chechen[0.040]; state[0.039]; military[0.039]; said[0.039]; said[0.039]; said[0.039]; Lebed[0.039]; sovereignty[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alexander Lebed\u001b[39m ==> ENTITY: \u001b[32mAlexander Lebed\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.043]; Russian[0.043]; president[0.043]; military[0.043]; Russia[0.042]; Russia[0.042]; Russia[0.042]; Russia[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Maskhadov[0.042]; separatist[0.041]; separatist[0.041]; Moscow[0.041]; Zelimkhan[0.041]; Chechen[0.041]; saying[0.040]; Yandarbiyev[0.040]; leader[0.040]; rebel[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.233:0.233[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; Chechen[0.043]; military[0.042]; region[0.042]; region[0.042]; region[0.042]; Russian[0.041]; Russian[0.041]; Russia[0.040]; Russia[0.040]; separatist[0.040]; separatist[0.040]; conflict[0.039]; conflict[0.039]; government[0.039]; Zelimkhan[0.039]; Yandarbiyev[0.039]; Moscow[0.039]; Moscow[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.050]; Chechnya[0.050]; region[0.045]; Russian[0.044]; conflict[0.043]; Moscow[0.042]; peace[0.042]; peace[0.042]; international[0.042]; Friday[0.042]; pro[0.041]; foreign[0.041]; insisted[0.041]; said[0.040]; involved[0.040]; internal[0.040]; Security[0.040]; Organisation[0.040]; viewed[0.040]; leaders[0.040]; defended[0.040]; criticism[0.039]; told[0.039]; Europe[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Flavio Cotti\u001b[39m ==> ENTITY: \u001b[32mFlavio Cotti\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.132:0.132[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCotti[0.046]; pro[0.045]; president[0.045]; chairman[0.044]; Swiss[0.044]; Swiss[0.044]; minister[0.043]; Bonn[0.043]; Organisation[0.042]; leaders[0.042]; leaders[0.042]; given[0.041]; Pro[0.041]; Vienna[0.040]; head[0.040]; Friday[0.040]; defended[0.040]; Security[0.040]; saying[0.040]; Europe[0.040]; German[0.040]; international[0.040]; Yandarbiyev[0.040]; Zelimkhan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.188:0.188[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.046]; Russia[0.046]; Russia[0.046]; Russian[0.044]; Moscow[0.044]; government[0.043]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; Chechnya[0.042]; region[0.042]; region[0.042]; president[0.041]; international[0.041]; Alexander[0.040]; round[0.039]; Chechen[0.039]; state[0.039]; military[0.039]; said[0.039]; said[0.039]; Lebed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cotti\u001b[39m ==> ENTITY: \u001b[32mFlavio Cotti\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.080:0.080[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfact[0.044]; member[0.044]; allowed[0.044]; issue[0.044]; possible[0.043]; challenge[0.042]; international[0.042]; small[0.042]; members[0.042]; situation[0.042]; said[0.042]; said[0.042]; framework[0.042]; maximum[0.041]; internal[0.041]; worked[0.041]; added[0.041]; state[0.041]; accorded[0.041]; restricted[0.040]; means[0.040]; autonomy[0.040]; body[0.040]; details[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cotti\u001b[39m ==> ENTITY: \u001b[32mFlavio Cotti\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.120:0.120[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCotti[0.046]; president[0.045]; Swiss[0.044]; Swiss[0.044]; government[0.044]; minister[0.044]; leaders[0.042]; given[0.042]; fact[0.042]; member[0.042]; allowed[0.042]; military[0.042]; rebel[0.042]; Pro[0.041]; leader[0.041]; Friday[0.040]; challenge[0.040]; saying[0.040]; international[0.040]; Yandarbiyev[0.040]; Zelimkhan[0.040]; members[0.040]; foreign[0.040]; Tim[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.226:0.226[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; Chechnya[0.046]; Chechen[0.043]; military[0.042]; region[0.042]; region[0.042]; region[0.042]; Russian[0.041]; Russian[0.041]; Russia[0.040]; Russia[0.040]; Russia[0.040]; separatist[0.040]; separatist[0.040]; conflict[0.040]; conflict[0.040]; government[0.040]; Zelimkhan[0.040]; Yandarbiyev[0.039]; Moscow[0.039]; declared[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Chechnya\u001b[39m ==> ENTITY: \u001b[32mChechnya\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChechnya[0.048]; Chechnya[0.048]; Chechnya[0.048]; Chechen[0.044]; military[0.043]; region[0.043]; region[0.043]; Russian[0.042]; Russia[0.041]; Russia[0.041]; Russia[0.041]; Russia[0.041]; separatist[0.041]; conflict[0.041]; conflict[0.041]; peace[0.040]; international[0.040]; borders[0.040]; Maskhadov[0.040]; state[0.039]; said[0.039]; said[0.039]; said[0.039]; outside[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = German ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.096:0.096[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurope[0.046]; region[0.045]; military[0.043]; government[0.043]; foreign[0.042]; international[0.042]; Swiss[0.042]; Swiss[0.042]; Alexander[0.042]; viewed[0.042]; powers[0.041]; Russian[0.041]; Russian[0.041]; Vienna[0.041]; Bonn[0.041]; president[0.041]; conflict[0.041]; pro[0.041]; diplomat[0.041]; leaders[0.040]; leaders[0.040]; Friday[0.040]; based[0.040]; involved[0.040]; \t\n\r [=======================================>......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 2s745ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4227/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1127testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moslem ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.118:0.118[0]; log p(e|m)= -1.214:-1.214[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoslem[0.049]; Moslem[0.049]; Bosnia[0.047]; Bosnia[0.047]; Serbs[0.044]; Serb[0.043]; Sarajevo[0.043]; believed[0.042]; troops[0.042]; received[0.041]; said[0.041]; said[0.041]; said[0.041]; village[0.041]; village[0.041]; Tuzla[0.041]; republic[0.040]; movement[0.039]; establish[0.039]; Saturday[0.039]; men[0.038]; men[0.038]; set[0.038]; fact[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Moslems\u001b[39m ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.135:0.135[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoslem[0.048]; Moslem[0.048]; Moslems[0.044]; Moslems[0.044]; order[0.042]; troops[0.042]; troops[0.042]; troops[0.042]; believed[0.042]; territory[0.041]; said[0.041]; said[0.041]; Colonel[0.041]; week[0.041]; war[0.040]; village[0.040]; village[0.040]; village[0.040]; month[0.040]; Thursday[0.040]; houses[0.040]; brought[0.039]; men[0.039]; men[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sarajevo\u001b[39m ==> ENTITY: \u001b[32mSarajevo\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.002:-0.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBosnia[0.049]; Bosnia[0.049]; Bosnian[0.044]; Bosnian[0.044]; Serbs[0.043]; war[0.042]; Serb[0.042]; Serb[0.042]; Serb[0.042]; troops[0.041]; troops[0.041]; Nato[0.041]; Nato[0.041]; Nato[0.041]; Nato[0.041]; Nato[0.041]; territory[0.041]; Tuzla[0.040]; violent[0.039]; establish[0.039]; tense[0.039]; month[0.039]; Saturday[0.039]; Lieutenant[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSerb\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSerbs\u001b[39m <---> \u001b[32mSerbia\u001b[39m\t\nSCORES: global= 0.257:0.246[\u001b[31m0.011\u001b[39m]; local()= 0.169:0.148[\u001b[31m0.021\u001b[39m]; log p(e|m)= -0.027:-2.172[\u001b[31m2.144\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBosnia[0.047]; Bosnia[0.047]; Serb[0.046]; Serbs[0.046]; territory[0.046]; troops[0.045]; Bosnian[0.045]; republic[0.044]; republic[0.044]; Sarajevo[0.044]; Moslems[0.043]; Tuzla[0.042]; village[0.042]; village[0.042]; village[0.042]; village[0.042]; village[0.042]; Moslem[0.042]; Moslem[0.042]; Moslem[0.042]; Moslem[0.042]; violent[0.042]; Colonel[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.076:0.076[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nweek[0.044]; month[0.044]; territory[0.043]; order[0.043]; earlier[0.042]; war[0.042]; troops[0.042]; troops[0.042]; Thursday[0.042]; Fights[0.042]; Moslems[0.042]; Moslems[0.042]; Moslems[0.042]; establish[0.042]; entered[0.041]; Reuters[0.041]; Bosnian[0.041]; Bosnian[0.041]; began[0.040]; Moslem[0.040]; intentions[0.040]; helicopters[0.040]; Saturday[0.040]; brought[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moslem ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.124:0.124[0]; log p(e|m)= -1.214:-1.214[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoslem[0.047]; Moslem[0.047]; Moslem[0.047]; Bosnia[0.046]; Bosnia[0.046]; Bosnian[0.044]; Serbs[0.043]; Serb[0.042]; Serb[0.042]; Sarajevo[0.042]; believed[0.041]; troops[0.041]; received[0.040]; said[0.040]; said[0.040]; said[0.040]; village[0.040]; village[0.040]; village[0.040]; village[0.040]; Tuzla[0.040]; republic[0.039]; republic[0.039]; movement[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moslem ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.139:0.139[0]; log p(e|m)= -1.214:-1.214[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoslem[0.051]; Moslem[0.051]; Bosnia[0.049]; Bosnian[0.047]; Bosnian[0.047]; Moslems[0.046]; Moslems[0.046]; Moslems[0.046]; Serbs[0.046]; Serb[0.045]; Serb[0.045]; Serb[0.045]; Sarajevo[0.045]; order[0.044]; war[0.044]; believed[0.044]; month[0.044]; territory[0.044]; troops[0.043]; troops[0.043]; troops[0.043]; received[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.060:0.060[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nroad[0.044]; month[0.044]; week[0.043]; troops[0.043]; troops[0.043]; troops[0.043]; reporter[0.043]; said[0.043]; Thursday[0.042]; earlier[0.042]; helicopters[0.042]; war[0.042]; rebuild[0.041]; territory[0.041]; entered[0.041]; establish[0.041]; houses[0.041]; ground[0.041]; Moslem[0.040]; republic[0.040]; seen[0.040]; restored[0.040]; driven[0.040]; morning[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tuzla ==> ENTITY: \u001b[32mTuzla\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.167:0.167[0]; log p(e|m)= -0.030:-0.030[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBosnia[0.045]; Bosnia[0.045]; village[0.044]; village[0.044]; village[0.044]; village[0.044]; village[0.044]; Sarajevo[0.043]; Bosnian[0.041]; Bosnian[0.041]; troops[0.041]; troops[0.041]; Serb[0.041]; Serb[0.041]; Serb[0.041]; territory[0.041]; Nato[0.040]; Nato[0.040]; Nato[0.040]; Nato[0.040]; Nato[0.040]; Nato[0.040]; Serbs[0.040]; Thursday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moslem ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.157:0.157[0]; log p(e|m)= -1.214:-1.214[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoslem[0.050]; Moslem[0.050]; Moslem[0.050]; Bosnia[0.049]; Bosnia[0.049]; Bosnian[0.046]; Bosnian[0.046]; Moslems[0.046]; Moslems[0.046]; Serbs[0.046]; Serb[0.044]; Serb[0.044]; Sarajevo[0.044]; war[0.044]; believed[0.043]; month[0.043]; territory[0.043]; troops[0.043]; received[0.043]; said[0.042]; said[0.042]; said[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mahala\u001b[39m ==> ENTITY: \u001b[32mMahala\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBosnia[0.058]; Bosnia[0.058]; village[0.051]; village[0.051]; village[0.051]; village[0.051]; Mahala[0.049]; Mahala[0.049]; Mahala[0.049]; Mahala[0.049]; Sarajevo[0.049]; Bosnian[0.047]; Serbs[0.047]; Tuzla[0.046]; Serb[0.045]; Serb[0.045]; territory[0.043]; Nato[0.041]; Nato[0.041]; Nato[0.041]; Nato[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Moslems\u001b[39m ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoslem[0.049]; Moslems[0.045]; Moslems[0.045]; order[0.043]; troops[0.043]; troops[0.043]; troops[0.043]; territory[0.042]; said[0.041]; said[0.041]; Colonel[0.041]; week[0.041]; war[0.041]; village[0.041]; village[0.041]; village[0.041]; month[0.041]; Thursday[0.041]; houses[0.040]; brought[0.040]; men[0.040]; restored[0.040]; republic[0.040]; earlier[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mahala\u001b[39m ==> ENTITY: \u001b[32mMahala\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBosnia[0.056]; Bosnia[0.056]; village[0.049]; village[0.049]; Mahala[0.048]; Mahala[0.048]; Mahala[0.048]; Sarajevo[0.047]; Serbs[0.045]; Tuzla[0.044]; Serb[0.043]; Nato[0.039]; Nato[0.039]; Nato[0.039]; Nato[0.039]; Moslem[0.039]; Moslem[0.039]; Moslem[0.039]; violent[0.039]; republic[0.039]; believed[0.039]; establish[0.039]; tense[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NATO\u001b[39m ==> ENTITY: \u001b[32mNATO\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.096:0.096[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNato[0.047]; Nato[0.047]; troops[0.045]; troops[0.045]; helicopters[0.044]; war[0.044]; Bosnian[0.042]; Bosnian[0.042]; month[0.042]; Thursday[0.042]; Serb[0.042]; Serb[0.042]; territory[0.041]; earlier[0.040]; week[0.040]; police[0.040]; police[0.040]; overhead[0.039]; order[0.039]; flew[0.039]; republic[0.039]; entered[0.039]; Fights[0.039]; Reuters[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Moslems\u001b[39m ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoslem[0.048]; Moslem[0.048]; Moslems[0.044]; Moslems[0.044]; order[0.042]; troops[0.042]; troops[0.042]; troops[0.042]; believed[0.042]; territory[0.041]; said[0.041]; said[0.041]; Colonel[0.040]; week[0.040]; war[0.040]; Bosnia[0.040]; village[0.040]; village[0.040]; village[0.040]; village[0.040]; month[0.040]; Thursday[0.040]; houses[0.040]; brought[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NATO\u001b[39m ==> ENTITY: \u001b[32mNATO\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.111:0.111[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNato[0.046]; Nato[0.046]; Nato[0.046]; Bosnia[0.045]; Bosnia[0.045]; troops[0.044]; war[0.042]; monitoring[0.042]; Bosnian[0.041]; Bosnian[0.041]; Sarajevo[0.041]; month[0.041]; Serb[0.040]; Serb[0.040]; Serb[0.040]; territory[0.040]; identities[0.040]; said[0.040]; said[0.040]; said[0.040]; Serbs[0.040]; spokesman[0.039]; Lieutenant[0.039]; report[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bosnia ==> ENTITY: \u001b[32mBosnia and Herzegovina\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.187:0.187[0]; log p(e|m)= -0.345:-0.345[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBosnia[0.056]; Sarajevo[0.049]; territory[0.049]; Bosnian[0.049]; troops[0.048]; Serbs[0.048]; Nato[0.048]; Nato[0.048]; Nato[0.048]; Nato[0.048]; Serb[0.047]; Serb[0.047]; republic[0.047]; republic[0.047]; Moslem[0.046]; Moslem[0.046]; Moslem[0.046]; Moslem[0.046]; Moslems[0.046]; reports[0.045]; report[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mahala\u001b[39m ==> ENTITY: \u001b[32mMahala\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nvillage[0.048]; village[0.048]; village[0.048]; Mahala[0.047]; Mahala[0.047]; Bosnian[0.045]; Bosnian[0.045]; Tuzla[0.044]; Serb[0.042]; Serb[0.042]; territory[0.041]; houses[0.040]; Nato[0.039]; Nato[0.039]; Nato[0.039]; Moslem[0.039]; Moslems[0.039]; Moslems[0.039]; Moslems[0.039]; republic[0.038]; entered[0.038]; police[0.038]; police[0.038]; establish[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mahala\u001b[39m ==> ENTITY: \u001b[32mMahala\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.172:0.172[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBosnia[0.050]; Bosnia[0.050]; village[0.044]; village[0.044]; village[0.044]; village[0.044]; village[0.044]; Mahala[0.043]; Mahala[0.043]; Mahala[0.043]; Mahala[0.043]; Mahala[0.043]; Sarajevo[0.042]; Bosnian[0.041]; Bosnian[0.041]; Serbs[0.041]; Tuzla[0.040]; Serb[0.039]; Serb[0.039]; Serb[0.039]; territory[0.037]; houses[0.036]; republic[0.036]; republic[0.036]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSerb\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSerbs\u001b[39m <---> \u001b[32mSerbia\u001b[39m\t\nSCORES: global= 0.257:0.244[\u001b[31m0.012\u001b[39m]; local()= 0.152:0.122[\u001b[31m0.030\u001b[39m]; log p(e|m)= -0.027:-2.172[\u001b[31m2.144\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSerb[0.053]; troops[0.052]; troops[0.052]; troops[0.052]; Bosnian[0.051]; Bosnian[0.051]; territory[0.051]; Sarajevo[0.051]; village[0.051]; village[0.051]; village[0.051]; war[0.050]; republic[0.049]; Tuzla[0.049]; Moslem[0.048]; Moslems[0.048]; Moslems[0.048]; Moslems[0.048]; Colonel[0.047]; Thursday[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NATO\u001b[39m ==> ENTITY: \u001b[32mNATO\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.122:0.122[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNato[0.045]; Nato[0.045]; Nato[0.045]; Nato[0.045]; Bosnia[0.045]; Bosnia[0.045]; troops[0.044]; troops[0.044]; war[0.042]; monitoring[0.042]; Bosnian[0.041]; Bosnian[0.041]; Sarajevo[0.040]; month[0.040]; Thursday[0.040]; Serb[0.040]; Serb[0.040]; Serb[0.040]; territory[0.039]; identities[0.039]; said[0.039]; said[0.039]; said[0.039]; Serbs[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Serbs ==> ENTITY: \u001b[32mSerbs\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.168:0.168[0]; log p(e|m)= -0.028:-0.028[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBosnia[0.048]; Bosnia[0.048]; territory[0.045]; Bosnian[0.045]; Bosnian[0.045]; Serb[0.044]; Serb[0.044]; republic[0.044]; republic[0.044]; Sarajevo[0.044]; troops[0.043]; said[0.043]; said[0.043]; said[0.043]; war[0.043]; Moslems[0.042]; Moslems[0.042]; Tuzla[0.042]; Moslem[0.041]; Moslem[0.041]; Moslem[0.041]; Moslem[0.041]; report[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NATO\u001b[39m ==> ENTITY: \u001b[32mNATO\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.110:0.110[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNato[0.047]; Nato[0.047]; Nato[0.047]; Bosnia[0.046]; Bosnia[0.046]; troops[0.045]; monitoring[0.043]; Sarajevo[0.042]; Serb[0.041]; said[0.041]; said[0.041]; said[0.041]; Serbs[0.041]; spokesman[0.040]; Lieutenant[0.040]; report[0.040]; received[0.040]; monitors[0.040]; confrontation[0.039]; Tuzla[0.039]; days[0.039]; republic[0.039]; reports[0.039]; Colonel[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bosnia ==> ENTITY: \u001b[32mBosnia and Herzegovina\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.152:0.152[0]; log p(e|m)= -0.345:-0.345[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBosnia[0.052]; Sarajevo[0.045]; Serbs[0.044]; Serb[0.043]; republic[0.043]; troops[0.043]; Moslem[0.042]; Moslem[0.042]; Moslem[0.042]; reports[0.041]; report[0.041]; village[0.041]; village[0.041]; village[0.041]; Tuzla[0.040]; men[0.040]; men[0.040]; men[0.040]; establish[0.040]; Nato[0.040]; Nato[0.040]; Nato[0.040]; Nato[0.040]; monitoring[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bosnian ==> ENTITY: \u001b[32mBosnia and Herzegovina\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.187:0.187[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSarajevo[0.047]; territory[0.047]; Bosnian[0.047]; Serbs[0.045]; war[0.045]; Serb[0.045]; Serb[0.045]; republic[0.044]; republic[0.044]; troops[0.044]; troops[0.044]; troops[0.044]; Moslem[0.043]; Moslem[0.043]; village[0.042]; village[0.042]; village[0.042]; Tuzla[0.042]; Moslems[0.042]; Moslems[0.042]; Moslems[0.042]; said[0.041]; said[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NATO\u001b[39m ==> ENTITY: \u001b[32mNATO\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.096:0.096[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNato[0.047]; Nato[0.047]; troops[0.045]; troops[0.045]; helicopters[0.044]; war[0.043]; Bosnian[0.042]; Bosnian[0.042]; month[0.042]; Thursday[0.042]; Serb[0.041]; Serb[0.041]; territory[0.041]; identities[0.041]; earlier[0.040]; week[0.040]; police[0.040]; police[0.040]; Tuzla[0.040]; overhead[0.039]; order[0.039]; flew[0.039]; republic[0.039]; entered[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mahala\u001b[39m ==> ENTITY: \u001b[32mMahala\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.168:0.168[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBosnia[0.058]; village[0.051]; village[0.051]; village[0.051]; village[0.051]; Mahala[0.050]; Mahala[0.050]; Mahala[0.050]; Mahala[0.050]; Sarajevo[0.049]; Bosnian[0.048]; Bosnian[0.048]; Serbs[0.047]; Tuzla[0.046]; Serb[0.045]; Serb[0.045]; Serb[0.045]; territory[0.043]; houses[0.042]; republic[0.041]; republic[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mahala\u001b[39m ==> ENTITY: \u001b[32mMahala\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBosnia[0.057]; Bosnia[0.057]; village[0.050]; village[0.050]; village[0.050]; village[0.050]; village[0.050]; Mahala[0.048]; Mahala[0.048]; Mahala[0.048]; Mahala[0.048]; Sarajevo[0.048]; Bosnian[0.046]; Bosnian[0.046]; Serbs[0.046]; Tuzla[0.045]; Serb[0.044]; Serb[0.044]; Serb[0.044]; territory[0.042]; houses[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NATO\u001b[39m ==> ENTITY: \u001b[32mNATO\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.106:0.106[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNato[0.046]; Nato[0.046]; Nato[0.046]; troops[0.044]; troops[0.044]; troops[0.044]; helicopters[0.043]; war[0.043]; Bosnian[0.041]; Bosnian[0.041]; Sarajevo[0.041]; month[0.041]; Thursday[0.041]; Serb[0.041]; Serb[0.041]; territory[0.040]; identities[0.040]; said[0.040]; said[0.040]; Lieutenant[0.040]; earlier[0.039]; week[0.039]; police[0.039]; police[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = NATO\u001b[39m ==> ENTITY: \u001b[32mNATO\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.105:0.105[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNato[0.047]; Nato[0.047]; Nato[0.047]; Bosnia[0.046]; Bosnia[0.046]; troops[0.045]; monitoring[0.043]; Sarajevo[0.042]; Serb[0.041]; identities[0.041]; said[0.041]; said[0.041]; said[0.041]; Serbs[0.040]; spokesman[0.040]; Lieutenant[0.040]; report[0.040]; received[0.040]; monitors[0.040]; confrontation[0.039]; Tuzla[0.039]; days[0.039]; republic[0.039]; reports[0.039]; \t\n\r [=======================================>......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 2s584ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4257/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1024testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNew York\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNew York\u001b[39m <---> \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.234:0.232[\u001b[31m0.002\u001b[39m]; local()= -0.037:-0.037[\u001b[32m0.000\u001b[39m]; log p(e|m)= -0.289:-1.790[\u001b[31m1.501\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYork[0.050]; loss[0.048]; loss[0.048]; loss[0.048]; loss[0.048]; loss[0.048]; loss[0.048]; loss[0.048]; loss[0.048]; loss[0.048]; widens[0.046]; Computer[0.044]; Revs[0.044]; Revs[0.044]; Numbers[0.044]; Avg[0.043]; Avg[0.043]; Aw[0.043]; Share[0.042]; Net[0.042]; Net[0.042]; Newsdesk[0.042]; \t\n\r [=======================================>......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 2s585ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4258/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 980testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.057:0.057[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhit[0.047]; selling[0.044]; gallon[0.043]; gallon[0.043]; volume[0.043]; early[0.043]; Day[0.042]; weekend[0.042]; September[0.041]; September[0.041]; September[0.041]; Texas[0.041]; losses[0.041]; York[0.041]; Friday[0.041]; profit[0.041]; low[0.040]; low[0.040]; session[0.040]; session[0.040]; session[0.040]; pre[0.040]; lower[0.040]; lower[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Texas ==> ENTITY: \u001b[32mTexas\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.057:0.057[0]; log p(e|m)= -0.161:-0.161[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nahead[0.045]; ahead[0.045]; hit[0.044]; early[0.043]; players[0.043]; weekend[0.043]; Friday[0.043]; Friday[0.043]; losses[0.043]; month[0.042]; shortly[0.041]; Day[0.041]; slim[0.041]; near[0.041]; heating[0.040]; heating[0.040]; heating[0.040]; heating[0.040]; said[0.040]; said[0.040]; said[0.040]; lower[0.040]; lower[0.040]; selling[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Labor Day\u001b[39m ==> ENTITY: \u001b[32mLabor Day\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.177:0.177[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSeptember[0.073]; September[0.073]; September[0.073]; month[0.049]; Friday[0.044]; Friday[0.044]; shortly[0.041]; weekend[0.039]; early[0.036]; amid[0.036]; losses[0.036]; close[0.035]; close[0.035]; close[0.035]; York[0.035]; near[0.035]; hit[0.035]; ahead[0.035]; ahead[0.035]; low[0.035]; low[0.035]; session[0.034]; session[0.034]; session[0.034]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Labor Day\u001b[39m ==> ENTITY: \u001b[32mLabor Day\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOctober[0.068]; October[0.068]; October[0.068]; month[0.050]; Monday[0.045]; Thursday[0.045]; Thursday[0.045]; Friday[0.045]; Day[0.040]; weekend[0.040]; holiday[0.038]; early[0.037]; start[0.037]; York[0.036]; ahead[0.035]; expiry[0.035]; closed[0.035]; said[0.034]; Labor[0.034]; left[0.034]; stood[0.034]; Desk[0.033]; cents[0.033]; players[0.033]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Labor Day\u001b[39m ==> ENTITY: \u001b[32mLabor Day\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSeptember[0.071]; October[0.065]; October[0.065]; October[0.065]; month[0.048]; Monday[0.043]; Thursday[0.043]; Thursday[0.043]; Friday[0.043]; Day[0.039]; weekend[0.038]; holiday[0.036]; early[0.035]; start[0.035]; York[0.034]; ahead[0.034]; expiry[0.034]; closed[0.033]; said[0.032]; said[0.032]; Labor[0.032]; left[0.032]; stood[0.032]; Desk[0.032]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.082:0.082[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSeptember[0.044]; September[0.044]; early[0.043]; session[0.043]; session[0.043]; session[0.043]; said[0.042]; said[0.042]; shortly[0.042]; close[0.042]; close[0.042]; close[0.042]; weekend[0.042]; hit[0.041]; ahead[0.041]; lackluster[0.041]; profit[0.041]; amid[0.040]; pre[0.040]; buying[0.040]; taking[0.040]; lingered[0.040]; cents[0.040]; cents[0.040]; \t\n\r [=======================================>......] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 2s569ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4264/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1076testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBenin\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBenin\u001b[39m <---> \u001b[32mBenin national football team\u001b[39m\t\nSCORES: global= 0.255:0.253[\u001b[31m0.003\u001b[39m]; local()= 0.130:0.183[\u001b[32m0.053\u001b[39m]; log p(e|m)= -0.420:-2.163[\u001b[31m1.743\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBenin[0.048]; Benin[0.048]; Soccer[0.048]; soccer[0.047]; Friday[0.046]; Cup[0.045]; Cup[0.045]; match[0.045]; drew[0.045]; African[0.044]; African[0.044]; Nations[0.044]; Nations[0.044]; preliminary[0.043]; won[0.043]; Mauritania[0.043]; Mauritania[0.043]; round[0.040]; Nouakchott[0.040]; second[0.040]; Draw[0.039]; aggregate[0.039]; leg[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nouakchott\u001b[39m ==> ENTITY: \u001b[32mNouakchott\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.118:0.118[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMauritania[0.053]; Mauritania[0.053]; leg[0.047]; Friday[0.047]; African[0.045]; African[0.045]; preliminary[0.044]; Soccer[0.043]; drew[0.043]; aggregate[0.043]; round[0.042]; soccer[0.042]; match[0.042]; Benin[0.042]; Benin[0.042]; Benin[0.042]; Cup[0.041]; Cup[0.041]; Nations[0.041]; Nations[0.041]; second[0.041]; Draw[0.040]; won[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBenin\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBenin\u001b[39m <---> \u001b[32mBenin national football team\u001b[39m\t\nSCORES: global= 0.255:0.253[\u001b[31m0.003\u001b[39m]; local()= 0.130:0.183[\u001b[32m0.053\u001b[39m]; log p(e|m)= -0.420:-2.163[\u001b[31m1.743\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBenin[0.048]; Benin[0.048]; Soccer[0.048]; soccer[0.047]; Friday[0.046]; Cup[0.045]; Cup[0.045]; match[0.045]; drew[0.045]; African[0.044]; African[0.044]; Nations[0.044]; Nations[0.044]; preliminary[0.043]; won[0.043]; Mauritania[0.043]; Mauritania[0.043]; round[0.040]; Nouakchott[0.040]; second[0.040]; Draw[0.039]; aggregate[0.039]; leg[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = African Nations Cup\u001b[39m ==> ENTITY: \u001b[32mAfrica Cup of Nations\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.054]; Soccer[0.051]; Cup[0.050]; match[0.050]; won[0.050]; Benin[0.049]; Benin[0.049]; Benin[0.049]; African[0.049]; Nations[0.048]; leg[0.048]; second[0.048]; Mauritania[0.047]; Mauritania[0.047]; preliminary[0.046]; round[0.046]; drew[0.045]; Draw[0.044]; Friday[0.044]; aggregate[0.043]; Nouakchott[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = African Nations Cup\u001b[39m ==> ENTITY: \u001b[32mAfrica Cup of Nations\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.054]; Soccer[0.051]; Cup[0.050]; match[0.050]; won[0.050]; Benin[0.049]; Benin[0.049]; Benin[0.049]; African[0.049]; Nations[0.048]; leg[0.048]; second[0.048]; Mauritania[0.047]; Mauritania[0.047]; preliminary[0.046]; round[0.046]; drew[0.045]; Draw[0.044]; Friday[0.044]; aggregate[0.043]; Nouakchott[0.043]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBenin\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBenin\u001b[39m <---> \u001b[32mBenin national football team\u001b[39m\t\nSCORES: global= 0.255:0.253[\u001b[31m0.003\u001b[39m]; local()= 0.130:0.183[\u001b[32m0.053\u001b[39m]; log p(e|m)= -0.420:-2.163[\u001b[31m1.743\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBenin[0.048]; Benin[0.048]; Soccer[0.048]; soccer[0.047]; Friday[0.046]; Cup[0.045]; Cup[0.045]; match[0.045]; drew[0.045]; African[0.044]; African[0.044]; Nations[0.044]; Nations[0.044]; preliminary[0.043]; won[0.043]; Mauritania[0.043]; Mauritania[0.043]; round[0.040]; Nouakchott[0.040]; second[0.040]; Draw[0.039]; aggregate[0.039]; leg[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mauritania ==> ENTITY: \u001b[32mMauritania national football team\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.223:0.223[0]; log p(e|m)= -2.146:-2.146[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMauritania[0.051]; soccer[0.046]; Cup[0.045]; Cup[0.045]; Nations[0.045]; Nations[0.045]; African[0.045]; African[0.045]; Benin[0.045]; Benin[0.045]; Benin[0.045]; Soccer[0.044]; match[0.044]; Nouakchott[0.044]; won[0.043]; preliminary[0.042]; drew[0.041]; leg[0.041]; round[0.040]; Friday[0.040]; second[0.040]; aggregate[0.040]; Draw[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Mauritania ==> ENTITY: \u001b[32mMauritania national football team\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.223:0.223[0]; log p(e|m)= -2.146:-2.146[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMauritania[0.051]; soccer[0.046]; Cup[0.045]; Cup[0.045]; Nations[0.045]; Nations[0.045]; African[0.045]; African[0.045]; Benin[0.045]; Benin[0.045]; Benin[0.045]; Soccer[0.044]; match[0.044]; Nouakchott[0.044]; won[0.043]; preliminary[0.042]; drew[0.041]; leg[0.041]; round[0.040]; Friday[0.040]; second[0.040]; aggregate[0.040]; Draw[0.040]; \t\n\r [========================================>.....] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 2s527ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4272/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 972testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.198:0.198[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.055]; Australia[0.055]; cricket[0.052]; match[0.049]; Cricket[0.049]; tournament[0.047]; Colombo[0.047]; day[0.046]; overs[0.045]; overs[0.045]; overs[0.045]; wickets[0.045]; World[0.044]; Lanka[0.043]; Lanka[0.043]; Lanka[0.043]; Friday[0.042]; Wickets[0.042]; Scores[0.041]; beat[0.041]; Series[0.041]; Beat[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sri Lanka ==> ENTITY: \u001b[32mSri Lanka national cricket team\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.208:0.208[0]; log p(e|m)= -2.263:-2.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.046]; Australia[0.045]; Australia[0.045]; Australia[0.045]; Colombo[0.044]; Lanka[0.044]; Lanka[0.044]; Cricket[0.044]; match[0.044]; overs[0.043]; overs[0.043]; overs[0.043]; tournament[0.043]; wickets[0.042]; World[0.042]; day[0.042]; Sri[0.039]; Sri[0.039]; Friday[0.039]; beat[0.039]; Wickets[0.038]; Scores[0.037]; Series[0.037]; Beat[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sri Lanka ==> ENTITY: \u001b[32mSri Lanka national cricket team\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.208:0.208[0]; log p(e|m)= -2.263:-2.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.046]; Australia[0.045]; Australia[0.045]; Australia[0.045]; Colombo[0.044]; Lanka[0.044]; Lanka[0.044]; Cricket[0.044]; match[0.044]; overs[0.043]; overs[0.043]; overs[0.043]; tournament[0.043]; wickets[0.042]; World[0.042]; day[0.042]; Sri[0.039]; Sri[0.039]; Friday[0.039]; beat[0.039]; Wickets[0.038]; Scores[0.037]; Series[0.037]; Beat[0.036]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.198:0.198[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.055]; Australia[0.055]; cricket[0.052]; match[0.049]; Cricket[0.049]; tournament[0.047]; Colombo[0.047]; day[0.046]; overs[0.045]; overs[0.045]; overs[0.045]; wickets[0.045]; World[0.044]; Lanka[0.043]; Lanka[0.043]; Lanka[0.043]; Friday[0.042]; Wickets[0.042]; Scores[0.041]; beat[0.041]; Series[0.041]; Beat[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Colombo ==> ENTITY: \u001b[32mColombo\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.173:0.173[0]; log p(e|m)= -0.194:-0.194[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.046]; Lanka[0.045]; Lanka[0.045]; Lanka[0.045]; Cricket[0.045]; match[0.043]; Australia[0.043]; Australia[0.043]; Australia[0.043]; tournament[0.042]; Sri[0.042]; Sri[0.042]; Sri[0.042]; day[0.041]; World[0.041]; overs[0.040]; overs[0.040]; overs[0.040]; Series[0.040]; Scores[0.040]; wickets[0.039]; Friday[0.039]; Wickets[0.038]; beat[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Singer World Series\u001b[39m ==> ENTITY: \u001b[32mSinger World Series\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.047]; Australia[0.047]; Australia[0.047]; cricket[0.046]; Colombo[0.045]; Cricket[0.045]; Lanka[0.044]; Lanka[0.044]; Lanka[0.044]; match[0.042]; Sri[0.041]; Sri[0.041]; Sri[0.041]; overs[0.041]; overs[0.041]; overs[0.041]; wickets[0.040]; tournament[0.040]; Friday[0.040]; Scores[0.038]; Wickets[0.038]; day[0.037]; beat[0.036]; Beat[0.034]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australia ==> ENTITY: \u001b[32mAustralia national cricket team\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.198:0.198[0]; log p(e|m)= -3.507:-3.507[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAustralia[0.055]; Australia[0.055]; cricket[0.052]; match[0.049]; Cricket[0.049]; tournament[0.047]; Colombo[0.047]; day[0.046]; overs[0.045]; overs[0.045]; overs[0.045]; wickets[0.045]; World[0.044]; Lanka[0.043]; Lanka[0.043]; Lanka[0.043]; Friday[0.042]; Wickets[0.042]; Scores[0.041]; beat[0.041]; Series[0.041]; Beat[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sri Lanka ==> ENTITY: \u001b[32mSri Lanka national cricket team\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.208:0.208[0]; log p(e|m)= -2.263:-2.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncricket[0.046]; Australia[0.045]; Australia[0.045]; Australia[0.045]; Colombo[0.044]; Lanka[0.044]; Lanka[0.044]; Cricket[0.044]; match[0.044]; overs[0.043]; overs[0.043]; overs[0.043]; tournament[0.043]; wickets[0.042]; World[0.042]; day[0.042]; Sri[0.039]; Sri[0.039]; Friday[0.039]; beat[0.039]; Wickets[0.038]; Scores[0.037]; Series[0.037]; Beat[0.036]; \t\n\r [========================================>.....] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 2s513ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4280/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1090testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Greater Milwaukee Open\u001b[39m ==> ENTITY: \u001b[32mU.S. Bank Championship in Milwaukee\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.048]; Golf[0.047]; Golf[0.047]; tournament[0.046]; round[0.046]; round[0.046]; Tour[0.046]; final[0.044]; course[0.043]; victory[0.043]; Park[0.043]; stroke[0.043]; strokes[0.043]; lead[0.043]; Saturday[0.042]; Saturday[0.042]; record[0.042]; record[0.042]; moved[0.041]; Deer[0.041]; Shot[0.041]; set[0.041]; Course[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Henke\u001b[39m ==> ENTITY: \u001b[32mNolan Henke\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nround[0.044]; Estes[0.044]; Parnevik[0.044]; Parnevik[0.044]; Henke[0.043]; course[0.043]; par[0.043]; par[0.043]; seventh[0.042]; Bob[0.042]; bogey[0.041]; birdied[0.041]; birdied[0.041]; seven[0.041]; game[0.041]; second[0.041]; really[0.041]; putted[0.041]; footer[0.040]; footer[0.040]; ve[0.040]; ve[0.040]; shot[0.040]; bogeying[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nolan Henke\u001b[39m ==> ENTITY: \u001b[32mNolan Henke\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.212:0.212[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmateur[0.044]; Golf[0.043]; Golf[0.043]; Open[0.043]; Open[0.043]; tournament[0.043]; tournament[0.043]; tournament[0.043]; Classic[0.043]; Tour[0.042]; Tour[0.042]; carded[0.041]; round[0.041]; round[0.041]; Championship[0.041]; victory[0.041]; victory[0.041]; professional[0.041]; professional[0.041]; Course[0.040]; Parnevik[0.040]; Parnevik[0.040]; Parnevik[0.040]; Parnevik[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = PGA Tour\u001b[39m ==> ENTITY: \u001b[32mPGA Tour\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.241:0.241[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTour[0.047]; tournament[0.044]; tournament[0.044]; Golf[0.044]; Golf[0.044]; fourth[0.043]; stroke[0.042]; Open[0.042]; Open[0.042]; final[0.042]; round[0.042]; round[0.042]; Classic[0.041]; hole[0.041]; Woods[0.041]; strokes[0.041]; professional[0.041]; Parnevik[0.040]; Parnevik[0.040]; Parnevik[0.040]; Parnevik[0.040]; entering[0.040]; Course[0.040]; career[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWorld Series of Golf\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mWGC-Bridgestone Invitational\u001b[39m <---> \u001b[32mWorld Series of Golf\u001b[39m\t\nSCORES: global= 0.284:0.283[\u001b[31m0.001\u001b[39m]; local()= 0.215:0.164[\u001b[31m0.051\u001b[39m]; log p(e|m)= -0.711:0.000[\u001b[32m0.711\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.048]; Golf[0.045]; Tour[0.044]; golfers[0.044]; second[0.044]; second[0.044]; week[0.043]; place[0.042]; skipped[0.042]; Steve[0.042]; Stuart[0.042]; Bob[0.042]; World[0.041]; Estes[0.041]; Stricker[0.041]; Series[0.041]; Appleby[0.040]; shot[0.040]; tied[0.039]; tied[0.039]; lead[0.039]; lead[0.039]; carded[0.039]; list[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Sweden ==> ENTITY: \u001b[32mSweden\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.085:0.085[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.046]; Tour[0.044]; final[0.043]; Golf[0.043]; Golf[0.043]; record[0.043]; record[0.043]; shot[0.043]; round[0.042]; round[0.042]; Open[0.042]; Open[0.042]; Jesper[0.042]; million[0.042]; tied[0.042]; victory[0.041]; tying[0.041]; entering[0.040]; Shot[0.040]; Park[0.040]; lead[0.039]; Takes[0.039]; Saturday[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Parnevik\u001b[39m ==> ENTITY: \u001b[32mJesper Parnevik\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nParnevik[0.044]; Parnevik[0.044]; professional[0.044]; professional[0.044]; career[0.044]; Tour[0.043]; tournament[0.043]; tournament[0.043]; round[0.042]; round[0.042]; Amateur[0.042]; Championship[0.042]; Woods[0.042]; Woods[0.042]; Swede[0.041]; victory[0.041]; course[0.041]; winning[0.041]; second[0.040]; Classic[0.040]; cut[0.040]; Tiger[0.039]; fourth[0.039]; game[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mWorld Series of Golf\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mWGC-Bridgestone Invitational\u001b[39m <---> \u001b[32mWorld Series of Golf\u001b[39m\t\nSCORES: global= 0.284:0.283[\u001b[31m0.000\u001b[39m]; local()= 0.221:0.177[\u001b[31m0.044\u001b[39m]; log p(e|m)= -0.711:0.000[\u001b[32m0.711\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.047]; Golf[0.045]; Tour[0.044]; golfers[0.043]; second[0.043]; second[0.043]; week[0.042]; hole[0.042]; hole[0.042]; place[0.042]; skipped[0.042]; Steve[0.041]; Stuart[0.041]; Bob[0.041]; seventh[0.041]; World[0.041]; Estes[0.041]; Stricker[0.040]; Series[0.040]; Appleby[0.040]; shot[0.040]; par[0.039]; tied[0.039]; tied[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tiger Woods\u001b[39m ==> ENTITY: \u001b[32mTiger Woods\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.250:0.250[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.048]; Woods[0.046]; tournament[0.044]; tournament[0.044]; tournament[0.044]; Amateur[0.043]; Open[0.043]; Tour[0.042]; Tour[0.042]; fourth[0.041]; round[0.041]; round[0.041]; round[0.041]; round[0.041]; professional[0.041]; professional[0.041]; Championship[0.041]; second[0.040]; Course[0.040]; final[0.040]; Parnevik[0.039]; Parnevik[0.039]; Parnevik[0.039]; Park[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Loren Roberts\u001b[39m ==> ENTITY: \u001b[32mLoren Roberts\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.234:0.234[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.045]; Open[0.045]; Golf[0.044]; Golf[0.044]; Tour[0.044]; Tour[0.044]; tournament[0.043]; tournament[0.043]; tournament[0.043]; round[0.042]; round[0.042]; Classic[0.041]; fourth[0.041]; stroke[0.041]; Woods[0.041]; Woods[0.041]; final[0.040]; Course[0.040]; Parnevik[0.040]; Parnevik[0.040]; Parnevik[0.040]; Parnevik[0.040]; professional[0.040]; professional[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Greater Milwaukee Open\u001b[39m ==> ENTITY: \u001b[32mU.S. Bank Championship in Milwaukee\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.045]; Golf[0.044]; Golf[0.044]; tournament[0.044]; tournament[0.044]; round[0.043]; round[0.043]; Classic[0.043]; Tour[0.043]; Tour[0.043]; cut[0.041]; final[0.041]; fourth[0.041]; course[0.041]; victory[0.041]; victory[0.041]; Park[0.041]; Woods[0.041]; stroke[0.040]; strokes[0.040]; lead[0.040]; professional[0.039]; Saturday[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Henke\u001b[39m ==> ENTITY: \u001b[32mNolan Henke\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.048]; round[0.046]; Estes[0.045]; Parnevik[0.045]; Henke[0.045]; par[0.044]; par[0.044]; seventh[0.044]; lead[0.044]; Series[0.043]; Series[0.043]; Bob[0.043]; bogey[0.043]; birdied[0.043]; Waldorf[0.043]; Appleby[0.043]; place[0.043]; game[0.042]; Stricker[0.042]; second[0.042]; second[0.042]; really[0.042]; putted[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Milwaukee ==> ENTITY: \u001b[32mMilwaukee\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.088:0.088[0]; log p(e|m)= -0.282:-0.282[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrecord[0.046]; record[0.046]; Milwaukee[0.046]; Milwaukee[0.046]; victory[0.045]; tying[0.045]; tournament[0.044]; Wisconsin[0.044]; final[0.044]; Saturday[0.043]; Saturday[0.043]; round[0.043]; round[0.043]; Park[0.043]; lead[0.043]; Greg[0.043]; tied[0.043]; led[0.043]; moved[0.042]; Brown[0.042]; million[0.042]; Takes[0.042]; Deer[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kraft\u001b[39m ==> ENTITY: \u001b[32mGreg Kraft\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.121:0.121[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTour[0.045]; golfers[0.045]; tournament[0.045]; tied[0.043]; tied[0.043]; Stricker[0.042]; Golf[0.042]; Golf[0.042]; Series[0.042]; Series[0.042]; Waldorf[0.041]; Steve[0.041]; second[0.041]; second[0.041]; lead[0.041]; lead[0.041]; really[0.041]; Bob[0.041]; place[0.040]; pars[0.040]; carded[0.040]; week[0.040]; skipped[0.040]; group[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Parnevik\u001b[39m ==> ENTITY: \u001b[32mJesper Parnevik\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.188:0.188[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nParnevik[0.045]; Parnevik[0.045]; professional[0.045]; professional[0.045]; tournament[0.044]; round[0.043]; round[0.043]; round[0.043]; Amateur[0.043]; Championship[0.043]; Woods[0.043]; Swede[0.042]; course[0.042]; winning[0.041]; second[0.041]; game[0.040]; birdied[0.039]; birdied[0.039]; ninth[0.039]; seven[0.039]; hole[0.039]; hole[0.039]; hole[0.039]; seventh[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Greg Kraft\u001b[39m ==> ENTITY: \u001b[32mGreg Kraft\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTour[0.044]; Tour[0.044]; tournament[0.043]; tournament[0.043]; tournament[0.043]; professional[0.043]; professional[0.043]; Championship[0.043]; record[0.042]; record[0.042]; Amateur[0.042]; round[0.042]; round[0.042]; cut[0.042]; tied[0.041]; Open[0.041]; Open[0.041]; Golf[0.041]; Golf[0.041]; course[0.041]; Shot[0.041]; Classic[0.040]; Course[0.040]; Tiger[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Parnevik\u001b[39m ==> ENTITY: \u001b[32mJesper Parnevik\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.044]; Golf[0.044]; Open[0.043]; Open[0.043]; Parnevik[0.043]; Parnevik[0.043]; Parnevik[0.043]; professional[0.043]; career[0.043]; Tour[0.042]; Tour[0.042]; tournament[0.042]; tournament[0.042]; stroke[0.042]; round[0.041]; round[0.041]; Woods[0.041]; strokes[0.040]; victory[0.040]; victory[0.040]; course[0.040]; Sweden[0.039]; Classic[0.039]; cut[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Stuart Appleby\u001b[39m ==> ENTITY: \u001b[32mStuart Appleby\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ngolfers[0.048]; tournament[0.048]; Golf[0.048]; Golf[0.048]; skipped[0.045]; Tour[0.045]; second[0.044]; second[0.044]; carded[0.043]; shot[0.043]; seventh[0.043]; place[0.042]; lead[0.042]; lead[0.042]; Series[0.042]; Series[0.042]; tied[0.042]; tied[0.042]; shots[0.042]; Bob[0.042]; Stricker[0.042]; list[0.042]; Estes[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Woods\u001b[39m ==> ENTITY: \u001b[32mTiger Woods\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.244:0.244[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.048]; Woods[0.046]; tournament[0.044]; tournament[0.044]; tournament[0.044]; Amateur[0.043]; Tiger[0.042]; Tour[0.042]; fourth[0.041]; round[0.041]; round[0.041]; round[0.041]; professional[0.041]; professional[0.041]; Championship[0.041]; second[0.041]; course[0.041]; Course[0.040]; Parnevik[0.040]; Parnevik[0.040]; Parnevik[0.040]; Park[0.040]; winning[0.039]; Classic[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bob Estes\u001b[39m ==> ENTITY: \u001b[32mBob Estes\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nround[0.045]; tournament[0.044]; Golf[0.044]; Golf[0.044]; Tour[0.043]; place[0.043]; golfers[0.043]; par[0.042]; par[0.042]; list[0.042]; Steve[0.042]; good[0.041]; tied[0.041]; tied[0.041]; Series[0.041]; Series[0.041]; second[0.041]; second[0.041]; today[0.040]; World[0.040]; World[0.040]; Waldorf[0.040]; Stuart[0.040]; carded[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = PGA Tour\u001b[39m ==> ENTITY: \u001b[32mPGA Tour\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.250:0.250[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTour[0.048]; Championship[0.046]; tournament[0.045]; tournament[0.045]; tournament[0.045]; Golf[0.045]; Amateur[0.044]; fourth[0.044]; stroke[0.044]; Open[0.044]; final[0.043]; round[0.043]; round[0.043]; round[0.043]; seventh[0.042]; Classic[0.042]; hole[0.042]; hole[0.042]; hole[0.042]; ninth[0.042]; Woods[0.042]; Woods[0.042]; strokes[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Parnevik\u001b[39m ==> ENTITY: \u001b[32mJesper Parnevik\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.043]; Golf[0.043]; Open[0.043]; Open[0.043]; Parnevik[0.042]; Parnevik[0.042]; Parnevik[0.042]; professional[0.042]; career[0.042]; Tour[0.042]; Tour[0.042]; tournament[0.042]; tournament[0.042]; tournament[0.042]; stroke[0.041]; round[0.041]; round[0.041]; Woods[0.041]; Woods[0.041]; strokes[0.040]; victory[0.040]; victory[0.040]; course[0.040]; Sweden[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = PGA Tour\u001b[39m ==> ENTITY: \u001b[32mPGA Tour\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.046]; Golf[0.046]; Golf[0.046]; list[0.045]; second[0.044]; second[0.044]; golfers[0.043]; week[0.043]; skipped[0.042]; World[0.042]; World[0.042]; money[0.042]; place[0.042]; Appleby[0.041]; carded[0.041]; Stricker[0.040]; Series[0.040]; Series[0.040]; Estes[0.040]; Bob[0.040]; Steve[0.039]; lead[0.038]; lead[0.038]; shots[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Parnevik\u001b[39m ==> ENTITY: \u001b[32mJesper Parnevik\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.181:0.181[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.045]; Golf[0.045]; Open[0.045]; Open[0.045]; Parnevik[0.044]; Parnevik[0.044]; Parnevik[0.044]; Tour[0.044]; tournament[0.044]; round[0.043]; strokes[0.042]; victory[0.042]; course[0.041]; Sweden[0.041]; Park[0.040]; final[0.040]; Greg[0.040]; Shot[0.039]; Course[0.039]; record[0.039]; record[0.039]; Takes[0.039]; hole[0.039]; Jesper[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Wisconsin ==> ENTITY: \u001b[32mWisconsin\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.063:0.063[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.047]; record[0.044]; record[0.044]; victory[0.043]; Saturday[0.043]; Saturday[0.043]; round[0.043]; round[0.043]; tying[0.043]; Tour[0.042]; final[0.042]; Brown[0.042]; Milwaukee[0.041]; Milwaukee[0.041]; Milwaukee[0.041]; entering[0.041]; tied[0.040]; Greg[0.040]; moved[0.040]; Park[0.040]; led[0.040]; Golf[0.039]; Golf[0.039]; Roberts[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Parnevik\u001b[39m ==> ENTITY: \u001b[32mJesper Parnevik\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nParnevik[0.048]; Parnevik[0.048]; round[0.046]; round[0.046]; round[0.046]; Swede[0.046]; course[0.045]; second[0.044]; game[0.043]; birdied[0.042]; birdied[0.042]; ninth[0.042]; seven[0.042]; hole[0.042]; hole[0.042]; hole[0.042]; hole[0.042]; hole[0.042]; seventh[0.042]; seventh[0.042]; bogey[0.042]; free[0.042]; driver[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Duffy Waldorf\u001b[39m ==> ENTITY: \u001b[32mDuffy Waldorf\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.118:0.118[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.046]; Golf[0.046]; tournament[0.045]; golfers[0.045]; Tour[0.044]; skipped[0.042]; Series[0.042]; Series[0.042]; tied[0.042]; tied[0.042]; Estes[0.041]; really[0.041]; list[0.041]; World[0.041]; World[0.041]; Stricker[0.040]; carded[0.040]; make[0.040]; Jasper[0.040]; second[0.040]; second[0.040]; pars[0.040]; said[0.040]; Appleby[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.S. Amateur Championship\u001b[39m ==> ENTITY: \u001b[32mUnited States Amateur Championship (golf)\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.221:0.221[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.046]; tournament[0.046]; Woods[0.044]; Woods[0.044]; Golf[0.044]; round[0.043]; round[0.043]; round[0.043]; professional[0.043]; professional[0.043]; Tour[0.042]; Championship[0.041]; Classic[0.041]; Greg[0.040]; Course[0.040]; winning[0.040]; seven[0.040]; course[0.040]; Park[0.040]; career[0.040]; second[0.040]; Tiger[0.039]; Brown[0.039]; Tuesday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Brown Deer Park Golf Course\u001b[39m ==> ENTITY: \u001b[32mBrown Deer Park Golf Course\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWoods[0.047]; Woods[0.047]; Golf[0.046]; Championship[0.045]; course[0.044]; Amateur[0.044]; tournament[0.044]; tournament[0.044]; tournament[0.044]; hole[0.043]; Tiger[0.043]; Open[0.043]; Open[0.043]; Tour[0.043]; Tour[0.043]; lead[0.043]; final[0.042]; tied[0.042]; round[0.042]; round[0.042]; strokes[0.042]; cut[0.042]; stroke[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Steve Stricker\u001b[39m ==> ENTITY: \u001b[32mSteve Stricker\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.251:0.251[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.046]; Golf[0.046]; tournament[0.045]; round[0.044]; Tour[0.044]; golfers[0.043]; second[0.043]; second[0.043]; Appleby[0.042]; tied[0.041]; tied[0.041]; week[0.041]; shot[0.041]; par[0.041]; seventh[0.041]; carded[0.040]; place[0.040]; birdied[0.040]; lead[0.040]; lead[0.040]; hole[0.040]; hole[0.040]; skipped[0.040]; hit[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jesper Parnevik\u001b[39m ==> ENTITY: \u001b[32mJesper Parnevik\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.192:0.192[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGolf[0.045]; Golf[0.045]; Open[0.045]; Open[0.045]; Parnevik[0.044]; Parnevik[0.044]; Parnevik[0.044]; Tour[0.044]; tournament[0.043]; stroke[0.043]; round[0.042]; round[0.042]; strokes[0.042]; victory[0.041]; course[0.041]; Sweden[0.040]; Park[0.040]; final[0.039]; Greg[0.039]; Shot[0.039]; Course[0.039]; record[0.038]; record[0.038]; carded[0.038]; \t\n\r [========================================>.....] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 2s355ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4311/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1091testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Monica\u001b[39m ==> ENTITY: \u001b[32mMonica Seles\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.216:0.216[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSeles[0.048]; Seles[0.048]; Seles[0.048]; Seles[0.048]; match[0.045]; singles[0.045]; round[0.045]; matches[0.045]; champion[0.044]; Open[0.044]; semifinals[0.043]; won[0.043]; win[0.043]; victory[0.042]; opening[0.042]; draw[0.042]; wrist[0.041]; Davenport[0.041]; injury[0.041]; Lindsay[0.041]; threesome[0.041]; debut[0.040]; Monica[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = King\u001b[39m ==> ENTITY: \u001b[32mBillie Jean King\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.167:0.167[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.044]; Davenport[0.043]; match[0.043]; singles[0.043]; champion[0.043]; won[0.042]; matches[0.042]; Seles[0.042]; Seles[0.042]; Seles[0.042]; Seles[0.042]; Lindsay[0.042]; team[0.042]; Arantxa[0.041]; round[0.041]; win[0.041]; semifinals[0.041]; Mary[0.041]; Mary[0.041]; King[0.040]; King[0.040]; King[0.040]; King[0.040]; King[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = King\u001b[39m ==> ENTITY: \u001b[32mBillie Jean King\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.150:0.150[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.045]; Davenport[0.044]; match[0.044]; champion[0.044]; Seles[0.043]; Seles[0.043]; Lindsay[0.043]; team[0.043]; Arantxa[0.042]; win[0.042]; Mary[0.041]; Mary[0.041]; King[0.041]; King[0.041]; King[0.041]; King[0.041]; Olympic[0.041]; Joe[0.041]; Joe[0.041]; Monica[0.041]; coach[0.040]; veteran[0.040]; good[0.039]; got[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mSpain\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSpain\u001b[39m <---> \u001b[32mSpain Fed Cup team\u001b[39m\t\nSCORES: global= 0.242:0.221[\u001b[31m0.022\u001b[39m]; local()= 0.148:0.220[\u001b[32m0.073\u001b[39m]; log p(e|m)= -0.189:-4.962[\u001b[31m4.773\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.049]; Spain[0.049]; matches[0.044]; Cup[0.043]; Cup[0.043]; team[0.043]; team[0.043]; Tennis[0.042]; international[0.041]; Netherlands[0.041]; final[0.041]; Open[0.041]; Open[0.041]; women[0.041]; Americans[0.040]; semifinals[0.040]; Germany[0.040]; draw[0.040]; France[0.040]; Belgium[0.040]; champion[0.040]; round[0.040]; round[0.040]; competition[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.047:0.047[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.049]; matches[0.048]; matches[0.048]; final[0.047]; final[0.047]; Spain[0.047]; won[0.047]; number[0.046]; month[0.045]; Austria[0.045]; singles[0.045]; debut[0.045]; win[0.045]; semifinals[0.045]; semifinals[0.045]; Germany[0.045]; Cup[0.044]; Cup[0.044]; draw[0.044]; court[0.044]; States[0.043]; captain[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.070:0.070[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.044]; team[0.044]; competition[0.044]; Open[0.043]; Open[0.043]; matches[0.043]; hit[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; international[0.042]; host[0.042]; semifinals[0.041]; Germany[0.040]; final[0.040]; final[0.040]; champion[0.040]; Cup[0.040]; Cup[0.040]; round[0.040]; round[0.040]; hosts[0.040]; draw[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Seles\u001b[39m ==> ENTITY: \u001b[32mMonica Seles\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.218:0.218[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSeles[0.046]; Seles[0.046]; Seles[0.046]; Arantxa[0.046]; match[0.044]; singles[0.043]; round[0.043]; matches[0.043]; champion[0.043]; Open[0.042]; semifinals[0.041]; won[0.041]; win[0.041]; victory[0.041]; opening[0.040]; draw[0.040]; wrist[0.040]; Davenport[0.040]; injury[0.040]; duo[0.039]; Lindsay[0.039]; Vicario[0.039]; threesome[0.039]; debut[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Fed Cup ==> ENTITY: \u001b[32mFed Cup\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.187:0.187[0]; log p(e|m)= -0.311:-0.311[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.046]; Cup[0.045]; match[0.044]; team[0.044]; singles[0.043]; won[0.043]; semifinals[0.042]; Spain[0.042]; Open[0.042]; champion[0.042]; win[0.041]; Arantxa[0.041]; victory[0.041]; Austria[0.041]; draw[0.041]; Japan[0.041]; Japan[0.041]; number[0.041]; Seles[0.040]; Seles[0.040]; Seles[0.040]; Seles[0.040]; squad[0.040]; goal[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Arantxa Sanchez Vicario\u001b[39m ==> ENTITY: \u001b[32mArantxa Sánchez Vicario\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.045]; match[0.045]; Seles[0.045]; Seles[0.045]; Davenport[0.044]; champion[0.042]; Spain[0.042]; Cup[0.042]; Olympic[0.042]; Mary[0.042]; Mary[0.042]; win[0.042]; Fernandez[0.041]; Fernandez[0.041]; Fernandez[0.041]; Martinez[0.041]; Lindsay[0.040]; injury[0.040]; withdraw[0.039]; coach[0.039]; veteran[0.039]; injuries[0.039]; Joe[0.039]; Joe[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Conchita Martinez\u001b[39m ==> ENTITY: \u001b[32mConchita Martínez\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArantxa[0.048]; Open[0.046]; Seles[0.044]; match[0.044]; Fernandez[0.043]; Fernandez[0.043]; Fernandez[0.043]; Vicario[0.043]; Davenport[0.042]; champion[0.042]; Spain[0.042]; Mary[0.041]; Mary[0.041]; Olympic[0.041]; Cup[0.041]; Lindsay[0.040]; coach[0.040]; injuries[0.040]; wrist[0.040]; Sanchez[0.040]; injury[0.040]; threesome[0.039]; Joe[0.039]; Joe[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Billie Jean King\u001b[39m ==> ENTITY: \u001b[32mBillie Jean King\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncourt[0.044]; singles[0.044]; won[0.043]; matches[0.043]; Seles[0.043]; Seles[0.043]; Seles[0.043]; team[0.042]; final[0.042]; final[0.042]; host[0.042]; round[0.042]; round[0.042]; win[0.042]; semifinals[0.041]; semifinals[0.041]; King[0.041]; King[0.041]; victory[0.041]; month[0.041]; Monica[0.040]; Monica[0.040]; number[0.040]; lost[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. Open ==> ENTITY: \u001b[32mUS Open (tennis)\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.189:0.189[0]; log p(e|m)= -1.238:-1.238[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTennis[0.048]; final[0.044]; final[0.044]; matches[0.044]; competition[0.044]; Open[0.043]; Open[0.043]; round[0.043]; round[0.043]; international[0.042]; host[0.042]; semifinals[0.042]; champion[0.042]; draw[0.041]; Cup[0.040]; Cup[0.040]; hosts[0.040]; Richard[0.040]; Germany[0.039]; team[0.039]; team[0.039]; Americans[0.039]; women[0.039]; lost[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Fed Cup ==> ENTITY: \u001b[32mFed Cup\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.188:0.188[0]; log p(e|m)= -0.311:-0.311[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.046]; Cup[0.045]; Cup[0.045]; match[0.044]; team[0.043]; singles[0.043]; won[0.043]; final[0.042]; semifinals[0.042]; Spain[0.042]; Open[0.042]; win[0.041]; victory[0.041]; Austria[0.041]; draw[0.040]; Japan[0.040]; Japan[0.040]; number[0.040]; Seles[0.040]; Seles[0.040]; Seles[0.040]; lost[0.040]; squad[0.040]; goal[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mary Joe Fernandez\u001b[39m ==> ENTITY: \u001b[32mMary Joe Fernández\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.206:0.206[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; Arantxa[0.045]; match[0.044]; champion[0.043]; won[0.043]; singles[0.043]; Davenport[0.043]; semifinals[0.042]; win[0.042]; victory[0.042]; Open[0.042]; Martinez[0.042]; Fernandez[0.041]; Fernandez[0.041]; Vicario[0.041]; Monica[0.040]; Monica[0.040]; Olympic[0.040]; round[0.040]; team[0.040]; Seles[0.040]; Seles[0.040]; Seles[0.040]; Seles[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = King\u001b[39m ==> ENTITY: \u001b[32mBillie Jean King\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.166:0.166[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.044]; Davenport[0.043]; match[0.043]; singles[0.043]; champion[0.043]; won[0.042]; matches[0.042]; Seles[0.042]; Seles[0.042]; Seles[0.042]; Seles[0.042]; Lindsay[0.042]; team[0.042]; Arantxa[0.041]; round[0.041]; win[0.041]; semifinals[0.041]; Mary[0.041]; Mary[0.041]; King[0.040]; King[0.040]; King[0.040]; King[0.040]; victory[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mFrance\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFrance\u001b[39m <---> \u001b[32mFrance Fed Cup team\u001b[39m\t\nSCORES: global= 0.241:0.219[\u001b[31m0.022\u001b[39m]; local()= 0.140:0.168[\u001b[32m0.027\u001b[39m]; log p(e|m)= -0.163:-5.521[\u001b[31m5.359\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.046]; team[0.046]; matches[0.045]; Belgium[0.045]; Spain[0.044]; Spain[0.044]; Spain[0.044]; Spain[0.044]; Austria[0.044]; Germany[0.043]; Open[0.043]; Open[0.043]; international[0.043]; Republic[0.043]; semifinals[0.043]; Cup[0.043]; Cup[0.043]; Cup[0.043]; final[0.043]; final[0.043]; champion[0.043]; Netherlands[0.042]; Tennis[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = King\u001b[39m ==> ENTITY: \u001b[32mBillie Jean King\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.136:0.136[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.046]; Davenport[0.045]; match[0.045]; champion[0.045]; Seles[0.044]; Lindsay[0.044]; Arantxa[0.043]; Mary[0.043]; King[0.042]; King[0.042]; Olympic[0.042]; Joe[0.042]; coach[0.041]; good[0.040]; got[0.040]; Fernandez[0.040]; Fernandez[0.040]; injuries[0.040]; forced[0.040]; right[0.040]; right[0.040]; said[0.039]; said[0.039]; duo[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mUnited States\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUnited States\u001b[39m <---> \u001b[32mUnited States Fed Cup team\u001b[39m\t\nSCORES: global= 0.239:0.219[\u001b[31m0.021\u001b[39m]; local()= 0.036:0.184[\u001b[32m0.148\u001b[39m]; log p(e|m)= -0.074:-5.298[\u001b[31m5.225\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.045]; won[0.045]; matches[0.045]; Cup[0.043]; Cup[0.043]; Cup[0.043]; Spain[0.043]; final[0.043]; final[0.043]; Austria[0.041]; host[0.041]; win[0.041]; semifinals[0.041]; semifinals[0.041]; captain[0.041]; singles[0.040]; victory[0.040]; draw[0.040]; France[0.040]; opening[0.040]; blanked[0.040]; debut[0.040]; Joe[0.040]; round[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Monica Seles\u001b[39m ==> ENTITY: \u001b[32mMonica Seles\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSeles[0.047]; Seles[0.047]; match[0.044]; singles[0.044]; round[0.044]; matches[0.044]; Open[0.043]; final[0.043]; semifinals[0.042]; won[0.042]; win[0.042]; victory[0.041]; opening[0.041]; draw[0.041]; wrist[0.041]; injury[0.040]; debut[0.040]; Monica[0.039]; Fernandez[0.039]; Fernandez[0.039]; defeat[0.039]; number[0.039]; Salzburg[0.039]; Austria[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Fernandez\u001b[39m ==> ENTITY: \u001b[32mMary Joe Fernández\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.203:0.203[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; Arantxa[0.045]; match[0.044]; champion[0.043]; won[0.043]; singles[0.043]; Davenport[0.043]; semifinals[0.042]; win[0.042]; victory[0.042]; Open[0.042]; Martinez[0.042]; Fernandez[0.042]; Fernandez[0.042]; Vicario[0.041]; Monica[0.040]; Monica[0.040]; Olympic[0.040]; team[0.040]; Seles[0.040]; Seles[0.040]; Seles[0.040]; Seles[0.040]; Lindsay[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. Open ==> ENTITY: \u001b[32mUS Open (tennis)\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.182:0.182[0]; log p(e|m)= -1.238:-1.238[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.045]; match[0.044]; Open[0.044]; Arantxa[0.043]; win[0.043]; won[0.043]; singles[0.042]; Seles[0.042]; Seles[0.042]; Seles[0.042]; Seles[0.042]; semifinals[0.042]; victory[0.042]; Davenport[0.042]; champion[0.042]; threesome[0.041]; number[0.040]; withdraw[0.040]; team[0.040]; Cup[0.040]; Cup[0.040]; Lindsay[0.040]; Olympic[0.040]; left[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Fed Cup ==> ENTITY: \u001b[32mFed Cup\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.190:0.190[0]; log p(e|m)= -0.311:-0.311[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.046]; Tennis[0.045]; team[0.043]; team[0.043]; Cup[0.043]; competition[0.043]; final[0.042]; Spain[0.042]; Spain[0.042]; Spain[0.042]; Open[0.042]; Open[0.042]; Belgium[0.041]; champion[0.041]; draw[0.041]; Japan[0.041]; semifinals[0.041]; Teams[0.041]; Fed[0.040]; Czech[0.040]; Netherlands[0.040]; international[0.040]; Germany[0.040]; France[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.067:0.067[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncompetition[0.044]; team[0.044]; team[0.044]; Open[0.044]; Open[0.044]; matches[0.043]; hit[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; international[0.042]; finalists[0.042]; Germany[0.041]; Teams[0.041]; champion[0.040]; semifinals[0.040]; Cup[0.040]; Cup[0.040]; round[0.040]; round[0.040]; hosts[0.040]; draw[0.040]; Netherlands[0.040]; States[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.245:0.245[0]; local()= 0.044:0.044[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.045]; matches[0.044]; Spain[0.044]; squad[0.043]; won[0.043]; number[0.043]; match[0.043]; final[0.042]; month[0.042]; Austria[0.042]; singles[0.042]; debut[0.042]; win[0.042]; semifinals[0.042]; Cup[0.040]; Cup[0.040]; Cup[0.040]; draw[0.040]; court[0.040]; States[0.040]; captain[0.040]; blanked[0.040]; host[0.040]; opening[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = King\u001b[39m ==> ENTITY: \u001b[32mBillie Jean King\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.044]; Davenport[0.044]; match[0.043]; singles[0.043]; champion[0.043]; won[0.042]; matches[0.042]; Seles[0.042]; Seles[0.042]; Seles[0.042]; Lindsay[0.042]; team[0.042]; Arantxa[0.041]; win[0.041]; Mary[0.041]; Mary[0.041]; King[0.040]; King[0.040]; King[0.040]; King[0.040]; King[0.040]; Olympic[0.040]; Joe[0.040]; Joe[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Seles\u001b[39m ==> ENTITY: \u001b[32mMonica Seles\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.196:0.196[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSeles[0.047]; Arantxa[0.047]; match[0.045]; singles[0.044]; matches[0.044]; champion[0.044]; Open[0.044]; won[0.042]; win[0.042]; wrist[0.041]; Davenport[0.041]; injury[0.041]; Martinez[0.041]; duo[0.041]; Lindsay[0.040]; Vicario[0.040]; threesome[0.040]; Monica[0.040]; Fernandez[0.039]; Fernandez[0.039]; Fernandez[0.039]; number[0.039]; Olympic[0.039]; lead[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mUnited States\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUnited States\u001b[39m <---> \u001b[32mUnited States Fed Cup team\u001b[39m\t\nSCORES: global= 0.242:0.219[\u001b[31m0.023\u001b[39m]; local()= 0.067:0.193[\u001b[32m0.125\u001b[39m]; log p(e|m)= -0.074:-5.298[\u001b[31m5.225\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.045]; team[0.045]; matches[0.045]; Cup[0.043]; Cup[0.043]; Spain[0.043]; Spain[0.043]; Spain[0.043]; final[0.043]; final[0.043]; international[0.042]; Tennis[0.041]; Open[0.041]; Open[0.041]; semifinals[0.041]; Germany[0.041]; competition[0.041]; Richard[0.040]; draw[0.040]; Saturday[0.040]; France[0.040]; Netherlands[0.040]; champion[0.039]; hosts[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Salzburg ==> ENTITY: \u001b[32mSalzburg\u001b[39m\t\nSCORES: global= 0.248:0.248[0]; local()= 0.058:0.058[0]; log p(e|m)= -0.120:-0.120[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.047]; team[0.045]; Austria[0.044]; final[0.043]; final[0.043]; won[0.043]; draw[0.042]; semifinals[0.042]; semifinals[0.042]; win[0.042]; Cup[0.041]; Cup[0.041]; Cup[0.041]; King[0.041]; King[0.041]; King[0.041]; King[0.041]; April[0.040]; United[0.040]; victory[0.040]; July[0.040]; injury[0.040]; debut[0.040]; defeat[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = 1997 Fed Cup\u001b[39m ==> ENTITY: \u001b[32m1997 Fed Cup\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCup[0.044]; final[0.044]; Spain[0.043]; Spain[0.043]; Spain[0.043]; Belgium[0.042]; France[0.042]; Open[0.042]; Open[0.042]; round[0.042]; round[0.042]; semifinals[0.042]; matches[0.042]; Netherlands[0.041]; Teams[0.041]; Germany[0.041]; international[0.041]; Japan[0.041]; competition[0.041]; draw[0.041]; team[0.041]; team[0.041]; Saturday[0.040]; Tennis[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Seles\u001b[39m ==> ENTITY: \u001b[32mMonica Seles\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSeles[0.047]; Seles[0.047]; match[0.045]; singles[0.044]; round[0.044]; matches[0.044]; Open[0.044]; semifinals[0.042]; won[0.042]; win[0.042]; victory[0.042]; opening[0.041]; draw[0.041]; wrist[0.041]; injury[0.041]; debut[0.040]; Monica[0.040]; Monica[0.040]; Fernandez[0.039]; Fernandez[0.039]; number[0.039]; Salzburg[0.039]; Austria[0.039]; lead[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBelgium\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBelgium\u001b[39m <---> \u001b[32mBelgium Fed Cup team\u001b[39m\t\nSCORES: global= 0.247:0.224[\u001b[31m0.023\u001b[39m]; local()= 0.161:0.198[\u001b[32m0.037\u001b[39m]; log p(e|m)= -0.242:-4.343[\u001b[31m4.101\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.046]; matches[0.045]; Tennis[0.044]; Netherlands[0.043]; Germany[0.043]; Spain[0.042]; Spain[0.042]; Spain[0.042]; Spain[0.042]; Open[0.041]; Open[0.041]; finalists[0.041]; Cup[0.041]; Cup[0.041]; international[0.041]; draw[0.041]; team[0.041]; team[0.041]; final[0.041]; final[0.041]; Czech[0.040]; host[0.040]; champion[0.040]; competition[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = King\u001b[39m ==> ENTITY: \u001b[32mBillie Jean King\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.044]; court[0.043]; Davenport[0.043]; match[0.043]; singles[0.043]; champion[0.043]; won[0.042]; matches[0.042]; Seles[0.042]; Seles[0.042]; Seles[0.042]; Seles[0.042]; Lindsay[0.042]; team[0.042]; round[0.041]; win[0.041]; semifinals[0.041]; Mary[0.041]; Mary[0.041]; King[0.040]; King[0.040]; King[0.040]; King[0.040]; victory[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Spain ==> ENTITY: \u001b[32mSpain\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.149:0.149[0]; log p(e|m)= -0.189:-0.189[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSpain[0.050]; Spain[0.050]; matches[0.044]; Cup[0.043]; Cup[0.043]; team[0.043]; team[0.043]; Tennis[0.042]; international[0.042]; Netherlands[0.041]; Open[0.041]; Open[0.041]; women[0.041]; semifinals[0.041]; Germany[0.040]; draw[0.040]; France[0.040]; Belgium[0.040]; champion[0.040]; round[0.040]; round[0.040]; competition[0.039]; Republic[0.039]; Richard[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.067:0.067[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.048]; team[0.048]; matches[0.046]; Cup[0.043]; Cup[0.043]; Tennis[0.043]; competition[0.042]; final[0.042]; finalists[0.042]; Saturday[0.042]; travels[0.042]; semifinals[0.041]; round[0.041]; round[0.041]; March[0.040]; Richard[0.040]; champion[0.040]; Life[0.040]; visiting[0.039]; July[0.039]; draw[0.039]; based[0.039]; hosts[0.039]; States[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Davenport ==> ENTITY: \u001b[32mLindsay Davenport\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.175:0.175[0]; log p(e|m)= -2.882:-2.882[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLindsay[0.045]; matches[0.045]; match[0.045]; team[0.045]; Joe[0.044]; Joe[0.044]; Open[0.043]; Mary[0.043]; Mary[0.043]; champion[0.042]; Arantxa[0.042]; squad[0.042]; Seles[0.041]; Seles[0.041]; Monica[0.041]; win[0.040]; Cup[0.040]; Vicario[0.040]; Martinez[0.040]; Fernandez[0.039]; Fernandez[0.039]; Fernandez[0.039]; veteran[0.039]; injury[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mary Joe Fernandez\u001b[39m ==> ENTITY: \u001b[32mMary Joe Fernández\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nArantxa[0.046]; match[0.046]; champion[0.044]; Davenport[0.044]; win[0.043]; Open[0.043]; Martinez[0.043]; Fernandez[0.043]; Fernandez[0.043]; Vicario[0.042]; Monica[0.041]; Olympic[0.041]; team[0.041]; Seles[0.041]; Seles[0.041]; Lindsay[0.040]; number[0.040]; Joe[0.040]; Sanchez[0.040]; Spain[0.040]; coach[0.040]; Mary[0.040]; injuries[0.039]; lead[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Atlantic City ==> ENTITY: \u001b[32mAtlantic City, New Jersey\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.078:-0.078[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.048]; team[0.048]; blanked[0.043]; lost[0.042]; round[0.042]; round[0.042]; round[0.042]; won[0.042]; Cup[0.042]; Cup[0.042]; Saturday[0.041]; matches[0.041]; matches[0.041]; host[0.041]; semifinals[0.041]; semifinals[0.041]; final[0.040]; final[0.040]; September[0.040]; September[0.040]; opening[0.040]; Billie[0.040]; try[0.040]; debut[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nagoya ==> ENTITY: \u001b[32mNagoya\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.117:0.117[0]; log p(e|m)= -0.083:-0.083[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJapan[0.044]; Japan[0.044]; Cup[0.044]; Cup[0.044]; Cup[0.044]; team[0.044]; United[0.043]; squad[0.042]; matches[0.042]; match[0.042]; City[0.042]; world[0.042]; April[0.041]; September[0.041]; month[0.040]; number[0.040]; win[0.040]; injury[0.040]; draw[0.040]; semifinals[0.040]; court[0.040]; won[0.040]; opening[0.039]; captain[0.039]; \t\n\r [========================================>.....] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 2s106ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4361/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 969testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sarajevo\u001b[39m ==> ENTITY: \u001b[32mSarajevo\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.109:0.109[0]; log p(e|m)= -0.002:-0.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.049]; Sarajevo[0.047]; Sarajevo[0.047]; Games[0.046]; medal[0.045]; stadium[0.045]; athletes[0.043]; International[0.041]; leading[0.041]; competition[0.041]; record[0.041]; Amateur[0.041]; Athletics[0.040]; world[0.040]; war[0.040]; champion[0.040]; September[0.040]; long[0.040]; silver[0.040]; Federation[0.040]; Meeting[0.039]; Athletic[0.039]; winner[0.039]; Friday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Atlanta Games\u001b[39m ==> ENTITY: \u001b[32m1996 Summer Olympics\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.051]; medal[0.045]; athletes[0.044]; Final[0.043]; competition[0.043]; Amateur[0.042]; Athletics[0.042]; final[0.042]; Milan[0.042]; Milan[0.042]; champion[0.042]; International[0.041]; Meet[0.041]; record[0.041]; holder[0.040]; hold[0.040]; stadium[0.040]; jump[0.040]; jump[0.040]; jump[0.040]; Meeting[0.040]; Athletic[0.040]; world[0.040]; women[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Milan ==> ENTITY: \u001b[32mMilan\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.059:0.059[0]; log p(e|m)= -0.301:-0.301[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMilan[0.049]; women[0.044]; final[0.043]; stadium[0.043]; Grand[0.042]; men[0.042]; long[0.042]; Bosnia[0.042]; Edwards[0.042]; Edwards[0.042]; allowed[0.042]; Final[0.041]; grand[0.041]; Balkans[0.041]; hold[0.041]; Friday[0.041]; war[0.041]; war[0.041]; said[0.040]; days[0.040]; quoted[0.040]; make[0.040]; make[0.040]; winner[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sarajevo\u001b[39m ==> ENTITY: \u001b[32mSarajevo\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.138:0.138[0]; log p(e|m)= -0.002:-0.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.048]; stadium[0.047]; Bosnia[0.046]; Sarajevo[0.045]; Sarajevo[0.045]; Sarajevo[0.045]; Games[0.044]; medal[0.043]; Balkans[0.042]; athletes[0.042]; war[0.041]; war[0.041]; competition[0.040]; record[0.040]; Grand[0.040]; trip[0.039]; Amateur[0.039]; grand[0.039]; Milan[0.039]; Milan[0.039]; final[0.039]; humanitarian[0.039]; champion[0.039]; long[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jonathan Edwards ==> ENTITY: \u001b[32mJonathan Edwards (athlete)\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.171:0.171[0]; log p(e|m)= -1.415:-1.415[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.046]; Games[0.045]; medal[0.043]; final[0.043]; Edwards[0.043]; Edwards[0.043]; Athletics[0.043]; September[0.042]; athletes[0.042]; record[0.042]; Friday[0.042]; world[0.042]; holder[0.041]; Atlanta[0.041]; jump[0.041]; jump[0.041]; jump[0.041]; schedule[0.041]; long[0.040]; women[0.040]; Harrison[0.040]; Harrison[0.040]; champion[0.040]; Kenny[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sarajevo\u001b[39m ==> ENTITY: \u001b[32mSarajevo\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.002:-0.002[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBosnia[0.048]; Sarajevo[0.047]; Balkans[0.046]; stadium[0.044]; war[0.043]; war[0.043]; Games[0.042]; medal[0.042]; trip[0.041]; Milan[0.041]; Milan[0.041]; final[0.041]; humanitarian[0.041]; leading[0.041]; saying[0.041]; Final[0.040]; athletes[0.040]; called[0.040]; devastated[0.040]; long[0.040]; Grand[0.039]; days[0.039]; quoted[0.039]; going[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bosnia ==> ENTITY: \u001b[32mBosnia and Herzegovina\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.113:0.113[0]; log p(e|m)= -0.345:-0.345[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBalkans[0.046]; Sarajevo[0.046]; Sarajevo[0.046]; Sarajevo[0.046]; stadium[0.044]; war[0.043]; war[0.043]; athletes[0.042]; women[0.041]; humanitarian[0.041]; competition[0.041]; Olympic[0.041]; winner[0.041]; September[0.040]; men[0.040]; final[0.040]; schedule[0.040]; type[0.040]; Federation[0.040]; Grand[0.040]; champion[0.040]; Friday[0.040]; record[0.040]; devastated[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Harrison\u001b[39m ==> ENTITY: \u001b[32mKenny Harrison\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrecord[0.047]; Harrison[0.045]; Olympic[0.045]; jump[0.044]; jump[0.044]; jump[0.044]; Kenny[0.043]; world[0.042]; September[0.042]; triple[0.042]; triple[0.042]; holder[0.041]; Games[0.041]; Friday[0.041]; long[0.040]; Athletic[0.040]; stadium[0.040]; Athletics[0.040]; Edwards[0.039]; Edwards[0.039]; Edwards[0.039]; champion[0.039]; medal[0.039]; Jonathan[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Balkans\u001b[39m ==> ENTITY: \u001b[32mBalkans\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.108:0.108[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBosnia[0.051]; war[0.046]; war[0.046]; Sarajevo[0.044]; Sarajevo[0.044]; Sarajevo[0.044]; humanitarian[0.043]; women[0.042]; September[0.042]; leading[0.042]; men[0.041]; saying[0.041]; Friday[0.041]; devastated[0.040]; International[0.040]; said[0.040]; Federation[0.040]; world[0.040]; days[0.039]; refusing[0.039]; Grand[0.039]; Solidarity[0.039]; athletes[0.039]; trip[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = International Amateur Athletic Federation\u001b[39m ==> ENTITY: \u001b[32mInternational Association of Athletics Federations\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.161:0.161[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOlympic[0.048]; athletes[0.045]; Games[0.045]; Athletics[0.044]; medal[0.044]; record[0.044]; world[0.042]; holder[0.042]; Balkans[0.041]; competition[0.041]; September[0.041]; final[0.041]; champion[0.041]; days[0.040]; long[0.040]; jump[0.040]; jump[0.040]; jump[0.040]; Final[0.040]; schedule[0.040]; meeting[0.040]; Meet[0.040]; stadium[0.040]; Meeting[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Monte Carlo ==> ENTITY: \u001b[32mMonte Carlo\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.079:0.079[0]; log p(e|m)= -0.816:-0.816[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nchampion[0.044]; Friday[0.044]; Jonathan[0.044]; men[0.044]; Amateur[0.043]; usable[0.043]; world[0.043]; Edwards[0.043]; Edwards[0.043]; Edwards[0.043]; record[0.042]; meeting[0.041]; International[0.041]; winner[0.041]; hold[0.041]; competition[0.041]; Meeting[0.040]; Olympic[0.040]; triple[0.040]; triple[0.040]; Harrison[0.040]; Harrison[0.040]; September[0.040]; holder[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Milan ==> ENTITY: \u001b[32mMilan\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.058:0.058[0]; log p(e|m)= -0.301:-0.301[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMilan[0.049]; women[0.044]; September[0.043]; final[0.043]; stadium[0.043]; Grand[0.042]; men[0.042]; long[0.042]; Bosnia[0.042]; Edwards[0.041]; Edwards[0.041]; competition[0.041]; allowed[0.041]; Athletic[0.041]; Final[0.041]; grand[0.041]; Balkans[0.041]; hold[0.041]; Amateur[0.041]; Friday[0.040]; war[0.040]; war[0.040]; said[0.040]; days[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kenny Harrison\u001b[39m ==> ENTITY: \u001b[32mKenny Harrison\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrecord[0.047]; Harrison[0.045]; Olympic[0.045]; jump[0.044]; jump[0.044]; jump[0.044]; athletes[0.044]; world[0.042]; September[0.042]; triple[0.042]; triple[0.042]; holder[0.041]; Games[0.041]; Friday[0.041]; long[0.040]; Athletic[0.040]; stadium[0.040]; Athletics[0.040]; Balkans[0.040]; Edwards[0.039]; Edwards[0.039]; Edwards[0.039]; champion[0.039]; medal[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIAAF Grand Prix Final\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIAAF Grand Prix Final\u001b[39m <---> \u001b[32m1996 IAAF Grand Prix Final\u001b[39m\t\nSCORES: global= 0.246:0.232[\u001b[31m0.014\u001b[39m]; local()= 0.117:0.110[\u001b[31m0.007\u001b[39m]; log p(e|m)= -0.282:-1.839[\u001b[31m1.556\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMilan[0.047]; Milan[0.047]; stadium[0.045]; International[0.044]; Sarajevo[0.043]; Sarajevo[0.043]; Games[0.043]; final[0.043]; athletes[0.042]; grand[0.042]; Federation[0.042]; Athletic[0.041]; silver[0.041]; Friday[0.041]; medal[0.041]; winner[0.041]; jump[0.040]; jump[0.040]; Balkans[0.040]; Amateur[0.040]; hold[0.039]; long[0.039]; women[0.039]; schedule[0.039]; \t\n\r [=========================================>....] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 2s30ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4376/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1160testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brussels ==> ENTITY: \u001b[32mBrussels\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.157:0.157[0]; log p(e|m)= -0.025:-0.025[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.049]; Belgian[0.045]; Saturday[0.043]; town[0.042]; Liege[0.042]; Liege[0.042]; Liege[0.042]; eastern[0.042]; bus[0.042]; bus[0.042]; official[0.042]; Thursday[0.041]; Thursday[0.041]; arriving[0.041]; Friday[0.041]; got[0.041]; Severine[0.041]; trip[0.041]; home[0.041]; said[0.040]; said[0.040]; days[0.040]; far[0.040]; took[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Liege\u001b[39m ==> ENTITY: \u001b[32mLiège\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.171:0.171[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.048]; Belgium[0.048]; town[0.044]; Belgian[0.044]; Liege[0.044]; Liege[0.044]; eastern[0.044]; Brussels[0.042]; Severine[0.042]; Friday[0.042]; Thursday[0.041]; Thursday[0.041]; arriving[0.041]; Marc[0.040]; days[0.040]; Saturday[0.040]; took[0.040]; said[0.040]; said[0.040]; village[0.040]; trip[0.040]; reported[0.039]; home[0.039]; home[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nandrin\u001b[39m ==> ENTITY: \u001b[32mNandrin\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.096:0.096[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.045]; town[0.045]; Liege[0.044]; Liege[0.044]; Liege[0.044]; Saturday[0.044]; Friday[0.042]; village[0.042]; eastern[0.042]; home[0.042]; home[0.042]; Thursday[0.041]; Thursday[0.041]; far[0.041]; trip[0.041]; said[0.041]; said[0.041]; Severine[0.041]; abuse[0.040]; official[0.040]; girls[0.039]; Late[0.039]; reporting[0.039]; reported[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Liege\u001b[39m ==> ENTITY: \u001b[32mLiège\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.050]; Belgium[0.050]; town[0.046]; Belgian[0.045]; Liege[0.045]; Liege[0.045]; eastern[0.045]; Brussels[0.044]; Severine[0.044]; Friday[0.043]; Thursday[0.043]; Thursday[0.043]; arriving[0.043]; Marc[0.042]; days[0.042]; weeks[0.042]; Saturday[0.041]; took[0.041]; said[0.041]; said[0.041]; village[0.041]; trip[0.041]; reported[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Marc Dutroux\u001b[39m ==> ENTITY: \u001b[32mMarc Dutroux\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nscandal[0.045]; Belgium[0.044]; abuse[0.044]; case[0.044]; police[0.043]; police[0.043]; police[0.043]; murder[0.042]; sex[0.042]; reportedly[0.042]; suspected[0.041]; Liege[0.041]; Liege[0.041]; Liege[0.041]; theft[0.041]; rocked[0.041]; days[0.040]; Severine[0.040]; far[0.040]; said[0.040]; home[0.040]; home[0.040]; kidnap[0.040]; stolen[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgian ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.162:0.162[0]; log p(e|m)= -0.124:-0.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.055]; eastern[0.043]; said[0.043]; said[0.043]; Liege[0.043]; Liege[0.043]; Liege[0.043]; arriving[0.042]; Brussels[0.042]; home[0.042]; home[0.042]; town[0.041]; official[0.041]; Severine[0.041]; trip[0.041]; told[0.040]; far[0.040]; Reuters[0.040]; Friday[0.040]; Thursday[0.040]; Thursday[0.040]; teenagers[0.039]; reportedly[0.039]; reported[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Liege\u001b[39m ==> ENTITY: \u001b[32mLiège\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgium[0.050]; Belgium[0.050]; town[0.046]; Belgian[0.045]; Liege[0.045]; Liege[0.045]; eastern[0.045]; Brussels[0.044]; Severine[0.044]; Friday[0.043]; Thursday[0.043]; Thursday[0.043]; arriving[0.043]; Marc[0.042]; days[0.042]; weeks[0.042]; Saturday[0.041]; took[0.041]; said[0.041]; said[0.041]; village[0.041]; trip[0.041]; reported[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.138:0.138[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nreporting[0.045]; reported[0.045]; told[0.044]; eastern[0.044]; reportedly[0.043]; statement[0.043]; official[0.043]; days[0.042]; said[0.042]; said[0.042]; Thursday[0.042]; Thursday[0.042]; case[0.041]; Friday[0.041]; scandal[0.041]; past[0.041]; far[0.040]; link[0.040]; shopping[0.040]; shopping[0.040]; Late[0.040]; trip[0.040]; declined[0.040]; seen[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgium ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.125:0.125[0]; log p(e|m)= -0.242:-0.242[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsex[0.044]; Liege[0.044]; Liege[0.044]; Liege[0.044]; weeks[0.043]; said[0.043]; eastern[0.043]; Marc[0.042]; arriving[0.042]; scandal[0.042]; Friday[0.041]; Severine[0.041]; Thursday[0.041]; Thursday[0.041]; case[0.041]; official[0.041]; past[0.041]; link[0.041]; took[0.041]; seen[0.041]; declined[0.040]; got[0.040]; home[0.040]; home[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgium ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.242:-0.242[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelgian[0.046]; Liege[0.045]; Liege[0.045]; Liege[0.045]; eastern[0.043]; Brussels[0.042]; arriving[0.042]; days[0.042]; took[0.042]; ago[0.041]; Friday[0.041]; Saturday[0.041]; trip[0.041]; Thursday[0.041]; Thursday[0.041]; official[0.041]; got[0.041]; Severine[0.040]; missing[0.040]; missing[0.040]; home[0.040]; reportedly[0.040]; far[0.040]; Reuters[0.040]; \t\n\r [=========================================>....] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s982ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4386/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1046testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = United Nations\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.113:0.113[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.048]; Israel[0.048]; Palestinian[0.045]; Palestinian[0.045]; Israeli[0.045]; Council[0.043]; Council[0.043]; Council[0.043]; peace[0.043]; peace[0.043]; negotiations[0.042]; David[0.040]; Thursday[0.040]; Security[0.040]; Security[0.040]; President[0.040]; mission[0.040]; action[0.040]; charge[0.039]; Germany[0.039]; Responding[0.039]; said[0.038]; letter[0.038]; bulldozing[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.048]; Israel[0.048]; Israeli[0.045]; Israeli[0.045]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Council[0.044]; Council[0.044]; Council[0.044]; Security[0.043]; Security[0.043]; Jerusalem[0.043]; President[0.043]; action[0.043]; David[0.043]; views[0.043]; added[0.042]; session[0.042]; mission[0.042]; said[0.042]; said[0.042]; said[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.200:0.200[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.052]; Israel[0.052]; Israeli[0.047]; Israeli[0.047]; Palestinian[0.045]; Jerusalem[0.044]; Jerusalem[0.044]; Jerusalem[0.044]; international[0.041]; international[0.041]; settlers[0.041]; Holy[0.040]; Canada[0.040]; community[0.040]; status[0.039]; Tuesday[0.039]; represents[0.038]; said[0.038]; al[0.038]; million[0.038]; recently[0.038]; Organisation[0.038]; agreements[0.038]; law[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.226:0.226[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israel[0.050]; Israeli[0.046]; Israeli[0.046]; Israeli[0.046]; Israeli[0.046]; Palestinian[0.043]; Palestinian[0.043]; Jerusalem[0.042]; Jerusalem[0.042]; Jerusalem[0.042]; international[0.041]; international[0.041]; Holy[0.039]; settlers[0.039]; negotiations[0.039]; agreements[0.039]; Peleg[0.039]; David[0.038]; discuss[0.038]; Canada[0.038]; saying[0.038]; Tuesday[0.037]; peace[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Palestinian ==> ENTITY: \u001b[32mState of Palestine\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.264:0.264[0]; log p(e|m)= -2.847:-2.847[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.048]; Israel[0.048]; Israel[0.048]; Palestinian[0.046]; Palestinian[0.046]; Israeli[0.045]; Israeli[0.045]; Israeli[0.045]; Jerusalem[0.043]; Jerusalem[0.043]; international[0.040]; international[0.040]; Marwan[0.040]; peace[0.040]; peace[0.040]; agreements[0.039]; community[0.039]; resolutions[0.038]; status[0.038]; settlers[0.038]; Peleg[0.038]; Organisation[0.038]; Council[0.038]; negotiations[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Security Council\u001b[39m ==> ENTITY: \u001b[32mUnited Nations Security Council\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.137:0.137[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.046]; Israel[0.046]; Israel[0.046]; President[0.044]; Council[0.043]; Council[0.043]; Israeli[0.043]; Israeli[0.043]; Security[0.043]; action[0.042]; Palestinian[0.041]; Palestinian[0.041]; Palestinian[0.041]; negotiations[0.040]; Thursday[0.040]; discuss[0.040]; added[0.040]; United[0.040]; Responding[0.040]; Jilani[0.040]; Jerusalem[0.040]; said[0.039]; said[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Middle East\u001b[39m ==> ENTITY: \u001b[32mMiddle East\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.114:0.114[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.048]; Israel[0.048]; Palestinian[0.046]; Palestinian[0.046]; Palestinian[0.046]; Israeli[0.045]; Israeli[0.045]; taken[0.041]; day[0.040]; President[0.040]; action[0.040]; added[0.040]; United[0.040]; mission[0.040]; peace[0.040]; peace[0.040]; said[0.040]; said[0.040]; Security[0.040]; Security[0.040]; Marwan[0.040]; Germany[0.039]; concern[0.039]; concern[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mOld City\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mJerusalem\u001b[39m <---> \u001b[32mOld City (Jerusalem)\u001b[39m\t\nSCORES: global= 0.259:0.251[\u001b[31m0.009\u001b[39m]; local()= 0.200:0.225[\u001b[32m0.025\u001b[39m]; log p(e|m)= -2.180:-0.179[\u001b[32m2.001\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.047]; Israel[0.047]; Jerusalem[0.045]; Jerusalem[0.045]; Jerusalem[0.045]; Israeli[0.045]; Israeli[0.045]; Israeli[0.045]; Israeli[0.045]; Palestinian[0.044]; Palestinian[0.044]; building[0.041]; settlers[0.039]; based[0.039]; al[0.039]; demolished[0.039]; City[0.039]; city[0.038]; walls[0.038]; walls[0.038]; Holy[0.038]; Marwan[0.038]; build[0.037]; community[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jerusalem ==> ENTITY: \u001b[32mJerusalem\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.186:0.186[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJerusalem[0.047]; Israel[0.047]; Israel[0.047]; Israel[0.047]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; Israeli[0.043]; Holy[0.042]; Palestinian[0.042]; Palestinian[0.042]; Palestinian[0.042]; Peleg[0.041]; David[0.041]; settlers[0.041]; walls[0.039]; Old[0.039]; Marwan[0.039]; community[0.039]; Tuesday[0.038]; agreements[0.038]; revival[0.038]; peace[0.038]; status[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.075:0.075[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ninternational[0.048]; international[0.048]; units[0.044]; Tuesday[0.043]; City[0.043]; recently[0.043]; settlers[0.042]; based[0.042]; position[0.042]; practices[0.041]; expect[0.041]; million[0.041]; build[0.041]; plans[0.041]; recent[0.041]; Society[0.040]; said[0.040]; community[0.040]; saying[0.040]; building[0.040]; city[0.040]; Organisation[0.040]; crews[0.040]; represents[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Germany ==> ENTITY: \u001b[32mGermany\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.037:0.037[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.045]; Israel[0.045]; Israel[0.045]; added[0.044]; day[0.042]; Nations[0.042]; charge[0.042]; charge[0.042]; President[0.042]; United[0.042]; said[0.041]; said[0.041]; said[0.041]; David[0.041]; East[0.041]; effort[0.041]; maintenance[0.041]; Israeli[0.041]; Israeli[0.041]; action[0.040]; session[0.040]; Security[0.040]; Security[0.040]; status[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.159:0.159[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.054]; Israel[0.054]; Palestinian[0.047]; Palestinian[0.047]; Peleg[0.045]; David[0.045]; peace[0.042]; peace[0.042]; bulldozing[0.041]; Thursday[0.040]; bulldozers[0.040]; said[0.040]; Germany[0.040]; Council[0.039]; Council[0.039]; Council[0.039]; day[0.039]; United[0.038]; concerned[0.038]; expressed[0.038]; expressed[0.038]; urged[0.038]; President[0.038]; East[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.167:0.167[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.053]; Israeli[0.048]; Israeli[0.048]; Palestinian[0.046]; Palestinian[0.046]; David[0.043]; President[0.042]; Germany[0.042]; negotiations[0.041]; United[0.041]; Peleg[0.041]; Security[0.040]; Security[0.040]; peace[0.039]; peace[0.039]; Thursday[0.039]; said[0.039]; said[0.039]; Nations[0.039]; Council[0.039]; Council[0.039]; Council[0.039]; asked[0.039]; asked[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.216:0.216[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israel[0.050]; Israeli[0.046]; Israeli[0.046]; Israeli[0.046]; Israeli[0.046]; Palestinian[0.043]; Palestinian[0.043]; Jerusalem[0.043]; Jerusalem[0.043]; international[0.041]; international[0.041]; Holy[0.040]; settlers[0.040]; Germany[0.040]; negotiations[0.039]; agreements[0.039]; Peleg[0.039]; David[0.038]; discuss[0.038]; Tuesday[0.038]; Security[0.037]; peace[0.037]; President[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.210:0.210[0]; log p(e|m)= -0.213:-0.213[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.051]; Israel[0.051]; Israel[0.051]; Israeli[0.046]; Israeli[0.046]; Palestinian[0.044]; Palestinian[0.044]; Palestinian[0.044]; Jerusalem[0.041]; Jerusalem[0.041]; David[0.040]; President[0.040]; Holy[0.040]; Germany[0.040]; negotiations[0.039]; agreements[0.039]; United[0.039]; Peleg[0.039]; discuss[0.038]; Security[0.038]; Security[0.038]; peace[0.037]; peace[0.037]; Thursday[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Palestinian ==> ENTITY: \u001b[32mState of Palestine\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.223:0.223[0]; log p(e|m)= -2.847:-2.847[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israel[0.050]; Palestinian[0.048]; Palestinian[0.048]; Israeli[0.046]; Israeli[0.046]; Jerusalem[0.045]; Marwan[0.042]; peace[0.041]; peace[0.041]; Peleg[0.039]; Council[0.039]; Council[0.039]; Council[0.039]; negotiations[0.039]; destruction[0.039]; Germany[0.039]; Jilani[0.039]; said[0.039]; said[0.039]; said[0.039]; President[0.038]; David[0.038]; urged[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.210:0.210[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.052]; Israel[0.052]; Israeli[0.047]; Israeli[0.047]; Palestinian[0.044]; Palestinian[0.044]; Jerusalem[0.044]; Jerusalem[0.044]; Jerusalem[0.044]; international[0.041]; international[0.041]; settlers[0.040]; Holy[0.039]; Canada[0.039]; community[0.039]; status[0.039]; building[0.039]; Tuesday[0.038]; represents[0.038]; said[0.038]; said[0.038]; al[0.038]; million[0.038]; Marwan[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.220:0.220[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.051]; Israel[0.051]; Israeli[0.046]; Israeli[0.046]; Israeli[0.046]; Palestinian[0.044]; Palestinian[0.044]; Jerusalem[0.043]; Jerusalem[0.043]; Jerusalem[0.043]; international[0.040]; international[0.040]; settlers[0.040]; Holy[0.039]; Canada[0.039]; community[0.039]; status[0.038]; Tuesday[0.038]; represents[0.038]; added[0.038]; said[0.038]; said[0.038]; said[0.038]; al[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Security Council\u001b[39m ==> ENTITY: \u001b[32mUnited Nations Security Council\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.129:0.129[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.046]; Israel[0.046]; President[0.045]; Council[0.044]; Council[0.044]; Israeli[0.044]; Security[0.044]; action[0.042]; Palestinian[0.042]; Palestinian[0.042]; negotiations[0.041]; Thursday[0.041]; United[0.040]; Responding[0.040]; said[0.040]; said[0.040]; Peleg[0.040]; charge[0.040]; affaires[0.040]; process[0.040]; process[0.040]; peace[0.040]; peace[0.040]; mission[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = David Peleg\u001b[39m ==> ENTITY: \u001b[32mDavid Peleg\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.047]; Israel[0.047]; Israel[0.047]; Israel[0.047]; Israeli[0.044]; Israeli[0.044]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; Jerusalem[0.042]; Jerusalem[0.042]; affaires[0.041]; affaires[0.041]; Responding[0.039]; United[0.039]; Security[0.039]; Security[0.039]; President[0.038]; Council[0.038]; Council[0.038]; Council[0.038]; Organisation[0.038]; care[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Jerusalem ==> ENTITY: \u001b[32mJerusalem\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.132:-0.132[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJerusalem[0.048]; Israel[0.048]; Israeli[0.045]; Israeli[0.045]; Israeli[0.045]; Holy[0.044]; Palestinian[0.043]; settlers[0.042]; building[0.042]; demolished[0.042]; al[0.042]; walls[0.041]; walls[0.041]; Old[0.040]; city[0.040]; community[0.040]; million[0.040]; Tuesday[0.040]; called[0.039]; agreements[0.039]; revival[0.039]; status[0.039]; plans[0.039]; represents[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Palestinian ==> ENTITY: \u001b[32mState of Palestine\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.202:0.202[0]; log p(e|m)= -2.847:-2.847[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.051]; Israel[0.051]; Palestinian[0.049]; Israeli[0.048]; Israeli[0.048]; peace[0.042]; peace[0.042]; Peleg[0.040]; Council[0.040]; Council[0.040]; Council[0.040]; negotiations[0.040]; Germany[0.040]; said[0.040]; said[0.040]; President[0.039]; David[0.039]; urged[0.039]; views[0.039]; United[0.039]; concerned[0.039]; East[0.038]; Middle[0.038]; mission[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = U.N.\u001b[39m ==> ENTITY: \u001b[32mUnited Nations\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.131:0.131[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.045]; Israel[0.045]; international[0.045]; international[0.045]; resolutions[0.044]; saying[0.044]; Israeli[0.042]; Israeli[0.042]; Israeli[0.042]; Palestinian[0.041]; Palestinian[0.041]; Canada[0.041]; Jerusalem[0.041]; Jerusalem[0.041]; Jerusalem[0.041]; accordance[0.041]; agreements[0.040]; added[0.040]; recent[0.040]; Organisation[0.040]; said[0.040]; said[0.040]; letter[0.039]; position[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israeli ==> ENTITY: \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.230:0.230[0]; log p(e|m)= -0.348:-0.348[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.050]; Israel[0.050]; Israel[0.050]; Israel[0.050]; Israeli[0.045]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; Palestinian[0.043]; Jerusalem[0.042]; Jerusalem[0.042]; Peleg[0.041]; David[0.041]; international[0.039]; settlers[0.039]; peace[0.038]; peace[0.038]; bulldozing[0.037]; community[0.037]; represents[0.037]; added[0.037]; said[0.037]; said[0.037]; said[0.037]; \t\n\r [=========================================>....] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s879ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4410/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 950testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland national football team\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.149:0.149[0]; log p(e|m)= -2.781:-2.781[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.049]; England[0.049]; Soccer[0.044]; footballer[0.043]; Newcastle[0.043]; World[0.042]; Cup[0.042]; captain[0.042]; captain[0.042]; qualifier[0.042]; European[0.042]; Alan[0.041]; David[0.041]; Hoddle[0.041]; London[0.041]; coach[0.040]; championship[0.040]; captaincy[0.040]; June[0.040]; Shearer[0.040]; Shearer[0.040]; Shearer[0.040]; Adams[0.040]; Adams[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shearer\u001b[39m ==> ENTITY: \u001b[32mAlan Shearer\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlan[0.046]; Alan[0.046]; Shearer[0.045]; team[0.045]; team[0.045]; team[0.045]; Southampton[0.045]; squad[0.044]; Newcastle[0.044]; Hoddle[0.043]; skipper[0.041]; injury[0.041]; withdrew[0.041]; mates[0.040]; difficult[0.040]; Sheringham[0.040]; probably[0.040]; wanted[0.039]; mate[0.039]; really[0.039]; job[0.039]; Ferdinand[0.039]; Euro[0.039]; partner[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moldova ==> ENTITY: \u001b[32mMoldova national football team\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.154:0.154[0]; log p(e|m)= -2.419:-2.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.050]; England[0.050]; England[0.050]; Soccer[0.047]; qualifier[0.046]; championship[0.046]; European[0.044]; Cup[0.044]; footballer[0.043]; Sunday[0.043]; Southampton[0.043]; World[0.042]; David[0.042]; coach[0.042]; June[0.042]; Friday[0.041]; led[0.041]; captain[0.041]; captain[0.041]; Newcastle[0.041]; people[0.041]; Hoddle[0.041]; Hoddle[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = David Platt ==> ENTITY: \u001b[32mDavid Platt (footballer)\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.159:0.159[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.050]; England[0.050]; England[0.050]; Southampton[0.047]; Alan[0.045]; Alan[0.045]; Alan[0.045]; Tony[0.044]; Blackburn[0.044]; Platt[0.043]; Newcastle[0.043]; Hoddle[0.042]; Hoddle[0.042]; footballer[0.042]; takes[0.042]; takes[0.042]; joined[0.041]; Soccer[0.041]; European[0.041]; June[0.041]; captain[0.041]; captain[0.041]; opening[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Newcastle ==> ENTITY: \u001b[32mNewcastle United F.C.\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.147:0.147[0]; log p(e|m)= -2.489:-2.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.045]; England[0.045]; England[0.045]; Blackburn[0.044]; Southampton[0.044]; Soccer[0.042]; Alan[0.042]; Alan[0.042]; David[0.042]; captain[0.042]; captain[0.042]; Cup[0.042]; London[0.042]; footballer[0.041]; Glenn[0.041]; Hoddle[0.040]; Hoddle[0.040]; joined[0.040]; Tony[0.040]; coach[0.040]; Shearer[0.039]; Shearer[0.039]; Shearer[0.039]; captaincy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alan\u001b[39m ==> ENTITY: \u001b[32mAlan Shearer\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlan[0.046]; David[0.046]; Shearer[0.045]; Shearer[0.045]; team[0.045]; team[0.045]; Southampton[0.045]; England[0.045]; squad[0.045]; Newcastle[0.045]; Hoddle[0.044]; Hoddle[0.044]; Blackburn[0.043]; Glenn[0.043]; saw[0.042]; captain[0.042]; qualifier[0.042]; joined[0.042]; skipper[0.042]; Tony[0.041]; injury[0.041]; Cup[0.041]; withdrew[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Newcastle ==> ENTITY: \u001b[32mNewcastle United F.C.\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.147:0.147[0]; log p(e|m)= -2.489:-2.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nteam[0.047]; team[0.047]; team[0.047]; squad[0.046]; Alan[0.045]; injury[0.043]; Hoddle[0.043]; Shearer[0.042]; Shearer[0.042]; withdrew[0.041]; Friday[0.041]; replaced[0.040]; plays[0.040]; people[0.040]; come[0.040]; Sheringham[0.040]; knows[0.040]; knows[0.040]; days[0.040]; job[0.040]; wanted[0.040]; spoke[0.039]; probably[0.039]; Ferdinand[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Glenn Hoddle\u001b[39m ==> ENTITY: \u001b[32mGlenn Hoddle\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.170:0.170[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.046]; England[0.046]; England[0.046]; Southampton[0.045]; Hoddle[0.043]; Blackburn[0.042]; Alan[0.042]; Alan[0.042]; Alan[0.042]; David[0.042]; Newcastle[0.042]; joined[0.041]; team[0.041]; European[0.041]; footballer[0.041]; Platt[0.040]; Platt[0.040]; saw[0.040]; Tony[0.040]; named[0.040]; captain[0.039]; captain[0.039]; Soccer[0.039]; coach[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Platt ==> ENTITY: \u001b[32mDavid Platt (footballer)\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.159:0.159[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.048]; England[0.048]; England[0.048]; Southampton[0.045]; Alan[0.043]; Alan[0.043]; Alan[0.043]; David[0.043]; Tony[0.042]; Blackburn[0.042]; Platt[0.041]; Newcastle[0.041]; Hoddle[0.040]; Hoddle[0.040]; footballer[0.040]; takes[0.040]; takes[0.040]; joined[0.040]; Soccer[0.039]; European[0.039]; June[0.039]; captain[0.039]; captain[0.039]; opening[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tony Adams\u001b[39m ==> ENTITY: \u001b[32mTony Adams (footballer)\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAdams[0.060]; Glenn[0.051]; David[0.049]; Alan[0.045]; Alan[0.045]; Platt[0.043]; Platt[0.043]; England[0.042]; England[0.042]; England[0.042]; footballer[0.040]; Southampton[0.040]; Blackburn[0.039]; Hoddle[0.039]; Hoddle[0.039]; Shearer[0.038]; Shearer[0.038]; Shearer[0.038]; Newcastle[0.038]; Sunday[0.038]; captain[0.038]; captain[0.038]; skipper[0.038]; joined[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Euro 96\u001b[39m ==> ENTITY: \u001b[32mUEFA Euro 1996\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.127:0.127[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsquad[0.047]; team[0.046]; team[0.046]; team[0.046]; injury[0.044]; withdrew[0.043]; Hoddle[0.042]; Newcastle[0.042]; Ferdinand[0.041]; Shearer[0.041]; Shearer[0.041]; probably[0.041]; conclusion[0.041]; plays[0.041]; Friday[0.040]; really[0.040]; mates[0.040]; replaced[0.040]; days[0.040]; partner[0.040]; post[0.039]; come[0.039]; situation[0.039]; said[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.084:0.084[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.046]; England[0.046]; England[0.046]; Southampton[0.044]; Blackburn[0.043]; Soccer[0.043]; Cup[0.043]; qualifier[0.042]; footballer[0.042]; Newcastle[0.042]; David[0.041]; Alan[0.041]; Sunday[0.041]; Adams[0.041]; Adams[0.041]; Friday[0.040]; European[0.040]; championship[0.040]; opening[0.040]; saw[0.040]; Hoddle[0.040]; Tony[0.040]; named[0.040]; June[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hoddle\u001b[39m ==> ENTITY: \u001b[32mGlenn Hoddle\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.169:0.169[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; Southampton[0.045]; Hoddle[0.044]; Blackburn[0.043]; Alan[0.043]; Alan[0.043]; David[0.042]; Newcastle[0.042]; Newcastle[0.042]; squad[0.042]; team[0.041]; team[0.041]; team[0.041]; European[0.041]; Platt[0.041]; Platt[0.041]; saw[0.041]; withdrew[0.041]; Tony[0.040]; replaced[0.040]; Euro[0.040]; captain[0.040]; Glenn[0.040]; probably[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEuropean\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUEFA\u001b[39m <---> \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.244:0.242[\u001b[31m0.002\u001b[39m]; local()= 0.112:0.082[\u001b[31m0.030\u001b[39m]; log p(e|m)= -3.576:-0.256[\u001b[32m3.319\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.047]; Cup[0.046]; England[0.046]; England[0.046]; England[0.046]; qualifier[0.045]; championship[0.044]; world[0.044]; footballer[0.044]; World[0.043]; Blackburn[0.043]; people[0.043]; million[0.043]; million[0.043]; miss[0.043]; Moldova[0.043]; Southampton[0.042]; Newcastle[0.042]; opening[0.042]; make[0.042]; Sunday[0.042]; said[0.041]; said[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Les Ferdinand\u001b[39m ==> ENTITY: \u001b[32mLes Ferdinand\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.125:0.125[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlan[0.046]; team[0.044]; team[0.044]; team[0.044]; Hoddle[0.044]; Newcastle[0.043]; squad[0.043]; Shearer[0.042]; Shearer[0.042]; mates[0.042]; injury[0.041]; mate[0.041]; partner[0.041]; Sheringham[0.041]; probably[0.041]; withdrew[0.040]; knows[0.040]; knows[0.040]; plays[0.040]; Teddy[0.040]; replaced[0.040]; really[0.040]; wanted[0.040]; Euro[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Blackburn ==> ENTITY: \u001b[32mBlackburn Rovers F.C.\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.160:0.160[0]; log p(e|m)= -2.733:-2.733[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.045]; England[0.045]; Southampton[0.045]; David[0.044]; Newcastle[0.044]; Alan[0.043]; Alan[0.043]; Alan[0.043]; team[0.042]; team[0.042]; footballer[0.042]; Cup[0.041]; qualifier[0.041]; Tony[0.041]; joined[0.040]; London[0.040]; million[0.040]; million[0.040]; plays[0.040]; Shearer[0.040]; Shearer[0.040]; Hoddle[0.040]; Hoddle[0.040]; June[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Teddy Sheringham\u001b[39m ==> ENTITY: \u001b[32mTeddy Sheringham\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.144:0.144[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlan[0.047]; Alan[0.047]; team[0.044]; team[0.044]; team[0.044]; withdrew[0.043]; Hoddle[0.043]; squad[0.042]; Shearer[0.042]; Shearer[0.042]; injury[0.042]; partner[0.042]; Ferdinand[0.042]; wanted[0.042]; Newcastle[0.041]; probably[0.041]; mates[0.040]; sure[0.040]; sure[0.040]; mate[0.040]; really[0.040]; days[0.039]; Euro[0.039]; post[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shearer\u001b[39m ==> ENTITY: \u001b[32mAlan Shearer\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.175:0.175[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlan[0.044]; Alan[0.044]; Alan[0.044]; David[0.044]; Shearer[0.043]; Shearer[0.043]; Southampton[0.043]; England[0.043]; England[0.043]; England[0.043]; footballer[0.043]; Newcastle[0.042]; Hoddle[0.041]; Hoddle[0.041]; Blackburn[0.041]; Soccer[0.041]; Glenn[0.040]; saw[0.040]; captain[0.040]; captain[0.040]; qualifier[0.040]; joined[0.040]; skipper[0.039]; Tony[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Adams\u001b[39m ==> ENTITY: \u001b[32mTony Adams (footballer)\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTony[0.059]; Adams[0.058]; Glenn[0.049]; David[0.048]; Alan[0.044]; Alan[0.044]; Alan[0.044]; Platt[0.042]; Platt[0.042]; England[0.040]; England[0.040]; England[0.040]; footballer[0.039]; Southampton[0.039]; Blackburn[0.038]; Hoddle[0.037]; Hoddle[0.037]; Shearer[0.037]; Shearer[0.037]; Shearer[0.037]; Newcastle[0.037]; Sunday[0.037]; captain[0.037]; captain[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shearer\u001b[39m ==> ENTITY: \u001b[32mAlan Shearer\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlan[0.045]; David[0.045]; Shearer[0.044]; Shearer[0.044]; England[0.044]; England[0.044]; England[0.044]; footballer[0.044]; Newcastle[0.043]; Hoddle[0.042]; Soccer[0.041]; Glenn[0.041]; captain[0.041]; captain[0.041]; qualifier[0.040]; joined[0.040]; Tony[0.040]; Cup[0.040]; million[0.040]; million[0.040]; championship[0.040]; pounds[0.040]; coach[0.039]; miss[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alan Shearer\u001b[39m ==> ENTITY: \u001b[32mAlan Shearer\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDavid[0.045]; Shearer[0.044]; Shearer[0.044]; Southampton[0.044]; England[0.043]; England[0.043]; England[0.043]; footballer[0.043]; Newcastle[0.043]; Hoddle[0.042]; Blackburn[0.042]; Soccer[0.041]; Glenn[0.041]; saw[0.041]; captain[0.041]; captain[0.041]; qualifier[0.040]; joined[0.040]; skipper[0.040]; Tony[0.040]; Cup[0.040]; million[0.040]; million[0.040]; championship[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alan\u001b[39m ==> ENTITY: \u001b[32mAlan Shearer\u001b[39m\t\nSCORES: global= 0.308:0.308[0]; local()= 0.176:0.176[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlan[0.044]; David[0.044]; Shearer[0.043]; Shearer[0.043]; Shearer[0.043]; team[0.043]; team[0.043]; team[0.043]; Southampton[0.043]; England[0.043]; squad[0.043]; Newcastle[0.043]; Hoddle[0.042]; Hoddle[0.042]; Blackburn[0.041]; Glenn[0.041]; saw[0.040]; captain[0.040]; qualifier[0.040]; skipper[0.040]; injury[0.039]; Cup[0.039]; withdrew[0.039]; mates[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Southampton ==> ENTITY: \u001b[32mSouthampton F.C.\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.156:0.156[0]; log p(e|m)= -1.149:-1.149[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNewcastle[0.048]; England[0.048]; England[0.048]; Blackburn[0.047]; Hoddle[0.046]; Hoddle[0.046]; team[0.043]; team[0.043]; Shearer[0.043]; Shearer[0.043]; Shearer[0.043]; footballer[0.043]; London[0.043]; David[0.043]; Alan[0.042]; Alan[0.042]; Alan[0.042]; Cup[0.042]; Platt[0.041]; Platt[0.041]; opening[0.041]; saw[0.041]; joined[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland national football team\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.165:0.165[0]; log p(e|m)= -2.781:-2.781[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.049]; England[0.049]; Southampton[0.044]; Blackburn[0.043]; footballer[0.043]; Soccer[0.043]; Newcastle[0.042]; Cup[0.042]; captain[0.041]; captain[0.041]; qualifier[0.041]; Alan[0.041]; Alan[0.041]; Alan[0.041]; David[0.041]; Hoddle[0.041]; Hoddle[0.041]; European[0.040]; London[0.040]; Shearer[0.040]; Shearer[0.040]; Shearer[0.040]; skipper[0.040]; coach[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Shearer\u001b[39m ==> ENTITY: \u001b[32mAlan Shearer\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlan[0.046]; Shearer[0.045]; team[0.045]; team[0.045]; team[0.045]; squad[0.045]; Newcastle[0.045]; Hoddle[0.044]; injury[0.041]; withdrew[0.041]; mates[0.041]; difficult[0.041]; Sheringham[0.041]; probably[0.040]; wanted[0.040]; mate[0.040]; really[0.040]; job[0.040]; Ferdinand[0.039]; Euro[0.039]; partner[0.039]; situation[0.039]; said[0.039]; Teddy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland national football team\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.147:0.147[0]; log p(e|m)= -2.781:-2.781[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.053]; England[0.053]; Southampton[0.048]; footballer[0.047]; Soccer[0.046]; Newcastle[0.046]; Blackburn[0.046]; Cup[0.046]; qualifier[0.045]; captain[0.045]; captain[0.045]; named[0.045]; Alan[0.044]; David[0.044]; Hoddle[0.044]; European[0.044]; London[0.044]; skipper[0.043]; coach[0.043]; championship[0.043]; captaincy[0.043]; June[0.043]; \t\n\r [=========================================>....] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s761ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4437/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1002testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bucharest ==> ENTITY: \u001b[32mBucharest\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.098:0.098[0]; log p(e|m)= -0.018:-0.018[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRomania[0.046]; Romania[0.046]; major[0.043]; market[0.043]; past[0.042]; rate[0.042]; rate[0.042]; National[0.042]; week[0.042]; week[0.042]; week[0.042]; percent[0.041]; percent[0.041]; percent[0.041]; rises[0.041]; Thursday[0.041]; days[0.041]; exceeded[0.041]; lifted[0.041]; rates[0.040]; rates[0.040]; Bank[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Romania ==> ENTITY: \u001b[32mRomania\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.085:0.085[0]; log p(e|m)= -0.212:-0.212[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRomania[0.051]; Bucharest[0.046]; week[0.044]; week[0.044]; week[0.044]; entered[0.043]; major[0.042]; lei[0.042]; lei[0.042]; Thursday[0.041]; meet[0.041]; National[0.041]; days[0.041]; Bank[0.041]; minimum[0.041]; bids[0.041]; bids[0.041]; said[0.040]; said[0.040]; lifted[0.040]; keen[0.039]; past[0.039]; rate[0.039]; rate[0.039]; \t\n\r [=========================================>....] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s751ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4439/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1117testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.050]; Russian[0.047]; Napoleon[0.046]; Napoleon[0.046]; ceded[0.046]; Borodino[0.044]; War[0.044]; Prince[0.044]; defeated[0.043]; defeated[0.043]; Emperor[0.043]; Emperor[0.043]; ending[0.043]; Roman[0.043]; war[0.043]; west[0.042]; December[0.042]; Treaty[0.042]; army[0.042]; army[0.042]; Red[0.042]; miles[0.041]; entered[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Napoleon ==> ENTITY: \u001b[32mNapoleon\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.149:0.149[0]; log p(e|m)= -0.088:-0.088[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNapoleon[0.049]; France[0.044]; France[0.044]; war[0.043]; Borodino[0.043]; War[0.043]; Henry[0.042]; French[0.042]; Italy[0.042]; ceded[0.042]; Roman[0.041]; Charles[0.041]; Charles[0.041]; Russian[0.041]; army[0.040]; army[0.040]; Emperor[0.040]; Emperor[0.040]; freedom[0.040]; Treaty[0.040]; independence[0.040]; Moscow[0.040]; Moscow[0.040]; miles[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Cape Town\u001b[39m ==> ENTITY: \u001b[32mCape Town\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.113:0.113[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.047]; South[0.047]; won[0.044]; won[0.044]; won[0.044]; Taylor[0.043]; Bishop[0.042]; Anglicans[0.042]; Archbishop[0.042]; King[0.042]; century[0.041]; day[0.041]; Africa[0.041]; Africa[0.041]; announced[0.041]; winning[0.041]; championship[0.041]; London[0.040]; events[0.040]; began[0.040]; close[0.040]; John[0.040]; world[0.039]; world[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.194:0.194[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.048]; France[0.048]; English[0.045]; Napoleon[0.045]; Napoleon[0.045]; French[0.045]; Spanish[0.042]; Europe[0.042]; Italy[0.041]; ceded[0.041]; Orleans[0.041]; monarchs[0.040]; Charles[0.040]; Charles[0.040]; England[0.040]; Elizabeth[0.040]; independence[0.040]; history[0.040]; Anne[0.040]; Alsace[0.039]; born[0.039]; War[0.039]; Breisach[0.039]; Strasbourg[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAustrians\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustria\u001b[39m <---> \u001b[32mAustrian Empire\u001b[39m\t\nSCORES: global= 0.263:0.252[\u001b[31m0.012\u001b[39m]; local()= 0.096:0.116[\u001b[32m0.020\u001b[39m]; log p(e|m)= -1.088:-1.625[\u001b[31m0.537\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNapoleon[0.046]; born[0.044]; born[0.044]; born[0.044]; Henry[0.043]; actor[0.043]; British[0.043]; government[0.042]; James[0.042]; Italian[0.042]; independence[0.042]; gave[0.041]; granted[0.041]; Emperor[0.041]; war[0.041]; Lawrence[0.041]; seized[0.041]; freedom[0.040]; imposed[0.040]; powers[0.040]; films[0.040]; Lords[0.040]; unions[0.040]; liberation[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mChina\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mChina\u001b[39m <---> \u001b[32mQing Dynasty\u001b[39m\t\nSCORES: global= 0.259:0.224[\u001b[31m0.035\u001b[39m]; local()= 0.090:0.080[\u001b[31m0.010\u001b[39m]; log p(e|m)= -0.157:-4.828[\u001b[31m4.672\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.047]; Brazil[0.047]; Portugal[0.046]; Emperor[0.045]; war[0.044]; Peking[0.043]; appearances[0.043]; British[0.042]; government[0.042]; week[0.041]; December[0.041]; officially[0.041]; signing[0.040]; Napoleon[0.040]; independence[0.040]; freedom[0.040]; ended[0.040]; imposed[0.040]; entered[0.040]; stage[0.039]; stage[0.039]; gave[0.039]; James[0.039]; films[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEngland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEngland\u001b[39m <---> \u001b[32mKingdom of England\u001b[39m\t\nSCORES: global= 0.257:0.232[\u001b[31m0.025\u001b[39m]; local()= 0.115:0.162[\u001b[32m0.047\u001b[39m]; log p(e|m)= -0.190:-3.912[\u001b[31m3.722\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmonarchs[0.045]; Queen[0.045]; second[0.043]; France[0.042]; France[0.042]; Charles[0.042]; Charles[0.042]; London[0.042]; English[0.042]; history[0.042]; Italy[0.042]; Europe[0.042]; queen[0.042]; Prince[0.042]; major[0.041]; ceded[0.041]; September[0.041]; September[0.041]; Henry[0.041]; Boleyn[0.040]; got[0.040]; Elizabeth[0.040]; born[0.040]; northern[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.189:0.189[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.048]; Napoleon[0.045]; Napoleon[0.045]; England[0.045]; Italy[0.044]; Europe[0.043]; French[0.042]; French[0.042]; monarchs[0.042]; Portugal[0.042]; second[0.042]; week[0.041]; Brazil[0.041]; Brazil[0.041]; ceded[0.041]; Roman[0.040]; Alsace[0.040]; northern[0.040]; Spanish[0.040]; Treaty[0.040]; War[0.040]; independence[0.040]; Queen[0.039]; Breisach[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Baudouin I\u001b[39m ==> ENTITY: \u001b[32mBaudouin of Belgium\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.135:0.135[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nthrone[0.046]; Hermann[0.044]; King[0.043]; Belgian[0.043]; born[0.043]; born[0.043]; born[0.043]; born[0.043]; Cardinal[0.042]; father[0.042]; Alfred[0.042]; British[0.042]; Archbishop[0.041]; abdication[0.041]; Cold[0.041]; head[0.041]; announced[0.041]; paid[0.040]; Van[0.040]; Grand[0.040]; gave[0.040]; day[0.040]; Goering[0.040]; World[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leopold III ==> ENTITY: \u001b[32mLeopold III of Belgium\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.423:-0.423[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nthrone[0.047]; King[0.043]; Baudouin[0.043]; British[0.043]; Hermann[0.043]; abdication[0.042]; German[0.042]; Belgian[0.042]; crown[0.042]; gave[0.042]; day[0.041]; retirement[0.041]; War[0.041]; Anne[0.041]; Cardinal[0.041]; Alfred[0.041]; born[0.041]; born[0.041]; born[0.041]; Lawrence[0.041]; Grand[0.041]; father[0.041]; James[0.041]; appearances[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Africa ==> ENTITY: \u001b[32mSouth Africa\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.113:0.113[0]; log p(e|m)= -0.263:-0.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.045]; Africa[0.044]; day[0.043]; winning[0.043]; won[0.043]; won[0.043]; won[0.043]; championship[0.042]; Europe[0.042]; Europe[0.042]; leading[0.042]; Taylor[0.042]; Russia[0.042]; World[0.042]; signed[0.041]; world[0.041]; world[0.041]; Cape[0.041]; Belarus[0.041]; announced[0.040]; States[0.040]; Uzbekistan[0.040]; century[0.040]; Kazakhstan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kehl\u001b[39m ==> ENTITY: \u001b[32mKehl\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.184:0.184[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.046]; northern[0.045]; Breisach[0.045]; France[0.045]; France[0.045]; Strasbourg[0.044]; War[0.044]; Charles[0.044]; Charles[0.044]; Napoleon[0.044]; Napoleon[0.044]; troops[0.044]; west[0.043]; besieging[0.043]; Alsace[0.043]; Freiburg[0.043]; French[0.043]; French[0.043]; army[0.042]; army[0.042]; week[0.042]; Treaty[0.042]; Baden[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kazakhstan ==> ENTITY: \u001b[32mKazakhstan\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.146:0.146[0]; log p(e|m)= -0.414:-0.414[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwon[0.046]; won[0.046]; Russia[0.045]; Uzbekistan[0.045]; Belarus[0.044]; Tajikistan[0.044]; Russian[0.042]; Armenia[0.042]; Soviet[0.042]; Europe[0.042]; Europe[0.042]; Stars[0.042]; half[0.041]; republics[0.041]; Alan[0.041]; American[0.040]; time[0.040]; John[0.040]; British[0.039]; acclaim[0.039]; South[0.039]; South[0.039]; flag[0.039]; Iraq[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Armenia ==> ENTITY: \u001b[32mArmenia\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.431:-0.431[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.045]; Kazakhstan[0.044]; Uzbekistan[0.044]; republics[0.043]; Soviet[0.043]; Russia[0.043]; Russian[0.043]; city[0.042]; flag[0.042]; Tajikistan[0.042]; Iraq[0.042]; won[0.042]; won[0.042]; Europe[0.041]; Europe[0.041]; Anglicans[0.041]; food[0.041]; nearly[0.041]; restaurants[0.040]; forces[0.040]; American[0.040]; South[0.040]; help[0.039]; signed[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Turin ==> ENTITY: \u001b[32mTurin\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.049:-0.049[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItaly[0.048]; northern[0.044]; besieging[0.043]; September[0.043]; September[0.043]; Napoleon[0.042]; Napoleon[0.042]; France[0.042]; France[0.042]; French[0.042]; French[0.042]; army[0.041]; army[0.041]; Breisach[0.041]; troops[0.041]; England[0.041]; Orleans[0.041]; west[0.041]; Borodino[0.040]; Alsace[0.040]; Europe[0.040]; battle[0.040]; miles[0.040]; born[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.050]; Russian[0.047]; Napoleon[0.046]; Napoleon[0.046]; ceded[0.046]; Borodino[0.044]; War[0.044]; Prince[0.044]; defeated[0.043]; defeated[0.043]; Emperor[0.043]; Emperor[0.043]; ending[0.043]; Roman[0.043]; war[0.043]; west[0.042]; December[0.042]; Treaty[0.042]; army[0.042]; army[0.042]; Red[0.042]; miles[0.041]; entered[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBrazil\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBrazil\u001b[39m <---> \u001b[32mEmpire of Brazil\u001b[39m\t\nSCORES: global= 0.261:0.229[\u001b[31m0.033\u001b[39m]; local()= 0.135:0.144[\u001b[32m0.009\u001b[39m]; log p(e|m)= -0.213:-4.605[\u001b[31m4.392\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.048]; Portugal[0.044]; Italy[0.044]; independence[0.043]; Pedro[0.043]; France[0.042]; France[0.042]; government[0.042]; Napoleon[0.042]; Napoleon[0.042]; Spanish[0.042]; northern[0.042]; west[0.041]; ceded[0.041]; proclaimed[0.041]; ending[0.041]; granted[0.041]; Treaty[0.040]; week[0.040]; freedom[0.040]; China[0.040]; miles[0.040]; Red[0.039]; born[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.075:0.075[0]; log p(e|m)= -0.074:-0.074[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwon[0.046]; won[0.046]; won[0.046]; championship[0.043]; Russia[0.043]; Belarus[0.043]; winning[0.043]; World[0.042]; Grand[0.042]; Europe[0.041]; Europe[0.041]; day[0.041]; republics[0.041]; events[0.041]; Uzbekistan[0.041]; Kazakhstan[0.041]; world[0.040]; world[0.040]; British[0.040]; South[0.040]; South[0.040]; John[0.040]; Soviet[0.039]; began[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kuwaiti ==> ENTITY: \u001b[32mKuwait\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.041:-0.041[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIraq[0.048]; food[0.044]; Uzbekistan[0.043]; Gulf[0.043]; Saudi[0.043]; forces[0.043]; won[0.042]; won[0.042]; won[0.042]; championship[0.042]; Kazakhstan[0.042]; British[0.041]; began[0.041]; States[0.041]; winning[0.041]; announced[0.041]; common[0.040]; day[0.040]; Europe[0.040]; Europe[0.040]; Belarus[0.040]; face[0.040]; World[0.040]; events[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Peking Protocol\u001b[39m ==> ENTITY: \u001b[32mBoxer Protocol\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.112:0.112[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.048]; British[0.047]; paid[0.046]; granted[0.045]; freedom[0.045]; indemnity[0.045]; war[0.044]; imposed[0.044]; Cardinal[0.044]; government[0.044]; December[0.043]; powers[0.043]; Alfred[0.043]; Days[0.043]; Portugal[0.043]; Emperor[0.042]; Cold[0.042]; foreigners[0.042]; Italian[0.042]; ended[0.042]; Brazil[0.042]; signing[0.042]; director[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Elia Kazan\u001b[39m ==> ENTITY: \u001b[32mElia Kazan\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.130:0.130[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nactor[0.047]; James[0.045]; Allen[0.045]; films[0.045]; Lawrence[0.044]; Anthony[0.043]; Alex[0.043]; stage[0.042]; stage[0.042]; director[0.042]; Anne[0.041]; Cardinal[0.041]; House[0.041]; King[0.041]; appearances[0.040]; Alfred[0.040]; Days[0.040]; Red[0.040]; screen[0.040]; screen[0.040]; Italian[0.040]; Cold[0.039]; British[0.039]; indemnity[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = War of Spanish Succession\u001b[39m ==> ENTITY: \u001b[32mWar of the Spanish Succession\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.152:0.152[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmonarchs[0.046]; Napoleon[0.045]; Napoleon[0.045]; France[0.043]; France[0.043]; Queen[0.043]; Italy[0.042]; Roman[0.042]; ceded[0.042]; Pedro[0.042]; French[0.042]; French[0.042]; England[0.041]; Freiburg[0.041]; Treaty[0.041]; granted[0.041]; Breisach[0.040]; Strasbourg[0.040]; battle[0.040]; Europe[0.040]; queen[0.040]; Prince[0.040]; army[0.040]; army[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = World War Two\u001b[39m ==> ENTITY: \u001b[32mWorld War II\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.081:0.081[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.048]; States[0.044]; German[0.043]; campaign[0.042]; Italian[0.042]; Scottish[0.042]; world[0.042]; world[0.042]; London[0.042]; Days[0.041]; Sir[0.041]; James[0.041]; South[0.041]; South[0.041]; airforce[0.041]; bombing[0.041]; people[0.041]; began[0.041]; King[0.041]; day[0.041]; Lawrence[0.041]; gave[0.040]; secure[0.040]; United[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mGerman\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mGermany\u001b[39m <---> \u001b[32mNazi Germany\u001b[39m\t\nSCORES: global= 0.250:0.221[\u001b[31m0.029\u001b[39m]; local()= 0.128:0.087[\u001b[31m0.041\u001b[39m]; log p(e|m)= -1.047:-2.551[\u001b[31m1.504\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWar[0.045]; throne[0.044]; King[0.044]; British[0.043]; born[0.043]; born[0.043]; born[0.043]; Italian[0.042]; films[0.042]; actor[0.041]; States[0.041]; world[0.041]; world[0.041]; Hermann[0.041]; World[0.041]; Baudouin[0.041]; physicist[0.041]; people[0.041]; Grand[0.041]; South[0.041]; South[0.041]; campaign[0.040]; enthroned[0.040]; gave[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Europe ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.120:0.120[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.044]; Italy[0.044]; Roman[0.043]; monarchs[0.043]; history[0.043]; destroyed[0.043]; northern[0.043]; France[0.043]; France[0.043]; historical[0.041]; War[0.041]; heavy[0.041]; Breisach[0.041]; English[0.041]; Napoleon[0.041]; Russian[0.041]; major[0.041]; September[0.041]; September[0.041]; ceded[0.041]; ceased[0.041]; Spanish[0.040]; second[0.040]; besieging[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mFreiburg\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mFribourg\u001b[39m <---> \u001b[32mFreiburg im Breisgau\u001b[39m\t\nSCORES: global= 0.259:0.257[\u001b[31m0.001\u001b[39m]; local()= 0.200:0.149[\u001b[31m0.051\u001b[39m]; log p(e|m)= -2.254:-0.207[\u001b[32m2.047\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBreisach[0.048]; France[0.047]; France[0.047]; Italy[0.045]; Europe[0.045]; west[0.044]; French[0.044]; French[0.044]; northern[0.044]; Strasbourg[0.044]; Napoleon[0.043]; Napoleon[0.043]; Alsace[0.043]; battle[0.042]; heavy[0.042]; England[0.042]; got[0.042]; got[0.042]; Treaty[0.042]; Baden[0.042]; capture[0.042]; English[0.041]; miles[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.168:0.168[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.050]; Soviet[0.048]; Belarus[0.047]; republics[0.047]; Kazakhstan[0.046]; American[0.045]; Europe[0.045]; Europe[0.045]; Armenia[0.045]; British[0.044]; Uzbekistan[0.043]; Tajikistan[0.043]; historian[0.043]; authorities[0.042]; nearly[0.042]; States[0.041]; won[0.041]; won[0.041]; gave[0.041]; signed[0.041]; Berlin[0.041]; John[0.041]; modern[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.069:0.069[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwon[0.045]; Stars[0.044]; Belarus[0.044]; nearly[0.044]; leading[0.044]; American[0.044]; Russia[0.043]; British[0.043]; Kazakhstan[0.042]; Europe[0.041]; Europe[0.041]; John[0.041]; acclaim[0.041]; Uzbekistan[0.041]; time[0.041]; Soviet[0.041]; ordered[0.041]; ending[0.040]; events[0.040]; Alan[0.040]; Taylor[0.040]; Iraq[0.040]; Russian[0.040]; formally[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Pedro I ==> ENTITY: \u001b[32mPedro I of Brazil\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.150:0.150[0]; log p(e|m)= -0.066:-0.066[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNapoleon[0.045]; Napoleon[0.045]; Portugal[0.043]; Spanish[0.043]; Brazil[0.042]; Brazil[0.042]; Roman[0.042]; granted[0.042]; ceded[0.042]; Freiburg[0.041]; Italy[0.041]; signing[0.041]; Charles[0.041]; Charles[0.041]; freedom[0.041]; northern[0.041]; Emperor[0.041]; Emperor[0.041]; attempted[0.041]; proclaimed[0.041]; Breisach[0.041]; born[0.041]; seized[0.040]; Succession[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Iraq ==> ENTITY: \u001b[32mIraq\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.110:0.110[0]; log p(e|m)= -0.254:-0.254[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.045]; Kuwaiti[0.044]; forces[0.044]; Uzbekistan[0.043]; army[0.043]; Saudi[0.042]; won[0.042]; won[0.042]; won[0.042]; championship[0.042]; Gulf[0.042]; bombing[0.042]; States[0.042]; secure[0.041]; Europe[0.041]; Europe[0.041]; winning[0.041]; republics[0.041]; Kazakhstan[0.040]; campaign[0.040]; gave[0.040]; signed[0.040]; leading[0.040]; Belarus[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italy ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.206:-0.206[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.045]; northern[0.044]; Napoleon[0.044]; Napoleon[0.044]; Europe[0.044]; monarchs[0.043]; ceded[0.043]; Turin[0.043]; France[0.042]; France[0.042]; Portugal[0.042]; Roman[0.042]; Brazil[0.042]; Brazil[0.042]; troops[0.041]; second[0.040]; Treaty[0.040]; Spanish[0.040]; War[0.040]; Queen[0.040]; week[0.040]; French[0.039]; French[0.039]; history[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.061:0.061[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.051]; Ice[0.046]; appearances[0.044]; Lords[0.043]; imposed[0.043]; James[0.042]; Rising[0.042]; granted[0.042]; Act[0.042]; Lawrence[0.041]; actor[0.041]; China[0.040]; films[0.040]; ended[0.040]; got[0.040]; paid[0.040]; signing[0.040]; Days[0.040]; Allen[0.040]; House[0.040]; Zapata[0.040]; Sir[0.040]; stage[0.040]; stage[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Viva Zapata\u001b[39m ==> ENTITY: \u001b[32mViva Zapata!\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nfilms[0.045]; actor[0.045]; James[0.044]; Hermann[0.043]; Alex[0.043]; Kazan[0.043]; Anthony[0.043]; screen[0.042]; screen[0.042]; Allen[0.042]; Cardinal[0.042]; Days[0.041]; gave[0.041]; gave[0.041]; Alfred[0.041]; paid[0.041]; World[0.041]; Lawrence[0.041]; attempted[0.041]; director[0.041]; King[0.040]; great[0.040]; Cold[0.040]; signing[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCharles\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mCharles VIII of France\u001b[39m <---> \u001b[32mCharles VI, Holy Roman Emperor\u001b[39m\t\nSCORES: global= 0.254:0.250[\u001b[31m0.004\u001b[39m]; local()= 0.172:0.140[\u001b[31m0.032\u001b[39m]; log p(e|m)= -1.609:-0.409[\u001b[32m1.200\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNapoleon[0.045]; Napoleon[0.045]; monarchs[0.044]; French[0.044]; French[0.044]; France[0.043]; France[0.043]; queen[0.043]; Italy[0.043]; Charles[0.042]; ceded[0.041]; Prince[0.041]; Elizabeth[0.041]; Alsace[0.041]; Orleans[0.041]; Roman[0.041]; Freiburg[0.040]; wife[0.040]; Strasbourg[0.040]; Emperor[0.040]; Emperor[0.040]; Breisach[0.040]; England[0.040]; Europe[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hermann Goering\u001b[39m ==> ENTITY: \u001b[32mHermann Göring\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.067:0.067[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nforces[0.047]; German[0.044]; airforce[0.043]; Cardinal[0.043]; gave[0.043]; War[0.043]; encircling[0.042]; Alfred[0.042]; secure[0.042]; announced[0.042]; Cold[0.041]; head[0.041]; British[0.041]; Grand[0.041]; Blitz[0.041]; help[0.041]; Van[0.040]; succeeded[0.040]; Wolsey[0.040]; campaign[0.040]; championship[0.040]; appearances[0.040]; pay[0.040]; Anglicans[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scottish ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.108:0.108[0]; log p(e|m)= -0.480:-0.480[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJames[0.046]; throne[0.045]; King[0.045]; Stewart[0.043]; born[0.043]; born[0.043]; Allen[0.043]; London[0.042]; father[0.042]; world[0.041]; world[0.041]; South[0.041]; South[0.041]; people[0.041]; Italian[0.041]; German[0.041]; began[0.041]; Lawrence[0.040]; enthroned[0.040]; Anne[0.040]; States[0.040]; face[0.040]; Bishop[0.040]; War[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Europe ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurope[0.050]; Russia[0.042]; British[0.042]; won[0.042]; won[0.042]; modern[0.042]; Soviet[0.042]; Alan[0.042]; John[0.042]; American[0.042]; world[0.041]; nearly[0.041]; States[0.041]; events[0.041]; century[0.041]; century[0.041]; republics[0.041]; Africa[0.041]; Africa[0.041]; Belarus[0.040]; United[0.040]; Russian[0.040]; Stars[0.040]; forces[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mRussian\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRussia\u001b[39m <---> \u001b[32mRussian Empire\u001b[39m\t\nSCORES: global= 0.255:0.229[\u001b[31m0.025\u001b[39m]; local()= 0.140:0.147[\u001b[32m0.007\u001b[39m]; log p(e|m)= -1.053:-3.058[\u001b[31m2.005\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEnglish[0.045]; Moscow[0.045]; Moscow[0.045]; ceded[0.043]; Europe[0.043]; Napoleon[0.043]; Napoleon[0.043]; born[0.042]; Spanish[0.042]; northern[0.041]; England[0.041]; Treaty[0.041]; Roman[0.041]; Italy[0.041]; War[0.041]; monarchs[0.041]; freedom[0.040]; Borodino[0.040]; French[0.040]; French[0.040]; Charles[0.040]; Charles[0.040]; Emperor[0.040]; Emperor[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = American ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.058:0.058[0]; log p(e|m)= -0.130:-0.130[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.046]; Europe[0.046]; Europe[0.046]; John[0.044]; British[0.044]; Belarus[0.043]; acclaim[0.043]; nearly[0.043]; Soviet[0.042]; Russia[0.042]; Kazakhstan[0.042]; Alan[0.042]; Uzbekistan[0.042]; historian[0.041]; Taylor[0.040]; republics[0.040]; formally[0.040]; century[0.039]; century[0.039]; common[0.039]; Berlin[0.039]; Armenia[0.039]; modern[0.039]; Stars[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAustrians\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustria\u001b[39m <---> \u001b[32mAustrians\u001b[39m\t\nSCORES: global= 0.260:0.246[\u001b[31m0.014\u001b[39m]; local()= 0.155:0.148[\u001b[31m0.007\u001b[39m]; log p(e|m)= -1.088:-0.439[\u001b[32m0.649\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmonarchs[0.045]; Napoleon[0.045]; Napoleon[0.045]; Europe[0.045]; Breisach[0.044]; Italy[0.043]; ceded[0.043]; Queen[0.042]; queen[0.042]; France[0.042]; France[0.042]; northern[0.042]; born[0.041]; England[0.041]; Prince[0.041]; Henry[0.041]; Treaty[0.040]; War[0.040]; Roman[0.040]; Alsace[0.040]; history[0.040]; Charles[0.039]; Charles[0.039]; Elizabeth[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBrazil\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBrazil\u001b[39m <---> \u001b[32mEmpire of Brazil\u001b[39m\t\nSCORES: global= 0.261:0.228[\u001b[31m0.032\u001b[39m]; local()= 0.127:0.139[\u001b[32m0.013\u001b[39m]; log p(e|m)= -0.213:-4.605[\u001b[31m4.392\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBrazil[0.048]; Portugal[0.044]; Italy[0.044]; independence[0.043]; Pedro[0.043]; France[0.042]; France[0.042]; government[0.042]; Napoleon[0.042]; Napoleon[0.042]; Spanish[0.042]; northern[0.042]; west[0.041]; ceded[0.041]; proclaimed[0.041]; ending[0.041]; granted[0.041]; Treaty[0.040]; week[0.040]; freedom[0.040]; China[0.040]; miles[0.040]; imposed[0.039]; Red[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scottish ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.082:0.082[0]; log p(e|m)= -0.480:-0.480[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLords[0.044]; Napoleon[0.044]; Napoleon[0.044]; Charles[0.044]; Charles[0.044]; born[0.043]; ceded[0.042]; Henry[0.042]; officially[0.042]; independence[0.042]; Succession[0.042]; Russian[0.042]; Sir[0.041]; Campbell[0.041]; France[0.041]; France[0.041]; west[0.041]; Italian[0.041]; Liberal[0.041]; Treaty[0.040]; government[0.040]; Italy[0.040]; granted[0.040]; minister[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Anne Boleyn\u001b[39m ==> ENTITY: \u001b[32mAnne Boleyn\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.120:0.120[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nQueen[0.045]; Charles[0.045]; Charles[0.045]; Elizabeth[0.044]; wife[0.043]; queen[0.043]; Henry[0.043]; England[0.042]; monarchs[0.042]; second[0.042]; Prince[0.042]; major[0.041]; Breisach[0.041]; Daughter[0.041]; historical[0.041]; Emperor[0.041]; London[0.040]; September[0.040]; September[0.040]; Duke[0.040]; Vi[0.040]; arts[0.039]; got[0.039]; English[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Soviet ==> ENTITY: \u001b[32mSoviet Union\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.098:-0.098[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.047]; Russian[0.046]; republics[0.045]; Belarus[0.045]; Kazakhstan[0.043]; Armenia[0.043]; Europe[0.041]; Europe[0.041]; Uzbekistan[0.041]; Tajikistan[0.041]; American[0.041]; ordered[0.041]; authorities[0.041]; nearly[0.041]; forces[0.041]; Iraq[0.041]; army[0.040]; States[0.040]; British[0.040]; won[0.040]; won[0.040]; Berlin[0.040]; century[0.040]; century[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = On the Waterfront\u001b[39m ==> ENTITY: \u001b[32mOn the Waterfront\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nactor[0.046]; James[0.044]; films[0.044]; Kazan[0.043]; appearances[0.043]; screen[0.043]; screen[0.043]; Hermann[0.043]; Anthony[0.042]; director[0.042]; Anne[0.042]; stage[0.041]; stage[0.041]; Alex[0.041]; Cardinal[0.041]; drive[0.041]; Allen[0.041]; Days[0.040]; Alfred[0.040]; Rising[0.040]; Lawrence[0.040]; Italian[0.040]; leading[0.040]; great[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Anthony Quayle\u001b[39m ==> ENTITY: \u001b[32mAnthony Quayle\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.123:0.123[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJames[0.047]; actor[0.046]; Lawrence[0.044]; Allen[0.044]; British[0.043]; Hermann[0.043]; appearances[0.042]; films[0.042]; Anne[0.042]; director[0.042]; Cardinal[0.041]; War[0.041]; King[0.041]; screen[0.041]; screen[0.041]; Alfred[0.041]; Alex[0.040]; stage[0.040]; stage[0.040]; Cold[0.040]; Van[0.040]; Rising[0.040]; began[0.040]; indemnity[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.089:0.089[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.049]; English[0.045]; major[0.045]; signed[0.043]; Queen[0.043]; calendar[0.043]; second[0.043]; history[0.042]; Europe[0.042]; September[0.042]; September[0.042]; Orleans[0.041]; born[0.041]; ceased[0.041]; politics[0.041]; Prince[0.040]; defeated[0.040]; great[0.040]; Following[0.040]; Henry[0.040]; Italy[0.039]; Anne[0.039]; Boleyn[0.039]; monarchs[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Alsace\u001b[39m ==> ENTITY: \u001b[32mAlsace\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.179:0.179[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.047]; France[0.047]; Breisach[0.046]; northern[0.046]; French[0.045]; French[0.045]; Napoleon[0.045]; Napoleon[0.045]; Strasbourg[0.044]; Europe[0.043]; miles[0.043]; destroyed[0.043]; Spanish[0.043]; War[0.042]; ceded[0.042]; Pedro[0.042]; west[0.042]; Borodino[0.042]; Kehl[0.042]; heavy[0.042]; Anne[0.042]; born[0.042]; Baden[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Desmond Tutu\u001b[39m ==> ENTITY: \u001b[32mDesmond Tutu\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAfrica[0.044]; Africa[0.044]; Anglicans[0.043]; Bishop[0.043]; John[0.043]; British[0.043]; Archbishop[0.043]; world[0.042]; world[0.042]; Alan[0.042]; Hermann[0.042]; King[0.041]; people[0.041]; South[0.041]; South[0.041]; World[0.041]; campaign[0.041]; Europe[0.041]; Europe[0.041]; States[0.041]; Jackie[0.040]; blockade[0.040]; Kuwaiti[0.040]; pledges[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = South Africa ==> ENTITY: \u001b[32mSouth Africa\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.116:0.116[0]; log p(e|m)= -0.263:-0.263[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSouth[0.047]; Africa[0.045]; won[0.045]; won[0.045]; won[0.045]; day[0.045]; winning[0.045]; championship[0.044]; Europe[0.044]; Europe[0.044]; leading[0.044]; Taylor[0.043]; World[0.043]; Russia[0.043]; world[0.043]; world[0.043]; Cape[0.042]; Belarus[0.042]; announced[0.042]; States[0.042]; Uzbekistan[0.042]; century[0.041]; Kazakhstan[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = French ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.803:-0.803[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.049]; France[0.049]; English[0.046]; Napoleon[0.045]; French[0.045]; Spanish[0.043]; Europe[0.042]; Italy[0.041]; ceded[0.041]; Orleans[0.041]; monarchs[0.041]; Charles[0.041]; Charles[0.041]; England[0.040]; Elizabeth[0.040]; history[0.040]; Anne[0.040]; Alsace[0.040]; born[0.040]; War[0.040]; Breisach[0.040]; Strasbourg[0.040]; Treaty[0.039]; northern[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Kutuzov\u001b[39m ==> ENTITY: \u001b[32mMikhail Kutuzov\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.125:0.125[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNapoleon[0.045]; Napoleon[0.045]; Borodino[0.043]; War[0.043]; signed[0.043]; Italy[0.043]; Russian[0.042]; Freiburg[0.042]; troops[0.041]; Moscow[0.041]; Moscow[0.041]; army[0.041]; army[0.041]; week[0.041]; Breisach[0.041]; December[0.041]; got[0.041]; got[0.041]; strike[0.041]; Roman[0.041]; ending[0.041]; Emperor[0.040]; Emperor[0.040]; Prince[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belarus ==> ENTITY: \u001b[32mBelarus\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.138:0.138[0]; log p(e|m)= -0.422:-0.422[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussia[0.046]; Soviet[0.044]; Kazakhstan[0.043]; republics[0.043]; Russian[0.043]; Europe[0.043]; Europe[0.043]; Uzbekistan[0.042]; Armenia[0.042]; authorities[0.042]; States[0.042]; won[0.041]; won[0.041]; agreement[0.041]; pledges[0.041]; army[0.040]; Taylor[0.040]; conflicts[0.040]; leading[0.040]; half[0.040]; Tajikistan[0.040]; signed[0.040]; ordered[0.040]; Alan[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Henry Campbell-Bannerman\u001b[39m ==> ENTITY: \u001b[32mHenry Campbell-Bannerman\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.115:0.115[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSir[0.046]; Lords[0.045]; Liberal[0.044]; minister[0.044]; government[0.044]; Scottish[0.043]; Freiburg[0.043]; Breisach[0.042]; politician[0.042]; Bannerman[0.042]; House[0.041]; General[0.041]; Charles[0.041]; Charles[0.041]; paid[0.041]; December[0.041]; granted[0.040]; Vi[0.040]; imposed[0.040]; week[0.040]; independence[0.040]; prime[0.040]; Kehl[0.040]; ceded[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Liberal ==> ENTITY: \u001b[32mLiberal Party (UK)\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.098:0.098[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLords[0.044]; Campbell[0.044]; freedom[0.043]; Scottish[0.043]; Sir[0.043]; politician[0.043]; France[0.043]; France[0.043]; defeated[0.043]; government[0.042]; Spanish[0.042]; House[0.042]; proclaimed[0.041]; Charles[0.041]; Charles[0.041]; minister[0.041]; Strasbourg[0.041]; Italian[0.041]; got[0.040]; got[0.040]; unions[0.040]; paid[0.040]; entered[0.040]; leading[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Uzbekistan ==> ENTITY: \u001b[32mUzbekistan\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.169:0.169[0]; log p(e|m)= -0.389:-0.389[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKazakhstan[0.048]; Tajikistan[0.046]; Belarus[0.044]; Soviet[0.044]; Russia[0.044]; republics[0.043]; won[0.043]; won[0.043]; authorities[0.042]; Russian[0.042]; Armenia[0.042]; American[0.042]; Europe[0.041]; Europe[0.041]; flag[0.040]; pay[0.040]; head[0.040]; Iraq[0.040]; British[0.040]; ending[0.039]; South[0.039]; great[0.039]; nearly[0.039]; indefinitely[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Treaty of Baden\u001b[39m ==> ENTITY: \u001b[32mTreaty of Baden\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrench[0.044]; French[0.044]; France[0.044]; France[0.044]; Spanish[0.043]; Napoleon[0.043]; Napoleon[0.043]; Strasbourg[0.042]; Breisach[0.042]; ceased[0.042]; War[0.042]; Freiburg[0.042]; Italy[0.041]; ceded[0.041]; England[0.041]; Charles[0.041]; Charles[0.041]; Queen[0.041]; Roman[0.041]; Kehl[0.040]; monarchs[0.040]; ending[0.040]; Alsace[0.040]; English[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = British ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.104:0.104[0]; log p(e|m)= -0.489:-0.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJames[0.046]; films[0.046]; actor[0.044]; London[0.043]; Allen[0.043]; China[0.042]; World[0.042]; born[0.042]; born[0.042]; born[0.042]; born[0.042]; Lawrence[0.042]; Days[0.042]; Sir[0.041]; Anthony[0.041]; Anne[0.041]; Rising[0.041]; screen[0.040]; screen[0.040]; King[0.040]; director[0.040]; War[0.040]; freedom[0.040]; war[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ice Cold in Alex\u001b[39m ==> ENTITY: \u001b[32mIce Cold in Alex\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.112:0.112[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWar[0.044]; drive[0.044]; Hermann[0.044]; screen[0.043]; screen[0.043]; British[0.043]; Rising[0.043]; films[0.042]; Blitz[0.042]; Cardinal[0.042]; campaign[0.041]; German[0.041]; London[0.041]; Alfred[0.041]; Van[0.041]; actor[0.041]; King[0.041]; World[0.041]; Days[0.041]; war[0.040]; Red[0.040]; appearances[0.040]; indemnity[0.040]; James[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = English ==> ENTITY: \u001b[32mEngland\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.110:0.110[0]; log p(e|m)= -1.355:-1.355[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.049]; monarchs[0.044]; Spanish[0.044]; Queen[0.044]; Charles[0.043]; Charles[0.043]; London[0.042]; major[0.042]; born[0.041]; arts[0.041]; Henry[0.041]; history[0.041]; Europe[0.041]; queen[0.041]; Elizabeth[0.041]; France[0.041]; France[0.041]; Russian[0.040]; French[0.040]; French[0.040]; Prince[0.040]; ceded[0.040]; Orleans[0.040]; northern[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Strasbourg ==> ENTITY: \u001b[32mStrasbourg\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.148:0.148[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBreisach[0.045]; France[0.044]; France[0.044]; Roman[0.043]; Freiburg[0.042]; troops[0.042]; heavy[0.042]; army[0.042]; army[0.042]; northern[0.042]; Europe[0.042]; Napoleon[0.041]; Napoleon[0.041]; second[0.041]; battle[0.041]; ceded[0.041]; French[0.041]; French[0.041]; December[0.040]; monarchs[0.040]; week[0.040]; defeated[0.040]; defeated[0.040]; Baden[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Anglicans\u001b[39m ==> ENTITY: \u001b[32mAnglicanism\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.089:0.089[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurope[0.046]; Europe[0.046]; Bishop[0.046]; Archbishop[0.044]; Cape[0.043]; Africa[0.043]; Africa[0.043]; John[0.043]; South[0.042]; South[0.042]; Tutu[0.042]; British[0.041]; London[0.041]; States[0.041]; head[0.040]; Iraq[0.040]; Belarus[0.040]; World[0.040]; framework[0.040]; world[0.040]; world[0.040]; Scottish[0.039]; modern[0.039]; announced[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Saudi\u001b[39m ==> ENTITY: \u001b[32mSaudi Arabia\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.132:0.132[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nKuwaiti[0.047]; Iraq[0.046]; Uzbekistan[0.043]; Gulf[0.043]; airforce[0.043]; Kazakhstan[0.042]; British[0.042]; secure[0.042]; won[0.042]; won[0.042]; won[0.042]; forces[0.041]; Europe[0.041]; Europe[0.041]; announced[0.041]; States[0.041]; day[0.041]; Africa[0.040]; Africa[0.040]; championship[0.040]; Belarus[0.040]; World[0.040]; Tajikistan[0.040]; common[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Elizabeth I\u001b[39m ==> ENTITY: \u001b[32mElizabeth I of England\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.146:0.146[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nQueen[0.046]; England[0.044]; Charles[0.044]; Charles[0.044]; monarchs[0.044]; Boleyn[0.043]; queen[0.043]; London[0.041]; Spanish[0.041]; Prince[0.041]; Anne[0.041]; Duke[0.041]; Emperor[0.041]; Henry[0.041]; signed[0.041]; Orleans[0.041]; historical[0.041]; English[0.041]; wife[0.040]; Italy[0.040]; Roman[0.040]; arts[0.040]; great[0.040]; second[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Tajikistan ==> ENTITY: \u001b[32mTajikistan\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.368:-0.368[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nUzbekistan[0.048]; Kazakhstan[0.046]; won[0.045]; won[0.045]; Belarus[0.044]; republics[0.043]; Soviet[0.043]; Russia[0.043]; Russian[0.043]; Armenia[0.042]; Europe[0.042]; Europe[0.042]; American[0.041]; head[0.041]; Iraq[0.040]; nearly[0.040]; presence[0.040]; flag[0.039]; States[0.039]; face[0.039]; events[0.039]; South[0.039]; Stripes[0.039]; gave[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Napoleon ==> ENTITY: \u001b[32mNapoleon\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.163:0.163[0]; log p(e|m)= -0.088:-0.088[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNapoleon[0.051]; France[0.046]; France[0.046]; Borodino[0.044]; Henry[0.044]; Charles[0.044]; Charles[0.044]; War[0.044]; French[0.043]; French[0.043]; Italy[0.043]; troops[0.043]; ceded[0.043]; Roman[0.043]; Russian[0.042]; Europe[0.042]; army[0.042]; army[0.042]; Emperor[0.042]; Emperor[0.042]; freedom[0.042]; Treaty[0.041]; independence[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEugene\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mEugene, Oregon\u001b[39m <---> \u001b[32mPrince Eugene of Savoy\u001b[39m\t\nSCORES: global= 0.249:0.233[\u001b[31m0.016\u001b[39m]; local()= -0.003:0.171[\u001b[32m0.174\u001b[39m]; log p(e|m)= -0.392:-4.828[\u001b[31m4.437\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHenry[0.045]; Napoleon[0.045]; Napoleon[0.045]; Charles[0.043]; Charles[0.043]; Italy[0.043]; France[0.042]; France[0.042]; Breisach[0.042]; week[0.041]; northern[0.041]; Elizabeth[0.041]; Borodino[0.041]; besieging[0.041]; troops[0.041]; War[0.041]; Prince[0.041]; monarchs[0.041]; Kehl[0.040]; ceded[0.040]; English[0.040]; French[0.040]; French[0.040]; Duke[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mTransvaal\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mSouth African Republic\u001b[39m <---> \u001b[32mTransvaal Province\u001b[39m\t\nSCORES: global= 0.253:0.251[\u001b[31m0.002\u001b[39m]; local()= 0.088:0.096[\u001b[32m0.008\u001b[39m]; log p(e|m)= -0.652:-1.366[\u001b[31m0.714\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nweek[0.044]; west[0.043]; got[0.043]; got[0.043]; independence[0.042]; great[0.042]; granted[0.042]; born[0.042]; born[0.042]; defeated[0.042]; leading[0.042]; Lords[0.042]; prime[0.041]; War[0.041]; Campbell[0.041]; House[0.041]; Henry[0.041]; politician[0.041]; December[0.041]; seized[0.041]; officially[0.041]; imposed[0.041]; gave[0.041]; Sir[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Giuseppe Garibaldi\u001b[39m ==> ENTITY: \u001b[32mGiuseppe Garibaldi\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.091:0.091[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nItalian[0.045]; Napoleon[0.045]; Napoleon[0.045]; Pedro[0.044]; war[0.043]; December[0.043]; Zapata[0.042]; Kazan[0.041]; Russian[0.041]; government[0.041]; army[0.041]; Naples[0.041]; freedom[0.041]; week[0.041]; strike[0.041]; battle[0.041]; Borodino[0.041]; heavy[0.041]; gave[0.040]; Henry[0.040]; cost[0.040]; Portugal[0.040]; British[0.040]; officially[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Stars and Stripes ==> ENTITY: \u001b[32mFlag of the United States\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.085:0.085[0]; log p(e|m)= -0.144:-0.144[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nflag[0.048]; Iraq[0.045]; American[0.045]; British[0.043]; John[0.042]; nearly[0.042]; Gulf[0.042]; won[0.042]; time[0.042]; headquarters[0.041]; shaped[0.041]; blockade[0.041]; Alan[0.041]; Europe[0.041]; Europe[0.041]; modern[0.041]; events[0.041]; gave[0.041]; help[0.041]; pay[0.041]; lowered[0.040]; save[0.040]; signed[0.040]; forces[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Roman ==> ENTITY: \u001b[32mRoman Empire\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.143:0.143[0]; log p(e|m)= -1.343:-1.343[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmonarchs[0.046]; northern[0.046]; history[0.044]; Napoleon[0.043]; Napoleon[0.043]; Spanish[0.043]; England[0.042]; Italy[0.042]; Europe[0.042]; troops[0.042]; army[0.042]; army[0.042]; queen[0.041]; Emperor[0.041]; Emperor[0.041]; English[0.041]; ceded[0.040]; west[0.040]; Portugal[0.040]; Queen[0.040]; battle[0.040]; occurred[0.040]; France[0.040]; France[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Lawrence of Arabia\u001b[39m ==> ENTITY: \u001b[32mLawrence of Arabia (film)\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.101:0.101[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.048]; James[0.045]; actor[0.045]; films[0.044]; war[0.044]; World[0.043]; Hermann[0.043]; King[0.042]; Days[0.042]; War[0.042]; London[0.041]; Anthony[0.041]; bombing[0.041]; Cardinal[0.041]; leading[0.041]; airforce[0.040]; Alfred[0.040]; screen[0.040]; screen[0.040]; Rising[0.040]; campaign[0.040]; paid[0.040]; began[0.039]; Thousand[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = James Alfred Van Allen\u001b[39m ==> ENTITY: \u001b[32mJames Van Allen\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.078:0.078[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nphysicist[0.048]; Hermann[0.045]; Cardinal[0.043]; began[0.043]; Grand[0.042]; radiation[0.042]; campaign[0.042]; King[0.042]; succeeded[0.042]; Cold[0.041]; discovered[0.041]; British[0.041]; indemnity[0.041]; paid[0.041]; Goering[0.041]; world[0.041]; world[0.041]; director[0.041]; Rising[0.041]; Austrians[0.041]; Sir[0.040]; great[0.040]; stage[0.040]; stage[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = France ==> ENTITY: \u001b[32mFrance\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.190:0.190[0]; log p(e|m)= -0.163:-0.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.048]; Napoleon[0.045]; Napoleon[0.045]; England[0.045]; Italy[0.044]; Europe[0.043]; French[0.042]; French[0.042]; monarchs[0.042]; Portugal[0.042]; second[0.042]; week[0.041]; Brazil[0.041]; Brazil[0.041]; ceded[0.040]; government[0.040]; Roman[0.040]; Alsace[0.040]; northern[0.040]; Spanish[0.040]; Treaty[0.040]; War[0.040]; independence[0.039]; Breisach[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = British ==> ENTITY: \u001b[32mUnited Kingdom\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.103:0.103[0]; log p(e|m)= -0.489:-0.489[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAmerican[0.049]; Alan[0.045]; Europe[0.045]; Europe[0.045]; John[0.044]; Taylor[0.042]; States[0.041]; world[0.041]; world[0.041]; republics[0.041]; historian[0.041]; Russian[0.041]; century[0.041]; century[0.041]; South[0.041]; South[0.041]; Iraq[0.041]; nearly[0.041]; Africa[0.040]; Africa[0.040]; modern[0.040]; Italian[0.040]; acclaim[0.040]; Russia[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mOrleans\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mOrléans\u001b[39m <---> \u001b[32mHouse of Orléans\u001b[39m\t\nSCORES: global= 0.261:0.228[\u001b[31m0.033\u001b[39m]; local()= 0.183:0.158[\u001b[31m0.025\u001b[39m]; log p(e|m)= -1.470:-3.101[\u001b[31m1.631\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nFrance[0.046]; France[0.046]; Napoleon[0.043]; Napoleon[0.043]; historical[0.043]; west[0.043]; French[0.043]; French[0.043]; Prince[0.042]; Charles[0.042]; Charles[0.042]; monarchs[0.042]; northern[0.041]; Queen[0.041]; ceded[0.040]; Alsace[0.040]; major[0.040]; Elizabeth[0.040]; Strasbourg[0.040]; queen[0.040]; calendar[0.040]; Duke[0.040]; history[0.039]; Italy[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mBorodino\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mBattle of Borodino\u001b[39m <---> \u001b[32mBorodino (village), Mozhaysky District, Moscow Oblast\u001b[39m\t\nSCORES: global= 0.257:0.241[\u001b[31m0.017\u001b[39m]; local()= 0.110:0.080[\u001b[31m0.030\u001b[39m]; log p(e|m)= -0.414:-1.897[\u001b[31m1.483\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.044]; Moscow[0.043]; Moscow[0.043]; Napoleon[0.043]; Napoleon[0.043]; gave[0.042]; cost[0.042]; Roman[0.042]; granted[0.042]; army[0.042]; army[0.042]; heavy[0.042]; war[0.041]; War[0.041]; Charles[0.041]; Charles[0.041]; ceded[0.041]; battle[0.041]; December[0.041]; ceased[0.041]; destroyed[0.041]; government[0.041]; miles[0.040]; capture[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Reuters\u001b[39m ==> ENTITY: \u001b[32mReuters\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.028:0.028[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmajor[0.044]; London[0.044]; Elizabeth[0.044]; history[0.043]; northern[0.043]; troops[0.043]; Europe[0.043]; Anne[0.042]; calendar[0.042]; period[0.042]; monarchs[0.041]; Queen[0.041]; September[0.041]; September[0.041]; Prince[0.041]; Austrians[0.041]; second[0.041]; England[0.041]; historical[0.041]; wife[0.041]; army[0.040]; English[0.040]; Henry[0.040]; Daughter[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Henry VIII\u001b[39m ==> ENTITY: \u001b[32mHenry VIII of England\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nQueen[0.045]; monarchs[0.044]; England[0.044]; Charles[0.044]; Charles[0.044]; Elizabeth[0.044]; Boleyn[0.044]; queen[0.043]; Prince[0.043]; Anne[0.043]; wife[0.042]; London[0.041]; Duke[0.041]; born[0.041]; Roman[0.040]; English[0.040]; second[0.040]; September[0.040]; September[0.040]; France[0.040]; ceded[0.039]; major[0.039]; historical[0.039]; Succession[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.078:0.078[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.046]; won[0.044]; won[0.044]; James[0.043]; appearances[0.043]; airforce[0.043]; Grand[0.043]; Blitz[0.042]; bombing[0.042]; winning[0.041]; day[0.041]; championship[0.041]; announced[0.041]; South[0.041]; South[0.041]; people[0.041]; Anthony[0.041]; stage[0.041]; Alex[0.040]; pay[0.040]; King[0.040]; War[0.040]; began[0.040]; States[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Naples ==> ENTITY: \u001b[32mNaples\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.115:0.115[0]; log p(e|m)= -0.182:-0.182[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNapoleon[0.048]; Napoleon[0.048]; Italian[0.045]; Giuseppe[0.043]; Garibaldi[0.043]; west[0.042]; paid[0.042]; battle[0.041]; Emperor[0.041]; Anthony[0.041]; miles[0.041]; powers[0.041]; granted[0.041]; prime[0.041]; government[0.041]; independence[0.040]; entered[0.040]; British[0.040]; Henry[0.040]; December[0.040]; Portugal[0.040]; seized[0.040]; war[0.040]; Waterfront[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Italian ==> ENTITY: \u001b[32mItaly\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.131:0.131[0]; log p(e|m)= -0.759:-0.759[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGiuseppe[0.045]; Anthony[0.044]; independence[0.043]; Naples[0.043]; Napoleon[0.043]; Napoleon[0.043]; liberation[0.042]; British[0.042]; Portugal[0.042]; born[0.042]; born[0.042]; born[0.042]; government[0.041]; Brazil[0.041]; Brazil[0.041]; actor[0.041]; war[0.041]; Lawrence[0.041]; Arabia[0.041]; proclaimed[0.040]; freedom[0.040]; powers[0.040]; granted[0.039]; gave[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mCharles VI\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mCharles VIII of France\u001b[39m <---> \u001b[32mCharles VI, Holy Roman Emperor\u001b[39m\t\nSCORES: global= 0.254:0.250[\u001b[31m0.004\u001b[39m]; local()= 0.176:0.147[\u001b[31m0.029\u001b[39m]; log p(e|m)= -1.609:-0.409[\u001b[32m1.200\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNapoleon[0.045]; Napoleon[0.045]; monarchs[0.043]; French[0.043]; French[0.043]; France[0.043]; France[0.043]; queen[0.043]; Italy[0.043]; Queen[0.043]; Charles[0.042]; ceded[0.041]; Prince[0.041]; Elizabeth[0.041]; Alsace[0.041]; September[0.041]; Orleans[0.040]; Roman[0.040]; Freiburg[0.040]; wife[0.040]; Strasbourg[0.040]; Emperor[0.040]; Emperor[0.040]; Breisach[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.101:0.101[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBritish[0.051]; James[0.045]; London[0.044]; Allen[0.044]; Lawrence[0.043]; World[0.043]; imposed[0.042]; War[0.042]; Scottish[0.042]; day[0.041]; began[0.041]; appearances[0.041]; King[0.040]; actor[0.040]; world[0.040]; world[0.040]; Stewart[0.040]; Quayle[0.040]; films[0.040]; born[0.040]; born[0.040]; born[0.040]; born[0.040]; paid[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belgian ==> ENTITY: \u001b[32mBelgium\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.124:-0.124[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGrand[0.048]; Prix[0.046]; racing[0.045]; German[0.043]; British[0.042]; South[0.042]; Baudouin[0.042]; World[0.041]; Anne[0.041]; Italian[0.041]; Africa[0.041]; people[0.041]; Van[0.041]; won[0.041]; born[0.041]; born[0.041]; born[0.041]; born[0.041]; Arabia[0.040]; actor[0.040]; James[0.040]; world[0.040]; world[0.040]; appearances[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jackie Stewart\u001b[39m ==> ENTITY: \u001b[32mJackie Stewart\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.140:0.140[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScottish[0.045]; racing[0.045]; winning[0.044]; driver[0.044]; won[0.043]; won[0.043]; appearances[0.043]; James[0.043]; Alan[0.043]; Alex[0.042]; John[0.042]; Hermann[0.041]; championship[0.041]; born[0.041]; born[0.041]; Taylor[0.041]; British[0.041]; World[0.040]; United[0.040]; began[0.040]; Grand[0.040]; Cardinal[0.040]; campaign[0.039]; secure[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Breisach\u001b[39m ==> ENTITY: \u001b[32mBreisach\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.162:0.162[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRoman[0.044]; December[0.044]; War[0.044]; Holy[0.043]; French[0.043]; French[0.043]; ending[0.043]; Napoleon[0.042]; Napoleon[0.042]; Freiburg[0.042]; northern[0.041]; born[0.041]; Kehl[0.041]; Orleans[0.041]; proclaimed[0.041]; destroyed[0.041]; France[0.041]; France[0.041]; west[0.040]; heavy[0.040]; period[0.040]; Strasbourg[0.040]; besieging[0.040]; Spanish[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Europe ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.118:0.118[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEurope[0.052]; Russia[0.044]; British[0.044]; won[0.044]; won[0.044]; modern[0.044]; Soviet[0.044]; Alan[0.044]; John[0.044]; American[0.043]; nearly[0.043]; States[0.043]; events[0.043]; century[0.043]; century[0.043]; republics[0.043]; Africa[0.042]; Africa[0.042]; Belarus[0.042]; United[0.042]; Russian[0.042]; Stars[0.042]; forces[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.158:0.158[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBelarus[0.047]; Russian[0.045]; Soviet[0.044]; Kazakhstan[0.044]; won[0.043]; won[0.043]; republics[0.043]; flag[0.042]; Uzbekistan[0.042]; Europe[0.042]; Europe[0.042]; Armenia[0.042]; Stars[0.042]; American[0.041]; century[0.041]; century[0.041]; Tajikistan[0.040]; modern[0.040]; British[0.040]; United[0.039]; nearly[0.039]; States[0.039]; lowered[0.039]; gave[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Berlin ==> ENTITY: \u001b[32mBerlin\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.063:0.063[0]; log p(e|m)= -0.154:-0.154[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnearly[0.045]; Europe[0.044]; Europe[0.044]; Soviet[0.043]; Alan[0.043]; historian[0.043]; army[0.042]; city[0.042]; John[0.042]; blockade[0.042]; British[0.042]; American[0.041]; Taylor[0.041]; Iraq[0.041]; acclaim[0.041]; Russia[0.041]; century[0.041]; century[0.041]; Russian[0.041]; signed[0.041]; headquarters[0.041]; ordered[0.040]; ending[0.040]; Belarus[0.040]; \t\n\r [==========================================>...] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s276ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4531/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1155testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Paris ==> ENTITY: \u001b[32mParis\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.086:0.086[0]; log p(e|m)= -0.115:-0.115[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nhome[0.043]; Le[0.043]; west[0.043]; seven[0.042]; seven[0.042]; seven[0.042]; Algiers[0.042]; Matin[0.042]; Friday[0.042]; near[0.042]; Saturday[0.042]; forces[0.042]; people[0.042]; people[0.042]; people[0.042]; Liberte[0.042]; security[0.041]; town[0.041]; statement[0.041]; killed[0.041]; killed[0.041]; killed[0.041]; Algerian[0.041]; Algerian[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAlgiers\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAlgiers\u001b[39m <---> \u001b[32mAlgeria\u001b[39m\t\nSCORES: global= 0.269:0.268[\u001b[31m0.000\u001b[39m]; local()= 0.112:0.125[\u001b[32m0.013\u001b[39m]; log p(e|m)= -0.104:-2.749[\u001b[31m2.645\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntown[0.047]; town[0.047]; Algeria[0.046]; Algerian[0.046]; western[0.043]; estimated[0.042]; civilians[0.042]; near[0.042]; forces[0.041]; Haroun[0.041]; violence[0.041]; general[0.041]; Le[0.041]; early[0.040]; attacks[0.040]; past[0.040]; died[0.040]; bomb[0.040]; bomb[0.040]; bomb[0.040]; bomb[0.040]; bomb[0.040]; bomb[0.040]; seven[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Algerian ==> ENTITY: \u001b[32mAlgeria\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.134:0.134[0]; log p(e|m)= -0.209:-0.209[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlgeria[0.047]; government[0.043]; government[0.043]; Moslem[0.043]; Moslem[0.043]; Islamists[0.043]; violence[0.043]; Algiers[0.042]; western[0.042]; forces[0.042]; Liberte[0.041]; blamed[0.041]; civilians[0.041]; attacks[0.041]; estimated[0.041]; people[0.041]; people[0.041]; women[0.040]; town[0.040]; town[0.040]; authorities[0.040]; week[0.040]; Human[0.040]; radical[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Staoueli\u001b[39m ==> ENTITY: \u001b[32mStaouéli\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntown[0.050]; town[0.050]; Algiers[0.049]; Algiers[0.049]; Algeria[0.049]; Friday[0.046]; Saturday[0.045]; Le[0.045]; miles[0.045]; km[0.045]; Algerian[0.045]; Algerian[0.045]; Algerian[0.045]; Liberte[0.044]; near[0.044]; Paris[0.044]; week[0.044]; seven[0.044]; seven[0.044]; seven[0.044]; coastal[0.043]; coastal[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moslem ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.112:0.112[0]; log p(e|m)= -1.214:-1.214[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoslem[0.050]; western[0.045]; Islamists[0.044]; people[0.044]; Algeria[0.042]; town[0.042]; saying[0.042]; said[0.042]; National[0.042]; Algerian[0.041]; radical[0.041]; week[0.041]; civilians[0.041]; women[0.041]; August[0.041]; authorities[0.041]; government[0.041]; government[0.041]; appointed[0.040]; home[0.040]; estimated[0.040]; violence[0.040]; children[0.040]; forces[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Algeria ==> ENTITY: \u001b[32mAlgeria\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.337:-0.337[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlgerian[0.045]; Algerian[0.045]; Paris[0.044]; Algiers[0.043]; forces[0.043]; town[0.042]; west[0.042]; Liberte[0.042]; security[0.042]; home[0.042]; Friday[0.041]; people[0.041]; people[0.041]; people[0.041]; seven[0.041]; seven[0.041]; seven[0.041]; Le[0.040]; Matin[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; statement[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Algerian ==> ENTITY: \u001b[32mAlgeria\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.141:0.141[0]; log p(e|m)= -0.209:-0.209[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlgeria[0.049]; Algerian[0.044]; Algiers[0.044]; security[0.043]; forces[0.043]; Liberte[0.043]; town[0.042]; week[0.041]; people[0.041]; people[0.041]; people[0.041]; Paris[0.041]; home[0.041]; home[0.041]; killed[0.041]; killed[0.041]; killed[0.041]; seven[0.040]; seven[0.040]; seven[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Algeria ==> ENTITY: \u001b[32mAlgeria\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.133:0.133[0]; log p(e|m)= -0.337:-0.337[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlgerian[0.044]; government[0.043]; government[0.043]; western[0.043]; forces[0.043]; Islamists[0.043]; Algiers[0.042]; radical[0.042]; Moslem[0.042]; Moslem[0.042]; rebels[0.042]; guerrillas[0.041]; civilians[0.041]; August[0.041]; town[0.041]; week[0.041]; blamed[0.041]; violence[0.041]; shortly[0.041]; home[0.041]; early[0.041]; people[0.040]; general[0.040]; Haroun[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Le Matin ==> ENTITY: \u001b[32mLe Matin (France)\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.107:0.107[0]; log p(e|m)= -0.435:-0.435[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnewspaper[0.046]; newspaper[0.046]; newspaper[0.046]; Liberte[0.045]; Algeria[0.045]; newspapers[0.044]; Algerian[0.042]; Algerian[0.042]; Algerian[0.042]; Paris[0.042]; quoted[0.042]; statement[0.041]; Newspapers[0.041]; shortly[0.041]; Algiers[0.041]; Algiers[0.041]; week[0.040]; quoting[0.040]; Friday[0.039]; Bou[0.039]; km[0.039]; Haroun[0.038]; Saturday[0.038]; miles[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Algiers ==> ENTITY: \u001b[32mAlgiers\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.121:0.121[0]; log p(e|m)= -0.104:-0.104[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntown[0.047]; Algerian[0.046]; Algerian[0.046]; Algeria[0.044]; Paris[0.043]; near[0.042]; forces[0.042]; west[0.042]; Le[0.041]; Liberte[0.041]; shortly[0.041]; Saturday[0.041]; Friday[0.041]; bomb[0.040]; bomb[0.040]; bomb[0.040]; bomb[0.040]; bomb[0.040]; bomb[0.040]; seven[0.040]; seven[0.040]; seven[0.040]; home[0.040]; security[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moslem ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.257:0.257[0]; local()= 0.112:0.112[0]; log p(e|m)= -1.214:-1.214[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoslem[0.050]; western[0.045]; Islamists[0.044]; people[0.044]; Algeria[0.042]; town[0.042]; saying[0.042]; said[0.042]; National[0.042]; Algerian[0.041]; radical[0.041]; week[0.041]; civilians[0.041]; women[0.041]; August[0.041]; authorities[0.041]; government[0.041]; government[0.041]; appointed[0.040]; home[0.040]; estimated[0.040]; violence[0.040]; children[0.040]; forces[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Algerian ==> ENTITY: \u001b[32mAlgeria\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.209:-0.209[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAlgeria[0.053]; Algerian[0.048]; Algiers[0.047]; western[0.047]; security[0.047]; forces[0.047]; Liberte[0.046]; town[0.045]; town[0.045]; week[0.045]; people[0.045]; people[0.045]; people[0.045]; Paris[0.044]; home[0.044]; home[0.044]; killed[0.044]; killed[0.044]; killed[0.044]; seven[0.044]; seven[0.044]; seven[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Islamists\u001b[39m ==> ENTITY: \u001b[32mIslamism\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nwomen[0.045]; Moslem[0.045]; Moslem[0.045]; forces[0.044]; radical[0.044]; guerrillas[0.044]; government[0.043]; government[0.043]; bomb[0.042]; civilians[0.042]; people[0.042]; attacks[0.042]; rebels[0.041]; saying[0.041]; authorities[0.040]; quoted[0.040]; Algeria[0.040]; Algerian[0.040]; blamed[0.040]; appointed[0.040]; past[0.040]; said[0.040]; general[0.039]; Haroun[0.039]; \t\n\r [==========================================>...] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s217ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4544/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1027testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moslem ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.172:0.172[0]; log p(e|m)= -1.214:-1.214[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIslamic[0.050]; Egypt[0.044]; Egypt[0.044]; eastern[0.044]; al[0.043]; al[0.043]; al[0.043]; Cairo[0.042]; Cairo[0.042]; Al[0.042]; Islamiya[0.041]; said[0.041]; said[0.041]; said[0.041]; institutions[0.041]; individuals[0.041]; Sharqiyah[0.040]; Saadani[0.040]; outside[0.040]; prominent[0.040]; group[0.040]; government[0.040]; leader[0.040]; province[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Egypt ==> ENTITY: \u001b[32mEgypt\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.151:0.151[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEgypt[0.053]; Cairo[0.047]; Cairo[0.047]; al[0.045]; al[0.045]; al[0.045]; security[0.044]; Islamic[0.044]; Moslem[0.044]; Al[0.043]; Sharqiyah[0.043]; eastern[0.042]; government[0.042]; institutions[0.042]; officials[0.042]; detained[0.041]; newspapers[0.041]; newspapers[0.041]; newspapers[0.041]; province[0.041]; reported[0.041]; militants[0.041]; militants[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Islamic ==> ENTITY: \u001b[32mIslam\u001b[39m\t\nSCORES: global= 0.251:0.251[0]; local()= 0.122:0.122[0]; log p(e|m)= -0.212:-0.212[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nal[0.047]; Al[0.044]; Cairo[0.043]; militants[0.043]; security[0.043]; people[0.043]; government[0.042]; leader[0.042]; Akhbar[0.042]; establish[0.042]; prominent[0.041]; state[0.041]; said[0.041]; said[0.041]; satellite[0.041]; eastern[0.041]; institutions[0.041]; plotting[0.040]; Rami[0.040]; individuals[0.040]; Sharqiyah[0.040]; Mubarak[0.040]; officials[0.040]; city[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hosni Mubarak\u001b[39m ==> ENTITY: \u001b[32mHosni Mubarak\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.160:0.160[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCairo[0.046]; security[0.046]; government[0.044]; government[0.044]; Al[0.044]; militants[0.043]; al[0.043]; President[0.043]; Islamic[0.042]; people[0.042]; institutions[0.042]; strategic[0.041]; prominent[0.041]; State[0.041]; leader[0.041]; Rami[0.041]; officials[0.040]; Saadani[0.040]; said[0.040]; said[0.040]; said[0.040]; topple[0.040]; satellite[0.040]; state[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cairo ==> ENTITY: \u001b[32mCairo\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.181:0.181[0]; log p(e|m)= -0.152:-0.152[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCairo[0.049]; Egypt[0.046]; Egypt[0.046]; al[0.045]; al[0.045]; al[0.045]; Islamic[0.044]; Al[0.043]; Islamiya[0.042]; city[0.042]; said[0.041]; said[0.041]; said[0.041]; Sharqiyah[0.041]; institutions[0.040]; outside[0.040]; Moslem[0.040]; eastern[0.040]; Friday[0.039]; militants[0.039]; militants[0.039]; reported[0.038]; prominent[0.038]; details[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAl-Akhbar\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAl Akhbar (Lebanon)\u001b[39m <---> \u001b[32mAl Akhbar (Egypt)\u001b[39m\t\nSCORES: global= 0.292:0.286[\u001b[31m0.006\u001b[39m]; local()= 0.171:0.207[\u001b[32m0.036\u001b[39m]; log p(e|m)= 0.000:-1.720[\u001b[31m1.720\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nal[0.044]; al[0.044]; al[0.044]; Akhbar[0.044]; Egypt[0.044]; Egypt[0.044]; Cairo[0.042]; Cairo[0.042]; Hosni[0.042]; Islamiya[0.042]; Rami[0.042]; institutions[0.041]; Mubarak[0.041]; Islamic[0.041]; Islamic[0.041]; prominent[0.041]; newspapers[0.040]; newspapers[0.040]; newspapers[0.040]; said[0.040]; said[0.040]; said[0.040]; reported[0.040]; strategic[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cairo ==> ENTITY: \u001b[32mCairo\u001b[39m\t\nSCORES: global= 0.280:0.280[0]; local()= 0.195:0.195[0]; log p(e|m)= -0.152:-0.152[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCairo[0.049]; Egypt[0.045]; Egypt[0.045]; al[0.044]; al[0.044]; al[0.044]; Islamic[0.043]; Islamic[0.043]; Al[0.043]; Islamiya[0.042]; city[0.041]; available[0.041]; said[0.040]; said[0.040]; said[0.040]; Sharqiyah[0.040]; Hosni[0.040]; Mubarak[0.040]; institutions[0.039]; outside[0.039]; Moslem[0.039]; launched[0.039]; eastern[0.039]; establish[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Egypt ==> ENTITY: \u001b[32mEgypt\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.155:0.155[0]; log p(e|m)= -0.274:-0.274[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEgypt[0.051]; Cairo[0.045]; Cairo[0.045]; al[0.044]; al[0.044]; al[0.044]; Islamic[0.043]; Moslem[0.042]; Al[0.042]; Sharqiyah[0.041]; eastern[0.041]; government[0.041]; institutions[0.040]; detained[0.040]; newspapers[0.040]; newspapers[0.040]; province[0.040]; reported[0.040]; militants[0.040]; militants[0.040]; said[0.040]; said[0.040]; detain[0.040]; Rami[0.039]; \t\n\r [==========================================>...] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s176ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4552/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1134testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.141:0.141[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.051]; China[0.051]; national[0.046]; Qiang[0.043]; Wang[0.043]; Wang[0.043]; Xiao[0.043]; Donghai[0.042]; nation[0.042]; leadership[0.042]; Communist[0.041]; campaign[0.040]; campaign[0.040]; dissidents[0.040]; Manila[0.040]; York[0.040]; dissident[0.040]; sent[0.039]; appeared[0.039]; meant[0.039]; said[0.039]; said[0.039]; said[0.039]; Friday[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mXiao\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mDeng Xiaoping\u001b[39m <---> \u001b[32mXiao Qiang\u001b[39m\t\nSCORES: global= 0.259:0.223[\u001b[31m0.035\u001b[39m]; local()= 0.185:0.147[\u001b[31m0.037\u001b[39m]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.046]; China[0.046]; China[0.046]; government[0.044]; Communist[0.042]; Wang[0.042]; Wang[0.042]; Deng[0.041]; Xiao[0.041]; Xiao[0.041]; Xiao[0.041]; Amnesty[0.041]; national[0.041]; Xiaoping[0.041]; Party[0.041]; leader[0.041]; Human[0.041]; attend[0.040]; conference[0.040]; meant[0.040]; meant[0.040]; strengthen[0.040]; Qiang[0.040]; dozens[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mXiao\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mDeng Xiaoping\u001b[39m <---> \u001b[32mXiao Qiang\u001b[39m\t\nSCORES: global= 0.260:0.222[\u001b[31m0.038\u001b[39m]; local()= 0.193:0.129[\u001b[31m0.064\u001b[39m]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.047]; China[0.047]; Chen[0.043]; reform[0.043]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Xiao[0.042]; Xiao[0.042]; security[0.042]; ideological[0.042]; Hangzhou[0.041]; month[0.041]; encouraged[0.041]; education[0.040]; class[0.040]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; dissident[0.040]; issues[0.040]; released[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.160:0.160[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.049]; China[0.049]; China[0.049]; Hangzhou[0.043]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; month[0.042]; Xiao[0.042]; Xiao[0.042]; Xiao[0.042]; high[0.041]; International[0.040]; security[0.040]; government[0.040]; attend[0.039]; Manila[0.039]; authorities[0.039]; conference[0.039]; stamp[0.039]; dissident[0.039]; sent[0.039]; city[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.049]; China[0.049]; China[0.049]; national[0.045]; Qiang[0.042]; Wang[0.042]; Wang[0.042]; Xiao[0.041]; Xiao[0.041]; Xiao[0.041]; Xiao[0.041]; Xiaoping[0.041]; Deng[0.041]; period[0.041]; International[0.040]; government[0.040]; Communist[0.039]; campaign[0.039]; campaign[0.039]; Manila[0.039]; leader[0.039]; authorities[0.039]; conference[0.039]; York[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Communist Party ==> ENTITY: \u001b[32mCommunist Party of China\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.166:0.166[0]; log p(e|m)= -2.207:-2.207[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.046]; China[0.046]; China[0.046]; China[0.046]; leader[0.044]; leadership[0.043]; national[0.043]; government[0.043]; political[0.042]; group[0.041]; activist[0.041]; current[0.041]; Wang[0.041]; Wang[0.041]; Deng[0.040]; conference[0.040]; Xiao[0.040]; Xiao[0.040]; Xiao[0.040]; press[0.040]; independent[0.040]; executive[0.040]; sent[0.040]; nation[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.050]; China[0.050]; Hangzhou[0.043]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Chen[0.042]; month[0.042]; Xiao[0.042]; Xiao[0.042]; Xiao[0.042]; high[0.041]; International[0.041]; security[0.040]; government[0.040]; beat[0.039]; conference[0.039]; stamp[0.039]; dissident[0.039]; sent[0.039]; city[0.039]; officers[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mXiao\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mDeng Xiaoping\u001b[39m <---> \u001b[32mXiao Qiang\u001b[39m\t\nSCORES: global= 0.259:0.223[\u001b[31m0.036\u001b[39m]; local()= 0.193:0.143[\u001b[31m0.050\u001b[39m]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.050]; China[0.050]; China[0.050]; government[0.048]; Communist[0.046]; Wang[0.045]; Wang[0.045]; Wang[0.045]; Deng[0.045]; Xiao[0.045]; Xiao[0.045]; Amnesty[0.045]; national[0.045]; Xiaoping[0.045]; Party[0.045]; leader[0.044]; attend[0.044]; conference[0.044]; meant[0.043]; meant[0.043]; strengthen[0.043]; dozens[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.135:0.135[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.051]; China[0.051]; national[0.046]; Qiang[0.043]; Wang[0.043]; Wang[0.043]; Xiao[0.043]; Donghai[0.042]; nation[0.042]; leadership[0.042]; campaign[0.041]; dissidents[0.041]; Manila[0.040]; York[0.040]; dissident[0.040]; sent[0.040]; appeared[0.040]; said[0.039]; said[0.039]; said[0.039]; Friday[0.039]; Strike[0.039]; group[0.039]; current[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York-based ==> ENTITY: \u001b[32mNew York\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.056:0.056[0]; log p(e|m)= -0.017:-0.017[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHard[0.047]; group[0.044]; senior[0.044]; based[0.043]; Human[0.043]; leadership[0.043]; national[0.042]; transitions[0.042]; nation[0.042]; Friday[0.042]; transition[0.042]; showed[0.041]; appeared[0.041]; Saturday[0.041]; interview[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; director[0.040]; apparently[0.040]; going[0.040]; dozens[0.040]; activist[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hangzhou\u001b[39m ==> ENTITY: \u001b[32mHangzhou\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.153:0.153[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.048]; China[0.048]; China[0.048]; city[0.045]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Wang[0.042]; Xiao[0.042]; Xiao[0.042]; Xiao[0.042]; class[0.041]; Chen[0.041]; security[0.041]; month[0.040]; issues[0.039]; pay[0.039]; public[0.039]; family[0.039]; day[0.039]; potentially[0.038]; attack[0.038]; released[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Amnesty International\u001b[39m ==> ENTITY: \u001b[32mAmnesty International\u001b[39m\t\nSCORES: global= 0.279:0.279[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncampaign[0.044]; campaign[0.044]; activists[0.044]; rights[0.044]; rights[0.044]; national[0.042]; imprisoned[0.042]; political[0.041]; press[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; legitimate[0.041]; attend[0.041]; hundreds[0.041]; social[0.041]; issues[0.041]; thousands[0.041]; suggesting[0.040]; government[0.040]; detained[0.040]; interview[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Deng Xiaoping\u001b[39m ==> ENTITY: \u001b[32mDeng Xiaoping\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.198:0.198[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.050]; China[0.050]; China[0.050]; government[0.048]; leadership[0.048]; Communist[0.046]; Wang[0.045]; Wang[0.045]; national[0.045]; Party[0.045]; Xiao[0.045]; Xiao[0.045]; Xiao[0.045]; Xiao[0.045]; leader[0.044]; attend[0.044]; profound[0.044]; meant[0.043]; meant[0.043]; strengthen[0.043]; dozens[0.043]; Qiang[0.043]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mXiao\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mDeng Xiaoping\u001b[39m <---> \u001b[32mXiao Qiang\u001b[39m\t\nSCORES: global= 0.260:0.222[\u001b[31m0.037\u001b[39m]; local()= 0.177:0.117[\u001b[31m0.060\u001b[39m]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.047]; Chen[0.044]; reform[0.044]; Wang[0.043]; Wang[0.043]; Wang[0.043]; Wang[0.043]; Xiao[0.042]; security[0.042]; ideological[0.042]; Hangzhou[0.042]; month[0.041]; encouraged[0.041]; education[0.041]; class[0.041]; said[0.041]; said[0.041]; said[0.041]; dissident[0.041]; released[0.040]; euphemism[0.040]; attack[0.040]; coercive[0.040]; health[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Xiao Qiang\u001b[39m ==> ENTITY: \u001b[32mXiao Qiang\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.154:0.154[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nXiao[0.044]; China[0.043]; China[0.043]; China[0.043]; China[0.043]; Deng[0.043]; Wang[0.043]; Wang[0.043]; Human[0.043]; senior[0.042]; director[0.042]; Rights[0.042]; rights[0.041]; rights[0.041]; leadership[0.041]; Xiaoping[0.041]; aimed[0.041]; transitions[0.040]; dissident[0.040]; dissidents[0.040]; current[0.040]; transition[0.040]; based[0.040]; group[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.154:0.154[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.049]; China[0.049]; China[0.049]; Hangzhou[0.043]; Wang[0.042]; Wang[0.042]; Xiao[0.042]; Xiao[0.042]; Xiao[0.042]; Xiaoping[0.042]; period[0.041]; high[0.041]; International[0.040]; Deng[0.040]; security[0.040]; Party[0.040]; government[0.040]; Communist[0.040]; campaign[0.039]; attend[0.039]; Manila[0.039]; leader[0.039]; conference[0.039]; stamp[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mHuman Rights in China\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHuman rights in China\u001b[39m <---> \u001b[32mHuman Rights in China (organization)\u001b[39m\t\nSCORES: global= 0.295:0.288[\u001b[31m0.007\u001b[39m]; local()= 0.182:0.197[\u001b[32m0.014\u001b[39m]; log p(e|m)= 0.000:-0.068[\u001b[31m0.068\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.046]; China[0.046]; China[0.046]; activists[0.044]; executive[0.042]; dissidents[0.042]; social[0.042]; activist[0.042]; director[0.042]; rights[0.041]; rights[0.041]; York[0.041]; political[0.041]; leadership[0.041]; dissident[0.041]; authorities[0.040]; Xiao[0.040]; Xiao[0.040]; national[0.040]; group[0.040]; Wang[0.040]; Wang[0.040]; imprisoned[0.040]; dozens[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mXiao\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mDeng Xiaoping\u001b[39m <---> \u001b[32mXiao Qiang\u001b[39m\t\nSCORES: global= 0.259:0.223[\u001b[31m0.036\u001b[39m]; local()= 0.200:0.152[\u001b[31m0.047\u001b[39m]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.046]; China[0.046]; China[0.046]; China[0.046]; government[0.044]; reform[0.042]; Wang[0.041]; Wang[0.041]; Wang[0.041]; Deng[0.041]; Xiao[0.041]; Xiao[0.041]; Xiao[0.041]; Amnesty[0.041]; security[0.041]; Xiaoping[0.041]; ideological[0.041]; Hangzhou[0.041]; leader[0.040]; attend[0.040]; conference[0.040]; month[0.040]; meant[0.040]; education[0.040]; \t\n\r [==========================================>...] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s98ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4570/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1094testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = El Salvador ==> ENTITY: \u001b[32mEl Salvador national football team\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.162:0.162[0]; log p(e|m)= -2.146:-2.146[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.047]; Soccer[0.045]; Los[0.045]; Luis[0.045]; Salvador[0.045]; Salvador[0.045]; international[0.044]; States[0.043]; Wynalda[0.042]; El[0.042]; El[0.042]; Angeles[0.041]; minute[0.040]; kick[0.040]; Friday[0.040]; friendly[0.040]; Eric[0.039]; halftime[0.039]; penalty[0.039]; Lazo[0.039]; United[0.039]; beat[0.039]; Beat[0.039]; Joe[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = United States ==> ENTITY: \u001b[32mUnited States men's national soccer team\u001b[39m\t\nSCORES: global= 0.241:0.241[0]; local()= 0.133:0.133[0]; log p(e|m)= -4.343:-4.343[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.052]; Soccer[0.050]; international[0.046]; Wynalda[0.045]; Luis[0.042]; friendly[0.042]; Friday[0.042]; beat[0.042]; penalty[0.042]; kick[0.042]; Joe[0.041]; minute[0.041]; Eric[0.041]; halftime[0.041]; Attendance[0.040]; Salvador[0.040]; Salvador[0.040]; Salvador[0.040]; Scorers[0.039]; Moore[0.039]; Los[0.039]; Lazo[0.039]; Beat[0.038]; 3rd[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Eric Wynalda\u001b[39m ==> ENTITY: \u001b[32mEric Wynalda\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.110:0.110[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.049]; soccer[0.047]; States[0.043]; Luis[0.043]; international[0.043]; Joe[0.043]; United[0.043]; Moore[0.043]; minute[0.042]; halftime[0.041]; Scorers[0.041]; Attendance[0.041]; Salvador[0.041]; Salvador[0.041]; Salvador[0.041]; penalty[0.041]; kick[0.041]; Los[0.040]; 3rd[0.040]; beat[0.040]; Max[0.040]; Beat[0.039]; Angeles[0.039]; Friday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Joe-Max Moore\u001b[39m ==> ENTITY: \u001b[32mJoe-Max Moore\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.103:0.103[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.047]; soccer[0.046]; Moore[0.045]; Eric[0.043]; Wynalda[0.043]; international[0.043]; halftime[0.042]; States[0.042]; Luis[0.042]; Scorers[0.042]; Attendance[0.042]; United[0.042]; kick[0.041]; Salvador[0.041]; Salvador[0.041]; Salvador[0.041]; 3rd[0.041]; minute[0.041]; Angeles[0.040]; friendly[0.040]; Los[0.040]; Friday[0.038]; penalty[0.038]; Beat[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = El Salvador ==> ENTITY: \u001b[32mEl Salvador national football team\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.162:0.162[0]; log p(e|m)= -2.146:-2.146[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSalvador[0.048]; Salvador[0.048]; Los[0.046]; soccer[0.046]; Luis[0.045]; Soccer[0.045]; El[0.044]; El[0.044]; international[0.043]; States[0.043]; Wynalda[0.042]; Angeles[0.041]; Lazo[0.040]; minute[0.040]; kick[0.039]; Friday[0.039]; friendly[0.039]; Eric[0.039]; halftime[0.039]; penalty[0.039]; United[0.038]; beat[0.038]; Beat[0.038]; Joe[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Los Angeles ==> ENTITY: \u001b[32mLos Angeles\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.049:0.049[0]; log p(e|m)= -0.101:-0.101[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.050]; soccer[0.049]; international[0.046]; Wynalda[0.042]; States[0.042]; United[0.042]; Salvador[0.042]; Salvador[0.042]; Salvador[0.042]; Friday[0.042]; minute[0.042]; Luis[0.041]; Joe[0.041]; Attendance[0.041]; beat[0.040]; penalty[0.040]; Eric[0.040]; kick[0.040]; El[0.040]; El[0.040]; El[0.040]; Beat[0.039]; halftime[0.039]; friendly[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States men's national soccer team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.138:0.138[0]; log p(e|m)= -4.075:-4.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.051]; Soccer[0.050]; international[0.046]; Wynalda[0.045]; beat[0.043]; States[0.043]; Luis[0.042]; friendly[0.042]; Beat[0.042]; United[0.042]; Joe[0.041]; kick[0.041]; minute[0.040]; Attendance[0.040]; Friday[0.040]; Eric[0.040]; Moore[0.040]; Salvador[0.039]; Salvador[0.039]; Salvador[0.039]; Scorers[0.039]; penalty[0.039]; Los[0.039]; 3rd[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = El Salvador ==> ENTITY: \u001b[32mEl Salvador national football team\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.162:0.162[0]; log p(e|m)= -2.146:-2.146[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.047]; Soccer[0.045]; Los[0.045]; Luis[0.045]; Salvador[0.045]; Salvador[0.045]; international[0.044]; States[0.043]; Wynalda[0.042]; El[0.042]; El[0.042]; Angeles[0.041]; minute[0.040]; kick[0.040]; Friday[0.040]; friendly[0.040]; Eric[0.039]; halftime[0.039]; penalty[0.039]; Lazo[0.039]; United[0.039]; beat[0.039]; Beat[0.039]; Joe[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States men's national soccer team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.138:0.138[0]; log p(e|m)= -4.075:-4.075[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.051]; Soccer[0.050]; international[0.046]; Wynalda[0.045]; beat[0.043]; States[0.043]; Luis[0.042]; friendly[0.042]; Beat[0.042]; United[0.042]; Joe[0.041]; kick[0.041]; minute[0.040]; Attendance[0.040]; Friday[0.040]; Eric[0.040]; Moore[0.040]; Salvador[0.039]; Salvador[0.039]; Salvador[0.039]; Scorers[0.039]; penalty[0.039]; Los[0.039]; 3rd[0.039]; \t\n\r [==========================================>...] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s52ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4579/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1051testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Liechtenstein ==> ENTITY: \u001b[32mLiechtenstein national football team\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.166:0.166[0]; log p(e|m)= -1.995:-1.995[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLiechtenstein[0.048]; Ireland[0.047]; Ireland[0.047]; soccer[0.047]; qualifier[0.045]; Soccer[0.044]; Cup[0.044]; Cup[0.044]; Qualifier[0.044]; Republic[0.042]; European[0.042]; Saturday[0.041]; Eschen[0.041]; World[0.040]; World[0.040]; beat[0.040]; group[0.039]; halftime[0.038]; Harte[0.038]; Beat[0.038]; Neill[0.038]; Andy[0.038]; Keith[0.037]; Ian[0.037]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEschen\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUSV Eschen/Mauren\u001b[39m <---> \u001b[32mEschen\u001b[39m\t\nSCORES: global= 0.269:0.263[\u001b[31m0.006\u001b[39m]; local()= 0.120:0.049[\u001b[31m0.071\u001b[39m]; log p(e|m)= -2.323:0.000[\u001b[32m2.323\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.047]; Soccer[0.046]; Cup[0.045]; Cup[0.045]; qualifier[0.045]; Qualifier[0.045]; Liechtenstein[0.044]; Liechtenstein[0.044]; beat[0.042]; Saturday[0.042]; Andy[0.041]; Keith[0.041]; Attendance[0.041]; Niall[0.040]; Ian[0.040]; Scorers[0.040]; Neill[0.040]; halftime[0.040]; group[0.039]; European[0.039]; Republic[0.039]; Townsend[0.039]; Beat[0.039]; 7th[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ian Harte\u001b[39m ==> ENTITY: \u001b[32mIan Harte\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.100:0.100[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.046]; Ireland[0.046]; Andy[0.043]; Keith[0.043]; qualifier[0.042]; Neill[0.042]; Saturday[0.042]; Quinn[0.042]; Qualifier[0.042]; Soccer[0.042]; Niall[0.042]; Republic[0.042]; halftime[0.041]; Eschen[0.041]; Scorers[0.041]; Attendance[0.041]; European[0.041]; soccer[0.041]; Cup[0.040]; Cup[0.040]; beat[0.040]; group[0.040]; Liechtenstein[0.039]; Liechtenstein[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mEuropean\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mUEFA\u001b[39m <---> \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.242:0.239[\u001b[31m0.003\u001b[39m]; local()= 0.108:0.049[\u001b[31m0.060\u001b[39m]; log p(e|m)= -3.576:-0.256[\u001b[32m3.319\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.047]; Soccer[0.047]; Cup[0.045]; Cup[0.045]; qualifier[0.045]; Ireland[0.044]; Ireland[0.044]; Qualifier[0.044]; group[0.043]; Republic[0.041]; Liechtenstein[0.040]; Liechtenstein[0.040]; beat[0.040]; Saturday[0.040]; World[0.040]; World[0.040]; 7th[0.040]; Neill[0.039]; 5th[0.039]; Andy[0.039]; Eschen[0.039]; Attendance[0.039]; Keith[0.039]; Townsend[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andy Townsend\u001b[39m ==> ENTITY: \u001b[32mAndy Townsend\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.106:0.106[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.045]; Ireland[0.045]; Ian[0.044]; Soccer[0.043]; soccer[0.043]; Cup[0.043]; Cup[0.043]; Neill[0.042]; European[0.042]; Keith[0.042]; qualifier[0.042]; Qualifier[0.042]; World[0.041]; World[0.041]; Niall[0.041]; halftime[0.041]; Saturday[0.041]; Eschen[0.041]; Scorers[0.040]; Attendance[0.040]; Republic[0.040]; beat[0.040]; group[0.040]; Beat[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Liechtenstein ==> ENTITY: \u001b[32mLiechtenstein national football team\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.166:0.166[0]; log p(e|m)= -1.995:-1.995[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLiechtenstein[0.049]; Ireland[0.048]; Ireland[0.048]; qualifier[0.045]; soccer[0.045]; Qualifier[0.044]; Soccer[0.044]; Republic[0.043]; European[0.043]; Cup[0.041]; Cup[0.041]; Eschen[0.041]; World[0.040]; World[0.040]; beat[0.040]; Saturday[0.040]; group[0.039]; halftime[0.039]; Harte[0.039]; Beat[0.039]; Neill[0.038]; Andy[0.038]; Keith[0.038]; Ian[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Republic of Ireland ==> ENTITY: \u001b[32mRepublic of Ireland national football team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.153:0.153[0]; log p(e|m)= -1.461:-1.461[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.054]; soccer[0.045]; Soccer[0.045]; Qualifier[0.043]; Cup[0.043]; Cup[0.043]; Quinn[0.043]; qualifier[0.043]; European[0.042]; Neill[0.042]; Keith[0.042]; Harte[0.041]; Saturday[0.041]; Niall[0.041]; Andy[0.040]; Liechtenstein[0.040]; Liechtenstein[0.040]; Ian[0.040]; World[0.039]; World[0.039]; 5th[0.039]; group[0.039]; 7th[0.038]; beat[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Niall Quinn\u001b[39m ==> ENTITY: \u001b[32mNiall Quinn\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.128:0.128[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.047]; Ireland[0.047]; Neill[0.044]; Ian[0.044]; Keith[0.043]; 7th[0.043]; 5th[0.043]; Andy[0.043]; World[0.042]; World[0.042]; Saturday[0.042]; Cup[0.041]; Cup[0.041]; Qualifier[0.041]; soccer[0.041]; European[0.041]; Soccer[0.040]; qualifier[0.040]; Republic[0.040]; group[0.040]; Harte[0.040]; halftime[0.039]; Eschen[0.039]; Scorers[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIreland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mRepublic of Ireland\u001b[39m <---> \u001b[32mRepublic of Ireland national football team\u001b[39m\t\nSCORES: global= 0.240:0.232[\u001b[31m0.008\u001b[39m]; local()= 0.116:0.159[\u001b[32m0.043\u001b[39m]; log p(e|m)= -1.505:-4.423[\u001b[31m2.918\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIreland[0.055]; Cup[0.044]; Cup[0.044]; soccer[0.043]; Republic[0.043]; Soccer[0.042]; Neill[0.042]; qualifier[0.042]; European[0.042]; Ian[0.042]; Qualifier[0.041]; Keith[0.041]; Quinn[0.041]; Saturday[0.041]; Scorers[0.040]; beat[0.040]; Niall[0.040]; Liechtenstein[0.040]; Liechtenstein[0.040]; World[0.040]; World[0.040]; Andy[0.040]; Harte[0.040]; group[0.039]; \t\n\r [===========================================>..] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 1s | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4590/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1124testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Seoul ==> ENTITY: \u001b[32mSeoul\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.161:0.161[0]; log p(e|m)= -0.038:-0.038[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHyundai[0.044]; Hyundai[0.044]; won[0.044]; Korean[0.043]; Korean[0.043]; games[0.043]; games[0.043]; games[0.043]; games[0.043]; Samsung[0.043]; Samsung[0.043]; played[0.042]; played[0.042]; South[0.041]; Haitai[0.041]; place[0.040]; pro[0.040]; pro[0.040]; Lotte[0.040]; Friday[0.040]; Friday[0.040]; Baseball[0.040]; winning[0.039]; Standings[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = South Korean\u001b[39m ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbaseball[0.048]; baseball[0.048]; Korean[0.047]; Baseball[0.047]; Hyundai[0.046]; Hyundai[0.046]; games[0.045]; games[0.045]; games[0.045]; games[0.045]; Seoul[0.045]; Haitai[0.043]; Samsung[0.043]; Samsung[0.043]; place[0.042]; played[0.042]; played[0.042]; won[0.042]; Lotte[0.041]; Hanwha[0.039]; Standings[0.039]; winning[0.039]; percentage[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hanwha ==> ENTITY: \u001b[32mHanwha Eagles\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.172:0.172[0]; log p(e|m)= -0.693:-0.693[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbaseball[0.047]; baseball[0.047]; Baseball[0.045]; Haitai[0.044]; games[0.044]; games[0.044]; games[0.044]; games[0.044]; Seoul[0.042]; won[0.041]; Samsung[0.041]; Samsung[0.041]; Korean[0.041]; Korean[0.041]; Lotte[0.041]; Hyundai[0.041]; Hyundai[0.041]; played[0.040]; played[0.040]; South[0.039]; winning[0.039]; Gb[0.038]; Lg[0.038]; Standings[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = S. Korean\u001b[39m ==> ENTITY: \u001b[32mSouth Korea\u001b[39m\t\nSCORES: global= 0.304:0.304[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSeoul[0.048]; Korean[0.047]; Hyundai[0.046]; Hyundai[0.046]; Samsung[0.042]; Samsung[0.042]; games[0.042]; games[0.042]; games[0.042]; games[0.042]; played[0.042]; played[0.042]; Haitai[0.042]; baseball[0.041]; baseball[0.041]; South[0.040]; place[0.040]; won[0.040]; Hanwha[0.040]; Lotte[0.040]; Gb[0.039]; Lg[0.039]; Friday[0.039]; Friday[0.039]; \t\n\r [===========================================>..] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 985ms | Step: 5ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4594/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1068testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Glasgow ==> ENTITY: \u001b[32mGlasgow\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.196:0.196[0]; log p(e|m)= -0.155:-0.155[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nScottish[0.045]; Scottish[0.045]; Fife[0.044]; Greenock[0.043]; Dumbarton[0.043]; Inverness[0.043]; Dundee[0.043]; league[0.043]; Airdrieonians[0.042]; matches[0.042]; Stirling[0.042]; Stirling[0.042]; League[0.041]; Cowdenbeath[0.041]; Ross[0.041]; Clydebank[0.041]; Hamilton[0.041]; Clyde[0.040]; Stranraer[0.040]; Partick[0.040]; County[0.040]; Queen[0.040]; Queen[0.040]; Falkirk[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mStranraer\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mStranraer F.C.\u001b[39m <---> \u001b[32mStranraer\u001b[39m\t\nSCORES: global= 0.284:0.276[\u001b[31m0.008\u001b[39m]; local()= 0.239:0.149[\u001b[31m0.089\u001b[39m]; log p(e|m)= -1.030:0.000[\u001b[32m1.030\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.045]; Livingston[0.043]; League[0.043]; Stenhousemuir[0.042]; Cowdenbeath[0.042]; Greenock[0.042]; Inverness[0.042]; Dundee[0.042]; Brechin[0.042]; Dumbarton[0.042]; Glasgow[0.042]; Division[0.041]; Division[0.041]; Division[0.041]; Johnstone[0.041]; Stirling[0.041]; Stirling[0.041]; Fife[0.041]; Thistle[0.041]; Partick[0.041]; Scottish[0.041]; Scottish[0.041]; Ross[0.041]; league[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAyr\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAyr\u001b[39m <---> \u001b[32mAyr United F.C.\u001b[39m\t\nSCORES: global= 0.267:0.255[\u001b[31m0.011\u001b[39m]; local()= 0.185:0.236[\u001b[32m0.051\u001b[39m]; log p(e|m)= -0.066:-3.352[\u001b[31m3.286\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.043]; Cowdenbeath[0.043]; Stenhousemuir[0.043]; Thistle[0.043]; Dundee[0.042]; League[0.042]; Fife[0.042]; matches[0.042]; Scottish[0.042]; Scottish[0.042]; Partick[0.042]; Livingston[0.042]; Greenock[0.042]; Inverness[0.042]; Clydebank[0.041]; Stirling[0.041]; Stirling[0.041]; league[0.041]; Dumbarton[0.041]; Stranraer[0.041]; Falkirk[0.041]; Brechin[0.041]; Forfar[0.041]; Arbroath[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ross County\u001b[39m ==> ENTITY: \u001b[32mRoss County F.C.\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.044]; League[0.043]; Fife[0.043]; Livingston[0.043]; Division[0.043]; Division[0.043]; Division[0.043]; league[0.042]; matches[0.042]; Scottish[0.042]; Scottish[0.042]; Inverness[0.042]; Dundee[0.042]; Cowdenbeath[0.042]; Stirling[0.041]; Stirling[0.041]; Greenock[0.041]; Glasgow[0.041]; Hamilton[0.041]; Stenhousemuir[0.041]; Park[0.040]; Falkirk[0.040]; Partick[0.040]; Clyde[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Partick ==> ENTITY: \u001b[32mPartick Thistle F.C.\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.229:0.229[0]; log p(e|m)= -1.284:-1.284[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; Cowdenbeath[0.044]; Stenhousemuir[0.044]; Livingston[0.044]; Dundee[0.043]; League[0.043]; matches[0.042]; Glasgow[0.042]; Falkirk[0.042]; league[0.042]; Greenock[0.041]; Thistle[0.041]; Dumbarton[0.041]; Inverness[0.041]; Stranraer[0.041]; Stirling[0.041]; Stirling[0.041]; Brechin[0.041]; Clyde[0.041]; Scottish[0.040]; Scottish[0.040]; Fife[0.040]; Ross[0.040]; Johnstone[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Livingston ==> ENTITY: \u001b[32mLivingston F.C.\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.194:0.194[0]; log p(e|m)= -1.864:-1.864[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; Fife[0.044]; League[0.043]; Inverness[0.043]; league[0.042]; Dundee[0.042]; Stenhousemuir[0.042]; Stirling[0.042]; Stirling[0.042]; Greenock[0.042]; County[0.042]; matches[0.042]; Dumbarton[0.041]; Cowdenbeath[0.041]; Falkirk[0.041]; Brechin[0.041]; Albion[0.041]; Scottish[0.041]; Scottish[0.041]; Glasgow[0.041]; Thistle[0.041]; Alloa[0.040]; Ross[0.040]; Partick[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Inverness Thistle\u001b[39m ==> ENTITY: \u001b[32mInverness Thistle F.C.\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; Livingston[0.044]; Stenhousemuir[0.043]; Dundee[0.043]; Cowdenbeath[0.043]; matches[0.042]; League[0.042]; Stirling[0.042]; Stirling[0.042]; Ross[0.042]; Scottish[0.042]; Scottish[0.042]; Glasgow[0.041]; league[0.041]; Greenock[0.041]; Park[0.041]; Alloa[0.041]; Falkirk[0.041]; Partick[0.041]; Clydebank[0.041]; Dumbarton[0.040]; Ayr[0.040]; Stranraer[0.040]; Forfar[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clyde ==> ENTITY: \u001b[32mClyde F.C.\u001b[39m\t\nSCORES: global= 0.287:0.287[0]; local()= 0.209:0.209[0]; log p(e|m)= -0.983:-0.983[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; County[0.044]; League[0.043]; Livingston[0.043]; Dundee[0.043]; Stenhousemuir[0.042]; Cowdenbeath[0.042]; Park[0.042]; Stirling[0.042]; Stirling[0.042]; Glasgow[0.041]; Partick[0.041]; Greenock[0.041]; matches[0.041]; Fife[0.041]; Thistle[0.041]; Brechin[0.041]; Dumbarton[0.041]; Inverness[0.041]; Alloa[0.041]; Ross[0.040]; Falkirk[0.040]; league[0.040]; Clydebank[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clydebank ==> ENTITY: \u001b[32mClydebank F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.237:0.237[0]; log p(e|m)= -1.328:-1.328[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; Cowdenbeath[0.044]; Dundee[0.043]; Stenhousemuir[0.043]; Greenock[0.043]; Fife[0.042]; Stirling[0.042]; Stirling[0.042]; Glasgow[0.042]; Inverness[0.041]; matches[0.041]; Alloa[0.041]; Brechin[0.041]; Partick[0.041]; Dumbarton[0.041]; Arbroath[0.041]; league[0.041]; League[0.041]; Livingston[0.041]; Stranraer[0.041]; Scottish[0.041]; Scottish[0.041]; Forfar[0.041]; Thistle[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = St Mirren\u001b[39m ==> ENTITY: \u001b[32mSt. Mirren F.C.\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.044]; matches[0.044]; league[0.044]; League[0.044]; Stenhousemuir[0.043]; Dundee[0.043]; Cowdenbeath[0.043]; Livingston[0.042]; Partick[0.042]; Park[0.041]; Brechin[0.041]; Albion[0.041]; Stirling[0.041]; Stirling[0.041]; Glasgow[0.041]; Soccer[0.041]; Thistle[0.041]; Falkirk[0.041]; Dumbarton[0.040]; Greenock[0.040]; Fife[0.040]; Scottish[0.040]; Scottish[0.040]; Stranraer[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cowdenbeath ==> ENTITY: \u001b[32mCowdenbeath F.C.\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.223:0.223[0]; log p(e|m)= -0.247:-0.247[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.045]; Stenhousemuir[0.045]; Fife[0.043]; Dundee[0.043]; Livingston[0.043]; Brechin[0.042]; matches[0.042]; Inverness[0.042]; Scottish[0.042]; Scottish[0.042]; League[0.041]; Greenock[0.041]; Partick[0.041]; Glasgow[0.041]; Dumbarton[0.041]; Stranraer[0.041]; Stirling[0.041]; Stirling[0.041]; league[0.041]; Division[0.040]; Division[0.040]; Division[0.040]; Thistle[0.040]; Park[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scottish ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.168:0.168[0]; log p(e|m)= -0.480:-0.480[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.047]; Scottish[0.046]; Stenhousemuir[0.045]; League[0.045]; Queen[0.044]; Queen[0.044]; Cowdenbeath[0.044]; Livingston[0.044]; matches[0.044]; Dundee[0.044]; league[0.044]; Fife[0.043]; Greenock[0.043]; Glasgow[0.043]; Inverness[0.043]; Falkirk[0.043]; Ross[0.042]; Soccer[0.042]; Park[0.042]; Dumbarton[0.042]; Thistle[0.042]; Partick[0.042]; Brechin[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Forfar ==> ENTITY: \u001b[32mForfar Athletic F.C.\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.228:0.228[0]; log p(e|m)= -1.766:-1.766[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStenhousemuir[0.044]; Cowdenbeath[0.043]; Airdrieonians[0.043]; Inverness[0.042]; Scottish[0.042]; Scottish[0.042]; Arbroath[0.042]; Fife[0.042]; Dundee[0.042]; Greenock[0.042]; league[0.042]; Brechin[0.042]; Alloa[0.042]; Partick[0.041]; Stranraer[0.041]; League[0.041]; Stirling[0.041]; Stirling[0.041]; Clydebank[0.041]; Thistle[0.041]; Dumbarton[0.041]; Falkirk[0.040]; matches[0.040]; Johnstone[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Airdrieonians\u001b[39m ==> ENTITY: \u001b[32mAirdrieonians F.C.\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCowdenbeath[0.044]; Stenhousemuir[0.043]; Livingston[0.043]; Dundee[0.043]; Brechin[0.042]; Thistle[0.042]; Inverness[0.042]; matches[0.042]; Scottish[0.042]; Scottish[0.042]; Glasgow[0.042]; Partick[0.042]; League[0.042]; Stirling[0.042]; Stirling[0.042]; Stranraer[0.041]; league[0.041]; Forfar[0.041]; Clydebank[0.041]; Greenock[0.041]; Falkirk[0.041]; Alloa[0.040]; Dumbarton[0.040]; Fife[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAlloa\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAlloa\u001b[39m <---> \u001b[32mAlloa Athletic F.C.\u001b[39m\t\nSCORES: global= 0.289:0.285[\u001b[31m0.004\u001b[39m]; local()= 0.172:0.215[\u001b[32m0.043\u001b[39m]; log p(e|m)= 0.000:-0.794[\u001b[31m0.794\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.044]; Cowdenbeath[0.043]; Stenhousemuir[0.043]; Stirling[0.043]; Stirling[0.043]; Livingston[0.043]; Dundee[0.043]; League[0.043]; league[0.042]; Greenock[0.042]; Brechin[0.042]; Inverness[0.042]; matches[0.042]; Glasgow[0.041]; Thistle[0.041]; Arbroath[0.041]; Partick[0.041]; Dumbarton[0.041]; Falkirk[0.041]; Stranraer[0.041]; Fife[0.040]; Forfar[0.040]; Scottish[0.040]; Scottish[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dumbarton ==> ENTITY: \u001b[32mDumbarton F.C.\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.218:0.218[0]; log p(e|m)= -1.255:-1.255[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStenhousemuir[0.046]; Fife[0.045]; Cowdenbeath[0.045]; Airdrieonians[0.045]; Stranraer[0.044]; Partick[0.044]; Scottish[0.044]; Scottish[0.044]; Arbroath[0.044]; Greenock[0.044]; matches[0.043]; Division[0.043]; Division[0.043]; Division[0.043]; league[0.043]; Alloa[0.043]; League[0.043]; Brechin[0.043]; Inverness[0.042]; Falkirk[0.042]; Clyde[0.042]; Dundee[0.042]; Livingston[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Arbroath ==> ENTITY: \u001b[32mArbroath F.C.\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.194:0.194[0]; log p(e|m)= -0.860:-0.860[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStenhousemuir[0.044]; Dundee[0.044]; Airdrieonians[0.043]; Fife[0.043]; Livingston[0.043]; Cowdenbeath[0.042]; Scottish[0.042]; Scottish[0.042]; Greenock[0.042]; league[0.042]; Inverness[0.042]; League[0.042]; Brechin[0.041]; Dumbarton[0.041]; Ross[0.041]; Alloa[0.041]; Partick[0.041]; Glasgow[0.041]; Forfar[0.041]; Stranraer[0.041]; matches[0.040]; Stirling[0.040]; Stirling[0.040]; Park[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Berwick ==> ENTITY: \u001b[32mBerwick Rangers F.C.\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.182:0.182[0]; log p(e|m)= -3.689:-3.689[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLivingston[0.044]; Airdrieonians[0.044]; League[0.044]; Park[0.043]; Fife[0.043]; league[0.043]; matches[0.042]; Dundee[0.042]; Stirling[0.042]; Stirling[0.042]; Cowdenbeath[0.042]; Stenhousemuir[0.042]; Greenock[0.042]; Inverness[0.041]; Ross[0.041]; Clyde[0.041]; Scottish[0.041]; Scottish[0.041]; Brechin[0.040]; Hamilton[0.040]; Queen[0.040]; Queen[0.040]; Partick[0.040]; Glasgow[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = East Fife\u001b[39m ==> ENTITY: \u001b[32mEast Fife F.C.\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.207:0.207[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.047]; Cowdenbeath[0.046]; Stenhousemuir[0.046]; Dundee[0.044]; matches[0.044]; Alloa[0.044]; Dumbarton[0.044]; Glasgow[0.043]; Partick[0.043]; Inverness[0.043]; Brechin[0.043]; Clydebank[0.043]; Stirling[0.043]; Stirling[0.043]; Arbroath[0.043]; Livingston[0.043]; Stranraer[0.043]; Greenock[0.043]; Scottish[0.043]; Scottish[0.043]; Falkirk[0.042]; League[0.042]; Park[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scottish ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.168:0.168[0]; log p(e|m)= -0.480:-0.480[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.047]; Scottish[0.046]; Stenhousemuir[0.045]; League[0.045]; Queen[0.044]; Queen[0.044]; Cowdenbeath[0.044]; Livingston[0.044]; matches[0.044]; Dundee[0.044]; league[0.044]; Fife[0.043]; Greenock[0.043]; Glasgow[0.043]; Inverness[0.043]; Falkirk[0.043]; Ross[0.042]; Soccer[0.042]; Park[0.042]; Dumbarton[0.042]; Thistle[0.042]; Partick[0.042]; Brechin[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Greenock Morton\u001b[39m ==> ENTITY: \u001b[32mGreenock Morton F.C.\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.216:0.216[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.044]; Cowdenbeath[0.043]; Stenhousemuir[0.043]; League[0.043]; Scottish[0.043]; Scottish[0.043]; Partick[0.042]; matches[0.042]; league[0.042]; Livingston[0.042]; Dundee[0.042]; Glasgow[0.041]; Brechin[0.041]; Fife[0.041]; Inverness[0.041]; Stranraer[0.041]; Forfar[0.041]; Dumbarton[0.041]; Clydebank[0.041]; Stirling[0.041]; Stirling[0.041]; Alloa[0.040]; Arbroath[0.040]; Thistle[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Stenhousemuir\u001b[39m ==> ENTITY: \u001b[32mStenhousemuir F.C.\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.199:0.199[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; League[0.045]; Park[0.043]; league[0.043]; matches[0.042]; Stirling[0.042]; Stirling[0.042]; Cowdenbeath[0.042]; Dumbarton[0.042]; Fife[0.042]; Livingston[0.042]; Inverness[0.041]; Thistle[0.041]; Alloa[0.041]; Scottish[0.041]; Scottish[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Dundee[0.041]; Partick[0.041]; Greenock[0.041]; Johnstone[0.040]; Falkirk[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Montrose ==> ENTITY: \u001b[32mMontrose F.C.\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.198:0.198[0]; log p(e|m)= -2.180:-2.180[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.045]; Cowdenbeath[0.044]; Greenock[0.043]; Stenhousemuir[0.043]; Brechin[0.043]; Inverness[0.042]; Scottish[0.042]; Scottish[0.042]; Fife[0.042]; Park[0.042]; Alloa[0.042]; Stirling[0.041]; Stirling[0.041]; Livingston[0.041]; matches[0.041]; Forfar[0.041]; League[0.041]; Dumbarton[0.041]; Thistle[0.041]; Dundee[0.040]; Ross[0.040]; Partick[0.040]; Stranraer[0.040]; Arbroath[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brechin ==> ENTITY: \u001b[32mBrechin City F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.195:0.195[0]; log p(e|m)= -1.917:-1.917[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; Forfar[0.044]; Ayr[0.044]; Stenhousemuir[0.044]; Arbroath[0.042]; Cowdenbeath[0.042]; Alloa[0.042]; Dumbarton[0.042]; League[0.042]; Partick[0.042]; Stranraer[0.042]; Inverness[0.041]; Scottish[0.041]; Scottish[0.041]; Dundee[0.041]; Thistle[0.041]; Greenock[0.041]; Livingston[0.041]; Fife[0.040]; Stirling[0.040]; Stirling[0.040]; Clydebank[0.040]; Park[0.040]; Falkirk[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = St Johnstone\u001b[39m ==> ENTITY: \u001b[32mSt. Johnstone F.C.\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.197:0.197[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStenhousemuir[0.045]; Airdrieonians[0.045]; League[0.044]; Cowdenbeath[0.044]; Livingston[0.043]; Dundee[0.043]; league[0.042]; Partick[0.042]; Stirling[0.041]; Stirling[0.041]; Ross[0.041]; Dumbarton[0.041]; Inverness[0.041]; Falkirk[0.041]; Albion[0.041]; matches[0.041]; Park[0.041]; Thistle[0.041]; Alloa[0.041]; Scottish[0.041]; Scottish[0.041]; Brechin[0.041]; Glasgow[0.040]; Forfar[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = East Stirling\u001b[39m ==> ENTITY: \u001b[32mEast Stirlingshire F.C.\u001b[39m\t\nSCORES: global= 0.309:0.309[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; Cowdenbeath[0.044]; Stenhousemuir[0.043]; matches[0.043]; league[0.042]; Dundee[0.042]; Partick[0.042]; League[0.042]; Fife[0.042]; Stranraer[0.042]; Brechin[0.041]; Inverness[0.041]; Livingston[0.041]; Glasgow[0.041]; Scottish[0.041]; Scottish[0.041]; Dumbarton[0.041]; Greenock[0.041]; Thistle[0.041]; Park[0.041]; Alloa[0.041]; Forfar[0.040]; Clydebank[0.040]; Stirling[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Falkirk ==> ENTITY: \u001b[32mFalkirk F.C.\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.196:0.196[0]; log p(e|m)= -1.204:-1.204[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStenhousemuir[0.047]; Cowdenbeath[0.046]; Airdrieonians[0.046]; League[0.045]; Fife[0.044]; Scottish[0.044]; Scottish[0.044]; Ayr[0.044]; Alloa[0.044]; Stranraer[0.043]; Brechin[0.043]; Arbroath[0.043]; Inverness[0.043]; Division[0.043]; Division[0.043]; Division[0.043]; Dundee[0.042]; Partick[0.042]; Clydebank[0.042]; Forfar[0.042]; Livingston[0.042]; Dumbarton[0.042]; league[0.042]; \t\n\r [===========================================>..] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 831ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4624/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1016testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = New York ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.041:0.041[0]; log p(e|m)= -1.790:-1.790[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nships[0.047]; ship[0.046]; Boston[0.044]; coast[0.043]; Titanic[0.043]; Titanic[0.043]; Titanic[0.043]; Friday[0.042]; Canada[0.041]; said[0.041]; said[0.041]; sea[0.041]; scrapped[0.041]; piece[0.041]; piece[0.041]; based[0.041]; ocean[0.041]; ocean[0.041]; expedition[0.040]; expedition[0.040]; hull[0.040]; hull[0.040]; lifted[0.040]; Newfoundland[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Titanic ==> ENTITY: \u001b[32mRMS Titanic\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.141:0.141[0]; log p(e|m)= -0.671:-0.671[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTitanic[0.049]; Titanic[0.049]; ship[0.046]; ships[0.045]; hull[0.042]; hull[0.042]; crews[0.042]; piece[0.041]; piece[0.041]; piece[0.041]; represents[0.041]; spokeswoman[0.041]; stuck[0.040]; giant[0.040]; Friday[0.040]; said[0.040]; said[0.040]; said[0.040]; lifted[0.040]; expedition[0.040]; expedition[0.040]; problems[0.040]; failure[0.040]; coast[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Titanic ==> ENTITY: \u001b[32mRMS Titanic\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.156:0.156[0]; log p(e|m)= -0.671:-0.671[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nship[0.046]; ships[0.045]; ships[0.045]; sank[0.044]; crew[0.043]; unsinkable[0.043]; hull[0.042]; hulled[0.042]; crews[0.041]; wreck[0.041]; piece[0.041]; piece[0.041]; returned[0.041]; Thursday[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; board[0.041]; cruise[0.040]; stuck[0.040]; April[0.040]; iceberg[0.040]; burst[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Boston-based\u001b[39m ==> ENTITY: \u001b[32mBoston\u001b[39m\t\nSCORES: global= 0.305:0.305[0]; local()= 0.060:0.060[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nYork[0.046]; Canada[0.044]; based[0.044]; April[0.044]; Rms[0.044]; Equipment[0.043]; Friday[0.042]; Newfoundland[0.042]; scrapped[0.041]; Erin[0.041]; said[0.041]; said[0.041]; said[0.041]; said[0.041]; cables[0.041]; cables[0.041]; ship[0.041]; coast[0.041]; attached[0.041]; mission[0.041]; diesel[0.040]; steel[0.040]; steel[0.040]; haul[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = RMS Titanic\u001b[39m ==> ENTITY: \u001b[32mRMS Titanic\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.133:0.133[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTitanic[0.051]; Titanic[0.051]; ship[0.046]; ships[0.043]; expedition[0.041]; expedition[0.041]; Equipment[0.041]; said[0.041]; said[0.041]; said[0.041]; coast[0.041]; hull[0.041]; hull[0.041]; scrapped[0.041]; Canada[0.040]; ocean[0.040]; ocean[0.040]; stuck[0.040]; spokeswoman[0.040]; ton[0.040]; Erin[0.040]; failure[0.040]; cables[0.039]; giant[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.064:0.064[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNewfoundland[0.049]; Friday[0.046]; April[0.045]; coast[0.044]; Boston[0.043]; York[0.043]; mission[0.042]; based[0.042]; scrapped[0.042]; ships[0.041]; filled[0.041]; ship[0.041]; expedition[0.041]; expedition[0.041]; said[0.040]; said[0.040]; said[0.040]; said[0.040]; means[0.040]; haul[0.040]; stuck[0.040]; fell[0.040]; fell[0.040]; fell[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mNewfoundland\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mNewfoundland and Labrador\u001b[39m <---> \u001b[32mNewfoundland\u001b[39m\t\nSCORES: global= 0.259:0.257[\u001b[31m0.003\u001b[39m]; local()= 0.088:0.144[\u001b[32m0.057\u001b[39m]; log p(e|m)= -1.328:-0.730[\u001b[32m0.598\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncoast[0.050]; Canada[0.048]; ships[0.044]; unsinkable[0.043]; ship[0.043]; ocean[0.042]; ocean[0.042]; sea[0.042]; Titanic[0.042]; Titanic[0.042]; Titanic[0.042]; Titanic[0.042]; Friday[0.041]; hull[0.041]; hull[0.041]; feet[0.041]; Boston[0.041]; crews[0.040]; expedition[0.040]; expedition[0.040]; based[0.039]; means[0.039]; ton[0.039]; mission[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Titanic ==> ENTITY: \u001b[32mRMS Titanic\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.143:0.143[0]; log p(e|m)= -0.671:-0.671[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nTitanic[0.052]; Titanic[0.052]; ship[0.048]; ships[0.047]; hull[0.044]; hull[0.044]; piece[0.043]; piece[0.043]; represents[0.043]; spokeswoman[0.043]; giant[0.042]; Friday[0.042]; said[0.042]; said[0.042]; lifted[0.042]; expedition[0.042]; expedition[0.042]; problems[0.042]; failure[0.041]; coast[0.041]; scrapped[0.041]; Canada[0.041]; efforts[0.041]; \t\n\r [===========================================>..] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 792ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4632/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1130testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.050]; Russian[0.047]; Russian[0.047]; Russia[0.047]; military[0.045]; servicemen[0.045]; servicemen[0.045]; outside[0.044]; Interfax[0.044]; place[0.043]; officials[0.043]; officials[0.043]; took[0.043]; Saturday[0.042]; miles[0.042]; attack[0.042]; attack[0.042]; capital[0.042]; soldiers[0.041]; seized[0.041]; aircraft[0.041]; seizing[0.041]; number[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Moscow ==> ENTITY: \u001b[32mMoscow\u001b[39m\t\nSCORES: global= 0.268:0.268[0]; local()= 0.123:0.123[0]; log p(e|m)= -0.106:-0.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.050]; Russian[0.047]; Russian[0.047]; Russia[0.047]; military[0.045]; servicemen[0.045]; servicemen[0.045]; outside[0.044]; Interfax[0.044]; place[0.043]; officials[0.043]; officials[0.043]; took[0.043]; Saturday[0.042]; miles[0.042]; attack[0.042]; attack[0.042]; capital[0.042]; soldiers[0.041]; seized[0.041]; aircraft[0.041]; seizing[0.041]; number[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russia ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.128:0.128[0]; log p(e|m)= -0.203:-0.203[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.046]; Russian[0.046]; Moscow[0.045]; Moscow[0.045]; military[0.044]; Interfax[0.043]; servicemen[0.043]; servicemen[0.043]; officials[0.042]; officials[0.042]; soldiers[0.042]; agency[0.041]; said[0.040]; said[0.040]; aircraft[0.040]; took[0.040]; Kalashnikov[0.040]; news[0.040]; guns[0.040]; Saturday[0.040]; place[0.040]; outside[0.039]; aimed[0.039]; number[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mKalashnikov\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mMikhail Kalashnikov\u001b[39m <---> \u001b[32mAK-47\u001b[39m\t\nSCORES: global= 0.254:0.249[\u001b[31m0.005\u001b[39m]; local()= 0.126:0.115[\u001b[31m0.011\u001b[39m]; log p(e|m)= -1.355:-0.131[\u001b[32m1.224\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nrifles[0.048]; guns[0.048]; gun[0.047]; Russian[0.047]; Russian[0.047]; Russia[0.046]; Moscow[0.045]; Moscow[0.045]; use[0.044]; Posad[0.043]; attackers[0.043]; attack[0.042]; attack[0.042]; gunmen[0.042]; aimed[0.042]; soldiers[0.041]; military[0.041]; outside[0.041]; installation[0.041]; assault[0.041]; Interfax[0.041]; guards[0.040]; attacked[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.259:0.259[0]; local()= 0.121:0.121[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.049]; Russia[0.048]; Moscow[0.045]; Moscow[0.045]; military[0.044]; servicemen[0.043]; servicemen[0.043]; officials[0.042]; officials[0.042]; soldiers[0.042]; aircraft[0.040]; Interfax[0.040]; crimes[0.040]; number[0.040]; seized[0.040]; Kalashnikov[0.040]; guns[0.040]; said[0.039]; said[0.039]; growing[0.039]; guards[0.039]; took[0.039]; committed[0.039]; place[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Sergiyev Posad\u001b[39m ==> ENTITY: \u001b[32mSergiyev Posad\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.134:0.134[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nMoscow[0.045]; Moscow[0.045]; officials[0.043]; officials[0.043]; Russia[0.043]; disappeared[0.043]; Russian[0.043]; Russian[0.043]; seized[0.042]; Interfax[0.042]; military[0.042]; servicemen[0.042]; servicemen[0.042]; agency[0.042]; killed[0.042]; killed[0.042]; growing[0.040]; said[0.040]; said[0.040]; took[0.040]; capital[0.040]; km[0.039]; soldiers[0.039]; crimes[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Interfax\u001b[39m ==> ENTITY: \u001b[32mInterfax\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnews[0.048]; Moscow[0.048]; Moscow[0.048]; Russian[0.047]; Russian[0.047]; Russia[0.044]; agency[0.044]; quoted[0.043]; unidentified[0.043]; officials[0.042]; officials[0.042]; saying[0.042]; Posad[0.040]; military[0.040]; Kalashnikov[0.039]; aimed[0.039]; said[0.039]; said[0.039]; anti[0.039]; seizing[0.038]; servicemen[0.038]; servicemen[0.038]; km[0.038]; miles[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Russian ==> ENTITY: \u001b[32mRussia\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.127:0.127[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nRussian[0.049]; Russia[0.048]; Moscow[0.046]; Moscow[0.046]; military[0.045]; servicemen[0.044]; servicemen[0.044]; officials[0.042]; officials[0.042]; soldiers[0.042]; aircraft[0.040]; Interfax[0.040]; number[0.040]; seized[0.040]; Kalashnikov[0.040]; guns[0.040]; said[0.039]; said[0.039]; guards[0.039]; took[0.039]; place[0.039]; outside[0.039]; attacked[0.039]; quoted[0.039]; \t\n\r [===========================================>..] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 746ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4640/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1086testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Estonia ==> ENTITY: \u001b[32mEstonia national football team\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.142:0.142[0]; log p(e|m)= -1.924:-1.924[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEstonia[0.051]; soccer[0.050]; qualifier[0.050]; Qualifier[0.048]; Belarus[0.048]; Belarus[0.048]; Soccer[0.048]; European[0.048]; Cup[0.047]; Cup[0.047]; Vladimir[0.047]; World[0.045]; World[0.045]; group[0.044]; beat[0.044]; Makovsky[0.044]; Minsk[0.042]; Saturday[0.042]; halftime[0.041]; Scorer[0.041]; Beat[0.041]; Attendance[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Minsk ==> ENTITY: \u001b[32mMinsk\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.103:0.103[0]; log p(e|m)= -0.071:-0.071[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.050]; Belarus[0.049]; Belarus[0.049]; Vladimir[0.049]; soccer[0.049]; Saturday[0.047]; Cup[0.046]; Cup[0.046]; Estonia[0.046]; Estonia[0.046]; beat[0.045]; qualifier[0.045]; Beat[0.045]; European[0.045]; World[0.044]; World[0.044]; Qualifier[0.044]; Attendance[0.043]; group[0.043]; halftime[0.042]; Scorer[0.041]; Makovsky[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = European ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.242:0.242[0]; local()= 0.031:0.031[0]; log p(e|m)= -0.256:-0.256[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nsoccer[0.052]; Soccer[0.051]; Cup[0.049]; Cup[0.049]; qualifier[0.049]; Qualifier[0.048]; group[0.047]; Estonia[0.047]; Estonia[0.047]; Belarus[0.046]; Belarus[0.046]; beat[0.045]; Saturday[0.045]; World[0.044]; World[0.044]; Attendance[0.043]; Scorer[0.042]; Minsk[0.042]; Makovsky[0.041]; Vladimir[0.041]; Beat[0.041]; halftime[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Estonia ==> ENTITY: \u001b[32mEstonia national football team\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.142:0.142[0]; log p(e|m)= -1.924:-1.924[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEstonia[0.051]; soccer[0.050]; qualifier[0.050]; Qualifier[0.048]; Belarus[0.048]; Belarus[0.048]; Soccer[0.048]; European[0.048]; Cup[0.047]; Cup[0.047]; Vladimir[0.047]; World[0.045]; World[0.045]; group[0.044]; beat[0.044]; Makovsky[0.044]; Minsk[0.042]; Saturday[0.042]; halftime[0.041]; Scorer[0.041]; Beat[0.041]; Attendance[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belarus ==> ENTITY: \u001b[32mBelarus national football team\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.149:0.149[0]; log p(e|m)= -2.025:-2.025[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWorld[0.055]; World[0.055]; Belarus[0.051]; Estonia[0.049]; Estonia[0.049]; qualifier[0.048]; Qualifier[0.048]; soccer[0.047]; European[0.046]; Soccer[0.046]; beat[0.044]; Cup[0.044]; Cup[0.044]; Vladimir[0.044]; group[0.044]; Minsk[0.043]; Scorer[0.042]; Saturday[0.042]; Beat[0.040]; halftime[0.040]; Makovsky[0.038]; Attendance[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Belarus ==> ENTITY: \u001b[32mBelarus national football team\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.149:0.149[0]; log p(e|m)= -2.025:-2.025[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWorld[0.055]; World[0.055]; Belarus[0.051]; Estonia[0.049]; Estonia[0.049]; qualifier[0.048]; Qualifier[0.048]; soccer[0.047]; European[0.046]; Soccer[0.046]; beat[0.044]; Cup[0.044]; Cup[0.044]; Vladimir[0.044]; group[0.044]; Minsk[0.043]; Scorer[0.042]; Saturday[0.042]; Beat[0.040]; halftime[0.040]; Makovsky[0.038]; Attendance[0.038]; \t\n\r [===========================================>..] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 707ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4648/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 947testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.253:0.253[0]; local()= 0.047:0.047[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSurrey[0.046]; Essex[0.045]; Somerset[0.044]; Somerset[0.044]; Leicestershire[0.044]; Leicestershire[0.044]; Leicestershire[0.044]; match[0.043]; bowling[0.042]; Nottinghamshire[0.042]; Derbyshire[0.042]; Cricket[0.042]; Kent[0.042]; championship[0.041]; runs[0.040]; runs[0.040]; Road[0.040]; beat[0.040]; West[0.040]; Friday[0.039]; title[0.039]; county[0.039]; lost[0.039]; opening[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland cricket team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.174:0.174[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; England[0.047]; Warwickshire[0.043]; season[0.043]; match[0.043]; match[0.043]; Oval[0.043]; Essex[0.043]; Worcestershire[0.042]; Surrey[0.041]; Surrey[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; day[0.041]; Headingley[0.041]; second[0.041]; spinner[0.041]; captain[0.040]; Yorkshire[0.040]; Yorkshire[0.040]; Australian[0.040]; defeat[0.039]; Hussain[0.039]; Hussain[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Yorkshire ==> ENTITY: \u001b[32mYorkshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.210:0.210[0]; log p(e|m)= -1.778:-1.778[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeicestershire[0.046]; England[0.046]; England[0.046]; England[0.046]; Essex[0.046]; Essex[0.046]; bowling[0.044]; Somerset[0.044]; Somerset[0.044]; Nottinghamshire[0.044]; Headingley[0.044]; season[0.044]; Surrey[0.043]; match[0.043]; match[0.043]; Yorkshire[0.043]; Oval[0.042]; bowled[0.042]; championship[0.041]; Caddick[0.041]; balls[0.041]; Kent[0.041]; second[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Warwickshire ==> ENTITY: \u001b[32mWarwickshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.178:0.178[0]; log p(e|m)= -1.109:-1.109[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.045]; England[0.045]; Worcestershire[0.044]; Essex[0.044]; second[0.042]; Derbyshire[0.042]; Derbyshire[0.042]; match[0.042]; match[0.042]; Kent[0.041]; Surrey[0.041]; Surrey[0.041]; balls[0.041]; Chris[0.041]; Chris[0.041]; innings[0.041]; innings[0.041]; innings[0.041]; innings[0.041]; Mark[0.041]; Yorkshire[0.040]; runs[0.040]; Oval[0.040]; seven[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leicestershire ==> ENTITY: \u001b[32mLeicestershire County Cricket Club\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.190:0.190[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbowling[0.044]; bowled[0.044]; Leicestershire[0.044]; Leicestershire[0.044]; England[0.042]; England[0.042]; Essex[0.042]; Essex[0.042]; Essex[0.042]; match[0.042]; match[0.042]; Somerset[0.042]; Somerset[0.042]; Somerset[0.042]; Nottinghamshire[0.041]; Derbyshire[0.041]; Surrey[0.041]; innings[0.040]; innings[0.040]; innings[0.040]; innings[0.040]; Headingley[0.040]; Kent[0.040]; rounder[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Martin McCague\u001b[39m ==> ENTITY: \u001b[32mMartin McCague\u001b[39m\t\nSCORES: global= 0.294:0.294[0]; local()= 0.120:0.120[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; match[0.046]; Worcestershire[0.044]; took[0.043]; took[0.043]; took[0.043]; bowler[0.042]; Kent[0.042]; Kent[0.042]; Derbyshire[0.042]; day[0.042]; Nottinghamshire[0.041]; Australian[0.041]; innings[0.041]; innings[0.041]; innings[0.041]; Tim[0.040]; opening[0.040]; Adams[0.040]; runs[0.040]; championship[0.040]; second[0.040]; seeing[0.040]; stepped[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Surrey ==> ENTITY: \u001b[32mSurrey County Cricket Club\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.199:0.199[0]; log p(e|m)= -1.796:-1.796[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEssex[0.045]; Essex[0.045]; Leicestershire[0.044]; Leicestershire[0.044]; Leicestershire[0.044]; England[0.044]; Somerset[0.044]; Somerset[0.044]; Somerset[0.044]; Nottinghamshire[0.042]; Derbyshire[0.042]; Kent[0.042]; bowling[0.041]; match[0.041]; match[0.041]; rounder[0.040]; Cricket[0.040]; West[0.040]; second[0.040]; Hussain[0.039]; county[0.039]; bowled[0.039]; London[0.039]; Caddick[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Oval ==> ENTITY: \u001b[32mThe Oval\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.203:0.203[0]; log p(e|m)= -2.163:-2.163[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.046]; England[0.046]; England[0.046]; Surrey[0.043]; Surrey[0.043]; match[0.043]; match[0.043]; Warwickshire[0.043]; Essex[0.043]; Essex[0.043]; balls[0.042]; second[0.041]; Headingley[0.041]; Derbyshire[0.041]; Worcestershire[0.041]; Australian[0.041]; day[0.040]; innings[0.040]; innings[0.040]; innings[0.040]; Yorkshire[0.039]; Yorkshire[0.039]; runs[0.039]; season[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Somerset ==> ENTITY: \u001b[32mSomerset County Cricket Club\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.199:0.199[0]; log p(e|m)= -2.419:-2.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.046]; Essex[0.045]; Somerset[0.045]; Somerset[0.045]; Leicestershire[0.043]; Leicestershire[0.043]; Leicestershire[0.043]; bowling[0.043]; Surrey[0.043]; match[0.042]; Nottinghamshire[0.042]; Derbyshire[0.041]; Kent[0.041]; Cricket[0.041]; bowled[0.041]; rounder[0.040]; county[0.040]; Caddick[0.040]; innings[0.040]; innings[0.040]; runs[0.039]; runs[0.039]; Andy[0.039]; Road[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Essex ==> ENTITY: \u001b[32mEssex County Cricket Club\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.214:0.214[0]; log p(e|m)= -2.343:-2.343[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.044]; England[0.044]; Essex[0.044]; Essex[0.044]; Leicestershire[0.044]; Somerset[0.043]; Somerset[0.043]; bowling[0.043]; match[0.042]; match[0.042]; Surrey[0.042]; Surrey[0.042]; Nottinghamshire[0.042]; Derbyshire[0.042]; Headingley[0.041]; Kent[0.041]; Oval[0.041]; Yorkshire[0.040]; Yorkshire[0.040]; bowled[0.040]; spinner[0.040]; innings[0.040]; innings[0.040]; innings[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kent ==> ENTITY: \u001b[32mKent County Cricket Club\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.168:0.168[0]; log p(e|m)= -2.513:-2.513[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.049]; Worcestershire[0.046]; Kent[0.045]; Derbyshire[0.044]; bowler[0.044]; match[0.044]; Nottinghamshire[0.044]; second[0.042]; Martin[0.042]; day[0.041]; Adams[0.041]; took[0.041]; took[0.041]; took[0.041]; runs[0.040]; innings[0.040]; innings[0.040]; innings[0.040]; opening[0.040]; Chris[0.040]; Paul[0.040]; stumps[0.039]; Tim[0.039]; Johnson[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Tom Moody\u001b[39m ==> ENTITY: \u001b[32mTom Moody\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.141:0.141[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; Oval[0.044]; bowler[0.044]; Worcestershire[0.043]; match[0.042]; match[0.042]; England[0.042]; England[0.042]; England[0.042]; innings[0.042]; innings[0.042]; innings[0.042]; Nottinghamshire[0.042]; captain[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Thursday[0.041]; Australian[0.041]; Surrey[0.040]; Surrey[0.040]; Kent[0.040]; Kent[0.040]; seven[0.040]; Friday[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chris Adams ==> ENTITY: \u001b[32mChris Adams (cricketer)\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.167:0.167[0]; log p(e|m)= -1.053:-1.053[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; bowler[0.044]; Worcestershire[0.042]; stumps[0.042]; captain[0.042]; England[0.042]; England[0.042]; England[0.042]; Nottinghamshire[0.042]; match[0.042]; match[0.042]; Australian[0.041]; innings[0.041]; innings[0.041]; innings[0.041]; Surrey[0.041]; Surrey[0.041]; Derbyshire[0.041]; Derbyshire[0.041]; Martin[0.041]; runs[0.041]; Kent[0.041]; Kent[0.041]; second[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Essex ==> ENTITY: \u001b[32mEssex County Cricket Club\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.211:0.211[0]; log p(e|m)= -2.343:-2.343[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.045]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; Essex[0.044]; Leicestershire[0.044]; Somerset[0.043]; match[0.042]; match[0.042]; Surrey[0.042]; Surrey[0.042]; Headingley[0.041]; Oval[0.041]; Yorkshire[0.040]; Yorkshire[0.040]; bowled[0.040]; spinner[0.040]; innings[0.040]; innings[0.040]; innings[0.040]; season[0.040]; day[0.039]; Caddick[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Andy Caddick\u001b[39m ==> ENTITY: \u001b[32mAndrew Caddick\u001b[39m\t\nSCORES: global= 0.310:0.310[0]; local()= 0.182:0.182[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSomerset[0.046]; Somerset[0.046]; Somerset[0.046]; Essex[0.045]; Essex[0.045]; Essex[0.045]; bowling[0.045]; Leicestershire[0.044]; Leicestershire[0.044]; Derbyshire[0.044]; Nottinghamshire[0.043]; England[0.043]; England[0.043]; second[0.043]; spinner[0.043]; rounder[0.043]; bowled[0.042]; Surrey[0.042]; Indian[0.042]; match[0.042]; match[0.042]; Kent[0.042]; Headingley[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nottinghamshire ==> ENTITY: \u001b[32mNottinghamshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.264:0.264[0]; local()= 0.203:0.203[0]; log p(e|m)= -1.106:-1.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEssex[0.046]; Essex[0.046]; England[0.045]; England[0.045]; Somerset[0.045]; Somerset[0.045]; Somerset[0.045]; Leicestershire[0.045]; Leicestershire[0.045]; Leicestershire[0.045]; bowled[0.044]; bowling[0.043]; match[0.043]; match[0.043]; Derbyshire[0.043]; Surrey[0.043]; Cricket[0.042]; Kent[0.042]; second[0.042]; Caddick[0.041]; rounder[0.041]; Headingley[0.041]; Yorkshire[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland cricket team\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.179:0.179[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.052]; Warwickshire[0.048]; match[0.047]; match[0.047]; Oval[0.047]; Essex[0.047]; Worcestershire[0.046]; Surrey[0.045]; Surrey[0.045]; Derbyshire[0.045]; Derbyshire[0.045]; day[0.045]; second[0.045]; season[0.045]; spinner[0.045]; Nottinghamshire[0.044]; Kent[0.044]; captain[0.044]; Yorkshire[0.044]; Australian[0.044]; Mark[0.043]; balls[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Essex ==> ENTITY: \u001b[32mEssex County Cricket Club\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.210:0.210[0]; log p(e|m)= -2.343:-2.343[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.046]; Essex[0.046]; Leicestershire[0.046]; Leicestershire[0.046]; Leicestershire[0.046]; Somerset[0.045]; Somerset[0.045]; Somerset[0.045]; bowling[0.045]; match[0.044]; Surrey[0.044]; Nottinghamshire[0.044]; Derbyshire[0.044]; rounder[0.043]; Cricket[0.043]; Kent[0.043]; bowled[0.042]; innings[0.041]; innings[0.041]; innings[0.041]; Caddick[0.041]; second[0.041]; Hussain[0.041]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mHeadingley\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mHeadingley Stadium\u001b[39m <---> \u001b[32mHeadingley\u001b[39m\t\nSCORES: global= 0.273:0.258[\u001b[31m0.014\u001b[39m]; local()= 0.194:0.174[\u001b[31m0.020\u001b[39m]; log p(e|m)= -0.302:-0.571[\u001b[31m0.268\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.045]; England[0.045]; England[0.045]; Oval[0.044]; Leicestershire[0.043]; match[0.043]; match[0.043]; Somerset[0.042]; Somerset[0.042]; Essex[0.042]; Essex[0.042]; Nottinghamshire[0.042]; bowling[0.041]; Yorkshire[0.041]; Yorkshire[0.041]; second[0.041]; Kent[0.040]; Road[0.040]; season[0.040]; Surrey[0.040]; bowled[0.039]; innings[0.039]; innings[0.039]; innings[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Mark Butcher\u001b[39m ==> ENTITY: \u001b[32mMark Butcher\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.158:0.158[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.044]; England[0.044]; match[0.044]; match[0.044]; Warwickshire[0.043]; spinner[0.042]; Surrey[0.042]; Surrey[0.042]; Essex[0.042]; Derbyshire[0.042]; Derbyshire[0.042]; captain[0.041]; Worcestershire[0.041]; run[0.041]; Nottinghamshire[0.041]; innings[0.041]; innings[0.041]; innings[0.041]; innings[0.041]; runs[0.041]; seven[0.041]; Oval[0.040]; Chris[0.040]; Chris[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = West Indian ==> ENTITY: \u001b[32mWest Indies cricket team\u001b[39m\t\nSCORES: global= 0.266:0.266[0]; local()= 0.147:0.147[0]; log p(e|m)= -1.019:-1.019[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.046]; Essex[0.044]; match[0.043]; Cricket[0.043]; bowling[0.043]; bowled[0.042]; Surrey[0.042]; Somerset[0.042]; Somerset[0.042]; Leicestershire[0.042]; Leicestershire[0.042]; Leicestershire[0.042]; rounder[0.041]; Kent[0.041]; lost[0.041]; opening[0.041]; innings[0.041]; innings[0.041]; runs[0.041]; runs[0.041]; Derbyshire[0.040]; Nottinghamshire[0.040]; took[0.040]; Caddick[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Surrey ==> ENTITY: \u001b[32mSurrey County Cricket Club\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.198:0.198[0]; log p(e|m)= -1.796:-1.796[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.045]; Essex[0.045]; Essex[0.045]; Surrey[0.044]; England[0.044]; England[0.044]; England[0.044]; Worcestershire[0.044]; Derbyshire[0.042]; Oval[0.042]; match[0.041]; match[0.041]; Yorkshire[0.041]; Yorkshire[0.041]; Headingley[0.041]; second[0.040]; Hussain[0.039]; Hussain[0.039]; seven[0.039]; Adams[0.039]; spinner[0.039]; balls[0.039]; day[0.039]; captain[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nasser Hussain\u001b[39m ==> ENTITY: \u001b[32mNasser Hussain\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.151:0.151[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; England[0.047]; England[0.047]; captain[0.045]; Oval[0.044]; match[0.044]; match[0.044]; Hussain[0.044]; bowling[0.044]; Surrey[0.044]; Surrey[0.044]; Kent[0.043]; Essex[0.043]; Essex[0.043]; Essex[0.043]; Leicestershire[0.042]; Somerset[0.042]; Somerset[0.042]; Headingley[0.042]; Andy[0.042]; day[0.042]; Caddick[0.042]; Derbyshire[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Grace Road\u001b[39m ==> ENTITY: \u001b[32mGrace Road\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.205:0.205[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.044]; match[0.044]; Leicestershire[0.044]; Leicestershire[0.044]; Leicestershire[0.044]; England[0.043]; England[0.043]; Somerset[0.042]; Somerset[0.042]; Somerset[0.042]; Headingley[0.042]; second[0.042]; Essex[0.041]; Essex[0.041]; Essex[0.041]; Nottinghamshire[0.041]; season[0.041]; Cricket[0.041]; rounder[0.040]; bowled[0.040]; Derbyshire[0.040]; Kent[0.040]; spot[0.040]; bowling[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Derbyshire ==> ENTITY: \u001b[32mDerbyshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.210:0.210[0]; log p(e|m)= -1.483:-1.483[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.046]; England[0.045]; England[0.045]; England[0.045]; Worcestershire[0.044]; Nottinghamshire[0.044]; Derbyshire[0.043]; Surrey[0.043]; bowler[0.043]; Kent[0.042]; Kent[0.042]; match[0.042]; match[0.042]; Martin[0.041]; Paul[0.041]; second[0.040]; Adams[0.040]; Mark[0.040]; runs[0.039]; innings[0.039]; innings[0.039]; innings[0.039]; Chris[0.039]; Thursday[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Derbyshire ==> ENTITY: \u001b[32mDerbyshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.281:0.281[0]; local()= 0.212:0.212[0]; log p(e|m)= -1.483:-1.483[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.046]; Essex[0.046]; Essex[0.046]; Somerset[0.045]; Somerset[0.045]; Somerset[0.045]; Leicestershire[0.045]; Leicestershire[0.045]; Leicestershire[0.045]; Nottinghamshire[0.045]; bowling[0.045]; Surrey[0.044]; Cricket[0.044]; Kent[0.043]; rounder[0.043]; match[0.043]; Caddick[0.042]; bowled[0.042]; Road[0.041]; second[0.040]; runs[0.040]; runs[0.040]; Andy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Yorkshire ==> ENTITY: \u001b[32mYorkshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.213:0.213[0]; log p(e|m)= -1.778:-1.778[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.045]; England[0.044]; England[0.044]; England[0.044]; England[0.044]; Essex[0.044]; Essex[0.044]; Somerset[0.043]; Derbyshire[0.042]; Headingley[0.042]; Surrey[0.042]; Surrey[0.042]; Yorkshire[0.042]; match[0.041]; match[0.041]; Oval[0.041]; bowled[0.041]; Caddick[0.040]; balls[0.040]; second[0.039]; innings[0.039]; innings[0.039]; day[0.039]; seven[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kent ==> ENTITY: \u001b[32mKent County Cricket Club\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.187:0.187[0]; log p(e|m)= -2.513:-2.513[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; England[0.047]; Warwickshire[0.046]; Surrey[0.045]; Worcestershire[0.044]; Kent[0.044]; Derbyshire[0.043]; Derbyshire[0.043]; bowler[0.043]; match[0.043]; Nottinghamshire[0.042]; second[0.041]; Martin[0.040]; Mark[0.040]; day[0.040]; Adams[0.039]; seven[0.039]; took[0.039]; took[0.039]; took[0.039]; runs[0.039]; innings[0.039]; innings[0.039]; innings[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Somerset ==> ENTITY: \u001b[32mSomerset County Cricket Club\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.227:0.227[0]; log p(e|m)= -2.419:-2.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.051]; England[0.051]; Essex[0.050]; Essex[0.050]; Essex[0.050]; Somerset[0.050]; Somerset[0.050]; Leicestershire[0.048]; Leicestershire[0.048]; bowling[0.048]; Surrey[0.047]; match[0.047]; match[0.047]; Nottinghamshire[0.046]; Derbyshire[0.046]; Kent[0.046]; bowled[0.046]; Headingley[0.046]; second[0.045]; county[0.044]; Caddick[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Surrey ==> ENTITY: \u001b[32mSurrey County Cricket Club\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.190:0.190[0]; log p(e|m)= -1.796:-1.796[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.046]; Essex[0.046]; Surrey[0.045]; England[0.045]; England[0.045]; Worcestershire[0.044]; Nottinghamshire[0.043]; Derbyshire[0.042]; Derbyshire[0.042]; Kent[0.042]; Oval[0.042]; championship[0.042]; match[0.041]; match[0.041]; Yorkshire[0.041]; second[0.040]; title[0.040]; seven[0.039]; Adams[0.039]; spinner[0.039]; balls[0.039]; day[0.039]; captain[0.039]; Mark[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Derbyshire ==> ENTITY: \u001b[32mDerbyshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.210:0.210[0]; log p(e|m)= -1.483:-1.483[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.048]; England[0.047]; England[0.047]; England[0.047]; Worcestershire[0.046]; Nottinghamshire[0.045]; Derbyshire[0.045]; Surrey[0.045]; Surrey[0.045]; bowler[0.044]; Kent[0.044]; Oval[0.043]; match[0.043]; match[0.043]; Yorkshire[0.042]; Paul[0.042]; spinner[0.042]; second[0.041]; Adams[0.041]; Mark[0.041]; captain[0.041]; runs[0.040]; balls[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hussain\u001b[39m ==> ENTITY: \u001b[32mNasser Hussain\u001b[39m\t\nSCORES: global= 0.300:0.300[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.045]; England[0.045]; England[0.045]; captain[0.043]; Oval[0.043]; match[0.043]; match[0.043]; Hussain[0.042]; bowling[0.042]; Surrey[0.042]; Nasser[0.041]; Kent[0.041]; Essex[0.041]; Essex[0.041]; Leicestershire[0.041]; Somerset[0.041]; Somerset[0.041]; Headingley[0.040]; Andy[0.040]; day[0.040]; Caddick[0.040]; innings[0.040]; innings[0.040]; innings[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland cricket team\u001b[39m\t\nSCORES: global= 0.252:0.252[0]; local()= 0.185:0.185[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.049]; Leicestershire[0.045]; Leicestershire[0.045]; match[0.045]; match[0.045]; bowling[0.044]; Essex[0.044]; Essex[0.044]; Essex[0.044]; Somerset[0.043]; Somerset[0.043]; Somerset[0.043]; Surrey[0.043]; second[0.043]; Derbyshire[0.043]; rounder[0.042]; day[0.042]; Headingley[0.042]; Indian[0.042]; Nottinghamshire[0.042]; Kent[0.042]; bowled[0.042]; season[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Kent ==> ENTITY: \u001b[32mKent County Cricket Club\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.204:0.204[0]; log p(e|m)= -2.513:-2.513[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; Essex[0.047]; Essex[0.047]; Somerset[0.046]; Somerset[0.046]; Somerset[0.046]; Surrey[0.046]; Leicestershire[0.045]; Leicestershire[0.045]; Leicestershire[0.045]; bowling[0.043]; Derbyshire[0.043]; match[0.043]; match[0.043]; Nottinghamshire[0.043]; Cricket[0.042]; rounder[0.042]; bowled[0.042]; Yorkshire[0.041]; Headingley[0.041]; second[0.041]; time[0.041]; county[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chris Lewis ==> ENTITY: \u001b[32mChris Lewis (cricketer)\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.145:0.145[0]; log p(e|m)= -0.761:-0.761[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; England[0.047]; England[0.047]; balls[0.046]; Mark[0.044]; match[0.044]; match[0.044]; Chris[0.044]; Peter[0.044]; day[0.043]; season[0.043]; spinner[0.043]; Australian[0.043]; Warwickshire[0.043]; Headingley[0.043]; Essex[0.043]; captain[0.042]; Worcestershire[0.042]; Surrey[0.042]; Surrey[0.042]; avoiding[0.042]; took[0.041]; took[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leicestershire ==> ENTITY: \u001b[32mLeicestershire County Cricket Club\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.175:0.175[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbowling[0.046]; Leicestershire[0.045]; Leicestershire[0.045]; Essex[0.043]; match[0.043]; Somerset[0.043]; Somerset[0.043]; Nottinghamshire[0.043]; Derbyshire[0.042]; Surrey[0.042]; innings[0.042]; innings[0.042]; Kent[0.042]; rounder[0.041]; Cricket[0.041]; runs[0.040]; victory[0.040]; short[0.040]; lost[0.040]; beat[0.040]; Friday[0.039]; Phil[0.039]; county[0.039]; time[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Phil Simmons\u001b[39m ==> ENTITY: \u001b[32mPhil Simmons\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.127:0.127[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.045]; bowling[0.045]; Cricket[0.044]; Essex[0.043]; Leicestershire[0.043]; Leicestershire[0.043]; Leicestershire[0.043]; bowled[0.042]; match[0.042]; Indian[0.042]; Somerset[0.042]; Somerset[0.042]; rounder[0.042]; Nottinghamshire[0.042]; time[0.041]; Surrey[0.041]; Kent[0.041]; innings[0.041]; innings[0.041]; Andy[0.040]; Derbyshire[0.040]; Caddick[0.039]; short[0.039]; county[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Peter Such\u001b[39m ==> ENTITY: \u001b[32mPeter Such\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.043]; match[0.043]; Essex[0.043]; Essex[0.043]; Essex[0.043]; Surrey[0.043]; Surrey[0.043]; Headingley[0.042]; bowling[0.042]; Kent[0.042]; Nottinghamshire[0.042]; Derbyshire[0.042]; Andy[0.041]; Leicestershire[0.041]; Yorkshire[0.041]; Yorkshire[0.041]; Somerset[0.041]; Somerset[0.041]; second[0.041]; bowled[0.041]; got[0.041]; England[0.041]; England[0.041]; England[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Nottinghamshire ==> ENTITY: \u001b[32mNottinghamshire County Cricket Club\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.187:0.187[0]; log p(e|m)= -1.106:-1.106[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.049]; England[0.049]; Worcestershire[0.048]; bowler[0.047]; match[0.047]; stumps[0.047]; Derbyshire[0.047]; Derbyshire[0.047]; Surrey[0.046]; Kent[0.046]; Kent[0.046]; second[0.045]; Mark[0.044]; Adams[0.044]; Chris[0.044]; innings[0.044]; innings[0.044]; innings[0.044]; runs[0.044]; Martin[0.044]; day[0.043]; seven[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Such\u001b[39m ==> ENTITY: \u001b[32mPeter Such\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.148:0.148[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; Mark[0.043]; match[0.043]; match[0.043]; Essex[0.043]; Essex[0.043]; Surrey[0.043]; Surrey[0.043]; Headingley[0.042]; Peter[0.042]; Worcestershire[0.042]; Derbyshire[0.042]; Andy[0.041]; Yorkshire[0.041]; Yorkshire[0.041]; Somerset[0.041]; second[0.041]; got[0.041]; England[0.041]; England[0.041]; England[0.041]; spinner[0.041]; Caddick[0.040]; captain[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Paul Johnson ==> ENTITY: \u001b[32mPaul Johnson (cricketer)\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.103:0.103[0]; log p(e|m)= -2.957:-2.957[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.049]; championship[0.044]; bowler[0.044]; second[0.043]; took[0.043]; took[0.043]; took[0.043]; Chris[0.042]; match[0.042]; Martin[0.042]; title[0.041]; defeat[0.041]; seven[0.041]; runs[0.041]; held[0.040]; innings[0.040]; innings[0.040]; innings[0.040]; reached[0.040]; Kent[0.040]; Kent[0.040]; Tim[0.040]; Australian[0.040]; Moody[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Simmons\u001b[39m ==> ENTITY: \u001b[32mPhil Simmons\u001b[39m\t\nSCORES: global= 0.302:0.302[0]; local()= 0.142:0.142[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.044]; England[0.044]; bowling[0.044]; Essex[0.043]; Essex[0.043]; Essex[0.043]; Leicestershire[0.042]; bowled[0.042]; match[0.042]; match[0.042]; balls[0.042]; Somerset[0.041]; Somerset[0.041]; Nottinghamshire[0.041]; time[0.041]; Surrey[0.040]; Kent[0.040]; leaving[0.040]; innings[0.040]; innings[0.040]; innings[0.040]; spinner[0.040]; day[0.040]; Andy[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Somerset ==> ENTITY: \u001b[32mSomerset County Cricket Club\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.220:0.220[0]; log p(e|m)= -2.419:-2.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; England[0.047]; Somerset[0.046]; Somerset[0.046]; Essex[0.046]; Essex[0.046]; Leicestershire[0.044]; Leicestershire[0.044]; Leicestershire[0.044]; bowling[0.044]; Surrey[0.043]; match[0.043]; match[0.043]; Nottinghamshire[0.043]; Derbyshire[0.042]; Kent[0.042]; Cricket[0.042]; bowled[0.042]; Headingley[0.042]; county[0.041]; second[0.041]; rounder[0.041]; Caddick[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Worcestershire ==> ENTITY: \u001b[32mWorcestershire County Cricket Club\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.193:0.193[0]; log p(e|m)= -0.916:-0.916[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWarwickshire[0.044]; England[0.044]; England[0.044]; England[0.044]; bowler[0.043]; Nottinghamshire[0.043]; innings[0.042]; innings[0.042]; innings[0.042]; Surrey[0.042]; Surrey[0.042]; Derbyshire[0.042]; Derbyshire[0.042]; stumps[0.041]; match[0.041]; match[0.041]; Mark[0.041]; spinner[0.041]; runs[0.040]; second[0.040]; Chris[0.040]; Chris[0.040]; Oval[0.040]; balls[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAustralian\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAustralia national cricket team\u001b[39m <---> \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.252:0.243[\u001b[31m0.009\u001b[39m]; local()= 0.143:0.077[\u001b[31m0.066\u001b[39m]; log p(e|m)= -3.612:-0.491[\u001b[32m3.121\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.046]; England[0.046]; England[0.046]; Oval[0.044]; match[0.043]; match[0.043]; bowler[0.043]; captain[0.042]; Warwickshire[0.042]; Mark[0.042]; day[0.042]; Martin[0.042]; second[0.041]; seven[0.041]; Surrey[0.041]; Surrey[0.041]; Paul[0.040]; Worcestershire[0.040]; Kent[0.040]; Kent[0.040]; Nottinghamshire[0.039]; Tim[0.039]; Chris[0.039]; Chris[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = England ==> ENTITY: \u001b[32mEngland cricket team\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.184:0.184[0]; log p(e|m)= -3.297:-3.297[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEngland[0.047]; England[0.047]; season[0.043]; Leicestershire[0.043]; match[0.043]; match[0.043]; match[0.043]; Oval[0.043]; bowling[0.042]; Essex[0.042]; Essex[0.042]; Somerset[0.041]; Somerset[0.041]; Surrey[0.041]; day[0.040]; Headingley[0.040]; second[0.040]; spinner[0.040]; Nottinghamshire[0.040]; bowled[0.040]; captain[0.040]; Yorkshire[0.040]; Yorkshire[0.040]; Andy[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Leicestershire ==> ENTITY: \u001b[32mLeicestershire County Cricket Club\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.177:0.177[0]; log p(e|m)= -1.047:-1.047[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nbowling[0.045]; bowled[0.045]; Leicestershire[0.045]; Leicestershire[0.045]; England[0.043]; Essex[0.042]; match[0.042]; Somerset[0.042]; Somerset[0.042]; Somerset[0.042]; Nottinghamshire[0.042]; Derbyshire[0.041]; Surrey[0.041]; innings[0.041]; innings[0.041]; Kent[0.041]; rounder[0.041]; Cricket[0.041]; runs[0.040]; runs[0.040]; victory[0.039]; short[0.039]; Andy[0.039]; lost[0.039]; \t\n\r [============================================>.] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 475ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4695/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1029testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Michigan National Bank\u001b[39m ==> ENTITY: \u001b[32mMichigan National Bank\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.017:0.017[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBank[0.048]; Conc[0.044]; Michigan[0.044]; Amt[0.043]; Nic[0.043]; Sp[0.043]; Corp[0.042]; Series[0.042]; approx[0.042]; Mgr[0.042]; Bldg[0.041]; Bldg[0.041]; 1st[0.041]; Sr[0.041]; Competitive[0.041]; Qual[0.040]; Kenny[0.040]; Entry[0.040]; Na[0.040]; Na[0.040]; Na[0.040]; Detroit[0.040]; Detroit[0.040]; Detroit[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Detroit ==> ENTITY: \u001b[32mDetroit\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.057:0.057[0]; log p(e|m)= -0.217:-0.217[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDetroit[0.048]; Detroit[0.048]; National[0.045]; Michigan[0.044]; Michigan[0.044]; Authority[0.042]; Building[0.042]; Bay[0.042]; Bay[0.042]; Kenny[0.041]; Bid[0.041]; Entry[0.041]; Nic[0.041]; Mich[0.041]; Sure[0.041]; Competitive[0.041]; Bldg[0.041]; Bldg[0.041]; days[0.040]; Bank[0.040]; Bank[0.040]; 1st[0.039]; Time[0.039]; Delivery[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Detroit ==> ENTITY: \u001b[32mDetroit\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.050:0.050[0]; log p(e|m)= -0.217:-0.217[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDetroit[0.048]; Detroit[0.048]; National[0.045]; Michigan[0.045]; Michigan[0.045]; Authority[0.043]; Building[0.042]; Bay[0.042]; Kenny[0.041]; Bid[0.041]; Entry[0.041]; Nic[0.041]; Mich[0.041]; Sure[0.041]; Competitive[0.041]; Bldg[0.041]; Bldg[0.041]; days[0.040]; Bank[0.040]; Bank[0.040]; 1st[0.039]; Time[0.039]; Delivery[0.038]; Paying[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Detroit ==> ENTITY: \u001b[32mDetroit\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.042:0.042[0]; log p(e|m)= -0.217:-0.217[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDetroit[0.048]; Detroit[0.048]; National[0.045]; Michigan[0.045]; Michigan[0.045]; Authority[0.043]; Building[0.042]; Bay[0.042]; Kenny[0.041]; Bid[0.041]; Entry[0.041]; Nic[0.041]; Sure[0.041]; Competitive[0.041]; Bldg[0.041]; days[0.040]; Bank[0.040]; Bank[0.040]; 1st[0.039]; Time[0.039]; Delivery[0.039]; Paying[0.039]; Enhancements[0.038]; Series[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = NYC ==> ENTITY: \u001b[32mNew York City\u001b[39m\t\nSCORES: global= 0.267:0.267[0]; local()= 0.008:0.008[0]; log p(e|m)= -0.386:-0.386[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nNational[0.045]; Time[0.045]; Bay[0.043]; Bay[0.043]; Authority[0.043]; Bank[0.042]; Bank[0.042]; Date[0.042]; Detroit[0.042]; Detroit[0.042]; Detroit[0.042]; Building[0.042]; 1st[0.041]; Bid[0.041]; Corp[0.041]; Michigan[0.041]; Michigan[0.041]; Exempt[0.041]; Orders[0.040]; Delivery[0.040]; Agent[0.040]; Tax[0.040]; Paying[0.040]; Qual[0.040]; \t\n\r [============================================>.] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 453ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4700/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 978testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Ottawa ==> ENTITY: \u001b[32mOttawa\u001b[39m\t\nSCORES: global= 0.272:0.272[0]; local()= 0.142:0.142[0]; log p(e|m)= -0.350:-0.350[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.049]; Canada[0.049]; Canada[0.049]; Canada[0.049]; Canadian[0.046]; July[0.042]; Government[0.041]; June[0.041]; Friday[0.041]; Notes[0.040]; August[0.040]; Govt[0.040]; government[0.040]; government[0.040]; Treasury[0.040]; Bank[0.040]; week[0.040]; week[0.040]; bank[0.039]; said[0.039]; figures[0.039]; assets[0.039]; fall[0.039]; dollars[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bank of Canada\u001b[39m ==> ENTITY: \u001b[32mBank of Canada\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.193:0.193[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.049]; Canada[0.049]; Canada[0.049]; Canadian[0.047]; Canadian[0.047]; dollar[0.047]; Wk[0.047]; Ottawa[0.046]; government[0.045]; government[0.045]; currency[0.045]; bank[0.045]; dollars[0.045]; cash[0.044]; cash[0.044]; cash[0.044]; Treasury[0.044]; Government[0.044]; June[0.043]; June[0.043]; securities[0.043]; August[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canadian ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.160:0.160[0]; log p(e|m)= -0.412:-0.412[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.057]; Canada[0.057]; Canada[0.057]; Canadian[0.052]; Ottawa[0.050]; dollars[0.047]; dollars[0.047]; dollar[0.047]; cash[0.044]; cash[0.044]; government[0.043]; Bank[0.042]; week[0.042]; Government[0.042]; said[0.041]; outstanding[0.041]; assets[0.041]; assets[0.041]; June[0.041]; June[0.041]; Bureau[0.041]; ended[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.171:0.171[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.052]; Canada[0.052]; Canada[0.052]; Canadian[0.047]; Canadian[0.047]; Ottawa[0.046]; Friday[0.042]; week[0.041]; week[0.041]; July[0.040]; cash[0.040]; cash[0.040]; cash[0.040]; August[0.039]; fall[0.039]; June[0.039]; government[0.039]; government[0.039]; Bank[0.039]; foreign[0.038]; ended[0.038]; Notes[0.037]; outstanding[0.037]; dollars[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canadian ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.164:0.164[0]; log p(e|m)= -0.412:-0.412[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.052]; Canada[0.052]; Canada[0.052]; Canadian[0.048]; Ottawa[0.046]; Ottawa[0.046]; dollars[0.043]; dollars[0.043]; dollar[0.043]; cash[0.040]; cash[0.040]; government[0.039]; Bank[0.039]; week[0.038]; week[0.038]; Government[0.038]; said[0.038]; outstanding[0.038]; assets[0.038]; assets[0.038]; June[0.038]; June[0.038]; Bureau[0.038]; ended[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.284:0.284[0]; local()= 0.191:0.191[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.051]; Canada[0.051]; Canada[0.051]; Canadian[0.047]; Canadian[0.047]; Ottawa[0.045]; Ottawa[0.045]; August[0.040]; July[0.040]; July[0.040]; July[0.040]; June[0.039]; June[0.039]; cash[0.039]; cash[0.039]; cash[0.039]; week[0.039]; week[0.039]; dollars[0.039]; dollars[0.039]; government[0.039]; government[0.039]; Bank[0.038]; fall[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Canada ==> ENTITY: \u001b[32mCanada\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.147:0.147[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCanada[0.053]; Canada[0.053]; Canada[0.053]; Ottawa[0.047]; Friday[0.043]; week[0.042]; week[0.042]; July[0.041]; cash[0.040]; cash[0.040]; cash[0.040]; August[0.040]; fall[0.040]; June[0.040]; government[0.040]; government[0.040]; Bank[0.040]; foreign[0.039]; ended[0.039]; Notes[0.038]; outstanding[0.038]; dollars[0.038]; said[0.037]; figures[0.037]; \t\n\r [============================================>.] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 418ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4707/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1114testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Bonn\u001b[39m ==> ENTITY: \u001b[32mBonn\u001b[39m\t\nSCORES: global= 0.256:0.256[0]; local()= 0.057:0.057[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.049]; Soccer[0.047]; German[0.047]; German[0.047]; Neunkirchen[0.045]; Cup[0.044]; Cup[0.044]; Borussia[0.044]; Saturday[0.044]; second[0.044]; Karlsruhe[0.044]; Luebeck[0.044]; Duisburg[0.043]; Hansa[0.043]; Rostock[0.043]; extra[0.043]; round[0.042]; Round[0.042]; Second[0.041]; time[0.041]; Pauli[0.040]; Results[0.040]; Results[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Karlsruhe ==> ENTITY: \u001b[32mKarlsruher SC\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.170:0.170[0]; log p(e|m)= -2.226:-2.226[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nSoccer[0.051]; matches[0.049]; Borussia[0.047]; Cup[0.046]; Cup[0.046]; German[0.045]; German[0.045]; Duisburg[0.044]; Hansa[0.044]; round[0.044]; Saturday[0.044]; second[0.044]; Bonn[0.043]; Neunkirchen[0.042]; Luebeck[0.042]; Rostock[0.042]; Round[0.042]; Results[0.041]; Results[0.041]; extra[0.040]; time[0.040]; Second[0.039]; Pauli[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Duisburg ==> ENTITY: \u001b[32mMSV Duisburg\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.154:0.154[0]; log p(e|m)= -1.590:-1.590[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.051]; Soccer[0.047]; Borussia[0.047]; German[0.046]; German[0.046]; Bonn[0.045]; Karlsruhe[0.045]; Cup[0.045]; Cup[0.045]; Luebeck[0.045]; second[0.044]; Rostock[0.044]; extra[0.043]; time[0.043]; Hansa[0.043]; round[0.043]; Round[0.043]; Neunkirchen[0.042]; Saturday[0.041]; Second[0.039]; Results[0.038]; Results[0.038]; Pauli[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = St Pauli ==> ENTITY: \u001b[32mFC St. Pauli\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.146:0.146[0]; log p(e|m)= -0.117:-0.117[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.053]; Borussia[0.048]; Duisburg[0.047]; Soccer[0.046]; Cup[0.045]; Cup[0.045]; Karlsruhe[0.045]; second[0.045]; German[0.044]; German[0.044]; Bonn[0.044]; Rostock[0.044]; Neunkirchen[0.043]; round[0.043]; Saturday[0.042]; Hansa[0.042]; Luebeck[0.042]; time[0.041]; Round[0.041]; extra[0.041]; Second[0.039]; Results[0.039]; Results[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = German Cup\u001b[39m ==> ENTITY: \u001b[32mDFB-Pokal\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.053]; Soccer[0.052]; second[0.050]; Borussia[0.049]; Karlsruhe[0.048]; Cup[0.048]; Duisburg[0.047]; round[0.047]; German[0.046]; extra[0.046]; Round[0.046]; time[0.046]; Saturday[0.045]; Rostock[0.044]; Neunkirchen[0.044]; Bonn[0.043]; Luebeck[0.043]; Second[0.042]; Hansa[0.042]; Results[0.040]; Results[0.040]; Pauli[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Borussia Neunkirchen\u001b[39m ==> ENTITY: \u001b[32mBorussia Neunkirchen\u001b[39m\t\nSCORES: global= 0.295:0.295[0]; local()= 0.164:0.164[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.052]; German[0.049]; German[0.049]; Karlsruhe[0.049]; Soccer[0.049]; Duisburg[0.047]; Rostock[0.047]; round[0.046]; Bonn[0.046]; Cup[0.046]; Cup[0.046]; Luebeck[0.046]; second[0.044]; Round[0.044]; Saturday[0.043]; Second[0.043]; Results[0.043]; Results[0.043]; Hansa[0.042]; extra[0.042]; time[0.042]; Pauli[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hansa Rostock\u001b[39m ==> ENTITY: \u001b[32mF.C. Hansa Rostock\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.165:0.165[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.051]; Soccer[0.051]; Borussia[0.050]; Duisburg[0.049]; German[0.048]; German[0.048]; second[0.048]; Karlsruhe[0.047]; Luebeck[0.046]; Bonn[0.046]; Cup[0.045]; Cup[0.045]; Saturday[0.045]; Neunkirchen[0.044]; extra[0.044]; round[0.044]; time[0.043]; Round[0.043]; Second[0.042]; Pauli[0.041]; Results[0.040]; Results[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = German Cup\u001b[39m ==> ENTITY: \u001b[32mDFB-Pokal\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.143:0.143[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatches[0.053]; Soccer[0.052]; second[0.050]; Borussia[0.049]; Karlsruhe[0.048]; Cup[0.048]; Duisburg[0.047]; round[0.047]; German[0.046]; extra[0.046]; Round[0.046]; time[0.046]; Saturday[0.045]; Rostock[0.044]; Neunkirchen[0.044]; Bonn[0.043]; Luebeck[0.043]; Second[0.042]; Hansa[0.042]; Results[0.040]; Results[0.040]; Pauli[0.039]; \t\n\r [============================================>.] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 379ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4715/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1066testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = London ==> ENTITY: \u001b[32mLondon\u001b[39m\t\nSCORES: global= 0.246:0.246[0]; local()= 0.026:0.026[0]; log p(e|m)= -0.091:-0.091[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLeague[0.048]; matches[0.048]; league[0.047]; Division[0.045]; Soccer[0.045]; Airdrieonians[0.045]; played[0.043]; East[0.043]; won[0.042]; goals[0.041]; goals[0.041]; Saturday[0.041]; Dundee[0.040]; lost[0.040]; Scottish[0.040]; Scottish[0.040]; standings[0.040]; points[0.039]; Stirling[0.039]; Standings[0.039]; Falkirk[0.039]; Fife[0.039]; Partick[0.039]; drawn[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Berwick ==> ENTITY: \u001b[32mBerwick Rangers F.C.\u001b[39m\t\nSCORES: global= 0.254:0.254[0]; local()= 0.151:0.151[0]; log p(e|m)= -3.689:-3.689[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLivingston[0.045]; Park[0.044]; Fife[0.044]; Stirling[0.043]; Stirling[0.043]; Cowdenbeath[0.043]; Stenhousemuir[0.043]; Inverness[0.042]; Ross[0.042]; Clyde[0.042]; Brechin[0.041]; Hamilton[0.041]; Queen[0.041]; Queen[0.041]; East[0.041]; East[0.041]; County[0.040]; Dumbarton[0.040]; Forfar[0.040]; South[0.040]; Division[0.040]; Division[0.040]; Thistle[0.040]; Alloa[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Livingston ==> ENTITY: \u001b[32mLivingston F.C.\u001b[39m\t\nSCORES: global= 0.278:0.278[0]; local()= 0.190:0.190[0]; log p(e|m)= -1.864:-1.864[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.047]; Fife[0.045]; Dundee[0.043]; Stenhousemuir[0.043]; Stirling[0.043]; Greenock[0.043]; matches[0.042]; Dumbarton[0.042]; Cowdenbeath[0.042]; Falkirk[0.042]; Brechin[0.042]; Albion[0.042]; played[0.041]; Partick[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Stranraer[0.041]; Forfar[0.040]; won[0.040]; Clydebank[0.040]; Clyde[0.040]; Hamilton[0.039]; East[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Falkirk ==> ENTITY: \u001b[32mFalkirk F.C.\u001b[39m\t\nSCORES: global= 0.271:0.271[0]; local()= 0.190:0.190[0]; log p(e|m)= -1.204:-1.204[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStenhousemuir[0.046]; Airdrieonians[0.046]; League[0.044]; Fife[0.044]; Scottish[0.044]; Scottish[0.044]; Ayr[0.043]; Stranraer[0.043]; Brechin[0.043]; Division[0.042]; Division[0.042]; Dundee[0.042]; Partick[0.042]; Clydebank[0.042]; Livingston[0.041]; league[0.041]; Stirling[0.041]; Greenock[0.039]; matches[0.039]; standings[0.039]; Standings[0.039]; played[0.039]; won[0.039]; Johnstone[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Inverness Thistle\u001b[39m ==> ENTITY: \u001b[32mInverness Thistle F.C.\u001b[39m\t\nSCORES: global= 0.289:0.289[0]; local()= 0.120:0.120[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCowdenbeath[0.060]; Stirling[0.060]; Ross[0.059]; Park[0.058]; Alloa[0.058]; Dumbarton[0.057]; Stranraer[0.057]; Forfar[0.056]; Brechin[0.056]; Arbroath[0.056]; Albion[0.056]; Clyde[0.055]; Berwick[0.054]; Montrose[0.054]; Division[0.054]; East[0.052]; County[0.049]; Queen[0.047]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = St Johnstone\u001b[39m ==> ENTITY: \u001b[32mSt. Johnstone F.C.\u001b[39m\t\nSCORES: global= 0.301:0.301[0]; local()= 0.194:0.194[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStenhousemuir[0.046]; Airdrieonians[0.046]; League[0.045]; Livingston[0.044]; Dundee[0.043]; league[0.043]; Partick[0.043]; Stirling[0.042]; goals[0.042]; goals[0.042]; Falkirk[0.042]; matches[0.042]; Scottish[0.041]; Scottish[0.041]; played[0.041]; Greenock[0.040]; Fife[0.040]; Division[0.040]; Division[0.040]; Soccer[0.040]; Morton[0.040]; Saturday[0.040]; won[0.039]; Clydebank[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Brechin ==> ENTITY: \u001b[32mBrechin City F.C.\u001b[39m\t\nSCORES: global= 0.276:0.276[0]; local()= 0.185:0.185[0]; log p(e|m)= -1.917:-1.917[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nForfar[0.046]; Ayr[0.045]; Stenhousemuir[0.045]; Arbroath[0.043]; Cowdenbeath[0.043]; Alloa[0.043]; Dumbarton[0.043]; Partick[0.043]; Stranraer[0.043]; Inverness[0.043]; Thistle[0.042]; Livingston[0.042]; Fife[0.041]; Stirling[0.041]; Clydebank[0.041]; Park[0.041]; Falkirk[0.041]; Division[0.041]; Division[0.041]; Albion[0.039]; Berwick[0.039]; Clyde[0.039]; Hamilton[0.038]; Montrose[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Arbroath ==> ENTITY: \u001b[32mArbroath F.C.\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.160:0.160[0]; log p(e|m)= -0.860:-0.860[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStenhousemuir[0.046]; Cowdenbeath[0.045]; Inverness[0.044]; Brechin[0.044]; Dumbarton[0.044]; Ross[0.043]; Alloa[0.043]; Forfar[0.043]; Stranraer[0.043]; Stirling[0.042]; Park[0.042]; Ayr[0.042]; Division[0.042]; Clyde[0.042]; Thistle[0.041]; Albion[0.041]; Montrose[0.040]; East[0.040]; Hamilton[0.040]; Berwick[0.039]; Queen[0.039]; Queen[0.039]; South[0.039]; County[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Stenhousemuir\u001b[39m ==> ENTITY: \u001b[32mStenhousemuir F.C.\u001b[39m\t\nSCORES: global= 0.299:0.299[0]; local()= 0.174:0.174[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.048]; Park[0.044]; Stirling[0.044]; Cowdenbeath[0.043]; Dumbarton[0.043]; Fife[0.043]; Livingston[0.043]; Alloa[0.042]; Division[0.042]; Division[0.042]; Partick[0.042]; Johnstone[0.042]; Falkirk[0.042]; Brechin[0.041]; Stranraer[0.041]; Forfar[0.041]; Clydebank[0.041]; Berwick[0.040]; Albion[0.040]; Ayr[0.040]; Arbroath[0.040]; Clyde[0.038]; Hamilton[0.038]; East[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Partick ==> ENTITY: \u001b[32mPartick Thistle F.C.\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.221:0.221[0]; log p(e|m)= -1.284:-1.284[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; Stenhousemuir[0.044]; Livingston[0.044]; Dundee[0.043]; League[0.043]; matches[0.043]; Falkirk[0.042]; league[0.042]; Greenock[0.042]; Dumbarton[0.042]; Stranraer[0.041]; Stirling[0.041]; Brechin[0.041]; Clyde[0.041]; played[0.041]; Scottish[0.041]; Scottish[0.041]; Fife[0.041]; Johnstone[0.041]; Clydebank[0.041]; Soccer[0.040]; Division[0.039]; Division[0.039]; Hamilton[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = St Mirren\u001b[39m ==> ENTITY: \u001b[32mSt. Mirren F.C.\u001b[39m\t\nSCORES: global= 0.303:0.303[0]; local()= 0.202:0.202[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.045]; matches[0.045]; league[0.044]; League[0.044]; Stenhousemuir[0.044]; played[0.044]; Dundee[0.044]; Livingston[0.043]; Partick[0.042]; Stirling[0.041]; won[0.041]; Soccer[0.041]; Falkirk[0.041]; Greenock[0.041]; Fife[0.041]; Scottish[0.041]; Scottish[0.041]; Johnstone[0.040]; Clydebank[0.040]; goals[0.040]; goals[0.040]; Saturday[0.039]; Division[0.039]; Division[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scottish ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.250:0.250[0]; local()= 0.120:0.120[0]; log p(e|m)= -0.480:-0.480[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.047]; Scottish[0.046]; League[0.045]; matches[0.044]; Dundee[0.044]; league[0.044]; played[0.044]; Fife[0.043]; Greenock[0.043]; Falkirk[0.043]; Soccer[0.042]; Partick[0.042]; won[0.042]; Stirling[0.042]; Division[0.041]; Clydebank[0.040]; Johnstone[0.040]; London[0.039]; goals[0.039]; goals[0.039]; Standings[0.039]; Saturday[0.038]; standings[0.038]; East[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = East Stirling\u001b[39m ==> ENTITY: \u001b[32mEast Stirlingshire F.C.\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.155:0.155[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCowdenbeath[0.070]; Inverness[0.066]; Dumbarton[0.065]; Thistle[0.065]; Park[0.065]; Alloa[0.065]; Forfar[0.064]; Division[0.063]; Albion[0.063]; Clyde[0.062]; Arbroath[0.062]; Ross[0.061]; Montrose[0.060]; Berwick[0.058]; Queen[0.058]; County[0.053]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clyde ==> ENTITY: \u001b[32mClyde F.C.\u001b[39m\t\nSCORES: global= 0.286:0.286[0]; local()= 0.196:0.196[0]; log p(e|m)= -0.983:-0.983[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nLivingston[0.044]; Stenhousemuir[0.044]; Cowdenbeath[0.043]; Park[0.043]; Stirling[0.043]; Stirling[0.043]; Partick[0.043]; Fife[0.042]; Thistle[0.042]; Brechin[0.042]; Dumbarton[0.042]; Inverness[0.042]; Alloa[0.042]; Clydebank[0.041]; Albion[0.041]; Stranraer[0.041]; Forfar[0.041]; Montrose[0.041]; Arbroath[0.040]; Ayr[0.040]; South[0.040]; Division[0.040]; Division[0.040]; Hamilton[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = East Fife\u001b[39m ==> ENTITY: \u001b[32mEast Fife F.C.\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.219:0.219[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; Stenhousemuir[0.045]; Dundee[0.043]; matches[0.043]; Dumbarton[0.043]; Partick[0.042]; Brechin[0.042]; Clydebank[0.042]; Stirling[0.042]; Livingston[0.042]; Stranraer[0.042]; Greenock[0.042]; Scottish[0.042]; played[0.042]; Falkirk[0.041]; League[0.041]; Division[0.041]; Division[0.041]; Division[0.041]; Johnstone[0.041]; league[0.040]; Ayr[0.040]; Clyde[0.039]; won[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Airdrieonians\u001b[39m ==> ENTITY: \u001b[32mAirdrieonians F.C.\u001b[39m\t\nSCORES: global= 0.296:0.296[0]; local()= 0.190:0.190[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStenhousemuir[0.044]; Livingston[0.044]; Dundee[0.044]; matches[0.043]; Scottish[0.043]; Scottish[0.043]; Partick[0.043]; League[0.043]; Stirling[0.043]; Stranraer[0.042]; league[0.042]; Clydebank[0.042]; Greenock[0.042]; Falkirk[0.042]; Fife[0.041]; played[0.041]; Johnstone[0.041]; Division[0.041]; Division[0.041]; won[0.040]; Soccer[0.040]; Ayr[0.040]; Morton[0.040]; lost[0.039]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAlloa\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAlloa\u001b[39m <---> \u001b[32mAlloa Athletic F.C.\u001b[39m\t\nSCORES: global= 0.289:0.284[\u001b[31m0.005\u001b[39m]; local()= 0.159:0.184[\u001b[32m0.024\u001b[39m]; log p(e|m)= 0.000:-0.794[\u001b[31m0.794\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCowdenbeath[0.050]; Stenhousemuir[0.049]; Stirling[0.049]; Brechin[0.048]; Inverness[0.048]; Thistle[0.047]; Arbroath[0.047]; Dumbarton[0.047]; Stranraer[0.046]; Forfar[0.046]; Park[0.046]; Ross[0.045]; Albion[0.045]; Clyde[0.045]; Hamilton[0.045]; Ayr[0.045]; Division[0.044]; Montrose[0.044]; East[0.042]; Berwick[0.042]; Queen[0.040]; County[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Forfar ==> ENTITY: \u001b[32mForfar Athletic F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.197:0.197[0]; log p(e|m)= -1.766:-1.766[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStenhousemuir[0.046]; Cowdenbeath[0.045]; Inverness[0.044]; Arbroath[0.044]; Brechin[0.044]; Alloa[0.044]; Stranraer[0.043]; Stirling[0.043]; Thistle[0.043]; Dumbarton[0.043]; Livingston[0.041]; Park[0.041]; Division[0.041]; Division[0.041]; Ayr[0.040]; Ross[0.040]; Clyde[0.040]; Montrose[0.040]; Queen[0.040]; Queen[0.040]; Berwick[0.039]; Albion[0.039]; Hamilton[0.039]; County[0.038]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mAyr\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mAyr\u001b[39m <---> \u001b[32mAyr United F.C.\u001b[39m\t\nSCORES: global= 0.265:0.255[\u001b[31m0.010\u001b[39m]; local()= 0.165:0.216[\u001b[32m0.051\u001b[39m]; log p(e|m)= -0.066:-3.352[\u001b[31m3.286\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.045]; Cowdenbeath[0.044]; Stenhousemuir[0.044]; Dundee[0.044]; Fife[0.043]; Partick[0.043]; Livingston[0.043]; Clydebank[0.042]; Stirling[0.042]; Dumbarton[0.042]; Stranraer[0.042]; Falkirk[0.042]; Brechin[0.042]; Forfar[0.042]; Arbroath[0.042]; Alloa[0.041]; Albion[0.041]; Berwick[0.040]; Johnstone[0.040]; Clyde[0.040]; Division[0.039]; Division[0.039]; Hamilton[0.038]; East[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Clydebank ==> ENTITY: \u001b[32mClydebank F.C.\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.226:0.226[0]; log p(e|m)= -1.328:-1.328[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.047]; Dundee[0.044]; Stenhousemuir[0.044]; Greenock[0.043]; Fife[0.043]; Stirling[0.043]; played[0.042]; matches[0.042]; Brechin[0.042]; Partick[0.042]; league[0.042]; League[0.042]; Livingston[0.042]; Stranraer[0.042]; Scottish[0.041]; Scottish[0.041]; Johnstone[0.041]; Falkirk[0.041]; Clyde[0.041]; Ayr[0.040]; Division[0.039]; Division[0.039]; Hamilton[0.039]; Soccer[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Ross County\u001b[39m ==> ENTITY: \u001b[32mRoss County F.C.\u001b[39m\t\nSCORES: global= 0.292:0.292[0]; local()= 0.156:0.156[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nDivision[0.072]; Inverness[0.071]; Cowdenbeath[0.070]; Stirling[0.070]; Park[0.068]; Alloa[0.067]; Dumbarton[0.066]; Thistle[0.066]; Albion[0.066]; Arbroath[0.066]; Forfar[0.065]; Montrose[0.065]; East[0.064]; Berwick[0.063]; Queen[0.060]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scottish ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.249:0.249[0]; local()= 0.106:0.106[0]; log p(e|m)= -0.480:-0.480[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.047]; Scottish[0.046]; League[0.045]; matches[0.044]; Dundee[0.044]; league[0.044]; played[0.044]; Greenock[0.043]; Falkirk[0.043]; Soccer[0.043]; Partick[0.042]; won[0.042]; Stirling[0.042]; Division[0.041]; Clydebank[0.040]; Johnstone[0.040]; London[0.040]; goals[0.039]; goals[0.039]; Standings[0.039]; Saturday[0.039]; standings[0.039]; Morton[0.038]; lost[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Greenock Morton\u001b[39m ==> ENTITY: \u001b[32mGreenock Morton F.C.\u001b[39m\t\nSCORES: global= 0.291:0.291[0]; local()= 0.206:0.206[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; League[0.044]; Scottish[0.044]; Scottish[0.044]; Partick[0.044]; matches[0.044]; league[0.044]; played[0.044]; Livingston[0.043]; Dundee[0.043]; Fife[0.043]; Clydebank[0.042]; Stirling[0.042]; Falkirk[0.041]; Johnstone[0.041]; Soccer[0.041]; Division[0.041]; Division[0.041]; won[0.040]; Saturday[0.038]; goals[0.038]; goals[0.038]; lost[0.038]; standings[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dumbarton ==> ENTITY: \u001b[32mDumbarton F.C.\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.191:0.191[0]; log p(e|m)= -1.255:-1.255[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStenhousemuir[0.046]; Fife[0.045]; Cowdenbeath[0.045]; Stranraer[0.044]; Partick[0.043]; Arbroath[0.043]; Division[0.043]; Division[0.043]; Alloa[0.042]; Brechin[0.042]; Inverness[0.042]; Clyde[0.042]; Livingston[0.041]; Stirling[0.041]; Stirling[0.041]; Forfar[0.041]; Thistle[0.041]; Park[0.040]; Ross[0.040]; Hamilton[0.039]; Ayr[0.039]; County[0.039]; Queen[0.039]; Queen[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Montrose ==> ENTITY: \u001b[32mMontrose F.C.\u001b[39m\t\nSCORES: global= 0.262:0.262[0]; local()= 0.173:0.173[0]; log p(e|m)= -2.180:-2.180[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCowdenbeath[0.054]; Brechin[0.053]; Inverness[0.052]; Park[0.052]; Alloa[0.052]; Stirling[0.052]; Forfar[0.051]; Dumbarton[0.051]; Thistle[0.050]; Ross[0.050]; Stranraer[0.050]; Arbroath[0.050]; Clyde[0.050]; East[0.048]; Queen[0.048]; County[0.048]; Division[0.048]; Berwick[0.048]; Hamilton[0.047]; Albion[0.046]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Stranraer ==> ENTITY: \u001b[32mStranraer F.C.\u001b[39m\t\nSCORES: global= 0.283:0.283[0]; local()= 0.219:0.219[0]; log p(e|m)= -1.030:-1.030[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.046]; Livingston[0.045]; Stenhousemuir[0.044]; Cowdenbeath[0.043]; Inverness[0.043]; Brechin[0.043]; Dumbarton[0.043]; Division[0.042]; Division[0.042]; Stirling[0.042]; Fife[0.042]; Partick[0.042]; Park[0.041]; Falkirk[0.041]; Alloa[0.041]; Arbroath[0.041]; Forfar[0.041]; Ayr[0.041]; Clydebank[0.040]; Hamilton[0.040]; Clyde[0.040]; Berwick[0.040]; Albion[0.039]; Montrose[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Cowdenbeath ==> ENTITY: \u001b[32mCowdenbeath F.C.\u001b[39m\t\nSCORES: global= 0.277:0.277[0]; local()= 0.186:0.186[0]; log p(e|m)= -0.247:-0.247[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nStenhousemuir[0.049]; Livingston[0.046]; Brechin[0.046]; Inverness[0.045]; Dumbarton[0.045]; Stranraer[0.044]; Stirling[0.044]; Division[0.044]; Thistle[0.044]; Park[0.044]; Alloa[0.044]; Ross[0.044]; Arbroath[0.043]; Forfar[0.043]; Ayr[0.043]; Hamilton[0.043]; Albion[0.043]; Clyde[0.043]; East[0.042]; Berwick[0.041]; Montrose[0.040]; South[0.040]; County[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Dundee ==> ENTITY: \u001b[32mDundee F.C.\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.186:0.186[0]; log p(e|m)= -1.366:-1.366[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nAirdrieonians[0.047]; Livingston[0.044]; League[0.043]; Fife[0.043]; Greenock[0.043]; Scottish[0.043]; Scottish[0.043]; league[0.043]; played[0.043]; matches[0.043]; Falkirk[0.042]; Partick[0.042]; Stirling[0.042]; Clydebank[0.041]; Soccer[0.041]; London[0.041]; Ayr[0.040]; Saturday[0.040]; South[0.040]; East[0.039]; won[0.039]; Queen[0.039]; Division[0.039]; Division[0.039]; \t\n\r [============================================>.] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 222ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4746/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1142testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Beijing ==> ENTITY: \u001b[32mBeijing\u001b[39m\t\nSCORES: global= 0.275:0.275[0]; local()= 0.149:0.149[0]; log p(e|m)= -0.042:-0.042[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.047]; China[0.047]; Chinese[0.045]; Wuhan[0.044]; Wuhan[0.044]; central[0.044]; commercial[0.042]; September[0.042]; million[0.041]; Hubei[0.041]; cities[0.041]; cities[0.041]; city[0.041]; city[0.041]; capital[0.041]; Saturday[0.041]; Daily[0.041]; use[0.041]; use[0.041]; province[0.039]; entering[0.039]; said[0.038]; said[0.038]; said[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = China Daily\u001b[39m ==> ENTITY: \u001b[32mChina Daily\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.185:0.185[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChinese[0.051]; newspaper[0.051]; newspaper[0.051]; China[0.051]; Beijing[0.050]; Guangzhou[0.046]; commercial[0.045]; said[0.045]; said[0.045]; said[0.045]; said[0.045]; Guangdong[0.045]; cities[0.044]; cities[0.044]; Wuhan[0.044]; Wuhan[0.044]; Hubei[0.043]; September[0.043]; southern[0.042]; province[0.042]; province[0.042]; departments[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Guangzhou\u001b[39m ==> ENTITY: \u001b[32mGuangzhou\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.124:0.124[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGuangzhou[0.047]; China[0.047]; province[0.047]; province[0.047]; southern[0.046]; Guangdong[0.046]; Hubei[0.044]; Wuhan[0.043]; Wuhan[0.043]; central[0.043]; city[0.043]; city[0.043]; containers[0.042]; containers[0.042]; containers[0.042]; containers[0.042]; containers[0.042]; containers[0.042]; entering[0.042]; capital[0.042]; capital[0.042]; industrial[0.041]; Saturday[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wuhan\u001b[39m ==> ENTITY: \u001b[32mWuhan\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.229:0.229[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWuhan[0.045]; Beijing[0.045]; China[0.044]; Guangzhou[0.044]; Guangzhou[0.044]; Chinese[0.044]; Hubei[0.043]; Guangdong[0.043]; cities[0.042]; cities[0.042]; capital[0.041]; capital[0.041]; city[0.041]; city[0.041]; central[0.041]; southern[0.041]; September[0.040]; efforts[0.040]; province[0.040]; province[0.040]; Daily[0.039]; million[0.039]; commercial[0.039]; Saturday[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Guangdong\u001b[39m ==> ENTITY: \u001b[32mGuangdong\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWuhan[0.048]; Wuhan[0.048]; Guangzhou[0.048]; Guangzhou[0.048]; China[0.047]; province[0.047]; province[0.047]; southern[0.047]; Hubei[0.045]; central[0.043]; Saturday[0.042]; Daily[0.042]; capital[0.041]; capital[0.041]; city[0.041]; city[0.041]; September[0.041]; use[0.041]; uses[0.040]; ones[0.040]; industrial[0.040]; punish[0.039]; day[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Chinese ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.261:0.261[0]; local()= 0.137:0.137[0]; log p(e|m)= -1.419:-1.419[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.050]; China[0.050]; Beijing[0.046]; cities[0.044]; cities[0.044]; province[0.043]; Hubei[0.042]; Wuhan[0.041]; Wuhan[0.041]; Daily[0.041]; use[0.041]; use[0.041]; ones[0.041]; central[0.040]; said[0.040]; said[0.040]; said[0.040]; capital[0.040]; newspaper[0.040]; newspaper[0.040]; punish[0.040]; million[0.039]; commercial[0.039]; entering[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Hubei\u001b[39m ==> ENTITY: \u001b[32mHubei\u001b[39m\t\nSCORES: global= 0.282:0.282[0]; local()= 0.216:0.216[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChinese[0.045]; Guangdong[0.044]; Wuhan[0.044]; Wuhan[0.044]; China[0.044]; China[0.044]; Beijing[0.044]; southern[0.044]; Guangzhou[0.043]; Guangzhou[0.043]; province[0.043]; province[0.043]; central[0.042]; city[0.040]; city[0.040]; cities[0.040]; cities[0.040]; efforts[0.040]; capital[0.039]; capital[0.039]; Daily[0.039]; September[0.038]; use[0.038]; use[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Wuhan\u001b[39m ==> ENTITY: \u001b[32mWuhan\u001b[39m\t\nSCORES: global= 0.288:0.288[0]; local()= 0.233:0.233[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nWuhan[0.045]; Beijing[0.045]; China[0.044]; China[0.044]; Guangzhou[0.044]; Guangzhou[0.044]; Chinese[0.044]; Hubei[0.042]; Guangdong[0.042]; cities[0.041]; cities[0.041]; capital[0.041]; capital[0.041]; city[0.041]; city[0.041]; central[0.041]; southern[0.041]; September[0.040]; efforts[0.040]; province[0.040]; province[0.040]; sell[0.039]; Daily[0.039]; million[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Guangzhou\u001b[39m ==> ENTITY: \u001b[32mGuangzhou\u001b[39m\t\nSCORES: global= 0.273:0.273[0]; local()= 0.125:0.125[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nGuangzhou[0.052]; province[0.052]; province[0.052]; southern[0.051]; Guangdong[0.051]; Hubei[0.049]; Wuhan[0.048]; Wuhan[0.048]; central[0.047]; city[0.047]; city[0.047]; containers[0.046]; containers[0.046]; containers[0.046]; containers[0.046]; containers[0.046]; entering[0.046]; capital[0.046]; capital[0.046]; industrial[0.045]; Saturday[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = China ==> ENTITY: \u001b[32mChina\u001b[39m\t\nSCORES: global= 0.265:0.265[0]; local()= 0.144:0.144[0]; log p(e|m)= -0.157:-0.157[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nChina[0.053]; Chinese[0.050]; Beijing[0.049]; city[0.045]; city[0.045]; cities[0.045]; cities[0.045]; Wuhan[0.044]; September[0.044]; Hubei[0.043]; province[0.043]; efforts[0.043]; Daily[0.042]; commercial[0.042]; central[0.041]; capital[0.041]; Saturday[0.041]; departments[0.041]; industrial[0.041]; said[0.041]; said[0.041]; said[0.041]; entering[0.040]; \t\n\r [============================================>.] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 174ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4756/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1049testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Bulgaria ==> ENTITY: \u001b[32mBulgaria national football team\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.103:0.103[0]; log p(e|m)= -2.718:-2.718[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nBulgaria[0.057]; European[0.055]; European[0.055]; Soccer[0.054]; qualifier[0.053]; Qualifier[0.051]; Israel[0.049]; Israel[0.049]; Israel[0.049]; championship[0.048]; group[0.048]; Saturday[0.046]; halftime[0.045]; Herzliya[0.045]; Result[0.045]; Scorers[0.043]; Attendance[0.042]; Beat[0.042]; Haim[0.042]; Nir[0.041]; Hajaj[0.040]; \t\n\n====> \u001b[31mINCORRECT ANNOTATION\u001b[39m : mention = \u001b[36mIsrael\u001b[39m ==> ENTITIES (OURS/GOLD): \u001b[31mIsrael national football team\u001b[39m <---> \u001b[32mIsrael\u001b[39m\t\nSCORES: global= 0.258:0.252[\u001b[31m0.006\u001b[39m]; local()= 0.164:0.109[\u001b[31m0.055\u001b[39m]; log p(e|m)= -3.576:-0.213[\u001b[32m3.362\u001b[39m]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.064]; Israel[0.064]; Soccer[0.055]; qualifier[0.053]; Nir[0.053]; Herzliya[0.053]; Haim[0.051]; European[0.050]; European[0.050]; Qualifier[0.049]; Bulgaria[0.049]; Bulgaria[0.049]; championship[0.048]; Saturday[0.047]; Hajaj[0.047]; group[0.045]; halftime[0.044]; Attendance[0.044]; Beat[0.042]; Scorers[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Nir Sivilia\u001b[39m ==> ENTITY: \u001b[32mNir Sivilia\u001b[39m\t\nSCORES: global= 0.270:0.270[0]; local()= 0.147:0.147[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nHerzliya[0.061]; Israel[0.060]; Israel[0.060]; Israel[0.060]; Haim[0.055]; Hajaj[0.054]; qualifier[0.049]; Qualifier[0.049]; Soccer[0.048]; halftime[0.048]; Scorers[0.047]; Attendance[0.047]; Bulgaria[0.047]; Bulgaria[0.047]; European[0.045]; European[0.045]; Result[0.045]; championship[0.045]; Saturday[0.044]; Beat[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = European ==> ENTITY: \u001b[32mEurope\u001b[39m\t\nSCORES: global= 0.241:0.241[0]; local()= 0.010:0.010[0]; log p(e|m)= -0.256:-0.256[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nEuropean[0.059]; Soccer[0.057]; qualifier[0.055]; Bulgaria[0.054]; Bulgaria[0.054]; Qualifier[0.054]; group[0.052]; championship[0.052]; Israel[0.050]; Israel[0.050]; Israel[0.050]; Saturday[0.049]; Attendance[0.048]; Herzliya[0.047]; Result[0.046]; Beat[0.046]; halftime[0.045]; Scorers[0.045]; Nir[0.044]; Haim[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Israel ==> ENTITY: \u001b[32mIsrael national football team\u001b[39m\t\nSCORES: global= 0.258:0.258[0]; local()= 0.164:0.164[0]; log p(e|m)= -3.576:-3.576[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nIsrael[0.064]; Israel[0.064]; Soccer[0.055]; qualifier[0.053]; Nir[0.053]; Herzliya[0.053]; Haim[0.051]; European[0.050]; European[0.050]; Qualifier[0.049]; Bulgaria[0.049]; Bulgaria[0.049]; championship[0.048]; Saturday[0.047]; Hajaj[0.047]; group[0.045]; halftime[0.044]; Attendance[0.044]; Beat[0.042]; Scorers[0.042]; \t\n\r [============================================>.] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 139ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4763/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1088testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Hong Kong ==> ENTITY: \u001b[32mHong Kong\u001b[39m\t\nSCORES: global= 0.243:0.243[0]; local()= 0.072:0.072[0]; log p(e|m)= -0.224:-0.224[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmatch[0.045]; Kong[0.044]; Jansher[0.044]; Jansher[0.044]; Jansher[0.044]; Hong[0.043]; game[0.043]; Squash[0.043]; World[0.042]; player[0.042]; Open[0.042]; return[0.041]; Pakistani[0.041]; Australian[0.041]; winning[0.041]; number[0.040]; Hill[0.040]; Hill[0.040]; Hill[0.040]; Hill[0.040]; Hill[0.040]; world[0.040]; semifinals[0.040]; career[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jansher\u001b[39m ==> ENTITY: \u001b[32mJansher Khan\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.210:0.210[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJansher[0.046]; Jansher[0.046]; Jansher[0.046]; Jansher[0.046]; squash[0.043]; squash[0.043]; squash[0.043]; player[0.042]; Open[0.042]; Squash[0.042]; tennis[0.041]; time[0.041]; eighth[0.041]; second[0.041]; players[0.041]; final[0.041]; title[0.041]; played[0.041]; played[0.041]; today[0.040]; fit[0.039]; think[0.039]; generally[0.039]; added[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jansher\u001b[39m ==> ENTITY: \u001b[32mJansher Khan\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.214:0.214[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJansher[0.046]; Jansher[0.046]; Jansher[0.046]; squash[0.043]; squash[0.043]; tournament[0.043]; Open[0.042]; Open[0.042]; Squash[0.042]; tennis[0.042]; eighth[0.041]; second[0.041]; fourth[0.041]; players[0.041]; win[0.041]; final[0.041]; title[0.041]; played[0.041]; game[0.040]; semifinals[0.040]; beat[0.039]; beat[0.039]; beat[0.039]; think[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Peter Nicol\u001b[39m ==> ENTITY: \u001b[32mPeter Nicol\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.204:0.204[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.044]; Jansher[0.044]; Jansher[0.044]; Jansher[0.044]; Jansher[0.044]; Squash[0.043]; squash[0.043]; squash[0.043]; squash[0.043]; win[0.042]; title[0.042]; tennis[0.042]; Open[0.041]; Open[0.041]; Scotland[0.041]; players[0.041]; fourth[0.041]; second[0.040]; sponsors[0.040]; Professional[0.040]; played[0.040]; semifinals[0.040]; final[0.039]; eighth[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Professional Squash Association\u001b[39m ==> ENTITY: \u001b[32mProfessional Squash Association\u001b[39m\t\nSCORES: global= 0.293:0.293[0]; local()= 0.145:0.145[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntournament[0.047]; tennis[0.045]; player[0.043]; players[0.043]; Jansher[0.043]; Jansher[0.043]; Jansher[0.043]; Jansher[0.043]; Jansher[0.043]; sponsors[0.042]; Open[0.042]; Open[0.042]; squash[0.041]; squash[0.041]; squash[0.041]; win[0.041]; semifinals[0.041]; deal[0.040]; title[0.040]; played[0.040]; final[0.039]; today[0.039]; banned[0.039]; build[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jansher\u001b[39m ==> ENTITY: \u001b[32mJansher Khan\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.195:0.195[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJansher[0.047]; Jansher[0.047]; Jansher[0.047]; Jansher[0.047]; Pakistani[0.044]; squash[0.044]; player[0.043]; player[0.043]; time[0.042]; played[0.041]; today[0.041]; career[0.041]; game[0.040]; game[0.040]; fit[0.040]; generally[0.040]; won[0.040]; added[0.040]; return[0.040]; Australian[0.039]; upset[0.039]; claimed[0.039]; winning[0.039]; bit[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Eyles\u001b[39m ==> ENTITY: \u001b[32mRodney Eyles\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.173:0.173[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJansher[0.045]; Jansher[0.045]; Jansher[0.045]; Jansher[0.045]; Australian[0.045]; Open[0.043]; Open[0.043]; tournament[0.043]; final[0.042]; title[0.042]; semifinals[0.041]; Rodney[0.041]; played[0.041]; fourth[0.041]; Nicol[0.041]; eighth[0.041]; squash[0.040]; Eyles[0.040]; Eyles[0.040]; Eyles[0.040]; Peter[0.040]; second[0.039]; game[0.039]; win[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Scotland ==> ENTITY: \u001b[32mScotland\u001b[39m\t\nSCORES: global= 0.237:0.237[0]; local()= 0.045:0.045[0]; log p(e|m)= -0.229:-0.229[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nplayed[0.047]; players[0.046]; tournament[0.046]; game[0.046]; second[0.045]; fourth[0.045]; eighth[0.045]; final[0.045]; win[0.045]; title[0.043]; Jansher[0.043]; Jansher[0.043]; Jansher[0.043]; Jansher[0.043]; Squash[0.043]; semifinals[0.042]; Nicol[0.042]; Australian[0.042]; plays[0.042]; biggest[0.041]; tennis[0.041]; build[0.041]; Peter[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jansher\u001b[39m ==> ENTITY: \u001b[32mJansher Khan\u001b[39m\t\nSCORES: global= 0.307:0.307[0]; local()= 0.230:0.230[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJansher[0.045]; Jansher[0.045]; Jansher[0.045]; Jansher[0.045]; Jansher[0.045]; squash[0.042]; squash[0.042]; squash[0.042]; tournament[0.042]; player[0.041]; Open[0.041]; Open[0.041]; Squash[0.041]; tennis[0.041]; time[0.041]; eighth[0.040]; second[0.040]; fourth[0.040]; players[0.040]; win[0.040]; final[0.040]; title[0.040]; played[0.040]; game[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jansher\u001b[39m ==> ENTITY: \u001b[32mJansher Khan\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJansher[0.047]; Jansher[0.047]; Khan[0.046]; Pakistani[0.045]; match[0.044]; squash[0.044]; player[0.043]; Open[0.043]; Squash[0.043]; world[0.043]; number[0.043]; World[0.042]; game[0.041]; semifinals[0.041]; Saturday[0.040]; return[0.040]; Australian[0.040]; claimed[0.040]; winning[0.039]; Champion[0.038]; said[0.038]; good[0.038]; knew[0.037]; decision[0.037]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jansher Khan\u001b[39m ==> ENTITY: \u001b[32mJansher Khan\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.187:0.187[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJansher[0.049]; Jansher[0.049]; Pakistani[0.047]; match[0.046]; squash[0.046]; player[0.045]; Open[0.045]; Squash[0.045]; world[0.045]; number[0.045]; World[0.044]; career[0.043]; game[0.042]; semifinals[0.042]; Saturday[0.042]; return[0.042]; Australian[0.042]; claimed[0.041]; winning[0.041]; suspensions[0.040]; Champion[0.040]; said[0.040]; good[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Eyles\u001b[39m ==> ENTITY: \u001b[32mRodney Eyles\u001b[39m\t\nSCORES: global= 0.298:0.298[0]; local()= 0.180:0.180[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJansher[0.048]; Jansher[0.048]; Jansher[0.048]; Jansher[0.048]; Australian[0.048]; Squash[0.047]; players[0.046]; tennis[0.046]; Open[0.046]; Open[0.046]; tournament[0.046]; final[0.045]; title[0.045]; semifinals[0.044]; Rodney[0.044]; played[0.044]; fourth[0.044]; Nicol[0.044]; eighth[0.044]; Professional[0.043]; squash[0.043]; squash[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australian ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.080:0.080[0]; log p(e|m)= -0.491:-0.491[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ntennis[0.063]; tournament[0.057]; Open[0.056]; Open[0.056]; player[0.055]; Peter[0.052]; players[0.052]; final[0.051]; played[0.051]; second[0.051]; semifinals[0.051]; win[0.051]; game[0.051]; title[0.051]; Squash[0.051]; fourth[0.050]; squash[0.050]; squash[0.050]; squash[0.050]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jansher\u001b[39m ==> ENTITY: \u001b[32mJansher Khan\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.186:0.186[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJansher[0.047]; Jansher[0.047]; Jansher[0.047]; Jansher[0.047]; Pakistani[0.044]; squash[0.044]; squash[0.044]; player[0.043]; player[0.043]; time[0.042]; played[0.041]; today[0.041]; career[0.041]; game[0.040]; fit[0.040]; generally[0.040]; won[0.040]; added[0.040]; return[0.039]; Australian[0.039]; upset[0.039]; bit[0.038]; suspensions[0.038]; rhythm[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jansher\u001b[39m ==> ENTITY: \u001b[32mJansher Khan\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.213:0.213[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJansher[0.047]; Jansher[0.047]; Jansher[0.047]; squash[0.043]; tournament[0.043]; Open[0.042]; Open[0.042]; Squash[0.042]; eighth[0.042]; second[0.042]; fourth[0.042]; win[0.041]; final[0.041]; title[0.041]; played[0.041]; game[0.040]; semifinals[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; like[0.039]; Australian[0.039]; great[0.039]; action[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Australian ==> ENTITY: \u001b[32mAustralia\u001b[39m\t\nSCORES: global= 0.244:0.244[0]; local()= 0.060:0.060[0]; log p(e|m)= -0.491:-0.491[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nOpen[0.053]; match[0.053]; player[0.052]; number[0.048]; semifinals[0.048]; world[0.048]; game[0.048]; Squash[0.048]; Champion[0.048]; career[0.047]; squash[0.047]; Jansher[0.047]; Jansher[0.047]; Jansher[0.047]; Saturday[0.046]; Khan[0.046]; Anthony[0.046]; World[0.046]; winning[0.046]; referee[0.045]; Pakistani[0.045]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Rodney Eyles\u001b[39m ==> ENTITY: \u001b[32mRodney Eyles\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJansher[0.048]; Jansher[0.048]; Jansher[0.048]; Jansher[0.048]; Jansher[0.048]; Jansher[0.048]; Australian[0.048]; Squash[0.046]; player[0.046]; players[0.046]; tennis[0.046]; Open[0.045]; Open[0.045]; tournament[0.045]; final[0.045]; title[0.044]; semifinals[0.044]; played[0.043]; fourth[0.043]; Nicol[0.043]; eighth[0.043]; Professional[0.043]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jansher\u001b[39m ==> ENTITY: \u001b[32mJansher Khan\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.211:0.211[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJansher[0.048]; Jansher[0.048]; Jansher[0.048]; Jansher[0.048]; Khan[0.047]; Pakistani[0.045]; match[0.045]; squash[0.045]; player[0.044]; Open[0.044]; Squash[0.043]; world[0.043]; number[0.043]; World[0.042]; played[0.042]; career[0.042]; game[0.041]; game[0.041]; semifinals[0.041]; generally[0.041]; won[0.041]; Saturday[0.040]; return[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Jansher\u001b[39m ==> ENTITY: \u001b[32mJansher Khan\u001b[39m\t\nSCORES: global= 0.306:0.306[0]; local()= 0.206:0.206[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJansher[0.047]; Jansher[0.047]; Jansher[0.047]; squash[0.044]; tournament[0.044]; Open[0.043]; Open[0.043]; eighth[0.042]; second[0.042]; fourth[0.042]; win[0.042]; final[0.042]; title[0.041]; played[0.041]; game[0.040]; semifinals[0.040]; beat[0.040]; beat[0.040]; beat[0.040]; like[0.039]; Australian[0.039]; great[0.039]; action[0.038]; Scotland[0.038]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Eyles\u001b[39m ==> ENTITY: \u001b[32mRodney Eyles\u001b[39m\t\nSCORES: global= 0.297:0.297[0]; local()= 0.183:0.183[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nJansher[0.046]; Jansher[0.046]; Jansher[0.046]; Jansher[0.046]; Jansher[0.046]; Jansher[0.046]; Australian[0.046]; Squash[0.044]; player[0.044]; players[0.044]; tennis[0.044]; Open[0.043]; Open[0.043]; tournament[0.043]; final[0.043]; title[0.042]; semifinals[0.042]; Rodney[0.042]; played[0.041]; fourth[0.041]; Nicol[0.041]; eighth[0.041]; Professional[0.041]; \t\n\r [============================================>.] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b ETA: 39ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4783/4791 \r\n\t\n\u001b[34m============================================\u001b[39m\t\n\u001b[34m============ DOC : 1150testa ================\u001b[39m\t\n\u001b[34m============================================\u001b[39m\t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Detroit ==> ENTITY: \u001b[32mDetroit\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.035:0.035[0]; log p(e|m)= -0.217:-0.217[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \ncars[0.043]; launched[0.043]; Officials[0.043]; Saturday[0.043]; air[0.043]; air[0.043]; air[0.043]; air[0.043]; federal[0.042]; million[0.042]; number[0.042]; plan[0.042]; auto[0.042]; agency[0.042]; agency[0.042]; agencies[0.041]; enforcement[0.041]; automotive[0.041]; work[0.040]; reducing[0.040]; bags[0.040]; bags[0.040]; bags[0.040]; force[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Interstate 95 ==> ENTITY: \u001b[32mInterstate 95\u001b[39m\t\nSCORES: global= 0.247:0.247[0]; local()= 0.139:0.139[0]; log p(e|m)= -1.146:-1.146[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInterstates[0.050]; Interstate[0.050]; interstate[0.049]; portion[0.046]; East[0.045]; highways[0.045]; miles[0.044]; Coast[0.044]; Officials[0.043]; busiest[0.043]; nation[0.043]; driving[0.042]; involves[0.042]; trucks[0.042]; trucks[0.042]; billboards[0.041]; hour[0.041]; accident[0.041]; work[0.041]; agencies[0.041]; occurred[0.041]; reduced[0.041]; away[0.041]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = California ==> ENTITY: \u001b[32mCalifornia\u001b[39m\t\nSCORES: global= 0.263:0.263[0]; local()= 0.070:0.070[0]; log p(e|m)= -0.087:-0.087[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nnation[0.049]; miles[0.049]; industry[0.048]; industry[0.048]; Coast[0.047]; seat[0.047]; East[0.045]; National[0.045]; interstate[0.045]; adults[0.045]; percent[0.045]; highways[0.045]; Interstate[0.045]; Interstate[0.045]; trucks[0.045]; size[0.044]; position[0.044]; Interstates[0.044]; weekend[0.044]; portion[0.044]; driving[0.044]; Midwest[0.044]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Interstate 5 ==> ENTITY: \u001b[32mInterstate 5\u001b[39m\t\nSCORES: global= 0.255:0.255[0]; local()= 0.136:0.136[0]; log p(e|m)= -0.543:-0.543[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInterstates[0.048]; Interstate[0.047]; interstate[0.046]; highways[0.045]; portion[0.044]; California[0.043]; Coast[0.042]; miles[0.042]; East[0.041]; driving[0.041]; nation[0.041]; billboards[0.041]; hour[0.041]; agencies[0.041]; involves[0.041]; reduced[0.040]; funded[0.040]; Midwest[0.040]; pediatricians[0.040]; busiest[0.040]; Officials[0.040]; auto[0.040]; Kids[0.039]; car[0.039]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Labor Day\u001b[39m ==> ENTITY: \u001b[32mLabor Day\u001b[39m\t\nSCORES: global= 0.269:0.269[0]; local()= 0.104:0.104[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nweekend[0.046]; National[0.044]; Coast[0.043]; work[0.043]; agencies[0.043]; busiest[0.042]; portion[0.042]; force[0.042]; nation[0.042]; East[0.042]; miles[0.042]; Interstate[0.041]; Interstate[0.041]; hour[0.041]; interstate[0.041]; involves[0.041]; said[0.041]; said[0.041]; said[0.041]; enforcement[0.041]; Interstates[0.041]; campaign[0.041]; number[0.040]; media[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = Midwest ==> ENTITY: \u001b[32mMidwestern United States\u001b[39m\t\nSCORES: global= 0.274:0.274[0]; local()= 0.099:0.099[0]; log p(e|m)= -0.031:-0.031[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nCoast[0.046]; nation[0.043]; East[0.043]; industry[0.043]; industry[0.043]; portion[0.042]; interstate[0.042]; funded[0.042]; boards[0.042]; miles[0.042]; National[0.042]; Labor[0.041]; adults[0.041]; Interstates[0.041]; hour[0.041]; introducing[0.041]; percent[0.041]; children[0.041]; children[0.041]; children[0.041]; highways[0.041]; Officials[0.041]; enforcement[0.041]; California[0.040]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : \u001b[33mmention = Interstates 80\u001b[39m ==> ENTITY: \u001b[32mInterstate 80\u001b[39m\t\nSCORES: global= 0.290:0.290[0]; local()= 0.142:0.142[0]; log p(e|m)= 0.000:0.000[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nInterstate[0.050]; Interstate[0.050]; California[0.050]; portion[0.050]; interstate[0.048]; highways[0.048]; miles[0.047]; Coast[0.045]; East[0.045]; nation[0.045]; National[0.045]; reduced[0.044]; involves[0.044]; driving[0.044]; funded[0.044]; Midwest[0.044]; pediatricians[0.044]; Officials[0.043]; Kids[0.043]; automotive[0.043]; injuries[0.043]; weekend[0.042]; \t\n\n====> \u001b[32mCORRECT ANNOTATION\u001b[39m : mention = U.S. ==> ENTITY: \u001b[32mUnited States\u001b[39m\t\nSCORES: global= 0.260:0.260[0]; local()= 0.054:0.054[0]; log p(e|m)= -0.140:-0.140[0]\t\n\nTop context words (sorted by attention weight, only non-zero weights - top R words): \nmillion[0.048]; number[0.047]; launched[0.044]; law[0.043]; cars[0.042]; federal[0.042]; air[0.041]; air[0.041]; air[0.041]; ads[0.041]; agencies[0.041]; Officials[0.041]; agency[0.041]; agency[0.041]; Saturday[0.041]; Detroit[0.041]; enforcement[0.041]; media[0.041]; plan[0.041]; trucks[0.040]; campaign[0.040]; said[0.040]; government[0.040]; aimed[0.040]; \t\n\r [=============================================>] \b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b Tot: 22s803ms | Step: 4ms\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b 4791/4791 \n\r==> time to test 1 sample = 4.7597889610582ms\t\n==> \u001b[31maida-A\u001b[39m aida-A ; EPOCH = 307: Micro recall = 91.00% ; Micro F1 = \u001b[32m91.01%\u001b[39m\t\n ===> entity frequency stats :\t\nfreq = 11-20 : num = 5 ; correctly classified = 4 ; perc = 80.00\t\nfreq = 2-5 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\nfreq = 50+ : num = 4590 ; correctly classified = 4315 ; perc = 94.01\t\nfreq = 1 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\nfreq = 21-50 : num = 38 ; correctly classified = 36 ; perc = 94.74\t\nfreq = 0 : num = 6 ; correctly classified = 5 ; perc = 83.33\t\nfreq = 6-10 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\n\t\n ===> entity p(e|m) stats :\t\np(e|m) = <=0.001 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\np(e|m) = 0.01-0.03 : num = 174 ; correctly classified = 158 ; perc = 90.80\t\np(e|m) = 0.003-0.01 : num = 55 ; correctly classified = 42 ; perc = 76.36\t\np(e|m) = 0.3+ : num = 3740 ; correctly classified = 3589 ; perc = 95.96\t\np(e|m) = 0.1-0.3 : num = 387 ; correctly classified = 342 ; perc = 88.37\t\np(e|m) = 0.03-0.1 : num = 283 ; correctly classified = 229 ; perc = 80.92\t\np(e|m) = 0.001-0.003 : num = 0 ; correctly classified = 0 ; perc = 0.00\t\n num_mentions_w/o_gold_ent_in_candidates = 152 total num mentions in dataset = 4791\t\n percentage_mentions_w/o_gold_ent_in_candidates = 3.17%; percentage_mentions_solved : both_pem_ours = 71.45%; only_pem_not_ours = 2.38%; only_ours_not_pem = 19.56%; not_ours_not_pem = 3.44%\t\n" }, { "path": "deep-ed/deep-ed-master/utils/logger.lua", "content": "--[[ Logger: a simple class to log symbols during training,\n and automate plot generation\nExample:\n logger = optim.Logger('somefile.log') -- file to save stuff\n for i = 1,N do -- log some symbols during\n train_error = ... -- training/testing\n test_error = ...\n logger:add{['training error'] = train_error,\n ['test error'] = test_error}\n end\n logger:style{['training error'] = '-', -- define styles for plots\n ['test error'] = '-'}\n logger:plot() -- and plot\n---- OR ---\n logger = optim.Logger('somefile.log') -- file to save stuff\n logger:setNames{'training error', 'test error'}\n for i = 1,N do -- log some symbols during\n train_error = ... -- training/testing\n test_error = ...\n logger:add{train_error, test_error}\n end\n logger:style{'-', '-'} -- define styles for plots\n logger:plot() -- and plot\n]]\nrequire 'xlua'\nlocal Logger = torch.class('Logger')\n\nfunction Logger:__init(filename, timestamp)\n if filename then\n self.name = filename\n os.execute('mkdir ' .. (sys.uname() ~= 'windows' and '-p ' or '') .. ' \"' .. paths.dirname(filename) .. '\"')\n if timestamp then\n -- append timestamp to create unique log file\n filename = filename .. '-'..os.date(\"%Y_%m_%d_%X\")\n end\n self.file = io.open(filename,'w')\n self.epsfile = self.name .. '.eps'\n else\n self.file = io.stdout\n self.name = 'stdout'\n print(' warning: no path provided, logging to std out')\n end\n self.empty = true\n self.symbols = {}\n self.styles = {}\n self.names = {}\n self.idx = {}\n self.figure = nil\n self.showPlot = false\n self.plotRawCmd = nil\n self.defaultStyle = '+'\nend\n\nfunction Logger:setNames(names)\n self.names = names\n self.empty = false\n self.nsymbols = #names\n for k,key in pairs(names) do\n self.file:write(key .. '\\t')\n self.symbols[k] = {}\n self.styles[k] = {self.defaultStyle}\n self.idx[key] = k\n end\n self.file:write('\\n')\n self.file:flush()\nend\n\nfunction Logger:add(symbols)\n -- (1) first time ? print symbols' names on first row\n if self.empty then\n self.empty = false\n self.nsymbols = #symbols\n for k,val in pairs(symbols) do\n self.file:write(k .. '\\t')\n self.symbols[k] = {}\n self.styles[k] = {self.defaultStyle}\n self.names[k] = k\n end\n self.idx = self.names\n self.file:write('\\n')\n end\n -- (2) print all symbols on one row\n for k,val in pairs(symbols) do\n if type(val) == 'number' then\n self.file:write(string.format('%11.4e',val) .. '\\t')\n elseif type(val) == 'string' then\n self.file:write(val .. '\\t')\n else\n xlua.error('can only log numbers and strings', 'Logger')\n end\n end\n self.file:write('\\n')\n self.file:flush()\n -- (3) save symbols in internal table\n for k,val in pairs(symbols) do\n table.insert(self.symbols[k], val)\n end\nend\n\nfunction Logger:style(symbols)\n for name,style in pairs(symbols) do\n if type(style) == 'string' then\n self.styles[name] = {style}\n elseif type(style) == 'table' then\n self.styles[name] = style\n else\n xlua.error('style should be a string or a table of strings','Logger')\n end\n end\nend\n\nfunction Logger:plot(ylabel, xlabel)\n if not xlua.require('gnuplot') then\n if not self.warned then\n print(' warning: cannot plot with this version of Torch')\n self.warned = true\n end\n return\n end\n local plotit = false\n local plots = {}\n local plotsymbol =\n function(name,list)\n if #list > 1 then\n local nelts = #list\n local plot_y = torch.Tensor(nelts)\n for i = 1,nelts do\n plot_y[i] = list[i]\n end\n for _,style in ipairs(self.styles[name]) do\n table.insert(plots, {self.names[name], plot_y, style})\n end\n plotit = true\n end\n end\n\n -- plot all symbols\n for name,list in pairs(self.symbols) do\n plotsymbol(name,list)\n end\n\n if plotit then\n if self.showPlot then\n self.figure = gnuplot.figure(self.figure)\n gnuplot.plot(plots)\n gnuplot.xlabel(xlabel)\n gnuplot.ylabel(ylabel)\n if self.plotRawCmd then gnuplot.raw(self.plotRawCmd) end\n gnuplot.grid('on')\n gnuplot.title(banner)\n end\n\n if self.epsfile then\n os.execute('rm -f \"' .. self.epsfile .. '\"')\n local epsfig = gnuplot.epsfigure(self.epsfile)\n gnuplot.plot(plots)\n gnuplot.xlabel(xlabel)\n gnuplot.ylabel(ylabel) \n if self.plotRawCmd then gnuplot.raw(self.plotRawCmd) end\n gnuplot.grid('on')\n gnuplot.title(banner)\n gnuplot.plotflush()\n gnuplot.close(epsfig)\n end\n end\nend" }, { "path": "deep-ed/deep-ed-master/utils/optim/adadelta_mem.lua", "content": "-- Memory optimized implementation of optim.adadelta\n\n--[[ ADADELTA implementation for SGD http://arxiv.org/abs/1212.5701\nARGS:\n- `opfunc` : a function that takes a single input (X), the point of\n evaluation, and returns f(X) and df/dX\n- `x` : the initial point\n- `config` : a table of hyper-parameters\n- `config.rho` : interpolation parameter\n- `config.eps` : for numerical stability\n- `config.weightDecay` : weight decay\n- `state` : a table describing the state of the optimizer; after each\n call the state is modified\n- `state.paramVariance` : vector of temporal variances of parameters\n- `state.accDelta` : vector of accummulated delta of gradients\nRETURN:\n- `x` : the new x vector\n- `f(x)` : the function, evaluated before the update\n]]\nfunction adadelta_mem(opfunc, x, config, state)\n -- (0) get/update state\n if config == nil and state == nil then\n print('no state table, ADADELTA initializing')\n end\n local config = config or {}\n local state = state or config\n local rho = config.rho or 0.9\n local eps = config.eps or 1e-6\n local wd = config.weightDecay or 0\n state.evalCounter = state.evalCounter or 0\n -- (1) evaluate f(x) and df/dx\n local fx,dfdx = opfunc(x)\n\n -- (2) weight decay\n if wd ~= 0 then\n dfdx:add(wd, x)\n end\n\n -- (3) parameter update\n if not state.paramVariance then\n state.paramVariance = torch.Tensor():typeAs(x):resizeAs(dfdx):zero()\n state.delta = torch.Tensor():typeAs(x):resizeAs(dfdx):zero()\n state.accDelta = torch.Tensor():typeAs(x):resizeAs(dfdx):zero()\n end\n state.paramVariance:mul(rho):addcmul(1-rho,dfdx,dfdx)\n state.paramVariance:add(eps):sqrt() -- std\n state.delta:copy(state.accDelta):add(eps):sqrt():cdiv(state.paramVariance):cmul(dfdx)\n x:add(-1, state.delta)\n state.accDelta:mul(rho):addcmul(1-rho, state.delta, state.delta)\n state.paramVariance:cmul(state.paramVariance):csub(eps) -- recompute variance again\n\n -- (4) update evaluation counter\n state.evalCounter = state.evalCounter + 1\n\n -- return x*, f(x) before optimization\n return x,{fx}\nend" }, { "path": "deep-ed/deep-ed-master/utils/optim/adagrad_mem.lua", "content": "-- Memory optimized implementation of optim.adagrad\n\n--[[ ADAGRAD implementation for SGD\nARGS:\n- `opfunc` : a function that takes a single input (X), the point of\n evaluation, and returns f(X) and df/dX\n- `x` : the initial point\n- `state` : a table describing the state of the optimizer; after each\n call the state is modified\n- `state.learningRate` : learning rate\n- `state.paramVariance` : vector of temporal variances of parameters\n- `state.weightDecay` : scalar that controls weight decay\nRETURN:\n- `x` : the new x vector\n- `f(x)` : the function, evaluated before the update\n]]\n\nfunction adagrad_mem(opfunc, x, config, state)\n -- (0) get/update state\n if config == nil and state == nil then\n print('no state table, ADAGRAD initializing')\n end\n local config = config or {}\n local state = state or config\n local lr = config.learningRate or 1e-3\n local lrd = config.learningRateDecay or 0\n local wd = config.weightDecay or 0\n state.evalCounter = state.evalCounter or 0\n local nevals = state.evalCounter\n\n -- (1) evaluate f(x) and df/dx\n local fx,dfdx = opfunc(x)\n\n -- (2) weight decay with a single parameter\n if wd ~= 0 then\n dfdx:add(wd, x)\n end\n\n -- (3) learning rate decay (annealing)\n local clr = lr / (1 + nevals*lrd)\n\n -- (4) parameter update with single or individual learning rates\n if not state.paramVariance then\n state.paramVariance = torch.Tensor():typeAs(x):resizeAs(dfdx):zero()\n end\n \n state.paramVariance:addcmul(1,dfdx,dfdx)\n state.paramVariance:add(1e-10)\n state.paramVariance:sqrt() -- Keeps the std\n x:addcdiv(-clr, dfdx, state.paramVariance)\n state.paramVariance:cmul(state.paramVariance) -- Keeps the variance again\n state.paramVariance:add(-1e-10)\n\n\n -- (5) update evaluation counter\n state.evalCounter = state.evalCounter + 1\n\n -- return x*, f(x) before optimization\n return x,{fx}\nend" }, { "path": "deep-ed/deep-ed-master/utils/optim/rmsprop_mem.lua", "content": "-- Memory optimized implementation of optim.rmsprop\n\n--[[ An implementation of RMSprop\nARGS:\n- 'opfunc' : a function that takes a single input (X), the point\n of a evaluation, and returns f(X) and df/dX\n- 'x' : the initial point\n- 'config` : a table with configuration parameters for the optimizer\n- 'config.learningRate' : learning rate\n- 'config.alpha' : smoothing constant\n- 'config.epsilon' : value with which to initialise m\n- 'config.weightDecay' : weight decay\n- 'state' : a table describing the state of the optimizer;\n after each call the state is modified\n- 'state.m' : leaky sum of squares of parameter gradients,\n- 'state.tmp' : and the square root (with epsilon smoothing)\nRETURN:\n- `x` : the new x vector\n- `f(x)` : the function, evaluated before the update\n]]\n\nfunction rmsprop_mem(opfunc, x, config, state)\n -- (0) get/update state\n local config = config or {}\n local state = state or config\n local lr = config.learningRate or 1e-2\n local alpha = config.alpha or 0.99\n local epsilon = config.epsilon or 1e-8\n local wd = config.weightDecay or 0\n local mfill = config.initialMean or 0\n\n -- (1) evaluate f(x) and df/dx\n local fx, dfdx = opfunc(x)\n\n -- (2) weight decay\n if wd ~= 0 then\n dfdx:add(wd, x)\n end\n\n -- (3) initialize mean square values and square gradient storage\n if not state.m then\n state.m = torch.Tensor():typeAs(x):resizeAs(dfdx):fill(mfill)\n end\n\n -- (4) calculate new (leaky) mean squared values\n state.m:mul(alpha)\n state.m:addcmul(1.0-alpha, dfdx, dfdx)\n\n -- (5) perform update\n state.m:add(epsilon)\n state.m:sqrt()\n x:addcdiv(-lr, dfdx, state.m)\n state.m:cmul(state.m)\n state.m:add(-epsilon)\n\n -- return x*, f(x) before optimization\n return x, {fx}\nend" }, { "path": "deep-ed/deep-ed-master/utils/utils.lua", "content": "function topk(one_dim_tensor, k) \n local bestk, indices = torch.topk(one_dim_tensor, k, true)\n local sorted, newindices = torch.sort(bestk, true)\n local oldindices = torch.LongTensor(k)\n for i = 1,k do\n oldindices[i] = indices[newindices[i]]\n end\n return sorted, oldindices\nend\n\n\nfunction list_with_scores_to_str(list, scores)\n local str = ''\n for i,v in pairs(list) do\n str = str .. list[i] .. '[' .. string.format(\"%.2f\", scores[i]) .. ']; '\n end\n return str\nend\n\nfunction table_len(t)\n local count = 0\n for _ in pairs(t) do count = count + 1 end\n return count\nend\n\n\nfunction split(inputstr, sep)\n if sep == nil then\n sep = \"%s\"\n end\n local t={} ; i=1\n for str in string.gmatch(inputstr, \"([^\"..sep..\"]+)\") do\n t[i] = str\n i = i + 1\n end\n return t\nend\n-- Unit test:\nassert(6 == #split('aa_bb cc__dd ee _ _ __ff' , '_ '))\n\n\n\nfunction correct_type(data) \n if opt.type == 'float' then return data:float()\n elseif opt.type == 'double' then return data:double()\n elseif string.find(opt.type, 'cuda') then return data:cuda()\n else print('Unsuported type')\n end\nend\n\n-- color fonts:\nfunction red(s) \n return '\\27[31m' .. s .. '\\27[39m'\nend\n\nfunction green(s)\n return '\\27[32m' .. s .. '\\27[39m'\nend\n\nfunction yellow(s)\n return '\\27[33m' .. s .. '\\27[39m'\nend\n\nfunction blue(s)\n return '\\27[34m' .. s .. '\\27[39m'\nend\n\nfunction violet(s)\n return '\\27[35m' .. s .. '\\27[39m'\nend\n\nfunction skyblue(s)\n return '\\27[36m' .. s .. '\\27[39m'\nend\n\n\n\nfunction split_in_words(inputstr)\n local words = {}\n for word in inputstr:gmatch(\"%w+\") do table.insert(words, word) end\n return words\nend\n\n\nfunction first_letter_to_uppercase(s)\n return s:sub(1,1):upper() .. s:sub(2)\nend\n\nfunction modify_uppercase_phrase(s)\n if (s == s:upper()) then\n local words = split_in_words(s:lower())\n local res = {}\n for _,w in pairs(words) do\n table.insert(res, first_letter_to_uppercase(w))\n end\n return table.concat(res, ' ') \n else\n return s\n end\nend\n\nfunction blue_num_str(n)\n return blue(string.format(\"%.3f\", n))\nend\n\n\n\nfunction string_starts(s, m)\n return string.sub(s,1,string.len(m)) == m\nend\n\n-- trim:\nfunction trim1(s)\n return (s:gsub(\"^%s*(.-)%s*$\", \"%1\"))\nend\n\n\nfunction nice_print_red_green(a,b)\n local s = string.format(\"%.3f\", a) .. ':' .. string.format(\"%.3f\", b) .. '['\n if a > b then\n return s .. red(string.format(\"%.3f\", a-b)) .. ']'\n elseif a < b then\n return s .. green(string.format(\"%.3f\", b-a)) .. ']'\n else\n return s .. '0]'\n end\nend\n" }, { "path": "deep-ed/deep-ed-master/words/load_w_freq_and_vecs.lua", "content": "-- Loads all common words in both Wikipedia and Word2vec/Glove , their unigram frequencies and their pre-trained Word2Vec embeddings.\n\n-- To load this as a standalone file do:\n-- th> opt = {word_vecs = 'w2v', root_data_dir = '$DATA_PATH'}\n-- th> dofile 'words/load_w_freq_and_vecs.lua'\n\nif not opt then\n cmd = torch.CmdLine()\n cmd:option('-root_data_dir', '', 'Root path of the data, $DATA_PATH.')\n cmd:text()\n opt = cmd:parse(arg or {})\n assert(opt.root_data_dir ~= '', 'Specify a valid root_data_dir path argument.') \nend\n\n\ndefault_path = opt.root_data_dir .. 'basic_data/wordEmbeddings/'\n\ntds = tds or require 'tds' \ntorch.setdefaulttensortype('torch.FloatTensor')\nif not list_with_scores_to_str then\n dofile 'utils/utils.lua'\nend\nif not is_stop_word_or_number then\n dofile 'words/stop_words.lua'\nend\n\nassert(opt, 'Define opt')\nassert(opt.word_vecs, 'Define opt.word_vecs')\n\nprint('==> Loading common w2v + top freq list of words')\n\nlocal output_t7filename = opt.root_data_dir .. 'generated/common_top_words_freq_vectors_' .. opt.word_vecs .. '.t7'\n\nif paths.filep(output_t7filename) then\n print(' ---> from t7 file.')\n common_w2v_freq_words = torch.load(output_t7filename)\n \nelse\n print(' ---> t7 file NOT found. Loading from disk instead (slower). Out file = ' .. output_t7filename)\n local freq_words = tds.Hash()\n \n print(' word freq index ...')\n local num_freq_words = 1\n local w_freq_file = opt.root_data_dir .. 'generated/word_wiki_freq.txt'\n for line in io.lines(w_freq_file) do\n local parts = split(line, '\\t')\n local w = parts[1]\n local w_f = tonumber(parts[2])\n if not is_stop_word_or_number(w) then\n freq_words[w] = w_f\n num_freq_words = num_freq_words + 1\n end\n end\n\n common_w2v_freq_words = tds.Hash()\n\n print(' word vectors index ...')\n if opt.word_vecs == 'glove' then\n w2v_txtfilename = default_path .. 'Glove/glove.840B.300d.txt'\n local line_num = 0\n for line in io.lines(w2v_txtfilename) do\n line_num = line_num + 1\n if line_num % 200000 == 0 then\n print(' Processed ' .. line_num)\n end\n local parts = split(line, ' ')\n local w = parts[1]\n if freq_words[w] then\n common_w2v_freq_words[w] = 1\n end\n end\n\n else\n assert(opt.word_vecs == 'w2v')\n w2v_binfilename = default_path .. 'Word2Vec/GoogleNews-vectors-negative300.bin' \n local word_vecs_size = 300\n file = torch.DiskFile(w2v_binfilename,'r')\n file:ascii()\n local vocab_size = file:readInt()\n local size = file:readInt()\n assert(size == word_vecs_size, 'Wrong size : ' .. size .. ' vs ' .. word_vecs_size)\n\n function read_string_w2v(file) \n local str = {}\n while true do\n local char = file:readChar()\n if char == 32 or char == 10 or char == 0 then\n break\n else\n str[#str+1] = char\n end\n end\n str = torch.CharStorage(str)\n return str:string()\n end\n\n --Reading Contents\n file:binary()\n local line_num = 0\n for i = 1,vocab_size do\n line_num = line_num + 1\n if line_num % 200000 == 0 then\n print('Processed ' .. line_num)\n end\n local w = read_string_w2v(file)\n local v = torch.FloatTensor(file:readFloat(word_vecs_size))\n if freq_words[w] then\n common_w2v_freq_words[w] = 1\n end\n end\n end\n \n print('Writing t7 File for future usage. Next time loading will be faster!')\n torch.save(output_t7filename, common_w2v_freq_words)\nend\n\n-- Now load the freq and w2v indexes\ndofile 'words/w_freq/w_freq_index.lua'\n" }, { "path": "deep-ed/deep-ed-master/words/stop_words.lua", "content": "all_stop_words = { ['a'] = 1, ['about'] = 1, ['above'] = 1, ['across'] = 1, ['after'] = 1, ['afterwards'] = 1, ['again'] = 1, ['against'] = 1, ['all'] = 1, ['almost'] = 1, ['alone'] = 1, ['along'] = 1, ['already'] = 1, ['also'] = 1, ['although'] = 1, ['always'] = 1, ['am'] = 1, ['among'] = 1, ['amongst'] = 1, ['amoungst'] = 1, ['amount'] = 1, ['an'] = 1, ['and'] = 1, ['another'] = 1, ['any'] = 1, ['anyhow'] = 1, ['anyone'] = 1, ['anything'] = 1, ['anyway'] = 1, ['anywhere'] = 1, ['are'] = 1, ['around'] = 1, ['as'] = 1, ['at'] = 1, ['back'] = 1, ['be'] = 1, ['became'] = 1, ['because'] = 1, ['become'] = 1, ['becomes'] = 1, ['becoming'] = 1, ['been'] = 1, ['before'] = 1, ['beforehand'] = 1, ['behind'] = 1, ['being'] = 1, ['below'] = 1, ['beside'] = 1, ['besides'] = 1, ['between'] = 1, ['beyond'] = 1, ['both'] = 1, ['bottom'] = 1, ['but'] = 1, ['by'] = 1, ['call'] = 1, ['can'] = 1, ['cannot'] = 1, ['cant'] = 1, ['dont'] = 1, ['co'] = 1, ['con'] = 1, ['could'] = 1, ['couldnt'] = 1, ['cry'] = 1, ['de'] = 1, ['describe'] = 1, ['detail'] = 1, ['do'] = 1, ['done'] = 1, ['down'] = 1, ['due'] = 1, ['during'] = 1, ['each'] = 1, ['eg'] = 1, ['eight'] = 1, ['either'] = 1, ['eleven'] = 1, ['else'] = 1, ['elsewhere'] = 1, ['empty'] = 1, ['enough'] = 1, ['etc'] = 1, ['even'] = 1, ['ever'] = 1, ['every'] = 1, ['everyone'] = 1, ['everything'] = 1, ['everywhere'] = 1, ['except'] = 1, ['few'] = 1, ['fifteen'] = 1, ['fify'] = 1, ['fill'] = 1, ['find'] = 1, ['fire'] = 1, ['first'] = 1, ['five'] = 1, ['for'] = 1, ['former'] = 1, ['formerly'] = 1, ['forty'] = 1, ['found'] = 1, ['four'] = 1, ['from'] = 1, ['front'] = 1, ['full'] = 1, ['further'] = 1, ['get'] = 1, ['give'] = 1, ['go'] = 1, ['had'] = 1, ['has'] = 1, ['hasnt'] = 1, ['have'] = 1, ['he'] = 1, ['hence'] = 1, ['her'] = 1, ['here'] = 1, ['hereafter'] = 1, ['hereby'] = 1, ['herein'] = 1, ['hereupon'] = 1, ['hers'] = 1, ['herself'] = 1, ['him'] = 1, ['himself'] = 1, ['his'] = 1, ['how'] = 1, ['however'] = 1, ['hundred'] = 1, ['i'] = 1, ['ie'] = 1, ['if'] = 1, ['in'] = 1, ['inc'] = 1, ['indeed'] = 1, ['interest'] = 1, ['into'] = 1, ['is'] = 1, ['it'] = 1, ['its'] = 1, ['itself'] = 1, ['keep'] = 1, ['last'] = 1, ['latter'] = 1, ['latterly'] = 1, ['least'] = 1, ['less'] = 1, ['ltd'] = 1, ['made'] = 1, ['many'] = 1, ['may'] = 1, ['me'] = 1, ['meanwhile'] = 1, ['might'] = 1, ['mill'] = 1, ['mine'] = 1, ['more'] = 1, ['moreover'] = 1, ['most'] = 1, ['mostly'] = 1, ['move'] = 1, ['much'] = 1, ['must'] = 1, ['my'] = 1, ['myself'] = 1, ['name'] = 1, ['namely'] = 1, ['neither'] = 1, ['never'] = 1, ['nevertheless'] = 1, ['next'] = 1, ['nine'] = 1, ['no'] = 1, ['nobody'] = 1, ['none'] = 1, ['noone'] = 1, ['nor'] = 1, ['not'] = 1, ['nothing'] = 1, ['now'] = 1, ['nowhere'] = 1, ['of'] = 1, ['off'] = 1, ['often'] = 1, ['on'] = 1, ['once'] = 1, ['one'] = 1, ['only'] = 1, ['onto'] = 1, ['or'] = 1, ['other'] = 1, ['others'] = 1, ['otherwise'] = 1, ['our'] = 1, ['ours'] = 1, ['ourselves'] = 1, ['out'] = 1, ['over'] = 1, ['own'] = 1, ['part'] = 1, ['per'] = 1, ['perhaps'] = 1, ['please'] = 1, ['put'] = 1, ['rather'] = 1, ['re'] = 1, ['same'] = 1, ['see'] = 1, ['seem'] = 1, ['seemed'] = 1, ['seeming'] = 1, ['seems'] = 1, ['serious'] = 1, ['several'] = 1, ['she'] = 1, ['should'] = 1, ['show'] = 1, ['side'] = 1, ['since'] = 1, ['sincere'] = 1, ['six'] = 1, ['sixty'] = 1, ['so'] = 1, ['some'] = 1, ['somehow'] = 1, ['someone'] = 1, ['something'] = 1, ['sometime'] = 1, ['sometimes'] = 1, ['somewhere'] = 1, ['still'] = 1, ['such'] = 1, ['system'] = 1, ['take'] = 1, ['ten'] = 1, ['than'] = 1, ['that'] = 1, ['the'] = 1, ['their'] = 1, ['them'] = 1, ['themselves'] = 1, ['then'] = 1, ['thence'] = 1, ['there'] = 1, ['thereafter'] = 1, ['thereby'] = 1, ['therefore'] = 1, ['therein'] = 1, ['thereupon'] = 1, ['these'] = 1, ['they'] = 1, ['thick'] = 1, ['thin'] = 1, ['third'] = 1, ['this'] = 1, ['those'] = 1, ['though'] = 1, ['three'] = 1, ['through'] = 1, ['throughout'] = 1, ['thru'] = 1, ['thus'] = 1, ['to'] = 1, ['together'] = 1, ['too'] = 1, ['top'] = 1, ['toward'] = 1, ['towards'] = 1, ['twelve'] = 1, ['twenty'] = 1, ['two'] = 1, ['un'] = 1, ['under'] = 1, ['until'] = 1, ['up'] = 1, ['upon'] = 1, ['us'] = 1, ['very'] = 1, ['via'] = 1, ['was'] = 1, ['we'] = 1, ['well'] = 1, ['were'] = 1, ['what'] = 1, ['whatever'] = 1, ['when'] = 1, ['whence'] = 1, ['whenever'] = 1, ['where'] = 1, ['whereafter'] = 1, ['whereas'] = 1, ['whereby'] = 1, ['wherein'] = 1, ['whereupon'] = 1, ['wherever'] = 1, ['whether'] = 1, ['which'] = 1, ['while'] = 1, ['whither'] = 1, ['who'] = 1, ['whoever'] = 1, ['whole'] = 1, ['whom'] = 1, ['whose'] = 1, ['why'] = 1, ['will'] = 1, ['with'] = 1, ['within'] = 1, ['without'] = 1, ['would'] = 1, ['yet'] = 1, ['you'] = 1, ['your'] = 1, ['yours'] = 1, ['yourself'] = 1, ['yourselves'] = 1, ['st'] = 1, ['years'] = 1, ['yourselves'] = 1, ['new'] = 1, ['used'] = 1, ['known'] = 1, ['year'] = 1, ['later'] = 1, ['including'] = 1, ['used'] = 1, ['end'] = 1, ['did'] = 1, ['just'] = 1, ['best'] = 1, ['using'] = 1}\n\nfunction is_stop_word_or_number(w)\n assert(type(w) == 'string', w)\n if (all_stop_words[w:lower()]) or tonumber(w) or w:len() <= 1 then\n return true\n else\n return false\n end\nend" }, { "path": "deep-ed/deep-ed-master/words/w2v/glove_reader.lua", "content": "local M = torch.zeros(total_num_words(), word_vecs_size):float()\n\n--Reading Contents\nfor line in io.lines(w2v_txtfilename) do\n local parts = split(line, ' ')\n local w = parts[1]\n local w_id = get_id_from_word(w)\n if w_id ~= unk_w_id then\n for i=2, #parts do\n M[w_id][i-1] = tonumber(parts[i])\n end\n end\nend\n\nreturn M\n\n\n" }, { "path": "deep-ed/deep-ed-master/words/w2v/w2v.lua", "content": "-- Loads pre-trained word embeddings from either Word2Vec or Glove\n\nassert(get_id_from_word)\nassert(common_w2v_freq_words)\nassert(total_num_words)\n\nword_vecs_size = 300\n\n-- Loads pre-trained glove or word2vec embeddings: \nif opt.word_vecs == 'glove' then\n -- Glove downloaded from: http://nlp.stanford.edu/projects/glove/\n w2v_txtfilename = default_path .. 'Glove/glove.840B.300d.txt'\n w2v_t7filename = opt.root_data_dir .. 'generated/glove.840B.300d.t7'\n w2v_reader = 'words/w2v/glove_reader.lua'\nelseif opt.word_vecs == 'w2v' then\n -- Word2Vec downloaded from: https://code.google.com/archive/p/word2vec/\n w2v_binfilename = default_path .. 'Word2Vec/GoogleNews-vectors-negative300.bin'\n w2v_t7filename = opt.root_data_dir .. 'generated/GoogleNews-vectors-negative300.t7'\n w2v_reader = 'words/w2v/word2vec_reader.lua'\nend\n\n---------------------- Code: -----------------------\nw2vutils = {}\n\nprint('==> Loading ' .. opt.word_vecs .. ' vectors')\nif not paths.filep(w2v_t7filename) then\n print(' ---> t7 file NOT found. Loading w2v from the bin/txt file instead (slower).')\n w2vutils.M = require(w2v_reader)\n print('Writing t7 File for future usage. Next time Word2Vec loading will be faster!')\n torch.save(w2v_t7filename, w2vutils.M)\nelse\n print(' ---> from t7 file.')\n w2vutils.M = torch.load(w2v_t7filename)\nend\n\n-- Move the word embedding matrix on the GPU if we do some training. \n-- In this way we can perform word embedding lookup much faster.\nif opt and string.find(opt.type, 'cuda') then\n w2vutils.M = w2vutils.M:cuda()\nend\n\n---------- Define additional functions -----------------\n-- word -> vec\nw2vutils.get_w_vec = function (self,word)\n local w_id = get_id_from_word(word)\n return w2vutils.M[w_id]:clone()\nend\n\n-- word_id -> vec\nw2vutils.get_w_vec_from_id = function (self,w_id)\n return w2vutils.M[w_id]:clone()\nend\n\nw2vutils.lookup_w_vecs = function (self,word_id_tensor)\n assert(word_id_tensor:dim() <= 2, 'Only word id tensors w/ 1 or 2 dimensions are supported.')\n local output = torch.FloatTensor()\n local word_ids = word_id_tensor:long()\n if opt and string.find(opt.type, 'cuda') then\n output = output:cuda()\n word_ids = word_ids:cuda()\n end\n \n if word_ids:dim() == 2 then\n output:index(w2vutils.M, 1, word_ids:view(-1))\n output = output:view(word_ids:size(1), word_ids:size(2), w2vutils.M:size(2))\n elseif word_ids:dim() == 1 then\n output:index(w2vutils.M, 1, word_ids)\n output = output:view(word_ids:size(1), w2vutils.M:size(2))\n end\n\n return output\nend\n\n-- Normalize word vectors to have norm 1 .\nw2vutils.renormalize = function (self)\n w2vutils.M[unk_w_id]:mul(0)\n w2vutils.M[unk_w_id]:add(1)\n w2vutils.M:cdiv(w2vutils.M:norm(2,2):expand(w2vutils.M:size()))\n local x = w2vutils.M:norm(2,2):view(-1) - 1\n assert(x:norm() < 0.1, x:norm())\n assert(w2vutils.M[100]:norm() < 1.001 and w2vutils.M[100]:norm() > 0.99)\n w2vutils.M[unk_w_id]:mul(0)\nend\n\nw2vutils:renormalize()\n\nprint(' Done reading w2v data. Word vocab size = ' .. w2vutils.M:size(1))\n\n-- Phrase embedding using average of vectors of words in the phrase\nw2vutils.phrase_avg_vec = function(self, phrase)\n local words = split_in_words(phrase)\n local num_words = table_len(words)\n local num_existent_words = 0\n local vec = torch.zeros(word_vecs_size)\n for i = 1,num_words do\n local w = words[i]\n local w_id = get_id_from_word(w)\n if w_id ~= unk_w_id then\n vec:add(w2vutils:get_w_vec_from_id(w_id))\n num_existent_words = num_existent_words + 1\n end\n end\n if (num_existent_words > 0) then\n vec:div(num_existent_words)\n end\n return vec\nend\n\nw2vutils.top_k_closest_words = function (self,vec, k, mat)\n local k = k or 1\n vec = vec:float()\n local distances = torch.mv(mat, vec)\n local best_scores, best_word_ids = topk(distances, k)\n local returnwords = {}\n local returndistances = {}\n for i = 1,k do\n local w = get_word_from_id(best_word_ids[i])\n if is_stop_word_or_number(w) then\n table.insert(returnwords, red(w))\n else\n table.insert(returnwords, w)\n end\n assert(best_scores[i] == distances[best_word_ids[i]], best_scores[i] .. ' ' .. distances[best_word_ids[i]])\n table.insert(returndistances, distances[best_word_ids[i]])\n end\n return returnwords, returndistances\nend\n\nw2vutils.most_similar2word = function(self, word, k)\n local k = k or 1\n local v = w2vutils:get_w_vec(word)\n neighbors, scores = w2vutils:top_k_closest_words(v, k, w2vutils.M)\n print('To word ' .. skyblue(word) .. ' : ' .. list_with_scores_to_str(neighbors, scores))\nend\n\nw2vutils.most_similar2vec = function(self, vec, k)\n local k = k or 1\n neighbors, scores = w2vutils:top_k_closest_words(vec, k, w2vutils.M)\n print(list_with_scores_to_str(neighbors, scores))\nend\n\n\n--------------------- Unit tests ----------------------------------------\nlocal unit_tests = opt.unit_tests or false\nif (unit_tests) then\n print('\\nWord to word similarity test:')\n w2vutils:most_similar2word('nice', 5)\n w2vutils:most_similar2word('france', 5)\n w2vutils:most_similar2word('hello', 5)\nend\n\n-- Computes for each word w : \\sum_v exp() and \\sum_v \nw2vutils.total_word_correlation = function(self, k, j)\n local exp_Z = torch.zeros(w2vutils.M:narrow(1, 1, j):size(1))\n \n local sum_t = w2vutils.M:narrow(1, 1, j):sum(1) -- 1 x d\n local sum_Z = (w2vutils.M:narrow(1, 1, j) * sum_t:t()):view(-1) -- num_w\n\n print(red('Top words by sum_Z:'))\n best_sum_Z, best_word_ids = topk(sum_Z, k)\n for i = 1,k do\n local w = get_word_from_id(best_word_ids[i])\n assert(best_sum_Z[i] == sum_Z[best_word_ids[i]])\n print(w .. ' [' .. best_sum_Z[i] .. ']; ')\n end\n\n print('\\n' .. red('Bottom words by sum_Z:'))\n best_sum_Z, best_word_ids = topk(- sum_Z, k)\n for i = 1,k do\n local w = get_word_from_id(best_word_ids[i])\n assert(best_sum_Z[i] == - sum_Z[best_word_ids[i]])\n print(w .. ' [' .. sum_Z[best_word_ids[i]] .. ']; ')\n end\nend\n\n\n-- Plot with gnuplot:\n-- set palette model RGB defined ( 0 'white', 1 'pink', 2 'green' , 3 'blue', 4 'red' )\n-- plot 'tsne-w2v-vecs.txt_1000' using 1:2:3 with labels offset 0,1, '' using 1:2:4 w points pt 7 ps 2 palette\nw2vutils.tsne = function(self, num_rand_words)\n local topic1 = {'japan', 'china', 'france', 'switzerland', 'romania', 'india', 'australia', 'country', 'city', 'tokyo', 'nation', 'capital', 'continent', 'europe', 'asia', 'earth', 'america'}\n local topic2 = {'football', 'striker', 'goalkeeper', 'basketball', 'coach', 'championship', 'cup',\n 'soccer', 'player', 'captain', 'qualifier', 'goal', 'under-21', 'halftime', 'standings', 'basketball', \n 'games', 'league', 'rugby', 'hockey', 'fifa', 'fans', 'maradona', 'mutu', 'hagi', 'beckham', 'injury', 'game', \n 'kick', 'penalty'}\n local topic_avg = {'japan national football team', 'germany national football team', \n 'china national football team', 'brazil soccer', 'japan soccer', 'germany soccer', 'china soccer', \n 'fc barcelona', 'real madrid'} \n \n local stop_words_array = {}\n for w,_ in pairs(stop_words) do\n table.insert(stop_words_array, w)\n end\n \n local topic1_len = table_len(topic1)\n local topic2_len = table_len(topic2)\n local topic_avg_len = table_len(topic_avg)\n local stop_words_len = table_len(stop_words_array)\n \n torch.setdefaulttensortype('torch.DoubleTensor')\n w2vutils.M = w2vutils.M:double()\n \n local tensor = torch.zeros(num_rand_words + stop_words_len + topic1_len + topic2_len + topic_avg_len, word_vecs_size)\n local tensor_w_ids = torch.zeros(num_rand_words)\n local tensor_colors = torch.zeros(tensor:size(1))\n \n for i = 1,num_rand_words do\n tensor_w_ids[i] = math.random(1,25000)\n tensor_colors[i] = 0\n tensor[i]:copy(w2vutils.M[tensor_w_ids[i]])\n end\n \n for i = 1, stop_words_len do\n tensor_colors[num_rand_words + i] = 1\n tensor[num_rand_words + i]:copy(w2vutils:phrase_avg_vec(stop_words_array[i])) \n end\n \n for i = 1, topic1_len do\n tensor_colors[num_rand_words + stop_words_len + i] = 2\n tensor[num_rand_words + stop_words_len + i]:copy(w2vutils:phrase_avg_vec(topic1[i])) \n end\n \n for i = 1, topic2_len do\n tensor_colors[num_rand_words + stop_words_len + topic1_len + i] = 3\n tensor[num_rand_words + stop_words_len + topic1_len + i]:copy(w2vutils:phrase_avg_vec(topic2[i])) \n end\n\n for i = 1, topic_avg_len do\n tensor_colors[num_rand_words + stop_words_len + topic1_len + topic2_len + i] = 4\n tensor[num_rand_words + stop_words_len + topic1_len + topic2_len + i]:copy(w2vutils:phrase_avg_vec(topic_avg[i])) \n end\n \n local manifold = require 'manifold'\n opts = {ndims = 2, perplexity = 30, pca = 50, use_bh = false}\n mapped_x1 = manifold.embedding.tsne(tensor, opts)\n assert(mapped_x1:size(1) == tensor:size(1) and mapped_x1:size(2) == 2)\n ouf_vecs = assert(io.open('tsne-w2v-vecs.txt_' .. num_rand_words, \"w\"))\n for i = 1,mapped_x1:size(1) do\n local w = nil\n if tensor_colors[i] == 0 then \n w = get_word_from_id(tensor_w_ids[i])\n elseif tensor_colors[i] == 1 then \n w = stop_words_array[i - num_rand_words]:gsub(' ', '-')\n elseif tensor_colors[i] == 2 then \n w = topic1[i - num_rand_words - stop_words_len]:gsub(' ', '-')\n elseif tensor_colors[i] == 3 then \n w = topic2[i - num_rand_words - stop_words_len - topic1_len]:gsub(' ', '-')\n elseif tensor_colors[i] == 4 then \n w = topic_avg[i - num_rand_words - stop_words_len - topic1_len - topic2_len]:gsub(' ', '-') \n end\n assert(w)\n \n local v = mapped_x1[i]\n for j = 1,2 do\n ouf_vecs:write(v[j] .. ' ')\n end\n ouf_vecs:write(w .. ' ' .. tensor_colors[i] .. '\\n')\n end\n io.close(ouf_vecs)\n print(' DONE')\nend\n" }, { "path": "deep-ed/deep-ed-master/words/w2v/word2vec_reader.lua", "content": "-- Adapted from https://github.com/rotmanmi/word2vec.torch\nfunction read_string_w2v(file) \n local str = {}\n while true do\n local char = file:readChar()\n if char == 32 or char == 10 or char == 0 then\n break\n else\n str[#str+1] = char\n end\n end\n str = torch.CharStorage(str)\n return str:string()\nend\n\n\nfile = torch.DiskFile(w2v_binfilename,'r')\n\n--Reading Header\nfile:ascii()\nlocal vocab_size = file:readInt()\nlocal size = file:readInt()\nassert(size == word_vecs_size, 'Wrong size : ' .. size .. ' vs ' .. word_vecs_size)\n\nlocal M = torch.zeros(total_num_words(), word_vecs_size):float()\n\n--Reading Contents\nfile:binary()\nlocal num_phrases = 0\nfor i = 1,vocab_size do\n local w = read_string_w2v(file)\n local v = torch.FloatTensor(file:readFloat(word_vecs_size))\n local w_id = get_id_from_word(w)\n if w_id ~= unk_w_id then\n M[w_id]:copy(v)\n end\nend\n\nprint('Num words = ' .. total_num_words() .. '. Num phrases = ' .. num_phrases)\n\nreturn M\n" }, { "path": "deep-ed/deep-ed-master/words/w_freq/w_freq_gen.lua", "content": "-- Computes an unigram frequency of each word in the Wikipedia corpus\n\nif not opt then\n cmd = torch.CmdLine()\n cmd:option('-root_data_dir', '', 'Root path of the data, $DATA_PATH.')\n cmd:text()\n opt = cmd:parse(arg or {})\n assert(opt.root_data_dir ~= '', 'Specify a valid root_data_dir path argument.') \nend\n\n\nrequire 'torch'\ndofile 'utils/utils.lua'\n\ntds = tds or require 'tds' \n\nword_freqs = tds.Hash()\n\nlocal num_lines = 0\nit, _ = io.open(opt.root_data_dir .. 'generated/wiki_canonical_words.txt')\nline = it:read()\n\nwhile (line) do\n num_lines = num_lines + 1\n if num_lines % 100000 == 0 then\n print('Processed ' .. num_lines .. ' lines. ')\n end\n \n local parts = split(line , '\\t')\n local words = split(parts[3], ' ')\n for _,w in pairs(words) do\n if (not word_freqs[w]) then\n word_freqs[w] = 0\n end\n word_freqs[w] = word_freqs[w] + 1\n end\n line = it:read()\nend\n\n\n-- Writing word frequencies\nprint('Sorting and writing')\nsorted_word_freq = {}\nfor w,freq in pairs(word_freqs) do\n if freq >= 10 then\n table.insert(sorted_word_freq, {w = w, freq = freq})\n end\nend\n\ntable.sort(sorted_word_freq, function(a,b) return a.freq > b.freq end)\n\nout_file = opt.root_data_dir .. 'generated/word_wiki_freq.txt'\nouf = assert(io.open(out_file, \"w\"))\ntotal_freq = 0\nfor _,x in pairs(sorted_word_freq) do\n ouf:write(x.w .. '\\t' .. x.freq .. '\\n')\n total_freq = total_freq + x.freq\nend\nouf:flush()\nio.close(ouf)\n\nprint('Total freq = ' .. total_freq .. '\\n')\n" }, { "path": "deep-ed/deep-ed-master/words/w_freq/w_freq_index.lua", "content": "-- Loads all words and their frequencies and IDs from a dictionary.\nassert(common_w2v_freq_words)\nif not opt.unig_power then\n opt.unig_power = 0.6\nend\n \nprint('==> Loading word freq map with unig power ' .. red(opt.unig_power))\nlocal w_freq_file = opt.root_data_dir .. 'generated/word_wiki_freq.txt'\n\nlocal w_freq = {}\nw_freq.id2word = tds.Hash()\nw_freq.word2id = tds.Hash()\n\nw_freq.w_f_start = tds.Hash()\nw_freq.w_f_end = tds.Hash()\nw_freq.total_freq = 0.0\n\nw_freq.w_f_at_unig_power_start = tds.Hash()\nw_freq.w_f_at_unig_power_end = tds.Hash()\nw_freq.total_freq_at_unig_power = 0.0\n\n-- UNK word id\nunk_w_id = 1\nw_freq.word2id['UNK_W'] = unk_w_id \nw_freq.id2word[unk_w_id] = 'UNK_W'\n\nlocal tmp_wid = 1\nfor line in io.lines(w_freq_file) do \n local parts = split(line, '\\t')\n local w = parts[1]\n if common_w2v_freq_words[w] then\n tmp_wid = tmp_wid + 1\n local w_id = tmp_wid\n w_freq.id2word[w_id] = w\n w_freq.word2id[w] = w_id\n\n local w_f = tonumber(parts[2])\n assert(w_f)\n if w_f < 100 then\n w_f = 100\n end\n w_freq.w_f_start[w_id] = w_freq.total_freq\n w_freq.total_freq = w_freq.total_freq + w_f\n w_freq.w_f_end[w_id] = w_freq.total_freq\n\n w_freq.w_f_at_unig_power_start[w_id] = w_freq.total_freq_at_unig_power\n w_freq.total_freq_at_unig_power = w_freq.total_freq_at_unig_power + math.pow(w_f, opt.unig_power)\n w_freq.w_f_at_unig_power_end[w_id] = w_freq.total_freq_at_unig_power\n end\nend\n\nw_freq.total_num_words = tmp_wid\n\nprint(' Done loading word freq index. Num words = ' .. w_freq.total_num_words .. '; total freq = ' .. w_freq.total_freq)\n\n--------------------------------------------\ntotal_num_words = function()\n return w_freq.total_num_words\nend\n\ncontains_w_id = function(w_id)\n assert(w_id >= 1 and w_id <= total_num_words(), w_id)\n return (w_id ~= unk_w_id)\nend\n\n-- id -> word\nget_word_from_id = function(w_id)\n assert(w_id >= 1 and w_id <= total_num_words(), w_id)\n return w_freq.id2word[w_id]\nend\n\n-- word -> id\nget_id_from_word = function(w)\n local w_id = w_freq.word2id[w]\n if w_id == nil then\n return unk_w_id\n end\n return w_id\nend\n\ncontains_w = function(w)\n return contains_w_id(get_id_from_word(w))\nend\n\n-- word frequency:\nfunction get_w_id_freq(w_id)\n assert(contains_w_id(w_id), w_id)\n return w_freq.w_f_end[w_id] - w_freq.w_f_start[w_id] + 1\nend\n\n-- p(w) prior:\nfunction get_w_id_unigram(w_id)\n return get_w_id_freq(w_id) / w_freq.total_freq\nend\n\nfunction get_w_tensor_log_unigram(vec_w_ids)\n assert(vec_w_ids:dim() == 2)\n local v = torch.zeros(vec_w_ids:size(1), vec_w_ids:size(2))\n for i= 1,vec_w_ids:size(1) do\n for j = 1,vec_w_ids:size(2) do\n v[i][j] = math.log(get_w_id_unigram(vec_w_ids[i][j]))\n end\n end\n return v \nend\n\n\nif (opt.unit_tests) then\n print(get_w_id_unigram(get_id_from_word('the')))\n print(get_w_id_unigram(get_id_from_word('of')))\n print(get_w_id_unigram(get_id_from_word('and')))\n print(get_w_id_unigram(get_id_from_word('romania')))\nend\n\n-- Generates an random word sampled from the word unigram frequency.\nlocal function random_w_id(total_freq, w_f_start, w_f_end)\n local j = math.random() * total_freq\n local i_start = 2\n local i_end = total_num_words()\n\n while i_start <= i_end do\n local i_mid = math.floor((i_start + i_end) / 2)\n local w_id_mid = i_mid\n if w_f_start[w_id_mid] <= j and j <= w_f_end[w_id_mid] then\n return w_id_mid\n elseif (w_f_start[w_id_mid] > j) then\n i_end = i_mid - 1\n elseif (w_f_end[w_id_mid] < j) then\n i_start = i_mid + 1\n end\n end\n print(red('Binary search error !!'))\nend\n\n-- Frequent word subsampling procedure from the Word2Vec paper.\nfunction random_unigram_at_unig_power_w_id()\n return random_w_id(w_freq.total_freq_at_unig_power, w_freq.w_f_at_unig_power_start, w_freq.w_f_at_unig_power_end)\nend\n\n\nfunction get_w_unnorm_unigram_at_power(w_id)\n return math.pow(get_w_id_unigram(w_id), opt.unig_power)\nend\n\n\nfunction unit_test_random_unigram_at_unig_power_w_id(k_samples)\n local empirical_dist = {}\n for i=1,k_samples do\n local w_id = random_unigram_at_unig_power_w_id()\n assert(w_id ~= unk_w_id)\n if not empirical_dist[w_id] then\n empirical_dist[w_id] = 0\n end\n empirical_dist[w_id] = empirical_dist[w_id] + 1\n end\n print('Now sorting ..')\n local sorted_empirical_dist = {}\n for k,v in pairs(empirical_dist) do\n table.insert(sorted_empirical_dist, {w_id = k, f = v})\n end\n table.sort(sorted_empirical_dist, function(a,b) return a.f > b.f end)\n\n local str = ''\n for i = 1,math.min(100, table_len(sorted_empirical_dist)) do\n str = str .. get_word_from_id(sorted_empirical_dist[i].w_id) .. '{' .. sorted_empirical_dist[i].f .. '}; '\n end\n print('Unit test random sampling: ' .. str)\nend\n" }, { "path": "gerbil-SpotWrapNifWS4Test/.gitignore", "content": "*.class\n*.cache\n.classpath\n.project\ntarget\n.settings\ngerbil_keys.properties\n" }, { "path": "gerbil-SpotWrapNifWS4Test/Dockerfile", "content": "FROM tomcat:8.0.36-jre8\n\n# remove the default tomcat application\nRUN rm -rf /usr/local/tomcat/webapps/ROOT /usr/local/tomcat/webapps/ROOT.war\n\nADD target/gerbil-spotWrapNifWS4Test-0.0.1-SNAPSHOT.war /usr/local/tomcat/webapps/ROOT.war\n\n" }, { "path": "gerbil-SpotWrapNifWS4Test/LICENSE", "content": " GNU AFFERO GENERAL PUBLIC LICENSE\n Version 3, 19 November 2007\n\n Copyright (C) 2007 Free Software Foundation, Inc. \n Everyone is permitted to copy and distribute verbatim copies\n of this license document, but changing it is not allowed.\n\n Preamble\n\n The GNU Affero General Public License is a free, copyleft license for\nsoftware and other kinds of works, specifically designed to ensure\ncooperation with the community in the case of network server software.\n\n The licenses for most software and other practical works are designed\nto take away your freedom to share and change the works. By contrast,\nour General Public Licenses are intended to guarantee your freedom to\nshare and change all versions of a program--to make sure it remains free\nsoftware for all its users.\n\n When we speak of free software, we are referring to freedom, not\nprice. Our General Public Licenses are designed to make sure that you\nhave the freedom to distribute copies of free software (and charge for\nthem if you wish), that you receive source code or can get it if you\nwant it, that you can change the software or use pieces of it in new\nfree programs, and that you know you can do these things.\n\n Developers that use our General Public Licenses protect your rights\nwith two steps: (1) assert copyright on the software, and (2) offer\nyou this License which gives you legal permission to copy, distribute\nand/or modify the software.\n\n A secondary benefit of defending all users' freedom is that\nimprovements made in alternate versions of the program, if they\nreceive widespread use, become available for other developers to\nincorporate. Many developers of free software are heartened and\nencouraged by the resulting cooperation. However, in the case of\nsoftware used on network servers, this result may fail to come about.\nThe GNU General Public License permits making a modified version and\nletting the public access it on a server without ever releasing its\nsource code to the public.\n\n The GNU Affero General Public License is designed specifically to\nensure that, in such cases, the modified source code becomes available\nto the community. It requires the operator of a network server to\nprovide the source code of the modified version running there to the\nusers of that server. Therefore, public use of a modified version, on\na publicly accessible server, gives the public access to the source\ncode of the modified version.\n\n An older license, called the Affero General Public License and\npublished by Affero, was designed to accomplish similar goals. This is\na different license, not a version of the Affero GPL, but Affero has\nreleased a new version of the Affero GPL which permits relicensing under\nthis license.\n\n The precise terms and conditions for copying, distribution and\nmodification follow.\n\n TERMS AND CONDITIONS\n\n 0. Definitions.\n\n \"This License\" refers to version 3 of the GNU Affero General Public License.\n\n \"Copyright\" also means copyright-like laws that apply to other kinds of\nworks, such as semiconductor masks.\n\n \"The Program\" refers to any copyrightable work licensed under this\nLicense. Each licensee is addressed as \"you\". \"Licensees\" and\n\"recipients\" may be individuals or organizations.\n\n To \"modify\" a work means to copy from or adapt all or part of the work\nin a fashion requiring copyright permission, other than the making of an\nexact copy. The resulting work is called a \"modified version\" of the\nearlier work or a work \"based on\" the earlier work.\n\n A \"covered work\" means either the unmodified Program or a work based\non the Program.\n\n To \"propagate\" a work means to do anything with it that, without\npermission, would make you directly or secondarily liable for\ninfringement under applicable copyright law, except executing it on a\ncomputer or modifying a private copy. Propagation includes copying,\ndistribution (with or without modification), making available to the\npublic, and in some countries other activities as well.\n\n To \"convey\" a work means any kind of propagation that enables other\nparties to make or receive copies. Mere interaction with a user through\na computer network, with no transfer of a copy, is not conveying.\n\n An interactive user interface displays \"Appropriate Legal Notices\"\nto the extent that it includes a convenient and prominently visible\nfeature that (1) displays an appropriate copyright notice, and (2)\ntells the user that there is no warranty for the work (except to the\nextent that warranties are provided), that licensees may convey the\nwork under this License, and how to view a copy of this License. If\nthe interface presents a list of user commands or options, such as a\nmenu, a prominent item in the list meets this criterion.\n\n 1. Source Code.\n\n The \"source code\" for a work means the preferred form of the work\nfor making modifications to it. \"Object code\" means any non-source\nform of a work.\n\n A \"Standard Interface\" means an interface that either is an official\nstandard defined by a recognized standards body, or, in the case of\ninterfaces specified for a particular programming language, one that\nis widely used among developers working in that language.\n\n The \"System Libraries\" of an executable work include anything, other\nthan the work as a whole, that (a) is included in the normal form of\npackaging a Major Component, but which is not part of that Major\nComponent, and (b) serves only to enable use of the work with that\nMajor Component, or to implement a Standard Interface for which an\nimplementation is available to the public in source code form. A\n\"Major Component\", in this context, means a major essential component\n(kernel, window system, and so on) of the specific operating system\n(if any) on which the executable work runs, or a compiler used to\nproduce the work, or an object code interpreter used to run it.\n\n The \"Corresponding Source\" for a work in object code form means all\nthe source code needed to generate, install, and (for an executable\nwork) run the object code and to modify the work, including scripts to\ncontrol those activities. However, it does not include the work's\nSystem Libraries, or general-purpose tools or generally available free\nprograms which are used unmodified in performing those activities but\nwhich are not part of the work. For example, Corresponding Source\nincludes interface definition files associated with source files for\nthe work, and the source code for shared libraries and dynamically\nlinked subprograms that the work is specifically designed to require,\nsuch as by intimate data communication or control flow between those\nsubprograms and other parts of the work.\n\n The Corresponding Source need not include anything that users\ncan regenerate automatically from other parts of the Corresponding\nSource.\n\n The Corresponding Source for a work in source code form is that\nsame work.\n\n 2. Basic Permissions.\n\n All rights granted under this License are granted for the term of\ncopyright on the Program, and are irrevocable provided the stated\nconditions are met. This License explicitly affirms your unlimited\npermission to run the unmodified Program. The output from running a\ncovered work is covered by this License only if the output, given its\ncontent, constitutes a covered work. This License acknowledges your\nrights of fair use or other equivalent, as provided by copyright law.\n\n You may make, run and propagate covered works that you do not\nconvey, without conditions so long as your license otherwise remains\nin force. You may convey covered works to others for the sole purpose\nof having them make modifications exclusively for you, or provide you\nwith facilities for running those works, provided that you comply with\nthe terms of this License in conveying all material for which you do\nnot control copyright. Those thus making or running the covered works\nfor you must do so exclusively on your behalf, under your direction\nand control, on terms that prohibit them from making any copies of\nyour copyrighted material outside their relationship with you.\n\n Conveying under any other circumstances is permitted solely under\nthe conditions stated below. Sublicensing is not allowed; section 10\nmakes it unnecessary.\n\n 3. Protecting Users' Legal Rights From Anti-Circumvention Law.\n\n No covered work shall be deemed part of an effective technological\nmeasure under any applicable law fulfilling obligations under article\n11 of the WIPO copyright treaty adopted on 20 December 1996, or\nsimilar laws prohibiting or restricting circumvention of such\nmeasures.\n\n When you convey a covered work, you waive any legal power to forbid\ncircumvention of technological measures to the extent such circumvention\nis effected by exercising rights under this License with respect to\nthe covered work, and you disclaim any intention to limit operation or\nmodification of the work as a means of enforcing, against the work's\nusers, your or third parties' legal rights to forbid circumvention of\ntechnological measures.\n\n 4. Conveying Verbatim Copies.\n\n You may convey verbatim copies of the Program's source code as you\nreceive it, in any medium, provided that you conspicuously and\nappropriately publish on each copy an appropriate copyright notice;\nkeep intact all notices stating that this License and any\nnon-permissive terms added in accord with section 7 apply to the code;\nkeep intact all notices of the absence of any warranty; and give all\nrecipients a copy of this License along with the Program.\n\n You may charge any price or no price for each copy that you convey,\nand you may offer support or warranty protection for a fee.\n\n 5. Conveying Modified Source Versions.\n\n You may convey a work based on the Program, or the modifications to\nproduce it from the Program, in the form of source code under the\nterms of section 4, provided that you also meet all of these conditions:\n\n a) The work must carry prominent notices stating that you modified\n it, and giving a relevant date.\n\n b) The work must carry prominent notices stating that it is\n released under this License and any conditions added under section\n 7. This requirement modifies the requirement in section 4 to\n \"keep intact all notices\".\n\n c) You must license the entire work, as a whole, under this\n License to anyone who comes into possession of a copy. This\n License will therefore apply, along with any applicable section 7\n additional terms, to the whole of the work, and all its parts,\n regardless of how they are packaged. This License gives no\n permission to license the work in any other way, but it does not\n invalidate such permission if you have separately received it.\n\n d) If the work has interactive user interfaces, each must display\n Appropriate Legal Notices; however, if the Program has interactive\n interfaces that do not display Appropriate Legal Notices, your\n work need not make them do so.\n\n A compilation of a covered work with other separate and independent\nworks, which are not by their nature extensions of the covered work,\nand which are not combined with it such as to form a larger program,\nin or on a volume of a storage or distribution medium, is called an\n\"aggregate\" if the compilation and its resulting copyright are not\nused to limit the access or legal rights of the compilation's users\nbeyond what the individual works permit. Inclusion of a covered work\nin an aggregate does not cause this License to apply to the other\nparts of the aggregate.\n\n 6. Conveying Non-Source Forms.\n\n You may convey a covered work in object code form under the terms\nof sections 4 and 5, provided that you also convey the\nmachine-readable Corresponding Source under the terms of this License,\nin one of these ways:\n\n a) Convey the object code in, or embodied in, a physical product\n (including a physical distribution medium), accompanied by the\n Corresponding Source fixed on a durable physical medium\n customarily used for software interchange.\n\n b) Convey the object code in, or embodied in, a physical product\n (including a physical distribution medium), accompanied by a\n written offer, valid for at least three years and valid for as\n long as you offer spare parts or customer support for that product\n model, to give anyone who possesses the object code either (1) a\n copy of the Corresponding Source for all the software in the\n product that is covered by this License, on a durable physical\n medium customarily used for software interchange, for a price no\n more than your reasonable cost of physically performing this\n conveying of source, or (2) access to copy the\n Corresponding Source from a network server at no charge.\n\n c) Convey individual copies of the object code with a copy of the\n written offer to provide the Corresponding Source. This\n alternative is allowed only occasionally and noncommercially, and\n only if you received the object code with such an offer, in accord\n with subsection 6b.\n\n d) Convey the object code by offering access from a designated\n place (gratis or for a charge), and offer equivalent access to the\n Corresponding Source in the same way through the same place at no\n further charge. You need not require recipients to copy the\n Corresponding Source along with the object code. If the place to\n copy the object code is a network server, the Corresponding Source\n may be on a different server (operated by you or a third party)\n that supports equivalent copying facilities, provided you maintain\n clear directions next to the object code saying where to find the\n Corresponding Source. Regardless of what server hosts the\n Corresponding Source, you remain obligated to ensure that it is\n available for as long as needed to satisfy these requirements.\n\n e) Convey the object code using peer-to-peer transmission, provided\n you inform other peers where the object code and Corresponding\n Source of the work are being offered to the general public at no\n charge under subsection 6d.\n\n A separable portion of the object code, whose source code is excluded\nfrom the Corresponding Source as a System Library, need not be\nincluded in conveying the object code work.\n\n A \"User Product\" is either (1) a \"consumer product\", which means any\ntangible personal property which is normally used for personal, family,\nor household purposes, or (2) anything designed or sold for incorporation\ninto a dwelling. In determining whether a product is a consumer product,\ndoubtful cases shall be resolved in favor of coverage. For a particular\nproduct received by a particular user, \"normally used\" refers to a\ntypical or common use of that class of product, regardless of the status\nof the particular user or of the way in which the particular user\nactually uses, or expects or is expected to use, the product. A product\nis a consumer product regardless of whether the product has substantial\ncommercial, industrial or non-consumer uses, unless such uses represent\nthe only significant mode of use of the product.\n\n \"Installation Information\" for a User Product means any methods,\nprocedures, authorization keys, or other information required to install\nand execute modified versions of a covered work in that User Product from\na modified version of its Corresponding Source. The information must\nsuffice to ensure that the continued functioning of the modified object\ncode is in no case prevented or interfered with solely because\nmodification has been made.\n\n If you convey an object code work under this section in, or with, or\nspecifically for use in, a User Product, and the conveying occurs as\npart of a transaction in which the right of possession and use of the\nUser Product is transferred to the recipient in perpetuity or for a\nfixed term (regardless of how the transaction is characterized), the\nCorresponding Source conveyed under this section must be accompanied\nby the Installation Information. But this requirement does not apply\nif neither you nor any third party retains the ability to install\nmodified object code on the User Product (for example, the work has\nbeen installed in ROM).\n\n The requirement to provide Installation Information does not include a\nrequirement to continue to provide support service, warranty, or updates\nfor a work that has been modified or installed by the recipient, or for\nthe User Product in which it has been modified or installed. Access to a\nnetwork may be denied when the modification itself materially and\nadversely affects the operation of the network or violates the rules and\nprotocols for communication across the network.\n\n Corresponding Source conveyed, and Installation Information provided,\nin accord with this section must be in a format that is publicly\ndocumented (and with an implementation available to the public in\nsource code form), and must require no special password or key for\nunpacking, reading or copying.\n\n 7. Additional Terms.\n\n \"Additional permissions\" are terms that supplement the terms of this\nLicense by making exceptions from one or more of its conditions.\nAdditional permissions that are applicable to the entire Program shall\nbe treated as though they were included in this License, to the extent\nthat they are valid under applicable law. If additional permissions\napply only to part of the Program, that part may be used separately\nunder those permissions, but the entire Program remains governed by\nthis License without regard to the additional permissions.\n\n When you convey a copy of a covered work, you may at your option\nremove any additional permissions from that copy, or from any part of\nit. (Additional permissions may be written to require their own\nremoval in certain cases when you modify the work.) You may place\nadditional permissions on material, added by you to a covered work,\nfor which you have or can give appropriate copyright permission.\n\n Notwithstanding any other provision of this License, for material you\nadd to a covered work, you may (if authorized by the copyright holders of\nthat material) supplement the terms of this License with terms:\n\n a) Disclaiming warranty or limiting liability differently from the\n terms of sections 15 and 16 of this License; or\n\n b) Requiring preservation of specified reasonable legal notices or\n author attributions in that material or in the Appropriate Legal\n Notices displayed by works containing it; or\n\n c) Prohibiting misrepresentation of the origin of that material, or\n requiring that modified versions of such material be marked in\n reasonable ways as different from the original version; or\n\n d) Limiting the use for publicity purposes of names of licensors or\n authors of the material; or\n\n e) Declining to grant rights under trademark law for use of some\n trade names, trademarks, or service marks; or\n\n f) Requiring indemnification of licensors and authors of that\n material by anyone who conveys the material (or modified versions of\n it) with contractual assumptions of liability to the recipient, for\n any liability that these contractual assumptions directly impose on\n those licensors and authors.\n\n All other non-permissive additional terms are considered \"further\nrestrictions\" within the meaning of section 10. If the Program as you\nreceived it, or any part of it, contains a notice stating that it is\ngoverned by this License along with a term that is a further\nrestriction, you may remove that term. If a license document contains\na further restriction but permits relicensing or conveying under this\nLicense, you may add to a covered work material governed by the terms\nof that license document, provided that the further restriction does\nnot survive such relicensing or conveying.\n\n If you add terms to a covered work in accord with this section, you\nmust place, in the relevant source files, a statement of the\nadditional terms that apply to those files, or a notice indicating\nwhere to find the applicable terms.\n\n Additional terms, permissive or non-permissive, may be stated in the\nform of a separately written license, or stated as exceptions;\nthe above requirements apply either way.\n\n 8. Termination.\n\n You may not propagate or modify a covered work except as expressly\nprovided under this License. Any attempt otherwise to propagate or\nmodify it is void, and will automatically terminate your rights under\nthis License (including any patent licenses granted under the third\nparagraph of section 11).\n\n However, if you cease all violation of this License, then your\nlicense from a particular copyright holder is reinstated (a)\nprovisionally, unless and until the copyright holder explicitly and\nfinally terminates your license, and (b) permanently, if the copyright\nholder fails to notify you of the violation by some reasonable means\nprior to 60 days after the cessation.\n\n Moreover, your license from a particular copyright holder is\nreinstated permanently if the copyright holder notifies you of the\nviolation by some reasonable means, this is the first time you have\nreceived notice of violation of this License (for any work) from that\ncopyright holder, and you cure the violation prior to 30 days after\nyour receipt of the notice.\n\n Termination of your rights under this section does not terminate the\nlicenses of parties who have received copies or rights from you under\nthis License. If your rights have been terminated and not permanently\nreinstated, you do not qualify to receive new licenses for the same\nmaterial under section 10.\n\n 9. Acceptance Not Required for Having Copies.\n\n You are not required to accept this License in order to receive or\nrun a copy of the Program. Ancillary propagation of a covered work\noccurring solely as a consequence of using peer-to-peer transmission\nto receive a copy likewise does not require acceptance. However,\nnothing other than this License grants you permission to propagate or\nmodify any covered work. These actions infringe copyright if you do\nnot accept this License. Therefore, by modifying or propagating a\ncovered work, you indicate your acceptance of this License to do so.\n\n 10. Automatic Licensing of Downstream Recipients.\n\n Each time you convey a covered work, the recipient automatically\nreceives a license from the original licensors, to run, modify and\npropagate that work, subject to this License. You are not responsible\nfor enforcing compliance by third parties with this License.\n\n An \"entity transaction\" is a transaction transferring control of an\norganization, or substantially all assets of one, or subdividing an\norganization, or merging organizations. If propagation of a covered\nwork results from an entity transaction, each party to that\ntransaction who receives a copy of the work also receives whatever\nlicenses to the work the party's predecessor in interest had or could\ngive under the previous paragraph, plus a right to possession of the\nCorresponding Source of the work from the predecessor in interest, if\nthe predecessor has it or can get it with reasonable efforts.\n\n You may not impose any further restrictions on the exercise of the\nrights granted or affirmed under this License. For example, you may\nnot impose a license fee, royalty, or other charge for exercise of\nrights granted under this License, and you may not initiate litigation\n(including a cross-claim or counterclaim in a lawsuit) alleging that\nany patent claim is infringed by making, using, selling, offering for\nsale, or importing the Program or any portion of it.\n\n 11. Patents.\n\n A \"contributor\" is a copyright holder who authorizes use under this\nLicense of the Program or a work on which the Program is based. The\nwork thus licensed is called the contributor's \"contributor version\".\n\n A contributor's \"essential patent claims\" are all patent claims\nowned or controlled by the contributor, whether already acquired or\nhereafter acquired, that would be infringed by some manner, permitted\nby this License, of making, using, or selling its contributor version,\nbut do not include claims that would be infringed only as a\nconsequence of further modification of the contributor version. For\npurposes of this definition, \"control\" includes the right to grant\npatent sublicenses in a manner consistent with the requirements of\nthis License.\n\n Each contributor grants you a non-exclusive, worldwide, royalty-free\npatent license under the contributor's essential patent claims, to\nmake, use, sell, offer for sale, import and otherwise run, modify and\npropagate the contents of its contributor version.\n\n In the following three paragraphs, a \"patent license\" is any express\nagreement or commitment, however denominated, not to enforce a patent\n(such as an express permission to practice a patent or covenant not to\nsue for patent infringement). To \"grant\" such a patent license to a\nparty means to make such an agreement or commitment not to enforce a\npatent against the party.\n\n If you convey a covered work, knowingly relying on a patent license,\nand the Corresponding Source of the work is not available for anyone\nto copy, free of charge and under the terms of this License, through a\npublicly available network server or other readily accessible means,\nthen you must either (1) cause the Corresponding Source to be so\navailable, or (2) arrange to deprive yourself of the benefit of the\npatent license for this particular work, or (3) arrange, in a manner\nconsistent with the requirements of this License, to extend the patent\nlicense to downstream recipients. \"Knowingly relying\" means you have\nactual knowledge that, but for the patent license, your conveying the\ncovered work in a country, or your recipient's use of the covered work\nin a country, would infringe one or more identifiable patents in that\ncountry that you have reason to believe are valid.\n\n If, pursuant to or in connection with a single transaction or\narrangement, you convey, or propagate by procuring conveyance of, a\ncovered work, and grant a patent license to some of the parties\nreceiving the covered work authorizing them to use, propagate, modify\nor convey a specific copy of the covered work, then the patent license\nyou grant is automatically extended to all recipients of the covered\nwork and works based on it.\n\n A patent license is \"discriminatory\" if it does not include within\nthe scope of its coverage, prohibits the exercise of, or is\nconditioned on the non-exercise of one or more of the rights that are\nspecifically granted under this License. You may not convey a covered\nwork if you are a party to an arrangement with a third party that is\nin the business of distributing software, under which you make payment\nto the third party based on the extent of your activity of conveying\nthe work, and under which the third party grants, to any of the\nparties who would receive the covered work from you, a discriminatory\npatent license (a) in connection with copies of the covered work\nconveyed by you (or copies made from those copies), or (b) primarily\nfor and in connection with specific products or compilations that\ncontain the covered work, unless you entered into that arrangement,\nor that patent license was granted, prior to 28 March 2007.\n\n Nothing in this License shall be construed as excluding or limiting\nany implied license or other defenses to infringement that may\notherwise be available to you under applicable patent law.\n\n 12. No Surrender of Others' Freedom.\n\n If conditions are imposed on you (whether by court order, agreement or\notherwise) that contradict the conditions of this License, they do not\nexcuse you from the conditions of this License. If you cannot convey a\ncovered work so as to satisfy simultaneously your obligations under this\nLicense and any other pertinent obligations, then as a consequence you may\nnot convey it at all. For example, if you agree to terms that obligate you\nto collect a royalty for further conveying from those to whom you convey\nthe Program, the only way you could satisfy both those terms and this\nLicense would be to refrain entirely from conveying the Program.\n\n 13. Remote Network Interaction; Use with the GNU General Public License.\n\n Notwithstanding any other provision of this License, if you modify the\nProgram, your modified version must prominently offer all users\ninteracting with it remotely through a computer network (if your version\nsupports such interaction) an opportunity to receive the Corresponding\nSource of your version by providing access to the Corresponding Source\nfrom a network server at no charge, through some standard or customary\nmeans of facilitating copying of software. This Corresponding Source\nshall include the Corresponding Source for any work covered by version 3\nof the GNU General Public License that is incorporated pursuant to the\nfollowing paragraph.\n\n Notwithstanding any other provision of this License, you have\npermission to link or combine any covered work with a work licensed\nunder version 3 of the GNU General Public License into a single\ncombined work, and to convey the resulting work. The terms of this\nLicense will continue to apply to the part which is the covered work,\nbut the work with which it is combined will remain governed by version\n3 of the GNU General Public License.\n\n 14. Revised Versions of this License.\n\n The Free Software Foundation may publish revised and/or new versions of\nthe GNU Affero General Public License from time to time. Such new versions\nwill be similar in spirit to the present version, but may differ in detail to\naddress new problems or concerns.\n\n Each version is given a distinguishing version number. If the\nProgram specifies that a certain numbered version of the GNU Affero General\nPublic License \"or any later version\" applies to it, you have the\noption of following the terms and conditions either of that numbered\nversion or of any later version published by the Free Software\nFoundation. If the Program does not specify a version number of the\nGNU Affero General Public License, you may choose any version ever published\nby the Free Software Foundation.\n\n If the Program specifies that a proxy can decide which future\nversions of the GNU Affero General Public License can be used, that proxy's\npublic statement of acceptance of a version permanently authorizes you\nto choose that version for the Program.\n\n Later license versions may give you additional or different\npermissions. However, no additional obligations are imposed on any\nauthor or copyright holder as a result of your choosing to follow a\nlater version.\n\n 15. Disclaimer of Warranty.\n\n THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY\nAPPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT\nHOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM \"AS IS\" WITHOUT WARRANTY\nOF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO,\nTHE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR\nPURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM\nIS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF\nALL NECESSARY SERVICING, REPAIR OR CORRECTION.\n\n 16. Limitation of Liability.\n\n IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING\nWILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS\nTHE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY\nGENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE\nUSE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF\nDATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD\nPARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS),\nEVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF\nSUCH DAMAGES.\n\n 17. Interpretation of Sections 15 and 16.\n\n If the disclaimer of warranty and limitation of liability provided\nabove cannot be given local legal effect according to their terms,\nreviewing courts shall apply local law that most closely approximates\nan absolute waiver of all civil liability in connection with the\nProgram, unless a warranty or assumption of liability accompanies a\ncopy of the Program in return for a fee.\n\n END OF TERMS AND CONDITIONS\n\n How to Apply These Terms to Your New Programs\n\n If you develop a new program, and you want it to be of the greatest\npossible use to the public, the best way to achieve this is to make it\nfree software which everyone can redistribute and change under these terms.\n\n To do so, attach the following notices to the program. It is safest\nto attach them to the start of each source file to most effectively\nstate the exclusion of warranty; and each file should have at least\nthe \"copyright\" line and a pointer to where the full notice is found.\n\n \n Copyright (C) \n\n This program is free software: you can redistribute it and/or modify\n it under the terms of the GNU Affero General Public License as published\n by the Free Software Foundation, either version 3 of the License, or\n (at your option) any later version.\n\n This program is distributed in the hope that it will be useful,\n but WITHOUT ANY WARRANTY; without even the implied warranty of\n MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the\n GNU Affero General Public License for more details.\n\n You should have received a copy of the GNU Affero General Public License\n along with this program. If not, see .\n\nAlso add information on how to contact you by electronic and paper mail.\n\n If your software can interact with users remotely through a computer\nnetwork, you should also make sure that it provides a way for users to\nget its source. For example, if your program is a web application, its\ninterface could display a \"Source\" link that leads users to an archive\nof the code. There are many ways you could offer source, and different\nsolutions will be better for different programs; see section 13 for the\nspecific requirements.\n\n You should also get your employer (if you work as a programmer) or school,\nif any, to sign a \"copyright disclaimer\" for the program, if necessary.\nFor more information on this, and how to apply and follow the GNU AGPL, see\n.\n" }, { "path": "gerbil-SpotWrapNifWS4Test/Makefile", "content": "default: build dockerize\n\nbuild:\n mvn clean package -U\n\ndockerize:\n docker build -t git.project-hobbit.eu:4567/gerbil/spotwrapnifws4test .\n\npush:\n docker push git.project-hobbit.eu:4567/gerbil/spotwrapnifws4test" }, { "path": "gerbil-SpotWrapNifWS4Test/README.md", "content": "Gerbil\n========\nGeneral Entity Annotator Benchmark\n\nThis branch is part of the Gerbil project. It contains a very simple Webservice that wraps the DBpedia Spotlight Webservice and implements the NIF communication. We use it to test the NIF based Webservice annotator adapter that is part of Gerbil.\n" }, { "path": "gerbil-SpotWrapNifWS4Test/curlExample.sh", "content": "curl -d \"@example.ttl\" -H \"Content-Type: application/x-turtle\" http://localhost:8080/gerbil-spotWrapNifWS4Test/spotlight \n" }, { "path": "gerbil-SpotWrapNifWS4Test/docker-compose.yml", "content": "version: '2'\nservices:\n # spotwrapnifws4test .\n nifws4test:\n build: .\n restart: always\n container_name: nifws4test\n ports:\n - \"8080:8080\"\n" }, { "path": "gerbil-SpotWrapNifWS4Test/example.ttl", "content": "@prefix rdf: .\n@prefix rdfs: .\n@prefix xsd: .\n@prefix nif: .\n@prefix itsrdf: .\n\n\n a nif:RFC5147String , nif:String , nif:Context ;\n nif:beginIndex \"0\"^^xsd:nonNegativeInteger ;\n nif:endIndex \"146\"^^xsd:nonNegativeInteger ;\n nif:isString \"Florence May Harding studied at a school in Sydney, and with Douglas Robert Dundas , but in effect had no formal training in either botany or art.\"@en .\n\n\n a nif:RFC5147String , nif:String ;\n nif:anchorOf \"Sydney\"@en ;\n nif:beginIndex \"44\"^^xsd:nonNegativeInteger ;\n nif:endIndex \"50\"^^xsd:nonNegativeInteger ;\n nif:referenceContext ;\n itsrdf:taIdentRef . \n" }, { "path": "gerbil-SpotWrapNifWS4Test/my_notes/messages_format", "content": "Request: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_117, text=NYMEX natgas ends sharply lower on weather outlook. NEW YORK 1996-12-06 NYMEX Henry Hub natgas futures settled significantly lower Friday, pressured early by profit taking and driven even lower late by the National Weather Service's bearish six to 10 day forecast, sources said. January ended 29.7 cents lower at $ 3.487 per million British thermal units after dipping to a low of $ 3.46. Feb settled down 22 cents at $ 3.186. Most others also were down. \"Weather forecasts have been sketchy. Now the National Weather Service is calling for above-normal temperatures in more than half of the U.S.,\" one futures trader said. In its forecast, the NWS said it expects above-normal temperatures \"over the lower 48 states\" from December 12 through December 16. With more room to the downside anticipated early next week, traders said support in January was at $ 3.47, then $ 3.35. The next backstops were seen at $ 3.11 and $ 3.04, the low set on November 21. Resistance was pegged at the new contract high of $ 3.80. In the cash market, Gulf Coast prices were around $ 3.60 shortly before nomination deadlines. Midcontinent prices were similarly lower in the $ 3.40s. New York city gate gas slipped into the $ 4.40s, down almost 15 cents. NYMEX said an estimated 35,662 Hub contracts traded, down from Thursday's revised tally of 43,955. NYMEX Alberta natgas remained untraded, with January settling at $ 1.65, off 10 cents from Thursday. Physical prices for the weekend at the AECO storage hub were also down about 10 cents in the C$ 1.92-1.97 per gigajoule, or $ 1.52-1.56 per mmBtu range, pressured by unseasonably mild weather in western Canada. NYMEX Permian natgas, also untraded, ended 10 cents lower at $ 2.90. In congruence with futures, Permian cash prices for the weekend fell more than 10 cents to the high - $ 3.40s. On the KCBT, January finished 26.5 cents lower at $ 3.35 after dipping to a low of $ 3.33 earlier in the session. February was down 22 cents at the close, while other deferreds were 4.5 to nine cents lower. The East / West spread narrowed by 3.2 cents to 13.7 cents (NYMEX premium). Physical prices at Waha for the weekend lost more than 15 cents to the low-to-mid $ 3.50s as milder weather moved into the Southwest.-- H McCulloch, New York Power Desk+212-859-1628, markings=[]]\n2018-02-09 00:50:43,779 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_117, text=NYMEX natgas ends sharply lower on weather outlook. NEW YORK 1996-12-06 NYMEX Henry Hub natgas futures settled significantly lower Friday, pressured early by profit taking and driven even lower late by the National Weather Service's bearish six to 10 day forecast, sources said. January ended 29.7 cents lower at $ 3.487 per million British thermal units after dipping to a low of $ 3.46. Feb settled down 22 cents at $ 3.186. Most others also were down. \"Weather forecasts have been sketchy. Now the National Weather Service is calling for above-normal temperatures in more than half of the U.S.,\" one futures trader said. In its forecast, the NWS said it expects above-normal temperatures \"over the lower 48 states\" from December 12 through December 16. With more room to the downside anticipated early next week, traders said support in January was at $ 3.47, then $ 3.35. The next backstops were seen at $ 3.11 and $ 3.04, the low set on November 21. Resistance was pegged at the new contract high of $ 3.80. In the cash market, Gulf Coast prices were around $ 3.60 shortly before nomination deadlines. Midcontinent prices were similarly lower in the $ 3.40s. New York city gate gas slipped into the $ 4.40s, down almost 15 cents. NYMEX said an estimated 35,662 Hub contracts traded, down from Thursday's revised tally of 43,955. NYMEX Alberta natgas remained untraded, with January settling at $ 1.65, off 10 cents from Thursday. Physical prices for the weekend at the AECO storage hub were also down about 10 cents in the C$ 1.92-1.97 per gigajoule, or $ 1.52-1.56 per mmBtu range, pressured by unseasonably mild weather in western Canada. NYMEX Permian natgas, also untraded, ended 10 cents lower at $ 2.90. In congruence with futures, Permian cash prices for the weekend fell more than 10 cents to the high - $ 3.40s. On the KCBT, January finished 26.5 cents lower at $ 3.35 after dipping to a low of $ 3.33 earlier in the session. February was down 22 cents at the close, while other deferreds were 4.5 to nine cents lower. The East / West spread narrowed by 3.2 cents to 13.7 cents (NYMEX premium). Physical prices at Waha for the weekend lost more than 15 cents to the low-to-mid $ 3.50s as milder weather moved into the Southwest.-- H McCulloch, New York Power Desk+212-859-1628, markings=[(0, 5, [http://dbpedia.org/resource/New_York_Mercantile_Exchange], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/Company]), (52, 8, [http://dbpedia.org/resource/New_York_City], a []), (72, 5, [http://dbpedia.org/resource/New_York_Mercantile_Exchange], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/Company]), (78, 9, [http://dbpedia.org/resource/Henry_Hub], a []), (206, 24, [http://dbpedia.org/resource/National_Weather_Service], a [http://dbpedia.org/ontology/Agent, Schema:Organization, Schema:GovernmentOrganization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/GovernmentAgency]), (233, 7, [http://dbpedia.org/resource/Market_sentiment], a []), (333, 7, [http://dbpedia.org/resource/United_Kingdom], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (501, 24, [http://dbpedia.org/resource/National_Weather_Service], a [http://dbpedia.org/ontology/Agent, Schema:Organization, Schema:GovernmentOrganization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/GovernmentAgency]), (1033, 10, [http://dbpedia.org/resource/Gulf_Coast_of_the_United_States], a []), (1107, 12, [http://dbpedia.org/resource/Midcontinent_Communications], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/Company]), (1164, 8, [http://dbpedia.org/resource/New_York_City], a []), (1235, 5, [http://dbpedia.org/resource/New_York_Mercantile_Exchange], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/Company]), (1334, 5, [http://dbpedia.org/resource/New_York_Mercantile_Exchange], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/Company]), (1340, 7, [http://dbpedia.org/resource/Alberta], a []), (1474, 4, [http://dbpedia.org/resource/AECO_Records], a []), (1528, 2, [http://dbpedia.org/resource/Canadian_dollar], a [http://dbpedia.org/ontology/Currency]), (1545, 9, [http://dbpedia.org/resource/Joule], a []), (1575, 5, [http://dbpedia.org/resource/British_thermal_unit], a []), (1638, 6, [http://dbpedia.org/resource/Canada], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (1646, 5, [http://dbpedia.org/resource/New_York_Mercantile_Exchange], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/Company]), (1652, 7, [http://dbpedia.org/resource/Permian], a []), (1743, 7, [http://dbpedia.org/resource/Permian], a []), (2093, 5, [http://dbpedia.org/resource/New_York_Mercantile_Exchange], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/Company]), (2258, 14, [http://dbpedia.org/resource/New_York_Power], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/SportsTeam, http://dbpedia.org/ontology/Organisation, Schema:SportsTeam, http://dbpedia.org/ontology/SoccerClub])]]\nFeb 09, 2018 12:50:43 AM org.restlet.engine.log.LogFilter afterHandle\nINFO: 2018-02-09\t00:50:43\t127.0.0.1\t-\t127.0.0.1\t1235\tPOST\t/gerbil-spotWrapNifWS4Test/spotlight\t-\t200\t17494\t2896\t266\thttp://localhost:1235\tApache-HttpClient/4.5.2 (Java/1.8.0_151)\t-\n2018-02-09 00:50:43,785 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nRequest: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_118, text=U.S. barges lightly quoted on call session. ST. LOUIS 1996-12-06 U.S. barge rates were lightly quoted Friday on the St. Louis Merchants Exchange call session. No barges traded versus no trades Thursday. - Two barges next week Illinois bid at a steady 130 percent of tariff, offered at 135 percent. - One barge, week of December 15, lower Ohio bid 2-1/2 points higher at 105 percent, no offer. Two barges, week of January 5, Illinois, offered five points lower at 195 percent, bid at 150 percent. - Five barges, 30-day open, mid-Mississippi (McGregor and south) bid at 160 percent, offered at 170 percent, no comparisons. - 36 barges, two each week May-August, Illinois, offered at 130 percent of tariff, no bid or comparison. - 36 barges, two each week May-August, mid-Mississippi offered at a steady 135 percent, bid at 120 percent (basis one each week).-- Chicago newsdesk 312-408 8720, markings=[]]\n2018-02-09 00:50:44,001 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_118, text=U.S. barges lightly quoted on call session. ST. LOUIS 1996-12-06 U.S. barge rates were lightly quoted Friday on the St. Louis Merchants Exchange call session. No barges traded versus no trades Thursday. - Two barges next week Illinois bid at a steady 130 percent of tariff, offered at 135 percent. - One barge, week of December 15, lower Ohio bid 2-1/2 points higher at 105 percent, no offer. Two barges, week of January 5, Illinois, offered five points lower at 195 percent, bid at 150 percent. - Five barges, 30-day open, mid-Mississippi (McGregor and south) bid at 160 percent, offered at 170 percent, no comparisons. - 36 barges, two each week May-August, Illinois, offered at 130 percent of tariff, no bid or comparison. - 36 barges, two each week May-August, mid-Mississippi offered at a steady 135 percent, bid at 120 percent (basis one each week).-- Chicago newsdesk 312-408 8720, markings=[(70, 5, [http://dbpedia.org/resource/Barge], a []), (126, 18, [http://dbpedia.org/resource/Merchants_Exchange_Building_(St._Louis)], a []), (226, 8, [http://dbpedia.org/resource/Illinois], a [Schema:AdministrativeArea, Schema:Place, http://dbpedia.org/ontology/AdministrativeRegion, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Region, http://dbpedia.org/ontology/PopulatedPlace]), (304, 5, [http://dbpedia.org/resource/Barge], a []), (338, 4, [http://dbpedia.org/resource/Ohio], a [Schema:AdministrativeArea, Schema:Place, http://dbpedia.org/ontology/AdministrativeRegion, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Region, http://dbpedia.org/ontology/PopulatedPlace]), (424, 8, [http://dbpedia.org/resource/Illinois], a [Schema:AdministrativeArea, Schema:Place, http://dbpedia.org/ontology/AdministrativeRegion, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Region, http://dbpedia.org/ontology/PopulatedPlace]), (660, 8, [http://dbpedia.org/resource/Illinois], a [Schema:AdministrativeArea, Schema:Place, http://dbpedia.org/ontology/AdministrativeRegion, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Region, http://dbpedia.org/ontology/PopulatedPlace]), (858, 7, [http://dbpedia.org/resource/Chicago], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace])]]\nFeb 09, 2018 12:50:44 AM org.restlet.engine.log.LogFilter afterHandle\nINFO: 2018-02-09\t00:50:44\t127.0.0.1\t-\t127.0.0.1\t1235\tPOST\t/gerbil-spotWrapNifWS4Test/spotlight\t-\t200\t6691\t1487\t219\thttp://localhost:1235\tApache-HttpClient/4.5.2 (Java/1.8.0_151)\t-\n2018-02-09 00:50:44,007 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nRequest: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_119, text=CBOT grain / oilseed receipts and shipments. CHICAGO 1996-12-06 Grain and soybean receipts and shipments, in bushels, at delivery locations for the previous trading day, according to the Chicago Board of Trade - Receipts Shipments Wheat Chicago 0 0 St. Louis 21,346 0 Toledo 61,514 0 Corn Chicago 78,056 0 St. Louis 217,092 75,810 Toledo 285,505 561,287 Oats Chicago 0 0 Minneapolis 306,364 153,231 Soybeans Chicago 8,674 484,018 St. Louis 253,821 223,172 Toledo 64,334 160,476 ((Chicago Newsdesk 312-408-8720)), markings=[]]\n2018-02-09 00:50:44,229 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_119, text=CBOT grain / oilseed receipts and shipments. CHICAGO 1996-12-06 Grain and soybean receipts and shipments, in bushels, at delivery locations for the previous trading day, according to the Chicago Board of Trade - Receipts Shipments Wheat Chicago 0 0 St. Louis 21,346 0 Toledo 61,514 0 Corn Chicago 78,056 0 St. Louis 217,092 75,810 Toledo 285,505 561,287 Oats Chicago 0 0 Minneapolis 306,364 153,231 Soybeans Chicago 8,674 484,018 St. Louis 253,821 223,172 Toledo 64,334 160,476 ((Chicago Newsdesk 312-408-8720)), markings=[(0, 4, [http://dbpedia.org/resource/Chicago_Board_of_Trade], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/Company]), (13, 7, [http://dbpedia.org/resource/Vegetable_oil], a []), (45, 7, [http://dbpedia.org/resource/Chicago], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (74, 7, [http://dbpedia.org/resource/Soybean], a [http://dbpedia.org/ontology/Plant, http://dbpedia.org/ontology/Eukaryote, http://dbpedia.org/ontology/Species]), (109, 7, [http://dbpedia.org/resource/Bushel], a []), (187, 7, [http://dbpedia.org/resource/Chicago], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (237, 7, [http://dbpedia.org/resource/Chicago], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (268, 6, [http://dbpedia.org/resource/Toledo,_Ohio], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (289, 7, [http://dbpedia.org/resource/Chicago], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (331, 6, [http://dbpedia.org/resource/Toledo,_Ohio], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (359, 7, [http://dbpedia.org/resource/Chicago], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (371, 11, [http://dbpedia.org/resource/Minneapolis], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (399, 8, [http://dbpedia.org/resource/Soybean], a [http://dbpedia.org/ontology/Plant, http://dbpedia.org/ontology/Eukaryote, http://dbpedia.org/ontology/Species]), (408, 7, [http://dbpedia.org/resource/Chicago], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (456, 6, [http://dbpedia.org/resource/Toledo,_Ohio], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (480, 7, [http://dbpedia.org/resource/Chicago], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace])]]\nFeb 09, 2018 12:50:44 AM org.restlet.engine.log.LogFilter afterHandle\nINFO: 2018-02-09\t00:50:44\t127.0.0.1\t-\t127.0.0.1\t1235\tPOST\t/gerbil-spotWrapNifWS4Test/spotlight\t-\t200\t11832\t1111\t225\thttp://localhost:1235\tApache-HttpClient/4.5.2 (Java/1.8.0_151)\t-\n2018-02-09 00:50:44,235 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nRequest: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_120, text=Clinton to have more news conferences in 2nd term. WASHINGTON 1996-12-06 President Clinton aims to hold more news conferences in his second term and will have one Dec. 13, the White House said Friday. The president had only two formal, full-blown news conferences last year, one in January and one after he won re-election in November, although he had various other limited sessions with the press. White House spokesman Mike McCurry said Clinton \"plans to have regular news conferences\" during his second term. But when asked how frequent these would be, he was evasive, saying, \"periodic, occasional.\" \"He enjoys the give and take\" with reporters, the spokesman added., markings=[]]\n2018-02-09 00:50:44,440 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_120, text=Clinton to have more news conferences in 2nd term. WASHINGTON 1996-12-06 President Clinton aims to hold more news conferences in his second term and will have one Dec. 13, the White House said Friday. The president had only two formal, full-blown news conferences last year, one in January and one after he won re-election in November, although he had various other limited sessions with the press. White House spokesman Mike McCurry said Clinton \"plans to have regular news conferences\" during his second term. But when asked how frequent these would be, he was evasive, saying, \"periodic, occasional.\" \"He enjoys the give and take\" with reporters, the spokesman added., markings=[(51, 10, [http://dbpedia.org/resource/Washington,_D.C.], a [Schema:Place, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (176, 11, [http://dbpedia.org/resource/White_House], a [Schema:Place, http://dbpedia.org/ontology/ArchitecturalStructure, http://dbpedia.org/ontology/Building, http://dbpedia.org/ontology/Place]), (399, 11, [http://dbpedia.org/resource/White_House], a [Schema:Place, http://dbpedia.org/ontology/ArchitecturalStructure, http://dbpedia.org/ontology/Building, http://dbpedia.org/ontology/Place]), (421, 12, [http://dbpedia.org/resource/Mike_McCurry_(press_secretary)], a [http://dbpedia.org/ontology/Agent, Schema:Person, http://dbpedia.org/ontology/Person, http://dbpedia.org/ontology/Politician, Http://xmlns.com/foaf/0.1/Person])]]\nFeb 09, 2018 12:50:44 AM org.restlet.engine.log.LogFilter afterHandle\nINFO: 2018-02-09\t00:50:44\t127.0.0.1\t-\t127.0.0.1\t1235\tPOST\t/gerbil-spotWrapNifWS4Test/spotlight\t-\t200\t3982\t1276\t206\thttp://localhost:1235\tApache-HttpClient/4.5.2 (Java/1.8.0_151)\t-\n2018-02-09 00:50:44,446 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nRequest: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_121, text=Action Performance to acquire firms. TEMPE, Ariz. 1996-12-06 Action Performance Cos Inc said Friday it has agreed to acquire Motorsport Traditions Ltd and Creative Marketing & Promotions Inc for about $ 13 million in cash and stock. The two firms to be acquired have about $ 25 million in annual revenues from the design, manufacture and sale and distribution of licensed motorsports products. The deal is expected to close by the end of the year subject to due diligence and other customary closing conditions., markings=[]]\n2018-02-09 00:50:44,673 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_121, text=Action Performance to acquire firms. TEMPE, Ariz. 1996-12-06 Action Performance Cos Inc said Friday it has agreed to acquire Motorsport Traditions Ltd and Creative Marketing & Promotions Inc for about $ 13 million in cash and stock. The two firms to be acquired have about $ 25 million in annual revenues from the design, manufacture and sale and distribution of licensed motorsports products. The deal is expected to close by the end of the year subject to due diligence and other customary closing conditions., markings=[]]\nFeb 09, 2018 12:50:44 AM org.restlet.engine.log.LogFilter afterHandle\nINFO: 2018-02-09\t00:50:44\t127.0.0.1\t-\t127.0.0.1\t1235\tPOST\t/gerbil-spotWrapNifWS4Test/spotlight\t-\t200\t1111\t1111\t229\thttp://localhost:1235\tApache-HttpClient/4.5.2 (Java/1.8.0_151)\t-\n2018-02-09 00:50:44,679 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nRequest: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_122, text=Half of dog bites provoked, says American vet. CHICAGO 1996-12-06 As many as 1 million dog bites are recorded in the United States every year and half of them are provoked by humans, a veterinarian told fellow animal doctors on Friday. The Humane Society of the United States estimates that between 500,000 and one million bites are delivered by dogs each year, more than half of which are suffered by children. \"Most bites can be prevented by teaching children how to respect a dog,\" Michael Cornwell of the Glencoe Animal Hospital in Columbus, Ohio, told the annual meeting of the American Veterinary Medical Association. \"Let's not let our kids jump on them or crawl on them. Dogs and children do n't have to have an interaction. Let's respect their territories,\" he said. Cornwell said 50 percent of reported bites were provoked by a person and 60 percent were suffered by children. He also estimated that only 25 percent of bites were reported because medical attention was not needed. Don Rieck, president of the National Animal Control Association, said aggressiveness in dogs was related more to gender than breed and a male dog that had not been neutered was three times more likely to bite than an unspayed female. The five breeds credited with the most incidents were chow chows, Rottweilers, German shepherds, cocker spaniels and Dalmatians. \"The trends in dog bites by particular breeds have more to do with fad pets owned by individuals who need to have something unique. Speaking strictly of dogs, 15 years ago the macho fad pet was a Doberman. Today, Rottweilers are on the way up,\" Rieck said. If approached by a stray dog, children should be taught to stand still with fists folded underneath the neck, elbows in, and gaze forward until the dog goes away., markings=[]]\n2018-02-09 00:50:44,928 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_122, text=Half of dog bites provoked, says American vet. CHICAGO 1996-12-06 As many as 1 million dog bites are recorded in the United States every year and half of them are provoked by humans, a veterinarian told fellow animal doctors on Friday. The Humane Society of the United States estimates that between 500,000 and one million bites are delivered by dogs each year, more than half of which are suffered by children. \"Most bites can be prevented by teaching children how to respect a dog,\" Michael Cornwell of the Glencoe Animal Hospital in Columbus, Ohio, told the annual meeting of the American Veterinary Medical Association. \"Let's not let our kids jump on them or crawl on them. Dogs and children do n't have to have an interaction. Let's respect their territories,\" he said. Cornwell said 50 percent of reported bites were provoked by a person and 60 percent were suffered by children. He also estimated that only 25 percent of bites were reported because medical attention was not needed. Don Rieck, president of the National Animal Control Association, said aggressiveness in dogs was related more to gender than breed and a male dog that had not been neutered was three times more likely to bite than an unspayed female. The five breeds credited with the most incidents were chow chows, Rottweilers, German shepherds, cocker spaniels and Dalmatians. \"The trends in dog bites by particular breeds have more to do with fad pets owned by individuals who need to have something unique. Speaking strictly of dogs, 15 years ago the macho fad pet was a Doberman. Today, Rottweilers are on the way up,\" Rieck said. If approached by a stray dog, children should be taught to stand still with fists folded underneath the neck, elbows in, and gaze forward until the dog goes away., markings=[(8, 3, [http://dbpedia.org/resource/Dog], a [http://dbpedia.org/ontology/Eukaryote, http://dbpedia.org/ontology/Species, http://dbpedia.org/ontology/Mammal, http://dbpedia.org/ontology/Animal]), (33, 8, [http://dbpedia.org/resource/United_States], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (47, 7, [http://dbpedia.org/resource/Chicago], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (87, 3, [http://dbpedia.org/resource/Dog], a [http://dbpedia.org/ontology/Eukaryote, http://dbpedia.org/ontology/Species, http://dbpedia.org/ontology/Mammal, http://dbpedia.org/ontology/Animal]), (117, 13, [http://dbpedia.org/resource/United_States], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (185, 12, [http://dbpedia.org/resource/Veterinary_physician], a []), (240, 14, [http://dbpedia.org/resource/Humane_Society_of_the_United_States], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Non-ProfitOrganisation, http://dbpedia.org/ontology/Organisation]), (262, 13, [http://dbpedia.org/resource/United_States], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (479, 3, [http://dbpedia.org/resource/Dog], a [http://dbpedia.org/ontology/Eukaryote, http://dbpedia.org/ontology/Species, http://dbpedia.org/ontology/Mammal, http://dbpedia.org/ontology/Animal]), (509, 7, [http://dbpedia.org/resource/Glencoe,_Illinois], a [Schema:Place, http://dbpedia.org/ontology/Village, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (536, 14, [http://dbpedia.org/resource/Columbus,_Ohio], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (583, 39, [http://dbpedia.org/resource/American_Veterinary_Medical_Association], a []), (1116, 5, [http://dbpedia.org/resource/Dog_breed], a []), (1133, 3, [http://dbpedia.org/resource/Dog], a [http://dbpedia.org/ontology/Eukaryote, http://dbpedia.org/ontology/Species, http://dbpedia.org/ontology/Mammal, http://dbpedia.org/ontology/Animal]), (1208, 8, [http://dbpedia.org/resource/Neutering], a []), (1279, 10, [http://dbpedia.org/resource/Chow_Chow], a []), (1291, 11, [http://dbpedia.org/resource/Rottweiler], a []), (1304, 16, [http://dbpedia.org/resource/German_Shepherd], a []), (1322, 15, [http://dbpedia.org/resource/Cocker_Spaniel], a []), (1369, 3, [http://dbpedia.org/resource/Dog], a [http://dbpedia.org/ontology/Eukaryote, http://dbpedia.org/ontology/Species, http://dbpedia.org/ontology/Mammal, http://dbpedia.org/ontology/Animal]), (1530, 5, [http://dbpedia.org/resource/Machismo], a []), (1567, 11, [http://dbpedia.org/resource/Rottweiler], a []), (1636, 3, [http://dbpedia.org/resource/Dog], a [http://dbpedia.org/ontology/Eukaryote, http://dbpedia.org/ontology/Species, http://dbpedia.org/ontology/Mammal, http://dbpedia.org/ontology/Animal]), (1759, 3, [http://dbpedia.org/resource/Dog], a [http://dbpedia.org/ontology/Eukaryote, http://dbpedia.org/ontology/Species, http://dbpedia.org/ontology/Mammal, http://dbpedia.org/ontology/Animal])]]\nFeb 09, 2018 12:50:44 AM org.restlet.engine.log.LogFilter afterHandle\nINFO: 2018-02-09\t00:50:44\t127.0.0.1\t-\t127.0.0.1\t1235\tPOST\t/gerbil-spotWrapNifWS4Test/spotlight\t-\t200\t17192\t2381\t253\thttp://localhost:1235\tApache-HttpClient/4.5.2 (Java/1.8.0_151)\t-\n2018-02-09 00:50:44,947 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nRequest: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_123, text=Iowa-S Minn feedlot cattle market not tested- USDA. DES MOINES 1996-12-06 Steers and heifers were not tested, compared with Thursday's close, USDA said. Reported sales for Fri- None. Week to Date - None. ((Chicago newsdesk 312-408-8720)), markings=[]]\n2018-02-09 00:50:45,146 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_123, text=Iowa-S Minn feedlot cattle market not tested- USDA. DES MOINES 1996-12-06 Steers and heifers were not tested, compared with Thursday's close, USDA said. Reported sales for Fri- None. Week to Date - None. ((Chicago newsdesk 312-408-8720)), markings=[(12, 7, [http://dbpedia.org/resource/Feedlot], a []), (46, 4, [http://dbpedia.org/resource/United_States_Department_of_Agriculture], a [http://dbpedia.org/ontology/Agent, Schema:Organization, Schema:GovernmentOrganization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/GovernmentAgency]), (52, 3, [http://dbpedia.org/resource/Data_Encryption_Standard], a []), (142, 4, [http://dbpedia.org/resource/United_States_Department_of_Agriculture], a [http://dbpedia.org/ontology/Agent, Schema:Organization, Schema:GovernmentOrganization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/GovernmentAgency]), (206, 7, [http://dbpedia.org/resource/Chicago], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace])]]\nFeb 09, 2018 12:50:45 AM org.restlet.engine.log.LogFilter afterHandle\nINFO: 2018-02-09\t00:50:45\t127.0.0.1\t-\t127.0.0.1\t1235\tPOST\t/gerbil-spotWrapNifWS4Test/spotlight\t-\t200\t4024\t837\t201\thttp://localhost:1235\tApache-HttpClient/4.5.2 (Java/1.8.0_151)\t-\n2018-02-09 00:50:45,151 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nRequest: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_124, text=Nebraska fed cattle roundup - USDA. OMAHA 1996-12-06 Slaughter steers and heifers were not established Thursday. Demand limited. Seller interest light. - USDA Thursday 200 Last Week Holiday Last Year N / A Week to Date 3,500 Sm Pd Lst Wk 800 Sm Pd Lst Yr N / A Dressed Basis Delivered not well tested. Dressed Basis Steers : Few Select and Choice 2-3, 1200-1300 lbs 112.00 ; load early 114.00. Dressed Basis Heifers : Few Select and Choice 2-3, 1100-1200 lbs 112.00., markings=[]]\n2018-02-09 00:50:45,348 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_124, text=Nebraska fed cattle roundup - USDA. OMAHA 1996-12-06 Slaughter steers and heifers were not established Thursday. Demand limited. Seller interest light. - USDA Thursday 200 Last Week Holiday Last Year N / A Week to Date 3,500 Sm Pd Lst Wk 800 Sm Pd Lst Yr N / A Dressed Basis Delivered not well tested. Dressed Basis Steers : Few Select and Choice 2-3, 1200-1300 lbs 112.00 ; load early 114.00. Dressed Basis Heifers : Few Select and Choice 2-3, 1100-1200 lbs 112.00., markings=[(0, 8, [http://dbpedia.org/resource/Nebraska], a [Schema:AdministrativeArea, Schema:Place, http://dbpedia.org/ontology/AdministrativeRegion, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Region, http://dbpedia.org/ontology/PopulatedPlace]), (30, 4, [http://dbpedia.org/resource/United_States_Department_of_Agriculture], a [http://dbpedia.org/ontology/Agent, Schema:Organization, Schema:GovernmentOrganization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/GovernmentAgency]), (154, 4, [http://dbpedia.org/resource/United_States_Department_of_Agriculture], a [http://dbpedia.org/ontology/Agent, Schema:Organization, Schema:GovernmentOrganization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/GovernmentAgency]), (362, 3, [http://dbpedia.org/resource/Pound_(mass)], a []), (455, 3, [http://dbpedia.org/resource/Pound_(mass)], a [])]]\nFeb 09, 2018 12:50:45 AM org.restlet.engine.log.LogFilter afterHandle\nINFO: 2018-02-09\t00:50:45\t127.0.0.1\t-\t127.0.0.1\t1235\tPOST\t/gerbil-spotWrapNifWS4Test/spotlight\t-\t200\t4270\t1066\t200\thttp://localhost:1235\tApache-HttpClient/4.5.2 (Java/1.8.0_151)\t-\n2018-02-09 00:50:45,355 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nRequest: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_125, text=Four Africans said to vie for top U.N. post. Evelyn Leopold UNITED NATIONS 1996-12-06 Four African states are ready to nominate candidates for the post of U.N. secretary-general on Friday now that Boutros Boutros-Ghali has temporarily put aside his bid for re-election. The nominees, according to diplomats, are : Kofi Annan of Ghana, the U.N. undersecretary-general for peacekeeping ; Ahmedou Ould Abdallah of Mauritania, the former U.N. special envoy for Burundi ; Amara Essy of the Ivory Coast, its foreign minister and the U.N. General Assembly president in 1994-95 ; and Hamid Algabid of Niger, the secretary-general of the Organisation of the Islamic Conference. Representatives of the U.N. missions of Ghana, the Ivory Coast, Mauritania and Niger have scheduled a meeting with Security Council president Paolo Fulci of Italy to hand in the nominations in writing, the envoys said. It was not known if other candidates would step forward. Diplomats said General Joseph Garba of Nigeria, a U.N. General Assembly president in 1989-90, was putting forth his own candidacy without being nominated by his country. Boutros-Ghali on Wednesday opened the door for other Africans to contest his job by saying he was suspending temporarily his candidacy but was not withdrawing completely from the race. His supporters said this meant he remained a candidate in case the race reached an impasse. The United States on Nov. 19 vetoed his bid for re-election while the other 14 Security Council members supported him. The move by the African states means that the council could begin voting on candidates next week, a procedure that could either result in a decision or turn into a bitter fight with vetoes against each nominee. The Security Council has to vote on a new secretary-general and then seek the endorsement of the 185-members General Assembly before December 31 when Boutros-Ghali's term expires., markings=[]]\n2018-02-09 00:50:45,604 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_125, text=Four Africans said to vie for top U.N. post. Evelyn Leopold UNITED NATIONS 1996-12-06 Four African states are ready to nominate candidates for the post of U.N. secretary-general on Friday now that Boutros Boutros-Ghali has temporarily put aside his bid for re-election. The nominees, according to diplomats, are : Kofi Annan of Ghana, the U.N. undersecretary-general for peacekeeping ; Ahmedou Ould Abdallah of Mauritania, the former U.N. special envoy for Burundi ; Amara Essy of the Ivory Coast, its foreign minister and the U.N. General Assembly president in 1994-95 ; and Hamid Algabid of Niger, the secretary-general of the Organisation of the Islamic Conference. Representatives of the U.N. missions of Ghana, the Ivory Coast, Mauritania and Niger have scheduled a meeting with Security Council president Paolo Fulci of Italy to hand in the nominations in writing, the envoys said. It was not known if other candidates would step forward. Diplomats said General Joseph Garba of Nigeria, a U.N. General Assembly president in 1989-90, was putting forth his own candidacy without being nominated by his country. Boutros-Ghali on Wednesday opened the door for other Africans to contest his job by saying he was suspending temporarily his candidacy but was not withdrawing completely from the race. His supporters said this meant he remained a candidate in case the race reached an impasse. The United States on Nov. 19 vetoed his bid for re-election while the other 14 Security Council members supported him. The move by the African states means that the council could begin voting on candidates next week, a procedure that could either result in a decision or turn into a bitter fight with vetoes against each nominee. The Security Council has to vote on a new secretary-general and then seek the endorsement of the 185-members General Assembly before December 31 when Boutros-Ghali's term expires., markings=[(34, 4, [http://dbpedia.org/resource/United_Nations], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (60, 14, [http://dbpedia.org/resource/United_Nations], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (155, 22, [http://dbpedia.org/resource/Secretary-General_of_the_United_Nations], a []), (197, 21, [http://dbpedia.org/resource/Boutros_Boutros-Ghali], a [http://dbpedia.org/ontology/Agent, Schema:Person, http://dbpedia.org/ontology/Person, http://dbpedia.org/ontology/OfficeHolder, Http://xmlns.com/foaf/0.1/Person]), (314, 10, [http://dbpedia.org/resource/Kofi_Annan], a [http://dbpedia.org/ontology/Agent, Schema:Person, http://dbpedia.org/ontology/Person, http://dbpedia.org/ontology/OfficeHolder, Http://xmlns.com/foaf/0.1/Person]), (328, 5, [http://dbpedia.org/resource/Ghana], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (339, 4, [http://dbpedia.org/resource/United_Nations], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (371, 12, [http://dbpedia.org/resource/Peacekeeping], a []), (386, 21, [http://dbpedia.org/resource/Ahmedou_Ould-Abdallah], a [http://dbpedia.org/ontology/Agent, Schema:Person, http://dbpedia.org/ontology/Person, Http://xmlns.com/foaf/0.1/Person]), (411, 10, [http://dbpedia.org/resource/Mauritania], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (434, 4, [http://dbpedia.org/resource/United_Nations], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (457, 7, [http://dbpedia.org/resource/Burundi], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (467, 10, [http://dbpedia.org/resource/Amara_Essy], a [http://dbpedia.org/ontology/Agent, Schema:Person, http://dbpedia.org/ontology/Person, http://dbpedia.org/ontology/OfficeHolder, Http://xmlns.com/foaf/0.1/Person]), (485, 11, [http://dbpedia.org/resource/Ivory_Coast], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (527, 4, [http://dbpedia.org/resource/United_Nations], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (532, 26, [http://dbpedia.org/resource/President_of_the_United_Nations_General_Assembly], a []), (576, 13, [http://dbpedia.org/resource/Hamid_Algabid], a []), (593, 5, [http://dbpedia.org/resource/Niger], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (692, 4, [http://dbpedia.org/resource/United_Nations], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (709, 5, [http://dbpedia.org/resource/Ghana], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (720, 11, [http://dbpedia.org/resource/Ivory_Coast], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (733, 10, [http://dbpedia.org/resource/Mauritania], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (748, 5, [http://dbpedia.org/resource/Niger], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (784, 26, [http://dbpedia.org/resource/President_of_the_United_Nations_Security_Council], a []), (826, 5, [http://dbpedia.org/resource/Italy], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (968, 12, [http://dbpedia.org/resource/Joseph_Nanven_Garba], a []), (984, 7, [http://dbpedia.org/resource/Nigeria], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (995, 4, [http://dbpedia.org/resource/United_Nations], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (1000, 26, [http://dbpedia.org/resource/President_of_the_United_Nations_General_Assembly], a []), (1115, 13, [http://dbpedia.org/resource/Boutros_Boutros-Ghali], a [http://dbpedia.org/ontology/Agent, Schema:Person, http://dbpedia.org/ontology/Person, http://dbpedia.org/ontology/OfficeHolder, Http://xmlns.com/foaf/0.1/Person]), (1396, 13, [http://dbpedia.org/resource/United_States], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (1471, 16, [http://dbpedia.org/resource/United_Nations_Security_Council], a []), (1726, 16, [http://dbpedia.org/resource/United_Nations_Security_Council], a []), (1831, 16, [http://dbpedia.org/resource/United_Nations_General_Assembly], a []), (1872, 13, [http://dbpedia.org/resource/Boutros_Boutros-Ghali], a [http://dbpedia.org/ontology/Agent, Schema:Person, http://dbpedia.org/ontology/Person, http://dbpedia.org/ontology/OfficeHolder, Http://xmlns.com/foaf/0.1/Person])]]\nFeb 09, 2018 12:50:45 AM org.restlet.engine.log.LogFilter afterHandle\nINFO: 2018-02-09\t00:50:45\t127.0.0.1\t-\t127.0.0.1\t1235\tPOST\t/gerbil-spotWrapNifWS4Test/spotlight\t-\t200\t24995\t2503\t253\thttp://localhost:1235\tApache-HttpClient/4.5.2 (Java/1.8.0_151)\t-\n2018-02-09 00:50:45,614 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nRequest: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_126, text=Spain's police seize petrol bombs, arrest five. MADRID 1996-12-06 Spanish police said on Friday they had arrested five people and seized more than 90 petrol bombs during disturbances after a protest in the Basque country against Spain's constitution. Hooded protesters threw burning bottles and other objects at police in Pamplona after the protest organised by Herri Batasuna, the political wing of Basque separatist group ETA. Police also confiscated eight kg (18 lb) of screws, balaclavas and spray paint cans. The protest, which attracted several thousand supporters, coincided with the 18th anniversary of Spain's constitution., markings=[]]\n2018-02-09 00:50:45,836 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_126, text=Spain's police seize petrol bombs, arrest five. MADRID 1996-12-06 Spanish police said on Friday they had arrested five people and seized more than 90 petrol bombs during disturbances after a protest in the Basque country against Spain's constitution. Hooded protesters threw burning bottles and other objects at police in Pamplona after the protest organised by Herri Batasuna, the political wing of Basque separatist group ETA. Police also confiscated eight kg (18 lb) of screws, balaclavas and spray paint cans. The protest, which attracted several thousand supporters, coincided with the 18th anniversary of Spain's constitution., markings=[(0, 5, [http://dbpedia.org/resource/Spain], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (21, 12, [http://dbpedia.org/resource/Molotov_cocktail], a []), (48, 6, [http://dbpedia.org/resource/Madrid], a [Schema:Place, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (66, 7, [http://dbpedia.org/resource/Spain], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (150, 12, [http://dbpedia.org/resource/Molotov_cocktail], a []), (206, 14, [http://dbpedia.org/resource/Basque_Country_(autonomous_community)], a [Schema:Place, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (229, 5, [http://dbpedia.org/resource/Spain], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (322, 8, [http://dbpedia.org/resource/Pamplona], a [Schema:Place, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (362, 14, [http://dbpedia.org/resource/Batasuna], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/PoliticalParty]), (400, 6, [http://dbpedia.org/resource/Basque_nationalism], a []), (424, 3, [http://dbpedia.org/resource/ETA], a []), (466, 2, [http://dbpedia.org/resource/Pound_(mass)], a []), (481, 10, [http://dbpedia.org/resource/Balaclava_(clothing)], a []), (496, 11, [http://dbpedia.org/resource/Spray_painting], a []), (611, 5, [http://dbpedia.org/resource/Spain], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country])]]\nFeb 09, 2018 12:50:45 AM org.restlet.engine.log.LogFilter afterHandle\nINFO: 2018-02-09\t00:50:45\t127.0.0.1\t-\t127.0.0.1\t1235\tPOST\t/gerbil-spotWrapNifWS4Test/spotlight\t-\t200\t10188\t1232\t226\thttp://localhost:1235\tApache-HttpClient/4.5.2 (Java/1.8.0_151)\t-\n2018-02-09 00:50:45,843 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nRequest: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_127, text=Mussolini's granddaughter rejoins far-right party. ROME 1996-12-06 Alessandra Mussolini, the granddaughter of Italy's Fascist dictator Benito Mussolini, said on Friday she had rejoined the far-right National Alliance (AN) party she quit over policy differences last month. \"I've gone back,\" she told a radio show shortly after AN leader Gianfranco Fini, who was being interviewed on the programme, said the row had been resolved. \"He did n't want to lose me and I did n't want to lose him.\" Fini told state radio RAI he met Mussolini thanks to the good offices of Giuseppe Tatarella, AN's leader in the Chamber of Deputies (lower house), and had overcome their differences. Mussolini, 33, resigned from the parliamentary party group for what she said were strictly political reasons. The fiery politician, who is also a niece of screen star Sophia Loren, had accused AN leaders of stifling internal party debate. Mussolini, who sits in the Chamber, told La Stampa newspaper last month after quitting AN's parliamentary party that she was considering joining the neo-fascist Social Movement (MS-Fiamma) formed by some of the Duce's World War Two followers., markings=[]]\n2018-02-09 00:50:46,058 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_127, text=Mussolini's granddaughter rejoins far-right party. ROME 1996-12-06 Alessandra Mussolini, the granddaughter of Italy's Fascist dictator Benito Mussolini, said on Friday she had rejoined the far-right National Alliance (AN) party she quit over policy differences last month. \"I've gone back,\" she told a radio show shortly after AN leader Gianfranco Fini, who was being interviewed on the programme, said the row had been resolved. \"He did n't want to lose me and I did n't want to lose him.\" Fini told state radio RAI he met Mussolini thanks to the good offices of Giuseppe Tatarella, AN's leader in the Chamber of Deputies (lower house), and had overcome their differences. Mussolini, 33, resigned from the parliamentary party group for what she said were strictly political reasons. The fiery politician, who is also a niece of screen star Sophia Loren, had accused AN leaders of stifling internal party debate. Mussolini, who sits in the Chamber, told La Stampa newspaper last month after quitting AN's parliamentary party that she was considering joining the neo-fascist Social Movement (MS-Fiamma) formed by some of the Duce's World War Two followers., markings=[(34, 9, [http://dbpedia.org/resource/Far-right_politics], a []), (67, 20, [http://dbpedia.org/resource/Alessandra_Mussolini], a [http://dbpedia.org/ontology/Agent, Schema:Person, http://dbpedia.org/ontology/Person, http://dbpedia.org/ontology/OfficeHolder, Http://xmlns.com/foaf/0.1/Person]), (118, 7, [http://dbpedia.org/resource/Italian_Fascism], a []), (135, 16, [http://dbpedia.org/resource/Benito_Mussolini], a [http://dbpedia.org/ontology/Agent, Schema:Person, http://dbpedia.org/ontology/Person, http://dbpedia.org/ontology/President, http://dbpedia.org/ontology/Politician, Http://xmlns.com/foaf/0.1/Person]), (189, 9, [http://dbpedia.org/resource/Far-right_politics], a []), (199, 17, [http://dbpedia.org/resource/National_Alliance_(Italy)], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/PoliticalParty]), (337, 15, [http://dbpedia.org/resource/Gianfranco_Fini], a [http://dbpedia.org/ontology/Agent, Schema:Person, http://dbpedia.org/ontology/Person, http://dbpedia.org/ontology/OfficeHolder, Http://xmlns.com/foaf/0.1/Person]), (507, 9, [http://dbpedia.org/resource/RAI], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/Company]), (564, 18, [http://dbpedia.org/resource/Giuseppe_Tatarella], a [http://dbpedia.org/ontology/Agent, Schema:Person, http://dbpedia.org/ontology/Person, http://dbpedia.org/ontology/OfficeHolder, Http://xmlns.com/foaf/0.1/Person]), (841, 12, [http://dbpedia.org/resource/Sophia_Loren], a [http://dbpedia.org/ontology/Agent, Schema:Person, http://dbpedia.org/ontology/Person, Http://xmlns.com/foaf/0.1/Person]), (954, 9, [http://dbpedia.org/resource/La_Stampa], a [http://dbpedia.org/ontology/Work, http://dbpedia.org/ontology/PeriodicalLiterature, http://dbpedia.org/ontology/Newspaper, Schema:CreativeWork, http://dbpedia.org/ontology/WrittenWork]), (1062, 11, [http://dbpedia.org/resource/Neo-fascism], a []), (1124, 4, [http://dbpedia.org/resource/Benito_Mussolini], a [http://dbpedia.org/ontology/Agent, Schema:Person, http://dbpedia.org/ontology/Person, http://dbpedia.org/ontology/President, http://dbpedia.org/ontology/Politician, Http://xmlns.com/foaf/0.1/Person])]]\nFeb 09, 2018 12:50:46 AM org.restlet.engine.log.LogFilter afterHandle\nINFO: 2018-02-09\t00:50:46\t127.0.0.1\t-\t127.0.0.1\t1235\tPOST\t/gerbil-spotWrapNifWS4Test/spotlight\t-\t200\t10166\t1761\t216\thttp://localhost:1235\tApache-HttpClient/4.5.2 (Java/1.8.0_151)\t-\n2018-02-09 00:50:46,064 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nRequest: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_128, text=German Santa in bank nearly gets arrested. HANOVER, Germany 1996-12-06 A Santa Claus distributing presents to workers in a German bank on Friday nearly ended up behind bars when a passing police patrol thought he was a robber in disguise. The man, doing his rounds in the northern city of Hanover on the day when German children traditionally receive small presents from Saint Nicholas, convinced police eventually that he was genuine., markings=[]]\n2018-02-09 00:50:46,293 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_128, text=German Santa in bank nearly gets arrested. HANOVER, Germany 1996-12-06 A Santa Claus distributing presents to workers in a German bank on Friday nearly ended up behind bars when a passing police patrol thought he was a robber in disguise. The man, doing his rounds in the northern city of Hanover on the day when German children traditionally receive small presents from Saint Nicholas, convinced police eventually that he was genuine., markings=[(0, 6, [http://dbpedia.org/resource/Germany], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (16, 4, [http://dbpedia.org/resource/Bank], a []), (43, 7, [http://dbpedia.org/resource/Hanover], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (52, 7, [http://dbpedia.org/resource/Germany], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (73, 11, [http://dbpedia.org/resource/Santa_Claus], a []), (123, 6, [http://dbpedia.org/resource/Germany], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (130, 4, [http://dbpedia.org/resource/Bank], a []), (289, 7, [http://dbpedia.org/resource/Hanover], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (313, 6, [http://dbpedia.org/resource/Germany], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (371, 14, [http://dbpedia.org/resource/Santa_Claus], a [])]]\nFeb 09, 2018 12:50:46 AM org.restlet.engine.log.LogFilter afterHandle\nINFO: 2018-02-09\t00:50:46\t127.0.0.1\t-\t127.0.0.1\t1235\tPOST\t/gerbil-spotWrapNifWS4Test/spotlight\t-\t200\t7155\t1035\t231\thttp://localhost:1235\tApache-HttpClient/4.5.2 (Java/1.8.0_151)\t-\n2018-02-09 00:50:46,298 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nRequest: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_129, text=Italy commission concludes 1997 budget examination. ROME 1996-12-06 The Italian upper house Senate budget commission has concluded its examination of Italy's 1997 budget, and it will approve the measure officially by Saturday. From Tuesday, the full assembly of the Senate will start its examination of the financial package.-- Milan newsroom+392 66129502, markings=[]]\n2018-02-09 00:50:46,522 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_129, text=Italy commission concludes 1997 budget examination. ROME 1996-12-06 The Italian upper house Senate budget commission has concluded its examination of Italy's 1997 budget, and it will approve the measure officially by Saturday. From Tuesday, the full assembly of the Senate will start its examination of the financial package.-- Milan newsroom+392 66129502, markings=[(0, 5, [http://dbpedia.org/resource/Italy], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (72, 7, [http://dbpedia.org/resource/Italy], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (80, 11, [http://dbpedia.org/resource/Upper_house], a []), (92, 6, [http://dbpedia.org/resource/Senate_of_the_Republic_(Italy)], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/Legislature]), (150, 5, [http://dbpedia.org/resource/Italy], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (266, 6, [http://dbpedia.org/resource/Senate_of_the_Republic_(Italy)], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/Legislature]), (328, 5, [http://dbpedia.org/resource/Milan], a [Schema:Place, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace])]]\nFeb 09, 2018 12:50:46 AM org.restlet.engine.log.LogFilter afterHandle\nINFO: 2018-02-09\t00:50:46\t127.0.0.1\t-\t127.0.0.1\t1235\tPOST\t/gerbil-spotWrapNifWS4Test/spotlight\t-\t200\t5464\t955\t227\thttp://localhost:1235\tApache-HttpClient/4.5.2 (Java/1.8.0_151)\t-\n2018-02-09 00:50:46,527 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nRequest: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_130, text=EU, Poland agree on oil import tariffs. BRUSSELS 1996-12-06 The European Union and Poland have resolved disagreements over a new Polish oil import regime, the European Commission said on Friday. The EU had objected to increases in Polish tariffs on imports of gasoline and gasoil products introduced on January 1, 1996, saying they contravened levels envisaged in the so-called Europe Agreement between the EU and Poland. The increases were aimed at protecting the Polish market while helping to modernise the local oil industry. \"The EU and Poland have now reached a final settlement regarding issues related to the Polish import regime in the oils sector,\" the Commission said in a statement. Under the agreement, Poland will abolish all oil import tariffs by 2001, remove all oil price controls and end quantitative restrictions on imports by January 1, 1997. The agreement includes the early privatisation and modernisation of Polish oil refineries, which will be obliged to offer equal treatment to all buyers. The EU and Poland will monitor the settlement at six-monthly meetings., markings=[]]\n2018-02-09 00:50:46,764 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_130, text=EU, Poland agree on oil import tariffs. BRUSSELS 1996-12-06 The European Union and Poland have resolved disagreements over a new Polish oil import regime, the European Commission said on Friday. The EU had objected to increases in Polish tariffs on imports of gasoline and gasoil products introduced on January 1, 1996, saying they contravened levels envisaged in the so-called Europe Agreement between the EU and Poland. The increases were aimed at protecting the Polish market while helping to modernise the local oil industry. \"The EU and Poland have now reached a final settlement regarding issues related to the Polish import regime in the oils sector,\" the Commission said in a statement. Under the agreement, Poland will abolish all oil import tariffs by 2001, remove all oil price controls and end quantitative restrictions on imports by January 1, 1997. The agreement includes the early privatisation and modernisation of Polish oil refineries, which will be obliged to offer equal treatment to all buyers. The EU and Poland will monitor the settlement at six-monthly meetings., markings=[(0, 2, [http://dbpedia.org/resource/European_Union], a []), (4, 6, [http://dbpedia.org/resource/Poland], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (31, 7, [http://dbpedia.org/resource/Tariff], a []), (40, 8, [http://dbpedia.org/resource/Brussels], a [Schema:Place, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (64, 14, [http://dbpedia.org/resource/European_Union], a []), (83, 6, [http://dbpedia.org/resource/Poland], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (129, 6, [http://dbpedia.org/resource/Poland], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (159, 19, [http://dbpedia.org/resource/European_Commission], a []), (199, 2, [http://dbpedia.org/resource/European_Union], a []), (231, 6, [http://dbpedia.org/resource/Poland], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (238, 7, [http://dbpedia.org/resource/Tariff], a []), (260, 8, [http://dbpedia.org/resource/Gasoline], a []), (273, 6, [http://dbpedia.org/resource/Fuel_oil], a []), (378, 16, [http://dbpedia.org/resource/European_Union_Association_Agreement], a []), (407, 2, [http://dbpedia.org/resource/European_Union], a []), (414, 6, [http://dbpedia.org/resource/Poland], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (465, 6, [http://dbpedia.org/resource/Poland], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (535, 2, [http://dbpedia.org/resource/European_Union], a []), (542, 6, [http://dbpedia.org/resource/Poland], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (617, 6, [http://dbpedia.org/resource/Poland], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (716, 6, [http://dbpedia.org/resource/Poland], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (751, 7, [http://dbpedia.org/resource/Tariff], a []), (783, 14, [http://dbpedia.org/resource/Price_controls], a []), (896, 13, [http://dbpedia.org/resource/Privatization], a []), (931, 6, [http://dbpedia.org/resource/Poland], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (938, 14, [http://dbpedia.org/resource/Oil_refinery], a []), (1020, 2, [http://dbpedia.org/resource/European_Union], a []), (1027, 6, [http://dbpedia.org/resource/Poland], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country])]]\nFeb 09, 2018 12:50:46 AM org.restlet.engine.log.LogFilter afterHandle\nINFO: 2018-02-09\t00:50:46\t127.0.0.1\t-\t127.0.0.1\t1235\tPOST\t/gerbil-spotWrapNifWS4Test/spotlight\t-\t200\t18019\t1690\t239\thttp://localhost:1235\tApache-HttpClient/4.5.2 (Java/1.8.0_151)\t-\n2018-02-09 00:50:46,773 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nRequest: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_131, text=Hindu party forces India parliament to adjourn. NEW DELHI 1996-12-06 Hindu nationalists forced adjournment of India's lower house of parliament on Friday, in protest against a proposal to observe a minute's silence over the destruction of a mosque by a Hindu mob in 1992. Members of the Hindu nationalist Bharatiya Janata Party (BJP) shouted pro-Hindu slogans in the house after a communist deputy made the proposal in remembrance of the Babri mosque, which was razed on December 6, 1992. The house was first adjourned for two hours. When it reconvened, BJP deputies resumed the slogan-shouting, and deputy speaker Suraj Bhan suspended work until Monday. The destruction of the 16th-century mosque in the northern Indian town of Ayodhya triggered nationwide Hindu-Moslem violence in which more than 3,000 people were killed. Indian officials blame revenge-minded Moslem underworld gangs in Bombay for a string of bombings in the city three months later that killed 260 people. The BJP backs a hardline Hindu campaign to build a temple at the site of the mosque, which Hindus believe was the birthplace of the Lord Rama. The campaign catapulted BJP from the political fringe to become India's main opposition party in 1991., markings=[]]\n2018-02-09 00:50:47,012 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_131, text=Hindu party forces India parliament to adjourn. NEW DELHI 1996-12-06 Hindu nationalists forced adjournment of India's lower house of parliament on Friday, in protest against a proposal to observe a minute's silence over the destruction of a mosque by a Hindu mob in 1992. Members of the Hindu nationalist Bharatiya Janata Party (BJP) shouted pro-Hindu slogans in the house after a communist deputy made the proposal in remembrance of the Babri mosque, which was razed on December 6, 1992. The house was first adjourned for two hours. When it reconvened, BJP deputies resumed the slogan-shouting, and deputy speaker Suraj Bhan suspended work until Monday. The destruction of the 16th-century mosque in the northern Indian town of Ayodhya triggered nationwide Hindu-Moslem violence in which more than 3,000 people were killed. Indian officials blame revenge-minded Moslem underworld gangs in Bombay for a string of bombings in the city three months later that killed 260 people. The BJP backs a hardline Hindu campaign to build a temple at the site of the mosque, which Hindus believe was the birthplace of the Lord Rama. The campaign catapulted BJP from the political fringe to become India's main opposition party in 1991., markings=[(0, 5, [http://dbpedia.org/resource/Hindu_nationalism], a []), (19, 5, [http://dbpedia.org/resource/India], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (69, 18, [http://dbpedia.org/resource/Hindu_nationalism], a []), (110, 5, [http://dbpedia.org/resource/India], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (241, 6, [http://dbpedia.org/resource/Babri_Mosque], a []), (253, 5, [http://dbpedia.org/resource/Hindu_nationalism], a []), (287, 17, [http://dbpedia.org/resource/Hindu_nationalism], a []), (305, 22, [http://dbpedia.org/resource/Bharatiya_Janata_Party], a []), (329, 3, [http://dbpedia.org/resource/Bharatiya_Janata_Party], a []), (438, 12, [http://dbpedia.org/resource/Babri_Mosque], a []), (554, 3, [http://dbpedia.org/resource/Bharatiya_Janata_Party], a []), (691, 6, [http://dbpedia.org/resource/Babri_Mosque], a []), (714, 6, [http://dbpedia.org/resource/India], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (729, 7, [http://dbpedia.org/resource/Ayodhya], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (825, 6, [http://dbpedia.org/resource/India], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (863, 6, [http://dbpedia.org/resource/Muslim], a []), (890, 6, [http://dbpedia.org/resource/Mumbai], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (981, 3, [http://dbpedia.org/resource/Bharatiya_Janata_Party], a []), (1002, 5, [http://dbpedia.org/resource/Hindu_nationalism], a []), (1054, 6, [http://dbpedia.org/resource/Babri_Mosque], a []), (1068, 6, [http://dbpedia.org/resource/Hindu], a []), (1109, 9, [http://dbpedia.org/resource/Rama], a []), (1144, 3, [http://dbpedia.org/resource/Bharatiya_Janata_Party], a []), (1184, 5, [http://dbpedia.org/resource/India], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country])]]\nFeb 09, 2018 12:50:47 AM org.restlet.engine.log.LogFilter afterHandle\nINFO: 2018-02-09\t00:50:47\t127.0.0.1\t-\t127.0.0.1\t1235\tPOST\t/gerbil-spotWrapNifWS4Test/spotlight\t-\t200\t15469\t1824\t242\thttp://localhost:1235\tApache-HttpClient/4.5.2 (Java/1.8.0_151)\t-\n2018-02-09 00:50:47,021 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nRequest: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_132, text=Indian Sept crude oil output falls to 2.6 mln T. NEW DELHI 1996-12-06 India's crude petroleum output fell to 2.56 million tonnes in September from 2.83 million in the same month in 1995, the government said on Friday. STEEL OUTPUT Sept Sept Apr-Sept Apr-Sept 1996 1995 1996 1995 Crude petroleum 2,557 2,832 15,838 17,648 Petroleum products 4,605 5,110 30,589 29,328 Note - Figures are in thousands of tonnes and preliminary., markings=[]]\n2018-02-09 00:50:47,232 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_132, text=Indian Sept crude oil output falls to 2.6 mln T. NEW DELHI 1996-12-06 India's crude petroleum output fell to 2.56 million tonnes in September from 2.83 million in the same month in 1995, the government said on Friday. STEEL OUTPUT Sept Sept Apr-Sept Apr-Sept 1996 1995 1996 1995 Crude petroleum 2,557 2,832 15,838 17,648 Petroleum products 4,605 5,110 30,589 29,328 Note - Figures are in thousands of tonnes and preliminary., markings=[(0, 6, [http://dbpedia.org/resource/India], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (12, 9, [http://dbpedia.org/resource/Petroleum], a []), (70, 5, [http://dbpedia.org/resource/India], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (84, 9, [http://dbpedia.org/resource/Petroleum], a []), (285, 9, [http://dbpedia.org/resource/Petroleum], a [])]]\nFeb 09, 2018 12:50:47 AM org.restlet.engine.log.LogFilter afterHandle\nINFO: 2018-02-09\t00:50:47\t127.0.0.1\t-\t127.0.0.1\t1235\tPOST\t/gerbil-spotWrapNifWS4Test/spotlight\t-\t200\t3886\t1024\t215\thttp://localhost:1235\tApache-HttpClient/4.5.2 (Java/1.8.0_151)\t-\n2018-02-09 00:50:47,238 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nRequest: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_133, text=LUXEMBOURG CHRISTMAS MARKET GOES ON WORLD WIDE WEB. BRUSSELS 1996-12-06 Luxembourg's traditional Christmas market, which starts on Saturday and runs to December 24, has taken to the world wide web as a way of publicising its activities. The web site (http://www.pt.lu/infoweb/kreschtmaart) gives details of the market's concert programme as well as its various retailers.-- Brussels Newsroom+32 2 287 6810, Fax+32 2 230 7710, markings=[]]\n2018-02-09 00:50:47,437 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_133, text=LUXEMBOURG CHRISTMAS MARKET GOES ON WORLD WIDE WEB. BRUSSELS 1996-12-06 Luxembourg's traditional Christmas market, which starts on Saturday and runs to December 24, has taken to the world wide web as a way of publicising its activities. The web site (http://www.pt.lu/infoweb/kreschtmaart) gives details of the market's concert programme as well as its various retailers.-- Brussels Newsroom+32 2 287 6810, Fax+32 2 230 7710, markings=[(0, 10, [http://dbpedia.org/resource/Luxembourg], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (28, 4, [http://dbpedia.org/resource/Geostationary_Operational_Environmental_Satellite], a []), (52, 8, [http://dbpedia.org/resource/Brussels], a [Schema:Place, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (72, 10, [http://dbpedia.org/resource/Luxembourg], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (97, 16, [http://dbpedia.org/resource/Christmas_market], a []), (374, 8, [http://dbpedia.org/resource/Brussels], a [Schema:Place, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace])]]\nFeb 09, 2018 12:50:47 AM org.restlet.engine.log.LogFilter afterHandle\nINFO: 2018-02-09\t00:50:47\t127.0.0.1\t-\t127.0.0.1\t1235\tPOST\t/gerbil-spotWrapNifWS4Test/spotlight\t-\t200\t4747\t1024\t203\thttp://localhost:1235\tApache-HttpClient/4.5.2 (Java/1.8.0_151)\t-\n2018-02-09 00:50:47,442 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nRequest: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_134, text=London coal / ore fixtures. LONDON 1996-12-06 COAL - Lantau Peak - 120,000 tones coal Hay Point or Newcastle / Kaohsiung 20-30/1 $ 5.25 and $ 5.85 fio respectively 40,000-35,000 / 28,000 shinc China Steel. Royal Clipper - 77,000 tonnes coal Maracaibo / Fos 19/12-2/1 $ 9.90 fio 20,000 shinc / 25,000 shinc Coe and Clerici. ORE - IMC TBN - 70,000 tonnes Dampier / Kaohsiung 20-30/12 $ 5.25 fio 35,000 shinc / 30,000 shinc China Steel., markings=[]]\n2018-02-09 00:50:47,671 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_134, text=London coal / ore fixtures. LONDON 1996-12-06 COAL - Lantau Peak - 120,000 tones coal Hay Point or Newcastle / Kaohsiung 20-30/1 $ 5.25 and $ 5.85 fio respectively 40,000-35,000 / 28,000 shinc China Steel. Royal Clipper - 77,000 tonnes coal Maracaibo / Fos 19/12-2/1 $ 9.90 fio 20,000 shinc / 25,000 shinc Coe and Clerici. ORE - IMC TBN - 70,000 tonnes Dampier / Kaohsiung 20-30/12 $ 5.25 fio 35,000 shinc / 30,000 shinc China Steel., markings=[(0, 6, [http://dbpedia.org/resource/London], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (7, 4, [http://dbpedia.org/resource/Coal], a []), (53, 11, [http://dbpedia.org/resource/Lantau_Peak], a [Schema:Place, Schema:Mountain, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/NaturalPlace, http://dbpedia.org/ontology/Mountain]), (81, 4, [http://dbpedia.org/resource/Coal], a []), (86, 9, [http://dbpedia.org/resource/Hay_Point,_Queensland], a [Schema:AdministrativeArea, Schema:Place, http://dbpedia.org/ontology/AdministrativeRegion, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Region, http://dbpedia.org/ontology/PopulatedPlace]), (99, 9, [http://dbpedia.org/resource/Newcastle_United_F.C.], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/SportsTeam, http://dbpedia.org/ontology/Organisation, Schema:SportsTeam, http://dbpedia.org/ontology/SoccerClub]), (111, 9, [http://dbpedia.org/resource/Kaohsiung], a [Schema:Place, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (126, 2, [http://dbpedia.org/resource/1], a [http://dbpedia.org/ontology/TimePeriod, http://dbpedia.org/ontology/Year]), (193, 5, [http://dbpedia.org/resource/China], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (206, 13, [http://dbpedia.org/resource/Royal_Clipper], a [http://dbpedia.org/ontology/Ship, Schema:Product, http://dbpedia.org/ontology/MeanOfTransportation]), (236, 4, [http://dbpedia.org/resource/Coal], a []), (241, 9, [http://dbpedia.org/resource/Maracaibo], a [Schema:Place, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (253, 3, [http://dbpedia.org/resource/Marseille], a [Schema:Place, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (259, 3, [http://dbpedia.org/resource/12], a [http://dbpedia.org/ontology/TimePeriod, http://dbpedia.org/ontology/Year]), (264, 2, [http://dbpedia.org/resource/1], a [http://dbpedia.org/ontology/TimePeriod, http://dbpedia.org/ontology/Year]), (333, 3, [http://dbpedia.org/resource/Trinity_Broadcasting_Network], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Broadcaster, Schema:TelevisionStation, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/TelevisionStation]), (353, 7, [http://dbpedia.org/resource/Dampier,_Western_Australia], a [Schema:AdministrativeArea, Schema:Place, http://dbpedia.org/ontology/AdministrativeRegion, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Region, http://dbpedia.org/ontology/PopulatedPlace]), (363, 9, [http://dbpedia.org/resource/Kaohsiung], a [Schema:Place, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (378, 3, [http://dbpedia.org/resource/12], a [http://dbpedia.org/ontology/TimePeriod, http://dbpedia.org/ontology/Year]), (421, 5, [http://dbpedia.org/resource/China], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country])]]\nFeb 09, 2018 12:50:47 AM org.restlet.engine.log.LogFilter afterHandle\nINFO: 2018-02-09\t00:50:47\t127.0.0.1\t-\t127.0.0.1\t1235\tPOST\t/gerbil-spotWrapNifWS4Test/spotlight\t-\t200\t13793\t1033\t232\thttp://localhost:1235\tApache-HttpClient/4.5.2 (Java/1.8.0_151)\t-\n2018-02-09 00:50:47,678 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nRequest: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_135, text=UK bookmakers lengthen Conservative victory odds. LONDON 1996-12-06 UK bookmakers William Hill said on Friday they have lengthened the odds of a Conservative victory in the next general election from 9-4 to 5-2. William Hill said the odds were the longest they had been for six months. The Labour opposition are now 1-4 favourites, it said. The election must be held by May.-- London Newsroom+44 171 542-7768, markings=[]]\n2018-02-09 00:50:47,908 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_135, text=UK bookmakers lengthen Conservative victory odds. LONDON 1996-12-06 UK bookmakers William Hill said on Friday they have lengthened the odds of a Conservative victory in the next general election from 9-4 to 5-2. William Hill said the odds were the longest they had been for six months. The Labour opposition are now 1-4 favourites, it said. The election must be held by May.-- London Newsroom+44 171 542-7768, markings=[(0, 2, [http://dbpedia.org/resource/United_Kingdom], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (3, 10, [http://dbpedia.org/resource/Bookmaker], a []), (23, 12, [http://dbpedia.org/resource/Conservative_Party_(UK)], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/PoliticalParty]), (68, 2, [http://dbpedia.org/resource/United_Kingdom], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (71, 10, [http://dbpedia.org/resource/Bookmaker], a []), (82, 12, [http://dbpedia.org/resource/William_Hill_(bookmaker)], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/Company]), (145, 12, [http://dbpedia.org/resource/Conservative_Party_(UK)], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/PoliticalParty]), (178, 16, [http://dbpedia.org/resource/United_Kingdom_general_election,_2010], a [Schema:Event, http://dbpedia.org/ontology/Election, http://dbpedia.org/ontology/Event]), (212, 12, [http://dbpedia.org/resource/William_Hill_(bookmaker)], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/Company]), (290, 6, [http://dbpedia.org/resource/Labour_Party_(UK)], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/PoliticalParty]), (377, 6, [http://dbpedia.org/resource/London], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (392, 3, [http://dbpedia.org/resource/Telephone_numbers_in_the_United_Kingdom], a [])]]\nFeb 09, 2018 12:50:47 AM org.restlet.engine.log.LogFilter afterHandle\nINFO: 2018-02-09\t00:50:47\t127.0.0.1\t-\t127.0.0.1\t1235\tPOST\t/gerbil-spotWrapNifWS4Test/spotlight\t-\t200\t8635\t1008\t233\thttp://localhost:1235\tApache-HttpClient/4.5.2 (Java/1.8.0_151)\t-\n2018-02-09 00:50:47,914 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nRequest: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_136, text=Italy tops week of meagre bond returns - Salomon. LONDON 1996-12-06 High-flying Italy topped the league in a week of meagre returns on government bonds, Salomon Brothers said on Friday. In local currency terms, Italian BTPs offered returns of 0.85 percent in the week ended on Thursday, with fellow high-yielder Sweden close behind on 0.80 percent. The weekly government bond index rose 0.07 percent in local currency terms. France managed third place with 0.68 percent in the 16-nation world government bond index. Canada's were the worst performing bonds. They lost 2.21 percent, depressed by a wave of new Canadian supply. Returns on Treasuries were also in negative territory at minus 0.24 percent, the poorest result after Canada and British gilts which lost 0.28 percent. Australia was the only dollar-bloc country in the table to eke out a positive return, albeit a paltry 0.02 percent. German Bunds were not much better, offering returns of 0.05 percent, while Japanese government bonds managed a 0.38 percent gain. Spanish bonds, which had been top performers in Salomon Brothers ' league table for November as a whole, turned in a more subdued weekly performance with a return of only 0.27 percent. In U.S. dollar terms, Japan was the only country to give positive returns at 1.35 percent. France lost 0.37 percent, followed by Italy on minus 0.59 percent. The biggest losers in dollar terms were British gilts, which shed 3.41 percent, Canada with minus 3.03 percent and Australia at minus 1.54 percent. Salomon's bond index is calculated using all government bonds with over one year to maturity, weighted for market capitalisation. Only bonds freely available to institutional investors and with a certain minimum amount outstanding are included. Returns take account of price moves and accrued interest.-- Stephen Nisbet, International Bonds+44 171 6320, markings=[]]\n2018-02-09 00:50:48,226 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_136, text=Italy tops week of meagre bond returns - Salomon. LONDON 1996-12-06 High-flying Italy topped the league in a week of meagre returns on government bonds, Salomon Brothers said on Friday. In local currency terms, Italian BTPs offered returns of 0.85 percent in the week ended on Thursday, with fellow high-yielder Sweden close behind on 0.80 percent. The weekly government bond index rose 0.07 percent in local currency terms. France managed third place with 0.68 percent in the 16-nation world government bond index. Canada's were the worst performing bonds. They lost 2.21 percent, depressed by a wave of new Canadian supply. Returns on Treasuries were also in negative territory at minus 0.24 percent, the poorest result after Canada and British gilts which lost 0.28 percent. Australia was the only dollar-bloc country in the table to eke out a positive return, albeit a paltry 0.02 percent. German Bunds were not much better, offering returns of 0.05 percent, while Japanese government bonds managed a 0.38 percent gain. Spanish bonds, which had been top performers in Salomon Brothers ' league table for November as a whole, turned in a more subdued weekly performance with a return of only 0.27 percent. In U.S. dollar terms, Japan was the only country to give positive returns at 1.35 percent. France lost 0.37 percent, followed by Italy on minus 0.59 percent. The biggest losers in dollar terms were British gilts, which shed 3.41 percent, Canada with minus 3.03 percent and Australia at minus 1.54 percent. Salomon's bond index is calculated using all government bonds with over one year to maturity, weighted for market capitalisation. Only bonds freely available to institutional investors and with a certain minimum amount outstanding are included. Returns take account of price moves and accrued interest.-- Stephen Nisbet, International Bonds+44 171 6320, markings=[(0, 5, [http://dbpedia.org/resource/Italy], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (80, 5, [http://dbpedia.org/resource/Italy], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (135, 16, [http://dbpedia.org/resource/Government_bond], a []), (153, 16, [http://dbpedia.org/resource/Salomon_Brothers], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/Company]), (211, 7, [http://dbpedia.org/resource/Italians], a [http://dbpedia.org/ontology/EthnicGroup]), (219, 4, [http://dbpedia.org/resource/Volume_(thermodynamics)], a []), (312, 6, [http://dbpedia.org/resource/Sweden], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (360, 15, [http://dbpedia.org/resource/Government_bond], a []), (425, 6, [http://dbpedia.org/resource/France], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (493, 15, [http://dbpedia.org/resource/Government_bond], a []), (516, 6, [http://dbpedia.org/resource/Canada], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (609, 8, [http://dbpedia.org/resource/Canada], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (728, 6, [http://dbpedia.org/resource/Canada], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (739, 7, [http://dbpedia.org/resource/United_Kingdom], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (747, 5, [http://dbpedia.org/resource/Gilt-edged_securities], a []), (778, 9, [http://dbpedia.org/resource/Australia], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (894, 6, [http://dbpedia.org/resource/Germans], a [http://dbpedia.org/ontology/EthnicGroup]), (969, 8, [http://dbpedia.org/resource/Japan], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (978, 16, [http://dbpedia.org/resource/Government_bond], a []), (1024, 7, [http://dbpedia.org/resource/Spanish_language], a [Schema:Language, http://dbpedia.org/ontology/Language]), (1072, 16, [http://dbpedia.org/resource/Salomon_Brothers], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/Company]), (1212, 11, [http://dbpedia.org/resource/United_States_dollar], a [http://dbpedia.org/ontology/Currency]), (1231, 5, [http://dbpedia.org/resource/Economy_of_Japan], a []), (1300, 6, [http://dbpedia.org/resource/France], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (1338, 5, [http://dbpedia.org/resource/Italy], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (1407, 7, [http://dbpedia.org/resource/United_Kingdom], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (1415, 5, [http://dbpedia.org/resource/Gilt-edged_securities], a []), (1447, 6, [http://dbpedia.org/resource/Canada], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (1482, 9, [http://dbpedia.org/resource/Australia], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (1560, 16, [http://dbpedia.org/resource/Government_bond], a []), (1622, 21, [http://dbpedia.org/resource/Market_capitalization], a []), (1676, 23, [http://dbpedia.org/resource/Institutional_investor], a []), (1800, 16, [http://dbpedia.org/resource/Accrued_interest], a []), (1855, 3, [http://dbpedia.org/resource/Telephone_numbers_in_the_United_Kingdom], a [])]]\nFeb 09, 2018 12:50:48 AM org.restlet.engine.log.LogFilter afterHandle\nINFO: 2018-02-09\t00:50:48\t127.0.0.1\t-\t127.0.0.1\t1235\tPOST\t/gerbil-spotWrapNifWS4Test/spotlight\t-\t200\t23078\t2469\t315\thttp://localhost:1235\tApache-HttpClient/4.5.2 (Java/1.8.0_151)\t-\n2018-02-09 00:50:48,234 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nRequest: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_137, text=OPEC basket price $ 24.20 on Thursday. LONDON 1996-12-06 The price of the OPEC basket of seven crudes stood at $ 24.20 a barrel on Thursday, against $ 23.47 on Wednesday, the OPECNA news agency said, quoting the OPEC secretariat. The basket comprises Algeria's Saharan Blend, Indonesia's Minas, Nigeria's Bonny Light, Saudi Arabia's Arabian Light, Dubai of the UAE, Venezuela's Tia Juana and Mexico's Isthmus.-- London Newsroom+44 171 542 7630, markings=[]]\n2018-02-09 00:50:48,469 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_137, text=OPEC basket price $ 24.20 on Thursday. LONDON 1996-12-06 The price of the OPEC basket of seven crudes stood at $ 24.20 a barrel on Thursday, against $ 23.47 on Wednesday, the OPECNA news agency said, quoting the OPEC secretariat. The basket comprises Algeria's Saharan Blend, Indonesia's Minas, Nigeria's Bonny Light, Saudi Arabia's Arabian Light, Dubai of the UAE, Venezuela's Tia Juana and Mexico's Isthmus.-- London Newsroom+44 171 542 7630, markings=[(0, 11, [http://dbpedia.org/resource/OPEC_Reference_Basket], a []), (74, 11, [http://dbpedia.org/resource/OPEC_Reference_Basket], a []), (182, 11, [http://dbpedia.org/resource/Reuters], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/Company]), (212, 4, [http://dbpedia.org/resource/OPEC], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation]), (251, 7, [http://dbpedia.org/resource/Algeria], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (261, 7, [http://dbpedia.org/resource/Sahara], a [http://dbpedia.org/ontology/Place, Schema:Place]), (276, 9, [http://dbpedia.org/resource/Indonesia], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (295, 7, [http://dbpedia.org/resource/Nigeria], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (318, 12, [http://dbpedia.org/resource/Saudi_Arabia], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (333, 7, [http://dbpedia.org/resource/Arabian_Peninsula], a []), (348, 5, [http://dbpedia.org/resource/Dubai], a [Schema:Place, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (361, 3, [http://dbpedia.org/resource/United_Arab_Emirates], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (366, 9, [http://dbpedia.org/resource/Venezuela], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (392, 6, [http://dbpedia.org/resource/Mexico], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (412, 6, [http://dbpedia.org/resource/London], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (427, 3, [http://dbpedia.org/resource/Telephone_numbers_in_the_United_Kingdom], a [])]]\nFeb 09, 2018 12:50:48 AM org.restlet.engine.log.LogFilter afterHandle\nINFO: 2018-02-09\t00:50:48\t127.0.0.1\t-\t127.0.0.1\t1235\tPOST\t/gerbil-spotWrapNifWS4Test/spotlight\t-\t200\t11050\t1043\t238\thttp://localhost:1235\tApache-HttpClient/4.5.2 (Java/1.8.0_151)\t-\n2018-02-09 00:50:48,478 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nRequest: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_138, text=Relations between Clarke, Major good - spokesman. LONDON 1996-12-06 Relations between Chancellor of the Exchequer Kenneth Clarke and Prime Minister John Major are good despite media reports of a rift over European policy, a spokesman for Major's office said on Friday. Asked about the reports, the spokesman said : \"Relations are good.\" Asked about Major's mood after a day of media speculation about his political fortunes, the spokesman said : \"He is resolute. He is getting on with the job.\" The spokesman said he was not aware of any meetings overnight between Clarke and Major, nor of any talks between the prime minister and parliamentary business managers. Both Major and Clarke were in their constituencies on Friday., markings=[]]\n2018-02-09 00:50:48,680 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_138, text=Relations between Clarke, Major good - spokesman. LONDON 1996-12-06 Relations between Chancellor of the Exchequer Kenneth Clarke and Prime Minister John Major are good despite media reports of a rift over European policy, a spokesman for Major's office said on Friday. Asked about the reports, the spokesman said : \"Relations are good.\" Asked about Major's mood after a day of media speculation about his political fortunes, the spokesman said : \"He is resolute. He is getting on with the job.\" The spokesman said he was not aware of any meetings overnight between Clarke and Major, nor of any talks between the prime minister and parliamentary business managers. Both Major and Clarke were in their constituencies on Friday., markings=[(104, 9, [http://dbpedia.org/resource/Chancellor_of_the_Exchequer], a []), (114, 14, [http://dbpedia.org/resource/Kenneth_Clarke], a [http://dbpedia.org/ontology/Agent, Schema:Person, http://dbpedia.org/ontology/Person, http://dbpedia.org/ontology/OfficeHolder, Http://xmlns.com/foaf/0.1/Person]), (133, 14, [http://dbpedia.org/resource/Prime_Minister_of_the_United_Kingdom], a []), (148, 10, [http://dbpedia.org/resource/John_Major], a [http://dbpedia.org/ontology/Agent, Schema:Person, http://dbpedia.org/ontology/Person, http://dbpedia.org/ontology/OfficeHolder, Http://xmlns.com/foaf/0.1/Person])]]\nFeb 09, 2018 12:50:48 AM org.restlet.engine.log.LogFilter afterHandle\nINFO: 2018-02-09\t00:50:48\t127.0.0.1\t-\t127.0.0.1\t1235\tPOST\t/gerbil-spotWrapNifWS4Test/spotlight\t-\t200\t3803\t1329\t205\thttp://localhost:1235\tApache-HttpClient/4.5.2 (Java/1.8.0_151)\t-\n2018-02-09 00:50:48,685 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nRequest: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_139, text=Two dead after executive jet crashes in Newfoundland. STEPHENVILLE, Newfoundland 1996-12-06 Two people were killed when an executive jet en route to Ireland from Michigan crashed on approach to an airport in Stephenville, Newfoundland, on Friday, authorities said. The pilot and co-pilot, the only two aboard, were killed in the crash of the Learjet 36, airport manager David Snow said in a telephone interview. Snow said the plane last reported to air traffic control at about 3 A.M. local time / 1:30 A.M. EST (0630 GMT) when it began its final approach about 10 miles (16 km) from the airport in this east coast Canadian province. That was the last communication the aircraft made with the airport, he added. \"We considered it as being missing until about 0600 (4:30 A.M. EST) (0930 GMT). That's when the wreckage was discovered,\" Snow said. He said the cargo flight originated in Grand Rapids, Michigan, and was due to stop at Stephenville for refueling before going to Shannon, Ireland. The cause of the crash was not yet known. Investigators were due to fly to Stephenville later on Friday., markings=[]]\n2018-02-09 00:50:48,929 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_139, text=Two dead after executive jet crashes in Newfoundland. STEPHENVILLE, Newfoundland 1996-12-06 Two people were killed when an executive jet en route to Ireland from Michigan crashed on approach to an airport in Stephenville, Newfoundland, on Friday, authorities said. The pilot and co-pilot, the only two aboard, were killed in the crash of the Learjet 36, airport manager David Snow said in a telephone interview. Snow said the plane last reported to air traffic control at about 3 A.M. local time / 1:30 A.M. EST (0630 GMT) when it began its final approach about 10 miles (16 km) from the airport in this east coast Canadian province. That was the last communication the aircraft made with the airport, he added. \"We considered it as being missing until about 0600 (4:30 A.M. EST) (0930 GMT). That's when the wreckage was discovered,\" Snow said. He said the cargo flight originated in Grand Rapids, Michigan, and was due to stop at Stephenville for refueling before going to Shannon, Ireland. The cause of the crash was not yet known. Investigators were due to fly to Stephenville later on Friday., markings=[(25, 3, [http://dbpedia.org/resource/Jet_aircraft], a []), (40, 12, [http://dbpedia.org/resource/Newfoundland_and_Labrador], a []), (54, 12, [http://dbpedia.org/resource/Stephenville,_Newfoundland_and_Labrador], a [Schema:Place, http://dbpedia.org/ontology/Town, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (68, 12, [http://dbpedia.org/resource/Newfoundland_and_Labrador], a []), (133, 3, [http://dbpedia.org/resource/Jet_aircraft], a []), (149, 7, [http://dbpedia.org/resource/Ireland], a [Schema:Place, http://dbpedia.org/ontology/Island, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace]), (162, 8, [http://dbpedia.org/resource/Michigan], a [Schema:AdministrativeArea, Schema:Place, http://dbpedia.org/ontology/AdministrativeRegion, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Region, http://dbpedia.org/ontology/PopulatedPlace]), (197, 7, [http://dbpedia.org/resource/Airport], a []), (208, 26, [http://dbpedia.org/resource/Stephenville,_Newfoundland_and_Labrador], a [Schema:Place, http://dbpedia.org/ontology/Town, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (342, 7, [http://dbpedia.org/resource/Learjet], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/Company]), (354, 7, [http://dbpedia.org/resource/Airport], a []), (370, 10, [http://dbpedia.org/resource/David_Snow], a []), (449, 19, [http://dbpedia.org/resource/Air_traffic_control], a []), (480, 3, [http://dbpedia.org/resource/AM_broadcasting], a []), (503, 3, [http://dbpedia.org/resource/AM_broadcasting], a []), (508, 3, [http://dbpedia.org/resource/Eastern_Time_Zone], a []), (518, 3, [http://dbpedia.org/resource/Greenwich_Mean_Time], a []), (588, 7, [http://dbpedia.org/resource/Airport], a []), (615, 8, [http://dbpedia.org/resource/Canada], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (693, 7, [http://dbpedia.org/resource/Airport], a []), (770, 3, [http://dbpedia.org/resource/AM_broadcasting], a []), (775, 3, [http://dbpedia.org/resource/Eastern_Time_Zone], a []), (786, 3, [http://dbpedia.org/resource/Greenwich_Mean_Time], a []), (884, 22, [http://dbpedia.org/resource/Grand_Rapids,_Michigan], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (931, 12, [http://dbpedia.org/resource/Stephenville,_Newfoundland_and_Labrador], a [Schema:Place, http://dbpedia.org/ontology/Town, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (974, 16, [http://dbpedia.org/resource/Shannon_Airport], a [Schema:Place, http://dbpedia.org/ontology/ArchitecturalStructure, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Airport, http://dbpedia.org/ontology/Infrastructure, Schema:Airport]), (1067, 12, [http://dbpedia.org/resource/Stephenville,_Newfoundland_and_Labrador], a [Schema:Place, http://dbpedia.org/ontology/Town, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace])]]\nFeb 09, 2018 12:50:48 AM org.restlet.engine.log.LogFilter afterHandle\nINFO: 2018-02-09\t00:50:48\t127.0.0.1\t-\t127.0.0.1\t1235\tPOST\t/gerbil-spotWrapNifWS4Test/spotlight\t-\t200\t17660\t1700\t248\thttp://localhost:1235\tApache-HttpClient/4.5.2 (Java/1.8.0_151)\t-\n2018-02-09 00:50:48,948 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nRequest: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_140, text=PLO says Arafat, Netanyahu could meet Saturday. JERUSALEM 1996-12-06 PLO negotiators said on Friday Palestinian President Yasser Arafat, Israeli Prime Minister Benjamin Netanyahu and Egyptian President Hosni Mubarak might all meet on Saturday to try to clinch a deal on Israel's handover of Hebron to the PLO. \"It is very possible that Arafat and Netanyahu will meet in Cairo on Saturday. There is work on arranging such a meeting hosted by President Mubarak,\" one PLO official, who requested anonymity, told Reuters. Israeli officials said no meeting had yet been set. Arafat's adviser Nabil Abu Rdainah said : \"President Arafat is ready to meet Prime Minister Netanyahu but no time or date has been set for such a meeting yet.\" President Arafat's position is clear that such a meeting should come after successful negotiations so that the meeting would have positive results. Especially since the Hebron issue has not been agreed yet and the crucial disputed issues have not been resolved. \"But Rdainah said Arafat would go to Cairo on Saturday for talks with Mubarak. Both Arafat and Netanyahu have expressed willingness to meet. They last met in Washington after clashes in September that killed 60 Palestinians and 15 Israelis. The violence was spurred by Israel's opening an entrance to a tunnel near Moslem sites in Jerusalem. The Palestine Liberation Organisation (PLO) negotiators said the last two weeks of talks with Israel on implementing the long-delayed handover of most of Hebron to PLO rule had been\" meaningless \", necessitating an Arafat-Netanyahu meeting. Mubarak's adviser Osama el-Baz said on Thursday there were efforts to arrange a meeting between the Israeli and Palestinian leaders. Palestinian Authority Secretary General Ahmed Abdel-Rahman said on Thursday he understood it could be held in Cairo either on Friday or Sunday. Abdel-Rahman had said on Thursday he did not think Saturday would be the date because it is the Jewish sabbath. But the Jewish sabbath ends at sundown, so a night meeting would not interfere with the religious observance., markings=[]]\n2018-02-09 00:50:49,191 DEBUG [org.aksw.gerbil.ws4test.SpotlightResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_140, text=PLO says Arafat, Netanyahu could meet Saturday. JERUSALEM 1996-12-06 PLO negotiators said on Friday Palestinian President Yasser Arafat, Israeli Prime Minister Benjamin Netanyahu and Egyptian President Hosni Mubarak might all meet on Saturday to try to clinch a deal on Israel's handover of Hebron to the PLO. \"It is very possible that Arafat and Netanyahu will meet in Cairo on Saturday. There is work on arranging such a meeting hosted by President Mubarak,\" one PLO official, who requested anonymity, told Reuters. Israeli officials said no meeting had yet been set. Arafat's adviser Nabil Abu Rdainah said : \"President Arafat is ready to meet Prime Minister Netanyahu but no time or date has been set for such a meeting yet.\" President Arafat's position is clear that such a meeting should come after successful negotiations so that the meeting would have positive results. Especially since the Hebron issue has not been agreed yet and the crucial disputed issues have not been resolved. \"But Rdainah said Arafat would go to Cairo on Saturday for talks with Mubarak. Both Arafat and Netanyahu have expressed willingness to meet. They last met in Washington after clashes in September that killed 60 Palestinians and 15 Israelis. The violence was spurred by Israel's opening an entrance to a tunnel near Moslem sites in Jerusalem. The Palestine Liberation Organisation (PLO) negotiators said the last two weeks of talks with Israel on implementing the long-delayed handover of most of Hebron to PLO rule had been\" meaningless \", necessitating an Arafat-Netanyahu meeting. Mubarak's adviser Osama el-Baz said on Thursday there were efforts to arrange a meeting between the Israeli and Palestinian leaders. Palestinian Authority Secretary General Ahmed Abdel-Rahman said on Thursday he understood it could be held in Cairo either on Friday or Sunday. Abdel-Rahman had said on Thursday he did not think Saturday would be the date because it is the Jewish sabbath. But the Jewish sabbath ends at sundown, so a night meeting would not interfere with the religious observance., markings=[(0, 3, [http://dbpedia.org/resource/Palestine_Liberation_Organization], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/PoliticalParty]), (48, 9, [http://dbpedia.org/resource/Jerusalem], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (69, 3, [http://dbpedia.org/resource/Palestine_Liberation_Organization], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/PoliticalParty]), (100, 21, [http://dbpedia.org/resource/President_of_the_Palestinian_National_Authority], a []), (122, 13, [http://dbpedia.org/resource/Yasser_Arafat], a [http://dbpedia.org/ontology/Agent, Schema:Person, http://dbpedia.org/ontology/Person, http://dbpedia.org/ontology/President, http://dbpedia.org/ontology/Politician, Http://xmlns.com/foaf/0.1/Person]), (137, 22, [http://dbpedia.org/resource/Prime_Minister_of_Israel], a []), (160, 18, [http://dbpedia.org/resource/Benjamin_Netanyahu], a [http://dbpedia.org/ontology/Agent, Schema:Person, http://dbpedia.org/ontology/Person, http://dbpedia.org/ontology/OfficeHolder, Http://xmlns.com/foaf/0.1/Person]), (183, 8, [http://dbpedia.org/resource/Egypt], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (202, 13, [http://dbpedia.org/resource/Hosni_Mubarak], a [http://dbpedia.org/ontology/Agent, Schema:Person, http://dbpedia.org/ontology/Person, http://dbpedia.org/ontology/OfficeHolder, Http://xmlns.com/foaf/0.1/Person]), (270, 6, [http://dbpedia.org/resource/Israel], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (291, 6, [http://dbpedia.org/resource/Hebron], a [Schema:Place, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (305, 3, [http://dbpedia.org/resource/Palestine_Liberation_Organization], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/PoliticalParty]), (370, 5, [http://dbpedia.org/resource/Cairo], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (465, 3, [http://dbpedia.org/resource/Palestine_Liberation_Organization], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/PoliticalParty]), (509, 7, [http://dbpedia.org/resource/Reuters], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/Company]), (518, 7, [http://dbpedia.org/resource/Israel], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (647, 14, [http://dbpedia.org/resource/Prime_Minister_of_Israel], a []), (899, 6, [http://dbpedia.org/resource/Hebron], a [Schema:Place, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (1029, 5, [http://dbpedia.org/resource/Cairo], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (1150, 10, [http://dbpedia.org/resource/Washington,_D.C.], a [Schema:Place, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (1203, 12, [http://dbpedia.org/resource/Palestinian_National_Authority], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (1261, 6, [http://dbpedia.org/resource/Israel], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (1307, 6, [http://dbpedia.org/resource/Muslim], a []), (1323, 9, [http://dbpedia.org/resource/Jerusalem], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (1338, 33, [http://dbpedia.org/resource/Palestine_Liberation_Organization], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/PoliticalParty]), (1373, 3, [http://dbpedia.org/resource/Palestine_Liberation_Organization], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/PoliticalParty]), (1428, 6, [http://dbpedia.org/resource/Israel], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (1488, 6, [http://dbpedia.org/resource/Hebron], a [Schema:Place, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (1498, 3, [http://dbpedia.org/resource/Palestine_Liberation_Organization], a [http://dbpedia.org/ontology/Agent, Schema:Organization, http://dbpedia.org/ontology/Organisation, http://dbpedia.org/ontology/PoliticalParty]), (1593, 12, [http://dbpedia.org/resource/Osama_El-Baz], a []), (1675, 7, [http://dbpedia.org/resource/Israel], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (1687, 11, [http://dbpedia.org/resource/Palestinian_National_Authority], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (1708, 21, [http://dbpedia.org/resource/Palestinian_National_Authority], a [Schema:Place, http://dbpedia.org/ontology/Country, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/PopulatedPlace, Schema:Country]), (1818, 5, [http://dbpedia.org/resource/Cairo], a [Schema:Place, http://dbpedia.org/ontology/City, Schema:City, http://dbpedia.org/ontology/Place, http://dbpedia.org/ontology/Settlement, http://dbpedia.org/ontology/PopulatedPlace]), (1948, 14, [http://dbpedia.org/resource/Shabbat], a []), (1972, 14, [http://dbpedia.org/resource/Shabbat], a [])]]\nFeb 09, 2018 12:50:49 AM org.restlet.engine.log.LogFilter afterHandle\nINFO: 2018-02-09\t00:50:49\t127.0.0.1\t-\t127.0.0.1\t1235\tPOST\t/gerbil-spotWrapNifWS4Test/spotlight\t-\t200\t26359\t2682\t249\thttp://localhost:1235\tApache-HttpClient/4.5.2 (Java/1.8.0_151)\t-\n\n" }, { "path": "gerbil-SpotWrapNifWS4Test/my_notes/python_server_format", "content": "\n\n\nfor TEST-A uri=http://www.aksw.org/gerbil/NifWebService/request_215, was the last one\n\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - BELARUS BEAT ESTONIA IN WORLD CUP QUALIFIER. MINSK 1996-08-31 Belarus beat Estonia 1-0 (halftime 1-0) in a World Cup soccer European group 4 qualifier on Saturday. Scorer : Vladimir Makovsky (35th) Attendance : 6,000\"}'\ntext: SOCCER - BELARUS BEAT ESTONIA IN WORLD CUP QUALIFIER. MINSK 1996-08-31 Belarus beat Estonia 1-0 (halftime 1-0) in a World Cup soccer European group 4 qualifier on Saturday. Scorer : Vladimir Makovsky (35th) Attendance : 6,000\ntype(text): \nspans: []\ntype(spans): \n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"Pirelli cables look to tap Chinese growth. David Jones MILAN 1996-08-30 Italian tyre and cables giant Pirelli on Friday announced its long-awaited move into China with a cables joint venture set to capitalise on the rapidly-growing Chinese telecommunications market. Pirelli is linking with Hong Kong-based group CITIC Pacific in a venture to be called the Wuxi Tong Ling Company Ltd, which will operate in partnership with a local industrial company at its existing factory in Wuxi, Jiangsu province, Shanghai. The partners will invest around $ 30 million in the existing copper cable plant at Wuxi to update technology and include optic fibre production, with annual turnover expected to reach $ 60 million within the next few years. The move marks a further move by Pirelli\\\\u0027s cables division to expand in the fast-growing Far Eastern developing markets with the group already present in Indonesia, India and Malaysia. \\\\\"This is really positive news for Pirelli, and I expect that it will produce one of the best half-year results in late September compared to other industrial Italian companies,\\\\\" said analyst Paula Buratti at Indosuez. She emphasised that the move was positive because Pirelli will have management control of the Chinese venture, and it also showed another example of Pirelli exporting its technical know-how to developing markets. Pirelli shares reacted favourable even though talks had been underway for some time and news about a venture had been widely expected. The shares rose 0.2 percent to 2,555 lire by 1350 GMT in an easier Milan stock market. This will be Pirelli\\\\u0027s first industrial involvement in a Chinese market where demand for telecommunication networks is expected to grow to 80-100 million new lines between 1996 and 2000, doubling demand for optical cables. China\\\\u0027s second largest telecoms operator Unicom already has a mandate from the central government to establish 15 million new phone lines by the year 2000, which will necessitate new trunk line systems and local distribution networks. \\\\\"The starting of this production base in China has for our group an undoubted strategic value, representing an important enhancement of our presence in Asia,\\\\\" said Pirelli SpA chairman and chief executive officer Marco Tonchetti Provera. Pirelli Cables has global sales of over $ 3.5 billion, and has become a large supplier of optic cables and systems to major telecoms carriers in the U.S., Europe and the Far East. CITIC Pacific is a major Hong Kong-listed company focusing on infrastruture, trading, distribution and property, with 28 percent of its 1995 profits coming from telecoms. It has investments in several industrial joint ventures in China.\"}'\ntext: Pirelli cables look to tap Chinese growth. David Jones MILAN 1996-08-30 Italian tyre and cables giant Pirelli on Friday announced its long-awaited move into China with a cables joint venture set to capitalise on the rapidly-growing Chinese telecommunications market. Pirelli is linking with Hong Kong-based group CITIC Pacific in a venture to be called the Wuxi Tong Ling Company Ltd, which will operate in partnership with a local industrial company at its existing factory in Wuxi, Jiangsu province, Shanghai. The partners will invest around $ 30 million in the existing copper cable plant at Wuxi to update technology and include optic fibre production, with annual turnover expected to reach $ 60 million within the next few years. The move marks a further move by Pirelli's cables division to expand in the fast-growing Far Eastern developing markets with the group already present in Indonesia, India and Malaysia. \"This is really positive news for Pirelli, and I expect that it will produce one of the best half-year results in late September compared to other industrial Italian companies,\" said analyst Paula Buratti at Indosuez. She emphasised that the move was positive because Pirelli will have management control of the Chinese venture, and it also showed another example of Pirelli exporting its technical know-how to developing markets. Pirelli shares reacted favourable even though talks had been underway for some time and news about a venture had been widely expected. The shares rose 0.2 percent to 2,555 lire by 1350 GMT in an easier Milan stock market. This will be Pirelli's first industrial involvement in a Chinese market where demand for telecommunication networks is expected to grow to 80-100 million new lines between 1996 and 2000, doubling demand for optical cables. China's second largest telecoms operator Unicom already has a mandate from the central government to establish 15 million new phone lines by the year 2000, which will necessitate new trunk line systems and local distribution networks. \"The starting of this production base in China has for our group an undoubted strategic value, representing an important enhancement of our presence in Asia,\" said Pirelli SpA chairman and chief executive officer Marco Tonchetti Provera. Pirelli Cables has global sales of over $ 3.5 billion, and has become a large supplier of optic cables and systems to major telecoms carriers in the U.S., Europe and the Far East. CITIC Pacific is a major Hong Kong-listed company focusing on infrastruture, trading, distribution and property, with 28 percent of its 1995 profits coming from telecoms. It has investments in several industrial joint ventures in China.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"Dutch bond futures revival delayed - EOE. AMSTERDAM 1996-08-30 A broad attempt to spur activity in Dutch bond futures has been delayed to give participants a chance to become familiar with the trading system, the European Options Exchange (EOE) said on Friday. Market-making in the rarely-traded FTO contract was expected to begin today, but an EOE spokesman said the 10 banks and brokers involved in the initiative needed time to get accustomed to changes in the electronic trading system. \\\\\"It\\\\u0027s not ready yet. We found it wise to take some time between the commitment to start and the actual start,\\\\\" EOE spokesman Lex van Drooge told Reuters. He said no date had been fixed yet for the start of price making in the 10-year contract, but the EOE had agreed to speak again to the participants in one to two weeks. Investors in Dutch bonds currently use German bond futures to hedge their portfolios because the FTO contract is so illiquid. A limited attempt to reinvigorate the contract two years ago failed.-- Amsterdam newsroom+31 20 504 5000, Fax+31 20 504 5040\"}'\ntext: Dutch bond futures revival delayed - EOE. AMSTERDAM 1996-08-30 A broad attempt to spur activity in Dutch bond futures has been delayed to give participants a chance to become familiar with the trading system, the European Options Exchange (EOE) said on Friday. Market-making in the rarely-traded FTO contract was expected to begin today, but an EOE spokesman said the 10 banks and brokers involved in the initiative needed time to get accustomed to changes in the electronic trading system. \"It's not ready yet. We found it wise to take some time between the commitment to start and the actual start,\" EOE spokesman Lex van Drooge told Reuters. He said no date had been fixed yet for the start of price making in the 10-year contract, but the EOE had agreed to speak again to the participants in one to two weeks. Investors in Dutch bonds currently use German bond futures to hedge their portfolios because the FTO contract is so illiquid. A limited attempt to reinvigorate the contract two years ago failed.-- Amsterdam newsroom+31 20 504 5000, Fax+31 20 504 5040\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"OSCE defends record over Chechnya peace mission. BONN 1996-08-30 The head of an international mediating mission defended its record on Friday in the face of criticism by pro-Moscow leaders in breakway Chechnya and insisted it was doing its best to bring peace to the region. Flavio Cotti, the chairman of the Organisation for Security and Cooperation in Europe (OSCE), told German radio the Vienna-based body viewed the conflict in Chechnya as an internal Russian problem. \\\\\"The OSCE is completely involved. But one must not forget that the OSCE only has limited powers there,\\\\\" said Cotti, who is also the Swiss foreign minister. \\\\\"Our mission in Chechnya has done all it can within the given limitations.\\\\\" Pro-Moscow leaders in Chechnya have criticised Tim Guldimann, the Swiss diplomat who heads the OSCE Chechnya mission, saying he was biased toward Zelimkhan Yandarbiyev, president of the self-declared separatist government. Russian peacemaker Alexander Lebed and Chechen separatist military leader Aslan Maskhadov started a new round of peace talks on Friday just outside the rebel region. Cotti said Chechnya must remain part of Russia, but the solution to the conflict would be to accord the region maximum autonomy within Russia\\\\u0027s borders. \\\\\"There is no doubt that Chechnya, according to OSCE principles, belongs to a state called Russia,\\\\\" he said, pointing out that Russia was an OSCE member and it was not the organisation\\\\u0027s policy to challenge members \\\\u0027 sovereignty. He added that the OSCE was the only international body which has been allowed into the Chechnya to monitor the human rights situation there, but that its means were restricted by the fact that the conflict was a \\\\\"internal issue\\\\\". \\\\\"We have a small concept, the details of which have yet to be worked out. Chechnya must be accorded the maximum autonomy possible within the framework of Russian integrity,\\\\\" said Cotti.\"}'\ntext: OSCE defends record over Chechnya peace mission. BONN 1996-08-30 The head of an international mediating mission defended its record on Friday in the face of criticism by pro-Moscow leaders in breakway Chechnya and insisted it was doing its best to bring peace to the region. Flavio Cotti, the chairman of the Organisation for Security and Cooperation in Europe (OSCE), told German radio the Vienna-based body viewed the conflict in Chechnya as an internal Russian problem. \"The OSCE is completely involved. But one must not forget that the OSCE only has limited powers there,\" said Cotti, who is also the Swiss foreign minister. \"Our mission in Chechnya has done all it can within the given limitations.\" Pro-Moscow leaders in Chechnya have criticised Tim Guldimann, the Swiss diplomat who heads the OSCE Chechnya mission, saying he was biased toward Zelimkhan Yandarbiyev, president of the self-declared separatist government. Russian peacemaker Alexander Lebed and Chechen separatist military leader Aslan Maskhadov started a new round of peace talks on Friday just outside the rebel region. Cotti said Chechnya must remain part of Russia, but the solution to the conflict would be to accord the region maximum autonomy within Russia's borders. \"There is no doubt that Chechnya, according to OSCE principles, belongs to a state called Russia,\" he said, pointing out that Russia was an OSCE member and it was not the organisation's policy to challenge members ' sovereignty. He added that the OSCE was the only international body which has been allowed into the Chechnya to monitor the human rights situation there, but that its means were restricted by the fact that the conflict was a \"internal issue\". \"We have a small concept, the details of which have yet to be worked out. Chechnya must be accorded the maximum autonomy possible within the framework of Russian integrity,\" said Cotti.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"Dutch say no reason to reopen El Al carqo enquiry. THE HAGUE 1996-08-30 The Dutch transport minister Annemarie Jorritsma told the country \\\\u0027 second chamber that there is no further need to investigate the 1992 crash of an El Al freighter which left 43 dead in an Amsterdam suburb. She said that a request from her ministry for the aircraft\\\\u0027s waybill documentation and further information about the contents of its hold had been complied with by El Al\\\\u0027s head office in Tel Aviv. The Dutch transport ministry had come in for pressure from a cross-section of Dutch members of parliaments in May this year, some of whom believed the aircraft had been carrying unlisted, dangerous goods. Others said they thought the aircraft was loaded with too much airfreight. Jorritsma said the latest evidence from El Al in no way supported the allegations, and added there is no justification for a further investigation into the incident.-- Air Cargo Newsroom Tel+44 171 542 8982 Fax+44 171 542 5017\"}'\ntext: Dutch say no reason to reopen El Al carqo enquiry. THE HAGUE 1996-08-30 The Dutch transport minister Annemarie Jorritsma told the country ' second chamber that there is no further need to investigate the 1992 crash of an El Al freighter which left 43 dead in an Amsterdam suburb. She said that a request from her ministry for the aircraft's waybill documentation and further information about the contents of its hold had been complied with by El Al's head office in Tel Aviv. The Dutch transport ministry had come in for pressure from a cross-section of Dutch members of parliaments in May this year, some of whom believed the aircraft had been carrying unlisted, dangerous goods. Others said they thought the aircraft was loaded with too much airfreight. Jorritsma said the latest evidence from El Al in no way supported the allegations, and added there is no justification for a further investigation into the incident.-- Air Cargo Newsroom Tel+44 171 542 8982 Fax+44 171 542 5017\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"Armenians, Azeris hold peace talks in Germany. BONN 1996-08-30 Representatives from Armenia and Azerbaijan held talks earlier this week in Germany on bringing a lasting peace to the disputed Nagorno-Karabakh region, a diplomatic source close to the talks said on Friday. The source, who spoke on condition of anonymity, said Azerbaijani presidential adviser Vafa Gulizade and his Armenian counterpart Zhirayr Liparityan met to discuss the disputed enclave on Wednesday and had now flown home. An uneasy ceasefire has prevailed in Nagorno-Karabakh, which represents around 20 percent of Azeri territory, since May 1994 after ethnic Armenians drove Azeris out of the region. The conflict, which began in 1988, claimed over 10,000 lives. \\\\\"The main subject (of the talks) was the search for a peaceful solution for Nagorno-Karabakh,\\\\\" the source said. He declined to reveal any more details about the content of the talks or their exact location in Germany. Azerbaijan has said it is prepared to grant autonomy to Nagorno-Karabakh if Armenian forces pull out, but will not accept Armenia\\\\u0027s demands for the independence of the enclave. Russia\\\\u0027s Interfax news agency reported on Tuesday the officials had departed for negotiations in Germany, adding that face-to-face talks between the two sides first took place last December in Amsterdam. Interfax said the discussions were being held in parallel with peace talks mediated by the Organisation for Security and Cooperation in Europe (OSCE) and the broad-based Minsk Group of countries led by Russia and Finland.\"}'\ntext: Armenians, Azeris hold peace talks in Germany. BONN 1996-08-30 Representatives from Armenia and Azerbaijan held talks earlier this week in Germany on bringing a lasting peace to the disputed Nagorno-Karabakh region, a diplomatic source close to the talks said on Friday. The source, who spoke on condition of anonymity, said Azerbaijani presidential adviser Vafa Gulizade and his Armenian counterpart Zhirayr Liparityan met to discuss the disputed enclave on Wednesday and had now flown home. An uneasy ceasefire has prevailed in Nagorno-Karabakh, which represents around 20 percent of Azeri territory, since May 1994 after ethnic Armenians drove Azeris out of the region. The conflict, which began in 1988, claimed over 10,000 lives. \"The main subject (of the talks) was the search for a peaceful solution for Nagorno-Karabakh,\" the source said. He declined to reveal any more details about the content of the talks or their exact location in Germany. Azerbaijan has said it is prepared to grant autonomy to Nagorno-Karabakh if Armenian forces pull out, but will not accept Armenia's demands for the independence of the enclave. Russia's Interfax news agency reported on Tuesday the officials had departed for negotiations in Germany, adding that face-to-face talks between the two sides first took place last December in Amsterdam. Interfax said the discussions were being held in parallel with peace talks mediated by the Organisation for Security and Cooperation in Europe (OSCE) and the broad-based Minsk Group of countries led by Russia and Finland.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"Sombre mood on Arctic island after plane crash. Rolf Soderlind LONGYEAR, Norway 1996-08-30 The windblown, chilly streets of this tiny Arctic town are all but deserted and flags are flying at half-mast beneath a brooding, clouded sky. Longyear is a town in mourning, a close-knit community that has been shattered. Disaster struck on Thursday when a Russian airliner bringing coal miners to work crashed as it came in to land at the airport, killing all 141 people on board. \\\\\"It\\\\u0027s a sight I will never forget. I will remember it for the rest of my life,\\\\\" said Stig Onarheim. He was one of a handful of rescuers who raced to the scene of the crash in a helicopter on Thursday, hoping in vain to find survivors. The plane smashed into a snow-capped mountain on the Arctic island of Spitzbergen on Thursday, just east of Longyear. \\\\\"Imagine a big plane with a lot of luggage and people on board. Think of all that mixed together, with twisted, wrecked parts on the slope,\\\\\" Onarheim, 29, told Reuters. Police and local officials have sealed off the crash site, protecting it from intrusive reporters and from the polar bears that roam freely across the icy expanses. The dead were all Russians and Ukrainians, coming to work in the mining towns of Barentsburg and Pyramiden. Longyear is a Norwegian settlement of just over 1,000 people, but it also feels the loss keenly. \\\\\"I have trouble finding the words to express my grief. It\\\\u0027s a tragedy for everyone. We know many of the people who live in Barentsburg, some of them could have been on the plane,\\\\\" said Johan Sletten, 52. Sletten, a caretaker who has lived on the island for 30 years, said the Norwegian and Russian communities visit frequently, competing at soccer in the summer and with snow-scooter races in the winter. Teenage shop assistant Heidi Groenstein was blunter. \\\\\"I\\\\u0027m glad it was not a Norwegian plane,\\\\\" she said. \\\\\"Just think of it-- a mining village where so many workers die. They must be having a tough time of it now.\\\\\" Barentsburg, just a few hours ride by snow-scooter or 15 minutes by helicopter from Longyear, has asked to be left alone with its grief and told reporters to stay away. Around 100 Russian and Ukrainian miners were waiting in Longyear to fly home on the plane that crashed. They were given shelter in the town\\\\u0027s church overnight and ate a sombre breakfast before getting on a bus for the airport. Another plane had been sent from Moscow to pick them up. At this time of year, the only colour in Longyear comes from the brightly-painted wooden houses. Everything else is muddy, the waters of the fjord leaden. Winter is in the air. Barentsburg is an even grimmer place, a run-down testament to the hardships of the new Russia. Spitzbergen lies some 500 miles (800 km) off the northern tip of Norway and endures one of the most extreme climates on the planet. Inhabited by fewer than 3,000 people in total, it sees the sun for 24 hours a day during summer and is plunged into round-the-clock darkness in the winter months. The terrain is mountainous, the only roads are dirt tracks. Norway rules the island group under the terms of a 1920s international treaty which gave many other nations the right to establish setttlements and exploit the coal that is still mined there. Only Russia has chosen to do so.\"}'\ntext: Sombre mood on Arctic island after plane crash. Rolf Soderlind LONGYEAR, Norway 1996-08-30 The windblown, chilly streets of this tiny Arctic town are all but deserted and flags are flying at half-mast beneath a brooding, clouded sky. Longyear is a town in mourning, a close-knit community that has been shattered. Disaster struck on Thursday when a Russian airliner bringing coal miners to work crashed as it came in to land at the airport, killing all 141 people on board. \"It's a sight I will never forget. I will remember it for the rest of my life,\" said Stig Onarheim. He was one of a handful of rescuers who raced to the scene of the crash in a helicopter on Thursday, hoping in vain to find survivors. The plane smashed into a snow-capped mountain on the Arctic island of Spitzbergen on Thursday, just east of Longyear. \"Imagine a big plane with a lot of luggage and people on board. Think of all that mixed together, with twisted, wrecked parts on the slope,\" Onarheim, 29, told Reuters. Police and local officials have sealed off the crash site, protecting it from intrusive reporters and from the polar bears that roam freely across the icy expanses. The dead were all Russians and Ukrainians, coming to work in the mining towns of Barentsburg and Pyramiden. Longyear is a Norwegian settlement of just over 1,000 people, but it also feels the loss keenly. \"I have trouble finding the words to express my grief. It's a tragedy for everyone. We know many of the people who live in Barentsburg, some of them could have been on the plane,\" said Johan Sletten, 52. Sletten, a caretaker who has lived on the island for 30 years, said the Norwegian and Russian communities visit frequently, competing at soccer in the summer and with snow-scooter races in the winter. Teenage shop assistant Heidi Groenstein was blunter. \"I'm glad it was not a Norwegian plane,\" she said. \"Just think of it-- a mining village where so many workers die. They must be having a tough time of it now.\" Barentsburg, just a few hours ride by snow-scooter or 15 minutes by helicopter from Longyear, has asked to be left alone with its grief and told reporters to stay away. Around 100 Russian and Ukrainian miners were waiting in Longyear to fly home on the plane that crashed. They were given shelter in the town's church overnight and ate a sombre breakfast before getting on a bus for the airport. Another plane had been sent from Moscow to pick them up. At this time of year, the only colour in Longyear comes from the brightly-painted wooden houses. Everything else is muddy, the waters of the fjord leaden. Winter is in the air. Barentsburg is an even grimmer place, a run-down testament to the hardships of the new Russia. Spitzbergen lies some 500 miles (800 km) off the northern tip of Norway and endures one of the most extreme climates on the planet. Inhabited by fewer than 3,000 people in total, it sees the sun for 24 hours a day during summer and is plunged into round-the-clock darkness in the winter months. The terrain is mountainous, the only roads are dirt tracks. Norway rules the island group under the terms of a 1920s international treaty which gave many other nations the right to establish setttlements and exploit the coal that is still mined there. Only Russia has chosen to do so.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"Ericsson says wins 1.2 bln SKR China order. STOCKHOLM 1996-08-30 Swedish telecoms group LM Ericsson AB said on Friday it won an order worth 1.2 billion crowns for a fixed public telecoms network in the Guangdong province of China. Ericsson said in a statement the order was from the Guangdong Post and Telecommunications Administration (GPTA). The order included AXE switching equipment, ISDN equipment, Intelligent Network (IN) products, broad-band multi-media communication network products, services and training, Ericsson spokesman Per Zetterquist told Reuters. Deliveries are due to be completed by 1999, the company said.-- Stockholm newsroom+46-8-700 1017\"}'\ntext: Ericsson says wins 1.2 bln SKR China order. STOCKHOLM 1996-08-30 Swedish telecoms group LM Ericsson AB said on Friday it won an order worth 1.2 billion crowns for a fixed public telecoms network in the Guangdong province of China. Ericsson said in a statement the order was from the Guangdong Post and Telecommunications Administration (GPTA). The order included AXE switching equipment, ISDN equipment, Intelligent Network (IN) products, broad-band multi-media communication network products, services and training, Ericsson spokesman Per Zetterquist told Reuters. Deliveries are due to be completed by 1999, the company said.-- Stockholm newsroom+46-8-700 1017\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"HK has infrastructure in place for post-97 - Tsang. AUCKLAND 1996-08-30 Hong Kong Financial Secretary Donald Tsang said on Friday that the territory had the \\\\\"infrastructural hardware\\\\\" to make a success of its future under Chinese sovereignty from mid-1997. \\\\\"We have the largest and most efficient port on the South China coast ; we have the best transport and telecommunications infrastructure in the world ; and we are investing in this hardware on an enormous scale,\\\\\" Tsang said in a speech to Auckland during a visit to New Zealand. Hong Kong also had the necessary \\\\\"constitutional infrastructure\\\\\" in place, with the promise of autonomy in running its affairs after the handover from Britain to China. \\\\\"What this means in practice is that Hong Kong will go on raising its own taxes, issuing its own currency, setting its own expenditure priorities and managing its own enormous financial reserves,\\\\\" Tsang said. He acknowledged that many Hong Kong people had decided to seek their future elsewhere and others were sure to follow in the next nine months. \\\\\"But for the great majority of us, Hong Kong is our home and Hong Kong\\\\u0027s future is our future.\\\\\"-- Wellington newsroom 64 4 473-4746\"}'\ntext: HK has infrastructure in place for post-97 - Tsang. AUCKLAND 1996-08-30 Hong Kong Financial Secretary Donald Tsang said on Friday that the territory had the \"infrastructural hardware\" to make a success of its future under Chinese sovereignty from mid-1997. \"We have the largest and most efficient port on the South China coast ; we have the best transport and telecommunications infrastructure in the world ; and we are investing in this hardware on an enormous scale,\" Tsang said in a speech to Auckland during a visit to New Zealand. Hong Kong also had the necessary \"constitutional infrastructure\" in place, with the promise of autonomy in running its affairs after the handover from Britain to China. \"What this means in practice is that Hong Kong will go on raising its own taxes, issuing its own currency, setting its own expenditure priorities and managing its own enormous financial reserves,\" Tsang said. He acknowledged that many Hong Kong people had decided to seek their future elsewhere and others were sure to follow in the next nine months. \"But for the great majority of us, Hong Kong is our home and Hong Kong's future is our future.\"-- Wellington newsroom 64 4 473-4746\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"PRESS DIGEST - Indonesian newspapers - August 30. Following is a summary of major Indonesian political and business stories in leading newspapers, prepared by Reuters in Jakarta. Reuters has not checked the stories and does not guarantee their accuracy. Telephone : (6221) 384-6364. Fax : (6221) 344-8404. - - - - KOMPAS Indonesian President Suharto has asked businessmen to share their experiences with each other in an effort to boost the country\\\\u0027s exports. - - - - JAKARTA POST Speaker of the House of Representatives Wahono has called on those serving in high state institutions to direct their efforts in the coming years towards dismantling all barriers to social justice. An agreement to bring peace to the southern Philippines is set to be initialed on Friday after delegates from the Philippine government and the Moro National Liberation Front (MNLF) concluded negotiations on the treaty which is set to end almost 25 years of conflict in the region. - - - - MEDIA INDONESIA Around 2,000 of Indonesia\\\\u0027s controversial Timor national car made by Kia Motor Corp of South Korea arrived at Jakarta\\\\u0027s Tanjung Priok port on Thursday. The cars will be jointly marketed by Kia and PT Timor Putra Nasional, controlled by a son of President Suharto, which plans next year to start assembling the vehicles in Indonesia. - - - - REPUBLIKA The Central Jakarta District Court has started to hear the suit filed by ousted Indonesian Democratic Party (PDI) leader Megawati Sukarnoputri against the government and party rivals after the parties failed to reach an out-of-court settlement. Megawati has sued the defendants over a government-backed rebel congress which ousted her last June.\"}'\ntext: PRESS DIGEST - Indonesian newspapers - August 30. Following is a summary of major Indonesian political and business stories in leading newspapers, prepared by Reuters in Jakarta. Reuters has not checked the stories and does not guarantee their accuracy. Telephone : (6221) 384-6364. Fax : (6221) 344-8404. - - - - KOMPAS Indonesian President Suharto has asked businessmen to share their experiences with each other in an effort to boost the country's exports. - - - - JAKARTA POST Speaker of the House of Representatives Wahono has called on those serving in high state institutions to direct their efforts in the coming years towards dismantling all barriers to social justice. An agreement to bring peace to the southern Philippines is set to be initialed on Friday after delegates from the Philippine government and the Moro National Liberation Front (MNLF) concluded negotiations on the treaty which is set to end almost 25 years of conflict in the region. - - - - MEDIA INDONESIA Around 2,000 of Indonesia's controversial Timor national car made by Kia Motor Corp of South Korea arrived at Jakarta's Tanjung Priok port on Thursday. The cars will be jointly marketed by Kia and PT Timor Putra Nasional, controlled by a son of President Suharto, which plans next year to start assembling the vehicles in Indonesia. - - - - REPUBLIKA The Central Jakarta District Court has started to hear the suit filed by ousted Indonesian Democratic Party (PDI) leader Megawati Sukarnoputri against the government and party rivals after the parties failed to reach an out-of-court settlement. Megawati has sued the defendants over a government-backed rebel congress which ousted her last June.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"Jeans Mate Corp - 6mth parent forecast. TOKYO 1996-08-30 Six months to August 20, 1996 (in billions of yen unless specified) LATEST PREVIOUS ACTUAL (Parent) FORECAST FORECAST YEAR-AGO Sales 9.06 9.31 8.42 Current 818 million 979 million 882 million Net 415 million 490 million 412 million NOTE - Jeans Mate Corp is the full company name.\"}'\ntext: Jeans Mate Corp - 6mth parent forecast. TOKYO 1996-08-30 Six months to August 20, 1996 (in billions of yen unless specified) LATEST PREVIOUS ACTUAL (Parent) FORECAST FORECAST YEAR-AGO Sales 9.06 9.31 8.42 Current 818 million 979 million 882 million Net 415 million 490 million 412 million NOTE - Jeans Mate Corp is the full company name.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"Apic Yamada - 6mth parent forecast. TOKYO 1996-08-30 Six months to September 30, 1996 (in billions of yen unless specified) LATEST PREVIOUS ACTUAL (Parent) FORECAST FORECAST YEAR-AGO Sales 12.50 13.00 11.27 Current 1.30 1.35 1.09 Net 650 million 680 million 600 million NOTE - Apic Yamada Corp is a leading manufacturer of semiconductor leadframes.\"}'\ntext: Apic Yamada - 6mth parent forecast. TOKYO 1996-08-30 Six months to September 30, 1996 (in billions of yen unless specified) LATEST PREVIOUS ACTUAL (Parent) FORECAST FORECAST YEAR-AGO Sales 12.50 13.00 11.27 Current 1.30 1.35 1.09 Net 650 million 680 million 600 million NOTE - Apic Yamada Corp is a leading manufacturer of semiconductor leadframes.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"Bootleg brew kills 35 in China, police nab suspects. BEIJING 1996-08-30 Police in southwest China have arrested 30 people suspected of making and selling homemade alcohol that killed 35 people and poisoned 157, the Xinhua news agency said on Friday. A group of farmers in Huize county in the southwestern province of Yunnan were arrested for blending alcohol with methanol and selling the toxic liquor to local residents, the agency said. Between late June and July, a total of 192 people were poisoned by the toxic liquor, and 35 of them died and six were left severely handicapped, it said. Local authorities launched an investigation after they received reports of several similar deaths in the area, it said. Post-mortem examinations showed they were all caused by methanol poisoning. Police had confiscated the remainder of the poisonous liquor, Xinhua said. It gave no further details.\"}'\ntext: Bootleg brew kills 35 in China, police nab suspects. BEIJING 1996-08-30 Police in southwest China have arrested 30 people suspected of making and selling homemade alcohol that killed 35 people and poisoned 157, the Xinhua news agency said on Friday. A group of farmers in Huize county in the southwestern province of Yunnan were arrested for blending alcohol with methanol and selling the toxic liquor to local residents, the agency said. Between late June and July, a total of 192 people were poisoned by the toxic liquor, and 35 of them died and six were left severely handicapped, it said. Local authorities launched an investigation after they received reports of several similar deaths in the area, it said. Post-mortem examinations showed they were all caused by methanol poisoning. Police had confiscated the remainder of the poisonous liquor, Xinhua said. It gave no further details.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"Singapore hangs Thai drug trafficker. SINGAPORE 1996-08-30 Singapore hanged a Thai farmer at Changi Prison on Friday for drug trafficking, the Central Narcotics Bureau (CNB) said. Jeerasak Densakul, 24, was arrested in 1995 when he was found with 11 slabs of cannabis weighing 2.2 kg (4.8 pounds), the CNB said. Singapore has a mandatory death sentence for anyone over 18 years of age found guilty of trafficking in more than 15 grams (half an ounce) of heroin, 30 grams (an ounce) of morphine or 500 grams (18 oz) of cannabis or marijuana. Of the nearly 270 people hanged for various crimes in Singapore since 1975, almost half have been for drug-related charges.\"}'\ntext: Singapore hangs Thai drug trafficker. SINGAPORE 1996-08-30 Singapore hanged a Thai farmer at Changi Prison on Friday for drug trafficking, the Central Narcotics Bureau (CNB) said. Jeerasak Densakul, 24, was arrested in 1995 when he was found with 11 slabs of cannabis weighing 2.2 kg (4.8 pounds), the CNB said. Singapore has a mandatory death sentence for anyone over 18 years of age found guilty of trafficking in more than 15 grams (half an ounce) of heroin, 30 grams (an ounce) of morphine or 500 grams (18 oz) of cannabis or marijuana. Of the nearly 270 people hanged for various crimes in Singapore since 1975, almost half have been for drug-related charges.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"Arafat goes to Nablus ahead of cabinet meeting. NABLUS, West Bank 1996-08-30 Palestinian President Yasser Arafat arrived in the West Bank self-rule enclave of Nablus from Ramallah on Friday, witnesses said. His aides said Arafat would hold the weekly meeting of the Palestinian self-rule Authority\\\\u0027s cabinet in Nablus on Saturday. In Jerusalem, Israeli security forces were bracing for thousands of Palestinians expected to answer Arafat\\\\u0027s call earlier this week to come to the city holy to Moslems, Arabs and Jews to pray in protest against Israel\\\\u0027s settlement policy in the West Bank and delay in peace negotiations. Palestinians want Arab East Jerusalem as the capital of a future independent state. Israel, which captured and annexed East Jerusalem in 1967, says it will never cede any part of the city. Arafat, who made an interim peace deal with Israel in 1993 that set up self-rule, says he will only visit Jerusalem once Israeli occupation has ended.\"}'\ntext: Arafat goes to Nablus ahead of cabinet meeting. NABLUS, West Bank 1996-08-30 Palestinian President Yasser Arafat arrived in the West Bank self-rule enclave of Nablus from Ramallah on Friday, witnesses said. His aides said Arafat would hold the weekly meeting of the Palestinian self-rule Authority's cabinet in Nablus on Saturday. In Jerusalem, Israeli security forces were bracing for thousands of Palestinians expected to answer Arafat's call earlier this week to come to the city holy to Moslems, Arabs and Jews to pray in protest against Israel's settlement policy in the West Bank and delay in peace negotiations. Palestinians want Arab East Jerusalem as the capital of a future independent state. Israel, which captured and annexed East Jerusalem in 1967, says it will never cede any part of the city. Arafat, who made an interim peace deal with Israel in 1993 that set up self-rule, says he will only visit Jerusalem once Israeli occupation has ended.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"U.N. Council concerned about Israeli bulldozers. UNITED NATIONS 1996-08-29 Security Council members expressed concern on Thursday that Israel\\\\u0027s bulldozing of a Palestinian day-care centre for the disabled might further injure the Middle East peace process. Responding to a letter from the the Palestinian U.N. observer mission, Security Council President Tono Eitel of Germany said that members asked him to convey their views to Israel\\\\u0027s charge d\\\\u0027affaires, David Peleg. \\\\\"The members expressed their concern about the maintenance of the peace process and they urged that no action be taken that would have a negative impact on the negotiations,\\\\\" Eitel said after an informal council session. \\\\\"They asked me to call in the Israeli charge d\\\\u0027affaires and discuss the matter with him,\\\\\" he added. The Palestinian letter from Marwan Jilani said the destruction of the Jerusalem centre was an effort by Israel to \\\\\"alter the character, demographic composition and status of the Holy City of Jerusalem\\\\\" and violated agreements between Israel and the Palestinian Liberation Organisation. \\\\\"This most recent measure represents a revival of old, malicious plans to confiscate the land and build units for Israeli settlers within the walls of the Old City.\\\\\" \\\\\"We expect the international community to take a clear and firm position, based on international law and in accordance with U.N. resolutions, against all such Israeli violations and illegal practices,\\\\\" he said. On Tuesday, Israeli crews hoisted a bulldozer over the walls of Jerusalem\\\\u0027s old city and demolished the centre, saying it was being restored without a building permit. Canada had recently donated $ 30 million to the centre, called the Burj al-Laqlaq Society.\"}'\ntext: U.N. Council concerned about Israeli bulldozers. UNITED NATIONS 1996-08-29 Security Council members expressed concern on Thursday that Israel's bulldozing of a Palestinian day-care centre for the disabled might further injure the Middle East peace process. Responding to a letter from the the Palestinian U.N. observer mission, Security Council President Tono Eitel of Germany said that members asked him to convey their views to Israel's charge d'affaires, David Peleg. \"The members expressed their concern about the maintenance of the peace process and they urged that no action be taken that would have a negative impact on the negotiations,\" Eitel said after an informal council session. \"They asked me to call in the Israeli charge d'affaires and discuss the matter with him,\" he added. The Palestinian letter from Marwan Jilani said the destruction of the Jerusalem centre was an effort by Israel to \"alter the character, demographic composition and status of the Holy City of Jerusalem\" and violated agreements between Israel and the Palestinian Liberation Organisation. \"This most recent measure represents a revival of old, malicious plans to confiscate the land and build units for Israeli settlers within the walls of the Old City.\" \"We expect the international community to take a clear and firm position, based on international law and in accordance with U.N. resolutions, against all such Israeli violations and illegal practices,\" he said. On Tuesday, Israeli crews hoisted a bulldozer over the walls of Jerusalem's old city and demolished the centre, saying it was being restored without a building permit. Canada had recently donated $ 30 million to the centre, called the Burj al-Laqlaq Society.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - REAL SCRAPE 1-1 DRAW IN SCRAPPY OPENING MATCH. LA CORUNA, Spain 1996-08-31 A late goal by newly-signed defender Roberto Carlos saved the blushes of Real Madrid coach Fabio Capello and his multi-billion peseta line-up in the opening game of the Spanish championship on Saturday. The Brazilian\\\\u0027s 79th-minute effort was enough to earn Real a point from a scrappy 1-1 draw at fellow title contenders Deportivo Coruna. Deportivo started strongly, taking the lead midway through the first half when former Auxerre playmaker Corentine Martins headed home a corner after a flick-on by Brazilian-born Spanish international midfielder Donato. Real looked to be in deep trouble shortly after the break when Luis Milla was sent off for committing two bookable offences in as many minutes. But Deportivo were unable to capitalise on their numerical advantage, and were themselves reduced to ten men when Armando Alvarez was sent off 15 minutes from time. Shortly afterwards Roberto Carlos found space in the home defence and equalised for Real with a shot that was deflected past despairing Deportivo \\\\u0027 keeper Jacques Songo\\\\u0027o. In a frantic final five minutes there were chances at both ends, and Donato, who had earlier been booked, was sent off for protesting about the incursion of Real players at a free kick. Before the match Deportivo chairman Augusto Lendoiro said he would ignore a FIFA decision banning Brazilian midfielder Mauro Silva from playing in the match for failing to join his national side\\\\u0027s tour of Europe. In the event, coach John Toshack decided not to use Silva, who had claimed he did not join the Brazil squad because he had lost his passport.\"}'\ntext: SOCCER - REAL SCRAPE 1-1 DRAW IN SCRAPPY OPENING MATCH. LA CORUNA, Spain 1996-08-31 A late goal by newly-signed defender Roberto Carlos saved the blushes of Real Madrid coach Fabio Capello and his multi-billion peseta line-up in the opening game of the Spanish championship on Saturday. The Brazilian's 79th-minute effort was enough to earn Real a point from a scrappy 1-1 draw at fellow title contenders Deportivo Coruna. Deportivo started strongly, taking the lead midway through the first half when former Auxerre playmaker Corentine Martins headed home a corner after a flick-on by Brazilian-born Spanish international midfielder Donato. Real looked to be in deep trouble shortly after the break when Luis Milla was sent off for committing two bookable offences in as many minutes. But Deportivo were unable to capitalise on their numerical advantage, and were themselves reduced to ten men when Armando Alvarez was sent off 15 minutes from time. Shortly afterwards Roberto Carlos found space in the home defence and equalised for Real with a shot that was deflected past despairing Deportivo ' keeper Jacques Songo'o. In a frantic final five minutes there were chances at both ends, and Donato, who had earlier been booked, was sent off for protesting about the incursion of Real players at a free kick. Before the match Deportivo chairman Augusto Lendoiro said he would ignore a FIFA decision banning Brazilian midfielder Mauro Silva from playing in the match for failing to join his national side's tour of Europe. In the event, coach John Toshack decided not to use Silva, who had claimed he did not join the Brazil squad because he had lost his passport.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"RUGBY LEAGUE - WIGAN BEAT BRADFORD 42-36 IN SEMIFINAL. WIGAN, England 1996-08-31 Result of English rugby league premiership semifinal played on Saturday : Wigan 42 Bradford Bulls 36\"}'\ntext: RUGBY LEAGUE - WIGAN BEAT BRADFORD 42-36 IN SEMIFINAL. WIGAN, England 1996-08-31 Result of English rugby league premiership semifinal played on Saturday : Wigan 42 Bradford Bulls 36\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - ISRAEL BEAT BULGARIA IN EUROPEAN UNDER-21 QUALIFIER. HERZLIYA, Israel 1996-08-31 Result of European under-21 championship group 5 qualifier on Saturday : Israel 2, Bulgaria 0 (halftime 0-0) Scorers : Haim Hajaj (47th), Nir Sivilia (57th). Attendance : 2,000.\"}'\ntext: SOCCER - ISRAEL BEAT BULGARIA IN EUROPEAN UNDER-21 QUALIFIER. HERZLIYA, Israel 1996-08-31 Result of European under-21 championship group 5 qualifier on Saturday : Israel 2, Bulgaria 0 (halftime 0-0) Scorers : Haim Hajaj (47th), Nir Sivilia (57th). Attendance : 2,000.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - IRISH ERASE PAINFUL MEMORIES WITH 5-0 WIN. ESCHEN, Liechtenstein 1996-08-31 The Republic of Ireland\\\\u0027s new-look side dispelled painful memories of their last visit to Liechtenstein by beating the Alpine part-timers 5-0 in a World Cup qualifier on Saturday. The Irish, under new manager Mick McCarthy, took a 4-0 lead within 20 minutes through captain Andy Townsend, 20-year-old Norwich striker Keith O\\\\u0027Neill, Sunderland forward Niall Quinn and teenager Ian Harte. Quinn added his second and Ireland\\\\u0027s fifth just after the hour to complete the rout and give the Irish their biggest-ever away win. The result helped erase memories of Ireland\\\\u0027s visit to the Eschen stadium 14 months ago, when Jack Charlton\\\\u0027s side were held to a frustrating 0-0 draw which ultimately cost them a place in the European championship finals.\"}'\ntext: SOCCER - IRISH ERASE PAINFUL MEMORIES WITH 5-0 WIN. ESCHEN, Liechtenstein 1996-08-31 The Republic of Ireland's new-look side dispelled painful memories of their last visit to Liechtenstein by beating the Alpine part-timers 5-0 in a World Cup qualifier on Saturday. The Irish, under new manager Mick McCarthy, took a 4-0 lead within 20 minutes through captain Andy Townsend, 20-year-old Norwich striker Keith O'Neill, Sunderland forward Niall Quinn and teenager Ian Harte. Quinn added his second and Ireland's fifth just after the hour to complete the rout and give the Irish their biggest-ever away win. The result helped erase memories of Ireland's visit to the Eschen stadium 14 months ago, when Jack Charlton's side were held to a frustrating 0-0 draw which ultimately cost them a place in the European championship finals.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - IRELAND BEAT LIECHTENSTEIN 5-0 IN WORLD CUP QUALIFIER. ESCHEN 1996-08-31 The Republic of Ireland beat Liechtenstein 5-0 (halftime 4-0) in a World Cup soccer European group 8 qualifier on Saturday. Scorers : Andy Townsend (5th), Keith O\\\\u0027Neill (7th), Niall Quinn (11th, 61st), Ian Harte (19th). Attendance : 3,900\"}'\ntext: SOCCER - IRELAND BEAT LIECHTENSTEIN 5-0 IN WORLD CUP QUALIFIER. ESCHEN 1996-08-31 The Republic of Ireland beat Liechtenstein 5-0 (halftime 4-0) in a World Cup soccer European group 8 qualifier on Saturday. Scorers : Andy Townsend (5th), Keith O'Neill (7th), Niall Quinn (11th, 61st), Ian Harte (19th). Attendance : 3,900\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"GOLF - BRITISH MASTERS FINAL SCORES. NORTHAMPTON, England 1996-08-31 Leading scores after the final round of the British Masters golf tournament on Saturday (British unless stated) : 284 Robert Allenby (Australia) 69 71 71 73, Miguel Angel Martin (Spain) 75 70 71 68 (Allenby won at first play-off hole) 285 Costantino Rocca (Italy) 71 73 72 69 286 Miguel Angel Jimenez (Spain) 74 72 73 67 287 Ian Woosnam 70 76 71 70 288 Jose Coceres (Argentina) 69 78 71 70 289 Joakim Haeggman (Sweden) 71 77 70 71, Antoine Lebouc (France) 74 73 70 72 290 Colin Montgomerie 68 76 77 69, Robert Coles 74 76 71 69, Philip Walton (Ireland) 71 74 74 71, Peter Mitchell 74 71 74 71, Klas Eriksson (Sweden) 71 75 72 72, Pedro Linhart (Spain) 72 73 67 78 291 Phillip Price 72 76 74 69, Adam Hunter 70 79 73 69, Peter O\\\\u0027Malley (Australia) 71 73 75 72, Mark Roe 69 71 78 73, Mike Clayton (Australia) 69 76 73 73 292 Iain Pyman 71 75 75 71, David Gilford 69 74 77 72, Peter Hedblom (Sweden) 70 75 75 72, Stephen McAllister 73 76 69 74.\"}'\ntext: GOLF - BRITISH MASTERS FINAL SCORES. NORTHAMPTON, England 1996-08-31 Leading scores after the final round of the British Masters golf tournament on Saturday (British unless stated) : 284 Robert Allenby (Australia) 69 71 71 73, Miguel Angel Martin (Spain) 75 70 71 68 (Allenby won at first play-off hole) 285 Costantino Rocca (Italy) 71 73 72 69 286 Miguel Angel Jimenez (Spain) 74 72 73 67 287 Ian Woosnam 70 76 71 70 288 Jose Coceres (Argentina) 69 78 71 70 289 Joakim Haeggman (Sweden) 71 77 70 71, Antoine Lebouc (France) 74 73 70 72 290 Colin Montgomerie 68 76 77 69, Robert Coles 74 76 71 69, Philip Walton (Ireland) 71 74 74 71, Peter Mitchell 74 71 74 71, Klas Eriksson (Sweden) 71 75 72 72, Pedro Linhart (Spain) 72 73 67 78 291 Phillip Price 72 76 74 69, Adam Hunter 70 79 73 69, Peter O'Malley (Australia) 71 73 75 72, Mark Roe 69 71 78 73, Mike Clayton (Australia) 69 76 73 73 292 Iain Pyman 71 75 75 71, David Gilford 69 74 77 72, Peter Hedblom (Sweden) 70 75 75 72, Stephen McAllister 73 76 69 74.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - SLOVAKIA BEAT FAROES IN WORLD CUP QUALIFIER. TOFTIR, Faroe Islands 1996-08-31 Slovakia beat the Faroe Islands 2-1 (halftime 1-0) in their World Cup soccer European group six qualifying match on Saturday. Scorers : Faroe Islands - Jan Allan Mueller (60th minute) Slovakia - Lubomir Moravcik (13th), Peter Dubovsky (88th) Attendance : 1,445.\"}'\ntext: SOCCER - SLOVAKIA BEAT FAROES IN WORLD CUP QUALIFIER. TOFTIR, Faroe Islands 1996-08-31 Slovakia beat the Faroe Islands 2-1 (halftime 1-0) in their World Cup soccer European group six qualifying match on Saturday. Scorers : Faroe Islands - Jan Allan Mueller (60th minute) Slovakia - Lubomir Moravcik (13th), Peter Dubovsky (88th) Attendance : 1,445.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"CRICKET - ENGLAND BEAT PAKISTAN BY 107 RUNS IN SECOND ONE-DAYER. BIRMINGHAM, England 1996-08-31 England beat Pakistan by 107 runs in the second one-day international at Edgbaston on Saturday to take the series 2-0. Scores : England 292-8 innings closed (N. Knight 113), Pakistan 185 (Ijaz Ahmed 79 ; A. Hollioake 4-23)\"}'\ntext: CRICKET - ENGLAND BEAT PAKISTAN BY 107 RUNS IN SECOND ONE-DAYER. BIRMINGHAM, England 1996-08-31 England beat Pakistan by 107 runs in the second one-day international at Edgbaston on Saturday to take the series 2-0. Scores : England 292-8 innings closed (N. Knight 113), Pakistan 185 (Ijaz Ahmed 79 ; A. Hollioake 4-23)\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"CYCLING - TOUR OF NETHERLANDS FINAL RESULTS / STANDINGS. LANDGRAAF, Netherlands 1996-08-31 Leading results of the 205-km sixth and final stage of the Tour of the Netherlands between Roermond and Landgraaf on Saturday : 1. Olaf Ludwig (Germany) Telekom 4 hours 48 mins 2 seconds 2. Giovanni Lombardi (Italy) Polti 5 seconds behind 3. Tristan Hoffman (Netherlands) TVM same time 4. Erik Breukink (Netherlands) Rabobank 8 seconds 5. Jesper Skibby (Denmark) TVM 9 6. Vyacheslav Ekimov (Russia) Rabobank same time 7. Luca Pavanello (Italy) Aki 11 8. Eleuterio Anguita (Spain) MX Onda 9. Michael Andersson (Sweden) Telekom 10. Johan Capiot (Belgium) Collstrop all same time Final overall placings (after six stages) : 1. Rolf Sorensen (Denmark) Rabobank 20:36:54 2. Lance Armstrong (U.S.) Motorola 2 seconds behind 3. Ekimov 1:7 4. Marco Lietti (Italy) MG-Technogym 1:16 5. Erik Dekker (Netherlands) Rabobank 1:23 6. Ludwig 1:25 6. Breukink same time 8. Maarten den Bakker (Netherlands) TVM 1:33 9. Andersson 1:34 10. Skibby 1:45\"}'\ntext: CYCLING - TOUR OF NETHERLANDS FINAL RESULTS / STANDINGS. LANDGRAAF, Netherlands 1996-08-31 Leading results of the 205-km sixth and final stage of the Tour of the Netherlands between Roermond and Landgraaf on Saturday : 1. Olaf Ludwig (Germany) Telekom 4 hours 48 mins 2 seconds 2. Giovanni Lombardi (Italy) Polti 5 seconds behind 3. Tristan Hoffman (Netherlands) TVM same time 4. Erik Breukink (Netherlands) Rabobank 8 seconds 5. Jesper Skibby (Denmark) TVM 9 6. Vyacheslav Ekimov (Russia) Rabobank same time 7. Luca Pavanello (Italy) Aki 11 8. Eleuterio Anguita (Spain) MX Onda 9. Michael Andersson (Sweden) Telekom 10. Johan Capiot (Belgium) Collstrop all same time Final overall placings (after six stages) : 1. Rolf Sorensen (Denmark) Rabobank 20:36:54 2. Lance Armstrong (U.S.) Motorola 2 seconds behind 3. Ekimov 1:7 4. Marco Lietti (Italy) MG-Technogym 1:16 5. Erik Dekker (Netherlands) Rabobank 1:23 6. Ludwig 1:25 6. Breukink same time 8. Maarten den Bakker (Netherlands) TVM 1:33 9. Andersson 1:34 10. Skibby 1:45\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"CRICKET - ENGLAND V PAKISTAN ONE-DAY SCOREBOARD. BIRMINGHAM, England 1996-08-31 Scoreboard of the second one-day cricket match between England and Pakistan on Saturday : England N. Knight st Moin Khan b Saqlain Mushtaq 113 A. Stewart b Mushtaq Ahmed 46 M. Atherton lbw b Mushtaq Ahmed 1 G. Thorpe lbw b Ata-ur-Rehman 21 M. Maynard run out 1 R. Irani not out 45 A. Hollioake run out 15 D. Gough run out 0 R. Croft b Waqar Younis 15 D. Headley not out 3 Extras (lb-25 w-4 nb-3) 32 Total (for 8 wickets, innings closed) 292 Fall : 1-103 2-105 3-163 4-168 5-221 6-257 7-257 8-286. Did Not Bat : A. Mullally. Bowling : Wasim Akram 10-0-50-0, Waqar Younis 9-0-54-1, Ata-ur-Rehman 6-0-40-1, Saqlain Mushtaq 10-0-59-1, Mushtaq Ahmed 10-0-33-2, Aamir Sohail 5-0-31-0. pakistan Saeed Anwar c Stewart b Gough 33 Aamir Sohail c Croft b Gough 0 Moin Khan lbw b Mullally 0 Ijaz Ahmed b Croft 79 Inzamam-ul-Haq c Thorpe b Croft 6 Salim Malik c Stewart b Hollioake 23 Wasim Akram c Knight b Hollioake 21 Mushtaq Ahmed not out 14 Saqlain Mushtaq b Hollioake 0 Waqar Younis lbw b Gough 4 Ata-ur-Rehman c Knight b Hollioake 2 Extras (lb-2 nb-1) 3 Total (37.5 overs) 185 Fall of wickets : 1-1 2-6 3-54 4-104 5-137 6-164 7-164 8-168 9-177. Bowling : Gough 8-0-39-3, Mullally 6-0-30-1, Headley 7-0-32-0, Irani 2-0-22-0, Croft 8-0-37-2, Hollioake 6.5-1-23-4.\"}'\ntext: CRICKET - ENGLAND V PAKISTAN ONE-DAY SCOREBOARD. BIRMINGHAM, England 1996-08-31 Scoreboard of the second one-day cricket match between England and Pakistan on Saturday : England N. Knight st Moin Khan b Saqlain Mushtaq 113 A. Stewart b Mushtaq Ahmed 46 M. Atherton lbw b Mushtaq Ahmed 1 G. Thorpe lbw b Ata-ur-Rehman 21 M. Maynard run out 1 R. Irani not out 45 A. Hollioake run out 15 D. Gough run out 0 R. Croft b Waqar Younis 15 D. Headley not out 3 Extras (lb-25 w-4 nb-3) 32 Total (for 8 wickets, innings closed) 292 Fall : 1-103 2-105 3-163 4-168 5-221 6-257 7-257 8-286. Did Not Bat : A. Mullally. Bowling : Wasim Akram 10-0-50-0, Waqar Younis 9-0-54-1, Ata-ur-Rehman 6-0-40-1, Saqlain Mushtaq 10-0-59-1, Mushtaq Ahmed 10-0-33-2, Aamir Sohail 5-0-31-0. pakistan Saeed Anwar c Stewart b Gough 33 Aamir Sohail c Croft b Gough 0 Moin Khan lbw b Mullally 0 Ijaz Ahmed b Croft 79 Inzamam-ul-Haq c Thorpe b Croft 6 Salim Malik c Stewart b Hollioake 23 Wasim Akram c Knight b Hollioake 21 Mushtaq Ahmed not out 14 Saqlain Mushtaq b Hollioake 0 Waqar Younis lbw b Gough 4 Ata-ur-Rehman c Knight b Hollioake 2 Extras (lb-2 nb-1) 3 Total (37.5 overs) 185 Fall of wickets : 1-1 2-6 3-54 4-104 5-137 6-164 7-164 8-168 9-177. Bowling : Gough 8-0-39-3, Mullally 6-0-30-1, Headley 7-0-32-0, Irani 2-0-22-0, Croft 8-0-37-2, Hollioake 6.5-1-23-4.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"CYCLING - WORLD TRACK CHAMPIONSHIP RESULTS. MANCHESTER, England 1996-08-31 Results at the world track cycling championships on Saturday : Women\\\\u0027s 3,000 metres individual pursuit qualifying round (fastest eight to quarter finals) : 1. Antonella Bellutti (Italy) 3:31.526 (world record) 2. Marion Clignet (France) 3:31.674 3. Lucy Tyler-Sharman (Australia) 3:31.830 4. Yvonne McGregor (Britain) 3:41.823 5. Natalia Karimova (Russia) 3:45.061 6. Svetlana Samokhalova (Russia) 3:46.216 7 Jane Quigley (U.S.) 3:46.493 8. Rasa Mazeikyte (Lithuania) 3:46.834 9. Tatian Stiajkina (Ukraine) 3:52.204 World 4,000 metres team pursuit semifinals : Italy (Adler Capelli, Cristiano Citto, Andrea Collinelli, Mauro Trentino) 4:00.958 (world record) beat Russia (Anton Chantyr, Edouard Gritsoun, Nikolai Kouznetsov) 4:06.534. France (Cyril Bos, Philippe Ermenault, Jean-Michel Monin, Francis Moreau) 4:05.104 beat Germany (Guido Fulst, Danilo Hondo, Thorsten Rund, Heiko Szonn) 4:05.463 Germany take the bronze medal as fastest losing semifinalist. Women\\\\u0027s world 500 metres time trial final : 1. Felicia Ballanger (France) 34.829 2. Annett Neumann (Germany) 35.202 3. Michelle Ferris (Australia) 35.694 4. Magali Faure (France) 35.888 5. Olga Slioussareva (Russia) 36.170 6. Oksana Grichina (Russia) 36.242 7. Tanya Dubnicoff (Canada) 36.307 8. Kathrin Freitag (Germany) 36.491 9. Donna Wynd (New Zealand) 36.831 10. Mira Kasslin (Finland) 37.273 11. Wendy Everson (Britain) 37.624 12. Giovanna Troldi (Italy) 38.285 13. Rita Razmaite (Lithuania) 38.546 World 4,000 metres team pursuit championship final : Italy (Adler Capelli, Cristiano Citto, Andrea Collinelli, Mauro Trentino) 4:02.752 beat France (Cyril Bos, Philippe Ermenault, Jean-Michel Monin, Francis Moreau) 4:04.539 World sprint championship quarter finals (best of three matches) Florian Rousseau (France) beat Ainars Kiksis (Latvia) 2-0 Darryn Hill (Australia) beat Christian Arrue (U.S.) 2-0 Roberto Chiappa (Italy) beat Frederic Magne (France) 2-0 Marty Nothstein (U.S.) beat Pavel Buran (Czech Republic) 2-0 Women\\\\u0027s world 3,000 metres individual pursuit championship quarter-finals : Marion Clignet (France) 3:30.974 (World Record) beat Jane Quigley (USA) 3:42.852 Natalia Karimova (Russia) 3:40.036 beat Yvonne McGregor (Britain) 3:43.078 Lucy Tyler-Sharman (Australia) 3:35.087 beat Svetlana Samokhvalova (Russia) 3:45.011 Antonella Bellutti (Italy) 3:32.174 caught and eliminated Rasa Mazeikyte (Lithuania)\"}'\ntext: CYCLING - WORLD TRACK CHAMPIONSHIP RESULTS. MANCHESTER, England 1996-08-31 Results at the world track cycling championships on Saturday : Women's 3,000 metres individual pursuit qualifying round (fastest eight to quarter finals) : 1. Antonella Bellutti (Italy) 3:31.526 (world record) 2. Marion Clignet (France) 3:31.674 3. Lucy Tyler-Sharman (Australia) 3:31.830 4. Yvonne McGregor (Britain) 3:41.823 5. Natalia Karimova (Russia) 3:45.061 6. Svetlana Samokhalova (Russia) 3:46.216 7 Jane Quigley (U.S.) 3:46.493 8. Rasa Mazeikyte (Lithuania) 3:46.834 9. Tatian Stiajkina (Ukraine) 3:52.204 World 4,000 metres team pursuit semifinals : Italy (Adler Capelli, Cristiano Citto, Andrea Collinelli, Mauro Trentino) 4:00.958 (world record) beat Russia (Anton Chantyr, Edouard Gritsoun, Nikolai Kouznetsov) 4:06.534. France (Cyril Bos, Philippe Ermenault, Jean-Michel Monin, Francis Moreau) 4:05.104 beat Germany (Guido Fulst, Danilo Hondo, Thorsten Rund, Heiko Szonn) 4:05.463 Germany take the bronze medal as fastest losing semifinalist. Women's world 500 metres time trial final : 1. Felicia Ballanger (France) 34.829 2. Annett Neumann (Germany) 35.202 3. Michelle Ferris (Australia) 35.694 4. Magali Faure (France) 35.888 5. Olga Slioussareva (Russia) 36.170 6. Oksana Grichina (Russia) 36.242 7. Tanya Dubnicoff (Canada) 36.307 8. Kathrin Freitag (Germany) 36.491 9. Donna Wynd (New Zealand) 36.831 10. Mira Kasslin (Finland) 37.273 11. Wendy Everson (Britain) 37.624 12. Giovanna Troldi (Italy) 38.285 13. Rita Razmaite (Lithuania) 38.546 World 4,000 metres team pursuit championship final : Italy (Adler Capelli, Cristiano Citto, Andrea Collinelli, Mauro Trentino) 4:02.752 beat France (Cyril Bos, Philippe Ermenault, Jean-Michel Monin, Francis Moreau) 4:04.539 World sprint championship quarter finals (best of three matches) Florian Rousseau (France) beat Ainars Kiksis (Latvia) 2-0 Darryn Hill (Australia) beat Christian Arrue (U.S.) 2-0 Roberto Chiappa (Italy) beat Frederic Magne (France) 2-0 Marty Nothstein (U.S.) beat Pavel Buran (Czech Republic) 2-0 Women's world 3,000 metres individual pursuit championship quarter-finals : Marion Clignet (France) 3:30.974 (World Record) beat Jane Quigley (USA) 3:42.852 Natalia Karimova (Russia) 3:40.036 beat Yvonne McGregor (Britain) 3:43.078 Lucy Tyler-Sharman (Australia) 3:35.087 beat Svetlana Samokhvalova (Russia) 3:45.011 Antonella Bellutti (Italy) 3:32.174 caught and eliminated Rasa Mazeikyte (Lithuania)\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"CRICKET - PAKISTAN WIN TOSS, PUT ENGLAND IN TO BAT. BIRMINGHAM, England 1996-08-31 Pakistan won the toss and put England in to bat in the second limited overs cricket international at Edgbaston on Saturday. Surrey all-rounder Adam Hollioake was making his England debut, replacing Lancashire batsman Graham Lloyd, with seamer Peter Martin again being omitted from the 13. Pakistan kept the side who lost to England by five wickets at Old Trafford on Thursday in the first of the three one-day matches. Teams : England : Mike Atherton (captain), Nick Knight, Alec Stewart, Graham Thorpe, Matthew Maynard, Adam Hollioake, Ronnie Irani, Robert Croft, Darren Gough, Dean Headley, Alan Mullally. Pakistan : Aamir Sohail, Saeed Anwar, Ijaz Ahmed, Salim Malik, Inzamam-ul-Haq, Wasim Akram (captain), Moin Khan, Saqlain Mushtaq, Mushtaq Ahmed, Waqar Younis, Ata-ur-Rehman.\"}'\ntext: CRICKET - PAKISTAN WIN TOSS, PUT ENGLAND IN TO BAT. BIRMINGHAM, England 1996-08-31 Pakistan won the toss and put England in to bat in the second limited overs cricket international at Edgbaston on Saturday. Surrey all-rounder Adam Hollioake was making his England debut, replacing Lancashire batsman Graham Lloyd, with seamer Peter Martin again being omitted from the 13. Pakistan kept the side who lost to England by five wickets at Old Trafford on Thursday in the first of the three one-day matches. Teams : England : Mike Atherton (captain), Nick Knight, Alec Stewart, Graham Thorpe, Matthew Maynard, Adam Hollioake, Ronnie Irani, Robert Croft, Darren Gough, Dean Headley, Alan Mullally. Pakistan : Aamir Sohail, Saeed Anwar, Ijaz Ahmed, Salim Malik, Inzamam-ul-Haq, Wasim Akram (captain), Moin Khan, Saqlain Mushtaq, Mushtaq Ahmed, Waqar Younis, Ata-ur-Rehman.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"BASEBALL - GONZALEZ HOMERS TWICE AS RANGERS BEAT INDIANS. ARLINGTON, Texas 1996-08-31 Juan Gonzalez homered twice and Ivan Rodriguez added a two-run shot as the Texas Rangers defeated the Cleveland Indians 5-3 in a matchup of division leaders Friday. Rodriguez\\\\u0027s 18th homer, off Chad Ogea (7-5) in the first, gave Texas a 2-0 lead. One out later, Gonzalez smacked his 40th homer, extending his hitting streak to 20 games. Gonzalez, who hit in 21 straight games earlier this season, joined Mickey Rivers as the only players in Texas history with two 20-game streaks in the same year. Gonzalez hit his second homer in the third for his fifth multi-homer game of the season. Gonzalez has three 40-homer seasons and his 121 RBI broke Ruben Sierra\\\\u0027s team record of 119 set in 1989. The Indians had their four-game winning streak stopped. \\\\\"It\\\\u0027s not something I\\\\u0027m going to try to explain,\\\\\" said Texas manager Johnny Oates about his team winning seven of the 10 meetings from Cleveland this season. \\\\\"We\\\\u0027ve got two more regular season games against them and we might get lucky enough or unlucky enough to play them in the post-season.\\\\\" Roger Pavlik (15-7) gave up three runs and seven hits in 6 1/3 innings and became the fourth 15-game winner in the American League. Jeff Russell pitched two perfect innings for his third save. Brian Giles and Jim Thome homered for Cleveland. Cleveland\\\\u0027s lead over the White Sox in the American League Central dropped to nine games. Texas\\\\u0027s lead over Seattle in the West increased to six. At California, Tino Martinez\\\\u0027s two-run homer keyed a three-run first and Andy Pettitte became the league\\\\u0027s first 19-game winner as the New York Yankees beat the Angels 6-2. New York snapped a season-high five-game losing streak and also got homers from Mariano Duncan, Darryl Strawberry and Jim Leyritz. Pettite (19-7) allowed two runs and eight hits over eight innings with a walk and seven strikeouts. He improved to 12-2 following Yankees \\\\u0027 losses. Mariano Rivera pitched a scoreless ninth, striking out two. Ex-Yankee Randy Velarde hit his 11th homer, his most at any professional level. In Seattle, Pete Incaviglia\\\\u0027s grand slam with one out in the sixth snapped a tie and lifted the Baltimore Orioles past the Seattle Mariners, 5-2. It was Incaviglia\\\\u0027s sixth grand slam and 200th homer of his career. Baltimore\\\\u0027s Eddie Murray cracked his 20th homer of the season and 499th of his career. Jay Buhner hit his 38th homer and Edgar Martinez his 23rd for Seattle. The Orioles remained tied with the White Sox for the American League wild card with the Mariners a game back. In Toronto, Kevin Tapani (12-8) allowed two runs and six hits over 7 1/3 innings and Frank Thomas hit his 29th homer and drove in three runs as the Chicago White Sox cruised to an 11-2 victory over the Blue Jays. Thomas, Harold Baines and Robin Ventura each collected three hits. Baines homered and scored three runs. Danny Tartabull added two hits and three RBI as all Chicago starters got at least one hit. In Oakland, Dave Telgheder scattered seven hits over eight innings and Mark McGwire hit his major-league leading 45th homer and drove in three runs as the Athetlics blanked the Boston Red Sox 7-0. Telgheder (2-5) snapped a personal three-game losing streak. Buddy Groom pitched a perfect ninth inning. McGwire singled home a run to spark a three-run sixth and capped the scoring with a two-run homer in the seventh. The loss was Boston\\\\u0027s seventh in its last 29 games. In Detroit, Todd Van Poppel pitched a five-hitter for his first career shutout and Tony Clark homered to cap a four-run first inning as the Tigers blanked the Kansas City Royals 4-0. Van Poppel (3-6) walked two and struck out two in defeating the Royals for the second time this week. He threw 108 pitches as he lowered his ERA from 8.08 to 7.24. Kansas City has scored only one run in two games. In Milwaukee, Marc Newfield homered off Jose Parra (5-4) leading off the bottom of the 12th as the Brewers rallied for a 5-4 victory over the Minnesota Twins. Milwaukee has won 10 of its last 15. Bob Wickman (6-1) pitched 2 2/3 hitless innings for the win, his second for the Brewers. Matt Lawton hit a three-run homer off closer Mike Fetters with one out in the ninth to give Minnesota a 4-2 lead. But Milwaukee tied it up in the bottom of the ninth on pinch-hitter Dave Nilsson\\\\u0027s two-run double.\"}'\ntext: BASEBALL - GONZALEZ HOMERS TWICE AS RANGERS BEAT INDIANS. ARLINGTON, Texas 1996-08-31 Juan Gonzalez homered twice and Ivan Rodriguez added a two-run shot as the Texas Rangers defeated the Cleveland Indians 5-3 in a matchup of division leaders Friday. Rodriguez's 18th homer, off Chad Ogea (7-5) in the first, gave Texas a 2-0 lead. One out later, Gonzalez smacked his 40th homer, extending his hitting streak to 20 games. Gonzalez, who hit in 21 straight games earlier this season, joined Mickey Rivers as the only players in Texas history with two 20-game streaks in the same year. Gonzalez hit his second homer in the third for his fifth multi-homer game of the season. Gonzalez has three 40-homer seasons and his 121 RBI broke Ruben Sierra's team record of 119 set in 1989. The Indians had their four-game winning streak stopped. \"It's not something I'm going to try to explain,\" said Texas manager Johnny Oates about his team winning seven of the 10 meetings from Cleveland this season. \"We've got two more regular season games against them and we might get lucky enough or unlucky enough to play them in the post-season.\" Roger Pavlik (15-7) gave up three runs and seven hits in 6 1/3 innings and became the fourth 15-game winner in the American League. Jeff Russell pitched two perfect innings for his third save. Brian Giles and Jim Thome homered for Cleveland. Cleveland's lead over the White Sox in the American League Central dropped to nine games. Texas's lead over Seattle in the West increased to six. At California, Tino Martinez's two-run homer keyed a three-run first and Andy Pettitte became the league's first 19-game winner as the New York Yankees beat the Angels 6-2. New York snapped a season-high five-game losing streak and also got homers from Mariano Duncan, Darryl Strawberry and Jim Leyritz. Pettite (19-7) allowed two runs and eight hits over eight innings with a walk and seven strikeouts. He improved to 12-2 following Yankees ' losses. Mariano Rivera pitched a scoreless ninth, striking out two. Ex-Yankee Randy Velarde hit his 11th homer, his most at any professional level. In Seattle, Pete Incaviglia's grand slam with one out in the sixth snapped a tie and lifted the Baltimore Orioles past the Seattle Mariners, 5-2. It was Incaviglia's sixth grand slam and 200th homer of his career. Baltimore's Eddie Murray cracked his 20th homer of the season and 499th of his career. Jay Buhner hit his 38th homer and Edgar Martinez his 23rd for Seattle. The Orioles remained tied with the White Sox for the American League wild card with the Mariners a game back. In Toronto, Kevin Tapani (12-8) allowed two runs and six hits over 7 1/3 innings and Frank Thomas hit his 29th homer and drove in three runs as the Chicago White Sox cruised to an 11-2 victory over the Blue Jays. Thomas, Harold Baines and Robin Ventura each collected three hits. Baines homered and scored three runs. Danny Tartabull added two hits and three RBI as all Chicago starters got at least one hit. In Oakland, Dave Telgheder scattered seven hits over eight innings and Mark McGwire hit his major-league leading 45th homer and drove in three runs as the Athetlics blanked the Boston Red Sox 7-0. Telgheder (2-5) snapped a personal three-game losing streak. Buddy Groom pitched a perfect ninth inning. McGwire singled home a run to spark a three-run sixth and capped the scoring with a two-run homer in the seventh. The loss was Boston's seventh in its last 29 games. In Detroit, Todd Van Poppel pitched a five-hitter for his first career shutout and Tony Clark homered to cap a four-run first inning as the Tigers blanked the Kansas City Royals 4-0. Van Poppel (3-6) walked two and struck out two in defeating the Royals for the second time this week. He threw 108 pitches as he lowered his ERA from 8.08 to 7.24. Kansas City has scored only one run in two games. In Milwaukee, Marc Newfield homered off Jose Parra (5-4) leading off the bottom of the 12th as the Brewers rallied for a 5-4 victory over the Minnesota Twins. Milwaukee has won 10 of its last 15. Bob Wickman (6-1) pitched 2 2/3 hitless innings for the win, his second for the Brewers. Matt Lawton hit a three-run homer off closer Mike Fetters with one out in the ninth to give Minnesota a 4-2 lead. But Milwaukee tied it up in the bottom of the ninth on pinch-hitter Dave Nilsson's two-run double.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"CRICKET - ESSEX AND KENT MADE TO SWEAT IN TITLE RACE. LONDON 1996-08-31 Essex and Kent both face tense finishes on Monday as they attempt to keep pace with title hopefuls Derbyshire and Surrey, convincing three-day victors on Saturday, in the English county championship run-in. Essex need another 148 with five wickets in hand to beat Yorkshire after a maiden first-class century from Richard Kettleborough transformed a match which his side had seemed certain to lose. Kent will also need to keep their nerve against struggling Nottinghamshire who will enter the final day 137 ahead with four wickets left in a relatively low-scoring match at Tunbridge Wells. Derbyshire, nine-wicket winners over Worcestershire, and Surrey, who thrashed Warwickshire by an innings and 164 runs, can instead take the day off along with rivals Leicestershire, who beat Somerset inside two days. Warwickshire captain Tim Munton is tipping Surrey to emerge on top, impressed by the positive influence of Australian coach Dave Gilbert, but Derbyshire\\\\u0027s Australian captain Dean Jones is conceding nothing as his side chase their first title for 60 years. \\\\\"We took three absolutely brilliant catches in this match and our catching all season has been pretty impressive. Our catching will win or lose us the championship,\\\\\" he said.\"}'\ntext: CRICKET - ESSEX AND KENT MADE TO SWEAT IN TITLE RACE. LONDON 1996-08-31 Essex and Kent both face tense finishes on Monday as they attempt to keep pace with title hopefuls Derbyshire and Surrey, convincing three-day victors on Saturday, in the English county championship run-in. Essex need another 148 with five wickets in hand to beat Yorkshire after a maiden first-class century from Richard Kettleborough transformed a match which his side had seemed certain to lose. Kent will also need to keep their nerve against struggling Nottinghamshire who will enter the final day 137 ahead with four wickets left in a relatively low-scoring match at Tunbridge Wells. Derbyshire, nine-wicket winners over Worcestershire, and Surrey, who thrashed Warwickshire by an innings and 164 runs, can instead take the day off along with rivals Leicestershire, who beat Somerset inside two days. Warwickshire captain Tim Munton is tipping Surrey to emerge on top, impressed by the positive influence of Australian coach Dave Gilbert, but Derbyshire's Australian captain Dean Jones is conceding nothing as his side chase their first title for 60 years. \"We took three absolutely brilliant catches in this match and our catching all season has been pretty impressive. Our catching will win or lose us the championship,\" he said.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"GOLF - LEADING MONEY WINNERS ON EUROPEAN TOUR. LONDON 1996-08-31 Leading money winners on the European Tour after the British Masters won by Robert Allenby on Saturday (British unless stated) : 1. Ian Woosnam 510,258 pounds sterling 2. Colin Montgomerie 442,201 3. Robert Allenby (Australia) 407,748 4. Lee Westwood 301,972 5. Costantino Rocca (Italy) 297,157 6. Mark McNulty (Zimbabwe) 254,247 7. Andrew Coltart 248,142 8. Wayne Riley (Australia) 239,733 9. Raymond Russell 234,330 10. Paul Lawrie 211,420 11. Stephen Ames (Trinidad) 211,175 12. Frank Nobilo (New Zealand) 209,412 13. Paul McGinley (Ireland) 208,978 14. Padraig Harrington (Ireland) 202,593 15. Retief Goosen (South Africa) 195,283 16. Miguel Angel Jimenez (Spain) 184,180 17. Peter Mitchell 183,704 18. Miguel Angel Martin (Spain) 182,533 19. Jonathan Lomas 181,005 20. Paul Broadhurst 176,780\"}'\ntext: GOLF - LEADING MONEY WINNERS ON EUROPEAN TOUR. LONDON 1996-08-31 Leading money winners on the European Tour after the British Masters won by Robert Allenby on Saturday (British unless stated) : 1. Ian Woosnam 510,258 pounds sterling 2. Colin Montgomerie 442,201 3. Robert Allenby (Australia) 407,748 4. Lee Westwood 301,972 5. Costantino Rocca (Italy) 297,157 6. Mark McNulty (Zimbabwe) 254,247 7. Andrew Coltart 248,142 8. Wayne Riley (Australia) 239,733 9. Raymond Russell 234,330 10. Paul Lawrie 211,420 11. Stephen Ames (Trinidad) 211,175 12. Frank Nobilo (New Zealand) 209,412 13. Paul McGinley (Ireland) 208,978 14. Padraig Harrington (Ireland) 202,593 15. Retief Goosen (South Africa) 195,283 16. Miguel Angel Jimenez (Spain) 184,180 17. Peter Mitchell 183,704 18. Miguel Angel Martin (Spain) 182,533 19. Jonathan Lomas 181,005 20. Paul Broadhurst 176,780\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"RUGBY UNION - ENGLISH, SCOTTISH AND WELSH RESULTS. LONDON 1996-08-31 Results of English, Scottish and Welsh rugby union matches on Saturday : English National League one Harlequins 75 Gloucester 19 London Irish 27 Bristol 28 Northampton 46 West Hartlepool 20 Orrell 13 Bath 56 Sale 31 Wasps 33 Saracens 25 Leicester 23 Welsh division one Bridgend 13 Llanelli 9 Dunvant 21 Ebbw Vale 10 Treorchy 17 Newbridge 23 Newport 29 Caerphilly 10 Swansea 49 Cardiff 23 Scottish premier league division one Boroughmuir 20 Hawick 23 Currie 45 Heriot\\\\u0027s F.P. 5 Jed-Forest 17 Watsonians 54 Melrose 107 Stirling County 10\"}'\ntext: RUGBY UNION - ENGLISH, SCOTTISH AND WELSH RESULTS. LONDON 1996-08-31 Results of English, Scottish and Welsh rugby union matches on Saturday : English National League one Harlequins 75 Gloucester 19 London Irish 27 Bristol 28 Northampton 46 West Hartlepool 20 Orrell 13 Bath 56 Sale 31 Wasps 33 Saracens 25 Leicester 23 Welsh division one Bridgend 13 Llanelli 9 Dunvant 21 Ebbw Vale 10 Treorchy 17 Newbridge 23 Newport 29 Caerphilly 10 Swansea 49 Cardiff 23 Scottish premier league division one Boroughmuir 20 Hawick 23 Currie 45 Heriot's F.P. 5 Jed-Forest 17 Watsonians 54 Melrose 107 Stirling County 10\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - WALES BEAT SAN MARINO IN WORLD CUP QUALIFIER. CARDIFF 1996-08-31 Wales beat San Marino 6-0 (halftime 4-0) in a World Cup soccer European group 7 qualifier on Saturday. Scorers : Dean Saunders (2nd minute, 75th), Mark Hughes (25th, 54th), Andy Melville (33rd), John Robinson (45th). Attendance : 15,150\"}'\ntext: SOCCER - WALES BEAT SAN MARINO IN WORLD CUP QUALIFIER. CARDIFF 1996-08-31 Wales beat San Marino 6-0 (halftime 4-0) in a World Cup soccer European group 7 qualifier on Saturday. Scorers : Dean Saunders (2nd minute, 75th), Mark Hughes (25th, 54th), Andy Melville (33rd), John Robinson (45th). Attendance : 15,150\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - UKRAINE BEAT NORTHERN IRELAND IN WORLD CUP QUALIFIER. BELFAST 1996-08-31 Ukraine beat Northern Ireland 1-0 (halftime 0-0) in a World Cup soccer European group nine qualifier on Saturday. Scorer : Sergei Rebrov (79th minute) Attendance : 9,358\"}'\ntext: SOCCER - UKRAINE BEAT NORTHERN IRELAND IN WORLD CUP QUALIFIER. BELFAST 1996-08-31 Ukraine beat Northern Ireland 1-0 (halftime 0-0) in a World Cup soccer European group nine qualifier on Saturday. Scorer : Sergei Rebrov (79th minute) Attendance : 9,358\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"RUGBY UNION - LYNAGH SEALS VICTORY OVER DWYER AND LEICESTER. LONDON 1996-08-31 Former Wallaby captain Michael Lynagh began his career in English club rugby in impeccable fashion on Saturday to frustrate his old coach Bob Dwyer on his league coaching debut with Leicester. Lynagh kicked five penalties and a conversion from his six attempts at goal to steer his multi-national Saracens side to a 25-23 home win and offer millionaire backer Nigel Wray an early return on his big investment in the north London club. French centre Philippe Sella also enjoyed a good game alongside Lynagh, although the home team scored only one try through England scrum-half Kyran Bracken. The new French connection at Harlequins also made a good start, Laurent Cabannes and Laurent Benezech scoring a try apiece in their side\\\\u0027s 75-19 victory over Gloucester. Former England captain Will Carling, handed the kicking duties, finished with 20 points. With the first day of the league season briefly shifting the spotlight away from the discord between the clubs and the Rugby Football Union, there were also emphatic victories for champions Bath, 56-13 winners over Orrell, and Northampton and narrow successes for Wasps and Bristol.\"}'\ntext: RUGBY UNION - LYNAGH SEALS VICTORY OVER DWYER AND LEICESTER. LONDON 1996-08-31 Former Wallaby captain Michael Lynagh began his career in English club rugby in impeccable fashion on Saturday to frustrate his old coach Bob Dwyer on his league coaching debut with Leicester. Lynagh kicked five penalties and a conversion from his six attempts at goal to steer his multi-national Saracens side to a 25-23 home win and offer millionaire backer Nigel Wray an early return on his big investment in the north London club. French centre Philippe Sella also enjoyed a good game alongside Lynagh, although the home team scored only one try through England scrum-half Kyran Bracken. The new French connection at Harlequins also made a good start, Laurent Cabannes and Laurent Benezech scoring a try apiece in their side's 75-19 victory over Gloucester. Former England captain Will Carling, handed the kicking duties, finished with 20 points. With the first day of the league season briefly shifting the spotlight away from the discord between the clubs and the Rugby Football Union, there were also emphatic victories for champions Bath, 56-13 winners over Orrell, and Northampton and narrow successes for Wasps and Bristol.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - SCOTTISH LEAGUE STANDINGS. LONDON 1996-08-31 Scottish league standings after Saturday\\\\u0027s matches (tabulated - played, won, drawn, lost, goals for, goals against, points) : Division one Greenock Morton 3 2 0 1 5 2 6 Dundee 3 1 2 0 3 2 5 St Johnstone 2 1 1 0 3 0 4 St Mirren 3 1 1 1 5 4 4 Airdrieonians 2 1 1 0 1 0 4 Falkirk 3 1 1 1 1 1 4 Clydebank 2 1 0 1 1 3 3 Partick 3 0 2 1 1 2 2 Stirling 3 0 1 2 1 3 1 East Fife 2 0 1 1 0 4 1 Division two Livingston 3 3 0 0 6 2 9 Queen of South 3 2 0 1 5 4 6 Ayr 3 1 2 0 8 2 5 Stenhousemuir 3 1 1 1 6 1 4 Hamilton 3 1 1 1 3 2 4 Stranraer 3 1 1 1 3 3 4 Brechin 3 0 3 0 2 2 3 Clyde 3 1 0 2 2 5 3 Dumbarton 3 0 2 1 3 4 2 Berwick 3 0 0 3 1 14 0 Division three Albion 3 3 0 0 5 0 9 Forfar 3 2 0 1 7 4 6 Cowdenbeath 3 2 0 1 4 3 6 Arbroath 3 1 2 0 4 2 5 Alloa 3 1 1 1 3 3 4 Queen\\\\u0027s Park 3 1 1 1 6 8 4 Montrose 3 1 0 2 3 4 3 Inverness Thistle 3 1 0 2 3 6 3 East Stirling 3 0 2 1 3 4 2 Ross County 3 0 0 3 3 7 0\"}'\ntext: SOCCER - SCOTTISH LEAGUE STANDINGS. LONDON 1996-08-31 Scottish league standings after Saturday's matches (tabulated - played, won, drawn, lost, goals for, goals against, points) : Division one Greenock Morton 3 2 0 1 5 2 6 Dundee 3 1 2 0 3 2 5 St Johnstone 2 1 1 0 3 0 4 St Mirren 3 1 1 1 5 4 4 Airdrieonians 2 1 1 0 1 0 4 Falkirk 3 1 1 1 1 1 4 Clydebank 2 1 0 1 1 3 3 Partick 3 0 2 1 1 2 2 Stirling 3 0 1 2 1 3 1 East Fife 2 0 1 1 0 4 1 Division two Livingston 3 3 0 0 6 2 9 Queen of South 3 2 0 1 5 4 6 Ayr 3 1 2 0 8 2 5 Stenhousemuir 3 1 1 1 6 1 4 Hamilton 3 1 1 1 3 2 4 Stranraer 3 1 1 1 3 3 4 Brechin 3 0 3 0 2 2 3 Clyde 3 1 0 2 2 5 3 Dumbarton 3 0 2 1 3 4 2 Berwick 3 0 0 3 1 14 0 Division three Albion 3 3 0 0 5 0 9 Forfar 3 2 0 1 7 4 6 Cowdenbeath 3 2 0 1 4 3 6 Arbroath 3 1 2 0 4 2 5 Alloa 3 1 1 1 3 3 4 Queen's Park 3 1 1 1 6 8 4 Montrose 3 1 0 2 3 4 3 Inverness Thistle 3 1 0 2 3 6 3 East Stirling 3 0 2 1 3 4 2 Ross County 3 0 0 3 3 7 0\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - ENGLISH LEAGUE STANDINGS. LONDON 1996-08-31 English soccer league standings after Saturday\\\\u0027s matches (tabulated - played, won, drawn, lost, goals for, goals against, points) : Division one Stoke 4 3 1 0 7 4 10 Barnsley 3 3 0 0 8 2 9 Norwich 4 3 0 1 5 3 9 Tranmere 4 2 1 1 6 4 7 Bolton 3 2 1 0 5 2 7 Queens Park Rangers 3 2 1 0 5 3 7 Wolverhampton 4 2 1 1 5 3 7 Swindon 4 2 1 1 5 4 7 Bradford 4 2 0 2 4 3 6 Portsmouth 4 2 0 2 4 5 6 Ipswich 4 1 2 1 9 7 5 Crystal Palace 4 1 2 1 4 3 5 Port Vale 4 1 2 1 4 4 5 Birmingham 2 1 1 0 5 4 4 Reading 4 1 1 2 5 10 4 Huddersfield 3 1 1 1 4 4 4 Oxford 4 1 0 3 6 5 3 Manchester City 3 1 0 2 2 3 3 West Bromwich 3 0 2 1 2 3 2 Oldham 4 0 1 3 5 9 1 Sheffield United 2 0 1 1 4 5 1 Grimsby 4 0 1 3 4 8 1 Southend 4 0 1 3 2 10 1 Charlton 2 0 1 1 1 3 1 Division two Plymouth 4 3 1 0 10 6 10 Brentford 4 3 1 0 9 3 10 Bury 4 3 1 0 8 2 10 Chesterfield 4 3 0 1 4 2 9 Millwall 4 2 1 1 7 5 7 Shrewsbury 4 2 1 1 6 6 7 Blackpool 4 2 1 1 3 2 7 York 4 2 0 2 6 6 6 Burnley 4 2 0 2 6 7 6 Bournemouth 4 2 0 2 5 5 6 Watford 4 2 0 2 4 5 6 Bristol Rovers 3 1 2 0 2 1 5 Peterborough 3 1 1 1 4 4 4 Preston 4 1 1 2 4 5 4 Crewe 4 1 1 2 4 6 4 Gillingham 4 1 1 2 4 6 4 Notts County 3 1 1 1 2 2 4 Bristol City 4 1 0 3 7 8 3 Luton 4 1 0 3 4 10 3 Wycombe 4 0 3 1 2 3 3 Wrexham 2 0 2 0 5 5 2 Stockport 4 0 2 2 1 3 2 Rotherham 4 0 1 3 3 6 1 Walsall 3 0 1 2 2 4 1 Division three Wigan 4 3 1 0 9 4 10 Fulham 4 3 0 1 5 3 9 Hull 4 2 2 0 4 2 8 Hartlepool 4 2 1 1 6 5 7 Torquay 4 2 1 1 5 3 7 Cardiff 4 2 1 1 3 2 7 Scunthorpe 4 2 1 1 3 3 7 Carlisle 4 2 1 1 2 1 7 Scarborough 4 1 3 0 5 3 6 Northampton 4 1 2 1 6 4 5 Lincoln 4 1 2 1 5 5 5 Barnet 4 1 2 1 4 2 5 Exeter 4 1 2 1 4 5 5 Cambridge United 4 1 2 1 3 4 5 Darlington 4 1 1 2 9 8 4 Chester 4 1 1 2 6 7 4 Doncaster 4 1 1 2 4 5 4 Leyton Orient 4 1 1 2 3 3 4 Brighton 4 1 1 2 3 6 4 Hereford 4 1 1 2 2 3 4 Swansea 4 1 0 3 4 9 3 Colchester 4 0 3 1 2 4 3 Rochdale 4 0 2 2 2 4 2 Mansfield 4 0 2 2 2 6 2\"}'\ntext: SOCCER - ENGLISH LEAGUE STANDINGS. LONDON 1996-08-31 English soccer league standings after Saturday's matches (tabulated - played, won, drawn, lost, goals for, goals against, points) : Division one Stoke 4 3 1 0 7 4 10 Barnsley 3 3 0 0 8 2 9 Norwich 4 3 0 1 5 3 9 Tranmere 4 2 1 1 6 4 7 Bolton 3 2 1 0 5 2 7 Queens Park Rangers 3 2 1 0 5 3 7 Wolverhampton 4 2 1 1 5 3 7 Swindon 4 2 1 1 5 4 7 Bradford 4 2 0 2 4 3 6 Portsmouth 4 2 0 2 4 5 6 Ipswich 4 1 2 1 9 7 5 Crystal Palace 4 1 2 1 4 3 5 Port Vale 4 1 2 1 4 4 5 Birmingham 2 1 1 0 5 4 4 Reading 4 1 1 2 5 10 4 Huddersfield 3 1 1 1 4 4 4 Oxford 4 1 0 3 6 5 3 Manchester City 3 1 0 2 2 3 3 West Bromwich 3 0 2 1 2 3 2 Oldham 4 0 1 3 5 9 1 Sheffield United 2 0 1 1 4 5 1 Grimsby 4 0 1 3 4 8 1 Southend 4 0 1 3 2 10 1 Charlton 2 0 1 1 1 3 1 Division two Plymouth 4 3 1 0 10 6 10 Brentford 4 3 1 0 9 3 10 Bury 4 3 1 0 8 2 10 Chesterfield 4 3 0 1 4 2 9 Millwall 4 2 1 1 7 5 7 Shrewsbury 4 2 1 1 6 6 7 Blackpool 4 2 1 1 3 2 7 York 4 2 0 2 6 6 6 Burnley 4 2 0 2 6 7 6 Bournemouth 4 2 0 2 5 5 6 Watford 4 2 0 2 4 5 6 Bristol Rovers 3 1 2 0 2 1 5 Peterborough 3 1 1 1 4 4 4 Preston 4 1 1 2 4 5 4 Crewe 4 1 1 2 4 6 4 Gillingham 4 1 1 2 4 6 4 Notts County 3 1 1 1 2 2 4 Bristol City 4 1 0 3 7 8 3 Luton 4 1 0 3 4 10 3 Wycombe 4 0 3 1 2 3 3 Wrexham 2 0 2 0 5 5 2 Stockport 4 0 2 2 1 3 2 Rotherham 4 0 1 3 3 6 1 Walsall 3 0 1 2 2 4 1 Division three Wigan 4 3 1 0 9 4 10 Fulham 4 3 0 1 5 3 9 Hull 4 2 2 0 4 2 8 Hartlepool 4 2 1 1 6 5 7 Torquay 4 2 1 1 5 3 7 Cardiff 4 2 1 1 3 2 7 Scunthorpe 4 2 1 1 3 3 7 Carlisle 4 2 1 1 2 1 7 Scarborough 4 1 3 0 5 3 6 Northampton 4 1 2 1 6 4 5 Lincoln 4 1 2 1 5 5 5 Barnet 4 1 2 1 4 2 5 Exeter 4 1 2 1 4 5 5 Cambridge United 4 1 2 1 3 4 5 Darlington 4 1 1 2 9 8 4 Chester 4 1 1 2 6 7 4 Doncaster 4 1 1 2 4 5 4 Leyton Orient 4 1 1 2 3 3 4 Brighton 4 1 1 2 3 6 4 Hereford 4 1 1 2 2 3 4 Swansea 4 1 0 3 4 9 3 Colchester 4 0 3 1 2 4 3 Rochdale 4 0 2 2 2 4 2 Mansfield 4 0 2 2 2 6 2\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - SCOTTISH LEAGUE RESULTS. GLASGOW 1996-08-31 Results of Scottish league matches on Saturday : Division one Greenock Morton 1 Falkirk 0 Partick 1 St Mirren 1 Stirling 1 Dundee 1 Postponed : East Fife v Clydebank, St Johnstone v Airdrieonians. Division two Ayr 6 Berwick 0 Clyde 0 Queen of South 2 Dumbarton 1 Brechin 1 Livingston 1 Hamilton 0 Stenhousemuir 0 Stranraer 1 Division three Albion 2 Cowdenbeath 0 Arbroath 0 East Stirling 0 Inverness Thistle 1 Alloa 0 Montrose 2 Ross County 1 Queen\\\\u0027s Park 1 Forfar 4\"}'\ntext: SOCCER - SCOTTISH LEAGUE RESULTS. GLASGOW 1996-08-31 Results of Scottish league matches on Saturday : Division one Greenock Morton 1 Falkirk 0 Partick 1 St Mirren 1 Stirling 1 Dundee 1 Postponed : East Fife v Clydebank, St Johnstone v Airdrieonians. Division two Ayr 6 Berwick 0 Clyde 0 Queen of South 2 Dumbarton 1 Brechin 1 Livingston 1 Hamilton 0 Stenhousemuir 0 Stranraer 1 Division three Albion 2 Cowdenbeath 0 Arbroath 0 East Stirling 0 Inverness Thistle 1 Alloa 0 Montrose 2 Ross County 1 Queen's Park 1 Forfar 4\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - ENGLISH LEAGUE RESULTS. LONDON 1996-08-31 Results of English soccer matches on Saturday : Division one Bradford 1 Tranmere 0 Grimsby 0 Portsmouth 1 Huddersfield 1 Crystal Palace 1 Norwich 1 Wolverhampton 0 Oldham 3 Ipswich 3 Port Vale 2 Oxford 0 Reading 2 Stoke 2 Southend 1 Swindon 3 Postponed : Birmingham v Barnsley, Manchester City v Charlton Playing Sunday : Queens Park Rangers v Bolton Division two Blackpool 0 Wycombe 0 Bournemouth 1 Peterborough 2 Bristol Rovers 1 Stockport 1 Bury 4 Bristol City 0 Crewe 0 Watford 2 Gillingham 0 Chesterfield 1 Luton 1 Rotherham 0 Millwall 2 Burnley 1 Notts County 0 York 1 Shrewsbury 0 Brentford3 Postponed : Walsall v Wrexham Division three Brighton 1 Scunthorpe 1 Cambridge United 0 Cardiff 2 Colchester 1 Hereford 1 Doncaster 3 Darlington 2 Fulham 1 Carlisle 0 Hull 0 Barnet 0 Leyton Orient 2 Hartlepool 0 Mansfield 0 Rochdale 0 Scarborough 1 Northampton 1 Torquay 2 Exeter 0 Wigan 4 Chester 2\"}'\ntext: SOCCER - ENGLISH LEAGUE RESULTS. LONDON 1996-08-31 Results of English soccer matches on Saturday : Division one Bradford 1 Tranmere 0 Grimsby 0 Portsmouth 1 Huddersfield 1 Crystal Palace 1 Norwich 1 Wolverhampton 0 Oldham 3 Ipswich 3 Port Vale 2 Oxford 0 Reading 2 Stoke 2 Southend 1 Swindon 3 Postponed : Birmingham v Barnsley, Manchester City v Charlton Playing Sunday : Queens Park Rangers v Bolton Division two Blackpool 0 Wycombe 0 Bournemouth 1 Peterborough 2 Bristol Rovers 1 Stockport 1 Bury 4 Bristol City 0 Crewe 0 Watford 2 Gillingham 0 Chesterfield 1 Luton 1 Rotherham 0 Millwall 2 Burnley 1 Notts County 0 York 1 Shrewsbury 0 Brentford3 Postponed : Walsall v Wrexham Division three Brighton 1 Scunthorpe 1 Cambridge United 0 Cardiff 2 Colchester 1 Hereford 1 Doncaster 3 Darlington 2 Fulham 1 Carlisle 0 Hull 0 Barnet 0 Leyton Orient 2 Hartlepool 0 Mansfield 0 Rochdale 0 Scarborough 1 Northampton 1 Torquay 2 Exeter 0 Wigan 4 Chester 2\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"CRICKET - ENGLISH COUNTY CHAMPIONSHIP SCORES. LONDON 1996-08-31 Results and close scores of four-day English county championship matches on Saturday : At Portsmouth : Middlesex beat Hampshire by 188 runs. Middlesex 199 and 426, Hampshire 232 and 205 (A. Fraser 5-79, P. Tufnell 4-39). Middlesex 20 points, Hampshire 5. At Chester-le-Street : Glamorgan beat Durham by 141 runs. Glamorgan 259 and 207, Durham 114 and 211. Glamorgan 22 points, Durham 4. At Chesterfield : Derbyshire beat Worcestershire by nine wickets. Worcestershire 238 and 303 (K. Spiring 130 not out, S. Rhodes 57 ; P. DeFreitas 4-70), Derbyshire 471 and 71-1. Derbyshire 24 points, Worcestershire 5. At The Oval (London) : Surrey beat Warwickshire by an innings and 164 runs. Warwickshire 195 and 109 (J. Benjamin 4-17, M. Bicknell 4-38), Surrey 468 (C. Lewis 94, M. Butcher 70, G. Kersey 63, J. Ratcliffe 63, D. Bicknell 55). Surrey 24 points, Warwickshire 2. At Headingley (Leeds) : Yorkshire 290 and 329 (R. Kettleborough 108, G. Hamilton 61 ; P. Such 8-118), Essex 372 and 100-5. At Hove : Sussex 363 and 144, Lancashire 218 and 53-0. At Tunbridge Wells : Nottinghamshire 214 and 167-6 (C. Tolley 64 not out), Kent 244 (C. Hooper 58 ; C. Tolley 4-68, K. Evans 4-71) At Bristol : Gloucestershire 183 and 249 (J. Russell 75), Northamptonshire 190 and 218-9.\"}'\ntext: CRICKET - ENGLISH COUNTY CHAMPIONSHIP SCORES. LONDON 1996-08-31 Results and close scores of four-day English county championship matches on Saturday : At Portsmouth : Middlesex beat Hampshire by 188 runs. Middlesex 199 and 426, Hampshire 232 and 205 (A. Fraser 5-79, P. Tufnell 4-39). Middlesex 20 points, Hampshire 5. At Chester-le-Street : Glamorgan beat Durham by 141 runs. Glamorgan 259 and 207, Durham 114 and 211. Glamorgan 22 points, Durham 4. At Chesterfield : Derbyshire beat Worcestershire by nine wickets. Worcestershire 238 and 303 (K. Spiring 130 not out, S. Rhodes 57 ; P. DeFreitas 4-70), Derbyshire 471 and 71-1. Derbyshire 24 points, Worcestershire 5. At The Oval (London) : Surrey beat Warwickshire by an innings and 164 runs. Warwickshire 195 and 109 (J. Benjamin 4-17, M. Bicknell 4-38), Surrey 468 (C. Lewis 94, M. Butcher 70, G. Kersey 63, J. Ratcliffe 63, D. Bicknell 55). Surrey 24 points, Warwickshire 2. At Headingley (Leeds) : Yorkshire 290 and 329 (R. Kettleborough 108, G. Hamilton 61 ; P. Such 8-118), Essex 372 and 100-5. At Hove : Sussex 363 and 144, Lancashire 218 and 53-0. At Tunbridge Wells : Nottinghamshire 214 and 167-6 (C. Tolley 64 not out), Kent 244 (C. Hooper 58 ; C. Tolley 4-68, K. Evans 4-71) At Bristol : Gloucestershire 183 and 249 (J. Russell 75), Northamptonshire 190 and 218-9.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"MOTOR RACING - LEADING QUALIFIERS FOR VANCOUVER INDYCAR RACE. VANCOUVER 1996-08-31 Top ten drivers in grid for Sunday\\\\u0027s Vancouver IndyCar race after final qualifying on Saturday (tabulate by driver, country, chassis, motor and lap times in seconds) : 1. Alex Zanardi (Italy), Reynard Honda, 53.980 (113.576 mph / 182.778 kph) 2. Michael Andretti (U.S.), Lola Ford Cosworth, 54.483 3. Bobby Rahal (U.S.), Reynard Mercedes-Benz, 54.507 4. Bryan Herta (U.S.), Reynard Mercedes-Benz, 54.578 5. Jimmy Vasser (U.S.), Reynard Honda, 54.617 6. Paul Tracy (Canada), Penske Mercedes-Benz, 54.620 7. Al Unser Jr (U.S.), Penske Mercedes-Benz, 54.683 8. Andre Ribeiro (Brazil), Lola Honda, 54.750 9. Mauricio Gugelmin (Brazil), Reynard Ford Cosworth, 54.762 10. Gil de Ferran (Brazil), Reynard Honda, 54.774\"}'\ntext: MOTOR RACING - LEADING QUALIFIERS FOR VANCOUVER INDYCAR RACE. VANCOUVER 1996-08-31 Top ten drivers in grid for Sunday's Vancouver IndyCar race after final qualifying on Saturday (tabulate by driver, country, chassis, motor and lap times in seconds) : 1. Alex Zanardi (Italy), Reynard Honda, 53.980 (113.576 mph / 182.778 kph) 2. Michael Andretti (U.S.), Lola Ford Cosworth, 54.483 3. Bobby Rahal (U.S.), Reynard Mercedes-Benz, 54.507 4. Bryan Herta (U.S.), Reynard Mercedes-Benz, 54.578 5. Jimmy Vasser (U.S.), Reynard Honda, 54.617 6. Paul Tracy (Canada), Penske Mercedes-Benz, 54.620 7. Al Unser Jr (U.S.), Penske Mercedes-Benz, 54.683 8. Andre Ribeiro (Brazil), Lola Honda, 54.750 9. Mauricio Gugelmin (Brazil), Reynard Ford Cosworth, 54.762 10. Gil de Ferran (Brazil), Reynard Honda, 54.774\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - CANADA BEAT PANAMA 3-1 IN WORLD CUP QUALIFIER. EDMONTON 1996-08-31 Canada beat Panama 3-1 (halftime 2-0) in their CONCACAF semifinal phase qualifying match for the 1998 World Cup on Friday. Scorers : Canada - Aunger (41st min, pen), Paul Peschisolido (42nd), Carlo Corrazin (87th) Panama - Jorge Luis Dely Valdes (50th) Attendance : 9,402\"}'\ntext: SOCCER - CANADA BEAT PANAMA 3-1 IN WORLD CUP QUALIFIER. EDMONTON 1996-08-31 Canada beat Panama 3-1 (halftime 2-0) in their CONCACAF semifinal phase qualifying match for the 1998 World Cup on Friday. Scorers : Canada - Aunger (41st min, pen), Paul Peschisolido (42nd), Carlo Corrazin (87th) Panama - Jorge Luis Dely Valdes (50th) Attendance : 9,402\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"RUGBY UNION - SPRINGBOKS FINALLY BREAK ALL BLACK SPELL. Andy Colquhoun JOHANNESBURG 1996-08-31 South Africa managed to avoid a fifth successive defeat in 1996 at the hands of the All Blacks with an emphatic 32-22 victory in front of an ecstatic Ellis Park crowd on Saturday. They scored three tries in recording their highest total against New Zealand, salvaging some pride in a season in which the world champions have lost five out of eight tests. It also ended a run of nine successive victories this year for New Zealand but arrived too late to prevent a 2-1 series defeat and an historic first All Black series triumph on South African soil. Springbok scrum-half Joost van der Westhuizen was his side\\\\u0027s inspiration, scoring their opening try and making the third for flanker Andre Venter from a quickly taken penalty to give his side a 29-8 lead after 54 minutes. Fullback Andre Joubert scored the other, scorching in from 40 metres at the start of the second half to add to his three long-range penalties. The All Blacks salvaged some pride by scoring two tries from centre Walter Little and scrum-half Justin Marshall in the final five minutes to close a gap which at one point stood at 24 points. But they generally endured an off-day, highlighted by recalled fly-half Andrew Mehrtens who missed five out of eight kicks at goal. Recalled fly-half Henry Honiball kicked the Springboks into a 6-0 lead after 10 minutes only to see Andrew Mehrtens launch a penalty from eight metres inside his own half to narrow the gap. Mehrtens missed three further penalties and a conversion in the first 40 minutes which could have put his side ahead, but it was the Springboks who looked the more dangerous. Their promise was realised when Joubert made a 40-metre break in the 25th minute and, although winger Pieter Hendriks appeared to knock on Joubert\\\\u0027s reverse pass, Welsh referee Derek Bevan allowed Van der Westhuizen to pick up and score under the posts. Honiball converted and Joubert kicked a penalty before All Black hooker Sean Fitzpatrick scored a try from close range on the stroke of half-time to narrow the lead to 16-8 and hint at a comeback. Instead Joubert kicked another long penalty and then raced around the outside of the defence to score the Springboks \\\\u0027 second try. A quick penalty from Van der Westhuizen five metres from the All Black line set up the third try for Venter five minutes later and when Joubert kicked his third penalty the Springboks held an unassailable 32-8 lead going into the last quarter. When the All Blacks did break through, it was too late. Centre Walter Little followed up Mehrtens \\\\u0027 kick to score under the posts and scrum-half Justin Marshall forced himself over from a ruck close to the line in injury-time to give them some consolation. South Africa - 15 - Andre Joubert, 14 - Justin Swart, 13 - Japie Mulder (Joel Stransky, 48 mins) 12 - Danie van Schalkwyk, 11 - Pieter Hendriks ; 10 - Henry Honiball, 9 - Joost van der Westhuizen ; 8 - Gary Teichmann (captain), 7 - Andre Venter (Wayne Fyvie, 75), 6 - Ruben Kruge, 5 - Mark Andrews (Fritz van Heerden, 39), 4 - Kobus Wiese, 3 - Marius Hurter, 2 - James Dalton, 1 - Dawie Theron (Garry Pagel, 66). New Zealand - 15 - Christian Cullen (Alama Ieremia, 70), 14 - Jeff Wilson, 13 - Walter Little, 12 - Frank Bunce, 11 - Glen Osborne ; 10 - Andrew Mehrtens, 9 - Justin Marshall ; 8 - Zinzan Brooke, 7 - Josh Kronfeld, 6 - Michael Jones (Glenn Taylor, 53), 5 - Robin Brooke, 4 - Ian Jones, 3 - Olo Brown, 2 - Sean Fitzpatrick (captain), 1 - Craig Dowd.\"}'\ntext: RUGBY UNION - SPRINGBOKS FINALLY BREAK ALL BLACK SPELL. Andy Colquhoun JOHANNESBURG 1996-08-31 South Africa managed to avoid a fifth successive defeat in 1996 at the hands of the All Blacks with an emphatic 32-22 victory in front of an ecstatic Ellis Park crowd on Saturday. They scored three tries in recording their highest total against New Zealand, salvaging some pride in a season in which the world champions have lost five out of eight tests. It also ended a run of nine successive victories this year for New Zealand but arrived too late to prevent a 2-1 series defeat and an historic first All Black series triumph on South African soil. Springbok scrum-half Joost van der Westhuizen was his side's inspiration, scoring their opening try and making the third for flanker Andre Venter from a quickly taken penalty to give his side a 29-8 lead after 54 minutes. Fullback Andre Joubert scored the other, scorching in from 40 metres at the start of the second half to add to his three long-range penalties. The All Blacks salvaged some pride by scoring two tries from centre Walter Little and scrum-half Justin Marshall in the final five minutes to close a gap which at one point stood at 24 points. But they generally endured an off-day, highlighted by recalled fly-half Andrew Mehrtens who missed five out of eight kicks at goal. Recalled fly-half Henry Honiball kicked the Springboks into a 6-0 lead after 10 minutes only to see Andrew Mehrtens launch a penalty from eight metres inside his own half to narrow the gap. Mehrtens missed three further penalties and a conversion in the first 40 minutes which could have put his side ahead, but it was the Springboks who looked the more dangerous. Their promise was realised when Joubert made a 40-metre break in the 25th minute and, although winger Pieter Hendriks appeared to knock on Joubert's reverse pass, Welsh referee Derek Bevan allowed Van der Westhuizen to pick up and score under the posts. Honiball converted and Joubert kicked a penalty before All Black hooker Sean Fitzpatrick scored a try from close range on the stroke of half-time to narrow the lead to 16-8 and hint at a comeback. Instead Joubert kicked another long penalty and then raced around the outside of the defence to score the Springboks ' second try. A quick penalty from Van der Westhuizen five metres from the All Black line set up the third try for Venter five minutes later and when Joubert kicked his third penalty the Springboks held an unassailable 32-8 lead going into the last quarter. When the All Blacks did break through, it was too late. Centre Walter Little followed up Mehrtens ' kick to score under the posts and scrum-half Justin Marshall forced himself over from a ruck close to the line in injury-time to give them some consolation. South Africa - 15 - Andre Joubert, 14 - Justin Swart, 13 - Japie Mulder (Joel Stransky, 48 mins) 12 - Danie van Schalkwyk, 11 - Pieter Hendriks ; 10 - Henry Honiball, 9 - Joost van der Westhuizen ; 8 - Gary Teichmann (captain), 7 - Andre Venter (Wayne Fyvie, 75), 6 - Ruben Kruge, 5 - Mark Andrews (Fritz van Heerden, 39), 4 - Kobus Wiese, 3 - Marius Hurter, 2 - James Dalton, 1 - Dawie Theron (Garry Pagel, 66). New Zealand - 15 - Christian Cullen (Alama Ieremia, 70), 14 - Jeff Wilson, 13 - Walter Little, 12 - Frank Bunce, 11 - Glen Osborne ; 10 - Andrew Mehrtens, 9 - Justin Marshall ; 8 - Zinzan Brooke, 7 - Josh Kronfeld, 6 - Michael Jones (Glenn Taylor, 53), 5 - Robin Brooke, 4 - Ian Jones, 3 - Olo Brown, 2 - Sean Fitzpatrick (captain), 1 - Craig Dowd.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"RUGBY UNION - SOUTH AFRICA BEAT ALL BLACKS 32-22. JOHANNESBURG 1996-08-31 South Africa beat New Zealand 32-22 (haltime 16-8) in the final test match of their three-test series at Ellis Park on Saturday. Scorers : South Africa - Tries : Joost van der Westhuizen (2), Andre Joubert. Conversion : Henry Honiball. Penalties : Honiball (2), Joubert (3). New Zealand - Tries : Sean Fitzpatrick, Walter Little, Justin Marshall. Conversions : Andrew Mehrtens (2). Penalties : Mehrtens. New Zealand win test series 2-1.\"}'\ntext: RUGBY UNION - SOUTH AFRICA BEAT ALL BLACKS 32-22. JOHANNESBURG 1996-08-31 South Africa beat New Zealand 32-22 (haltime 16-8) in the final test match of their three-test series at Ellis Park on Saturday. Scorers : South Africa - Tries : Joost van der Westhuizen (2), Andre Joubert. Conversion : Henry Honiball. Penalties : Honiball (2), Joubert (3). New Zealand - Tries : Sean Fitzpatrick, Walter Little, Justin Marshall. Conversions : Andrew Mehrtens (2). Penalties : Mehrtens. New Zealand win test series 2-1.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - MAURITANIA DISSOLVES NATIONAL TEAM AFTER CUP EXIT. NOUAKCHOTT 1996-08-31 Mauritania\\\\u0027s soccer federation dissolved the national team and suspended this season\\\\u0027s domestic championship on Saturday in the wake of the country\\\\u0027s failure to qualify for the African Nations \\\\u0027 Cup. \\\\\"Since Mauritania has been eliminated on all fronts and the next commitments are not for another two years, we have reason to take a break,\\\\\" federation president Mohamed Lemine Cheiguer said. The North Africans were held to a goalless draw by Benin on Friday after losing the first leg of their qualifying tie 4-1.\"}'\ntext: SOCCER - MAURITANIA DISSOLVES NATIONAL TEAM AFTER CUP EXIT. NOUAKCHOTT 1996-08-31 Mauritania's soccer federation dissolved the national team and suspended this season's domestic championship on Saturday in the wake of the country's failure to qualify for the African Nations ' Cup. \"Since Mauritania has been eliminated on all fronts and the next commitments are not for another two years, we have reason to take a break,\" federation president Mohamed Lemine Cheiguer said. The North Africans were held to a goalless draw by Benin on Friday after losing the first leg of their qualifying tie 4-1.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - MAURITANIA DRAW WITH BENIN IN AFRICAN NATIONS CUP. NOUAKCHOTT 1996-08-31 Mauritania drew 0-0 with Benin in their African Nations Cup preliminary round, second leg soccer match on Friday. Benin won 4-1 on aggregate.\"}'\ntext: SOCCER - MAURITANIA DRAW WITH BENIN IN AFRICAN NATIONS CUP. NOUAKCHOTT 1996-08-31 Mauritania drew 0-0 with Benin in their African Nations Cup preliminary round, second leg soccer match on Friday. Benin won 4-1 on aggregate.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - YUGOSLAV LEAGUE RESULTS / STANDINGS. BELGRADE 1996-08-31 Results of Yugoslav league soccer matches played on Saturday : Division A Hajduk 2 Proleter (Z) 0 Zemun 1 Rad (B) 0 Borac 1 Mladost (L) 2 Cukaricki 1 Vojvodina 0 Buducnost 1 Red Star 3 Partizan 6 Becej 0 Standings (tabulate under won, drawn, lost, goals for, goals against, points) : Red Star 4 4 0 0 9 3 12 Partizan 4 3 1 0 13 3 10 Mladost (L) 4 2 1 1 8 5 7 Vojvodina 4 2 1 1 5 3 7 Becej 4 2 1 1 5 7 7 Hajduk 4 2 0 2 5 3 6 Cukaricki 4 2 0 2 6 6 6 Zemun 4 1 2 1 3 3 5 Rad (B) 4 1 1 2 2 3 4 Buducnost 4 1 0 3 4 8 3 Proleter (Z) 4 0 1 3 2 9 1 Borac 4 0 0 4 1 10 0 Division B Sloboda 4 Mladost (BJ) 0 Buducnost (V) 0 OFK Beograd 1 Rudar 0 OFK Kikinda 1 Obilic 2 Zeleznik (B) 0 Sutjeska 1 Loznica 0 Radnicki (N) - Spartak (to be played on Sunday) Standings : Obilic 4 4 0 0 10 1 12 OFK Kikinda 4 3 0 1 8 3 9 Sutjeska 4 3 0 1 7 5 9 Loznica 4 2 0 2 7 4 6 OFK Beograd 4 1 3 0 5 4 6 Buducnost (V) 4 2 0 2 4 5 6 Sloboda 4 1 1 2 8 8 4 Spartak 3 1 1 1 3 3 4 Radnicki (N) 3 1 0 2 5 6 3 Zeleznik (B) 4 1 0 3 4 7 3 Rudar 4 1 0 3 1 7 3 Mladost (BJ) 4 0 1 3 2 10 1\"}'\ntext: SOCCER - YUGOSLAV LEAGUE RESULTS / STANDINGS. BELGRADE 1996-08-31 Results of Yugoslav league soccer matches played on Saturday : Division A Hajduk 2 Proleter (Z) 0 Zemun 1 Rad (B) 0 Borac 1 Mladost (L) 2 Cukaricki 1 Vojvodina 0 Buducnost 1 Red Star 3 Partizan 6 Becej 0 Standings (tabulate under won, drawn, lost, goals for, goals against, points) : Red Star 4 4 0 0 9 3 12 Partizan 4 3 1 0 13 3 10 Mladost (L) 4 2 1 1 8 5 7 Vojvodina 4 2 1 1 5 3 7 Becej 4 2 1 1 5 7 7 Hajduk 4 2 0 2 5 3 6 Cukaricki 4 2 0 2 6 6 6 Zemun 4 1 2 1 3 3 5 Rad (B) 4 1 1 2 2 3 4 Buducnost 4 1 0 3 4 8 3 Proleter (Z) 4 0 1 3 2 9 1 Borac 4 0 0 4 1 10 0 Division B Sloboda 4 Mladost (BJ) 0 Buducnost (V) 0 OFK Beograd 1 Rudar 0 OFK Kikinda 1 Obilic 2 Zeleznik (B) 0 Sutjeska 1 Loznica 0 Radnicki (N) - Spartak (to be played on Sunday) Standings : Obilic 4 4 0 0 10 1 12 OFK Kikinda 4 3 0 1 8 3 9 Sutjeska 4 3 0 1 7 5 9 Loznica 4 2 0 2 7 4 6 OFK Beograd 4 1 3 0 5 4 6 Buducnost (V) 4 2 0 2 4 5 6 Sloboda 4 1 1 2 8 8 4 Spartak 3 1 1 1 3 3 4 Radnicki (N) 3 1 0 2 5 6 3 Zeleznik (B) 4 1 0 3 4 7 3 Rudar 4 1 0 3 1 7 3 Mladost (BJ) 4 0 1 3 2 10 1\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - INCE EXPOSED BY GASCOIGNE\\\\u0027S LATEST PRANK. CHISINAU, Moldova 1996-08-31 England\\\\u0027s irrepressible midfielder Paul Gascoigne was up to his old tricks on Saturday, pulling down his team mate Paul Ince\\\\u0027s trousers in front of an astonished crowd in Moldova. Ince was clambering over a wall at the Republican stadium in Chisinau as Glenn Hoddle\\\\u0027s England players tried to escape heavy rain during an under-21 clash. Gascoigne, whose compulsive practical joking has landed him in trouble in the past, tugged down the Inter Milan player\\\\u0027s trousers in front of a group of press photographers. Hoddle, coaching the side for the first time, declined to comment on the incident. England face Moldova in a World Cup qualifier in the same stadium on Sunday.\"}'\ntext: SOCCER - INCE EXPOSED BY GASCOIGNE'S LATEST PRANK. CHISINAU, Moldova 1996-08-31 England's irrepressible midfielder Paul Gascoigne was up to his old tricks on Saturday, pulling down his team mate Paul Ince's trousers in front of an astonished crowd in Moldova. Ince was clambering over a wall at the Republican stadium in Chisinau as Glenn Hoddle's England players tried to escape heavy rain during an under-21 clash. Gascoigne, whose compulsive practical joking has landed him in trouble in the past, tugged down the Inter Milan player's trousers in front of a group of press photographers. Hoddle, coaching the side for the first time, declined to comment on the incident. England face Moldova in a World Cup qualifier in the same stadium on Sunday.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"BASKETBALL - TROFEJ BEOGRAD TOURNAMENT RESULTS. BELGRADE 1996-08-31 Results in the Trofej Beograd 96 international basketball tournament on Saturday : Fifth place : Benetton (Italy) 92 Dinamo (Russia) 81 (halftime 50-28) Third place : Alba (Germany) 75 Red Star (Yugoslavia) 70 (42-41)\"}'\ntext: BASKETBALL - TROFEJ BEOGRAD TOURNAMENT RESULTS. BELGRADE 1996-08-31 Results in the Trofej Beograd 96 international basketball tournament on Saturday : Fifth place : Benetton (Italy) 92 Dinamo (Russia) 81 (halftime 50-28) Third place : Alba (Germany) 75 Red Star (Yugoslavia) 70 (42-41)\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"BASKETBALLSOCCER - TROFEJ BEOGRAD TOURNAMENT RESULTS. BELGRADE 1996-08-31 Results in the Trofej Beograd 96 international basketball tournament on Saturday : Fifth place : Benetton (Italy) 92 Dinamo (Russia) 81 (halftime 50-28) Third place : Alba (Germany) 75 Red Star (Yugoslavia) 70 (42-41)\"}'\ntext: BASKETBALLSOCCER - TROFEJ BEOGRAD TOURNAMENT RESULTS. BELGRADE 1996-08-31 Results in the Trofej Beograd 96 international basketball tournament on Saturday : Fifth place : Benetton (Italy) 92 Dinamo (Russia) 81 (halftime 50-28) Third place : Alba (Germany) 75 Red Star (Yugoslavia) 70 (42-41)\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - ROMANIA BEAT LITHUANIA IN WORLD CUP QUALIFIER. BUCHAREST 1996-08-31 Romania beat Lithuania 3-0 (halftime 1-0) in a World Cup soccer European group 8 qualifier on Saturday. Scorers : Romania - Viorel Moldovan (21st minute), Dan Petrescu (65th), Constantin Galca (77th) Attendence : 9,000\"}'\ntext: SOCCER - ROMANIA BEAT LITHUANIA IN WORLD CUP QUALIFIER. BUCHAREST 1996-08-31 Romania beat Lithuania 3-0 (halftime 1-0) in a World Cup soccer European group 8 qualifier on Saturday. Scorers : Romania - Viorel Moldovan (21st minute), Dan Petrescu (65th), Constantin Galca (77th) Attendence : 9,000\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - ARMENIA AND PORTUGAL DRAW 0-0 IN WORLD CUP QUALIFIER. YEREVAN 1996-08-31 Armenia and Portugal drew 0-0 in a World Cup soccer European group 9 qualifier on Saturday. Attendance : 5,000\"}'\ntext: SOCCER - ARMENIA AND PORTUGAL DRAW 0-0 IN WORLD CUP QUALIFIER. YEREVAN 1996-08-31 Armenia and Portugal drew 0-0 in a World Cup soccer European group 9 qualifier on Saturday. Attendance : 5,000\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - AZERBAIJAN BEAT SWITZERLAND IN WORLD CUP QUALIFIER. BAKU 1996-08-31 Azerbaijan beat Switzerland 1-0 (halftime 1-0) in their World Cup soccer European group three qualifying match on Saturday. Scorer : Vidadi Rzayev (28th) Attendance : 20,000\"}'\ntext: SOCCER - AZERBAIJAN BEAT SWITZERLAND IN WORLD CUP QUALIFIER. BAKU 1996-08-31 Azerbaijan beat Switzerland 1-0 (halftime 1-0) in their World Cup soccer European group three qualifying match on Saturday. Scorer : Vidadi Rzayev (28th) Attendance : 20,000\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"BASKETBALL - BENETTON BEAT DINAMO 92-81. BELGRADE 1996-08-31 Benetton of Italy beat Dinamo of Russia 92-81 (halftime 50-28) in a fifth place play-off in the Trofej Beograd 96 international basketball tournament on Saturday.\"}'\ntext: BASKETBALL - BENETTON BEAT DINAMO 92-81. BELGRADE 1996-08-31 Benetton of Italy beat Dinamo of Russia 92-81 (halftime 50-28) in a fifth place play-off in the Trofej Beograd 96 international basketball tournament on Saturday.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - SWEDEN BEAT LATVIA IN EUROPEAN UNDER-21 QUALIFIER. RIGA 1996-08-31 Sweden beat Latvia 2-0 (halftime 0-0) in a European under-21 soccer championship qualifier on Saturday. Scorers : Joakim Persson 81st minute, Daniel Andersson (89th) Attendance : 300\"}'\ntext: SOCCER - SWEDEN BEAT LATVIA IN EUROPEAN UNDER-21 QUALIFIER. RIGA 1996-08-31 Sweden beat Latvia 2-0 (halftime 0-0) in a European under-21 soccer championship qualifier on Saturday. Scorers : Joakim Persson 81st minute, Daniel Andersson (89th) Attendance : 300\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - BELARUS BEAT ESTONIA IN WORLD CUP QUALIFIER. MINSK 1996-08-31 Belarus beat Estonia 1-0 (halftime 1-0) in a World Cup soccer European group 4 qualifier on Saturday. Scorer : Vladimir Makovsky (35th) Attendance : 6,000\"}'\ntext: SOCCER - BELARUS BEAT ESTONIA IN WORLD CUP QUALIFIER. MINSK 1996-08-31 Belarus beat Estonia 1-0 (halftime 1-0) in a World Cup soccer European group 4 qualifier on Saturday. Scorer : Vladimir Makovsky (35th) Attendance : 6,000\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - ENGLAND BEAT MOLDOVA IN UNDER-21 QUALIFIER. CHISINAU 1996-08-31 England beat Moldova 2-0 (halftime 1-0) in a European Under-21 soccer championship group 2 qualifier on Saturday. Scorers : Bruce Dyer (39th minute), Darren Eadie (53rd) Attendance : 850\"}'\ntext: SOCCER - ENGLAND BEAT MOLDOVA IN UNDER-21 QUALIFIER. CHISINAU 1996-08-31 England beat Moldova 2-0 (halftime 1-0) in a European Under-21 soccer championship group 2 qualifier on Saturday. Scorers : Bruce Dyer (39th minute), Darren Eadie (53rd) Attendance : 850\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SQUASH - HILL BRANDS WORLD CHAMPION JANSHER A CHEAT. HONG KONG 1996-08-31 Controversial Australian Anthony Hill called Jansher Khan a cheat during his acrimonious defeat by the world number one in the Hong Kong Open semifinals on Saturday. The match boiled over when Hill made to walk off court after what he claimed was a game-winning point in the third. When the referee called Jansher\\\\u0027s return good and the decision was accepted by the player, Hill shrieked at the Pakistani : \\\\\"You cheat.\\\\\" \\\\\"He was standing right there and knew the shot was down so I called him a cheat,\\\\\" said Hill, whose squash career has been blighted by fines and suspensions for unacceptable behaviour. \\\\\"He knew it was down and accepted what I called him because of that.\\\\\" Hill won the game on the next point and said later that Jansher was generally honest on court but played by the referee\\\\u0027s decision. The Australian had upset Jansher\\\\u0027s rhythm with his mixture of gamesmanship and fluent stroke-making but eventually succumbed 15-7 17-15 14-15 15-8. \\\\\"I changed my strategy against him today and had him rattled,\\\\\" he added. He is not as fit as he used to be be but was too good for me in the end. I shook him a bit but he will come out next time and be a better player for it-- that\\\\u0027s Jansher. \\\\\"Jansher said that he was disturbed to be called a cheat.\\\\\" What he did was bad for squash, bad for the crowd and bad for the sponsors. \\\\\"We are trying to build up squash like tennis and players should not say things like that.\\\\\" I think the Professional Squash Association should look into this matter and deal with it properly. I am not calling for him to be banned but they have to take some action. \\\\\"Jansher, bidding for an eighth Hong Kong Open title, plays second-seeded Australian Rodney Eyles in the final. Eyles played his best squash of the tournament to beat fourth-seeded Peter Nicol of Scotland 15-10 8-15 15-10 15-4. Eyles, who defeated Jansher in the semifinals of the Portuguese Open in 1993, said that he would like to win for the good of the game.\\\\\" I have nothing against Jansher but it will be great if I could beat him, \\\\\"said Eyles.\\\\\" My biggest problem against Jansher is concentration. I want to beat him so badly that I just cannot get it together. \\\\\"\"}'\ntext: SQUASH - HILL BRANDS WORLD CHAMPION JANSHER A CHEAT. HONG KONG 1996-08-31 Controversial Australian Anthony Hill called Jansher Khan a cheat during his acrimonious defeat by the world number one in the Hong Kong Open semifinals on Saturday. The match boiled over when Hill made to walk off court after what he claimed was a game-winning point in the third. When the referee called Jansher's return good and the decision was accepted by the player, Hill shrieked at the Pakistani : \"You cheat.\" \"He was standing right there and knew the shot was down so I called him a cheat,\" said Hill, whose squash career has been blighted by fines and suspensions for unacceptable behaviour. \"He knew it was down and accepted what I called him because of that.\" Hill won the game on the next point and said later that Jansher was generally honest on court but played by the referee's decision. The Australian had upset Jansher's rhythm with his mixture of gamesmanship and fluent stroke-making but eventually succumbed 15-7 17-15 14-15 15-8. \"I changed my strategy against him today and had him rattled,\" he added. He is not as fit as he used to be be but was too good for me in the end. I shook him a bit but he will come out next time and be a better player for it-- that's Jansher. \"Jansher said that he was disturbed to be called a cheat.\" What he did was bad for squash, bad for the crowd and bad for the sponsors. \"We are trying to build up squash like tennis and players should not say things like that.\" I think the Professional Squash Association should look into this matter and deal with it properly. I am not calling for him to be banned but they have to take some action. \"Jansher, bidding for an eighth Hong Kong Open title, plays second-seeded Australian Rodney Eyles in the final. Eyles played his best squash of the tournament to beat fourth-seeded Peter Nicol of Scotland 15-10 8-15 15-10 15-4. Eyles, who defeated Jansher in the semifinals of the Portuguese Open in 1993, said that he would like to win for the good of the game.\" I have nothing against Jansher but it will be great if I could beat him, \"said Eyles.\" My biggest problem against Jansher is concentration. I want to beat him so badly that I just cannot get it together. \"\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SQUASH - HONG KONG OPEN SEMIFINAL RESULTS. HONG KONG 1996-08-31 Semifinal results in the Hong Kong Open on Saturday (prefix number denotes seeding) : 1 - Jansher Khan (Pakistan) beat Anthony Hill (Australia) 15-7 17-15 14-15 15-8 2 - Rodney Eyles (Australia) beat 4 - Peter Nicol (Scotland) 15-10 8-15 15-10 15-4\"}'\ntext: SQUASH - HONG KONG OPEN SEMIFINAL RESULTS. HONG KONG 1996-08-31 Semifinal results in the Hong Kong Open on Saturday (prefix number denotes seeding) : 1 - Jansher Khan (Pakistan) beat Anthony Hill (Australia) 15-7 17-15 14-15 15-8 2 - Rodney Eyles (Australia) beat 4 - Peter Nicol (Scotland) 15-10 8-15 15-10 15-4\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"GOLF - PARNEVIK TAKES ONE-SHOT LEAD AT GREATER MILWAUKEE OPEN. MILWAUKEE, Wisconsin 1996-08-31 Jesper Parnevik of Sweden fired a course record-tying eight-under-par 63 Saturday to take a one-shot lead into the final round of the $ 1.2 million Greater Milwaukee Open. Parnevik, who is seeking his first PGA Tour victory, moved to 19-under 194 for the tournament. Parnevik tied the 18-hole record set by Loren Roberts in 1994 at the Brown Deer Park Golf Course and also equalled Saturday by Greg Kraft. Nolan Henke, who led by two strokes entering the third round, carded a four-under 67 and was one stroke back at 18-under 195. He is striving for his fourth career PGA Tour victory and first since the 1993 BellSouth Classic. Tiger Woods, who made the cut in his first tournament as a professional, shot a two-over-par 73 and was four under for the tournament. The 20-year-old Woods, who turned professional Tuesday after winning an unprecedented third successive U.S. Amateur Championship, struggled on the front nine, bogeying the first and seventh holes and double-bogeying the par-four, 359-yard ninth hole. After bogeying the 10th hole to move to four-over for the round, he rallied for birdies on 15 and 18. After Parnevik started off his round by parring the first hole and bogeying the second, the Swede birdied six of the next seven holes. Parnevik continued to storm through the course, birdying three holes on the back nine, including two from 12 feet out on the 15th and 17th holes. \\\\\"I ca n\\\\u0027t remember when I\\\\u0027ve putted this well,\\\\\" said Parnevik. \\\\\"I was disappointed when a 12-footer did n\\\\u0027t go in.\\\\\" My game feels very good. I\\\\u0027ve been fading my driver but today whenever I set up for a fade it went straight. Whenever everyone\\\\u0027s making birdies, you never want to be even par. \\\\\"Henke had a bogey-free round and birdied four holes, including a 45-footer on the par-three 215-yard seventh hole and one from 12 feet out on the 12th hole.\\\\\" I did n\\\\u0027t hit it very well today, \\\\\"said Henke.\\\\\" Jasper blew right by me. Once he did, I knew I had to make birdies just to keep up. I made some really good pars. I basically picked up where I left off yesterday afternoon. \\\\\"Right now, I ca n\\\\u0027t put my finger on what\\\\u0027s wrong.\\\\\" Bob Estes shot a 67 for sole possession of third place at 15-under. Steve Stricker, who tied for second at last week\\\\u0027s World Series of Golf, and Stuart Appleby were both five shots off the lead at 14 under. Duffy Waldorf, who also tied for second at the World Series of Golf, carded a 70 to lead a group of six golfers at 13 under, including Kraft. The top four on the PGA Tour money list all skipped the tournament.\"}'\ntext: GOLF - PARNEVIK TAKES ONE-SHOT LEAD AT GREATER MILWAUKEE OPEN. MILWAUKEE, Wisconsin 1996-08-31 Jesper Parnevik of Sweden fired a course record-tying eight-under-par 63 Saturday to take a one-shot lead into the final round of the $ 1.2 million Greater Milwaukee Open. Parnevik, who is seeking his first PGA Tour victory, moved to 19-under 194 for the tournament. Parnevik tied the 18-hole record set by Loren Roberts in 1994 at the Brown Deer Park Golf Course and also equalled Saturday by Greg Kraft. Nolan Henke, who led by two strokes entering the third round, carded a four-under 67 and was one stroke back at 18-under 195. He is striving for his fourth career PGA Tour victory and first since the 1993 BellSouth Classic. Tiger Woods, who made the cut in his first tournament as a professional, shot a two-over-par 73 and was four under for the tournament. The 20-year-old Woods, who turned professional Tuesday after winning an unprecedented third successive U.S. Amateur Championship, struggled on the front nine, bogeying the first and seventh holes and double-bogeying the par-four, 359-yard ninth hole. After bogeying the 10th hole to move to four-over for the round, he rallied for birdies on 15 and 18. After Parnevik started off his round by parring the first hole and bogeying the second, the Swede birdied six of the next seven holes. Parnevik continued to storm through the course, birdying three holes on the back nine, including two from 12 feet out on the 15th and 17th holes. \"I ca n't remember when I've putted this well,\" said Parnevik. \"I was disappointed when a 12-footer did n't go in.\" My game feels very good. I've been fading my driver but today whenever I set up for a fade it went straight. Whenever everyone's making birdies, you never want to be even par. \"Henke had a bogey-free round and birdied four holes, including a 45-footer on the par-three 215-yard seventh hole and one from 12 feet out on the 12th hole.\" I did n't hit it very well today, \"said Henke.\" Jasper blew right by me. Once he did, I knew I had to make birdies just to keep up. I made some really good pars. I basically picked up where I left off yesterday afternoon. \"Right now, I ca n't put my finger on what's wrong.\" Bob Estes shot a 67 for sole possession of third place at 15-under. Steve Stricker, who tied for second at last week's World Series of Golf, and Stuart Appleby were both five shots off the lead at 14 under. Duffy Waldorf, who also tied for second at the World Series of Golf, carded a 70 to lead a group of six golfers at 13 under, including Kraft. The top four on the PGA Tour money list all skipped the tournament.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"TENNIS - SPAIN, U.S. TEAMS OPEN ON ROAD FOR 1997 FED CUP. Richard Finn NEW YORK 1996-08-31 This year\\\\u0027s Fed Cup finalists-- defending champion Spain and the United States-- will hit the road to open the 1997 women\\\\u0027s international team competition, based on the draw conducted Saturday at the U.S. Open. Spain travels to Belgium, while the U.S. team heads to the Netherlands for first-round matches March 1-2. The other two first-round ties will pit hosts Germany against the Czech Republic and visiting France against Japan. The semifinals are July 19-20, and the final September 27- 28. Life on the road this year did not slow the Americans, who will try to avenge their 3-2 defeat in the final last year when they host Spain on September 28-29 in Atlantic City. \\\\\"Last year we stood on the court after we had lost and we put out hands together and made it our committment to bring back the Cup,\\\\\" U.S. captain Billie Jean King said at the draw. \\\\\"That is our sole goal.\\\\\" The United States edged Austria in Salzburg 3-2 in the opening round in April, and then blanked Japan 5-0 in Nagoya last month in the semifinals. The victory against Japan marked the Fed Cup debut of Monica Seles, who became a naturalised U.S. citizen in 1994. Seles easily won both her singles matches and King is counting on the co-world number one to lead the team again. \\\\\"I told Monica we need her if we want to win,\\\\\" King said. Seles\\\\u0027s sore left shoulder and a wrist injury to Fed Cup veteran Mary Joe Fernandez have forced King to take a wait and see attitude regarding her squad for the best-of-five match. Fernandez was forced to withdraw from the U.S. Open. \\\\\"We will wait until the last minute so we check with everybody and their injuries,\\\\\" said King. \\\\\"What we like would be Seles, (Olympic champion Lindsay) Davenport and Mary Joe Fernandez.\\\\\" If she can get that threesome together, King will feel good about her chances against the Spain\\\\u0027s formidable duo of Arantxa Sanchez Vicario and Conchita Martinez. \\\\\"To be a great coach you have to have the right horses and I got the right horses,\\\\\" said King.\"}'\ntext: TENNIS - SPAIN, U.S. TEAMS OPEN ON ROAD FOR 1997 FED CUP. Richard Finn NEW YORK 1996-08-31 This year's Fed Cup finalists-- defending champion Spain and the United States-- will hit the road to open the 1997 women's international team competition, based on the draw conducted Saturday at the U.S. Open. Spain travels to Belgium, while the U.S. team heads to the Netherlands for first-round matches March 1-2. The other two first-round ties will pit hosts Germany against the Czech Republic and visiting France against Japan. The semifinals are July 19-20, and the final September 27- 28. Life on the road this year did not slow the Americans, who will try to avenge their 3-2 defeat in the final last year when they host Spain on September 28-29 in Atlantic City. \"Last year we stood on the court after we had lost and we put out hands together and made it our committment to bring back the Cup,\" U.S. captain Billie Jean King said at the draw. \"That is our sole goal.\" The United States edged Austria in Salzburg 3-2 in the opening round in April, and then blanked Japan 5-0 in Nagoya last month in the semifinals. The victory against Japan marked the Fed Cup debut of Monica Seles, who became a naturalised U.S. citizen in 1994. Seles easily won both her singles matches and King is counting on the co-world number one to lead the team again. \"I told Monica we need her if we want to win,\" King said. Seles's sore left shoulder and a wrist injury to Fed Cup veteran Mary Joe Fernandez have forced King to take a wait and see attitude regarding her squad for the best-of-five match. Fernandez was forced to withdraw from the U.S. Open. \"We will wait until the last minute so we check with everybody and their injuries,\" said King. \"What we like would be Seles, (Olympic champion Lindsay) Davenport and Mary Joe Fernandez.\" If she can get that threesome together, King will feel good about her chances against the Spain's formidable duo of Arantxa Sanchez Vicario and Conchita Martinez. \"To be a great coach you have to have the right horses and I got the right horses,\" said King.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"TENNIS - SATURDAY\\\\u0027S RESULTS FROM THE U.S. OPEN. NEW YORK 1996-08-31 Results from the U.S. Open Tennis Championships at the National Tennis Centre on Saturday (prefix number denotes seeding) : Women\\\\u0027s singles, third round 1 - Steffi Graf (Germany) beat Natasha Zvereva (Belarus) 6-4 6-2 16 - Martina Hingis (Switzerland) beat Naoko Kijimuta (Japan) 6-2 6-2 Judith Wiesner (Austria) beat Petra Langrova (Czech Republic) 6 - 2 6-0 Men\\\\u0027s singles, third round 13 - Thomas Enqvist (Sweden) beat Pablo Campana (Ecuador) 6-4 6-4 6-2\"}'\ntext: TENNIS - SATURDAY'S RESULTS FROM THE U.S. OPEN. NEW YORK 1996-08-31 Results from the U.S. Open Tennis Championships at the National Tennis Centre on Saturday (prefix number denotes seeding) : Women's singles, third round 1 - Steffi Graf (Germany) beat Natasha Zvereva (Belarus) 6-4 6-2 16 - Martina Hingis (Switzerland) beat Naoko Kijimuta (Japan) 6-2 6-2 Judith Wiesner (Austria) beat Petra Langrova (Czech Republic) 6 - 2 6-0 Men's singles, third round 13 - Thomas Enqvist (Sweden) beat Pablo Campana (Ecuador) 6-4 6-4 6-2\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"TENNIS - DRAW FOR 1997 FED CUP WOMEN\\\\u0027S TEAM TOURNAMENT. NEW YORK 1996-08-31 Draw for the women\\\\u0027s 1997 Fed Cup team tennis championships, as conducted at the U.S. Open on Saturday : World Group I, first round (March 1-2) United States at Netherlands Czech Republic at Germany France at Japan Spain at Belgium (semifinals July 19-20, and finals September 27-28) World Group II, first round (March 1-2) Austria at Croatia Switzerland at Slovak Republic Argentina at South Korea Australia at South Africa\"}'\ntext: TENNIS - DRAW FOR 1997 FED CUP WOMEN'S TEAM TOURNAMENT. NEW YORK 1996-08-31 Draw for the women's 1997 Fed Cup team tennis championships, as conducted at the U.S. Open on Saturday : World Group I, first round (March 1-2) United States at Netherlands Czech Republic at Germany France at Japan Spain at Belgium (semifinals July 19-20, and finals September 27-28) World Group II, first round (March 1-2) Austria at Croatia Switzerland at Slovak Republic Argentina at South Korea Australia at South Africa\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - U.S. BEAT EL SALVADOR 3-1. LOS ANGELES 1996-08-30 The United States beat El Salvador 3-1 (halftime 1-0) in an international soccer friendly on Friday. Scorers : U.S. - Joe-Max Moore (3rd minute, 88th on penalty kick), Eric Wynalda (61st) El Salvador - Luis Lazo (61st) Attendance - 18,661\"}'\ntext: SOCCER - U.S. BEAT EL SALVADOR 3-1. LOS ANGELES 1996-08-30 The United States beat El Salvador 3-1 (halftime 1-0) in an international soccer friendly on Friday. Scorers : U.S. - Joe-Max Moore (3rd minute, 88th on penalty kick), Eric Wynalda (61st) El Salvador - Luis Lazo (61st) Attendance - 18,661\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"BASEBALL - MAJOR LEAGUE STANDINGS AFTER FRIDAY\\\\u0027S GAMES. NEW YORK 1996-08-31 Major League Baseball standings after games played on Friday (tabulate under won, lost, winning percentage and games behind) : AMERICAN LEAGUE EASTERN DIVISION W L PCT GB NEW YORK 75 59.560 - BALTIMORE 71 63.530 4 BOSTON 69 66.511 6 1/2 TORONTO 63 72.467 12 1/2 DETROIT 49 86.363 26 1/2 CENTRAL DIVISION CLEVELAND 80 54.597 - CHICAGO 72 64.529 9 MINNESOTA 67 68.496 13 1/2 MILWAUKEE 65 71.478 16 KANSAS CITY 61 75.449 20 WESTERN DIVISION TEXAS 76 58.567 - SEATTLE 70 64.522 6 OAKLAND 65 72.474 12 1/2 CALIFORNIA 62 73.459 14 1/2 SATURDAY, AUGUST 31 SCHEDULE KANSAS CITY AT DETROIT BALTIMORE AT SEATTLE CHICAGO AT TORONTO MINNESOTA AT MILWAUKEE CLEVELAND AT TEXAS BOSTON AT OAKLAND NEW YORK AT CALIFORNIA NATIONAL LEAGUE EASTERN DIVISION W L PCT GB ATLANTA 84 50.627 - MONTREAL 71 62.534 12 1/2 FLORIDA 65 70.481 19 1/2 NEW YORK 59 76.437 25 1/2 PHILADELPHIA 54 81.400 30 1/2 CENTRAL DIVISION HOUSTON 73 63.537 - ST LOUIS 70 65.519 2 1/2 CHICAGO 66 67.496 5 1/2 CINCINNATI 66 68.493 6 PITTSBURGH 56 78.418 16 WESTERN DIVISION SAN DIEGO 76 60.559 - LOS ANGELES 73 61.545 2 COLORADO 70 66.515 6 SAN FRANCISCO 58 74.439 16 SATURDAY, AUGUST 31 SCHEDULE ATLANTA AT CHICAGO HOUSTON AT PITTSBURGH SAN FRANCISCO AT NEW YORK FLORIDA AT CINCINNATI LOS ANGELES AT PHILADELPHIA SAN DIEGO AT MONTREAL COLORADO AT ST LOUIS\"}'\ntext: BASEBALL - MAJOR LEAGUE STANDINGS AFTER FRIDAY'S GAMES. NEW YORK 1996-08-31 Major League Baseball standings after games played on Friday (tabulate under won, lost, winning percentage and games behind) : AMERICAN LEAGUE EASTERN DIVISION W L PCT GB NEW YORK 75 59.560 - BALTIMORE 71 63.530 4 BOSTON 69 66.511 6 1/2 TORONTO 63 72.467 12 1/2 DETROIT 49 86.363 26 1/2 CENTRAL DIVISION CLEVELAND 80 54.597 - CHICAGO 72 64.529 9 MINNESOTA 67 68.496 13 1/2 MILWAUKEE 65 71.478 16 KANSAS CITY 61 75.449 20 WESTERN DIVISION TEXAS 76 58.567 - SEATTLE 70 64.522 6 OAKLAND 65 72.474 12 1/2 CALIFORNIA 62 73.459 14 1/2 SATURDAY, AUGUST 31 SCHEDULE KANSAS CITY AT DETROIT BALTIMORE AT SEATTLE CHICAGO AT TORONTO MINNESOTA AT MILWAUKEE CLEVELAND AT TEXAS BOSTON AT OAKLAND NEW YORK AT CALIFORNIA NATIONAL LEAGUE EASTERN DIVISION W L PCT GB ATLANTA 84 50.627 - MONTREAL 71 62.534 12 1/2 FLORIDA 65 70.481 19 1/2 NEW YORK 59 76.437 25 1/2 PHILADELPHIA 54 81.400 30 1/2 CENTRAL DIVISION HOUSTON 73 63.537 - ST LOUIS 70 65.519 2 1/2 CHICAGO 66 67.496 5 1/2 CINCINNATI 66 68.493 6 PITTSBURGH 56 78.418 16 WESTERN DIVISION SAN DIEGO 76 60.559 - LOS ANGELES 73 61.545 2 COLORADO 70 66.515 6 SAN FRANCISCO 58 74.439 16 SATURDAY, AUGUST 31 SCHEDULE ATLANTA AT CHICAGO HOUSTON AT PITTSBURGH SAN FRANCISCO AT NEW YORK FLORIDA AT CINCINNATI LOS ANGELES AT PHILADELPHIA SAN DIEGO AT MONTREAL COLORADO AT ST LOUIS\nspans: []\n\n\n\n\n" }, { "path": "gerbil-SpotWrapNifWS4Test/my_notes/python_server_format_ed", "content": "\n\n[[Node: IteratorGetNext = IteratorGetNext[output_shapes=[ \n[?], [?,?], [?], [?,?,?], [?,?], [?,?], [?,?], [?], [?,?,?], [?,?,?], [?,?,?], [?,?], \nchunk_id, words, words_len, chars, chars_len, begin_span, end_span, spans_len, cand_entities, cand_entities_scores, cand_entities_labels, cand_entities_len, \n\n[?,?], [?], [?,?], [?,?]\nground_truth, ground_truth_len, begin_gm, end_gm\n\n\noutput_types=\nDT_STRING, DT_INT64, DT_INT64, \tDT_INT64, DT_INT64, \tDT_INT64, \tDT_INT64, \tDT_INT64, \tDT_INT64, \tDT_FLOAT, \t\tDT_INT64, \t\tDT_INT64, \nchunk_id, words, words_len, chars, chars_len, \tbegin_span, \tend_span, \tspans_len, \tcand_entities, cand_entities_scores, cand_entities_labels, cand_entities_len, \n\nDT_INT64, \tDT_INT64, \t\tDT_INT64, \tDT_INT64\nground_truth, ground_truth_len, \tbegin_gm, \tend_gm\n\n _device=\"/job:localhost/replica:0/task:0/device:CPU:0\"](IteratorFromStringHandle)]]\n\n\n\n\ncurl -d '{\"spans\":[{\"start\":41,\"length\":6},{\"start\":127,\"length\":11},{\"start\":165,\"length\":9},{\"start\":9,\"length\":7},{\"start\":81,\"length\":7},{\"start\":143,\"length\":17},{\"start\":112,\"length\":11}],\"text\":\"SOCCER - SPANISH FIRST DIVISION SUMMARY. MADRID 1996-12-07 Summary of Saturday\\\\u0027s Spanish first division match : Real Madrid 2 (Davor Suker 24, Predrag Mijatovic 48) Barcelona 0. Halftime 1-0. Attendance 106,000.\"}' localhost:5555\n\ntext: SOCCER - SPANISH FIRST DIVISION SUMMARY. MADRID 1996-12-07 Summary of Saturday\\u0027s Spanish first division match : Real Madrid 2 (Davor Suker 24, Predrag Mijatovic 48) Barcelona 0. Halftime 1-0. Attendance 106,000.\ntype(text): \nspans: [{'start': 41, 'length': 6}, {'start': 127, 'length': 11}, {'start': 165, 'length': 9}, {'start': 9, 'length': 7}, {'start': 81, 'length': 7}, {'start': 143, 'length': 17}, {'start': 112, 'length': 11}]\ntype(spans): \nspans: [(41, 6), (127, 11), (165, 9), (9, 7), (81, 7), (143, 17), (112, 11)]\ntype(spans[0][0]): \ntype(spans[0][1]): \n\n\n\n\n127.0.0.1 - - [09/Feb/2018 13:43:03] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":439,\"length\":7},{\"start\":176,\"length\":11},{\"start\":83,\"length\":7},{\"start\":51,\"length\":10},{\"start\":421,\"length\":12},{\"start\":399,\"length\":11},{\"start\":0,\"length\":7}],\"text\":\"Clinton to have more news conferences in 2nd term. WASHINGTON 1996-12-06 President Clinton aims to hold more news conferences in his second term and will have one Dec. 13, the White House said Friday. The president had only two formal, full-blown news conferences last year, one in January and one after he won re-election in November, although he had various other limited sessions with the press. White House spokesman Mike McCurry said Clinton \\\\\"plans to have regular news conferences\\\\\" during his second term. But when asked how frequent these would be, he was evasive, saying, \\\\\"periodic, occasional.\\\\\" \\\\\"He enjoys the give and take\\\\\" with reporters, the spokesman added.\"}'\ntext: Clinton to have more news conferences in 2nd term. WASHINGTON 1996-12-06 President Clinton aims to hold more news conferences in his second term and will have one Dec. 13, the White House said Friday. The president had only two formal, full-blown news conferences last year, one in January and one after he won re-election in November, although he had various other limited sessions with the press. White House spokesman Mike McCurry said Clinton \"plans to have regular news conferences\" during his second term. But when asked how frequent these would be, he was evasive, saying, \"periodic, occasional.\" \"He enjoys the give and take\" with reporters, the spokesman added.\ntype(text): \nspans: [{'start': 439, 'length': 7}, {'start': 176, 'length': 11}, {'start': 83, 'length': 7}, {'start': 51, 'length': 10}, {'start': 421, 'length': 12}, {'start': 399, 'length': 11}, {'start': 0, 'length': 7}]\ntype(spans): \nspans: [(439, 7), (176, 11), (83, 7), (51, 10), (421, 12), (399, 11), (0, 7)]\n\n\n\n\n\n\\n\n\n127.0.0.1 - - [09/Feb/2018 12:41:16] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":35,\"length\":3},{\"start\":106,\"length\":4},{\"start\":24,\"length\":7},{\"start\":9,\"length\":6},{\"start\":83,\"length\":11},{\"start\":76,\"length\":6},{\"start\":113,\"length\":12},{\"start\":58,\"length\":6},{\"start\":135,\"length\":31}],\"text\":\"BOXING - SCHULZ DEFEATS RIBALTA IN IBF HEAVYWEIGHT FIGHT. VIENNA 1996-12-07 German Axel Schulz outpointed Cuba\\\\u0027s Jose Ribalta in their International Boxing Federation non-title 10-round heavyweight fight on Saturday.\"}'\ntext: BOXING - SCHULZ DEFEATS RIBALTA IN IBF HEAVYWEIGHT FIGHT. VIENNA 1996-12-07 German Axel Schulz outpointed Cuba's Jose Ribalta in their International Boxing Federation non-title 10-round heavyweight fight on Saturday.\ntype(text): \nspans: [{'start': 35, 'length': 3}, {'start': 106, 'length': 4}, {'start': 24, 'length': 7}, {'start': 9, 'length': 6}, {'start': 83, 'length': 11}, {'start': 76, 'length': 6}, {'start': 113, 'length': 12}, {'start': 58, 'length': 6}, {'start': 135, 'length': 31}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:16] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":41,\"length\":6},{\"start\":127,\"length\":11},{\"start\":165,\"length\":9},{\"start\":9,\"length\":7},{\"start\":81,\"length\":7},{\"start\":143,\"length\":17},{\"start\":112,\"length\":11}],\"text\":\"SOCCER - SPANISH FIRST DIVISION SUMMARY. MADRID 1996-12-07 Summary of Saturday\\\\u0027s Spanish first division match : Real Madrid 2 (Davor Suker 24, Predrag Mijatovic 48) Barcelona 0. Halftime 1-0. Attendance 106,000.\"}'\ntext: SOCCER - SPANISH FIRST DIVISION SUMMARY. MADRID 1996-12-07 Summary of Saturday's Spanish first division match : Real Madrid 2 (Davor Suker 24, Predrag Mijatovic 48) Barcelona 0. Halftime 1-0. Attendance 106,000.\ntype(text): \nspans: [{'start': 41, 'length': 6}, {'start': 127, 'length': 11}, {'start': 165, 'length': 9}, {'start': 9, 'length': 7}, {'start': 81, 'length': 7}, {'start': 143, 'length': 17}, {'start': 112, 'length': 11}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:16] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":566,\"length\":9},{\"start\":193,\"length\":5},{\"start\":75,\"length\":11},{\"start\":294,\"length\":9},{\"start\":607,\"length\":16},{\"start\":9,\"length\":6},{\"start\":180,\"length\":9},{\"start\":646,\"length\":9},{\"start\":112,\"length\":11},{\"start\":546,\"length\":11},{\"start\":457,\"length\":9},{\"start\":765,\"length\":7},{\"start\":57,\"length\":6},{\"start\":89,\"length\":6},{\"start\":128,\"length\":17},{\"start\":246,\"length\":4},{\"start\":747,\"length\":9},{\"start\":479,\"length\":8},{\"start\":502,\"length\":5},{\"start\":51,\"length\":4},{\"start\":369,\"length\":4}],\"text\":\"SOCCER - BALKAN STRIKE FORCE WIN OLD FIRM GAME FOR REAL. MADRID 1996-12-07 Real Madrid\\\\u0027s Balkan strike force of Davor Suker and Predrag Mijatovic shot their side to a 2-0 win over Barcelona in Spain\\\\u0027s old firm game on Saturday. The result leaves Real on 38 points after 16 games, four ahead of Barcelona. With just one league match scheduled before the New Year break, Real are also assured of spending Christmas ahead of their arch-rivals. A mix-up in the Barcelona defence let Croatian international Suker in midway through the first half, and Montenegrin striker Mijatovic made it 2-0 after fine work by Clarence Seedorf just after the break. Barcelona fought back strongly but were twice denied by the woodwork on an unusually quiet night for Brazilian striker Ronaldo.\"}'\ntext: SOCCER - BALKAN STRIKE FORCE WIN OLD FIRM GAME FOR REAL. MADRID 1996-12-07 Real Madrid's Balkan strike force of Davor Suker and Predrag Mijatovic shot their side to a 2-0 win over Barcelona in Spain's old firm game on Saturday. The result leaves Real on 38 points after 16 games, four ahead of Barcelona. With just one league match scheduled before the New Year break, Real are also assured of spending Christmas ahead of their arch-rivals. A mix-up in the Barcelona defence let Croatian international Suker in midway through the first half, and Montenegrin striker Mijatovic made it 2-0 after fine work by Clarence Seedorf just after the break. Barcelona fought back strongly but were twice denied by the woodwork on an unusually quiet night for Brazilian striker Ronaldo.\ntype(text): \nspans: [{'start': 566, 'length': 9}, {'start': 193, 'length': 5}, {'start': 75, 'length': 11}, {'start': 294, 'length': 9}, {'start': 607, 'length': 16}, {'start': 9, 'length': 6}, {'start': 180, 'length': 9}, {'start': 646, 'length': 9}, {'start': 112, 'length': 11}, {'start': 546, 'length': 11}, {'start': 457, 'length': 9}, {'start': 765, 'length': 7}, {'start': 57, 'length': 6}, {'start': 89, 'length': 6}, {'start': 128, 'length': 17}, {'start': 246, 'length': 4}, {'start': 747, 'length': 9}, {'start': 479, 'length': 8}, {'start': 502, 'length': 5}, {'start': 51, 'length': 4}, {'start': 369, 'length': 4}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:16] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":400,\"length\":8},{\"start\":17,\"length\":8},{\"start\":552,\"length\":14},{\"start\":341,\"length\":3},{\"start\":56,\"length\":9},{\"start\":791,\"length\":10},{\"start\":306,\"length\":9},{\"start\":627,\"length\":10},{\"start\":278,\"length\":7},{\"start\":259,\"length\":9},{\"start\":106,\"length\":15},{\"start\":657,\"length\":9},{\"start\":607,\"length\":3},{\"start\":761,\"length\":24},{\"start\":290,\"length\":7},{\"start\":74,\"length\":7},{\"start\":476,\"length\":9},{\"start\":172,\"length\":13},{\"start\":730,\"length\":8},{\"start\":617,\"length\":4},{\"start\":86,\"length\":8},{\"start\":211,\"length\":8},{\"start\":143,\"length\":5},{\"start\":35,\"length\":9},{\"start\":816,\"length\":18},{\"start\":9,\"length\":3},{\"start\":688,\"length\":8}],\"text\":\"SOCCER - PSV HIT VOLENDAM FOR SIX. AMSTERDAM 1996-12-07 Brazilian striker Marcelo and Yugoslav midfielder Zeljko Petrovic each scored twice as Dutch first division leaders PSV Eindhoven romped to a 6-0 win over Volendam on Saturday. Their other marksmen were Brazilian defender Vampeta and Belgian striker Luc Nilis, his 14th of the season. PSV, well on the way to their 14th league title, outgunned Volendam in every department of the game. They boast a nine-point lead over Feyenoord, who have two games in hand, and are 16 points clear of champions Ajax Amsterdam, who have played 18 matches compared to PSV\\\\u0027s 19. Ajax face AZ Alkmaar away on Sunday and Feyenoord, eliminated from the UEFA Cup after losing 4-2 on aggregate to Tenerife on Tuesday, travel to De Graafschap Doetinchem. The Doetinchem side, dubbed \\\\\"The Super Peasants\\\\\", are one of the surprise packages of the season. They are fourth in the table.\"}'\ntext: SOCCER - PSV HIT VOLENDAM FOR SIX. AMSTERDAM 1996-12-07 Brazilian striker Marcelo and Yugoslav midfielder Zeljko Petrovic each scored twice as Dutch first division leaders PSV Eindhoven romped to a 6-0 win over Volendam on Saturday. Their other marksmen were Brazilian defender Vampeta and Belgian striker Luc Nilis, his 14th of the season. PSV, well on the way to their 14th league title, outgunned Volendam in every department of the game. They boast a nine-point lead over Feyenoord, who have two games in hand, and are 16 points clear of champions Ajax Amsterdam, who have played 18 matches compared to PSV's 19. Ajax face AZ Alkmaar away on Sunday and Feyenoord, eliminated from the UEFA Cup after losing 4-2 on aggregate to Tenerife on Tuesday, travel to De Graafschap Doetinchem. The Doetinchem side, dubbed \"The Super Peasants\", are one of the surprise packages of the season. They are fourth in the table.\ntype(text): \nspans: [{'start': 400, 'length': 8}, {'start': 17, 'length': 8}, {'start': 552, 'length': 14}, {'start': 341, 'length': 3}, {'start': 56, 'length': 9}, {'start': 791, 'length': 10}, {'start': 306, 'length': 9}, {'start': 627, 'length': 10}, {'start': 278, 'length': 7}, {'start': 259, 'length': 9}, {'start': 106, 'length': 15}, {'start': 657, 'length': 9}, {'start': 607, 'length': 3}, {'start': 761, 'length': 24}, {'start': 290, 'length': 7}, {'start': 74, 'length': 7}, {'start': 476, 'length': 9}, {'start': 172, 'length': 13}, {'start': 730, 'length': 8}, {'start': 617, 'length': 4}, {'start': 86, 'length': 8}, {'start': 211, 'length': 8}, {'start': 143, 'length': 5}, {'start': 35, 'length': 9}, {'start': 816, 'length': 18}, {'start': 9, 'length': 3}, {'start': 688, 'length': 8}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:16] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":644,\"length\":8},{\"start\":838,\"length\":8},{\"start\":756,\"length\":8},{\"start\":52,\"length\":6},{\"start\":464,\"length\":10},{\"start\":141,\"length\":9},{\"start\":364,\"length\":15},{\"start\":588,\"length\":10},{\"start\":398,\"length\":15},{\"start\":696,\"length\":8},{\"start\":241,\"length\":11},{\"start\":127,\"length\":11},{\"start\":272,\"length\":9},{\"start\":96,\"length\":7},{\"start\":783,\"length\":7},{\"start\":9,\"length\":7},{\"start\":335,\"length\":10},{\"start\":729,\"length\":8},{\"start\":561,\"length\":8},{\"start\":432,\"length\":13},{\"start\":865,\"length\":11},{\"start\":528,\"length\":14},{\"start\":493,\"length\":16},{\"start\":617,\"length\":8},{\"start\":301,\"length\":16},{\"start\":671,\"length\":6},{\"start\":809,\"length\":10}],\"text\":\"SOCCER - SPANISH FIRST DIVISION RESULT / STANDINGS. MADRID 1996-12-07 Result of Saturday\\\\u0027s only Spanish first division match : Real Madrid 2 Barcelona 0 Standings (tabulate under games played, won, drawn, lost, goals for, against, points) : Real Madrid 16 11 5 0 32 12 38 Barcelona 16 10 4 2 46 21 34 Deportivo Coruna 15 9 6 0 23 7 33 Real Betis 15 8 5 2 28 13 29 Atletico Madrid 15 8 3 4 26 17 27 Athletic Bilbao 15 7 4 4 28 22 25 Real Sociedad 15 7 3 5 20 18 24 Valladolid 15 7 3 5 19 18 24 Racing Santander 15 5 7 3 15 15 22 Rayo Vallecano 15 5 5 5 21 19 20 Valencia 15 6 2 7 23 22 20 Celta Vigo 15 5 5 5 17 17 20 Tenerife 15 5 4 6 23 17 19 Espanyol 15 4 4 7 17 20 16 Oviedo 15 4 4 7 17 21 16 Sporting Gijon 15 4 4 7 15 22 16 Logrones 15 4 3 8 11 33 15 Zaragoza 15 2 8 5 18 23 14 Sevilla 15 4 2 9 13 20 14 Compostela 15 3 4 8 13 28 13 Hercules 15 2 2 11 11 29 8 Extremadura 15 1 3 11 8 30 6\"}'\ntext: SOCCER - SPANISH FIRST DIVISION RESULT / STANDINGS. MADRID 1996-12-07 Result of Saturday's only Spanish first division match : Real Madrid 2 Barcelona 0 Standings (tabulate under games played, won, drawn, lost, goals for, against, points) : Real Madrid 16 11 5 0 32 12 38 Barcelona 16 10 4 2 46 21 34 Deportivo Coruna 15 9 6 0 23 7 33 Real Betis 15 8 5 2 28 13 29 Atletico Madrid 15 8 3 4 26 17 27 Athletic Bilbao 15 7 4 4 28 22 25 Real Sociedad 15 7 3 5 20 18 24 Valladolid 15 7 3 5 19 18 24 Racing Santander 15 5 7 3 15 15 22 Rayo Vallecano 15 5 5 5 21 19 20 Valencia 15 6 2 7 23 22 20 Celta Vigo 15 5 5 5 17 17 20 Tenerife 15 5 4 6 23 17 19 Espanyol 15 4 4 7 17 20 16 Oviedo 15 4 4 7 17 21 16 Sporting Gijon 15 4 4 7 15 22 16 Logrones 15 4 3 8 11 33 15 Zaragoza 15 2 8 5 18 23 14 Sevilla 15 4 2 9 13 20 14 Compostela 15 3 4 8 13 28 13 Hercules 15 2 2 11 11 29 8 Extremadura 15 1 3 11 8 30 6\ntype(text): \nspans: [{'start': 644, 'length': 8}, {'start': 838, 'length': 8}, {'start': 756, 'length': 8}, {'start': 52, 'length': 6}, {'start': 464, 'length': 10}, {'start': 141, 'length': 9}, {'start': 364, 'length': 15}, {'start': 588, 'length': 10}, {'start': 398, 'length': 15}, {'start': 696, 'length': 8}, {'start': 241, 'length': 11}, {'start': 127, 'length': 11}, {'start': 272, 'length': 9}, {'start': 96, 'length': 7}, {'start': 783, 'length': 7}, {'start': 9, 'length': 7}, {'start': 335, 'length': 10}, {'start': 729, 'length': 8}, {'start': 561, 'length': 8}, {'start': 432, 'length': 13}, {'start': 865, 'length': 11}, {'start': 528, 'length': 14}, {'start': 493, 'length': 16}, {'start': 617, 'length': 8}, {'start': 301, 'length': 16}, {'start': 671, 'length': 6}, {'start': 809, 'length': 10}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:16] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":310,\"length\":5},{\"start\":510,\"length\":19},{\"start\":417,\"length\":8},{\"start\":77,\"length\":13},{\"start\":699,\"length\":8},{\"start\":274,\"length\":7},{\"start\":20,\"length\":8},{\"start\":1381,\"length\":5},{\"start\":1094,\"length\":7},{\"start\":843,\"length\":8},{\"start\":1238,\"length\":7},{\"start\":1327,\"length\":14},{\"start\":640,\"length\":7},{\"start\":167,\"length\":10},{\"start\":220,\"length\":8},{\"start\":9,\"length\":10},{\"start\":125,\"length\":7},{\"start\":59,\"length\":6},{\"start\":1113,\"length\":5},{\"start\":860,\"length\":7},{\"start\":1062,\"length\":8},{\"start\":991,\"length\":7},{\"start\":1017,\"length\":9},{\"start\":353,\"length\":11},{\"start\":248,\"length\":5},{\"start\":49,\"length\":8},{\"start\":1199,\"length\":12},{\"start\":1139,\"length\":7}],\"text\":\"SOCCER - ENGLISHMAN CHARLTON IS MADE AN HONORARY IRISHMAN. DUBLIN 1996-12-07 Jack Charlton\\\\u0027s relationship with the people of Ireland was cemented on Saturday when the Englishman was officially declared one of their own. Charlton, 61, and his wife, Peggy, became citizens of Ireland when they formally received Irish passports from deputy Prime Minister Dick Spring who said the honour had been made in recognition of Charlton\\\\u0027s achievements as the national soccer manager. \\\\\"The years I spent as manager of the Republic of Ireland were the best years of my life. It all culminated in the fact that I now have lots of great, great friends in Ireland. That is why this is so emotional a night for me,\\\\\" Charlton said. \\\\\"It was the joy that we all had over the period, that I shared with people that I grew to love, that I treasure most,\\\\\" he added. Charlton managed Ireland for 93 matches, during which time they lost only 17 times in almost 10 years until he resigned in December 1995. He guided Ireland to two successive World Cup finals tournaments and to the 1988 European championship finals in Germany, after the Irish beat a well-fancied England team 1-0 in their group qualifier. The lanky former Leeds United defender did not make his England debut until the age of 30 but eventually won 35 caps and was a key member of the 1966 World Cup winning team with his younger brother, Bobby.\"}'\ntext: SOCCER - ENGLISHMAN CHARLTON IS MADE AN HONORARY IRISHMAN. DUBLIN 1996-12-07 Jack Charlton's relationship with the people of Ireland was cemented on Saturday when the Englishman was officially declared one of their own. Charlton, 61, and his wife, Peggy, became citizens of Ireland when they formally received Irish passports from deputy Prime Minister Dick Spring who said the honour had been made in recognition of Charlton's achievements as the national soccer manager. \"The years I spent as manager of the Republic of Ireland were the best years of my life. It all culminated in the fact that I now have lots of great, great friends in Ireland. That is why this is so emotional a night for me,\" Charlton said. \"It was the joy that we all had over the period, that I shared with people that I grew to love, that I treasure most,\" he added. Charlton managed Ireland for 93 matches, during which time they lost only 17 times in almost 10 years until he resigned in December 1995. He guided Ireland to two successive World Cup finals tournaments and to the 1988 European championship finals in Germany, after the Irish beat a well-fancied England team 1-0 in their group qualifier. The lanky former Leeds United defender did not make his England debut until the age of 30 but eventually won 35 caps and was a key member of the 1966 World Cup winning team with his younger brother, Bobby.\ntype(text): \nspans: [{'start': 310, 'length': 5}, {'start': 510, 'length': 19}, {'start': 417, 'length': 8}, {'start': 77, 'length': 13}, {'start': 699, 'length': 8}, {'start': 274, 'length': 7}, {'start': 20, 'length': 8}, {'start': 1381, 'length': 5}, {'start': 1094, 'length': 7}, {'start': 843, 'length': 8}, {'start': 1238, 'length': 7}, {'start': 1327, 'length': 14}, {'start': 640, 'length': 7}, {'start': 167, 'length': 10}, {'start': 220, 'length': 8}, {'start': 9, 'length': 10}, {'start': 125, 'length': 7}, {'start': 59, 'length': 6}, {'start': 1113, 'length': 5}, {'start': 860, 'length': 7}, {'start': 1062, 'length': 8}, {'start': 991, 'length': 7}, {'start': 1017, 'length': 9}, {'start': 353, 'length': 11}, {'start': 248, 'length': 5}, {'start': 49, 'length': 8}, {'start': 1199, 'length': 12}, {'start': 1139, 'length': 7}]\ntype(spans): \n\n\n\n\n\n\n\n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":584,\"length\":6},{\"start\":639,\"length\":4},{\"start\":332,\"length\":7},{\"start\":688,\"length\":4},{\"start\":381,\"length\":6},{\"start\":224,\"length\":6},{\"start\":250,\"length\":8},{\"start\":68,\"length\":6},{\"start\":277,\"length\":10},{\"start\":485,\"length\":4},{\"start\":185,\"length\":19},{\"start\":34,\"length\":5},{\"start\":9,\"length\":6},{\"start\":508,\"length\":6},{\"start\":609,\"length\":11},{\"start\":406,\"length\":8},{\"start\":663,\"length\":5},{\"start\":557,\"length\":8},{\"start\":358,\"length\":4},{\"start\":457,\"length\":9},{\"start\":307,\"length\":6},{\"start\":533,\"length\":4},{\"start\":433,\"length\":5}],\"text\":\"SOCCER - FRENCH LEAGUE STANDINGS. PARIS 1996-12-06 Standings in the French first division after Friday\\\\u0027s matches (tabulate under played, won, drawn, lost, goals for, against, points) : Paris Saint-Germain 21 12 6 3 34 15 42 Monaco 20 12 5 3 36 16 41 Bordeaux 20 9 7 4 29 21 34 Strasbourg 20 11 1 8 27 27 34 Bastia 20 9 6 5 27 22 33 Auxerre 20 8 8 4 26 12 32 Metz 20 8 7 5 21 16 31 Nantes 21 7 9 5 41 25 30 Guingamp 20 7 7 6 18 18 28 Lille 20 7 7 6 22 28 28 Marseille 20 6 8 6 18 17 26 Lyon 20 6 8 6 24 31 26 Rennes 20 7 4 9 23 28 25 Lens 21 7 4 10 25 34 25 Le Havre 20 5 7 8 20 21 22 Cannes 20 5 7 8 13 22 22 Montpellier 20 3 9 8 17 24 18 Caen 20 3 7 10 12 23 16 Nancy 21 3 7 11 14 26 16 Nice 20 3 4 13 17 38 13\"}'\ntext: SOCCER - FRENCH LEAGUE STANDINGS. PARIS 1996-12-06 Standings in the French first division after Friday's matches (tabulate under played, won, drawn, lost, goals for, against, points) : Paris Saint-Germain 21 12 6 3 34 15 42 Monaco 20 12 5 3 36 16 41 Bordeaux 20 9 7 4 29 21 34 Strasbourg 20 11 1 8 27 27 34 Bastia 20 9 6 5 27 22 33 Auxerre 20 8 8 4 26 12 32 Metz 20 8 7 5 21 16 31 Nantes 21 7 9 5 41 25 30 Guingamp 20 7 7 6 18 18 28 Lille 20 7 7 6 22 28 28 Marseille 20 6 8 6 18 17 26 Lyon 20 6 8 6 24 31 26 Rennes 20 7 4 9 23 28 25 Lens 21 7 4 10 25 34 25 Le Havre 20 5 7 8 20 21 22 Cannes 20 5 7 8 13 22 22 Montpellier 20 3 9 8 17 24 18 Caen 20 3 7 10 12 23 16 Nancy 21 3 7 11 14 26 16 Nice 20 3 4 13 17 38 13\ntype(text): \nspans: [{'start': 584, 'length': 6}, {'start': 639, 'length': 4}, {'start': 332, 'length': 7}, {'start': 688, 'length': 4}, {'start': 381, 'length': 6}, {'start': 224, 'length': 6}, {'start': 250, 'length': 8}, {'start': 68, 'length': 6}, {'start': 277, 'length': 10}, {'start': 485, 'length': 4}, {'start': 185, 'length': 19}, {'start': 34, 'length': 5}, {'start': 9, 'length': 6}, {'start': 508, 'length': 6}, {'start': 609, 'length': 11}, {'start': 406, 'length': 8}, {'start': 663, 'length': 5}, {'start': 557, 'length': 8}, {'start': 358, 'length': 4}, {'start': 457, 'length': 9}, {'start': 307, 'length': 6}, {'start': 533, 'length': 4}, {'start': 433, 'length': 5}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":32,\"length\":5},{\"start\":60,\"length\":6},{\"start\":118,\"length\":16},{\"start\":9,\"length\":6},{\"start\":137,\"length\":5},{\"start\":102,\"length\":4},{\"start\":109,\"length\":6}],\"text\":\"SOCCER - FRENCH LEAGUE RESULTS. PARIS 1996-12-06 Results of French first division matches on Friday : Lens 0 Nantes 4 Paris St Germain 1 Nancy 2\"}'\ntext: SOCCER - FRENCH LEAGUE RESULTS. PARIS 1996-12-06 Results of French first division matches on Friday : Lens 0 Nantes 4 Paris St Germain 1 Nancy 2\ntype(text): \nspans: [{'start': 32, 'length': 5}, {'start': 60, 'length': 6}, {'start': 118, 'length': 16}, {'start': 9, 'length': 6}, {'start': 137, 'length': 5}, {'start': 102, 'length': 4}, {'start': 109, 'length': 6}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":188,\"length\":7},{\"start\":434,\"length\":7},{\"start\":251,\"length\":13},{\"start\":93,\"length\":6},{\"start\":137,\"length\":9},{\"start\":445,\"length\":6},{\"start\":42,\"length\":4},{\"start\":202,\"length\":7},{\"start\":9,\"length\":6},{\"start\":466,\"length\":13},{\"start\":390,\"length\":8},{\"start\":377,\"length\":6},{\"start\":366,\"length\":4},{\"start\":279,\"length\":11},{\"start\":127,\"length\":6},{\"start\":168,\"length\":16},{\"start\":354,\"length\":5},{\"start\":268,\"length\":4},{\"start\":294,\"length\":8},{\"start\":341,\"length\":9},{\"start\":157,\"length\":4}],\"text\":\"SOCCER - GERMAN FIRST DIVISION SUMMARIES. BONN 1996-12-06 Summaries of matches played in the German first division on Friday : Bochum 2 (Stickroth 30th pen, Wosz 89th) Bayer Leverkusen 2 (Kirsten 18th, Ramelow 56th). Halftime 1-1. Attendance : 24,602 Werder Bremen 1 (Bode 85th) 1860 Munich 1 (Bormirow 8th). Halftime 0-1. Attendance 33,000 Karlsruhe 3 (Reich 29th, Carl 44th, Dundee 69th) Freiburg 0. Halftime 2-0. Attendance 33,000 Schalke 2 (Mulder 2nd and 27th) Hansa Rostock 0. Halftime 2-0. Attendance 29,300\"}'\ntext: SOCCER - GERMAN FIRST DIVISION SUMMARIES. BONN 1996-12-06 Summaries of matches played in the German first division on Friday : Bochum 2 (Stickroth 30th pen, Wosz 89th) Bayer Leverkusen 2 (Kirsten 18th, Ramelow 56th). Halftime 1-1. Attendance : 24,602 Werder Bremen 1 (Bode 85th) 1860 Munich 1 (Bormirow 8th). Halftime 0-1. Attendance 33,000 Karlsruhe 3 (Reich 29th, Carl 44th, Dundee 69th) Freiburg 0. Halftime 2-0. Attendance 33,000 Schalke 2 (Mulder 2nd and 27th) Hansa Rostock 0. Halftime 2-0. Attendance 29,300\ntype(text): \nspans: [{'start': 188, 'length': 7}, {'start': 434, 'length': 7}, {'start': 251, 'length': 13}, {'start': 93, 'length': 6}, {'start': 137, 'length': 9}, {'start': 445, 'length': 6}, {'start': 42, 'length': 4}, {'start': 202, 'length': 7}, {'start': 9, 'length': 6}, {'start': 466, 'length': 13}, {'start': 390, 'length': 8}, {'start': 377, 'length': 6}, {'start': 366, 'length': 4}, {'start': 279, 'length': 11}, {'start': 127, 'length': 6}, {'start': 168, 'length': 16}, {'start': 354, 'length': 5}, {'start': 268, 'length': 4}, {'start': 294, 'length': 8}, {'start': 341, 'length': 9}, {'start': 157, 'length': 4}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":208,\"length\":9},{\"start\":227,\"length\":18},{\"start\":115,\"length\":14},{\"start\":55,\"length\":7},{\"start\":9,\"length\":14},{\"start\":47,\"length\":6},{\"start\":160,\"length\":16},{\"start\":273,\"length\":4},{\"start\":178,\"length\":7},{\"start\":247,\"length\":6},{\"start\":192,\"length\":14},{\"start\":260,\"length\":11}],\"text\":\"TENNIS - GRAND SLAM CUP QUARTER-FINAL RESULTS. MUNICH, Germany 1996-12-06 Quarter-final results at the $ 6 million Grand Slam Cup tennis tournament on Friday : Goran Ivanisevic (Croatia) beat Mark Woodforde (Australia) 6-4 6-4 Yevgeny Kafelnikov (Russia) beat Jim Courier (U.S.) 2-6 6-4 8-6\"}'\ntext: TENNIS - GRAND SLAM CUP QUARTER-FINAL RESULTS. MUNICH, Germany 1996-12-06 Quarter-final results at the $ 6 million Grand Slam Cup tennis tournament on Friday : Goran Ivanisevic (Croatia) beat Mark Woodforde (Australia) 6-4 6-4 Yevgeny Kafelnikov (Russia) beat Jim Courier (U.S.) 2-6 6-4 8-6\ntype(text): \nspans: [{'start': 208, 'length': 9}, {'start': 227, 'length': 18}, {'start': 115, 'length': 14}, {'start': 55, 'length': 7}, {'start': 9, 'length': 14}, {'start': 47, 'length': 6}, {'start': 160, 'length': 16}, {'start': 273, 'length': 4}, {'start': 178, 'length': 7}, {'start': 247, 'length': 6}, {'start': 192, 'length': 14}, {'start': 260, 'length': 11}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":367,\"length\":8},{\"start\":679,\"length\":10},{\"start\":1281,\"length\":8},{\"start\":633,\"length\":7},{\"start\":1085,\"length\":6},{\"start\":1144,\"length\":13},{\"start\":952,\"length\":17},{\"start\":744,\"length\":15},{\"start\":405,\"length\":27},{\"start\":1124,\"length\":5},{\"start\":710,\"length\":11},{\"start\":1231,\"length\":10},{\"start\":1060,\"length\":9},{\"start\":1041,\"length\":5},{\"start\":822,\"length\":11},{\"start\":918,\"length\":14},{\"start\":944,\"length\":5},{\"start\":1007,\"length\":10},{\"start\":723,\"length\":5},{\"start\":1019,\"length\":5},{\"start\":910,\"length\":5},{\"start\":769,\"length\":16},{\"start\":89,\"length\":5},{\"start\":1209,\"length\":8},{\"start\":52,\"length\":6},{\"start\":308,\"length\":10},{\"start\":984,\"length\":5},{\"start\":691,\"length\":9},{\"start\":320,\"length\":5},{\"start\":1243,\"length\":5},{\"start\":971,\"length\":11},{\"start\":874,\"length\":16},{\"start\":1186,\"length\":7},{\"start\":70,\"length\":8},{\"start\":1167,\"length\":17},{\"start\":795,\"length\":17},{\"start\":148,\"length\":7},{\"start\":548,\"length\":7},{\"start\":1159,\"length\":5},{\"start\":216,\"length\":11},{\"start\":892,\"length\":16},{\"start\":1196,\"length\":11},{\"start\":814,\"length\":5},{\"start\":862,\"length\":9},{\"start\":1049,\"length\":9},{\"start\":617,\"length\":10},{\"start\":1251,\"length\":12},{\"start\":1113,\"length\":9},{\"start\":934,\"length\":8},{\"start\":787,\"length\":5},{\"start\":161,\"length\":9},{\"start\":1027,\"length\":12},{\"start\":761,\"length\":5},{\"start\":576,\"length\":7},{\"start\":1265,\"length\":14},{\"start\":702,\"length\":5},{\"start\":1101,\"length\":10},{\"start\":1220,\"length\":9},{\"start\":291,\"length\":8},{\"start\":454,\"length\":8},{\"start\":846,\"length\":14},{\"start\":33,\"length\":8},{\"start\":1072,\"length\":11},{\"start\":468,\"length\":8},{\"start\":1093,\"length\":5},{\"start\":112,\"length\":17},{\"start\":45,\"length\":5},{\"start\":9,\"length\":4},{\"start\":835,\"length\":8}],\"text\":\"SOCCER - WEAH HEAD-BUTT DEPRIVES PORTUGAL OF COSTA. LISBON 1996-12-06 Portugal called up Porto central defender Joao Manuel Pinto on Friday to face Germany in a World Cup qualifier in place of injured club colleague Jorge Costa, who is still nursing a broken nose after being head-butted by Liberian striker Georg Weah. Costa has not played since being struck by the AC Milan forward after a bad-tempered European Champions \\\\u0027 League game on November 27. Portugal lead European qualifying group nine with seven points from four games, one more than Ukraine and three more than Germany, who have only played twice. The Portuguese host Germany on December 14. Squad : Goalkeepers - Vitor Baia (Barcelona, Spain), Rui Correia (Braga) : Defenders - Paulinho Santos (Porto), Sergio Conceicao (Porto), Joao Manuel Pinto (Porto), Oceano Cruz (Sporting), Fernando Couto (Barcelona), Helder Cristovao (Deportivo Coruna, Spain), Dimas Teixeira (Juventus, Italy), Carlos Secretario (Real Madrid, Spain) : Midfielders - Rui Barros (Porto), Jose Barroso (Porto), Luis Figo (Barcelona), Paulo Bento (Oviedo, Spain), Jose Taira (Salamanca, Spain) : Forwards - Antonio Folha (Porto), Joao Vieira Pinto (Benfica), Paulo Alves (Sporting), Rui Costa (Fiorentina, Italy), Jorge Cadete (Celtic Glasgow, Scotland).\"}'\ntext: SOCCER - WEAH HEAD-BUTT DEPRIVES PORTUGAL OF COSTA. LISBON 1996-12-06 Portugal called up Porto central defender Joao Manuel Pinto on Friday to face Germany in a World Cup qualifier in place of injured club colleague Jorge Costa, who is still nursing a broken nose after being head-butted by Liberian striker Georg Weah. Costa has not played since being struck by the AC Milan forward after a bad-tempered European Champions ' League game on November 27. Portugal lead European qualifying group nine with seven points from four games, one more than Ukraine and three more than Germany, who have only played twice. The Portuguese host Germany on December 14. Squad : Goalkeepers - Vitor Baia (Barcelona, Spain), Rui Correia (Braga) : Defenders - Paulinho Santos (Porto), Sergio Conceicao (Porto), Joao Manuel Pinto (Porto), Oceano Cruz (Sporting), Fernando Couto (Barcelona), Helder Cristovao (Deportivo Coruna, Spain), Dimas Teixeira (Juventus, Italy), Carlos Secretario (Real Madrid, Spain) : Midfielders - Rui Barros (Porto), Jose Barroso (Porto), Luis Figo (Barcelona), Paulo Bento (Oviedo, Spain), Jose Taira (Salamanca, Spain) : Forwards - Antonio Folha (Porto), Joao Vieira Pinto (Benfica), Paulo Alves (Sporting), Rui Costa (Fiorentina, Italy), Jorge Cadete (Celtic Glasgow, Scotland).\ntype(text): \nspans: [{'start': 367, 'length': 8}, {'start': 679, 'length': 10}, {'start': 1281, 'length': 8}, {'start': 633, 'length': 7}, {'start': 1085, 'length': 6}, {'start': 1144, 'length': 13}, {'start': 952, 'length': 17}, {'start': 744, 'length': 15}, {'start': 405, 'length': 27}, {'start': 1124, 'length': 5}, {'start': 710, 'length': 11}, {'start': 1231, 'length': 10}, {'start': 1060, 'length': 9}, {'start': 1041, 'length': 5}, {'start': 822, 'length': 11}, {'start': 918, 'length': 14}, {'start': 944, 'length': 5}, {'start': 1007, 'length': 10}, {'start': 723, 'length': 5}, {'start': 1019, 'length': 5}, {'start': 910, 'length': 5}, {'start': 769, 'length': 16}, {'start': 89, 'length': 5}, {'start': 1209, 'length': 8}, {'start': 52, 'length': 6}, {'start': 308, 'length': 10}, {'start': 984, 'length': 5}, {'start': 691, 'length': 9}, {'start': 320, 'length': 5}, {'start': 1243, 'length': 5}, {'start': 971, 'length': 11}, {'start': 874, 'length': 16}, {'start': 1186, 'length': 7}, {'start': 70, 'length': 8}, {'start': 1167, 'length': 17}, {'start': 795, 'length': 17}, {'start': 148, 'length': 7}, {'start': 548, 'length': 7}, {'start': 1159, 'length': 5}, {'start': 216, 'length': 11}, {'start': 892, 'length': 16}, {'start': 1196, 'length': 11}, {'start': 814, 'length': 5}, {'start': 862, 'length': 9}, {'start': 1049, 'length': 9}, {'start': 617, 'length': 10}, {'start': 1251, 'length': 12}, {'start': 1113, 'length': 9}, {'start': 934, 'length': 8}, {'start': 787, 'length': 5}, {'start': 161, 'length': 9}, {'start': 1027, 'length': 12}, {'start': 761, 'length': 5}, {'start': 576, 'length': 7}, {'start': 1265, 'length': 14}, {'start': 702, 'length': 5}, {'start': 1101, 'length': 10}, {'start': 1220, 'length': 9}, {'start': 291, 'length': 8}, {'start': 454, 'length': 8}, {'start': 846, 'length': 14}, {'start': 33, 'length': 8}, {'start': 1072, 'length': 11}, {'start': 468, 'length': 8}, {'start': 1093, 'length': 5}, {'start': 112, 'length': 17}, {'start': 45, 'length': 5}, {'start': 9, 'length': 4}, {'start': 835, 'length': 8}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":979,\"length\":14},{\"start\":568,\"length\":11},{\"start\":1178,\"length\":17},{\"start\":1102,\"length\":10},{\"start\":1061,\"length\":15},{\"start\":257,\"length\":9},{\"start\":1201,\"length\":13},{\"start\":1278,\"length\":12},{\"start\":1240,\"length\":13},{\"start\":1136,\"length\":13},{\"start\":1296,\"length\":13},{\"start\":517,\"length\":11},{\"start\":1155,\"length\":17},{\"start\":1021,\"length\":14},{\"start\":902,\"length\":14},{\"start\":798,\"length\":16},{\"start\":744,\"length\":11},{\"start\":851,\"length\":10},{\"start\":142,\"length\":11},{\"start\":786,\"length\":7},{\"start\":1220,\"length\":14},{\"start\":660,\"length\":11},{\"start\":819,\"length\":9},{\"start\":1041,\"length\":14},{\"start\":52,\"length\":6},{\"start\":833,\"length\":13},{\"start\":770,\"length\":11},{\"start\":922,\"length\":14},{\"start\":70,\"length\":12},{\"start\":867,\"length\":16},{\"start\":183,\"length\":11},{\"start\":599,\"length\":11},{\"start\":942,\"length\":13},{\"start\":41,\"length\":9},{\"start\":696,\"length\":9},{\"start\":706,\"length\":9},{\"start\":546,\"length\":11},{\"start\":756,\"length\":9},{\"start\":105,\"length\":7},{\"start\":580,\"length\":9},{\"start\":961,\"length\":12},{\"start\":999,\"length\":16},{\"start\":1082,\"length\":14},{\"start\":1259,\"length\":13},{\"start\":238,\"length\":11},{\"start\":635,\"length\":9},{\"start\":529,\"length\":11},{\"start\":201,\"length\":9},{\"start\":1118,\"length\":12},{\"start\":158,\"length\":9},{\"start\":888,\"length\":8},{\"start\":650,\"length\":9},{\"start\":677,\"length\":9},{\"start\":27,\"length\":11}],\"text\":\"SOCCER SHOWCASE-BETTING ON REAL MADRID V BARCELONA. MADRID 1996-12-06 William Hill betting on Saturday\\\\u0027s Spanish first division match between Real Madrid and Barcelona : To win : 6-5 Real Madrid ; 7-4 Barcelona Draw : 9-4 Correct score : Real Madrid to win Barcelona to win 1-0 13-2 1-0 15-2 2-0 9-1 2-0 12-1 2-1 8-1 2-1 10-1 3-0 20-1 3-0 28-1 3-1 16-1 3-1 22-1 3-2 25-1 3-2 25-1 4-0 50-1 4-0 100-1 4-1 40-1 4-1 80-1 4-2 50-1 4-2 80-1 Draw : 0-0 8-1 1-1 11-2 2-2 14-1 3-3 50-1 Double result : half-time full-time 5-2 Real Madrid Real Madrid 14-1 Real Madrid Draw 28-1 Real Madrid Barcelona 5-1 Draw Real Madrid 4-1 Draw Draw 11-2 Draw Barcelona 25-1 Barcelona Real Madrid 14-1 Barcelona Draw 4-1 Barcelona Barcelona First goalscorer of match : Real Madrid Barcelona 9-2 Davor Suker 9-2 Ronaldo 5-1 Pedrag Mijatovic 7-1 Luis Figo 7-1 Raul Gonzalez 7-1 Juan Pizzi 12-1 Fernando Redondo 9-1 Giovanni 14-1 Victor Sanchez 12-1 Guillermo Amor 16-1 Jose Amavisca 14-1 Roger Garcia 16-1 Manolo Sanchis 14-1 Gheorghe Popescu 16-1 Roberto Carlos 16-1 JosepGuardiola 20-1 Fernando Hierro 20-1 Ivan de laPena 20-1 Luis Milla 25-1 Luis Enrique 33-1 Fernando Sanz 25-1 AbelardoFernandez 40-1 Carlos Secretario 28-1 Sergi Barjuan 40-1 Rafael Alkorta 33-1 Albert Ferrer 40-1 Chendo Porlan 33-1 Miguel Nadal 40-1 Laurent Blanc\"}'\ntext: SOCCER SHOWCASE-BETTING ON REAL MADRID V BARCELONA. MADRID 1996-12-06 William Hill betting on Saturday's Spanish first division match between Real Madrid and Barcelona : To win : 6-5 Real Madrid ; 7-4 Barcelona Draw : 9-4 Correct score : Real Madrid to win Barcelona to win 1-0 13-2 1-0 15-2 2-0 9-1 2-0 12-1 2-1 8-1 2-1 10-1 3-0 20-1 3-0 28-1 3-1 16-1 3-1 22-1 3-2 25-1 3-2 25-1 4-0 50-1 4-0 100-1 4-1 40-1 4-1 80-1 4-2 50-1 4-2 80-1 Draw : 0-0 8-1 1-1 11-2 2-2 14-1 3-3 50-1 Double result : half-time full-time 5-2 Real Madrid Real Madrid 14-1 Real Madrid Draw 28-1 Real Madrid Barcelona 5-1 Draw Real Madrid 4-1 Draw Draw 11-2 Draw Barcelona 25-1 Barcelona Real Madrid 14-1 Barcelona Draw 4-1 Barcelona Barcelona First goalscorer of match : Real Madrid Barcelona 9-2 Davor Suker 9-2 Ronaldo 5-1 Pedrag Mijatovic 7-1 Luis Figo 7-1 Raul Gonzalez 7-1 Juan Pizzi 12-1 Fernando Redondo 9-1 Giovanni 14-1 Victor Sanchez 12-1 Guillermo Amor 16-1 Jose Amavisca 14-1 Roger Garcia 16-1 Manolo Sanchis 14-1 Gheorghe Popescu 16-1 Roberto Carlos 16-1 JosepGuardiola 20-1 Fernando Hierro 20-1 Ivan de laPena 20-1 Luis Milla 25-1 Luis Enrique 33-1 Fernando Sanz 25-1 AbelardoFernandez 40-1 Carlos Secretario 28-1 Sergi Barjuan 40-1 Rafael Alkorta 33-1 Albert Ferrer 40-1 Chendo Porlan 33-1 Miguel Nadal 40-1 Laurent Blanc\ntype(text): \nspans: [{'start': 979, 'length': 14}, {'start': 568, 'length': 11}, {'start': 1178, 'length': 17}, {'start': 1102, 'length': 10}, {'start': 1061, 'length': 15}, {'start': 257, 'length': 9}, {'start': 1201, 'length': 13}, {'start': 1278, 'length': 12}, {'start': 1240, 'length': 13}, {'start': 1136, 'length': 13}, {'start': 1296, 'length': 13}, {'start': 517, 'length': 11}, {'start': 1155, 'length': 17}, {'start': 1021, 'length': 14}, {'start': 902, 'length': 14}, {'start': 798, 'length': 16}, {'start': 744, 'length': 11}, {'start': 851, 'length': 10}, {'start': 142, 'length': 11}, {'start': 786, 'length': 7}, {'start': 1220, 'length': 14}, {'start': 660, 'length': 11}, {'start': 819, 'length': 9}, {'start': 1041, 'length': 14}, {'start': 52, 'length': 6}, {'start': 833, 'length': 13}, {'start': 770, 'length': 11}, {'start': 922, 'length': 14}, {'start': 70, 'length': 12}, {'start': 867, 'length': 16}, {'start': 183, 'length': 11}, {'start': 599, 'length': 11}, {'start': 942, 'length': 13}, {'start': 41, 'length': 9}, {'start': 696, 'length': 9}, {'start': 706, 'length': 9}, {'start': 546, 'length': 11}, {'start': 756, 'length': 9}, {'start': 105, 'length': 7}, {'start': 580, 'length': 9}, {'start': 961, 'length': 12}, {'start': 999, 'length': 16}, {'start': 1082, 'length': 14}, {'start': 1259, 'length': 13}, {'start': 238, 'length': 11}, {'start': 635, 'length': 9}, {'start': 529, 'length': 11}, {'start': 201, 'length': 9}, {'start': 1118, 'length': 12}, {'start': 158, 'length': 9}, {'start': 888, 'length': 8}, {'start': 650, 'length': 9}, {'start': 677, 'length': 9}, {'start': 27, 'length': 11}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":204,\"length\":21},{\"start\":249,\"length\":8},{\"start\":76,\"length\":7},{\"start\":58,\"length\":6},{\"start\":236,\"length\":6},{\"start\":133,\"length\":25},{\"start\":356,\"length\":7}],\"text\":\"SOCCER SHOWCASE-FANS FACE BREATHALYSER TESTS, PAPER SAYS. MADRID 1996-12-06 Spanish police will breathalyse fans at the gates of the Santiago Bernabeu stadium and ban drunk supporters from Saturday\\\\u0027s big Real Madrid-Barcelona game, the Madrid daily El Mundo said on Friday. Although there are no known precedents in the country, the action is envisaged in Spanish legislation governing sports events. Tickets for the game stipulate that supporters will be barred if they are \\\\\"under the effects of alcohol\\\\\".\"}'\ntext: SOCCER SHOWCASE-FANS FACE BREATHALYSER TESTS, PAPER SAYS. MADRID 1996-12-06 Spanish police will breathalyse fans at the gates of the Santiago Bernabeu stadium and ban drunk supporters from Saturday's big Real Madrid-Barcelona game, the Madrid daily El Mundo said on Friday. Although there are no known precedents in the country, the action is envisaged in Spanish legislation governing sports events. Tickets for the game stipulate that supporters will be barred if they are \"under the effects of alcohol\".\ntype(text): \nspans: [{'start': 204, 'length': 21}, {'start': 249, 'length': 8}, {'start': 76, 'length': 7}, {'start': 58, 'length': 6}, {'start': 236, 'length': 6}, {'start': 133, 'length': 25}, {'start': 356, 'length': 7}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":304,\"length\":10},{\"start\":834,\"length\":11},{\"start\":210,\"length\":11},{\"start\":752,\"length\":7},{\"start\":270,\"length\":16},{\"start\":9,\"length\":7},{\"start\":497,\"length\":14},{\"start\":807,\"length\":8},{\"start\":43,\"length\":6},{\"start\":665,\"length\":8},{\"start\":725,\"length\":8},{\"start\":530,\"length\":8},{\"start\":401,\"length\":13},{\"start\":557,\"length\":10},{\"start\":640,\"length\":6},{\"start\":241,\"length\":9},{\"start\":78,\"length\":7},{\"start\":367,\"length\":15},{\"start\":613,\"length\":8},{\"start\":698,\"length\":8},{\"start\":586,\"length\":8},{\"start\":433,\"length\":10},{\"start\":333,\"length\":15},{\"start\":462,\"length\":16},{\"start\":778,\"length\":10}],\"text\":\"SOCCER - SPANISH FIRST DIVISION STANDINGS. MADRID 1996-12-06 Standings in the Spanish first division ahead of this weekend\\\\u0027s games. (tabulate under games played, won, drawn, lost, goals for, against, points) : Real Madrid 15 10 5 0 31 12 35 Barcelona 15 10 4 1 46 19 34 Deportivo Coruna 15 9 6 0 23 7 33 Real Betis 15 8 5 2 28 13 29 Atletico Madrid 15 8 3 4 26 17 27 Athletic Bilbao 15 7 4 4 28 22 25 Real Sociedad 15 7 3 5 20 18 24 Valladolid 15 7 3 5 19 18 24 Racing Santander 15 5 7 3 15 15 22 Rayo Vallecano 15 5 5 5 21 19 20 Valencia 15 6 2 7 23 22 20 Celta Vigo 15 5 5 5 17 17 20 Tenerife 15 5 4 6 23 17 19 Espanyol 15 4 4 7 17 20 16 Oviedo 15 4 4 7 17 21 16 Sporting Gijon 15 4 4 7 15 22 16 Logrones 15 4 3 8 11 33 15 Zaragoza 15 2 8 5 18 23 14 Sevilla 15 4 2 9 13 20 14 Compostela 15 3 4 8 13 28 13 Hercules 15 2 2 11 11 29 8 Extremadura 15 1 3 11 8 30 6\"}'\ntext: SOCCER - SPANISH FIRST DIVISION STANDINGS. MADRID 1996-12-06 Standings in the Spanish first division ahead of this weekend's games. (tabulate under games played, won, drawn, lost, goals for, against, points) : Real Madrid 15 10 5 0 31 12 35 Barcelona 15 10 4 1 46 19 34 Deportivo Coruna 15 9 6 0 23 7 33 Real Betis 15 8 5 2 28 13 29 Atletico Madrid 15 8 3 4 26 17 27 Athletic Bilbao 15 7 4 4 28 22 25 Real Sociedad 15 7 3 5 20 18 24 Valladolid 15 7 3 5 19 18 24 Racing Santander 15 5 7 3 15 15 22 Rayo Vallecano 15 5 5 5 21 19 20 Valencia 15 6 2 7 23 22 20 Celta Vigo 15 5 5 5 17 17 20 Tenerife 15 5 4 6 23 17 19 Espanyol 15 4 4 7 17 20 16 Oviedo 15 4 4 7 17 21 16 Sporting Gijon 15 4 4 7 15 22 16 Logrones 15 4 3 8 11 33 15 Zaragoza 15 2 8 5 18 23 14 Sevilla 15 4 2 9 13 20 14 Compostela 15 3 4 8 13 28 13 Hercules 15 2 2 11 11 29 8 Extremadura 15 1 3 11 8 30 6\ntype(text): \nspans: [{'start': 304, 'length': 10}, {'start': 834, 'length': 11}, {'start': 210, 'length': 11}, {'start': 752, 'length': 7}, {'start': 270, 'length': 16}, {'start': 9, 'length': 7}, {'start': 497, 'length': 14}, {'start': 807, 'length': 8}, {'start': 43, 'length': 6}, {'start': 665, 'length': 8}, {'start': 725, 'length': 8}, {'start': 530, 'length': 8}, {'start': 401, 'length': 13}, {'start': 557, 'length': 10}, {'start': 640, 'length': 6}, {'start': 241, 'length': 9}, {'start': 78, 'length': 7}, {'start': 367, 'length': 15}, {'start': 613, 'length': 8}, {'start': 698, 'length': 8}, {'start': 586, 'length': 8}, {'start': 433, 'length': 10}, {'start': 333, 'length': 15}, {'start': 462, 'length': 16}, {'start': 778, 'length': 10}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":9,\"length\":5},{\"start\":88,\"length\":15},{\"start\":298,\"length\":7},{\"start\":327,\"length\":15},{\"start\":216,\"length\":10},{\"start\":123,\"length\":16},{\"start\":353,\"length\":18},{\"start\":58,\"length\":6},{\"start\":188,\"length\":9},{\"start\":41,\"length\":9},{\"start\":151,\"length\":15},{\"start\":76,\"length\":5},{\"start\":29,\"length\":7}],\"text\":\"SOCCER - SPAIN PICK UNCAPPED ARMANDO FOR WORLD CUP CLASH. MADRID 1996-12-06 Spain coach Javier Clemente has added uncapped Deportivo Coruna midfielder Armando Alvarez to his squad for the World Cup qualifier against Yugoslavia on December 14. \\\\\"I do n\\\\u0027t believe it... I thought it was a joke,\\\\\" said Armando who replaces injured Atletico Madrid playmaker Jose Luis Caminero.\"}'\ntext: SOCCER - SPAIN PICK UNCAPPED ARMANDO FOR WORLD CUP CLASH. MADRID 1996-12-06 Spain coach Javier Clemente has added uncapped Deportivo Coruna midfielder Armando Alvarez to his squad for the World Cup qualifier against Yugoslavia on December 14. \"I do n't believe it... I thought it was a joke,\" said Armando who replaces injured Atletico Madrid playmaker Jose Luis Caminero.\ntype(text): \nspans: [{'start': 9, 'length': 5}, {'start': 88, 'length': 15}, {'start': 298, 'length': 7}, {'start': 327, 'length': 15}, {'start': 216, 'length': 10}, {'start': 123, 'length': 16}, {'start': 353, 'length': 18}, {'start': 58, 'length': 6}, {'start': 188, 'length': 9}, {'start': 41, 'length': 9}, {'start': 151, 'length': 15}, {'start': 76, 'length': 5}, {'start': 29, 'length': 7}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":1401,\"length\":13},{\"start\":389,\"length\":5},{\"start\":836,\"length\":8},{\"start\":1257,\"length\":4},{\"start\":1341,\"length\":6},{\"start\":1328,\"length\":9},{\"start\":45,\"length\":4},{\"start\":75,\"length\":14},{\"start\":236,\"length\":11},{\"start\":958,\"length\":27},{\"start\":9,\"length\":4},{\"start\":1150,\"length\":5},{\"start\":435,\"length\":8},{\"start\":830,\"length\":4},{\"start\":338,\"length\":4},{\"start\":1301,\"length\":5},{\"start\":133,\"length\":8},{\"start\":39,\"length\":4},{\"start\":1039,\"length\":5},{\"start\":416,\"length\":4},{\"start\":518,\"length\":9},{\"start\":274,\"length\":7},{\"start\":1114,\"length\":5},{\"start\":19,\"length\":9},{\"start\":292,\"length\":20},{\"start\":724,\"length\":5},{\"start\":142,\"length\":11},{\"start\":222,\"length\":5},{\"start\":1263,\"length\":4},{\"start\":1008,\"length\":4},{\"start\":794,\"length\":4},{\"start\":1381,\"length\":8},{\"start\":694,\"length\":6},{\"start\":61,\"length\":4}],\"text\":\"SOCCER - FIFA BOSS HAVELANGE STANDS BY WEAH. ROME 1996-12-06 FIFA chairman Joao Havelange said on Friday he would personally present AC Milan George Weah with world soccer\\\\u0027s fair play award despite the striker\\\\u0027s attack on Porto captain Jorge Costa. In an interview with the Italian newspaper Gazzetta dello Sport, he was quoted as saying Weah had been provoked into the assault which left Costa with a broken nose. \\\\\"FIFA has named the Liberian for its 1996 Fair Play award and it is not going to change its decision,\\\\\" Havelange said. \\\\\"A reaction, provoked, cannot erase 10 years of loyalty everywhere and in every competition.\\\\\" I will be happy to give him the award personally on January 20 in Lisbon and I\\\\u0027m confident that Costa himself will be there beside me on that day to shake his hand. \\\\\"Weah was suspended for one match by UEFA, European soccer\\\\u0027s governing body, pending a fuller investigation. The incident took place in the players \\\\u0027 tunnel after a European Champions \\\\u0027 League match on November 20. Weah has admitted head butting Costa but said he reacted to racist taunts. He has offered to apologise if Costa acknowledges the provocation. Costa, who needed surgery on his nose, has not accepted the offer and was reported to be considering suing Weah. Weah served out his suspension during Milan\\\\u0027s 2-1 home defeat by Rosenborg of Norway on Wednesday. The defeat put the Italians out of the Europoean Cup.\"}'\ntext: SOCCER - FIFA BOSS HAVELANGE STANDS BY WEAH. ROME 1996-12-06 FIFA chairman Joao Havelange said on Friday he would personally present AC Milan George Weah with world soccer's fair play award despite the striker's attack on Porto captain Jorge Costa. In an interview with the Italian newspaper Gazzetta dello Sport, he was quoted as saying Weah had been provoked into the assault which left Costa with a broken nose. \"FIFA has named the Liberian for its 1996 Fair Play award and it is not going to change its decision,\" Havelange said. \"A reaction, provoked, cannot erase 10 years of loyalty everywhere and in every competition.\" I will be happy to give him the award personally on January 20 in Lisbon and I'm confident that Costa himself will be there beside me on that day to shake his hand. \"Weah was suspended for one match by UEFA, European soccer's governing body, pending a fuller investigation. The incident took place in the players ' tunnel after a European Champions ' League match on November 20. Weah has admitted head butting Costa but said he reacted to racist taunts. He has offered to apologise if Costa acknowledges the provocation. Costa, who needed surgery on his nose, has not accepted the offer and was reported to be considering suing Weah. Weah served out his suspension during Milan's 2-1 home defeat by Rosenborg of Norway on Wednesday. The defeat put the Italians out of the Europoean Cup.\ntype(text): \nspans: [{'start': 1401, 'length': 13}, {'start': 389, 'length': 5}, {'start': 836, 'length': 8}, {'start': 1257, 'length': 4}, {'start': 1341, 'length': 6}, {'start': 1328, 'length': 9}, {'start': 45, 'length': 4}, {'start': 75, 'length': 14}, {'start': 236, 'length': 11}, {'start': 958, 'length': 27}, {'start': 9, 'length': 4}, {'start': 1150, 'length': 5}, {'start': 435, 'length': 8}, {'start': 830, 'length': 4}, {'start': 338, 'length': 4}, {'start': 1301, 'length': 5}, {'start': 133, 'length': 8}, {'start': 39, 'length': 4}, {'start': 1039, 'length': 5}, {'start': 416, 'length': 4}, {'start': 518, 'length': 9}, {'start': 274, 'length': 7}, {'start': 1114, 'length': 5}, {'start': 19, 'length': 9}, {'start': 292, 'length': 20}, {'start': 724, 'length': 5}, {'start': 142, 'length': 11}, {'start': 222, 'length': 5}, {'start': 1263, 'length': 4}, {'start': 1008, 'length': 4}, {'start': 794, 'length': 4}, {'start': 1381, 'length': 8}, {'start': 694, 'length': 6}, {'start': 61, 'length': 4}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":409,\"length\":7},{\"start\":17,\"length\":17},{\"start\":74,\"length\":17},{\"start\":52,\"length\":6},{\"start\":43,\"length\":7},{\"start\":299,\"length\":7},{\"start\":442,\"length\":12},{\"start\":248,\"length\":8},{\"start\":179,\"length\":8},{\"start\":220,\"length\":7},{\"start\":540,\"length\":17}],\"text\":\"GUNMEN WOUND TWO MANCHESTER UNITED FANS IN AUSTRIA. VIENNA 1996-12-06 Two Manchester United soccer fans were wounded by unidentified gunmen on Friday and taken to hospital in the Austrian capital, police said. \\\\\"The four Britons were shot at from a Mercedes car at around 1 a.m.,\\\\\" a spokeswoman told Reuters. The two men were hit in the pelvis and leg. Police said their lives were not in danger. The fans, in Austria to watch their team play Rapid Vienna last Wednesday, may have been involved in a pub brawl earlier, the spokeswoman said. Manchester United won 2-0.\"}'\ntext: GUNMEN WOUND TWO MANCHESTER UNITED FANS IN AUSTRIA. VIENNA 1996-12-06 Two Manchester United soccer fans were wounded by unidentified gunmen on Friday and taken to hospital in the Austrian capital, police said. \"The four Britons were shot at from a Mercedes car at around 1 a.m.,\" a spokeswoman told Reuters. The two men were hit in the pelvis and leg. Police said their lives were not in danger. The fans, in Austria to watch their team play Rapid Vienna last Wednesday, may have been involved in a pub brawl earlier, the spokeswoman said. Manchester United won 2-0.\ntype(text): \nspans: [{'start': 409, 'length': 7}, {'start': 17, 'length': 17}, {'start': 74, 'length': 17}, {'start': 52, 'length': 6}, {'start': 43, 'length': 7}, {'start': 299, 'length': 7}, {'start': 442, 'length': 12}, {'start': 248, 'length': 8}, {'start': 179, 'length': 8}, {'start': 220, 'length': 7}, {'start': 540, 'length': 17}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":1855,\"length\":5},{\"start\":1367,\"length\":12},{\"start\":3301,\"length\":6},{\"start\":1463,\"length\":10},{\"start\":2280,\"length\":5},{\"start\":1393,\"length\":11},{\"start\":2105,\"length\":6},{\"start\":2903,\"length\":15},{\"start\":2939,\"length\":6},{\"start\":190,\"length\":8},{\"start\":1475,\"length\":16},{\"start\":1724,\"length\":11},{\"start\":1227,\"length\":4},{\"start\":2593,\"length\":15},{\"start\":1946,\"length\":6},{\"start\":2466,\"length\":12},{\"start\":1115,\"length\":12},{\"start\":1642,\"length\":9},{\"start\":1523,\"length\":8},{\"start\":169,\"length\":3},{\"start\":1453,\"length\":9},{\"start\":1869,\"length\":7},{\"start\":1812,\"length\":15},{\"start\":1142,\"length\":7},{\"start\":2218,\"length\":5},{\"start\":1067,\"length\":5},{\"start\":3096,\"length\":7},{\"start\":9,\"length\":7},{\"start\":3317,\"length\":15},{\"start\":2297,\"length\":12},{\"start\":842,\"length\":10},{\"start\":825,\"length\":7},{\"start\":365,\"length\":14},{\"start\":1044,\"length\":7},{\"start\":298,\"length\":7},{\"start\":2440,\"length\":5},{\"start\":2090,\"length\":13},{\"start\":357,\"length\":7},{\"start\":1538,\"length\":7},{\"start\":1608,\"length\":18},{\"start\":3122,\"length\":13},{\"start\":3112,\"length\":9},{\"start\":706,\"length\":10},{\"start\":3361,\"length\":13},{\"start\":1129,\"length\":5},{\"start\":1779,\"length\":14},{\"start\":1971,\"length\":6},{\"start\":2724,\"length\":10},{\"start\":807,\"length\":10},{\"start\":1560,\"length\":5},{\"start\":1412,\"length\":4},{\"start\":1879,\"length\":6},{\"start\":1707,\"length\":8},{\"start\":2757,\"length\":20},{\"start\":785,\"length\":20},{\"start\":2071,\"length\":9},{\"start\":3345,\"length\":7},{\"start\":1894,\"length\":15},{\"start\":2782,\"length\":5},{\"start\":553,\"length\":8},{\"start\":2538,\"length\":14},{\"start\":1290,\"length\":5},{\"start\":1081,\"length\":13},{\"start\":2744,\"length\":5},{\"start\":905,\"length\":19},{\"start\":1496,\"length\":15},{\"start\":330,\"length\":5},{\"start\":534,\"length\":8},{\"start\":3191,\"length\":5},{\"start\":176,\"length\":7},{\"start\":496,\"length\":8},{\"start\":2202,\"length\":14},{\"start\":2528,\"length\":9},{\"start\":1800,\"length\":11},{\"start\":2448,\"length\":6},{\"start\":2873,\"length\":9},{\"start\":730,\"length\":8},{\"start\":1267,\"length\":7},{\"start\":518,\"length\":8},{\"start\":2025,\"length\":6},{\"start\":778,\"length\":6},{\"start\":1350,\"length\":5},{\"start\":624,\"length\":6},{\"start\":54,\"length\":4},{\"start\":2788,\"length\":11},{\"start\":1279,\"length\":9},{\"start\":1933,\"length\":6},{\"start\":2682,\"length\":8},{\"start\":2231,\"length\":9},{\"start\":1657,\"length\":9},{\"start\":251,\"length\":7},{\"start\":70,\"length\":7},{\"start\":228,\"length\":7},{\"start\":336,\"length\":16},{\"start\":2251,\"length\":5},{\"start\":2414,\"length\":8},{\"start\":2972,\"length\":7},{\"start\":439,\"length\":8},{\"start\":1763,\"length\":5},{\"start\":960,\"length\":15},{\"start\":2633,\"length\":9},{\"start\":1329,\"length\":19},{\"start\":2957,\"length\":13},{\"start\":691,\"length\":5},{\"start\":2986,\"length\":14},{\"start\":1170,\"length\":5},{\"start\":635,\"length\":5},{\"start\":2576,\"length\":8},{\"start\":3269,\"length\":5},{\"start\":78,\"length\":7},{\"start\":391,\"length\":7},{\"start\":2649,\"length\":8},{\"start\":2492,\"length\":15},{\"start\":929,\"length\":15}],\"text\":\"SOCCER - ITALIAN FIRST DIVISION MATCHES THIS WEEKEND. ROME 1996-12-06 Italian Serie A games to be played on Sunday (league positions in parentheses, all kick- off times GMT) : Bologna (4) v Piacenza (13) 1330 Along with leaders Vicenza, fourth-placed Bologna represent the biggest surprise of this Italian autumn. Led as usual by Swede Kennet Andersson and Russian Igor Kolyvanov in attack, Bologna can expect a tough home match against a Piacenza side still exultant after a 3-2 league win over AC Milan last Sunday. Cagliari (16) v Reggiana (18) 1530 Cagliari start favourite in this relegation scrap following draws with Napoli and Inter in last two outings but will be without suspended Swiss defender Ramon Vega. Bottom team Reggiana are also without a suspended defender, German Dietmar Beiersdorfer. Fiorentina (10) v Perugia (8) 1330 Fiorentina will be without three suspended players-- defenders Daniele Carnasciali and Lorenzo Amoruso and midfielder Emiliano Bigica-- for a difficult home match against unpredictable, attack-oriented Perugia led by in-form Croat striker Milan Rapajic and the experienced Fausto Pizzi. Lazio (12) v AS Roma (7) 1930 Poor man\\\\u0027s Roman derby in what has been a miserable season for both Rome teams, already eliminated from the Italian and UEFA Cups. Lazio have injury doubts about striker Pierluigi Casiraghi, Czech midfielder Pavel Nedved and defender Paolo Negro, while Roma present a full strength side led by Argentine Abel Balbo, Marco Delvecchio and Francesco Totti in attack. AC Milan (9) v Udinese (11) 1330 Can Milan sink any further? Following their midweek Champions \\\\u0027 League elimination by Norwegian side Rosenborg, a morale-boosting win is badly needed. Liberian striker George Weah makes a welcome return for Milan alongside Roberto Baggio, with Montenegrin Dejan Savicevic in midfield. Good news for Milan is that Udinese\\\\u0027s German striker Oliver Bierhoff is out through injury. Napoli (5) v Verona (17) 1330 In-form Napoli should prove too strong for second from bottom Verona despite the absence of their suspended Argentine defender Roberto Ayala. Verona\\\\u0027s slim chances have been further reduced by a knee injury to their experienced midfielder Eugenio Corini. Parma (14) v Atlalanta (15) 1330 Parma may field new signing, Croat midfielder Mario Stanic, in an attempt to lift a miserable season which has seen them go without a win since a 1-0 triumph over Cagliari eight weeks ago. Parma\\\\u0027s French midfielder Daniel Bravo and defender Fabio Cannavaro are suspended while Argentine Nestor Sensini is out through injury. Atalanta look to Filippo Inzaghi, scorer of eight goals. Sampdoria (6) v Juventus (3) 1330 All-conquering Juventus field their most recent signing, Portuguese defender Dimas, while Alessandro Del Piero and Croat Alen Boksic lead the attack. The new world club champions may prove too strong for a Sampdoria side led by captain Roberto Mancini but missing injured French midfielder Pierre Laigle. Vicenza (1) v Internazionale (2) 1330 Not exactly a clash of the titans but an intriuguing match nonetheless. Full strength Vicenza, led by Uruguayan Marcelo Otero, may continue their surprise run at the top against an Inter side that has been less than impressive in three successive home draws. Inter will be without suspended French defender Joceyln Angloma and injured Chilean striker Ivan Zamorano.\"}'\ntext: SOCCER - ITALIAN FIRST DIVISION MATCHES THIS WEEKEND. ROME 1996-12-06 Italian Serie A games to be played on Sunday (league positions in parentheses, all kick- off times GMT) : Bologna (4) v Piacenza (13) 1330 Along with leaders Vicenza, fourth-placed Bologna represent the biggest surprise of this Italian autumn. Led as usual by Swede Kennet Andersson and Russian Igor Kolyvanov in attack, Bologna can expect a tough home match against a Piacenza side still exultant after a 3-2 league win over AC Milan last Sunday. Cagliari (16) v Reggiana (18) 1530 Cagliari start favourite in this relegation scrap following draws with Napoli and Inter in last two outings but will be without suspended Swiss defender Ramon Vega. Bottom team Reggiana are also without a suspended defender, German Dietmar Beiersdorfer. Fiorentina (10) v Perugia (8) 1330 Fiorentina will be without three suspended players-- defenders Daniele Carnasciali and Lorenzo Amoruso and midfielder Emiliano Bigica-- for a difficult home match against unpredictable, attack-oriented Perugia led by in-form Croat striker Milan Rapajic and the experienced Fausto Pizzi. Lazio (12) v AS Roma (7) 1930 Poor man's Roman derby in what has been a miserable season for both Rome teams, already eliminated from the Italian and UEFA Cups. Lazio have injury doubts about striker Pierluigi Casiraghi, Czech midfielder Pavel Nedved and defender Paolo Negro, while Roma present a full strength side led by Argentine Abel Balbo, Marco Delvecchio and Francesco Totti in attack. AC Milan (9) v Udinese (11) 1330 Can Milan sink any further? Following their midweek Champions ' League elimination by Norwegian side Rosenborg, a morale-boosting win is badly needed. Liberian striker George Weah makes a welcome return for Milan alongside Roberto Baggio, with Montenegrin Dejan Savicevic in midfield. Good news for Milan is that Udinese's German striker Oliver Bierhoff is out through injury. Napoli (5) v Verona (17) 1330 In-form Napoli should prove too strong for second from bottom Verona despite the absence of their suspended Argentine defender Roberto Ayala. Verona's slim chances have been further reduced by a knee injury to their experienced midfielder Eugenio Corini. Parma (14) v Atlalanta (15) 1330 Parma may field new signing, Croat midfielder Mario Stanic, in an attempt to lift a miserable season which has seen them go without a win since a 1-0 triumph over Cagliari eight weeks ago. Parma's French midfielder Daniel Bravo and defender Fabio Cannavaro are suspended while Argentine Nestor Sensini is out through injury. Atalanta look to Filippo Inzaghi, scorer of eight goals. Sampdoria (6) v Juventus (3) 1330 All-conquering Juventus field their most recent signing, Portuguese defender Dimas, while Alessandro Del Piero and Croat Alen Boksic lead the attack. The new world club champions may prove too strong for a Sampdoria side led by captain Roberto Mancini but missing injured French midfielder Pierre Laigle. Vicenza (1) v Internazionale (2) 1330 Not exactly a clash of the titans but an intriuguing match nonetheless. Full strength Vicenza, led by Uruguayan Marcelo Otero, may continue their surprise run at the top against an Inter side that has been less than impressive in three successive home draws. Inter will be without suspended French defender Joceyln Angloma and injured Chilean striker Ivan Zamorano.\ntype(text): \nspans: [{'start': 1855, 'length': 5}, {'start': 1367, 'length': 12}, {'start': 3301, 'length': 6}, {'start': 1463, 'length': 10}, {'start': 2280, 'length': 5}, {'start': 1393, 'length': 11}, {'start': 2105, 'length': 6}, {'start': 2903, 'length': 15}, {'start': 2939, 'length': 6}, {'start': 190, 'length': 8}, {'start': 1475, 'length': 16}, {'start': 1724, 'length': 11}, {'start': 1227, 'length': 4}, {'start': 2593, 'length': 15}, {'start': 1946, 'length': 6}, {'start': 2466, 'length': 12}, {'start': 1115, 'length': 12}, {'start': 1642, 'length': 9}, {'start': 1523, 'length': 8}, {'start': 169, 'length': 3}, {'start': 1453, 'length': 9}, {'start': 1869, 'length': 7}, {'start': 1812, 'length': 15}, {'start': 1142, 'length': 7}, {'start': 2218, 'length': 5}, {'start': 1067, 'length': 5}, {'start': 3096, 'length': 7}, {'start': 9, 'length': 7}, {'start': 3317, 'length': 15}, {'start': 2297, 'length': 12}, {'start': 842, 'length': 10}, {'start': 825, 'length': 7}, {'start': 365, 'length': 14}, {'start': 1044, 'length': 7}, {'start': 298, 'length': 7}, {'start': 2440, 'length': 5}, {'start': 2090, 'length': 13}, {'start': 357, 'length': 7}, {'start': 1538, 'length': 7}, {'start': 1608, 'length': 18}, {'start': 3122, 'length': 13}, {'start': 3112, 'length': 9}, {'start': 706, 'length': 10}, {'start': 3361, 'length': 13}, {'start': 1129, 'length': 5}, {'start': 1779, 'length': 14}, {'start': 1971, 'length': 6}, {'start': 2724, 'length': 10}, {'start': 807, 'length': 10}, {'start': 1560, 'length': 5}, {'start': 1412, 'length': 4}, {'start': 1879, 'length': 6}, {'start': 1707, 'length': 8}, {'start': 2757, 'length': 20}, {'start': 785, 'length': 20}, {'start': 2071, 'length': 9}, {'start': 3345, 'length': 7}, {'start': 1894, 'length': 15}, {'start': 2782, 'length': 5}, {'start': 553, 'length': 8}, {'start': 2538, 'length': 14}, {'start': 1290, 'length': 5}, {'start': 1081, 'length': 13}, {'start': 2744, 'length': 5}, {'start': 905, 'length': 19}, {'start': 1496, 'length': 15}, {'start': 330, 'length': 5}, {'start': 534, 'length': 8}, {'start': 3191, 'length': 5}, {'start': 176, 'length': 7}, {'start': 496, 'length': 8}, {'start': 2202, 'length': 14}, {'start': 2528, 'length': 9}, {'start': 1800, 'length': 11}, {'start': 2448, 'length': 6}, {'start': 2873, 'length': 9}, {'start': 730, 'length': 8}, {'start': 1267, 'length': 7}, {'start': 518, 'length': 8}, {'start': 2025, 'length': 6}, {'start': 778, 'length': 6}, {'start': 1350, 'length': 5}, {'start': 624, 'length': 6}, {'start': 54, 'length': 4}, {'start': 2788, 'length': 11}, {'start': 1279, 'length': 9}, {'start': 1933, 'length': 6}, {'start': 2682, 'length': 8}, {'start': 2231, 'length': 9}, {'start': 1657, 'length': 9}, {'start': 251, 'length': 7}, {'start': 70, 'length': 7}, {'start': 228, 'length': 7}, {'start': 336, 'length': 16}, {'start': 2251, 'length': 5}, {'start': 2414, 'length': 8}, {'start': 2972, 'length': 7}, {'start': 439, 'length': 8}, {'start': 1763, 'length': 5}, {'start': 960, 'length': 15}, {'start': 2633, 'length': 9}, {'start': 1329, 'length': 19}, {'start': 2957, 'length': 13}, {'start': 691, 'length': 5}, {'start': 2986, 'length': 14}, {'start': 1170, 'length': 5}, {'start': 635, 'length': 5}, {'start': 2576, 'length': 8}, {'start': 3269, 'length': 5}, {'start': 78, 'length': 7}, {'start': 391, 'length': 7}, {'start': 2649, 'length': 8}, {'start': 2492, 'length': 15}, {'start': 929, 'length': 15}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":152,\"length\":18},{\"start\":453,\"length\":8},{\"start\":464,\"length\":15},{\"start\":408,\"length\":5},{\"start\":208,\"length\":9},{\"start\":252,\"length\":12},{\"start\":117,\"length\":9},{\"start\":238,\"length\":9},{\"start\":13,\"length\":10},{\"start\":172,\"length\":5},{\"start\":220,\"length\":13},{\"start\":358,\"length\":17},{\"start\":268,\"length\":11},{\"start\":323,\"length\":9},{\"start\":347,\"length\":8},{\"start\":282,\"length\":15},{\"start\":140,\"length\":7},{\"start\":32,\"length\":8},{\"start\":64,\"length\":10},{\"start\":129,\"length\":9},{\"start\":492,\"length\":11},{\"start\":377,\"length\":10},{\"start\":302,\"length\":16},{\"start\":392,\"length\":14}],\"text\":\"BASKETBALL - EUROLEAGUE RESULT. BRUSSELS 1996-12-06 Result of a EuroLeague basketball match on Thursday : Group B In Charleroi : Charleroi (Belgium) 75 Estudiantes Madrid (Spain) 82 (34-35) Leading scorers : Charleroi - Eric Cleymans 18, Ron Ellis 18, Jacques Stas 14 Estudiantes - Harper Williams 20, Chadler Thompson 17, Juan Aisa 14 Group D In Belgrade : Partizan Belgrade (Yugoslavia) 78 Kinder Bologna (Italy) 70 (halftime 44-35) Leading scorers : Partizan - Dejan Koturovic 21 Kinder - Zoran Savic 18\"}'\ntext: BASKETBALL - EUROLEAGUE RESULT. BRUSSELS 1996-12-06 Result of a EuroLeague basketball match on Thursday : Group B In Charleroi : Charleroi (Belgium) 75 Estudiantes Madrid (Spain) 82 (34-35) Leading scorers : Charleroi - Eric Cleymans 18, Ron Ellis 18, Jacques Stas 14 Estudiantes - Harper Williams 20, Chadler Thompson 17, Juan Aisa 14 Group D In Belgrade : Partizan Belgrade (Yugoslavia) 78 Kinder Bologna (Italy) 70 (halftime 44-35) Leading scorers : Partizan - Dejan Koturovic 21 Kinder - Zoran Savic 18\ntype(text): \nspans: [{'start': 152, 'length': 18}, {'start': 453, 'length': 8}, {'start': 464, 'length': 15}, {'start': 408, 'length': 5}, {'start': 208, 'length': 9}, {'start': 252, 'length': 12}, {'start': 117, 'length': 9}, {'start': 238, 'length': 9}, {'start': 13, 'length': 10}, {'start': 172, 'length': 5}, {'start': 220, 'length': 13}, {'start': 358, 'length': 17}, {'start': 268, 'length': 11}, {'start': 323, 'length': 9}, {'start': 347, 'length': 8}, {'start': 282, 'length': 15}, {'start': 140, 'length': 7}, {'start': 32, 'length': 8}, {'start': 64, 'length': 10}, {'start': 129, 'length': 9}, {'start': 492, 'length': 11}, {'start': 377, 'length': 10}, {'start': 302, 'length': 16}, {'start': 392, 'length': 14}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":398,\"length\":5},{\"start\":463,\"length\":11},{\"start\":52,\"length\":6},{\"start\":359,\"length\":11},{\"start\":9,\"length\":5},{\"start\":260,\"length\":10},{\"start\":348,\"length\":10},{\"start\":478,\"length\":7},{\"start\":94,\"length\":12},{\"start\":314,\"length\":10},{\"start\":441,\"length\":8},{\"start\":650,\"length\":5},{\"start\":426,\"length\":11},{\"start\":286,\"length\":22},{\"start\":504,\"length\":5},{\"start\":60,\"length\":5}],\"text\":\"SQUASH - EYLES WITHIN SIGHT OF FIFTH TITLE OF YEAR. BOMBAY, India 1996-12-06 World number two Rodney Eyles moved within sight of his fifth title of the year on Friday when he hurried in only 40 minutes to the final of the richest squash tournament outside the World Open, the $ 105,000 Mahindra International. The Australian brushed aside unseeded Englishman Mark Cairns 15-7 15-6 15-8. Top-seeded Eyles now meets titleholder Peter Nicol of Scotland who overcame Simon Parke of England 15-7 15-12 15-12. Nicol was full of praise for his opponent who has battled testicular cancer to return to the circuit. \\\\\"He\\\\u0027s a remarkably courageous player,\\\\\" said Nicol.\"}'\ntext: SQUASH - EYLES WITHIN SIGHT OF FIFTH TITLE OF YEAR. BOMBAY, India 1996-12-06 World number two Rodney Eyles moved within sight of his fifth title of the year on Friday when he hurried in only 40 minutes to the final of the richest squash tournament outside the World Open, the $ 105,000 Mahindra International. The Australian brushed aside unseeded Englishman Mark Cairns 15-7 15-6 15-8. Top-seeded Eyles now meets titleholder Peter Nicol of Scotland who overcame Simon Parke of England 15-7 15-12 15-12. Nicol was full of praise for his opponent who has battled testicular cancer to return to the circuit. \"He's a remarkably courageous player,\" said Nicol.\ntype(text): \nspans: [{'start': 398, 'length': 5}, {'start': 463, 'length': 11}, {'start': 52, 'length': 6}, {'start': 359, 'length': 11}, {'start': 9, 'length': 5}, {'start': 260, 'length': 10}, {'start': 348, 'length': 10}, {'start': 478, 'length': 7}, {'start': 94, 'length': 12}, {'start': 314, 'length': 10}, {'start': 441, 'length': 8}, {'start': 650, 'length': 5}, {'start': 426, 'length': 11}, {'start': 286, 'length': 22}, {'start': 504, 'length': 5}, {'start': 60, 'length': 5}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":186,\"length\":11},{\"start\":255,\"length\":11},{\"start\":309,\"length\":5},{\"start\":158,\"length\":11},{\"start\":105,\"length\":22},{\"start\":199,\"length\":7},{\"start\":51,\"length\":6},{\"start\":171,\"length\":8},{\"start\":239,\"length\":9},{\"start\":225,\"length\":12},{\"start\":301,\"length\":5},{\"start\":59,\"length\":5},{\"start\":268,\"length\":7},{\"start\":9,\"length\":22}],\"text\":\"SQUASH - MAHINDRA INTERNATIONAL SEMIFINAL RESULTS. BOMBAY, India 1996-12-06 Results of semifinals in the Mahindra International squash tournament on Friday : Peter Nicol (Scotland) beat Simon Parke (England) 15-7 15-12 15-12 Rodney Eyles (Australia) beat Mark Cairns (England) 15-7 15-6 15-8. Final : Nicol v Eyles, on Saturday.\"}'\ntext: SQUASH - MAHINDRA INTERNATIONAL SEMIFINAL RESULTS. BOMBAY, India 1996-12-06 Results of semifinals in the Mahindra International squash tournament on Friday : Peter Nicol (Scotland) beat Simon Parke (England) 15-7 15-12 15-12 Rodney Eyles (Australia) beat Mark Cairns (England) 15-7 15-6 15-8. Final : Nicol v Eyles, on Saturday.\ntype(text): \nspans: [{'start': 186, 'length': 11}, {'start': 255, 'length': 11}, {'start': 309, 'length': 5}, {'start': 158, 'length': 11}, {'start': 105, 'length': 22}, {'start': 199, 'length': 7}, {'start': 51, 'length': 6}, {'start': 171, 'length': 8}, {'start': 239, 'length': 9}, {'start': 225, 'length': 12}, {'start': 301, 'length': 5}, {'start': 59, 'length': 5}, {'start': 268, 'length': 7}, {'start': 9, 'length': 22}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":1359,\"length\":3},{\"start\":294,\"length\":14},{\"start\":20,\"length\":8},{\"start\":929,\"length\":13},{\"start\":418,\"length\":11},{\"start\":311,\"length\":25},{\"start\":1401,\"length\":21},{\"start\":54,\"length\":12},{\"start\":338,\"length\":3},{\"start\":433,\"length\":13},{\"start\":181,\"length\":4},{\"start\":778,\"length\":7},{\"start\":761,\"length\":11},{\"start\":31,\"length\":4},{\"start\":1367,\"length\":4},{\"start\":46,\"length\":6},{\"start\":684,\"length\":8},{\"start\":157,\"length\":12},{\"start\":1378,\"length\":20}],\"text\":\"GUNMEN KILL FOUR IN S.AFRICA\\\\u0027S ZULU PROVINCE. DURBAN, South Africa 1996-12-06 At least four people have been shot dead in two suspected political attacks in South Africa\\\\u0027s volatile Zulu heartland, police said on Friday. A police spokesman said two youths believed to be supporters of President Nelson Mandela\\\\u0027s African National Congress (ANC) had been killed when unknown gunmen opened fire at the rural settlement of Izingolweni on KwaZulu-Natal province\\\\u0027s south coast on Thursday night. The victims were 18 and 20, he said, adding one other youth had been wounded in the shooting. In another attack, also on the province\\\\u0027s south coast on Thursday night, two men were shot dead near Umkomaas. \\\\\"We suspect that these killings are linked to politics,\\\\\" spokesman Bala Naidoo told Reuters. There had been no arrests. The killings came just hours after violence monitors said they were not optimistic of a peaceful festive season in KwaZulu-Natal and pointed the south coast region where 18 people were massacred last Christmas as one of potential hot spots. They said the recent lull in political feuding could be upset as thousands of migrant workers, some tense with grudges brewed in the cities and keen to settle old scores, flock back to their home villages. More than 14,000 people have lost their lives in over a decade of political turf wars between the ANC and Zulu Chief Mangosuthu Buthelezi\\\\u0027s Inkatha Freedom Party in the province.\"}'\ntext: GUNMEN KILL FOUR IN S.AFRICA'S ZULU PROVINCE. DURBAN, South Africa 1996-12-06 At least four people have been shot dead in two suspected political attacks in South Africa's volatile Zulu heartland, police said on Friday. A police spokesman said two youths believed to be supporters of President Nelson Mandela's African National Congress (ANC) had been killed when unknown gunmen opened fire at the rural settlement of Izingolweni on KwaZulu-Natal province's south coast on Thursday night. The victims were 18 and 20, he said, adding one other youth had been wounded in the shooting. In another attack, also on the province's south coast on Thursday night, two men were shot dead near Umkomaas. \"We suspect that these killings are linked to politics,\" spokesman Bala Naidoo told Reuters. There had been no arrests. The killings came just hours after violence monitors said they were not optimistic of a peaceful festive season in KwaZulu-Natal and pointed the south coast region where 18 people were massacred last Christmas as one of potential hot spots. They said the recent lull in political feuding could be upset as thousands of migrant workers, some tense with grudges brewed in the cities and keen to settle old scores, flock back to their home villages. More than 14,000 people have lost their lives in over a decade of political turf wars between the ANC and Zulu Chief Mangosuthu Buthelezi's Inkatha Freedom Party in the province.\ntype(text): \nspans: [{'start': 1359, 'length': 3}, {'start': 294, 'length': 14}, {'start': 20, 'length': 8}, {'start': 929, 'length': 13}, {'start': 418, 'length': 11}, {'start': 311, 'length': 25}, {'start': 1401, 'length': 21}, {'start': 54, 'length': 12}, {'start': 338, 'length': 3}, {'start': 433, 'length': 13}, {'start': 181, 'length': 4}, {'start': 778, 'length': 7}, {'start': 761, 'length': 11}, {'start': 31, 'length': 4}, {'start': 1367, 'length': 4}, {'start': 46, 'length': 6}, {'start': 684, 'length': 8}, {'start': 157, 'length': 12}, {'start': 1378, 'length': 20}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":1909,\"length\":8},{\"start\":1705,\"length\":7},{\"start\":962,\"length\":15},{\"start\":1245,\"length\":5},{\"start\":1361,\"length\":8},{\"start\":1126,\"length\":8},{\"start\":1647,\"length\":6},{\"start\":849,\"length\":5},{\"start\":1330,\"length\":4},{\"start\":14,\"length\":5},{\"start\":179,\"length\":5},{\"start\":1229,\"length\":14},{\"start\":736,\"length\":8},{\"start\":1840,\"length\":14},{\"start\":149,\"length\":18},{\"start\":47,\"length\":16},{\"start\":0,\"length\":5},{\"start\":82,\"length\":5},{\"start\":485,\"length\":13},{\"start\":591,\"length\":5},{\"start\":321,\"length\":18},{\"start\":1575,\"length\":14},{\"start\":1684,\"length\":15},{\"start\":2037,\"length\":6},{\"start\":947,\"length\":8},{\"start\":1753,\"length\":4},{\"start\":1159,\"length\":6},{\"start\":712,\"length\":6},{\"start\":692,\"length\":9},{\"start\":1816,\"length\":5},{\"start\":98,\"length\":12},{\"start\":1438,\"length\":11},{\"start\":620,\"length\":7},{\"start\":1886,\"length\":17},{\"start\":397,\"length\":8},{\"start\":204,\"length\":13},{\"start\":1066,\"length\":10},{\"start\":1716,\"length\":8},{\"start\":1662,\"length\":2},{\"start\":1042,\"length\":13},{\"start\":1856,\"length\":13},{\"start\":27,\"length\":8},{\"start\":1206,\"length\":4},{\"start\":629,\"length\":5},{\"start\":64,\"length\":6},{\"start\":983,\"length\":12},{\"start\":267,\"length\":5}],\"text\":\"HAVEL PRAISES CZECH NATIVE ALBRIGHT AS FRIEND. Klara Gajduskova PRAGUE 1996-12-06 Czech President Vaclav Havel on Friday welcomed the appointment of Madeleine Albright, who is of Czech extraction, as the United States \\\\u0027 first woman Secretary of State. In a statement Havel, who is recovering from cancer surgery, said : \\\\\"Madeleine Albright is a distinguished friend, a tested diplomat, and a true American of fine origins.\\\\\" \\\\\"I look forward to continuing our good relations... with the United States and with the first woman ever to hold the position of Secretary of State. I wish her well,\\\\\" Havel said in a statement to Reuters. Havel, who helped lead the \\\\\"velvet revolution\\\\\" that ousted the Communist regime in Prague in 1989, invited Albright, then working for a private foreign policy think tank, to advise his new democratic government in 1990. Havel had a small malignant tumour removed from his lung on Monday and is recovering in hospital. Albright, born Marie Korbelova to a Czechoslovak diplomat in 1937, fled with her family to the United States after the Communists came to power in a coup in 1948. As an academic, Albright studied and lectured on Europe\\\\u0027s 20th century problems before becoming U.S. ambassador to the United Nations. Czech diplomats, seeking to have their country included in the expected expansion of NATO, praised the selection of Albright, known to be a strong supporter of alliance\\\\u0027s integration of former Soveit-bloc countries. \\\\\"The nomination... is a clear signal that one key of the lines of foreign policy will be the strengthening of the trans-Atlantic cooperation, a creation of strategic partnership between Europe and the US,\\\\\" Foreign Minister Josef Zieleniec told Reuters. \\\\\"(Albright) is a convinced advocate of NATO enlargement and of stabilisation of security structures.\\\\\" Czech ambassador to the United Nations, Karel Kovanda, told the daily Mlada Fronta Dnes that Albright \\\\\"is a little light in our diplomatic heaven,\\\\\" but warned against expecting her to exert any influence in favour of the Czechs.\"}'\ntext: HAVEL PRAISES CZECH NATIVE ALBRIGHT AS FRIEND. Klara Gajduskova PRAGUE 1996-12-06 Czech President Vaclav Havel on Friday welcomed the appointment of Madeleine Albright, who is of Czech extraction, as the United States ' first woman Secretary of State. In a statement Havel, who is recovering from cancer surgery, said : \"Madeleine Albright is a distinguished friend, a tested diplomat, and a true American of fine origins.\" \"I look forward to continuing our good relations... with the United States and with the first woman ever to hold the position of Secretary of State. I wish her well,\" Havel said in a statement to Reuters. Havel, who helped lead the \"velvet revolution\" that ousted the Communist regime in Prague in 1989, invited Albright, then working for a private foreign policy think tank, to advise his new democratic government in 1990. Havel had a small malignant tumour removed from his lung on Monday and is recovering in hospital. Albright, born Marie Korbelova to a Czechoslovak diplomat in 1937, fled with her family to the United States after the Communists came to power in a coup in 1948. As an academic, Albright studied and lectured on Europe's 20th century problems before becoming U.S. ambassador to the United Nations. Czech diplomats, seeking to have their country included in the expected expansion of NATO, praised the selection of Albright, known to be a strong supporter of alliance's integration of former Soveit-bloc countries. \"The nomination... is a clear signal that one key of the lines of foreign policy will be the strengthening of the trans-Atlantic cooperation, a creation of strategic partnership between Europe and the US,\" Foreign Minister Josef Zieleniec told Reuters. \"(Albright) is a convinced advocate of NATO enlargement and of stabilisation of security structures.\" Czech ambassador to the United Nations, Karel Kovanda, told the daily Mlada Fronta Dnes that Albright \"is a little light in our diplomatic heaven,\" but warned against expecting her to exert any influence in favour of the Czechs.\ntype(text): \nspans: [{'start': 1909, 'length': 8}, {'start': 1705, 'length': 7}, {'start': 962, 'length': 15}, {'start': 1245, 'length': 5}, {'start': 1361, 'length': 8}, {'start': 1126, 'length': 8}, {'start': 1647, 'length': 6}, {'start': 849, 'length': 5}, {'start': 1330, 'length': 4}, {'start': 14, 'length': 5}, {'start': 179, 'length': 5}, {'start': 1229, 'length': 14}, {'start': 736, 'length': 8}, {'start': 1840, 'length': 14}, {'start': 149, 'length': 18}, {'start': 47, 'length': 16}, {'start': 0, 'length': 5}, {'start': 82, 'length': 5}, {'start': 485, 'length': 13}, {'start': 591, 'length': 5}, {'start': 321, 'length': 18}, {'start': 1575, 'length': 14}, {'start': 1684, 'length': 15}, {'start': 2037, 'length': 6}, {'start': 947, 'length': 8}, {'start': 1753, 'length': 4}, {'start': 1159, 'length': 6}, {'start': 712, 'length': 6}, {'start': 692, 'length': 9}, {'start': 1816, 'length': 5}, {'start': 98, 'length': 12}, {'start': 1438, 'length': 11}, {'start': 620, 'length': 7}, {'start': 1886, 'length': 17}, {'start': 397, 'length': 8}, {'start': 204, 'length': 13}, {'start': 1066, 'length': 10}, {'start': 1716, 'length': 8}, {'start': 1662, 'length': 2}, {'start': 1042, 'length': 13}, {'start': 1856, 'length': 13}, {'start': 27, 'length': 8}, {'start': 1206, 'length': 4}, {'start': 629, 'length': 5}, {'start': 64, 'length': 6}, {'start': 983, 'length': 12}, {'start': 267, 'length': 5}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":681,\"length\":6},{\"start\":500,\"length\":4},{\"start\":566,\"length\":7},{\"start\":320,\"length\":19},{\"start\":692,\"length\":5},{\"start\":636,\"length\":9},{\"start\":675,\"length\":4},{\"start\":274,\"length\":4},{\"start\":63,\"length\":13},{\"start\":100,\"length\":21},{\"start\":454,\"length\":14},{\"start\":214,\"length\":23},{\"start\":588,\"length\":10},{\"start\":246,\"length\":26},{\"start\":435,\"length\":7},{\"start\":42,\"length\":9},{\"start\":0,\"length\":13},{\"start\":724,\"length\":18}],\"text\":\"RADIO ROMANIA AFTERNOON HEALINES AT 4 PM. BUCHAREST 1996-12-06 Radio Romania news headlines : * The Democratic Convention signed an agreement on government and parliamentary support with its coalition partners the Social Democratic Union and the Hungarian Democratic Union (UDMR). The ceremony was attended by President Emil Constantinescu. * The three parties in the government coalition have committed themselves to a real reform of Romania\\\\u0027s economy, Constantinescu said after the ceremony. * The UDMR wants to contribute to social reform and economic revival in Romania, union leader Marko Bela said. * The international airport in Timisoara and the domestic airports in Arad, Oradea and Sibiu were closed due to fog.-- Bucharest Newsroom 40-1 3120264\"}'\ntext: RADIO ROMANIA AFTERNOON HEALINES AT 4 PM. BUCHAREST 1996-12-06 Radio Romania news headlines : * The Democratic Convention signed an agreement on government and parliamentary support with its coalition partners the Social Democratic Union and the Hungarian Democratic Union (UDMR). The ceremony was attended by President Emil Constantinescu. * The three parties in the government coalition have committed themselves to a real reform of Romania's economy, Constantinescu said after the ceremony. * The UDMR wants to contribute to social reform and economic revival in Romania, union leader Marko Bela said. * The international airport in Timisoara and the domestic airports in Arad, Oradea and Sibiu were closed due to fog.-- Bucharest Newsroom 40-1 3120264\ntype(text): \nspans: [{'start': 681, 'length': 6}, {'start': 500, 'length': 4}, {'start': 566, 'length': 7}, {'start': 320, 'length': 19}, {'start': 692, 'length': 5}, {'start': 636, 'length': 9}, {'start': 675, 'length': 4}, {'start': 274, 'length': 4}, {'start': 63, 'length': 13}, {'start': 100, 'length': 21}, {'start': 454, 'length': 14}, {'start': 214, 'length': 23}, {'start': 588, 'length': 10}, {'start': 246, 'length': 26}, {'start': 435, 'length': 7}, {'start': 42, 'length': 9}, {'start': 0, 'length': 13}, {'start': 724, 'length': 18}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":770,\"length\":4},{\"start\":112,\"length\":28},{\"start\":876,\"length\":5},{\"start\":925,\"length\":7},{\"start\":248,\"length\":15},{\"start\":1499,\"length\":5},{\"start\":613,\"length\":3},{\"start\":1046,\"length\":9},{\"start\":752,\"length\":5},{\"start\":1654,\"length\":16},{\"start\":1452,\"length\":3},{\"start\":0,\"length\":5},{\"start\":51,\"length\":6},{\"start\":1294,\"length\":5},{\"start\":1218,\"length\":14},{\"start\":990,\"length\":5},{\"start\":1178,\"length\":5},{\"start\":1708,\"length\":15},{\"start\":230,\"length\":3},{\"start\":854,\"length\":3},{\"start\":650,\"length\":12},{\"start\":428,\"length\":9},{\"start\":948,\"length\":17},{\"start\":1384,\"length\":5},{\"start\":488,\"length\":7},{\"start\":142,\"length\":3}],\"text\":\"CZECH VICE-PM SEES WIDER DEBATE AT PARTY CONGRESS. PRAGUE 1996-12-06 Saturday\\\\u0027s national congress of the ruling Czech Civic Democratic Party (ODS) will discuss making the party more efficient and transparent, Foreign Minister and ODS vice-chairman Josef Zieleniec, said on Friday. \\\\\"Modernisation and more profesionalism of the party\\\\u0027s structure, having financing of the party be more transparent... are absolutely fundamental,\\\\\" Zieleniec, who is also vice-premier in the government, told Reuters. He said after June general elections in which the ruling three-party coalition lost its parliamentary majority, the ODS executive, led by Prime Minister Vaclav Klaus, had developed proposals on these subjects to present at the congress on Saturday in the Czech second city Brno. \\\\\"I am convinced, that the congress will tackle these proposals,\\\\\" he said. The ODS, a party in which Klaus often tries to emulate the style of former British Prime Minister Margaret Thatcher, has been in control of Czech politics since winning general elections in 1992. Zieleniec in the summer led calls for the party and its leadership to listen to more diverse opinions, a thinly-veiled criticism of Klaus who has spearheaded the country\\\\u0027s post-Communist economic reforms. The party, led by the vigorously-confident Klaus, took 32 of 81 seats after late November runoff elections to the new upper house of Czech parliament. But after the first round vote a week before, the ODS had the potential to win as many 79 seats. Klaus and his coalition lost its majority in parliament in June lower house elections after the left-wing opposition consolidated, putting the centre-left Social Democrats in a strong second-place position.-- Prague Newsroom 42-2-2423-0003\"}'\ntext: CZECH VICE-PM SEES WIDER DEBATE AT PARTY CONGRESS. PRAGUE 1996-12-06 Saturday's national congress of the ruling Czech Civic Democratic Party (ODS) will discuss making the party more efficient and transparent, Foreign Minister and ODS vice-chairman Josef Zieleniec, said on Friday. \"Modernisation and more profesionalism of the party's structure, having financing of the party be more transparent... are absolutely fundamental,\" Zieleniec, who is also vice-premier in the government, told Reuters. He said after June general elections in which the ruling three-party coalition lost its parliamentary majority, the ODS executive, led by Prime Minister Vaclav Klaus, had developed proposals on these subjects to present at the congress on Saturday in the Czech second city Brno. \"I am convinced, that the congress will tackle these proposals,\" he said. The ODS, a party in which Klaus often tries to emulate the style of former British Prime Minister Margaret Thatcher, has been in control of Czech politics since winning general elections in 1992. Zieleniec in the summer led calls for the party and its leadership to listen to more diverse opinions, a thinly-veiled criticism of Klaus who has spearheaded the country's post-Communist economic reforms. The party, led by the vigorously-confident Klaus, took 32 of 81 seats after late November runoff elections to the new upper house of Czech parliament. But after the first round vote a week before, the ODS had the potential to win as many 79 seats. Klaus and his coalition lost its majority in parliament in June lower house elections after the left-wing opposition consolidated, putting the centre-left Social Democrats in a strong second-place position.-- Prague Newsroom 42-2-2423-0003\ntype(text): \nspans: [{'start': 770, 'length': 4}, {'start': 112, 'length': 28}, {'start': 876, 'length': 5}, {'start': 925, 'length': 7}, {'start': 248, 'length': 15}, {'start': 1499, 'length': 5}, {'start': 613, 'length': 3}, {'start': 1046, 'length': 9}, {'start': 752, 'length': 5}, {'start': 1654, 'length': 16}, {'start': 1452, 'length': 3}, {'start': 0, 'length': 5}, {'start': 51, 'length': 6}, {'start': 1294, 'length': 5}, {'start': 1218, 'length': 14}, {'start': 990, 'length': 5}, {'start': 1178, 'length': 5}, {'start': 1708, 'length': 15}, {'start': 230, 'length': 3}, {'start': 854, 'length': 3}, {'start': 650, 'length': 12}, {'start': 428, 'length': 9}, {'start': 948, 'length': 17}, {'start': 1384, 'length': 5}, {'start': 488, 'length': 7}, {'start': 142, 'length': 3}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":1979,\"length\":6},{\"start\":64,\"length\":6},{\"start\":2278,\"length\":5},{\"start\":109,\"length\":5},{\"start\":2183,\"length\":3},{\"start\":267,\"length\":13},{\"start\":682,\"length\":15},{\"start\":2398,\"length\":5},{\"start\":1792,\"length\":6},{\"start\":1006,\"length\":4},{\"start\":414,\"length\":6},{\"start\":935,\"length\":6},{\"start\":438,\"length\":5},{\"start\":82,\"length\":6},{\"start\":47,\"length\":16},{\"start\":1042,\"length\":6},{\"start\":823,\"length\":6},{\"start\":190,\"length\":9},{\"start\":1069,\"length\":6},{\"start\":481,\"length\":6},{\"start\":2243,\"length\":5},{\"start\":971,\"length\":4},{\"start\":1435,\"length\":6},{\"start\":1448,\"length\":6},{\"start\":299,\"length\":14},{\"start\":644,\"length\":11},{\"start\":1156,\"length\":6},{\"start\":0,\"length\":6},{\"start\":1289,\"length\":13},{\"start\":613,\"length\":6},{\"start\":2164,\"length\":17},{\"start\":2350,\"length\":6},{\"start\":1625,\"length\":11},{\"start\":1063,\"length\":4},{\"start\":31,\"length\":5},{\"start\":2001,\"length\":5},{\"start\":1198,\"length\":5},{\"start\":1932,\"length\":6},{\"start\":724,\"length\":29},{\"start\":2128,\"length\":9},{\"start\":524,\"length\":11},{\"start\":1245,\"length\":14},{\"start\":784,\"length\":6},{\"start\":1515,\"length\":11},{\"start\":1167,\"length\":11},{\"start\":162,\"length\":6},{\"start\":2428,\"length\":6},{\"start\":1011,\"length\":6},{\"start\":877,\"length\":5},{\"start\":169,\"length\":4}],\"text\":\"POLAND GOT MONEY FROM POST-WAR SWISS ACCOUNTS. Marcin Grajewski WARSAW 1996-12-06 Poland said on Friday that Swiss bank accounts, which in many cases belonged to Polish Jews who died in the Holocaust, were used in debt settlements between the two countries after the World War Two. Foreign Minister Dariusz Rosati, unveiling first findings of a special government commission, said that in 1970s the then communist Poland received 460,000 Swiss francs from the accounts. \\\\\"In 1970s, Poland received from unclaimed accounts in Switzerland a sum of 460,000 francs. What was its right (to the money)...I do not know,\\\\\" Rosati told a news conference. Switzerland stands accused by Senator Alfonse D\\\\u0027Amato, chairman of the powerful U.S. Senate Banking Committee, of agreeing to give money to Poland from unclaimed bank accounts of Polish citizens, as part of an accord on compensating Swiss nationals whose assets had been seized in communist Poland. Many of these citizens were Jews murdered during the war, when Nazi German invaders killed most of Poland\\\\u0027s 3.5 million Jews. Rosati did not say whether the payment in 1970s was part of the 1949 agreement between Warsaw and Switzerland on compensation to Swiss citizens whose assets were seized by the Soviet-imposed communists authorities after World War Two. \\\\\"I expect that the commission will finish gathering information within two to three weeks and then more details will be provided,\\\\\" Rosati said. Rosati confirmed that the 1949 agreement had provided for granting Switzerland about 53 million francs and most of this sum was repaid with coal exports. He said, however, that Switzerland did get about 16,000 francs from the so-called \\\\\"dead accounts\\\\\" as part of the compensation. \\\\\"About 16,000 francs were seized from accounts of four or five Polish citizens, whose data we do not precisely know. The issue is of moral and legal nature, because its financial significance is small,\\\\\" Rosati said. Under pressure from international Jewish organisations, Swiss government has devised a plan to pay out millions of dollars in unclaimed bank accounts as a conciliatory gesture toward Holocaust victims. The conservative Radical Democrats (FDP) have said they would ask parliament next week to order Swiss banks to put some 40 million Swiss francs ($ 31 million) in dormant wealth into a fund earmarked for Jewish groups and charitable organisations. But Swiss banks and the country\\\\u0027s Jewish community voiced doubts whether the plan would work.\"}'\ntext: POLAND GOT MONEY FROM POST-WAR SWISS ACCOUNTS. Marcin Grajewski WARSAW 1996-12-06 Poland said on Friday that Swiss bank accounts, which in many cases belonged to Polish Jews who died in the Holocaust, were used in debt settlements between the two countries after the World War Two. Foreign Minister Dariusz Rosati, unveiling first findings of a special government commission, said that in 1970s the then communist Poland received 460,000 Swiss francs from the accounts. \"In 1970s, Poland received from unclaimed accounts in Switzerland a sum of 460,000 francs. What was its right (to the money)...I do not know,\" Rosati told a news conference. Switzerland stands accused by Senator Alfonse D'Amato, chairman of the powerful U.S. Senate Banking Committee, of agreeing to give money to Poland from unclaimed bank accounts of Polish citizens, as part of an accord on compensating Swiss nationals whose assets had been seized in communist Poland. Many of these citizens were Jews murdered during the war, when Nazi German invaders killed most of Poland's 3.5 million Jews. Rosati did not say whether the payment in 1970s was part of the 1949 agreement between Warsaw and Switzerland on compensation to Swiss citizens whose assets were seized by the Soviet-imposed communists authorities after World War Two. \"I expect that the commission will finish gathering information within two to three weeks and then more details will be provided,\" Rosati said. Rosati confirmed that the 1949 agreement had provided for granting Switzerland about 53 million francs and most of this sum was repaid with coal exports. He said, however, that Switzerland did get about 16,000 francs from the so-called \"dead accounts\" as part of the compensation. \"About 16,000 francs were seized from accounts of four or five Polish citizens, whose data we do not precisely know. The issue is of moral and legal nature, because its financial significance is small,\" Rosati said. Under pressure from international Jewish organisations, Swiss government has devised a plan to pay out millions of dollars in unclaimed bank accounts as a conciliatory gesture toward Holocaust victims. The conservative Radical Democrats (FDP) have said they would ask parliament next week to order Swiss banks to put some 40 million Swiss francs ($ 31 million) in dormant wealth into a fund earmarked for Jewish groups and charitable organisations. But Swiss banks and the country's Jewish community voiced doubts whether the plan would work.\ntype(text): \nspans: [{'start': 1979, 'length': 6}, {'start': 64, 'length': 6}, {'start': 2278, 'length': 5}, {'start': 109, 'length': 5}, {'start': 2183, 'length': 3}, {'start': 267, 'length': 13}, {'start': 682, 'length': 15}, {'start': 2398, 'length': 5}, {'start': 1792, 'length': 6}, {'start': 1006, 'length': 4}, {'start': 414, 'length': 6}, {'start': 935, 'length': 6}, {'start': 438, 'length': 5}, {'start': 82, 'length': 6}, {'start': 47, 'length': 16}, {'start': 1042, 'length': 6}, {'start': 823, 'length': 6}, {'start': 190, 'length': 9}, {'start': 1069, 'length': 6}, {'start': 481, 'length': 6}, {'start': 2243, 'length': 5}, {'start': 971, 'length': 4}, {'start': 1435, 'length': 6}, {'start': 1448, 'length': 6}, {'start': 299, 'length': 14}, {'start': 644, 'length': 11}, {'start': 1156, 'length': 6}, {'start': 0, 'length': 6}, {'start': 1289, 'length': 13}, {'start': 613, 'length': 6}, {'start': 2164, 'length': 17}, {'start': 2350, 'length': 6}, {'start': 1625, 'length': 11}, {'start': 1063, 'length': 4}, {'start': 31, 'length': 5}, {'start': 2001, 'length': 5}, {'start': 1198, 'length': 5}, {'start': 1932, 'length': 6}, {'start': 724, 'length': 29}, {'start': 2128, 'length': 9}, {'start': 524, 'length': 11}, {'start': 1245, 'length': 14}, {'start': 784, 'length': 6}, {'start': 1515, 'length': 11}, {'start': 1167, 'length': 11}, {'start': 162, 'length': 6}, {'start': 2428, 'length': 6}, {'start': 1011, 'length': 6}, {'start': 877, 'length': 5}, {'start': 169, 'length': 4}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":56,\"length\":6},{\"start\":317,\"length\":16},{\"start\":3473,\"length\":6},{\"start\":3207,\"length\":6},{\"start\":1704,\"length\":6},{\"start\":2910,\"length\":11},{\"start\":1730,\"length\":27},{\"start\":0,\"length\":16},{\"start\":2927,\"length\":6},{\"start\":1900,\"length\":11},{\"start\":3428,\"length\":8},{\"start\":1330,\"length\":6},{\"start\":3481,\"length\":11},{\"start\":2048,\"length\":6},{\"start\":2538,\"length\":5},{\"start\":2949,\"length\":6},{\"start\":1888,\"length\":9},{\"start\":2997,\"length\":9},{\"start\":883,\"length\":6},{\"start\":2428,\"length\":14},{\"start\":2685,\"length\":6},{\"start\":2642,\"length\":5},{\"start\":2582,\"length\":7},{\"start\":777,\"length\":11},{\"start\":1656,\"length\":8},{\"start\":2514,\"length\":6},{\"start\":1371,\"length\":7},{\"start\":1939,\"length\":6},{\"start\":1990,\"length\":2},{\"start\":3214,\"length\":10},{\"start\":1290,\"length\":6},{\"start\":1813,\"length\":5},{\"start\":1132,\"length\":7},{\"start\":2487,\"length\":5},{\"start\":1759,\"length\":3},{\"start\":1148,\"length\":6},{\"start\":2887,\"length\":5},{\"start\":1392,\"length\":11},{\"start\":74,\"length\":6},{\"start\":3342,\"length\":6},{\"start\":1383,\"length\":7},{\"start\":615,\"length\":6},{\"start\":88,\"length\":6},{\"start\":3503,\"length\":6},{\"start\":42,\"length\":13},{\"start\":1966,\"length\":22},{\"start\":3032,\"length\":6},{\"start\":3402,\"length\":7},{\"start\":1624,\"length\":6},{\"start\":1804,\"length\":4},{\"start\":1995,\"length\":11},{\"start\":3065,\"length\":8},{\"start\":3710,\"length\":15},{\"start\":1671,\"length\":9},{\"start\":642,\"length\":30},{\"start\":339,\"length\":7},{\"start\":1409,\"length\":6}],\"text\":\"INTERVIEW-ZYWIEC SEES NO BIG 97 NET RISE. Steven Silber WARSAW 1996-12-06 Polish brewer Zywiec\\\\u0027s 1996 profit slump may last into next year due in part to hefty depreciation charges, but recent high investment should help the firm defend its 10-percent market share, the firm\\\\u0027s chief executive said. Company President Jean van Boxmeer told Reuters in an interview on Friday that the firm, whose net profit fell 77 percent in the first 10 months of 1996 despite a 30-percent rise in sales, might only post slightly better profits in 1997 before having a chance to make a more significant turnaround. So far this year Zywiec, whose full name is Zaklady Piwowarskie w Zywcu SA, has netted six million zlotys on sales of 224 million zlotys. It has produced 1.5 million hectolitres. Van Boxmeer would not say how much higher 1997 profits or market share could be but said sales of leading Polish brewers should rise as the country\\\\u0027s young urban professionals gradually switch from vodka to beer. \\\\\"The perspective on growth is such that reasonably we can think that somewhere between 65 and 80 litres per year is certainly reachable,\\\\\" van Boxmeer said on Polish per-capita beer consumption, currently around 40 litres. He said the 65-80-litre level could be reached in the next ten years and make Poland, with its 40-million population, Europe\\\\u0027s third largest beer market after Germany and Britain. Van Boxmeer said Poland\\\\u0027s top five brewers, which produce about 55 percent of the country\\\\u0027s beer, could all raise market share as some of the numerous small brewers fall to competition from the large brewers with foreign investors. Zywiec is 31.8-percent owned by Heineken while Carlsberg has the same amount in Okocim. Earlier this year South African Breweries Ltd (SAB) bought strategic stakes in the unlisted Lech and Tychy brewers, which together hold more than 20 percent of the market, and Australia\\\\u0027s Brewpole BV has a controlling stake in Poland\\\\u0027s larges t brewery, Elbrewery Company Ltd. (EB). Van Boxmeer said the tough competition had prevented Zywiec from raising prices in line with inflation, which had added to the pressure on the firm\\\\u0027s margins. He said advertising costs would also increase in the fight for market share. But he said the company\\\\u0027s investment of more than $ 100 million already this decade, largely in production, would help position it to compete with such competitors as brewers from the neighbouring Czech Republic. Some analysts say cheaper but high-quality Czech imports could invade Poland once tariffs for CEFTA countries are lifted in 1998, but van Boxmeer says such a threat might be exaggerated despite the Czech beer market\\\\u0027s overcapacity. \\\\\"I think Polish consumers in general are quite proud of their beers-- and I\\\\u0027m speaking about all the brands-- and as we make good beers... I think that this fidelity to our beers is a factor which can limit the Czech beers,\\\\\" he said. Van Boxmeer said Zywiec had its eye on Okocim, which has said it would start producing Carlsberg beer next year, but that Zywiec\\\\u0027s potential production of Heineken was a medium-term possibility rather than a short-term one. He said his firm would be better off concentrating on its leading brand, Zywiec Full Light, which accounts for 85 percent of sales and is the country\\\\u0027s largest-selling brand. \\\\\"You will not win the war of the Polish beer market with imported international brands,\\\\\" van Boxmeer said, adding that Heineken would remain an up-market import in Poland. Van Boxmeer also said Zywiec would be boosted by its recent shedding of soft drinks which only accounted for about three percent of the firm\\\\u0027s overall sales and for which 7.6 million zlotys in provisions had already been made.-- Warsaw Newsroom+48 22 653 9700\"}'\ntext: INTERVIEW-ZYWIEC SEES NO BIG 97 NET RISE. Steven Silber WARSAW 1996-12-06 Polish brewer Zywiec's 1996 profit slump may last into next year due in part to hefty depreciation charges, but recent high investment should help the firm defend its 10-percent market share, the firm's chief executive said. Company President Jean van Boxmeer told Reuters in an interview on Friday that the firm, whose net profit fell 77 percent in the first 10 months of 1996 despite a 30-percent rise in sales, might only post slightly better profits in 1997 before having a chance to make a more significant turnaround. So far this year Zywiec, whose full name is Zaklady Piwowarskie w Zywcu SA, has netted six million zlotys on sales of 224 million zlotys. It has produced 1.5 million hectolitres. Van Boxmeer would not say how much higher 1997 profits or market share could be but said sales of leading Polish brewers should rise as the country's young urban professionals gradually switch from vodka to beer. \"The perspective on growth is such that reasonably we can think that somewhere between 65 and 80 litres per year is certainly reachable,\" van Boxmeer said on Polish per-capita beer consumption, currently around 40 litres. He said the 65-80-litre level could be reached in the next ten years and make Poland, with its 40-million population, Europe's third largest beer market after Germany and Britain. Van Boxmeer said Poland's top five brewers, which produce about 55 percent of the country's beer, could all raise market share as some of the numerous small brewers fall to competition from the large brewers with foreign investors. Zywiec is 31.8-percent owned by Heineken while Carlsberg has the same amount in Okocim. Earlier this year South African Breweries Ltd (SAB) bought strategic stakes in the unlisted Lech and Tychy brewers, which together hold more than 20 percent of the market, and Australia's Brewpole BV has a controlling stake in Poland's larges t brewery, Elbrewery Company Ltd. (EB). Van Boxmeer said the tough competition had prevented Zywiec from raising prices in line with inflation, which had added to the pressure on the firm's margins. He said advertising costs would also increase in the fight for market share. But he said the company's investment of more than $ 100 million already this decade, largely in production, would help position it to compete with such competitors as brewers from the neighbouring Czech Republic. Some analysts say cheaper but high-quality Czech imports could invade Poland once tariffs for CEFTA countries are lifted in 1998, but van Boxmeer says such a threat might be exaggerated despite the Czech beer market's overcapacity. \"I think Polish consumers in general are quite proud of their beers-- and I'm speaking about all the brands-- and as we make good beers... I think that this fidelity to our beers is a factor which can limit the Czech beers,\" he said. Van Boxmeer said Zywiec had its eye on Okocim, which has said it would start producing Carlsberg beer next year, but that Zywiec's potential production of Heineken was a medium-term possibility rather than a short-term one. He said his firm would be better off concentrating on its leading brand, Zywiec Full Light, which accounts for 85 percent of sales and is the country's largest-selling brand. \"You will not win the war of the Polish beer market with imported international brands,\" van Boxmeer said, adding that Heineken would remain an up-market import in Poland. Van Boxmeer also said Zywiec would be boosted by its recent shedding of soft drinks which only accounted for about three percent of the firm's overall sales and for which 7.6 million zlotys in provisions had already been made.-- Warsaw Newsroom+48 22 653 9700\ntype(text): \nspans: [{'start': 56, 'length': 6}, {'start': 317, 'length': 16}, {'start': 3473, 'length': 6}, {'start': 3207, 'length': 6}, {'start': 1704, 'length': 6}, {'start': 2910, 'length': 11}, {'start': 1730, 'length': 27}, {'start': 0, 'length': 16}, {'start': 2927, 'length': 6}, {'start': 1900, 'length': 11}, {'start': 3428, 'length': 8}, {'start': 1330, 'length': 6}, {'start': 3481, 'length': 11}, {'start': 2048, 'length': 6}, {'start': 2538, 'length': 5}, {'start': 2949, 'length': 6}, {'start': 1888, 'length': 9}, {'start': 2997, 'length': 9}, {'start': 883, 'length': 6}, {'start': 2428, 'length': 14}, {'start': 2685, 'length': 6}, {'start': 2642, 'length': 5}, {'start': 2582, 'length': 7}, {'start': 777, 'length': 11}, {'start': 1656, 'length': 8}, {'start': 2514, 'length': 6}, {'start': 1371, 'length': 7}, {'start': 1939, 'length': 6}, {'start': 1990, 'length': 2}, {'start': 3214, 'length': 10}, {'start': 1290, 'length': 6}, {'start': 1813, 'length': 5}, {'start': 1132, 'length': 7}, {'start': 2487, 'length': 5}, {'start': 1759, 'length': 3}, {'start': 1148, 'length': 6}, {'start': 2887, 'length': 5}, {'start': 1392, 'length': 11}, {'start': 74, 'length': 6}, {'start': 3342, 'length': 6}, {'start': 1383, 'length': 7}, {'start': 615, 'length': 6}, {'start': 88, 'length': 6}, {'start': 3503, 'length': 6}, {'start': 42, 'length': 13}, {'start': 1966, 'length': 22}, {'start': 3032, 'length': 6}, {'start': 3402, 'length': 7}, {'start': 1624, 'length': 6}, {'start': 1804, 'length': 4}, {'start': 1995, 'length': 11}, {'start': 3065, 'length': 8}, {'start': 3710, 'length': 15}, {'start': 1671, 'length': 9}, {'start': 642, 'length': 30}, {'start': 339, 'length': 7}, {'start': 1409, 'length': 6}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":509,\"length\":16},{\"start\":0,\"length\":5},{\"start\":1030,\"length\":5},{\"start\":134,\"length\":12},{\"start\":118,\"length\":5},{\"start\":287,\"length\":5},{\"start\":49,\"length\":6},{\"start\":933,\"length\":5},{\"start\":736,\"length\":5}],\"text\":\"HAVEL HAS TRAECHEOTOMY AFTER CONDITIONS WORSENS. PRAGUE 1996-12-06 Doctors performed an emergency tracheotomy to help Czech President Vaclav Havel breathe after cancer surgery on his lungs earlier this week, a spokesman said on Friday. He said that the procedure to insert a device into Havel\\\\u0027s throat, done after his breathing worsened on Thursday, had helped, and the president\\\\u0027s condition significantly improved. \\\\\"A worsening in the president\\\\u0027s lung functions took place yesterday,\\\\\" presidential spokesman Ladlislav Spacek said in a statement. \\\\\"A tracheotomy was performed and supportive breathing was installed through the help of a breathing device,\\\\\" he said. \\\\\"After these steps, the president\\\\u0027s condition signigicantly improved.\\\\\" Havel has been recovering from surgery on Monday which removed a small malignant tumour and half of his right lung. Doctors after the operation said that they had caught the cancer early, and that Havel could fully recover from the surgery within six weeks. His spokesman said on Thursday that Havel, 60 and a heavy smoker, had also developed a slight case of pneumonia in the left lung.\"}'\ntext: HAVEL HAS TRAECHEOTOMY AFTER CONDITIONS WORSENS. PRAGUE 1996-12-06 Doctors performed an emergency tracheotomy to help Czech President Vaclav Havel breathe after cancer surgery on his lungs earlier this week, a spokesman said on Friday. He said that the procedure to insert a device into Havel's throat, done after his breathing worsened on Thursday, had helped, and the president's condition significantly improved. \"A worsening in the president's lung functions took place yesterday,\" presidential spokesman Ladlislav Spacek said in a statement. \"A tracheotomy was performed and supportive breathing was installed through the help of a breathing device,\" he said. \"After these steps, the president's condition signigicantly improved.\" Havel has been recovering from surgery on Monday which removed a small malignant tumour and half of his right lung. Doctors after the operation said that they had caught the cancer early, and that Havel could fully recover from the surgery within six weeks. His spokesman said on Thursday that Havel, 60 and a heavy smoker, had also developed a slight case of pneumonia in the left lung.\ntype(text): \nspans: [{'start': 509, 'length': 16}, {'start': 0, 'length': 5}, {'start': 1030, 'length': 5}, {'start': 134, 'length': 12}, {'start': 118, 'length': 5}, {'start': 287, 'length': 5}, {'start': 49, 'length': 6}, {'start': 933, 'length': 5}, {'start': 736, 'length': 5}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":48,\"length\":6},{\"start\":307,\"length\":3},{\"start\":73,\"length\":23},{\"start\":165,\"length\":4},{\"start\":0,\"length\":5},{\"start\":316,\"length\":7},{\"start\":70,\"length\":2}],\"text\":\"UK-US open skies talks end, no date to restart. LONDON 1996-12-06 The UK Department of Transport on Friday said that the latest round of \\\\\"open skies\\\\\" talks with the U.S. had ended with no deal on liberalising the transatlantic flight market and no date set for when talks would restart. A spokesman for the DOT told Reuters \\\\\"We have had talks towards concluding a new air service agreement which would produce liberalisation... useful progress was made on a number of issues, but not all. No date has been set for further talks.\\\\\"\"}'\ntext: UK-US open skies talks end, no date to restart. LONDON 1996-12-06 The UK Department of Transport on Friday said that the latest round of \"open skies\" talks with the U.S. had ended with no deal on liberalising the transatlantic flight market and no date set for when talks would restart. A spokesman for the DOT told Reuters \"We have had talks towards concluding a new air service agreement which would produce liberalisation... useful progress was made on a number of issues, but not all. No date has been set for further talks.\"\ntype(text): \nspans: [{'start': 48, 'length': 6}, {'start': 307, 'length': 3}, {'start': 73, 'length': 23}, {'start': 165, 'length': 4}, {'start': 0, 'length': 5}, {'start': 316, 'length': 7}, {'start': 70, 'length': 2}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":37,\"length\":6},{\"start\":102,\"length\":6},{\"start\":55,\"length\":16},{\"start\":29,\"length\":6},{\"start\":0,\"length\":13},{\"start\":285,\"length\":7}],\"text\":\"Tambang Timah at $ 15.575 in London. LONDON 1996-12-07 PT Tambang Timah closed at $ 15.575 per GDR in London on Friday. It recorded the day\\\\u0027s low of $ 15.475 and the day\\\\u0027s high of $ 15.725. It closed at $ 15.80 on Thursday. One Global Depository Receipt represents 10 common shares.-- Jakarta newsroom+6221 384-6364\"}'\ntext: Tambang Timah at $ 15.575 in London. LONDON 1996-12-07 PT Tambang Timah closed at $ 15.575 per GDR in London on Friday. It recorded the day's low of $ 15.475 and the day's high of $ 15.725. It closed at $ 15.80 on Thursday. One Global Depository Receipt represents 10 common shares.-- Jakarta newsroom+6221 384-6364\ntype(text): \nspans: [{'start': 37, 'length': 6}, {'start': 102, 'length': 6}, {'start': 55, 'length': 16}, {'start': 29, 'length': 6}, {'start': 0, 'length': 13}, {'start': 285, 'length': 7}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":270,\"length\":7},{\"start\":26,\"length\":6},{\"start\":44,\"length\":27},{\"start\":0,\"length\":6},{\"start\":18,\"length\":6},{\"start\":99,\"length\":6},{\"start\":73,\"length\":6}],\"text\":\"Telkom at $ 35 in London. LONDON 1996-12-07 PT Telekomunikasi Indonesia (Telkom) closed at $ 35 in London on Friday. It recorded the day\\\\u0027s low of $ 34.475 and the day\\\\u0027s high of $ 35.375. Its previous close on Thursday as $ 35.63. One ADS represents 20 ordinary shares-- Jakarta newsroom+6221 384-6364.\"}'\ntext: Telkom at $ 35 in London. LONDON 1996-12-07 PT Telekomunikasi Indonesia (Telkom) closed at $ 35 in London on Friday. It recorded the day's low of $ 34.475 and the day's high of $ 35.375. Its previous close on Thursday as $ 35.63. One ADS represents 20 ordinary shares-- Jakarta newsroom+6221 384-6364.\ntype(text): \nspans: [{'start': 270, 'length': 7}, {'start': 26, 'length': 6}, {'start': 44, 'length': 27}, {'start': 0, 'length': 6}, {'start': 18, 'length': 6}, {'start': 99, 'length': 6}, {'start': 73, 'length': 6}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":177,\"length\":7},{\"start\":126,\"length\":21},{\"start\":41,\"length\":7},{\"start\":711,\"length\":16},{\"start\":19,\"length\":10},{\"start\":698,\"length\":7}],\"text\":\"Woman charged over N. Ireland arms find. BELFAST 1996-12-06 A woman was charged on Friday with terrorist offences after three Irish Republican Army mortar bombs were found in a Belfast house, police said. Police said the bombs were found hidden with incendiaries and ammunition that were blocked up behind a kitchen wall. The 35-year-old woman was charged with possession of explosives with intent to endanger life and making a house available for the purpose of terrorism, police said. She will appear in court on Saturday. Her name was not released. Security forces said the bombs may have been intended for use in a pre-Christmas bombing campaign by the guerrilla group that is battling to oust Britain from Northern Ireland.\"}'\ntext: Woman charged over N. Ireland arms find. BELFAST 1996-12-06 A woman was charged on Friday with terrorist offences after three Irish Republican Army mortar bombs were found in a Belfast house, police said. Police said the bombs were found hidden with incendiaries and ammunition that were blocked up behind a kitchen wall. The 35-year-old woman was charged with possession of explosives with intent to endanger life and making a house available for the purpose of terrorism, police said. She will appear in court on Saturday. Her name was not released. Security forces said the bombs may have been intended for use in a pre-Christmas bombing campaign by the guerrilla group that is battling to oust Britain from Northern Ireland.\ntype(text): \nspans: [{'start': 177, 'length': 7}, {'start': 126, 'length': 21}, {'start': 41, 'length': 7}, {'start': 711, 'length': 16}, {'start': 19, 'length': 10}, {'start': 698, 'length': 7}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":0,\"length\":7},{\"start\":67,\"length\":6},{\"start\":673,\"length\":22},{\"start\":535,\"length\":14},{\"start\":33,\"length\":8},{\"start\":2091,\"length\":15},{\"start\":491,\"length\":7},{\"start\":232,\"length\":7},{\"start\":1405,\"length\":15},{\"start\":1962,\"length\":15},{\"start\":2319,\"length\":20},{\"start\":2175,\"length\":17},{\"start\":155,\"length\":14},{\"start\":916,\"length\":2},{\"start\":2291,\"length\":16},{\"start\":211,\"length\":17},{\"start\":2557,\"length\":22},{\"start\":909,\"length\":2},{\"start\":363,\"length\":8},{\"start\":242,\"length\":33},{\"start\":1453,\"length\":8},{\"start\":663,\"length\":7},{\"start\":2369,\"length\":8},{\"start\":507,\"length\":13},{\"start\":1255,\"length\":24},{\"start\":334,\"length\":18},{\"start\":1378,\"length\":8},{\"start\":1007,\"length\":2},{\"start\":1111,\"length\":6},{\"start\":1445,\"length\":6},{\"start\":1664,\"length\":15},{\"start\":967,\"length\":14},{\"start\":2111,\"length\":8},{\"start\":89,\"length\":7},{\"start\":2017,\"length\":5},{\"start\":1014,\"length\":4},{\"start\":1358,\"length\":15},{\"start\":187,\"length\":19},{\"start\":2165,\"length\":6},{\"start\":52,\"length\":14},{\"start\":617,\"length\":4},{\"start\":2054,\"length\":14},{\"start\":1472,\"length\":9},{\"start\":2351,\"length\":15},{\"start\":2515,\"length\":4}],\"text\":\"Britain sets conditions to clear American alliance. Edna Fernandes LONDON 1996-12-06 The British government warned Friday that it would refer the proposed trans-Atlantic alliance between British Airways Plc and American Airlines to Britain\\\\u0027s Monopolies and Mergers Commission unless the carriers complied with a number of conditions. Trade and Industry Secretary Ian Lang added that even if the conditions were met by both airlines, final clearance would hinge on an open skies deal between Britain and the United States to liberalise trans-Atlantic air traffic, which would create greater competition on the routes. Lang said he supported conditions proposed by Britain\\\\u0027s Office of Fair Trading, which was asked to examine the case last month. \\\\\"I agree... that without suitable undertakings the alliance would be likely to lead to a significant loss of actual and potential passengers, on those routes where BA and AA currently compete and for all passengers on the trans-Atlantic market route between the UK and U.S.,\\\\\" he said. His comments came just minutes after the latest set of open skies talks ended in London with no deal signed. Industry sources said there was no new date for fresh talks and blamed the deadlock on uncertainty over whether the British Airways-American deal would be cleared. The conditions for clearance of the alliance were that British Airways and American drop 168 slots at London Heathrow airport, the busiest in Europe. American\\\\u0027s parent, AMR Corp., said it did not view the terms as a \\\\\"deal breaker.\\\\\" However, it called the conditions \\\\\"more severe\\\\\" than those imposed by other regulatory authorities on similar airline alliances. British Airways\\\\u0027s initial response was that \\\\\"unconditional divestiture of slots is unprecedented and if done it must be on the basis of fair market value.\\\\\" It added that it would be \\\\\"prepared to take reasonable steps to assist the introduction of additional competition.\\\\\" The government also wants British Airways to drop a clause in its agreement with USAir that bars it from competing on trans-Atlantic routes, and said both British Airways and American should be prepared to reduce services on the London to Dallas-Fort Worth route in the event that a new entrant wishes to enter. It also suggested losing some slots on the London-to-Boston route. The Office of Fair Trade called for British Airways / American to allow third-party access to their joint frequent flyer programme where the applicant does not have access to an equivalent programme. Lang said responses should be made to the Office of Fair Trading by Jan. 10, 1997.\"}'\ntext: Britain sets conditions to clear American alliance. Edna Fernandes LONDON 1996-12-06 The British government warned Friday that it would refer the proposed trans-Atlantic alliance between British Airways Plc and American Airlines to Britain's Monopolies and Mergers Commission unless the carriers complied with a number of conditions. Trade and Industry Secretary Ian Lang added that even if the conditions were met by both airlines, final clearance would hinge on an open skies deal between Britain and the United States to liberalise trans-Atlantic air traffic, which would create greater competition on the routes. Lang said he supported conditions proposed by Britain's Office of Fair Trading, which was asked to examine the case last month. \"I agree... that without suitable undertakings the alliance would be likely to lead to a significant loss of actual and potential passengers, on those routes where BA and AA currently compete and for all passengers on the trans-Atlantic market route between the UK and U.S.,\" he said. His comments came just minutes after the latest set of open skies talks ended in London with no deal signed. Industry sources said there was no new date for fresh talks and blamed the deadlock on uncertainty over whether the British Airways-American deal would be cleared. The conditions for clearance of the alliance were that British Airways and American drop 168 slots at London Heathrow airport, the busiest in Europe. American's parent, AMR Corp., said it did not view the terms as a \"deal breaker.\" However, it called the conditions \"more severe\" than those imposed by other regulatory authorities on similar airline alliances. British Airways's initial response was that \"unconditional divestiture of slots is unprecedented and if done it must be on the basis of fair market value.\" It added that it would be \"prepared to take reasonable steps to assist the introduction of additional competition.\" The government also wants British Airways to drop a clause in its agreement with USAir that bars it from competing on trans-Atlantic routes, and said both British Airways and American should be prepared to reduce services on the London to Dallas-Fort Worth route in the event that a new entrant wishes to enter. It also suggested losing some slots on the London-to-Boston route. The Office of Fair Trade called for British Airways / American to allow third-party access to their joint frequent flyer programme where the applicant does not have access to an equivalent programme. Lang said responses should be made to the Office of Fair Trading by Jan. 10, 1997.\ntype(text): \nspans: [{'start': 0, 'length': 7}, {'start': 67, 'length': 6}, {'start': 673, 'length': 22}, {'start': 535, 'length': 14}, {'start': 33, 'length': 8}, {'start': 2091, 'length': 15}, {'start': 491, 'length': 7}, {'start': 232, 'length': 7}, {'start': 1405, 'length': 15}, {'start': 1962, 'length': 15}, {'start': 2319, 'length': 20}, {'start': 2175, 'length': 17}, {'start': 155, 'length': 14}, {'start': 916, 'length': 2}, {'start': 2291, 'length': 16}, {'start': 211, 'length': 17}, {'start': 2557, 'length': 22}, {'start': 909, 'length': 2}, {'start': 363, 'length': 8}, {'start': 242, 'length': 33}, {'start': 1453, 'length': 8}, {'start': 663, 'length': 7}, {'start': 2369, 'length': 8}, {'start': 507, 'length': 13}, {'start': 1255, 'length': 24}, {'start': 334, 'length': 18}, {'start': 1378, 'length': 8}, {'start': 1007, 'length': 2}, {'start': 1111, 'length': 6}, {'start': 1445, 'length': 6}, {'start': 1664, 'length': 15}, {'start': 967, 'length': 14}, {'start': 2111, 'length': 8}, {'start': 89, 'length': 7}, {'start': 2017, 'length': 5}, {'start': 1014, 'length': 4}, {'start': 1358, 'length': 15}, {'start': 187, 'length': 19}, {'start': 2165, 'length': 6}, {'start': 52, 'length': 14}, {'start': 617, 'length': 4}, {'start': 2054, 'length': 14}, {'start': 1472, 'length': 9}, {'start': 2351, 'length': 15}, {'start': 2515, 'length': 4}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":529,\"length\":5},{\"start\":1254,\"length\":9},{\"start\":1526,\"length\":3},{\"start\":1488,\"length\":4},{\"start\":583,\"length\":6},{\"start\":665,\"length\":5},{\"start\":539,\"length\":7},{\"start\":1242,\"length\":6},{\"start\":1042,\"length\":8},{\"start\":1503,\"length\":7},{\"start\":33,\"length\":3},{\"start\":50,\"length\":6},{\"start\":1398,\"length\":3},{\"start\":1139,\"length\":3},{\"start\":385,\"length\":5},{\"start\":594,\"length\":5},{\"start\":139,\"length\":3},{\"start\":1234,\"length\":4}],\"text\":\"Med oil products mostly lower as Elf strike ends. LONDON 1996-12-06 Mediterranean oil products were steady to mostly lower on Friday after Elf refinery workers voted to end their nine-day strike. Gas oil erased Thursday\\\\u0027s gains, plunging $ 5.50 a tonne in line with the screen. Volume was very thin and market remained long, with premiums down $ 1 at about high cif quotes+$0.50 basis Genoa. \\\\\"The sharp moves on the screen make everyone nervous,\\\\\" a trader said. Trades were discussed in 0.2, 0.5 and one percent heating oil into Syria and Lebanon and there were fresh inquiries from France and Spain for low sulphur diesel. Interest remains focussed on a tender by India for a second purchase of high speed diesel for January delivery. Fuel oil lost ground sharply with weaker crude, but also suffered from some pricing pressure. High sulphur cracked fuel lost about $ 3 to $ 109-111 fob Med with several cargoes threatening to overhang the market. The chance of material heading north, talked earlier this week, may be in jeopardy now since American fuel oil is expected to head transatlantic following outages at two coking units in the U.S. Up to 165,000 tonnes of fuel will have to find a new home and with the arbitrage from the U.S. to Europe open Rotterdam is a prime candidate. Low sulphur prices were lower with cif Med pegged in the mid to low $ 140s. Gasoline prices fell after striking Elf refinery workers voted to go back to work, traders said. But an open arbitrage to the U.S. and tight Italian supplies after Elf scooped up Med material over the last week, continued to underpin prices into next week.\"}'\ntext: Med oil products mostly lower as Elf strike ends. LONDON 1996-12-06 Mediterranean oil products were steady to mostly lower on Friday after Elf refinery workers voted to end their nine-day strike. Gas oil erased Thursday's gains, plunging $ 5.50 a tonne in line with the screen. Volume was very thin and market remained long, with premiums down $ 1 at about high cif quotes+$0.50 basis Genoa. \"The sharp moves on the screen make everyone nervous,\" a trader said. Trades were discussed in 0.2, 0.5 and one percent heating oil into Syria and Lebanon and there were fresh inquiries from France and Spain for low sulphur diesel. Interest remains focussed on a tender by India for a second purchase of high speed diesel for January delivery. Fuel oil lost ground sharply with weaker crude, but also suffered from some pricing pressure. High sulphur cracked fuel lost about $ 3 to $ 109-111 fob Med with several cargoes threatening to overhang the market. The chance of material heading north, talked earlier this week, may be in jeopardy now since American fuel oil is expected to head transatlantic following outages at two coking units in the U.S. Up to 165,000 tonnes of fuel will have to find a new home and with the arbitrage from the U.S. to Europe open Rotterdam is a prime candidate. Low sulphur prices were lower with cif Med pegged in the mid to low $ 140s. Gasoline prices fell after striking Elf refinery workers voted to go back to work, traders said. But an open arbitrage to the U.S. and tight Italian supplies after Elf scooped up Med material over the last week, continued to underpin prices into next week.\ntype(text): \nspans: [{'start': 529, 'length': 5}, {'start': 1254, 'length': 9}, {'start': 1526, 'length': 3}, {'start': 1488, 'length': 4}, {'start': 583, 'length': 6}, {'start': 665, 'length': 5}, {'start': 539, 'length': 7}, {'start': 1242, 'length': 6}, {'start': 1042, 'length': 8}, {'start': 1503, 'length': 7}, {'start': 33, 'length': 3}, {'start': 50, 'length': 6}, {'start': 1398, 'length': 3}, {'start': 1139, 'length': 3}, {'start': 385, 'length': 5}, {'start': 594, 'length': 5}, {'start': 139, 'length': 3}, {'start': 1234, 'length': 4}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":807,\"length\":7},{\"start\":1239,\"length\":11},{\"start\":301,\"length\":7},{\"start\":35,\"length\":6},{\"start\":175,\"length\":7},{\"start\":265,\"length\":19},{\"start\":725,\"length\":5},{\"start\":317,\"length\":10},{\"start\":918,\"length\":8},{\"start\":26,\"length\":7}],\"text\":\"New meningitis scare hits Britain. LONDON 1996-12-06 A boy has died from meningitis and a girl from the same school has contracted the disease in the second such scare to hit Britain in as many weeks, health authorities said on Friday. The 16-year-old who attended Sale Grammar School in the northern England city of Manchester died less than a day after becoming ill. The 15-year-old girl is also suffering from the disease and hospital officials described her condition as serious. \\\\\"At the moment there is no evidence the two cases are linked. However, we are assuming they are as a precaution for the time being,\\\\\" a spokeswoman said. The more than 1,260 students at the school are being given antibiotics as a precaution. Wales grappled with its own cluster of meningitis cases on a university campus in Cardiff. At least two people have died and hundreds have been vaccinated in an effort to contain the virus. In Scotland, eight people have died and hundreds more are fighting a widespread food-poisoning outbreak. A health authority spokeswoman said 78 people suspected of having the disease, including 64 confirmed cases, were still being treated. Three were listed in poor condition. More than 290 people have reported symptoms in Lanarkshire county, the worst-hit area, since the outbreak first came to light after people ate tainted meat pies at a pensioners \\\\u0027 lunch.\"}'\ntext: New meningitis scare hits Britain. LONDON 1996-12-06 A boy has died from meningitis and a girl from the same school has contracted the disease in the second such scare to hit Britain in as many weeks, health authorities said on Friday. The 16-year-old who attended Sale Grammar School in the northern England city of Manchester died less than a day after becoming ill. The 15-year-old girl is also suffering from the disease and hospital officials described her condition as serious. \"At the moment there is no evidence the two cases are linked. However, we are assuming they are as a precaution for the time being,\" a spokeswoman said. The more than 1,260 students at the school are being given antibiotics as a precaution. Wales grappled with its own cluster of meningitis cases on a university campus in Cardiff. At least two people have died and hundreds have been vaccinated in an effort to contain the virus. In Scotland, eight people have died and hundreds more are fighting a widespread food-poisoning outbreak. A health authority spokeswoman said 78 people suspected of having the disease, including 64 confirmed cases, were still being treated. Three were listed in poor condition. More than 290 people have reported symptoms in Lanarkshire county, the worst-hit area, since the outbreak first came to light after people ate tainted meat pies at a pensioners ' lunch.\ntype(text): \nspans: [{'start': 807, 'length': 7}, {'start': 1239, 'length': 11}, {'start': 301, 'length': 7}, {'start': 35, 'length': 6}, {'start': 175, 'length': 7}, {'start': 265, 'length': 19}, {'start': 725, 'length': 5}, {'start': 317, 'length': 10}, {'start': 918, 'length': 8}, {'start': 26, 'length': 7}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":300,\"length\":5},{\"start\":729,\"length\":5},{\"start\":137,\"length\":12},{\"start\":1001,\"length\":16},{\"start\":50,\"length\":6},{\"start\":863,\"length\":5},{\"start\":1077,\"length\":5},{\"start\":436,\"length\":5},{\"start\":68,\"length\":7},{\"start\":8,\"length\":20},{\"start\":794,\"length\":16},{\"start\":475,\"length\":5},{\"start\":157,\"length\":10},{\"start\":748,\"length\":5},{\"start\":0,\"length\":5},{\"start\":91,\"length\":10},{\"start\":454,\"length\":14}],\"text\":\"Major\\\\u0027s office-Conservatives still have majority. LONDON 1996-12-06 British Prime Minister John Major\\\\u0027s office said on Friday that rebel Conservative MP Sir John Gorst had not \\\\\"resigned the whip\\\\\" (quit the parliamentary party) and the government still had a majority in the 651-seat parliament. \\\\\"He (Gorst) isreserving the right not to cooperate, but he has not resigned the whip. The government still has a majority,\\\\\" a spokesman from Major\\\\u0027s office in Downing Street said. Gorst\\\\u0027s office said later the MP would not feel himself obliged to vote with the government. He said at one point during a press conference : \\\\\"I have seen my whip (party manager) for next week which, of course, does n\\\\u0027t mean very much to me now.\\\\\" Before Gorst\\\\u0027s statement, Major had a one-seat majority in the 651-seat House of Commons lower house of parliament. In his formal statement, Gorst said : \\\\\"I am today withdrawing my cooperation from the government and shall not treat the\\\\\" whip \\\\u0027 as either a summons to attend the House of Commons or as placing me under any obligation to vote as advised. \\\\\"Gorst resigned over a hospital closure in his constituency.\"}'\ntext: Major's office-Conservatives still have majority. LONDON 1996-12-06 British Prime Minister John Major's office said on Friday that rebel Conservative MP Sir John Gorst had not \"resigned the whip\" (quit the parliamentary party) and the government still had a majority in the 651-seat parliament. \"He (Gorst) isreserving the right not to cooperate, but he has not resigned the whip. The government still has a majority,\" a spokesman from Major's office in Downing Street said. Gorst's office said later the MP would not feel himself obliged to vote with the government. He said at one point during a press conference : \"I have seen my whip (party manager) for next week which, of course, does n't mean very much to me now.\" Before Gorst's statement, Major had a one-seat majority in the 651-seat House of Commons lower house of parliament. In his formal statement, Gorst said : \"I am today withdrawing my cooperation from the government and shall not treat the\" whip ' as either a summons to attend the House of Commons or as placing me under any obligation to vote as advised. \"Gorst resigned over a hospital closure in his constituency.\ntype(text): \nspans: [{'start': 300, 'length': 5}, {'start': 729, 'length': 5}, {'start': 137, 'length': 12}, {'start': 1001, 'length': 16}, {'start': 50, 'length': 6}, {'start': 863, 'length': 5}, {'start': 1077, 'length': 5}, {'start': 436, 'length': 5}, {'start': 68, 'length': 7}, {'start': 8, 'length': 20}, {'start': 794, 'length': 16}, {'start': 475, 'length': 5}, {'start': 157, 'length': 10}, {'start': 748, 'length': 5}, {'start': 0, 'length': 5}, {'start': 91, 'length': 10}, {'start': 454, 'length': 14}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":646,\"length\":13},{\"start\":89,\"length\":23},{\"start\":262,\"length\":3},{\"start\":449,\"length\":8},{\"start\":497,\"length\":4},{\"start\":0,\"length\":15},{\"start\":752,\"length\":15},{\"start\":43,\"length\":6},{\"start\":399,\"length\":22},{\"start\":425,\"length\":9},{\"start\":522,\"length\":24},{\"start\":462,\"length\":33},{\"start\":635,\"length\":6}],\"text\":\"Electronic Data bags flight data contract. LONDON 1996-12-06 Information technology firm Electronic Data Systems said on Friday it had bagged a contract for the first air traffic control project being funded under the Private Finance Initiative. In a statement, EDS said the contract would be in the region of 50 million stg. The contract involved upgrading the flight data processing system at the Oceanic Control Centre in Prestwick in south west Scotland for National Air Traffic Services Ltd (NATS), subsidiary of the Civil Aviation Authority. The system is responsible for the control of aircraft flying transatlantic routes from Europe and North America. The system, which would use satellite technology, is scheduled to enter service in 2000.-- London Newsroom+44-171-542 7717\"}'\ntext: Electronic Data bags flight data contract. LONDON 1996-12-06 Information technology firm Electronic Data Systems said on Friday it had bagged a contract for the first air traffic control project being funded under the Private Finance Initiative. In a statement, EDS said the contract would be in the region of 50 million stg. The contract involved upgrading the flight data processing system at the Oceanic Control Centre in Prestwick in south west Scotland for National Air Traffic Services Ltd (NATS), subsidiary of the Civil Aviation Authority. The system is responsible for the control of aircraft flying transatlantic routes from Europe and North America. The system, which would use satellite technology, is scheduled to enter service in 2000.-- London Newsroom+44-171-542 7717\ntype(text): \nspans: [{'start': 646, 'length': 13}, {'start': 89, 'length': 23}, {'start': 262, 'length': 3}, {'start': 449, 'length': 8}, {'start': 497, 'length': 4}, {'start': 0, 'length': 15}, {'start': 752, 'length': 15}, {'start': 43, 'length': 6}, {'start': 399, 'length': 22}, {'start': 425, 'length': 9}, {'start': 522, 'length': 24}, {'start': 462, 'length': 33}, {'start': 635, 'length': 6}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":0,\"length\":4},{\"start\":112,\"length\":12},{\"start\":169,\"length\":9},{\"start\":63,\"length\":9},{\"start\":283,\"length\":16},{\"start\":153,\"length\":11},{\"start\":393,\"length\":15},{\"start\":310,\"length\":22},{\"start\":223,\"length\":11},{\"start\":34,\"length\":21}],\"text\":\"RTRS - Cricket - Play restarts in Australia-West Indies match. MELBOURNE 1996-12-06 Play restarted in the first World Series limited overs match between West Indies and Australia after a rain delay of 50 minutes on Friday. West Indies resumed their innings on 53 for two with opener Sherwin Campbell on 25 and Shivnarine Chanderpaul 10. Rain earlier delayed the start of play by 30 minutes.-- Sydney Newsroom 61-2 9373-1800\"}'\ntext: RTRS - Cricket - Play restarts in Australia-West Indies match. MELBOURNE 1996-12-06 Play restarted in the first World Series limited overs match between West Indies and Australia after a rain delay of 50 minutes on Friday. West Indies resumed their innings on 53 for two with opener Sherwin Campbell on 25 and Shivnarine Chanderpaul 10. Rain earlier delayed the start of play by 30 minutes.-- Sydney Newsroom 61-2 9373-1800\ntype(text): \nspans: [{'start': 0, 'length': 4}, {'start': 112, 'length': 12}, {'start': 169, 'length': 9}, {'start': 63, 'length': 9}, {'start': 283, 'length': 16}, {'start': 153, 'length': 11}, {'start': 393, 'length': 15}, {'start': 310, 'length': 22}, {'start': 223, 'length': 11}, {'start': 34, 'length': 21}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":215,\"length\":11},{\"start\":77,\"length\":8},{\"start\":48,\"length\":7},{\"start\":199,\"length\":8},{\"start\":10,\"length\":8},{\"start\":57,\"length\":8},{\"start\":91,\"length\":11},{\"start\":24,\"length\":11}],\"text\":\"Cricket - Pakistan beat New Zealand by 46 runs. SIALKOT, Pakistan 1996-12-06 Pakistan beat New Zealand by 46 runs on Friday to take an unbeatable 2-0 lead in the three-match one-day series. Scores : Pakistan 277-9, New Zealand 231\"}'\ntext: Cricket - Pakistan beat New Zealand by 46 runs. SIALKOT, Pakistan 1996-12-06 Pakistan beat New Zealand by 46 runs on Friday to take an unbeatable 2-0 lead in the three-match one-day series. Scores : Pakistan 277-9, New Zealand 231\ntype(text): \nspans: [{'start': 215, 'length': 11}, {'start': 77, 'length': 8}, {'start': 48, 'length': 7}, {'start': 199, 'length': 8}, {'start': 10, 'length': 8}, {'start': 57, 'length': 8}, {'start': 91, 'length': 11}, {'start': 24, 'length': 11}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":119,\"length\":8},{\"start\":50,\"length\":8},{\"start\":0,\"length\":13},{\"start\":780,\"length\":3},{\"start\":705,\"length\":2},{\"start\":847,\"length\":8},{\"start\":830,\"length\":13},{\"start\":786,\"length\":4},{\"start\":70,\"length\":13},{\"start\":672,\"length\":8},{\"start\":731,\"length\":8}],\"text\":\"Manitoba Pork forward contract PM prices - Dec 6. WINNIPEG 1996-12-06 Manitoba Pork closing forward contract prices in Canadian dollars per hundred lbs (Cwt) for Dec 6 including minimum guaranteed price-- CONTRACT PREVIOUS CLOSE PM CLOSE PM CLOSING RANGE DATE PM CLOSE FIXED MINIMUM AT 1230 CST Feb 97 79.94 79.67 75.55 77.01-81.80 Mar 97 76.37 76.12 72.02 73.47-78.24 Apr 97 74.13 74.69 70.59 72.04-76.81 May 97 76.51 77.07 72.97 74.42-79.19 Jun 97 77.53 77.24 73.17 74.62-79.35 Jul 97 74.45 74.01 69.95 71.39-76.12 Aug 97 72.41 72.07 68.01 69.45-74.18 Sep 97 69.18 69.24 65.17 66.61-71.34 Oct 97 68.00 68.05 63.98 65.43-70.16 Nov 97 68.00 68.05 63.98 65.43-70.16 Note : Manitoba Government Price Index (C$ per cwt) - Dec 4 87.16 Manitoba\\\\u0027s Hog Price Range : 84.00-86.00 per cwt CAN / U.S. DOLLAR EXCHANGE RATE : 1.3570 Source : Manitoba Pork. ((Winnipeg bureau 204-947-3548))\"}'\ntext: Manitoba Pork forward contract PM prices - Dec 6. WINNIPEG 1996-12-06 Manitoba Pork closing forward contract prices in Canadian dollars per hundred lbs (Cwt) for Dec 6 including minimum guaranteed price-- CONTRACT PREVIOUS CLOSE PM CLOSE PM CLOSING RANGE DATE PM CLOSE FIXED MINIMUM AT 1230 CST Feb 97 79.94 79.67 75.55 77.01-81.80 Mar 97 76.37 76.12 72.02 73.47-78.24 Apr 97 74.13 74.69 70.59 72.04-76.81 May 97 76.51 77.07 72.97 74.42-79.19 Jun 97 77.53 77.24 73.17 74.62-79.35 Jul 97 74.45 74.01 69.95 71.39-76.12 Aug 97 72.41 72.07 68.01 69.45-74.18 Sep 97 69.18 69.24 65.17 66.61-71.34 Oct 97 68.00 68.05 63.98 65.43-70.16 Nov 97 68.00 68.05 63.98 65.43-70.16 Note : Manitoba Government Price Index (C$ per cwt) - Dec 4 87.16 Manitoba's Hog Price Range : 84.00-86.00 per cwt CAN / U.S. DOLLAR EXCHANGE RATE : 1.3570 Source : Manitoba Pork. ((Winnipeg bureau 204-947-3548))\ntype(text): \nspans: [{'start': 119, 'length': 8}, {'start': 50, 'length': 8}, {'start': 0, 'length': 13}, {'start': 780, 'length': 3}, {'start': 705, 'length': 2}, {'start': 847, 'length': 8}, {'start': 830, 'length': 13}, {'start': 786, 'length': 4}, {'start': 70, 'length': 13}, {'start': 672, 'length': 8}, {'start': 731, 'length': 8}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":0,\"length\":8},{\"start\":408,\"length\":13},{\"start\":257,\"length\":11},{\"start\":394,\"length\":9},{\"start\":43,\"length\":8},{\"start\":428,\"length\":15},{\"start\":67,\"length\":20},{\"start\":147,\"length\":8}],\"text\":\"Canadian West Coast Vessel Loadings - CWB. WINNIPEG 1996-12-06 The Canadian Wheat Board reported six ships loading, 10 waiting and four due at the Canadian West Coast, as of Friday. The longest wait to load on the West Coast was 13 days. Two ship loaded in Thunder Bay, one waited and seven were due. Two ships loaded on the East Coast, three waited to load, six were due. Port Loading Waiting Vancouver 5 7 Prince Rupert 1 3 ((Gilbert Le Gras 204 947 3548))\"}'\ntext: Canadian West Coast Vessel Loadings - CWB. WINNIPEG 1996-12-06 The Canadian Wheat Board reported six ships loading, 10 waiting and four due at the Canadian West Coast, as of Friday. The longest wait to load on the West Coast was 13 days. Two ship loaded in Thunder Bay, one waited and seven were due. Two ships loaded on the East Coast, three waited to load, six were due. Port Loading Waiting Vancouver 5 7 Prince Rupert 1 3 ((Gilbert Le Gras 204 947 3548))\ntype(text): \nspans: [{'start': 0, 'length': 8}, {'start': 408, 'length': 13}, {'start': 257, 'length': 11}, {'start': 394, 'length': 9}, {'start': 43, 'length': 8}, {'start': 428, 'length': 15}, {'start': 67, 'length': 20}, {'start': 147, 'length': 8}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":101,\"length\":25},{\"start\":39,\"length\":8},{\"start\":0,\"length\":8},{\"start\":89,\"length\":8}],\"text\":\"New York timecharter fixtures - Dec 6. NEW YORK 1996-12-06 No new fixtures reported from New York.-- New York Commodities Desk+1 212 859 1640\"}'\ntext: New York timecharter fixtures - Dec 6. NEW YORK 1996-12-06 No new fixtures reported from New York.-- New York Commodities Desk+1 212 859 1640\ntype(text): \nspans: [{'start': 101, 'length': 25}, {'start': 39, 'length': 8}, {'start': 0, 'length': 8}, {'start': 89, 'length': 8}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":157,\"length\":11},{\"start\":172,\"length\":20},{\"start\":111,\"length\":9},{\"start\":0,\"length\":8},{\"start\":101,\"length\":7},{\"start\":46,\"length\":8},{\"start\":72,\"length\":14}],\"text\":\"New York coal / ore / scrap fixtures - Dec 6. NEW YORK 1996-12-06 ORE - Maritime Queen 70,000 tonnes Dampier / Kaohsiung 20-30/12 $ 5.25 fio 35,000 / 30,000 China Steel.-- New York Commodities Desk+1 212 859 1640\"}'\ntext: New York coal / ore / scrap fixtures - Dec 6. NEW YORK 1996-12-06 ORE - Maritime Queen 70,000 tonnes Dampier / Kaohsiung 20-30/12 $ 5.25 fio 35,000 / 30,000 China Steel.-- New York Commodities Desk+1 212 859 1640\ntype(text): \nspans: [{'start': 157, 'length': 11}, {'start': 172, 'length': 20}, {'start': 111, 'length': 9}, {'start': 0, 'length': 8}, {'start': 101, 'length': 7}, {'start': 46, 'length': 8}, {'start': 72, 'length': 14}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":119,\"length\":6},{\"start\":49,\"length\":8},{\"start\":145,\"length\":20},{\"start\":136,\"length\":5},{\"start\":101,\"length\":6},{\"start\":44,\"length\":3}],\"text\":\"Clean tankers fixtures and enquiries - 2321 GMT. NEW YORK 1996-12-06 FIXTURES - WESTERN HEMISPHERE - Danila 28.5 16/12 Caribs / up W224 Mobil.-- New York Commodities Desk, 212-859-1640\"}'\ntext: Clean tankers fixtures and enquiries - 2321 GMT. NEW YORK 1996-12-06 FIXTURES - WESTERN HEMISPHERE - Danila 28.5 16/12 Caribs / up W224 Mobil.-- New York Commodities Desk, 212-859-1640\ntype(text): \nspans: [{'start': 119, 'length': 6}, {'start': 49, 'length': 8}, {'start': 145, 'length': 20}, {'start': 136, 'length': 5}, {'start': 101, 'length': 6}, {'start': 44, 'length': 3}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":112,\"length\":10},{\"start\":230,\"length\":10},{\"start\":205,\"length\":5},{\"start\":48,\"length\":8},{\"start\":147,\"length\":13},{\"start\":125,\"length\":7},{\"start\":265,\"length\":7},{\"start\":78,\"length\":7},{\"start\":284,\"length\":6},{\"start\":163,\"length\":6},{\"start\":212,\"length\":7},{\"start\":192,\"length\":7},{\"start\":311,\"length\":25},{\"start\":88,\"length\":13},{\"start\":258,\"length\":5},{\"start\":43,\"length\":3},{\"start\":179,\"length\":10},{\"start\":140,\"length\":5},{\"start\":243,\"length\":7},{\"start\":302,\"length\":5},{\"start\":293,\"length\":3}],\"text\":\"Dirty tanker fixtures and enquiries - 2317 GMT. NEW YORK 1996-12-06 MIDEAST / RED SEA - Thai Resource 264 31/12 Ras Tanura / Red Sea W46.50 Mobil. MEDITERRANEAN - Lula I 85 25/12 Sidi Kreir / Augusta W100 Exxon. Spetses 139 17/12 Sidi Kreir / Augusta W97.50 Exxon. Mesipia 77.5 17/12 Bajaia / Fos W105 Exxon.-- New York Commodities Desk+1 212 859 1640\"}'\ntext: Dirty tanker fixtures and enquiries - 2317 GMT. NEW YORK 1996-12-06 MIDEAST / RED SEA - Thai Resource 264 31/12 Ras Tanura / Red Sea W46.50 Mobil. MEDITERRANEAN - Lula I 85 25/12 Sidi Kreir / Augusta W100 Exxon. Spetses 139 17/12 Sidi Kreir / Augusta W97.50 Exxon. Mesipia 77.5 17/12 Bajaia / Fos W105 Exxon.-- New York Commodities Desk+1 212 859 1640\ntype(text): \nspans: [{'start': 112, 'length': 10}, {'start': 230, 'length': 10}, {'start': 205, 'length': 5}, {'start': 48, 'length': 8}, {'start': 147, 'length': 13}, {'start': 125, 'length': 7}, {'start': 265, 'length': 7}, {'start': 78, 'length': 7}, {'start': 284, 'length': 6}, {'start': 163, 'length': 6}, {'start': 212, 'length': 7}, {'start': 192, 'length': 7}, {'start': 311, 'length': 25}, {'start': 88, 'length': 13}, {'start': 258, 'length': 5}, {'start': 43, 'length': 3}, {'start': 179, 'length': 10}, {'start': 140, 'length': 5}, {'start': 243, 'length': 7}, {'start': 302, 'length': 5}, {'start': 293, 'length': 3}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":965,\"length\":8},{\"start\":1039,\"length\":21},{\"start\":881,\"length\":13},{\"start\":920,\"length\":4},{\"start\":689,\"length\":9},{\"start\":30,\"length\":4},{\"start\":221,\"length\":8},{\"start\":50,\"length\":8},{\"start\":0,\"length\":3},{\"start\":1234,\"length\":11},{\"start\":70,\"length\":13},{\"start\":299,\"length\":14}],\"text\":\"NYC Jan refunding has its 1st Euro floating rate. NEW YORK 1996-12-06 New York City on Friday said that it planned a $ 775 million refunding for January that will include its first floating rate issue of taxable debt for European investors. A city official, who declined to be named, explained that Goldman, Sachs, which this summer was demoted to the second tier of the syndicate, proposed the floating rate issue and as a result was promoted to book runner for this offering. By selling the floating rate debt, the city hopes to establish a benchmark, the city official said, adding that it needed a large deal to accomplish this objective. The city in late June sold its first issue of Euronotes, a strategy that it says saved it $ 500,000 in interest costs, and it has been trying to build on this strategy of expanding the pool of potential investors since then. In November, New York City said it became the first U.S. municipality to offer bonds for sale in European markets by competitive bidding as it listed taxable bonds on the London Stock Exchange. The refunding planned for January also includes a $ 475 million tax-exempt offering. No specific date in January has been selected for the debt sale, the official added.-- Joan Gralla, 212-859-1654\"}'\ntext: NYC Jan refunding has its 1st Euro floating rate. NEW YORK 1996-12-06 New York City on Friday said that it planned a $ 775 million refunding for January that will include its first floating rate issue of taxable debt for European investors. A city official, who declined to be named, explained that Goldman, Sachs, which this summer was demoted to the second tier of the syndicate, proposed the floating rate issue and as a result was promoted to book runner for this offering. By selling the floating rate debt, the city hopes to establish a benchmark, the city official said, adding that it needed a large deal to accomplish this objective. The city in late June sold its first issue of Euronotes, a strategy that it says saved it $ 500,000 in interest costs, and it has been trying to build on this strategy of expanding the pool of potential investors since then. In November, New York City said it became the first U.S. municipality to offer bonds for sale in European markets by competitive bidding as it listed taxable bonds on the London Stock Exchange. The refunding planned for January also includes a $ 475 million tax-exempt offering. No specific date in January has been selected for the debt sale, the official added.-- Joan Gralla, 212-859-1654\ntype(text): \nspans: [{'start': 965, 'length': 8}, {'start': 1039, 'length': 21}, {'start': 881, 'length': 13}, {'start': 920, 'length': 4}, {'start': 689, 'length': 9}, {'start': 30, 'length': 4}, {'start': 221, 'length': 8}, {'start': 50, 'length': 8}, {'start': 0, 'length': 3}, {'start': 1234, 'length': 11}, {'start': 70, 'length': 13}, {'start': 299, 'length': 14}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":0,\"length\":4},{\"start\":212,\"length\":4},{\"start\":40,\"length\":10}],\"text\":\"USDA gross cutout hide and offal value. DES MOINES 1996-12-06 The hide and offal value from a typical slaughter steer for Friday was estimated at $ 9.54 per cwt live, dn 0.05 when compared to Thursday\\\\u0027s value. - USDA\"}'\ntext: USDA gross cutout hide and offal value. DES MOINES 1996-12-06 The hide and offal value from a typical slaughter steer for Friday was estimated at $ 9.54 per cwt live, dn 0.05 when compared to Thursday's value. - USDA\ntype(text): \nspans: [{'start': 0, 'length': 4}, {'start': 212, 'length': 4}, {'start': 40, 'length': 10}]\ntype(spans): \n127.0.0.1 - - [09/Feb/2018 12:41:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[{\"start\":2860,\"length\":7},{\"start\":2277,\"length\":7},{\"start\":2209,\"length\":5},{\"start\":2952,\"length\":8},{\"start\":1336,\"length\":8},{\"start\":232,\"length\":19},{\"start\":2074,\"length\":8},{\"start\":2928,\"length\":7},{\"start\":1738,\"length\":9},{\"start\":0,\"length\":7},{\"start\":924,\"length\":7},{\"start\":2198,\"length\":6},{\"start\":2349,\"length\":8},{\"start\":3132,\"length\":8},{\"start\":1566,\"length\":8},{\"start\":2727,\"length\":16},{\"start\":445,\"length\":8},{\"start\":253,\"length\":8},{\"start\":3047,\"length\":8},{\"start\":1758,\"length\":13},{\"start\":2458,\"length\":9},{\"start\":2161,\"length\":6},{\"start\":1974,\"length\":4},{\"start\":2880,\"length\":8},{\"start\":2124,\"length\":6},{\"start\":2132,\"length\":10},{\"start\":1639,\"length\":9},{\"start\":1577,\"length\":6},{\"start\":25,\"length\":8},{\"start\":1961,\"length\":11},{\"start\":1826,\"length\":5},{\"start\":105,\"length\":11},{\"start\":196,\"length\":8},{\"start\":3078,\"length\":7},{\"start\":368,\"length\":11},{\"start\":1426,\"length\":6},{\"start\":1452,\"length\":15},{\"start\":1414,\"length\":9},{\"start\":758,\"length\":7},{\"start\":433,\"length\":7},{\"start\":2979,\"length\":23},{\"start\":1836,\"length\":6},{\"start\":903,\"length\":7},{\"start\":816,\"length\":8},{\"start\":2153,\"length\":6},{\"start\":3307,\"length\":7},{\"start\":2144,\"length\":7},{\"start\":968,\"length\":8},{\"start\":1815,\"length\":9},{\"start\":3010,\"length\":8},{\"start\":682,\"length\":8},{\"start\":42,\"length\":17},{\"start\":2187,\"length\":9},{\"start\":1603,\"length\":12},{\"start\":1292,\"length\":17},{\"start\":339,\"length\":7},{\"start\":3418,\"length\":8},{\"start\":3438,\"length\":16},{\"start\":80,\"length\":19},{\"start\":2236,\"length\":10},{\"start\":2554,\"length\":7},{\"start\":1935,\"length\":8},{\"start\":569,\"length\":7},{\"start\":856,\"length\":6},{\"start\":60,\"length\":8},{\"start\":2647,\"length\":8},{\"start\":1585,\"length\":13},{\"start\":2169,\"length\":9},{\"start\":3276,\"length\":7},{\"start\":1314,\"length\":18},{\"start\":2216,\"length\":11},{\"start\":2180,\"length\":5},{\"start\":2329,\"length\":7},{\"start\":1800,\"length\":13},{\"start\":166,\"length\":26}],\"text\":\"Wall St speculates about Santa Fe savior. Brendan Intindola NEW YORK 1996-12-06 Homestake Mining Co tops Wall Street\\\\u0027s list as the most likely white knight buyer for Santa Fe Pacific Gold Corp if Santa Fe rejects unsolicited suitor Newmont Mining Corp. Santa Fe is so far mum on the more than $ 2 billion stock swap takeover proposal from Newmont, announced Thursday. Wall Street, since the bid, has speculated that any deal between Newmont and Santa Fe would be a \\\\\"bear hug,\\\\\" or a reluctantly negotiated agreement where the buyer is not necessarily a friendly suitor. Newmont said the companies have had previous contact, though declined to detail the encounters. Analysts predict Santa Fe will go to the highest bidder, and that if a rival buyer is found, Newmont may not be able to match its offer. They said the Santa Fe deal, which includes desirable Nevada mining territory, would only payoff for Newmont longer term. Newmont, in fact, will not benefit from the Santa Fe acquisition on an earnings basis for at least two years, which also limits its capacity to raise its offer. Any deal, friendly or hostile, would almost assuredly be a stock swap, which is necessary to preserve the tax-free, pooling-of-interest accounting, they said. Analysts and arbitrageurs immediately ruled out Barrick Gold Corp and Bre-X Minerals Ltd as Santa Fe saviors because they are locked in negotiations over their splitting Indonesia\\\\u0027s Busang vast gold deposit. Placer Dome Inc too was considered unlikley because it is focusing on geographic expansion in areas that do match Santa Fe\\\\u0027s Nevada, South America and Central Asia presence, they said. A Homestake spokesman was not immediately available to comment on speculation that it tops the list. Homestake, based in San Francisco, operates gold mines in the United States, Australia, Chile and Canada. Earnings in 1995 were $ 0.22 per share, or $ 30.3 million, on revenues of $ 746.3 million. Santa Fe is headquartered Albuquerque, N.M. and reported 1995 earnings of $ 0.30 per share, or $ 40 million, on revenues of $ 350 million. Santa Fe has mining and exploration operations in Nevada, California, Montana, Canada, Brazil, Australia, Chile, Kazakstan, Mexico and Ghana. PaineWebber analyst Marc Cohen said he lowered his rating on Newmont to neutral from attractive today because if Newmont merged with Santa Fe, investors would have to wait until the second half of 1998 to realize earnings accretion. \\\\\"I think Homestake could come in as a white knight, but how much is someone willing to come in above the Newmont number. One would have to outbid by at least 15 percent, but there is going to be a (Santa Fe) deal with someone,\\\\\" he said. \\\\\"Longer term, two to three years out, (a Newmont-Santa Fe deal) is positive, it does all the right things. But in the near-term it is, at worst, neutral,\\\\\" the analyst added. Newmont proposed to Santa Fe a stock-swap merger at a ratio of 0.33 Newmont shares for each Santa Fe shares. In Friday New York Stock Exchange trade, Newmonth was off 1/2 to 46-5/8 while Santa Fe added 1/4 to 15-1/8. \\\\\"Newmont said it wants to discuss a friendly deal with Santa Fe, which is almost always a euphemism for \\\\u0027 We have more money in our pocket, \\\\u0027\\\\\" said an arb, referring to a possible sweetened bid from Newmont. Two other arbs called Newmont\\\\u0027s move a \\\\\"a 32 cent bid\\\\\" because there is no formal tender offer, only the proposal letter \\\\\"mailed\\\\\" to Santa Fe\\\\u0027s board.-- Wall Street Desk, 212-859-1734.\"}'\ntext: Wall St speculates about Santa Fe savior. Brendan Intindola NEW YORK 1996-12-06 Homestake Mining Co tops Wall Street's list as the most likely white knight buyer for Santa Fe Pacific Gold Corp if Santa Fe rejects unsolicited suitor Newmont Mining Corp. Santa Fe is so far mum on the more than $ 2 billion stock swap takeover proposal from Newmont, announced Thursday. Wall Street, since the bid, has speculated that any deal between Newmont and Santa Fe would be a \"bear hug,\" or a reluctantly negotiated agreement where the buyer is not necessarily a friendly suitor. Newmont said the companies have had previous contact, though declined to detail the encounters. Analysts predict Santa Fe will go to the highest bidder, and that if a rival buyer is found, Newmont may not be able to match its offer. They said the Santa Fe deal, which includes desirable Nevada mining territory, would only payoff for Newmont longer term. Newmont, in fact, will not benefit from the Santa Fe acquisition on an earnings basis for at least two years, which also limits its capacity to raise its offer. Any deal, friendly or hostile, would almost assuredly be a stock swap, which is necessary to preserve the tax-free, pooling-of-interest accounting, they said. Analysts and arbitrageurs immediately ruled out Barrick Gold Corp and Bre-X Minerals Ltd as Santa Fe saviors because they are locked in negotiations over their splitting Indonesia's Busang vast gold deposit. Placer Dome Inc too was considered unlikley because it is focusing on geographic expansion in areas that do match Santa Fe's Nevada, South America and Central Asia presence, they said. A Homestake spokesman was not immediately available to comment on speculation that it tops the list. Homestake, based in San Francisco, operates gold mines in the United States, Australia, Chile and Canada. Earnings in 1995 were $ 0.22 per share, or $ 30.3 million, on revenues of $ 746.3 million. Santa Fe is headquartered Albuquerque, N.M. and reported 1995 earnings of $ 0.30 per share, or $ 40 million, on revenues of $ 350 million. Santa Fe has mining and exploration operations in Nevada, California, Montana, Canada, Brazil, Australia, Chile, Kazakstan, Mexico and Ghana. PaineWebber analyst Marc Cohen said he lowered his rating on Newmont to neutral from attractive today because if Newmont merged with Santa Fe, investors would have to wait until the second half of 1998 to realize earnings accretion. \"I think Homestake could come in as a white knight, but how much is someone willing to come in above the Newmont number. One would have to outbid by at least 15 percent, but there is going to be a (Santa Fe) deal with someone,\" he said. \"Longer term, two to three years out, (a Newmont-Santa Fe deal) is positive, it does all the right things. But in the near-term it is, at worst, neutral,\" the analyst added. Newmont proposed to Santa Fe a stock-swap merger at a ratio of 0.33 Newmont shares for each Santa Fe shares. In Friday New York Stock Exchange trade, Newmonth was off 1/2 to 46-5/8 while Santa Fe added 1/4 to 15-1/8. \"Newmont said it wants to discuss a friendly deal with Santa Fe, which is almost always a euphemism for ' We have more money in our pocket, '\" said an arb, referring to a possible sweetened bid from Newmont. Two other arbs called Newmont's move a \"a 32 cent bid\" because there is no formal tender offer, only the proposal letter \"mailed\" to Santa Fe's board.-- Wall Street Desk, 212-859-1734.\n\n" }, { "path": "gerbil-SpotWrapNifWS4Test/my_notes/python_server_format_el", "content": "\n\n\nfor TEST-A uri=http://www.aksw.org/gerbil/NifWebService/request_215, was the last one\n\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - BELARUS BEAT ESTONIA IN WORLD CUP QUALIFIER. MINSK 1996-08-31 Belarus beat Estonia 1-0 (halftime 1-0) in a World Cup soccer European group 4 qualifier on Saturday. Scorer : Vladimir Makovsky (35th) Attendance : 6,000\"}'\ntext: SOCCER - BELARUS BEAT ESTONIA IN WORLD CUP QUALIFIER. MINSK 1996-08-31 Belarus beat Estonia 1-0 (halftime 1-0) in a World Cup soccer European group 4 qualifier on Saturday. Scorer : Vladimir Makovsky (35th) Attendance : 6,000\ntype(text): \nspans: []\ntype(spans): \n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"Pirelli cables look to tap Chinese growth. David Jones MILAN 1996-08-30 Italian tyre and cables giant Pirelli on Friday announced its long-awaited move into China with a cables joint venture set to capitalise on the rapidly-growing Chinese telecommunications market. Pirelli is linking with Hong Kong-based group CITIC Pacific in a venture to be called the Wuxi Tong Ling Company Ltd, which will operate in partnership with a local industrial company at its existing factory in Wuxi, Jiangsu province, Shanghai. The partners will invest around $ 30 million in the existing copper cable plant at Wuxi to update technology and include optic fibre production, with annual turnover expected to reach $ 60 million within the next few years. The move marks a further move by Pirelli\\\\u0027s cables division to expand in the fast-growing Far Eastern developing markets with the group already present in Indonesia, India and Malaysia. \\\\\"This is really positive news for Pirelli, and I expect that it will produce one of the best half-year results in late September compared to other industrial Italian companies,\\\\\" said analyst Paula Buratti at Indosuez. She emphasised that the move was positive because Pirelli will have management control of the Chinese venture, and it also showed another example of Pirelli exporting its technical know-how to developing markets. Pirelli shares reacted favourable even though talks had been underway for some time and news about a venture had been widely expected. The shares rose 0.2 percent to 2,555 lire by 1350 GMT in an easier Milan stock market. This will be Pirelli\\\\u0027s first industrial involvement in a Chinese market where demand for telecommunication networks is expected to grow to 80-100 million new lines between 1996 and 2000, doubling demand for optical cables. China\\\\u0027s second largest telecoms operator Unicom already has a mandate from the central government to establish 15 million new phone lines by the year 2000, which will necessitate new trunk line systems and local distribution networks. \\\\\"The starting of this production base in China has for our group an undoubted strategic value, representing an important enhancement of our presence in Asia,\\\\\" said Pirelli SpA chairman and chief executive officer Marco Tonchetti Provera. Pirelli Cables has global sales of over $ 3.5 billion, and has become a large supplier of optic cables and systems to major telecoms carriers in the U.S., Europe and the Far East. CITIC Pacific is a major Hong Kong-listed company focusing on infrastruture, trading, distribution and property, with 28 percent of its 1995 profits coming from telecoms. It has investments in several industrial joint ventures in China.\"}'\ntext: Pirelli cables look to tap Chinese growth. David Jones MILAN 1996-08-30 Italian tyre and cables giant Pirelli on Friday announced its long-awaited move into China with a cables joint venture set to capitalise on the rapidly-growing Chinese telecommunications market. Pirelli is linking with Hong Kong-based group CITIC Pacific in a venture to be called the Wuxi Tong Ling Company Ltd, which will operate in partnership with a local industrial company at its existing factory in Wuxi, Jiangsu province, Shanghai. The partners will invest around $ 30 million in the existing copper cable plant at Wuxi to update technology and include optic fibre production, with annual turnover expected to reach $ 60 million within the next few years. The move marks a further move by Pirelli's cables division to expand in the fast-growing Far Eastern developing markets with the group already present in Indonesia, India and Malaysia. \"This is really positive news for Pirelli, and I expect that it will produce one of the best half-year results in late September compared to other industrial Italian companies,\" said analyst Paula Buratti at Indosuez. She emphasised that the move was positive because Pirelli will have management control of the Chinese venture, and it also showed another example of Pirelli exporting its technical know-how to developing markets. Pirelli shares reacted favourable even though talks had been underway for some time and news about a venture had been widely expected. The shares rose 0.2 percent to 2,555 lire by 1350 GMT in an easier Milan stock market. This will be Pirelli's first industrial involvement in a Chinese market where demand for telecommunication networks is expected to grow to 80-100 million new lines between 1996 and 2000, doubling demand for optical cables. China's second largest telecoms operator Unicom already has a mandate from the central government to establish 15 million new phone lines by the year 2000, which will necessitate new trunk line systems and local distribution networks. \"The starting of this production base in China has for our group an undoubted strategic value, representing an important enhancement of our presence in Asia,\" said Pirelli SpA chairman and chief executive officer Marco Tonchetti Provera. Pirelli Cables has global sales of over $ 3.5 billion, and has become a large supplier of optic cables and systems to major telecoms carriers in the U.S., Europe and the Far East. CITIC Pacific is a major Hong Kong-listed company focusing on infrastruture, trading, distribution and property, with 28 percent of its 1995 profits coming from telecoms. It has investments in several industrial joint ventures in China.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"Dutch bond futures revival delayed - EOE. AMSTERDAM 1996-08-30 A broad attempt to spur activity in Dutch bond futures has been delayed to give participants a chance to become familiar with the trading system, the European Options Exchange (EOE) said on Friday. Market-making in the rarely-traded FTO contract was expected to begin today, but an EOE spokesman said the 10 banks and brokers involved in the initiative needed time to get accustomed to changes in the electronic trading system. \\\\\"It\\\\u0027s not ready yet. We found it wise to take some time between the commitment to start and the actual start,\\\\\" EOE spokesman Lex van Drooge told Reuters. He said no date had been fixed yet for the start of price making in the 10-year contract, but the EOE had agreed to speak again to the participants in one to two weeks. Investors in Dutch bonds currently use German bond futures to hedge their portfolios because the FTO contract is so illiquid. A limited attempt to reinvigorate the contract two years ago failed.-- Amsterdam newsroom+31 20 504 5000, Fax+31 20 504 5040\"}'\ntext: Dutch bond futures revival delayed - EOE. AMSTERDAM 1996-08-30 A broad attempt to spur activity in Dutch bond futures has been delayed to give participants a chance to become familiar with the trading system, the European Options Exchange (EOE) said on Friday. Market-making in the rarely-traded FTO contract was expected to begin today, but an EOE spokesman said the 10 banks and brokers involved in the initiative needed time to get accustomed to changes in the electronic trading system. \"It's not ready yet. We found it wise to take some time between the commitment to start and the actual start,\" EOE spokesman Lex van Drooge told Reuters. He said no date had been fixed yet for the start of price making in the 10-year contract, but the EOE had agreed to speak again to the participants in one to two weeks. Investors in Dutch bonds currently use German bond futures to hedge their portfolios because the FTO contract is so illiquid. A limited attempt to reinvigorate the contract two years ago failed.-- Amsterdam newsroom+31 20 504 5000, Fax+31 20 504 5040\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"OSCE defends record over Chechnya peace mission. BONN 1996-08-30 The head of an international mediating mission defended its record on Friday in the face of criticism by pro-Moscow leaders in breakway Chechnya and insisted it was doing its best to bring peace to the region. Flavio Cotti, the chairman of the Organisation for Security and Cooperation in Europe (OSCE), told German radio the Vienna-based body viewed the conflict in Chechnya as an internal Russian problem. \\\\\"The OSCE is completely involved. But one must not forget that the OSCE only has limited powers there,\\\\\" said Cotti, who is also the Swiss foreign minister. \\\\\"Our mission in Chechnya has done all it can within the given limitations.\\\\\" Pro-Moscow leaders in Chechnya have criticised Tim Guldimann, the Swiss diplomat who heads the OSCE Chechnya mission, saying he was biased toward Zelimkhan Yandarbiyev, president of the self-declared separatist government. Russian peacemaker Alexander Lebed and Chechen separatist military leader Aslan Maskhadov started a new round of peace talks on Friday just outside the rebel region. Cotti said Chechnya must remain part of Russia, but the solution to the conflict would be to accord the region maximum autonomy within Russia\\\\u0027s borders. \\\\\"There is no doubt that Chechnya, according to OSCE principles, belongs to a state called Russia,\\\\\" he said, pointing out that Russia was an OSCE member and it was not the organisation\\\\u0027s policy to challenge members \\\\u0027 sovereignty. He added that the OSCE was the only international body which has been allowed into the Chechnya to monitor the human rights situation there, but that its means were restricted by the fact that the conflict was a \\\\\"internal issue\\\\\". \\\\\"We have a small concept, the details of which have yet to be worked out. Chechnya must be accorded the maximum autonomy possible within the framework of Russian integrity,\\\\\" said Cotti.\"}'\ntext: OSCE defends record over Chechnya peace mission. BONN 1996-08-30 The head of an international mediating mission defended its record on Friday in the face of criticism by pro-Moscow leaders in breakway Chechnya and insisted it was doing its best to bring peace to the region. Flavio Cotti, the chairman of the Organisation for Security and Cooperation in Europe (OSCE), told German radio the Vienna-based body viewed the conflict in Chechnya as an internal Russian problem. \"The OSCE is completely involved. But one must not forget that the OSCE only has limited powers there,\" said Cotti, who is also the Swiss foreign minister. \"Our mission in Chechnya has done all it can within the given limitations.\" Pro-Moscow leaders in Chechnya have criticised Tim Guldimann, the Swiss diplomat who heads the OSCE Chechnya mission, saying he was biased toward Zelimkhan Yandarbiyev, president of the self-declared separatist government. Russian peacemaker Alexander Lebed and Chechen separatist military leader Aslan Maskhadov started a new round of peace talks on Friday just outside the rebel region. Cotti said Chechnya must remain part of Russia, but the solution to the conflict would be to accord the region maximum autonomy within Russia's borders. \"There is no doubt that Chechnya, according to OSCE principles, belongs to a state called Russia,\" he said, pointing out that Russia was an OSCE member and it was not the organisation's policy to challenge members ' sovereignty. He added that the OSCE was the only international body which has been allowed into the Chechnya to monitor the human rights situation there, but that its means were restricted by the fact that the conflict was a \"internal issue\". \"We have a small concept, the details of which have yet to be worked out. Chechnya must be accorded the maximum autonomy possible within the framework of Russian integrity,\" said Cotti.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"Dutch say no reason to reopen El Al carqo enquiry. THE HAGUE 1996-08-30 The Dutch transport minister Annemarie Jorritsma told the country \\\\u0027 second chamber that there is no further need to investigate the 1992 crash of an El Al freighter which left 43 dead in an Amsterdam suburb. She said that a request from her ministry for the aircraft\\\\u0027s waybill documentation and further information about the contents of its hold had been complied with by El Al\\\\u0027s head office in Tel Aviv. The Dutch transport ministry had come in for pressure from a cross-section of Dutch members of parliaments in May this year, some of whom believed the aircraft had been carrying unlisted, dangerous goods. Others said they thought the aircraft was loaded with too much airfreight. Jorritsma said the latest evidence from El Al in no way supported the allegations, and added there is no justification for a further investigation into the incident.-- Air Cargo Newsroom Tel+44 171 542 8982 Fax+44 171 542 5017\"}'\ntext: Dutch say no reason to reopen El Al carqo enquiry. THE HAGUE 1996-08-30 The Dutch transport minister Annemarie Jorritsma told the country ' second chamber that there is no further need to investigate the 1992 crash of an El Al freighter which left 43 dead in an Amsterdam suburb. She said that a request from her ministry for the aircraft's waybill documentation and further information about the contents of its hold had been complied with by El Al's head office in Tel Aviv. The Dutch transport ministry had come in for pressure from a cross-section of Dutch members of parliaments in May this year, some of whom believed the aircraft had been carrying unlisted, dangerous goods. Others said they thought the aircraft was loaded with too much airfreight. Jorritsma said the latest evidence from El Al in no way supported the allegations, and added there is no justification for a further investigation into the incident.-- Air Cargo Newsroom Tel+44 171 542 8982 Fax+44 171 542 5017\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"Armenians, Azeris hold peace talks in Germany. BONN 1996-08-30 Representatives from Armenia and Azerbaijan held talks earlier this week in Germany on bringing a lasting peace to the disputed Nagorno-Karabakh region, a diplomatic source close to the talks said on Friday. The source, who spoke on condition of anonymity, said Azerbaijani presidential adviser Vafa Gulizade and his Armenian counterpart Zhirayr Liparityan met to discuss the disputed enclave on Wednesday and had now flown home. An uneasy ceasefire has prevailed in Nagorno-Karabakh, which represents around 20 percent of Azeri territory, since May 1994 after ethnic Armenians drove Azeris out of the region. The conflict, which began in 1988, claimed over 10,000 lives. \\\\\"The main subject (of the talks) was the search for a peaceful solution for Nagorno-Karabakh,\\\\\" the source said. He declined to reveal any more details about the content of the talks or their exact location in Germany. Azerbaijan has said it is prepared to grant autonomy to Nagorno-Karabakh if Armenian forces pull out, but will not accept Armenia\\\\u0027s demands for the independence of the enclave. Russia\\\\u0027s Interfax news agency reported on Tuesday the officials had departed for negotiations in Germany, adding that face-to-face talks between the two sides first took place last December in Amsterdam. Interfax said the discussions were being held in parallel with peace talks mediated by the Organisation for Security and Cooperation in Europe (OSCE) and the broad-based Minsk Group of countries led by Russia and Finland.\"}'\ntext: Armenians, Azeris hold peace talks in Germany. BONN 1996-08-30 Representatives from Armenia and Azerbaijan held talks earlier this week in Germany on bringing a lasting peace to the disputed Nagorno-Karabakh region, a diplomatic source close to the talks said on Friday. The source, who spoke on condition of anonymity, said Azerbaijani presidential adviser Vafa Gulizade and his Armenian counterpart Zhirayr Liparityan met to discuss the disputed enclave on Wednesday and had now flown home. An uneasy ceasefire has prevailed in Nagorno-Karabakh, which represents around 20 percent of Azeri territory, since May 1994 after ethnic Armenians drove Azeris out of the region. The conflict, which began in 1988, claimed over 10,000 lives. \"The main subject (of the talks) was the search for a peaceful solution for Nagorno-Karabakh,\" the source said. He declined to reveal any more details about the content of the talks or their exact location in Germany. Azerbaijan has said it is prepared to grant autonomy to Nagorno-Karabakh if Armenian forces pull out, but will not accept Armenia's demands for the independence of the enclave. Russia's Interfax news agency reported on Tuesday the officials had departed for negotiations in Germany, adding that face-to-face talks between the two sides first took place last December in Amsterdam. Interfax said the discussions were being held in parallel with peace talks mediated by the Organisation for Security and Cooperation in Europe (OSCE) and the broad-based Minsk Group of countries led by Russia and Finland.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"Sombre mood on Arctic island after plane crash. Rolf Soderlind LONGYEAR, Norway 1996-08-30 The windblown, chilly streets of this tiny Arctic town are all but deserted and flags are flying at half-mast beneath a brooding, clouded sky. Longyear is a town in mourning, a close-knit community that has been shattered. Disaster struck on Thursday when a Russian airliner bringing coal miners to work crashed as it came in to land at the airport, killing all 141 people on board. \\\\\"It\\\\u0027s a sight I will never forget. I will remember it for the rest of my life,\\\\\" said Stig Onarheim. He was one of a handful of rescuers who raced to the scene of the crash in a helicopter on Thursday, hoping in vain to find survivors. The plane smashed into a snow-capped mountain on the Arctic island of Spitzbergen on Thursday, just east of Longyear. \\\\\"Imagine a big plane with a lot of luggage and people on board. Think of all that mixed together, with twisted, wrecked parts on the slope,\\\\\" Onarheim, 29, told Reuters. Police and local officials have sealed off the crash site, protecting it from intrusive reporters and from the polar bears that roam freely across the icy expanses. The dead were all Russians and Ukrainians, coming to work in the mining towns of Barentsburg and Pyramiden. Longyear is a Norwegian settlement of just over 1,000 people, but it also feels the loss keenly. \\\\\"I have trouble finding the words to express my grief. It\\\\u0027s a tragedy for everyone. We know many of the people who live in Barentsburg, some of them could have been on the plane,\\\\\" said Johan Sletten, 52. Sletten, a caretaker who has lived on the island for 30 years, said the Norwegian and Russian communities visit frequently, competing at soccer in the summer and with snow-scooter races in the winter. Teenage shop assistant Heidi Groenstein was blunter. \\\\\"I\\\\u0027m glad it was not a Norwegian plane,\\\\\" she said. \\\\\"Just think of it-- a mining village where so many workers die. They must be having a tough time of it now.\\\\\" Barentsburg, just a few hours ride by snow-scooter or 15 minutes by helicopter from Longyear, has asked to be left alone with its grief and told reporters to stay away. Around 100 Russian and Ukrainian miners were waiting in Longyear to fly home on the plane that crashed. They were given shelter in the town\\\\u0027s church overnight and ate a sombre breakfast before getting on a bus for the airport. Another plane had been sent from Moscow to pick them up. At this time of year, the only colour in Longyear comes from the brightly-painted wooden houses. Everything else is muddy, the waters of the fjord leaden. Winter is in the air. Barentsburg is an even grimmer place, a run-down testament to the hardships of the new Russia. Spitzbergen lies some 500 miles (800 km) off the northern tip of Norway and endures one of the most extreme climates on the planet. Inhabited by fewer than 3,000 people in total, it sees the sun for 24 hours a day during summer and is plunged into round-the-clock darkness in the winter months. The terrain is mountainous, the only roads are dirt tracks. Norway rules the island group under the terms of a 1920s international treaty which gave many other nations the right to establish setttlements and exploit the coal that is still mined there. Only Russia has chosen to do so.\"}'\ntext: Sombre mood on Arctic island after plane crash. Rolf Soderlind LONGYEAR, Norway 1996-08-30 The windblown, chilly streets of this tiny Arctic town are all but deserted and flags are flying at half-mast beneath a brooding, clouded sky. Longyear is a town in mourning, a close-knit community that has been shattered. Disaster struck on Thursday when a Russian airliner bringing coal miners to work crashed as it came in to land at the airport, killing all 141 people on board. \"It's a sight I will never forget. I will remember it for the rest of my life,\" said Stig Onarheim. He was one of a handful of rescuers who raced to the scene of the crash in a helicopter on Thursday, hoping in vain to find survivors. The plane smashed into a snow-capped mountain on the Arctic island of Spitzbergen on Thursday, just east of Longyear. \"Imagine a big plane with a lot of luggage and people on board. Think of all that mixed together, with twisted, wrecked parts on the slope,\" Onarheim, 29, told Reuters. Police and local officials have sealed off the crash site, protecting it from intrusive reporters and from the polar bears that roam freely across the icy expanses. The dead were all Russians and Ukrainians, coming to work in the mining towns of Barentsburg and Pyramiden. Longyear is a Norwegian settlement of just over 1,000 people, but it also feels the loss keenly. \"I have trouble finding the words to express my grief. It's a tragedy for everyone. We know many of the people who live in Barentsburg, some of them could have been on the plane,\" said Johan Sletten, 52. Sletten, a caretaker who has lived on the island for 30 years, said the Norwegian and Russian communities visit frequently, competing at soccer in the summer and with snow-scooter races in the winter. Teenage shop assistant Heidi Groenstein was blunter. \"I'm glad it was not a Norwegian plane,\" she said. \"Just think of it-- a mining village where so many workers die. They must be having a tough time of it now.\" Barentsburg, just a few hours ride by snow-scooter or 15 minutes by helicopter from Longyear, has asked to be left alone with its grief and told reporters to stay away. Around 100 Russian and Ukrainian miners were waiting in Longyear to fly home on the plane that crashed. They were given shelter in the town's church overnight and ate a sombre breakfast before getting on a bus for the airport. Another plane had been sent from Moscow to pick them up. At this time of year, the only colour in Longyear comes from the brightly-painted wooden houses. Everything else is muddy, the waters of the fjord leaden. Winter is in the air. Barentsburg is an even grimmer place, a run-down testament to the hardships of the new Russia. Spitzbergen lies some 500 miles (800 km) off the northern tip of Norway and endures one of the most extreme climates on the planet. Inhabited by fewer than 3,000 people in total, it sees the sun for 24 hours a day during summer and is plunged into round-the-clock darkness in the winter months. The terrain is mountainous, the only roads are dirt tracks. Norway rules the island group under the terms of a 1920s international treaty which gave many other nations the right to establish setttlements and exploit the coal that is still mined there. Only Russia has chosen to do so.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"Ericsson says wins 1.2 bln SKR China order. STOCKHOLM 1996-08-30 Swedish telecoms group LM Ericsson AB said on Friday it won an order worth 1.2 billion crowns for a fixed public telecoms network in the Guangdong province of China. Ericsson said in a statement the order was from the Guangdong Post and Telecommunications Administration (GPTA). The order included AXE switching equipment, ISDN equipment, Intelligent Network (IN) products, broad-band multi-media communication network products, services and training, Ericsson spokesman Per Zetterquist told Reuters. Deliveries are due to be completed by 1999, the company said.-- Stockholm newsroom+46-8-700 1017\"}'\ntext: Ericsson says wins 1.2 bln SKR China order. STOCKHOLM 1996-08-30 Swedish telecoms group LM Ericsson AB said on Friday it won an order worth 1.2 billion crowns for a fixed public telecoms network in the Guangdong province of China. Ericsson said in a statement the order was from the Guangdong Post and Telecommunications Administration (GPTA). The order included AXE switching equipment, ISDN equipment, Intelligent Network (IN) products, broad-band multi-media communication network products, services and training, Ericsson spokesman Per Zetterquist told Reuters. Deliveries are due to be completed by 1999, the company said.-- Stockholm newsroom+46-8-700 1017\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"HK has infrastructure in place for post-97 - Tsang. AUCKLAND 1996-08-30 Hong Kong Financial Secretary Donald Tsang said on Friday that the territory had the \\\\\"infrastructural hardware\\\\\" to make a success of its future under Chinese sovereignty from mid-1997. \\\\\"We have the largest and most efficient port on the South China coast ; we have the best transport and telecommunications infrastructure in the world ; and we are investing in this hardware on an enormous scale,\\\\\" Tsang said in a speech to Auckland during a visit to New Zealand. Hong Kong also had the necessary \\\\\"constitutional infrastructure\\\\\" in place, with the promise of autonomy in running its affairs after the handover from Britain to China. \\\\\"What this means in practice is that Hong Kong will go on raising its own taxes, issuing its own currency, setting its own expenditure priorities and managing its own enormous financial reserves,\\\\\" Tsang said. He acknowledged that many Hong Kong people had decided to seek their future elsewhere and others were sure to follow in the next nine months. \\\\\"But for the great majority of us, Hong Kong is our home and Hong Kong\\\\u0027s future is our future.\\\\\"-- Wellington newsroom 64 4 473-4746\"}'\ntext: HK has infrastructure in place for post-97 - Tsang. AUCKLAND 1996-08-30 Hong Kong Financial Secretary Donald Tsang said on Friday that the territory had the \"infrastructural hardware\" to make a success of its future under Chinese sovereignty from mid-1997. \"We have the largest and most efficient port on the South China coast ; we have the best transport and telecommunications infrastructure in the world ; and we are investing in this hardware on an enormous scale,\" Tsang said in a speech to Auckland during a visit to New Zealand. Hong Kong also had the necessary \"constitutional infrastructure\" in place, with the promise of autonomy in running its affairs after the handover from Britain to China. \"What this means in practice is that Hong Kong will go on raising its own taxes, issuing its own currency, setting its own expenditure priorities and managing its own enormous financial reserves,\" Tsang said. He acknowledged that many Hong Kong people had decided to seek their future elsewhere and others were sure to follow in the next nine months. \"But for the great majority of us, Hong Kong is our home and Hong Kong's future is our future.\"-- Wellington newsroom 64 4 473-4746\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"PRESS DIGEST - Indonesian newspapers - August 30. Following is a summary of major Indonesian political and business stories in leading newspapers, prepared by Reuters in Jakarta. Reuters has not checked the stories and does not guarantee their accuracy. Telephone : (6221) 384-6364. Fax : (6221) 344-8404. - - - - KOMPAS Indonesian President Suharto has asked businessmen to share their experiences with each other in an effort to boost the country\\\\u0027s exports. - - - - JAKARTA POST Speaker of the House of Representatives Wahono has called on those serving in high state institutions to direct their efforts in the coming years towards dismantling all barriers to social justice. An agreement to bring peace to the southern Philippines is set to be initialed on Friday after delegates from the Philippine government and the Moro National Liberation Front (MNLF) concluded negotiations on the treaty which is set to end almost 25 years of conflict in the region. - - - - MEDIA INDONESIA Around 2,000 of Indonesia\\\\u0027s controversial Timor national car made by Kia Motor Corp of South Korea arrived at Jakarta\\\\u0027s Tanjung Priok port on Thursday. The cars will be jointly marketed by Kia and PT Timor Putra Nasional, controlled by a son of President Suharto, which plans next year to start assembling the vehicles in Indonesia. - - - - REPUBLIKA The Central Jakarta District Court has started to hear the suit filed by ousted Indonesian Democratic Party (PDI) leader Megawati Sukarnoputri against the government and party rivals after the parties failed to reach an out-of-court settlement. Megawati has sued the defendants over a government-backed rebel congress which ousted her last June.\"}'\ntext: PRESS DIGEST - Indonesian newspapers - August 30. Following is a summary of major Indonesian political and business stories in leading newspapers, prepared by Reuters in Jakarta. Reuters has not checked the stories and does not guarantee their accuracy. Telephone : (6221) 384-6364. Fax : (6221) 344-8404. - - - - KOMPAS Indonesian President Suharto has asked businessmen to share their experiences with each other in an effort to boost the country's exports. - - - - JAKARTA POST Speaker of the House of Representatives Wahono has called on those serving in high state institutions to direct their efforts in the coming years towards dismantling all barriers to social justice. An agreement to bring peace to the southern Philippines is set to be initialed on Friday after delegates from the Philippine government and the Moro National Liberation Front (MNLF) concluded negotiations on the treaty which is set to end almost 25 years of conflict in the region. - - - - MEDIA INDONESIA Around 2,000 of Indonesia's controversial Timor national car made by Kia Motor Corp of South Korea arrived at Jakarta's Tanjung Priok port on Thursday. The cars will be jointly marketed by Kia and PT Timor Putra Nasional, controlled by a son of President Suharto, which plans next year to start assembling the vehicles in Indonesia. - - - - REPUBLIKA The Central Jakarta District Court has started to hear the suit filed by ousted Indonesian Democratic Party (PDI) leader Megawati Sukarnoputri against the government and party rivals after the parties failed to reach an out-of-court settlement. Megawati has sued the defendants over a government-backed rebel congress which ousted her last June.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"Jeans Mate Corp - 6mth parent forecast. TOKYO 1996-08-30 Six months to August 20, 1996 (in billions of yen unless specified) LATEST PREVIOUS ACTUAL (Parent) FORECAST FORECAST YEAR-AGO Sales 9.06 9.31 8.42 Current 818 million 979 million 882 million Net 415 million 490 million 412 million NOTE - Jeans Mate Corp is the full company name.\"}'\ntext: Jeans Mate Corp - 6mth parent forecast. TOKYO 1996-08-30 Six months to August 20, 1996 (in billions of yen unless specified) LATEST PREVIOUS ACTUAL (Parent) FORECAST FORECAST YEAR-AGO Sales 9.06 9.31 8.42 Current 818 million 979 million 882 million Net 415 million 490 million 412 million NOTE - Jeans Mate Corp is the full company name.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"Apic Yamada - 6mth parent forecast. TOKYO 1996-08-30 Six months to September 30, 1996 (in billions of yen unless specified) LATEST PREVIOUS ACTUAL (Parent) FORECAST FORECAST YEAR-AGO Sales 12.50 13.00 11.27 Current 1.30 1.35 1.09 Net 650 million 680 million 600 million NOTE - Apic Yamada Corp is a leading manufacturer of semiconductor leadframes.\"}'\ntext: Apic Yamada - 6mth parent forecast. TOKYO 1996-08-30 Six months to September 30, 1996 (in billions of yen unless specified) LATEST PREVIOUS ACTUAL (Parent) FORECAST FORECAST YEAR-AGO Sales 12.50 13.00 11.27 Current 1.30 1.35 1.09 Net 650 million 680 million 600 million NOTE - Apic Yamada Corp is a leading manufacturer of semiconductor leadframes.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"Bootleg brew kills 35 in China, police nab suspects. BEIJING 1996-08-30 Police in southwest China have arrested 30 people suspected of making and selling homemade alcohol that killed 35 people and poisoned 157, the Xinhua news agency said on Friday. A group of farmers in Huize county in the southwestern province of Yunnan were arrested for blending alcohol with methanol and selling the toxic liquor to local residents, the agency said. Between late June and July, a total of 192 people were poisoned by the toxic liquor, and 35 of them died and six were left severely handicapped, it said. Local authorities launched an investigation after they received reports of several similar deaths in the area, it said. Post-mortem examinations showed they were all caused by methanol poisoning. Police had confiscated the remainder of the poisonous liquor, Xinhua said. It gave no further details.\"}'\ntext: Bootleg brew kills 35 in China, police nab suspects. BEIJING 1996-08-30 Police in southwest China have arrested 30 people suspected of making and selling homemade alcohol that killed 35 people and poisoned 157, the Xinhua news agency said on Friday. A group of farmers in Huize county in the southwestern province of Yunnan were arrested for blending alcohol with methanol and selling the toxic liquor to local residents, the agency said. Between late June and July, a total of 192 people were poisoned by the toxic liquor, and 35 of them died and six were left severely handicapped, it said. Local authorities launched an investigation after they received reports of several similar deaths in the area, it said. Post-mortem examinations showed they were all caused by methanol poisoning. Police had confiscated the remainder of the poisonous liquor, Xinhua said. It gave no further details.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"Singapore hangs Thai drug trafficker. SINGAPORE 1996-08-30 Singapore hanged a Thai farmer at Changi Prison on Friday for drug trafficking, the Central Narcotics Bureau (CNB) said. Jeerasak Densakul, 24, was arrested in 1995 when he was found with 11 slabs of cannabis weighing 2.2 kg (4.8 pounds), the CNB said. Singapore has a mandatory death sentence for anyone over 18 years of age found guilty of trafficking in more than 15 grams (half an ounce) of heroin, 30 grams (an ounce) of morphine or 500 grams (18 oz) of cannabis or marijuana. Of the nearly 270 people hanged for various crimes in Singapore since 1975, almost half have been for drug-related charges.\"}'\ntext: Singapore hangs Thai drug trafficker. SINGAPORE 1996-08-30 Singapore hanged a Thai farmer at Changi Prison on Friday for drug trafficking, the Central Narcotics Bureau (CNB) said. Jeerasak Densakul, 24, was arrested in 1995 when he was found with 11 slabs of cannabis weighing 2.2 kg (4.8 pounds), the CNB said. Singapore has a mandatory death sentence for anyone over 18 years of age found guilty of trafficking in more than 15 grams (half an ounce) of heroin, 30 grams (an ounce) of morphine or 500 grams (18 oz) of cannabis or marijuana. Of the nearly 270 people hanged for various crimes in Singapore since 1975, almost half have been for drug-related charges.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"Arafat goes to Nablus ahead of cabinet meeting. NABLUS, West Bank 1996-08-30 Palestinian President Yasser Arafat arrived in the West Bank self-rule enclave of Nablus from Ramallah on Friday, witnesses said. His aides said Arafat would hold the weekly meeting of the Palestinian self-rule Authority\\\\u0027s cabinet in Nablus on Saturday. In Jerusalem, Israeli security forces were bracing for thousands of Palestinians expected to answer Arafat\\\\u0027s call earlier this week to come to the city holy to Moslems, Arabs and Jews to pray in protest against Israel\\\\u0027s settlement policy in the West Bank and delay in peace negotiations. Palestinians want Arab East Jerusalem as the capital of a future independent state. Israel, which captured and annexed East Jerusalem in 1967, says it will never cede any part of the city. Arafat, who made an interim peace deal with Israel in 1993 that set up self-rule, says he will only visit Jerusalem once Israeli occupation has ended.\"}'\ntext: Arafat goes to Nablus ahead of cabinet meeting. NABLUS, West Bank 1996-08-30 Palestinian President Yasser Arafat arrived in the West Bank self-rule enclave of Nablus from Ramallah on Friday, witnesses said. His aides said Arafat would hold the weekly meeting of the Palestinian self-rule Authority's cabinet in Nablus on Saturday. In Jerusalem, Israeli security forces were bracing for thousands of Palestinians expected to answer Arafat's call earlier this week to come to the city holy to Moslems, Arabs and Jews to pray in protest against Israel's settlement policy in the West Bank and delay in peace negotiations. Palestinians want Arab East Jerusalem as the capital of a future independent state. Israel, which captured and annexed East Jerusalem in 1967, says it will never cede any part of the city. Arafat, who made an interim peace deal with Israel in 1993 that set up self-rule, says he will only visit Jerusalem once Israeli occupation has ended.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"U.N. Council concerned about Israeli bulldozers. UNITED NATIONS 1996-08-29 Security Council members expressed concern on Thursday that Israel\\\\u0027s bulldozing of a Palestinian day-care centre for the disabled might further injure the Middle East peace process. Responding to a letter from the the Palestinian U.N. observer mission, Security Council President Tono Eitel of Germany said that members asked him to convey their views to Israel\\\\u0027s charge d\\\\u0027affaires, David Peleg. \\\\\"The members expressed their concern about the maintenance of the peace process and they urged that no action be taken that would have a negative impact on the negotiations,\\\\\" Eitel said after an informal council session. \\\\\"They asked me to call in the Israeli charge d\\\\u0027affaires and discuss the matter with him,\\\\\" he added. The Palestinian letter from Marwan Jilani said the destruction of the Jerusalem centre was an effort by Israel to \\\\\"alter the character, demographic composition and status of the Holy City of Jerusalem\\\\\" and violated agreements between Israel and the Palestinian Liberation Organisation. \\\\\"This most recent measure represents a revival of old, malicious plans to confiscate the land and build units for Israeli settlers within the walls of the Old City.\\\\\" \\\\\"We expect the international community to take a clear and firm position, based on international law and in accordance with U.N. resolutions, against all such Israeli violations and illegal practices,\\\\\" he said. On Tuesday, Israeli crews hoisted a bulldozer over the walls of Jerusalem\\\\u0027s old city and demolished the centre, saying it was being restored without a building permit. Canada had recently donated $ 30 million to the centre, called the Burj al-Laqlaq Society.\"}'\ntext: U.N. Council concerned about Israeli bulldozers. UNITED NATIONS 1996-08-29 Security Council members expressed concern on Thursday that Israel's bulldozing of a Palestinian day-care centre for the disabled might further injure the Middle East peace process. Responding to a letter from the the Palestinian U.N. observer mission, Security Council President Tono Eitel of Germany said that members asked him to convey their views to Israel's charge d'affaires, David Peleg. \"The members expressed their concern about the maintenance of the peace process and they urged that no action be taken that would have a negative impact on the negotiations,\" Eitel said after an informal council session. \"They asked me to call in the Israeli charge d'affaires and discuss the matter with him,\" he added. The Palestinian letter from Marwan Jilani said the destruction of the Jerusalem centre was an effort by Israel to \"alter the character, demographic composition and status of the Holy City of Jerusalem\" and violated agreements between Israel and the Palestinian Liberation Organisation. \"This most recent measure represents a revival of old, malicious plans to confiscate the land and build units for Israeli settlers within the walls of the Old City.\" \"We expect the international community to take a clear and firm position, based on international law and in accordance with U.N. resolutions, against all such Israeli violations and illegal practices,\" he said. On Tuesday, Israeli crews hoisted a bulldozer over the walls of Jerusalem's old city and demolished the centre, saying it was being restored without a building permit. Canada had recently donated $ 30 million to the centre, called the Burj al-Laqlaq Society.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - REAL SCRAPE 1-1 DRAW IN SCRAPPY OPENING MATCH. LA CORUNA, Spain 1996-08-31 A late goal by newly-signed defender Roberto Carlos saved the blushes of Real Madrid coach Fabio Capello and his multi-billion peseta line-up in the opening game of the Spanish championship on Saturday. The Brazilian\\\\u0027s 79th-minute effort was enough to earn Real a point from a scrappy 1-1 draw at fellow title contenders Deportivo Coruna. Deportivo started strongly, taking the lead midway through the first half when former Auxerre playmaker Corentine Martins headed home a corner after a flick-on by Brazilian-born Spanish international midfielder Donato. Real looked to be in deep trouble shortly after the break when Luis Milla was sent off for committing two bookable offences in as many minutes. But Deportivo were unable to capitalise on their numerical advantage, and were themselves reduced to ten men when Armando Alvarez was sent off 15 minutes from time. Shortly afterwards Roberto Carlos found space in the home defence and equalised for Real with a shot that was deflected past despairing Deportivo \\\\u0027 keeper Jacques Songo\\\\u0027o. In a frantic final five minutes there were chances at both ends, and Donato, who had earlier been booked, was sent off for protesting about the incursion of Real players at a free kick. Before the match Deportivo chairman Augusto Lendoiro said he would ignore a FIFA decision banning Brazilian midfielder Mauro Silva from playing in the match for failing to join his national side\\\\u0027s tour of Europe. In the event, coach John Toshack decided not to use Silva, who had claimed he did not join the Brazil squad because he had lost his passport.\"}'\ntext: SOCCER - REAL SCRAPE 1-1 DRAW IN SCRAPPY OPENING MATCH. LA CORUNA, Spain 1996-08-31 A late goal by newly-signed defender Roberto Carlos saved the blushes of Real Madrid coach Fabio Capello and his multi-billion peseta line-up in the opening game of the Spanish championship on Saturday. The Brazilian's 79th-minute effort was enough to earn Real a point from a scrappy 1-1 draw at fellow title contenders Deportivo Coruna. Deportivo started strongly, taking the lead midway through the first half when former Auxerre playmaker Corentine Martins headed home a corner after a flick-on by Brazilian-born Spanish international midfielder Donato. Real looked to be in deep trouble shortly after the break when Luis Milla was sent off for committing two bookable offences in as many minutes. But Deportivo were unable to capitalise on their numerical advantage, and were themselves reduced to ten men when Armando Alvarez was sent off 15 minutes from time. Shortly afterwards Roberto Carlos found space in the home defence and equalised for Real with a shot that was deflected past despairing Deportivo ' keeper Jacques Songo'o. In a frantic final five minutes there were chances at both ends, and Donato, who had earlier been booked, was sent off for protesting about the incursion of Real players at a free kick. Before the match Deportivo chairman Augusto Lendoiro said he would ignore a FIFA decision banning Brazilian midfielder Mauro Silva from playing in the match for failing to join his national side's tour of Europe. In the event, coach John Toshack decided not to use Silva, who had claimed he did not join the Brazil squad because he had lost his passport.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"RUGBY LEAGUE - WIGAN BEAT BRADFORD 42-36 IN SEMIFINAL. WIGAN, England 1996-08-31 Result of English rugby league premiership semifinal played on Saturday : Wigan 42 Bradford Bulls 36\"}'\ntext: RUGBY LEAGUE - WIGAN BEAT BRADFORD 42-36 IN SEMIFINAL. WIGAN, England 1996-08-31 Result of English rugby league premiership semifinal played on Saturday : Wigan 42 Bradford Bulls 36\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - ISRAEL BEAT BULGARIA IN EUROPEAN UNDER-21 QUALIFIER. HERZLIYA, Israel 1996-08-31 Result of European under-21 championship group 5 qualifier on Saturday : Israel 2, Bulgaria 0 (halftime 0-0) Scorers : Haim Hajaj (47th), Nir Sivilia (57th). Attendance : 2,000.\"}'\ntext: SOCCER - ISRAEL BEAT BULGARIA IN EUROPEAN UNDER-21 QUALIFIER. HERZLIYA, Israel 1996-08-31 Result of European under-21 championship group 5 qualifier on Saturday : Israel 2, Bulgaria 0 (halftime 0-0) Scorers : Haim Hajaj (47th), Nir Sivilia (57th). Attendance : 2,000.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - IRISH ERASE PAINFUL MEMORIES WITH 5-0 WIN. ESCHEN, Liechtenstein 1996-08-31 The Republic of Ireland\\\\u0027s new-look side dispelled painful memories of their last visit to Liechtenstein by beating the Alpine part-timers 5-0 in a World Cup qualifier on Saturday. The Irish, under new manager Mick McCarthy, took a 4-0 lead within 20 minutes through captain Andy Townsend, 20-year-old Norwich striker Keith O\\\\u0027Neill, Sunderland forward Niall Quinn and teenager Ian Harte. Quinn added his second and Ireland\\\\u0027s fifth just after the hour to complete the rout and give the Irish their biggest-ever away win. The result helped erase memories of Ireland\\\\u0027s visit to the Eschen stadium 14 months ago, when Jack Charlton\\\\u0027s side were held to a frustrating 0-0 draw which ultimately cost them a place in the European championship finals.\"}'\ntext: SOCCER - IRISH ERASE PAINFUL MEMORIES WITH 5-0 WIN. ESCHEN, Liechtenstein 1996-08-31 The Republic of Ireland's new-look side dispelled painful memories of their last visit to Liechtenstein by beating the Alpine part-timers 5-0 in a World Cup qualifier on Saturday. The Irish, under new manager Mick McCarthy, took a 4-0 lead within 20 minutes through captain Andy Townsend, 20-year-old Norwich striker Keith O'Neill, Sunderland forward Niall Quinn and teenager Ian Harte. Quinn added his second and Ireland's fifth just after the hour to complete the rout and give the Irish their biggest-ever away win. The result helped erase memories of Ireland's visit to the Eschen stadium 14 months ago, when Jack Charlton's side were held to a frustrating 0-0 draw which ultimately cost them a place in the European championship finals.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - IRELAND BEAT LIECHTENSTEIN 5-0 IN WORLD CUP QUALIFIER. ESCHEN 1996-08-31 The Republic of Ireland beat Liechtenstein 5-0 (halftime 4-0) in a World Cup soccer European group 8 qualifier on Saturday. Scorers : Andy Townsend (5th), Keith O\\\\u0027Neill (7th), Niall Quinn (11th, 61st), Ian Harte (19th). Attendance : 3,900\"}'\ntext: SOCCER - IRELAND BEAT LIECHTENSTEIN 5-0 IN WORLD CUP QUALIFIER. ESCHEN 1996-08-31 The Republic of Ireland beat Liechtenstein 5-0 (halftime 4-0) in a World Cup soccer European group 8 qualifier on Saturday. Scorers : Andy Townsend (5th), Keith O'Neill (7th), Niall Quinn (11th, 61st), Ian Harte (19th). Attendance : 3,900\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"GOLF - BRITISH MASTERS FINAL SCORES. NORTHAMPTON, England 1996-08-31 Leading scores after the final round of the British Masters golf tournament on Saturday (British unless stated) : 284 Robert Allenby (Australia) 69 71 71 73, Miguel Angel Martin (Spain) 75 70 71 68 (Allenby won at first play-off hole) 285 Costantino Rocca (Italy) 71 73 72 69 286 Miguel Angel Jimenez (Spain) 74 72 73 67 287 Ian Woosnam 70 76 71 70 288 Jose Coceres (Argentina) 69 78 71 70 289 Joakim Haeggman (Sweden) 71 77 70 71, Antoine Lebouc (France) 74 73 70 72 290 Colin Montgomerie 68 76 77 69, Robert Coles 74 76 71 69, Philip Walton (Ireland) 71 74 74 71, Peter Mitchell 74 71 74 71, Klas Eriksson (Sweden) 71 75 72 72, Pedro Linhart (Spain) 72 73 67 78 291 Phillip Price 72 76 74 69, Adam Hunter 70 79 73 69, Peter O\\\\u0027Malley (Australia) 71 73 75 72, Mark Roe 69 71 78 73, Mike Clayton (Australia) 69 76 73 73 292 Iain Pyman 71 75 75 71, David Gilford 69 74 77 72, Peter Hedblom (Sweden) 70 75 75 72, Stephen McAllister 73 76 69 74.\"}'\ntext: GOLF - BRITISH MASTERS FINAL SCORES. NORTHAMPTON, England 1996-08-31 Leading scores after the final round of the British Masters golf tournament on Saturday (British unless stated) : 284 Robert Allenby (Australia) 69 71 71 73, Miguel Angel Martin (Spain) 75 70 71 68 (Allenby won at first play-off hole) 285 Costantino Rocca (Italy) 71 73 72 69 286 Miguel Angel Jimenez (Spain) 74 72 73 67 287 Ian Woosnam 70 76 71 70 288 Jose Coceres (Argentina) 69 78 71 70 289 Joakim Haeggman (Sweden) 71 77 70 71, Antoine Lebouc (France) 74 73 70 72 290 Colin Montgomerie 68 76 77 69, Robert Coles 74 76 71 69, Philip Walton (Ireland) 71 74 74 71, Peter Mitchell 74 71 74 71, Klas Eriksson (Sweden) 71 75 72 72, Pedro Linhart (Spain) 72 73 67 78 291 Phillip Price 72 76 74 69, Adam Hunter 70 79 73 69, Peter O'Malley (Australia) 71 73 75 72, Mark Roe 69 71 78 73, Mike Clayton (Australia) 69 76 73 73 292 Iain Pyman 71 75 75 71, David Gilford 69 74 77 72, Peter Hedblom (Sweden) 70 75 75 72, Stephen McAllister 73 76 69 74.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - SLOVAKIA BEAT FAROES IN WORLD CUP QUALIFIER. TOFTIR, Faroe Islands 1996-08-31 Slovakia beat the Faroe Islands 2-1 (halftime 1-0) in their World Cup soccer European group six qualifying match on Saturday. Scorers : Faroe Islands - Jan Allan Mueller (60th minute) Slovakia - Lubomir Moravcik (13th), Peter Dubovsky (88th) Attendance : 1,445.\"}'\ntext: SOCCER - SLOVAKIA BEAT FAROES IN WORLD CUP QUALIFIER. TOFTIR, Faroe Islands 1996-08-31 Slovakia beat the Faroe Islands 2-1 (halftime 1-0) in their World Cup soccer European group six qualifying match on Saturday. Scorers : Faroe Islands - Jan Allan Mueller (60th minute) Slovakia - Lubomir Moravcik (13th), Peter Dubovsky (88th) Attendance : 1,445.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"CRICKET - ENGLAND BEAT PAKISTAN BY 107 RUNS IN SECOND ONE-DAYER. BIRMINGHAM, England 1996-08-31 England beat Pakistan by 107 runs in the second one-day international at Edgbaston on Saturday to take the series 2-0. Scores : England 292-8 innings closed (N. Knight 113), Pakistan 185 (Ijaz Ahmed 79 ; A. Hollioake 4-23)\"}'\ntext: CRICKET - ENGLAND BEAT PAKISTAN BY 107 RUNS IN SECOND ONE-DAYER. BIRMINGHAM, England 1996-08-31 England beat Pakistan by 107 runs in the second one-day international at Edgbaston on Saturday to take the series 2-0. Scores : England 292-8 innings closed (N. Knight 113), Pakistan 185 (Ijaz Ahmed 79 ; A. Hollioake 4-23)\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"CYCLING - TOUR OF NETHERLANDS FINAL RESULTS / STANDINGS. LANDGRAAF, Netherlands 1996-08-31 Leading results of the 205-km sixth and final stage of the Tour of the Netherlands between Roermond and Landgraaf on Saturday : 1. Olaf Ludwig (Germany) Telekom 4 hours 48 mins 2 seconds 2. Giovanni Lombardi (Italy) Polti 5 seconds behind 3. Tristan Hoffman (Netherlands) TVM same time 4. Erik Breukink (Netherlands) Rabobank 8 seconds 5. Jesper Skibby (Denmark) TVM 9 6. Vyacheslav Ekimov (Russia) Rabobank same time 7. Luca Pavanello (Italy) Aki 11 8. Eleuterio Anguita (Spain) MX Onda 9. Michael Andersson (Sweden) Telekom 10. Johan Capiot (Belgium) Collstrop all same time Final overall placings (after six stages) : 1. Rolf Sorensen (Denmark) Rabobank 20:36:54 2. Lance Armstrong (U.S.) Motorola 2 seconds behind 3. Ekimov 1:7 4. Marco Lietti (Italy) MG-Technogym 1:16 5. Erik Dekker (Netherlands) Rabobank 1:23 6. Ludwig 1:25 6. Breukink same time 8. Maarten den Bakker (Netherlands) TVM 1:33 9. Andersson 1:34 10. Skibby 1:45\"}'\ntext: CYCLING - TOUR OF NETHERLANDS FINAL RESULTS / STANDINGS. LANDGRAAF, Netherlands 1996-08-31 Leading results of the 205-km sixth and final stage of the Tour of the Netherlands between Roermond and Landgraaf on Saturday : 1. Olaf Ludwig (Germany) Telekom 4 hours 48 mins 2 seconds 2. Giovanni Lombardi (Italy) Polti 5 seconds behind 3. Tristan Hoffman (Netherlands) TVM same time 4. Erik Breukink (Netherlands) Rabobank 8 seconds 5. Jesper Skibby (Denmark) TVM 9 6. Vyacheslav Ekimov (Russia) Rabobank same time 7. Luca Pavanello (Italy) Aki 11 8. Eleuterio Anguita (Spain) MX Onda 9. Michael Andersson (Sweden) Telekom 10. Johan Capiot (Belgium) Collstrop all same time Final overall placings (after six stages) : 1. Rolf Sorensen (Denmark) Rabobank 20:36:54 2. Lance Armstrong (U.S.) Motorola 2 seconds behind 3. Ekimov 1:7 4. Marco Lietti (Italy) MG-Technogym 1:16 5. Erik Dekker (Netherlands) Rabobank 1:23 6. Ludwig 1:25 6. Breukink same time 8. Maarten den Bakker (Netherlands) TVM 1:33 9. Andersson 1:34 10. Skibby 1:45\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"CRICKET - ENGLAND V PAKISTAN ONE-DAY SCOREBOARD. BIRMINGHAM, England 1996-08-31 Scoreboard of the second one-day cricket match between England and Pakistan on Saturday : England N. Knight st Moin Khan b Saqlain Mushtaq 113 A. Stewart b Mushtaq Ahmed 46 M. Atherton lbw b Mushtaq Ahmed 1 G. Thorpe lbw b Ata-ur-Rehman 21 M. Maynard run out 1 R. Irani not out 45 A. Hollioake run out 15 D. Gough run out 0 R. Croft b Waqar Younis 15 D. Headley not out 3 Extras (lb-25 w-4 nb-3) 32 Total (for 8 wickets, innings closed) 292 Fall : 1-103 2-105 3-163 4-168 5-221 6-257 7-257 8-286. Did Not Bat : A. Mullally. Bowling : Wasim Akram 10-0-50-0, Waqar Younis 9-0-54-1, Ata-ur-Rehman 6-0-40-1, Saqlain Mushtaq 10-0-59-1, Mushtaq Ahmed 10-0-33-2, Aamir Sohail 5-0-31-0. pakistan Saeed Anwar c Stewart b Gough 33 Aamir Sohail c Croft b Gough 0 Moin Khan lbw b Mullally 0 Ijaz Ahmed b Croft 79 Inzamam-ul-Haq c Thorpe b Croft 6 Salim Malik c Stewart b Hollioake 23 Wasim Akram c Knight b Hollioake 21 Mushtaq Ahmed not out 14 Saqlain Mushtaq b Hollioake 0 Waqar Younis lbw b Gough 4 Ata-ur-Rehman c Knight b Hollioake 2 Extras (lb-2 nb-1) 3 Total (37.5 overs) 185 Fall of wickets : 1-1 2-6 3-54 4-104 5-137 6-164 7-164 8-168 9-177. Bowling : Gough 8-0-39-3, Mullally 6-0-30-1, Headley 7-0-32-0, Irani 2-0-22-0, Croft 8-0-37-2, Hollioake 6.5-1-23-4.\"}'\ntext: CRICKET - ENGLAND V PAKISTAN ONE-DAY SCOREBOARD. BIRMINGHAM, England 1996-08-31 Scoreboard of the second one-day cricket match between England and Pakistan on Saturday : England N. Knight st Moin Khan b Saqlain Mushtaq 113 A. Stewart b Mushtaq Ahmed 46 M. Atherton lbw b Mushtaq Ahmed 1 G. Thorpe lbw b Ata-ur-Rehman 21 M. Maynard run out 1 R. Irani not out 45 A. Hollioake run out 15 D. Gough run out 0 R. Croft b Waqar Younis 15 D. Headley not out 3 Extras (lb-25 w-4 nb-3) 32 Total (for 8 wickets, innings closed) 292 Fall : 1-103 2-105 3-163 4-168 5-221 6-257 7-257 8-286. Did Not Bat : A. Mullally. Bowling : Wasim Akram 10-0-50-0, Waqar Younis 9-0-54-1, Ata-ur-Rehman 6-0-40-1, Saqlain Mushtaq 10-0-59-1, Mushtaq Ahmed 10-0-33-2, Aamir Sohail 5-0-31-0. pakistan Saeed Anwar c Stewart b Gough 33 Aamir Sohail c Croft b Gough 0 Moin Khan lbw b Mullally 0 Ijaz Ahmed b Croft 79 Inzamam-ul-Haq c Thorpe b Croft 6 Salim Malik c Stewart b Hollioake 23 Wasim Akram c Knight b Hollioake 21 Mushtaq Ahmed not out 14 Saqlain Mushtaq b Hollioake 0 Waqar Younis lbw b Gough 4 Ata-ur-Rehman c Knight b Hollioake 2 Extras (lb-2 nb-1) 3 Total (37.5 overs) 185 Fall of wickets : 1-1 2-6 3-54 4-104 5-137 6-164 7-164 8-168 9-177. Bowling : Gough 8-0-39-3, Mullally 6-0-30-1, Headley 7-0-32-0, Irani 2-0-22-0, Croft 8-0-37-2, Hollioake 6.5-1-23-4.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"CYCLING - WORLD TRACK CHAMPIONSHIP RESULTS. MANCHESTER, England 1996-08-31 Results at the world track cycling championships on Saturday : Women\\\\u0027s 3,000 metres individual pursuit qualifying round (fastest eight to quarter finals) : 1. Antonella Bellutti (Italy) 3:31.526 (world record) 2. Marion Clignet (France) 3:31.674 3. Lucy Tyler-Sharman (Australia) 3:31.830 4. Yvonne McGregor (Britain) 3:41.823 5. Natalia Karimova (Russia) 3:45.061 6. Svetlana Samokhalova (Russia) 3:46.216 7 Jane Quigley (U.S.) 3:46.493 8. Rasa Mazeikyte (Lithuania) 3:46.834 9. Tatian Stiajkina (Ukraine) 3:52.204 World 4,000 metres team pursuit semifinals : Italy (Adler Capelli, Cristiano Citto, Andrea Collinelli, Mauro Trentino) 4:00.958 (world record) beat Russia (Anton Chantyr, Edouard Gritsoun, Nikolai Kouznetsov) 4:06.534. France (Cyril Bos, Philippe Ermenault, Jean-Michel Monin, Francis Moreau) 4:05.104 beat Germany (Guido Fulst, Danilo Hondo, Thorsten Rund, Heiko Szonn) 4:05.463 Germany take the bronze medal as fastest losing semifinalist. Women\\\\u0027s world 500 metres time trial final : 1. Felicia Ballanger (France) 34.829 2. Annett Neumann (Germany) 35.202 3. Michelle Ferris (Australia) 35.694 4. Magali Faure (France) 35.888 5. Olga Slioussareva (Russia) 36.170 6. Oksana Grichina (Russia) 36.242 7. Tanya Dubnicoff (Canada) 36.307 8. Kathrin Freitag (Germany) 36.491 9. Donna Wynd (New Zealand) 36.831 10. Mira Kasslin (Finland) 37.273 11. Wendy Everson (Britain) 37.624 12. Giovanna Troldi (Italy) 38.285 13. Rita Razmaite (Lithuania) 38.546 World 4,000 metres team pursuit championship final : Italy (Adler Capelli, Cristiano Citto, Andrea Collinelli, Mauro Trentino) 4:02.752 beat France (Cyril Bos, Philippe Ermenault, Jean-Michel Monin, Francis Moreau) 4:04.539 World sprint championship quarter finals (best of three matches) Florian Rousseau (France) beat Ainars Kiksis (Latvia) 2-0 Darryn Hill (Australia) beat Christian Arrue (U.S.) 2-0 Roberto Chiappa (Italy) beat Frederic Magne (France) 2-0 Marty Nothstein (U.S.) beat Pavel Buran (Czech Republic) 2-0 Women\\\\u0027s world 3,000 metres individual pursuit championship quarter-finals : Marion Clignet (France) 3:30.974 (World Record) beat Jane Quigley (USA) 3:42.852 Natalia Karimova (Russia) 3:40.036 beat Yvonne McGregor (Britain) 3:43.078 Lucy Tyler-Sharman (Australia) 3:35.087 beat Svetlana Samokhvalova (Russia) 3:45.011 Antonella Bellutti (Italy) 3:32.174 caught and eliminated Rasa Mazeikyte (Lithuania)\"}'\ntext: CYCLING - WORLD TRACK CHAMPIONSHIP RESULTS. MANCHESTER, England 1996-08-31 Results at the world track cycling championships on Saturday : Women's 3,000 metres individual pursuit qualifying round (fastest eight to quarter finals) : 1. Antonella Bellutti (Italy) 3:31.526 (world record) 2. Marion Clignet (France) 3:31.674 3. Lucy Tyler-Sharman (Australia) 3:31.830 4. Yvonne McGregor (Britain) 3:41.823 5. Natalia Karimova (Russia) 3:45.061 6. Svetlana Samokhalova (Russia) 3:46.216 7 Jane Quigley (U.S.) 3:46.493 8. Rasa Mazeikyte (Lithuania) 3:46.834 9. Tatian Stiajkina (Ukraine) 3:52.204 World 4,000 metres team pursuit semifinals : Italy (Adler Capelli, Cristiano Citto, Andrea Collinelli, Mauro Trentino) 4:00.958 (world record) beat Russia (Anton Chantyr, Edouard Gritsoun, Nikolai Kouznetsov) 4:06.534. France (Cyril Bos, Philippe Ermenault, Jean-Michel Monin, Francis Moreau) 4:05.104 beat Germany (Guido Fulst, Danilo Hondo, Thorsten Rund, Heiko Szonn) 4:05.463 Germany take the bronze medal as fastest losing semifinalist. Women's world 500 metres time trial final : 1. Felicia Ballanger (France) 34.829 2. Annett Neumann (Germany) 35.202 3. Michelle Ferris (Australia) 35.694 4. Magali Faure (France) 35.888 5. Olga Slioussareva (Russia) 36.170 6. Oksana Grichina (Russia) 36.242 7. Tanya Dubnicoff (Canada) 36.307 8. Kathrin Freitag (Germany) 36.491 9. Donna Wynd (New Zealand) 36.831 10. Mira Kasslin (Finland) 37.273 11. Wendy Everson (Britain) 37.624 12. Giovanna Troldi (Italy) 38.285 13. Rita Razmaite (Lithuania) 38.546 World 4,000 metres team pursuit championship final : Italy (Adler Capelli, Cristiano Citto, Andrea Collinelli, Mauro Trentino) 4:02.752 beat France (Cyril Bos, Philippe Ermenault, Jean-Michel Monin, Francis Moreau) 4:04.539 World sprint championship quarter finals (best of three matches) Florian Rousseau (France) beat Ainars Kiksis (Latvia) 2-0 Darryn Hill (Australia) beat Christian Arrue (U.S.) 2-0 Roberto Chiappa (Italy) beat Frederic Magne (France) 2-0 Marty Nothstein (U.S.) beat Pavel Buran (Czech Republic) 2-0 Women's world 3,000 metres individual pursuit championship quarter-finals : Marion Clignet (France) 3:30.974 (World Record) beat Jane Quigley (USA) 3:42.852 Natalia Karimova (Russia) 3:40.036 beat Yvonne McGregor (Britain) 3:43.078 Lucy Tyler-Sharman (Australia) 3:35.087 beat Svetlana Samokhvalova (Russia) 3:45.011 Antonella Bellutti (Italy) 3:32.174 caught and eliminated Rasa Mazeikyte (Lithuania)\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"CRICKET - PAKISTAN WIN TOSS, PUT ENGLAND IN TO BAT. BIRMINGHAM, England 1996-08-31 Pakistan won the toss and put England in to bat in the second limited overs cricket international at Edgbaston on Saturday. Surrey all-rounder Adam Hollioake was making his England debut, replacing Lancashire batsman Graham Lloyd, with seamer Peter Martin again being omitted from the 13. Pakistan kept the side who lost to England by five wickets at Old Trafford on Thursday in the first of the three one-day matches. Teams : England : Mike Atherton (captain), Nick Knight, Alec Stewart, Graham Thorpe, Matthew Maynard, Adam Hollioake, Ronnie Irani, Robert Croft, Darren Gough, Dean Headley, Alan Mullally. Pakistan : Aamir Sohail, Saeed Anwar, Ijaz Ahmed, Salim Malik, Inzamam-ul-Haq, Wasim Akram (captain), Moin Khan, Saqlain Mushtaq, Mushtaq Ahmed, Waqar Younis, Ata-ur-Rehman.\"}'\ntext: CRICKET - PAKISTAN WIN TOSS, PUT ENGLAND IN TO BAT. BIRMINGHAM, England 1996-08-31 Pakistan won the toss and put England in to bat in the second limited overs cricket international at Edgbaston on Saturday. Surrey all-rounder Adam Hollioake was making his England debut, replacing Lancashire batsman Graham Lloyd, with seamer Peter Martin again being omitted from the 13. Pakistan kept the side who lost to England by five wickets at Old Trafford on Thursday in the first of the three one-day matches. Teams : England : Mike Atherton (captain), Nick Knight, Alec Stewart, Graham Thorpe, Matthew Maynard, Adam Hollioake, Ronnie Irani, Robert Croft, Darren Gough, Dean Headley, Alan Mullally. Pakistan : Aamir Sohail, Saeed Anwar, Ijaz Ahmed, Salim Malik, Inzamam-ul-Haq, Wasim Akram (captain), Moin Khan, Saqlain Mushtaq, Mushtaq Ahmed, Waqar Younis, Ata-ur-Rehman.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"BASEBALL - GONZALEZ HOMERS TWICE AS RANGERS BEAT INDIANS. ARLINGTON, Texas 1996-08-31 Juan Gonzalez homered twice and Ivan Rodriguez added a two-run shot as the Texas Rangers defeated the Cleveland Indians 5-3 in a matchup of division leaders Friday. Rodriguez\\\\u0027s 18th homer, off Chad Ogea (7-5) in the first, gave Texas a 2-0 lead. One out later, Gonzalez smacked his 40th homer, extending his hitting streak to 20 games. Gonzalez, who hit in 21 straight games earlier this season, joined Mickey Rivers as the only players in Texas history with two 20-game streaks in the same year. Gonzalez hit his second homer in the third for his fifth multi-homer game of the season. Gonzalez has three 40-homer seasons and his 121 RBI broke Ruben Sierra\\\\u0027s team record of 119 set in 1989. The Indians had their four-game winning streak stopped. \\\\\"It\\\\u0027s not something I\\\\u0027m going to try to explain,\\\\\" said Texas manager Johnny Oates about his team winning seven of the 10 meetings from Cleveland this season. \\\\\"We\\\\u0027ve got two more regular season games against them and we might get lucky enough or unlucky enough to play them in the post-season.\\\\\" Roger Pavlik (15-7) gave up three runs and seven hits in 6 1/3 innings and became the fourth 15-game winner in the American League. Jeff Russell pitched two perfect innings for his third save. Brian Giles and Jim Thome homered for Cleveland. Cleveland\\\\u0027s lead over the White Sox in the American League Central dropped to nine games. Texas\\\\u0027s lead over Seattle in the West increased to six. At California, Tino Martinez\\\\u0027s two-run homer keyed a three-run first and Andy Pettitte became the league\\\\u0027s first 19-game winner as the New York Yankees beat the Angels 6-2. New York snapped a season-high five-game losing streak and also got homers from Mariano Duncan, Darryl Strawberry and Jim Leyritz. Pettite (19-7) allowed two runs and eight hits over eight innings with a walk and seven strikeouts. He improved to 12-2 following Yankees \\\\u0027 losses. Mariano Rivera pitched a scoreless ninth, striking out two. Ex-Yankee Randy Velarde hit his 11th homer, his most at any professional level. In Seattle, Pete Incaviglia\\\\u0027s grand slam with one out in the sixth snapped a tie and lifted the Baltimore Orioles past the Seattle Mariners, 5-2. It was Incaviglia\\\\u0027s sixth grand slam and 200th homer of his career. Baltimore\\\\u0027s Eddie Murray cracked his 20th homer of the season and 499th of his career. Jay Buhner hit his 38th homer and Edgar Martinez his 23rd for Seattle. The Orioles remained tied with the White Sox for the American League wild card with the Mariners a game back. In Toronto, Kevin Tapani (12-8) allowed two runs and six hits over 7 1/3 innings and Frank Thomas hit his 29th homer and drove in three runs as the Chicago White Sox cruised to an 11-2 victory over the Blue Jays. Thomas, Harold Baines and Robin Ventura each collected three hits. Baines homered and scored three runs. Danny Tartabull added two hits and three RBI as all Chicago starters got at least one hit. In Oakland, Dave Telgheder scattered seven hits over eight innings and Mark McGwire hit his major-league leading 45th homer and drove in three runs as the Athetlics blanked the Boston Red Sox 7-0. Telgheder (2-5) snapped a personal three-game losing streak. Buddy Groom pitched a perfect ninth inning. McGwire singled home a run to spark a three-run sixth and capped the scoring with a two-run homer in the seventh. The loss was Boston\\\\u0027s seventh in its last 29 games. In Detroit, Todd Van Poppel pitched a five-hitter for his first career shutout and Tony Clark homered to cap a four-run first inning as the Tigers blanked the Kansas City Royals 4-0. Van Poppel (3-6) walked two and struck out two in defeating the Royals for the second time this week. He threw 108 pitches as he lowered his ERA from 8.08 to 7.24. Kansas City has scored only one run in two games. In Milwaukee, Marc Newfield homered off Jose Parra (5-4) leading off the bottom of the 12th as the Brewers rallied for a 5-4 victory over the Minnesota Twins. Milwaukee has won 10 of its last 15. Bob Wickman (6-1) pitched 2 2/3 hitless innings for the win, his second for the Brewers. Matt Lawton hit a three-run homer off closer Mike Fetters with one out in the ninth to give Minnesota a 4-2 lead. But Milwaukee tied it up in the bottom of the ninth on pinch-hitter Dave Nilsson\\\\u0027s two-run double.\"}'\ntext: BASEBALL - GONZALEZ HOMERS TWICE AS RANGERS BEAT INDIANS. ARLINGTON, Texas 1996-08-31 Juan Gonzalez homered twice and Ivan Rodriguez added a two-run shot as the Texas Rangers defeated the Cleveland Indians 5-3 in a matchup of division leaders Friday. Rodriguez's 18th homer, off Chad Ogea (7-5) in the first, gave Texas a 2-0 lead. One out later, Gonzalez smacked his 40th homer, extending his hitting streak to 20 games. Gonzalez, who hit in 21 straight games earlier this season, joined Mickey Rivers as the only players in Texas history with two 20-game streaks in the same year. Gonzalez hit his second homer in the third for his fifth multi-homer game of the season. Gonzalez has three 40-homer seasons and his 121 RBI broke Ruben Sierra's team record of 119 set in 1989. The Indians had their four-game winning streak stopped. \"It's not something I'm going to try to explain,\" said Texas manager Johnny Oates about his team winning seven of the 10 meetings from Cleveland this season. \"We've got two more regular season games against them and we might get lucky enough or unlucky enough to play them in the post-season.\" Roger Pavlik (15-7) gave up three runs and seven hits in 6 1/3 innings and became the fourth 15-game winner in the American League. Jeff Russell pitched two perfect innings for his third save. Brian Giles and Jim Thome homered for Cleveland. Cleveland's lead over the White Sox in the American League Central dropped to nine games. Texas's lead over Seattle in the West increased to six. At California, Tino Martinez's two-run homer keyed a three-run first and Andy Pettitte became the league's first 19-game winner as the New York Yankees beat the Angels 6-2. New York snapped a season-high five-game losing streak and also got homers from Mariano Duncan, Darryl Strawberry and Jim Leyritz. Pettite (19-7) allowed two runs and eight hits over eight innings with a walk and seven strikeouts. He improved to 12-2 following Yankees ' losses. Mariano Rivera pitched a scoreless ninth, striking out two. Ex-Yankee Randy Velarde hit his 11th homer, his most at any professional level. In Seattle, Pete Incaviglia's grand slam with one out in the sixth snapped a tie and lifted the Baltimore Orioles past the Seattle Mariners, 5-2. It was Incaviglia's sixth grand slam and 200th homer of his career. Baltimore's Eddie Murray cracked his 20th homer of the season and 499th of his career. Jay Buhner hit his 38th homer and Edgar Martinez his 23rd for Seattle. The Orioles remained tied with the White Sox for the American League wild card with the Mariners a game back. In Toronto, Kevin Tapani (12-8) allowed two runs and six hits over 7 1/3 innings and Frank Thomas hit his 29th homer and drove in three runs as the Chicago White Sox cruised to an 11-2 victory over the Blue Jays. Thomas, Harold Baines and Robin Ventura each collected three hits. Baines homered and scored three runs. Danny Tartabull added two hits and three RBI as all Chicago starters got at least one hit. In Oakland, Dave Telgheder scattered seven hits over eight innings and Mark McGwire hit his major-league leading 45th homer and drove in three runs as the Athetlics blanked the Boston Red Sox 7-0. Telgheder (2-5) snapped a personal three-game losing streak. Buddy Groom pitched a perfect ninth inning. McGwire singled home a run to spark a three-run sixth and capped the scoring with a two-run homer in the seventh. The loss was Boston's seventh in its last 29 games. In Detroit, Todd Van Poppel pitched a five-hitter for his first career shutout and Tony Clark homered to cap a four-run first inning as the Tigers blanked the Kansas City Royals 4-0. Van Poppel (3-6) walked two and struck out two in defeating the Royals for the second time this week. He threw 108 pitches as he lowered his ERA from 8.08 to 7.24. Kansas City has scored only one run in two games. In Milwaukee, Marc Newfield homered off Jose Parra (5-4) leading off the bottom of the 12th as the Brewers rallied for a 5-4 victory over the Minnesota Twins. Milwaukee has won 10 of its last 15. Bob Wickman (6-1) pitched 2 2/3 hitless innings for the win, his second for the Brewers. Matt Lawton hit a three-run homer off closer Mike Fetters with one out in the ninth to give Minnesota a 4-2 lead. But Milwaukee tied it up in the bottom of the ninth on pinch-hitter Dave Nilsson's two-run double.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"CRICKET - ESSEX AND KENT MADE TO SWEAT IN TITLE RACE. LONDON 1996-08-31 Essex and Kent both face tense finishes on Monday as they attempt to keep pace with title hopefuls Derbyshire and Surrey, convincing three-day victors on Saturday, in the English county championship run-in. Essex need another 148 with five wickets in hand to beat Yorkshire after a maiden first-class century from Richard Kettleborough transformed a match which his side had seemed certain to lose. Kent will also need to keep their nerve against struggling Nottinghamshire who will enter the final day 137 ahead with four wickets left in a relatively low-scoring match at Tunbridge Wells. Derbyshire, nine-wicket winners over Worcestershire, and Surrey, who thrashed Warwickshire by an innings and 164 runs, can instead take the day off along with rivals Leicestershire, who beat Somerset inside two days. Warwickshire captain Tim Munton is tipping Surrey to emerge on top, impressed by the positive influence of Australian coach Dave Gilbert, but Derbyshire\\\\u0027s Australian captain Dean Jones is conceding nothing as his side chase their first title for 60 years. \\\\\"We took three absolutely brilliant catches in this match and our catching all season has been pretty impressive. Our catching will win or lose us the championship,\\\\\" he said.\"}'\ntext: CRICKET - ESSEX AND KENT MADE TO SWEAT IN TITLE RACE. LONDON 1996-08-31 Essex and Kent both face tense finishes on Monday as they attempt to keep pace with title hopefuls Derbyshire and Surrey, convincing three-day victors on Saturday, in the English county championship run-in. Essex need another 148 with five wickets in hand to beat Yorkshire after a maiden first-class century from Richard Kettleborough transformed a match which his side had seemed certain to lose. Kent will also need to keep their nerve against struggling Nottinghamshire who will enter the final day 137 ahead with four wickets left in a relatively low-scoring match at Tunbridge Wells. Derbyshire, nine-wicket winners over Worcestershire, and Surrey, who thrashed Warwickshire by an innings and 164 runs, can instead take the day off along with rivals Leicestershire, who beat Somerset inside two days. Warwickshire captain Tim Munton is tipping Surrey to emerge on top, impressed by the positive influence of Australian coach Dave Gilbert, but Derbyshire's Australian captain Dean Jones is conceding nothing as his side chase their first title for 60 years. \"We took three absolutely brilliant catches in this match and our catching all season has been pretty impressive. Our catching will win or lose us the championship,\" he said.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"GOLF - LEADING MONEY WINNERS ON EUROPEAN TOUR. LONDON 1996-08-31 Leading money winners on the European Tour after the British Masters won by Robert Allenby on Saturday (British unless stated) : 1. Ian Woosnam 510,258 pounds sterling 2. Colin Montgomerie 442,201 3. Robert Allenby (Australia) 407,748 4. Lee Westwood 301,972 5. Costantino Rocca (Italy) 297,157 6. Mark McNulty (Zimbabwe) 254,247 7. Andrew Coltart 248,142 8. Wayne Riley (Australia) 239,733 9. Raymond Russell 234,330 10. Paul Lawrie 211,420 11. Stephen Ames (Trinidad) 211,175 12. Frank Nobilo (New Zealand) 209,412 13. Paul McGinley (Ireland) 208,978 14. Padraig Harrington (Ireland) 202,593 15. Retief Goosen (South Africa) 195,283 16. Miguel Angel Jimenez (Spain) 184,180 17. Peter Mitchell 183,704 18. Miguel Angel Martin (Spain) 182,533 19. Jonathan Lomas 181,005 20. Paul Broadhurst 176,780\"}'\ntext: GOLF - LEADING MONEY WINNERS ON EUROPEAN TOUR. LONDON 1996-08-31 Leading money winners on the European Tour after the British Masters won by Robert Allenby on Saturday (British unless stated) : 1. Ian Woosnam 510,258 pounds sterling 2. Colin Montgomerie 442,201 3. Robert Allenby (Australia) 407,748 4. Lee Westwood 301,972 5. Costantino Rocca (Italy) 297,157 6. Mark McNulty (Zimbabwe) 254,247 7. Andrew Coltart 248,142 8. Wayne Riley (Australia) 239,733 9. Raymond Russell 234,330 10. Paul Lawrie 211,420 11. Stephen Ames (Trinidad) 211,175 12. Frank Nobilo (New Zealand) 209,412 13. Paul McGinley (Ireland) 208,978 14. Padraig Harrington (Ireland) 202,593 15. Retief Goosen (South Africa) 195,283 16. Miguel Angel Jimenez (Spain) 184,180 17. Peter Mitchell 183,704 18. Miguel Angel Martin (Spain) 182,533 19. Jonathan Lomas 181,005 20. Paul Broadhurst 176,780\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"RUGBY UNION - ENGLISH, SCOTTISH AND WELSH RESULTS. LONDON 1996-08-31 Results of English, Scottish and Welsh rugby union matches on Saturday : English National League one Harlequins 75 Gloucester 19 London Irish 27 Bristol 28 Northampton 46 West Hartlepool 20 Orrell 13 Bath 56 Sale 31 Wasps 33 Saracens 25 Leicester 23 Welsh division one Bridgend 13 Llanelli 9 Dunvant 21 Ebbw Vale 10 Treorchy 17 Newbridge 23 Newport 29 Caerphilly 10 Swansea 49 Cardiff 23 Scottish premier league division one Boroughmuir 20 Hawick 23 Currie 45 Heriot\\\\u0027s F.P. 5 Jed-Forest 17 Watsonians 54 Melrose 107 Stirling County 10\"}'\ntext: RUGBY UNION - ENGLISH, SCOTTISH AND WELSH RESULTS. LONDON 1996-08-31 Results of English, Scottish and Welsh rugby union matches on Saturday : English National League one Harlequins 75 Gloucester 19 London Irish 27 Bristol 28 Northampton 46 West Hartlepool 20 Orrell 13 Bath 56 Sale 31 Wasps 33 Saracens 25 Leicester 23 Welsh division one Bridgend 13 Llanelli 9 Dunvant 21 Ebbw Vale 10 Treorchy 17 Newbridge 23 Newport 29 Caerphilly 10 Swansea 49 Cardiff 23 Scottish premier league division one Boroughmuir 20 Hawick 23 Currie 45 Heriot's F.P. 5 Jed-Forest 17 Watsonians 54 Melrose 107 Stirling County 10\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - WALES BEAT SAN MARINO IN WORLD CUP QUALIFIER. CARDIFF 1996-08-31 Wales beat San Marino 6-0 (halftime 4-0) in a World Cup soccer European group 7 qualifier on Saturday. Scorers : Dean Saunders (2nd minute, 75th), Mark Hughes (25th, 54th), Andy Melville (33rd), John Robinson (45th). Attendance : 15,150\"}'\ntext: SOCCER - WALES BEAT SAN MARINO IN WORLD CUP QUALIFIER. CARDIFF 1996-08-31 Wales beat San Marino 6-0 (halftime 4-0) in a World Cup soccer European group 7 qualifier on Saturday. Scorers : Dean Saunders (2nd minute, 75th), Mark Hughes (25th, 54th), Andy Melville (33rd), John Robinson (45th). Attendance : 15,150\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - UKRAINE BEAT NORTHERN IRELAND IN WORLD CUP QUALIFIER. BELFAST 1996-08-31 Ukraine beat Northern Ireland 1-0 (halftime 0-0) in a World Cup soccer European group nine qualifier on Saturday. Scorer : Sergei Rebrov (79th minute) Attendance : 9,358\"}'\ntext: SOCCER - UKRAINE BEAT NORTHERN IRELAND IN WORLD CUP QUALIFIER. BELFAST 1996-08-31 Ukraine beat Northern Ireland 1-0 (halftime 0-0) in a World Cup soccer European group nine qualifier on Saturday. Scorer : Sergei Rebrov (79th minute) Attendance : 9,358\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"RUGBY UNION - LYNAGH SEALS VICTORY OVER DWYER AND LEICESTER. LONDON 1996-08-31 Former Wallaby captain Michael Lynagh began his career in English club rugby in impeccable fashion on Saturday to frustrate his old coach Bob Dwyer on his league coaching debut with Leicester. Lynagh kicked five penalties and a conversion from his six attempts at goal to steer his multi-national Saracens side to a 25-23 home win and offer millionaire backer Nigel Wray an early return on his big investment in the north London club. French centre Philippe Sella also enjoyed a good game alongside Lynagh, although the home team scored only one try through England scrum-half Kyran Bracken. The new French connection at Harlequins also made a good start, Laurent Cabannes and Laurent Benezech scoring a try apiece in their side\\\\u0027s 75-19 victory over Gloucester. Former England captain Will Carling, handed the kicking duties, finished with 20 points. With the first day of the league season briefly shifting the spotlight away from the discord between the clubs and the Rugby Football Union, there were also emphatic victories for champions Bath, 56-13 winners over Orrell, and Northampton and narrow successes for Wasps and Bristol.\"}'\ntext: RUGBY UNION - LYNAGH SEALS VICTORY OVER DWYER AND LEICESTER. LONDON 1996-08-31 Former Wallaby captain Michael Lynagh began his career in English club rugby in impeccable fashion on Saturday to frustrate his old coach Bob Dwyer on his league coaching debut with Leicester. Lynagh kicked five penalties and a conversion from his six attempts at goal to steer his multi-national Saracens side to a 25-23 home win and offer millionaire backer Nigel Wray an early return on his big investment in the north London club. French centre Philippe Sella also enjoyed a good game alongside Lynagh, although the home team scored only one try through England scrum-half Kyran Bracken. The new French connection at Harlequins also made a good start, Laurent Cabannes and Laurent Benezech scoring a try apiece in their side's 75-19 victory over Gloucester. Former England captain Will Carling, handed the kicking duties, finished with 20 points. With the first day of the league season briefly shifting the spotlight away from the discord between the clubs and the Rugby Football Union, there were also emphatic victories for champions Bath, 56-13 winners over Orrell, and Northampton and narrow successes for Wasps and Bristol.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - SCOTTISH LEAGUE STANDINGS. LONDON 1996-08-31 Scottish league standings after Saturday\\\\u0027s matches (tabulated - played, won, drawn, lost, goals for, goals against, points) : Division one Greenock Morton 3 2 0 1 5 2 6 Dundee 3 1 2 0 3 2 5 St Johnstone 2 1 1 0 3 0 4 St Mirren 3 1 1 1 5 4 4 Airdrieonians 2 1 1 0 1 0 4 Falkirk 3 1 1 1 1 1 4 Clydebank 2 1 0 1 1 3 3 Partick 3 0 2 1 1 2 2 Stirling 3 0 1 2 1 3 1 East Fife 2 0 1 1 0 4 1 Division two Livingston 3 3 0 0 6 2 9 Queen of South 3 2 0 1 5 4 6 Ayr 3 1 2 0 8 2 5 Stenhousemuir 3 1 1 1 6 1 4 Hamilton 3 1 1 1 3 2 4 Stranraer 3 1 1 1 3 3 4 Brechin 3 0 3 0 2 2 3 Clyde 3 1 0 2 2 5 3 Dumbarton 3 0 2 1 3 4 2 Berwick 3 0 0 3 1 14 0 Division three Albion 3 3 0 0 5 0 9 Forfar 3 2 0 1 7 4 6 Cowdenbeath 3 2 0 1 4 3 6 Arbroath 3 1 2 0 4 2 5 Alloa 3 1 1 1 3 3 4 Queen\\\\u0027s Park 3 1 1 1 6 8 4 Montrose 3 1 0 2 3 4 3 Inverness Thistle 3 1 0 2 3 6 3 East Stirling 3 0 2 1 3 4 2 Ross County 3 0 0 3 3 7 0\"}'\ntext: SOCCER - SCOTTISH LEAGUE STANDINGS. LONDON 1996-08-31 Scottish league standings after Saturday's matches (tabulated - played, won, drawn, lost, goals for, goals against, points) : Division one Greenock Morton 3 2 0 1 5 2 6 Dundee 3 1 2 0 3 2 5 St Johnstone 2 1 1 0 3 0 4 St Mirren 3 1 1 1 5 4 4 Airdrieonians 2 1 1 0 1 0 4 Falkirk 3 1 1 1 1 1 4 Clydebank 2 1 0 1 1 3 3 Partick 3 0 2 1 1 2 2 Stirling 3 0 1 2 1 3 1 East Fife 2 0 1 1 0 4 1 Division two Livingston 3 3 0 0 6 2 9 Queen of South 3 2 0 1 5 4 6 Ayr 3 1 2 0 8 2 5 Stenhousemuir 3 1 1 1 6 1 4 Hamilton 3 1 1 1 3 2 4 Stranraer 3 1 1 1 3 3 4 Brechin 3 0 3 0 2 2 3 Clyde 3 1 0 2 2 5 3 Dumbarton 3 0 2 1 3 4 2 Berwick 3 0 0 3 1 14 0 Division three Albion 3 3 0 0 5 0 9 Forfar 3 2 0 1 7 4 6 Cowdenbeath 3 2 0 1 4 3 6 Arbroath 3 1 2 0 4 2 5 Alloa 3 1 1 1 3 3 4 Queen's Park 3 1 1 1 6 8 4 Montrose 3 1 0 2 3 4 3 Inverness Thistle 3 1 0 2 3 6 3 East Stirling 3 0 2 1 3 4 2 Ross County 3 0 0 3 3 7 0\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - ENGLISH LEAGUE STANDINGS. LONDON 1996-08-31 English soccer league standings after Saturday\\\\u0027s matches (tabulated - played, won, drawn, lost, goals for, goals against, points) : Division one Stoke 4 3 1 0 7 4 10 Barnsley 3 3 0 0 8 2 9 Norwich 4 3 0 1 5 3 9 Tranmere 4 2 1 1 6 4 7 Bolton 3 2 1 0 5 2 7 Queens Park Rangers 3 2 1 0 5 3 7 Wolverhampton 4 2 1 1 5 3 7 Swindon 4 2 1 1 5 4 7 Bradford 4 2 0 2 4 3 6 Portsmouth 4 2 0 2 4 5 6 Ipswich 4 1 2 1 9 7 5 Crystal Palace 4 1 2 1 4 3 5 Port Vale 4 1 2 1 4 4 5 Birmingham 2 1 1 0 5 4 4 Reading 4 1 1 2 5 10 4 Huddersfield 3 1 1 1 4 4 4 Oxford 4 1 0 3 6 5 3 Manchester City 3 1 0 2 2 3 3 West Bromwich 3 0 2 1 2 3 2 Oldham 4 0 1 3 5 9 1 Sheffield United 2 0 1 1 4 5 1 Grimsby 4 0 1 3 4 8 1 Southend 4 0 1 3 2 10 1 Charlton 2 0 1 1 1 3 1 Division two Plymouth 4 3 1 0 10 6 10 Brentford 4 3 1 0 9 3 10 Bury 4 3 1 0 8 2 10 Chesterfield 4 3 0 1 4 2 9 Millwall 4 2 1 1 7 5 7 Shrewsbury 4 2 1 1 6 6 7 Blackpool 4 2 1 1 3 2 7 York 4 2 0 2 6 6 6 Burnley 4 2 0 2 6 7 6 Bournemouth 4 2 0 2 5 5 6 Watford 4 2 0 2 4 5 6 Bristol Rovers 3 1 2 0 2 1 5 Peterborough 3 1 1 1 4 4 4 Preston 4 1 1 2 4 5 4 Crewe 4 1 1 2 4 6 4 Gillingham 4 1 1 2 4 6 4 Notts County 3 1 1 1 2 2 4 Bristol City 4 1 0 3 7 8 3 Luton 4 1 0 3 4 10 3 Wycombe 4 0 3 1 2 3 3 Wrexham 2 0 2 0 5 5 2 Stockport 4 0 2 2 1 3 2 Rotherham 4 0 1 3 3 6 1 Walsall 3 0 1 2 2 4 1 Division three Wigan 4 3 1 0 9 4 10 Fulham 4 3 0 1 5 3 9 Hull 4 2 2 0 4 2 8 Hartlepool 4 2 1 1 6 5 7 Torquay 4 2 1 1 5 3 7 Cardiff 4 2 1 1 3 2 7 Scunthorpe 4 2 1 1 3 3 7 Carlisle 4 2 1 1 2 1 7 Scarborough 4 1 3 0 5 3 6 Northampton 4 1 2 1 6 4 5 Lincoln 4 1 2 1 5 5 5 Barnet 4 1 2 1 4 2 5 Exeter 4 1 2 1 4 5 5 Cambridge United 4 1 2 1 3 4 5 Darlington 4 1 1 2 9 8 4 Chester 4 1 1 2 6 7 4 Doncaster 4 1 1 2 4 5 4 Leyton Orient 4 1 1 2 3 3 4 Brighton 4 1 1 2 3 6 4 Hereford 4 1 1 2 2 3 4 Swansea 4 1 0 3 4 9 3 Colchester 4 0 3 1 2 4 3 Rochdale 4 0 2 2 2 4 2 Mansfield 4 0 2 2 2 6 2\"}'\ntext: SOCCER - ENGLISH LEAGUE STANDINGS. LONDON 1996-08-31 English soccer league standings after Saturday's matches (tabulated - played, won, drawn, lost, goals for, goals against, points) : Division one Stoke 4 3 1 0 7 4 10 Barnsley 3 3 0 0 8 2 9 Norwich 4 3 0 1 5 3 9 Tranmere 4 2 1 1 6 4 7 Bolton 3 2 1 0 5 2 7 Queens Park Rangers 3 2 1 0 5 3 7 Wolverhampton 4 2 1 1 5 3 7 Swindon 4 2 1 1 5 4 7 Bradford 4 2 0 2 4 3 6 Portsmouth 4 2 0 2 4 5 6 Ipswich 4 1 2 1 9 7 5 Crystal Palace 4 1 2 1 4 3 5 Port Vale 4 1 2 1 4 4 5 Birmingham 2 1 1 0 5 4 4 Reading 4 1 1 2 5 10 4 Huddersfield 3 1 1 1 4 4 4 Oxford 4 1 0 3 6 5 3 Manchester City 3 1 0 2 2 3 3 West Bromwich 3 0 2 1 2 3 2 Oldham 4 0 1 3 5 9 1 Sheffield United 2 0 1 1 4 5 1 Grimsby 4 0 1 3 4 8 1 Southend 4 0 1 3 2 10 1 Charlton 2 0 1 1 1 3 1 Division two Plymouth 4 3 1 0 10 6 10 Brentford 4 3 1 0 9 3 10 Bury 4 3 1 0 8 2 10 Chesterfield 4 3 0 1 4 2 9 Millwall 4 2 1 1 7 5 7 Shrewsbury 4 2 1 1 6 6 7 Blackpool 4 2 1 1 3 2 7 York 4 2 0 2 6 6 6 Burnley 4 2 0 2 6 7 6 Bournemouth 4 2 0 2 5 5 6 Watford 4 2 0 2 4 5 6 Bristol Rovers 3 1 2 0 2 1 5 Peterborough 3 1 1 1 4 4 4 Preston 4 1 1 2 4 5 4 Crewe 4 1 1 2 4 6 4 Gillingham 4 1 1 2 4 6 4 Notts County 3 1 1 1 2 2 4 Bristol City 4 1 0 3 7 8 3 Luton 4 1 0 3 4 10 3 Wycombe 4 0 3 1 2 3 3 Wrexham 2 0 2 0 5 5 2 Stockport 4 0 2 2 1 3 2 Rotherham 4 0 1 3 3 6 1 Walsall 3 0 1 2 2 4 1 Division three Wigan 4 3 1 0 9 4 10 Fulham 4 3 0 1 5 3 9 Hull 4 2 2 0 4 2 8 Hartlepool 4 2 1 1 6 5 7 Torquay 4 2 1 1 5 3 7 Cardiff 4 2 1 1 3 2 7 Scunthorpe 4 2 1 1 3 3 7 Carlisle 4 2 1 1 2 1 7 Scarborough 4 1 3 0 5 3 6 Northampton 4 1 2 1 6 4 5 Lincoln 4 1 2 1 5 5 5 Barnet 4 1 2 1 4 2 5 Exeter 4 1 2 1 4 5 5 Cambridge United 4 1 2 1 3 4 5 Darlington 4 1 1 2 9 8 4 Chester 4 1 1 2 6 7 4 Doncaster 4 1 1 2 4 5 4 Leyton Orient 4 1 1 2 3 3 4 Brighton 4 1 1 2 3 6 4 Hereford 4 1 1 2 2 3 4 Swansea 4 1 0 3 4 9 3 Colchester 4 0 3 1 2 4 3 Rochdale 4 0 2 2 2 4 2 Mansfield 4 0 2 2 2 6 2\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - SCOTTISH LEAGUE RESULTS. GLASGOW 1996-08-31 Results of Scottish league matches on Saturday : Division one Greenock Morton 1 Falkirk 0 Partick 1 St Mirren 1 Stirling 1 Dundee 1 Postponed : East Fife v Clydebank, St Johnstone v Airdrieonians. Division two Ayr 6 Berwick 0 Clyde 0 Queen of South 2 Dumbarton 1 Brechin 1 Livingston 1 Hamilton 0 Stenhousemuir 0 Stranraer 1 Division three Albion 2 Cowdenbeath 0 Arbroath 0 East Stirling 0 Inverness Thistle 1 Alloa 0 Montrose 2 Ross County 1 Queen\\\\u0027s Park 1 Forfar 4\"}'\ntext: SOCCER - SCOTTISH LEAGUE RESULTS. GLASGOW 1996-08-31 Results of Scottish league matches on Saturday : Division one Greenock Morton 1 Falkirk 0 Partick 1 St Mirren 1 Stirling 1 Dundee 1 Postponed : East Fife v Clydebank, St Johnstone v Airdrieonians. Division two Ayr 6 Berwick 0 Clyde 0 Queen of South 2 Dumbarton 1 Brechin 1 Livingston 1 Hamilton 0 Stenhousemuir 0 Stranraer 1 Division three Albion 2 Cowdenbeath 0 Arbroath 0 East Stirling 0 Inverness Thistle 1 Alloa 0 Montrose 2 Ross County 1 Queen's Park 1 Forfar 4\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - ENGLISH LEAGUE RESULTS. LONDON 1996-08-31 Results of English soccer matches on Saturday : Division one Bradford 1 Tranmere 0 Grimsby 0 Portsmouth 1 Huddersfield 1 Crystal Palace 1 Norwich 1 Wolverhampton 0 Oldham 3 Ipswich 3 Port Vale 2 Oxford 0 Reading 2 Stoke 2 Southend 1 Swindon 3 Postponed : Birmingham v Barnsley, Manchester City v Charlton Playing Sunday : Queens Park Rangers v Bolton Division two Blackpool 0 Wycombe 0 Bournemouth 1 Peterborough 2 Bristol Rovers 1 Stockport 1 Bury 4 Bristol City 0 Crewe 0 Watford 2 Gillingham 0 Chesterfield 1 Luton 1 Rotherham 0 Millwall 2 Burnley 1 Notts County 0 York 1 Shrewsbury 0 Brentford3 Postponed : Walsall v Wrexham Division three Brighton 1 Scunthorpe 1 Cambridge United 0 Cardiff 2 Colchester 1 Hereford 1 Doncaster 3 Darlington 2 Fulham 1 Carlisle 0 Hull 0 Barnet 0 Leyton Orient 2 Hartlepool 0 Mansfield 0 Rochdale 0 Scarborough 1 Northampton 1 Torquay 2 Exeter 0 Wigan 4 Chester 2\"}'\ntext: SOCCER - ENGLISH LEAGUE RESULTS. LONDON 1996-08-31 Results of English soccer matches on Saturday : Division one Bradford 1 Tranmere 0 Grimsby 0 Portsmouth 1 Huddersfield 1 Crystal Palace 1 Norwich 1 Wolverhampton 0 Oldham 3 Ipswich 3 Port Vale 2 Oxford 0 Reading 2 Stoke 2 Southend 1 Swindon 3 Postponed : Birmingham v Barnsley, Manchester City v Charlton Playing Sunday : Queens Park Rangers v Bolton Division two Blackpool 0 Wycombe 0 Bournemouth 1 Peterborough 2 Bristol Rovers 1 Stockport 1 Bury 4 Bristol City 0 Crewe 0 Watford 2 Gillingham 0 Chesterfield 1 Luton 1 Rotherham 0 Millwall 2 Burnley 1 Notts County 0 York 1 Shrewsbury 0 Brentford3 Postponed : Walsall v Wrexham Division three Brighton 1 Scunthorpe 1 Cambridge United 0 Cardiff 2 Colchester 1 Hereford 1 Doncaster 3 Darlington 2 Fulham 1 Carlisle 0 Hull 0 Barnet 0 Leyton Orient 2 Hartlepool 0 Mansfield 0 Rochdale 0 Scarborough 1 Northampton 1 Torquay 2 Exeter 0 Wigan 4 Chester 2\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"CRICKET - ENGLISH COUNTY CHAMPIONSHIP SCORES. LONDON 1996-08-31 Results and close scores of four-day English county championship matches on Saturday : At Portsmouth : Middlesex beat Hampshire by 188 runs. Middlesex 199 and 426, Hampshire 232 and 205 (A. Fraser 5-79, P. Tufnell 4-39). Middlesex 20 points, Hampshire 5. At Chester-le-Street : Glamorgan beat Durham by 141 runs. Glamorgan 259 and 207, Durham 114 and 211. Glamorgan 22 points, Durham 4. At Chesterfield : Derbyshire beat Worcestershire by nine wickets. Worcestershire 238 and 303 (K. Spiring 130 not out, S. Rhodes 57 ; P. DeFreitas 4-70), Derbyshire 471 and 71-1. Derbyshire 24 points, Worcestershire 5. At The Oval (London) : Surrey beat Warwickshire by an innings and 164 runs. Warwickshire 195 and 109 (J. Benjamin 4-17, M. Bicknell 4-38), Surrey 468 (C. Lewis 94, M. Butcher 70, G. Kersey 63, J. Ratcliffe 63, D. Bicknell 55). Surrey 24 points, Warwickshire 2. At Headingley (Leeds) : Yorkshire 290 and 329 (R. Kettleborough 108, G. Hamilton 61 ; P. Such 8-118), Essex 372 and 100-5. At Hove : Sussex 363 and 144, Lancashire 218 and 53-0. At Tunbridge Wells : Nottinghamshire 214 and 167-6 (C. Tolley 64 not out), Kent 244 (C. Hooper 58 ; C. Tolley 4-68, K. Evans 4-71) At Bristol : Gloucestershire 183 and 249 (J. Russell 75), Northamptonshire 190 and 218-9.\"}'\ntext: CRICKET - ENGLISH COUNTY CHAMPIONSHIP SCORES. LONDON 1996-08-31 Results and close scores of four-day English county championship matches on Saturday : At Portsmouth : Middlesex beat Hampshire by 188 runs. Middlesex 199 and 426, Hampshire 232 and 205 (A. Fraser 5-79, P. Tufnell 4-39). Middlesex 20 points, Hampshire 5. At Chester-le-Street : Glamorgan beat Durham by 141 runs. Glamorgan 259 and 207, Durham 114 and 211. Glamorgan 22 points, Durham 4. At Chesterfield : Derbyshire beat Worcestershire by nine wickets. Worcestershire 238 and 303 (K. Spiring 130 not out, S. Rhodes 57 ; P. DeFreitas 4-70), Derbyshire 471 and 71-1. Derbyshire 24 points, Worcestershire 5. At The Oval (London) : Surrey beat Warwickshire by an innings and 164 runs. Warwickshire 195 and 109 (J. Benjamin 4-17, M. Bicknell 4-38), Surrey 468 (C. Lewis 94, M. Butcher 70, G. Kersey 63, J. Ratcliffe 63, D. Bicknell 55). Surrey 24 points, Warwickshire 2. At Headingley (Leeds) : Yorkshire 290 and 329 (R. Kettleborough 108, G. Hamilton 61 ; P. Such 8-118), Essex 372 and 100-5. At Hove : Sussex 363 and 144, Lancashire 218 and 53-0. At Tunbridge Wells : Nottinghamshire 214 and 167-6 (C. Tolley 64 not out), Kent 244 (C. Hooper 58 ; C. Tolley 4-68, K. Evans 4-71) At Bristol : Gloucestershire 183 and 249 (J. Russell 75), Northamptonshire 190 and 218-9.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"MOTOR RACING - LEADING QUALIFIERS FOR VANCOUVER INDYCAR RACE. VANCOUVER 1996-08-31 Top ten drivers in grid for Sunday\\\\u0027s Vancouver IndyCar race after final qualifying on Saturday (tabulate by driver, country, chassis, motor and lap times in seconds) : 1. Alex Zanardi (Italy), Reynard Honda, 53.980 (113.576 mph / 182.778 kph) 2. Michael Andretti (U.S.), Lola Ford Cosworth, 54.483 3. Bobby Rahal (U.S.), Reynard Mercedes-Benz, 54.507 4. Bryan Herta (U.S.), Reynard Mercedes-Benz, 54.578 5. Jimmy Vasser (U.S.), Reynard Honda, 54.617 6. Paul Tracy (Canada), Penske Mercedes-Benz, 54.620 7. Al Unser Jr (U.S.), Penske Mercedes-Benz, 54.683 8. Andre Ribeiro (Brazil), Lola Honda, 54.750 9. Mauricio Gugelmin (Brazil), Reynard Ford Cosworth, 54.762 10. Gil de Ferran (Brazil), Reynard Honda, 54.774\"}'\ntext: MOTOR RACING - LEADING QUALIFIERS FOR VANCOUVER INDYCAR RACE. VANCOUVER 1996-08-31 Top ten drivers in grid for Sunday's Vancouver IndyCar race after final qualifying on Saturday (tabulate by driver, country, chassis, motor and lap times in seconds) : 1. Alex Zanardi (Italy), Reynard Honda, 53.980 (113.576 mph / 182.778 kph) 2. Michael Andretti (U.S.), Lola Ford Cosworth, 54.483 3. Bobby Rahal (U.S.), Reynard Mercedes-Benz, 54.507 4. Bryan Herta (U.S.), Reynard Mercedes-Benz, 54.578 5. Jimmy Vasser (U.S.), Reynard Honda, 54.617 6. Paul Tracy (Canada), Penske Mercedes-Benz, 54.620 7. Al Unser Jr (U.S.), Penske Mercedes-Benz, 54.683 8. Andre Ribeiro (Brazil), Lola Honda, 54.750 9. Mauricio Gugelmin (Brazil), Reynard Ford Cosworth, 54.762 10. Gil de Ferran (Brazil), Reynard Honda, 54.774\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - CANADA BEAT PANAMA 3-1 IN WORLD CUP QUALIFIER. EDMONTON 1996-08-31 Canada beat Panama 3-1 (halftime 2-0) in their CONCACAF semifinal phase qualifying match for the 1998 World Cup on Friday. Scorers : Canada - Aunger (41st min, pen), Paul Peschisolido (42nd), Carlo Corrazin (87th) Panama - Jorge Luis Dely Valdes (50th) Attendance : 9,402\"}'\ntext: SOCCER - CANADA BEAT PANAMA 3-1 IN WORLD CUP QUALIFIER. EDMONTON 1996-08-31 Canada beat Panama 3-1 (halftime 2-0) in their CONCACAF semifinal phase qualifying match for the 1998 World Cup on Friday. Scorers : Canada - Aunger (41st min, pen), Paul Peschisolido (42nd), Carlo Corrazin (87th) Panama - Jorge Luis Dely Valdes (50th) Attendance : 9,402\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"RUGBY UNION - SPRINGBOKS FINALLY BREAK ALL BLACK SPELL. Andy Colquhoun JOHANNESBURG 1996-08-31 South Africa managed to avoid a fifth successive defeat in 1996 at the hands of the All Blacks with an emphatic 32-22 victory in front of an ecstatic Ellis Park crowd on Saturday. They scored three tries in recording their highest total against New Zealand, salvaging some pride in a season in which the world champions have lost five out of eight tests. It also ended a run of nine successive victories this year for New Zealand but arrived too late to prevent a 2-1 series defeat and an historic first All Black series triumph on South African soil. Springbok scrum-half Joost van der Westhuizen was his side\\\\u0027s inspiration, scoring their opening try and making the third for flanker Andre Venter from a quickly taken penalty to give his side a 29-8 lead after 54 minutes. Fullback Andre Joubert scored the other, scorching in from 40 metres at the start of the second half to add to his three long-range penalties. The All Blacks salvaged some pride by scoring two tries from centre Walter Little and scrum-half Justin Marshall in the final five minutes to close a gap which at one point stood at 24 points. But they generally endured an off-day, highlighted by recalled fly-half Andrew Mehrtens who missed five out of eight kicks at goal. Recalled fly-half Henry Honiball kicked the Springboks into a 6-0 lead after 10 minutes only to see Andrew Mehrtens launch a penalty from eight metres inside his own half to narrow the gap. Mehrtens missed three further penalties and a conversion in the first 40 minutes which could have put his side ahead, but it was the Springboks who looked the more dangerous. Their promise was realised when Joubert made a 40-metre break in the 25th minute and, although winger Pieter Hendriks appeared to knock on Joubert\\\\u0027s reverse pass, Welsh referee Derek Bevan allowed Van der Westhuizen to pick up and score under the posts. Honiball converted and Joubert kicked a penalty before All Black hooker Sean Fitzpatrick scored a try from close range on the stroke of half-time to narrow the lead to 16-8 and hint at a comeback. Instead Joubert kicked another long penalty and then raced around the outside of the defence to score the Springboks \\\\u0027 second try. A quick penalty from Van der Westhuizen five metres from the All Black line set up the third try for Venter five minutes later and when Joubert kicked his third penalty the Springboks held an unassailable 32-8 lead going into the last quarter. When the All Blacks did break through, it was too late. Centre Walter Little followed up Mehrtens \\\\u0027 kick to score under the posts and scrum-half Justin Marshall forced himself over from a ruck close to the line in injury-time to give them some consolation. South Africa - 15 - Andre Joubert, 14 - Justin Swart, 13 - Japie Mulder (Joel Stransky, 48 mins) 12 - Danie van Schalkwyk, 11 - Pieter Hendriks ; 10 - Henry Honiball, 9 - Joost van der Westhuizen ; 8 - Gary Teichmann (captain), 7 - Andre Venter (Wayne Fyvie, 75), 6 - Ruben Kruge, 5 - Mark Andrews (Fritz van Heerden, 39), 4 - Kobus Wiese, 3 - Marius Hurter, 2 - James Dalton, 1 - Dawie Theron (Garry Pagel, 66). New Zealand - 15 - Christian Cullen (Alama Ieremia, 70), 14 - Jeff Wilson, 13 - Walter Little, 12 - Frank Bunce, 11 - Glen Osborne ; 10 - Andrew Mehrtens, 9 - Justin Marshall ; 8 - Zinzan Brooke, 7 - Josh Kronfeld, 6 - Michael Jones (Glenn Taylor, 53), 5 - Robin Brooke, 4 - Ian Jones, 3 - Olo Brown, 2 - Sean Fitzpatrick (captain), 1 - Craig Dowd.\"}'\ntext: RUGBY UNION - SPRINGBOKS FINALLY BREAK ALL BLACK SPELL. Andy Colquhoun JOHANNESBURG 1996-08-31 South Africa managed to avoid a fifth successive defeat in 1996 at the hands of the All Blacks with an emphatic 32-22 victory in front of an ecstatic Ellis Park crowd on Saturday. They scored three tries in recording their highest total against New Zealand, salvaging some pride in a season in which the world champions have lost five out of eight tests. It also ended a run of nine successive victories this year for New Zealand but arrived too late to prevent a 2-1 series defeat and an historic first All Black series triumph on South African soil. Springbok scrum-half Joost van der Westhuizen was his side's inspiration, scoring their opening try and making the third for flanker Andre Venter from a quickly taken penalty to give his side a 29-8 lead after 54 minutes. Fullback Andre Joubert scored the other, scorching in from 40 metres at the start of the second half to add to his three long-range penalties. The All Blacks salvaged some pride by scoring two tries from centre Walter Little and scrum-half Justin Marshall in the final five minutes to close a gap which at one point stood at 24 points. But they generally endured an off-day, highlighted by recalled fly-half Andrew Mehrtens who missed five out of eight kicks at goal. Recalled fly-half Henry Honiball kicked the Springboks into a 6-0 lead after 10 minutes only to see Andrew Mehrtens launch a penalty from eight metres inside his own half to narrow the gap. Mehrtens missed three further penalties and a conversion in the first 40 minutes which could have put his side ahead, but it was the Springboks who looked the more dangerous. Their promise was realised when Joubert made a 40-metre break in the 25th minute and, although winger Pieter Hendriks appeared to knock on Joubert's reverse pass, Welsh referee Derek Bevan allowed Van der Westhuizen to pick up and score under the posts. Honiball converted and Joubert kicked a penalty before All Black hooker Sean Fitzpatrick scored a try from close range on the stroke of half-time to narrow the lead to 16-8 and hint at a comeback. Instead Joubert kicked another long penalty and then raced around the outside of the defence to score the Springboks ' second try. A quick penalty from Van der Westhuizen five metres from the All Black line set up the third try for Venter five minutes later and when Joubert kicked his third penalty the Springboks held an unassailable 32-8 lead going into the last quarter. When the All Blacks did break through, it was too late. Centre Walter Little followed up Mehrtens ' kick to score under the posts and scrum-half Justin Marshall forced himself over from a ruck close to the line in injury-time to give them some consolation. South Africa - 15 - Andre Joubert, 14 - Justin Swart, 13 - Japie Mulder (Joel Stransky, 48 mins) 12 - Danie van Schalkwyk, 11 - Pieter Hendriks ; 10 - Henry Honiball, 9 - Joost van der Westhuizen ; 8 - Gary Teichmann (captain), 7 - Andre Venter (Wayne Fyvie, 75), 6 - Ruben Kruge, 5 - Mark Andrews (Fritz van Heerden, 39), 4 - Kobus Wiese, 3 - Marius Hurter, 2 - James Dalton, 1 - Dawie Theron (Garry Pagel, 66). New Zealand - 15 - Christian Cullen (Alama Ieremia, 70), 14 - Jeff Wilson, 13 - Walter Little, 12 - Frank Bunce, 11 - Glen Osborne ; 10 - Andrew Mehrtens, 9 - Justin Marshall ; 8 - Zinzan Brooke, 7 - Josh Kronfeld, 6 - Michael Jones (Glenn Taylor, 53), 5 - Robin Brooke, 4 - Ian Jones, 3 - Olo Brown, 2 - Sean Fitzpatrick (captain), 1 - Craig Dowd.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"RUGBY UNION - SOUTH AFRICA BEAT ALL BLACKS 32-22. JOHANNESBURG 1996-08-31 South Africa beat New Zealand 32-22 (haltime 16-8) in the final test match of their three-test series at Ellis Park on Saturday. Scorers : South Africa - Tries : Joost van der Westhuizen (2), Andre Joubert. Conversion : Henry Honiball. Penalties : Honiball (2), Joubert (3). New Zealand - Tries : Sean Fitzpatrick, Walter Little, Justin Marshall. Conversions : Andrew Mehrtens (2). Penalties : Mehrtens. New Zealand win test series 2-1.\"}'\ntext: RUGBY UNION - SOUTH AFRICA BEAT ALL BLACKS 32-22. JOHANNESBURG 1996-08-31 South Africa beat New Zealand 32-22 (haltime 16-8) in the final test match of their three-test series at Ellis Park on Saturday. Scorers : South Africa - Tries : Joost van der Westhuizen (2), Andre Joubert. Conversion : Henry Honiball. Penalties : Honiball (2), Joubert (3). New Zealand - Tries : Sean Fitzpatrick, Walter Little, Justin Marshall. Conversions : Andrew Mehrtens (2). Penalties : Mehrtens. New Zealand win test series 2-1.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - MAURITANIA DISSOLVES NATIONAL TEAM AFTER CUP EXIT. NOUAKCHOTT 1996-08-31 Mauritania\\\\u0027s soccer federation dissolved the national team and suspended this season\\\\u0027s domestic championship on Saturday in the wake of the country\\\\u0027s failure to qualify for the African Nations \\\\u0027 Cup. \\\\\"Since Mauritania has been eliminated on all fronts and the next commitments are not for another two years, we have reason to take a break,\\\\\" federation president Mohamed Lemine Cheiguer said. The North Africans were held to a goalless draw by Benin on Friday after losing the first leg of their qualifying tie 4-1.\"}'\ntext: SOCCER - MAURITANIA DISSOLVES NATIONAL TEAM AFTER CUP EXIT. NOUAKCHOTT 1996-08-31 Mauritania's soccer federation dissolved the national team and suspended this season's domestic championship on Saturday in the wake of the country's failure to qualify for the African Nations ' Cup. \"Since Mauritania has been eliminated on all fronts and the next commitments are not for another two years, we have reason to take a break,\" federation president Mohamed Lemine Cheiguer said. The North Africans were held to a goalless draw by Benin on Friday after losing the first leg of their qualifying tie 4-1.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - MAURITANIA DRAW WITH BENIN IN AFRICAN NATIONS CUP. NOUAKCHOTT 1996-08-31 Mauritania drew 0-0 with Benin in their African Nations Cup preliminary round, second leg soccer match on Friday. Benin won 4-1 on aggregate.\"}'\ntext: SOCCER - MAURITANIA DRAW WITH BENIN IN AFRICAN NATIONS CUP. NOUAKCHOTT 1996-08-31 Mauritania drew 0-0 with Benin in their African Nations Cup preliminary round, second leg soccer match on Friday. Benin won 4-1 on aggregate.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:14] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - YUGOSLAV LEAGUE RESULTS / STANDINGS. BELGRADE 1996-08-31 Results of Yugoslav league soccer matches played on Saturday : Division A Hajduk 2 Proleter (Z) 0 Zemun 1 Rad (B) 0 Borac 1 Mladost (L) 2 Cukaricki 1 Vojvodina 0 Buducnost 1 Red Star 3 Partizan 6 Becej 0 Standings (tabulate under won, drawn, lost, goals for, goals against, points) : Red Star 4 4 0 0 9 3 12 Partizan 4 3 1 0 13 3 10 Mladost (L) 4 2 1 1 8 5 7 Vojvodina 4 2 1 1 5 3 7 Becej 4 2 1 1 5 7 7 Hajduk 4 2 0 2 5 3 6 Cukaricki 4 2 0 2 6 6 6 Zemun 4 1 2 1 3 3 5 Rad (B) 4 1 1 2 2 3 4 Buducnost 4 1 0 3 4 8 3 Proleter (Z) 4 0 1 3 2 9 1 Borac 4 0 0 4 1 10 0 Division B Sloboda 4 Mladost (BJ) 0 Buducnost (V) 0 OFK Beograd 1 Rudar 0 OFK Kikinda 1 Obilic 2 Zeleznik (B) 0 Sutjeska 1 Loznica 0 Radnicki (N) - Spartak (to be played on Sunday) Standings : Obilic 4 4 0 0 10 1 12 OFK Kikinda 4 3 0 1 8 3 9 Sutjeska 4 3 0 1 7 5 9 Loznica 4 2 0 2 7 4 6 OFK Beograd 4 1 3 0 5 4 6 Buducnost (V) 4 2 0 2 4 5 6 Sloboda 4 1 1 2 8 8 4 Spartak 3 1 1 1 3 3 4 Radnicki (N) 3 1 0 2 5 6 3 Zeleznik (B) 4 1 0 3 4 7 3 Rudar 4 1 0 3 1 7 3 Mladost (BJ) 4 0 1 3 2 10 1\"}'\ntext: SOCCER - YUGOSLAV LEAGUE RESULTS / STANDINGS. BELGRADE 1996-08-31 Results of Yugoslav league soccer matches played on Saturday : Division A Hajduk 2 Proleter (Z) 0 Zemun 1 Rad (B) 0 Borac 1 Mladost (L) 2 Cukaricki 1 Vojvodina 0 Buducnost 1 Red Star 3 Partizan 6 Becej 0 Standings (tabulate under won, drawn, lost, goals for, goals against, points) : Red Star 4 4 0 0 9 3 12 Partizan 4 3 1 0 13 3 10 Mladost (L) 4 2 1 1 8 5 7 Vojvodina 4 2 1 1 5 3 7 Becej 4 2 1 1 5 7 7 Hajduk 4 2 0 2 5 3 6 Cukaricki 4 2 0 2 6 6 6 Zemun 4 1 2 1 3 3 5 Rad (B) 4 1 1 2 2 3 4 Buducnost 4 1 0 3 4 8 3 Proleter (Z) 4 0 1 3 2 9 1 Borac 4 0 0 4 1 10 0 Division B Sloboda 4 Mladost (BJ) 0 Buducnost (V) 0 OFK Beograd 1 Rudar 0 OFK Kikinda 1 Obilic 2 Zeleznik (B) 0 Sutjeska 1 Loznica 0 Radnicki (N) - Spartak (to be played on Sunday) Standings : Obilic 4 4 0 0 10 1 12 OFK Kikinda 4 3 0 1 8 3 9 Sutjeska 4 3 0 1 7 5 9 Loznica 4 2 0 2 7 4 6 OFK Beograd 4 1 3 0 5 4 6 Buducnost (V) 4 2 0 2 4 5 6 Sloboda 4 1 1 2 8 8 4 Spartak 3 1 1 1 3 3 4 Radnicki (N) 3 1 0 2 5 6 3 Zeleznik (B) 4 1 0 3 4 7 3 Rudar 4 1 0 3 1 7 3 Mladost (BJ) 4 0 1 3 2 10 1\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - INCE EXPOSED BY GASCOIGNE\\\\u0027S LATEST PRANK. CHISINAU, Moldova 1996-08-31 England\\\\u0027s irrepressible midfielder Paul Gascoigne was up to his old tricks on Saturday, pulling down his team mate Paul Ince\\\\u0027s trousers in front of an astonished crowd in Moldova. Ince was clambering over a wall at the Republican stadium in Chisinau as Glenn Hoddle\\\\u0027s England players tried to escape heavy rain during an under-21 clash. Gascoigne, whose compulsive practical joking has landed him in trouble in the past, tugged down the Inter Milan player\\\\u0027s trousers in front of a group of press photographers. Hoddle, coaching the side for the first time, declined to comment on the incident. England face Moldova in a World Cup qualifier in the same stadium on Sunday.\"}'\ntext: SOCCER - INCE EXPOSED BY GASCOIGNE'S LATEST PRANK. CHISINAU, Moldova 1996-08-31 England's irrepressible midfielder Paul Gascoigne was up to his old tricks on Saturday, pulling down his team mate Paul Ince's trousers in front of an astonished crowd in Moldova. Ince was clambering over a wall at the Republican stadium in Chisinau as Glenn Hoddle's England players tried to escape heavy rain during an under-21 clash. Gascoigne, whose compulsive practical joking has landed him in trouble in the past, tugged down the Inter Milan player's trousers in front of a group of press photographers. Hoddle, coaching the side for the first time, declined to comment on the incident. England face Moldova in a World Cup qualifier in the same stadium on Sunday.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"BASKETBALL - TROFEJ BEOGRAD TOURNAMENT RESULTS. BELGRADE 1996-08-31 Results in the Trofej Beograd 96 international basketball tournament on Saturday : Fifth place : Benetton (Italy) 92 Dinamo (Russia) 81 (halftime 50-28) Third place : Alba (Germany) 75 Red Star (Yugoslavia) 70 (42-41)\"}'\ntext: BASKETBALL - TROFEJ BEOGRAD TOURNAMENT RESULTS. BELGRADE 1996-08-31 Results in the Trofej Beograd 96 international basketball tournament on Saturday : Fifth place : Benetton (Italy) 92 Dinamo (Russia) 81 (halftime 50-28) Third place : Alba (Germany) 75 Red Star (Yugoslavia) 70 (42-41)\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"BASKETBALLSOCCER - TROFEJ BEOGRAD TOURNAMENT RESULTS. BELGRADE 1996-08-31 Results in the Trofej Beograd 96 international basketball tournament on Saturday : Fifth place : Benetton (Italy) 92 Dinamo (Russia) 81 (halftime 50-28) Third place : Alba (Germany) 75 Red Star (Yugoslavia) 70 (42-41)\"}'\ntext: BASKETBALLSOCCER - TROFEJ BEOGRAD TOURNAMENT RESULTS. BELGRADE 1996-08-31 Results in the Trofej Beograd 96 international basketball tournament on Saturday : Fifth place : Benetton (Italy) 92 Dinamo (Russia) 81 (halftime 50-28) Third place : Alba (Germany) 75 Red Star (Yugoslavia) 70 (42-41)\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - ROMANIA BEAT LITHUANIA IN WORLD CUP QUALIFIER. BUCHAREST 1996-08-31 Romania beat Lithuania 3-0 (halftime 1-0) in a World Cup soccer European group 8 qualifier on Saturday. Scorers : Romania - Viorel Moldovan (21st minute), Dan Petrescu (65th), Constantin Galca (77th) Attendence : 9,000\"}'\ntext: SOCCER - ROMANIA BEAT LITHUANIA IN WORLD CUP QUALIFIER. BUCHAREST 1996-08-31 Romania beat Lithuania 3-0 (halftime 1-0) in a World Cup soccer European group 8 qualifier on Saturday. Scorers : Romania - Viorel Moldovan (21st minute), Dan Petrescu (65th), Constantin Galca (77th) Attendence : 9,000\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - ARMENIA AND PORTUGAL DRAW 0-0 IN WORLD CUP QUALIFIER. YEREVAN 1996-08-31 Armenia and Portugal drew 0-0 in a World Cup soccer European group 9 qualifier on Saturday. Attendance : 5,000\"}'\ntext: SOCCER - ARMENIA AND PORTUGAL DRAW 0-0 IN WORLD CUP QUALIFIER. YEREVAN 1996-08-31 Armenia and Portugal drew 0-0 in a World Cup soccer European group 9 qualifier on Saturday. Attendance : 5,000\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - AZERBAIJAN BEAT SWITZERLAND IN WORLD CUP QUALIFIER. BAKU 1996-08-31 Azerbaijan beat Switzerland 1-0 (halftime 1-0) in their World Cup soccer European group three qualifying match on Saturday. Scorer : Vidadi Rzayev (28th) Attendance : 20,000\"}'\ntext: SOCCER - AZERBAIJAN BEAT SWITZERLAND IN WORLD CUP QUALIFIER. BAKU 1996-08-31 Azerbaijan beat Switzerland 1-0 (halftime 1-0) in their World Cup soccer European group three qualifying match on Saturday. Scorer : Vidadi Rzayev (28th) Attendance : 20,000\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"BASKETBALL - BENETTON BEAT DINAMO 92-81. BELGRADE 1996-08-31 Benetton of Italy beat Dinamo of Russia 92-81 (halftime 50-28) in a fifth place play-off in the Trofej Beograd 96 international basketball tournament on Saturday.\"}'\ntext: BASKETBALL - BENETTON BEAT DINAMO 92-81. BELGRADE 1996-08-31 Benetton of Italy beat Dinamo of Russia 92-81 (halftime 50-28) in a fifth place play-off in the Trofej Beograd 96 international basketball tournament on Saturday.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - SWEDEN BEAT LATVIA IN EUROPEAN UNDER-21 QUALIFIER. RIGA 1996-08-31 Sweden beat Latvia 2-0 (halftime 0-0) in a European under-21 soccer championship qualifier on Saturday. Scorers : Joakim Persson 81st minute, Daniel Andersson (89th) Attendance : 300\"}'\ntext: SOCCER - SWEDEN BEAT LATVIA IN EUROPEAN UNDER-21 QUALIFIER. RIGA 1996-08-31 Sweden beat Latvia 2-0 (halftime 0-0) in a European under-21 soccer championship qualifier on Saturday. Scorers : Joakim Persson 81st minute, Daniel Andersson (89th) Attendance : 300\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - BELARUS BEAT ESTONIA IN WORLD CUP QUALIFIER. MINSK 1996-08-31 Belarus beat Estonia 1-0 (halftime 1-0) in a World Cup soccer European group 4 qualifier on Saturday. Scorer : Vladimir Makovsky (35th) Attendance : 6,000\"}'\ntext: SOCCER - BELARUS BEAT ESTONIA IN WORLD CUP QUALIFIER. MINSK 1996-08-31 Belarus beat Estonia 1-0 (halftime 1-0) in a World Cup soccer European group 4 qualifier on Saturday. Scorer : Vladimir Makovsky (35th) Attendance : 6,000\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - ENGLAND BEAT MOLDOVA IN UNDER-21 QUALIFIER. CHISINAU 1996-08-31 England beat Moldova 2-0 (halftime 1-0) in a European Under-21 soccer championship group 2 qualifier on Saturday. Scorers : Bruce Dyer (39th minute), Darren Eadie (53rd) Attendance : 850\"}'\ntext: SOCCER - ENGLAND BEAT MOLDOVA IN UNDER-21 QUALIFIER. CHISINAU 1996-08-31 England beat Moldova 2-0 (halftime 1-0) in a European Under-21 soccer championship group 2 qualifier on Saturday. Scorers : Bruce Dyer (39th minute), Darren Eadie (53rd) Attendance : 850\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SQUASH - HILL BRANDS WORLD CHAMPION JANSHER A CHEAT. HONG KONG 1996-08-31 Controversial Australian Anthony Hill called Jansher Khan a cheat during his acrimonious defeat by the world number one in the Hong Kong Open semifinals on Saturday. The match boiled over when Hill made to walk off court after what he claimed was a game-winning point in the third. When the referee called Jansher\\\\u0027s return good and the decision was accepted by the player, Hill shrieked at the Pakistani : \\\\\"You cheat.\\\\\" \\\\\"He was standing right there and knew the shot was down so I called him a cheat,\\\\\" said Hill, whose squash career has been blighted by fines and suspensions for unacceptable behaviour. \\\\\"He knew it was down and accepted what I called him because of that.\\\\\" Hill won the game on the next point and said later that Jansher was generally honest on court but played by the referee\\\\u0027s decision. The Australian had upset Jansher\\\\u0027s rhythm with his mixture of gamesmanship and fluent stroke-making but eventually succumbed 15-7 17-15 14-15 15-8. \\\\\"I changed my strategy against him today and had him rattled,\\\\\" he added. He is not as fit as he used to be be but was too good for me in the end. I shook him a bit but he will come out next time and be a better player for it-- that\\\\u0027s Jansher. \\\\\"Jansher said that he was disturbed to be called a cheat.\\\\\" What he did was bad for squash, bad for the crowd and bad for the sponsors. \\\\\"We are trying to build up squash like tennis and players should not say things like that.\\\\\" I think the Professional Squash Association should look into this matter and deal with it properly. I am not calling for him to be banned but they have to take some action. \\\\\"Jansher, bidding for an eighth Hong Kong Open title, plays second-seeded Australian Rodney Eyles in the final. Eyles played his best squash of the tournament to beat fourth-seeded Peter Nicol of Scotland 15-10 8-15 15-10 15-4. Eyles, who defeated Jansher in the semifinals of the Portuguese Open in 1993, said that he would like to win for the good of the game.\\\\\" I have nothing against Jansher but it will be great if I could beat him, \\\\\"said Eyles.\\\\\" My biggest problem against Jansher is concentration. I want to beat him so badly that I just cannot get it together. \\\\\"\"}'\ntext: SQUASH - HILL BRANDS WORLD CHAMPION JANSHER A CHEAT. HONG KONG 1996-08-31 Controversial Australian Anthony Hill called Jansher Khan a cheat during his acrimonious defeat by the world number one in the Hong Kong Open semifinals on Saturday. The match boiled over when Hill made to walk off court after what he claimed was a game-winning point in the third. When the referee called Jansher's return good and the decision was accepted by the player, Hill shrieked at the Pakistani : \"You cheat.\" \"He was standing right there and knew the shot was down so I called him a cheat,\" said Hill, whose squash career has been blighted by fines and suspensions for unacceptable behaviour. \"He knew it was down and accepted what I called him because of that.\" Hill won the game on the next point and said later that Jansher was generally honest on court but played by the referee's decision. The Australian had upset Jansher's rhythm with his mixture of gamesmanship and fluent stroke-making but eventually succumbed 15-7 17-15 14-15 15-8. \"I changed my strategy against him today and had him rattled,\" he added. He is not as fit as he used to be be but was too good for me in the end. I shook him a bit but he will come out next time and be a better player for it-- that's Jansher. \"Jansher said that he was disturbed to be called a cheat.\" What he did was bad for squash, bad for the crowd and bad for the sponsors. \"We are trying to build up squash like tennis and players should not say things like that.\" I think the Professional Squash Association should look into this matter and deal with it properly. I am not calling for him to be banned but they have to take some action. \"Jansher, bidding for an eighth Hong Kong Open title, plays second-seeded Australian Rodney Eyles in the final. Eyles played his best squash of the tournament to beat fourth-seeded Peter Nicol of Scotland 15-10 8-15 15-10 15-4. Eyles, who defeated Jansher in the semifinals of the Portuguese Open in 1993, said that he would like to win for the good of the game.\" I have nothing against Jansher but it will be great if I could beat him, \"said Eyles.\" My biggest problem against Jansher is concentration. I want to beat him so badly that I just cannot get it together. \"\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SQUASH - HONG KONG OPEN SEMIFINAL RESULTS. HONG KONG 1996-08-31 Semifinal results in the Hong Kong Open on Saturday (prefix number denotes seeding) : 1 - Jansher Khan (Pakistan) beat Anthony Hill (Australia) 15-7 17-15 14-15 15-8 2 - Rodney Eyles (Australia) beat 4 - Peter Nicol (Scotland) 15-10 8-15 15-10 15-4\"}'\ntext: SQUASH - HONG KONG OPEN SEMIFINAL RESULTS. HONG KONG 1996-08-31 Semifinal results in the Hong Kong Open on Saturday (prefix number denotes seeding) : 1 - Jansher Khan (Pakistan) beat Anthony Hill (Australia) 15-7 17-15 14-15 15-8 2 - Rodney Eyles (Australia) beat 4 - Peter Nicol (Scotland) 15-10 8-15 15-10 15-4\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"GOLF - PARNEVIK TAKES ONE-SHOT LEAD AT GREATER MILWAUKEE OPEN. MILWAUKEE, Wisconsin 1996-08-31 Jesper Parnevik of Sweden fired a course record-tying eight-under-par 63 Saturday to take a one-shot lead into the final round of the $ 1.2 million Greater Milwaukee Open. Parnevik, who is seeking his first PGA Tour victory, moved to 19-under 194 for the tournament. Parnevik tied the 18-hole record set by Loren Roberts in 1994 at the Brown Deer Park Golf Course and also equalled Saturday by Greg Kraft. Nolan Henke, who led by two strokes entering the third round, carded a four-under 67 and was one stroke back at 18-under 195. He is striving for his fourth career PGA Tour victory and first since the 1993 BellSouth Classic. Tiger Woods, who made the cut in his first tournament as a professional, shot a two-over-par 73 and was four under for the tournament. The 20-year-old Woods, who turned professional Tuesday after winning an unprecedented third successive U.S. Amateur Championship, struggled on the front nine, bogeying the first and seventh holes and double-bogeying the par-four, 359-yard ninth hole. After bogeying the 10th hole to move to four-over for the round, he rallied for birdies on 15 and 18. After Parnevik started off his round by parring the first hole and bogeying the second, the Swede birdied six of the next seven holes. Parnevik continued to storm through the course, birdying three holes on the back nine, including two from 12 feet out on the 15th and 17th holes. \\\\\"I ca n\\\\u0027t remember when I\\\\u0027ve putted this well,\\\\\" said Parnevik. \\\\\"I was disappointed when a 12-footer did n\\\\u0027t go in.\\\\\" My game feels very good. I\\\\u0027ve been fading my driver but today whenever I set up for a fade it went straight. Whenever everyone\\\\u0027s making birdies, you never want to be even par. \\\\\"Henke had a bogey-free round and birdied four holes, including a 45-footer on the par-three 215-yard seventh hole and one from 12 feet out on the 12th hole.\\\\\" I did n\\\\u0027t hit it very well today, \\\\\"said Henke.\\\\\" Jasper blew right by me. Once he did, I knew I had to make birdies just to keep up. I made some really good pars. I basically picked up where I left off yesterday afternoon. \\\\\"Right now, I ca n\\\\u0027t put my finger on what\\\\u0027s wrong.\\\\\" Bob Estes shot a 67 for sole possession of third place at 15-under. Steve Stricker, who tied for second at last week\\\\u0027s World Series of Golf, and Stuart Appleby were both five shots off the lead at 14 under. Duffy Waldorf, who also tied for second at the World Series of Golf, carded a 70 to lead a group of six golfers at 13 under, including Kraft. The top four on the PGA Tour money list all skipped the tournament.\"}'\ntext: GOLF - PARNEVIK TAKES ONE-SHOT LEAD AT GREATER MILWAUKEE OPEN. MILWAUKEE, Wisconsin 1996-08-31 Jesper Parnevik of Sweden fired a course record-tying eight-under-par 63 Saturday to take a one-shot lead into the final round of the $ 1.2 million Greater Milwaukee Open. Parnevik, who is seeking his first PGA Tour victory, moved to 19-under 194 for the tournament. Parnevik tied the 18-hole record set by Loren Roberts in 1994 at the Brown Deer Park Golf Course and also equalled Saturday by Greg Kraft. Nolan Henke, who led by two strokes entering the third round, carded a four-under 67 and was one stroke back at 18-under 195. He is striving for his fourth career PGA Tour victory and first since the 1993 BellSouth Classic. Tiger Woods, who made the cut in his first tournament as a professional, shot a two-over-par 73 and was four under for the tournament. The 20-year-old Woods, who turned professional Tuesday after winning an unprecedented third successive U.S. Amateur Championship, struggled on the front nine, bogeying the first and seventh holes and double-bogeying the par-four, 359-yard ninth hole. After bogeying the 10th hole to move to four-over for the round, he rallied for birdies on 15 and 18. After Parnevik started off his round by parring the first hole and bogeying the second, the Swede birdied six of the next seven holes. Parnevik continued to storm through the course, birdying three holes on the back nine, including two from 12 feet out on the 15th and 17th holes. \"I ca n't remember when I've putted this well,\" said Parnevik. \"I was disappointed when a 12-footer did n't go in.\" My game feels very good. I've been fading my driver but today whenever I set up for a fade it went straight. Whenever everyone's making birdies, you never want to be even par. \"Henke had a bogey-free round and birdied four holes, including a 45-footer on the par-three 215-yard seventh hole and one from 12 feet out on the 12th hole.\" I did n't hit it very well today, \"said Henke.\" Jasper blew right by me. Once he did, I knew I had to make birdies just to keep up. I made some really good pars. I basically picked up where I left off yesterday afternoon. \"Right now, I ca n't put my finger on what's wrong.\" Bob Estes shot a 67 for sole possession of third place at 15-under. Steve Stricker, who tied for second at last week's World Series of Golf, and Stuart Appleby were both five shots off the lead at 14 under. Duffy Waldorf, who also tied for second at the World Series of Golf, carded a 70 to lead a group of six golfers at 13 under, including Kraft. The top four on the PGA Tour money list all skipped the tournament.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"TENNIS - SPAIN, U.S. TEAMS OPEN ON ROAD FOR 1997 FED CUP. Richard Finn NEW YORK 1996-08-31 This year\\\\u0027s Fed Cup finalists-- defending champion Spain and the United States-- will hit the road to open the 1997 women\\\\u0027s international team competition, based on the draw conducted Saturday at the U.S. Open. Spain travels to Belgium, while the U.S. team heads to the Netherlands for first-round matches March 1-2. The other two first-round ties will pit hosts Germany against the Czech Republic and visiting France against Japan. The semifinals are July 19-20, and the final September 27- 28. Life on the road this year did not slow the Americans, who will try to avenge their 3-2 defeat in the final last year when they host Spain on September 28-29 in Atlantic City. \\\\\"Last year we stood on the court after we had lost and we put out hands together and made it our committment to bring back the Cup,\\\\\" U.S. captain Billie Jean King said at the draw. \\\\\"That is our sole goal.\\\\\" The United States edged Austria in Salzburg 3-2 in the opening round in April, and then blanked Japan 5-0 in Nagoya last month in the semifinals. The victory against Japan marked the Fed Cup debut of Monica Seles, who became a naturalised U.S. citizen in 1994. Seles easily won both her singles matches and King is counting on the co-world number one to lead the team again. \\\\\"I told Monica we need her if we want to win,\\\\\" King said. Seles\\\\u0027s sore left shoulder and a wrist injury to Fed Cup veteran Mary Joe Fernandez have forced King to take a wait and see attitude regarding her squad for the best-of-five match. Fernandez was forced to withdraw from the U.S. Open. \\\\\"We will wait until the last minute so we check with everybody and their injuries,\\\\\" said King. \\\\\"What we like would be Seles, (Olympic champion Lindsay) Davenport and Mary Joe Fernandez.\\\\\" If she can get that threesome together, King will feel good about her chances against the Spain\\\\u0027s formidable duo of Arantxa Sanchez Vicario and Conchita Martinez. \\\\\"To be a great coach you have to have the right horses and I got the right horses,\\\\\" said King.\"}'\ntext: TENNIS - SPAIN, U.S. TEAMS OPEN ON ROAD FOR 1997 FED CUP. Richard Finn NEW YORK 1996-08-31 This year's Fed Cup finalists-- defending champion Spain and the United States-- will hit the road to open the 1997 women's international team competition, based on the draw conducted Saturday at the U.S. Open. Spain travels to Belgium, while the U.S. team heads to the Netherlands for first-round matches March 1-2. The other two first-round ties will pit hosts Germany against the Czech Republic and visiting France against Japan. The semifinals are July 19-20, and the final September 27- 28. Life on the road this year did not slow the Americans, who will try to avenge their 3-2 defeat in the final last year when they host Spain on September 28-29 in Atlantic City. \"Last year we stood on the court after we had lost and we put out hands together and made it our committment to bring back the Cup,\" U.S. captain Billie Jean King said at the draw. \"That is our sole goal.\" The United States edged Austria in Salzburg 3-2 in the opening round in April, and then blanked Japan 5-0 in Nagoya last month in the semifinals. The victory against Japan marked the Fed Cup debut of Monica Seles, who became a naturalised U.S. citizen in 1994. Seles easily won both her singles matches and King is counting on the co-world number one to lead the team again. \"I told Monica we need her if we want to win,\" King said. Seles's sore left shoulder and a wrist injury to Fed Cup veteran Mary Joe Fernandez have forced King to take a wait and see attitude regarding her squad for the best-of-five match. Fernandez was forced to withdraw from the U.S. Open. \"We will wait until the last minute so we check with everybody and their injuries,\" said King. \"What we like would be Seles, (Olympic champion Lindsay) Davenport and Mary Joe Fernandez.\" If she can get that threesome together, King will feel good about her chances against the Spain's formidable duo of Arantxa Sanchez Vicario and Conchita Martinez. \"To be a great coach you have to have the right horses and I got the right horses,\" said King.\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"TENNIS - SATURDAY\\\\u0027S RESULTS FROM THE U.S. OPEN. NEW YORK 1996-08-31 Results from the U.S. Open Tennis Championships at the National Tennis Centre on Saturday (prefix number denotes seeding) : Women\\\\u0027s singles, third round 1 - Steffi Graf (Germany) beat Natasha Zvereva (Belarus) 6-4 6-2 16 - Martina Hingis (Switzerland) beat Naoko Kijimuta (Japan) 6-2 6-2 Judith Wiesner (Austria) beat Petra Langrova (Czech Republic) 6 - 2 6-0 Men\\\\u0027s singles, third round 13 - Thomas Enqvist (Sweden) beat Pablo Campana (Ecuador) 6-4 6-4 6-2\"}'\ntext: TENNIS - SATURDAY'S RESULTS FROM THE U.S. OPEN. NEW YORK 1996-08-31 Results from the U.S. Open Tennis Championships at the National Tennis Centre on Saturday (prefix number denotes seeding) : Women's singles, third round 1 - Steffi Graf (Germany) beat Natasha Zvereva (Belarus) 6-4 6-2 16 - Martina Hingis (Switzerland) beat Naoko Kijimuta (Japan) 6-2 6-2 Judith Wiesner (Austria) beat Petra Langrova (Czech Republic) 6 - 2 6-0 Men's singles, third round 13 - Thomas Enqvist (Sweden) beat Pablo Campana (Ecuador) 6-4 6-4 6-2\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"TENNIS - DRAW FOR 1997 FED CUP WOMEN\\\\u0027S TEAM TOURNAMENT. NEW YORK 1996-08-31 Draw for the women\\\\u0027s 1997 Fed Cup team tennis championships, as conducted at the U.S. Open on Saturday : World Group I, first round (March 1-2) United States at Netherlands Czech Republic at Germany France at Japan Spain at Belgium (semifinals July 19-20, and finals September 27-28) World Group II, first round (March 1-2) Austria at Croatia Switzerland at Slovak Republic Argentina at South Korea Australia at South Africa\"}'\ntext: TENNIS - DRAW FOR 1997 FED CUP WOMEN'S TEAM TOURNAMENT. NEW YORK 1996-08-31 Draw for the women's 1997 Fed Cup team tennis championships, as conducted at the U.S. Open on Saturday : World Group I, first round (March 1-2) United States at Netherlands Czech Republic at Germany France at Japan Spain at Belgium (semifinals July 19-20, and finals September 27-28) World Group II, first round (March 1-2) Austria at Croatia Switzerland at Slovak Republic Argentina at South Korea Australia at South Africa\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"SOCCER - U.S. BEAT EL SALVADOR 3-1. LOS ANGELES 1996-08-30 The United States beat El Salvador 3-1 (halftime 1-0) in an international soccer friendly on Friday. Scorers : U.S. - Joe-Max Moore (3rd minute, 88th on penalty kick), Eric Wynalda (61st) El Salvador - Luis Lazo (61st) Attendance - 18,661\"}'\ntext: SOCCER - U.S. BEAT EL SALVADOR 3-1. LOS ANGELES 1996-08-30 The United States beat El Salvador 3-1 (halftime 1-0) in an international soccer friendly on Friday. Scorers : U.S. - Joe-Max Moore (3rd minute, 88th on penalty kick), Eric Wynalda (61st) El Salvador - Luis Lazo (61st) Attendance - 18,661\nspans: []\n127.0.0.1 - - [09/Feb/2018 12:05:15] \"POST / HTTP/1.1\" 200 -\nreceived post body: b'{\"spans\":[],\"text\":\"BASEBALL - MAJOR LEAGUE STANDINGS AFTER FRIDAY\\\\u0027S GAMES. NEW YORK 1996-08-31 Major League Baseball standings after games played on Friday (tabulate under won, lost, winning percentage and games behind) : AMERICAN LEAGUE EASTERN DIVISION W L PCT GB NEW YORK 75 59.560 - BALTIMORE 71 63.530 4 BOSTON 69 66.511 6 1/2 TORONTO 63 72.467 12 1/2 DETROIT 49 86.363 26 1/2 CENTRAL DIVISION CLEVELAND 80 54.597 - CHICAGO 72 64.529 9 MINNESOTA 67 68.496 13 1/2 MILWAUKEE 65 71.478 16 KANSAS CITY 61 75.449 20 WESTERN DIVISION TEXAS 76 58.567 - SEATTLE 70 64.522 6 OAKLAND 65 72.474 12 1/2 CALIFORNIA 62 73.459 14 1/2 SATURDAY, AUGUST 31 SCHEDULE KANSAS CITY AT DETROIT BALTIMORE AT SEATTLE CHICAGO AT TORONTO MINNESOTA AT MILWAUKEE CLEVELAND AT TEXAS BOSTON AT OAKLAND NEW YORK AT CALIFORNIA NATIONAL LEAGUE EASTERN DIVISION W L PCT GB ATLANTA 84 50.627 - MONTREAL 71 62.534 12 1/2 FLORIDA 65 70.481 19 1/2 NEW YORK 59 76.437 25 1/2 PHILADELPHIA 54 81.400 30 1/2 CENTRAL DIVISION HOUSTON 73 63.537 - ST LOUIS 70 65.519 2 1/2 CHICAGO 66 67.496 5 1/2 CINCINNATI 66 68.493 6 PITTSBURGH 56 78.418 16 WESTERN DIVISION SAN DIEGO 76 60.559 - LOS ANGELES 73 61.545 2 COLORADO 70 66.515 6 SAN FRANCISCO 58 74.439 16 SATURDAY, AUGUST 31 SCHEDULE ATLANTA AT CHICAGO HOUSTON AT PITTSBURGH SAN FRANCISCO AT NEW YORK FLORIDA AT CINCINNATI LOS ANGELES AT PHILADELPHIA SAN DIEGO AT MONTREAL COLORADO AT ST LOUIS\"}'\ntext: BASEBALL - MAJOR LEAGUE STANDINGS AFTER FRIDAY'S GAMES. NEW YORK 1996-08-31 Major League Baseball standings after games played on Friday (tabulate under won, lost, winning percentage and games behind) : AMERICAN LEAGUE EASTERN DIVISION W L PCT GB NEW YORK 75 59.560 - BALTIMORE 71 63.530 4 BOSTON 69 66.511 6 1/2 TORONTO 63 72.467 12 1/2 DETROIT 49 86.363 26 1/2 CENTRAL DIVISION CLEVELAND 80 54.597 - CHICAGO 72 64.529 9 MINNESOTA 67 68.496 13 1/2 MILWAUKEE 65 71.478 16 KANSAS CITY 61 75.449 20 WESTERN DIVISION TEXAS 76 58.567 - SEATTLE 70 64.522 6 OAKLAND 65 72.474 12 1/2 CALIFORNIA 62 73.459 14 1/2 SATURDAY, AUGUST 31 SCHEDULE KANSAS CITY AT DETROIT BALTIMORE AT SEATTLE CHICAGO AT TORONTO MINNESOTA AT MILWAUKEE CLEVELAND AT TEXAS BOSTON AT OAKLAND NEW YORK AT CALIFORNIA NATIONAL LEAGUE EASTERN DIVISION W L PCT GB ATLANTA 84 50.627 - MONTREAL 71 62.534 12 1/2 FLORIDA 65 70.481 19 1/2 NEW YORK 59 76.437 25 1/2 PHILADELPHIA 54 81.400 30 1/2 CENTRAL DIVISION HOUSTON 73 63.537 - ST LOUIS 70 65.519 2 1/2 CHICAGO 66 67.496 5 1/2 CINCINNATI 66 68.493 6 PITTSBURGH 56 78.418 16 WESTERN DIVISION SAN DIEGO 76 60.559 - LOS ANGELES 73 61.545 2 COLORADO 70 66.515 6 SAN FRANCISCO 58 74.439 16 SATURDAY, AUGUST 31 SCHEDULE ATLANTA AT CHICAGO HOUSTON AT PITTSBURGH SAN FRANCISCO AT NEW YORK FLORIDA AT CINCINNATI LOS ANGELES AT PHILADELPHIA SAN DIEGO AT MONTREAL COLORADO AT ST LOUIS\nspans: []\n\n\n\n\n" }, { "path": "gerbil-SpotWrapNifWS4Test/pom.xml", "content": "\n\t4.0.0\n\torg.aksw\n\tgerbil-spotWrapNifWS4Test\n\t0.0.1-SNAPSHOT\n\tNIF Webservice for testing the General Entity Annotator Benchmark\n\tThis branch of the GERBIL project is simply for testing purposes.\n\n\t\n\t\t1.7\n\t\t1.7.5\n\t\tUTF-8\n\t\n\n\t\n\t\t\n\t\t\n\t\t\tlocal repository\n\t\t\tfile://${project.basedir}/repository\n\t\t\n\t\t\n\t\t\tmaven.aksw.internal\n\t\t\tUniversity Leipzig, AKSW Maven2 Repository\n\t\t\thttp://maven.aksw.org/repository/internal\n\t\t\n\t\t\n\t\t\tmaven.aksw.snapshots\n\t\t\tUniversity Leipzig, AKSW Maven2 Repository\n\t\t\thttp://maven.aksw.org/repository/snapshots\n\t\t\n\t\n\n\t\n\t\t\n\t\t\torg.aksw\n\t\t\tgerbil.nif.transfer\n\t\t\t1.1.0-SNAPSHOT\n\t\t\n\t\t\n\t\t\n\t\t\torg.restlet\n\t\t\torg.restlet\n\t\t\t2.2.1\n\t\t\n\t\t\n\t\t\torg.restlet\n\t\t\torg.restlet.ext.servlet\n\t\t\t2.2.1\n\t\t\n\t\t\n\t\t\n\t\t\torg.apache.httpcomponents\n\t\t\thttpclient\n\t\t\t4.3.5\n\t\t\n\t\t\n\t\t\n\t\t\tcommons-io\n\t\t\tcommons-io\n\t\t\t2.7\n\t\t\n\t\t\n\t\t\n\t\t\tcom.googlecode.json-simple\n\t\t\tjson-simple\n\t\t\t1.1.1\n\t\t\n \n \n com.carrotsearch\n hppc\n 0.5.3\n \n\t\t\n\t\t\n\t\t\torg.apache.jena\n\t\t\tjena-core\n\t\t\t2.11.1\n\t\t\n\t\t\n\t\t\torg.apache.jena\n\t\t\tjena-arq\n\t\t\t2.11.1\n\t\t\n\n\t\t\n\t\t\n\t\t\torg.slf4j\n\t\t\tslf4j-api\n\t\t\t${slf4j.version}\n\t\t\n\t\t\n\t\t\n\t\t\torg.slf4j\n\t\t\tslf4j-log4j12\n\t\t\t${slf4j.version}\n\t\t\n\n\t\t\n\t\t\n \t\t\tcom.google.code.gson\n \t\t\tgson\n \t\t\t2.3.1\n\t\t\n\n\t\n\n\t\n\t\t\n\t\t\t\n\t\t\t\torg.apache.maven.plugins\n\t\t\t\tmaven-compiler-plugin\n\t\t\t\t3.1\n\t\t\t\t\n\t\t\t\t\t${java.version}\n\t\t\t\t\t${java.version}\n\t\t\t\t\n\t\t\t\n\t\t\n\t\n\twar\n\n" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/aksw/gerbil.nif.transfer/1.1.0-SNAPSHOT/gerbil.nif.transfer-1.1.0-SNAPSHOT-javadoc.jar.md5", "content": "d8cb94795d15a968a097109b2739f6f5" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/aksw/gerbil.nif.transfer/1.1.0-SNAPSHOT/gerbil.nif.transfer-1.1.0-SNAPSHOT-javadoc.jar.sha1", "content": "0b44e6e4b45bdbaf175e6cef700fef1f971542b9" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/aksw/gerbil.nif.transfer/1.1.0-SNAPSHOT/gerbil.nif.transfer-1.1.0-SNAPSHOT-sources.jar.md5", "content": "a622142dc8912285288b320b5bf51a3b" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/aksw/gerbil.nif.transfer/1.1.0-SNAPSHOT/gerbil.nif.transfer-1.1.0-SNAPSHOT-sources.jar.sha1", "content": "ea177f3f37dfe9ab6ab032b4689ed3f4e90c69c6" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/aksw/gerbil.nif.transfer/1.1.0-SNAPSHOT/gerbil.nif.transfer-1.1.0-SNAPSHOT.jar.md5", "content": "2ba688c9e81ac41ebf8948bd6c2ec228" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/aksw/gerbil.nif.transfer/1.1.0-SNAPSHOT/gerbil.nif.transfer-1.1.0-SNAPSHOT.jar.sha1", "content": "c9d243a0a93e95804bbeb740b4f756f3eb4b9a71" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/aksw/gerbil.nif.transfer/1.1.0-SNAPSHOT/gerbil.nif.transfer-1.1.0-SNAPSHOT.pom", "content": "\n\n\t4.0.0\n\torg.aksw\n\tgerbil.nif.transfer\n\t1.1.0-SNAPSHOT\n\tNIF transfer library for the General Entity Annotator Benchmark\n\tThis project contains classes for transferring documents using NIF.\n\n\t\n\t\t1.7\n\t\t1.7.5\n\t\t4.11\n\t\t2.13.0\n\t\tUTF-8\n\t\n\n\t\n\t\t\n\t\t\tmaven.aksw.internal\n\t\t\tUniversity Leipzig, AKSW Maven2 Repository\n\t\t\thttp://maven.aksw.org/repository/internal\n\t\t\n\t\t\n\t\t\tmaven.aksw.snapshots\n\t\t\tUniversity Leipzig, AKSW Maven2 Repository\n\t\t\thttp://maven.aksw.org/repository/snapshots\n\t\t\n\t\n\n\t\n\t\t\n\t\t\n\t\t\torg.apache.jena\n\t\t\tjena-core\n\t\t\t${jena.version}\n\t\t\n\t\t\n\t\t\torg.apache.jena\n\t\t\tjena-arq\n\t\t\t${jena.version}\n\t\t\n\n\t\t\n\t\t\tcommons-io\n\t\t\tcommons-io\n\t\t\t2.4\n\t\t\n\n\t\t\n\t\t\n\t\t\torg.slf4j\n\t\t\tslf4j-api\n\t\t\t${slf4j.version}\n\t\t\n\t\t\n\t\t\n\t\t\torg.slf4j\n\t\t\tslf4j-log4j12\n\t\t\t${slf4j.version}\n\t\t\n\n\t\t\n\t\t\n\t\t\tjunit\n\t\t\tjunit\n\t\t\t${junit.version}\n\t\t\ttest\n\t\t\n\t\n\n\t\n\t\t\n\t\t\t\n\t\t\t\torg.apache.maven.plugins\n\t\t\t\tmaven-compiler-plugin\n\t\t\t\t3.1\n\t\t\t\t\n\t\t\t\t\t${java.version}\n\t\t\t\t\t${java.version}\n\t\t\t\t\n\t\t\t\n\t\t\t\n\t\t\t\n\t\t\t\torg.apache.maven.plugins\n\t\t\t\tmaven-javadoc-plugin\n\t\t\t\t2.10.1\n\t\t\t\t\n\t\t\t\t\tprivate\n\t\t\t\t\ttrue\n\t\t\t\t\n\t\t\t\t\n\t\t\t\t\t\n\t\t\t\t\t\tpackage\n\t\t\t\t\t\t\n\t\t\t\t\t\t\tjar\n\t\t\t\t\t\t\n\t\t\t\t\t\n\t\t\t\t\n\t\t\t\n\t\t\t\n\t\t\t\n\t\t\t\torg.apache.maven.plugins\n\t\t\t\tmaven-source-plugin\n\t\t\t\t2.4\n\t\t\t\t\n\t\t\t\t\t\n\t\t\t\t\t\tattach-sources\n\t\t\t\t\t\t\n\t\t\t\t\t\t\tjar\n\t\t\t\t\t\t\n\t\t\t\t\t\n\t\t\t\t\n\t\t\t\n\t\t\t\n\t\t\t\n\t\t\t\tcom.mycila\n\t\t\t\tlicense-maven-plugin\n\t\t\t\t2.6\n\t\t\t\t\n\t\t\t\t\t
src/main/resources/license_template.txt
\n\t\t\t\t\t\n\t\t\t\t\t\tAgile Knowledge Engineering and Semantic Web (AKSW)\n\t\t\t\t\t\t${project.inceptionYear}\n\t\t\t\t\t\tusbeck@informatik.uni-leipzig.de\n\t\t\t\t\t\n\t\t\t\t\t\n\t\t\t\t\t\t**/README\n\t\t\t\t\t\t**/LICENSE\n\t\t\t\t\t\t**/install_to_gerbil.sh\n\t\t\t\t\t\tdiagrams/**\n\t\t\t\t\t\trepository/**\n\t\t\t\t\t\tgerbil_data/**\n\t\t\t\t\t\tsrc/test/resources/**\n\t\t\t\t\t\tsrc/main/resources/**\n\t\t\t\t\t\n\t\t\t\t\n\t\t\t\n\t\t\n\t\n" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/aksw/gerbil.nif.transfer/1.1.0-SNAPSHOT/gerbil.nif.transfer-1.1.0-SNAPSHOT.pom.md5", "content": "bd93e677c0a2041a4e2d205ff00eaf30" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/aksw/gerbil.nif.transfer/1.1.0-SNAPSHOT/gerbil.nif.transfer-1.1.0-SNAPSHOT.pom.sha1", "content": "303d1b06abe9fa32d1c3470e977882590202a475" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/aksw/gerbil.nif.transfer/1.1.0-SNAPSHOT/maven-metadata-local.xml", "content": "\n\n org.aksw\n gerbil.nif.transfer\n 1.1.0-SNAPSHOT\n \n \n true\n \n 20150520153444\n \n \n javadoc\n jar\n 1.1.0-SNAPSHOT\n 20150520153444\n \n \n sources\n jar\n 1.1.0-SNAPSHOT\n 20150520153444\n \n \n jar\n 1.1.0-SNAPSHOT\n 20150520153444\n \n \n pom\n 1.1.0-SNAPSHOT\n 20150520153444\n \n \n \n\n" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/aksw/gerbil.nif.transfer/1.1.0-SNAPSHOT/maven-metadata-local.xml.md5", "content": "8d28899dd51b52fc95bce996f12f0fb1" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/aksw/gerbil.nif.transfer/1.1.0-SNAPSHOT/maven-metadata-local.xml.sha1", "content": "8cdcb67c4c86e693884264f9cb92f04ae0b9b788" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/aksw/gerbil.nif.transfer/maven-metadata-local.xml", "content": "\n\n org.aksw\n gerbil.nif.transfer\n \n 1.2.0\n \n 1.1.0-SNAPSHOT\n 1.2.0-SNAPSHOT\n 1.2.0\n \n 20151104111943\n \n\n" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/aksw/gerbil.nif.transfer/maven-metadata-local.xml.md5", "content": "0b6e4e09b60735f465f60ce53de21b22" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/aksw/gerbil.nif.transfer/maven-metadata-local.xml.sha1", "content": "af04aebdca0e9affc3dd1c4f714097b4a8f4b8b0" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/restlet/org.restlet/2.2.1/org.restlet-2.2.1.jar.md5", "content": "f0a69fc7748a4859f185869a803fd31d" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/restlet/org.restlet/2.2.1/org.restlet-2.2.1.jar.sha1", "content": "e5ef2d645af76d7787f4fa5c360d4916864795f3" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/restlet/org.restlet/2.2.1/org.restlet-2.2.1.pom", "content": "\n\n\n 4.0.0\n org.restlet\n org.restlet\n 2.2.1\n POM was created from install:install-file\n\n" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/restlet/org.restlet/2.2.1/org.restlet-2.2.1.pom.md5", "content": "fa265a18a02fe6b1afee5d59ef09f9b2" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/restlet/org.restlet/2.2.1/org.restlet-2.2.1.pom.sha1", "content": "bf2f49d20b9b7708c5fdf548d9d5403e95e5824d" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/restlet/org.restlet/maven-metadata-local.xml", "content": "\n\n\n org.restlet\n org.restlet\n 2.2.1\n \n \n 2.2.1\n \n 20140804130630\n \n\n" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/restlet/org.restlet/maven-metadata-local.xml.md5", "content": "08a59985f1be54d59c6a67bc52f71769" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/restlet/org.restlet/maven-metadata-local.xml.sha1", "content": "b7656b95e6a29dd775f74fdbf58044f9c7122cbd" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/restlet/org.restlet.ext.servlet/2.2.1/org.restlet.ext.servlet-2.2.1.jar.md5", "content": "83e5a0a3c182e3a0b54bb1c9ea89a4ba" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/restlet/org.restlet.ext.servlet/2.2.1/org.restlet.ext.servlet-2.2.1.jar.sha1", "content": "4f1f9af4af0c187ffdcbad3f7b20df3158a2b860" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/restlet/org.restlet.ext.servlet/2.2.1/org.restlet.ext.servlet-2.2.1.pom", "content": "\n\n\n 4.0.0\n org.restlet\n org.restlet.ext.servlet\n 2.2.1\n POM was created from install:install-file\n\n" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/restlet/org.restlet.ext.servlet/2.2.1/org.restlet.ext.servlet-2.2.1.pom.md5", "content": "119436ba0b07f90b0ee6bed3bb5a38c9" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/restlet/org.restlet.ext.servlet/2.2.1/org.restlet.ext.servlet-2.2.1.pom.sha1", "content": "20160a0a755fcbd86a1a3a0ef6092247a67b0b56" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/restlet/org.restlet.ext.servlet/maven-metadata-local.xml", "content": "\n\n\n org.restlet\n org.restlet.ext.servlet\n 2.2.1\n \n \n 2.2.1\n \n 20140804132542\n \n\n" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/restlet/org.restlet.ext.servlet/maven-metadata-local.xml.md5", "content": "b25961e20f84c3eb27a617f9698731fd" }, { "path": "gerbil-SpotWrapNifWS4Test/repository/org/restlet/org.restlet.ext.servlet/maven-metadata-local.xml.sha1", "content": "27494e8d89841244a60eab9037402baf55f2eb35" }, { "path": "gerbil-SpotWrapNifWS4Test/src/main/java/org/aksw/gerbil/ws4test/EDResource.java", "content": "\npackage org.aksw.gerbil.ws4test;\n\nimport java.io.IOException;\nimport java.io.InputStream;\nimport java.io.Reader;\nimport java.util.ArrayList;\nimport java.util.List;\n\nimport org.aksw.gerbil.transfer.nif.Document;\nimport org.aksw.gerbil.transfer.nif.Marking;\nimport org.aksw.gerbil.transfer.nif.Span;\nimport org.aksw.gerbil.transfer.nif.TurtleNIFDocumentCreator;\nimport org.aksw.gerbil.transfer.nif.TurtleNIFDocumentParser;\nimport org.aksw.gerbil.transfer.nif.data.NamedEntity;\nimport org.restlet.representation.Representation;\nimport org.restlet.resource.Post;\nimport org.restlet.resource.ServerResource;\nimport org.slf4j.Logger;\nimport org.slf4j.LoggerFactory;\nimport org.apache.commons.io.IOUtils;\nimport org.apache.http.HttpEntity;\nimport org.apache.http.StatusLine;\nimport org.apache.http.client.HttpClient;\nimport org.apache.http.client.methods.CloseableHttpResponse;\nimport org.apache.http.client.methods.HttpPost;\nimport org.apache.http.entity.StringEntity;\nimport org.apache.http.impl.client.HttpClients;\nimport org.apache.http.util.EntityUtils;\n\nimport com.google.gson.Gson;\nimport com.google.gson.GsonBuilder;\nimport com.google.gson.JsonArray;\nimport com.google.gson.JsonElement;\nimport com.google.gson.JsonObject;\nimport com.google.gson.JsonParser;\n\n\n\npublic class EDResource extends ServerResource {\n\n\tprivate static final Logger LOGGER = LoggerFactory.getLogger(EDResource.class);\n\n private String clusterServiceURL = \"http://localhost:5555/\";\n \n private Gson gson = new GsonBuilder().create();\n private HttpClient client = HttpClients.createDefault(); \n \n @Post\n public String accept(Representation request) {\n \tSystem.out.println(\"-------------------------------------------------\");\n Reader inputReader;\n try {\n inputReader = request.getReader();\n } catch (IOException e) {\n System.err.println(\"Exception while reading request.\" + e.getMessage());\n return \"\";\n }\n \n List entities = sendRequestToCluster(\"dummy text\", new ArrayList()); \n return \"dummy response\";\n }\n \n \n public List sendRequestToCluster(String text , List markings) {\n \tList entities = new ArrayList(markings.size());\n \t\n \t//JsonObject out = null;\n \tJsonArray out = null;\n \ttry {\n\t\t\tout = queryJson(text, markings, clusterServiceURL);\n\t\t} catch (IOException e) {\n\t\t\te.printStackTrace();\n\t\t}\n \t\n \t// process the response\n \t/*if (out!= null) {\n \t\tJsonArray outMentionsJson = out.getAsJsonArray(\"annotations\");\n \t\tfor (JsonElement je : outMentionsJson) {\n \t\t\tJsonObject mentionJson = je.getAsJsonObject();\n \t\t\tint start = mentionJson.get(\"start\").getAsInt();\n \t\t\tint length = mentionJson.get(\"length\").getAsInt();\n \t\t\tString entity = mentionJson.get(\"entity\").getAsString();\n \t\t\tentities.add(new NamedEntity(start, length, entity)); \t\t\n \t\t}\n \t}*/\n \tSystem.out.println(\"Json response from NN: \"+out);\n \tif (out!= null) {\n \t\t//JsonArray outMentionsJson = out.getAsJsonArray(\"annotations\");\n \t\tfor (JsonElement je : out) {\n \t\t\tJsonArray mentionJson = je.getAsJsonArray();\n \t\t\tint start = mentionJson.get(0).getAsInt();\n \t\t\tint length = mentionJson.get(1).getAsInt();\n \t\t\tString entity = mentionJson.get(2).getAsString();\n \t\t\tentities.add(new NamedEntity(start, length, \"http://dbpedia.org/resource/\" + entity)); \t\t\n \t\t}\n \t}\n \tSystem.out.println(entities);\n \treturn entities;\n }\n \n \n private JsonArray queryJson(String text, List markings, String url) throws IOException {\n\n JsonObject parameters = new JsonObject();\n\n if (markings != null) {\n JsonArray mentionsJson = new JsonArray();\n for (Marking m : markings) {\n \tSpan sp = (Span) m;\n JsonObject mentionJson = new JsonObject();\n mentionJson.addProperty(\"start\", sp.getStartPosition());\n mentionJson.addProperty(\"length\", sp.getLength());\n mentionsJson.add(mentionJson);\n }\n parameters.add(\"spans\", mentionsJson);\n }\n parameters.addProperty(\"text\", text);\n\n HttpPost request = new HttpPost(url);\n request.addHeader(\"Content-Type\", \"application/json\");\n request.setEntity(new StringEntity(gson.toJson(parameters), \"UTF8\"));\n request.addHeader(\"Accept\", \"application/json\");\n\n CloseableHttpResponse response = (CloseableHttpResponse) client.execute(request);\n InputStream is = null;\n HttpEntity entity = null;\n try {\n StatusLine status = response.getStatusLine();\n if ((status.getStatusCode() < 200) || (status.getStatusCode() >= 300)) {\n entity = response.getEntity();\n System.err.println(\"The response had a wrong status: \\\"\" + status.toString() + \"\\\". Content of response: \\\"\"\n + IOUtils.toString(entity.getContent()) + \"\\\". Returning null.\");\n return null;\n }\n entity = response.getEntity();\n is = entity.getContent();\n //return new JsonParser().parse(IOUtils.toString(is)).getAsJsonObject();\n return new JsonParser().parse(IOUtils.toString(is)).getAsJsonArray();\n } catch (Exception e) {\n \tSystem.err.println(\"Couldn't request annotation for given text. Returning null.\" + e.getMessage());\n } finally {\n IOUtils.closeQuietly(is);\n if (entity != null) {\n EntityUtils.consume(entity);\n }\n if (response != null) {\n response.close();\n }\n }\n return null;\n } \n}\n" }, { "path": "gerbil-SpotWrapNifWS4Test/src/main/java/org/aksw/gerbil/ws4test/LocalIntermediateWebserver.java", "content": "package org.aksw.gerbil.ws4test;\n\nimport java.io.IOException;\nimport java.io.InputStream;\nimport java.io.Reader;\nimport java.util.ArrayList;\nimport java.util.List;\n\nimport org.aksw.gerbil.transfer.nif.Document;\nimport org.aksw.gerbil.transfer.nif.Marking;\nimport org.aksw.gerbil.transfer.nif.Span;\nimport org.aksw.gerbil.transfer.nif.TurtleNIFDocumentCreator;\nimport org.aksw.gerbil.transfer.nif.TurtleNIFDocumentParser;\nimport org.aksw.gerbil.transfer.nif.data.NamedEntity;\nimport org.restlet.representation.Representation;\nimport org.restlet.resource.Post;\nimport org.restlet.resource.ServerResource;\nimport org.apache.commons.io.IOUtils;\nimport org.apache.http.HttpEntity;\nimport org.apache.http.StatusLine;\nimport org.apache.http.client.HttpClient;\nimport org.apache.http.client.methods.CloseableHttpResponse;\nimport org.apache.http.client.methods.HttpPost;\nimport org.apache.http.entity.StringEntity;\nimport org.apache.http.impl.client.HttpClients;\nimport org.apache.http.util.EntityUtils;\n\nimport com.google.gson.Gson;\nimport com.google.gson.GsonBuilder;\nimport com.google.gson.JsonArray;\nimport com.google.gson.JsonElement;\nimport com.google.gson.JsonObject;\nimport com.google.gson.JsonParser;\n\n\n\npublic class LocalIntermediateWebserver extends ServerResource {\n\n private TurtleNIFDocumentParser parser = new TurtleNIFDocumentParser();\n private TurtleNIFDocumentCreator creator = new TurtleNIFDocumentCreator();\n\n private String clusterServiceURL = \"http://localhost:5555/cluster_ed\";\n \n private Gson gson = new GsonBuilder().create();\n private HttpClient client = HttpClients.createDefault(); \n \n @Post\n public String accept(Representation request) {\n \tSystem.out.println(\"-------------------------------------------------\");\n Reader inputReader;\n try {\n inputReader = request.getReader();\n } catch (IOException e) {\n System.err.println(\"Exception while reading request.\" + e.getMessage());\n return \"\";\n }\n Document document;\n try {\n document = parser.getDocumentFromNIFReader(inputReader);\n } catch (Exception e) {\n \tSystem.err.println(\"Exception while reading request.\" + e.getMessage());\n return \"\";\n }\n\n List markings = document.getMarkings();\n String text = document.getText(); \n\n //System.out.println(\"num spots = \" + markings.size());\n List entities = sendRequestToCluster(text, markings); \n //System.out.println(\" Solution size = \" + entities.size()); \n \n // ... this new list is added to the document and the document is\n // send back to GERBIL\n document.setMarkings(entities);\n String nifDocument = creator.getDocumentAsNIFString(document);\n return nifDocument;\n }\n \n \n public List sendRequestToCluster(String text , List markings) {\n \tList entities = new ArrayList(markings.size());\n \t\n \tJsonObject out = null;\n \ttry {\n\t\t\tout = queryJson(text, markings, clusterServiceURL);\n\t\t} catch (IOException e) {\n\t\t\te.printStackTrace();\n\t\t}\n \t\n \t// process the response\n \t/*if (out!= null) {\n \t\tJsonArray outMentionsJson = out.getAsJsonArray(\"annotations\");\n \t\tfor (JsonElement je : outMentionsJson) {\n \t\t\tJsonObject mentionJson = je.getAsJsonObject();\n \t\t\tint start = mentionJson.get(\"start\").getAsInt();\n \t\t\tint length = mentionJson.get(\"length\").getAsInt();\n \t\t\tString entity = mentionJson.get(\"entity\").getAsString();\n \t\t\tentities.add(new NamedEntity(start, length, entity)); \t\t\n \t\t}\n \t}*/\n \t\n \treturn entities;\n }\n \n \n private JsonObject queryJson(String text, List markings, String url) throws IOException {\n\n JsonObject parameters = new JsonObject();\n\n if (markings != null) {\n JsonArray mentionsJson = new JsonArray();\n for (Marking m : markings) {\n \tSpan sp = (Span) m;\n JsonObject mentionJson = new JsonObject();\n mentionJson.addProperty(\"start\", sp.getStartPosition());\n mentionJson.addProperty(\"length\", sp.getLength());\n mentionsJson.add(mentionJson);\n }\n parameters.add(\"spans\", mentionsJson);\n }\n parameters.addProperty(\"text\", text);\n\n HttpPost request = new HttpPost(url);\n request.addHeader(\"Content-Type\", \"application/json\");\n request.setEntity(new StringEntity(gson.toJson(parameters), \"UTF8\"));\n request.addHeader(\"Accept\", \"application/json\");\n\n CloseableHttpResponse response = (CloseableHttpResponse) client.execute(request);\n InputStream is = null;\n HttpEntity entity = null;\n try {\n StatusLine status = response.getStatusLine();\n if ((status.getStatusCode() < 200) || (status.getStatusCode() >= 300)) {\n entity = response.getEntity();\n System.err.println(\"The response had a wrong status: \\\"\" + status.toString() + \"\\\". Content of response: \\\"\"\n + IOUtils.toString(entity.getContent()) + \"\\\". Returning null.\");\n return null;\n }\n entity = response.getEntity();\n is = entity.getContent();\n return new JsonParser().parse(IOUtils.toString(is)).getAsJsonObject();\n } catch (Exception e) {\n \tSystem.err.println(\"Couldn't request annotation for given text. Returning null.\" + e.getMessage());\n } finally {\n IOUtils.closeQuietly(is);\n if (entity != null) {\n EntityUtils.consume(entity);\n }\n if (response != null) {\n response.close();\n }\n }\n return null;\n } \n}\n" }, { "path": "gerbil-SpotWrapNifWS4Test/src/main/java/org/aksw/gerbil/ws4test/MyResource.java", "content": "package org.aksw.gerbil.ws4test;\n\nimport java.io.IOException;\nimport java.io.InputStream;\nimport java.io.Reader;\nimport java.util.ArrayList;\nimport java.util.List;\n\nimport org.aksw.gerbil.transfer.nif.Document;\nimport org.aksw.gerbil.transfer.nif.Marking;\nimport org.aksw.gerbil.transfer.nif.Span;\nimport org.aksw.gerbil.transfer.nif.TurtleNIFDocumentCreator;\nimport org.aksw.gerbil.transfer.nif.TurtleNIFDocumentParser;\nimport org.aksw.gerbil.transfer.nif.data.NamedEntity;\nimport org.restlet.representation.Representation;\nimport org.restlet.resource.Post;\nimport org.restlet.resource.ServerResource;\nimport org.slf4j.Logger;\nimport org.slf4j.LoggerFactory;\nimport org.apache.commons.io.IOUtils;\nimport org.apache.http.HttpEntity;\nimport org.apache.http.StatusLine;\nimport org.apache.http.client.HttpClient;\nimport org.apache.http.client.methods.CloseableHttpResponse;\nimport org.apache.http.client.methods.HttpPost;\nimport org.apache.http.entity.StringEntity;\nimport org.apache.http.impl.client.HttpClients;\nimport org.apache.http.util.EntityUtils;\n\nimport com.google.gson.Gson;\nimport com.google.gson.GsonBuilder;\nimport com.google.gson.JsonArray;\nimport com.google.gson.JsonElement;\nimport com.google.gson.JsonObject;\nimport com.google.gson.JsonParser;\n\n\n\npublic class MyResource extends ServerResource {\n\n\tprivate static final Logger LOGGER = LoggerFactory.getLogger(MyResource.class);\n private TurtleNIFDocumentParser parser = new TurtleNIFDocumentParser();\n private TurtleNIFDocumentCreator creator = new TurtleNIFDocumentCreator();\n\n private String clusterServiceURL = \"http://localhost:5555/\";\n \n private Gson gson = new GsonBuilder().create();\n private HttpClient client = HttpClients.createDefault(); \n \n @Post\n public String accept(Representation request) {\n \tSystem.out.println(\"-------------------------------------------------\");\n Reader inputReader;\n try {\n inputReader = request.getReader();\n } catch (IOException e) {\n System.err.println(\"Exception while reading request.\" + e.getMessage());\n return \"\";\n }\n //LOGGER.debug(\"New post received: \" + );\n Document document;\n try {\n document = parser.getDocumentFromNIFReader(inputReader);\n } catch (Exception e) {\n \tSystem.err.println(\"Exception while converting to DocumentNIFReader.\" + e.getMessage());\n return \"\";\n }\n\n LOGGER.debug(\"New post from Gerbil: \" + document.toString());\n List markings = document.getMarkings();\n String text = document.getText(); \n \t\n List entities = sendRequestToCluster(text, markings); \n \n document.setMarkings(entities);\n String nifDocument = creator.getDocumentAsNIFString(document);\n \n LOGGER.debug(\"ELResource now returning the response to gerbil.\");\n return nifDocument;\n }\n \n \n public List sendRequestToCluster(String text , List markings) {\n \tList entities = new ArrayList(markings.size());\n \t\n \t//JsonObject out = null;\n \tJsonArray out = null;\n \ttry {\n\t\t\tout = queryJson(text, markings, clusterServiceURL);\n\t\t} catch (IOException e) {\n\t\t\te.printStackTrace();\n\t\t}\n \t\n \t// process the response\n \t/*if (out!= null) {\n \t\tJsonArray outMentionsJson = out.getAsJsonArray(\"annotations\");\n \t\tfor (JsonElement je : outMentionsJson) {\n \t\t\tJsonObject mentionJson = je.getAsJsonObject();\n \t\t\tint start = mentionJson.get(\"start\").getAsInt();\n \t\t\tint length = mentionJson.get(\"length\").getAsInt();\n \t\t\tString entity = mentionJson.get(\"entity\").getAsString();\n \t\t\tentities.add(new NamedEntity(start, length, entity)); \t\t\n \t\t}\n \t}*/\n \t//System.out.println(\"Json response from NN: \"+out);\n \tif (out!= null) {\n \t\t//JsonArray outMentionsJson = out.getAsJsonArray(\"annotations\");\n \t\tfor (JsonElement je : out) {\n \t\t\tJsonArray mentionJson = je.getAsJsonArray();\n \t\t\tint start = mentionJson.get(0).getAsInt();\n \t\t\tint length = mentionJson.get(1).getAsInt();\n \t\t\tString entity = mentionJson.get(2).getAsString();\n \t\t\tentities.add(new NamedEntity(start, length, \"http://dbpedia.org/resource/\" + entity)); \t\t\n \t\t}\n \t}\n \tSystem.out.println(\"my NN response: \" + entities);\n \treturn entities;\n }\n \n \n private JsonArray queryJson(String text, List markings, String url) throws IOException {\n\n JsonObject parameters = new JsonObject();\n\n if (markings != null) {\n JsonArray mentionsJson = new JsonArray();\n for (Marking m : markings) {\n \tSpan sp = (Span) m;\n JsonObject mentionJson = new JsonObject();\n mentionJson.addProperty(\"start\", sp.getStartPosition());\n mentionJson.addProperty(\"length\", sp.getLength());\n mentionsJson.add(mentionJson);\n }\n parameters.add(\"spans\", mentionsJson);\n }\n parameters.addProperty(\"text\", text);\n\n HttpPost request = new HttpPost(url);\n request.addHeader(\"Content-Type\", \"application/json\");\n request.setEntity(new StringEntity(gson.toJson(parameters), \"UTF8\"));\n request.addHeader(\"Accept\", \"application/json\");\n\n CloseableHttpResponse response = (CloseableHttpResponse) client.execute(request);\n InputStream is = null;\n HttpEntity entity = null;\n try {\n StatusLine status = response.getStatusLine();\n if ((status.getStatusCode() < 200) || (status.getStatusCode() >= 300)) {\n entity = response.getEntity();\n System.err.println(\"The response had a wrong status: \\\"\" + status.toString() + \"\\\". Content of response: \\\"\"\n + IOUtils.toString(entity.getContent()) + \"\\\". Returning null.\");\n return null;\n }\n entity = response.getEntity();\n is = entity.getContent();\n //return new JsonParser().parse(IOUtils.toString(is)).getAsJsonObject();\n return new JsonParser().parse(IOUtils.toString(is)).getAsJsonArray();\n } catch (Exception e) {\n \tSystem.err.println(\"Couldn't request annotation for given text. Returning null.\" + e.getMessage());\n } finally {\n IOUtils.closeQuietly(is);\n if (entity != null) {\n EntityUtils.consume(entity);\n }\n if (response != null) {\n response.close();\n }\n }\n return null;\n } \n}\n" }, { "path": "gerbil-SpotWrapNifWS4Test/src/main/java/org/aksw/gerbil/ws4test/SpotlightClient.java", "content": "package org.aksw.gerbil.ws4test;\n\nimport java.io.BufferedReader;\nimport java.io.DataOutputStream;\nimport java.io.IOException;\nimport java.io.InputStream;\nimport java.io.InputStreamReader;\nimport java.io.Reader;\nimport java.io.UnsupportedEncodingException;\nimport java.net.HttpURLConnection;\nimport java.net.URL;\nimport java.net.URLDecoder;\nimport java.net.URLEncoder;\nimport java.util.ArrayList;\nimport java.util.HashSet;\nimport java.util.List;\nimport java.util.Set;\n\nimport org.aksw.gerbil.transfer.nif.Document;\nimport org.aksw.gerbil.transfer.nif.Span;\nimport org.aksw.gerbil.transfer.nif.data.DocumentImpl;\nimport org.aksw.gerbil.transfer.nif.data.SpanImpl;\nimport org.aksw.gerbil.transfer.nif.data.TypedNamedEntity;\nimport org.apache.commons.io.IOUtils;\nimport org.json.simple.JSONArray;\nimport org.json.simple.JSONObject;\nimport org.json.simple.parser.JSONParser;\nimport org.json.simple.parser.ParseException;\nimport org.slf4j.Logger;\nimport org.slf4j.LoggerFactory;\n\nimport com.carrotsearch.hppc.ObjectObjectOpenHashMap;\n\n/**\n * Client of DBpedia Spotlight. This implementation is mainly based on the implementations of the\n * BAT-Framework and the HAWK project.\n *\n * @author Michael Röder \n *\n */\npublic class SpotlightClient {\n\n private static final Logger LOGGER = LoggerFactory.getLogger(SpotlightClient.class);\n\n private static final String TYPE_PREFIX_URI_MAPPING[][] = new String[][] {\n {\"freebase\", \"http://rdf.freebase.com/ns/\"}, {\"dbpedia\", \"http://dbpedia.org/ontology/\"}};\n\n private static final String DEFAULT_REQUEST_URL = \"http://model.dbpedia-spotlight.org:2222/rest/\";\n // private static final double DEFAULT_MIN_CONFIDENCE = -1;\n // private static final int DEFAULT_MIN_SUPPORT = -1;\n\n private static final String ANNOTATE_RESOURCE = \"annotate\";\n private static final String SPOT_RESOURCE = \"spot\";\n private static final String DISAMBIGUATE_RESOURCE = \"disambiguate\";\n\n private final String serviceURL;\n // private double minConfidence = 0.2;\n // private int minSupport = 20;\n private final ObjectObjectOpenHashMap typePrefixToUriMapping;\n\n public SpotlightClient() {\n this(DEFAULT_REQUEST_URL/* , DEFAULT_MIN_CONFIDENCE, DEFAULT_MIN_SUPPORT */);\n }\n\n // public SpotlightClient(String serviceURL, double minConfidence, int\n // minSupport) {\n // super();\n // this.serviceURL = serviceURL.endsWith(\"/\") ? serviceURL : (serviceURL +\n // \"/\");\n // this.minConfidence = minConfidence;\n // this.minSupport = minSupport;\n //\n // typePrefixToUriMapping = new ObjectObjectOpenHashMap();\n // for (int i = 0; i < TYPE_PREFIX_URI_MAPPING.length; ++i) {\n // typePrefixToUriMapping.put(TYPE_PREFIX_URI_MAPPING[i][0],\n // TYPE_PREFIX_URI_MAPPING[i][1]);\n // }\n // }\n\n public SpotlightClient(final String serviceURL) {\n this.serviceURL = serviceURL.endsWith(\"/\") ? serviceURL : (serviceURL + \"/\");\n\n typePrefixToUriMapping = new ObjectObjectOpenHashMap();\n for (int i = 0; i < TYPE_PREFIX_URI_MAPPING.length; ++i) {\n typePrefixToUriMapping.put(TYPE_PREFIX_URI_MAPPING[i][0], TYPE_PREFIX_URI_MAPPING[i][1]);\n }\n }\n\n protected String request(final String inputText, final String requestUrl) throws IOException {\n final String parameters = \"text=\" + URLEncoder.encode(inputText, \"UTF-8\");\n // parameters += \"&confidence=\" + minConfidence;\n // parameters += \"&support=\" + minSupport;\n\n final URL url = new URL(requestUrl);\n final HttpURLConnection connection = (HttpURLConnection) url.openConnection();\n connection.setRequestMethod(\"POST\");\n connection.setDoOutput(true);\n connection.setDoInput(true);\n connection.setUseCaches(false);\n connection.setRequestProperty(\"Accept\", \"application/json\");\n connection.setRequestProperty(\"Content-Type\",\n \"application/x-www-form-urlencoded;charset=UTF-8\");\n connection.setRequestProperty(\"Content-Length\", String.valueOf(parameters.length()));\n\n final DataOutputStream wr = new DataOutputStream(connection.getOutputStream());\n wr.writeBytes(parameters);\n wr.flush();\n\n final StringBuilder sb = new StringBuilder();\n InputStream is = null;\n Reader ir = null;\n BufferedReader br = null;\n try {\n is = connection.getInputStream();\n ir = new InputStreamReader(is);\n br = new BufferedReader(ir);\n\n while (br.ready()) {\n sb.append(br.readLine());\n }\n } finally {\n IOUtils.closeQuietly(br);\n IOUtils.closeQuietly(ir);\n IOUtils.closeQuietly(is);\n connection.disconnect();\n }\n return sb.toString();\n }\n\n public List annotateSavely(final Document document) {\n try {\n return annotate(document);\n } catch (final IOException e) {\n LOGGER.error(\"Error while requesting DBpedia Spotlight to annotate text. Returning null.\", e);\n return null;\n }\n }\n\n public List annotate(final Document document) throws IOException {\n final String response = request(document.getText(), serviceURL + ANNOTATE_RESOURCE);\n return parseAnnotationResponse(response);\n }\n\n protected List parseAnnotationResponse(final String response) {\n final List markings = new ArrayList();\n\n final JSONParser parser = new JSONParser();\n JSONObject jsonObject = null;\n try {\n jsonObject = (JSONObject) parser.parse(response);\n } catch (final ParseException e) {\n LOGGER.error(\"Error while parsing DBpedia Spotlight response. Returning null.\", e);\n return null;\n }\n\n final JSONArray resources = (JSONArray) jsonObject.get(\"Resources\");\n JSONObject resource;\n int start;\n int length;\n String uri = null;\n Set types;\n String typeStrings[], uriParts[];\n if (resources != null) {\n for (final Object res : resources.toArray()) {\n resource = (JSONObject) res;\n start = Integer.parseInt((String) resource.get(\"@offset\"));\n length = ((String) resource.get(\"@surfaceForm\")).length();\n try {\n uri = URLDecoder.decode((String) resource.get(\"@URI\"), \"UTF-8\");\n } catch (final UnsupportedEncodingException e) {\n LOGGER.error(\"Error while parsing DBpedia Spotlight response. Returning null.\", e);\n return null;\n }\n // create Types set\n typeStrings = ((String) resource.get(\"@types\")).split(\",\");\n types = new HashSet(typeStrings.length);\n for (int i = 0; i < typeStrings.length; ++i) {\n uriParts = typeStrings[i].split(\":\");\n uriParts[0] = uriParts[0].toLowerCase();\n if (typePrefixToUriMapping.containsKey(uriParts[0])) {\n types.add(typePrefixToUriMapping.get(uriParts[0]) + uriParts[1]);\n } else {\n types.add(typeStrings[i]);\n }\n }\n markings.add(new TypedNamedEntity(start, length, uri, types));\n }\n }\n\n return markings;\n }\n\n public List spotSavely(final Document document) {\n try {\n return spot(document);\n } catch (final IOException e) {\n LOGGER.error(\"Error while requesting DBpedia Spotlight to spot text. Returning null.\", e);\n return null;\n }\n }\n\n public List spot(final Document document) throws IOException {\n final String response = request(document.getText(), serviceURL + SPOT_RESOURCE);\n return parseSpottingResponse(response);\n }\n\n protected List parseSpottingResponse(final String response) {\n final List markings = new ArrayList();\n\n final JSONParser parser = new JSONParser();\n JSONObject jsonObject = null;\n try {\n jsonObject = (JSONObject) parser.parse(response);\n } catch (final ParseException e) {\n LOGGER.error(\"Error while parsing DBpedia Spotlight response. Returning null.\", e);\n return null;\n }\n\n jsonObject = (JSONObject) jsonObject.get(\"annotation\");\n final JSONArray resources = (JSONArray) jsonObject.get(\"surfaceForm\");\n JSONObject resource;\n int start;\n int length;\n if (resources != null) {\n for (final Object res : resources.toArray()) {\n resource = (JSONObject) res;\n start = Integer.parseInt((String) resource.get(\"@offset\"));\n length = ((String) resource.get(\"@name\")).length();\n markings.add(new SpanImpl(start, length));\n }\n }\n\n return markings;\n }\n\n public List disambiguate(final Document document) throws IOException {\n final String text = document.getText();\n final StringBuilder requestBuilder = new StringBuilder();\n requestBuilder.append(\"\");\n\n final List spans = document.getMarkings(Span.class);\n int start;\n for (final Span span : spans) {\n start = span.getStartPosition();\n requestBuilder.append(\"\");\n }\n requestBuilder.append(\"\");\n\n final String response = request(requestBuilder.toString(), serviceURL + DISAMBIGUATE_RESOURCE);\n LOGGER.error(response);\n return parseAnnotationResponse(response);\n }\n\n public List disambiguateSavely(final Document document) {\n try {\n return disambiguate(document);\n } catch (final IOException e) {\n LOGGER.error(\"Error while requesting DBpedia Spotlight to spot text. Returning null.\", e);\n return null;\n }\n }\n\n @SuppressWarnings({\"rawtypes\", \"unchecked\"})\n public static void main(final String args[]) {\n final SpotlightClient client = new SpotlightClient();\n List list;\n\n list = client.annotateSavely(new DocumentImpl(\n \"President Obama called Wednesday on Congress to extend a tax break for students included in last year's economic stimulus package, arguing that the policy provides more generous assistance.\"));\n for (int i = 0; i < list.size(); ++i) {\n System.out.println(list.get(i));\n }\n list = client.spotSavely(new DocumentImpl(\n \"President Obama called Wednesday on Congress to extend a tax break for students included in last year's economic stimulus package, arguing that the policy provides more generous assistance.\"));\n for (int i = 0; i < list.size(); ++i) {\n System.out.println(list.get(i));\n }\n list = client.disambiguateSavely(new DocumentImpl(\n \"President Obama called Wednesday on Congress to extend a tax break for students included in last year's economic stimulus package, arguing that the policy provides more generous assistance.\",\n list));\n for (int i = 0; i < list.size(); ++i) {\n System.out.println(list.get(i));\n }\n }\n}\n" }, { "path": "gerbil-SpotWrapNifWS4Test/src/main/java/org/aksw/gerbil/ws4test/SpotlightResource.java", "content": "package org.aksw.gerbil.ws4test;\n\nimport java.io.IOException;\nimport java.io.Reader;\nimport java.util.ArrayList;\n\nimport org.aksw.gerbil.transfer.nif.Document;\nimport org.aksw.gerbil.transfer.nif.Marking;\nimport org.aksw.gerbil.transfer.nif.TurtleNIFDocumentCreator;\nimport org.aksw.gerbil.transfer.nif.TurtleNIFDocumentParser;\nimport org.restlet.representation.Representation;\nimport org.restlet.resource.Post;\nimport org.restlet.resource.ServerResource;\nimport org.slf4j.Logger;\nimport org.slf4j.LoggerFactory;\n\npublic class SpotlightResource extends ServerResource {\n\n private static final Logger LOGGER = LoggerFactory.getLogger(SpotlightResource.class);\n\n private TurtleNIFDocumentParser parser = new TurtleNIFDocumentParser();\n private TurtleNIFDocumentCreator creator = new TurtleNIFDocumentCreator();\n private SpotlightClient client = new SpotlightClient();\n\n @Post\n public String accept(Representation request) {\n Reader inputReader;\n try {\n inputReader = request.getReader();\n } catch (IOException e) {\n LOGGER.error(\"Exception while reading request.\", e);\n return \"\";\n }\n Document document;\n try {\n document = parser.getDocumentFromNIFReader(inputReader);\n } catch (Exception e) {\n LOGGER.error(\"Exception while reading request.\", e);\n return \"\";\n }\n LOGGER.debug(\"Request: \" + document.toString());\n document.setMarkings(new ArrayList(client.annotateSavely(document)));\n LOGGER.debug(\"Result: \" + document.toString());\n String nifDocument = creator.getDocumentAsNIFString(document);\n return nifDocument;\n }\n}\n" }, { "path": "gerbil-SpotWrapNifWS4Test/src/main/java/org/aksw/gerbil/ws4test/TestApplication.java", "content": "/**\n * Copyright (C) 2014 Michael Röder (michael.roeder@unister.de)\n *\n * Permission is hereby granted, free of charge, to any person obtaining a copy of this software and\n * associated documentation files (the \"Software\"), to deal in the Software without restriction,\n * including without limitation the rights to use, copy, modify, merge, publish, distribute,\n * sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is\n * furnished to do so, subject to the following conditions:\n *\n * The above copyright notice and this permission notice shall be included in all copies or\n * substantial portions of the Software.\n *\n * THE SOFTWARE IS PROVIDED \"AS IS\", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT\n * NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND\n * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM,\n * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,\n * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.\n */\npackage org.aksw.gerbil.ws4test;\n\nimport org.restlet.Application;\nimport org.restlet.Restlet;\nimport org.restlet.routing.Router;\n\npublic class TestApplication extends Application {\n\n /**\n * Creates a root Restlet that will receive all incoming calls.\n */\n @Override\n public Restlet createInboundRoot() {\n final Router router = new Router(getContext());\n router.attach(\"/ed\", EDResource.class);\n router.attach(\"/myalgorithm\", MyResource.class);\n router.attach(\"/spotlight\", SpotlightResource.class);\n return router;\n }\n}\n" }, { "path": "gerbil-SpotWrapNifWS4Test/src/main/java/org/aksw/gerbil/ws4test/data_format", "content": "\n\ned\n\nRequest: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_228, text=SOCCER - PSV HIT VOLENDAM FOR SIX. AMSTERDAM 1996-12-07 Brazilian striker Marcelo and Yugoslav midfielder Zeljko Petrovic each scored twice as Dutch first division leaders PSV Eindhoven romped to a 6-0 win over Volendam on Saturday. Their other marksmen were Brazilian defender Vampeta and Belgian striker Luc Nilis, his 14th of the season. PSV, well on the way to their 14th league title, outgunned Volendam in every department of the game. They boast a nine-point lead over Feyenoord, who have two games in hand, and are 16 points clear of champions Ajax Amsterdam, who have played 18 matches compared to PSV's 19. Ajax face AZ Alkmaar away on Sunday and Feyenoord, eliminated from the UEFA Cup after losing 4-2 on aggregate to Tenerife on Tuesday, travel to De Graafschap Doetinchem. The Doetinchem side, dubbed \"The Super Peasants\", are one of the surprise packages of the season. They are fourth in the table., markings=[(400, 8), (17, 8), (552, 14), (341, 3), (56, 9), (791, 10), (306, 9), (627, 10), (278, 7), (259, 9), (106, 15), (657, 9), (607, 3), (761, 24), (290, 7), (74, 7), (476, 9), (172, 13), (730, 8), (617, 4), (86, 8), (211, 8), (143, 5), (35, 9), (816, 18), (9, 3), (688, 8)]]\n2018-02-05 20:05:08,337 DEBUG [org.aksw.gerbil.ws4test.EDResource] - \nResult: AnnotatedDocumentImpl [uri=http://www.aksw.org/gerbil/NifWebService/request_225, text=BOXING - SCHULZ DEFEATS RIBALTA IN IBF HEAVYWEIGHT FIGHT. VIENNA 1996-12-07 German Axel Schulz outpointed Cuba's Jose Ribalta in their International Boxing Federation non-title 10-round heavyweight fight on Saturday., markings=[(35, 3), (106, 4), (24, 7), (9, 6), (83, 11), (76, 6), (113, 12), (58, 6), (135, 31)]]\n2018-02-05 20:05:08,343 DEBUG [org.aksw.gerbil.ws4test.EDResource] - \n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n" }, { "path": "gerbil-SpotWrapNifWS4Test/src/main/resources/log4j.properties", "content": "# Direct log messages to stdout\nlog4j.rootLogger=WARN,stdout,CATALINA\n\nlog4j.appender.stdout=org.apache.log4j.ConsoleAppender\nlog4j.appender.stdout.layout=org.apache.log4j.PatternLayout\nlog4j.appender.stdout.layout.ConversionPattern=%d %p [%c] - <%m>%n\n\nlog4j.appender.CATALINA=org.apache.log4j.ConsoleAppender\nlog4j.appender.CATALINA.Encoding=UTF-8\nlog4j.appender.CATALINA.layout = org.apache.log4j.PatternLayout\nlog4j.appender.stdout.layout.ConversionPattern=%d %p [%c] - <%m>%n\n\nlog4j.category.org.aksw=DEBUG" }, { "path": "gerbil-SpotWrapNifWS4Test/src/main/webapp/WEB-INF/web.xml", "content": "\n\n Gerbil Spotlight Wrapping NIF Webservice 4 Testing\n \n \n org.restlet.application\n org.aksw.gerbil.ws4test.TestApplication\n \n\n \n \n RestletServlet\n org.restlet.ext.servlet.ServerServlet\n \n\n \n \n RestletServlet\n /*\n \n" }, { "path": "readme.md", "content": "# End-to-End Neural Entity Linking (CoNLL 2018, full paper)\n### Python source code\n\n\n1: Setting up the environment.\n```\ngit clone arxiv_url\ncd end2end_neural_el\npython3 -m pip install virtualenv\npython3 -m virtualenv end2end_neural_el_env\nsource end2end_neural_el_env/bin/activate\npip install -r requirements.txt\n```\n\nDownload the 'data' folder from [this link](https://drive.google.com/file/d/1OSKvIiXHVVaWUhQ1-fpvePTBQfgMT6Ps/view?usp=sharing), unzip it and place it under end2end_neural_el/ \nThese data are enough for running the pretrained models and reproducing the results. \nFor reproducing the results you can skip all the next steps and go directly to the \n[gerbil evaluation section](#gerbil-evaluation).\n\nIf you want to create new models with different preprocessing \n(or run the preprocessing steps described below)\nyou should also download the pre-trained Word2Vec vectors \n[GoogleNews-vectors-negative300.bin.gz](https://code.google.com/archive/p/word2vec/). \nUnzip it and place the bin file in the folder end2end_neural_el/data/basic_data/wordEmbeddings/Word2Vec.\n\nIf you also want to train your own entity vectors then follow the instructions from [here](https://github.com/dalab/deep-ed).\nThe 'basic_data' folder this time will be placed under end2end_neural_el/deep-ed/data/\n\n2: Preprocessing stage 1.\n\nTransform the AIDA, ACE2004, AQUAINT, MSNBC, CLUEWEB datasets to a common easy to use format.\nOnly AIDA train is used by our system for training, and the AIDA-TESTA for hyperparameter tuning.\n```\ncd code\npython -m preprocessing.prepro_aida\npython -m preprocessing.prepro_other_datasets\n```\n\n3: Preprocessing stage 2 (converting datasets to tfrecords).\n\nThis step requires the entity vectors and the word-embeddings to exist.\nAn essential part of our system are the entity vectors (the equivalent of word-embeddings for entities). \nYou can create your entity vectors by following the instructions of the [next chapter](#gerbil-evaluation), otherwise you can use\nthe provided pretrained ones. We have pretrained 502661 entity vectors. Specifically, \nwe have trained entity vectors for all the candidate entities from all possible spans of \nAIDA-TestA, AIDA-TestB, AIDA-Training 1, ACE2004, AQUAINT, MSNBC, Clueweb, DBpediaSpotlight, Derczynski, \nERD2014, GERDAQ-Dev, GERDAQ-Test, GERDAQ-TrainingA, GERDAQ-TrainingB, KORE50, \nMicroposts2016-Dev, Microposts2016-Test, Microposts2016-Train, N3-RSS-500, N3-Reuters-128,\nOKE 2015 Task1, OKE 2016 Task1, and the entity relatedness dataset of (Ceccarelli et al., 2013). In more detail, \nthis is done by considering all possible spans\nof the document as a candidate span and querying our p(e|m) dictionary for all the candidate entities\nfor this span (we keep only the top 30 for each candidate span). \n\n1 For AIDA-Training 10% of the candidate entities that are detected from the above\n method are missing from the current set of pretrained entities. So in case you want to evaluate\n the algorithm on AIDA-Training for EL this extra entities are needed.\n\nWe now encode the words and characters of the datasets to numbers, we find the candidate spans and \nfor each one the candidate entities (using the p(e|m) dictionary) and\nencode these to numbers as well.\n```\npython -m preprocessing.prepro_util\n```\n\n4: Training the Neural Network.\n\nBase Model + att + global for EL (according to the paper)\n```\nfor v in 1 2 3\ndo\nbsub -n 2 -W 24:00 -R \"rusage[mem=10000,ngpus_excl_p=1]\" python3 -m model.train \\\n --batch_size=4 --experiment_name=corefmerge \\\n --training_name=group_global/global_model_v$v \\\n --ent_vecs_regularization=l2dropout --evaluation_minutes=10 --nepoch_no_imprv=6 \\\n --span_emb=\"boundaries\" \\\n --dim_char=50 --hidden_size_char=50 --hidden_size_lstm=150 \\\n --nn_components=pem_lstm_attention_global \\\n --fast_evaluation=True --all_spans_training=True \\\n --attention_ent_vecs_no_regularization=True --final_score_ffnn=0_0 \\\n --attention_R=10 --attention_K=100 \\\n --train_datasets=aida_train \\\n --el_datasets=aida_dev_z_aida_test_z_aida_train --el_val_datasets=0 \\\n --global_thr=0.001 --global_score_ffnn=0_0 \ndone\n```\n\nBase Model for EL (according to the paper):\n```\nfor v in 1 2 3\ndo\nbsub -n 2 -W 24:00 -R \"rusage[mem=10000,ngpus_excl_p=1]\" python3 -m model.train \\\n --batch_size=4 --experiment_name=corefmerge \\\n --training_name=group_lstm/base_model_v$v \\\n --ent_vecs_regularization=l2dropout --evaluation_minutes=10 --nepoch_no_imprv=6 \\\n --span_emb=\"boundaries\" \\\n --dim_char=50 --hidden_size_char=50 --hidden_size_lstm=150 \\\n --nn_components=pem_lstm \\\n --fast_evaluation=True --all_spans_training=True \\\n --final_score_ffnn=0_0 \\\n --train_datasets=aida_train \\\n --el_datasets=aida_dev_z_aida_test_z_aida_train --el_val_datasets=0 \ndone\n```\n\n\nBase Model + att + global for ED (the only difference from the corresponding command\nfor the EL model is the absence of the argument --all_spans_training=True which means that we train\nexactly the same architecture but now only with the gold mentions).\n```\nfor v in 1 2 3\ndo\nbsub -n 2 -W 24:00 -R \"rusage[mem=10000,ngpus_excl_p=1]\" python3 -m model.train \\\n --batch_size=4 --experiment_name=corefmerge \\\n --training_name=group_global/global_model_v$v \\\n --ent_vecs_regularization=l2dropout --evaluation_minutes=10 --nepoch_no_imprv=6 \\\n --span_emb=\"boundaries\" \\\n --dim_char=50 --hidden_size_char=50 --hidden_size_lstm=150 \\\n --nn_components=pem_lstm_attention_global \\\n --fast_evaluation=True \\\n --attention_ent_vecs_no_regularization=True --final_score_ffnn=0_0 \\\n --attention_R=10 --attention_K=100 \\\n --train_datasets=aida_train \\\n --ed_datasets=aida_dev_z_aida_test_z_aida_train --ed_val_datasets=0 \\\n --global_thr=0.001 --global_score_ffnn=0_0 \ndone\n```\n\nWhen running multiple experiments with multiple different hyperparameters we want an easy way to evaluate\nthe different models. The following command goes through the log files of all the models and sorts them\nbased on the performance on a selected dataset (here the AIDA-TestA).\n```\ncd evaluation; python summarize_all_experiments.py --macro_or_micro=micro --dev_set=aida_dev --test_set=aida_test\n```\n\n# Gerbil evaluation \nOn one terminal run [Gerbil](https://github.com/dice-group/gerbil). Execute:\n```\ncd gerbil/ \n./start.sh\n```\nCaution: Gerbil might be incompatible with some java versions. With Java 8 it works.\n\nOn another terminal execute:\n```\ncd end2end_neural_el/gerbil-SpotWrapNifWS4Test/\nmvn clean -Dmaven.tomcat.port=1235 tomcat:run\n```\n\nOn a third terminal execute:\n```\ncd end2end_neural_el/code \nFor EL:\npython -m gerbil.server --training_name=base_att_global --experiment_name=paper_models \\\n --persons_coreference_merge=True --all_spans_training=True --entity_extension=extension_entities\n\nFor ED:\npython -m gerbil.server --training_name=base_att_global --experiment_name=paper_models \\\n --persons_coreference_merge=True --ed_mode --entity_extension=extension_entities\n \nFor EL using this system for the ED part and stanford NER for the NER subtask.\npython -m gerbil.server --training_name=base_att_global --experiment_name=paper_models \\ \n --persons_coreference_merge=True --ed_mode --entity_extension=extension_entities \\\n --el_with_stanfordner_and_our_ed=True\n```\nBy changing the training_name parameter you can try the different models \n(base_att_global, base_att, basemodel) and reproduce the results shown in the paper (tables 2, 7, and 8).\n\nOpen the url http://localhost:1234/gerbil\n- Configure experiment\n- In URI field write: http://localhost:1235/gerbil-spotWrapNifWS4Test/myalgorithm\n- Name: whatever you wish\n- Press add annotator button\n- Select the datasets that you want to evaluate the model on\n- Run experiment\n\n\n\nFor **local evaluation and for printing the annotated datasets** run the following command:\n(reproduces the results of the table 9 of the paper)\n```\nFor EL:\npython -m model.evaluate --training_name=base_att_global --experiment_name=paper_models --entity_extension=extension_entities --el_datasets=aida_dev_z_aida_test_z_aida_train_z_ace2004_z_aquaint_z_clueweb_z_msnbc_z_wikipedia --el_val_datasets=0 --ed_datasets=\"\" --ed_val_datasets=0 --all_spans_training=True\nFor ED:\npython -m model.evaluate --training_name=base_att_global --experiment_name=paper_models --entity_extension=extension_entities --ed_datasets=aida_dev_z_aida_test_z_aida_train_z_ace2004_z_aquaint_z_clueweb_z_msnbc_z_wikipedia --ed_val_datasets=0 --el_datasets=\"\" --el_val_datasets=0\n```\nThe local evaluation for EL is almost identical to gerbil scores but for ED it is higher. This should\nprobably be attributed to some parsing errors and different preprocessing when input is from gerbil in\ncomparison to local evaluation that is done on the official tokenized datasets.\n\n# Trying the system on random user input text\nIf you want to try the system on custom input it can be done in multiple ways but the simplest is\nfollowing: \nOn one terminal run the command (similar to the Gerbil evaluation)\n```\ncd end2end_neural_el/code \nFor EL:\npython -m gerbil.server --training_name=base_att_global --experiment_name=paper_models \\\n --persons_coreference_merge=True --all_spans_training=True --entity_extension=extension_entities\n\nFor ED:\npython -m gerbil.server --training_name=base_att_global --experiment_name=paper_models \\ \n --persons_coreference_merge=True --ed_mode --entity_extension=extension_entities\n```\nThis launches a server that expects to receive json objects of the following format:\n```\n{ \"text\": \"Obama will visit Germany and have a meeting with Merkel tomorrow.\", \n\"spans\": [{\"start\":0,\"length\":5}, {\"start\":17,\"length\":7}, {\"start\":49,\"length\":6}] }\n```\nFor the EL task of course the \"spans\" field will be an **empty array** e.g.\n```\n{ \"text\": \"Obama will visit Germany and have a meeting with Merkel tomorrow.\", \"spans\": [] }\n```\nIn another terminal you can submit your query in the following way:\nIn python:\n```\nimport requests, json\nmyjson = { \"text\": \"Obama will visit Germany and have a meeting with Merkel tomorrow.\", \"spans\": [{\"start\":0,\"length\":5}, {\"start\":17,\"length\":7}, {\"start\":49,\"length\":6}] }\nmyjson = { \"text\": \"Obama will visit Germany and have a meeting with Merkel tomorrow.\", \"spans\": [] }\nrequests.post(\"http://localhost:5555\", json=myjson)\n```\nFrom the terminal directly with curl command:\n```\ncurl -X POST --header 'Content-Type: application/json' --header 'Accept: application/json' -d \"{ \\\"text\\\": \\\"Obama will visit Germany and have a meeting with Merkel tomorrow.\\\", \\\"spans\\\": [{\\\"start\\\":0,\\\"length\\\":5}, {\\\"start\\\":17,\\\"length\\\":7}, {\\\"start\\\":49,\\\"length\\\":6}] }\" 'http://localhost:5555'\n```\nThe server's terminal prints the result on the screen e.g.\n```\n[(17, 7, 'Germany'), (0, 5, 'Barack_Obama'), (49, 6, 'Angela_Merkel')]\n```\nThe third value of the tuple is the wikipedia title. So to obtain a hyperlink add the prefix \n 'https://en.wikipedia.org/wiki/' e.g. https://en.wikipedia.org/wiki/Barack_Obama\n\nMoreover, you can evalute the jupyter notebook, which gives a usefull way to test on your own data. Find the following notebook in the main repo:\n*Examples _ End-to-End Neural Entity Linking.ipynb*\n\nOther ways to evaluate your corpus is to publish your documents in the supported formats i.e.\npublish them a) in the same format as the files in the data/new_datasets folder b) in the same \nformat as the AIDA dataset and then run the prepro_aida.py c) in the format of ace2004, aquaint, msnbc\netc datasets (look folder data/basic_data/test_datasets/wned-datasets) \n\n# Creating Entity Vectors\nDetailed instructions for creating entity vectors can be found [here](create_entity_vectors.md). \n\n" }, { "path": "requirements.txt", "content": "enum34==1.1.6\nrequests==2.18.4\nscipy==1.0.0\ntensorflow-tensorboard==0.4.0\nnumpy==1.14.0\ntermcolor==1.1.0\ntensorflow==2.5.0\ngensim==3.1.0\nnltk==3.4.5\n\n" } ]